diff options
author | Linus Torvalds <torvalds@linux-foundation.org> | 2016-08-02 20:11:27 (GMT) |
---|---|---|
committer | Linus Torvalds <torvalds@linux-foundation.org> | 2016-08-02 20:11:27 (GMT) |
commit | 221bb8a46e230b9824204ae86537183d9991ff2a (patch) | |
tree | 92510d72285b2285be7cb87288bf088cb28af4c1 /arch | |
parent | f7b32e4c021fd788f13f6785e17efbc3eb05b351 (diff) | |
parent | 23528bb21ee2c9b27f3feddd77a2a3351a8df148 (diff) | |
download | linux-221bb8a46e230b9824204ae86537183d9991ff2a.tar.xz |
Merge tag 'for-linus' of git://git.kernel.org/pub/scm/virt/kvm/kvm
Pull KVM updates from Paolo Bonzini:
- ARM: GICv3 ITS emulation and various fixes. Removal of the
old VGIC implementation.
- s390: support for trapping software breakpoints, nested
virtualization (vSIE), the STHYI opcode, initial extensions
for CPU model support.
- MIPS: support for MIPS64 hosts (32-bit guests only) and lots
of cleanups, preliminary to this and the upcoming support for
hardware virtualization extensions.
- x86: support for execute-only mappings in nested EPT; reduced
vmexit latency for TSC deadline timer (by about 30%) on Intel
hosts; support for more than 255 vCPUs.
- PPC: bugfixes.
* tag 'for-linus' of git://git.kernel.org/pub/scm/virt/kvm/kvm: (302 commits)
KVM: PPC: Introduce KVM_CAP_PPC_HTM
MIPS: Select HAVE_KVM for MIPS64_R{2,6}
MIPS: KVM: Reset CP0_PageMask during host TLB flush
MIPS: KVM: Fix ptr->int cast via KVM_GUEST_KSEGX()
MIPS: KVM: Sign extend MFC0/RDHWR results
MIPS: KVM: Fix 64-bit big endian dynamic translation
MIPS: KVM: Fail if ebase doesn't fit in CP0_EBase
MIPS: KVM: Use 64-bit CP0_EBase when appropriate
MIPS: KVM: Set CP0_Status.KX on MIPS64
MIPS: KVM: Make entry code MIPS64 friendly
MIPS: KVM: Use kmap instead of CKSEG0ADDR()
MIPS: KVM: Use virt_to_phys() to get commpage PFN
MIPS: Fix definition of KSEGX() for 64-bit
KVM: VMX: Add VMCS to CPU's loaded VMCSs before VMPTRLD
kvm: x86: nVMX: maintain internal copy of current VMCS
KVM: PPC: Book3S HV: Save/restore TM state in H_CEDE
KVM: PPC: Book3S HV: Pull out TM state save/restore into separate procedures
KVM: arm64: vgic-its: Simplify MAPI error handling
KVM: arm64: vgic-its: Make vgic_its_cmd_handle_mapi similar to other handlers
KVM: arm64: vgic-its: Turn device_id validation into generic ID validation
...
Diffstat (limited to 'arch')
131 files changed, 9022 insertions, 3507 deletions
diff --git a/arch/arm/include/asm/kvm_asm.h b/arch/arm/include/asm/kvm_asm.h index 3d5a5cd..58faff5 100644 --- a/arch/arm/include/asm/kvm_asm.h +++ b/arch/arm/include/asm/kvm_asm.h @@ -66,6 +66,8 @@ extern void __kvm_tlb_flush_vmid(struct kvm *kvm); extern int __kvm_vcpu_run(struct kvm_vcpu *vcpu); extern void __init_stage2_translation(void); + +extern void __kvm_hyp_reset(unsigned long); #endif #endif /* __ARM_KVM_ASM_H__ */ diff --git a/arch/arm/include/asm/kvm_host.h b/arch/arm/include/asm/kvm_host.h index 96387d4..de338d9 100644 --- a/arch/arm/include/asm/kvm_host.h +++ b/arch/arm/include/asm/kvm_host.h @@ -241,8 +241,7 @@ int kvm_arm_coproc_set_reg(struct kvm_vcpu *vcpu, const struct kvm_one_reg *); int handle_exit(struct kvm_vcpu *vcpu, struct kvm_run *run, int exception_index); -static inline void __cpu_init_hyp_mode(phys_addr_t boot_pgd_ptr, - phys_addr_t pgd_ptr, +static inline void __cpu_init_hyp_mode(phys_addr_t pgd_ptr, unsigned long hyp_stack_ptr, unsigned long vector_ptr) { @@ -251,18 +250,13 @@ static inline void __cpu_init_hyp_mode(phys_addr_t boot_pgd_ptr, * code. The init code doesn't need to preserve these * registers as r0-r3 are already callee saved according to * the AAPCS. - * Note that we slightly misuse the prototype by casing the + * Note that we slightly misuse the prototype by casting the * stack pointer to a void *. - * - * We don't have enough registers to perform the full init in - * one go. Install the boot PGD first, and then install the - * runtime PGD, stack pointer and vectors. The PGDs are always - * passed as the third argument, in order to be passed into - * r2-r3 to the init code (yes, this is compliant with the - * PCS!). - */ - kvm_call_hyp(NULL, 0, boot_pgd_ptr); + * The PGDs are always passed as the third argument, in order + * to be passed into r2-r3 to the init code (yes, this is + * compliant with the PCS!). + */ kvm_call_hyp((void*)hyp_stack_ptr, vector_ptr, pgd_ptr); } @@ -272,16 +266,13 @@ static inline void __cpu_init_stage2(void) kvm_call_hyp(__init_stage2_translation); } -static inline void __cpu_reset_hyp_mode(phys_addr_t boot_pgd_ptr, +static inline void __cpu_reset_hyp_mode(unsigned long vector_ptr, phys_addr_t phys_idmap_start) { - /* - * TODO - * kvm_call_reset(boot_pgd_ptr, phys_idmap_start); - */ + kvm_call_hyp((void *)virt_to_idmap(__kvm_hyp_reset), vector_ptr); } -static inline int kvm_arch_dev_ioctl_check_extension(long ext) +static inline int kvm_arch_dev_ioctl_check_extension(struct kvm *kvm, long ext) { return 0; } diff --git a/arch/arm/include/asm/kvm_hyp.h b/arch/arm/include/asm/kvm_hyp.h index f0e8607..6eaff28 100644 --- a/arch/arm/include/asm/kvm_hyp.h +++ b/arch/arm/include/asm/kvm_hyp.h @@ -25,9 +25,6 @@ #define __hyp_text __section(.hyp.text) notrace -#define kern_hyp_va(v) (v) -#define hyp_kern_va(v) (v) - #define __ACCESS_CP15(CRn, Op1, CRm, Op2) \ "mrc", "mcr", __stringify(p15, Op1, %0, CRn, CRm, Op2), u32 #define __ACCESS_CP15_64(Op1, CRm) \ diff --git a/arch/arm/include/asm/kvm_mmu.h b/arch/arm/include/asm/kvm_mmu.h index f9a6506..3bb803d 100644 --- a/arch/arm/include/asm/kvm_mmu.h +++ b/arch/arm/include/asm/kvm_mmu.h @@ -26,16 +26,7 @@ * We directly use the kernel VA for the HYP, as we can directly share * the mapping (HTTBR "covers" TTBR1). */ -#define HYP_PAGE_OFFSET_MASK UL(~0) -#define HYP_PAGE_OFFSET PAGE_OFFSET -#define KERN_TO_HYP(kva) (kva) - -/* - * Our virtual mapping for the boot-time MMU-enable code. Must be - * shared across all the page-tables. Conveniently, we use the vectors - * page, where no kernel data will ever be shared with HYP. - */ -#define TRAMPOLINE_VA UL(CONFIG_VECTORS_BASE) +#define kern_hyp_va(kva) (kva) /* * KVM_MMU_CACHE_MIN_PAGES is the number of stage2 page table translation levels. @@ -49,9 +40,8 @@ #include <asm/pgalloc.h> #include <asm/stage2_pgtable.h> -int create_hyp_mappings(void *from, void *to); +int create_hyp_mappings(void *from, void *to, pgprot_t prot); int create_hyp_io_mappings(void *from, void *to, phys_addr_t); -void free_boot_hyp_pgd(void); void free_hyp_pgds(void); void stage2_unmap_vm(struct kvm *kvm); @@ -65,7 +55,6 @@ int kvm_handle_guest_abort(struct kvm_vcpu *vcpu, struct kvm_run *run); void kvm_mmu_free_memory_caches(struct kvm_vcpu *vcpu); phys_addr_t kvm_mmu_get_httbr(void); -phys_addr_t kvm_mmu_get_boot_httbr(void); phys_addr_t kvm_get_idmap_vector(void); phys_addr_t kvm_get_idmap_start(void); int kvm_mmu_init(void); diff --git a/arch/arm/include/asm/pgtable.h b/arch/arm/include/asm/pgtable.h index d622040..a8d656d 100644 --- a/arch/arm/include/asm/pgtable.h +++ b/arch/arm/include/asm/pgtable.h @@ -97,7 +97,9 @@ extern pgprot_t pgprot_s2_device; #define PAGE_READONLY_EXEC _MOD_PROT(pgprot_user, L_PTE_USER | L_PTE_RDONLY) #define PAGE_KERNEL _MOD_PROT(pgprot_kernel, L_PTE_XN) #define PAGE_KERNEL_EXEC pgprot_kernel -#define PAGE_HYP _MOD_PROT(pgprot_kernel, L_PTE_HYP) +#define PAGE_HYP _MOD_PROT(pgprot_kernel, L_PTE_HYP | L_PTE_XN) +#define PAGE_HYP_EXEC _MOD_PROT(pgprot_kernel, L_PTE_HYP | L_PTE_RDONLY) +#define PAGE_HYP_RO _MOD_PROT(pgprot_kernel, L_PTE_HYP | L_PTE_RDONLY | L_PTE_XN) #define PAGE_HYP_DEVICE _MOD_PROT(pgprot_hyp_device, L_PTE_HYP) #define PAGE_S2 _MOD_PROT(pgprot_s2, L_PTE_S2_RDONLY) #define PAGE_S2_DEVICE _MOD_PROT(pgprot_s2_device, L_PTE_S2_RDONLY) diff --git a/arch/arm/include/asm/virt.h b/arch/arm/include/asm/virt.h index d4ceaf5..a2e75b8 100644 --- a/arch/arm/include/asm/virt.h +++ b/arch/arm/include/asm/virt.h @@ -80,6 +80,10 @@ static inline bool is_kernel_in_hyp_mode(void) return false; } +/* The section containing the hypervisor idmap text */ +extern char __hyp_idmap_text_start[]; +extern char __hyp_idmap_text_end[]; + /* The section containing the hypervisor text */ extern char __hyp_text_start[]; extern char __hyp_text_end[]; diff --git a/arch/arm/kvm/Kconfig b/arch/arm/kvm/Kconfig index 02abfff..95a0005 100644 --- a/arch/arm/kvm/Kconfig +++ b/arch/arm/kvm/Kconfig @@ -46,13 +46,6 @@ config KVM_ARM_HOST ---help--- Provides host support for ARM processors. -config KVM_NEW_VGIC - bool "New VGIC implementation" - depends on KVM - default y - ---help--- - uses the new VGIC implementation - source drivers/vhost/Kconfig endif # VIRTUALIZATION diff --git a/arch/arm/kvm/Makefile b/arch/arm/kvm/Makefile index a596b58..5e28df8 100644 --- a/arch/arm/kvm/Makefile +++ b/arch/arm/kvm/Makefile @@ -22,7 +22,6 @@ obj-y += kvm-arm.o init.o interrupts.o obj-y += arm.o handle_exit.o guest.o mmu.o emulate.o reset.o obj-y += coproc.o coproc_a15.o coproc_a7.o mmio.o psci.o perf.o -ifeq ($(CONFIG_KVM_NEW_VGIC),y) obj-y += $(KVM)/arm/vgic/vgic.o obj-y += $(KVM)/arm/vgic/vgic-init.o obj-y += $(KVM)/arm/vgic/vgic-irqfd.o @@ -30,9 +29,4 @@ obj-y += $(KVM)/arm/vgic/vgic-v2.o obj-y += $(KVM)/arm/vgic/vgic-mmio.o obj-y += $(KVM)/arm/vgic/vgic-mmio-v2.o obj-y += $(KVM)/arm/vgic/vgic-kvm-device.o -else -obj-y += $(KVM)/arm/vgic.o -obj-y += $(KVM)/arm/vgic-v2.o -obj-y += $(KVM)/arm/vgic-v2-emul.o -endif obj-y += $(KVM)/arm/arch_timer.o diff --git a/arch/arm/kvm/arm.c b/arch/arm/kvm/arm.c index f1bde7c..d94bb90 100644 --- a/arch/arm/kvm/arm.c +++ b/arch/arm/kvm/arm.c @@ -20,6 +20,7 @@ #include <linux/errno.h> #include <linux/err.h> #include <linux/kvm_host.h> +#include <linux/list.h> #include <linux/module.h> #include <linux/vmalloc.h> #include <linux/fs.h> @@ -122,7 +123,7 @@ int kvm_arch_init_vm(struct kvm *kvm, unsigned long type) if (ret) goto out_fail_alloc; - ret = create_hyp_mappings(kvm, kvm + 1); + ret = create_hyp_mappings(kvm, kvm + 1, PAGE_HYP); if (ret) goto out_free_stage2_pgd; @@ -201,7 +202,7 @@ int kvm_vm_ioctl_check_extension(struct kvm *kvm, long ext) r = KVM_MAX_VCPUS; break; default: - r = kvm_arch_dev_ioctl_check_extension(ext); + r = kvm_arch_dev_ioctl_check_extension(kvm, ext); break; } return r; @@ -239,7 +240,7 @@ struct kvm_vcpu *kvm_arch_vcpu_create(struct kvm *kvm, unsigned int id) if (err) goto free_vcpu; - err = create_hyp_mappings(vcpu, vcpu + 1); + err = create_hyp_mappings(vcpu, vcpu + 1, PAGE_HYP); if (err) goto vcpu_uninit; @@ -377,7 +378,7 @@ void force_vm_exit(const cpumask_t *mask) /** * need_new_vmid_gen - check that the VMID is still valid - * @kvm: The VM's VMID to checkt + * @kvm: The VM's VMID to check * * return true if there is a new generation of VMIDs being used * @@ -616,7 +617,7 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *run) * Enter the guest */ trace_kvm_entry(*vcpu_pc(vcpu)); - __kvm_guest_enter(); + guest_enter_irqoff(); vcpu->mode = IN_GUEST_MODE; ret = kvm_call_hyp(__kvm_vcpu_run, vcpu); @@ -642,14 +643,14 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *run) local_irq_enable(); /* - * We do local_irq_enable() before calling kvm_guest_exit() so + * We do local_irq_enable() before calling guest_exit() so * that if a timer interrupt hits while running the guest we * account that tick as being spent in the guest. We enable - * preemption after calling kvm_guest_exit() so that if we get + * preemption after calling guest_exit() so that if we get * preempted we make sure ticks after that is not counted as * guest time. */ - kvm_guest_exit(); + guest_exit(); trace_kvm_exit(ret, kvm_vcpu_trap_get_class(vcpu), *vcpu_pc(vcpu)); /* @@ -1039,7 +1040,6 @@ long kvm_arch_vm_ioctl(struct file *filp, static void cpu_init_hyp_mode(void *dummy) { - phys_addr_t boot_pgd_ptr; phys_addr_t pgd_ptr; unsigned long hyp_stack_ptr; unsigned long stack_page; @@ -1048,13 +1048,12 @@ static void cpu_init_hyp_mode(void *dummy) /* Switch from the HYP stub to our own HYP init vector */ __hyp_set_vectors(kvm_get_idmap_vector()); - boot_pgd_ptr = kvm_mmu_get_boot_httbr(); pgd_ptr = kvm_mmu_get_httbr(); stack_page = __this_cpu_read(kvm_arm_hyp_stack_page); hyp_stack_ptr = stack_page + PAGE_SIZE; vector_ptr = (unsigned long)kvm_ksym_ref(__kvm_hyp_vector); - __cpu_init_hyp_mode(boot_pgd_ptr, pgd_ptr, hyp_stack_ptr, vector_ptr); + __cpu_init_hyp_mode(pgd_ptr, hyp_stack_ptr, vector_ptr); __cpu_init_stage2(); kvm_arm_init_debug(); @@ -1076,15 +1075,9 @@ static void cpu_hyp_reinit(void) static void cpu_hyp_reset(void) { - phys_addr_t boot_pgd_ptr; - phys_addr_t phys_idmap_start; - - if (!is_kernel_in_hyp_mode()) { - boot_pgd_ptr = kvm_mmu_get_boot_httbr(); - phys_idmap_start = kvm_get_idmap_start(); - - __cpu_reset_hyp_mode(boot_pgd_ptr, phys_idmap_start); - } + if (!is_kernel_in_hyp_mode()) + __cpu_reset_hyp_mode(hyp_default_vectors, + kvm_get_idmap_start()); } static void _kvm_arch_hardware_enable(void *discard) @@ -1294,14 +1287,14 @@ static int init_hyp_mode(void) * Map the Hyp-code called directly from the host */ err = create_hyp_mappings(kvm_ksym_ref(__hyp_text_start), - kvm_ksym_ref(__hyp_text_end)); + kvm_ksym_ref(__hyp_text_end), PAGE_HYP_EXEC); if (err) { kvm_err("Cannot map world-switch code\n"); goto out_err; } err = create_hyp_mappings(kvm_ksym_ref(__start_rodata), - kvm_ksym_ref(__end_rodata)); + kvm_ksym_ref(__end_rodata), PAGE_HYP_RO); if (err) { kvm_err("Cannot map rodata section\n"); goto out_err; @@ -1312,7 +1305,8 @@ static int init_hyp_mode(void) */ for_each_possible_cpu(cpu) { char *stack_page = (char *)per_cpu(kvm_arm_hyp_stack_page, cpu); - err = create_hyp_mappings(stack_page, stack_page + PAGE_SIZE); + err = create_hyp_mappings(stack_page, stack_page + PAGE_SIZE, + PAGE_HYP); if (err) { kvm_err("Cannot map hyp stack\n"); @@ -1324,7 +1318,7 @@ static int init_hyp_mode(void) kvm_cpu_context_t *cpu_ctxt; cpu_ctxt = per_cpu_ptr(kvm_host_cpu_state, cpu); - err = create_hyp_mappings(cpu_ctxt, cpu_ctxt + 1); + err = create_hyp_mappings(cpu_ctxt, cpu_ctxt + 1, PAGE_HYP); if (err) { kvm_err("Cannot map host CPU state: %d\n", err); @@ -1332,10 +1326,6 @@ static int init_hyp_mode(void) } } -#ifndef CONFIG_HOTPLUG_CPU - free_boot_hyp_pgd(); -#endif - /* set size of VMID supported by CPU */ kvm_vmid_bits = kvm_get_vmid_bits(); kvm_info("%d-bit VMID\n", kvm_vmid_bits); diff --git a/arch/arm/kvm/emulate.c b/arch/arm/kvm/emulate.c index a494def..af93e3f 100644 --- a/arch/arm/kvm/emulate.c +++ b/arch/arm/kvm/emulate.c @@ -210,7 +210,7 @@ bool kvm_condition_valid(struct kvm_vcpu *vcpu) * @vcpu: The VCPU pointer * * When exceptions occur while instructions are executed in Thumb IF-THEN - * blocks, the ITSTATE field of the CPSR is not advanved (updated), so we have + * blocks, the ITSTATE field of the CPSR is not advanced (updated), so we have * to do this little bit of work manually. The fields map like this: * * IT[7:0] -> CPSR[26:25],CPSR[15:10] diff --git a/arch/arm/kvm/guest.c b/arch/arm/kvm/guest.c index 9093ed0..9aca920 100644 --- a/arch/arm/kvm/guest.c +++ b/arch/arm/kvm/guest.c @@ -182,7 +182,7 @@ unsigned long kvm_arm_num_regs(struct kvm_vcpu *vcpu) /** * kvm_arm_copy_reg_indices - get indices of all registers. * - * We do core registers right here, then we apppend coproc regs. + * We do core registers right here, then we append coproc regs. */ int kvm_arm_copy_reg_indices(struct kvm_vcpu *vcpu, u64 __user *uindices) { diff --git a/arch/arm/kvm/init.S b/arch/arm/kvm/init.S index 1f9ae17..bf89c91 100644 --- a/arch/arm/kvm/init.S +++ b/arch/arm/kvm/init.S @@ -32,23 +32,13 @@ * r2,r3 = Hypervisor pgd pointer * * The init scenario is: - * - We jump in HYP with four parameters: boot HYP pgd, runtime HYP pgd, - * runtime stack, runtime vectors - * - Enable the MMU with the boot pgd - * - Jump to a target into the trampoline page (remember, this is the same - * physical page!) - * - Now switch to the runtime pgd (same VA, and still the same physical - * page!) + * - We jump in HYP with 3 parameters: runtime HYP pgd, runtime stack, + * runtime vectors * - Invalidate TLBs * - Set stack and vectors + * - Setup the page tables + * - Enable the MMU * - Profit! (or eret, if you only care about the code). - * - * As we only have four registers available to pass parameters (and we - * need six), we split the init in two phases: - * - Phase 1: r0 = 0, r1 = 0, r2,r3 contain the boot PGD. - * Provides the basic HYP init, and enable the MMU. - * - Phase 2: r0 = ToS, r1 = vectors, r2,r3 contain the runtime PGD. - * Switches to the runtime PGD, set stack and vectors. */ .text @@ -68,8 +58,11 @@ __kvm_hyp_init: W(b) . __do_hyp_init: - cmp r0, #0 @ We have a SP? - bne phase2 @ Yes, second stage init + @ Set stack pointer + mov sp, r0 + + @ Set HVBAR to point to the HYP vectors + mcr p15, 4, r1, c12, c0, 0 @ HVBAR @ Set the HTTBR to point to the hypervisor PGD pointer passed mcrr p15, 4, rr_lo_hi(r2, r3), c2 @@ -114,34 +107,25 @@ __do_hyp_init: THUMB( ldr r2, =(HSCTLR_M | HSCTLR_A | HSCTLR_TE) ) orr r1, r1, r2 orr r0, r0, r1 - isb mcr p15, 4, r0, c1, c0, 0 @ HSCR + isb - @ End of init phase-1 eret -phase2: - @ Set stack pointer - mov sp, r0 - - @ Set HVBAR to point to the HYP vectors - mcr p15, 4, r1, c12, c0, 0 @ HVBAR - - @ Jump to the trampoline page - ldr r0, =TRAMPOLINE_VA - adr r1, target - bfi r0, r1, #0, #PAGE_SHIFT - ret r0 + @ r0 : stub vectors address +ENTRY(__kvm_hyp_reset) + /* We're now in idmap, disable MMU */ + mrc p15, 4, r1, c1, c0, 0 @ HSCTLR + ldr r2, =(HSCTLR_M | HSCTLR_A | HSCTLR_C | HSCTLR_I) + bic r1, r1, r2 + mcr p15, 4, r1, c1, c0, 0 @ HSCTLR -target: @ We're now in the trampoline code, switch page tables - mcrr p15, 4, rr_lo_hi(r2, r3), c2 + /* Install stub vectors */ + mcr p15, 4, r0, c12, c0, 0 @ HVBAR isb - @ Invalidate the old TLBs - mcr p15, 4, r0, c8, c7, 0 @ TLBIALLH - dsb ish - eret +ENDPROC(__kvm_hyp_reset) .ltorg diff --git a/arch/arm/kvm/mmu.c b/arch/arm/kvm/mmu.c index 45c43ae..bda27b6 100644 --- a/arch/arm/kvm/mmu.c +++ b/arch/arm/kvm/mmu.c @@ -32,8 +32,6 @@ #include "trace.h" -extern char __hyp_idmap_text_start[], __hyp_idmap_text_end[]; - static pgd_t *boot_hyp_pgd; static pgd_t *hyp_pgd; static pgd_t *merged_hyp_pgd; @@ -484,28 +482,6 @@ static void unmap_hyp_range(pgd_t *pgdp, phys_addr_t start, u64 size) } /** - * free_boot_hyp_pgd - free HYP boot page tables - * - * Free the HYP boot page tables. The bounce page is also freed. - */ -void free_boot_hyp_pgd(void) -{ - mutex_lock(&kvm_hyp_pgd_mutex); - - if (boot_hyp_pgd) { - unmap_hyp_range(boot_hyp_pgd, hyp_idmap_start, PAGE_SIZE); - unmap_hyp_range(boot_hyp_pgd, TRAMPOLINE_VA, PAGE_SIZE); - free_pages((unsigned long)boot_hyp_pgd, hyp_pgd_order); - boot_hyp_pgd = NULL; - } - - if (hyp_pgd) - unmap_hyp_range(hyp_pgd, TRAMPOLINE_VA, PAGE_SIZE); - - mutex_unlock(&kvm_hyp_pgd_mutex); -} - -/** * free_hyp_pgds - free Hyp-mode page tables * * Assumes hyp_pgd is a page table used strictly in Hyp-mode and @@ -519,15 +495,20 @@ void free_hyp_pgds(void) { unsigned long addr; - free_boot_hyp_pgd(); - mutex_lock(&kvm_hyp_pgd_mutex); + if (boot_hyp_pgd) { + unmap_hyp_range(boot_hyp_pgd, hyp_idmap_start, PAGE_SIZE); + free_pages((unsigned long)boot_hyp_pgd, hyp_pgd_order); + boot_hyp_pgd = NULL; + } + if (hyp_pgd) { + unmap_hyp_range(hyp_pgd, hyp_idmap_start, PAGE_SIZE); for (addr = PAGE_OFFSET; virt_addr_valid(addr); addr += PGDIR_SIZE) - unmap_hyp_range(hyp_pgd, KERN_TO_HYP(addr), PGDIR_SIZE); + unmap_hyp_range(hyp_pgd, kern_hyp_va(addr), PGDIR_SIZE); for (addr = VMALLOC_START; is_vmalloc_addr((void*)addr); addr += PGDIR_SIZE) - unmap_hyp_range(hyp_pgd, KERN_TO_HYP(addr), PGDIR_SIZE); + unmap_hyp_range(hyp_pgd, kern_hyp_va(addr), PGDIR_SIZE); free_pages((unsigned long)hyp_pgd, hyp_pgd_order); hyp_pgd = NULL; @@ -679,17 +660,18 @@ static phys_addr_t kvm_kaddr_to_phys(void *kaddr) * create_hyp_mappings - duplicate a kernel virtual address range in Hyp mode * @from: The virtual kernel start address of the range * @to: The virtual kernel end address of the range (exclusive) + * @prot: The protection to be applied to this range * * The same virtual address as the kernel virtual address is also used * in Hyp-mode mapping (modulo HYP_PAGE_OFFSET) to the same underlying * physical pages. */ -int create_hyp_mappings(void *from, void *to) +int create_hyp_mappings(void *from, void *to, pgprot_t prot) { phys_addr_t phys_addr; unsigned long virt_addr; - unsigned long start = KERN_TO_HYP((unsigned long)from); - unsigned long end = KERN_TO_HYP((unsigned long)to); + unsigned long start = kern_hyp_va((unsigned long)from); + unsigned long end = kern_hyp_va((unsigned long)to); if (is_kernel_in_hyp_mode()) return 0; @@ -704,7 +686,7 @@ int create_hyp_mappings(void *from, void *to) err = __create_hyp_mappings(hyp_pgd, virt_addr, virt_addr + PAGE_SIZE, __phys_to_pfn(phys_addr), - PAGE_HYP); + prot); if (err) return err; } @@ -723,8 +705,8 @@ int create_hyp_mappings(void *from, void *to) */ int create_hyp_io_mappings(void *from, void *to, phys_addr_t phys_addr) { - unsigned long start = KERN_TO_HYP((unsigned long)from); - unsigned long end = KERN_TO_HYP((unsigned long)to); + unsigned long start = kern_hyp_va((unsigned long)from); + unsigned long end = kern_hyp_va((unsigned long)to); if (is_kernel_in_hyp_mode()) return 0; @@ -1687,14 +1669,6 @@ phys_addr_t kvm_mmu_get_httbr(void) return virt_to_phys(hyp_pgd); } -phys_addr_t kvm_mmu_get_boot_httbr(void) -{ - if (__kvm_cpu_uses_extended_idmap()) - return virt_to_phys(merged_hyp_pgd); - else - return virt_to_phys(boot_hyp_pgd); -} - phys_addr_t kvm_get_idmap_vector(void) { return hyp_idmap_vector; @@ -1705,6 +1679,22 @@ phys_addr_t kvm_get_idmap_start(void) return hyp_idmap_start; } +static int kvm_map_idmap_text(pgd_t *pgd) +{ + int err; + + /* Create the idmap in the boot page tables */ + err = __create_hyp_mappings(pgd, + hyp_idmap_start, hyp_idmap_end, + __phys_to_pfn(hyp_idmap_start), + PAGE_HYP_EXEC); + if (err) + kvm_err("Failed to idmap %lx-%lx\n", + hyp_idmap_start, hyp_idmap_end); + + return err; +} + int kvm_mmu_init(void) { int err; @@ -1719,28 +1709,41 @@ int kvm_mmu_init(void) */ BUG_ON((hyp_idmap_start ^ (hyp_idmap_end - 1)) & PAGE_MASK); - hyp_pgd = (pgd_t *)__get_free_pages(GFP_KERNEL | __GFP_ZERO, hyp_pgd_order); - boot_hyp_pgd = (pgd_t *)__get_free_pages(GFP_KERNEL | __GFP_ZERO, hyp_pgd_order); + kvm_info("IDMAP page: %lx\n", hyp_idmap_start); + kvm_info("HYP VA range: %lx:%lx\n", + kern_hyp_va(PAGE_OFFSET), kern_hyp_va(~0UL)); - if (!hyp_pgd || !boot_hyp_pgd) { - kvm_err("Hyp mode PGD not allocated\n"); - err = -ENOMEM; + if (hyp_idmap_start >= kern_hyp_va(PAGE_OFFSET) && + hyp_idmap_start < kern_hyp_va(~0UL)) { + /* + * The idmap page is intersecting with the VA space, + * it is not safe to continue further. + */ + kvm_err("IDMAP intersecting with HYP VA, unable to continue\n"); + err = -EINVAL; goto out; } - /* Create the idmap in the boot page tables */ - err = __create_hyp_mappings(boot_hyp_pgd, - hyp_idmap_start, hyp_idmap_end, - __phys_to_pfn(hyp_idmap_start), - PAGE_HYP); - - if (err) { - kvm_err("Failed to idmap %lx-%lx\n", - hyp_idmap_start, hyp_idmap_end); + hyp_pgd = (pgd_t *)__get_free_pages(GFP_KERNEL | __GFP_ZERO, hyp_pgd_order); + if (!hyp_pgd) { + kvm_err("Hyp mode PGD not allocated\n"); + err = -ENOMEM; goto out; } if (__kvm_cpu_uses_extended_idmap()) { + boot_hyp_pgd = (pgd_t *)__get_free_pages(GFP_KERNEL | __GFP_ZERO, + hyp_pgd_order); + if (!boot_hyp_pgd) { + kvm_err("Hyp boot PGD not allocated\n"); + err = -ENOMEM; + goto out; + } + + err = kvm_map_idmap_text(boot_hyp_pgd); + if (err) + goto out; + merged_hyp_pgd = (pgd_t *)__get_free_page(GFP_KERNEL | __GFP_ZERO); if (!merged_hyp_pgd) { kvm_err("Failed to allocate extra HYP pgd\n"); @@ -1748,29 +1751,10 @@ int kvm_mmu_init(void) } __kvm_extend_hypmap(boot_hyp_pgd, hyp_pgd, merged_hyp_pgd, hyp_idmap_start); - return 0; - } - - /* Map the very same page at the trampoline VA */ - err = __create_hyp_mappings(boot_hyp_pgd, - TRAMPOLINE_VA, TRAMPOLINE_VA + PAGE_SIZE, - __phys_to_pfn(hyp_idmap_start), - PAGE_HYP); - if (err) { - kvm_err("Failed to map trampoline @%lx into boot HYP pgd\n", - TRAMPOLINE_VA); - goto out; - } - - /* Map the same page again into the runtime page tables */ - err = __create_hyp_mappings(hyp_pgd, - TRAMPOLINE_VA, TRAMPOLINE_VA + PAGE_SIZE, - __phys_to_pfn(hyp_idmap_start), - PAGE_HYP); - if (err) { - kvm_err("Failed to map trampoline @%lx into runtime HYP pgd\n", - TRAMPOLINE_VA); - goto out; + } else { + err = kvm_map_idmap_text(hyp_pgd); + if (err) + goto out; } return 0; diff --git a/arch/arm/kvm/reset.c b/arch/arm/kvm/reset.c index 0048b5a..4b5e802 100644 --- a/arch/arm/kvm/reset.c +++ b/arch/arm/kvm/reset.c @@ -52,7 +52,7 @@ static const struct kvm_irq_level cortexa_vtimer_irq = { * @vcpu: The VCPU pointer * * This function finds the right table above and sets the registers on the - * virtual CPU struct to their architectually defined reset values. + * virtual CPU struct to their architecturally defined reset values. */ int kvm_reset_vcpu(struct kvm_vcpu *vcpu) { diff --git a/arch/arm64/include/asm/cpufeature.h b/arch/arm64/include/asm/cpufeature.h index 49dd1bd..7099f26 100644 --- a/arch/arm64/include/asm/cpufeature.h +++ b/arch/arm64/include/asm/cpufeature.h @@ -36,8 +36,9 @@ #define ARM64_HAS_VIRT_HOST_EXTN 11 #define ARM64_WORKAROUND_CAVIUM_27456 12 #define ARM64_HAS_32BIT_EL0 13 +#define ARM64_HYP_OFFSET_LOW 14 -#define ARM64_NCAPS 14 +#define ARM64_NCAPS 15 #ifndef __ASSEMBLY__ diff --git a/arch/arm64/include/asm/kvm_arm.h b/arch/arm64/include/asm/kvm_arm.h index 2cdb6b5..4b5c977 100644 --- a/arch/arm64/include/asm/kvm_arm.h +++ b/arch/arm64/include/asm/kvm_arm.h @@ -178,7 +178,7 @@ /* Hyp System Trap Register */ #define HSTR_EL2_T(x) (1 << x) -/* Hyp Coproccessor Trap Register Shifts */ +/* Hyp Coprocessor Trap Register Shifts */ #define CPTR_EL2_TFP_SHIFT 10 /* Hyp Coprocessor Trap Register */ diff --git a/arch/arm64/include/asm/kvm_host.h b/arch/arm64/include/asm/kvm_host.h index 49095fc..3eda975 100644 --- a/arch/arm64/include/asm/kvm_host.h +++ b/arch/arm64/include/asm/kvm_host.h @@ -47,8 +47,7 @@ int __attribute_const__ kvm_target_cpu(void); int kvm_reset_vcpu(struct kvm_vcpu *vcpu); -int kvm_arch_dev_ioctl_check_extension(long ext); -unsigned long kvm_hyp_reset_entry(void); +int kvm_arch_dev_ioctl_check_extension(struct kvm *kvm, long ext); void __extended_idmap_trampoline(phys_addr_t boot_pgd, phys_addr_t idmap_start); struct kvm_arch { @@ -348,8 +347,7 @@ int kvm_perf_teardown(void); struct kvm_vcpu *kvm_mpidr_to_vcpu(struct kvm *kvm, unsigned long mpidr); -static inline void __cpu_init_hyp_mode(phys_addr_t boot_pgd_ptr, - phys_addr_t pgd_ptr, +static inline void __cpu_init_hyp_mode(phys_addr_t pgd_ptr, unsigned long hyp_stack_ptr, unsigned long vector_ptr) { @@ -357,19 +355,14 @@ static inline void __cpu_init_hyp_mode(phys_addr_t boot_pgd_ptr, * Call initialization code, and switch to the full blown * HYP code. */ - __kvm_call_hyp((void *)boot_pgd_ptr, pgd_ptr, - hyp_stack_ptr, vector_ptr); + __kvm_call_hyp((void *)pgd_ptr, hyp_stack_ptr, vector_ptr); } -static inline void __cpu_reset_hyp_mode(phys_addr_t boot_pgd_ptr, +void __kvm_hyp_teardown(void); +static inline void __cpu_reset_hyp_mode(unsigned long vector_ptr, phys_addr_t phys_idmap_start) { - /* - * Call reset code, and switch back to stub hyp vectors. - * Uses __kvm_call_hyp() to avoid kaslr's kvm_ksym_ref() translation. - */ - __kvm_call_hyp((void *)kvm_hyp_reset_entry(), - boot_pgd_ptr, phys_idmap_start); + kvm_call_hyp(__kvm_hyp_teardown, phys_idmap_start); } static inline void kvm_arch_hardware_unsetup(void) {} diff --git a/arch/arm64/include/asm/kvm_hyp.h b/arch/arm64/include/asm/kvm_hyp.h index 44eaff7..cff5105 100644 --- a/arch/arm64/include/asm/kvm_hyp.h +++ b/arch/arm64/include/asm/kvm_hyp.h @@ -25,29 +25,6 @@ #define __hyp_text __section(.hyp.text) notrace -static inline unsigned long __kern_hyp_va(unsigned long v) -{ - asm volatile(ALTERNATIVE("and %0, %0, %1", - "nop", - ARM64_HAS_VIRT_HOST_EXTN) - : "+r" (v) : "i" (HYP_PAGE_OFFSET_MASK)); - return v; -} - -#define kern_hyp_va(v) (typeof(v))(__kern_hyp_va((unsigned long)(v))) - -static inline unsigned long __hyp_kern_va(unsigned long v) -{ - u64 offset = PAGE_OFFSET - HYP_PAGE_OFFSET; - asm volatile(ALTERNATIVE("add %0, %0, %1", - "nop", - ARM64_HAS_VIRT_HOST_EXTN) - : "+r" (v) : "r" (offset)); - return v; -} - -#define hyp_kern_va(v) (typeof(v))(__hyp_kern_va((unsigned long)(v))) - #define read_sysreg_elx(r,nvh,vh) \ ({ \ u64 reg; \ diff --git a/arch/arm64/include/asm/kvm_mmu.h b/arch/arm64/include/asm/kvm_mmu.h index f05ac27..b6bb834 100644 --- a/arch/arm64/include/asm/kvm_mmu.h +++ b/arch/arm64/include/asm/kvm_mmu.h @@ -29,21 +29,48 @@ * * Instead, give the HYP mode its own VA region at a fixed offset from * the kernel by just masking the top bits (which are all ones for a - * kernel address). + * kernel address). We need to find out how many bits to mask. * - * ARMv8.1 (using VHE) does have a TTBR1_EL2, and doesn't use these - * macros (the entire kernel runs at EL2). + * We want to build a set of page tables that cover both parts of the + * idmap (the trampoline page used to initialize EL2), and our normal + * runtime VA space, at the same time. + * + * Given that the kernel uses VA_BITS for its entire address space, + * and that half of that space (VA_BITS - 1) is used for the linear + * mapping, we can also limit the EL2 space to (VA_BITS - 1). + * + * The main question is "Within the VA_BITS space, does EL2 use the + * top or the bottom half of that space to shadow the kernel's linear + * mapping?". As we need to idmap the trampoline page, this is + * determined by the range in which this page lives. + * + * If the page is in the bottom half, we have to use the top half. If + * the page is in the top half, we have to use the bottom half: + * + * T = __virt_to_phys(__hyp_idmap_text_start) + * if (T & BIT(VA_BITS - 1)) + * HYP_VA_MIN = 0 //idmap in upper half + * else + * HYP_VA_MIN = 1 << (VA_BITS - 1) + * HYP_VA_MAX = HYP_VA_MIN + (1 << (VA_BITS - 1)) - 1 + * + * This of course assumes that the trampoline page exists within the + * VA_BITS range. If it doesn't, then it means we're in the odd case + * where the kernel idmap (as well as HYP) uses more levels than the + * kernel runtime page tables (as seen when the kernel is configured + * for 4k pages, 39bits VA, and yet memory lives just above that + * limit, forcing the idmap to use 4 levels of page tables while the + * kernel itself only uses 3). In this particular case, it doesn't + * matter which side of VA_BITS we use, as we're guaranteed not to + * conflict with anything. + * + * When using VHE, there are no separate hyp mappings and all KVM + * functionality is already mapped as part of the main kernel + * mappings, and none of this applies in that case. */ -#define HYP_PAGE_OFFSET_SHIFT VA_BITS -#define HYP_PAGE_OFFSET_MASK ((UL(1) << HYP_PAGE_OFFSET_SHIFT) - 1) -#define HYP_PAGE_OFFSET (PAGE_OFFSET & HYP_PAGE_OFFSET_MASK) -/* - * Our virtual mapping for the idmap-ed MMU-enable code. Must be - * shared across all the page-tables. Conveniently, we use the last - * possible page, where no kernel mapping will ever exist. - */ -#define TRAMPOLINE_VA (HYP_PAGE_OFFSET_MASK & PAGE_MASK) +#define HYP_PAGE_OFFSET_HIGH_MASK ((UL(1) << VA_BITS) - 1) +#define HYP_PAGE_OFFSET_LOW_MASK ((UL(1) << (VA_BITS - 1)) - 1) #ifdef __ASSEMBLY__ @@ -53,13 +80,33 @@ /* * Convert a kernel VA into a HYP VA. * reg: VA to be converted. + * + * This generates the following sequences: + * - High mask: + * and x0, x0, #HYP_PAGE_OFFSET_HIGH_MASK + * nop + * - Low mask: + * and x0, x0, #HYP_PAGE_OFFSET_HIGH_MASK + * and x0, x0, #HYP_PAGE_OFFSET_LOW_MASK + * - VHE: + * nop + * nop + * + * The "low mask" version works because the mask is a strict subset of + * the "high mask", hence performing the first mask for nothing. + * Should be completely invisible on any viable CPU. */ .macro kern_hyp_va reg -alternative_if_not ARM64_HAS_VIRT_HOST_EXTN - and \reg, \reg, #HYP_PAGE_OFFSET_MASK +alternative_if_not ARM64_HAS_VIRT_HOST_EXTN + and \reg, \reg, #HYP_PAGE_OFFSET_HIGH_MASK alternative_else nop alternative_endif +alternative_if_not ARM64_HYP_OFFSET_LOW + nop +alternative_else + and \reg, \reg, #HYP_PAGE_OFFSET_LOW_MASK +alternative_endif .endm #else @@ -70,7 +117,22 @@ alternative_endif #include <asm/mmu_context.h> #include <asm/pgtable.h> -#define KERN_TO_HYP(kva) ((unsigned long)kva - PAGE_OFFSET + HYP_PAGE_OFFSET) +static inline unsigned long __kern_hyp_va(unsigned long v) +{ + asm volatile(ALTERNATIVE("and %0, %0, %1", + "nop", + ARM64_HAS_VIRT_HOST_EXTN) + : "+r" (v) + : "i" (HYP_PAGE_OFFSET_HIGH_MASK)); + asm volatile(ALTERNATIVE("nop", + "and %0, %0, %1", + ARM64_HYP_OFFSET_LOW) + : "+r" (v) + : "i" (HYP_PAGE_OFFSET_LOW_MASK)); + return v; +} + +#define kern_hyp_va(v) (typeof(v))(__kern_hyp_va((unsigned long)(v))) /* * We currently only support a 40bit IPA. @@ -81,9 +143,8 @@ alternative_endif #include <asm/stage2_pgtable.h> -int create_hyp_mappings(void *from, void *to); +int create_hyp_mappings(void *from, void *to, pgprot_t prot); int create_hyp_io_mappings(void *from, void *to, phys_addr_t); -void free_boot_hyp_pgd(void); void free_hyp_pgds(void); void stage2_unmap_vm(struct kvm *kvm); @@ -97,7 +158,6 @@ int kvm_handle_guest_abort(struct kvm_vcpu *vcpu, struct kvm_run *run); void kvm_mmu_free_memory_caches(struct kvm_vcpu *vcpu); phys_addr_t kvm_mmu_get_httbr(void); -phys_addr_t kvm_mmu_get_boot_httbr(void); phys_addr_t kvm_get_idmap_vector(void); phys_addr_t kvm_get_idmap_start(void); int kvm_mmu_init(void); diff --git a/arch/arm64/include/asm/pgtable-hwdef.h b/arch/arm64/include/asm/pgtable-hwdef.h index 2813748..c3ae239 100644 --- a/arch/arm64/include/asm/pgtable-hwdef.h +++ b/arch/arm64/include/asm/pgtable-hwdef.h @@ -164,6 +164,7 @@ #define PTE_CONT (_AT(pteval_t, 1) << 52) /* Contiguous range */ #define PTE_PXN (_AT(pteval_t, 1) << 53) /* Privileged XN */ #define PTE_UXN (_AT(pteval_t, 1) << 54) /* User XN */ +#define PTE_HYP_XN (_AT(pteval_t, 1) << 54) /* HYP XN */ /* * AttrIndx[2:0] encoding (mapping attributes defined in the MAIR* registers). diff --git a/arch/arm64/include/asm/pgtable-prot.h b/arch/arm64/include/asm/pgtable-prot.h index 29fcb33..39f5252 100644 --- a/arch/arm64/include/asm/pgtable-prot.h +++ b/arch/arm64/include/asm/pgtable-prot.h @@ -55,7 +55,9 @@ #define PAGE_KERNEL_EXEC __pgprot(_PAGE_DEFAULT | PTE_UXN | PTE_DIRTY | PTE_WRITE) #define PAGE_KERNEL_EXEC_CONT __pgprot(_PAGE_DEFAULT | PTE_UXN | PTE_DIRTY | PTE_WRITE | PTE_CONT) -#define PAGE_HYP __pgprot(_PAGE_DEFAULT | PTE_HYP) +#define PAGE_HYP __pgprot(_PAGE_DEFAULT | PTE_HYP | PTE_HYP_XN) +#define PAGE_HYP_EXEC __pgprot(_PAGE_DEFAULT | PTE_HYP | PTE_RDONLY) +#define PAGE_HYP_RO __pgprot(_PAGE_DEFAULT | PTE_HYP | PTE_RDONLY | PTE_HYP_XN) #define PAGE_HYP_DEVICE __pgprot(PROT_DEVICE_nGnRE | PTE_HYP) #define PAGE_S2 __pgprot(PROT_DEFAULT | PTE_S2_MEMATTR(MT_S2_NORMAL) | PTE_S2_RDONLY) diff --git a/arch/arm64/include/asm/virt.h b/arch/arm64/include/asm/virt.h index bbc6a8c..1788545 100644 --- a/arch/arm64/include/asm/virt.h +++ b/arch/arm64/include/asm/virt.h @@ -87,6 +87,10 @@ extern void verify_cpu_run_el(void); static inline void verify_cpu_run_el(void) {} #endif +/* The section containing the hypervisor idmap text */ +extern char __hyp_idmap_text_start[]; +extern char __hyp_idmap_text_end[]; + /* The section containing the hypervisor text */ extern char __hyp_text_start[]; extern char __hyp_text_end[]; diff --git a/arch/arm64/include/uapi/asm/kvm.h b/arch/arm64/include/uapi/asm/kvm.h index f209ea1..3051f86 100644 --- a/arch/arm64/include/uapi/asm/kvm.h +++ b/arch/arm64/include/uapi/asm/kvm.h @@ -87,9 +87,11 @@ struct kvm_regs { /* Supported VGICv3 address types */ #define KVM_VGIC_V3_ADDR_TYPE_DIST 2 #define KVM_VGIC_V3_ADDR_TYPE_REDIST 3 +#define KVM_VGIC_ITS_ADDR_TYPE 4 #define KVM_VGIC_V3_DIST_SIZE SZ_64K #define KVM_VGIC_V3_REDIST_SIZE (2 * SZ_64K) +#define KVM_VGIC_V3_ITS_SIZE (2 * SZ_64K) #define KVM_ARM_VCPU_POWER_OFF 0 /* CPU is started in OFF state */ #define KVM_ARM_VCPU_EL1_32BIT 1 /* CPU running a 32bit VM */ diff --git a/arch/arm64/kernel/cpufeature.c b/arch/arm64/kernel/cpufeature.c index 916d27a..62272ea 100644 --- a/arch/arm64/kernel/cpufeature.c +++ b/arch/arm64/kernel/cpufeature.c @@ -726,6 +726,19 @@ static bool runs_at_el2(const struct arm64_cpu_capabilities *entry, int __unused return is_kernel_in_hyp_mode(); } +static bool hyp_offset_low(const struct arm64_cpu_capabilities *entry, + int __unused) +{ + phys_addr_t idmap_addr = virt_to_phys(__hyp_idmap_text_start); + + /* + * Activate the lower HYP offset only if: + * - the idmap doesn't clash with it, + * - the kernel is not running at EL2. + */ + return idmap_addr > GENMASK(VA_BITS - 2, 0) && !is_kernel_in_hyp_mode(); +} + static const struct arm64_cpu_capabilities arm64_features[] = { { .desc = "GIC system register CPU interface", @@ -803,6 +816,12 @@ static const struct arm64_cpu_capabilities arm64_features[] = { .field_pos = ID_AA64PFR0_EL0_SHIFT, .min_field_value = ID_AA64PFR0_EL0_32BIT_64BIT, }, + { + .desc = "Reduced HYP mapping offset", + .capability = ARM64_HYP_OFFSET_LOW, + .def_scope = SCOPE_SYSTEM, + .matches = hyp_offset_low, + }, {}, }; diff --git a/arch/arm64/kvm/Kconfig b/arch/arm64/kvm/Kconfig index c4f26ef..9d2eff0 100644 --- a/arch/arm64/kvm/Kconfig +++ b/arch/arm64/kvm/Kconfig @@ -36,6 +36,7 @@ config KVM select HAVE_KVM_IRQFD select KVM_ARM_VGIC_V3 select KVM_ARM_PMU if HW_PERF_EVENTS + select HAVE_KVM_MSI ---help--- Support hosting virtualized guest machines. We don't support KVM with 16K page tables yet, due to the multiple @@ -54,13 +55,6 @@ config KVM_ARM_PMU Adds support for a virtual Performance Monitoring Unit (PMU) in virtual machines. -config KVM_NEW_VGIC - bool "New VGIC implementation" - depends on KVM - default y - ---help--- - uses the new VGIC implementation - source drivers/vhost/Kconfig endif # VIRTUALIZATION diff --git a/arch/arm64/kvm/Makefile b/arch/arm64/kvm/Makefile index a7a958c..a5b9664 100644 --- a/arch/arm64/kvm/Makefile +++ b/arch/arm64/kvm/Makefile @@ -20,7 +20,6 @@ kvm-$(CONFIG_KVM_ARM_HOST) += emulate.o inject_fault.o regmap.o kvm-$(CONFIG_KVM_ARM_HOST) += hyp.o hyp-init.o handle_exit.o kvm-$(CONFIG_KVM_ARM_HOST) += guest.o debug.o reset.o sys_regs.o sys_regs_generic_v8.o -ifeq ($(CONFIG_KVM_NEW_VGIC),y) kvm-$(CONFIG_KVM_ARM_HOST) += $(KVM)/arm/vgic/vgic.o kvm-$(CONFIG_KVM_ARM_HOST) += $(KVM)/arm/vgic/vgic-init.o kvm-$(CONFIG_KVM_ARM_HOST) += $(KVM)/arm/vgic/vgic-irqfd.o @@ -30,12 +29,6 @@ kvm-$(CONFIG_KVM_ARM_HOST) += $(KVM)/arm/vgic/vgic-mmio.o kvm-$(CONFIG_KVM_ARM_HOST) += $(KVM)/arm/vgic/vgic-mmio-v2.o kvm-$(CONFIG_KVM_ARM_HOST) += $(KVM)/arm/vgic/vgic-mmio-v3.o kvm-$(CONFIG_KVM_ARM_HOST) += $(KVM)/arm/vgic/vgic-kvm-device.o -else -kvm-$(CONFIG_KVM_ARM_HOST) += $(KVM)/arm/vgic.o -kvm-$(CONFIG_KVM_ARM_HOST) += $(KVM)/arm/vgic-v2.o -kvm-$(CONFIG_KVM_ARM_HOST) += $(KVM)/arm/vgic-v2-emul.o -kvm-$(CONFIG_KVM_ARM_HOST) += $(KVM)/arm/vgic-v3.o -kvm-$(CONFIG_KVM_ARM_HOST) += $(KVM)/arm/vgic-v3-emul.o -endif +kvm-$(CONFIG_KVM_ARM_HOST) += $(KVM)/arm/vgic/vgic-its.o kvm-$(CONFIG_KVM_ARM_HOST) += $(KVM)/arm/arch_timer.o kvm-$(CONFIG_KVM_ARM_PMU) += $(KVM)/arm/pmu.o diff --git a/arch/arm64/kvm/guest.c b/arch/arm64/kvm/guest.c index 32fad75..3f9e157 100644 --- a/arch/arm64/kvm/guest.c +++ b/arch/arm64/kvm/guest.c @@ -211,7 +211,7 @@ unsigned long kvm_arm_num_regs(struct kvm_vcpu *vcpu) /** * kvm_arm_copy_reg_indices - get indices of all registers. * - * We do core registers right here, then we apppend system regs. + * We do core registers right here, then we append system regs. */ int kvm_arm_copy_reg_indices(struct kvm_vcpu *vcpu, u64 __user *uindices) { diff --git a/arch/arm64/kvm/hyp-init.S b/arch/arm64/kvm/hyp-init.S index a873a6d..6b29d3d 100644 --- a/arch/arm64/kvm/hyp-init.S +++ b/arch/arm64/kvm/hyp-init.S @@ -53,10 +53,9 @@ __invalid: b . /* - * x0: HYP boot pgd - * x1: HYP pgd - * x2: HYP stack - * x3: HYP vectors + * x0: HYP pgd + * x1: HYP stack + * x2: HYP vectors */ __do_hyp_init: @@ -110,71 +109,27 @@ __do_hyp_init: msr sctlr_el2, x4 isb - /* Skip the trampoline dance if we merged the boot and runtime PGDs */ - cmp x0, x1 - b.eq merged - - /* MMU is now enabled. Get ready for the trampoline dance */ - ldr x4, =TRAMPOLINE_VA - adr x5, target - bfi x4, x5, #0, #PAGE_SHIFT - br x4 - -target: /* We're now in the trampoline code, switch page tables */ - msr ttbr0_el2, x1 - isb - - /* Invalidate the old TLBs */ - tlbi alle2 - dsb sy - -merged: /* Set the stack and new vectors */ + kern_hyp_va x1 + mov sp, x1 kern_hyp_va x2 - mov sp, x2 - kern_hyp_va x3 - msr vbar_el2, x3 + msr vbar_el2, x2 /* Hello, World! */ eret ENDPROC(__kvm_hyp_init) /* - * Reset kvm back to the hyp stub. This is the trampoline dance in - * reverse. If kvm used an extended idmap, __extended_idmap_trampoline - * calls this code directly in the idmap. In this case switching to the - * boot tables is a no-op. - * - * x0: HYP boot pgd - * x1: HYP phys_idmap_start + * Reset kvm back to the hyp stub. */ ENTRY(__kvm_hyp_reset) - /* We're in trampoline code in VA, switch back to boot page tables */ - msr ttbr0_el2, x0 - isb - - /* Ensure the PA branch doesn't find a stale tlb entry or stale code. */ - ic iallu - tlbi alle2 - dsb sy - isb - - /* Branch into PA space */ - adr x0, 1f - bfi x1, x0, #0, #PAGE_SHIFT - br x1 - /* We're now in idmap, disable MMU */ -1: mrs x0, sctlr_el2 + mrs x0, sctlr_el2 ldr x1, =SCTLR_ELx_FLAGS bic x0, x0, x1 // Clear SCTL_M and etc msr sctlr_el2, x0 isb - /* Invalidate the old TLBs */ - tlbi alle2 - dsb sy - /* Install stub vectors */ adr_l x0, __hyp_stub_vectors msr vbar_el2, x0 diff --git a/arch/arm64/kvm/hyp/entry.S b/arch/arm64/kvm/hyp/entry.S index 70254a6..ce9e5e5 100644 --- a/arch/arm64/kvm/hyp/entry.S +++ b/arch/arm64/kvm/hyp/entry.S @@ -164,22 +164,3 @@ alternative_endif eret ENDPROC(__fpsimd_guest_restore) - -/* - * When using the extended idmap, we don't have a trampoline page we can use - * while we switch pages tables during __kvm_hyp_reset. Accessing the idmap - * directly would be ideal, but if we're using the extended idmap then the - * idmap is located above HYP_PAGE_OFFSET, and the address will be masked by - * kvm_call_hyp using kern_hyp_va. - * - * x0: HYP boot pgd - * x1: HYP phys_idmap_start - */ -ENTRY(__extended_idmap_trampoline) - mov x4, x1 - adr_l x3, __kvm_hyp_reset - - /* insert __kvm_hyp_reset()s offset into phys_idmap_start */ - bfi x4, x3, #0, #PAGE_SHIFT - br x4 -ENDPROC(__extended_idmap_trampoline) diff --git a/arch/arm64/kvm/hyp/hyp-entry.S b/arch/arm64/kvm/hyp/hyp-entry.S index 2d87f36..f6d9694 100644 --- a/arch/arm64/kvm/hyp/hyp-entry.S +++ b/arch/arm64/kvm/hyp/hyp-entry.S @@ -62,6 +62,21 @@ ENTRY(__vhe_hyp_call) isb ret ENDPROC(__vhe_hyp_call) + +/* + * Compute the idmap address of __kvm_hyp_reset based on the idmap + * start passed as a parameter, and jump there. + * + * x0: HYP phys_idmap_start + */ +ENTRY(__kvm_hyp_teardown) + mov x4, x0 + adr_l x3, __kvm_hyp_reset + + /* insert __kvm_hyp_reset()s offset into phys_idmap_start */ + bfi x4, x3, #0, #PAGE_SHIFT + br x4 +ENDPROC(__kvm_hyp_teardown) el1_sync: // Guest trapped into EL2 save_x0_to_x3 diff --git a/arch/arm64/kvm/hyp/switch.c b/arch/arm64/kvm/hyp/switch.c index 4373997..ae7855f 100644 --- a/arch/arm64/kvm/hyp/switch.c +++ b/arch/arm64/kvm/hyp/switch.c @@ -299,9 +299,16 @@ static const char __hyp_panic_string[] = "HYP panic:\nPS:%08llx PC:%016llx ESR:% static void __hyp_text __hyp_call_panic_nvhe(u64 spsr, u64 elr, u64 par) { - unsigned long str_va = (unsigned long)__hyp_panic_string; + unsigned long str_va; - __hyp_do_panic(hyp_kern_va(str_va), + /* + * Force the panic string to be loaded from the literal pool, + * making sure it is a kernel address and not a PC-relative + * reference. + */ + asm volatile("ldr %0, =__hyp_panic_string" : "=r" (str_va)); + + __hyp_do_panic(str_va, spsr, elr, read_sysreg(esr_el2), read_sysreg_el2(far), read_sysreg(hpfar_el2), par, diff --git a/arch/arm64/kvm/reset.c b/arch/arm64/kvm/reset.c index b1ad730..5bc4608 100644 --- a/arch/arm64/kvm/reset.c +++ b/arch/arm64/kvm/reset.c @@ -65,7 +65,7 @@ static bool cpu_has_32bit_el1(void) * We currently assume that the number of HW registers is uniform * across all CPUs (see cpuinfo_sanity_check). */ -int kvm_arch_dev_ioctl_check_extension(long ext) +int kvm_arch_dev_ioctl_check_extension(struct kvm *kvm, long ext) { int r; @@ -86,6 +86,12 @@ int kvm_arch_dev_ioctl_check_extension(long ext) case KVM_CAP_VCPU_ATTRIBUTES: r = 1; break; + case KVM_CAP_MSI_DEVID: + if (!kvm) + r = -EINVAL; + else + r = kvm->arch.vgic.msis_require_devid; + break; default: r = 0; } @@ -98,7 +104,7 @@ int kvm_arch_dev_ioctl_check_extension(long ext) * @vcpu: The VCPU pointer * * This function finds the right table above and sets the registers on - * the virtual CPU struct to their architectually defined reset + * the virtual CPU struct to their architecturally defined reset * values. */ int kvm_reset_vcpu(struct kvm_vcpu *vcpu) @@ -132,31 +138,3 @@ int kvm_reset_vcpu(struct kvm_vcpu *vcpu) /* Reset timer */ return kvm_timer_vcpu_reset(vcpu, cpu_vtimer_irq); } - -extern char __hyp_idmap_text_start[]; - -unsigned long kvm_hyp_reset_entry(void) -{ - if (!__kvm_cpu_uses_extended_idmap()) { - unsigned long offset; - - /* - * Find the address of __kvm_hyp_reset() in the trampoline page. - * This is present in the running page tables, and the boot page - * tables, so we call the code here to start the trampoline - * dance in reverse. - */ - offset = (unsigned long)__kvm_hyp_reset - - ((unsigned long)__hyp_idmap_text_start & PAGE_MASK); - - return TRAMPOLINE_VA + offset; - } else { - /* - * KVM is running with merged page tables, which don't have the - * trampoline page mapped. We know the idmap is still mapped, - * but can't be called into directly. Use - * __extended_idmap_trampoline to do the call. - */ - return (unsigned long)kvm_ksym_ref(__extended_idmap_trampoline); - } -} diff --git a/arch/arm64/kvm/sys_regs.c b/arch/arm64/kvm/sys_regs.c index a57d650..b0b225c 100644 --- a/arch/arm64/kvm/sys_regs.c +++ b/arch/arm64/kvm/sys_regs.c @@ -1546,7 +1546,7 @@ static void unhandled_cp_access(struct kvm_vcpu *vcpu, struct sys_reg_params *params) { u8 hsr_ec = kvm_vcpu_trap_get_class(vcpu); - int cp; + int cp = -1; switch(hsr_ec) { case ESR_ELx_EC_CP15_32: @@ -1558,7 +1558,7 @@ static void unhandled_cp_access(struct kvm_vcpu *vcpu, cp = 14; break; default: - WARN_ON((cp = -1)); + WARN_ON(1); } kvm_err("Unsupported guest CP%d access at: %08lx\n", diff --git a/arch/mips/Kconfig b/arch/mips/Kconfig index ac91939..2986713 100644 --- a/arch/mips/Kconfig +++ b/arch/mips/Kconfig @@ -1488,6 +1488,7 @@ config CPU_MIPS64_R2 select CPU_SUPPORTS_HIGHMEM select CPU_SUPPORTS_HUGEPAGES select CPU_SUPPORTS_MSA + select HAVE_KVM help Choose this option to build a kernel for release 2 or later of the MIPS64 architecture. Many modern embedded systems with a 64-bit @@ -1505,6 +1506,7 @@ config CPU_MIPS64_R6 select CPU_SUPPORTS_MSA select GENERIC_CSUM select MIPS_O32_FP64_SUPPORT if MIPS32_O32 + select HAVE_KVM help Choose this option to build a kernel for release 6 or later of the MIPS64 architecture. New MIPS processors, starting with the Warrior diff --git a/arch/mips/include/asm/addrspace.h b/arch/mips/include/asm/addrspace.h index 3b0e51d..c5b04e7 100644 --- a/arch/mips/include/asm/addrspace.h +++ b/arch/mips/include/asm/addrspace.h @@ -45,7 +45,7 @@ /* * Returns the kernel segment base of a given address */ -#define KSEGX(a) ((_ACAST32_ (a)) & 0xe0000000) +#define KSEGX(a) ((_ACAST32_(a)) & _ACAST32_(0xe0000000)) /* * Returns the physical address of a CKSEGx / XKPHYS address diff --git a/arch/mips/include/asm/kvm_host.h b/arch/mips/include/asm/kvm_host.h index 36a391d..b54bcad 100644 --- a/arch/mips/include/asm/kvm_host.h +++ b/arch/mips/include/asm/kvm_host.h @@ -19,6 +19,9 @@ #include <linux/threads.h> #include <linux/spinlock.h> +#include <asm/inst.h> +#include <asm/mipsregs.h> + /* MIPS KVM register ids */ #define MIPS_CP0_32(_R, _S) \ (KVM_REG_MIPS_CP0 | KVM_REG_SIZE_U32 | (8 * (_R) + (_S))) @@ -53,6 +56,12 @@ #define KVM_REG_MIPS_CP0_CONFIG7 MIPS_CP0_32(16, 7) #define KVM_REG_MIPS_CP0_XCONTEXT MIPS_CP0_64(20, 0) #define KVM_REG_MIPS_CP0_ERROREPC MIPS_CP0_64(30, 0) +#define KVM_REG_MIPS_CP0_KSCRATCH1 MIPS_CP0_64(31, 2) +#define KVM_REG_MIPS_CP0_KSCRATCH2 MIPS_CP0_64(31, 3) +#define KVM_REG_MIPS_CP0_KSCRATCH3 MIPS_CP0_64(31, 4) +#define KVM_REG_MIPS_CP0_KSCRATCH4 MIPS_CP0_64(31, 5) +#define KVM_REG_MIPS_CP0_KSCRATCH5 MIPS_CP0_64(31, 6) +#define KVM_REG_MIPS_CP0_KSCRATCH6 MIPS_CP0_64(31, 7) #define KVM_MAX_VCPUS 1 @@ -65,8 +74,14 @@ -/* Special address that contains the comm page, used for reducing # of traps */ -#define KVM_GUEST_COMMPAGE_ADDR 0x0 +/* + * Special address that contains the comm page, used for reducing # of traps + * This needs to be within 32Kb of 0x0 (so the zero register can be used), but + * preferably not at 0x0 so that most kernel NULL pointer dereferences can be + * caught. + */ +#define KVM_GUEST_COMMPAGE_ADDR ((PAGE_SIZE > 0x8000) ? 0 : \ + (0x8000 - PAGE_SIZE)) #define KVM_GUEST_KERNEL_MODE(vcpu) ((kvm_read_c0_guest_status(vcpu->arch.cop0) & (ST0_EXL | ST0_ERL)) || \ ((kvm_read_c0_guest_status(vcpu->arch.cop0) & KSU_USER) == 0)) @@ -93,9 +108,6 @@ #define KVM_INVALID_ADDR 0xdeadbeef extern atomic_t kvm_mips_instance; -extern kvm_pfn_t (*kvm_mips_gfn_to_pfn)(struct kvm *kvm, gfn_t gfn); -extern void (*kvm_mips_release_pfn_clean)(kvm_pfn_t pfn); -extern bool (*kvm_mips_is_error_pfn)(kvm_pfn_t pfn); struct kvm_vm_stat { u32 remote_tlb_flush; @@ -126,28 +138,6 @@ struct kvm_vcpu_stat { u32 halt_wakeup; }; -enum kvm_mips_exit_types { - WAIT_EXITS, - CACHE_EXITS, - SIGNAL_EXITS, - INT_EXITS, - COP_UNUSABLE_EXITS, - TLBMOD_EXITS, - TLBMISS_LD_EXITS, - TLBMISS_ST_EXITS, - ADDRERR_ST_EXITS, - ADDRERR_LD_EXITS, - SYSCALL_EXITS, - RESVD_INST_EXITS, - BREAK_INST_EXITS, - TRAP_INST_EXITS, - MSA_FPE_EXITS, - FPE_EXITS, - MSA_DISABLED_EXITS, - FLUSH_DCACHE_EXITS, - MAX_KVM_MIPS_EXIT_TYPES -}; - struct kvm_arch_memory_slot { }; @@ -215,73 +205,6 @@ struct mips_coproc { #define MIPS_CP0_CONFIG4_SEL 4 #define MIPS_CP0_CONFIG5_SEL 5 -/* Config0 register bits */ -#define CP0C0_M 31 -#define CP0C0_K23 28 -#define CP0C0_KU 25 -#define CP0C0_MDU 20 -#define CP0C0_MM 17 -#define CP0C0_BM 16 -#define CP0C0_BE 15 -#define CP0C0_AT 13 -#define CP0C0_AR 10 -#define CP0C0_MT 7 -#define CP0C0_VI 3 -#define CP0C0_K0 0 - -/* Config1 register bits */ -#define CP0C1_M 31 -#define CP0C1_MMU 25 -#define CP0C1_IS 22 -#define CP0C1_IL 19 -#define CP0C1_IA 16 -#define CP0C1_DS 13 -#define CP0C1_DL 10 -#define CP0C1_DA 7 -#define CP0C1_C2 6 -#define CP0C1_MD 5 -#define CP0C1_PC 4 -#define CP0C1_WR 3 -#define CP0C1_CA 2 -#define CP0C1_EP 1 -#define CP0C1_FP 0 - -/* Config2 Register bits */ -#define CP0C2_M 31 -#define CP0C2_TU 28 -#define CP0C2_TS 24 -#define CP0C2_TL 20 -#define CP0C2_TA 16 -#define CP0C2_SU 12 -#define CP0C2_SS 8 -#define CP0C2_SL 4 -#define CP0C2_SA 0 - -/* Config3 Register bits */ -#define CP0C3_M 31 -#define CP0C3_ISA_ON_EXC 16 -#define CP0C3_ULRI 13 -#define CP0C3_DSPP 10 -#define CP0C3_LPA 7 -#define CP0C3_VEIC 6 -#define CP0C3_VInt 5 -#define CP0C3_SP 4 -#define CP0C3_MT 2 -#define CP0C3_SM 1 -#define CP0C3_TL 0 - -/* MMU types, the first four entries have the same layout as the - CP0C0_MT field. */ -enum mips_mmu_types { - MMU_TYPE_NONE, - MMU_TYPE_R4000, - MMU_TYPE_RESERVED, - MMU_TYPE_FMT, - MMU_TYPE_R3000, - MMU_TYPE_R6000, - MMU_TYPE_R8000 -}; - /* Resume Flags */ #define RESUME_FLAG_DR (1<<0) /* Reload guest nonvolatile state? */ #define RESUME_FLAG_HOST (1<<1) /* Resume host? */ @@ -298,11 +221,6 @@ enum emulation_result { EMULATE_PRIV_FAIL, }; -#define MIPS3_PG_G 0x00000001 /* Global; ignore ASID if in lo0 & lo1 */ -#define MIPS3_PG_V 0x00000002 /* Valid */ -#define MIPS3_PG_NV 0x00000000 -#define MIPS3_PG_D 0x00000004 /* Dirty */ - #define mips3_paddr_to_tlbpfn(x) \ (((unsigned long)(x) >> MIPS3_PG_SHIFT) & MIPS3_PG_FRAME) #define mips3_tlbpfn_to_paddr(x) \ @@ -313,13 +231,11 @@ enum emulation_result { #define VPN2_MASK 0xffffe000 #define KVM_ENTRYHI_ASID MIPS_ENTRYHI_ASID -#define TLB_IS_GLOBAL(x) (((x).tlb_lo0 & MIPS3_PG_G) && \ - ((x).tlb_lo1 & MIPS3_PG_G)) +#define TLB_IS_GLOBAL(x) ((x).tlb_lo[0] & (x).tlb_lo[1] & ENTRYLO_G) #define TLB_VPN2(x) ((x).tlb_hi & VPN2_MASK) #define TLB_ASID(x) ((x).tlb_hi & KVM_ENTRYHI_ASID) -#define TLB_IS_VALID(x, va) (((va) & (1 << PAGE_SHIFT)) \ - ? ((x).tlb_lo1 & MIPS3_PG_V) \ - : ((x).tlb_lo0 & MIPS3_PG_V)) +#define TLB_LO_IDX(x, va) (((va) >> PAGE_SHIFT) & 1) +#define TLB_IS_VALID(x, va) ((x).tlb_lo[TLB_LO_IDX(x, va)] & ENTRYLO_V) #define TLB_HI_VPN2_HIT(x, y) ((TLB_VPN2(x) & ~(x).tlb_mask) == \ ((y) & VPN2_MASK & ~(x).tlb_mask)) #define TLB_HI_ASID_HIT(x, y) (TLB_IS_GLOBAL(x) || \ @@ -328,26 +244,23 @@ enum emulation_result { struct kvm_mips_tlb { long tlb_mask; long tlb_hi; - long tlb_lo0; - long tlb_lo1; + long tlb_lo[2]; }; -#define KVM_MIPS_FPU_FPU 0x1 -#define KVM_MIPS_FPU_MSA 0x2 +#define KVM_MIPS_AUX_FPU 0x1 +#define KVM_MIPS_AUX_MSA 0x2 #define KVM_MIPS_GUEST_TLB_SIZE 64 struct kvm_vcpu_arch { - void *host_ebase, *guest_ebase; + void *guest_ebase; int (*vcpu_run)(struct kvm_run *run, struct kvm_vcpu *vcpu); unsigned long host_stack; unsigned long host_gp; /* Host CP0 registers used when handling exits from guest */ unsigned long host_cp0_badvaddr; - unsigned long host_cp0_cause; unsigned long host_cp0_epc; - unsigned long host_cp0_entryhi; - uint32_t guest_inst; + u32 host_cp0_cause; /* GPRS */ unsigned long gprs[32]; @@ -357,8 +270,8 @@ struct kvm_vcpu_arch { /* FPU State */ struct mips_fpu_struct fpu; - /* Which FPU state is loaded (KVM_MIPS_FPU_*) */ - unsigned int fpu_inuse; + /* Which auxiliary state is loaded (KVM_MIPS_AUX_*) */ + unsigned int aux_inuse; /* COP0 State */ struct mips_coproc *cop0; @@ -370,11 +283,11 @@ struct kvm_vcpu_arch { struct hrtimer comparecount_timer; /* Count timer control KVM register */ - uint32_t count_ctl; + u32 count_ctl; /* Count bias from the raw time */ - uint32_t count_bias; + u32 count_bias; /* Frequency of timer in Hz */ - uint32_t count_hz; + u32 count_hz; /* Dynamic nanosecond bias (multiple of count_period) to avoid overflow */ s64 count_dyn_bias; /* Resume time */ @@ -388,7 +301,7 @@ struct kvm_vcpu_arch { /* Bitmask of pending exceptions to be cleared */ unsigned long pending_exceptions_clr; - unsigned long pending_load_cause; + u32 pending_load_cause; /* Save/Restore the entryhi register when are are preempted/scheduled back in */ unsigned long preempt_entryhi; @@ -397,8 +310,8 @@ struct kvm_vcpu_arch { struct kvm_mips_tlb guest_tlb[KVM_MIPS_GUEST_TLB_SIZE]; /* Cached guest kernel/user ASIDs */ - uint32_t guest_user_asid[NR_CPUS]; - uint32_t guest_kernel_asid[NR_CPUS]; + u32 guest_user_asid[NR_CPUS]; + u32 guest_kernel_asid[NR_CPUS]; struct mm_struct guest_kernel_mm, guest_user_mm; int last_sched_cpu; @@ -408,6 +321,7 @@ struct kvm_vcpu_arch { u8 fpu_enabled; u8 msa_enabled; + u8 kscratch_enabled; }; @@ -461,6 +375,18 @@ struct kvm_vcpu_arch { #define kvm_write_c0_guest_config7(cop0, val) (cop0->reg[MIPS_CP0_CONFIG][7] = (val)) #define kvm_read_c0_guest_errorepc(cop0) (cop0->reg[MIPS_CP0_ERROR_PC][0]) #define kvm_write_c0_guest_errorepc(cop0, val) (cop0->reg[MIPS_CP0_ERROR_PC][0] = (val)) +#define kvm_read_c0_guest_kscratch1(cop0) (cop0->reg[MIPS_CP0_DESAVE][2]) +#define kvm_read_c0_guest_kscratch2(cop0) (cop0->reg[MIPS_CP0_DESAVE][3]) +#define kvm_read_c0_guest_kscratch3(cop0) (cop0->reg[MIPS_CP0_DESAVE][4]) +#define kvm_read_c0_guest_kscratch4(cop0) (cop0->reg[MIPS_CP0_DESAVE][5]) +#define kvm_read_c0_guest_kscratch5(cop0) (cop0->reg[MIPS_CP0_DESAVE][6]) +#define kvm_read_c0_guest_kscratch6(cop0) (cop0->reg[MIPS_CP0_DESAVE][7]) +#define kvm_write_c0_guest_kscratch1(cop0, val) (cop0->reg[MIPS_CP0_DESAVE][2] = (val)) +#define kvm_write_c0_guest_kscratch2(cop0, val) (cop0->reg[MIPS_CP0_DESAVE][3] = (val)) +#define kvm_write_c0_guest_kscratch3(cop0, val) (cop0->reg[MIPS_CP0_DESAVE][4] = (val)) +#define kvm_write_c0_guest_kscratch4(cop0, val) (cop0->reg[MIPS_CP0_DESAVE][5] = (val)) +#define kvm_write_c0_guest_kscratch5(cop0, val) (cop0->reg[MIPS_CP0_DESAVE][6] = (val)) +#define kvm_write_c0_guest_kscratch6(cop0, val) (cop0->reg[MIPS_CP0_DESAVE][7] = (val)) /* * Some of the guest registers may be modified asynchronously (e.g. from a @@ -474,7 +400,7 @@ static inline void _kvm_atomic_set_c0_guest_reg(unsigned long *reg, unsigned long temp; do { __asm__ __volatile__( - " .set mips3 \n" + " .set "MIPS_ISA_ARCH_LEVEL" \n" " " __LL "%0, %1 \n" " or %0, %2 \n" " " __SC "%0, %1 \n" @@ -490,7 +416,7 @@ static inline void _kvm_atomic_clear_c0_guest_reg(unsigned long *reg, unsigned long temp; do { __asm__ __volatile__( - " .set mips3 \n" + " .set "MIPS_ISA_ARCH_LEVEL" \n" " " __LL "%0, %1 \n" " and %0, %2 \n" " " __SC "%0, %1 \n" @@ -507,7 +433,7 @@ static inline void _kvm_atomic_change_c0_guest_reg(unsigned long *reg, unsigned long temp; do { __asm__ __volatile__( - " .set mips3 \n" + " .set "MIPS_ISA_ARCH_LEVEL" \n" " " __LL "%0, %1 \n" " and %0, %2 \n" " or %0, %3 \n" @@ -542,7 +468,7 @@ static inline void _kvm_atomic_change_c0_guest_reg(unsigned long *reg, static inline bool kvm_mips_guest_can_have_fpu(struct kvm_vcpu_arch *vcpu) { - return (!__builtin_constant_p(cpu_has_fpu) || cpu_has_fpu) && + return (!__builtin_constant_p(raw_cpu_has_fpu) || raw_cpu_has_fpu) && vcpu->fpu_enabled; } @@ -589,9 +515,11 @@ struct kvm_mips_callbacks { void (*dequeue_io_int)(struct kvm_vcpu *vcpu, struct kvm_mips_interrupt *irq); int (*irq_deliver)(struct kvm_vcpu *vcpu, unsigned int priority, - uint32_t cause); + u32 cause); int (*irq_clear)(struct kvm_vcpu *vcpu, unsigned int priority, - uint32_t cause); + u32 cause); + unsigned long (*num_regs)(struct kvm_vcpu *vcpu); + int (*copy_reg_indices)(struct kvm_vcpu *vcpu, u64 __user *indices); int (*get_one_reg)(struct kvm_vcpu *vcpu, const struct kvm_one_reg *reg, s64 *v); int (*set_one_reg)(struct kvm_vcpu *vcpu, @@ -605,8 +533,13 @@ int kvm_mips_emulation_init(struct kvm_mips_callbacks **install_callbacks); /* Debug: dump vcpu state */ int kvm_arch_vcpu_dump_regs(struct kvm_vcpu *vcpu); -/* Trampoline ASM routine to start running in "Guest" context */ -extern int __kvm_mips_vcpu_run(struct kvm_run *run, struct kvm_vcpu *vcpu); +extern int kvm_mips_handle_exit(struct kvm_run *run, struct kvm_vcpu *vcpu); + +/* Building of entry/exception code */ +int kvm_mips_entry_setup(void); +void *kvm_mips_build_vcpu_run(void *addr); +void *kvm_mips_build_exception(void *addr, void *handler); +void *kvm_mips_build_exit(void *addr); /* FPU/MSA context management */ void __kvm_save_fpu(struct kvm_vcpu_arch *vcpu); @@ -622,11 +555,11 @@ void kvm_drop_fpu(struct kvm_vcpu *vcpu); void kvm_lose_fpu(struct kvm_vcpu *vcpu); /* TLB handling */ -uint32_t kvm_get_kernel_asid(struct kvm_vcpu *vcpu); +u32 kvm_get_kernel_asid(struct kvm_vcpu *vcpu); -uint32_t kvm_get_user_asid(struct kvm_vcpu *vcpu); +u32 kvm_get_user_asid(struct kvm_vcpu *vcpu); -uint32_t kvm_get_commpage_asid (struct kvm_vcpu *vcpu); +u32 kvm_get_commpage_asid (struct kvm_vcpu *vcpu); extern int kvm_mips_handle_kseg0_tlb_fault(unsigned long badbaddr, struct kvm_vcpu *vcpu); @@ -635,22 +568,24 @@ extern int kvm_mips_handle_commpage_tlb_fault(unsigned long badvaddr, struct kvm_vcpu *vcpu); extern int kvm_mips_handle_mapped_seg_tlb_fault(struct kvm_vcpu *vcpu, - struct kvm_mips_tlb *tlb, - unsigned long *hpa0, - unsigned long *hpa1); + struct kvm_mips_tlb *tlb); -extern enum emulation_result kvm_mips_handle_tlbmiss(unsigned long cause, - uint32_t *opc, +extern enum emulation_result kvm_mips_handle_tlbmiss(u32 cause, + u32 *opc, struct kvm_run *run, struct kvm_vcpu *vcpu); -extern enum emulation_result kvm_mips_handle_tlbmod(unsigned long cause, - uint32_t *opc, +extern enum emulation_result kvm_mips_handle_tlbmod(u32 cause, + u32 *opc, struct kvm_run *run, struct kvm_vcpu *vcpu); extern void kvm_mips_dump_host_tlbs(void); extern void kvm_mips_dump_guest_tlbs(struct kvm_vcpu *vcpu); +extern int kvm_mips_host_tlb_write(struct kvm_vcpu *vcpu, unsigned long entryhi, + unsigned long entrylo0, + unsigned long entrylo1, + int flush_dcache_mask); extern void kvm_mips_flush_host_tlb(int skip_kseg0); extern int kvm_mips_host_tlb_inv(struct kvm_vcpu *vcpu, unsigned long entryhi); @@ -667,90 +602,90 @@ extern void kvm_mips_vcpu_load(struct kvm_vcpu *vcpu, int cpu); extern void kvm_mips_vcpu_put(struct kvm_vcpu *vcpu); /* Emulation */ -uint32_t kvm_get_inst(uint32_t *opc, struct kvm_vcpu *vcpu); -enum emulation_result update_pc(struct kvm_vcpu *vcpu, uint32_t cause); +u32 kvm_get_inst(u32 *opc, struct kvm_vcpu *vcpu); +enum emulation_result update_pc(struct kvm_vcpu *vcpu, u32 cause); -extern enum emulation_result kvm_mips_emulate_inst(unsigned long cause, - uint32_t *opc, +extern enum emulation_result kvm_mips_emulate_inst(u32 cause, + u32 *opc, struct kvm_run *run, struct kvm_vcpu *vcpu); -extern enum emulation_result kvm_mips_emulate_syscall(unsigned long cause, - uint32_t *opc, +extern enum emulation_result kvm_mips_emulate_syscall(u32 cause, + u32 *opc, struct kvm_run *run, struct kvm_vcpu *vcpu); -extern enum emulation_result kvm_mips_emulate_tlbmiss_ld(unsigned long cause, - uint32_t *opc, +extern enum emulation_result kvm_mips_emulate_tlbmiss_ld(u32 cause, + u32 *opc, struct kvm_run *run, struct kvm_vcpu *vcpu); -extern enum emulation_result kvm_mips_emulate_tlbinv_ld(unsigned long cause, - uint32_t *opc, +extern enum emulation_result kvm_mips_emulate_tlbinv_ld(u32 cause, + u32 *opc, struct kvm_run *run, struct kvm_vcpu *vcpu); -extern enum emulation_result kvm_mips_emulate_tlbmiss_st(unsigned long cause, - uint32_t *opc, +extern enum emulation_result kvm_mips_emulate_tlbmiss_st(u32 cause, + u32 *opc, struct kvm_run *run, struct kvm_vcpu *vcpu); -extern enum emulation_result kvm_mips_emulate_tlbinv_st(unsigned long cause, - uint32_t *opc, +extern enum emulation_result kvm_mips_emulate_tlbinv_st(u32 cause, + u32 *opc, struct kvm_run *run, struct kvm_vcpu *vcpu); -extern enum emulation_result kvm_mips_emulate_tlbmod(unsigned long cause, - uint32_t *opc, +extern enum emulation_result kvm_mips_emulate_tlbmod(u32 cause, + u32 *opc, struct kvm_run *run, struct kvm_vcpu *vcpu); -extern enum emulation_result kvm_mips_emulate_fpu_exc(unsigned long cause, - uint32_t *opc, +extern enum emulation_result kvm_mips_emulate_fpu_exc(u32 cause, + u32 *opc, struct kvm_run *run, struct kvm_vcpu *vcpu); -extern enum emulation_result kvm_mips_handle_ri(unsigned long cause, - uint32_t *opc, +extern enum emulation_result kvm_mips_handle_ri(u32 cause, + u32 *opc, struct kvm_run *run, struct kvm_vcpu *vcpu); -extern enum emulation_result kvm_mips_emulate_ri_exc(unsigned long cause, - uint32_t *opc, +extern enum emulation_result kvm_mips_emulate_ri_exc(u32 cause, + u32 *opc, struct kvm_run *run, struct kvm_vcpu *vcpu); -extern enum emulation_result kvm_mips_emulate_bp_exc(unsigned long cause, - uint32_t *opc, +extern enum emulation_result kvm_mips_emulate_bp_exc(u32 cause, + u32 *opc, struct kvm_run *run, struct kvm_vcpu *vcpu); -extern enum emulation_result kvm_mips_emulate_trap_exc(unsigned long cause, - uint32_t *opc, +extern enum emulation_result kvm_mips_emulate_trap_exc(u32 cause, + u32 *opc, struct kvm_run *run, struct kvm_vcpu *vcpu); -extern enum emulation_result kvm_mips_emulate_msafpe_exc(unsigned long cause, - uint32_t *opc, +extern enum emulation_result kvm_mips_emulate_msafpe_exc(u32 cause, + u32 *opc, struct kvm_run *run, struct kvm_vcpu *vcpu); -extern enum emulation_result kvm_mips_emulate_fpe_exc(unsigned long cause, - uint32_t *opc, +extern enum emulation_result kvm_mips_emulate_fpe_exc(u32 cause, + u32 *opc, struct kvm_run *run, struct kvm_vcpu *vcpu); -extern enum emulation_result kvm_mips_emulate_msadis_exc(unsigned long cause, - uint32_t *opc, +extern enum emulation_result kvm_mips_emulate_msadis_exc(u32 cause, + u32 *opc, struct kvm_run *run, struct kvm_vcpu *vcpu); extern enum emulation_result kvm_mips_complete_mmio_load(struct kvm_vcpu *vcpu, struct kvm_run *run); -uint32_t kvm_mips_read_count(struct kvm_vcpu *vcpu); -void kvm_mips_write_count(struct kvm_vcpu *vcpu, uint32_t count); -void kvm_mips_write_compare(struct kvm_vcpu *vcpu, uint32_t compare, bool ack); +u32 kvm_mips_read_count(struct kvm_vcpu *vcpu); +void kvm_mips_write_count(struct kvm_vcpu *vcpu, u32 count); +void kvm_mips_write_compare(struct kvm_vcpu *vcpu, u32 compare, bool ack); void kvm_mips_init_count(struct kvm_vcpu *vcpu); int kvm_mips_set_count_ctl(struct kvm_vcpu *vcpu, s64 count_ctl); int kvm_mips_set_count_resume(struct kvm_vcpu *vcpu, s64 count_resume); @@ -759,27 +694,27 @@ void kvm_mips_count_enable_cause(struct kvm_vcpu *vcpu); void kvm_mips_count_disable_cause(struct kvm_vcpu *vcpu); enum hrtimer_restart kvm_mips_count_timeout(struct kvm_vcpu *vcpu); -enum emulation_result kvm_mips_check_privilege(unsigned long cause, - uint32_t *opc, +enum emulation_result kvm_mips_check_privilege(u32 cause, + u32 *opc, struct kvm_run *run, struct kvm_vcpu *vcpu); -enum emulation_result kvm_mips_emulate_cache(uint32_t inst, - uint32_t *opc, - uint32_t cause, +enum emulation_result kvm_mips_emulate_cache(union mips_instruction inst, + u32 *opc, + u32 cause, struct kvm_run *run, struct kvm_vcpu *vcpu); -enum emulation_result kvm_mips_emulate_CP0(uint32_t inst, - uint32_t *opc, - uint32_t cause, +enum emulation_result kvm_mips_emulate_CP0(union mips_instruction inst, + u32 *opc, + u32 cause, struct kvm_run *run, struct kvm_vcpu *vcpu); -enum emulation_result kvm_mips_emulate_store(uint32_t inst, - uint32_t cause, +enum emulation_result kvm_mips_emulate_store(union mips_instruction inst, + u32 cause, struct kvm_run *run, struct kvm_vcpu *vcpu); -enum emulation_result kvm_mips_emulate_load(uint32_t inst, - uint32_t cause, +enum emulation_result kvm_mips_emulate_load(union mips_instruction inst, + u32 cause, struct kvm_run *run, struct kvm_vcpu *vcpu); @@ -789,13 +724,13 @@ unsigned int kvm_mips_config4_wrmask(struct kvm_vcpu *vcpu); unsigned int kvm_mips_config5_wrmask(struct kvm_vcpu *vcpu); /* Dynamic binary translation */ -extern int kvm_mips_trans_cache_index(uint32_t inst, uint32_t *opc, - struct kvm_vcpu *vcpu); -extern int kvm_mips_trans_cache_va(uint32_t inst, uint32_t *opc, +extern int kvm_mips_trans_cache_index(union mips_instruction inst, + u32 *opc, struct kvm_vcpu *vcpu); +extern int kvm_mips_trans_cache_va(union mips_instruction inst, u32 *opc, struct kvm_vcpu *vcpu); -extern int kvm_mips_trans_mfc0(uint32_t inst, uint32_t *opc, +extern int kvm_mips_trans_mfc0(union mips_instruction inst, u32 *opc, struct kvm_vcpu *vcpu); -extern int kvm_mips_trans_mtc0(uint32_t inst, uint32_t *opc, +extern int kvm_mips_trans_mtc0(union mips_instruction inst, u32 *opc, struct kvm_vcpu *vcpu); /* Misc */ diff --git a/arch/mips/include/asm/mach-cavium-octeon/cpu-feature-overrides.h b/arch/mips/include/asm/mach-cavium-octeon/cpu-feature-overrides.h index d68e685..bd8b9bb 100644 --- a/arch/mips/include/asm/mach-cavium-octeon/cpu-feature-overrides.h +++ b/arch/mips/include/asm/mach-cavium-octeon/cpu-feature-overrides.h @@ -55,7 +55,7 @@ #define cpu_has_mipsmt 0 #define cpu_has_vint 0 #define cpu_has_veic 0 -#define cpu_hwrena_impl_bits 0xc0000000 +#define cpu_hwrena_impl_bits (MIPS_HWRENA_IMPL1 | MIPS_HWRENA_IMPL2) #define cpu_has_wsbh 1 #define cpu_has_rixi (cpu_data[0].cputype != CPU_CAVIUM_OCTEON) diff --git a/arch/mips/include/asm/mipsregs.h b/arch/mips/include/asm/mipsregs.h index e1ca65c..def9d8d 100644 --- a/arch/mips/include/asm/mipsregs.h +++ b/arch/mips/include/asm/mipsregs.h @@ -53,7 +53,7 @@ #define CP0_SEGCTL2 $5, 4 #define CP0_WIRED $6 #define CP0_INFO $7 -#define CP0_HWRENA $7, 0 +#define CP0_HWRENA $7 #define CP0_BADVADDR $8 #define CP0_BADINSTR $8, 1 #define CP0_COUNT $9 @@ -533,6 +533,7 @@ #define TX49_CONF_CWFON (_ULCAST_(1) << 27) /* Bits specific to the MIPS32/64 PRA. */ +#define MIPS_CONF_VI (_ULCAST_(1) << 3) #define MIPS_CONF_MT (_ULCAST_(7) << 7) #define MIPS_CONF_MT_TLB (_ULCAST_(1) << 7) #define MIPS_CONF_MT_FTLB (_ULCAST_(4) << 7) @@ -853,6 +854,24 @@ #define MIPS_CDMMBASE_ADDR_SHIFT 11 #define MIPS_CDMMBASE_ADDR_START 15 +/* RDHWR register numbers */ +#define MIPS_HWR_CPUNUM 0 /* CPU number */ +#define MIPS_HWR_SYNCISTEP 1 /* SYNCI step size */ +#define MIPS_HWR_CC 2 /* Cycle counter */ +#define MIPS_HWR_CCRES 3 /* Cycle counter resolution */ +#define MIPS_HWR_ULR 29 /* UserLocal */ +#define MIPS_HWR_IMPL1 30 /* Implementation dependent */ +#define MIPS_HWR_IMPL2 31 /* Implementation dependent */ + +/* Bits in HWREna register */ +#define MIPS_HWRENA_CPUNUM (_ULCAST_(1) << MIPS_HWR_CPUNUM) +#define MIPS_HWRENA_SYNCISTEP (_ULCAST_(1) << MIPS_HWR_SYNCISTEP) +#define MIPS_HWRENA_CC (_ULCAST_(1) << MIPS_HWR_CC) +#define MIPS_HWRENA_CCRES (_ULCAST_(1) << MIPS_HWR_CCRES) +#define MIPS_HWRENA_ULR (_ULCAST_(1) << MIPS_HWR_ULR) +#define MIPS_HWRENA_IMPL1 (_ULCAST_(1) << MIPS_HWR_IMPL1) +#define MIPS_HWRENA_IMPL2 (_ULCAST_(1) << MIPS_HWR_IMPL2) + /* * Bitfields in the TX39 family CP0 Configuration Register 3 */ diff --git a/arch/mips/include/asm/setup.h b/arch/mips/include/asm/setup.h index d7bfdeb..4f5279a 100644 --- a/arch/mips/include/asm/setup.h +++ b/arch/mips/include/asm/setup.h @@ -21,6 +21,7 @@ extern void *set_vi_handler(int n, vi_handler_t addr); extern void *set_except_vector(int n, void *addr); extern unsigned long ebase; +extern unsigned int hwrena; extern void per_cpu_trap_init(bool); extern void cpu_cache_init(void); diff --git a/arch/mips/include/asm/uasm.h b/arch/mips/include/asm/uasm.h index b6ecfee..f7929f6 100644 --- a/arch/mips/include/asm/uasm.h +++ b/arch/mips/include/asm/uasm.h @@ -104,8 +104,13 @@ Ip_u1s2(_bltz); Ip_u1s2(_bltzl); Ip_u1u2s3(_bne); Ip_u2s3u1(_cache); +Ip_u1u2(_cfc1); +Ip_u2u1(_cfcmsa); +Ip_u1u2(_ctc1); +Ip_u2u1(_ctcmsa); Ip_u2u1s3(_daddiu); Ip_u3u1u2(_daddu); +Ip_u1(_di); Ip_u2u1msbu3(_dins); Ip_u2u1msbu3(_dinsm); Ip_u1u2(_divu); @@ -141,6 +146,8 @@ Ip_u1(_mfhi); Ip_u1(_mflo); Ip_u1u2u3(_mtc0); Ip_u1u2u3(_mthc0); +Ip_u1(_mthi); +Ip_u1(_mtlo); Ip_u3u1u2(_mul); Ip_u3u1u2(_or); Ip_u2u1u3(_ori); diff --git a/arch/mips/include/uapi/asm/inst.h b/arch/mips/include/uapi/asm/inst.h index 8051f9a..77429d1 100644 --- a/arch/mips/include/uapi/asm/inst.h +++ b/arch/mips/include/uapi/asm/inst.h @@ -21,20 +21,20 @@ enum major_op { spec_op, bcond_op, j_op, jal_op, beq_op, bne_op, blez_op, bgtz_op, - addi_op, cbcond0_op = addi_op, addiu_op, slti_op, sltiu_op, + addi_op, pop10_op = addi_op, addiu_op, slti_op, sltiu_op, andi_op, ori_op, xori_op, lui_op, cop0_op, cop1_op, cop2_op, cop1x_op, beql_op, bnel_op, blezl_op, bgtzl_op, - daddi_op, cbcond1_op = daddi_op, daddiu_op, ldl_op, ldr_op, + daddi_op, pop30_op = daddi_op, daddiu_op, ldl_op, ldr_op, spec2_op, jalx_op, mdmx_op, msa_op = mdmx_op, spec3_op, lb_op, lh_op, lwl_op, lw_op, lbu_op, lhu_op, lwr_op, lwu_op, sb_op, sh_op, swl_op, sw_op, sdl_op, sdr_op, swr_op, cache_op, ll_op, lwc1_op, lwc2_op, bc6_op = lwc2_op, pref_op, - lld_op, ldc1_op, ldc2_op, beqzcjic_op = ldc2_op, ld_op, + lld_op, ldc1_op, ldc2_op, pop66_op = ldc2_op, ld_op, sc_op, swc1_op, swc2_op, balc6_op = swc2_op, major_3b_op, - scd_op, sdc1_op, sdc2_op, bnezcjialc_op = sdc2_op, sd_op + scd_op, sdc1_op, sdc2_op, pop76_op = sdc2_op, sd_op }; /* @@ -93,6 +93,50 @@ enum spec3_op { }; /* + * Bits 10-6 minor opcode for r6 spec mult/div encodings + */ +enum mult_op { + mult_mult_op = 0x0, + mult_mul_op = 0x2, + mult_muh_op = 0x3, +}; +enum multu_op { + multu_multu_op = 0x0, + multu_mulu_op = 0x2, + multu_muhu_op = 0x3, +}; +enum div_op { + div_div_op = 0x0, + div_div6_op = 0x2, + div_mod_op = 0x3, +}; +enum divu_op { + divu_divu_op = 0x0, + divu_divu6_op = 0x2, + divu_modu_op = 0x3, +}; +enum dmult_op { + dmult_dmult_op = 0x0, + dmult_dmul_op = 0x2, + dmult_dmuh_op = 0x3, +}; +enum dmultu_op { + dmultu_dmultu_op = 0x0, + dmultu_dmulu_op = 0x2, + dmultu_dmuhu_op = 0x3, +}; +enum ddiv_op { + ddiv_ddiv_op = 0x0, + ddiv_ddiv6_op = 0x2, + ddiv_dmod_op = 0x3, +}; +enum ddivu_op { + ddivu_ddivu_op = 0x0, + ddivu_ddivu6_op = 0x2, + ddivu_dmodu_op = 0x3, +}; + +/* * rt field of bcond opcodes. */ enum rt_op { @@ -103,7 +147,7 @@ enum rt_op { bltzal_op, bgezal_op, bltzall_op, bgezall_op, rt_op_0x14, rt_op_0x15, rt_op_0x16, rt_op_0x17, rt_op_0x18, rt_op_0x19, rt_op_0x1a, rt_op_0x1b, - bposge32_op, rt_op_0x1d, rt_op_0x1e, rt_op_0x1f + bposge32_op, rt_op_0x1d, rt_op_0x1e, synci_op }; /* @@ -238,6 +282,21 @@ enum bshfl_func { }; /* + * MSA minor opcodes. + */ +enum msa_func { + msa_elm_op = 0x19, +}; + +/* + * MSA ELM opcodes. + */ +enum msa_elm { + msa_ctc_op = 0x3e, + msa_cfc_op = 0x7e, +}; + +/* * func field for MSA MI10 format. */ enum msa_mi10_func { @@ -264,7 +323,7 @@ enum mm_major_op { mm_pool32b_op, mm_pool16b_op, mm_lhu16_op, mm_andi16_op, mm_addiu32_op, mm_lhu32_op, mm_sh32_op, mm_lh32_op, mm_pool32i_op, mm_pool16c_op, mm_lwsp16_op, mm_pool16d_op, - mm_ori32_op, mm_pool32f_op, mm_reserved1_op, mm_reserved2_op, + mm_ori32_op, mm_pool32f_op, mm_pool32s_op, mm_reserved2_op, mm_pool32c_op, mm_lwgp16_op, mm_lw16_op, mm_pool16e_op, mm_xori32_op, mm_jals32_op, mm_addiupc_op, mm_reserved3_op, mm_reserved4_op, mm_pool16f_op, mm_sb16_op, mm_beqz16_op, @@ -360,7 +419,10 @@ enum mm_32axf_minor_op { mm_mflo32_op = 0x075, mm_jalrhb_op = 0x07c, mm_tlbwi_op = 0x08d, + mm_mthi32_op = 0x0b5, mm_tlbwr_op = 0x0cd, + mm_mtlo32_op = 0x0f5, + mm_di_op = 0x11d, mm_jalrs_op = 0x13c, mm_jalrshb_op = 0x17c, mm_sync_op = 0x1ad, @@ -479,6 +541,13 @@ enum mm_32f_73_minor_op { }; /* + * (microMIPS) POOL32S minor opcodes. + */ +enum mm_32s_minor_op { + mm_32s_elm_op = 0x16, +}; + +/* * (microMIPS) POOL16C minor opcodes. */ enum mm_16c_minor_op { @@ -586,6 +655,36 @@ struct r_format { /* Register format */ ;)))))) }; +struct c0r_format { /* C0 register format */ + __BITFIELD_FIELD(unsigned int opcode : 6, + __BITFIELD_FIELD(unsigned int rs : 5, + __BITFIELD_FIELD(unsigned int rt : 5, + __BITFIELD_FIELD(unsigned int rd : 5, + __BITFIELD_FIELD(unsigned int z: 8, + __BITFIELD_FIELD(unsigned int sel : 3, + ;)))))) +}; + +struct mfmc0_format { /* MFMC0 register format */ + __BITFIELD_FIELD(unsigned int opcode : 6, + __BITFIELD_FIELD(unsigned int rs : 5, + __BITFIELD_FIELD(unsigned int rt : 5, + __BITFIELD_FIELD(unsigned int rd : 5, + __BITFIELD_FIELD(unsigned int re : 5, + __BITFIELD_FIELD(unsigned int sc : 1, + __BITFIELD_FIELD(unsigned int : 2, + __BITFIELD_FIELD(unsigned int sel : 3, + ;)))))))) +}; + +struct co_format { /* C0 CO format */ + __BITFIELD_FIELD(unsigned int opcode : 6, + __BITFIELD_FIELD(unsigned int co : 1, + __BITFIELD_FIELD(unsigned int code : 19, + __BITFIELD_FIELD(unsigned int func : 6, + ;)))) +}; + struct p_format { /* Performance counter format (R10000) */ __BITFIELD_FIELD(unsigned int opcode : 6, __BITFIELD_FIELD(unsigned int rs : 5, @@ -937,6 +1036,9 @@ union mips_instruction { struct u_format u_format; struct c_format c_format; struct r_format r_format; + struct c0r_format c0r_format; + struct mfmc0_format mfmc0_format; + struct co_format co_format; struct p_format p_format; struct f_format f_format; struct ma_format ma_format; diff --git a/arch/mips/kernel/asm-offsets.c b/arch/mips/kernel/asm-offsets.c index 1ea973b..fae2f94 100644 --- a/arch/mips/kernel/asm-offsets.c +++ b/arch/mips/kernel/asm-offsets.c @@ -339,71 +339,9 @@ void output_pm_defines(void) } #endif -void output_cpuinfo_defines(void) -{ - COMMENT(" MIPS cpuinfo offsets. "); - DEFINE(CPUINFO_SIZE, sizeof(struct cpuinfo_mips)); -#ifdef CONFIG_MIPS_ASID_BITS_VARIABLE - OFFSET(CPUINFO_ASID_MASK, cpuinfo_mips, asid_mask); -#endif -} - void output_kvm_defines(void) { COMMENT(" KVM/MIPS Specfic offsets. "); - DEFINE(VCPU_ARCH_SIZE, sizeof(struct kvm_vcpu_arch)); - OFFSET(VCPU_RUN, kvm_vcpu, run); - OFFSET(VCPU_HOST_ARCH, kvm_vcpu, arch); - - OFFSET(VCPU_HOST_EBASE, kvm_vcpu_arch, host_ebase); - OFFSET(VCPU_GUEST_EBASE, kvm_vcpu_arch, guest_ebase); - - OFFSET(VCPU_HOST_STACK, kvm_vcpu_arch, host_stack); - OFFSET(VCPU_HOST_GP, kvm_vcpu_arch, host_gp); - - OFFSET(VCPU_HOST_CP0_BADVADDR, kvm_vcpu_arch, host_cp0_badvaddr); - OFFSET(VCPU_HOST_CP0_CAUSE, kvm_vcpu_arch, host_cp0_cause); - OFFSET(VCPU_HOST_EPC, kvm_vcpu_arch, host_cp0_epc); - OFFSET(VCPU_HOST_ENTRYHI, kvm_vcpu_arch, host_cp0_entryhi); - - OFFSET(VCPU_GUEST_INST, kvm_vcpu_arch, guest_inst); - - OFFSET(VCPU_R0, kvm_vcpu_arch, gprs[0]); - OFFSET(VCPU_R1, kvm_vcpu_arch, gprs[1]); - OFFSET(VCPU_R2, kvm_vcpu_arch, gprs[2]); - OFFSET(VCPU_R3, kvm_vcpu_arch, gprs[3]); - OFFSET(VCPU_R4, kvm_vcpu_arch, gprs[4]); - OFFSET(VCPU_R5, kvm_vcpu_arch, gprs[5]); - OFFSET(VCPU_R6, kvm_vcpu_arch, gprs[6]); - OFFSET(VCPU_R7, kvm_vcpu_arch, gprs[7]); - OFFSET(VCPU_R8, kvm_vcpu_arch, gprs[8]); - OFFSET(VCPU_R9, kvm_vcpu_arch, gprs[9]); - OFFSET(VCPU_R10, kvm_vcpu_arch, gprs[10]); - OFFSET(VCPU_R11, kvm_vcpu_arch, gprs[11]); - OFFSET(VCPU_R12, kvm_vcpu_arch, gprs[12]); - OFFSET(VCPU_R13, kvm_vcpu_arch, gprs[13]); - OFFSET(VCPU_R14, kvm_vcpu_arch, gprs[14]); - OFFSET(VCPU_R15, kvm_vcpu_arch, gprs[15]); - OFFSET(VCPU_R16, kvm_vcpu_arch, gprs[16]); - OFFSET(VCPU_R17, kvm_vcpu_arch, gprs[17]); - OFFSET(VCPU_R18, kvm_vcpu_arch, gprs[18]); - OFFSET(VCPU_R19, kvm_vcpu_arch, gprs[19]); - OFFSET(VCPU_R20, kvm_vcpu_arch, gprs[20]); - OFFSET(VCPU_R21, kvm_vcpu_arch, gprs[21]); - OFFSET(VCPU_R22, kvm_vcpu_arch, gprs[22]); - OFFSET(VCPU_R23, kvm_vcpu_arch, gprs[23]); - OFFSET(VCPU_R24, kvm_vcpu_arch, gprs[24]); - OFFSET(VCPU_R25, kvm_vcpu_arch, gprs[25]); - OFFSET(VCPU_R26, kvm_vcpu_arch, gprs[26]); - OFFSET(VCPU_R27, kvm_vcpu_arch, gprs[27]); - OFFSET(VCPU_R28, kvm_vcpu_arch, gprs[28]); - OFFSET(VCPU_R29, kvm_vcpu_arch, gprs[29]); - OFFSET(VCPU_R30, kvm_vcpu_arch, gprs[30]); - OFFSET(VCPU_R31, kvm_vcpu_arch, gprs[31]); - OFFSET(VCPU_LO, kvm_vcpu_arch, lo); - OFFSET(VCPU_HI, kvm_vcpu_arch, hi); - OFFSET(VCPU_PC, kvm_vcpu_arch, pc); - BLANK(); OFFSET(VCPU_FPR0, kvm_vcpu_arch, fpu.fpr[0]); OFFSET(VCPU_FPR1, kvm_vcpu_arch, fpu.fpr[1]); @@ -441,14 +379,6 @@ void output_kvm_defines(void) OFFSET(VCPU_FCR31, kvm_vcpu_arch, fpu.fcr31); OFFSET(VCPU_MSA_CSR, kvm_vcpu_arch, fpu.msacsr); BLANK(); - - OFFSET(VCPU_COP0, kvm_vcpu_arch, cop0); - OFFSET(VCPU_GUEST_KERNEL_ASID, kvm_vcpu_arch, guest_kernel_asid); - OFFSET(VCPU_GUEST_USER_ASID, kvm_vcpu_arch, guest_user_asid); - - OFFSET(COP0_TLB_HI, mips_coproc, reg[MIPS_CP0_TLB_HI][0]); - OFFSET(COP0_STATUS, mips_coproc, reg[MIPS_CP0_STATUS][0]); - BLANK(); } #ifdef CONFIG_MIPS_CPS diff --git a/arch/mips/kernel/branch.c b/arch/mips/kernel/branch.c index 6dc3f1f..46c227f 100644 --- a/arch/mips/kernel/branch.c +++ b/arch/mips/kernel/branch.c @@ -790,7 +790,7 @@ int __compute_return_epc_for_insn(struct pt_regs *regs, epc += 4 + (insn.i_format.simmediate << 2); regs->cp0_epc = epc; break; - case beqzcjic_op: + case pop66_op: if (!cpu_has_mips_r6) { ret = -SIGILL; break; @@ -798,7 +798,7 @@ int __compute_return_epc_for_insn(struct pt_regs *regs, /* Compact branch: BEQZC || JIC */ regs->cp0_epc += 8; break; - case bnezcjialc_op: + case pop76_op: if (!cpu_has_mips_r6) { ret = -SIGILL; break; @@ -809,8 +809,8 @@ int __compute_return_epc_for_insn(struct pt_regs *regs, regs->cp0_epc += 8; break; #endif - case cbcond0_op: - case cbcond1_op: + case pop10_op: + case pop30_op: /* Only valid for MIPS R6 */ if (!cpu_has_mips_r6) { ret = -SIGILL; diff --git a/arch/mips/kernel/traps.c b/arch/mips/kernel/traps.c index 4a1712b..6fb4704 100644 --- a/arch/mips/kernel/traps.c +++ b/arch/mips/kernel/traps.c @@ -619,17 +619,17 @@ static int simulate_rdhwr(struct pt_regs *regs, int rd, int rt) perf_sw_event(PERF_COUNT_SW_EMULATION_FAULTS, 1, regs, 0); switch (rd) { - case 0: /* CPU number */ + case MIPS_HWR_CPUNUM: /* CPU number */ regs->regs[rt] = smp_processor_id(); return 0; - case 1: /* SYNCI length */ + case MIPS_HWR_SYNCISTEP: /* SYNCI length */ regs->regs[rt] = min(current_cpu_data.dcache.linesz, current_cpu_data.icache.linesz); return 0; - case 2: /* Read count register */ + case MIPS_HWR_CC: /* Read count register */ regs->regs[rt] = read_c0_count(); return 0; - case 3: /* Count register resolution */ + case MIPS_HWR_CCRES: /* Count register resolution */ switch (current_cpu_type()) { case CPU_20KC: case CPU_25KF: @@ -639,7 +639,7 @@ static int simulate_rdhwr(struct pt_regs *regs, int rd, int rt) regs->regs[rt] = 2; } return 0; - case 29: + case MIPS_HWR_ULR: /* Read UserLocal register */ regs->regs[rt] = ti->tp_value; return 0; default: @@ -1859,6 +1859,7 @@ void __noreturn nmi_exception_handler(struct pt_regs *regs) #define VECTORSPACING 0x100 /* for EI/VI mode */ unsigned long ebase; +EXPORT_SYMBOL_GPL(ebase); unsigned long exception_handlers[32]; unsigned long vi_handlers[64]; @@ -2063,16 +2064,22 @@ static void configure_status(void) status_set); } +unsigned int hwrena; +EXPORT_SYMBOL_GPL(hwrena); + /* configure HWRENA register */ static void configure_hwrena(void) { - unsigned int hwrena = cpu_hwrena_impl_bits; + hwrena = cpu_hwrena_impl_bits; if (cpu_has_mips_r2_r6) - hwrena |= 0x0000000f; + hwrena |= MIPS_HWRENA_CPUNUM | + MIPS_HWRENA_SYNCISTEP | + MIPS_HWRENA_CC | + MIPS_HWRENA_CCRES; if (!noulri && cpu_has_userlocal) - hwrena |= (1 << 29); + hwrena |= MIPS_HWRENA_ULR; if (hwrena) write_c0_hwrena(hwrena); diff --git a/arch/mips/kvm/Kconfig b/arch/mips/kvm/Kconfig index 2ae1282..7c56d6b 100644 --- a/arch/mips/kvm/Kconfig +++ b/arch/mips/kvm/Kconfig @@ -17,6 +17,7 @@ if VIRTUALIZATION config KVM tristate "Kernel-based Virtual Machine (KVM) support" depends on HAVE_KVM + select EXPORT_UASM select PREEMPT_NOTIFIERS select ANON_INODES select KVM_MMIO diff --git a/arch/mips/kvm/Makefile b/arch/mips/kvm/Makefile index 637ebbe..847429d 100644 --- a/arch/mips/kvm/Makefile +++ b/arch/mips/kvm/Makefile @@ -7,9 +7,10 @@ EXTRA_CFLAGS += -Ivirt/kvm -Iarch/mips/kvm common-objs-$(CONFIG_CPU_HAS_MSA) += msa.o -kvm-objs := $(common-objs-y) mips.o emulate.o locore.o \ +kvm-objs := $(common-objs-y) mips.o emulate.o entry.o \ interrupt.o stats.o commpage.o \ dyntrans.o trap_emul.o fpu.o +kvm-objs += mmu.o obj-$(CONFIG_KVM) += kvm.o obj-y += callback.o tlb.o diff --git a/arch/mips/kvm/commpage.c b/arch/mips/kvm/commpage.c index 2d6e976..a36b77e 100644 --- a/arch/mips/kvm/commpage.c +++ b/arch/mips/kvm/commpage.c @@ -4,7 +4,7 @@ * for more details. * * commpage, currently used for Virtual COP0 registers. - * Mapped into the guest kernel @ 0x0. + * Mapped into the guest kernel @ KVM_GUEST_COMMPAGE_ADDR. * * Copyright (C) 2012 MIPS Technologies, Inc. All rights reserved. * Authors: Sanjay Lal <sanjayl@kymasys.com> diff --git a/arch/mips/kvm/dyntrans.c b/arch/mips/kvm/dyntrans.c index f1527a4..d280894 100644 --- a/arch/mips/kvm/dyntrans.c +++ b/arch/mips/kvm/dyntrans.c @@ -11,6 +11,7 @@ #include <linux/errno.h> #include <linux/err.h> +#include <linux/highmem.h> #include <linux/kvm_host.h> #include <linux/module.h> #include <linux/vmalloc.h> @@ -20,125 +21,114 @@ #include "commpage.h" -#define SYNCI_TEMPLATE 0x041f0000 -#define SYNCI_BASE(x) (((x) >> 21) & 0x1f) -#define SYNCI_OFFSET ((x) & 0xffff) +/** + * kvm_mips_trans_replace() - Replace trapping instruction in guest memory. + * @vcpu: Virtual CPU. + * @opc: PC of instruction to replace. + * @replace: Instruction to write + */ +static int kvm_mips_trans_replace(struct kvm_vcpu *vcpu, u32 *opc, + union mips_instruction replace) +{ + unsigned long paddr, flags; + void *vaddr; + + if (KVM_GUEST_KSEGX((unsigned long)opc) == KVM_GUEST_KSEG0) { + paddr = kvm_mips_translate_guest_kseg0_to_hpa(vcpu, + (unsigned long)opc); + vaddr = kmap_atomic(pfn_to_page(PHYS_PFN(paddr))); + vaddr += paddr & ~PAGE_MASK; + memcpy(vaddr, (void *)&replace, sizeof(u32)); + local_flush_icache_range((unsigned long)vaddr, + (unsigned long)vaddr + 32); + kunmap_atomic(vaddr); + } else if (KVM_GUEST_KSEGX((unsigned long) opc) == KVM_GUEST_KSEG23) { + local_irq_save(flags); + memcpy((void *)opc, (void *)&replace, sizeof(u32)); + local_flush_icache_range((unsigned long)opc, + (unsigned long)opc + 32); + local_irq_restore(flags); + } else { + kvm_err("%s: Invalid address: %p\n", __func__, opc); + return -EFAULT; + } -#define LW_TEMPLATE 0x8c000000 -#define CLEAR_TEMPLATE 0x00000020 -#define SW_TEMPLATE 0xac000000 + return 0; +} -int kvm_mips_trans_cache_index(uint32_t inst, uint32_t *opc, +int kvm_mips_trans_cache_index(union mips_instruction inst, u32 *opc, struct kvm_vcpu *vcpu) { - int result = 0; - unsigned long kseg0_opc; - uint32_t synci_inst = 0x0; + union mips_instruction nop_inst = { 0 }; /* Replace the CACHE instruction, with a NOP */ - kseg0_opc = - CKSEG0ADDR(kvm_mips_translate_guest_kseg0_to_hpa - (vcpu, (unsigned long) opc)); - memcpy((void *)kseg0_opc, (void *)&synci_inst, sizeof(uint32_t)); - local_flush_icache_range(kseg0_opc, kseg0_opc + 32); - - return result; + return kvm_mips_trans_replace(vcpu, opc, nop_inst); } /* * Address based CACHE instructions are transformed into synci(s). A little * heavy for just D-cache invalidates, but avoids an expensive trap */ -int kvm_mips_trans_cache_va(uint32_t inst, uint32_t *opc, +int kvm_mips_trans_cache_va(union mips_instruction inst, u32 *opc, struct kvm_vcpu *vcpu) { - int result = 0; - unsigned long kseg0_opc; - uint32_t synci_inst = SYNCI_TEMPLATE, base, offset; - - base = (inst >> 21) & 0x1f; - offset = inst & 0xffff; - synci_inst |= (base << 21); - synci_inst |= offset; - - kseg0_opc = - CKSEG0ADDR(kvm_mips_translate_guest_kseg0_to_hpa - (vcpu, (unsigned long) opc)); - memcpy((void *)kseg0_opc, (void *)&synci_inst, sizeof(uint32_t)); - local_flush_icache_range(kseg0_opc, kseg0_opc + 32); - - return result; + union mips_instruction synci_inst = { 0 }; + + synci_inst.i_format.opcode = bcond_op; + synci_inst.i_format.rs = inst.i_format.rs; + synci_inst.i_format.rt = synci_op; + if (cpu_has_mips_r6) + synci_inst.i_format.simmediate = inst.spec3_format.simmediate; + else + synci_inst.i_format.simmediate = inst.i_format.simmediate; + + return kvm_mips_trans_replace(vcpu, opc, synci_inst); } -int kvm_mips_trans_mfc0(uint32_t inst, uint32_t *opc, struct kvm_vcpu *vcpu) +int kvm_mips_trans_mfc0(union mips_instruction inst, u32 *opc, + struct kvm_vcpu *vcpu) { - int32_t rt, rd, sel; - uint32_t mfc0_inst; - unsigned long kseg0_opc, flags; - - rt = (inst >> 16) & 0x1f; - rd = (inst >> 11) & 0x1f; - sel = inst & 0x7; + union mips_instruction mfc0_inst = { 0 }; + u32 rd, sel; - if ((rd == MIPS_CP0_ERRCTL) && (sel == 0)) { - mfc0_inst = CLEAR_TEMPLATE; - mfc0_inst |= ((rt & 0x1f) << 16); - } else { - mfc0_inst = LW_TEMPLATE; - mfc0_inst |= ((rt & 0x1f) << 16); - mfc0_inst |= offsetof(struct kvm_mips_commpage, - cop0.reg[rd][sel]); - } + rd = inst.c0r_format.rd; + sel = inst.c0r_format.sel; - if (KVM_GUEST_KSEGX(opc) == KVM_GUEST_KSEG0) { - kseg0_opc = - CKSEG0ADDR(kvm_mips_translate_guest_kseg0_to_hpa - (vcpu, (unsigned long) opc)); - memcpy((void *)kseg0_opc, (void *)&mfc0_inst, sizeof(uint32_t)); - local_flush_icache_range(kseg0_opc, kseg0_opc + 32); - } else if (KVM_GUEST_KSEGX((unsigned long) opc) == KVM_GUEST_KSEG23) { - local_irq_save(flags); - memcpy((void *)opc, (void *)&mfc0_inst, sizeof(uint32_t)); - local_flush_icache_range((unsigned long)opc, - (unsigned long)opc + 32); - local_irq_restore(flags); + if (rd == MIPS_CP0_ERRCTL && sel == 0) { + mfc0_inst.r_format.opcode = spec_op; + mfc0_inst.r_format.rd = inst.c0r_format.rt; + mfc0_inst.r_format.func = add_op; } else { - kvm_err("%s: Invalid address: %p\n", __func__, opc); - return -EFAULT; + mfc0_inst.i_format.opcode = lw_op; + mfc0_inst.i_format.rt = inst.c0r_format.rt; + mfc0_inst.i_format.simmediate = KVM_GUEST_COMMPAGE_ADDR | + offsetof(struct kvm_mips_commpage, cop0.reg[rd][sel]); +#ifdef CONFIG_CPU_BIG_ENDIAN + if (sizeof(vcpu->arch.cop0->reg[0][0]) == 8) + mfc0_inst.i_format.simmediate |= 4; +#endif } - return 0; + return kvm_mips_trans_replace(vcpu, opc, mfc0_inst); } -int kvm_mips_trans_mtc0(uint32_t inst, uint32_t *opc, struct kvm_vcpu *vcpu) +int kvm_mips_trans_mtc0(union mips_instruction inst, u32 *opc, + struct kvm_vcpu *vcpu) { - int32_t rt, rd, sel; - uint32_t mtc0_inst = SW_TEMPLATE; - unsigned long kseg0_opc, flags; - - rt = (inst >> 16) & 0x1f; - rd = (inst >> 11) & 0x1f; - sel = inst & 0x7; - - mtc0_inst |= ((rt & 0x1f) << 16); - mtc0_inst |= offsetof(struct kvm_mips_commpage, cop0.reg[rd][sel]); - - if (KVM_GUEST_KSEGX(opc) == KVM_GUEST_KSEG0) { - kseg0_opc = - CKSEG0ADDR(kvm_mips_translate_guest_kseg0_to_hpa - (vcpu, (unsigned long) opc)); - memcpy((void *)kseg0_opc, (void *)&mtc0_inst, sizeof(uint32_t)); - local_flush_icache_range(kseg0_opc, kseg0_opc + 32); - } else if (KVM_GUEST_KSEGX((unsigned long) opc) == KVM_GUEST_KSEG23) { - local_irq_save(flags); - memcpy((void *)opc, (void *)&mtc0_inst, sizeof(uint32_t)); - local_flush_icache_range((unsigned long)opc, - (unsigned long)opc + 32); - local_irq_restore(flags); - } else { - kvm_err("%s: Invalid address: %p\n", __func__, opc); - return -EFAULT; - } - - return 0; + union mips_instruction mtc0_inst = { 0 }; + u32 rd, sel; + + rd = inst.c0r_format.rd; + sel = inst.c0r_format.sel; + + mtc0_inst.i_format.opcode = sw_op; + mtc0_inst.i_format.rt = inst.c0r_format.rt; + mtc0_inst.i_format.simmediate = KVM_GUEST_COMMPAGE_ADDR | + offsetof(struct kvm_mips_commpage, cop0.reg[rd][sel]); +#ifdef CONFIG_CPU_BIG_ENDIAN + if (sizeof(vcpu->arch.cop0->reg[0][0]) == 8) + mtc0_inst.i_format.simmediate |= 4; +#endif + + return kvm_mips_trans_replace(vcpu, opc, mtc0_inst); } diff --git a/arch/mips/kvm/emulate.c b/arch/mips/kvm/emulate.c index 645c8a1..6eb52b9 100644 --- a/arch/mips/kvm/emulate.c +++ b/arch/mips/kvm/emulate.c @@ -52,7 +52,7 @@ unsigned long kvm_compute_return_epc(struct kvm_vcpu *vcpu, goto unaligned; /* Read the instruction */ - insn.word = kvm_get_inst((uint32_t *) epc, vcpu); + insn.word = kvm_get_inst((u32 *) epc, vcpu); if (insn.word == KVM_INVALID_INST) return KVM_INVALID_INST; @@ -161,9 +161,12 @@ unsigned long kvm_compute_return_epc(struct kvm_vcpu *vcpu, nextpc = epc; break; - case blez_op: /* not really i_format */ - case blezl_op: - /* rt field assumed to be zero */ + case blez_op: /* POP06 */ +#ifndef CONFIG_CPU_MIPSR6 + case blezl_op: /* removed in R6 */ +#endif + if (insn.i_format.rt != 0) + goto compact_branch; if ((long)arch->gprs[insn.i_format.rs] <= 0) epc = epc + 4 + (insn.i_format.simmediate << 2); else @@ -171,9 +174,12 @@ unsigned long kvm_compute_return_epc(struct kvm_vcpu *vcpu, nextpc = epc; break; - case bgtz_op: - case bgtzl_op: - /* rt field assumed to be zero */ + case bgtz_op: /* POP07 */ +#ifndef CONFIG_CPU_MIPSR6 + case bgtzl_op: /* removed in R6 */ +#endif + if (insn.i_format.rt != 0) + goto compact_branch; if ((long)arch->gprs[insn.i_format.rs] > 0) epc = epc + 4 + (insn.i_format.simmediate << 2); else @@ -185,6 +191,40 @@ unsigned long kvm_compute_return_epc(struct kvm_vcpu *vcpu, case cop1_op: kvm_err("%s: unsupported cop1_op\n", __func__); break; + +#ifdef CONFIG_CPU_MIPSR6 + /* R6 added the following compact branches with forbidden slots */ + case blezl_op: /* POP26 */ + case bgtzl_op: /* POP27 */ + /* only rt == 0 isn't compact branch */ + if (insn.i_format.rt != 0) + goto compact_branch; + break; + case pop10_op: + case pop30_op: + /* only rs == rt == 0 is reserved, rest are compact branches */ + if (insn.i_format.rs != 0 || insn.i_format.rt != 0) + goto compact_branch; + break; + case pop66_op: + case pop76_op: + /* only rs == 0 isn't compact branch */ + if (insn.i_format.rs != 0) + goto compact_branch; + break; +compact_branch: + /* + * If we've hit an exception on the forbidden slot, then + * the branch must not have been taken. + */ + epc += 8; + nextpc = epc; + break; +#else +compact_branch: + /* Compact branches not supported before R6 */ + break; +#endif } return nextpc; @@ -198,7 +238,7 @@ sigill: return nextpc; } -enum emulation_result update_pc(struct kvm_vcpu *vcpu, uint32_t cause) +enum emulation_result update_pc(struct kvm_vcpu *vcpu, u32 cause) { unsigned long branch_pc; enum emulation_result er = EMULATE_DONE; @@ -243,7 +283,7 @@ static inline int kvm_mips_count_disabled(struct kvm_vcpu *vcpu) * * Assumes !kvm_mips_count_disabled(@vcpu) (guest CP0_Count timer is running). */ -static uint32_t kvm_mips_ktime_to_count(struct kvm_vcpu *vcpu, ktime_t now) +static u32 kvm_mips_ktime_to_count(struct kvm_vcpu *vcpu, ktime_t now) { s64 now_ns, periods; u64 delta; @@ -300,11 +340,11 @@ static inline ktime_t kvm_mips_count_time(struct kvm_vcpu *vcpu) * * Returns: The current value of the guest CP0_Count register. */ -static uint32_t kvm_mips_read_count_running(struct kvm_vcpu *vcpu, ktime_t now) +static u32 kvm_mips_read_count_running(struct kvm_vcpu *vcpu, ktime_t now) { struct mips_coproc *cop0 = vcpu->arch.cop0; ktime_t expires, threshold; - uint32_t count, compare; + u32 count, compare; int running; /* Calculate the biased and scaled guest CP0_Count */ @@ -315,7 +355,7 @@ static uint32_t kvm_mips_read_count_running(struct kvm_vcpu *vcpu, ktime_t now) * Find whether CP0_Count has reached the closest timer interrupt. If * not, we shouldn't inject it. */ - if ((int32_t)(count - compare) < 0) + if ((s32)(count - compare) < 0) return count; /* @@ -360,7 +400,7 @@ static uint32_t kvm_mips_read_count_running(struct kvm_vcpu *vcpu, ktime_t now) * * Returns: The current guest CP0_Count value. */ -uint32_t kvm_mips_read_count(struct kvm_vcpu *vcpu) +u32 kvm_mips_read_count(struct kvm_vcpu *vcpu) { struct mips_coproc *cop0 = vcpu->arch.cop0; @@ -387,8 +427,7 @@ uint32_t kvm_mips_read_count(struct kvm_vcpu *vcpu) * * Returns: The ktime at the point of freeze. */ -static ktime_t kvm_mips_freeze_hrtimer(struct kvm_vcpu *vcpu, - uint32_t *count) +static ktime_t kvm_mips_freeze_hrtimer(struct kvm_vcpu *vcpu, u32 *count) { ktime_t now; @@ -419,16 +458,16 @@ static ktime_t kvm_mips_freeze_hrtimer(struct kvm_vcpu *vcpu, * Assumes !kvm_mips_count_disabled(@vcpu) (guest CP0_Count timer is running). */ static void kvm_mips_resume_hrtimer(struct kvm_vcpu *vcpu, - ktime_t now, uint32_t count) + ktime_t now, u32 count) { struct mips_coproc *cop0 = vcpu->arch.cop0; - uint32_t compare; + u32 compare; u64 delta; ktime_t expire; /* Calculate timeout (wrap 0 to 2^32) */ compare = kvm_read_c0_guest_compare(cop0); - delta = (u64)(uint32_t)(compare - count - 1) + 1; + delta = (u64)(u32)(compare - count - 1) + 1; delta = div_u64(delta * NSEC_PER_SEC, vcpu->arch.count_hz); expire = ktime_add_ns(now, delta); @@ -444,7 +483,7 @@ static void kvm_mips_resume_hrtimer(struct kvm_vcpu *vcpu, * * Sets the CP0_Count value and updates the timer accordingly. */ -void kvm_mips_write_count(struct kvm_vcpu *vcpu, uint32_t count) +void kvm_mips_write_count(struct kvm_vcpu *vcpu, u32 count) { struct mips_coproc *cop0 = vcpu->arch.cop0; ktime_t now; @@ -538,13 +577,13 @@ int kvm_mips_set_count_hz(struct kvm_vcpu *vcpu, s64 count_hz) * If @ack, atomically acknowledge any pending timer interrupt, otherwise ensure * any pending timer interrupt is preserved. */ -void kvm_mips_write_compare(struct kvm_vcpu *vcpu, uint32_t compare, bool ack) +void kvm_mips_write_compare(struct kvm_vcpu *vcpu, u32 compare, bool ack) { struct mips_coproc *cop0 = vcpu->arch.cop0; int dc; u32 old_compare = kvm_read_c0_guest_compare(cop0); ktime_t now; - uint32_t count; + u32 count; /* if unchanged, must just be an ack */ if (old_compare == compare) { @@ -585,7 +624,7 @@ void kvm_mips_write_compare(struct kvm_vcpu *vcpu, uint32_t compare, bool ack) static ktime_t kvm_mips_count_disable(struct kvm_vcpu *vcpu) { struct mips_coproc *cop0 = vcpu->arch.cop0; - uint32_t count; + u32 count; ktime_t now; /* Stop hrtimer */ @@ -632,7 +671,7 @@ void kvm_mips_count_disable_cause(struct kvm_vcpu *vcpu) void kvm_mips_count_enable_cause(struct kvm_vcpu *vcpu) { struct mips_coproc *cop0 = vcpu->arch.cop0; - uint32_t count; + u32 count; kvm_clear_c0_guest_cause(cop0, CAUSEF_DC); @@ -661,7 +700,7 @@ int kvm_mips_set_count_ctl(struct kvm_vcpu *vcpu, s64 count_ctl) s64 changed = count_ctl ^ vcpu->arch.count_ctl; s64 delta; ktime_t expire, now; - uint32_t count, compare; + u32 count, compare; /* Only allow defined bits to be changed */ if (changed & ~(s64)(KVM_REG_MIPS_COUNT_CTL_DC)) @@ -687,7 +726,7 @@ int kvm_mips_set_count_ctl(struct kvm_vcpu *vcpu, s64 count_ctl) */ count = kvm_read_c0_guest_count(cop0); compare = kvm_read_c0_guest_compare(cop0); - delta = (u64)(uint32_t)(compare - count - 1) + 1; + delta = (u64)(u32)(compare - count - 1) + 1; delta = div_u64(delta * NSEC_PER_SEC, vcpu->arch.count_hz); expire = ktime_add_ns(vcpu->arch.count_resume, delta); @@ -776,7 +815,7 @@ enum emulation_result kvm_mips_emul_wait(struct kvm_vcpu *vcpu) vcpu->arch.pending_exceptions); ++vcpu->stat.wait_exits; - trace_kvm_exit(vcpu, WAIT_EXITS); + trace_kvm_exit(vcpu, KVM_TRACE_EXIT_WAIT); if (!vcpu->arch.pending_exceptions) { vcpu->arch.wait = 1; kvm_vcpu_block(vcpu); @@ -801,9 +840,9 @@ enum emulation_result kvm_mips_emul_wait(struct kvm_vcpu *vcpu) enum emulation_result kvm_mips_emul_tlbr(struct kvm_vcpu *vcpu) { struct mips_coproc *cop0 = vcpu->arch.cop0; - uint32_t pc = vcpu->arch.pc; + unsigned long pc = vcpu->arch.pc; - kvm_err("[%#x] COP0_TLBR [%ld]\n", pc, kvm_read_c0_guest_index(cop0)); + kvm_err("[%#lx] COP0_TLBR [%ld]\n", pc, kvm_read_c0_guest_index(cop0)); return EMULATE_FAIL; } @@ -813,11 +852,11 @@ enum emulation_result kvm_mips_emul_tlbwi(struct kvm_vcpu *vcpu) struct mips_coproc *cop0 = vcpu->arch.cop0; int index = kvm_read_c0_guest_index(cop0); struct kvm_mips_tlb *tlb = NULL; - uint32_t pc = vcpu->arch.pc; + unsigned long pc = vcpu->arch.pc; if (index < 0 || index >= KVM_MIPS_GUEST_TLB_SIZE) { kvm_debug("%s: illegal index: %d\n", __func__, index); - kvm_debug("[%#x] COP0_TLBWI [%d] (entryhi: %#lx, entrylo0: %#lx entrylo1: %#lx, mask: %#lx)\n", + kvm_debug("[%#lx] COP0_TLBWI [%d] (entryhi: %#lx, entrylo0: %#lx entrylo1: %#lx, mask: %#lx)\n", pc, index, kvm_read_c0_guest_entryhi(cop0), kvm_read_c0_guest_entrylo0(cop0), kvm_read_c0_guest_entrylo1(cop0), @@ -834,10 +873,10 @@ enum emulation_result kvm_mips_emul_tlbwi(struct kvm_vcpu *vcpu) tlb->tlb_mask = kvm_read_c0_guest_pagemask(cop0); tlb->tlb_hi = kvm_read_c0_guest_entryhi(cop0); - tlb->tlb_lo0 = kvm_read_c0_guest_entrylo0(cop0); - tlb->tlb_lo1 = kvm_read_c0_guest_entrylo1(cop0); + tlb->tlb_lo[0] = kvm_read_c0_guest_entrylo0(cop0); + tlb->tlb_lo[1] = kvm_read_c0_guest_entrylo1(cop0); - kvm_debug("[%#x] COP0_TLBWI [%d] (entryhi: %#lx, entrylo0: %#lx entrylo1: %#lx, mask: %#lx)\n", + kvm_debug("[%#lx] COP0_TLBWI [%d] (entryhi: %#lx, entrylo0: %#lx entrylo1: %#lx, mask: %#lx)\n", pc, index, kvm_read_c0_guest_entryhi(cop0), kvm_read_c0_guest_entrylo0(cop0), kvm_read_c0_guest_entrylo1(cop0), @@ -851,7 +890,7 @@ enum emulation_result kvm_mips_emul_tlbwr(struct kvm_vcpu *vcpu) { struct mips_coproc *cop0 = vcpu->arch.cop0; struct kvm_mips_tlb *tlb = NULL; - uint32_t pc = vcpu->arch.pc; + unsigned long pc = vcpu->arch.pc; int index; get_random_bytes(&index, sizeof(index)); @@ -867,10 +906,10 @@ enum emulation_result kvm_mips_emul_tlbwr(struct kvm_vcpu *vcpu) tlb->tlb_mask = kvm_read_c0_guest_pagemask(cop0); tlb->tlb_hi = kvm_read_c0_guest_entryhi(cop0); - tlb->tlb_lo0 = kvm_read_c0_guest_entrylo0(cop0); - tlb->tlb_lo1 = kvm_read_c0_guest_entrylo1(cop0); + tlb->tlb_lo[0] = kvm_read_c0_guest_entrylo0(cop0); + tlb->tlb_lo[1] = kvm_read_c0_guest_entrylo1(cop0); - kvm_debug("[%#x] COP0_TLBWR[%d] (entryhi: %#lx, entrylo0: %#lx entrylo1: %#lx)\n", + kvm_debug("[%#lx] COP0_TLBWR[%d] (entryhi: %#lx, entrylo0: %#lx entrylo1: %#lx)\n", pc, index, kvm_read_c0_guest_entryhi(cop0), kvm_read_c0_guest_entrylo0(cop0), kvm_read_c0_guest_entrylo1(cop0)); @@ -882,14 +921,14 @@ enum emulation_result kvm_mips_emul_tlbp(struct kvm_vcpu *vcpu) { struct mips_coproc *cop0 = vcpu->arch.cop0; long entryhi = kvm_read_c0_guest_entryhi(cop0); - uint32_t pc = vcpu->arch.pc; + unsigned long pc = vcpu->arch.pc; int index = -1; index = kvm_mips_guest_tlb_lookup(vcpu, entryhi); kvm_write_c0_guest_index(cop0, index); - kvm_debug("[%#x] COP0_TLBP (entryhi: %#lx), index: %d\n", pc, entryhi, + kvm_debug("[%#lx] COP0_TLBP (entryhi: %#lx), index: %d\n", pc, entryhi, index); return EMULATE_DONE; @@ -922,8 +961,8 @@ unsigned int kvm_mips_config1_wrmask(struct kvm_vcpu *vcpu) */ unsigned int kvm_mips_config3_wrmask(struct kvm_vcpu *vcpu) { - /* Config4 is optional */ - unsigned int mask = MIPS_CONF_M; + /* Config4 and ULRI are optional */ + unsigned int mask = MIPS_CONF_M | MIPS_CONF3_ULRI; /* Permit MSA to be present if MSA is supported */ if (kvm_mips_guest_can_have_msa(&vcpu->arch)) @@ -942,7 +981,12 @@ unsigned int kvm_mips_config3_wrmask(struct kvm_vcpu *vcpu) unsigned int kvm_mips_config4_wrmask(struct kvm_vcpu *vcpu) { /* Config5 is optional */ - return MIPS_CONF_M; + unsigned int mask = MIPS_CONF_M; + + /* KScrExist */ + mask |= (unsigned int)vcpu->arch.kscratch_enabled << 16; + + return mask; } /** @@ -973,14 +1017,14 @@ unsigned int kvm_mips_config5_wrmask(struct kvm_vcpu *vcpu) return mask; } -enum emulation_result kvm_mips_emulate_CP0(uint32_t inst, uint32_t *opc, - uint32_t cause, struct kvm_run *run, +enum emulation_result kvm_mips_emulate_CP0(union mips_instruction inst, + u32 *opc, u32 cause, + struct kvm_run *run, struct kvm_vcpu *vcpu) { struct mips_coproc *cop0 = vcpu->arch.cop0; enum emulation_result er = EMULATE_DONE; - int32_t rt, rd, copz, sel, co_bit, op; - uint32_t pc = vcpu->arch.pc; + u32 rt, rd, sel; unsigned long curr_pc; /* @@ -992,16 +1036,8 @@ enum emulation_result kvm_mips_emulate_CP0(uint32_t inst, uint32_t *opc, if (er == EMULATE_FAIL) return er; - copz = (inst >> 21) & 0x1f; - rt = (inst >> 16) & 0x1f; - rd = (inst >> 11) & 0x1f; - sel = inst & 0x7; - co_bit = (inst >> 25) & 1; - - if (co_bit) { - op = (inst) & 0xff; - - switch (op) { + if (inst.co_format.co) { + switch (inst.co_format.func) { case tlbr_op: /* Read indexed TLB entry */ er = kvm_mips_emul_tlbr(vcpu); break; @@ -1020,47 +1056,58 @@ enum emulation_result kvm_mips_emulate_CP0(uint32_t inst, uint32_t *opc, case eret_op: er = kvm_mips_emul_eret(vcpu); goto dont_update_pc; - break; case wait_op: er = kvm_mips_emul_wait(vcpu); break; } } else { - switch (copz) { + rt = inst.c0r_format.rt; + rd = inst.c0r_format.rd; + sel = inst.c0r_format.sel; + + switch (inst.c0r_format.rs) { case mfc_op: #ifdef CONFIG_KVM_MIPS_DEBUG_COP0_COUNTERS cop0->stat[rd][sel]++; #endif /* Get reg */ if ((rd == MIPS_CP0_COUNT) && (sel == 0)) { - vcpu->arch.gprs[rt] = kvm_mips_read_count(vcpu); + vcpu->arch.gprs[rt] = + (s32)kvm_mips_read_count(vcpu); } else if ((rd == MIPS_CP0_ERRCTL) && (sel == 0)) { vcpu->arch.gprs[rt] = 0x0; #ifdef CONFIG_KVM_MIPS_DYN_TRANS kvm_mips_trans_mfc0(inst, opc, vcpu); #endif } else { - vcpu->arch.gprs[rt] = cop0->reg[rd][sel]; + vcpu->arch.gprs[rt] = (s32)cop0->reg[rd][sel]; #ifdef CONFIG_KVM_MIPS_DYN_TRANS kvm_mips_trans_mfc0(inst, opc, vcpu); #endif } - kvm_debug - ("[%#x] MFCz[%d][%d], vcpu->arch.gprs[%d]: %#lx\n", - pc, rd, sel, rt, vcpu->arch.gprs[rt]); - + trace_kvm_hwr(vcpu, KVM_TRACE_MFC0, + KVM_TRACE_COP0(rd, sel), + vcpu->arch.gprs[rt]); break; case dmfc_op: vcpu->arch.gprs[rt] = cop0->reg[rd][sel]; + + trace_kvm_hwr(vcpu, KVM_TRACE_DMFC0, + KVM_TRACE_COP0(rd, sel), + vcpu->arch.gprs[rt]); break; case mtc_op: #ifdef CONFIG_KVM_MIPS_DEBUG_COP0_COUNTERS cop0->stat[rd][sel]++; #endif + trace_kvm_hwr(vcpu, KVM_TRACE_MTC0, + KVM_TRACE_COP0(rd, sel), + vcpu->arch.gprs[rt]); + if ((rd == MIPS_CP0_TLB_INDEX) && (vcpu->arch.gprs[rt] >= KVM_MIPS_GUEST_TLB_SIZE)) { @@ -1078,16 +1125,15 @@ enum emulation_result kvm_mips_emulate_CP0(uint32_t inst, uint32_t *opc, kvm_err("MTCz, cop0->reg[EBASE]: %#lx\n", kvm_read_c0_guest_ebase(cop0)); } else if (rd == MIPS_CP0_TLB_HI && sel == 0) { - uint32_t nasid = + u32 nasid = vcpu->arch.gprs[rt] & KVM_ENTRYHI_ASID; if ((KSEGX(vcpu->arch.gprs[rt]) != CKSEG0) && ((kvm_read_c0_guest_entryhi(cop0) & KVM_ENTRYHI_ASID) != nasid)) { - kvm_debug("MTCz, change ASID from %#lx to %#lx\n", + trace_kvm_asid_change(vcpu, kvm_read_c0_guest_entryhi(cop0) - & KVM_ENTRYHI_ASID, - vcpu->arch.gprs[rt] - & KVM_ENTRYHI_ASID); + & KVM_ENTRYHI_ASID, + nasid); /* Blow away the shadow host TLBs */ kvm_mips_flush_host_tlb(1); @@ -1100,10 +1146,6 @@ enum emulation_result kvm_mips_emulate_CP0(uint32_t inst, uint32_t *opc, kvm_mips_write_count(vcpu, vcpu->arch.gprs[rt]); goto done; } else if ((rd == MIPS_CP0_COMPARE) && (sel == 0)) { - kvm_debug("[%#x] MTCz, COMPARE %#lx <- %#lx\n", - pc, kvm_read_c0_guest_compare(cop0), - vcpu->arch.gprs[rt]); - /* If we are writing to COMPARE */ /* Clear pending timer interrupt, if any */ kvm_mips_write_compare(vcpu, @@ -1155,7 +1197,7 @@ enum emulation_result kvm_mips_emulate_CP0(uint32_t inst, uint32_t *opc, * it first. */ if (change & ST0_CU1 && !(val & ST0_FR) && - vcpu->arch.fpu_inuse & KVM_MIPS_FPU_MSA) + vcpu->arch.aux_inuse & KVM_MIPS_AUX_MSA) kvm_lose_fpu(vcpu); /* @@ -1166,7 +1208,7 @@ enum emulation_result kvm_mips_emulate_CP0(uint32_t inst, uint32_t *opc, * the near future. */ if (change & ST0_CU1 && - vcpu->arch.fpu_inuse & KVM_MIPS_FPU_FPU) + vcpu->arch.aux_inuse & KVM_MIPS_AUX_FPU) change_c0_status(ST0_CU1, val); preempt_enable(); @@ -1201,7 +1243,7 @@ enum emulation_result kvm_mips_emulate_CP0(uint32_t inst, uint32_t *opc, * context is already loaded. */ if (change & MIPS_CONF5_FRE && - vcpu->arch.fpu_inuse & KVM_MIPS_FPU_FPU) + vcpu->arch.aux_inuse & KVM_MIPS_AUX_FPU) change_c0_config5(MIPS_CONF5_FRE, val); /* @@ -1211,7 +1253,7 @@ enum emulation_result kvm_mips_emulate_CP0(uint32_t inst, uint32_t *opc, * quickly enabled again in the near future. */ if (change & MIPS_CONF5_MSAEN && - vcpu->arch.fpu_inuse & KVM_MIPS_FPU_MSA) + vcpu->arch.aux_inuse & KVM_MIPS_AUX_MSA) change_c0_config5(MIPS_CONF5_MSAEN, val); @@ -1219,7 +1261,7 @@ enum emulation_result kvm_mips_emulate_CP0(uint32_t inst, uint32_t *opc, kvm_write_c0_guest_config5(cop0, val); } else if ((rd == MIPS_CP0_CAUSE) && (sel == 0)) { - uint32_t old_cause, new_cause; + u32 old_cause, new_cause; old_cause = kvm_read_c0_guest_cause(cop0); new_cause = vcpu->arch.gprs[rt]; @@ -1233,20 +1275,30 @@ enum emulation_result kvm_mips_emulate_CP0(uint32_t inst, uint32_t *opc, else kvm_mips_count_enable_cause(vcpu); } + } else if ((rd == MIPS_CP0_HWRENA) && (sel == 0)) { + u32 mask = MIPS_HWRENA_CPUNUM | + MIPS_HWRENA_SYNCISTEP | + MIPS_HWRENA_CC | + MIPS_HWRENA_CCRES; + + if (kvm_read_c0_guest_config3(cop0) & + MIPS_CONF3_ULRI) + mask |= MIPS_HWRENA_ULR; + cop0->reg[rd][sel] = vcpu->arch.gprs[rt] & mask; } else { cop0->reg[rd][sel] = vcpu->arch.gprs[rt]; #ifdef CONFIG_KVM_MIPS_DYN_TRANS kvm_mips_trans_mtc0(inst, opc, vcpu); #endif } - - kvm_debug("[%#x] MTCz, cop0->reg[%d][%d]: %#lx\n", pc, - rd, sel, cop0->reg[rd][sel]); break; case dmtc_op: kvm_err("!!!!!!![%#lx]dmtc_op: rt: %d, rd: %d, sel: %d!!!!!!\n", vcpu->arch.pc, rt, rd, sel); + trace_kvm_hwr(vcpu, KVM_TRACE_DMTC0, + KVM_TRACE_COP0(rd, sel), + vcpu->arch.gprs[rt]); er = EMULATE_FAIL; break; @@ -1258,7 +1310,7 @@ enum emulation_result kvm_mips_emulate_CP0(uint32_t inst, uint32_t *opc, vcpu->arch.gprs[rt] = kvm_read_c0_guest_status(cop0); /* EI */ - if (inst & 0x20) { + if (inst.mfmc0_format.sc) { kvm_debug("[%#lx] mfmc0_op: EI\n", vcpu->arch.pc); kvm_set_c0_guest_status(cop0, ST0_IE); @@ -1272,9 +1324,8 @@ enum emulation_result kvm_mips_emulate_CP0(uint32_t inst, uint32_t *opc, case wrpgpr_op: { - uint32_t css = - cop0->reg[MIPS_CP0_STATUS][2] & 0xf; - uint32_t pss = + u32 css = cop0->reg[MIPS_CP0_STATUS][2] & 0xf; + u32 pss = (cop0->reg[MIPS_CP0_STATUS][2] >> 6) & 0xf; /* * We don't support any shadow register sets, so @@ -1291,7 +1342,7 @@ enum emulation_result kvm_mips_emulate_CP0(uint32_t inst, uint32_t *opc, break; default: kvm_err("[%#lx]MachEmulateCP0: unsupported COP0, copz: 0x%x\n", - vcpu->arch.pc, copz); + vcpu->arch.pc, inst.c0r_format.rs); er = EMULATE_FAIL; break; } @@ -1312,13 +1363,14 @@ dont_update_pc: return er; } -enum emulation_result kvm_mips_emulate_store(uint32_t inst, uint32_t cause, +enum emulation_result kvm_mips_emulate_store(union mips_instruction inst, + u32 cause, struct kvm_run *run, struct kvm_vcpu *vcpu) { enum emulation_result er = EMULATE_DO_MMIO; - int32_t op, base, rt, offset; - uint32_t bytes; + u32 rt; + u32 bytes; void *data = run->mmio.data; unsigned long curr_pc; @@ -1331,12 +1383,9 @@ enum emulation_result kvm_mips_emulate_store(uint32_t inst, uint32_t cause, if (er == EMULATE_FAIL) return er; - rt = (inst >> 16) & 0x1f; - base = (inst >> 21) & 0x1f; - offset = inst & 0xffff; - op = (inst >> 26) & 0x3f; + rt = inst.i_format.rt; - switch (op) { + switch (inst.i_format.opcode) { case sb_op: bytes = 1; if (bytes > sizeof(run->mmio.data)) { @@ -1357,7 +1406,7 @@ enum emulation_result kvm_mips_emulate_store(uint32_t inst, uint32_t cause, *(u8 *) data = vcpu->arch.gprs[rt]; kvm_debug("OP_SB: eaddr: %#lx, gpr: %#lx, data: %#x\n", vcpu->arch.host_cp0_badvaddr, vcpu->arch.gprs[rt], - *(uint8_t *) data); + *(u8 *) data); break; @@ -1379,11 +1428,11 @@ enum emulation_result kvm_mips_emulate_store(uint32_t inst, uint32_t cause, run->mmio.is_write = 1; vcpu->mmio_needed = 1; vcpu->mmio_is_write = 1; - *(uint32_t *) data = vcpu->arch.gprs[rt]; + *(u32 *) data = vcpu->arch.gprs[rt]; kvm_debug("[%#lx] OP_SW: eaddr: %#lx, gpr: %#lx, data: %#x\n", vcpu->arch.pc, vcpu->arch.host_cp0_badvaddr, - vcpu->arch.gprs[rt], *(uint32_t *) data); + vcpu->arch.gprs[rt], *(u32 *) data); break; case sh_op: @@ -1404,15 +1453,16 @@ enum emulation_result kvm_mips_emulate_store(uint32_t inst, uint32_t cause, run->mmio.is_write = 1; vcpu->mmio_needed = 1; vcpu->mmio_is_write = 1; - *(uint16_t *) data = vcpu->arch.gprs[rt]; + *(u16 *) data = vcpu->arch.gprs[rt]; kvm_debug("[%#lx] OP_SH: eaddr: %#lx, gpr: %#lx, data: %#x\n", vcpu->arch.pc, vcpu->arch.host_cp0_badvaddr, - vcpu->arch.gprs[rt], *(uint32_t *) data); + vcpu->arch.gprs[rt], *(u32 *) data); break; default: - kvm_err("Store not yet supported"); + kvm_err("Store not yet supported (inst=0x%08x)\n", + inst.word); er = EMULATE_FAIL; break; } @@ -1424,18 +1474,16 @@ enum emulation_result kvm_mips_emulate_store(uint32_t inst, uint32_t cause, return er; } -enum emulation_result kvm_mips_emulate_load(uint32_t inst, uint32_t cause, - struct kvm_run *run, +enum emulation_result kvm_mips_emulate_load(union mips_instruction inst, + u32 cause, struct kvm_run *run, struct kvm_vcpu *vcpu) { enum emulation_result er = EMULATE_DO_MMIO; - int32_t op, base, rt, offset; - uint32_t bytes; + u32 op, rt; + u32 bytes; - rt = (inst >> 16) & 0x1f; - base = (inst >> 21) & 0x1f; - offset = inst & 0xffff; - op = (inst >> 26) & 0x3f; + rt = inst.i_format.rt; + op = inst.i_format.opcode; vcpu->arch.pending_load_cause = cause; vcpu->arch.io_gpr = rt; @@ -1521,7 +1569,8 @@ enum emulation_result kvm_mips_emulate_load(uint32_t inst, uint32_t cause, break; default: - kvm_err("Load not yet supported"); + kvm_err("Load not yet supported (inst=0x%08x)\n", + inst.word); er = EMULATE_FAIL; break; } @@ -1529,40 +1578,15 @@ enum emulation_result kvm_mips_emulate_load(uint32_t inst, uint32_t cause, return er; } -int kvm_mips_sync_icache(unsigned long va, struct kvm_vcpu *vcpu) -{ - unsigned long offset = (va & ~PAGE_MASK); - struct kvm *kvm = vcpu->kvm; - unsigned long pa; - gfn_t gfn; - kvm_pfn_t pfn; - - gfn = va >> PAGE_SHIFT; - - if (gfn >= kvm->arch.guest_pmap_npages) { - kvm_err("%s: Invalid gfn: %#llx\n", __func__, gfn); - kvm_mips_dump_host_tlbs(); - kvm_arch_vcpu_dump_regs(vcpu); - return -1; - } - pfn = kvm->arch.guest_pmap[gfn]; - pa = (pfn << PAGE_SHIFT) | offset; - - kvm_debug("%s: va: %#lx, unmapped: %#x\n", __func__, va, - CKSEG0ADDR(pa)); - - local_flush_icache_range(CKSEG0ADDR(pa), 32); - return 0; -} - -enum emulation_result kvm_mips_emulate_cache(uint32_t inst, uint32_t *opc, - uint32_t cause, +enum emulation_result kvm_mips_emulate_cache(union mips_instruction inst, + u32 *opc, u32 cause, struct kvm_run *run, struct kvm_vcpu *vcpu) { struct mips_coproc *cop0 = vcpu->arch.cop0; enum emulation_result er = EMULATE_DONE; - int32_t offset, cache, op_inst, op, base; + u32 cache, op_inst, op, base; + s16 offset; struct kvm_vcpu_arch *arch = &vcpu->arch; unsigned long va; unsigned long curr_pc; @@ -1576,9 +1600,12 @@ enum emulation_result kvm_mips_emulate_cache(uint32_t inst, uint32_t *opc, if (er == EMULATE_FAIL) return er; - base = (inst >> 21) & 0x1f; - op_inst = (inst >> 16) & 0x1f; - offset = (int16_t)inst; + base = inst.i_format.rs; + op_inst = inst.i_format.rt; + if (cpu_has_mips_r6) + offset = inst.spec3_format.simmediate; + else + offset = inst.i_format.simmediate; cache = op_inst & CacheOp_Cache; op = op_inst & CacheOp_Op; @@ -1634,7 +1661,6 @@ enum emulation_result kvm_mips_emulate_cache(uint32_t inst, uint32_t *opc, (cop0) & KVM_ENTRYHI_ASID)); if (index < 0) { - vcpu->arch.host_cp0_entryhi = (va & VPN2_MASK); vcpu->arch.host_cp0_badvaddr = va; vcpu->arch.pc = curr_pc; er = kvm_mips_emulate_tlbmiss_ld(cause, NULL, run, @@ -1659,9 +1685,7 @@ enum emulation_result kvm_mips_emulate_cache(uint32_t inst, uint32_t *opc, * We fault an entry from the guest tlb to the * shadow host TLB */ - kvm_mips_handle_mapped_seg_tlb_fault(vcpu, tlb, - NULL, - NULL); + kvm_mips_handle_mapped_seg_tlb_fault(vcpu, tlb); } } } else { @@ -1714,20 +1738,20 @@ dont_update_pc: return er; } -enum emulation_result kvm_mips_emulate_inst(unsigned long cause, uint32_t *opc, +enum emulation_result kvm_mips_emulate_inst(u32 cause, u32 *opc, struct kvm_run *run, struct kvm_vcpu *vcpu) { + union mips_instruction inst; enum emulation_result er = EMULATE_DONE; - uint32_t inst; /* Fetch the instruction. */ if (cause & CAUSEF_BD) opc += 1; - inst = kvm_get_inst(opc, vcpu); + inst.word = kvm_get_inst(opc, vcpu); - switch (((union mips_instruction)inst).r_format.opcode) { + switch (inst.r_format.opcode) { case cop0_op: er = kvm_mips_emulate_CP0(inst, opc, cause, run, vcpu); break; @@ -1744,15 +1768,31 @@ enum emulation_result kvm_mips_emulate_inst(unsigned long cause, uint32_t *opc, er = kvm_mips_emulate_load(inst, cause, run, vcpu); break; +#ifndef CONFIG_CPU_MIPSR6 case cache_op: ++vcpu->stat.cache_exits; - trace_kvm_exit(vcpu, CACHE_EXITS); + trace_kvm_exit(vcpu, KVM_TRACE_EXIT_CACHE); er = kvm_mips_emulate_cache(inst, opc, cause, run, vcpu); break; +#else + case spec3_op: + switch (inst.spec3_format.func) { + case cache6_op: + ++vcpu->stat.cache_exits; + trace_kvm_exit(vcpu, KVM_TRACE_EXIT_CACHE); + er = kvm_mips_emulate_cache(inst, opc, cause, run, + vcpu); + break; + default: + goto unknown; + }; + break; +unknown: +#endif default: kvm_err("Instruction emulation not supported (%p/%#x)\n", opc, - inst); + inst.word); kvm_arch_vcpu_dump_regs(vcpu); er = EMULATE_FAIL; break; @@ -1761,8 +1801,8 @@ enum emulation_result kvm_mips_emulate_inst(unsigned long cause, uint32_t *opc, return er; } -enum emulation_result kvm_mips_emulate_syscall(unsigned long cause, - uint32_t *opc, +enum emulation_result kvm_mips_emulate_syscall(u32 cause, + u32 *opc, struct kvm_run *run, struct kvm_vcpu *vcpu) { @@ -1796,8 +1836,8 @@ enum emulation_result kvm_mips_emulate_syscall(unsigned long cause, return er; } -enum emulation_result kvm_mips_emulate_tlbmiss_ld(unsigned long cause, - uint32_t *opc, +enum emulation_result kvm_mips_emulate_tlbmiss_ld(u32 cause, + u32 *opc, struct kvm_run *run, struct kvm_vcpu *vcpu) { @@ -1842,8 +1882,8 @@ enum emulation_result kvm_mips_emulate_tlbmiss_ld(unsigned long cause, return EMULATE_DONE; } -enum emulation_result kvm_mips_emulate_tlbinv_ld(unsigned long cause, - uint32_t *opc, +enum emulation_result kvm_mips_emulate_tlbinv_ld(u32 cause, + u32 *opc, struct kvm_run *run, struct kvm_vcpu *vcpu) { @@ -1888,8 +1928,8 @@ enum emulation_result kvm_mips_emulate_tlbinv_ld(unsigned long cause, return EMULATE_DONE; } -enum emulation_result kvm_mips_emulate_tlbmiss_st(unsigned long cause, - uint32_t *opc, +enum emulation_result kvm_mips_emulate_tlbmiss_st(u32 cause, + u32 *opc, struct kvm_run *run, struct kvm_vcpu *vcpu) { @@ -1932,8 +1972,8 @@ enum emulation_result kvm_mips_emulate_tlbmiss_st(unsigned long cause, return EMULATE_DONE; } -enum emulation_result kvm_mips_emulate_tlbinv_st(unsigned long cause, - uint32_t *opc, +enum emulation_result kvm_mips_emulate_tlbinv_st(u32 cause, + u32 *opc, struct kvm_run *run, struct kvm_vcpu *vcpu) { @@ -1977,7 +2017,7 @@ enum emulation_result kvm_mips_emulate_tlbinv_st(unsigned long cause, } /* TLBMOD: store into address matching TLB with Dirty bit off */ -enum emulation_result kvm_mips_handle_tlbmod(unsigned long cause, uint32_t *opc, +enum emulation_result kvm_mips_handle_tlbmod(u32 cause, u32 *opc, struct kvm_run *run, struct kvm_vcpu *vcpu) { @@ -2005,8 +2045,8 @@ enum emulation_result kvm_mips_handle_tlbmod(unsigned long cause, uint32_t *opc, return er; } -enum emulation_result kvm_mips_emulate_tlbmod(unsigned long cause, - uint32_t *opc, +enum emulation_result kvm_mips_emulate_tlbmod(u32 cause, + u32 *opc, struct kvm_run *run, struct kvm_vcpu *vcpu) { @@ -2048,8 +2088,8 @@ enum emulation_result kvm_mips_emulate_tlbmod(unsigned long cause, return EMULATE_DONE; } -enum emulation_result kvm_mips_emulate_fpu_exc(unsigned long cause, - uint32_t *opc, +enum emulation_result kvm_mips_emulate_fpu_exc(u32 cause, + u32 *opc, struct kvm_run *run, struct kvm_vcpu *vcpu) { @@ -2077,8 +2117,8 @@ enum emulation_result kvm_mips_emulate_fpu_exc(unsigned long cause, return EMULATE_DONE; } -enum emulation_result kvm_mips_emulate_ri_exc(unsigned long cause, - uint32_t *opc, +enum emulation_result kvm_mips_emulate_ri_exc(u32 cause, + u32 *opc, struct kvm_run *run, struct kvm_vcpu *vcpu) { @@ -2112,8 +2152,8 @@ enum emulation_result kvm_mips_emulate_ri_exc(unsigned long cause, return er; } -enum emulation_result kvm_mips_emulate_bp_exc(unsigned long cause, - uint32_t *opc, +enum emulation_result kvm_mips_emulate_bp_exc(u32 cause, + u32 *opc, struct kvm_run *run, struct kvm_vcpu *vcpu) { @@ -2147,8 +2187,8 @@ enum emulation_result kvm_mips_emulate_bp_exc(unsigned long cause, return er; } -enum emulation_result kvm_mips_emulate_trap_exc(unsigned long cause, - uint32_t *opc, +enum emulation_result kvm_mips_emulate_trap_exc(u32 cause, + u32 *opc, struct kvm_run *run, struct kvm_vcpu *vcpu) { @@ -2182,8 +2222,8 @@ enum emulation_result kvm_mips_emulate_trap_exc(unsigned long cause, return er; } -enum emulation_result kvm_mips_emulate_msafpe_exc(unsigned long cause, - uint32_t *opc, +enum emulation_result kvm_mips_emulate_msafpe_exc(u32 cause, + u32 *opc, struct kvm_run *run, struct kvm_vcpu *vcpu) { @@ -2217,8 +2257,8 @@ enum emulation_result kvm_mips_emulate_msafpe_exc(unsigned long cause, return er; } -enum emulation_result kvm_mips_emulate_fpe_exc(unsigned long cause, - uint32_t *opc, +enum emulation_result kvm_mips_emulate_fpe_exc(u32 cause, + u32 *opc, struct kvm_run *run, struct kvm_vcpu *vcpu) { @@ -2252,8 +2292,8 @@ enum emulation_result kvm_mips_emulate_fpe_exc(unsigned long cause, return er; } -enum emulation_result kvm_mips_emulate_msadis_exc(unsigned long cause, - uint32_t *opc, +enum emulation_result kvm_mips_emulate_msadis_exc(u32 cause, + u32 *opc, struct kvm_run *run, struct kvm_vcpu *vcpu) { @@ -2287,22 +2327,7 @@ enum emulation_result kvm_mips_emulate_msadis_exc(unsigned long cause, return er; } -/* ll/sc, rdhwr, sync emulation */ - -#define OPCODE 0xfc000000 -#define BASE 0x03e00000 -#define RT 0x001f0000 -#define OFFSET 0x0000ffff -#define LL 0xc0000000 -#define SC 0xe0000000 -#define SPEC0 0x00000000 -#define SPEC3 0x7c000000 -#define RD 0x0000f800 -#define FUNC 0x0000003f -#define SYNC 0x0000000f -#define RDHWR 0x0000003b - -enum emulation_result kvm_mips_handle_ri(unsigned long cause, uint32_t *opc, +enum emulation_result kvm_mips_handle_ri(u32 cause, u32 *opc, struct kvm_run *run, struct kvm_vcpu *vcpu) { @@ -2310,7 +2335,7 @@ enum emulation_result kvm_mips_handle_ri(unsigned long cause, uint32_t *opc, struct kvm_vcpu_arch *arch = &vcpu->arch; enum emulation_result er = EMULATE_DONE; unsigned long curr_pc; - uint32_t inst; + union mips_instruction inst; /* * Update PC and hold onto current PC in case there is @@ -2325,17 +2350,22 @@ enum emulation_result kvm_mips_handle_ri(unsigned long cause, uint32_t *opc, if (cause & CAUSEF_BD) opc += 1; - inst = kvm_get_inst(opc, vcpu); + inst.word = kvm_get_inst(opc, vcpu); - if (inst == KVM_INVALID_INST) { + if (inst.word == KVM_INVALID_INST) { kvm_err("%s: Cannot get inst @ %p\n", __func__, opc); return EMULATE_FAIL; } - if ((inst & OPCODE) == SPEC3 && (inst & FUNC) == RDHWR) { + if (inst.r_format.opcode == spec3_op && + inst.r_format.func == rdhwr_op && + inst.r_format.rs == 0 && + (inst.r_format.re >> 3) == 0) { int usermode = !KVM_GUEST_KERNEL_MODE(vcpu); - int rd = (inst & RD) >> 11; - int rt = (inst & RT) >> 16; + int rd = inst.r_format.rd; + int rt = inst.r_format.rt; + int sel = inst.r_format.re & 0x7; + /* If usermode, check RDHWR rd is allowed by guest HWREna */ if (usermode && !(kvm_read_c0_guest_hwrena(cop0) & BIT(rd))) { kvm_debug("RDHWR %#x disallowed by HWREna @ %p\n", @@ -2343,17 +2373,17 @@ enum emulation_result kvm_mips_handle_ri(unsigned long cause, uint32_t *opc, goto emulate_ri; } switch (rd) { - case 0: /* CPU number */ - arch->gprs[rt] = 0; + case MIPS_HWR_CPUNUM: /* CPU number */ + arch->gprs[rt] = vcpu->vcpu_id; break; - case 1: /* SYNCI length */ + case MIPS_HWR_SYNCISTEP: /* SYNCI length */ arch->gprs[rt] = min(current_cpu_data.dcache.linesz, current_cpu_data.icache.linesz); break; - case 2: /* Read count register */ - arch->gprs[rt] = kvm_mips_read_count(vcpu); + case MIPS_HWR_CC: /* Read count register */ + arch->gprs[rt] = (s32)kvm_mips_read_count(vcpu); break; - case 3: /* Count register resolution */ + case MIPS_HWR_CCRES: /* Count register resolution */ switch (current_cpu_data.cputype) { case CPU_20KC: case CPU_25KF: @@ -2363,7 +2393,7 @@ enum emulation_result kvm_mips_handle_ri(unsigned long cause, uint32_t *opc, arch->gprs[rt] = 2; } break; - case 29: + case MIPS_HWR_ULR: /* Read UserLocal register */ arch->gprs[rt] = kvm_read_c0_guest_userlocal(cop0); break; @@ -2371,8 +2401,12 @@ enum emulation_result kvm_mips_handle_ri(unsigned long cause, uint32_t *opc, kvm_debug("RDHWR %#x not supported @ %p\n", rd, opc); goto emulate_ri; } + + trace_kvm_hwr(vcpu, KVM_TRACE_RDHWR, KVM_TRACE_HWR(rd, sel), + vcpu->arch.gprs[rt]); } else { - kvm_debug("Emulate RI not supported @ %p: %#x\n", opc, inst); + kvm_debug("Emulate RI not supported @ %p: %#x\n", + opc, inst.word); goto emulate_ri; } @@ -2405,19 +2439,19 @@ enum emulation_result kvm_mips_complete_mmio_load(struct kvm_vcpu *vcpu, switch (run->mmio.len) { case 4: - *gpr = *(int32_t *) run->mmio.data; + *gpr = *(s32 *) run->mmio.data; break; case 2: if (vcpu->mmio_needed == 2) - *gpr = *(int16_t *) run->mmio.data; + *gpr = *(s16 *) run->mmio.data; else - *gpr = *(uint16_t *)run->mmio.data; + *gpr = *(u16 *)run->mmio.data; break; case 1: if (vcpu->mmio_needed == 2) - *gpr = *(int8_t *) run->mmio.data; + *gpr = *(s8 *) run->mmio.data; else *gpr = *(u8 *) run->mmio.data; break; @@ -2432,12 +2466,12 @@ done: return er; } -static enum emulation_result kvm_mips_emulate_exc(unsigned long cause, - uint32_t *opc, +static enum emulation_result kvm_mips_emulate_exc(u32 cause, + u32 *opc, struct kvm_run *run, struct kvm_vcpu *vcpu) { - uint32_t exccode = (cause >> CAUSEB_EXCCODE) & 0x1f; + u32 exccode = (cause >> CAUSEB_EXCCODE) & 0x1f; struct mips_coproc *cop0 = vcpu->arch.cop0; struct kvm_vcpu_arch *arch = &vcpu->arch; enum emulation_result er = EMULATE_DONE; @@ -2470,13 +2504,13 @@ static enum emulation_result kvm_mips_emulate_exc(unsigned long cause, return er; } -enum emulation_result kvm_mips_check_privilege(unsigned long cause, - uint32_t *opc, +enum emulation_result kvm_mips_check_privilege(u32 cause, + u32 *opc, struct kvm_run *run, struct kvm_vcpu *vcpu) { enum emulation_result er = EMULATE_DONE; - uint32_t exccode = (cause >> CAUSEB_EXCCODE) & 0x1f; + u32 exccode = (cause >> CAUSEB_EXCCODE) & 0x1f; unsigned long badvaddr = vcpu->arch.host_cp0_badvaddr; int usermode = !KVM_GUEST_KERNEL_MODE(vcpu); @@ -2566,18 +2600,18 @@ enum emulation_result kvm_mips_check_privilege(unsigned long cause, * (2) TLB entry is present in the Guest TLB but not in the shadow, in this * case we inject the TLB from the Guest TLB into the shadow host TLB */ -enum emulation_result kvm_mips_handle_tlbmiss(unsigned long cause, - uint32_t *opc, +enum emulation_result kvm_mips_handle_tlbmiss(u32 cause, + u32 *opc, struct kvm_run *run, struct kvm_vcpu *vcpu) { enum emulation_result er = EMULATE_DONE; - uint32_t exccode = (cause >> CAUSEB_EXCCODE) & 0x1f; + u32 exccode = (cause >> CAUSEB_EXCCODE) & 0x1f; unsigned long va = vcpu->arch.host_cp0_badvaddr; int index; - kvm_debug("kvm_mips_handle_tlbmiss: badvaddr: %#lx, entryhi: %#lx\n", - vcpu->arch.host_cp0_badvaddr, vcpu->arch.host_cp0_entryhi); + kvm_debug("kvm_mips_handle_tlbmiss: badvaddr: %#lx\n", + vcpu->arch.host_cp0_badvaddr); /* * KVM would not have got the exception if this entry was valid in the @@ -2620,13 +2654,12 @@ enum emulation_result kvm_mips_handle_tlbmiss(unsigned long cause, } } else { kvm_debug("Injecting hi: %#lx, lo0: %#lx, lo1: %#lx into shadow host TLB\n", - tlb->tlb_hi, tlb->tlb_lo0, tlb->tlb_lo1); + tlb->tlb_hi, tlb->tlb_lo[0], tlb->tlb_lo[1]); /* * OK we have a Guest TLB entry, now inject it into the * shadow host TLB */ - kvm_mips_handle_mapped_seg_tlb_fault(vcpu, tlb, NULL, - NULL); + kvm_mips_handle_mapped_seg_tlb_fault(vcpu, tlb); } } diff --git a/arch/mips/kvm/entry.c b/arch/mips/kvm/entry.c new file mode 100644 index 0000000..6a02b3a --- /dev/null +++ b/arch/mips/kvm/entry.c @@ -0,0 +1,701 @@ +/* + * This file is subject to the terms and conditions of the GNU General Public + * License. See the file "COPYING" in the main directory of this archive + * for more details. + * + * Generation of main entry point for the guest, exception handling. + * + * Copyright (C) 2012 MIPS Technologies, Inc. + * Authors: Sanjay Lal <sanjayl@kymasys.com> + * + * Copyright (C) 2016 Imagination Technologies Ltd. + */ + +#include <linux/kvm_host.h> +#include <asm/msa.h> +#include <asm/setup.h> +#include <asm/uasm.h> + +/* Register names */ +#define ZERO 0 +#define AT 1 +#define V0 2 +#define V1 3 +#define A0 4 +#define A1 5 + +#if _MIPS_SIM == _MIPS_SIM_ABI32 +#define T0 8 +#define T1 9 +#define T2 10 +#define T3 11 +#endif /* _MIPS_SIM == _MIPS_SIM_ABI32 */ + +#if _MIPS_SIM == _MIPS_SIM_ABI64 || _MIPS_SIM == _MIPS_SIM_NABI32 +#define T0 12 +#define T1 13 +#define T2 14 +#define T3 15 +#endif /* _MIPS_SIM == _MIPS_SIM_ABI64 || _MIPS_SIM == _MIPS_SIM_NABI32 */ + +#define S0 16 +#define S1 17 +#define T9 25 +#define K0 26 +#define K1 27 +#define GP 28 +#define SP 29 +#define RA 31 + +/* Some CP0 registers */ +#define C0_HWRENA 7, 0 +#define C0_BADVADDR 8, 0 +#define C0_ENTRYHI 10, 0 +#define C0_STATUS 12, 0 +#define C0_CAUSE 13, 0 +#define C0_EPC 14, 0 +#define C0_EBASE 15, 1 +#define C0_CONFIG5 16, 5 +#define C0_DDATA_LO 28, 3 +#define C0_ERROREPC 30, 0 + +#define CALLFRAME_SIZ 32 + +#ifdef CONFIG_64BIT +#define ST0_KX_IF_64 ST0_KX +#else +#define ST0_KX_IF_64 0 +#endif + +static unsigned int scratch_vcpu[2] = { C0_DDATA_LO }; +static unsigned int scratch_tmp[2] = { C0_ERROREPC }; + +enum label_id { + label_fpu_1 = 1, + label_msa_1, + label_return_to_host, + label_kernel_asid, + label_exit_common, +}; + +UASM_L_LA(_fpu_1) +UASM_L_LA(_msa_1) +UASM_L_LA(_return_to_host) +UASM_L_LA(_kernel_asid) +UASM_L_LA(_exit_common) + +static void *kvm_mips_build_enter_guest(void *addr); +static void *kvm_mips_build_ret_from_exit(void *addr); +static void *kvm_mips_build_ret_to_guest(void *addr); +static void *kvm_mips_build_ret_to_host(void *addr); + +/** + * kvm_mips_entry_setup() - Perform global setup for entry code. + * + * Perform global setup for entry code, such as choosing a scratch register. + * + * Returns: 0 on success. + * -errno on failure. + */ +int kvm_mips_entry_setup(void) +{ + /* + * We prefer to use KScratchN registers if they are available over the + * defaults above, which may not work on all cores. + */ + unsigned int kscratch_mask = cpu_data[0].kscratch_mask & 0xfc; + + /* Pick a scratch register for storing VCPU */ + if (kscratch_mask) { + scratch_vcpu[0] = 31; + scratch_vcpu[1] = ffs(kscratch_mask) - 1; + kscratch_mask &= ~BIT(scratch_vcpu[1]); + } + + /* Pick a scratch register to use as a temp for saving state */ + if (kscratch_mask) { + scratch_tmp[0] = 31; + scratch_tmp[1] = ffs(kscratch_mask) - 1; + kscratch_mask &= ~BIT(scratch_tmp[1]); + } + + return 0; +} + +static void kvm_mips_build_save_scratch(u32 **p, unsigned int tmp, + unsigned int frame) +{ + /* Save the VCPU scratch register value in cp0_epc of the stack frame */ + UASM_i_MFC0(p, tmp, scratch_vcpu[0], scratch_vcpu[1]); + UASM_i_SW(p, tmp, offsetof(struct pt_regs, cp0_epc), frame); + + /* Save the temp scratch register value in cp0_cause of stack frame */ + if (scratch_tmp[0] == 31) { + UASM_i_MFC0(p, tmp, scratch_tmp[0], scratch_tmp[1]); + UASM_i_SW(p, tmp, offsetof(struct pt_regs, cp0_cause), frame); + } +} + +static void kvm_mips_build_restore_scratch(u32 **p, unsigned int tmp, + unsigned int frame) +{ + /* + * Restore host scratch register values saved by + * kvm_mips_build_save_scratch(). + */ + UASM_i_LW(p, tmp, offsetof(struct pt_regs, cp0_epc), frame); + UASM_i_MTC0(p, tmp, scratch_vcpu[0], scratch_vcpu[1]); + + if (scratch_tmp[0] == 31) { + UASM_i_LW(p, tmp, offsetof(struct pt_regs, cp0_cause), frame); + UASM_i_MTC0(p, tmp, scratch_tmp[0], scratch_tmp[1]); + } +} + +/** + * build_set_exc_base() - Assemble code to write exception base address. + * @p: Code buffer pointer. + * @reg: Source register (generated code may set WG bit in @reg). + * + * Assemble code to modify the exception base address in the EBase register, + * using the appropriately sized access and setting the WG bit if necessary. + */ +static inline void build_set_exc_base(u32 **p, unsigned int reg) +{ + if (cpu_has_ebase_wg) { + /* Set WG so that all the bits get written */ + uasm_i_ori(p, reg, reg, MIPS_EBASE_WG); + UASM_i_MTC0(p, reg, C0_EBASE); + } else { + uasm_i_mtc0(p, reg, C0_EBASE); + } +} + +/** + * kvm_mips_build_vcpu_run() - Assemble function to start running a guest VCPU. + * @addr: Address to start writing code. + * + * Assemble the start of the vcpu_run function to run a guest VCPU. The function + * conforms to the following prototype: + * + * int vcpu_run(struct kvm_run *run, struct kvm_vcpu *vcpu); + * + * The exit from the guest and return to the caller is handled by the code + * generated by kvm_mips_build_ret_to_host(). + * + * Returns: Next address after end of written function. + */ +void *kvm_mips_build_vcpu_run(void *addr) +{ + u32 *p = addr; + unsigned int i; + + /* + * A0: run + * A1: vcpu + */ + + /* k0/k1 not being used in host kernel context */ + UASM_i_ADDIU(&p, K1, SP, -(int)sizeof(struct pt_regs)); + for (i = 16; i < 32; ++i) { + if (i == 24) + i = 28; + UASM_i_SW(&p, i, offsetof(struct pt_regs, regs[i]), K1); + } + + /* Save host status */ + uasm_i_mfc0(&p, V0, C0_STATUS); + UASM_i_SW(&p, V0, offsetof(struct pt_regs, cp0_status), K1); + + /* Save scratch registers, will be used to store pointer to vcpu etc */ + kvm_mips_build_save_scratch(&p, V1, K1); + + /* VCPU scratch register has pointer to vcpu */ + UASM_i_MTC0(&p, A1, scratch_vcpu[0], scratch_vcpu[1]); + + /* Offset into vcpu->arch */ + UASM_i_ADDIU(&p, K1, A1, offsetof(struct kvm_vcpu, arch)); + + /* + * Save the host stack to VCPU, used for exception processing + * when we exit from the Guest + */ + UASM_i_SW(&p, SP, offsetof(struct kvm_vcpu_arch, host_stack), K1); + + /* Save the kernel gp as well */ + UASM_i_SW(&p, GP, offsetof(struct kvm_vcpu_arch, host_gp), K1); + + /* + * Setup status register for running the guest in UM, interrupts + * are disabled + */ + UASM_i_LA(&p, K0, ST0_EXL | KSU_USER | ST0_BEV | ST0_KX_IF_64); + uasm_i_mtc0(&p, K0, C0_STATUS); + uasm_i_ehb(&p); + + /* load up the new EBASE */ + UASM_i_LW(&p, K0, offsetof(struct kvm_vcpu_arch, guest_ebase), K1); + build_set_exc_base(&p, K0); + + /* + * Now that the new EBASE has been loaded, unset BEV, set + * interrupt mask as it was but make sure that timer interrupts + * are enabled + */ + uasm_i_addiu(&p, K0, ZERO, ST0_EXL | KSU_USER | ST0_IE | ST0_KX_IF_64); + uasm_i_andi(&p, V0, V0, ST0_IM); + uasm_i_or(&p, K0, K0, V0); + uasm_i_mtc0(&p, K0, C0_STATUS); + uasm_i_ehb(&p); + + p = kvm_mips_build_enter_guest(p); + + return p; +} + +/** + * kvm_mips_build_enter_guest() - Assemble code to resume guest execution. + * @addr: Address to start writing code. + * + * Assemble the code to resume guest execution. This code is common between the + * initial entry into the guest from the host, and returning from the exit + * handler back to the guest. + * + * Returns: Next address after end of written function. + */ +static void *kvm_mips_build_enter_guest(void *addr) +{ + u32 *p = addr; + unsigned int i; + struct uasm_label labels[2]; + struct uasm_reloc relocs[2]; + struct uasm_label *l = labels; + struct uasm_reloc *r = relocs; + + memset(labels, 0, sizeof(labels)); + memset(relocs, 0, sizeof(relocs)); + + /* Set Guest EPC */ + UASM_i_LW(&p, T0, offsetof(struct kvm_vcpu_arch, pc), K1); + UASM_i_MTC0(&p, T0, C0_EPC); + + /* Set the ASID for the Guest Kernel */ + UASM_i_LW(&p, T0, offsetof(struct kvm_vcpu_arch, cop0), K1); + UASM_i_LW(&p, T0, offsetof(struct mips_coproc, reg[MIPS_CP0_STATUS][0]), + T0); + uasm_i_andi(&p, T0, T0, KSU_USER | ST0_ERL | ST0_EXL); + uasm_i_xori(&p, T0, T0, KSU_USER); + uasm_il_bnez(&p, &r, T0, label_kernel_asid); + UASM_i_ADDIU(&p, T1, K1, + offsetof(struct kvm_vcpu_arch, guest_kernel_asid)); + /* else user */ + UASM_i_ADDIU(&p, T1, K1, + offsetof(struct kvm_vcpu_arch, guest_user_asid)); + uasm_l_kernel_asid(&l, p); + + /* t1: contains the base of the ASID array, need to get the cpu id */ + /* smp_processor_id */ + uasm_i_lw(&p, T2, offsetof(struct thread_info, cpu), GP); + /* x4 */ + uasm_i_sll(&p, T2, T2, 2); + UASM_i_ADDU(&p, T3, T1, T2); + uasm_i_lw(&p, K0, 0, T3); +#ifdef CONFIG_MIPS_ASID_BITS_VARIABLE + /* x sizeof(struct cpuinfo_mips)/4 */ + uasm_i_addiu(&p, T3, ZERO, sizeof(struct cpuinfo_mips)/4); + uasm_i_mul(&p, T2, T2, T3); + + UASM_i_LA_mostly(&p, AT, (long)&cpu_data[0].asid_mask); + UASM_i_ADDU(&p, AT, AT, T2); + UASM_i_LW(&p, T2, uasm_rel_lo((long)&cpu_data[0].asid_mask), AT); + uasm_i_and(&p, K0, K0, T2); +#else + uasm_i_andi(&p, K0, K0, MIPS_ENTRYHI_ASID); +#endif + uasm_i_mtc0(&p, K0, C0_ENTRYHI); + uasm_i_ehb(&p); + + /* Disable RDHWR access */ + uasm_i_mtc0(&p, ZERO, C0_HWRENA); + + /* load the guest context from VCPU and return */ + for (i = 1; i < 32; ++i) { + /* Guest k0/k1 loaded later */ + if (i == K0 || i == K1) + continue; + UASM_i_LW(&p, i, offsetof(struct kvm_vcpu_arch, gprs[i]), K1); + } + +#ifndef CONFIG_CPU_MIPSR6 + /* Restore hi/lo */ + UASM_i_LW(&p, K0, offsetof(struct kvm_vcpu_arch, hi), K1); + uasm_i_mthi(&p, K0); + + UASM_i_LW(&p, K0, offsetof(struct kvm_vcpu_arch, lo), K1); + uasm_i_mtlo(&p, K0); +#endif + + /* Restore the guest's k0/k1 registers */ + UASM_i_LW(&p, K0, offsetof(struct kvm_vcpu_arch, gprs[K0]), K1); + UASM_i_LW(&p, K1, offsetof(struct kvm_vcpu_arch, gprs[K1]), K1); + + /* Jump to guest */ + uasm_i_eret(&p); + + uasm_resolve_relocs(relocs, labels); + + return p; +} + +/** + * kvm_mips_build_exception() - Assemble first level guest exception handler. + * @addr: Address to start writing code. + * @handler: Address of common handler (within range of @addr). + * + * Assemble exception vector code for guest execution. The generated vector will + * branch to the common exception handler generated by kvm_mips_build_exit(). + * + * Returns: Next address after end of written function. + */ +void *kvm_mips_build_exception(void *addr, void *handler) +{ + u32 *p = addr; + struct uasm_label labels[2]; + struct uasm_reloc relocs[2]; + struct uasm_label *l = labels; + struct uasm_reloc *r = relocs; + + memset(labels, 0, sizeof(labels)); + memset(relocs, 0, sizeof(relocs)); + + /* Save guest k1 into scratch register */ + UASM_i_MTC0(&p, K1, scratch_tmp[0], scratch_tmp[1]); + + /* Get the VCPU pointer from the VCPU scratch register */ + UASM_i_MFC0(&p, K1, scratch_vcpu[0], scratch_vcpu[1]); + UASM_i_ADDIU(&p, K1, K1, offsetof(struct kvm_vcpu, arch)); + + /* Save guest k0 into VCPU structure */ + UASM_i_SW(&p, K0, offsetof(struct kvm_vcpu_arch, gprs[K0]), K1); + + /* Branch to the common handler */ + uasm_il_b(&p, &r, label_exit_common); + uasm_i_nop(&p); + + uasm_l_exit_common(&l, handler); + uasm_resolve_relocs(relocs, labels); + + return p; +} + +/** + * kvm_mips_build_exit() - Assemble common guest exit handler. + * @addr: Address to start writing code. + * + * Assemble the generic guest exit handling code. This is called by the + * exception vectors (generated by kvm_mips_build_exception()), and calls + * kvm_mips_handle_exit(), then either resumes the guest or returns to the host + * depending on the return value. + * + * Returns: Next address after end of written function. + */ +void *kvm_mips_build_exit(void *addr) +{ + u32 *p = addr; + unsigned int i; + struct uasm_label labels[3]; + struct uasm_reloc relocs[3]; + struct uasm_label *l = labels; + struct uasm_reloc *r = relocs; + + memset(labels, 0, sizeof(labels)); + memset(relocs, 0, sizeof(relocs)); + + /* + * Generic Guest exception handler. We end up here when the guest + * does something that causes a trap to kernel mode. + * + * Both k0/k1 registers will have already been saved (k0 into the vcpu + * structure, and k1 into the scratch_tmp register). + * + * The k1 register will already contain the kvm_vcpu_arch pointer. + */ + + /* Start saving Guest context to VCPU */ + for (i = 0; i < 32; ++i) { + /* Guest k0/k1 saved later */ + if (i == K0 || i == K1) + continue; + UASM_i_SW(&p, i, offsetof(struct kvm_vcpu_arch, gprs[i]), K1); + } + +#ifndef CONFIG_CPU_MIPSR6 + /* We need to save hi/lo and restore them on the way out */ + uasm_i_mfhi(&p, T0); + UASM_i_SW(&p, T0, offsetof(struct kvm_vcpu_arch, hi), K1); + + uasm_i_mflo(&p, T0); + UASM_i_SW(&p, T0, offsetof(struct kvm_vcpu_arch, lo), K1); +#endif + + /* Finally save guest k1 to VCPU */ + uasm_i_ehb(&p); + UASM_i_MFC0(&p, T0, scratch_tmp[0], scratch_tmp[1]); + UASM_i_SW(&p, T0, offsetof(struct kvm_vcpu_arch, gprs[K1]), K1); + + /* Now that context has been saved, we can use other registers */ + + /* Restore vcpu */ + UASM_i_MFC0(&p, A1, scratch_vcpu[0], scratch_vcpu[1]); + uasm_i_move(&p, S1, A1); + + /* Restore run (vcpu->run) */ + UASM_i_LW(&p, A0, offsetof(struct kvm_vcpu, run), A1); + /* Save pointer to run in s0, will be saved by the compiler */ + uasm_i_move(&p, S0, A0); + + /* + * Save Host level EPC, BadVaddr and Cause to VCPU, useful to process + * the exception + */ + UASM_i_MFC0(&p, K0, C0_EPC); + UASM_i_SW(&p, K0, offsetof(struct kvm_vcpu_arch, pc), K1); + + UASM_i_MFC0(&p, K0, C0_BADVADDR); + UASM_i_SW(&p, K0, offsetof(struct kvm_vcpu_arch, host_cp0_badvaddr), + K1); + + uasm_i_mfc0(&p, K0, C0_CAUSE); + uasm_i_sw(&p, K0, offsetof(struct kvm_vcpu_arch, host_cp0_cause), K1); + + /* Now restore the host state just enough to run the handlers */ + + /* Switch EBASE to the one used by Linux */ + /* load up the host EBASE */ + uasm_i_mfc0(&p, V0, C0_STATUS); + + uasm_i_lui(&p, AT, ST0_BEV >> 16); + uasm_i_or(&p, K0, V0, AT); + + uasm_i_mtc0(&p, K0, C0_STATUS); + uasm_i_ehb(&p); + + UASM_i_LA_mostly(&p, K0, (long)&ebase); + UASM_i_LW(&p, K0, uasm_rel_lo((long)&ebase), K0); + build_set_exc_base(&p, K0); + + if (raw_cpu_has_fpu) { + /* + * If FPU is enabled, save FCR31 and clear it so that later + * ctc1's don't trigger FPE for pending exceptions. + */ + uasm_i_lui(&p, AT, ST0_CU1 >> 16); + uasm_i_and(&p, V1, V0, AT); + uasm_il_beqz(&p, &r, V1, label_fpu_1); + uasm_i_nop(&p); + uasm_i_cfc1(&p, T0, 31); + uasm_i_sw(&p, T0, offsetof(struct kvm_vcpu_arch, fpu.fcr31), + K1); + uasm_i_ctc1(&p, ZERO, 31); + uasm_l_fpu_1(&l, p); + } + + if (cpu_has_msa) { + /* + * If MSA is enabled, save MSACSR and clear it so that later + * instructions don't trigger MSAFPE for pending exceptions. + */ + uasm_i_mfc0(&p, T0, C0_CONFIG5); + uasm_i_ext(&p, T0, T0, 27, 1); /* MIPS_CONF5_MSAEN */ + uasm_il_beqz(&p, &r, T0, label_msa_1); + uasm_i_nop(&p); + uasm_i_cfcmsa(&p, T0, MSA_CSR); + uasm_i_sw(&p, T0, offsetof(struct kvm_vcpu_arch, fpu.msacsr), + K1); + uasm_i_ctcmsa(&p, MSA_CSR, ZERO); + uasm_l_msa_1(&l, p); + } + + /* Now that the new EBASE has been loaded, unset BEV and KSU_USER */ + uasm_i_addiu(&p, AT, ZERO, ~(ST0_EXL | KSU_USER | ST0_IE)); + uasm_i_and(&p, V0, V0, AT); + uasm_i_lui(&p, AT, ST0_CU0 >> 16); + uasm_i_or(&p, V0, V0, AT); + uasm_i_mtc0(&p, V0, C0_STATUS); + uasm_i_ehb(&p); + + /* Load up host GP */ + UASM_i_LW(&p, GP, offsetof(struct kvm_vcpu_arch, host_gp), K1); + + /* Need a stack before we can jump to "C" */ + UASM_i_LW(&p, SP, offsetof(struct kvm_vcpu_arch, host_stack), K1); + + /* Saved host state */ + UASM_i_ADDIU(&p, SP, SP, -(int)sizeof(struct pt_regs)); + + /* + * XXXKYMA do we need to load the host ASID, maybe not because the + * kernel entries are marked GLOBAL, need to verify + */ + + /* Restore host scratch registers, as we'll have clobbered them */ + kvm_mips_build_restore_scratch(&p, K0, SP); + + /* Restore RDHWR access */ + UASM_i_LA_mostly(&p, K0, (long)&hwrena); + uasm_i_lw(&p, K0, uasm_rel_lo((long)&hwrena), K0); + uasm_i_mtc0(&p, K0, C0_HWRENA); + + /* Jump to handler */ + /* + * XXXKYMA: not sure if this is safe, how large is the stack?? + * Now jump to the kvm_mips_handle_exit() to see if we can deal + * with this in the kernel + */ + UASM_i_LA(&p, T9, (unsigned long)kvm_mips_handle_exit); + uasm_i_jalr(&p, RA, T9); + UASM_i_ADDIU(&p, SP, SP, -CALLFRAME_SIZ); + + uasm_resolve_relocs(relocs, labels); + + p = kvm_mips_build_ret_from_exit(p); + + return p; +} + +/** + * kvm_mips_build_ret_from_exit() - Assemble guest exit return handler. + * @addr: Address to start writing code. + * + * Assemble the code to handle the return from kvm_mips_handle_exit(), either + * resuming the guest or returning to the host depending on the return value. + * + * Returns: Next address after end of written function. + */ +static void *kvm_mips_build_ret_from_exit(void *addr) +{ + u32 *p = addr; + struct uasm_label labels[2]; + struct uasm_reloc relocs[2]; + struct uasm_label *l = labels; + struct uasm_reloc *r = relocs; + + memset(labels, 0, sizeof(labels)); + memset(relocs, 0, sizeof(relocs)); + + /* Return from handler Make sure interrupts are disabled */ + uasm_i_di(&p, ZERO); + uasm_i_ehb(&p); + + /* + * XXXKYMA: k0/k1 could have been blown away if we processed + * an exception while we were handling the exception from the + * guest, reload k1 + */ + + uasm_i_move(&p, K1, S1); + UASM_i_ADDIU(&p, K1, K1, offsetof(struct kvm_vcpu, arch)); + + /* + * Check return value, should tell us if we are returning to the + * host (handle I/O etc)or resuming the guest + */ + uasm_i_andi(&p, T0, V0, RESUME_HOST); + uasm_il_bnez(&p, &r, T0, label_return_to_host); + uasm_i_nop(&p); + + p = kvm_mips_build_ret_to_guest(p); + + uasm_l_return_to_host(&l, p); + p = kvm_mips_build_ret_to_host(p); + + uasm_resolve_relocs(relocs, labels); + + return p; +} + +/** + * kvm_mips_build_ret_to_guest() - Assemble code to return to the guest. + * @addr: Address to start writing code. + * + * Assemble the code to handle return from the guest exit handler + * (kvm_mips_handle_exit()) back to the guest. + * + * Returns: Next address after end of written function. + */ +static void *kvm_mips_build_ret_to_guest(void *addr) +{ + u32 *p = addr; + + /* Put the saved pointer to vcpu (s1) back into the scratch register */ + UASM_i_MTC0(&p, S1, scratch_vcpu[0], scratch_vcpu[1]); + + /* Load up the Guest EBASE to minimize the window where BEV is set */ + UASM_i_LW(&p, T0, offsetof(struct kvm_vcpu_arch, guest_ebase), K1); + + /* Switch EBASE back to the one used by KVM */ + uasm_i_mfc0(&p, V1, C0_STATUS); + uasm_i_lui(&p, AT, ST0_BEV >> 16); + uasm_i_or(&p, K0, V1, AT); + uasm_i_mtc0(&p, K0, C0_STATUS); + uasm_i_ehb(&p); + build_set_exc_base(&p, T0); + + /* Setup status register for running guest in UM */ + uasm_i_ori(&p, V1, V1, ST0_EXL | KSU_USER | ST0_IE); + UASM_i_LA(&p, AT, ~(ST0_CU0 | ST0_MX)); + uasm_i_and(&p, V1, V1, AT); + uasm_i_mtc0(&p, V1, C0_STATUS); + uasm_i_ehb(&p); + + p = kvm_mips_build_enter_guest(p); + + return p; +} + +/** + * kvm_mips_build_ret_to_host() - Assemble code to return to the host. + * @addr: Address to start writing code. + * + * Assemble the code to handle return from the guest exit handler + * (kvm_mips_handle_exit()) back to the host, i.e. to the caller of the vcpu_run + * function generated by kvm_mips_build_vcpu_run(). + * + * Returns: Next address after end of written function. + */ +static void *kvm_mips_build_ret_to_host(void *addr) +{ + u32 *p = addr; + unsigned int i; + + /* EBASE is already pointing to Linux */ + UASM_i_LW(&p, K1, offsetof(struct kvm_vcpu_arch, host_stack), K1); + UASM_i_ADDIU(&p, K1, K1, -(int)sizeof(struct pt_regs)); + + /* + * r2/v0 is the return code, shift it down by 2 (arithmetic) + * to recover the err code + */ + uasm_i_sra(&p, K0, V0, 2); + uasm_i_move(&p, V0, K0); + + /* Load context saved on the host stack */ + for (i = 16; i < 31; ++i) { + if (i == 24) + i = 28; + UASM_i_LW(&p, i, offsetof(struct pt_regs, regs[i]), K1); + } + + /* Restore RDHWR access */ + UASM_i_LA_mostly(&p, K0, (long)&hwrena); + uasm_i_lw(&p, K0, uasm_rel_lo((long)&hwrena), K0); + uasm_i_mtc0(&p, K0, C0_HWRENA); + + /* Restore RA, which is the address we will return to */ + UASM_i_LW(&p, RA, offsetof(struct pt_regs, regs[RA]), K1); + uasm_i_jr(&p, RA); + uasm_i_nop(&p); + + return p; +} + diff --git a/arch/mips/kvm/fpu.S b/arch/mips/kvm/fpu.S index 531fbf5..16f17c6 100644 --- a/arch/mips/kvm/fpu.S +++ b/arch/mips/kvm/fpu.S @@ -14,13 +14,16 @@ #include <asm/mipsregs.h> #include <asm/regdef.h> +/* preprocessor replaces the fp in ".set fp=64" with $30 otherwise */ +#undef fp + .set noreorder .set noat LEAF(__kvm_save_fpu) .set push - .set mips64r2 SET_HARDFLOAT + .set fp=64 mfc0 t0, CP0_STATUS sll t0, t0, 5 # is Status.FR set? bgez t0, 1f # no: skip odd doubles @@ -63,8 +66,8 @@ LEAF(__kvm_save_fpu) LEAF(__kvm_restore_fpu) .set push - .set mips64r2 SET_HARDFLOAT + .set fp=64 mfc0 t0, CP0_STATUS sll t0, t0, 5 # is Status.FR set? bgez t0, 1f # no: skip odd doubles diff --git a/arch/mips/kvm/interrupt.c b/arch/mips/kvm/interrupt.c index 95f7906..ad28dac 100644 --- a/arch/mips/kvm/interrupt.c +++ b/arch/mips/kvm/interrupt.c @@ -22,12 +22,12 @@ #include "interrupt.h" -void kvm_mips_queue_irq(struct kvm_vcpu *vcpu, uint32_t priority) +void kvm_mips_queue_irq(struct kvm_vcpu *vcpu, unsigned int priority) { set_bit(priority, &vcpu->arch.pending_exceptions); } -void kvm_mips_dequeue_irq(struct kvm_vcpu *vcpu, uint32_t priority) +void kvm_mips_dequeue_irq(struct kvm_vcpu *vcpu, unsigned int priority) { clear_bit(priority, &vcpu->arch.pending_exceptions); } @@ -114,10 +114,10 @@ void kvm_mips_dequeue_io_int_cb(struct kvm_vcpu *vcpu, /* Deliver the interrupt of the corresponding priority, if possible. */ int kvm_mips_irq_deliver_cb(struct kvm_vcpu *vcpu, unsigned int priority, - uint32_t cause) + u32 cause) { int allowed = 0; - uint32_t exccode; + u32 exccode; struct kvm_vcpu_arch *arch = &vcpu->arch; struct mips_coproc *cop0 = vcpu->arch.cop0; @@ -196,12 +196,12 @@ int kvm_mips_irq_deliver_cb(struct kvm_vcpu *vcpu, unsigned int priority, } int kvm_mips_irq_clear_cb(struct kvm_vcpu *vcpu, unsigned int priority, - uint32_t cause) + u32 cause) { return 1; } -void kvm_mips_deliver_interrupts(struct kvm_vcpu *vcpu, uint32_t cause) +void kvm_mips_deliver_interrupts(struct kvm_vcpu *vcpu, u32 cause) { unsigned long *pending = &vcpu->arch.pending_exceptions; unsigned long *pending_clr = &vcpu->arch.pending_exceptions_clr; diff --git a/arch/mips/kvm/interrupt.h b/arch/mips/kvm/interrupt.h index 2143884..fb118a2 100644 --- a/arch/mips/kvm/interrupt.h +++ b/arch/mips/kvm/interrupt.h @@ -28,17 +28,13 @@ #define MIPS_EXC_MAX 12 /* XXXSL More to follow */ -extern char __kvm_mips_vcpu_run_end[]; -extern char mips32_exception[], mips32_exceptionEnd[]; -extern char mips32_GuestException[], mips32_GuestExceptionEnd[]; - #define C_TI (_ULCAST_(1) << 30) #define KVM_MIPS_IRQ_DELIVER_ALL_AT_ONCE (0) #define KVM_MIPS_IRQ_CLEAR_ALL_AT_ONCE (0) -void kvm_mips_queue_irq(struct kvm_vcpu *vcpu, uint32_t priority); -void kvm_mips_dequeue_irq(struct kvm_vcpu *vcpu, uint32_t priority); +void kvm_mips_queue_irq(struct kvm_vcpu *vcpu, unsigned int priority); +void kvm_mips_dequeue_irq(struct kvm_vcpu *vcpu, unsigned int priority); int kvm_mips_pending_timer(struct kvm_vcpu *vcpu); void kvm_mips_queue_timer_int_cb(struct kvm_vcpu *vcpu); @@ -48,7 +44,7 @@ void kvm_mips_queue_io_int_cb(struct kvm_vcpu *vcpu, void kvm_mips_dequeue_io_int_cb(struct kvm_vcpu *vcpu, struct kvm_mips_interrupt *irq); int kvm_mips_irq_deliver_cb(struct kvm_vcpu *vcpu, unsigned int priority, - uint32_t cause); + u32 cause); int kvm_mips_irq_clear_cb(struct kvm_vcpu *vcpu, unsigned int priority, - uint32_t cause); -void kvm_mips_deliver_interrupts(struct kvm_vcpu *vcpu, uint32_t cause); + u32 cause); +void kvm_mips_deliver_interrupts(struct kvm_vcpu *vcpu, u32 cause); diff --git a/arch/mips/kvm/locore.S b/arch/mips/kvm/locore.S deleted file mode 100644 index 828fcfc..0000000 --- a/arch/mips/kvm/locore.S +++ /dev/null @@ -1,605 +0,0 @@ -/* - * This file is subject to the terms and conditions of the GNU General Public - * License. See the file "COPYING" in the main directory of this archive - * for more details. - * - * Main entry point for the guest, exception handling. - * - * Copyright (C) 2012 MIPS Technologies, Inc. All rights reserved. - * Authors: Sanjay Lal <sanjayl@kymasys.com> - */ - -#include <asm/asm.h> -#include <asm/asmmacro.h> -#include <asm/regdef.h> -#include <asm/mipsregs.h> -#include <asm/stackframe.h> -#include <asm/asm-offsets.h> - -#define _C_LABEL(x) x -#define MIPSX(name) mips32_ ## name -#define CALLFRAME_SIZ 32 - -/* - * VECTOR - * exception vector entrypoint - */ -#define VECTOR(x, regmask) \ - .ent _C_LABEL(x),0; \ - EXPORT(x); - -#define VECTOR_END(x) \ - EXPORT(x); - -/* Overload, Danger Will Robinson!! */ -#define PT_HOST_USERLOCAL PT_EPC - -#define CP0_DDATA_LO $28,3 - -/* Resume Flags */ -#define RESUME_FLAG_HOST (1<<1) /* Resume host? */ - -#define RESUME_GUEST 0 -#define RESUME_HOST RESUME_FLAG_HOST - -/* - * __kvm_mips_vcpu_run: entry point to the guest - * a0: run - * a1: vcpu - */ - .set noreorder - -FEXPORT(__kvm_mips_vcpu_run) - /* k0/k1 not being used in host kernel context */ - INT_ADDIU k1, sp, -PT_SIZE - LONG_S $16, PT_R16(k1) - LONG_S $17, PT_R17(k1) - LONG_S $18, PT_R18(k1) - LONG_S $19, PT_R19(k1) - LONG_S $20, PT_R20(k1) - LONG_S $21, PT_R21(k1) - LONG_S $22, PT_R22(k1) - LONG_S $23, PT_R23(k1) - - LONG_S $28, PT_R28(k1) - LONG_S $29, PT_R29(k1) - LONG_S $30, PT_R30(k1) - LONG_S $31, PT_R31(k1) - - /* Save hi/lo */ - mflo v0 - LONG_S v0, PT_LO(k1) - mfhi v1 - LONG_S v1, PT_HI(k1) - - /* Save host status */ - mfc0 v0, CP0_STATUS - LONG_S v0, PT_STATUS(k1) - - /* Save DDATA_LO, will be used to store pointer to vcpu */ - mfc0 v1, CP0_DDATA_LO - LONG_S v1, PT_HOST_USERLOCAL(k1) - - /* DDATA_LO has pointer to vcpu */ - mtc0 a1, CP0_DDATA_LO - - /* Offset into vcpu->arch */ - INT_ADDIU k1, a1, VCPU_HOST_ARCH - - /* - * Save the host stack to VCPU, used for exception processing - * when we exit from the Guest - */ - LONG_S sp, VCPU_HOST_STACK(k1) - - /* Save the kernel gp as well */ - LONG_S gp, VCPU_HOST_GP(k1) - - /* - * Setup status register for running the guest in UM, interrupts - * are disabled - */ - li k0, (ST0_EXL | KSU_USER | ST0_BEV) - mtc0 k0, CP0_STATUS - ehb - - /* load up the new EBASE */ - LONG_L k0, VCPU_GUEST_EBASE(k1) - mtc0 k0, CP0_EBASE - - /* - * Now that the new EBASE has been loaded, unset BEV, set - * interrupt mask as it was but make sure that timer interrupts - * are enabled - */ - li k0, (ST0_EXL | KSU_USER | ST0_IE) - andi v0, v0, ST0_IM - or k0, k0, v0 - mtc0 k0, CP0_STATUS - ehb - - /* Set Guest EPC */ - LONG_L t0, VCPU_PC(k1) - mtc0 t0, CP0_EPC - -FEXPORT(__kvm_mips_load_asid) - /* Set the ASID for the Guest Kernel */ - PTR_L t0, VCPU_COP0(k1) - LONG_L t0, COP0_STATUS(t0) - andi t0, KSU_USER | ST0_ERL | ST0_EXL - xori t0, KSU_USER - bnez t0, 1f /* If kernel */ - INT_ADDIU t1, k1, VCPU_GUEST_KERNEL_ASID /* (BD) */ - INT_ADDIU t1, k1, VCPU_GUEST_USER_ASID /* else user */ -1: - /* t1: contains the base of the ASID array, need to get the cpu id */ - LONG_L t2, TI_CPU($28) /* smp_processor_id */ - INT_SLL t2, t2, 2 /* x4 */ - REG_ADDU t3, t1, t2 - LONG_L k0, (t3) -#ifdef CONFIG_MIPS_ASID_BITS_VARIABLE - li t3, CPUINFO_SIZE/4 - mul t2, t2, t3 /* x sizeof(struct cpuinfo_mips)/4 */ - LONG_L t2, (cpu_data + CPUINFO_ASID_MASK)(t2) - and k0, k0, t2 -#else - andi k0, k0, MIPS_ENTRYHI_ASID -#endif - mtc0 k0, CP0_ENTRYHI - ehb - - /* Disable RDHWR access */ - mtc0 zero, CP0_HWRENA - - .set noat - /* Now load up the Guest Context from VCPU */ - LONG_L $1, VCPU_R1(k1) - LONG_L $2, VCPU_R2(k1) - LONG_L $3, VCPU_R3(k1) - - LONG_L $4, VCPU_R4(k1) - LONG_L $5, VCPU_R5(k1) - LONG_L $6, VCPU_R6(k1) - LONG_L $7, VCPU_R7(k1) - - LONG_L $8, VCPU_R8(k1) - LONG_L $9, VCPU_R9(k1) - LONG_L $10, VCPU_R10(k1) - LONG_L $11, VCPU_R11(k1) - LONG_L $12, VCPU_R12(k1) - LONG_L $13, VCPU_R13(k1) - LONG_L $14, VCPU_R14(k1) - LONG_L $15, VCPU_R15(k1) - LONG_L $16, VCPU_R16(k1) - LONG_L $17, VCPU_R17(k1) - LONG_L $18, VCPU_R18(k1) - LONG_L $19, VCPU_R19(k1) - LONG_L $20, VCPU_R20(k1) - LONG_L $21, VCPU_R21(k1) - LONG_L $22, VCPU_R22(k1) - LONG_L $23, VCPU_R23(k1) - LONG_L $24, VCPU_R24(k1) - LONG_L $25, VCPU_R25(k1) - - /* k0/k1 loaded up later */ - - LONG_L $28, VCPU_R28(k1) - LONG_L $29, VCPU_R29(k1) - LONG_L $30, VCPU_R30(k1) - LONG_L $31, VCPU_R31(k1) - - /* Restore hi/lo */ - LONG_L k0, VCPU_LO(k1) - mtlo k0 - - LONG_L k0, VCPU_HI(k1) - mthi k0 - -FEXPORT(__kvm_mips_load_k0k1) - /* Restore the guest's k0/k1 registers */ - LONG_L k0, VCPU_R26(k1) - LONG_L k1, VCPU_R27(k1) - - /* Jump to guest */ - eret -EXPORT(__kvm_mips_vcpu_run_end) - -VECTOR(MIPSX(exception), unknown) -/* Find out what mode we came from and jump to the proper handler. */ - mtc0 k0, CP0_ERROREPC #01: Save guest k0 - ehb #02: - - mfc0 k0, CP0_EBASE #02: Get EBASE - INT_SRL k0, k0, 10 #03: Get rid of CPUNum - INT_SLL k0, k0, 10 #04 - LONG_S k1, 0x3000(k0) #05: Save k1 @ offset 0x3000 - INT_ADDIU k0, k0, 0x2000 #06: Exception handler is - # installed @ offset 0x2000 - j k0 #07: jump to the function - nop #08: branch delay slot -VECTOR_END(MIPSX(exceptionEnd)) -.end MIPSX(exception) - -/* - * Generic Guest exception handler. We end up here when the guest - * does something that causes a trap to kernel mode. - */ -NESTED (MIPSX(GuestException), CALLFRAME_SIZ, ra) - /* Get the VCPU pointer from DDTATA_LO */ - mfc0 k1, CP0_DDATA_LO - INT_ADDIU k1, k1, VCPU_HOST_ARCH - - /* Start saving Guest context to VCPU */ - LONG_S $0, VCPU_R0(k1) - LONG_S $1, VCPU_R1(k1) - LONG_S $2, VCPU_R2(k1) - LONG_S $3, VCPU_R3(k1) - LONG_S $4, VCPU_R4(k1) - LONG_S $5, VCPU_R5(k1) - LONG_S $6, VCPU_R6(k1) - LONG_S $7, VCPU_R7(k1) - LONG_S $8, VCPU_R8(k1) - LONG_S $9, VCPU_R9(k1) - LONG_S $10, VCPU_R10(k1) - LONG_S $11, VCPU_R11(k1) - LONG_S $12, VCPU_R12(k1) - LONG_S $13, VCPU_R13(k1) - LONG_S $14, VCPU_R14(k1) - LONG_S $15, VCPU_R15(k1) - LONG_S $16, VCPU_R16(k1) - LONG_S $17, VCPU_R17(k1) - LONG_S $18, VCPU_R18(k1) - LONG_S $19, VCPU_R19(k1) - LONG_S $20, VCPU_R20(k1) - LONG_S $21, VCPU_R21(k1) - LONG_S $22, VCPU_R22(k1) - LONG_S $23, VCPU_R23(k1) - LONG_S $24, VCPU_R24(k1) - LONG_S $25, VCPU_R25(k1) - - /* Guest k0/k1 saved later */ - - LONG_S $28, VCPU_R28(k1) - LONG_S $29, VCPU_R29(k1) - LONG_S $30, VCPU_R30(k1) - LONG_S $31, VCPU_R31(k1) - - .set at - - /* We need to save hi/lo and restore them on the way out */ - mfhi t0 - LONG_S t0, VCPU_HI(k1) - - mflo t0 - LONG_S t0, VCPU_LO(k1) - - /* Finally save guest k0/k1 to VCPU */ - mfc0 t0, CP0_ERROREPC - LONG_S t0, VCPU_R26(k1) - - /* Get GUEST k1 and save it in VCPU */ - PTR_LI t1, ~0x2ff - mfc0 t0, CP0_EBASE - and t0, t0, t1 - LONG_L t0, 0x3000(t0) - LONG_S t0, VCPU_R27(k1) - - /* Now that context has been saved, we can use other registers */ - - /* Restore vcpu */ - mfc0 a1, CP0_DDATA_LO - move s1, a1 - - /* Restore run (vcpu->run) */ - LONG_L a0, VCPU_RUN(a1) - /* Save pointer to run in s0, will be saved by the compiler */ - move s0, a0 - - /* - * Save Host level EPC, BadVaddr and Cause to VCPU, useful to - * process the exception - */ - mfc0 k0,CP0_EPC - LONG_S k0, VCPU_PC(k1) - - mfc0 k0, CP0_BADVADDR - LONG_S k0, VCPU_HOST_CP0_BADVADDR(k1) - - mfc0 k0, CP0_CAUSE - LONG_S k0, VCPU_HOST_CP0_CAUSE(k1) - - mfc0 k0, CP0_ENTRYHI - LONG_S k0, VCPU_HOST_ENTRYHI(k1) - - /* Now restore the host state just enough to run the handlers */ - - /* Switch EBASE to the one used by Linux */ - /* load up the host EBASE */ - mfc0 v0, CP0_STATUS - - or k0, v0, ST0_BEV - - mtc0 k0, CP0_STATUS - ehb - - LONG_L k0, VCPU_HOST_EBASE(k1) - mtc0 k0,CP0_EBASE - - /* - * If FPU is enabled, save FCR31 and clear it so that later ctc1's don't - * trigger FPE for pending exceptions. - */ - and v1, v0, ST0_CU1 - beqz v1, 1f - nop - .set push - SET_HARDFLOAT - cfc1 t0, fcr31 - sw t0, VCPU_FCR31(k1) - ctc1 zero,fcr31 - .set pop -1: - -#ifdef CONFIG_CPU_HAS_MSA - /* - * If MSA is enabled, save MSACSR and clear it so that later - * instructions don't trigger MSAFPE for pending exceptions. - */ - mfc0 t0, CP0_CONFIG3 - ext t0, t0, 28, 1 /* MIPS_CONF3_MSAP */ - beqz t0, 1f - nop - mfc0 t0, CP0_CONFIG5 - ext t0, t0, 27, 1 /* MIPS_CONF5_MSAEN */ - beqz t0, 1f - nop - _cfcmsa t0, MSA_CSR - sw t0, VCPU_MSA_CSR(k1) - _ctcmsa MSA_CSR, zero -1: -#endif - - /* Now that the new EBASE has been loaded, unset BEV and KSU_USER */ - and v0, v0, ~(ST0_EXL | KSU_USER | ST0_IE) - or v0, v0, ST0_CU0 - mtc0 v0, CP0_STATUS - ehb - - /* Load up host GP */ - LONG_L gp, VCPU_HOST_GP(k1) - - /* Need a stack before we can jump to "C" */ - LONG_L sp, VCPU_HOST_STACK(k1) - - /* Saved host state */ - INT_ADDIU sp, sp, -PT_SIZE - - /* - * XXXKYMA do we need to load the host ASID, maybe not because the - * kernel entries are marked GLOBAL, need to verify - */ - - /* Restore host DDATA_LO */ - LONG_L k0, PT_HOST_USERLOCAL(sp) - mtc0 k0, CP0_DDATA_LO - - /* Restore RDHWR access */ - PTR_LI k0, 0x2000000F - mtc0 k0, CP0_HWRENA - - /* Jump to handler */ -FEXPORT(__kvm_mips_jump_to_handler) - /* - * XXXKYMA: not sure if this is safe, how large is the stack?? - * Now jump to the kvm_mips_handle_exit() to see if we can deal - * with this in the kernel - */ - PTR_LA t9, kvm_mips_handle_exit - jalr.hb t9 - INT_ADDIU sp, sp, -CALLFRAME_SIZ /* BD Slot */ - - /* Return from handler Make sure interrupts are disabled */ - di - ehb - - /* - * XXXKYMA: k0/k1 could have been blown away if we processed - * an exception while we were handling the exception from the - * guest, reload k1 - */ - - move k1, s1 - INT_ADDIU k1, k1, VCPU_HOST_ARCH - - /* - * Check return value, should tell us if we are returning to the - * host (handle I/O etc)or resuming the guest - */ - andi t0, v0, RESUME_HOST - bnez t0, __kvm_mips_return_to_host - nop - -__kvm_mips_return_to_guest: - /* Put the saved pointer to vcpu (s1) back into the DDATA_LO Register */ - mtc0 s1, CP0_DDATA_LO - - /* Load up the Guest EBASE to minimize the window where BEV is set */ - LONG_L t0, VCPU_GUEST_EBASE(k1) - - /* Switch EBASE back to the one used by KVM */ - mfc0 v1, CP0_STATUS - or k0, v1, ST0_BEV - mtc0 k0, CP0_STATUS - ehb - mtc0 t0, CP0_EBASE - - /* Setup status register for running guest in UM */ - or v1, v1, (ST0_EXL | KSU_USER | ST0_IE) - and v1, v1, ~(ST0_CU0 | ST0_MX) - mtc0 v1, CP0_STATUS - ehb - - /* Set Guest EPC */ - LONG_L t0, VCPU_PC(k1) - mtc0 t0, CP0_EPC - - /* Set the ASID for the Guest Kernel */ - PTR_L t0, VCPU_COP0(k1) - LONG_L t0, COP0_STATUS(t0) - andi t0, KSU_USER | ST0_ERL | ST0_EXL - xori t0, KSU_USER - bnez t0, 1f /* If kernel */ - INT_ADDIU t1, k1, VCPU_GUEST_KERNEL_ASID /* (BD) */ - INT_ADDIU t1, k1, VCPU_GUEST_USER_ASID /* else user */ -1: - /* t1: contains the base of the ASID array, need to get the cpu id */ - LONG_L t2, TI_CPU($28) /* smp_processor_id */ - INT_SLL t2, t2, 2 /* x4 */ - REG_ADDU t3, t1, t2 - LONG_L k0, (t3) -#ifdef CONFIG_MIPS_ASID_BITS_VARIABLE - li t3, CPUINFO_SIZE/4 - mul t2, t2, t3 /* x sizeof(struct cpuinfo_mips)/4 */ - LONG_L t2, (cpu_data + CPUINFO_ASID_MASK)(t2) - and k0, k0, t2 -#else - andi k0, k0, MIPS_ENTRYHI_ASID -#endif - mtc0 k0, CP0_ENTRYHI - ehb - - /* Disable RDHWR access */ - mtc0 zero, CP0_HWRENA - - .set noat - /* load the guest context from VCPU and return */ - LONG_L $0, VCPU_R0(k1) - LONG_L $1, VCPU_R1(k1) - LONG_L $2, VCPU_R2(k1) - LONG_L $3, VCPU_R3(k1) - LONG_L $4, VCPU_R4(k1) - LONG_L $5, VCPU_R5(k1) - LONG_L $6, VCPU_R6(k1) - LONG_L $7, VCPU_R7(k1) - LONG_L $8, VCPU_R8(k1) - LONG_L $9, VCPU_R9(k1) - LONG_L $10, VCPU_R10(k1) - LONG_L $11, VCPU_R11(k1) - LONG_L $12, VCPU_R12(k1) - LONG_L $13, VCPU_R13(k1) - LONG_L $14, VCPU_R14(k1) - LONG_L $15, VCPU_R15(k1) - LONG_L $16, VCPU_R16(k1) - LONG_L $17, VCPU_R17(k1) - LONG_L $18, VCPU_R18(k1) - LONG_L $19, VCPU_R19(k1) - LONG_L $20, VCPU_R20(k1) - LONG_L $21, VCPU_R21(k1) - LONG_L $22, VCPU_R22(k1) - LONG_L $23, VCPU_R23(k1) - LONG_L $24, VCPU_R24(k1) - LONG_L $25, VCPU_R25(k1) - - /* $/k1 loaded later */ - LONG_L $28, VCPU_R28(k1) - LONG_L $29, VCPU_R29(k1) - LONG_L $30, VCPU_R30(k1) - LONG_L $31, VCPU_R31(k1) - -FEXPORT(__kvm_mips_skip_guest_restore) - LONG_L k0, VCPU_HI(k1) - mthi k0 - - LONG_L k0, VCPU_LO(k1) - mtlo k0 - - LONG_L k0, VCPU_R26(k1) - LONG_L k1, VCPU_R27(k1) - - eret - .set at - -__kvm_mips_return_to_host: - /* EBASE is already pointing to Linux */ - LONG_L k1, VCPU_HOST_STACK(k1) - INT_ADDIU k1,k1, -PT_SIZE - - /* Restore host DDATA_LO */ - LONG_L k0, PT_HOST_USERLOCAL(k1) - mtc0 k0, CP0_DDATA_LO - - /* - * r2/v0 is the return code, shift it down by 2 (arithmetic) - * to recover the err code - */ - INT_SRA k0, v0, 2 - move $2, k0 - - /* Load context saved on the host stack */ - LONG_L $16, PT_R16(k1) - LONG_L $17, PT_R17(k1) - LONG_L $18, PT_R18(k1) - LONG_L $19, PT_R19(k1) - LONG_L $20, PT_R20(k1) - LONG_L $21, PT_R21(k1) - LONG_L $22, PT_R22(k1) - LONG_L $23, PT_R23(k1) - - LONG_L $28, PT_R28(k1) - LONG_L $29, PT_R29(k1) - LONG_L $30, PT_R30(k1) - - LONG_L k0, PT_HI(k1) - mthi k0 - - LONG_L k0, PT_LO(k1) - mtlo k0 - - /* Restore RDHWR access */ - PTR_LI k0, 0x2000000F - mtc0 k0, CP0_HWRENA - - /* Restore RA, which is the address we will return to */ - LONG_L ra, PT_R31(k1) - j ra - nop - -VECTOR_END(MIPSX(GuestExceptionEnd)) -.end MIPSX(GuestException) - -MIPSX(exceptions): - #### - ##### The exception handlers. - ##### - .word _C_LABEL(MIPSX(GuestException)) # 0 - .word _C_LABEL(MIPSX(GuestException)) # 1 - .word _C_LABEL(MIPSX(GuestException)) # 2 - .word _C_LABEL(MIPSX(GuestException)) # 3 - .word _C_LABEL(MIPSX(GuestException)) # 4 - .word _C_LABEL(MIPSX(GuestException)) # 5 - .word _C_LABEL(MIPSX(GuestException)) # 6 - .word _C_LABEL(MIPSX(GuestException)) # 7 - .word _C_LABEL(MIPSX(GuestException)) # 8 - .word _C_LABEL(MIPSX(GuestException)) # 9 - .word _C_LABEL(MIPSX(GuestException)) # 10 - .word _C_LABEL(MIPSX(GuestException)) # 11 - .word _C_LABEL(MIPSX(GuestException)) # 12 - .word _C_LABEL(MIPSX(GuestException)) # 13 - .word _C_LABEL(MIPSX(GuestException)) # 14 - .word _C_LABEL(MIPSX(GuestException)) # 15 - .word _C_LABEL(MIPSX(GuestException)) # 16 - .word _C_LABEL(MIPSX(GuestException)) # 17 - .word _C_LABEL(MIPSX(GuestException)) # 18 - .word _C_LABEL(MIPSX(GuestException)) # 19 - .word _C_LABEL(MIPSX(GuestException)) # 20 - .word _C_LABEL(MIPSX(GuestException)) # 21 - .word _C_LABEL(MIPSX(GuestException)) # 22 - .word _C_LABEL(MIPSX(GuestException)) # 23 - .word _C_LABEL(MIPSX(GuestException)) # 24 - .word _C_LABEL(MIPSX(GuestException)) # 25 - .word _C_LABEL(MIPSX(GuestException)) # 26 - .word _C_LABEL(MIPSX(GuestException)) # 27 - .word _C_LABEL(MIPSX(GuestException)) # 28 - .word _C_LABEL(MIPSX(GuestException)) # 29 - .word _C_LABEL(MIPSX(GuestException)) # 30 - .word _C_LABEL(MIPSX(GuestException)) # 31 diff --git a/arch/mips/kvm/mips.c b/arch/mips/kvm/mips.c index 44da525..a6ea084 100644 --- a/arch/mips/kvm/mips.c +++ b/arch/mips/kvm/mips.c @@ -9,6 +9,7 @@ * Authors: Sanjay Lal <sanjayl@kymasys.com> */ +#include <linux/bitops.h> #include <linux/errno.h> #include <linux/err.h> #include <linux/kdebug.h> @@ -147,7 +148,7 @@ void kvm_mips_free_vcpus(struct kvm *kvm) /* Put the pages we reserved for the guest pmap */ for (i = 0; i < kvm->arch.guest_pmap_npages; i++) { if (kvm->arch.guest_pmap[i] != KVM_INVALID_PAGE) - kvm_mips_release_pfn_clean(kvm->arch.guest_pmap[i]); + kvm_release_pfn_clean(kvm->arch.guest_pmap[i]); } kfree(kvm->arch.guest_pmap); @@ -244,10 +245,27 @@ void kvm_arch_commit_memory_region(struct kvm *kvm, } } +static inline void dump_handler(const char *symbol, void *start, void *end) +{ + u32 *p; + + pr_debug("LEAF(%s)\n", symbol); + + pr_debug("\t.set push\n"); + pr_debug("\t.set noreorder\n"); + + for (p = start; p < (u32 *)end; ++p) + pr_debug("\t.word\t0x%08x\t\t# %p\n", *p, p); + + pr_debug("\t.set\tpop\n"); + + pr_debug("\tEND(%s)\n", symbol); +} + struct kvm_vcpu *kvm_arch_vcpu_create(struct kvm *kvm, unsigned int id) { - int err, size, offset; - void *gebase; + int err, size; + void *gebase, *p, *handler; int i; struct kvm_vcpu *vcpu = kzalloc(sizeof(struct kvm_vcpu), GFP_KERNEL); @@ -273,9 +291,6 @@ struct kvm_vcpu *kvm_arch_vcpu_create(struct kvm *kvm, unsigned int id) else size = 0x4000; - /* Save Linux EBASE */ - vcpu->arch.host_ebase = (void *)read_c0_ebase(); - gebase = kzalloc(ALIGN(size, PAGE_SIZE), GFP_KERNEL); if (!gebase) { @@ -285,44 +300,53 @@ struct kvm_vcpu *kvm_arch_vcpu_create(struct kvm *kvm, unsigned int id) kvm_debug("Allocated %d bytes for KVM Exception Handlers @ %p\n", ALIGN(size, PAGE_SIZE), gebase); + /* + * Check new ebase actually fits in CP0_EBase. The lack of a write gate + * limits us to the low 512MB of physical address space. If the memory + * we allocate is out of range, just give up now. + */ + if (!cpu_has_ebase_wg && virt_to_phys(gebase) >= 0x20000000) { + kvm_err("CP0_EBase.WG required for guest exception base %pK\n", + gebase); + err = -ENOMEM; + goto out_free_gebase; + } + /* Save new ebase */ vcpu->arch.guest_ebase = gebase; - /* Copy L1 Guest Exception handler to correct offset */ + /* Build guest exception vectors dynamically in unmapped memory */ + handler = gebase + 0x2000; /* TLB Refill, EXL = 0 */ - memcpy(gebase, mips32_exception, - mips32_exceptionEnd - mips32_exception); + kvm_mips_build_exception(gebase, handler); /* General Exception Entry point */ - memcpy(gebase + 0x180, mips32_exception, - mips32_exceptionEnd - mips32_exception); + kvm_mips_build_exception(gebase + 0x180, handler); /* For vectored interrupts poke the exception code @ all offsets 0-7 */ for (i = 0; i < 8; i++) { kvm_debug("L1 Vectored handler @ %p\n", gebase + 0x200 + (i * VECTORSPACING)); - memcpy(gebase + 0x200 + (i * VECTORSPACING), mips32_exception, - mips32_exceptionEnd - mips32_exception); + kvm_mips_build_exception(gebase + 0x200 + i * VECTORSPACING, + handler); } - /* General handler, relocate to unmapped space for sanity's sake */ - offset = 0x2000; - kvm_debug("Installing KVM Exception handlers @ %p, %#x bytes\n", - gebase + offset, - mips32_GuestExceptionEnd - mips32_GuestException); + /* General exit handler */ + p = handler; + p = kvm_mips_build_exit(p); - memcpy(gebase + offset, mips32_GuestException, - mips32_GuestExceptionEnd - mips32_GuestException); + /* Guest entry routine */ + vcpu->arch.vcpu_run = p; + p = kvm_mips_build_vcpu_run(p); -#ifdef MODULE - offset += mips32_GuestExceptionEnd - mips32_GuestException; - memcpy(gebase + offset, (char *)__kvm_mips_vcpu_run, - __kvm_mips_vcpu_run_end - (char *)__kvm_mips_vcpu_run); - vcpu->arch.vcpu_run = gebase + offset; -#else - vcpu->arch.vcpu_run = __kvm_mips_vcpu_run; -#endif + /* Dump the generated code */ + pr_debug("#include <asm/asm.h>\n"); + pr_debug("#include <asm/regdef.h>\n"); + pr_debug("\n"); + dump_handler("kvm_vcpu_run", vcpu->arch.vcpu_run, p); + dump_handler("kvm_gen_exc", gebase + 0x180, gebase + 0x200); + dump_handler("kvm_exit", gebase + 0x2000, vcpu->arch.vcpu_run); /* Invalidate the icache for these ranges */ local_flush_icache_range((unsigned long)gebase, @@ -408,17 +432,19 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *run) kvm_mips_deliver_interrupts(vcpu, kvm_read_c0_guest_cause(vcpu->arch.cop0)); - __kvm_guest_enter(); + guest_enter_irqoff(); /* Disable hardware page table walking while in guest */ htw_stop(); + trace_kvm_enter(vcpu); r = vcpu->arch.vcpu_run(run, vcpu); + trace_kvm_out(vcpu); /* Re-enable HTW before enabling interrupts */ htw_start(); - __kvm_guest_exit(); + guest_exit_irqoff(); local_irq_enable(); if (vcpu->sigset_active) @@ -507,8 +533,10 @@ static u64 kvm_mips_get_one_regs[] = { KVM_REG_MIPS_R30, KVM_REG_MIPS_R31, +#ifndef CONFIG_CPU_MIPSR6 KVM_REG_MIPS_HI, KVM_REG_MIPS_LO, +#endif KVM_REG_MIPS_PC, KVM_REG_MIPS_CP0_INDEX, @@ -539,6 +567,104 @@ static u64 kvm_mips_get_one_regs[] = { KVM_REG_MIPS_COUNT_HZ, }; +static u64 kvm_mips_get_one_regs_fpu[] = { + KVM_REG_MIPS_FCR_IR, + KVM_REG_MIPS_FCR_CSR, +}; + +static u64 kvm_mips_get_one_regs_msa[] = { + KVM_REG_MIPS_MSA_IR, + KVM_REG_MIPS_MSA_CSR, +}; + +static u64 kvm_mips_get_one_regs_kscratch[] = { + KVM_REG_MIPS_CP0_KSCRATCH1, + KVM_REG_MIPS_CP0_KSCRATCH2, + KVM_REG_MIPS_CP0_KSCRATCH3, + KVM_REG_MIPS_CP0_KSCRATCH4, + KVM_REG_MIPS_CP0_KSCRATCH5, + KVM_REG_MIPS_CP0_KSCRATCH6, +}; + +static unsigned long kvm_mips_num_regs(struct kvm_vcpu *vcpu) +{ + unsigned long ret; + + ret = ARRAY_SIZE(kvm_mips_get_one_regs); + if (kvm_mips_guest_can_have_fpu(&vcpu->arch)) { + ret += ARRAY_SIZE(kvm_mips_get_one_regs_fpu) + 48; + /* odd doubles */ + if (boot_cpu_data.fpu_id & MIPS_FPIR_F64) + ret += 16; + } + if (kvm_mips_guest_can_have_msa(&vcpu->arch)) + ret += ARRAY_SIZE(kvm_mips_get_one_regs_msa) + 32; + ret += __arch_hweight8(vcpu->arch.kscratch_enabled); + ret += kvm_mips_callbacks->num_regs(vcpu); + + return ret; +} + +static int kvm_mips_copy_reg_indices(struct kvm_vcpu *vcpu, u64 __user *indices) +{ + u64 index; + unsigned int i; + + if (copy_to_user(indices, kvm_mips_get_one_regs, + sizeof(kvm_mips_get_one_regs))) + return -EFAULT; + indices += ARRAY_SIZE(kvm_mips_get_one_regs); + + if (kvm_mips_guest_can_have_fpu(&vcpu->arch)) { + if (copy_to_user(indices, kvm_mips_get_one_regs_fpu, + sizeof(kvm_mips_get_one_regs_fpu))) + return -EFAULT; + indices += ARRAY_SIZE(kvm_mips_get_one_regs_fpu); + + for (i = 0; i < 32; ++i) { + index = KVM_REG_MIPS_FPR_32(i); + if (copy_to_user(indices, &index, sizeof(index))) + return -EFAULT; + ++indices; + + /* skip odd doubles if no F64 */ + if (i & 1 && !(boot_cpu_data.fpu_id & MIPS_FPIR_F64)) + continue; + + index = KVM_REG_MIPS_FPR_64(i); + if (copy_to_user(indices, &index, sizeof(index))) + return -EFAULT; + ++indices; + } + } + + if (kvm_mips_guest_can_have_msa(&vcpu->arch)) { + if (copy_to_user(indices, kvm_mips_get_one_regs_msa, + sizeof(kvm_mips_get_one_regs_msa))) + return -EFAULT; + indices += ARRAY_SIZE(kvm_mips_get_one_regs_msa); + + for (i = 0; i < 32; ++i) { + index = KVM_REG_MIPS_VEC_128(i); + if (copy_to_user(indices, &index, sizeof(index))) + return -EFAULT; + ++indices; + } + } + + for (i = 0; i < 6; ++i) { + if (!(vcpu->arch.kscratch_enabled & BIT(i + 2))) + continue; + + if (copy_to_user(indices, &kvm_mips_get_one_regs_kscratch[i], + sizeof(kvm_mips_get_one_regs_kscratch[i]))) + return -EFAULT; + ++indices; + } + + return kvm_mips_callbacks->copy_reg_indices(vcpu, indices); +} + static int kvm_mips_get_reg(struct kvm_vcpu *vcpu, const struct kvm_one_reg *reg) { @@ -554,12 +680,14 @@ static int kvm_mips_get_reg(struct kvm_vcpu *vcpu, case KVM_REG_MIPS_R0 ... KVM_REG_MIPS_R31: v = (long)vcpu->arch.gprs[reg->id - KVM_REG_MIPS_R0]; break; +#ifndef CONFIG_CPU_MIPSR6 case KVM_REG_MIPS_HI: v = (long)vcpu->arch.hi; break; case KVM_REG_MIPS_LO: v = (long)vcpu->arch.lo; break; +#endif case KVM_REG_MIPS_PC: v = (long)vcpu->arch.pc; break; @@ -688,17 +816,37 @@ static int kvm_mips_get_reg(struct kvm_vcpu *vcpu, case KVM_REG_MIPS_CP0_ERROREPC: v = (long)kvm_read_c0_guest_errorepc(cop0); break; + case KVM_REG_MIPS_CP0_KSCRATCH1 ... KVM_REG_MIPS_CP0_KSCRATCH6: + idx = reg->id - KVM_REG_MIPS_CP0_KSCRATCH1 + 2; + if (!(vcpu->arch.kscratch_enabled & BIT(idx))) + return -EINVAL; + switch (idx) { + case 2: + v = (long)kvm_read_c0_guest_kscratch1(cop0); + break; + case 3: + v = (long)kvm_read_c0_guest_kscratch2(cop0); + break; + case 4: + v = (long)kvm_read_c0_guest_kscratch3(cop0); + break; + case 5: + v = (long)kvm_read_c0_guest_kscratch4(cop0); + break; + case 6: + v = (long)kvm_read_c0_guest_kscratch5(cop0); + break; + case 7: + v = (long)kvm_read_c0_guest_kscratch6(cop0); + break; + } + break; /* registers to be handled specially */ - case KVM_REG_MIPS_CP0_COUNT: - case KVM_REG_MIPS_COUNT_CTL: - case KVM_REG_MIPS_COUNT_RESUME: - case KVM_REG_MIPS_COUNT_HZ: + default: ret = kvm_mips_callbacks->get_one_reg(vcpu, reg, &v); if (ret) return ret; break; - default: - return -EINVAL; } if ((reg->id & KVM_REG_SIZE_MASK) == KVM_REG_SIZE_U64) { u64 __user *uaddr64 = (u64 __user *)(long)reg->addr; @@ -755,12 +903,14 @@ static int kvm_mips_set_reg(struct kvm_vcpu *vcpu, case KVM_REG_MIPS_R1 ... KVM_REG_MIPS_R31: vcpu->arch.gprs[reg->id - KVM_REG_MIPS_R0] = v; break; +#ifndef CONFIG_CPU_MIPSR6 case KVM_REG_MIPS_HI: vcpu->arch.hi = v; break; case KVM_REG_MIPS_LO: vcpu->arch.lo = v; break; +#endif case KVM_REG_MIPS_PC: vcpu->arch.pc = v; break; @@ -859,22 +1009,34 @@ static int kvm_mips_set_reg(struct kvm_vcpu *vcpu, case KVM_REG_MIPS_CP0_ERROREPC: kvm_write_c0_guest_errorepc(cop0, v); break; + case KVM_REG_MIPS_CP0_KSCRATCH1 ... KVM_REG_MIPS_CP0_KSCRATCH6: + idx = reg->id - KVM_REG_MIPS_CP0_KSCRATCH1 + 2; + if (!(vcpu->arch.kscratch_enabled & BIT(idx))) + return -EINVAL; + switch (idx) { + case 2: + kvm_write_c0_guest_kscratch1(cop0, v); + break; + case 3: + kvm_write_c0_guest_kscratch2(cop0, v); + break; + case 4: + kvm_write_c0_guest_kscratch3(cop0, v); + break; + case 5: + kvm_write_c0_guest_kscratch4(cop0, v); + break; + case 6: + kvm_write_c0_guest_kscratch5(cop0, v); + break; + case 7: + kvm_write_c0_guest_kscratch6(cop0, v); + break; + } + break; /* registers to be handled specially */ - case KVM_REG_MIPS_CP0_COUNT: - case KVM_REG_MIPS_CP0_COMPARE: - case KVM_REG_MIPS_CP0_CAUSE: - case KVM_REG_MIPS_CP0_CONFIG: - case KVM_REG_MIPS_CP0_CONFIG1: - case KVM_REG_MIPS_CP0_CONFIG2: - case KVM_REG_MIPS_CP0_CONFIG3: - case KVM_REG_MIPS_CP0_CONFIG4: - case KVM_REG_MIPS_CP0_CONFIG5: - case KVM_REG_MIPS_COUNT_CTL: - case KVM_REG_MIPS_COUNT_RESUME: - case KVM_REG_MIPS_COUNT_HZ: - return kvm_mips_callbacks->set_one_reg(vcpu, reg, v); default: - return -EINVAL; + return kvm_mips_callbacks->set_one_reg(vcpu, reg, v); } return 0; } @@ -927,23 +1089,18 @@ long kvm_arch_vcpu_ioctl(struct file *filp, unsigned int ioctl, } case KVM_GET_REG_LIST: { struct kvm_reg_list __user *user_list = argp; - u64 __user *reg_dest; struct kvm_reg_list reg_list; unsigned n; if (copy_from_user(®_list, user_list, sizeof(reg_list))) return -EFAULT; n = reg_list.n; - reg_list.n = ARRAY_SIZE(kvm_mips_get_one_regs); + reg_list.n = kvm_mips_num_regs(vcpu); if (copy_to_user(user_list, ®_list, sizeof(reg_list))) return -EFAULT; if (n < reg_list.n) return -E2BIG; - reg_dest = user_list->reg; - if (copy_to_user(reg_dest, kvm_mips_get_one_regs, - sizeof(kvm_mips_get_one_regs))) - return -EFAULT; - return 0; + return kvm_mips_copy_reg_indices(vcpu, user_list->reg); } case KVM_NMI: /* Treat the NMI as a CPU reset */ @@ -1222,7 +1379,7 @@ int kvm_arch_vcpu_setup(struct kvm_vcpu *vcpu) static void kvm_mips_set_c0_status(void) { - uint32_t status = read_c0_status(); + u32 status = read_c0_status(); if (cpu_has_dsp) status |= (ST0_MX); @@ -1236,9 +1393,9 @@ static void kvm_mips_set_c0_status(void) */ int kvm_mips_handle_exit(struct kvm_run *run, struct kvm_vcpu *vcpu) { - uint32_t cause = vcpu->arch.host_cp0_cause; - uint32_t exccode = (cause >> CAUSEB_EXCCODE) & 0x1f; - uint32_t __user *opc = (uint32_t __user *) vcpu->arch.pc; + u32 cause = vcpu->arch.host_cp0_cause; + u32 exccode = (cause >> CAUSEB_EXCCODE) & 0x1f; + u32 __user *opc = (u32 __user *) vcpu->arch.pc; unsigned long badvaddr = vcpu->arch.host_cp0_badvaddr; enum emulation_result er = EMULATE_DONE; int ret = RESUME_GUEST; @@ -1260,6 +1417,7 @@ int kvm_mips_handle_exit(struct kvm_run *run, struct kvm_vcpu *vcpu) kvm_debug("kvm_mips_handle_exit: cause: %#x, PC: %p, kvm_run: %p, kvm_vcpu: %p\n", cause, opc, run, vcpu); + trace_kvm_exit(vcpu, exccode); /* * Do a privilege check, if in UM most of these exit conditions end up @@ -1279,7 +1437,6 @@ int kvm_mips_handle_exit(struct kvm_run *run, struct kvm_vcpu *vcpu) kvm_debug("[%d]EXCCODE_INT @ %p\n", vcpu->vcpu_id, opc); ++vcpu->stat.int_exits; - trace_kvm_exit(vcpu, INT_EXITS); if (need_resched()) cond_resched(); @@ -1291,7 +1448,6 @@ int kvm_mips_handle_exit(struct kvm_run *run, struct kvm_vcpu *vcpu) kvm_debug("EXCCODE_CPU: @ PC: %p\n", opc); ++vcpu->stat.cop_unusable_exits; - trace_kvm_exit(vcpu, COP_UNUSABLE_EXITS); ret = kvm_mips_callbacks->handle_cop_unusable(vcpu); /* XXXKYMA: Might need to return to user space */ if (run->exit_reason == KVM_EXIT_IRQ_WINDOW_OPEN) @@ -1300,7 +1456,6 @@ int kvm_mips_handle_exit(struct kvm_run *run, struct kvm_vcpu *vcpu) case EXCCODE_MOD: ++vcpu->stat.tlbmod_exits; - trace_kvm_exit(vcpu, TLBMOD_EXITS); ret = kvm_mips_callbacks->handle_tlb_mod(vcpu); break; @@ -1310,7 +1465,6 @@ int kvm_mips_handle_exit(struct kvm_run *run, struct kvm_vcpu *vcpu) badvaddr); ++vcpu->stat.tlbmiss_st_exits; - trace_kvm_exit(vcpu, TLBMISS_ST_EXITS); ret = kvm_mips_callbacks->handle_tlb_st_miss(vcpu); break; @@ -1319,61 +1473,51 @@ int kvm_mips_handle_exit(struct kvm_run *run, struct kvm_vcpu *vcpu) cause, opc, badvaddr); ++vcpu->stat.tlbmiss_ld_exits; - trace_kvm_exit(vcpu, TLBMISS_LD_EXITS); ret = kvm_mips_callbacks->handle_tlb_ld_miss(vcpu); break; case EXCCODE_ADES: ++vcpu->stat.addrerr_st_exits; - trace_kvm_exit(vcpu, ADDRERR_ST_EXITS); ret = kvm_mips_callbacks->handle_addr_err_st(vcpu); break; case EXCCODE_ADEL: ++vcpu->stat.addrerr_ld_exits; - trace_kvm_exit(vcpu, ADDRERR_LD_EXITS); ret = kvm_mips_callbacks->handle_addr_err_ld(vcpu); break; case EXCCODE_SYS: ++vcpu->stat.syscall_exits; - trace_kvm_exit(vcpu, SYSCALL_EXITS); ret = kvm_mips_callbacks->handle_syscall(vcpu); break; case EXCCODE_RI: ++vcpu->stat.resvd_inst_exits; - trace_kvm_exit(vcpu, RESVD_INST_EXITS); ret = kvm_mips_callbacks->handle_res_inst(vcpu); break; case EXCCODE_BP: ++vcpu->stat.break_inst_exits; - trace_kvm_exit(vcpu, BREAK_INST_EXITS); ret = kvm_mips_callbacks->handle_break(vcpu); break; case EXCCODE_TR: ++vcpu->stat.trap_inst_exits; - trace_kvm_exit(vcpu, TRAP_INST_EXITS); ret = kvm_mips_callbacks->handle_trap(vcpu); break; case EXCCODE_MSAFPE: ++vcpu->stat.msa_fpe_exits; - trace_kvm_exit(vcpu, MSA_FPE_EXITS); ret = kvm_mips_callbacks->handle_msa_fpe(vcpu); break; case EXCCODE_FPE: ++vcpu->stat.fpe_exits; - trace_kvm_exit(vcpu, FPE_EXITS); ret = kvm_mips_callbacks->handle_fpe(vcpu); break; case EXCCODE_MSADIS: ++vcpu->stat.msa_disabled_exits; - trace_kvm_exit(vcpu, MSA_DISABLED_EXITS); ret = kvm_mips_callbacks->handle_msa_disabled(vcpu); break; @@ -1400,11 +1544,13 @@ skip_emul: run->exit_reason = KVM_EXIT_INTR; ret = (-EINTR << 2) | RESUME_HOST; ++vcpu->stat.signal_exits; - trace_kvm_exit(vcpu, SIGNAL_EXITS); + trace_kvm_exit(vcpu, KVM_TRACE_EXIT_SIGNAL); } } if (ret == RESUME_GUEST) { + trace_kvm_reenter(vcpu); + /* * If FPU / MSA are enabled (i.e. the guest's FPU / MSA context * is live), restore FCR31 / MSACSR. @@ -1450,7 +1596,7 @@ void kvm_own_fpu(struct kvm_vcpu *vcpu) * not to clobber the status register directly via the commpage. */ if (cpu_has_msa && sr & ST0_CU1 && !(sr & ST0_FR) && - vcpu->arch.fpu_inuse & KVM_MIPS_FPU_MSA) + vcpu->arch.aux_inuse & KVM_MIPS_AUX_MSA) kvm_lose_fpu(vcpu); /* @@ -1465,9 +1611,12 @@ void kvm_own_fpu(struct kvm_vcpu *vcpu) enable_fpu_hazard(); /* If guest FPU state not active, restore it now */ - if (!(vcpu->arch.fpu_inuse & KVM_MIPS_FPU_FPU)) { + if (!(vcpu->arch.aux_inuse & KVM_MIPS_AUX_FPU)) { __kvm_restore_fpu(&vcpu->arch); - vcpu->arch.fpu_inuse |= KVM_MIPS_FPU_FPU; + vcpu->arch.aux_inuse |= KVM_MIPS_AUX_FPU; + trace_kvm_aux(vcpu, KVM_TRACE_AUX_RESTORE, KVM_TRACE_AUX_FPU); + } else { + trace_kvm_aux(vcpu, KVM_TRACE_AUX_ENABLE, KVM_TRACE_AUX_FPU); } preempt_enable(); @@ -1494,8 +1643,8 @@ void kvm_own_msa(struct kvm_vcpu *vcpu) * interacts with MSA state, so play it safe and save it first. */ if (!(sr & ST0_FR) && - (vcpu->arch.fpu_inuse & (KVM_MIPS_FPU_FPU | - KVM_MIPS_FPU_MSA)) == KVM_MIPS_FPU_FPU) + (vcpu->arch.aux_inuse & (KVM_MIPS_AUX_FPU | + KVM_MIPS_AUX_MSA)) == KVM_MIPS_AUX_FPU) kvm_lose_fpu(vcpu); change_c0_status(ST0_CU1 | ST0_FR, sr); @@ -1509,22 +1658,26 @@ void kvm_own_msa(struct kvm_vcpu *vcpu) set_c0_config5(MIPS_CONF5_MSAEN); enable_fpu_hazard(); - switch (vcpu->arch.fpu_inuse & (KVM_MIPS_FPU_FPU | KVM_MIPS_FPU_MSA)) { - case KVM_MIPS_FPU_FPU: + switch (vcpu->arch.aux_inuse & (KVM_MIPS_AUX_FPU | KVM_MIPS_AUX_MSA)) { + case KVM_MIPS_AUX_FPU: /* * Guest FPU state already loaded, only restore upper MSA state */ __kvm_restore_msa_upper(&vcpu->arch); - vcpu->arch.fpu_inuse |= KVM_MIPS_FPU_MSA; + vcpu->arch.aux_inuse |= KVM_MIPS_AUX_MSA; + trace_kvm_aux(vcpu, KVM_TRACE_AUX_RESTORE, KVM_TRACE_AUX_MSA); break; case 0: /* Neither FPU or MSA already active, restore full MSA state */ __kvm_restore_msa(&vcpu->arch); - vcpu->arch.fpu_inuse |= KVM_MIPS_FPU_MSA; + vcpu->arch.aux_inuse |= KVM_MIPS_AUX_MSA; if (kvm_mips_guest_has_fpu(&vcpu->arch)) - vcpu->arch.fpu_inuse |= KVM_MIPS_FPU_FPU; + vcpu->arch.aux_inuse |= KVM_MIPS_AUX_FPU; + trace_kvm_aux(vcpu, KVM_TRACE_AUX_RESTORE, + KVM_TRACE_AUX_FPU_MSA); break; default: + trace_kvm_aux(vcpu, KVM_TRACE_AUX_ENABLE, KVM_TRACE_AUX_MSA); break; } @@ -1536,13 +1689,15 @@ void kvm_own_msa(struct kvm_vcpu *vcpu) void kvm_drop_fpu(struct kvm_vcpu *vcpu) { preempt_disable(); - if (cpu_has_msa && vcpu->arch.fpu_inuse & KVM_MIPS_FPU_MSA) { + if (cpu_has_msa && vcpu->arch.aux_inuse & KVM_MIPS_AUX_MSA) { disable_msa(); - vcpu->arch.fpu_inuse &= ~KVM_MIPS_FPU_MSA; + trace_kvm_aux(vcpu, KVM_TRACE_AUX_DISCARD, KVM_TRACE_AUX_MSA); + vcpu->arch.aux_inuse &= ~KVM_MIPS_AUX_MSA; } - if (vcpu->arch.fpu_inuse & KVM_MIPS_FPU_FPU) { + if (vcpu->arch.aux_inuse & KVM_MIPS_AUX_FPU) { clear_c0_status(ST0_CU1 | ST0_FR); - vcpu->arch.fpu_inuse &= ~KVM_MIPS_FPU_FPU; + trace_kvm_aux(vcpu, KVM_TRACE_AUX_DISCARD, KVM_TRACE_AUX_FPU); + vcpu->arch.aux_inuse &= ~KVM_MIPS_AUX_FPU; } preempt_enable(); } @@ -1558,25 +1713,27 @@ void kvm_lose_fpu(struct kvm_vcpu *vcpu) */ preempt_disable(); - if (cpu_has_msa && vcpu->arch.fpu_inuse & KVM_MIPS_FPU_MSA) { + if (cpu_has_msa && vcpu->arch.aux_inuse & KVM_MIPS_AUX_MSA) { set_c0_config5(MIPS_CONF5_MSAEN); enable_fpu_hazard(); __kvm_save_msa(&vcpu->arch); + trace_kvm_aux(vcpu, KVM_TRACE_AUX_SAVE, KVM_TRACE_AUX_FPU_MSA); /* Disable MSA & FPU */ disable_msa(); - if (vcpu->arch.fpu_inuse & KVM_MIPS_FPU_FPU) { + if (vcpu->arch.aux_inuse & KVM_MIPS_AUX_FPU) { clear_c0_status(ST0_CU1 | ST0_FR); disable_fpu_hazard(); } - vcpu->arch.fpu_inuse &= ~(KVM_MIPS_FPU_FPU | KVM_MIPS_FPU_MSA); - } else if (vcpu->arch.fpu_inuse & KVM_MIPS_FPU_FPU) { + vcpu->arch.aux_inuse &= ~(KVM_MIPS_AUX_FPU | KVM_MIPS_AUX_MSA); + } else if (vcpu->arch.aux_inuse & KVM_MIPS_AUX_FPU) { set_c0_status(ST0_CU1); enable_fpu_hazard(); __kvm_save_fpu(&vcpu->arch); - vcpu->arch.fpu_inuse &= ~KVM_MIPS_FPU_FPU; + vcpu->arch.aux_inuse &= ~KVM_MIPS_AUX_FPU; + trace_kvm_aux(vcpu, KVM_TRACE_AUX_SAVE, KVM_TRACE_AUX_FPU); /* Disable FPU */ clear_c0_status(ST0_CU1 | ST0_FR); @@ -1638,6 +1795,10 @@ static int __init kvm_mips_init(void) { int ret; + ret = kvm_mips_entry_setup(); + if (ret) + return ret; + ret = kvm_init(NULL, sizeof(struct kvm_vcpu), 0, THIS_MODULE); if (ret) @@ -1645,18 +1806,6 @@ static int __init kvm_mips_init(void) register_die_notifier(&kvm_mips_csr_die_notifier); - /* - * On MIPS, kernel modules are executed from "mapped space", which - * requires TLBs. The TLB handling code is statically linked with - * the rest of the kernel (tlb.c) to avoid the possibility of - * double faulting. The issue is that the TLB code references - * routines that are part of the the KVM module, which are only - * available once the module is loaded. - */ - kvm_mips_gfn_to_pfn = gfn_to_pfn; - kvm_mips_release_pfn_clean = kvm_release_pfn_clean; - kvm_mips_is_error_pfn = is_error_pfn; - return 0; } @@ -1664,10 +1813,6 @@ static void __exit kvm_mips_exit(void) { kvm_exit(); - kvm_mips_gfn_to_pfn = NULL; - kvm_mips_release_pfn_clean = NULL; - kvm_mips_is_error_pfn = NULL; - unregister_die_notifier(&kvm_mips_csr_die_notifier); } diff --git a/arch/mips/kvm/mmu.c b/arch/mips/kvm/mmu.c new file mode 100644 index 0000000..57319ee --- /dev/null +++ b/arch/mips/kvm/mmu.c @@ -0,0 +1,375 @@ +/* + * This file is subject to the terms and conditions of the GNU General Public + * License. See the file "COPYING" in the main directory of this archive + * for more details. + * + * KVM/MIPS MMU handling in the KVM module. + * + * Copyright (C) 2012 MIPS Technologies, Inc. All rights reserved. + * Authors: Sanjay Lal <sanjayl@kymasys.com> + */ + +#include <linux/highmem.h> +#include <linux/kvm_host.h> +#include <asm/mmu_context.h> + +static u32 kvm_mips_get_kernel_asid(struct kvm_vcpu *vcpu) +{ + int cpu = smp_processor_id(); + + return vcpu->arch.guest_kernel_asid[cpu] & + cpu_asid_mask(&cpu_data[cpu]); +} + +static u32 kvm_mips_get_user_asid(struct kvm_vcpu *vcpu) +{ + int cpu = smp_processor_id(); + + return vcpu->arch.guest_user_asid[cpu] & + cpu_asid_mask(&cpu_data[cpu]); +} + +static int kvm_mips_map_page(struct kvm *kvm, gfn_t gfn) +{ + int srcu_idx, err = 0; + kvm_pfn_t pfn; + + if (kvm->arch.guest_pmap[gfn] != KVM_INVALID_PAGE) + return 0; + + srcu_idx = srcu_read_lock(&kvm->srcu); + pfn = gfn_to_pfn(kvm, gfn); + + if (is_error_pfn(pfn)) { + kvm_err("Couldn't get pfn for gfn %#llx!\n", gfn); + err = -EFAULT; + goto out; + } + + kvm->arch.guest_pmap[gfn] = pfn; +out: + srcu_read_unlock(&kvm->srcu, srcu_idx); + return err; +} + +/* Translate guest KSEG0 addresses to Host PA */ +unsigned long kvm_mips_translate_guest_kseg0_to_hpa(struct kvm_vcpu *vcpu, + unsigned long gva) +{ + gfn_t gfn; + unsigned long offset = gva & ~PAGE_MASK; + struct kvm *kvm = vcpu->kvm; + + if (KVM_GUEST_KSEGX(gva) != KVM_GUEST_KSEG0) { + kvm_err("%s/%p: Invalid gva: %#lx\n", __func__, + __builtin_return_address(0), gva); + return KVM_INVALID_PAGE; + } + + gfn = (KVM_GUEST_CPHYSADDR(gva) >> PAGE_SHIFT); + + if (gfn >= kvm->arch.guest_pmap_npages) { + kvm_err("%s: Invalid gfn: %#llx, GVA: %#lx\n", __func__, gfn, + gva); + return KVM_INVALID_PAGE; + } + + if (kvm_mips_map_page(vcpu->kvm, gfn) < 0) + return KVM_INVALID_ADDR; + + return (kvm->arch.guest_pmap[gfn] << PAGE_SHIFT) + offset; +} + +/* XXXKYMA: Must be called with interrupts disabled */ +int kvm_mips_handle_kseg0_tlb_fault(unsigned long badvaddr, + struct kvm_vcpu *vcpu) +{ + gfn_t gfn; + kvm_pfn_t pfn0, pfn1; + unsigned long vaddr = 0; + unsigned long entryhi = 0, entrylo0 = 0, entrylo1 = 0; + struct kvm *kvm = vcpu->kvm; + const int flush_dcache_mask = 0; + int ret; + + if (KVM_GUEST_KSEGX(badvaddr) != KVM_GUEST_KSEG0) { + kvm_err("%s: Invalid BadVaddr: %#lx\n", __func__, badvaddr); + kvm_mips_dump_host_tlbs(); + return -1; + } + + gfn = (KVM_GUEST_CPHYSADDR(badvaddr) >> PAGE_SHIFT); + if (gfn >= kvm->arch.guest_pmap_npages) { + kvm_err("%s: Invalid gfn: %#llx, BadVaddr: %#lx\n", __func__, + gfn, badvaddr); + kvm_mips_dump_host_tlbs(); + return -1; + } + vaddr = badvaddr & (PAGE_MASK << 1); + + if (kvm_mips_map_page(vcpu->kvm, gfn) < 0) + return -1; + + if (kvm_mips_map_page(vcpu->kvm, gfn ^ 0x1) < 0) + return -1; + + pfn0 = kvm->arch.guest_pmap[gfn & ~0x1]; + pfn1 = kvm->arch.guest_pmap[gfn | 0x1]; + + entrylo0 = mips3_paddr_to_tlbpfn(pfn0 << PAGE_SHIFT) | + ((_page_cachable_default >> _CACHE_SHIFT) << ENTRYLO_C_SHIFT) | + ENTRYLO_D | ENTRYLO_V; + entrylo1 = mips3_paddr_to_tlbpfn(pfn1 << PAGE_SHIFT) | + ((_page_cachable_default >> _CACHE_SHIFT) << ENTRYLO_C_SHIFT) | + ENTRYLO_D | ENTRYLO_V; + + preempt_disable(); + entryhi = (vaddr | kvm_mips_get_kernel_asid(vcpu)); + ret = kvm_mips_host_tlb_write(vcpu, entryhi, entrylo0, entrylo1, + flush_dcache_mask); + preempt_enable(); + + return ret; +} + +int kvm_mips_handle_mapped_seg_tlb_fault(struct kvm_vcpu *vcpu, + struct kvm_mips_tlb *tlb) +{ + unsigned long entryhi = 0, entrylo0 = 0, entrylo1 = 0; + struct kvm *kvm = vcpu->kvm; + kvm_pfn_t pfn0, pfn1; + int ret; + + if ((tlb->tlb_hi & VPN2_MASK) == 0) { + pfn0 = 0; + pfn1 = 0; + } else { + if (kvm_mips_map_page(kvm, mips3_tlbpfn_to_paddr(tlb->tlb_lo[0]) + >> PAGE_SHIFT) < 0) + return -1; + + if (kvm_mips_map_page(kvm, mips3_tlbpfn_to_paddr(tlb->tlb_lo[1]) + >> PAGE_SHIFT) < 0) + return -1; + + pfn0 = kvm->arch.guest_pmap[ + mips3_tlbpfn_to_paddr(tlb->tlb_lo[0]) >> PAGE_SHIFT]; + pfn1 = kvm->arch.guest_pmap[ + mips3_tlbpfn_to_paddr(tlb->tlb_lo[1]) >> PAGE_SHIFT]; + } + + /* Get attributes from the Guest TLB */ + entrylo0 = mips3_paddr_to_tlbpfn(pfn0 << PAGE_SHIFT) | + ((_page_cachable_default >> _CACHE_SHIFT) << ENTRYLO_C_SHIFT) | + (tlb->tlb_lo[0] & ENTRYLO_D) | + (tlb->tlb_lo[0] & ENTRYLO_V); + entrylo1 = mips3_paddr_to_tlbpfn(pfn1 << PAGE_SHIFT) | + ((_page_cachable_default >> _CACHE_SHIFT) << ENTRYLO_C_SHIFT) | + (tlb->tlb_lo[1] & ENTRYLO_D) | + (tlb->tlb_lo[1] & ENTRYLO_V); + + kvm_debug("@ %#lx tlb_lo0: 0x%08lx tlb_lo1: 0x%08lx\n", vcpu->arch.pc, + tlb->tlb_lo[0], tlb->tlb_lo[1]); + + preempt_disable(); + entryhi = (tlb->tlb_hi & VPN2_MASK) | (KVM_GUEST_KERNEL_MODE(vcpu) ? + kvm_mips_get_kernel_asid(vcpu) : + kvm_mips_get_user_asid(vcpu)); + ret = kvm_mips_host_tlb_write(vcpu, entryhi, entrylo0, entrylo1, + tlb->tlb_mask); + preempt_enable(); + + return ret; +} + +void kvm_get_new_mmu_context(struct mm_struct *mm, unsigned long cpu, + struct kvm_vcpu *vcpu) +{ + unsigned long asid = asid_cache(cpu); + + asid += cpu_asid_inc(); + if (!(asid & cpu_asid_mask(&cpu_data[cpu]))) { + if (cpu_has_vtag_icache) + flush_icache_all(); + + kvm_local_flush_tlb_all(); /* start new asid cycle */ + + if (!asid) /* fix version if needed */ + asid = asid_first_version(cpu); + } + + cpu_context(cpu, mm) = asid_cache(cpu) = asid; +} + +/** + * kvm_mips_migrate_count() - Migrate timer. + * @vcpu: Virtual CPU. + * + * Migrate CP0_Count hrtimer to the current CPU by cancelling and restarting it + * if it was running prior to being cancelled. + * + * Must be called when the VCPU is migrated to a different CPU to ensure that + * timer expiry during guest execution interrupts the guest and causes the + * interrupt to be delivered in a timely manner. + */ +static void kvm_mips_migrate_count(struct kvm_vcpu *vcpu) +{ + if (hrtimer_cancel(&vcpu->arch.comparecount_timer)) + hrtimer_restart(&vcpu->arch.comparecount_timer); +} + +/* Restore ASID once we are scheduled back after preemption */ +void kvm_arch_vcpu_load(struct kvm_vcpu *vcpu, int cpu) +{ + unsigned long asid_mask = cpu_asid_mask(&cpu_data[cpu]); + unsigned long flags; + int newasid = 0; + + kvm_debug("%s: vcpu %p, cpu: %d\n", __func__, vcpu, cpu); + + /* Allocate new kernel and user ASIDs if needed */ + + local_irq_save(flags); + + if ((vcpu->arch.guest_kernel_asid[cpu] ^ asid_cache(cpu)) & + asid_version_mask(cpu)) { + kvm_get_new_mmu_context(&vcpu->arch.guest_kernel_mm, cpu, vcpu); + vcpu->arch.guest_kernel_asid[cpu] = + vcpu->arch.guest_kernel_mm.context.asid[cpu]; + kvm_get_new_mmu_context(&vcpu->arch.guest_user_mm, cpu, vcpu); + vcpu->arch.guest_user_asid[cpu] = + vcpu->arch.guest_user_mm.context.asid[cpu]; + newasid++; + + kvm_debug("[%d]: cpu_context: %#lx\n", cpu, + cpu_context(cpu, current->mm)); + kvm_debug("[%d]: Allocated new ASID for Guest Kernel: %#x\n", + cpu, vcpu->arch.guest_kernel_asid[cpu]); + kvm_debug("[%d]: Allocated new ASID for Guest User: %#x\n", cpu, + vcpu->arch.guest_user_asid[cpu]); + } + + if (vcpu->arch.last_sched_cpu != cpu) { + kvm_debug("[%d->%d]KVM VCPU[%d] switch\n", + vcpu->arch.last_sched_cpu, cpu, vcpu->vcpu_id); + /* + * Migrate the timer interrupt to the current CPU so that it + * always interrupts the guest and synchronously triggers a + * guest timer interrupt. + */ + kvm_mips_migrate_count(vcpu); + } + + if (!newasid) { + /* + * If we preempted while the guest was executing, then reload + * the pre-empted ASID + */ + if (current->flags & PF_VCPU) { + write_c0_entryhi(vcpu->arch. + preempt_entryhi & asid_mask); + ehb(); + } + } else { + /* New ASIDs were allocated for the VM */ + + /* + * Were we in guest context? If so then the pre-empted ASID is + * no longer valid, we need to set it to what it should be based + * on the mode of the Guest (Kernel/User) + */ + if (current->flags & PF_VCPU) { + if (KVM_GUEST_KERNEL_MODE(vcpu)) + write_c0_entryhi(vcpu->arch. + guest_kernel_asid[cpu] & + asid_mask); + else + write_c0_entryhi(vcpu->arch. + guest_user_asid[cpu] & + asid_mask); + ehb(); + } + } + + /* restore guest state to registers */ + kvm_mips_callbacks->vcpu_set_regs(vcpu); + + local_irq_restore(flags); + +} + +/* ASID can change if another task is scheduled during preemption */ +void kvm_arch_vcpu_put(struct kvm_vcpu *vcpu) +{ + unsigned long flags; + int cpu; + + local_irq_save(flags); + + cpu = smp_processor_id(); + + vcpu->arch.preempt_entryhi = read_c0_entryhi(); + vcpu->arch.last_sched_cpu = cpu; + + /* save guest state in registers */ + kvm_mips_callbacks->vcpu_get_regs(vcpu); + + if (((cpu_context(cpu, current->mm) ^ asid_cache(cpu)) & + asid_version_mask(cpu))) { + kvm_debug("%s: Dropping MMU Context: %#lx\n", __func__, + cpu_context(cpu, current->mm)); + drop_mmu_context(current->mm, cpu); + } + write_c0_entryhi(cpu_asid(cpu, current->mm)); + ehb(); + + local_irq_restore(flags); +} + +u32 kvm_get_inst(u32 *opc, struct kvm_vcpu *vcpu) +{ + struct mips_coproc *cop0 = vcpu->arch.cop0; + unsigned long paddr, flags, vpn2, asid; + unsigned long va = (unsigned long)opc; + void *vaddr; + u32 inst; + int index; + + if (KVM_GUEST_KSEGX(va) < KVM_GUEST_KSEG0 || + KVM_GUEST_KSEGX(va) == KVM_GUEST_KSEG23) { + local_irq_save(flags); + index = kvm_mips_host_tlb_lookup(vcpu, va); + if (index >= 0) { + inst = *(opc); + } else { + vpn2 = va & VPN2_MASK; + asid = kvm_read_c0_guest_entryhi(cop0) & + KVM_ENTRYHI_ASID; + index = kvm_mips_guest_tlb_lookup(vcpu, vpn2 | asid); + if (index < 0) { + kvm_err("%s: get_user_failed for %p, vcpu: %p, ASID: %#lx\n", + __func__, opc, vcpu, read_c0_entryhi()); + kvm_mips_dump_host_tlbs(); + kvm_mips_dump_guest_tlbs(vcpu); + local_irq_restore(flags); + return KVM_INVALID_INST; + } + kvm_mips_handle_mapped_seg_tlb_fault(vcpu, + &vcpu->arch. + guest_tlb[index]); + inst = *(opc); + } + local_irq_restore(flags); + } else if (KVM_GUEST_KSEGX(va) == KVM_GUEST_KSEG0) { + paddr = kvm_mips_translate_guest_kseg0_to_hpa(vcpu, va); + vaddr = kmap_atomic(pfn_to_page(PHYS_PFN(paddr))); + vaddr += paddr & ~PAGE_MASK; + inst = *(u32 *)vaddr; + kunmap_atomic(vaddr); + } else { + kvm_err("%s: illegal address: %p\n", __func__, opc); + return KVM_INVALID_INST; + } + + return inst; +} diff --git a/arch/mips/kvm/stats.c b/arch/mips/kvm/stats.c index 888bb67..53f851a 100644 --- a/arch/mips/kvm/stats.c +++ b/arch/mips/kvm/stats.c @@ -11,27 +11,6 @@ #include <linux/kvm_host.h> -char *kvm_mips_exit_types_str[MAX_KVM_MIPS_EXIT_TYPES] = { - "WAIT", - "CACHE", - "Signal", - "Interrupt", - "COP0/1 Unusable", - "TLB Mod", - "TLB Miss (LD)", - "TLB Miss (ST)", - "Address Err (ST)", - "Address Error (LD)", - "System Call", - "Reserved Inst", - "Break Inst", - "Trap Inst", - "MSA FPE", - "FPE", - "MSA Disabled", - "D-Cache Flushes", -}; - char *kvm_cop0_str[N_MIPS_COPROC_REGS] = { "Index", "Random", diff --git a/arch/mips/kvm/tlb.c b/arch/mips/kvm/tlb.c index ed021ae..254377d 100644 --- a/arch/mips/kvm/tlb.c +++ b/arch/mips/kvm/tlb.c @@ -14,7 +14,7 @@ #include <linux/smp.h> #include <linux/mm.h> #include <linux/delay.h> -#include <linux/module.h> +#include <linux/export.h> #include <linux/kvm_host.h> #include <linux/srcu.h> @@ -24,6 +24,7 @@ #include <asm/pgtable.h> #include <asm/cacheflush.h> #include <asm/tlb.h> +#include <asm/tlbdebug.h> #undef CONFIG_MIPS_MT #include <asm/r4kcache.h> @@ -32,22 +33,10 @@ #define KVM_GUEST_PC_TLB 0 #define KVM_GUEST_SP_TLB 1 -#define PRIx64 "llx" - atomic_t kvm_mips_instance; EXPORT_SYMBOL_GPL(kvm_mips_instance); -/* These function pointers are initialized once the KVM module is loaded */ -kvm_pfn_t (*kvm_mips_gfn_to_pfn)(struct kvm *kvm, gfn_t gfn); -EXPORT_SYMBOL_GPL(kvm_mips_gfn_to_pfn); - -void (*kvm_mips_release_pfn_clean)(kvm_pfn_t pfn); -EXPORT_SYMBOL_GPL(kvm_mips_release_pfn_clean); - -bool (*kvm_mips_is_error_pfn)(kvm_pfn_t pfn); -EXPORT_SYMBOL_GPL(kvm_mips_is_error_pfn); - -uint32_t kvm_mips_get_kernel_asid(struct kvm_vcpu *vcpu) +static u32 kvm_mips_get_kernel_asid(struct kvm_vcpu *vcpu) { int cpu = smp_processor_id(); @@ -55,7 +44,7 @@ uint32_t kvm_mips_get_kernel_asid(struct kvm_vcpu *vcpu) cpu_asid_mask(&cpu_data[cpu]); } -uint32_t kvm_mips_get_user_asid(struct kvm_vcpu *vcpu) +static u32 kvm_mips_get_user_asid(struct kvm_vcpu *vcpu) { int cpu = smp_processor_id(); @@ -63,7 +52,7 @@ uint32_t kvm_mips_get_user_asid(struct kvm_vcpu *vcpu) cpu_asid_mask(&cpu_data[cpu]); } -inline uint32_t kvm_mips_get_commpage_asid(struct kvm_vcpu *vcpu) +inline u32 kvm_mips_get_commpage_asid(struct kvm_vcpu *vcpu) { return vcpu->kvm->arch.commpage_tlb; } @@ -72,50 +61,15 @@ inline uint32_t kvm_mips_get_commpage_asid(struct kvm_vcpu *vcpu) void kvm_mips_dump_host_tlbs(void) { - unsigned long old_entryhi; - unsigned long old_pagemask; - struct kvm_mips_tlb tlb; unsigned long flags; - int i; local_irq_save(flags); - old_entryhi = read_c0_entryhi(); - old_pagemask = read_c0_pagemask(); - kvm_info("HOST TLBs:\n"); - kvm_info("ASID: %#lx\n", read_c0_entryhi() & - cpu_asid_mask(¤t_cpu_data)); - - for (i = 0; i < current_cpu_data.tlbsize; i++) { - write_c0_index(i); - mtc0_tlbw_hazard(); - - tlb_read(); - tlbw_use_hazard(); + dump_tlb_regs(); + pr_info("\n"); + dump_tlb_all(); - tlb.tlb_hi = read_c0_entryhi(); - tlb.tlb_lo0 = read_c0_entrylo0(); - tlb.tlb_lo1 = read_c0_entrylo1(); - tlb.tlb_mask = read_c0_pagemask(); - - kvm_info("TLB%c%3d Hi 0x%08lx ", - (tlb.tlb_lo0 | tlb.tlb_lo1) & MIPS3_PG_V ? ' ' : '*', - i, tlb.tlb_hi); - kvm_info("Lo0=0x%09" PRIx64 " %c%c attr %lx ", - (uint64_t) mips3_tlbpfn_to_paddr(tlb.tlb_lo0), - (tlb.tlb_lo0 & MIPS3_PG_D) ? 'D' : ' ', - (tlb.tlb_lo0 & MIPS3_PG_G) ? 'G' : ' ', - (tlb.tlb_lo0 >> 3) & 7); - kvm_info("Lo1=0x%09" PRIx64 " %c%c attr %lx sz=%lx\n", - (uint64_t) mips3_tlbpfn_to_paddr(tlb.tlb_lo1), - (tlb.tlb_lo1 & MIPS3_PG_D) ? 'D' : ' ', - (tlb.tlb_lo1 & MIPS3_PG_G) ? 'G' : ' ', - (tlb.tlb_lo1 >> 3) & 7, tlb.tlb_mask); - } - write_c0_entryhi(old_entryhi); - write_c0_pagemask(old_pagemask); - mtc0_tlbw_hazard(); local_irq_restore(flags); } EXPORT_SYMBOL_GPL(kvm_mips_dump_host_tlbs); @@ -132,74 +86,24 @@ void kvm_mips_dump_guest_tlbs(struct kvm_vcpu *vcpu) for (i = 0; i < KVM_MIPS_GUEST_TLB_SIZE; i++) { tlb = vcpu->arch.guest_tlb[i]; kvm_info("TLB%c%3d Hi 0x%08lx ", - (tlb.tlb_lo0 | tlb.tlb_lo1) & MIPS3_PG_V ? ' ' : '*', + (tlb.tlb_lo[0] | tlb.tlb_lo[1]) & ENTRYLO_V + ? ' ' : '*', i, tlb.tlb_hi); - kvm_info("Lo0=0x%09" PRIx64 " %c%c attr %lx ", - (uint64_t) mips3_tlbpfn_to_paddr(tlb.tlb_lo0), - (tlb.tlb_lo0 & MIPS3_PG_D) ? 'D' : ' ', - (tlb.tlb_lo0 & MIPS3_PG_G) ? 'G' : ' ', - (tlb.tlb_lo0 >> 3) & 7); - kvm_info("Lo1=0x%09" PRIx64 " %c%c attr %lx sz=%lx\n", - (uint64_t) mips3_tlbpfn_to_paddr(tlb.tlb_lo1), - (tlb.tlb_lo1 & MIPS3_PG_D) ? 'D' : ' ', - (tlb.tlb_lo1 & MIPS3_PG_G) ? 'G' : ' ', - (tlb.tlb_lo1 >> 3) & 7, tlb.tlb_mask); + kvm_info("Lo0=0x%09llx %c%c attr %lx ", + (u64) mips3_tlbpfn_to_paddr(tlb.tlb_lo[0]), + (tlb.tlb_lo[0] & ENTRYLO_D) ? 'D' : ' ', + (tlb.tlb_lo[0] & ENTRYLO_G) ? 'G' : ' ', + (tlb.tlb_lo[0] & ENTRYLO_C) >> ENTRYLO_C_SHIFT); + kvm_info("Lo1=0x%09llx %c%c attr %lx sz=%lx\n", + (u64) mips3_tlbpfn_to_paddr(tlb.tlb_lo[1]), + (tlb.tlb_lo[1] & ENTRYLO_D) ? 'D' : ' ', + (tlb.tlb_lo[1] & ENTRYLO_G) ? 'G' : ' ', + (tlb.tlb_lo[1] & ENTRYLO_C) >> ENTRYLO_C_SHIFT, + tlb.tlb_mask); } } EXPORT_SYMBOL_GPL(kvm_mips_dump_guest_tlbs); -static int kvm_mips_map_page(struct kvm *kvm, gfn_t gfn) -{ - int srcu_idx, err = 0; - kvm_pfn_t pfn; - - if (kvm->arch.guest_pmap[gfn] != KVM_INVALID_PAGE) - return 0; - - srcu_idx = srcu_read_lock(&kvm->srcu); - pfn = kvm_mips_gfn_to_pfn(kvm, gfn); - - if (kvm_mips_is_error_pfn(pfn)) { - kvm_err("Couldn't get pfn for gfn %#" PRIx64 "!\n", gfn); - err = -EFAULT; - goto out; - } - - kvm->arch.guest_pmap[gfn] = pfn; -out: - srcu_read_unlock(&kvm->srcu, srcu_idx); - return err; -} - -/* Translate guest KSEG0 addresses to Host PA */ -unsigned long kvm_mips_translate_guest_kseg0_to_hpa(struct kvm_vcpu *vcpu, - unsigned long gva) -{ - gfn_t gfn; - uint32_t offset = gva & ~PAGE_MASK; - struct kvm *kvm = vcpu->kvm; - - if (KVM_GUEST_KSEGX(gva) != KVM_GUEST_KSEG0) { - kvm_err("%s/%p: Invalid gva: %#lx\n", __func__, - __builtin_return_address(0), gva); - return KVM_INVALID_PAGE; - } - - gfn = (KVM_GUEST_CPHYSADDR(gva) >> PAGE_SHIFT); - - if (gfn >= kvm->arch.guest_pmap_npages) { - kvm_err("%s: Invalid gfn: %#llx, GVA: %#lx\n", __func__, gfn, - gva); - return KVM_INVALID_PAGE; - } - - if (kvm_mips_map_page(vcpu->kvm, gfn) < 0) - return KVM_INVALID_ADDR; - - return (kvm->arch.guest_pmap[gfn] << PAGE_SHIFT) + offset; -} -EXPORT_SYMBOL_GPL(kvm_mips_translate_guest_kseg0_to_hpa); - /* XXXKYMA: Must be called with interrupts disabled */ /* set flush_dcache_mask == 0 if no dcache flush required */ int kvm_mips_host_tlb_write(struct kvm_vcpu *vcpu, unsigned long entryhi, @@ -243,12 +147,12 @@ int kvm_mips_host_tlb_write(struct kvm_vcpu *vcpu, unsigned long entryhi, /* Flush D-cache */ if (flush_dcache_mask) { - if (entrylo0 & MIPS3_PG_V) { + if (entrylo0 & ENTRYLO_V) { ++vcpu->stat.flush_dcache_exits; flush_data_cache_page((entryhi & VPN2_MASK) & ~flush_dcache_mask); } - if (entrylo1 & MIPS3_PG_V) { + if (entrylo1 & ENTRYLO_V) { ++vcpu->stat.flush_dcache_exits; flush_data_cache_page(((entryhi & VPN2_MASK) & ~flush_dcache_mask) | @@ -259,96 +163,35 @@ int kvm_mips_host_tlb_write(struct kvm_vcpu *vcpu, unsigned long entryhi, /* Restore old ASID */ write_c0_entryhi(old_entryhi); mtc0_tlbw_hazard(); - tlbw_use_hazard(); local_irq_restore(flags); return 0; } - -/* XXXKYMA: Must be called with interrupts disabled */ -int kvm_mips_handle_kseg0_tlb_fault(unsigned long badvaddr, - struct kvm_vcpu *vcpu) -{ - gfn_t gfn; - kvm_pfn_t pfn0, pfn1; - unsigned long vaddr = 0; - unsigned long entryhi = 0, entrylo0 = 0, entrylo1 = 0; - int even; - struct kvm *kvm = vcpu->kvm; - const int flush_dcache_mask = 0; - int ret; - - if (KVM_GUEST_KSEGX(badvaddr) != KVM_GUEST_KSEG0) { - kvm_err("%s: Invalid BadVaddr: %#lx\n", __func__, badvaddr); - kvm_mips_dump_host_tlbs(); - return -1; - } - - gfn = (KVM_GUEST_CPHYSADDR(badvaddr) >> PAGE_SHIFT); - if (gfn >= kvm->arch.guest_pmap_npages) { - kvm_err("%s: Invalid gfn: %#llx, BadVaddr: %#lx\n", __func__, - gfn, badvaddr); - kvm_mips_dump_host_tlbs(); - return -1; - } - even = !(gfn & 0x1); - vaddr = badvaddr & (PAGE_MASK << 1); - - if (kvm_mips_map_page(vcpu->kvm, gfn) < 0) - return -1; - - if (kvm_mips_map_page(vcpu->kvm, gfn ^ 0x1) < 0) - return -1; - - if (even) { - pfn0 = kvm->arch.guest_pmap[gfn]; - pfn1 = kvm->arch.guest_pmap[gfn ^ 0x1]; - } else { - pfn0 = kvm->arch.guest_pmap[gfn ^ 0x1]; - pfn1 = kvm->arch.guest_pmap[gfn]; - } - - entrylo0 = mips3_paddr_to_tlbpfn(pfn0 << PAGE_SHIFT) | (0x3 << 3) | - (1 << 2) | (0x1 << 1); - entrylo1 = mips3_paddr_to_tlbpfn(pfn1 << PAGE_SHIFT) | (0x3 << 3) | - (1 << 2) | (0x1 << 1); - - preempt_disable(); - entryhi = (vaddr | kvm_mips_get_kernel_asid(vcpu)); - ret = kvm_mips_host_tlb_write(vcpu, entryhi, entrylo0, entrylo1, - flush_dcache_mask); - preempt_enable(); - - return ret; -} -EXPORT_SYMBOL_GPL(kvm_mips_handle_kseg0_tlb_fault); +EXPORT_SYMBOL_GPL(kvm_mips_host_tlb_write); int kvm_mips_handle_commpage_tlb_fault(unsigned long badvaddr, struct kvm_vcpu *vcpu) { - kvm_pfn_t pfn0, pfn1; + kvm_pfn_t pfn; unsigned long flags, old_entryhi = 0, vaddr = 0; - unsigned long entrylo0 = 0, entrylo1 = 0; + unsigned long entrylo[2] = { 0, 0 }; + unsigned int pair_idx; - pfn0 = CPHYSADDR(vcpu->arch.kseg0_commpage) >> PAGE_SHIFT; - pfn1 = 0; - entrylo0 = mips3_paddr_to_tlbpfn(pfn0 << PAGE_SHIFT) | (0x3 << 3) | - (1 << 2) | (0x1 << 1); - entrylo1 = 0; + pfn = PFN_DOWN(virt_to_phys(vcpu->arch.kseg0_commpage)); + pair_idx = (badvaddr >> PAGE_SHIFT) & 1; + entrylo[pair_idx] = mips3_paddr_to_tlbpfn(pfn << PAGE_SHIFT) | + ((_page_cachable_default >> _CACHE_SHIFT) << ENTRYLO_C_SHIFT) | + ENTRYLO_D | ENTRYLO_V; local_irq_save(flags); old_entryhi = read_c0_entryhi(); vaddr = badvaddr & (PAGE_MASK << 1); write_c0_entryhi(vaddr | kvm_mips_get_kernel_asid(vcpu)); - mtc0_tlbw_hazard(); - write_c0_entrylo0(entrylo0); - mtc0_tlbw_hazard(); - write_c0_entrylo1(entrylo1); - mtc0_tlbw_hazard(); + write_c0_entrylo0(entrylo[0]); + write_c0_entrylo1(entrylo[1]); write_c0_index(kvm_mips_get_commpage_asid(vcpu)); mtc0_tlbw_hazard(); tlb_write_indexed(); - mtc0_tlbw_hazard(); tlbw_use_hazard(); kvm_debug("@ %#lx idx: %2d [entryhi(R): %#lx] entrylo0 (R): 0x%08lx, entrylo1(R): 0x%08lx\n", @@ -358,68 +201,12 @@ int kvm_mips_handle_commpage_tlb_fault(unsigned long badvaddr, /* Restore old ASID */ write_c0_entryhi(old_entryhi); mtc0_tlbw_hazard(); - tlbw_use_hazard(); local_irq_restore(flags); return 0; } EXPORT_SYMBOL_GPL(kvm_mips_handle_commpage_tlb_fault); -int kvm_mips_handle_mapped_seg_tlb_fault(struct kvm_vcpu *vcpu, - struct kvm_mips_tlb *tlb, - unsigned long *hpa0, - unsigned long *hpa1) -{ - unsigned long entryhi = 0, entrylo0 = 0, entrylo1 = 0; - struct kvm *kvm = vcpu->kvm; - kvm_pfn_t pfn0, pfn1; - int ret; - - if ((tlb->tlb_hi & VPN2_MASK) == 0) { - pfn0 = 0; - pfn1 = 0; - } else { - if (kvm_mips_map_page(kvm, mips3_tlbpfn_to_paddr(tlb->tlb_lo0) - >> PAGE_SHIFT) < 0) - return -1; - - if (kvm_mips_map_page(kvm, mips3_tlbpfn_to_paddr(tlb->tlb_lo1) - >> PAGE_SHIFT) < 0) - return -1; - - pfn0 = kvm->arch.guest_pmap[mips3_tlbpfn_to_paddr(tlb->tlb_lo0) - >> PAGE_SHIFT]; - pfn1 = kvm->arch.guest_pmap[mips3_tlbpfn_to_paddr(tlb->tlb_lo1) - >> PAGE_SHIFT]; - } - - if (hpa0) - *hpa0 = pfn0 << PAGE_SHIFT; - - if (hpa1) - *hpa1 = pfn1 << PAGE_SHIFT; - - /* Get attributes from the Guest TLB */ - entrylo0 = mips3_paddr_to_tlbpfn(pfn0 << PAGE_SHIFT) | (0x3 << 3) | - (tlb->tlb_lo0 & MIPS3_PG_D) | (tlb->tlb_lo0 & MIPS3_PG_V); - entrylo1 = mips3_paddr_to_tlbpfn(pfn1 << PAGE_SHIFT) | (0x3 << 3) | - (tlb->tlb_lo1 & MIPS3_PG_D) | (tlb->tlb_lo1 & MIPS3_PG_V); - - kvm_debug("@ %#lx tlb_lo0: 0x%08lx tlb_lo1: 0x%08lx\n", vcpu->arch.pc, - tlb->tlb_lo0, tlb->tlb_lo1); - - preempt_disable(); - entryhi = (tlb->tlb_hi & VPN2_MASK) | (KVM_GUEST_KERNEL_MODE(vcpu) ? - kvm_mips_get_kernel_asid(vcpu) : - kvm_mips_get_user_asid(vcpu)); - ret = kvm_mips_host_tlb_write(vcpu, entryhi, entrylo0, entrylo1, - tlb->tlb_mask); - preempt_enable(); - - return ret; -} -EXPORT_SYMBOL_GPL(kvm_mips_handle_mapped_seg_tlb_fault); - int kvm_mips_guest_tlb_lookup(struct kvm_vcpu *vcpu, unsigned long entryhi) { int i; @@ -435,7 +222,7 @@ int kvm_mips_guest_tlb_lookup(struct kvm_vcpu *vcpu, unsigned long entryhi) } kvm_debug("%s: entryhi: %#lx, index: %d lo0: %#lx, lo1: %#lx\n", - __func__, entryhi, index, tlb[i].tlb_lo0, tlb[i].tlb_lo1); + __func__, entryhi, index, tlb[i].tlb_lo[0], tlb[i].tlb_lo[1]); return index; } @@ -467,7 +254,6 @@ int kvm_mips_host_tlb_lookup(struct kvm_vcpu *vcpu, unsigned long vaddr) /* Restore old ASID */ write_c0_entryhi(old_entryhi); mtc0_tlbw_hazard(); - tlbw_use_hazard(); local_irq_restore(flags); @@ -498,21 +284,16 @@ int kvm_mips_host_tlb_inv(struct kvm_vcpu *vcpu, unsigned long va) if (idx > 0) { write_c0_entryhi(UNIQUE_ENTRYHI(idx)); - mtc0_tlbw_hazard(); - write_c0_entrylo0(0); - mtc0_tlbw_hazard(); - write_c0_entrylo1(0); mtc0_tlbw_hazard(); tlb_write_indexed(); - mtc0_tlbw_hazard(); + tlbw_use_hazard(); } write_c0_entryhi(old_entryhi); mtc0_tlbw_hazard(); - tlbw_use_hazard(); local_irq_restore(flags); @@ -540,61 +321,39 @@ void kvm_mips_flush_host_tlb(int skip_kseg0) /* Blast 'em all away. */ for (entry = 0; entry < maxentry; entry++) { write_c0_index(entry); - mtc0_tlbw_hazard(); if (skip_kseg0) { + mtc0_tlbr_hazard(); tlb_read(); - tlbw_use_hazard(); + tlb_read_hazard(); entryhi = read_c0_entryhi(); /* Don't blow away guest kernel entries */ if (KVM_GUEST_KSEGX(entryhi) == KVM_GUEST_KSEG0) continue; + + write_c0_pagemask(old_pagemask); } /* Make sure all entries differ. */ write_c0_entryhi(UNIQUE_ENTRYHI(entry)); - mtc0_tlbw_hazard(); write_c0_entrylo0(0); - mtc0_tlbw_hazard(); write_c0_entrylo1(0); mtc0_tlbw_hazard(); tlb_write_indexed(); - mtc0_tlbw_hazard(); + tlbw_use_hazard(); } - tlbw_use_hazard(); - write_c0_entryhi(old_entryhi); write_c0_pagemask(old_pagemask); mtc0_tlbw_hazard(); - tlbw_use_hazard(); local_irq_restore(flags); } EXPORT_SYMBOL_GPL(kvm_mips_flush_host_tlb); -void kvm_get_new_mmu_context(struct mm_struct *mm, unsigned long cpu, - struct kvm_vcpu *vcpu) -{ - unsigned long asid = asid_cache(cpu); - - asid += cpu_asid_inc(); - if (!(asid & cpu_asid_mask(&cpu_data[cpu]))) { - if (cpu_has_vtag_icache) - flush_icache_all(); - - kvm_local_flush_tlb_all(); /* start new asid cycle */ - - if (!asid) /* fix version if needed */ - asid = asid_first_version(cpu); - } - - cpu_context(cpu, mm) = asid_cache(cpu) = asid; -} - void kvm_local_flush_tlb_all(void) { unsigned long flags; @@ -614,185 +373,12 @@ void kvm_local_flush_tlb_all(void) write_c0_index(entry); mtc0_tlbw_hazard(); tlb_write_indexed(); + tlbw_use_hazard(); entry++; } - tlbw_use_hazard(); write_c0_entryhi(old_ctx); mtc0_tlbw_hazard(); local_irq_restore(flags); } EXPORT_SYMBOL_GPL(kvm_local_flush_tlb_all); - -/** - * kvm_mips_migrate_count() - Migrate timer. - * @vcpu: Virtual CPU. - * - * Migrate CP0_Count hrtimer to the current CPU by cancelling and restarting it - * if it was running prior to being cancelled. - * - * Must be called when the VCPU is migrated to a different CPU to ensure that - * timer expiry during guest execution interrupts the guest and causes the - * interrupt to be delivered in a timely manner. - */ -static void kvm_mips_migrate_count(struct kvm_vcpu *vcpu) -{ - if (hrtimer_cancel(&vcpu->arch.comparecount_timer)) - hrtimer_restart(&vcpu->arch.comparecount_timer); -} - -/* Restore ASID once we are scheduled back after preemption */ -void kvm_arch_vcpu_load(struct kvm_vcpu *vcpu, int cpu) -{ - unsigned long asid_mask = cpu_asid_mask(&cpu_data[cpu]); - unsigned long flags; - int newasid = 0; - - kvm_debug("%s: vcpu %p, cpu: %d\n", __func__, vcpu, cpu); - - /* Allocate new kernel and user ASIDs if needed */ - - local_irq_save(flags); - - if ((vcpu->arch.guest_kernel_asid[cpu] ^ asid_cache(cpu)) & - asid_version_mask(cpu)) { - kvm_get_new_mmu_context(&vcpu->arch.guest_kernel_mm, cpu, vcpu); - vcpu->arch.guest_kernel_asid[cpu] = - vcpu->arch.guest_kernel_mm.context.asid[cpu]; - kvm_get_new_mmu_context(&vcpu->arch.guest_user_mm, cpu, vcpu); - vcpu->arch.guest_user_asid[cpu] = - vcpu->arch.guest_user_mm.context.asid[cpu]; - newasid++; - - kvm_debug("[%d]: cpu_context: %#lx\n", cpu, - cpu_context(cpu, current->mm)); - kvm_debug("[%d]: Allocated new ASID for Guest Kernel: %#x\n", - cpu, vcpu->arch.guest_kernel_asid[cpu]); - kvm_debug("[%d]: Allocated new ASID for Guest User: %#x\n", cpu, - vcpu->arch.guest_user_asid[cpu]); - } - - if (vcpu->arch.last_sched_cpu != cpu) { - kvm_debug("[%d->%d]KVM VCPU[%d] switch\n", - vcpu->arch.last_sched_cpu, cpu, vcpu->vcpu_id); - /* - * Migrate the timer interrupt to the current CPU so that it - * always interrupts the guest and synchronously triggers a - * guest timer interrupt. - */ - kvm_mips_migrate_count(vcpu); - } - - if (!newasid) { - /* - * If we preempted while the guest was executing, then reload - * the pre-empted ASID - */ - if (current->flags & PF_VCPU) { - write_c0_entryhi(vcpu->arch. - preempt_entryhi & asid_mask); - ehb(); - } - } else { - /* New ASIDs were allocated for the VM */ - - /* - * Were we in guest context? If so then the pre-empted ASID is - * no longer valid, we need to set it to what it should be based - * on the mode of the Guest (Kernel/User) - */ - if (current->flags & PF_VCPU) { - if (KVM_GUEST_KERNEL_MODE(vcpu)) - write_c0_entryhi(vcpu->arch. - guest_kernel_asid[cpu] & - asid_mask); - else - write_c0_entryhi(vcpu->arch. - guest_user_asid[cpu] & - asid_mask); - ehb(); - } - } - - /* restore guest state to registers */ - kvm_mips_callbacks->vcpu_set_regs(vcpu); - - local_irq_restore(flags); - -} -EXPORT_SYMBOL_GPL(kvm_arch_vcpu_load); - -/* ASID can change if another task is scheduled during preemption */ -void kvm_arch_vcpu_put(struct kvm_vcpu *vcpu) -{ - unsigned long flags; - uint32_t cpu; - - local_irq_save(flags); - - cpu = smp_processor_id(); - - vcpu->arch.preempt_entryhi = read_c0_entryhi(); - vcpu->arch.last_sched_cpu = cpu; - - /* save guest state in registers */ - kvm_mips_callbacks->vcpu_get_regs(vcpu); - - if (((cpu_context(cpu, current->mm) ^ asid_cache(cpu)) & - asid_version_mask(cpu))) { - kvm_debug("%s: Dropping MMU Context: %#lx\n", __func__, - cpu_context(cpu, current->mm)); - drop_mmu_context(current->mm, cpu); - } - write_c0_entryhi(cpu_asid(cpu, current->mm)); - ehb(); - - local_irq_restore(flags); -} -EXPORT_SYMBOL_GPL(kvm_arch_vcpu_put); - -uint32_t kvm_get_inst(uint32_t *opc, struct kvm_vcpu *vcpu) -{ - struct mips_coproc *cop0 = vcpu->arch.cop0; - unsigned long paddr, flags, vpn2, asid; - uint32_t inst; - int index; - - if (KVM_GUEST_KSEGX((unsigned long) opc) < KVM_GUEST_KSEG0 || - KVM_GUEST_KSEGX((unsigned long) opc) == KVM_GUEST_KSEG23) { - local_irq_save(flags); - index = kvm_mips_host_tlb_lookup(vcpu, (unsigned long) opc); - if (index >= 0) { - inst = *(opc); - } else { - vpn2 = (unsigned long) opc & VPN2_MASK; - asid = kvm_read_c0_guest_entryhi(cop0) & - KVM_ENTRYHI_ASID; - index = kvm_mips_guest_tlb_lookup(vcpu, vpn2 | asid); - if (index < 0) { - kvm_err("%s: get_user_failed for %p, vcpu: %p, ASID: %#lx\n", - __func__, opc, vcpu, read_c0_entryhi()); - kvm_mips_dump_host_tlbs(); - local_irq_restore(flags); - return KVM_INVALID_INST; - } - kvm_mips_handle_mapped_seg_tlb_fault(vcpu, - &vcpu->arch. - guest_tlb[index], - NULL, NULL); - inst = *(opc); - } - local_irq_restore(flags); - } else if (KVM_GUEST_KSEGX(opc) == KVM_GUEST_KSEG0) { - paddr = - kvm_mips_translate_guest_kseg0_to_hpa(vcpu, - (unsigned long) opc); - inst = *(uint32_t *) CKSEG0ADDR(paddr); - } else { - kvm_err("%s: illegal address: %p\n", __func__, opc); - return KVM_INVALID_INST; - } - - return inst; -} -EXPORT_SYMBOL_GPL(kvm_get_inst); diff --git a/arch/mips/kvm/trace.h b/arch/mips/kvm/trace.h index bd6437f..c858cf1 100644 --- a/arch/mips/kvm/trace.h +++ b/arch/mips/kvm/trace.h @@ -17,8 +17,75 @@ #define TRACE_INCLUDE_PATH . #define TRACE_INCLUDE_FILE trace -/* Tracepoints for VM eists */ -extern char *kvm_mips_exit_types_str[MAX_KVM_MIPS_EXIT_TYPES]; +/* + * Tracepoints for VM enters + */ +DECLARE_EVENT_CLASS(kvm_transition, + TP_PROTO(struct kvm_vcpu *vcpu), + TP_ARGS(vcpu), + TP_STRUCT__entry( + __field(unsigned long, pc) + ), + + TP_fast_assign( + __entry->pc = vcpu->arch.pc; + ), + + TP_printk("PC: 0x%08lx", + __entry->pc) +); + +DEFINE_EVENT(kvm_transition, kvm_enter, + TP_PROTO(struct kvm_vcpu *vcpu), + TP_ARGS(vcpu)); + +DEFINE_EVENT(kvm_transition, kvm_reenter, + TP_PROTO(struct kvm_vcpu *vcpu), + TP_ARGS(vcpu)); + +DEFINE_EVENT(kvm_transition, kvm_out, + TP_PROTO(struct kvm_vcpu *vcpu), + TP_ARGS(vcpu)); + +/* The first 32 exit reasons correspond to Cause.ExcCode */ +#define KVM_TRACE_EXIT_INT 0 +#define KVM_TRACE_EXIT_TLBMOD 1 +#define KVM_TRACE_EXIT_TLBMISS_LD 2 +#define KVM_TRACE_EXIT_TLBMISS_ST 3 +#define KVM_TRACE_EXIT_ADDRERR_LD 4 +#define KVM_TRACE_EXIT_ADDRERR_ST 5 +#define KVM_TRACE_EXIT_SYSCALL 8 +#define KVM_TRACE_EXIT_BREAK_INST 9 +#define KVM_TRACE_EXIT_RESVD_INST 10 +#define KVM_TRACE_EXIT_COP_UNUSABLE 11 +#define KVM_TRACE_EXIT_TRAP_INST 13 +#define KVM_TRACE_EXIT_MSA_FPE 14 +#define KVM_TRACE_EXIT_FPE 15 +#define KVM_TRACE_EXIT_MSA_DISABLED 21 +/* Further exit reasons */ +#define KVM_TRACE_EXIT_WAIT 32 +#define KVM_TRACE_EXIT_CACHE 33 +#define KVM_TRACE_EXIT_SIGNAL 34 + +/* Tracepoints for VM exits */ +#define kvm_trace_symbol_exit_types \ + { KVM_TRACE_EXIT_INT, "Interrupt" }, \ + { KVM_TRACE_EXIT_TLBMOD, "TLB Mod" }, \ + { KVM_TRACE_EXIT_TLBMISS_LD, "TLB Miss (LD)" }, \ + { KVM_TRACE_EXIT_TLBMISS_ST, "TLB Miss (ST)" }, \ + { KVM_TRACE_EXIT_ADDRERR_LD, "Address Error (LD)" }, \ + { KVM_TRACE_EXIT_ADDRERR_ST, "Address Err (ST)" }, \ + { KVM_TRACE_EXIT_SYSCALL, "System Call" }, \ + { KVM_TRACE_EXIT_BREAK_INST, "Break Inst" }, \ + { KVM_TRACE_EXIT_RESVD_INST, "Reserved Inst" }, \ + { KVM_TRACE_EXIT_COP_UNUSABLE, "COP0/1 Unusable" }, \ + { KVM_TRACE_EXIT_TRAP_INST, "Trap Inst" }, \ + { KVM_TRACE_EXIT_MSA_FPE, "MSA FPE" }, \ + { KVM_TRACE_EXIT_FPE, "FPE" }, \ + { KVM_TRACE_EXIT_MSA_DISABLED, "MSA Disabled" }, \ + { KVM_TRACE_EXIT_WAIT, "WAIT" }, \ + { KVM_TRACE_EXIT_CACHE, "CACHE" }, \ + { KVM_TRACE_EXIT_SIGNAL, "Signal" } TRACE_EVENT(kvm_exit, TP_PROTO(struct kvm_vcpu *vcpu, unsigned int reason), @@ -34,10 +101,173 @@ TRACE_EVENT(kvm_exit, ), TP_printk("[%s]PC: 0x%08lx", - kvm_mips_exit_types_str[__entry->reason], + __print_symbolic(__entry->reason, + kvm_trace_symbol_exit_types), __entry->pc) ); +#define KVM_TRACE_MFC0 0 +#define KVM_TRACE_MTC0 1 +#define KVM_TRACE_DMFC0 2 +#define KVM_TRACE_DMTC0 3 +#define KVM_TRACE_RDHWR 4 + +#define KVM_TRACE_HWR_COP0 0 +#define KVM_TRACE_HWR_HWR 1 + +#define KVM_TRACE_COP0(REG, SEL) ((KVM_TRACE_HWR_COP0 << 8) | \ + ((REG) << 3) | (SEL)) +#define KVM_TRACE_HWR(REG, SEL) ((KVM_TRACE_HWR_HWR << 8) | \ + ((REG) << 3) | (SEL)) + +#define kvm_trace_symbol_hwr_ops \ + { KVM_TRACE_MFC0, "MFC0" }, \ + { KVM_TRACE_MTC0, "MTC0" }, \ + { KVM_TRACE_DMFC0, "DMFC0" }, \ + { KVM_TRACE_DMTC0, "DMTC0" }, \ + { KVM_TRACE_RDHWR, "RDHWR" } + +#define kvm_trace_symbol_hwr_cop \ + { KVM_TRACE_HWR_COP0, "COP0" }, \ + { KVM_TRACE_HWR_HWR, "HWR" } + +#define kvm_trace_symbol_hwr_regs \ + { KVM_TRACE_COP0( 0, 0), "Index" }, \ + { KVM_TRACE_COP0( 2, 0), "EntryLo0" }, \ + { KVM_TRACE_COP0( 3, 0), "EntryLo1" }, \ + { KVM_TRACE_COP0( 4, 0), "Context" }, \ + { KVM_TRACE_COP0( 4, 2), "UserLocal" }, \ + { KVM_TRACE_COP0( 5, 0), "PageMask" }, \ + { KVM_TRACE_COP0( 6, 0), "Wired" }, \ + { KVM_TRACE_COP0( 7, 0), "HWREna" }, \ + { KVM_TRACE_COP0( 8, 0), "BadVAddr" }, \ + { KVM_TRACE_COP0( 9, 0), "Count" }, \ + { KVM_TRACE_COP0(10, 0), "EntryHi" }, \ + { KVM_TRACE_COP0(11, 0), "Compare" }, \ + { KVM_TRACE_COP0(12, 0), "Status" }, \ + { KVM_TRACE_COP0(12, 1), "IntCtl" }, \ + { KVM_TRACE_COP0(12, 2), "SRSCtl" }, \ + { KVM_TRACE_COP0(13, 0), "Cause" }, \ + { KVM_TRACE_COP0(14, 0), "EPC" }, \ + { KVM_TRACE_COP0(15, 0), "PRId" }, \ + { KVM_TRACE_COP0(15, 1), "EBase" }, \ + { KVM_TRACE_COP0(16, 0), "Config" }, \ + { KVM_TRACE_COP0(16, 1), "Config1" }, \ + { KVM_TRACE_COP0(16, 2), "Config2" }, \ + { KVM_TRACE_COP0(16, 3), "Config3" }, \ + { KVM_TRACE_COP0(16, 4), "Config4" }, \ + { KVM_TRACE_COP0(16, 5), "Config5" }, \ + { KVM_TRACE_COP0(16, 7), "Config7" }, \ + { KVM_TRACE_COP0(26, 0), "ECC" }, \ + { KVM_TRACE_COP0(30, 0), "ErrorEPC" }, \ + { KVM_TRACE_COP0(31, 2), "KScratch1" }, \ + { KVM_TRACE_COP0(31, 3), "KScratch2" }, \ + { KVM_TRACE_COP0(31, 4), "KScratch3" }, \ + { KVM_TRACE_COP0(31, 5), "KScratch4" }, \ + { KVM_TRACE_COP0(31, 6), "KScratch5" }, \ + { KVM_TRACE_COP0(31, 7), "KScratch6" }, \ + { KVM_TRACE_HWR( 0, 0), "CPUNum" }, \ + { KVM_TRACE_HWR( 1, 0), "SYNCI_Step" }, \ + { KVM_TRACE_HWR( 2, 0), "CC" }, \ + { KVM_TRACE_HWR( 3, 0), "CCRes" }, \ + { KVM_TRACE_HWR(29, 0), "ULR" } + +TRACE_EVENT(kvm_hwr, + TP_PROTO(struct kvm_vcpu *vcpu, unsigned int op, unsigned int reg, + unsigned long val), + TP_ARGS(vcpu, op, reg, val), + TP_STRUCT__entry( + __field(unsigned long, val) + __field(u16, reg) + __field(u8, op) + ), + + TP_fast_assign( + __entry->val = val; + __entry->reg = reg; + __entry->op = op; + ), + + TP_printk("%s %s (%s:%u:%u) 0x%08lx", + __print_symbolic(__entry->op, + kvm_trace_symbol_hwr_ops), + __print_symbolic(__entry->reg, + kvm_trace_symbol_hwr_regs), + __print_symbolic(__entry->reg >> 8, + kvm_trace_symbol_hwr_cop), + (__entry->reg >> 3) & 0x1f, + __entry->reg & 0x7, + __entry->val) +); + +#define KVM_TRACE_AUX_RESTORE 0 +#define KVM_TRACE_AUX_SAVE 1 +#define KVM_TRACE_AUX_ENABLE 2 +#define KVM_TRACE_AUX_DISABLE 3 +#define KVM_TRACE_AUX_DISCARD 4 + +#define KVM_TRACE_AUX_FPU 1 +#define KVM_TRACE_AUX_MSA 2 +#define KVM_TRACE_AUX_FPU_MSA 3 + +#define kvm_trace_symbol_aux_op \ + { KVM_TRACE_AUX_RESTORE, "restore" }, \ + { KVM_TRACE_AUX_SAVE, "save" }, \ + { KVM_TRACE_AUX_ENABLE, "enable" }, \ + { KVM_TRACE_AUX_DISABLE, "disable" }, \ + { KVM_TRACE_AUX_DISCARD, "discard" } + +#define kvm_trace_symbol_aux_state \ + { KVM_TRACE_AUX_FPU, "FPU" }, \ + { KVM_TRACE_AUX_MSA, "MSA" }, \ + { KVM_TRACE_AUX_FPU_MSA, "FPU & MSA" } + +TRACE_EVENT(kvm_aux, + TP_PROTO(struct kvm_vcpu *vcpu, unsigned int op, + unsigned int state), + TP_ARGS(vcpu, op, state), + TP_STRUCT__entry( + __field(unsigned long, pc) + __field(u8, op) + __field(u8, state) + ), + + TP_fast_assign( + __entry->pc = vcpu->arch.pc; + __entry->op = op; + __entry->state = state; + ), + + TP_printk("%s %s PC: 0x%08lx", + __print_symbolic(__entry->op, + kvm_trace_symbol_aux_op), + __print_symbolic(__entry->state, + kvm_trace_symbol_aux_state), + __entry->pc) +); + +TRACE_EVENT(kvm_asid_change, + TP_PROTO(struct kvm_vcpu *vcpu, unsigned int old_asid, + unsigned int new_asid), + TP_ARGS(vcpu, old_asid, new_asid), + TP_STRUCT__entry( + __field(unsigned long, pc) + __field(u8, old_asid) + __field(u8, new_asid) + ), + + TP_fast_assign( + __entry->pc = vcpu->arch.pc; + __entry->old_asid = old_asid; + __entry->new_asid = new_asid; + ), + + TP_printk("PC: 0x%08lx old: 0x%02x new: 0x%02x", + __entry->pc, + __entry->old_asid, + __entry->new_asid) +); + #endif /* _TRACE_KVM_H */ /* This part must be outside protection */ diff --git a/arch/mips/kvm/trap_emul.c b/arch/mips/kvm/trap_emul.c index 6ba0faf..0915539 100644 --- a/arch/mips/kvm/trap_emul.c +++ b/arch/mips/kvm/trap_emul.c @@ -21,7 +21,7 @@ static gpa_t kvm_trap_emul_gva_to_gpa_cb(gva_t gva) { gpa_t gpa; - uint32_t kseg = KSEGX(gva); + gva_t kseg = KSEGX(gva); if ((kseg == CKSEG0) || (kseg == CKSEG1)) gpa = CPHYSADDR(gva); @@ -40,8 +40,8 @@ static int kvm_trap_emul_handle_cop_unusable(struct kvm_vcpu *vcpu) { struct mips_coproc *cop0 = vcpu->arch.cop0; struct kvm_run *run = vcpu->run; - uint32_t __user *opc = (uint32_t __user *) vcpu->arch.pc; - unsigned long cause = vcpu->arch.host_cp0_cause; + u32 __user *opc = (u32 __user *) vcpu->arch.pc; + u32 cause = vcpu->arch.host_cp0_cause; enum emulation_result er = EMULATE_DONE; int ret = RESUME_GUEST; @@ -87,15 +87,15 @@ static int kvm_trap_emul_handle_cop_unusable(struct kvm_vcpu *vcpu) static int kvm_trap_emul_handle_tlb_mod(struct kvm_vcpu *vcpu) { struct kvm_run *run = vcpu->run; - uint32_t __user *opc = (uint32_t __user *) vcpu->arch.pc; + u32 __user *opc = (u32 __user *) vcpu->arch.pc; unsigned long badvaddr = vcpu->arch.host_cp0_badvaddr; - unsigned long cause = vcpu->arch.host_cp0_cause; + u32 cause = vcpu->arch.host_cp0_cause; enum emulation_result er = EMULATE_DONE; int ret = RESUME_GUEST; if (KVM_GUEST_KSEGX(badvaddr) < KVM_GUEST_KSEG0 || KVM_GUEST_KSEGX(badvaddr) == KVM_GUEST_KSEG23) { - kvm_debug("USER/KSEG23 ADDR TLB MOD fault: cause %#lx, PC: %p, BadVaddr: %#lx\n", + kvm_debug("USER/KSEG23 ADDR TLB MOD fault: cause %#x, PC: %p, BadVaddr: %#lx\n", cause, opc, badvaddr); er = kvm_mips_handle_tlbmod(cause, opc, run, vcpu); @@ -111,14 +111,14 @@ static int kvm_trap_emul_handle_tlb_mod(struct kvm_vcpu *vcpu) * when we are not using HIGHMEM. Need to address this in a * HIGHMEM kernel */ - kvm_err("TLB MOD fault not handled, cause %#lx, PC: %p, BadVaddr: %#lx\n", + kvm_err("TLB MOD fault not handled, cause %#x, PC: %p, BadVaddr: %#lx\n", cause, opc, badvaddr); kvm_mips_dump_host_tlbs(); kvm_arch_vcpu_dump_regs(vcpu); run->exit_reason = KVM_EXIT_INTERNAL_ERROR; ret = RESUME_HOST; } else { - kvm_err("Illegal TLB Mod fault address , cause %#lx, PC: %p, BadVaddr: %#lx\n", + kvm_err("Illegal TLB Mod fault address , cause %#x, PC: %p, BadVaddr: %#lx\n", cause, opc, badvaddr); kvm_mips_dump_host_tlbs(); kvm_arch_vcpu_dump_regs(vcpu); @@ -128,59 +128,12 @@ static int kvm_trap_emul_handle_tlb_mod(struct kvm_vcpu *vcpu) return ret; } -static int kvm_trap_emul_handle_tlb_st_miss(struct kvm_vcpu *vcpu) -{ - struct kvm_run *run = vcpu->run; - uint32_t __user *opc = (uint32_t __user *) vcpu->arch.pc; - unsigned long badvaddr = vcpu->arch.host_cp0_badvaddr; - unsigned long cause = vcpu->arch.host_cp0_cause; - enum emulation_result er = EMULATE_DONE; - int ret = RESUME_GUEST; - - if (((badvaddr & PAGE_MASK) == KVM_GUEST_COMMPAGE_ADDR) - && KVM_GUEST_KERNEL_MODE(vcpu)) { - if (kvm_mips_handle_commpage_tlb_fault(badvaddr, vcpu) < 0) { - run->exit_reason = KVM_EXIT_INTERNAL_ERROR; - ret = RESUME_HOST; - } - } else if (KVM_GUEST_KSEGX(badvaddr) < KVM_GUEST_KSEG0 - || KVM_GUEST_KSEGX(badvaddr) == KVM_GUEST_KSEG23) { - kvm_debug("USER ADDR TLB LD fault: cause %#lx, PC: %p, BadVaddr: %#lx\n", - cause, opc, badvaddr); - er = kvm_mips_handle_tlbmiss(cause, opc, run, vcpu); - if (er == EMULATE_DONE) - ret = RESUME_GUEST; - else { - run->exit_reason = KVM_EXIT_INTERNAL_ERROR; - ret = RESUME_HOST; - } - } else if (KVM_GUEST_KSEGX(badvaddr) == KVM_GUEST_KSEG0) { - /* - * All KSEG0 faults are handled by KVM, as the guest kernel does - * not expect to ever get them - */ - if (kvm_mips_handle_kseg0_tlb_fault - (vcpu->arch.host_cp0_badvaddr, vcpu) < 0) { - run->exit_reason = KVM_EXIT_INTERNAL_ERROR; - ret = RESUME_HOST; - } - } else { - kvm_err("Illegal TLB LD fault address , cause %#lx, PC: %p, BadVaddr: %#lx\n", - cause, opc, badvaddr); - kvm_mips_dump_host_tlbs(); - kvm_arch_vcpu_dump_regs(vcpu); - run->exit_reason = KVM_EXIT_INTERNAL_ERROR; - ret = RESUME_HOST; - } - return ret; -} - -static int kvm_trap_emul_handle_tlb_ld_miss(struct kvm_vcpu *vcpu) +static int kvm_trap_emul_handle_tlb_miss(struct kvm_vcpu *vcpu, bool store) { struct kvm_run *run = vcpu->run; - uint32_t __user *opc = (uint32_t __user *) vcpu->arch.pc; + u32 __user *opc = (u32 __user *) vcpu->arch.pc; unsigned long badvaddr = vcpu->arch.host_cp0_badvaddr; - unsigned long cause = vcpu->arch.host_cp0_cause; + u32 cause = vcpu->arch.host_cp0_cause; enum emulation_result er = EMULATE_DONE; int ret = RESUME_GUEST; @@ -192,8 +145,8 @@ static int kvm_trap_emul_handle_tlb_ld_miss(struct kvm_vcpu *vcpu) } } else if (KVM_GUEST_KSEGX(badvaddr) < KVM_GUEST_KSEG0 || KVM_GUEST_KSEGX(badvaddr) == KVM_GUEST_KSEG23) { - kvm_debug("USER ADDR TLB ST fault: PC: %#lx, BadVaddr: %#lx\n", - vcpu->arch.pc, badvaddr); + kvm_debug("USER ADDR TLB %s fault: cause %#x, PC: %p, BadVaddr: %#lx\n", + store ? "ST" : "LD", cause, opc, badvaddr); /* * User Address (UA) fault, this could happen if @@ -213,14 +166,18 @@ static int kvm_trap_emul_handle_tlb_ld_miss(struct kvm_vcpu *vcpu) ret = RESUME_HOST; } } else if (KVM_GUEST_KSEGX(badvaddr) == KVM_GUEST_KSEG0) { + /* + * All KSEG0 faults are handled by KVM, as the guest kernel does + * not expect to ever get them + */ if (kvm_mips_handle_kseg0_tlb_fault (vcpu->arch.host_cp0_badvaddr, vcpu) < 0) { run->exit_reason = KVM_EXIT_INTERNAL_ERROR; ret = RESUME_HOST; } } else { - kvm_err("Illegal TLB ST fault address , cause %#lx, PC: %p, BadVaddr: %#lx\n", - cause, opc, badvaddr); + kvm_err("Illegal TLB %s fault address , cause %#x, PC: %p, BadVaddr: %#lx\n", + store ? "ST" : "LD", cause, opc, badvaddr); kvm_mips_dump_host_tlbs(); kvm_arch_vcpu_dump_regs(vcpu); run->exit_reason = KVM_EXIT_INTERNAL_ERROR; @@ -229,12 +186,22 @@ static int kvm_trap_emul_handle_tlb_ld_miss(struct kvm_vcpu *vcpu) return ret; } +static int kvm_trap_emul_handle_tlb_st_miss(struct kvm_vcpu *vcpu) +{ + return kvm_trap_emul_handle_tlb_miss(vcpu, true); +} + +static int kvm_trap_emul_handle_tlb_ld_miss(struct kvm_vcpu *vcpu) +{ + return kvm_trap_emul_handle_tlb_miss(vcpu, false); +} + static int kvm_trap_emul_handle_addr_err_st(struct kvm_vcpu *vcpu) { struct kvm_run *run = vcpu->run; - uint32_t __user *opc = (uint32_t __user *) vcpu->arch.pc; + u32 __user *opc = (u32 __user *) vcpu->arch.pc; unsigned long badvaddr = vcpu->arch.host_cp0_badvaddr; - unsigned long cause = vcpu->arch.host_cp0_cause; + u32 cause = vcpu->arch.host_cp0_cause; enum emulation_result er = EMULATE_DONE; int ret = RESUME_GUEST; @@ -251,7 +218,7 @@ static int kvm_trap_emul_handle_addr_err_st(struct kvm_vcpu *vcpu) ret = RESUME_HOST; } } else { - kvm_err("Address Error (STORE): cause %#lx, PC: %p, BadVaddr: %#lx\n", + kvm_err("Address Error (STORE): cause %#x, PC: %p, BadVaddr: %#lx\n", cause, opc, badvaddr); run->exit_reason = KVM_EXIT_INTERNAL_ERROR; ret = RESUME_HOST; @@ -262,9 +229,9 @@ static int kvm_trap_emul_handle_addr_err_st(struct kvm_vcpu *vcpu) static int kvm_trap_emul_handle_addr_err_ld(struct kvm_vcpu *vcpu) { struct kvm_run *run = vcpu->run; - uint32_t __user *opc = (uint32_t __user *) vcpu->arch.pc; + u32 __user *opc = (u32 __user *) vcpu->arch.pc; unsigned long badvaddr = vcpu->arch.host_cp0_badvaddr; - unsigned long cause = vcpu->arch.host_cp0_cause; + u32 cause = vcpu->arch.host_cp0_cause; enum emulation_result er = EMULATE_DONE; int ret = RESUME_GUEST; @@ -280,7 +247,7 @@ static int kvm_trap_emul_handle_addr_err_ld(struct kvm_vcpu *vcpu) ret = RESUME_HOST; } } else { - kvm_err("Address Error (LOAD): cause %#lx, PC: %p, BadVaddr: %#lx\n", + kvm_err("Address Error (LOAD): cause %#x, PC: %p, BadVaddr: %#lx\n", cause, opc, badvaddr); run->exit_reason = KVM_EXIT_INTERNAL_ERROR; ret = RESUME_HOST; @@ -292,8 +259,8 @@ static int kvm_trap_emul_handle_addr_err_ld(struct kvm_vcpu *vcpu) static int kvm_trap_emul_handle_syscall(struct kvm_vcpu *vcpu) { struct kvm_run *run = vcpu->run; - uint32_t __user *opc = (uint32_t __user *) vcpu->arch.pc; - unsigned long cause = vcpu->arch.host_cp0_cause; + u32 __user *opc = (u32 __user *) vcpu->arch.pc; + u32 cause = vcpu->arch.host_cp0_cause; enum emulation_result er = EMULATE_DONE; int ret = RESUME_GUEST; @@ -310,8 +277,8 @@ static int kvm_trap_emul_handle_syscall(struct kvm_vcpu *vcpu) static int kvm_trap_emul_handle_res_inst(struct kvm_vcpu *vcpu) { struct kvm_run *run = vcpu->run; - uint32_t __user *opc = (uint32_t __user *) vcpu->arch.pc; - unsigned long cause = vcpu->arch.host_cp0_cause; + u32 __user *opc = (u32 __user *) vcpu->arch.pc; + u32 cause = vcpu->arch.host_cp0_cause; enum emulation_result er = EMULATE_DONE; int ret = RESUME_GUEST; @@ -328,8 +295,8 @@ static int kvm_trap_emul_handle_res_inst(struct kvm_vcpu *vcpu) static int kvm_trap_emul_handle_break(struct kvm_vcpu *vcpu) { struct kvm_run *run = vcpu->run; - uint32_t __user *opc = (uint32_t __user *) vcpu->arch.pc; - unsigned long cause = vcpu->arch.host_cp0_cause; + u32 __user *opc = (u32 __user *) vcpu->arch.pc; + u32 cause = vcpu->arch.host_cp0_cause; enum emulation_result er = EMULATE_DONE; int ret = RESUME_GUEST; @@ -346,8 +313,8 @@ static int kvm_trap_emul_handle_break(struct kvm_vcpu *vcpu) static int kvm_trap_emul_handle_trap(struct kvm_vcpu *vcpu) { struct kvm_run *run = vcpu->run; - uint32_t __user *opc = (uint32_t __user *)vcpu->arch.pc; - unsigned long cause = vcpu->arch.host_cp0_cause; + u32 __user *opc = (u32 __user *)vcpu->arch.pc; + u32 cause = vcpu->arch.host_cp0_cause; enum emulation_result er = EMULATE_DONE; int ret = RESUME_GUEST; @@ -364,8 +331,8 @@ static int kvm_trap_emul_handle_trap(struct kvm_vcpu *vcpu) static int kvm_trap_emul_handle_msa_fpe(struct kvm_vcpu *vcpu) { struct kvm_run *run = vcpu->run; - uint32_t __user *opc = (uint32_t __user *)vcpu->arch.pc; - unsigned long cause = vcpu->arch.host_cp0_cause; + u32 __user *opc = (u32 __user *)vcpu->arch.pc; + u32 cause = vcpu->arch.host_cp0_cause; enum emulation_result er = EMULATE_DONE; int ret = RESUME_GUEST; @@ -382,8 +349,8 @@ static int kvm_trap_emul_handle_msa_fpe(struct kvm_vcpu *vcpu) static int kvm_trap_emul_handle_fpe(struct kvm_vcpu *vcpu) { struct kvm_run *run = vcpu->run; - uint32_t __user *opc = (uint32_t __user *)vcpu->arch.pc; - unsigned long cause = vcpu->arch.host_cp0_cause; + u32 __user *opc = (u32 __user *)vcpu->arch.pc; + u32 cause = vcpu->arch.host_cp0_cause; enum emulation_result er = EMULATE_DONE; int ret = RESUME_GUEST; @@ -407,8 +374,8 @@ static int kvm_trap_emul_handle_msa_disabled(struct kvm_vcpu *vcpu) { struct mips_coproc *cop0 = vcpu->arch.cop0; struct kvm_run *run = vcpu->run; - uint32_t __user *opc = (uint32_t __user *) vcpu->arch.pc; - unsigned long cause = vcpu->arch.host_cp0_cause; + u32 __user *opc = (u32 __user *) vcpu->arch.pc; + u32 cause = vcpu->arch.host_cp0_cause; enum emulation_result er = EMULATE_DONE; int ret = RESUME_GUEST; @@ -451,24 +418,41 @@ static int kvm_trap_emul_vm_init(struct kvm *kvm) static int kvm_trap_emul_vcpu_init(struct kvm_vcpu *vcpu) { + vcpu->arch.kscratch_enabled = 0xfc; + return 0; } static int kvm_trap_emul_vcpu_setup(struct kvm_vcpu *vcpu) { struct mips_coproc *cop0 = vcpu->arch.cop0; - uint32_t config1; + u32 config, config1; int vcpu_id = vcpu->vcpu_id; /* * Arch specific stuff, set up config registers properly so that the - * guest will come up as expected, for now we simulate a MIPS 24kc + * guest will come up as expected */ +#ifndef CONFIG_CPU_MIPSR6 + /* r2-r5, simulate a MIPS 24kc */ kvm_write_c0_guest_prid(cop0, 0x00019300); - /* Have config1, Cacheable, noncoherent, write-back, write allocate */ - kvm_write_c0_guest_config(cop0, MIPS_CONF_M | (0x3 << CP0C0_K0) | - (0x1 << CP0C0_AR) | - (MMU_TYPE_R4000 << CP0C0_MT)); +#else + /* r6+, simulate a generic QEMU machine */ + kvm_write_c0_guest_prid(cop0, 0x00010000); +#endif + /* + * Have config1, Cacheable, noncoherent, write-back, write allocate. + * Endianness, arch revision & virtually tagged icache should match + * host. + */ + config = read_c0_config() & MIPS_CONF_AR; + config |= MIPS_CONF_M | CONF_CM_CACHABLE_NONCOHERENT | MIPS_CONF_MT_TLB; +#ifdef CONFIG_CPU_BIG_ENDIAN + config |= CONF_BE; +#endif + if (cpu_has_vtag_icache) + config |= MIPS_CONF_VI; + kvm_write_c0_guest_config(cop0, config); /* Read the cache characteristics from the host Config1 Register */ config1 = (read_c0_config1() & ~0x7f); @@ -478,9 +462,8 @@ static int kvm_trap_emul_vcpu_setup(struct kvm_vcpu *vcpu) config1 |= ((KVM_MIPS_GUEST_TLB_SIZE - 1) << 25); /* We unset some bits that we aren't emulating */ - config1 &= - ~((1 << CP0C1_C2) | (1 << CP0C1_MD) | (1 << CP0C1_PC) | - (1 << CP0C1_WR) | (1 << CP0C1_CA)); + config1 &= ~(MIPS_CONF1_C2 | MIPS_CONF1_MD | MIPS_CONF1_PC | + MIPS_CONF1_WR | MIPS_CONF1_CA); kvm_write_c0_guest_config1(cop0, config1); /* Have config3, no tertiary/secondary caches implemented */ @@ -511,6 +494,17 @@ static int kvm_trap_emul_vcpu_setup(struct kvm_vcpu *vcpu) return 0; } +static unsigned long kvm_trap_emul_num_regs(struct kvm_vcpu *vcpu) +{ + return 0; +} + +static int kvm_trap_emul_copy_reg_indices(struct kvm_vcpu *vcpu, + u64 __user *indices) +{ + return 0; +} + static int kvm_trap_emul_get_one_reg(struct kvm_vcpu *vcpu, const struct kvm_one_reg *reg, s64 *v) @@ -660,6 +654,8 @@ static struct kvm_mips_callbacks kvm_trap_emul_callbacks = { .dequeue_io_int = kvm_mips_dequeue_io_int_cb, .irq_deliver = kvm_mips_irq_deliver_cb, .irq_clear = kvm_mips_irq_clear_cb, + .num_regs = kvm_trap_emul_num_regs, + .copy_reg_indices = kvm_trap_emul_copy_reg_indices, .get_one_reg = kvm_trap_emul_get_one_reg, .set_one_reg = kvm_trap_emul_set_one_reg, .vcpu_get_regs = kvm_trap_emul_vcpu_get_regs, diff --git a/arch/mips/math-emu/cp1emu.c b/arch/mips/math-emu/cp1emu.c index d96e912..6dc07fb 100644 --- a/arch/mips/math-emu/cp1emu.c +++ b/arch/mips/math-emu/cp1emu.c @@ -627,8 +627,8 @@ static int isBranchInstr(struct pt_regs *regs, struct mm_decoded_insn dec_insn, dec_insn.pc_inc + dec_insn.next_pc_inc; return 1; - case cbcond0_op: - case cbcond1_op: + case pop10_op: + case pop30_op: if (!cpu_has_mips_r6) break; if (insn.i_format.rt && !insn.i_format.rs) @@ -683,14 +683,14 @@ static int isBranchInstr(struct pt_regs *regs, struct mm_decoded_insn dec_insn, dec_insn.next_pc_inc; return 1; - case beqzcjic_op: + case pop66_op: if (!cpu_has_mips_r6) break; *contpc = regs->cp0_epc + dec_insn.pc_inc + dec_insn.next_pc_inc; return 1; - case bnezcjialc_op: + case pop76_op: if (!cpu_has_mips_r6) break; if (!insn.i_format.rs) diff --git a/arch/mips/mm/c-r4k.c b/arch/mips/mm/c-r4k.c index ef7f925..7a9c345 100644 --- a/arch/mips/mm/c-r4k.c +++ b/arch/mips/mm/c-r4k.c @@ -1206,7 +1206,7 @@ static void probe_pcache(void) c->icache.linesz; c->icache.waybit = __ffs(icache_size/c->icache.ways); - if (config & 0x8) /* VI bit */ + if (config & MIPS_CONF_VI) c->icache.flags |= MIPS_CACHE_VTAG; /* diff --git a/arch/mips/mm/uasm-micromips.c b/arch/mips/mm/uasm-micromips.c index d78178d..277cf52 100644 --- a/arch/mips/mm/uasm-micromips.c +++ b/arch/mips/mm/uasm-micromips.c @@ -53,8 +53,13 @@ static struct insn insn_table_MM[] = { { insn_bltzl, 0, 0 }, { insn_bne, M(mm_bne32_op, 0, 0, 0, 0, 0), RT | RS | BIMM }, { insn_cache, M(mm_pool32b_op, 0, 0, mm_cache_func, 0, 0), RT | RS | SIMM }, + { insn_cfc1, M(mm_pool32f_op, 0, 0, 0, mm_cfc1_op, mm_32f_73_op), RT | RS }, + { insn_cfcmsa, M(mm_pool32s_op, 0, msa_cfc_op, 0, 0, mm_32s_elm_op), RD | RE }, + { insn_ctc1, M(mm_pool32f_op, 0, 0, 0, mm_ctc1_op, mm_32f_73_op), RT | RS }, + { insn_ctcmsa, M(mm_pool32s_op, 0, msa_ctc_op, 0, 0, mm_32s_elm_op), RD | RE }, { insn_daddu, 0, 0 }, { insn_daddiu, 0, 0 }, + { insn_di, M(mm_pool32a_op, 0, 0, 0, mm_di_op, mm_pool32axf_op), RS }, { insn_divu, M(mm_pool32a_op, 0, 0, 0, mm_divu_op, mm_pool32axf_op), RT | RS }, { insn_dmfc0, 0, 0 }, { insn_dmtc0, 0, 0 }, @@ -84,6 +89,8 @@ static struct insn insn_table_MM[] = { { insn_mfhi, M(mm_pool32a_op, 0, 0, 0, mm_mfhi32_op, mm_pool32axf_op), RS }, { insn_mflo, M(mm_pool32a_op, 0, 0, 0, mm_mflo32_op, mm_pool32axf_op), RS }, { insn_mtc0, M(mm_pool32a_op, 0, 0, 0, mm_mtc0_op, mm_pool32axf_op), RT | RS | RD }, + { insn_mthi, M(mm_pool32a_op, 0, 0, 0, mm_mthi32_op, mm_pool32axf_op), RS }, + { insn_mtlo, M(mm_pool32a_op, 0, 0, 0, mm_mtlo32_op, mm_pool32axf_op), RS }, { insn_mul, M(mm_pool32a_op, 0, 0, 0, 0, mm_mul_op), RT | RS | RD }, { insn_or, M(mm_pool32a_op, 0, 0, 0, 0, mm_or32_op), RT | RS | RD }, { insn_ori, M(mm_ori32_op, 0, 0, 0, 0, 0), RT | RS | UIMM }, @@ -166,13 +173,15 @@ static void build_insn(u32 **buf, enum opcode opc, ...) op = ip->match; va_start(ap, opc); if (ip->fields & RS) { - if (opc == insn_mfc0 || opc == insn_mtc0) + if (opc == insn_mfc0 || opc == insn_mtc0 || + opc == insn_cfc1 || opc == insn_ctc1) op |= build_rt(va_arg(ap, u32)); else op |= build_rs(va_arg(ap, u32)); } if (ip->fields & RT) { - if (opc == insn_mfc0 || opc == insn_mtc0) + if (opc == insn_mfc0 || opc == insn_mtc0 || + opc == insn_cfc1 || opc == insn_ctc1) op |= build_rs(va_arg(ap, u32)); else op |= build_rt(va_arg(ap, u32)); diff --git a/arch/mips/mm/uasm-mips.c b/arch/mips/mm/uasm-mips.c index 9c2220a..cec5241 100644 --- a/arch/mips/mm/uasm-mips.c +++ b/arch/mips/mm/uasm-mips.c @@ -67,9 +67,14 @@ static struct insn insn_table[] = { #else { insn_cache, M6(cache_op, 0, 0, 0, cache6_op), RS | RT | SIMM9 }, #endif + { insn_cfc1, M(cop1_op, cfc_op, 0, 0, 0, 0), RT | RD }, + { insn_cfcmsa, M(msa_op, 0, msa_cfc_op, 0, 0, msa_elm_op), RD | RE }, + { insn_ctc1, M(cop1_op, ctc_op, 0, 0, 0, 0), RT | RD }, + { insn_ctcmsa, M(msa_op, 0, msa_ctc_op, 0, 0, msa_elm_op), RD | RE }, { insn_daddiu, M(daddiu_op, 0, 0, 0, 0, 0), RS | RT | SIMM }, { insn_daddu, M(spec_op, 0, 0, 0, 0, daddu_op), RS | RT | RD }, { insn_dinsm, M(spec3_op, 0, 0, 0, 0, dinsm_op), RS | RT | RD | RE }, + { insn_di, M(cop0_op, mfmc0_op, 0, 12, 0, 0), RT }, { insn_dins, M(spec3_op, 0, 0, 0, 0, dins_op), RS | RT | RD | RE }, { insn_divu, M(spec_op, 0, 0, 0, 0, divu_op), RS | RT }, { insn_dmfc0, M(cop0_op, dmfc_op, 0, 0, 0, 0), RT | RD | SET}, @@ -114,7 +119,13 @@ static struct insn insn_table[] = { { insn_mflo, M(spec_op, 0, 0, 0, 0, mflo_op), RD }, { insn_mtc0, M(cop0_op, mtc_op, 0, 0, 0, 0), RT | RD | SET}, { insn_mthc0, M(cop0_op, mthc0_op, 0, 0, 0, 0), RT | RD | SET}, + { insn_mthi, M(spec_op, 0, 0, 0, 0, mthi_op), RS }, + { insn_mtlo, M(spec_op, 0, 0, 0, 0, mtlo_op), RS }, +#ifndef CONFIG_CPU_MIPSR6 { insn_mul, M(spec2_op, 0, 0, 0, 0, mul_op), RS | RT | RD}, +#else + { insn_mul, M(spec_op, 0, 0, 0, mult_mul_op, mult_op), RS | RT | RD}, +#endif { insn_ori, M(ori_op, 0, 0, 0, 0, 0), RS | RT | UIMM }, { insn_or, M(spec_op, 0, 0, 0, 0, or_op), RS | RT | RD }, #ifndef CONFIG_CPU_MIPSR6 diff --git a/arch/mips/mm/uasm.c b/arch/mips/mm/uasm.c index ad718de..3e0282d 100644 --- a/arch/mips/mm/uasm.c +++ b/arch/mips/mm/uasm.c @@ -49,18 +49,19 @@ enum opcode { insn_invalid, insn_addiu, insn_addu, insn_and, insn_andi, insn_bbit0, insn_bbit1, insn_beq, insn_beql, insn_bgez, insn_bgezl, insn_bltz, insn_bltzl, - insn_bne, insn_cache, insn_daddiu, insn_daddu, insn_dins, insn_dinsm, - insn_divu, insn_dmfc0, insn_dmtc0, insn_drotr, insn_drotr32, insn_dsll, + insn_bne, insn_cache, insn_cfc1, insn_cfcmsa, insn_ctc1, insn_ctcmsa, + insn_daddiu, insn_daddu, insn_di, insn_dins, insn_dinsm, insn_divu, + insn_dmfc0, insn_dmtc0, insn_drotr, insn_drotr32, insn_dsll, insn_dsll32, insn_dsra, insn_dsrl, insn_dsrl32, insn_dsubu, insn_eret, insn_ext, insn_ins, insn_j, insn_jal, insn_jalr, insn_jr, insn_lb, insn_ld, insn_ldx, insn_lh, insn_ll, insn_lld, insn_lui, insn_lw, insn_lwx, insn_mfc0, insn_mfhc0, insn_mfhi, insn_mflo, insn_mtc0, - insn_mthc0, insn_mul, insn_or, insn_ori, insn_pref, insn_rfe, - insn_rotr, insn_sc, insn_scd, insn_sd, insn_sll, insn_sllv, insn_slt, - insn_sltiu, insn_sltu, insn_sra, insn_srl, insn_srlv, insn_subu, - insn_sw, insn_sync, insn_syscall, insn_tlbp, insn_tlbr, insn_tlbwi, - insn_tlbwr, insn_wait, insn_wsbh, insn_xor, insn_xori, insn_yield, - insn_lddir, insn_ldpte, + insn_mthc0, insn_mthi, insn_mtlo, insn_mul, insn_or, insn_ori, + insn_pref, insn_rfe, insn_rotr, insn_sc, insn_scd, insn_sd, insn_sll, + insn_sllv, insn_slt, insn_sltiu, insn_sltu, insn_sra, insn_srl, + insn_srlv, insn_subu, insn_sw, insn_sync, insn_syscall, insn_tlbp, + insn_tlbr, insn_tlbwi, insn_tlbwr, insn_wait, insn_wsbh, insn_xor, + insn_xori, insn_yield, insn_lddir, insn_ldpte, }; struct insn { @@ -268,10 +269,15 @@ I_u1s2(_bltz) I_u1s2(_bltzl) I_u1u2s3(_bne) I_u2s3u1(_cache) +I_u1u2(_cfc1) +I_u2u1(_cfcmsa) +I_u1u2(_ctc1) +I_u2u1(_ctcmsa) I_u1u2u3(_dmfc0) I_u1u2u3(_dmtc0) I_u2u1s3(_daddiu) I_u3u1u2(_daddu) +I_u1(_di); I_u1u2(_divu) I_u2u1u3(_dsll) I_u2u1u3(_dsll32) @@ -301,6 +307,8 @@ I_u1(_mfhi) I_u1(_mflo) I_u1u2u3(_mtc0) I_u1u2u3(_mthc0) +I_u1(_mthi) +I_u1(_mtlo) I_u3u1u2(_mul) I_u2u1u3(_ori) I_u3u1u2(_or) diff --git a/arch/powerpc/include/asm/hmi.h b/arch/powerpc/include/asm/hmi.h new file mode 100644 index 0000000..88b4901 --- /dev/null +++ b/arch/powerpc/include/asm/hmi.h @@ -0,0 +1,45 @@ +/* + * Hypervisor Maintenance Interrupt header file. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program. + * + * Copyright 2015 IBM Corporation + * Author: Mahesh Salgaonkar <mahesh@linux.vnet.ibm.com> + */ + +#ifndef __ASM_PPC64_HMI_H__ +#define __ASM_PPC64_HMI_H__ + +#ifdef CONFIG_PPC_BOOK3S_64 + +#define CORE_TB_RESYNC_REQ_BIT 63 +#define MAX_SUBCORE_PER_CORE 4 + +/* + * sibling_subcore_state structure is used to co-ordinate all threads + * during HMI to avoid TB corruption. This structure is allocated once + * per each core and shared by all threads on that core. + */ +struct sibling_subcore_state { + unsigned long flags; + u8 in_guest[MAX_SUBCORE_PER_CORE]; +}; + +extern void wait_for_subcore_guest_exit(void); +extern void wait_for_tb_resync(void); +#else +static inline void wait_for_subcore_guest_exit(void) { } +static inline void wait_for_tb_resync(void) { } +#endif +#endif /* __ASM_PPC64_HMI_H__ */ diff --git a/arch/powerpc/include/asm/paca.h b/arch/powerpc/include/asm/paca.h index ad171e9..148303e7 100644 --- a/arch/powerpc/include/asm/paca.h +++ b/arch/powerpc/include/asm/paca.h @@ -26,6 +26,7 @@ #include <asm/kvm_book3s_asm.h> #endif #include <asm/accounting.h> +#include <asm/hmi.h> register struct paca_struct *local_paca asm("r13"); @@ -182,6 +183,11 @@ struct paca_struct { */ u16 in_mce; u8 hmi_event_available; /* HMI event is available */ + /* + * Bitmap for sibling subcore status. See kvm/book3s_hv_ras.c for + * more details + */ + struct sibling_subcore_state *sibling_subcore_state; #endif /* Stuff for accurate time accounting */ diff --git a/arch/powerpc/kernel/Makefile b/arch/powerpc/kernel/Makefile index fe4c075..b2027a5 100644 --- a/arch/powerpc/kernel/Makefile +++ b/arch/powerpc/kernel/Makefile @@ -41,7 +41,7 @@ obj-$(CONFIG_VDSO32) += vdso32/ obj-$(CONFIG_HAVE_HW_BREAKPOINT) += hw_breakpoint.o obj-$(CONFIG_PPC_BOOK3S_64) += cpu_setup_ppc970.o cpu_setup_pa6t.o obj-$(CONFIG_PPC_BOOK3S_64) += cpu_setup_power.o -obj-$(CONFIG_PPC_BOOK3S_64) += mce.o mce_power.o +obj-$(CONFIG_PPC_BOOK3S_64) += mce.o mce_power.o hmi.o obj-$(CONFIG_PPC_BOOK3E_64) += exceptions-64e.o idle_book3e.o obj-$(CONFIG_PPC64) += vdso64/ obj-$(CONFIG_ALTIVEC) += vecemu.o diff --git a/arch/powerpc/kernel/exceptions-64s.S b/arch/powerpc/kernel/exceptions-64s.S index 6200e49..694def6 100644 --- a/arch/powerpc/kernel/exceptions-64s.S +++ b/arch/powerpc/kernel/exceptions-64s.S @@ -671,6 +671,8 @@ BEGIN_FTR_SECTION beq h_doorbell_common cmpwi r3,0xea0 beq h_virt_irq_common + cmpwi r3,0xe60 + beq hmi_exception_common FTR_SECTION_ELSE cmpwi r3,0xa00 beq doorbell_super_common @@ -1172,7 +1174,7 @@ fwnmi_data_area: .globl hmi_exception_early hmi_exception_early: - EXCEPTION_PROLOG_1(PACA_EXGEN, NOTEST, 0xe60) + EXCEPTION_PROLOG_1(PACA_EXGEN, KVMTEST, 0xe62) mr r10,r1 /* Save r1 */ ld r1,PACAEMERGSP(r13) /* Use emergency stack */ subi r1,r1,INT_FRAME_SIZE /* alloc stack frame */ diff --git a/arch/powerpc/kernel/hmi.c b/arch/powerpc/kernel/hmi.c new file mode 100644 index 0000000..e3f738e --- /dev/null +++ b/arch/powerpc/kernel/hmi.c @@ -0,0 +1,56 @@ +/* + * Hypervisor Maintenance Interrupt (HMI) handling. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program. + * + * Copyright 2015 IBM Corporation + * Author: Mahesh Salgaonkar <mahesh@linux.vnet.ibm.com> + */ + +#undef DEBUG + +#include <linux/types.h> +#include <linux/compiler.h> +#include <asm/paca.h> +#include <asm/hmi.h> + +void wait_for_subcore_guest_exit(void) +{ + int i; + + /* + * NULL bitmap pointer indicates that KVM module hasn't + * been loaded yet and hence no guests are running. + * If no KVM is in use, no need to co-ordinate among threads + * as all of them will always be in host and no one is going + * to modify TB other than the opal hmi handler. + * Hence, just return from here. + */ + if (!local_paca->sibling_subcore_state) + return; + + for (i = 0; i < MAX_SUBCORE_PER_CORE; i++) + while (local_paca->sibling_subcore_state->in_guest[i]) + cpu_relax(); +} + +void wait_for_tb_resync(void) +{ + if (!local_paca->sibling_subcore_state) + return; + + while (test_bit(CORE_TB_RESYNC_REQ_BIT, + &local_paca->sibling_subcore_state->flags)) + cpu_relax(); +} diff --git a/arch/powerpc/kernel/idle_book3s.S b/arch/powerpc/kernel/idle_book3s.S index 335eb6c..8a56a51 100644 --- a/arch/powerpc/kernel/idle_book3s.S +++ b/arch/powerpc/kernel/idle_book3s.S @@ -336,7 +336,9 @@ ALT_FTR_SECTION_END_NESTED_IFSET(CPU_FTR_ARCH_207S, 66); \ ld r2,PACATOC(r13); \ ld r1,PACAR1(r13); \ std r3,ORIG_GPR3(r1); /* Save original r3 */ \ - bl opal_rm_handle_hmi; \ + li r3,0; /* NULL argument */ \ + bl hmi_exception_realmode; \ + nop; \ ld r3,ORIG_GPR3(r1); /* Restore original r3 */ \ 20: nop; diff --git a/arch/powerpc/kernel/traps.c b/arch/powerpc/kernel/traps.c index f7e2f2e..2cb5892 100644 --- a/arch/powerpc/kernel/traps.c +++ b/arch/powerpc/kernel/traps.c @@ -61,6 +61,7 @@ #include <asm/tm.h> #include <asm/debug.h> #include <asm/asm-prototypes.h> +#include <asm/hmi.h> #include <sysdev/fsl_pci.h> #if defined(CONFIG_DEBUGGER) || defined(CONFIG_KEXEC) @@ -308,9 +309,13 @@ long hmi_exception_realmode(struct pt_regs *regs) { __this_cpu_inc(irq_stat.hmi_exceptions); + wait_for_subcore_guest_exit(); + if (ppc_md.hmi_exception_early) ppc_md.hmi_exception_early(regs); + wait_for_tb_resync(); + return 0; } diff --git a/arch/powerpc/kvm/book3s_hv.c b/arch/powerpc/kvm/book3s_hv.c index e20beae..2fd5580 100644 --- a/arch/powerpc/kvm/book3s_hv.c +++ b/arch/powerpc/kvm/book3s_hv.c @@ -52,6 +52,7 @@ #include <asm/switch_to.h> #include <asm/smp.h> #include <asm/dbell.h> +#include <asm/hmi.h> #include <linux/gfp.h> #include <linux/vmalloc.h> #include <linux/highmem.h> @@ -2522,7 +2523,7 @@ static noinline void kvmppc_run_core(struct kvmppc_vcore *vc) list_for_each_entry(pvc, &core_info.vcs[sub], preempt_list) spin_unlock(&pvc->lock); - kvm_guest_enter(); + guest_enter(); srcu_idx = srcu_read_lock(&vc->kvm->srcu); @@ -2570,7 +2571,7 @@ static noinline void kvmppc_run_core(struct kvmppc_vcore *vc) /* make sure updates to secondary vcpu structs are visible now */ smp_mb(); - kvm_guest_exit(); + guest_exit(); for (sub = 0; sub < core_info.n_subcores; ++sub) list_for_each_entry_safe(pvc, vcnext, &core_info.vcs[sub], @@ -3401,6 +3402,38 @@ static struct kvmppc_ops kvm_ops_hv = { .hcall_implemented = kvmppc_hcall_impl_hv, }; +static int kvm_init_subcore_bitmap(void) +{ + int i, j; + int nr_cores = cpu_nr_cores(); + struct sibling_subcore_state *sibling_subcore_state; + + for (i = 0; i < nr_cores; i++) { + int first_cpu = i * threads_per_core; + int node = cpu_to_node(first_cpu); + + /* Ignore if it is already allocated. */ + if (paca[first_cpu].sibling_subcore_state) + continue; + + sibling_subcore_state = + kmalloc_node(sizeof(struct sibling_subcore_state), + GFP_KERNEL, node); + if (!sibling_subcore_state) + return -ENOMEM; + + memset(sibling_subcore_state, 0, + sizeof(struct sibling_subcore_state)); + + for (j = 0; j < threads_per_core; j++) { + int cpu = first_cpu + j; + + paca[cpu].sibling_subcore_state = sibling_subcore_state; + } + } + return 0; +} + static int kvmppc_book3s_init_hv(void) { int r; @@ -3411,6 +3444,10 @@ static int kvmppc_book3s_init_hv(void) if (r < 0) return -ENODEV; + r = kvm_init_subcore_bitmap(); + if (r) + return r; + kvm_ops_hv.owner = THIS_MODULE; kvmppc_hv_ops = &kvm_ops_hv; diff --git a/arch/powerpc/kvm/book3s_hv_ras.c b/arch/powerpc/kvm/book3s_hv_ras.c index 93b5f5c..0fa70a9 100644 --- a/arch/powerpc/kvm/book3s_hv_ras.c +++ b/arch/powerpc/kvm/book3s_hv_ras.c @@ -13,6 +13,9 @@ #include <linux/kernel.h> #include <asm/opal.h> #include <asm/mce.h> +#include <asm/machdep.h> +#include <asm/cputhreads.h> +#include <asm/hmi.h> /* SRR1 bits for machine check on POWER7 */ #define SRR1_MC_LDSTERR (1ul << (63-42)) @@ -140,3 +143,176 @@ long kvmppc_realmode_machine_check(struct kvm_vcpu *vcpu) { return kvmppc_realmode_mc_power7(vcpu); } + +/* Check if dynamic split is in force and return subcore size accordingly. */ +static inline int kvmppc_cur_subcore_size(void) +{ + if (local_paca->kvm_hstate.kvm_split_mode) + return local_paca->kvm_hstate.kvm_split_mode->subcore_size; + + return threads_per_subcore; +} + +void kvmppc_subcore_enter_guest(void) +{ + int thread_id, subcore_id; + + thread_id = cpu_thread_in_core(local_paca->paca_index); + subcore_id = thread_id / kvmppc_cur_subcore_size(); + + local_paca->sibling_subcore_state->in_guest[subcore_id] = 1; +} + +void kvmppc_subcore_exit_guest(void) +{ + int thread_id, subcore_id; + + thread_id = cpu_thread_in_core(local_paca->paca_index); + subcore_id = thread_id / kvmppc_cur_subcore_size(); + + local_paca->sibling_subcore_state->in_guest[subcore_id] = 0; +} + +static bool kvmppc_tb_resync_required(void) +{ + if (test_and_set_bit(CORE_TB_RESYNC_REQ_BIT, + &local_paca->sibling_subcore_state->flags)) + return false; + + return true; +} + +static void kvmppc_tb_resync_done(void) +{ + clear_bit(CORE_TB_RESYNC_REQ_BIT, + &local_paca->sibling_subcore_state->flags); +} + +/* + * kvmppc_realmode_hmi_handler() is called only by primary thread during + * guest exit path. + * + * There are multiple reasons why HMI could occur, one of them is + * Timebase (TB) error. If this HMI is due to TB error, then TB would + * have been in stopped state. The opal hmi handler Will fix it and + * restore the TB value with host timebase value. For HMI caused due + * to non-TB errors, opal hmi handler will not touch/restore TB register + * and hence there won't be any change in TB value. + * + * Since we are not sure about the cause of this HMI, we can't be sure + * about the content of TB register whether it holds guest or host timebase + * value. Hence the idea is to resync the TB on every HMI, so that we + * know about the exact state of the TB value. Resync TB call will + * restore TB to host timebase. + * + * Things to consider: + * - On TB error, HMI interrupt is reported on all the threads of the core + * that has encountered TB error irrespective of split-core mode. + * - The very first thread on the core that get chance to fix TB error + * would rsync the TB with local chipTOD value. + * - The resync TB is a core level action i.e. it will sync all the TBs + * in that core independent of split-core mode. This means if we trigger + * TB sync from a thread from one subcore, it would affect TB values of + * sibling subcores of the same core. + * + * All threads need to co-ordinate before making opal hmi handler. + * All threads will use sibling_subcore_state->in_guest[] (shared by all + * threads in the core) in paca which holds information about whether + * sibling subcores are in Guest mode or host mode. The in_guest[] array + * is of size MAX_SUBCORE_PER_CORE=4, indexed using subcore id to set/unset + * subcore status. Only primary threads from each subcore is responsible + * to set/unset its designated array element while entering/exiting the + * guset. + * + * After invoking opal hmi handler call, one of the thread (of entire core) + * will need to resync the TB. Bit 63 from subcore state bitmap flags + * (sibling_subcore_state->flags) will be used to co-ordinate between + * primary threads to decide who takes up the responsibility. + * + * This is what we do: + * - Primary thread from each subcore tries to set resync required bit[63] + * of paca->sibling_subcore_state->flags. + * - The first primary thread that is able to set the flag takes the + * responsibility of TB resync. (Let us call it as thread leader) + * - All other threads which are in host will call + * wait_for_subcore_guest_exit() and wait for in_guest[0-3] from + * paca->sibling_subcore_state to get cleared. + * - All the primary thread will clear its subcore status from subcore + * state in_guest[] array respectively. + * - Once all primary threads clear in_guest[0-3], all of them will invoke + * opal hmi handler. + * - Now all threads will wait for TB resync to complete by invoking + * wait_for_tb_resync() except the thread leader. + * - Thread leader will do a TB resync by invoking opal_resync_timebase() + * call and the it will clear the resync required bit. + * - All other threads will now come out of resync wait loop and proceed + * with individual execution. + * - On return of this function, primary thread will signal all + * secondary threads to proceed. + * - All secondary threads will eventually call opal hmi handler on + * their exit path. + */ + +long kvmppc_realmode_hmi_handler(void) +{ + int ptid = local_paca->kvm_hstate.ptid; + bool resync_req; + + /* This is only called on primary thread. */ + BUG_ON(ptid != 0); + __this_cpu_inc(irq_stat.hmi_exceptions); + + /* + * By now primary thread has already completed guest->host + * partition switch but haven't signaled secondaries yet. + * All the secondary threads on this subcore is waiting + * for primary thread to signal them to go ahead. + * + * For threads from subcore which isn't in guest, they all will + * wait until all other subcores on this core exit the guest. + * + * Now set the resync required bit. If you are the first to + * set this bit then kvmppc_tb_resync_required() function will + * return true. For rest all other subcores + * kvmppc_tb_resync_required() will return false. + * + * If resync_req == true, then this thread is responsible to + * initiate TB resync after hmi handler has completed. + * All other threads on this core will wait until this thread + * clears the resync required bit flag. + */ + resync_req = kvmppc_tb_resync_required(); + + /* Reset the subcore status to indicate it has exited guest */ + kvmppc_subcore_exit_guest(); + + /* + * Wait for other subcores on this core to exit the guest. + * All the primary threads and threads from subcore that are + * not in guest will wait here until all subcores are out + * of guest context. + */ + wait_for_subcore_guest_exit(); + + /* + * At this point we are sure that primary threads from each + * subcore on this core have completed guest->host partition + * switch. Now it is safe to call HMI handler. + */ + if (ppc_md.hmi_exception_early) + ppc_md.hmi_exception_early(NULL); + + /* + * Check if this thread is responsible to resync TB. + * All other threads will wait until this thread completes the + * TB resync. + */ + if (resync_req) { + opal_resync_timebase(); + /* Reset TB resync req bit */ + kvmppc_tb_resync_done(); + } else { + wait_for_tb_resync(); + } + return 0; +} diff --git a/arch/powerpc/kvm/book3s_hv_rmhandlers.S b/arch/powerpc/kvm/book3s_hv_rmhandlers.S index 86f0cae..9756555 100644 --- a/arch/powerpc/kvm/book3s_hv_rmhandlers.S +++ b/arch/powerpc/kvm/book3s_hv_rmhandlers.S @@ -29,6 +29,7 @@ #include <asm/kvm_book3s_asm.h> #include <asm/book3s/64/mmu-hash.h> #include <asm/tm.h> +#include <asm/opal.h> #define VCPU_GPRS_TM(reg) (((reg) * ULONG_SIZE) + VCPU_GPR_TM) @@ -373,6 +374,18 @@ kvm_secondary_got_guest: lwsync std r0, HSTATE_KVM_VCORE(r13) + /* + * All secondaries exiting guest will fall through this path. + * Before proceeding, just check for HMI interrupt and + * invoke opal hmi handler. By now we are sure that the + * primary thread on this core/subcore has already made partition + * switch/TB resync and we are good to call opal hmi handler. + */ + cmpwi r12, BOOK3S_INTERRUPT_HMI + bne kvm_no_guest + + li r3,0 /* NULL argument */ + bl hmi_exception_realmode /* * At this point we have finished executing in the guest. * We need to wait for hwthread_req to become zero, since @@ -428,6 +441,22 @@ kvm_no_guest: */ kvm_unsplit_nap: /* + * When secondaries are napping in kvm_unsplit_nap() with + * hwthread_req = 1, HMI goes ignored even though subcores are + * already exited the guest. Hence HMI keeps waking up secondaries + * from nap in a loop and secondaries always go back to nap since + * no vcore is assigned to them. This makes impossible for primary + * thread to get hold of secondary threads resulting into a soft + * lockup in KVM path. + * + * Let us check if HMI is pending and handle it before we go to nap. + */ + cmpwi r12, BOOK3S_INTERRUPT_HMI + bne 55f + li r3, 0 /* NULL argument */ + bl hmi_exception_realmode +55: + /* * Ensure that secondary doesn't nap when it has * its vcore pointer set. */ @@ -601,6 +630,11 @@ BEGIN_FTR_SECTION mtspr SPRN_DPDES, r8 END_FTR_SECTION_IFSET(CPU_FTR_ARCH_207S) + /* Mark the subcore state as inside guest */ + bl kvmppc_subcore_enter_guest + nop + ld r5, HSTATE_KVM_VCORE(r13) + ld r4, HSTATE_KVM_VCPU(r13) li r0,1 stb r0,VCORE_IN_GUEST(r5) /* signal secondaries to continue */ @@ -655,112 +689,8 @@ END_FTR_SECTION_IFCLR(CPU_FTR_ARCH_207S) #ifdef CONFIG_PPC_TRANSACTIONAL_MEM BEGIN_FTR_SECTION - b skip_tm -END_FTR_SECTION_IFCLR(CPU_FTR_TM) - - /* Turn on TM/FP/VSX/VMX so we can restore them. */ - mfmsr r5 - li r6, MSR_TM >> 32 - sldi r6, r6, 32 - or r5, r5, r6 - ori r5, r5, MSR_FP - oris r5, r5, (MSR_VEC | MSR_VSX)@h - mtmsrd r5 - - /* - * The user may change these outside of a transaction, so they must - * always be context switched. - */ - ld r5, VCPU_TFHAR(r4) - ld r6, VCPU_TFIAR(r4) - ld r7, VCPU_TEXASR(r4) - mtspr SPRN_TFHAR, r5 - mtspr SPRN_TFIAR, r6 - mtspr SPRN_TEXASR, r7 - - ld r5, VCPU_MSR(r4) - rldicl. r5, r5, 64 - MSR_TS_S_LG, 62 - beq skip_tm /* TM not active in guest */ - - /* Make sure the failure summary is set, otherwise we'll program check - * when we trechkpt. It's possible that this might have been not set - * on a kvmppc_set_one_reg() call but we shouldn't let this crash the - * host. - */ - oris r7, r7, (TEXASR_FS)@h - mtspr SPRN_TEXASR, r7 - - /* - * We need to load up the checkpointed state for the guest. - * We need to do this early as it will blow away any GPRs, VSRs and - * some SPRs. - */ - - mr r31, r4 - addi r3, r31, VCPU_FPRS_TM - bl load_fp_state - addi r3, r31, VCPU_VRS_TM - bl load_vr_state - mr r4, r31 - lwz r7, VCPU_VRSAVE_TM(r4) - mtspr SPRN_VRSAVE, r7 - - ld r5, VCPU_LR_TM(r4) - lwz r6, VCPU_CR_TM(r4) - ld r7, VCPU_CTR_TM(r4) - ld r8, VCPU_AMR_TM(r4) - ld r9, VCPU_TAR_TM(r4) - mtlr r5 - mtcr r6 - mtctr r7 - mtspr SPRN_AMR, r8 - mtspr SPRN_TAR, r9 - - /* - * Load up PPR and DSCR values but don't put them in the actual SPRs - * till the last moment to avoid running with userspace PPR and DSCR for - * too long. - */ - ld r29, VCPU_DSCR_TM(r4) - ld r30, VCPU_PPR_TM(r4) - - std r2, PACATMSCRATCH(r13) /* Save TOC */ - - /* Clear the MSR RI since r1, r13 are all going to be foobar. */ - li r5, 0 - mtmsrd r5, 1 - - /* Load GPRs r0-r28 */ - reg = 0 - .rept 29 - ld reg, VCPU_GPRS_TM(reg)(r31) - reg = reg + 1 - .endr - - mtspr SPRN_DSCR, r29 - mtspr SPRN_PPR, r30 - - /* Load final GPRs */ - ld 29, VCPU_GPRS_TM(29)(r31) - ld 30, VCPU_GPRS_TM(30)(r31) - ld 31, VCPU_GPRS_TM(31)(r31) - - /* TM checkpointed state is now setup. All GPRs are now volatile. */ - TRECHKPT - - /* Now let's get back the state we need. */ - HMT_MEDIUM - GET_PACA(r13) - ld r29, HSTATE_DSCR(r13) - mtspr SPRN_DSCR, r29 - ld r4, HSTATE_KVM_VCPU(r13) - ld r1, HSTATE_HOST_R1(r13) - ld r2, PACATMSCRATCH(r13) - - /* Set the MSR RI since we have our registers back. */ - li r5, MSR_RI - mtmsrd r5, 1 -skip_tm: + bl kvmppc_restore_tm +END_FTR_SECTION_IFSET(CPU_FTR_TM) #endif /* Load guest PMU registers */ @@ -841,12 +771,6 @@ BEGIN_FTR_SECTION /* Skip next section on POWER7 */ b 8f END_FTR_SECTION_IFCLR(CPU_FTR_ARCH_207S) - /* Turn on TM so we can access TFHAR/TFIAR/TEXASR */ - mfmsr r8 - li r0, 1 - rldimi r8, r0, MSR_TM_LG, 63-MSR_TM_LG - mtmsrd r8 - /* Load up POWER8-specific registers */ ld r5, VCPU_IAMR(r4) lwz r6, VCPU_PSPB(r4) @@ -1436,106 +1360,8 @@ END_FTR_SECTION_IFCLR(CPU_FTR_ARCH_207S) #ifdef CONFIG_PPC_TRANSACTIONAL_MEM BEGIN_FTR_SECTION - b 2f -END_FTR_SECTION_IFCLR(CPU_FTR_TM) - /* Turn on TM. */ - mfmsr r8 - li r0, 1 - rldimi r8, r0, MSR_TM_LG, 63-MSR_TM_LG - mtmsrd r8 - - ld r5, VCPU_MSR(r9) - rldicl. r5, r5, 64 - MSR_TS_S_LG, 62 - beq 1f /* TM not active in guest. */ - - li r3, TM_CAUSE_KVM_RESCHED - - /* Clear the MSR RI since r1, r13 are all going to be foobar. */ - li r5, 0 - mtmsrd r5, 1 - - /* All GPRs are volatile at this point. */ - TRECLAIM(R3) - - /* Temporarily store r13 and r9 so we have some regs to play with */ - SET_SCRATCH0(r13) - GET_PACA(r13) - std r9, PACATMSCRATCH(r13) - ld r9, HSTATE_KVM_VCPU(r13) - - /* Get a few more GPRs free. */ - std r29, VCPU_GPRS_TM(29)(r9) - std r30, VCPU_GPRS_TM(30)(r9) - std r31, VCPU_GPRS_TM(31)(r9) - - /* Save away PPR and DSCR soon so don't run with user values. */ - mfspr r31, SPRN_PPR - HMT_MEDIUM - mfspr r30, SPRN_DSCR - ld r29, HSTATE_DSCR(r13) - mtspr SPRN_DSCR, r29 - - /* Save all but r9, r13 & r29-r31 */ - reg = 0 - .rept 29 - .if (reg != 9) && (reg != 13) - std reg, VCPU_GPRS_TM(reg)(r9) - .endif - reg = reg + 1 - .endr - /* ... now save r13 */ - GET_SCRATCH0(r4) - std r4, VCPU_GPRS_TM(13)(r9) - /* ... and save r9 */ - ld r4, PACATMSCRATCH(r13) - std r4, VCPU_GPRS_TM(9)(r9) - - /* Reload stack pointer and TOC. */ - ld r1, HSTATE_HOST_R1(r13) - ld r2, PACATOC(r13) - - /* Set MSR RI now we have r1 and r13 back. */ - li r5, MSR_RI - mtmsrd r5, 1 - - /* Save away checkpinted SPRs. */ - std r31, VCPU_PPR_TM(r9) - std r30, VCPU_DSCR_TM(r9) - mflr r5 - mfcr r6 - mfctr r7 - mfspr r8, SPRN_AMR - mfspr r10, SPRN_TAR - std r5, VCPU_LR_TM(r9) - stw r6, VCPU_CR_TM(r9) - std r7, VCPU_CTR_TM(r9) - std r8, VCPU_AMR_TM(r9) - std r10, VCPU_TAR_TM(r9) - - /* Restore r12 as trap number. */ - lwz r12, VCPU_TRAP(r9) - - /* Save FP/VSX. */ - addi r3, r9, VCPU_FPRS_TM - bl store_fp_state - addi r3, r9, VCPU_VRS_TM - bl store_vr_state - mfspr r6, SPRN_VRSAVE - stw r6, VCPU_VRSAVE_TM(r9) -1: - /* - * We need to save these SPRs after the treclaim so that the software - * error code is recorded correctly in the TEXASR. Also the user may - * change these outside of a transaction, so they must always be - * context switched. - */ - mfspr r5, SPRN_TFHAR - mfspr r6, SPRN_TFIAR - mfspr r7, SPRN_TEXASR - std r5, VCPU_TFHAR(r9) - std r6, VCPU_TFIAR(r9) - std r7, VCPU_TEXASR(r9) -2: + bl kvmppc_save_tm +END_FTR_SECTION_IFSET(CPU_FTR_TM) #endif /* Increment yield count if they have a VPA */ @@ -1683,6 +1509,23 @@ BEGIN_FTR_SECTION mtspr SPRN_DPDES, r8 END_FTR_SECTION_IFSET(CPU_FTR_ARCH_207S) + /* If HMI, call kvmppc_realmode_hmi_handler() */ + cmpwi r12, BOOK3S_INTERRUPT_HMI + bne 27f + bl kvmppc_realmode_hmi_handler + nop + li r12, BOOK3S_INTERRUPT_HMI + /* + * At this point kvmppc_realmode_hmi_handler would have resync-ed + * the TB. Hence it is not required to subtract guest timebase + * offset from timebase. So, skip it. + * + * Also, do not call kvmppc_subcore_exit_guest() because it has + * been invoked as part of kvmppc_realmode_hmi_handler(). + */ + b 30f + +27: /* Subtract timebase offset from timebase */ ld r8,VCORE_TB_OFFSET(r5) cmpdi r8,0 @@ -1698,8 +1541,13 @@ END_FTR_SECTION_IFSET(CPU_FTR_ARCH_207S) addis r8,r8,0x100 /* if so, increment upper 40 bits */ mtspr SPRN_TBU40,r8 +17: bl kvmppc_subcore_exit_guest + nop +30: ld r5,HSTATE_KVM_VCORE(r13) + ld r4,VCORE_KVM(r5) /* pointer to struct kvm */ + /* Reset PCR */ -17: ld r0, VCORE_PCR(r5) + ld r0, VCORE_PCR(r5) cmpdi r0, 0 beq 18f li r0, 0 @@ -2245,6 +2093,13 @@ _GLOBAL(kvmppc_h_cede) /* r3 = vcpu pointer, r11 = msr, r13 = paca */ /* save FP state */ bl kvmppc_save_fp +#ifdef CONFIG_PPC_TRANSACTIONAL_MEM +BEGIN_FTR_SECTION + ld r9, HSTATE_KVM_VCPU(r13) + bl kvmppc_save_tm +END_FTR_SECTION_IFSET(CPU_FTR_TM) +#endif + /* * Set DEC to the smaller of DEC and HDEC, so that we wake * no later than the end of our timeslice (HDEC interrupts @@ -2321,6 +2176,12 @@ kvm_end_cede: bl kvmhv_accumulate_time #endif +#ifdef CONFIG_PPC_TRANSACTIONAL_MEM +BEGIN_FTR_SECTION + bl kvmppc_restore_tm +END_FTR_SECTION_IFSET(CPU_FTR_TM) +#endif + /* load up FP state */ bl kvmppc_load_fp @@ -2461,6 +2322,8 @@ BEGIN_FTR_SECTION cmpwi r6, 3 /* hypervisor doorbell? */ beq 3f END_FTR_SECTION_IFSET(CPU_FTR_ARCH_207S) + cmpwi r6, 0xa /* Hypervisor maintenance ? */ + beq 4f li r3, 1 /* anything else, return 1 */ 0: blr @@ -2482,6 +2345,11 @@ END_FTR_SECTION_IFSET(CPU_FTR_ARCH_207S) li r3, -1 blr + /* Woken up due to Hypervisor maintenance interrupt */ +4: li r12, BOOK3S_INTERRUPT_HMI + li r3, 1 + blr + /* * Determine what sort of external interrupt is pending (if any). * Returns: @@ -2631,6 +2499,239 @@ END_FTR_SECTION_IFSET(CPU_FTR_ALTIVEC) mr r4,r31 blr +#ifdef CONFIG_PPC_TRANSACTIONAL_MEM +/* + * Save transactional state and TM-related registers. + * Called with r9 pointing to the vcpu struct. + * This can modify all checkpointed registers, but + * restores r1, r2 and r9 (vcpu pointer) before exit. + */ +kvmppc_save_tm: + mflr r0 + std r0, PPC_LR_STKOFF(r1) + + /* Turn on TM. */ + mfmsr r8 + li r0, 1 + rldimi r8, r0, MSR_TM_LG, 63-MSR_TM_LG + mtmsrd r8 + + ld r5, VCPU_MSR(r9) + rldicl. r5, r5, 64 - MSR_TS_S_LG, 62 + beq 1f /* TM not active in guest. */ + + std r1, HSTATE_HOST_R1(r13) + li r3, TM_CAUSE_KVM_RESCHED + + /* Clear the MSR RI since r1, r13 are all going to be foobar. */ + li r5, 0 + mtmsrd r5, 1 + + /* All GPRs are volatile at this point. */ + TRECLAIM(R3) + + /* Temporarily store r13 and r9 so we have some regs to play with */ + SET_SCRATCH0(r13) + GET_PACA(r13) + std r9, PACATMSCRATCH(r13) + ld r9, HSTATE_KVM_VCPU(r13) + + /* Get a few more GPRs free. */ + std r29, VCPU_GPRS_TM(29)(r9) + std r30, VCPU_GPRS_TM(30)(r9) + std r31, VCPU_GPRS_TM(31)(r9) + + /* Save away PPR and DSCR soon so don't run with user values. */ + mfspr r31, SPRN_PPR + HMT_MEDIUM + mfspr r30, SPRN_DSCR + ld r29, HSTATE_DSCR(r13) + mtspr SPRN_DSCR, r29 + + /* Save all but r9, r13 & r29-r31 */ + reg = 0 + .rept 29 + .if (reg != 9) && (reg != 13) + std reg, VCPU_GPRS_TM(reg)(r9) + .endif + reg = reg + 1 + .endr + /* ... now save r13 */ + GET_SCRATCH0(r4) + std r4, VCPU_GPRS_TM(13)(r9) + /* ... and save r9 */ + ld r4, PACATMSCRATCH(r13) + std r4, VCPU_GPRS_TM(9)(r9) + + /* Reload stack pointer and TOC. */ + ld r1, HSTATE_HOST_R1(r13) + ld r2, PACATOC(r13) + + /* Set MSR RI now we have r1 and r13 back. */ + li r5, MSR_RI + mtmsrd r5, 1 + + /* Save away checkpinted SPRs. */ + std r31, VCPU_PPR_TM(r9) + std r30, VCPU_DSCR_TM(r9) + mflr r5 + mfcr r6 + mfctr r7 + mfspr r8, SPRN_AMR + mfspr r10, SPRN_TAR + std r5, VCPU_LR_TM(r9) + stw r6, VCPU_CR_TM(r9) + std r7, VCPU_CTR_TM(r9) + std r8, VCPU_AMR_TM(r9) + std r10, VCPU_TAR_TM(r9) + + /* Restore r12 as trap number. */ + lwz r12, VCPU_TRAP(r9) + + /* Save FP/VSX. */ + addi r3, r9, VCPU_FPRS_TM + bl store_fp_state + addi r3, r9, VCPU_VRS_TM + bl store_vr_state + mfspr r6, SPRN_VRSAVE + stw r6, VCPU_VRSAVE_TM(r9) +1: + /* + * We need to save these SPRs after the treclaim so that the software + * error code is recorded correctly in the TEXASR. Also the user may + * change these outside of a transaction, so they must always be + * context switched. + */ + mfspr r5, SPRN_TFHAR + mfspr r6, SPRN_TFIAR + mfspr r7, SPRN_TEXASR + std r5, VCPU_TFHAR(r9) + std r6, VCPU_TFIAR(r9) + std r7, VCPU_TEXASR(r9) + + ld r0, PPC_LR_STKOFF(r1) + mtlr r0 + blr + +/* + * Restore transactional state and TM-related registers. + * Called with r4 pointing to the vcpu struct. + * This potentially modifies all checkpointed registers. + * It restores r1, r2, r4 from the PACA. + */ +kvmppc_restore_tm: + mflr r0 + std r0, PPC_LR_STKOFF(r1) + + /* Turn on TM/FP/VSX/VMX so we can restore them. */ + mfmsr r5 + li r6, MSR_TM >> 32 + sldi r6, r6, 32 + or r5, r5, r6 + ori r5, r5, MSR_FP + oris r5, r5, (MSR_VEC | MSR_VSX)@h + mtmsrd r5 + + /* + * The user may change these outside of a transaction, so they must + * always be context switched. + */ + ld r5, VCPU_TFHAR(r4) + ld r6, VCPU_TFIAR(r4) + ld r7, VCPU_TEXASR(r4) + mtspr SPRN_TFHAR, r5 + mtspr SPRN_TFIAR, r6 + mtspr SPRN_TEXASR, r7 + + ld r5, VCPU_MSR(r4) + rldicl. r5, r5, 64 - MSR_TS_S_LG, 62 + beqlr /* TM not active in guest */ + std r1, HSTATE_HOST_R1(r13) + + /* Make sure the failure summary is set, otherwise we'll program check + * when we trechkpt. It's possible that this might have been not set + * on a kvmppc_set_one_reg() call but we shouldn't let this crash the + * host. + */ + oris r7, r7, (TEXASR_FS)@h + mtspr SPRN_TEXASR, r7 + + /* + * We need to load up the checkpointed state for the guest. + * We need to do this early as it will blow away any GPRs, VSRs and + * some SPRs. + */ + + mr r31, r4 + addi r3, r31, VCPU_FPRS_TM + bl load_fp_state + addi r3, r31, VCPU_VRS_TM + bl load_vr_state + mr r4, r31 + lwz r7, VCPU_VRSAVE_TM(r4) + mtspr SPRN_VRSAVE, r7 + + ld r5, VCPU_LR_TM(r4) + lwz r6, VCPU_CR_TM(r4) + ld r7, VCPU_CTR_TM(r4) + ld r8, VCPU_AMR_TM(r4) + ld r9, VCPU_TAR_TM(r4) + mtlr r5 + mtcr r6 + mtctr r7 + mtspr SPRN_AMR, r8 + mtspr SPRN_TAR, r9 + + /* + * Load up PPR and DSCR values but don't put them in the actual SPRs + * till the last moment to avoid running with userspace PPR and DSCR for + * too long. + */ + ld r29, VCPU_DSCR_TM(r4) + ld r30, VCPU_PPR_TM(r4) + + std r2, PACATMSCRATCH(r13) /* Save TOC */ + + /* Clear the MSR RI since r1, r13 are all going to be foobar. */ + li r5, 0 + mtmsrd r5, 1 + + /* Load GPRs r0-r28 */ + reg = 0 + .rept 29 + ld reg, VCPU_GPRS_TM(reg)(r31) + reg = reg + 1 + .endr + + mtspr SPRN_DSCR, r29 + mtspr SPRN_PPR, r30 + + /* Load final GPRs */ + ld 29, VCPU_GPRS_TM(29)(r31) + ld 30, VCPU_GPRS_TM(30)(r31) + ld 31, VCPU_GPRS_TM(31)(r31) + + /* TM checkpointed state is now setup. All GPRs are now volatile. */ + TRECHKPT + + /* Now let's get back the state we need. */ + HMT_MEDIUM + GET_PACA(r13) + ld r29, HSTATE_DSCR(r13) + mtspr SPRN_DSCR, r29 + ld r4, HSTATE_KVM_VCPU(r13) + ld r1, HSTATE_HOST_R1(r13) + ld r2, PACATMSCRATCH(r13) + + /* Set the MSR RI since we have our registers back. */ + li r5, MSR_RI + mtmsrd r5, 1 + + ld r0, PPC_LR_STKOFF(r1) + mtlr r0 + blr +#endif + /* * We come here if we get any exception or interrupt while we are * executing host real mode code while in guest MMU context. diff --git a/arch/powerpc/kvm/book3s_pr.c b/arch/powerpc/kvm/book3s_pr.c index c4f7d6b..e76f79a 100644 --- a/arch/powerpc/kvm/book3s_pr.c +++ b/arch/powerpc/kvm/book3s_pr.c @@ -914,7 +914,7 @@ int kvmppc_handle_exit_pr(struct kvm_run *run, struct kvm_vcpu *vcpu, /* We get here with MSR.EE=1 */ trace_kvm_exit(exit_nr, vcpu); - kvm_guest_exit(); + guest_exit(); switch (exit_nr) { case BOOK3S_INTERRUPT_INST_STORAGE: @@ -1049,7 +1049,17 @@ int kvmppc_handle_exit_pr(struct kvm_run *run, struct kvm_vcpu *vcpu, int emul; program_interrupt: - flags = vcpu->arch.shadow_srr1 & 0x1f0000ull; + /* + * shadow_srr1 only contains valid flags if we came here via + * a program exception. The other exceptions (emulation assist, + * FP unavailable, etc.) do not provide flags in SRR1, so use + * an illegal-instruction exception when injecting a program + * interrupt into the guest. + */ + if (exit_nr == BOOK3S_INTERRUPT_PROGRAM) + flags = vcpu->arch.shadow_srr1 & 0x1f0000ull; + else + flags = SRR1_PROGILL; emul = kvmppc_get_last_inst(vcpu, INST_GENERIC, &last_inst); if (emul != EMULATE_DONE) { @@ -1531,7 +1541,7 @@ static int kvmppc_vcpu_run_pr(struct kvm_run *kvm_run, struct kvm_vcpu *vcpu) kvmppc_clear_debug(vcpu); - /* No need for kvm_guest_exit. It's done in handle_exit. + /* No need for guest_exit. It's done in handle_exit. We also get here with interrupts enabled. */ /* Make sure we save the guest FPU/Altivec/VSX state */ diff --git a/arch/powerpc/kvm/booke.c b/arch/powerpc/kvm/booke.c index 4afae69..02b4672 100644 --- a/arch/powerpc/kvm/booke.c +++ b/arch/powerpc/kvm/booke.c @@ -776,7 +776,7 @@ int kvmppc_vcpu_run(struct kvm_run *kvm_run, struct kvm_vcpu *vcpu) ret = __kvmppc_vcpu_run(kvm_run, vcpu); - /* No need for kvm_guest_exit. It's done in handle_exit. + /* No need for guest_exit. It's done in handle_exit. We also get here with interrupts enabled. */ /* Switch back to user space debug context */ @@ -1012,7 +1012,7 @@ int kvmppc_handle_exit(struct kvm_run *run, struct kvm_vcpu *vcpu, } trace_kvm_exit(exit_nr, vcpu); - __kvm_guest_exit(); + guest_exit_irqoff(); local_irq_enable(); diff --git a/arch/powerpc/kvm/emulate.c b/arch/powerpc/kvm/emulate.c index 5cc2e7a..b379146 100644 --- a/arch/powerpc/kvm/emulate.c +++ b/arch/powerpc/kvm/emulate.c @@ -302,7 +302,6 @@ int kvmppc_emulate_instruction(struct kvm_run *run, struct kvm_vcpu *vcpu) advance = 0; printk(KERN_ERR "Couldn't emulate instruction 0x%08x " "(op %d xop %d)\n", inst, get_op(inst), get_xop(inst)); - kvmppc_core_queue_program(vcpu, 0); } } diff --git a/arch/powerpc/kvm/mpic.c b/arch/powerpc/kvm/mpic.c index 6249cdc..ed38f81 100644 --- a/arch/powerpc/kvm/mpic.c +++ b/arch/powerpc/kvm/mpic.c @@ -1823,7 +1823,8 @@ int kvm_set_msi(struct kvm_kernel_irq_routing_entry *e, return 0; } -int kvm_set_routing_entry(struct kvm_kernel_irq_routing_entry *e, +int kvm_set_routing_entry(struct kvm *kvm, + struct kvm_kernel_irq_routing_entry *e, const struct kvm_irq_routing_entry *ue) { int r = -EINVAL; diff --git a/arch/powerpc/kvm/powerpc.c b/arch/powerpc/kvm/powerpc.c index 02416fe..6ce40dd 100644 --- a/arch/powerpc/kvm/powerpc.c +++ b/arch/powerpc/kvm/powerpc.c @@ -119,7 +119,7 @@ int kvmppc_prepare_to_enter(struct kvm_vcpu *vcpu) continue; } - __kvm_guest_enter(); + guest_enter_irqoff(); return 1; } @@ -588,6 +588,10 @@ int kvm_vm_ioctl_check_extension(struct kvm *kvm, long ext) r = 1; break; #endif + case KVM_CAP_PPC_HTM: + r = cpu_has_feature(CPU_FTR_TM_COMP) && + is_kvmppc_hv_enabled(kvm); + break; default: r = 0; break; diff --git a/arch/powerpc/platforms/powernv/opal-wrappers.S b/arch/powerpc/platforms/powernv/opal-wrappers.S index cf928bb..3d29d40 100644 --- a/arch/powerpc/platforms/powernv/opal-wrappers.S +++ b/arch/powerpc/platforms/powernv/opal-wrappers.S @@ -64,7 +64,6 @@ END_FTR_SECTION(0, 1); \ OPAL_BRANCH(opal_tracepoint_entry) \ mfcr r12; \ stw r12,8(r1); \ - std r1,PACAR1(r13); \ li r11,0; \ mfmsr r12; \ ori r11,r11,MSR_EE; \ @@ -127,7 +126,6 @@ opal_tracepoint_entry: mfcr r12 std r11,16(r1) stw r12,8(r1) - std r1,PACAR1(r13) li r11,0 mfmsr r12 ori r11,r11,MSR_EE diff --git a/arch/s390/hypfs/hypfs_diag.c b/arch/s390/hypfs/hypfs_diag.c index 67d43a0..28f03ca 100644 --- a/arch/s390/hypfs/hypfs_diag.c +++ b/arch/s390/hypfs/hypfs_diag.c @@ -19,29 +19,10 @@ #include <asm/ebcdic.h> #include "hypfs.h" -#define LPAR_NAME_LEN 8 /* lpar name len in diag 204 data */ -#define CPU_NAME_LEN 16 /* type name len of cpus in diag224 name table */ #define TMP_SIZE 64 /* size of temporary buffers */ #define DBFS_D204_HDR_VERSION 0 -/* diag 204 subcodes */ -enum diag204_sc { - SUBC_STIB4 = 4, - SUBC_RSI = 5, - SUBC_STIB6 = 6, - SUBC_STIB7 = 7 -}; - -/* The two available diag 204 data formats */ -enum diag204_format { - INFO_SIMPLE = 0, - INFO_EXT = 0x00010000 -}; - -/* bit is set in flags, when physical cpu info is included in diag 204 data */ -#define LPAR_PHYS_FLG 0x80 - static char *diag224_cpu_names; /* diag 224 name table */ static enum diag204_sc diag204_store_sc; /* used subcode for store */ static enum diag204_format diag204_info_type; /* used diag 204 data format */ @@ -53,7 +34,7 @@ static int diag204_buf_pages; /* number of pages for diag204 data */ static struct dentry *dbfs_d204_file; /* - * DIAG 204 data structures and member access functions. + * DIAG 204 member access functions. * * Since we have two different diag 204 data formats for old and new s390 * machines, we do not access the structs directly, but use getter functions for @@ -62,304 +43,173 @@ static struct dentry *dbfs_d204_file; /* Time information block */ -struct info_blk_hdr { - __u8 npar; - __u8 flags; - __u16 tslice; - __u16 phys_cpus; - __u16 this_part; - __u64 curtod; -} __attribute__ ((packed)); - -struct x_info_blk_hdr { - __u8 npar; - __u8 flags; - __u16 tslice; - __u16 phys_cpus; - __u16 this_part; - __u64 curtod1; - __u64 curtod2; - char reserved[40]; -} __attribute__ ((packed)); - static inline int info_blk_hdr__size(enum diag204_format type) { - if (type == INFO_SIMPLE) - return sizeof(struct info_blk_hdr); - else /* INFO_EXT */ - return sizeof(struct x_info_blk_hdr); + if (type == DIAG204_INFO_SIMPLE) + return sizeof(struct diag204_info_blk_hdr); + else /* DIAG204_INFO_EXT */ + return sizeof(struct diag204_x_info_blk_hdr); } static inline __u8 info_blk_hdr__npar(enum diag204_format type, void *hdr) { - if (type == INFO_SIMPLE) - return ((struct info_blk_hdr *)hdr)->npar; - else /* INFO_EXT */ - return ((struct x_info_blk_hdr *)hdr)->npar; + if (type == DIAG204_INFO_SIMPLE) + return ((struct diag204_info_blk_hdr *)hdr)->npar; + else /* DIAG204_INFO_EXT */ + return ((struct diag204_x_info_blk_hdr *)hdr)->npar; } static inline __u8 info_blk_hdr__flags(enum diag204_format type, void *hdr) { - if (type == INFO_SIMPLE) - return ((struct info_blk_hdr *)hdr)->flags; - else /* INFO_EXT */ - return ((struct x_info_blk_hdr *)hdr)->flags; + if (type == DIAG204_INFO_SIMPLE) + return ((struct diag204_info_blk_hdr *)hdr)->flags; + else /* DIAG204_INFO_EXT */ + return ((struct diag204_x_info_blk_hdr *)hdr)->flags; } static inline __u16 info_blk_hdr__pcpus(enum diag204_format type, void *hdr) { - if (type == INFO_SIMPLE) - return ((struct info_blk_hdr *)hdr)->phys_cpus; - else /* INFO_EXT */ - return ((struct x_info_blk_hdr *)hdr)->phys_cpus; + if (type == DIAG204_INFO_SIMPLE) + return ((struct diag204_info_blk_hdr *)hdr)->phys_cpus; + else /* DIAG204_INFO_EXT */ + return ((struct diag204_x_info_blk_hdr *)hdr)->phys_cpus; } /* Partition header */ -struct part_hdr { - __u8 pn; - __u8 cpus; - char reserved[6]; - char part_name[LPAR_NAME_LEN]; -} __attribute__ ((packed)); - -struct x_part_hdr { - __u8 pn; - __u8 cpus; - __u8 rcpus; - __u8 pflag; - __u32 mlu; - char part_name[LPAR_NAME_LEN]; - char lpc_name[8]; - char os_name[8]; - __u64 online_cs; - __u64 online_es; - __u8 upid; - char reserved1[3]; - __u32 group_mlu; - char group_name[8]; - char reserved2[32]; -} __attribute__ ((packed)); - static inline int part_hdr__size(enum diag204_format type) { - if (type == INFO_SIMPLE) - return sizeof(struct part_hdr); - else /* INFO_EXT */ - return sizeof(struct x_part_hdr); + if (type == DIAG204_INFO_SIMPLE) + return sizeof(struct diag204_part_hdr); + else /* DIAG204_INFO_EXT */ + return sizeof(struct diag204_x_part_hdr); } static inline __u8 part_hdr__rcpus(enum diag204_format type, void *hdr) { - if (type == INFO_SIMPLE) - return ((struct part_hdr *)hdr)->cpus; - else /* INFO_EXT */ - return ((struct x_part_hdr *)hdr)->rcpus; + if (type == DIAG204_INFO_SIMPLE) + return ((struct diag204_part_hdr *)hdr)->cpus; + else /* DIAG204_INFO_EXT */ + return ((struct diag204_x_part_hdr *)hdr)->rcpus; } static inline void part_hdr__part_name(enum diag204_format type, void *hdr, char *name) { - if (type == INFO_SIMPLE) - memcpy(name, ((struct part_hdr *)hdr)->part_name, - LPAR_NAME_LEN); - else /* INFO_EXT */ - memcpy(name, ((struct x_part_hdr *)hdr)->part_name, - LPAR_NAME_LEN); - EBCASC(name, LPAR_NAME_LEN); - name[LPAR_NAME_LEN] = 0; + if (type == DIAG204_INFO_SIMPLE) + memcpy(name, ((struct diag204_part_hdr *)hdr)->part_name, + DIAG204_LPAR_NAME_LEN); + else /* DIAG204_INFO_EXT */ + memcpy(name, ((struct diag204_x_part_hdr *)hdr)->part_name, + DIAG204_LPAR_NAME_LEN); + EBCASC(name, DIAG204_LPAR_NAME_LEN); + name[DIAG204_LPAR_NAME_LEN] = 0; strim(name); } -struct cpu_info { - __u16 cpu_addr; - char reserved1[2]; - __u8 ctidx; - __u8 cflag; - __u16 weight; - __u64 acc_time; - __u64 lp_time; -} __attribute__ ((packed)); - -struct x_cpu_info { - __u16 cpu_addr; - char reserved1[2]; - __u8 ctidx; - __u8 cflag; - __u16 weight; - __u64 acc_time; - __u64 lp_time; - __u16 min_weight; - __u16 cur_weight; - __u16 max_weight; - char reseved2[2]; - __u64 online_time; - __u64 wait_time; - __u32 pma_weight; - __u32 polar_weight; - char reserved3[40]; -} __attribute__ ((packed)); - /* CPU info block */ static inline int cpu_info__size(enum diag204_format type) { - if (type == INFO_SIMPLE) - return sizeof(struct cpu_info); - else /* INFO_EXT */ - return sizeof(struct x_cpu_info); + if (type == DIAG204_INFO_SIMPLE) + return sizeof(struct diag204_cpu_info); + else /* DIAG204_INFO_EXT */ + return sizeof(struct diag204_x_cpu_info); } static inline __u8 cpu_info__ctidx(enum diag204_format type, void *hdr) { - if (type == INFO_SIMPLE) - return ((struct cpu_info *)hdr)->ctidx; - else /* INFO_EXT */ - return ((struct x_cpu_info *)hdr)->ctidx; + if (type == DIAG204_INFO_SIMPLE) + return ((struct diag204_cpu_info *)hdr)->ctidx; + else /* DIAG204_INFO_EXT */ + return ((struct diag204_x_cpu_info *)hdr)->ctidx; } static inline __u16 cpu_info__cpu_addr(enum diag204_format type, void *hdr) { - if (type == INFO_SIMPLE) - return ((struct cpu_info *)hdr)->cpu_addr; - else /* INFO_EXT */ - return ((struct x_cpu_info *)hdr)->cpu_addr; + if (type == DIAG204_INFO_SIMPLE) + return ((struct diag204_cpu_info *)hdr)->cpu_addr; + else /* DIAG204_INFO_EXT */ + return ((struct diag204_x_cpu_info *)hdr)->cpu_addr; } static inline __u64 cpu_info__acc_time(enum diag204_format type, void *hdr) { - if (type == INFO_SIMPLE) - return ((struct cpu_info *)hdr)->acc_time; - else /* INFO_EXT */ - return ((struct x_cpu_info *)hdr)->acc_time; + if (type == DIAG204_INFO_SIMPLE) + return ((struct diag204_cpu_info *)hdr)->acc_time; + else /* DIAG204_INFO_EXT */ + return ((struct diag204_x_cpu_info *)hdr)->acc_time; } static inline __u64 cpu_info__lp_time(enum diag204_format type, void *hdr) { - if (type == INFO_SIMPLE) - return ((struct cpu_info *)hdr)->lp_time; - else /* INFO_EXT */ - return ((struct x_cpu_info *)hdr)->lp_time; + if (type == DIAG204_INFO_SIMPLE) + return ((struct diag204_cpu_info *)hdr)->lp_time; + else /* DIAG204_INFO_EXT */ + return ((struct diag204_x_cpu_info *)hdr)->lp_time; } static inline __u64 cpu_info__online_time(enum diag204_format type, void *hdr) { - if (type == INFO_SIMPLE) + if (type == DIAG204_INFO_SIMPLE) return 0; /* online_time not available in simple info */ - else /* INFO_EXT */ - return ((struct x_cpu_info *)hdr)->online_time; + else /* DIAG204_INFO_EXT */ + return ((struct diag204_x_cpu_info *)hdr)->online_time; } /* Physical header */ -struct phys_hdr { - char reserved1[1]; - __u8 cpus; - char reserved2[6]; - char mgm_name[8]; -} __attribute__ ((packed)); - -struct x_phys_hdr { - char reserved1[1]; - __u8 cpus; - char reserved2[6]; - char mgm_name[8]; - char reserved3[80]; -} __attribute__ ((packed)); - static inline int phys_hdr__size(enum diag204_format type) { - if (type == INFO_SIMPLE) - return sizeof(struct phys_hdr); - else /* INFO_EXT */ - return sizeof(struct x_phys_hdr); + if (type == DIAG204_INFO_SIMPLE) + return sizeof(struct diag204_phys_hdr); + else /* DIAG204_INFO_EXT */ + return sizeof(struct diag204_x_phys_hdr); } static inline __u8 phys_hdr__cpus(enum diag204_format type, void *hdr) { - if (type == INFO_SIMPLE) - return ((struct phys_hdr *)hdr)->cpus; - else /* INFO_EXT */ - return ((struct x_phys_hdr *)hdr)->cpus; + if (type == DIAG204_INFO_SIMPLE) + return ((struct diag204_phys_hdr *)hdr)->cpus; + else /* DIAG204_INFO_EXT */ + return ((struct diag204_x_phys_hdr *)hdr)->cpus; } /* Physical CPU info block */ -struct phys_cpu { - __u16 cpu_addr; - char reserved1[2]; - __u8 ctidx; - char reserved2[3]; - __u64 mgm_time; - char reserved3[8]; -} __attribute__ ((packed)); - -struct x_phys_cpu { - __u16 cpu_addr; - char reserved1[2]; - __u8 ctidx; - char reserved2[3]; - __u64 mgm_time; - char reserved3[80]; -} __attribute__ ((packed)); - static inline int phys_cpu__size(enum diag204_format type) { - if (type == INFO_SIMPLE) - return sizeof(struct phys_cpu); - else /* INFO_EXT */ - return sizeof(struct x_phys_cpu); + if (type == DIAG204_INFO_SIMPLE) + return sizeof(struct diag204_phys_cpu); + else /* DIAG204_INFO_EXT */ + return sizeof(struct diag204_x_phys_cpu); } static inline __u16 phys_cpu__cpu_addr(enum diag204_format type, void *hdr) { - if (type == INFO_SIMPLE) - return ((struct phys_cpu *)hdr)->cpu_addr; - else /* INFO_EXT */ - return ((struct x_phys_cpu *)hdr)->cpu_addr; + if (type == DIAG204_INFO_SIMPLE) + return ((struct diag204_phys_cpu *)hdr)->cpu_addr; + else /* DIAG204_INFO_EXT */ + return ((struct diag204_x_phys_cpu *)hdr)->cpu_addr; } static inline __u64 phys_cpu__mgm_time(enum diag204_format type, void *hdr) { - if (type == INFO_SIMPLE) - return ((struct phys_cpu *)hdr)->mgm_time; - else /* INFO_EXT */ - return ((struct x_phys_cpu *)hdr)->mgm_time; + if (type == DIAG204_INFO_SIMPLE) + return ((struct diag204_phys_cpu *)hdr)->mgm_time; + else /* DIAG204_INFO_EXT */ + return ((struct diag204_x_phys_cpu *)hdr)->mgm_time; } static inline __u64 phys_cpu__ctidx(enum diag204_format type, void *hdr) { - if (type == INFO_SIMPLE) - return ((struct phys_cpu *)hdr)->ctidx; - else /* INFO_EXT */ - return ((struct x_phys_cpu *)hdr)->ctidx; + if (type == DIAG204_INFO_SIMPLE) + return ((struct diag204_phys_cpu *)hdr)->ctidx; + else /* DIAG204_INFO_EXT */ + return ((struct diag204_x_phys_cpu *)hdr)->ctidx; } /* Diagnose 204 functions */ - -static inline int __diag204(unsigned long *subcode, unsigned long size, void *addr) -{ - register unsigned long _subcode asm("0") = *subcode; - register unsigned long _size asm("1") = size; - - asm volatile( - " diag %2,%0,0x204\n" - "0: nopr %%r7\n" - EX_TABLE(0b,0b) - : "+d" (_subcode), "+d" (_size) : "d" (addr) : "memory"); - *subcode = _subcode; - return _size; -} - -static int diag204(unsigned long subcode, unsigned long size, void *addr) -{ - diag_stat_inc(DIAG_STAT_X204); - size = __diag204(&subcode, size, addr); - if (subcode) - return -1; - return size; -} - /* * For the old diag subcode 4 with simple data format we have to use real * memory. If we use subcode 6 or 7 with extended data format, we can (and @@ -411,12 +261,12 @@ static void *diag204_get_buffer(enum diag204_format fmt, int *pages) *pages = diag204_buf_pages; return diag204_buf; } - if (fmt == INFO_SIMPLE) { + if (fmt == DIAG204_INFO_SIMPLE) { *pages = 1; return diag204_alloc_rbuf(); - } else {/* INFO_EXT */ - *pages = diag204((unsigned long)SUBC_RSI | - (unsigned long)INFO_EXT, 0, NULL); + } else {/* DIAG204_INFO_EXT */ + *pages = diag204((unsigned long)DIAG204_SUBC_RSI | + (unsigned long)DIAG204_INFO_EXT, 0, NULL); if (*pages <= 0) return ERR_PTR(-ENOSYS); else @@ -443,18 +293,18 @@ static int diag204_probe(void) void *buf; int pages, rc; - buf = diag204_get_buffer(INFO_EXT, &pages); + buf = diag204_get_buffer(DIAG204_INFO_EXT, &pages); if (!IS_ERR(buf)) { - if (diag204((unsigned long)SUBC_STIB7 | - (unsigned long)INFO_EXT, pages, buf) >= 0) { - diag204_store_sc = SUBC_STIB7; - diag204_info_type = INFO_EXT; + if (diag204((unsigned long)DIAG204_SUBC_STIB7 | + (unsigned long)DIAG204_INFO_EXT, pages, buf) >= 0) { + diag204_store_sc = DIAG204_SUBC_STIB7; + diag204_info_type = DIAG204_INFO_EXT; goto out; } - if (diag204((unsigned long)SUBC_STIB6 | - (unsigned long)INFO_EXT, pages, buf) >= 0) { - diag204_store_sc = SUBC_STIB6; - diag204_info_type = INFO_EXT; + if (diag204((unsigned long)DIAG204_SUBC_STIB6 | + (unsigned long)DIAG204_INFO_EXT, pages, buf) >= 0) { + diag204_store_sc = DIAG204_SUBC_STIB6; + diag204_info_type = DIAG204_INFO_EXT; goto out; } diag204_free_buffer(); @@ -462,15 +312,15 @@ static int diag204_probe(void) /* subcodes 6 and 7 failed, now try subcode 4 */ - buf = diag204_get_buffer(INFO_SIMPLE, &pages); + buf = diag204_get_buffer(DIAG204_INFO_SIMPLE, &pages); if (IS_ERR(buf)) { rc = PTR_ERR(buf); goto fail_alloc; } - if (diag204((unsigned long)SUBC_STIB4 | - (unsigned long)INFO_SIMPLE, pages, buf) >= 0) { - diag204_store_sc = SUBC_STIB4; - diag204_info_type = INFO_SIMPLE; + if (diag204((unsigned long)DIAG204_SUBC_STIB4 | + (unsigned long)DIAG204_INFO_SIMPLE, pages, buf) >= 0) { + diag204_store_sc = DIAG204_SUBC_STIB4; + diag204_info_type = DIAG204_INFO_SIMPLE; goto out; } else { rc = -ENOSYS; @@ -510,20 +360,6 @@ out: /* Diagnose 224 functions */ -static int diag224(void *ptr) -{ - int rc = -EOPNOTSUPP; - - diag_stat_inc(DIAG_STAT_X224); - asm volatile( - " diag %1,%2,0x224\n" - "0: lhi %0,0x0\n" - "1:\n" - EX_TABLE(0b,1b) - : "+d" (rc) :"d" (0), "d" (ptr) : "memory"); - return rc; -} - static int diag224_get_name_table(void) { /* memory must be below 2GB */ @@ -545,9 +381,9 @@ static void diag224_delete_name_table(void) static int diag224_idx2name(int index, char *name) { - memcpy(name, diag224_cpu_names + ((index + 1) * CPU_NAME_LEN), - CPU_NAME_LEN); - name[CPU_NAME_LEN] = 0; + memcpy(name, diag224_cpu_names + ((index + 1) * DIAG204_CPU_NAME_LEN), + DIAG204_CPU_NAME_LEN); + name[DIAG204_CPU_NAME_LEN] = 0; strim(name); return 0; } @@ -603,7 +439,7 @@ __init int hypfs_diag_init(void) pr_err("The hardware system does not support hypfs\n"); return -ENODATA; } - if (diag204_info_type == INFO_EXT) { + if (diag204_info_type == DIAG204_INFO_EXT) { rc = hypfs_dbfs_create_file(&dbfs_file_d204); if (rc) return rc; @@ -651,7 +487,7 @@ static int hypfs_create_cpu_files(struct dentry *cpus_dir, void *cpu_info) cpu_info__lp_time(diag204_info_type, cpu_info)); if (IS_ERR(rc)) return PTR_ERR(rc); - if (diag204_info_type == INFO_EXT) { + if (diag204_info_type == DIAG204_INFO_EXT) { rc = hypfs_create_u64(cpu_dir, "onlinetime", cpu_info__online_time(diag204_info_type, cpu_info)); @@ -667,12 +503,12 @@ static void *hypfs_create_lpar_files(struct dentry *systems_dir, void *part_hdr) { struct dentry *cpus_dir; struct dentry *lpar_dir; - char lpar_name[LPAR_NAME_LEN + 1]; + char lpar_name[DIAG204_LPAR_NAME_LEN + 1]; void *cpu_info; int i; part_hdr__part_name(diag204_info_type, part_hdr, lpar_name); - lpar_name[LPAR_NAME_LEN] = 0; + lpar_name[DIAG204_LPAR_NAME_LEN] = 0; lpar_dir = hypfs_mkdir(systems_dir, lpar_name); if (IS_ERR(lpar_dir)) return lpar_dir; @@ -755,7 +591,8 @@ int hypfs_diag_create_files(struct dentry *root) goto err_out; } } - if (info_blk_hdr__flags(diag204_info_type, time_hdr) & LPAR_PHYS_FLG) { + if (info_blk_hdr__flags(diag204_info_type, time_hdr) & + DIAG204_LPAR_PHYS_FLG) { ptr = hypfs_create_phys_files(root, part_hdr); if (IS_ERR(ptr)) { rc = PTR_ERR(ptr); diff --git a/arch/s390/include/asm/cpacf.h b/arch/s390/include/asm/cpacf.h index 1a82cf2..d28621d 100644 --- a/arch/s390/include/asm/cpacf.h +++ b/arch/s390/include/asm/cpacf.h @@ -20,6 +20,9 @@ #define CPACF_KMC 0xb92f /* MSA */ #define CPACF_KIMD 0xb93e /* MSA */ #define CPACF_KLMD 0xb93f /* MSA */ +#define CPACF_PCKMO 0xb928 /* MSA3 */ +#define CPACF_KMF 0xb92a /* MSA4 */ +#define CPACF_KMO 0xb92b /* MSA4 */ #define CPACF_PCC 0xb92c /* MSA4 */ #define CPACF_KMCTR 0xb92d /* MSA4 */ #define CPACF_PPNO 0xb93c /* MSA5 */ @@ -136,6 +139,7 @@ static inline void __cpacf_query(unsigned int opcode, unsigned char *status) register unsigned long r1 asm("1") = (unsigned long) status; asm volatile( + " spm 0\n" /* pckmo doesn't change the cc */ /* Parameter registers are ignored, but may not be 0 */ "0: .insn rrf,%[opc] << 16,2,2,2,0\n" " brc 1,0b\n" /* handle partial completion */ @@ -157,6 +161,12 @@ static inline int cpacf_query(unsigned int opcode, unsigned int func) if (!test_facility(17)) /* check for MSA */ return 0; break; + case CPACF_PCKMO: + if (!test_facility(76)) /* check for MSA3 */ + return 0; + break; + case CPACF_KMF: + case CPACF_KMO: case CPACF_PCC: case CPACF_KMCTR: if (!test_facility(77)) /* check for MSA4 */ diff --git a/arch/s390/include/asm/diag.h b/arch/s390/include/asm/diag.h index 86cae09..8acf482 100644 --- a/arch/s390/include/asm/diag.h +++ b/arch/s390/include/asm/diag.h @@ -78,4 +78,153 @@ struct diag210 { extern int diag210(struct diag210 *addr); +/* bit is set in flags, when physical cpu info is included in diag 204 data */ +#define DIAG204_LPAR_PHYS_FLG 0x80 +#define DIAG204_LPAR_NAME_LEN 8 /* lpar name len in diag 204 data */ +#define DIAG204_CPU_NAME_LEN 16 /* type name len of cpus in diag224 name table */ + +/* diag 204 subcodes */ +enum diag204_sc { + DIAG204_SUBC_STIB4 = 4, + DIAG204_SUBC_RSI = 5, + DIAG204_SUBC_STIB6 = 6, + DIAG204_SUBC_STIB7 = 7 +}; + +/* The two available diag 204 data formats */ +enum diag204_format { + DIAG204_INFO_SIMPLE = 0, + DIAG204_INFO_EXT = 0x00010000 +}; + +enum diag204_cpu_flags { + DIAG204_CPU_ONLINE = 0x20, + DIAG204_CPU_CAPPED = 0x40, +}; + +struct diag204_info_blk_hdr { + __u8 npar; + __u8 flags; + __u16 tslice; + __u16 phys_cpus; + __u16 this_part; + __u64 curtod; +} __packed; + +struct diag204_x_info_blk_hdr { + __u8 npar; + __u8 flags; + __u16 tslice; + __u16 phys_cpus; + __u16 this_part; + __u64 curtod1; + __u64 curtod2; + char reserved[40]; +} __packed; + +struct diag204_part_hdr { + __u8 pn; + __u8 cpus; + char reserved[6]; + char part_name[DIAG204_LPAR_NAME_LEN]; +} __packed; + +struct diag204_x_part_hdr { + __u8 pn; + __u8 cpus; + __u8 rcpus; + __u8 pflag; + __u32 mlu; + char part_name[DIAG204_LPAR_NAME_LEN]; + char lpc_name[8]; + char os_name[8]; + __u64 online_cs; + __u64 online_es; + __u8 upid; + __u8 reserved:3; + __u8 mtid:5; + char reserved1[2]; + __u32 group_mlu; + char group_name[8]; + char hardware_group_name[8]; + char reserved2[24]; +} __packed; + +struct diag204_cpu_info { + __u16 cpu_addr; + char reserved1[2]; + __u8 ctidx; + __u8 cflag; + __u16 weight; + __u64 acc_time; + __u64 lp_time; +} __packed; + +struct diag204_x_cpu_info { + __u16 cpu_addr; + char reserved1[2]; + __u8 ctidx; + __u8 cflag; + __u16 weight; + __u64 acc_time; + __u64 lp_time; + __u16 min_weight; + __u16 cur_weight; + __u16 max_weight; + char reseved2[2]; + __u64 online_time; + __u64 wait_time; + __u32 pma_weight; + __u32 polar_weight; + __u32 cpu_type_cap; + __u32 group_cpu_type_cap; + char reserved3[32]; +} __packed; + +struct diag204_phys_hdr { + char reserved1[1]; + __u8 cpus; + char reserved2[6]; + char mgm_name[8]; +} __packed; + +struct diag204_x_phys_hdr { + char reserved1[1]; + __u8 cpus; + char reserved2[6]; + char mgm_name[8]; + char reserved3[80]; +} __packed; + +struct diag204_phys_cpu { + __u16 cpu_addr; + char reserved1[2]; + __u8 ctidx; + char reserved2[3]; + __u64 mgm_time; + char reserved3[8]; +} __packed; + +struct diag204_x_phys_cpu { + __u16 cpu_addr; + char reserved1[2]; + __u8 ctidx; + char reserved2[1]; + __u16 weight; + __u64 mgm_time; + char reserved3[80]; +} __packed; + +struct diag204_x_part_block { + struct diag204_x_part_hdr hdr; + struct diag204_x_cpu_info cpus[]; +} __packed; + +struct diag204_x_phys_block { + struct diag204_x_phys_hdr hdr; + struct diag204_x_phys_cpu cpus[]; +} __packed; + +int diag204(unsigned long subcode, unsigned long size, void *addr); +int diag224(void *ptr); #endif /* _ASM_S390_DIAG_H */ diff --git a/arch/s390/include/asm/gmap.h b/arch/s390/include/asm/gmap.h index d054c1b..741ddba 100644 --- a/arch/s390/include/asm/gmap.h +++ b/arch/s390/include/asm/gmap.h @@ -10,14 +10,25 @@ /** * struct gmap_struct - guest address space + * @list: list head for the mm->context gmap list * @crst_list: list of all crst tables used in the guest address space * @mm: pointer to the parent mm_struct * @guest_to_host: radix tree with guest to host address translation * @host_to_guest: radix tree with pointer to segment table entries * @guest_table_lock: spinlock to protect all entries in the guest page table + * @ref_count: reference counter for the gmap structure * @table: pointer to the page directory * @asce: address space control element for gmap page table * @pfault_enabled: defines if pfaults are applicable for the guest + * @host_to_rmap: radix tree with gmap_rmap lists + * @children: list of shadow gmap structures + * @pt_list: list of all page tables used in the shadow guest address space + * @shadow_lock: spinlock to protect the shadow gmap list + * @parent: pointer to the parent gmap for shadow guest address spaces + * @orig_asce: ASCE for which the shadow page table has been created + * @edat_level: edat level to be used for the shadow translation + * @removed: flag to indicate if a shadow guest address space has been removed + * @initialized: flag to indicate if a shadow guest address space can be used */ struct gmap { struct list_head list; @@ -26,26 +37,64 @@ struct gmap { struct radix_tree_root guest_to_host; struct radix_tree_root host_to_guest; spinlock_t guest_table_lock; + atomic_t ref_count; unsigned long *table; unsigned long asce; unsigned long asce_end; void *private; bool pfault_enabled; + /* Additional data for shadow guest address spaces */ + struct radix_tree_root host_to_rmap; + struct list_head children; + struct list_head pt_list; + spinlock_t shadow_lock; + struct gmap *parent; + unsigned long orig_asce; + int edat_level; + bool removed; + bool initialized; }; /** + * struct gmap_rmap - reverse mapping for shadow page table entries + * @next: pointer to next rmap in the list + * @raddr: virtual rmap address in the shadow guest address space + */ +struct gmap_rmap { + struct gmap_rmap *next; + unsigned long raddr; +}; + +#define gmap_for_each_rmap(pos, head) \ + for (pos = (head); pos; pos = pos->next) + +#define gmap_for_each_rmap_safe(pos, n, head) \ + for (pos = (head); n = pos ? pos->next : NULL, pos; pos = n) + +/** * struct gmap_notifier - notify function block for page invalidation * @notifier_call: address of callback function */ struct gmap_notifier { struct list_head list; - void (*notifier_call)(struct gmap *gmap, unsigned long gaddr); + struct rcu_head rcu; + void (*notifier_call)(struct gmap *gmap, unsigned long start, + unsigned long end); }; -struct gmap *gmap_alloc(struct mm_struct *mm, unsigned long limit); -void gmap_free(struct gmap *gmap); +static inline int gmap_is_shadow(struct gmap *gmap) +{ + return !!gmap->parent; +} + +struct gmap *gmap_create(struct mm_struct *mm, unsigned long limit); +void gmap_remove(struct gmap *gmap); +struct gmap *gmap_get(struct gmap *gmap); +void gmap_put(struct gmap *gmap); + void gmap_enable(struct gmap *gmap); void gmap_disable(struct gmap *gmap); +struct gmap *gmap_get_enabled(void); int gmap_map_segment(struct gmap *gmap, unsigned long from, unsigned long to, unsigned long len); int gmap_unmap_segment(struct gmap *gmap, unsigned long to, unsigned long len); @@ -57,8 +106,29 @@ void gmap_discard(struct gmap *, unsigned long from, unsigned long to); void __gmap_zap(struct gmap *, unsigned long gaddr); void gmap_unlink(struct mm_struct *, unsigned long *table, unsigned long vmaddr); -void gmap_register_ipte_notifier(struct gmap_notifier *); -void gmap_unregister_ipte_notifier(struct gmap_notifier *); -int gmap_ipte_notify(struct gmap *, unsigned long start, unsigned long len); +int gmap_read_table(struct gmap *gmap, unsigned long gaddr, unsigned long *val); + +struct gmap *gmap_shadow(struct gmap *parent, unsigned long asce, + int edat_level); +int gmap_shadow_valid(struct gmap *sg, unsigned long asce, int edat_level); +int gmap_shadow_r2t(struct gmap *sg, unsigned long saddr, unsigned long r2t, + int fake); +int gmap_shadow_r3t(struct gmap *sg, unsigned long saddr, unsigned long r3t, + int fake); +int gmap_shadow_sgt(struct gmap *sg, unsigned long saddr, unsigned long sgt, + int fake); +int gmap_shadow_pgt(struct gmap *sg, unsigned long saddr, unsigned long pgt, + int fake); +int gmap_shadow_pgt_lookup(struct gmap *sg, unsigned long saddr, + unsigned long *pgt, int *dat_protection, int *fake); +int gmap_shadow_page(struct gmap *sg, unsigned long saddr, pte_t pte); + +void gmap_register_pte_notifier(struct gmap_notifier *); +void gmap_unregister_pte_notifier(struct gmap_notifier *); +void gmap_pte_notify(struct mm_struct *, unsigned long addr, pte_t *, + unsigned long bits); + +int gmap_mprotect_notify(struct gmap *, unsigned long start, + unsigned long len, int prot); #endif /* _ASM_S390_GMAP_H */ diff --git a/arch/s390/include/asm/kvm_host.h b/arch/s390/include/asm/kvm_host.h index ac82e8e..8e5daf7 100644 --- a/arch/s390/include/asm/kvm_host.h +++ b/arch/s390/include/asm/kvm_host.h @@ -43,6 +43,7 @@ /* s390-specific vcpu->requests bit members */ #define KVM_REQ_ENABLE_IBS 8 #define KVM_REQ_DISABLE_IBS 9 +#define KVM_REQ_ICPT_OPEREXC 10 #define SIGP_CTRL_C 0x80 #define SIGP_CTRL_SCN_MASK 0x3f @@ -145,7 +146,7 @@ struct kvm_s390_sie_block { __u64 cputm; /* 0x0028 */ __u64 ckc; /* 0x0030 */ __u64 epoch; /* 0x0038 */ - __u8 reserved40[4]; /* 0x0040 */ + __u32 svcc; /* 0x0040 */ #define LCTL_CR0 0x8000 #define LCTL_CR6 0x0200 #define LCTL_CR9 0x0040 @@ -154,6 +155,7 @@ struct kvm_s390_sie_block { #define LCTL_CR14 0x0002 __u16 lctl; /* 0x0044 */ __s16 icpua; /* 0x0046 */ +#define ICTL_OPEREXC 0x80000000 #define ICTL_PINT 0x20000000 #define ICTL_LPSW 0x00400000 #define ICTL_STCTL 0x00040000 @@ -166,6 +168,9 @@ struct kvm_s390_sie_block { #define ICPT_INST 0x04 #define ICPT_PROGI 0x08 #define ICPT_INSTPROGI 0x0C +#define ICPT_EXTINT 0x14 +#define ICPT_VALIDITY 0x20 +#define ICPT_STOP 0x28 #define ICPT_OPEREXC 0x2C #define ICPT_PARTEXEC 0x38 #define ICPT_IOINST 0x40 @@ -185,7 +190,9 @@ struct kvm_s390_sie_block { __u32 scaol; /* 0x0064 */ __u8 reserved68[4]; /* 0x0068 */ __u32 todpr; /* 0x006c */ - __u8 reserved70[32]; /* 0x0070 */ + __u8 reserved70[16]; /* 0x0070 */ + __u64 mso; /* 0x0080 */ + __u64 msl; /* 0x0088 */ psw_t gpsw; /* 0x0090 */ __u64 gg14; /* 0x00a0 */ __u64 gg15; /* 0x00a8 */ @@ -223,7 +230,7 @@ struct kvm_s390_sie_block { __u8 reserved1e6[2]; /* 0x01e6 */ __u64 itdba; /* 0x01e8 */ __u64 riccbd; /* 0x01f0 */ - __u8 reserved1f8[8]; /* 0x01f8 */ + __u64 gvrd; /* 0x01f8 */ } __attribute__((packed)); struct kvm_s390_itdb { @@ -256,6 +263,7 @@ struct kvm_vcpu_stat { u32 instruction_stctg; u32 exit_program_interruption; u32 exit_instr_and_program; + u32 exit_operation_exception; u32 deliver_external_call; u32 deliver_emergency_signal; u32 deliver_service_signal; @@ -278,7 +286,9 @@ struct kvm_vcpu_stat { u32 instruction_stsi; u32 instruction_stfl; u32 instruction_tprot; + u32 instruction_sie; u32 instruction_essa; + u32 instruction_sthyi; u32 instruction_sigp_sense; u32 instruction_sigp_sense_running; u32 instruction_sigp_external_call; @@ -541,12 +551,16 @@ struct kvm_guestdbg_info_arch { struct kvm_vcpu_arch { struct kvm_s390_sie_block *sie_block; + /* if vsie is active, currently executed shadow sie control block */ + struct kvm_s390_sie_block *vsie_block; unsigned int host_acrs[NUM_ACRS]; struct fpu host_fpregs; struct kvm_s390_local_interrupt local_int; struct hrtimer ckc_timer; struct kvm_s390_pgm_info pgm; struct gmap *gmap; + /* backup location for the currently enabled gmap when scheduled out */ + struct gmap *enabled_gmap; struct kvm_guestdbg_info_arch guestdbg; unsigned long pfault_token; unsigned long pfault_select; @@ -631,6 +645,14 @@ struct sie_page2 { u8 reserved900[0x1000 - 0x900]; /* 0x0900 */ } __packed; +struct kvm_s390_vsie { + struct mutex mutex; + struct radix_tree_root addr_to_page; + int page_count; + int next; + struct page *pages[KVM_MAX_VCPUS]; +}; + struct kvm_arch{ void *sca; int use_esca; @@ -646,15 +668,20 @@ struct kvm_arch{ int user_cpu_state_ctrl; int user_sigp; int user_stsi; + int user_instr0; struct s390_io_adapter *adapters[MAX_S390_IO_ADAPTERS]; wait_queue_head_t ipte_wq; int ipte_lock_count; struct mutex ipte_mutex; + struct ratelimit_state sthyi_limit; spinlock_t start_stop_lock; struct sie_page2 *sie_page2; struct kvm_s390_cpu_model model; struct kvm_s390_crypto crypto; + struct kvm_s390_vsie vsie; u64 epoch; + /* subset of available cpu features enabled by user space */ + DECLARE_BITMAP(cpu_feat, KVM_S390_VM_CPU_FEAT_NR_BITS); }; #define KVM_HVA_ERR_BAD (-1UL) diff --git a/arch/s390/include/asm/mmu.h b/arch/s390/include/asm/mmu.h index 1822643..6d39329 100644 --- a/arch/s390/include/asm/mmu.h +++ b/arch/s390/include/asm/mmu.h @@ -8,8 +8,9 @@ typedef struct { cpumask_t cpu_attach_mask; atomic_t flush_count; unsigned int flush_mm; - spinlock_t list_lock; + spinlock_t pgtable_lock; struct list_head pgtable_list; + spinlock_t gmap_lock; struct list_head gmap_list; unsigned long asce; unsigned long asce_limit; @@ -22,9 +23,11 @@ typedef struct { unsigned int use_skey:1; } mm_context_t; -#define INIT_MM_CONTEXT(name) \ - .context.list_lock = __SPIN_LOCK_UNLOCKED(name.context.list_lock), \ - .context.pgtable_list = LIST_HEAD_INIT(name.context.pgtable_list), \ +#define INIT_MM_CONTEXT(name) \ + .context.pgtable_lock = \ + __SPIN_LOCK_UNLOCKED(name.context.pgtable_lock), \ + .context.pgtable_list = LIST_HEAD_INIT(name.context.pgtable_list), \ + .context.gmap_lock = __SPIN_LOCK_UNLOCKED(name.context.gmap_lock), \ .context.gmap_list = LIST_HEAD_INIT(name.context.gmap_list), static inline int tprot(unsigned long addr) diff --git a/arch/s390/include/asm/mmu_context.h b/arch/s390/include/asm/mmu_context.h index f77c638..c6a088c 100644 --- a/arch/s390/include/asm/mmu_context.h +++ b/arch/s390/include/asm/mmu_context.h @@ -15,8 +15,9 @@ static inline int init_new_context(struct task_struct *tsk, struct mm_struct *mm) { - spin_lock_init(&mm->context.list_lock); + spin_lock_init(&mm->context.pgtable_lock); INIT_LIST_HEAD(&mm->context.pgtable_list); + spin_lock_init(&mm->context.gmap_lock); INIT_LIST_HEAD(&mm->context.gmap_list); cpumask_clear(&mm->context.cpu_attach_mask); atomic_set(&mm->context.flush_count, 0); diff --git a/arch/s390/include/asm/page.h b/arch/s390/include/asm/page.h index b2146c4..69b8a41 100644 --- a/arch/s390/include/asm/page.h +++ b/arch/s390/include/asm/page.h @@ -111,13 +111,14 @@ static inline unsigned char page_get_storage_key(unsigned long addr) static inline int page_reset_referenced(unsigned long addr) { - unsigned int ipm; + int cc; asm volatile( " rrbe 0,%1\n" " ipm %0\n" - : "=d" (ipm) : "a" (addr) : "cc"); - return !!(ipm & 0x20000000); + " srl %0,28\n" + : "=d" (cc) : "a" (addr) : "cc"); + return cc; } /* Bits int the storage key */ @@ -148,6 +149,8 @@ static inline int devmem_is_allowed(unsigned long pfn) #define virt_to_page(kaddr) pfn_to_page(__pa(kaddr) >> PAGE_SHIFT) #define page_to_phys(page) (page_to_pfn(page) << PAGE_SHIFT) #define virt_addr_valid(kaddr) pfn_valid(__pa(kaddr) >> PAGE_SHIFT) +#define pfn_to_virt(pfn) __va((pfn) << PAGE_SHIFT) +#define page_to_virt(page) pfn_to_virt(page_to_pfn(page)) #define VM_DATA_DEFAULT_FLAGS (VM_READ | VM_WRITE | \ VM_MAYREAD | VM_MAYWRITE | VM_MAYEXEC) diff --git a/arch/s390/include/asm/pgalloc.h b/arch/s390/include/asm/pgalloc.h index da34cb6..f4eb984 100644 --- a/arch/s390/include/asm/pgalloc.h +++ b/arch/s390/include/asm/pgalloc.h @@ -19,8 +19,10 @@ unsigned long *crst_table_alloc(struct mm_struct *); void crst_table_free(struct mm_struct *, unsigned long *); unsigned long *page_table_alloc(struct mm_struct *); +struct page *page_table_alloc_pgste(struct mm_struct *mm); void page_table_free(struct mm_struct *, unsigned long *); void page_table_free_rcu(struct mmu_gather *, unsigned long *, unsigned long); +void page_table_free_pgste(struct page *page); extern int page_table_allocate_pgste; static inline void clear_table(unsigned long *s, unsigned long val, size_t n) diff --git a/arch/s390/include/asm/pgtable.h b/arch/s390/include/asm/pgtable.h index 48d383a..72c7f60 100644 --- a/arch/s390/include/asm/pgtable.h +++ b/arch/s390/include/asm/pgtable.h @@ -277,6 +277,7 @@ static inline int is_module_addr(void *addr) /* Bits in the region table entry */ #define _REGION_ENTRY_ORIGIN ~0xfffUL/* region/segment table origin */ #define _REGION_ENTRY_PROTECT 0x200 /* region protection bit */ +#define _REGION_ENTRY_OFFSET 0xc0 /* region table offset */ #define _REGION_ENTRY_INVALID 0x20 /* invalid region table entry */ #define _REGION_ENTRY_TYPE_MASK 0x0c /* region/segment table type mask */ #define _REGION_ENTRY_TYPE_R1 0x0c /* region first table type */ @@ -364,6 +365,7 @@ static inline int is_module_addr(void *addr) #define PGSTE_GC_BIT 0x0002000000000000UL #define PGSTE_UC_BIT 0x0000800000000000UL /* user dirty (migration) */ #define PGSTE_IN_BIT 0x0000400000000000UL /* IPTE notify bit */ +#define PGSTE_VSIE_BIT 0x0000200000000000UL /* ref'd in a shadow table */ /* Guest Page State used for virtualization */ #define _PGSTE_GPS_ZERO 0x0000000080000000UL @@ -1002,15 +1004,26 @@ static inline int ptep_set_access_flags(struct vm_area_struct *vma, void ptep_set_pte_at(struct mm_struct *mm, unsigned long addr, pte_t *ptep, pte_t entry); void ptep_set_notify(struct mm_struct *mm, unsigned long addr, pte_t *ptep); -void ptep_notify(struct mm_struct *mm, unsigned long addr, pte_t *ptep); +void ptep_notify(struct mm_struct *mm, unsigned long addr, + pte_t *ptep, unsigned long bits); +int ptep_force_prot(struct mm_struct *mm, unsigned long gaddr, + pte_t *ptep, int prot, unsigned long bit); void ptep_zap_unused(struct mm_struct *mm, unsigned long addr, pte_t *ptep , int reset); void ptep_zap_key(struct mm_struct *mm, unsigned long addr, pte_t *ptep); +int ptep_shadow_pte(struct mm_struct *mm, unsigned long saddr, + pte_t *sptep, pte_t *tptep, pte_t pte); +void ptep_unshadow_pte(struct mm_struct *mm, unsigned long saddr, pte_t *ptep); bool test_and_clear_guest_dirty(struct mm_struct *mm, unsigned long address); int set_guest_storage_key(struct mm_struct *mm, unsigned long addr, unsigned char key, bool nq); -unsigned char get_guest_storage_key(struct mm_struct *mm, unsigned long addr); +int cond_set_guest_storage_key(struct mm_struct *mm, unsigned long addr, + unsigned char key, unsigned char *oldkey, + bool nq, bool mr, bool mc); +int reset_guest_reference_bit(struct mm_struct *mm, unsigned long addr); +int get_guest_storage_key(struct mm_struct *mm, unsigned long addr, + unsigned char *key); /* * Certain architectures need to do special things when PTEs diff --git a/arch/s390/include/asm/processor.h b/arch/s390/include/asm/processor.h index 0952920..0332317 100644 --- a/arch/s390/include/asm/processor.h +++ b/arch/s390/include/asm/processor.h @@ -112,6 +112,8 @@ struct thread_struct { unsigned long ksp; /* kernel stack pointer */ mm_segment_t mm_segment; unsigned long gmap_addr; /* address of last gmap fault. */ + unsigned int gmap_write_flag; /* gmap fault write indication */ + unsigned int gmap_int_code; /* int code of last gmap fault */ unsigned int gmap_pfault; /* signal of a pending guest pfault */ struct per_regs per_user; /* User specified PER registers */ struct per_event per_event; /* Cause of the last PER trap */ diff --git a/arch/s390/include/asm/sclp.h b/arch/s390/include/asm/sclp.h index e4f6f73..2ad9c20 100644 --- a/arch/s390/include/asm/sclp.h +++ b/arch/s390/include/asm/sclp.h @@ -32,12 +32,19 @@ struct sclp_core_entry { u8 reserved0; u8 : 4; u8 sief2 : 1; - u8 : 3; - u8 : 3; + u8 skey : 1; + u8 : 2; + u8 : 2; + u8 gpere : 1; u8 siif : 1; u8 sigpif : 1; u8 : 3; - u8 reserved2[10]; + u8 reserved2[3]; + u8 : 2; + u8 ib : 1; + u8 cei : 1; + u8 : 4; + u8 reserved3[6]; u8 type; u8 reserved1; } __attribute__((packed)); @@ -59,6 +66,15 @@ struct sclp_info { unsigned char has_hvs : 1; unsigned char has_esca : 1; unsigned char has_sief2 : 1; + unsigned char has_64bscao : 1; + unsigned char has_gpere : 1; + unsigned char has_cmma : 1; + unsigned char has_gsls : 1; + unsigned char has_ib : 1; + unsigned char has_cei : 1; + unsigned char has_pfmfi : 1; + unsigned char has_ibs : 1; + unsigned char has_skey : 1; unsigned int ibc; unsigned int mtid; unsigned int mtid_cp; @@ -101,5 +117,6 @@ int memcpy_hsa_kernel(void *dest, unsigned long src, size_t count); int memcpy_hsa_user(void __user *dest, unsigned long src, size_t count); void sclp_early_detect(void); void _sclp_print_early(const char *); +void sclp_ocf_cpc_name_copy(char *dst); #endif /* _ASM_S390_SCLP_H */ diff --git a/arch/s390/include/uapi/asm/kvm.h b/arch/s390/include/uapi/asm/kvm.h index 3b8e99e..a2ffec4 100644 --- a/arch/s390/include/uapi/asm/kvm.h +++ b/arch/s390/include/uapi/asm/kvm.h @@ -93,6 +93,47 @@ struct kvm_s390_vm_cpu_machine { __u64 fac_list[256]; }; +#define KVM_S390_VM_CPU_PROCESSOR_FEAT 2 +#define KVM_S390_VM_CPU_MACHINE_FEAT 3 + +#define KVM_S390_VM_CPU_FEAT_NR_BITS 1024 +#define KVM_S390_VM_CPU_FEAT_ESOP 0 +#define KVM_S390_VM_CPU_FEAT_SIEF2 1 +#define KVM_S390_VM_CPU_FEAT_64BSCAO 2 +#define KVM_S390_VM_CPU_FEAT_SIIF 3 +#define KVM_S390_VM_CPU_FEAT_GPERE 4 +#define KVM_S390_VM_CPU_FEAT_GSLS 5 +#define KVM_S390_VM_CPU_FEAT_IB 6 +#define KVM_S390_VM_CPU_FEAT_CEI 7 +#define KVM_S390_VM_CPU_FEAT_IBS 8 +#define KVM_S390_VM_CPU_FEAT_SKEY 9 +#define KVM_S390_VM_CPU_FEAT_CMMA 10 +#define KVM_S390_VM_CPU_FEAT_PFMFI 11 +#define KVM_S390_VM_CPU_FEAT_SIGPIF 12 +struct kvm_s390_vm_cpu_feat { + __u64 feat[16]; +}; + +#define KVM_S390_VM_CPU_PROCESSOR_SUBFUNC 4 +#define KVM_S390_VM_CPU_MACHINE_SUBFUNC 5 +/* for "test bit" instructions MSB 0 bit ordering, for "query" raw blocks */ +struct kvm_s390_vm_cpu_subfunc { + __u8 plo[32]; /* always */ + __u8 ptff[16]; /* with TOD-clock steering */ + __u8 kmac[16]; /* with MSA */ + __u8 kmc[16]; /* with MSA */ + __u8 km[16]; /* with MSA */ + __u8 kimd[16]; /* with MSA */ + __u8 klmd[16]; /* with MSA */ + __u8 pckmo[16]; /* with MSA3 */ + __u8 kmctr[16]; /* with MSA4 */ + __u8 kmf[16]; /* with MSA4 */ + __u8 kmo[16]; /* with MSA4 */ + __u8 pcc[16]; /* with MSA4 */ + __u8 ppno[16]; /* with MSA5 */ + __u8 reserved[1824]; +}; + /* kvm attributes for crypto */ #define KVM_S390_VM_CRYPTO_ENABLE_AES_KW 0 #define KVM_S390_VM_CRYPTO_ENABLE_DEA_KW 1 diff --git a/arch/s390/include/uapi/asm/sie.h b/arch/s390/include/uapi/asm/sie.h index 8fb5d4a..3ac6343 100644 --- a/arch/s390/include/uapi/asm/sie.h +++ b/arch/s390/include/uapi/asm/sie.h @@ -140,6 +140,7 @@ exit_code_ipa0(0xB2, 0x4c, "TAR"), \ exit_code_ipa0(0xB2, 0x50, "CSP"), \ exit_code_ipa0(0xB2, 0x54, "MVPG"), \ + exit_code_ipa0(0xB2, 0x56, "STHYI"), \ exit_code_ipa0(0xB2, 0x58, "BSG"), \ exit_code_ipa0(0xB2, 0x5a, "BSA"), \ exit_code_ipa0(0xB2, 0x5f, "CHSC"), \ diff --git a/arch/s390/kernel/diag.c b/arch/s390/kernel/diag.c index 48b37b8..a97354c 100644 --- a/arch/s390/kernel/diag.c +++ b/arch/s390/kernel/diag.c @@ -162,6 +162,30 @@ int diag14(unsigned long rx, unsigned long ry1, unsigned long subcode) } EXPORT_SYMBOL(diag14); +static inline int __diag204(unsigned long *subcode, unsigned long size, void *addr) +{ + register unsigned long _subcode asm("0") = *subcode; + register unsigned long _size asm("1") = size; + + asm volatile( + " diag %2,%0,0x204\n" + "0: nopr %%r7\n" + EX_TABLE(0b,0b) + : "+d" (_subcode), "+d" (_size) : "d" (addr) : "memory"); + *subcode = _subcode; + return _size; +} + +int diag204(unsigned long subcode, unsigned long size, void *addr) +{ + diag_stat_inc(DIAG_STAT_X204); + size = __diag204(&subcode, size, addr); + if (subcode) + return -1; + return size; +} +EXPORT_SYMBOL(diag204); + /* * Diagnose 210: Get information about a virtual device */ @@ -196,3 +220,18 @@ int diag210(struct diag210 *addr) return ccode; } EXPORT_SYMBOL(diag210); + +int diag224(void *ptr) +{ + int rc = -EOPNOTSUPP; + + diag_stat_inc(DIAG_STAT_X224); + asm volatile( + " diag %1,%2,0x224\n" + "0: lhi %0,0x0\n" + "1:\n" + EX_TABLE(0b,1b) + : "+d" (rc) :"d" (0), "d" (ptr) : "memory"); + return rc; +} +EXPORT_SYMBOL(diag224); diff --git a/arch/s390/kvm/Makefile b/arch/s390/kvm/Makefile index d42fa38..09a9e6d 100644 --- a/arch/s390/kvm/Makefile +++ b/arch/s390/kvm/Makefile @@ -12,6 +12,6 @@ common-objs = $(KVM)/kvm_main.o $(KVM)/eventfd.o $(KVM)/async_pf.o $(KVM)/irqch ccflags-y := -Ivirt/kvm -Iarch/s390/kvm kvm-objs := $(common-objs) kvm-s390.o intercept.o interrupt.o priv.o sigp.o -kvm-objs += diag.o gaccess.o guestdbg.o +kvm-objs += diag.o gaccess.o guestdbg.o sthyi.o vsie.o obj-$(CONFIG_KVM) += kvm.o diff --git a/arch/s390/kvm/diag.c b/arch/s390/kvm/diag.c index 1ea4095..ce865bd 100644 --- a/arch/s390/kvm/diag.c +++ b/arch/s390/kvm/diag.c @@ -212,6 +212,11 @@ static int __diag_virtio_hypercall(struct kvm_vcpu *vcpu) (vcpu->run->s.regs.gprs[1] != KVM_S390_VIRTIO_CCW_NOTIFY)) return -EOPNOTSUPP; + VCPU_EVENT(vcpu, 4, "diag 0x500 schid 0x%8.8x queue 0x%x cookie 0x%llx", + (u32) vcpu->run->s.regs.gprs[2], + (u32) vcpu->run->s.regs.gprs[3], + vcpu->run->s.regs.gprs[4]); + /* * The layout is as follows: * - gpr 2 contains the subchannel id (passed as addr) diff --git a/arch/s390/kvm/gaccess.c b/arch/s390/kvm/gaccess.c index 66938d2..5420020 100644 --- a/arch/s390/kvm/gaccess.c +++ b/arch/s390/kvm/gaccess.c @@ -8,6 +8,7 @@ #include <linux/vmalloc.h> #include <linux/err.h> #include <asm/pgtable.h> +#include <asm/gmap.h> #include "kvm-s390.h" #include "gaccess.h" #include <asm/switch_to.h> @@ -476,18 +477,73 @@ enum { FSI_FETCH = 2 /* Exception was due to fetch operation */ }; -static int get_vcpu_asce(struct kvm_vcpu *vcpu, union asce *asce, - ar_t ar, enum gacc_mode mode) +enum prot_type { + PROT_TYPE_LA = 0, + PROT_TYPE_KEYC = 1, + PROT_TYPE_ALC = 2, + PROT_TYPE_DAT = 3, +}; + +static int trans_exc(struct kvm_vcpu *vcpu, int code, unsigned long gva, + ar_t ar, enum gacc_mode mode, enum prot_type prot) { - int rc; - struct psw_bits psw = psw_bits(vcpu->arch.sie_block->gpsw); struct kvm_s390_pgm_info *pgm = &vcpu->arch.pgm; - struct trans_exc_code_bits *tec_bits; + struct trans_exc_code_bits *tec; memset(pgm, 0, sizeof(*pgm)); - tec_bits = (struct trans_exc_code_bits *)&pgm->trans_exc_code; - tec_bits->fsi = mode == GACC_STORE ? FSI_STORE : FSI_FETCH; - tec_bits->as = psw.as; + pgm->code = code; + tec = (struct trans_exc_code_bits *)&pgm->trans_exc_code; + + switch (code) { + case PGM_ASCE_TYPE: + case PGM_PAGE_TRANSLATION: + case PGM_REGION_FIRST_TRANS: + case PGM_REGION_SECOND_TRANS: + case PGM_REGION_THIRD_TRANS: + case PGM_SEGMENT_TRANSLATION: + /* + * op_access_id only applies to MOVE_PAGE -> set bit 61 + * exc_access_id has to be set to 0 for some instructions. Both + * cases have to be handled by the caller. We can always store + * exc_access_id, as it is undefined for non-ar cases. + */ + tec->addr = gva >> PAGE_SHIFT; + tec->fsi = mode == GACC_STORE ? FSI_STORE : FSI_FETCH; + tec->as = psw_bits(vcpu->arch.sie_block->gpsw).as; + /* FALL THROUGH */ + case PGM_ALEN_TRANSLATION: + case PGM_ALE_SEQUENCE: + case PGM_ASTE_VALIDITY: + case PGM_ASTE_SEQUENCE: + case PGM_EXTENDED_AUTHORITY: + pgm->exc_access_id = ar; + break; + case PGM_PROTECTION: + switch (prot) { + case PROT_TYPE_ALC: + tec->b60 = 1; + /* FALL THROUGH */ + case PROT_TYPE_DAT: + tec->b61 = 1; + tec->addr = gva >> PAGE_SHIFT; + tec->fsi = mode == GACC_STORE ? FSI_STORE : FSI_FETCH; + tec->as = psw_bits(vcpu->arch.sie_block->gpsw).as; + /* exc_access_id is undefined for most cases */ + pgm->exc_access_id = ar; + break; + default: /* LA and KEYC set b61 to 0, other params undefined */ + break; + } + break; + } + return code; +} + +static int get_vcpu_asce(struct kvm_vcpu *vcpu, union asce *asce, + unsigned long ga, ar_t ar, enum gacc_mode mode) +{ + int rc; + struct psw_bits psw = psw_bits(vcpu->arch.sie_block->gpsw); if (!psw.t) { asce->val = 0; @@ -510,21 +566,8 @@ static int get_vcpu_asce(struct kvm_vcpu *vcpu, union asce *asce, return 0; case PSW_AS_ACCREG: rc = ar_translation(vcpu, asce, ar, mode); - switch (rc) { - case PGM_ALEN_TRANSLATION: - case PGM_ALE_SEQUENCE: - case PGM_ASTE_VALIDITY: - case PGM_ASTE_SEQUENCE: - case PGM_EXTENDED_AUTHORITY: - vcpu->arch.pgm.exc_access_id = ar; - break; - case PGM_PROTECTION: - tec_bits->b60 = 1; - tec_bits->b61 = 1; - break; - } if (rc > 0) - pgm->code = rc; + return trans_exc(vcpu, rc, ga, ar, mode, PROT_TYPE_ALC); return rc; } return 0; @@ -729,40 +772,31 @@ static int low_address_protection_enabled(struct kvm_vcpu *vcpu, return 1; } -static int guest_page_range(struct kvm_vcpu *vcpu, unsigned long ga, +static int guest_page_range(struct kvm_vcpu *vcpu, unsigned long ga, ar_t ar, unsigned long *pages, unsigned long nr_pages, const union asce asce, enum gacc_mode mode) { - struct kvm_s390_pgm_info *pgm = &vcpu->arch.pgm; psw_t *psw = &vcpu->arch.sie_block->gpsw; - struct trans_exc_code_bits *tec_bits; - int lap_enabled, rc; + int lap_enabled, rc = 0; - tec_bits = (struct trans_exc_code_bits *)&pgm->trans_exc_code; lap_enabled = low_address_protection_enabled(vcpu, asce); while (nr_pages) { ga = kvm_s390_logical_to_effective(vcpu, ga); - tec_bits->addr = ga >> PAGE_SHIFT; - if (mode == GACC_STORE && lap_enabled && is_low_address(ga)) { - pgm->code = PGM_PROTECTION; - return pgm->code; - } + if (mode == GACC_STORE && lap_enabled && is_low_address(ga)) + return trans_exc(vcpu, PGM_PROTECTION, ga, ar, mode, + PROT_TYPE_LA); ga &= PAGE_MASK; if (psw_bits(*psw).t) { rc = guest_translate(vcpu, ga, pages, asce, mode); if (rc < 0) return rc; - if (rc == PGM_PROTECTION) - tec_bits->b61 = 1; - if (rc) - pgm->code = rc; } else { *pages = kvm_s390_real_to_abs(vcpu, ga); if (kvm_is_error_gpa(vcpu->kvm, *pages)) - pgm->code = PGM_ADDRESSING; + rc = PGM_ADDRESSING; } - if (pgm->code) - return pgm->code; + if (rc) + return trans_exc(vcpu, rc, ga, ar, mode, PROT_TYPE_DAT); ga += PAGE_SIZE; pages++; nr_pages--; @@ -783,7 +817,8 @@ int access_guest(struct kvm_vcpu *vcpu, unsigned long ga, ar_t ar, void *data, if (!len) return 0; - rc = get_vcpu_asce(vcpu, &asce, ar, mode); + ga = kvm_s390_logical_to_effective(vcpu, ga); + rc = get_vcpu_asce(vcpu, &asce, ga, ar, mode); if (rc) return rc; nr_pages = (((ga & ~PAGE_MASK) + len - 1) >> PAGE_SHIFT) + 1; @@ -795,7 +830,7 @@ int access_guest(struct kvm_vcpu *vcpu, unsigned long ga, ar_t ar, void *data, need_ipte_lock = psw_bits(*psw).t && !asce.r; if (need_ipte_lock) ipte_lock(vcpu); - rc = guest_page_range(vcpu, ga, pages, nr_pages, asce, mode); + rc = guest_page_range(vcpu, ga, ar, pages, nr_pages, asce, mode); for (idx = 0; idx < nr_pages && !rc; idx++) { gpa = *(pages + idx) + (ga & ~PAGE_MASK); _len = min(PAGE_SIZE - (gpa & ~PAGE_MASK), len); @@ -846,37 +881,28 @@ int access_guest_real(struct kvm_vcpu *vcpu, unsigned long gra, int guest_translate_address(struct kvm_vcpu *vcpu, unsigned long gva, ar_t ar, unsigned long *gpa, enum gacc_mode mode) { - struct kvm_s390_pgm_info *pgm = &vcpu->arch.pgm; psw_t *psw = &vcpu->arch.sie_block->gpsw; - struct trans_exc_code_bits *tec; union asce asce; int rc; gva = kvm_s390_logical_to_effective(vcpu, gva); - tec = (struct trans_exc_code_bits *)&pgm->trans_exc_code; - rc = get_vcpu_asce(vcpu, &asce, ar, mode); - tec->addr = gva >> PAGE_SHIFT; + rc = get_vcpu_asce(vcpu, &asce, gva, ar, mode); if (rc) return rc; if (is_low_address(gva) && low_address_protection_enabled(vcpu, asce)) { - if (mode == GACC_STORE) { - rc = pgm->code = PGM_PROTECTION; - return rc; - } + if (mode == GACC_STORE) + return trans_exc(vcpu, PGM_PROTECTION, gva, 0, + mode, PROT_TYPE_LA); } if (psw_bits(*psw).t && !asce.r) { /* Use DAT? */ rc = guest_translate(vcpu, gva, gpa, asce, mode); - if (rc > 0) { - if (rc == PGM_PROTECTION) - tec->b61 = 1; - pgm->code = rc; - } + if (rc > 0) + return trans_exc(vcpu, rc, gva, 0, mode, PROT_TYPE_DAT); } else { - rc = 0; *gpa = kvm_s390_real_to_abs(vcpu, gva); if (kvm_is_error_gpa(vcpu->kvm, *gpa)) - rc = pgm->code = PGM_ADDRESSING; + return trans_exc(vcpu, rc, gva, PGM_ADDRESSING, mode, 0); } return rc; @@ -915,20 +941,247 @@ int check_gva_range(struct kvm_vcpu *vcpu, unsigned long gva, ar_t ar, */ int kvm_s390_check_low_addr_prot_real(struct kvm_vcpu *vcpu, unsigned long gra) { - struct kvm_s390_pgm_info *pgm = &vcpu->arch.pgm; - psw_t *psw = &vcpu->arch.sie_block->gpsw; - struct trans_exc_code_bits *tec_bits; union ctlreg0 ctlreg0 = {.val = vcpu->arch.sie_block->gcr[0]}; if (!ctlreg0.lap || !is_low_address(gra)) return 0; + return trans_exc(vcpu, PGM_PROTECTION, gra, 0, GACC_STORE, PROT_TYPE_LA); +} - memset(pgm, 0, sizeof(*pgm)); - tec_bits = (struct trans_exc_code_bits *)&pgm->trans_exc_code; - tec_bits->fsi = FSI_STORE; - tec_bits->as = psw_bits(*psw).as; - tec_bits->addr = gra >> PAGE_SHIFT; - pgm->code = PGM_PROTECTION; +/** + * kvm_s390_shadow_tables - walk the guest page table and create shadow tables + * @sg: pointer to the shadow guest address space structure + * @saddr: faulting address in the shadow gmap + * @pgt: pointer to the page table address result + * @fake: pgt references contiguous guest memory block, not a pgtable + */ +static int kvm_s390_shadow_tables(struct gmap *sg, unsigned long saddr, + unsigned long *pgt, int *dat_protection, + int *fake) +{ + struct gmap *parent; + union asce asce; + union vaddress vaddr; + unsigned long ptr; + int rc; + + *fake = 0; + *dat_protection = 0; + parent = sg->parent; + vaddr.addr = saddr; + asce.val = sg->orig_asce; + ptr = asce.origin * 4096; + if (asce.r) { + *fake = 1; + asce.dt = ASCE_TYPE_REGION1; + } + switch (asce.dt) { + case ASCE_TYPE_REGION1: + if (vaddr.rfx01 > asce.tl && !asce.r) + return PGM_REGION_FIRST_TRANS; + break; + case ASCE_TYPE_REGION2: + if (vaddr.rfx) + return PGM_ASCE_TYPE; + if (vaddr.rsx01 > asce.tl) + return PGM_REGION_SECOND_TRANS; + break; + case ASCE_TYPE_REGION3: + if (vaddr.rfx || vaddr.rsx) + return PGM_ASCE_TYPE; + if (vaddr.rtx01 > asce.tl) + return PGM_REGION_THIRD_TRANS; + break; + case ASCE_TYPE_SEGMENT: + if (vaddr.rfx || vaddr.rsx || vaddr.rtx) + return PGM_ASCE_TYPE; + if (vaddr.sx01 > asce.tl) + return PGM_SEGMENT_TRANSLATION; + break; + } + + switch (asce.dt) { + case ASCE_TYPE_REGION1: { + union region1_table_entry rfte; - return pgm->code; + if (*fake) { + /* offset in 16EB guest memory block */ + ptr = ptr + ((unsigned long) vaddr.rsx << 53UL); + rfte.val = ptr; + goto shadow_r2t; + } + rc = gmap_read_table(parent, ptr + vaddr.rfx * 8, &rfte.val); + if (rc) + return rc; + if (rfte.i) + return PGM_REGION_FIRST_TRANS; + if (rfte.tt != TABLE_TYPE_REGION1) + return PGM_TRANSLATION_SPEC; + if (vaddr.rsx01 < rfte.tf || vaddr.rsx01 > rfte.tl) + return PGM_REGION_SECOND_TRANS; + if (sg->edat_level >= 1) + *dat_protection |= rfte.p; + ptr = rfte.rto << 12UL; +shadow_r2t: + rc = gmap_shadow_r2t(sg, saddr, rfte.val, *fake); + if (rc) + return rc; + /* fallthrough */ + } + case ASCE_TYPE_REGION2: { + union region2_table_entry rste; + + if (*fake) { + /* offset in 8PB guest memory block */ + ptr = ptr + ((unsigned long) vaddr.rtx << 42UL); + rste.val = ptr; + goto shadow_r3t; + } + rc = gmap_read_table(parent, ptr + vaddr.rsx * 8, &rste.val); + if (rc) + return rc; + if (rste.i) + return PGM_REGION_SECOND_TRANS; + if (rste.tt != TABLE_TYPE_REGION2) + return PGM_TRANSLATION_SPEC; + if (vaddr.rtx01 < rste.tf || vaddr.rtx01 > rste.tl) + return PGM_REGION_THIRD_TRANS; + if (sg->edat_level >= 1) + *dat_protection |= rste.p; + ptr = rste.rto << 12UL; +shadow_r3t: + rste.p |= *dat_protection; + rc = gmap_shadow_r3t(sg, saddr, rste.val, *fake); + if (rc) + return rc; + /* fallthrough */ + } + case ASCE_TYPE_REGION3: { + union region3_table_entry rtte; + + if (*fake) { + /* offset in 4TB guest memory block */ + ptr = ptr + ((unsigned long) vaddr.sx << 31UL); + rtte.val = ptr; + goto shadow_sgt; + } + rc = gmap_read_table(parent, ptr + vaddr.rtx * 8, &rtte.val); + if (rc) + return rc; + if (rtte.i) + return PGM_REGION_THIRD_TRANS; + if (rtte.tt != TABLE_TYPE_REGION3) + return PGM_TRANSLATION_SPEC; + if (rtte.cr && asce.p && sg->edat_level >= 2) + return PGM_TRANSLATION_SPEC; + if (rtte.fc && sg->edat_level >= 2) { + *dat_protection |= rtte.fc0.p; + *fake = 1; + ptr = rtte.fc1.rfaa << 31UL; + rtte.val = ptr; + goto shadow_sgt; + } + if (vaddr.sx01 < rtte.fc0.tf || vaddr.sx01 > rtte.fc0.tl) + return PGM_SEGMENT_TRANSLATION; + if (sg->edat_level >= 1) + *dat_protection |= rtte.fc0.p; + ptr = rtte.fc0.sto << 12UL; +shadow_sgt: + rtte.fc0.p |= *dat_protection; + rc = gmap_shadow_sgt(sg, saddr, rtte.val, *fake); + if (rc) + return rc; + /* fallthrough */ + } + case ASCE_TYPE_SEGMENT: { + union segment_table_entry ste; + + if (*fake) { + /* offset in 2G guest memory block */ + ptr = ptr + ((unsigned long) vaddr.sx << 20UL); + ste.val = ptr; + goto shadow_pgt; + } + rc = gmap_read_table(parent, ptr + vaddr.sx * 8, &ste.val); + if (rc) + return rc; + if (ste.i) + return PGM_SEGMENT_TRANSLATION; + if (ste.tt != TABLE_TYPE_SEGMENT) + return PGM_TRANSLATION_SPEC; + if (ste.cs && asce.p) + return PGM_TRANSLATION_SPEC; + *dat_protection |= ste.fc0.p; + if (ste.fc && sg->edat_level >= 1) { + *fake = 1; + ptr = ste.fc1.sfaa << 20UL; + ste.val = ptr; + goto shadow_pgt; + } + ptr = ste.fc0.pto << 11UL; +shadow_pgt: + ste.fc0.p |= *dat_protection; + rc = gmap_shadow_pgt(sg, saddr, ste.val, *fake); + if (rc) + return rc; + } + } + /* Return the parent address of the page table */ + *pgt = ptr; + return 0; +} + +/** + * kvm_s390_shadow_fault - handle fault on a shadow page table + * @vcpu: virtual cpu + * @sg: pointer to the shadow guest address space structure + * @saddr: faulting address in the shadow gmap + * + * Returns: - 0 if the shadow fault was successfully resolved + * - > 0 (pgm exception code) on exceptions while faulting + * - -EAGAIN if the caller can retry immediately + * - -EFAULT when accessing invalid guest addresses + * - -ENOMEM if out of memory + */ +int kvm_s390_shadow_fault(struct kvm_vcpu *vcpu, struct gmap *sg, + unsigned long saddr) +{ + union vaddress vaddr; + union page_table_entry pte; + unsigned long pgt; + int dat_protection, fake; + int rc; + + down_read(&sg->mm->mmap_sem); + /* + * We don't want any guest-2 tables to change - so the parent + * tables/pointers we read stay valid - unshadowing is however + * always possible - only guest_table_lock protects us. + */ + ipte_lock(vcpu); + + rc = gmap_shadow_pgt_lookup(sg, saddr, &pgt, &dat_protection, &fake); + if (rc) + rc = kvm_s390_shadow_tables(sg, saddr, &pgt, &dat_protection, + &fake); + + vaddr.addr = saddr; + if (fake) { + /* offset in 1MB guest memory block */ + pte.val = pgt + ((unsigned long) vaddr.px << 12UL); + goto shadow_page; + } + if (!rc) + rc = gmap_read_table(sg->parent, pgt + vaddr.px * 8, &pte.val); + if (!rc && pte.i) + rc = PGM_PAGE_TRANSLATION; + if (!rc && (pte.z || (pte.co && sg->edat_level < 1))) + rc = PGM_TRANSLATION_SPEC; +shadow_page: + pte.p |= dat_protection; + if (!rc) + rc = gmap_shadow_page(sg, saddr, __pte(pte.val)); + ipte_unlock(vcpu); + up_read(&sg->mm->mmap_sem); + return rc; } diff --git a/arch/s390/kvm/gaccess.h b/arch/s390/kvm/gaccess.h index df0a79d..8756569 100644 --- a/arch/s390/kvm/gaccess.h +++ b/arch/s390/kvm/gaccess.h @@ -361,4 +361,7 @@ void ipte_unlock(struct kvm_vcpu *vcpu); int ipte_lock_held(struct kvm_vcpu *vcpu); int kvm_s390_check_low_addr_prot_real(struct kvm_vcpu *vcpu, unsigned long gra); +int kvm_s390_shadow_fault(struct kvm_vcpu *vcpu, struct gmap *shadow, + unsigned long saddr); + #endif /* __KVM_S390_GACCESS_H */ diff --git a/arch/s390/kvm/guestdbg.c b/arch/s390/kvm/guestdbg.c index e8c6843..31a0533 100644 --- a/arch/s390/kvm/guestdbg.c +++ b/arch/s390/kvm/guestdbg.c @@ -439,6 +439,23 @@ exit_required: #define guest_per_enabled(vcpu) \ (vcpu->arch.sie_block->gpsw.mask & PSW_MASK_PER) +int kvm_s390_handle_per_ifetch_icpt(struct kvm_vcpu *vcpu) +{ + const u8 ilen = kvm_s390_get_ilen(vcpu); + struct kvm_s390_pgm_info pgm_info = { + .code = PGM_PER, + .per_code = PER_EVENT_IFETCH >> 24, + .per_address = __rewind_psw(vcpu->arch.sie_block->gpsw, ilen), + }; + + /* + * The PSW points to the next instruction, therefore the intercepted + * instruction generated a PER i-fetch event. PER address therefore + * points at the previous PSW address (could be an EXECUTE function). + */ + return kvm_s390_inject_prog_irq(vcpu, &pgm_info); +} + static void filter_guest_per_event(struct kvm_vcpu *vcpu) { u32 perc = vcpu->arch.sie_block->perc << 24; @@ -465,7 +482,7 @@ static void filter_guest_per_event(struct kvm_vcpu *vcpu) guest_perc &= ~PER_EVENT_IFETCH; /* All other PER events will be given to the guest */ - /* TODO: Check alterated address/address space */ + /* TODO: Check altered address/address space */ vcpu->arch.sie_block->perc = guest_perc >> 24; diff --git a/arch/s390/kvm/intercept.c b/arch/s390/kvm/intercept.c index 2521571..dfd0ca2 100644 --- a/arch/s390/kvm/intercept.c +++ b/arch/s390/kvm/intercept.c @@ -351,8 +351,26 @@ static int handle_partial_execution(struct kvm_vcpu *vcpu) return -EOPNOTSUPP; } +static int handle_operexc(struct kvm_vcpu *vcpu) +{ + vcpu->stat.exit_operation_exception++; + trace_kvm_s390_handle_operexc(vcpu, vcpu->arch.sie_block->ipa, + vcpu->arch.sie_block->ipb); + + if (vcpu->arch.sie_block->ipa == 0xb256 && + test_kvm_facility(vcpu->kvm, 74)) + return handle_sthyi(vcpu); + + if (vcpu->arch.sie_block->ipa == 0 && vcpu->kvm->arch.user_instr0) + return -EOPNOTSUPP; + + return kvm_s390_inject_program_int(vcpu, PGM_OPERATION); +} + int kvm_handle_sie_intercept(struct kvm_vcpu *vcpu) { + int rc, per_rc = 0; + if (kvm_is_ucontrol(vcpu->kvm)) return -EOPNOTSUPP; @@ -361,7 +379,8 @@ int kvm_handle_sie_intercept(struct kvm_vcpu *vcpu) case 0x18: return handle_noop(vcpu); case 0x04: - return handle_instruction(vcpu); + rc = handle_instruction(vcpu); + break; case 0x08: return handle_prog(vcpu); case 0x14: @@ -372,9 +391,19 @@ int kvm_handle_sie_intercept(struct kvm_vcpu *vcpu) return handle_validity(vcpu); case 0x28: return handle_stop(vcpu); + case 0x2c: + rc = handle_operexc(vcpu); + break; case 0x38: - return handle_partial_execution(vcpu); + rc = handle_partial_execution(vcpu); + break; default: return -EOPNOTSUPP; } + + /* process PER, also if the instrution is processed in user space */ + if (vcpu->arch.sie_block->icptstatus & 0x02 && + (!rc || rc == -EOPNOTSUPP)) + per_rc = kvm_s390_handle_per_ifetch_icpt(vcpu); + return per_rc ? per_rc : rc; } diff --git a/arch/s390/kvm/interrupt.c b/arch/s390/kvm/interrupt.c index 5a80af7..24524c0 100644 --- a/arch/s390/kvm/interrupt.c +++ b/arch/s390/kvm/interrupt.c @@ -28,9 +28,6 @@ #include "gaccess.h" #include "trace-s390.h" -#define IOINT_SCHID_MASK 0x0000ffff -#define IOINT_SSID_MASK 0x00030000 -#define IOINT_CSSID_MASK 0x03fc0000 #define PFAULT_INIT 0x0600 #define PFAULT_DONE 0x0680 #define VIRTIO_PARAM 0x0d00 @@ -821,7 +818,14 @@ static int __must_check __deliver_io(struct kvm_vcpu *vcpu, struct kvm_s390_interrupt_info, list); if (inti) { - VCPU_EVENT(vcpu, 4, "deliver: I/O 0x%llx", inti->type); + if (inti->type & KVM_S390_INT_IO_AI_MASK) + VCPU_EVENT(vcpu, 4, "%s", "deliver: I/O (AI)"); + else + VCPU_EVENT(vcpu, 4, "deliver: I/O %x ss %x schid %04x", + inti->io.subchannel_id >> 8, + inti->io.subchannel_id >> 1 & 0x3, + inti->io.subchannel_nr); + vcpu->stat.deliver_io_int++; trace_kvm_s390_deliver_interrupt(vcpu->vcpu_id, inti->type, @@ -991,6 +995,11 @@ void kvm_s390_vcpu_wakeup(struct kvm_vcpu *vcpu) swake_up(&vcpu->wq); vcpu->stat.halt_wakeup++; } + /* + * The VCPU might not be sleeping but is executing the VSIE. Let's + * kick it, so it leaves the SIE to process the request. + */ + kvm_s390_vsie_kick(vcpu); } enum hrtimer_restart kvm_s390_idle_wakeup(struct hrtimer *timer) @@ -1415,6 +1424,13 @@ static int __inject_io(struct kvm *kvm, struct kvm_s390_interrupt_info *inti) } fi->counters[FIRQ_CNTR_IO] += 1; + if (inti->type & KVM_S390_INT_IO_AI_MASK) + VM_EVENT(kvm, 4, "%s", "inject: I/O (AI)"); + else + VM_EVENT(kvm, 4, "inject: I/O %x ss %x schid %04x", + inti->io.subchannel_id >> 8, + inti->io.subchannel_id >> 1 & 0x3, + inti->io.subchannel_nr); isc = int_word_to_isc(inti->io.io_int_word); list = &fi->lists[FIRQ_LIST_IO_ISC_0 + isc]; list_add_tail(&inti->list, list); @@ -1531,13 +1547,6 @@ int kvm_s390_inject_vm(struct kvm *kvm, inti->mchk.mcic = s390int->parm64; break; case KVM_S390_INT_IO_MIN...KVM_S390_INT_IO_MAX: - if (inti->type & KVM_S390_INT_IO_AI_MASK) - VM_EVENT(kvm, 5, "%s", "inject: I/O (AI)"); - else - VM_EVENT(kvm, 5, "inject: I/O css %x ss %x schid %04x", - s390int->type & IOINT_CSSID_MASK, - s390int->type & IOINT_SSID_MASK, - s390int->type & IOINT_SCHID_MASK); inti->io.subchannel_id = s390int->parm >> 16; inti->io.subchannel_nr = s390int->parm & 0x0000ffffu; inti->io.io_int_parm = s390int->parm64 >> 32; @@ -2237,7 +2246,8 @@ static int set_adapter_int(struct kvm_kernel_irq_routing_entry *e, return ret; } -int kvm_set_routing_entry(struct kvm_kernel_irq_routing_entry *e, +int kvm_set_routing_entry(struct kvm *kvm, + struct kvm_kernel_irq_routing_entry *e, const struct kvm_irq_routing_entry *ue) { int ret; diff --git a/arch/s390/kvm/kvm-s390.c b/arch/s390/kvm/kvm-s390.c index 6f5c344..3f3ae48 100644 --- a/arch/s390/kvm/kvm-s390.c +++ b/arch/s390/kvm/kvm-s390.c @@ -21,11 +21,13 @@ #include <linux/init.h> #include <linux/kvm.h> #include <linux/kvm_host.h> +#include <linux/mman.h> #include <linux/module.h> #include <linux/random.h> #include <linux/slab.h> #include <linux/timer.h> #include <linux/vmalloc.h> +#include <linux/bitmap.h> #include <asm/asm-offsets.h> #include <asm/lowcore.h> #include <asm/stp.h> @@ -35,6 +37,8 @@ #include <asm/switch_to.h> #include <asm/isc.h> #include <asm/sclp.h> +#include <asm/cpacf.h> +#include <asm/timex.h> #include "kvm-s390.h" #include "gaccess.h" @@ -64,6 +68,7 @@ struct kvm_stats_debugfs_item debugfs_entries[] = { { "exit_pei", VCPU_STAT(exit_pei) }, { "exit_program_interruption", VCPU_STAT(exit_program_interruption) }, { "exit_instr_and_program_int", VCPU_STAT(exit_instr_and_program) }, + { "exit_operation_exception", VCPU_STAT(exit_operation_exception) }, { "halt_successful_poll", VCPU_STAT(halt_successful_poll) }, { "halt_attempted_poll", VCPU_STAT(halt_attempted_poll) }, { "halt_poll_invalid", VCPU_STAT(halt_poll_invalid) }, @@ -94,6 +99,8 @@ struct kvm_stats_debugfs_item debugfs_entries[] = { { "instruction_stsi", VCPU_STAT(instruction_stsi) }, { "instruction_stfl", VCPU_STAT(instruction_stfl) }, { "instruction_tprot", VCPU_STAT(instruction_tprot) }, + { "instruction_sthyi", VCPU_STAT(instruction_sthyi) }, + { "instruction_sie", VCPU_STAT(instruction_sie) }, { "instruction_sigp_sense", VCPU_STAT(instruction_sigp_sense) }, { "instruction_sigp_sense_running", VCPU_STAT(instruction_sigp_sense_running) }, { "instruction_sigp_external_call", VCPU_STAT(instruction_sigp_external_call) }, @@ -119,6 +126,11 @@ struct kvm_stats_debugfs_item debugfs_entries[] = { { NULL } }; +/* allow nested virtualization in KVM (if enabled by user space) */ +static int nested; +module_param(nested, int, S_IRUGO); +MODULE_PARM_DESC(nested, "Nested virtualization support"); + /* upper facilities limit for kvm */ unsigned long kvm_s390_fac_list_mask[16] = { 0xffe6000000000000UL, @@ -131,7 +143,13 @@ unsigned long kvm_s390_fac_list_mask_size(void) return ARRAY_SIZE(kvm_s390_fac_list_mask); } +/* available cpu features supported by kvm */ +static DECLARE_BITMAP(kvm_s390_available_cpu_feat, KVM_S390_VM_CPU_FEAT_NR_BITS); +/* available subfunctions indicated via query / "test bit" */ +static struct kvm_s390_vm_cpu_subfunc kvm_s390_available_subfunc; + static struct gmap_notifier gmap_notifier; +static struct gmap_notifier vsie_gmap_notifier; debug_info_t *kvm_s390_dbf; /* Section: not file related */ @@ -141,7 +159,8 @@ int kvm_arch_hardware_enable(void) return 0; } -static void kvm_gmap_notifier(struct gmap *gmap, unsigned long address); +static void kvm_gmap_notifier(struct gmap *gmap, unsigned long start, + unsigned long end); /* * This callback is executed during stop_machine(). All CPUs are therefore @@ -163,6 +182,8 @@ static int kvm_clock_sync(struct notifier_block *notifier, unsigned long val, vcpu->arch.sie_block->epoch -= *delta; if (vcpu->arch.cputm_enabled) vcpu->arch.cputm_start += *delta; + if (vcpu->arch.vsie_block) + vcpu->arch.vsie_block->epoch -= *delta; } } return NOTIFY_OK; @@ -175,7 +196,9 @@ static struct notifier_block kvm_clock_notifier = { int kvm_arch_hardware_setup(void) { gmap_notifier.notifier_call = kvm_gmap_notifier; - gmap_register_ipte_notifier(&gmap_notifier); + gmap_register_pte_notifier(&gmap_notifier); + vsie_gmap_notifier.notifier_call = kvm_s390_vsie_gmap_notifier; + gmap_register_pte_notifier(&vsie_gmap_notifier); atomic_notifier_chain_register(&s390_epoch_delta_notifier, &kvm_clock_notifier); return 0; @@ -183,11 +206,109 @@ int kvm_arch_hardware_setup(void) void kvm_arch_hardware_unsetup(void) { - gmap_unregister_ipte_notifier(&gmap_notifier); + gmap_unregister_pte_notifier(&gmap_notifier); + gmap_unregister_pte_notifier(&vsie_gmap_notifier); atomic_notifier_chain_unregister(&s390_epoch_delta_notifier, &kvm_clock_notifier); } +static void allow_cpu_feat(unsigned long nr) +{ + set_bit_inv(nr, kvm_s390_available_cpu_feat); +} + +static inline int plo_test_bit(unsigned char nr) +{ + register unsigned long r0 asm("0") = (unsigned long) nr | 0x100; + int cc = 3; /* subfunction not available */ + + asm volatile( + /* Parameter registers are ignored for "test bit" */ + " plo 0,0,0,0(0)\n" + " ipm %0\n" + " srl %0,28\n" + : "=d" (cc) + : "d" (r0) + : "cc"); + return cc == 0; +} + +static void kvm_s390_cpu_feat_init(void) +{ + int i; + + for (i = 0; i < 256; ++i) { + if (plo_test_bit(i)) + kvm_s390_available_subfunc.plo[i >> 3] |= 0x80 >> (i & 7); + } + + if (test_facility(28)) /* TOD-clock steering */ + ptff(kvm_s390_available_subfunc.ptff, + sizeof(kvm_s390_available_subfunc.ptff), + PTFF_QAF); + + if (test_facility(17)) { /* MSA */ + __cpacf_query(CPACF_KMAC, kvm_s390_available_subfunc.kmac); + __cpacf_query(CPACF_KMC, kvm_s390_available_subfunc.kmc); + __cpacf_query(CPACF_KM, kvm_s390_available_subfunc.km); + __cpacf_query(CPACF_KIMD, kvm_s390_available_subfunc.kimd); + __cpacf_query(CPACF_KLMD, kvm_s390_available_subfunc.klmd); + } + if (test_facility(76)) /* MSA3 */ + __cpacf_query(CPACF_PCKMO, kvm_s390_available_subfunc.pckmo); + if (test_facility(77)) { /* MSA4 */ + __cpacf_query(CPACF_KMCTR, kvm_s390_available_subfunc.kmctr); + __cpacf_query(CPACF_KMF, kvm_s390_available_subfunc.kmf); + __cpacf_query(CPACF_KMO, kvm_s390_available_subfunc.kmo); + __cpacf_query(CPACF_PCC, kvm_s390_available_subfunc.pcc); + } + if (test_facility(57)) /* MSA5 */ + __cpacf_query(CPACF_PPNO, kvm_s390_available_subfunc.ppno); + + if (MACHINE_HAS_ESOP) + allow_cpu_feat(KVM_S390_VM_CPU_FEAT_ESOP); + /* + * We need SIE support, ESOP (PROT_READ protection for gmap_shadow), + * 64bit SCAO (SCA passthrough) and IDTE (for gmap_shadow unshadowing). + */ + if (!sclp.has_sief2 || !MACHINE_HAS_ESOP || !sclp.has_64bscao || + !test_facility(3) || !nested) + return; + allow_cpu_feat(KVM_S390_VM_CPU_FEAT_SIEF2); + if (sclp.has_64bscao) + allow_cpu_feat(KVM_S390_VM_CPU_FEAT_64BSCAO); + if (sclp.has_siif) + allow_cpu_feat(KVM_S390_VM_CPU_FEAT_SIIF); + if (sclp.has_gpere) + allow_cpu_feat(KVM_S390_VM_CPU_FEAT_GPERE); + if (sclp.has_gsls) + allow_cpu_feat(KVM_S390_VM_CPU_FEAT_GSLS); + if (sclp.has_ib) + allow_cpu_feat(KVM_S390_VM_CPU_FEAT_IB); + if (sclp.has_cei) + allow_cpu_feat(KVM_S390_VM_CPU_FEAT_CEI); + if (sclp.has_ibs) + allow_cpu_feat(KVM_S390_VM_CPU_FEAT_IBS); + /* + * KVM_S390_VM_CPU_FEAT_SKEY: Wrong shadow of PTE.I bits will make + * all skey handling functions read/set the skey from the PGSTE + * instead of the real storage key. + * + * KVM_S390_VM_CPU_FEAT_CMMA: Wrong shadow of PTE.I bits will make + * pages being detected as preserved although they are resident. + * + * KVM_S390_VM_CPU_FEAT_PFMFI: Wrong shadow of PTE.I bits will + * have the same effect as for KVM_S390_VM_CPU_FEAT_SKEY. + * + * For KVM_S390_VM_CPU_FEAT_SKEY, KVM_S390_VM_CPU_FEAT_CMMA and + * KVM_S390_VM_CPU_FEAT_PFMFI, all PTE.I and PGSTE bits have to be + * correctly shadowed. We can do that for the PGSTE but not for PTE.I. + * + * KVM_S390_VM_CPU_FEAT_SIGPIF: Wrong SCB addresses in the SCA. We + * cannot easily shadow the SCA because of the ipte lock. + */ +} + int kvm_arch_init(void *opaque) { kvm_s390_dbf = debug_register("kvm-trace", 32, 1, 7 * sizeof(long)); @@ -199,6 +320,8 @@ int kvm_arch_init(void *opaque) return -ENOMEM; } + kvm_s390_cpu_feat_init(); + /* Register floating interrupt controller interface. */ return kvm_register_device_ops(&kvm_flic_ops, KVM_DEV_TYPE_FLIC); } @@ -244,6 +367,7 @@ int kvm_vm_ioctl_check_extension(struct kvm *kvm, long ext) case KVM_CAP_S390_USER_STSI: case KVM_CAP_S390_SKEYS: case KVM_CAP_S390_IRQ_STATE: + case KVM_CAP_S390_USER_INSTR0: r = 1; break; case KVM_CAP_S390_MEM_OP: @@ -251,8 +375,9 @@ int kvm_vm_ioctl_check_extension(struct kvm *kvm, long ext) break; case KVM_CAP_NR_VCPUS: case KVM_CAP_MAX_VCPUS: - r = sclp.has_esca ? KVM_S390_ESCA_CPU_SLOTS - : KVM_S390_BSCA_CPU_SLOTS; + r = KVM_S390_BSCA_CPU_SLOTS; + if (sclp.has_esca && sclp.has_64bscao) + r = KVM_S390_ESCA_CPU_SLOTS; break; case KVM_CAP_NR_MEMSLOTS: r = KVM_USER_MEM_SLOTS; @@ -335,6 +460,16 @@ out: return r; } +static void icpt_operexc_on_all_vcpus(struct kvm *kvm) +{ + unsigned int i; + struct kvm_vcpu *vcpu; + + kvm_for_each_vcpu(i, vcpu, kvm) { + kvm_s390_sync_request(KVM_REQ_ICPT_OPEREXC, vcpu); + } +} + static int kvm_vm_ioctl_enable_cap(struct kvm *kvm, struct kvm_enable_cap *cap) { int r; @@ -355,7 +490,7 @@ static int kvm_vm_ioctl_enable_cap(struct kvm *kvm, struct kvm_enable_cap *cap) break; case KVM_CAP_S390_VECTOR_REGISTERS: mutex_lock(&kvm->lock); - if (atomic_read(&kvm->online_vcpus)) { + if (kvm->created_vcpus) { r = -EBUSY; } else if (MACHINE_HAS_VX) { set_kvm_facility(kvm->arch.model.fac_mask, 129); @@ -370,7 +505,7 @@ static int kvm_vm_ioctl_enable_cap(struct kvm *kvm, struct kvm_enable_cap *cap) case KVM_CAP_S390_RI: r = -EINVAL; mutex_lock(&kvm->lock); - if (atomic_read(&kvm->online_vcpus)) { + if (kvm->created_vcpus) { r = -EBUSY; } else if (test_facility(64)) { set_kvm_facility(kvm->arch.model.fac_mask, 64); @@ -386,6 +521,12 @@ static int kvm_vm_ioctl_enable_cap(struct kvm *kvm, struct kvm_enable_cap *cap) kvm->arch.user_stsi = 1; r = 0; break; + case KVM_CAP_S390_USER_INSTR0: + VM_EVENT(kvm, 3, "%s", "ENABLE: CAP_S390_USER_INSTR0"); + kvm->arch.user_instr0 = 1; + icpt_operexc_on_all_vcpus(kvm); + r = 0; + break; default: r = -EINVAL; break; @@ -418,21 +559,23 @@ static int kvm_s390_set_mem_control(struct kvm *kvm, struct kvm_device_attr *att unsigned int idx; switch (attr->attr) { case KVM_S390_VM_MEM_ENABLE_CMMA: - /* enable CMMA only for z10 and later (EDAT_1) */ - ret = -EINVAL; - if (!MACHINE_IS_LPAR || !MACHINE_HAS_EDAT1) + ret = -ENXIO; + if (!sclp.has_cmma) break; ret = -EBUSY; VM_EVENT(kvm, 3, "%s", "ENABLE: CMMA support"); mutex_lock(&kvm->lock); - if (atomic_read(&kvm->online_vcpus) == 0) { + if (!kvm->created_vcpus) { kvm->arch.use_cmma = 1; ret = 0; } mutex_unlock(&kvm->lock); break; case KVM_S390_VM_MEM_CLR_CMMA: + ret = -ENXIO; + if (!sclp.has_cmma) + break; ret = -EINVAL; if (!kvm->arch.use_cmma) break; @@ -461,20 +604,20 @@ static int kvm_s390_set_mem_control(struct kvm *kvm, struct kvm_device_attr *att if (!new_limit) return -EINVAL; - /* gmap_alloc takes last usable address */ + /* gmap_create takes last usable address */ if (new_limit != KVM_S390_NO_MEM_LIMIT) new_limit -= 1; ret = -EBUSY; mutex_lock(&kvm->lock); - if (atomic_read(&kvm->online_vcpus) == 0) { - /* gmap_alloc will round the limit up */ - struct gmap *new = gmap_alloc(current->mm, new_limit); + if (!kvm->created_vcpus) { + /* gmap_create will round the limit up */ + struct gmap *new = gmap_create(current->mm, new_limit); if (!new) { ret = -ENOMEM; } else { - gmap_free(kvm->arch.gmap); + gmap_remove(kvm->arch.gmap); new->private = kvm; kvm->arch.gmap = new; ret = 0; @@ -644,7 +787,7 @@ static int kvm_s390_set_processor(struct kvm *kvm, struct kvm_device_attr *attr) int ret = 0; mutex_lock(&kvm->lock); - if (atomic_read(&kvm->online_vcpus)) { + if (kvm->created_vcpus) { ret = -EBUSY; goto out; } @@ -676,6 +819,39 @@ out: return ret; } +static int kvm_s390_set_processor_feat(struct kvm *kvm, + struct kvm_device_attr *attr) +{ + struct kvm_s390_vm_cpu_feat data; + int ret = -EBUSY; + + if (copy_from_user(&data, (void __user *)attr->addr, sizeof(data))) + return -EFAULT; + if (!bitmap_subset((unsigned long *) data.feat, + kvm_s390_available_cpu_feat, + KVM_S390_VM_CPU_FEAT_NR_BITS)) + return -EINVAL; + + mutex_lock(&kvm->lock); + if (!atomic_read(&kvm->online_vcpus)) { + bitmap_copy(kvm->arch.cpu_feat, (unsigned long *) data.feat, + KVM_S390_VM_CPU_FEAT_NR_BITS); + ret = 0; + } + mutex_unlock(&kvm->lock); + return ret; +} + +static int kvm_s390_set_processor_subfunc(struct kvm *kvm, + struct kvm_device_attr *attr) +{ + /* + * Once supported by kernel + hw, we have to store the subfunctions + * in kvm->arch and remember that user space configured them. + */ + return -ENXIO; +} + static int kvm_s390_set_cpu_model(struct kvm *kvm, struct kvm_device_attr *attr) { int ret = -ENXIO; @@ -684,6 +860,12 @@ static int kvm_s390_set_cpu_model(struct kvm *kvm, struct kvm_device_attr *attr) case KVM_S390_VM_CPU_PROCESSOR: ret = kvm_s390_set_processor(kvm, attr); break; + case KVM_S390_VM_CPU_PROCESSOR_FEAT: + ret = kvm_s390_set_processor_feat(kvm, attr); + break; + case KVM_S390_VM_CPU_PROCESSOR_SUBFUNC: + ret = kvm_s390_set_processor_subfunc(kvm, attr); + break; } return ret; } @@ -732,6 +914,50 @@ out: return ret; } +static int kvm_s390_get_processor_feat(struct kvm *kvm, + struct kvm_device_attr *attr) +{ + struct kvm_s390_vm_cpu_feat data; + + bitmap_copy((unsigned long *) data.feat, kvm->arch.cpu_feat, + KVM_S390_VM_CPU_FEAT_NR_BITS); + if (copy_to_user((void __user *)attr->addr, &data, sizeof(data))) + return -EFAULT; + return 0; +} + +static int kvm_s390_get_machine_feat(struct kvm *kvm, + struct kvm_device_attr *attr) +{ + struct kvm_s390_vm_cpu_feat data; + + bitmap_copy((unsigned long *) data.feat, + kvm_s390_available_cpu_feat, + KVM_S390_VM_CPU_FEAT_NR_BITS); + if (copy_to_user((void __user *)attr->addr, &data, sizeof(data))) + return -EFAULT; + return 0; +} + +static int kvm_s390_get_processor_subfunc(struct kvm *kvm, + struct kvm_device_attr *attr) +{ + /* + * Once we can actually configure subfunctions (kernel + hw support), + * we have to check if they were already set by user space, if so copy + * them from kvm->arch. + */ + return -ENXIO; +} + +static int kvm_s390_get_machine_subfunc(struct kvm *kvm, + struct kvm_device_attr *attr) +{ + if (copy_to_user((void __user *)attr->addr, &kvm_s390_available_subfunc, + sizeof(struct kvm_s390_vm_cpu_subfunc))) + return -EFAULT; + return 0; +} static int kvm_s390_get_cpu_model(struct kvm *kvm, struct kvm_device_attr *attr) { int ret = -ENXIO; @@ -743,6 +969,18 @@ static int kvm_s390_get_cpu_model(struct kvm *kvm, struct kvm_device_attr *attr) case KVM_S390_VM_CPU_MACHINE: ret = kvm_s390_get_machine(kvm, attr); break; + case KVM_S390_VM_CPU_PROCESSOR_FEAT: + ret = kvm_s390_get_processor_feat(kvm, attr); + break; + case KVM_S390_VM_CPU_MACHINE_FEAT: + ret = kvm_s390_get_machine_feat(kvm, attr); + break; + case KVM_S390_VM_CPU_PROCESSOR_SUBFUNC: + ret = kvm_s390_get_processor_subfunc(kvm, attr); + break; + case KVM_S390_VM_CPU_MACHINE_SUBFUNC: + ret = kvm_s390_get_machine_subfunc(kvm, attr); + break; } return ret; } @@ -803,6 +1041,8 @@ static int kvm_s390_vm_has_attr(struct kvm *kvm, struct kvm_device_attr *attr) switch (attr->attr) { case KVM_S390_VM_MEM_ENABLE_CMMA: case KVM_S390_VM_MEM_CLR_CMMA: + ret = sclp.has_cmma ? 0 : -ENXIO; + break; case KVM_S390_VM_MEM_LIMIT_SIZE: ret = 0; break; @@ -826,8 +1066,13 @@ static int kvm_s390_vm_has_attr(struct kvm *kvm, struct kvm_device_attr *attr) switch (attr->attr) { case KVM_S390_VM_CPU_PROCESSOR: case KVM_S390_VM_CPU_MACHINE: + case KVM_S390_VM_CPU_PROCESSOR_FEAT: + case KVM_S390_VM_CPU_MACHINE_FEAT: + case KVM_S390_VM_CPU_MACHINE_SUBFUNC: ret = 0; break; + /* configuring subfunctions is not supported yet */ + case KVM_S390_VM_CPU_PROCESSOR_SUBFUNC: default: ret = -ENXIO; break; @@ -858,7 +1103,6 @@ static long kvm_s390_get_skeys(struct kvm *kvm, struct kvm_s390_skeys *args) { uint8_t *keys; uint64_t hva; - unsigned long curkey; int i, r = 0; if (args->flags != 0) @@ -879,26 +1123,27 @@ static long kvm_s390_get_skeys(struct kvm *kvm, struct kvm_s390_skeys *args) if (!keys) return -ENOMEM; + down_read(¤t->mm->mmap_sem); for (i = 0; i < args->count; i++) { hva = gfn_to_hva(kvm, args->start_gfn + i); if (kvm_is_error_hva(hva)) { r = -EFAULT; - goto out; + break; } - curkey = get_guest_storage_key(current->mm, hva); - if (IS_ERR_VALUE(curkey)) { - r = curkey; - goto out; - } - keys[i] = curkey; + r = get_guest_storage_key(current->mm, hva, &keys[i]); + if (r) + break; + } + up_read(¤t->mm->mmap_sem); + + if (!r) { + r = copy_to_user((uint8_t __user *)args->skeydata_addr, keys, + sizeof(uint8_t) * args->count); + if (r) + r = -EFAULT; } - r = copy_to_user((uint8_t __user *)args->skeydata_addr, keys, - sizeof(uint8_t) * args->count); - if (r) - r = -EFAULT; -out: kvfree(keys); return r; } @@ -935,24 +1180,25 @@ static long kvm_s390_set_skeys(struct kvm *kvm, struct kvm_s390_skeys *args) if (r) goto out; + down_read(¤t->mm->mmap_sem); for (i = 0; i < args->count; i++) { hva = gfn_to_hva(kvm, args->start_gfn + i); if (kvm_is_error_hva(hva)) { r = -EFAULT; - goto out; + break; } /* Lowest order bit is reserved */ if (keys[i] & 0x01) { r = -EINVAL; - goto out; + break; } - r = set_guest_storage_key(current->mm, hva, - (unsigned long)keys[i], 0); + r = set_guest_storage_key(current->mm, hva, keys[i], 0); if (r) - goto out; + break; } + up_read(¤t->mm->mmap_sem); out: kvfree(keys); return r; @@ -1129,6 +1375,7 @@ static void sca_dispose(struct kvm *kvm) int kvm_arch_init_vm(struct kvm *kvm, unsigned long type) { + gfp_t alloc_flags = GFP_KERNEL; int i, rc; char debug_name[16]; static unsigned long sca_offset; @@ -1150,9 +1397,13 @@ int kvm_arch_init_vm(struct kvm *kvm, unsigned long type) rc = -ENOMEM; + ratelimit_state_init(&kvm->arch.sthyi_limit, 5 * HZ, 500); + kvm->arch.use_esca = 0; /* start with basic SCA */ + if (!sclp.has_64bscao) + alloc_flags |= GFP_DMA; rwlock_init(&kvm->arch.sca_lock); - kvm->arch.sca = (struct bsca_block *) get_zeroed_page(GFP_KERNEL); + kvm->arch.sca = (struct bsca_block *) get_zeroed_page(alloc_flags); if (!kvm->arch.sca) goto out_err; spin_lock(&kvm_lock); @@ -1189,6 +1440,9 @@ int kvm_arch_init_vm(struct kvm *kvm, unsigned long type) memcpy(kvm->arch.model.fac_list, kvm->arch.model.fac_mask, S390_ARCH_FAC_LIST_SIZE_BYTE); + set_kvm_facility(kvm->arch.model.fac_mask, 74); + set_kvm_facility(kvm->arch.model.fac_list, 74); + kvm->arch.model.cpuid = kvm_s390_get_initial_cpuid(); kvm->arch.model.ibc = sclp.ibc & 0x0fff; @@ -1212,7 +1466,7 @@ int kvm_arch_init_vm(struct kvm *kvm, unsigned long type) else kvm->arch.mem_limit = min_t(unsigned long, TASK_MAX_SIZE, sclp.hamax + 1); - kvm->arch.gmap = gmap_alloc(current->mm, kvm->arch.mem_limit - 1); + kvm->arch.gmap = gmap_create(current->mm, kvm->arch.mem_limit - 1); if (!kvm->arch.gmap) goto out_err; kvm->arch.gmap->private = kvm; @@ -1224,6 +1478,7 @@ int kvm_arch_init_vm(struct kvm *kvm, unsigned long type) kvm->arch.epoch = 0; spin_lock_init(&kvm->arch.start_stop_lock); + kvm_s390_vsie_init(kvm); KVM_EVENT(3, "vm 0x%pK created by pid %u", kvm, current->pid); return 0; @@ -1245,7 +1500,7 @@ void kvm_arch_vcpu_destroy(struct kvm_vcpu *vcpu) sca_del_vcpu(vcpu); if (kvm_is_ucontrol(vcpu->kvm)) - gmap_free(vcpu->arch.gmap); + gmap_remove(vcpu->arch.gmap); if (vcpu->kvm->arch.use_cmma) kvm_s390_vcpu_unsetup_cmma(vcpu); @@ -1278,16 +1533,17 @@ void kvm_arch_destroy_vm(struct kvm *kvm) debug_unregister(kvm->arch.dbf); free_page((unsigned long)kvm->arch.sie_page2); if (!kvm_is_ucontrol(kvm)) - gmap_free(kvm->arch.gmap); + gmap_remove(kvm->arch.gmap); kvm_s390_destroy_adapters(kvm); kvm_s390_clear_float_irqs(kvm); + kvm_s390_vsie_destroy(kvm); KVM_EVENT(3, "vm 0x%pK destroyed", kvm); } /* Section: vcpu related */ static int __kvm_ucontrol_vcpu_init(struct kvm_vcpu *vcpu) { - vcpu->arch.gmap = gmap_alloc(current->mm, -1UL); + vcpu->arch.gmap = gmap_create(current->mm, -1UL); if (!vcpu->arch.gmap) return -ENOMEM; vcpu->arch.gmap->private = vcpu->kvm; @@ -1396,7 +1652,7 @@ static int sca_can_add_vcpu(struct kvm *kvm, unsigned int id) if (id < KVM_S390_BSCA_CPU_SLOTS) return true; - if (!sclp.has_esca) + if (!sclp.has_esca || !sclp.has_64bscao) return false; mutex_lock(&kvm->lock); @@ -1537,7 +1793,7 @@ void kvm_arch_vcpu_load(struct kvm_vcpu *vcpu, int cpu) save_access_regs(vcpu->arch.host_acrs); restore_access_regs(vcpu->run->s.regs.acrs); - gmap_enable(vcpu->arch.gmap); + gmap_enable(vcpu->arch.enabled_gmap); atomic_or(CPUSTAT_RUNNING, &vcpu->arch.sie_block->cpuflags); if (vcpu->arch.cputm_enabled && !is_vcpu_idle(vcpu)) __start_cpu_timer_accounting(vcpu); @@ -1550,7 +1806,8 @@ void kvm_arch_vcpu_put(struct kvm_vcpu *vcpu) if (vcpu->arch.cputm_enabled && !is_vcpu_idle(vcpu)) __stop_cpu_timer_accounting(vcpu); atomic_andnot(CPUSTAT_RUNNING, &vcpu->arch.sie_block->cpuflags); - gmap_disable(vcpu->arch.gmap); + vcpu->arch.enabled_gmap = gmap_get_enabled(); + gmap_disable(vcpu->arch.enabled_gmap); /* Save guest register state */ save_fpu_regs(); @@ -1599,7 +1856,10 @@ void kvm_arch_vcpu_postcreate(struct kvm_vcpu *vcpu) vcpu->arch.gmap = vcpu->kvm->arch.gmap; sca_add_vcpu(vcpu); } - + if (test_kvm_facility(vcpu->kvm, 74) || vcpu->kvm->arch.user_instr0) + vcpu->arch.sie_block->ictl |= ICTL_OPEREXC; + /* make vcpu_load load the right gmap on the first trigger */ + vcpu->arch.enabled_gmap = vcpu->arch.gmap; } static void kvm_s390_vcpu_crypto_setup(struct kvm_vcpu *vcpu) @@ -1658,15 +1918,21 @@ int kvm_arch_vcpu_setup(struct kvm_vcpu *vcpu) kvm_s390_vcpu_setup_model(vcpu); - vcpu->arch.sie_block->ecb = 0x02; + /* pgste_set_pte has special handling for !MACHINE_HAS_ESOP */ + if (MACHINE_HAS_ESOP) + vcpu->arch.sie_block->ecb |= 0x02; if (test_kvm_facility(vcpu->kvm, 9)) vcpu->arch.sie_block->ecb |= 0x04; - if (test_kvm_facility(vcpu->kvm, 50) && test_kvm_facility(vcpu->kvm, 73)) + if (test_kvm_facility(vcpu->kvm, 73)) vcpu->arch.sie_block->ecb |= 0x10; - if (test_kvm_facility(vcpu->kvm, 8)) + if (test_kvm_facility(vcpu->kvm, 8) && sclp.has_pfmfi) vcpu->arch.sie_block->ecb2 |= 0x08; - vcpu->arch.sie_block->eca = 0xC1002000U; + vcpu->arch.sie_block->eca = 0x1002000U; + if (sclp.has_cei) + vcpu->arch.sie_block->eca |= 0x80000000U; + if (sclp.has_ib) + vcpu->arch.sie_block->eca |= 0x40000000U; if (sclp.has_siif) vcpu->arch.sie_block->eca |= 1; if (sclp.has_sigpif) @@ -1716,6 +1982,10 @@ struct kvm_vcpu *kvm_arch_vcpu_create(struct kvm *kvm, vcpu->arch.sie_block = &sie_page->sie_block; vcpu->arch.sie_block->itdba = (unsigned long) &sie_page->itdb; + /* the real guest size will always be smaller than msl */ + vcpu->arch.sie_block->mso = 0; + vcpu->arch.sie_block->msl = sclp.hamax; + vcpu->arch.sie_block->icpua = id; spin_lock_init(&vcpu->arch.local_int.lock); vcpu->arch.local_int.float_int = &kvm->arch.float_int; @@ -1784,16 +2054,25 @@ void kvm_s390_sync_request(int req, struct kvm_vcpu *vcpu) kvm_s390_vcpu_request(vcpu); } -static void kvm_gmap_notifier(struct gmap *gmap, unsigned long address) +static void kvm_gmap_notifier(struct gmap *gmap, unsigned long start, + unsigned long end) { - int i; struct kvm *kvm = gmap->private; struct kvm_vcpu *vcpu; + unsigned long prefix; + int i; + if (gmap_is_shadow(gmap)) + return; + if (start >= 1UL << 31) + /* We are only interested in prefix pages */ + return; kvm_for_each_vcpu(i, vcpu, kvm) { /* match against both prefix pages */ - if (kvm_s390_get_prefix(vcpu) == (address & ~0x1000UL)) { - VCPU_EVENT(vcpu, 2, "gmap notifier for %lx", address); + prefix = kvm_s390_get_prefix(vcpu); + if (prefix <= end && start <= prefix + 2*PAGE_SIZE - 1) { + VCPU_EVENT(vcpu, 2, "gmap notifier for %lx-%lx", + start, end); kvm_s390_sync_request(KVM_REQ_MMU_RELOAD, vcpu); } } @@ -2002,6 +2281,8 @@ int kvm_arch_vcpu_ioctl_set_guest_debug(struct kvm_vcpu *vcpu, if (dbg->control & ~VALID_GUESTDBG_FLAGS) return -EINVAL; + if (!sclp.has_gpere) + return -EINVAL; if (dbg->control & KVM_GUESTDBG_ENABLE) { vcpu->guest_debug = dbg->control; @@ -2070,16 +2351,16 @@ retry: return 0; /* * We use MMU_RELOAD just to re-arm the ipte notifier for the - * guest prefix page. gmap_ipte_notify will wait on the ptl lock. + * guest prefix page. gmap_mprotect_notify will wait on the ptl lock. * This ensures that the ipte instruction for this request has * already finished. We might race against a second unmapper that * wants to set the blocking bit. Lets just retry the request loop. */ if (kvm_check_request(KVM_REQ_MMU_RELOAD, vcpu)) { int rc; - rc = gmap_ipte_notify(vcpu->arch.gmap, - kvm_s390_get_prefix(vcpu), - PAGE_SIZE * 2); + rc = gmap_mprotect_notify(vcpu->arch.gmap, + kvm_s390_get_prefix(vcpu), + PAGE_SIZE * 2, PROT_WRITE); if (rc) return rc; goto retry; @@ -2108,6 +2389,11 @@ retry: goto retry; } + if (kvm_check_request(KVM_REQ_ICPT_OPEREXC, vcpu)) { + vcpu->arch.sie_block->ictl |= ICTL_OPEREXC; + goto retry; + } + /* nothing to do, just clear the request */ clear_bit(KVM_REQ_UNHALT, &vcpu->requests); @@ -2362,14 +2648,14 @@ static int __vcpu_run(struct kvm_vcpu *vcpu) * guest_enter and guest_exit should be no uaccess. */ local_irq_disable(); - __kvm_guest_enter(); + guest_enter_irqoff(); __disable_cpu_timer_accounting(vcpu); local_irq_enable(); exit_reason = sie64a(vcpu->arch.sie_block, vcpu->run->s.regs.gprs); local_irq_disable(); __enable_cpu_timer_accounting(vcpu); - __kvm_guest_exit(); + guest_exit_irqoff(); local_irq_enable(); vcpu->srcu_idx = srcu_read_lock(&vcpu->kvm->srcu); @@ -2598,6 +2884,8 @@ static void __disable_ibs_on_all_vcpus(struct kvm *kvm) static void __enable_ibs_on_vcpu(struct kvm_vcpu *vcpu) { + if (!sclp.has_ibs) + return; kvm_check_request(KVM_REQ_DISABLE_IBS, vcpu); kvm_s390_sync_request(KVM_REQ_ENABLE_IBS, vcpu); } diff --git a/arch/s390/kvm/kvm-s390.h b/arch/s390/kvm/kvm-s390.h index 8621ab0..b843286 100644 --- a/arch/s390/kvm/kvm-s390.h +++ b/arch/s390/kvm/kvm-s390.h @@ -56,7 +56,7 @@ static inline int is_vcpu_stopped(struct kvm_vcpu *vcpu) static inline int is_vcpu_idle(struct kvm_vcpu *vcpu) { - return atomic_read(&vcpu->arch.sie_block->cpuflags) & CPUSTAT_WAIT; + return test_bit(vcpu->vcpu_id, vcpu->arch.local_int.float_int->idle_mask); } static inline int kvm_is_ucontrol(struct kvm *kvm) @@ -175,6 +175,12 @@ static inline int set_kvm_facility(u64 *fac_list, unsigned long nr) return 0; } +static inline int test_kvm_cpu_feat(struct kvm *kvm, unsigned long nr) +{ + WARN_ON_ONCE(nr >= KVM_S390_VM_CPU_FEAT_NR_BITS); + return test_bit_inv(nr, kvm->arch.cpu_feat); +} + /* are cpu states controlled by user space */ static inline int kvm_s390_user_cpu_state_ctrl(struct kvm *kvm) { @@ -232,6 +238,8 @@ static inline void kvm_s390_forward_psw(struct kvm_vcpu *vcpu, int ilen) } static inline void kvm_s390_retry_instr(struct kvm_vcpu *vcpu) { + /* don't inject PER events if we re-execute the instruction */ + vcpu->arch.sie_block->icptstatus &= ~0x02; kvm_s390_rewind_psw(vcpu, kvm_s390_get_ilen(vcpu)); } @@ -246,10 +254,21 @@ int kvm_s390_handle_stctl(struct kvm_vcpu *vcpu); int kvm_s390_handle_lctl(struct kvm_vcpu *vcpu); int kvm_s390_handle_eb(struct kvm_vcpu *vcpu); +/* implemented in vsie.c */ +int kvm_s390_handle_vsie(struct kvm_vcpu *vcpu); +void kvm_s390_vsie_kick(struct kvm_vcpu *vcpu); +void kvm_s390_vsie_gmap_notifier(struct gmap *gmap, unsigned long start, + unsigned long end); +void kvm_s390_vsie_init(struct kvm *kvm); +void kvm_s390_vsie_destroy(struct kvm *kvm); + /* implemented in sigp.c */ int kvm_s390_handle_sigp(struct kvm_vcpu *vcpu); int kvm_s390_handle_sigp_pei(struct kvm_vcpu *vcpu); +/* implemented in sthyi.c */ +int handle_sthyi(struct kvm_vcpu *vcpu); + /* implemented in kvm-s390.c */ void kvm_s390_set_tod_clock(struct kvm *kvm, u64 tod); long kvm_arch_fault_in_page(struct kvm_vcpu *vcpu, gpa_t gpa, int writable); @@ -360,6 +379,7 @@ int kvm_s390_import_bp_data(struct kvm_vcpu *vcpu, struct kvm_guest_debug *dbg); void kvm_s390_clear_bp_data(struct kvm_vcpu *vcpu); void kvm_s390_prepare_debug_exit(struct kvm_vcpu *vcpu); +int kvm_s390_handle_per_ifetch_icpt(struct kvm_vcpu *vcpu); void kvm_s390_handle_per_event(struct kvm_vcpu *vcpu); /* support for Basic/Extended SCA handling */ diff --git a/arch/s390/kvm/priv.c b/arch/s390/kvm/priv.c index 95916fa..4616038 100644 --- a/arch/s390/kvm/priv.c +++ b/arch/s390/kvm/priv.c @@ -27,6 +27,7 @@ #include <asm/io.h> #include <asm/ptrace.h> #include <asm/compat.h> +#include <asm/sclp.h> #include "gaccess.h" #include "kvm-s390.h" #include "trace.h" @@ -152,30 +153,166 @@ static int handle_store_cpu_address(struct kvm_vcpu *vcpu) static int __skey_check_enable(struct kvm_vcpu *vcpu) { int rc = 0; + + trace_kvm_s390_skey_related_inst(vcpu); if (!(vcpu->arch.sie_block->ictl & (ICTL_ISKE | ICTL_SSKE | ICTL_RRBE))) return rc; rc = s390_enable_skey(); - VCPU_EVENT(vcpu, 3, "%s", "enabling storage keys for guest"); - trace_kvm_s390_skey_related_inst(vcpu); - vcpu->arch.sie_block->ictl &= ~(ICTL_ISKE | ICTL_SSKE | ICTL_RRBE); + VCPU_EVENT(vcpu, 3, "enabling storage keys for guest: %d", rc); + if (!rc) + vcpu->arch.sie_block->ictl &= ~(ICTL_ISKE | ICTL_SSKE | ICTL_RRBE); return rc; } - -static int handle_skey(struct kvm_vcpu *vcpu) +static int try_handle_skey(struct kvm_vcpu *vcpu) { - int rc = __skey_check_enable(vcpu); + int rc; + vcpu->stat.instruction_storage_key++; + rc = __skey_check_enable(vcpu); if (rc) return rc; - vcpu->stat.instruction_storage_key++; - + if (sclp.has_skey) { + /* with storage-key facility, SIE interprets it for us */ + kvm_s390_retry_instr(vcpu); + VCPU_EVENT(vcpu, 4, "%s", "retrying storage key operation"); + return -EAGAIN; + } if (vcpu->arch.sie_block->gpsw.mask & PSW_MASK_PSTATE) return kvm_s390_inject_program_int(vcpu, PGM_PRIVILEGED_OP); + return 0; +} - kvm_s390_retry_instr(vcpu); - VCPU_EVENT(vcpu, 4, "%s", "retrying storage key operation"); +static int handle_iske(struct kvm_vcpu *vcpu) +{ + unsigned long addr; + unsigned char key; + int reg1, reg2; + int rc; + + rc = try_handle_skey(vcpu); + if (rc) + return rc != -EAGAIN ? rc : 0; + + kvm_s390_get_regs_rre(vcpu, ®1, ®2); + + addr = vcpu->run->s.regs.gprs[reg2] & PAGE_MASK; + addr = kvm_s390_logical_to_effective(vcpu, addr); + addr = kvm_s390_real_to_abs(vcpu, addr); + addr = gfn_to_hva(vcpu->kvm, gpa_to_gfn(addr)); + if (kvm_is_error_hva(addr)) + return kvm_s390_inject_program_int(vcpu, PGM_ADDRESSING); + + down_read(¤t->mm->mmap_sem); + rc = get_guest_storage_key(current->mm, addr, &key); + up_read(¤t->mm->mmap_sem); + if (rc) + return kvm_s390_inject_program_int(vcpu, PGM_ADDRESSING); + vcpu->run->s.regs.gprs[reg1] &= ~0xff; + vcpu->run->s.regs.gprs[reg1] |= key; + return 0; +} + +static int handle_rrbe(struct kvm_vcpu *vcpu) +{ + unsigned long addr; + int reg1, reg2; + int rc; + + rc = try_handle_skey(vcpu); + if (rc) + return rc != -EAGAIN ? rc : 0; + + kvm_s390_get_regs_rre(vcpu, ®1, ®2); + + addr = vcpu->run->s.regs.gprs[reg2] & PAGE_MASK; + addr = kvm_s390_logical_to_effective(vcpu, addr); + addr = kvm_s390_real_to_abs(vcpu, addr); + addr = gfn_to_hva(vcpu->kvm, gpa_to_gfn(addr)); + if (kvm_is_error_hva(addr)) + return kvm_s390_inject_program_int(vcpu, PGM_ADDRESSING); + + down_read(¤t->mm->mmap_sem); + rc = reset_guest_reference_bit(current->mm, addr); + up_read(¤t->mm->mmap_sem); + if (rc < 0) + return kvm_s390_inject_program_int(vcpu, PGM_ADDRESSING); + + kvm_s390_set_psw_cc(vcpu, rc); + return 0; +} + +#define SSKE_NQ 0x8 +#define SSKE_MR 0x4 +#define SSKE_MC 0x2 +#define SSKE_MB 0x1 +static int handle_sske(struct kvm_vcpu *vcpu) +{ + unsigned char m3 = vcpu->arch.sie_block->ipb >> 28; + unsigned long start, end; + unsigned char key, oldkey; + int reg1, reg2; + int rc; + + rc = try_handle_skey(vcpu); + if (rc) + return rc != -EAGAIN ? rc : 0; + + if (!test_kvm_facility(vcpu->kvm, 8)) + m3 &= ~SSKE_MB; + if (!test_kvm_facility(vcpu->kvm, 10)) + m3 &= ~(SSKE_MC | SSKE_MR); + if (!test_kvm_facility(vcpu->kvm, 14)) + m3 &= ~SSKE_NQ; + + kvm_s390_get_regs_rre(vcpu, ®1, ®2); + + key = vcpu->run->s.regs.gprs[reg1] & 0xfe; + start = vcpu->run->s.regs.gprs[reg2] & PAGE_MASK; + start = kvm_s390_logical_to_effective(vcpu, start); + if (m3 & SSKE_MB) { + /* start already designates an absolute address */ + end = (start + (1UL << 20)) & ~((1UL << 20) - 1); + } else { + start = kvm_s390_real_to_abs(vcpu, start); + end = start + PAGE_SIZE; + } + + while (start != end) { + unsigned long addr = gfn_to_hva(vcpu->kvm, gpa_to_gfn(start)); + + if (kvm_is_error_hva(addr)) + return kvm_s390_inject_program_int(vcpu, PGM_ADDRESSING); + + down_read(¤t->mm->mmap_sem); + rc = cond_set_guest_storage_key(current->mm, addr, key, &oldkey, + m3 & SSKE_NQ, m3 & SSKE_MR, + m3 & SSKE_MC); + up_read(¤t->mm->mmap_sem); + if (rc < 0) + return kvm_s390_inject_program_int(vcpu, PGM_ADDRESSING); + start += PAGE_SIZE; + }; + + if (m3 & (SSKE_MC | SSKE_MR)) { + if (m3 & SSKE_MB) { + /* skey in reg1 is unpredictable */ + kvm_s390_set_psw_cc(vcpu, 3); + } else { + kvm_s390_set_psw_cc(vcpu, rc); + vcpu->run->s.regs.gprs[reg1] &= ~0xff00UL; + vcpu->run->s.regs.gprs[reg1] |= (u64) oldkey << 8; + } + } + if (m3 & SSKE_MB) { + if (psw_bits(vcpu->arch.sie_block->gpsw).eaba == PSW_AMODE_64BIT) + vcpu->run->s.regs.gprs[reg2] &= ~PAGE_MASK; + else + vcpu->run->s.regs.gprs[reg2] &= ~0xfffff000UL; + end = kvm_s390_logical_to_effective(vcpu, end); + vcpu->run->s.regs.gprs[reg2] |= end; + } return 0; } @@ -582,10 +719,11 @@ static const intercept_handler_t b2_handlers[256] = { [0x10] = handle_set_prefix, [0x11] = handle_store_prefix, [0x12] = handle_store_cpu_address, + [0x14] = kvm_s390_handle_vsie, [0x21] = handle_ipte_interlock, - [0x29] = handle_skey, - [0x2a] = handle_skey, - [0x2b] = handle_skey, + [0x29] = handle_iske, + [0x2a] = handle_rrbe, + [0x2b] = handle_sske, [0x2c] = handle_test_block, [0x30] = handle_io_inst, [0x31] = handle_io_inst, @@ -654,8 +792,10 @@ static int handle_epsw(struct kvm_vcpu *vcpu) static int handle_pfmf(struct kvm_vcpu *vcpu) { + bool mr = false, mc = false, nq; int reg1, reg2; unsigned long start, end; + unsigned char key; vcpu->stat.instruction_pfmf++; @@ -675,15 +815,27 @@ static int handle_pfmf(struct kvm_vcpu *vcpu) !test_kvm_facility(vcpu->kvm, 14)) return kvm_s390_inject_program_int(vcpu, PGM_SPECIFICATION); - /* No support for conditional-SSKE */ - if (vcpu->run->s.regs.gprs[reg1] & (PFMF_MR | PFMF_MC)) - return kvm_s390_inject_program_int(vcpu, PGM_SPECIFICATION); + /* Only provide conditional-SSKE support if enabled for the guest */ + if (vcpu->run->s.regs.gprs[reg1] & PFMF_SK && + test_kvm_facility(vcpu->kvm, 10)) { + mr = vcpu->run->s.regs.gprs[reg1] & PFMF_MR; + mc = vcpu->run->s.regs.gprs[reg1] & PFMF_MC; + } + nq = vcpu->run->s.regs.gprs[reg1] & PFMF_NQ; + key = vcpu->run->s.regs.gprs[reg1] & PFMF_KEY; start = vcpu->run->s.regs.gprs[reg2] & PAGE_MASK; start = kvm_s390_logical_to_effective(vcpu, start); + if (vcpu->run->s.regs.gprs[reg1] & PFMF_CF) { + if (kvm_s390_check_low_addr_prot_real(vcpu, start)) + return kvm_s390_inject_prog_irq(vcpu, &vcpu->arch.pgm); + } + switch (vcpu->run->s.regs.gprs[reg1] & PFMF_FSC) { case 0x00000000: + /* only 4k frames specify a real address */ + start = kvm_s390_real_to_abs(vcpu, start); end = (start + (1UL << 12)) & ~((1UL << 12) - 1); break; case 0x00001000: @@ -701,20 +853,11 @@ static int handle_pfmf(struct kvm_vcpu *vcpu) return kvm_s390_inject_program_int(vcpu, PGM_SPECIFICATION); } - if (vcpu->run->s.regs.gprs[reg1] & PFMF_CF) { - if (kvm_s390_check_low_addr_prot_real(vcpu, start)) - return kvm_s390_inject_prog_irq(vcpu, &vcpu->arch.pgm); - } - - while (start < end) { - unsigned long useraddr, abs_addr; + while (start != end) { + unsigned long useraddr; /* Translate guest address to host address */ - if ((vcpu->run->s.regs.gprs[reg1] & PFMF_FSC) == 0) - abs_addr = kvm_s390_real_to_abs(vcpu, start); - else - abs_addr = start; - useraddr = gfn_to_hva(vcpu->kvm, gpa_to_gfn(abs_addr)); + useraddr = gfn_to_hva(vcpu->kvm, gpa_to_gfn(start)); if (kvm_is_error_hva(useraddr)) return kvm_s390_inject_program_int(vcpu, PGM_ADDRESSING); @@ -728,16 +871,25 @@ static int handle_pfmf(struct kvm_vcpu *vcpu) if (rc) return rc; - if (set_guest_storage_key(current->mm, useraddr, - vcpu->run->s.regs.gprs[reg1] & PFMF_KEY, - vcpu->run->s.regs.gprs[reg1] & PFMF_NQ)) + down_read(¤t->mm->mmap_sem); + rc = cond_set_guest_storage_key(current->mm, useraddr, + key, NULL, nq, mr, mc); + up_read(¤t->mm->mmap_sem); + if (rc < 0) return kvm_s390_inject_program_int(vcpu, PGM_ADDRESSING); } start += PAGE_SIZE; } - if (vcpu->run->s.regs.gprs[reg1] & PFMF_FSC) - vcpu->run->s.regs.gprs[reg2] = end; + if (vcpu->run->s.regs.gprs[reg1] & PFMF_FSC) { + if (psw_bits(vcpu->arch.sie_block->gpsw).eaba == PSW_AMODE_64BIT) { + vcpu->run->s.regs.gprs[reg2] = end; + } else { + vcpu->run->s.regs.gprs[reg2] &= ~0xffffffffUL; + end = kvm_s390_logical_to_effective(vcpu, end); + vcpu->run->s.regs.gprs[reg2] |= end; + } + } return 0; } @@ -1033,7 +1185,15 @@ static int handle_sckpf(struct kvm_vcpu *vcpu) return 0; } +static int handle_ptff(struct kvm_vcpu *vcpu) +{ + /* we don't emulate any control instructions yet */ + kvm_s390_set_psw_cc(vcpu, 3); + return 0; +} + static const intercept_handler_t x01_handlers[256] = { + [0x04] = handle_ptff, [0x07] = handle_sckpf, }; diff --git a/arch/s390/kvm/sigp.c b/arch/s390/kvm/sigp.c index 28ea0ca..1a252f5 100644 --- a/arch/s390/kvm/sigp.c +++ b/arch/s390/kvm/sigp.c @@ -77,18 +77,18 @@ static int __sigp_conditional_emergency(struct kvm_vcpu *vcpu, const u64 psw_int_mask = PSW_MASK_IO | PSW_MASK_EXT; u16 p_asn, s_asn; psw_t *psw; - u32 flags; + bool idle; - flags = atomic_read(&dst_vcpu->arch.sie_block->cpuflags); + idle = is_vcpu_idle(vcpu); psw = &dst_vcpu->arch.sie_block->gpsw; p_asn = dst_vcpu->arch.sie_block->gcr[4] & 0xffff; /* Primary ASN */ s_asn = dst_vcpu->arch.sie_block->gcr[3] & 0xffff; /* Secondary ASN */ /* Inject the emergency signal? */ - if (!(flags & CPUSTAT_STOPPED) + if (!is_vcpu_stopped(vcpu) || (psw->mask & psw_int_mask) != psw_int_mask - || ((flags & CPUSTAT_WAIT) && psw->addr != 0) - || (!(flags & CPUSTAT_WAIT) && (asn == p_asn || asn == s_asn))) { + || (idle && psw->addr != 0) + || (!idle && (asn == p_asn || asn == s_asn))) { return __inject_sigp_emergency(vcpu, dst_vcpu); } else { *reg &= 0xffffffff00000000UL; diff --git a/arch/s390/kvm/sthyi.c b/arch/s390/kvm/sthyi.c new file mode 100644 index 0000000..bd98b7d --- /dev/null +++ b/arch/s390/kvm/sthyi.c @@ -0,0 +1,471 @@ +/* + * store hypervisor information instruction emulation functions. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License (version 2 only) + * as published by the Free Software Foundation. + * + * Copyright IBM Corp. 2016 + * Author(s): Janosch Frank <frankja@linux.vnet.ibm.com> + */ +#include <linux/kvm_host.h> +#include <linux/errno.h> +#include <linux/pagemap.h> +#include <linux/vmalloc.h> +#include <linux/ratelimit.h> + +#include <asm/kvm_host.h> +#include <asm/asm-offsets.h> +#include <asm/sclp.h> +#include <asm/diag.h> +#include <asm/sysinfo.h> +#include <asm/ebcdic.h> + +#include "kvm-s390.h" +#include "gaccess.h" +#include "trace.h" + +#define DED_WEIGHT 0xffff +/* + * CP and IFL as EBCDIC strings, SP/0x40 determines the end of string + * as they are justified with spaces. + */ +#define CP 0xc3d7404040404040UL +#define IFL 0xc9c6d34040404040UL + +enum hdr_flags { + HDR_NOT_LPAR = 0x10, + HDR_STACK_INCM = 0x20, + HDR_STSI_UNAV = 0x40, + HDR_PERF_UNAV = 0x80, +}; + +enum mac_validity { + MAC_NAME_VLD = 0x20, + MAC_ID_VLD = 0x40, + MAC_CNT_VLD = 0x80, +}; + +enum par_flag { + PAR_MT_EN = 0x80, +}; + +enum par_validity { + PAR_GRP_VLD = 0x08, + PAR_ID_VLD = 0x10, + PAR_ABS_VLD = 0x20, + PAR_WGHT_VLD = 0x40, + PAR_PCNT_VLD = 0x80, +}; + +struct hdr_sctn { + u8 infhflg1; + u8 infhflg2; /* reserved */ + u8 infhval1; /* reserved */ + u8 infhval2; /* reserved */ + u8 reserved[3]; + u8 infhygct; + u16 infhtotl; + u16 infhdln; + u16 infmoff; + u16 infmlen; + u16 infpoff; + u16 infplen; + u16 infhoff1; + u16 infhlen1; + u16 infgoff1; + u16 infglen1; + u16 infhoff2; + u16 infhlen2; + u16 infgoff2; + u16 infglen2; + u16 infhoff3; + u16 infhlen3; + u16 infgoff3; + u16 infglen3; + u8 reserved2[4]; +} __packed; + +struct mac_sctn { + u8 infmflg1; /* reserved */ + u8 infmflg2; /* reserved */ + u8 infmval1; + u8 infmval2; /* reserved */ + u16 infmscps; + u16 infmdcps; + u16 infmsifl; + u16 infmdifl; + char infmname[8]; + char infmtype[4]; + char infmmanu[16]; + char infmseq[16]; + char infmpman[4]; + u8 reserved[4]; +} __packed; + +struct par_sctn { + u8 infpflg1; + u8 infpflg2; /* reserved */ + u8 infpval1; + u8 infpval2; /* reserved */ + u16 infppnum; + u16 infpscps; + u16 infpdcps; + u16 infpsifl; + u16 infpdifl; + u16 reserved; + char infppnam[8]; + u32 infpwbcp; + u32 infpabcp; + u32 infpwbif; + u32 infpabif; + char infplgnm[8]; + u32 infplgcp; + u32 infplgif; +} __packed; + +struct sthyi_sctns { + struct hdr_sctn hdr; + struct mac_sctn mac; + struct par_sctn par; +} __packed; + +struct cpu_inf { + u64 lpar_cap; + u64 lpar_grp_cap; + u64 lpar_weight; + u64 all_weight; + int cpu_num_ded; + int cpu_num_shd; +}; + +struct lpar_cpu_inf { + struct cpu_inf cp; + struct cpu_inf ifl; +}; + +static inline u64 cpu_id(u8 ctidx, void *diag224_buf) +{ + return *((u64 *)(diag224_buf + (ctidx + 1) * DIAG204_CPU_NAME_LEN)); +} + +/* + * Scales the cpu capping from the lpar range to the one expected in + * sthyi data. + * + * diag204 reports a cap in hundredths of processor units. + * z/VM's range for one core is 0 - 0x10000. + */ +static u32 scale_cap(u32 in) +{ + return (0x10000 * in) / 100; +} + +static void fill_hdr(struct sthyi_sctns *sctns) +{ + sctns->hdr.infhdln = sizeof(sctns->hdr); + sctns->hdr.infmoff = sizeof(sctns->hdr); + sctns->hdr.infmlen = sizeof(sctns->mac); + sctns->hdr.infplen = sizeof(sctns->par); + sctns->hdr.infpoff = sctns->hdr.infhdln + sctns->hdr.infmlen; + sctns->hdr.infhtotl = sctns->hdr.infpoff + sctns->hdr.infplen; +} + +static void fill_stsi_mac(struct sthyi_sctns *sctns, + struct sysinfo_1_1_1 *sysinfo) +{ + if (stsi(sysinfo, 1, 1, 1)) + return; + + sclp_ocf_cpc_name_copy(sctns->mac.infmname); + + memcpy(sctns->mac.infmtype, sysinfo->type, sizeof(sctns->mac.infmtype)); + memcpy(sctns->mac.infmmanu, sysinfo->manufacturer, sizeof(sctns->mac.infmmanu)); + memcpy(sctns->mac.infmpman, sysinfo->plant, sizeof(sctns->mac.infmpman)); + memcpy(sctns->mac.infmseq, sysinfo->sequence, sizeof(sctns->mac.infmseq)); + + sctns->mac.infmval1 |= MAC_ID_VLD | MAC_NAME_VLD; +} + +static void fill_stsi_par(struct sthyi_sctns *sctns, + struct sysinfo_2_2_2 *sysinfo) +{ + if (stsi(sysinfo, 2, 2, 2)) + return; + + sctns->par.infppnum = sysinfo->lpar_number; + memcpy(sctns->par.infppnam, sysinfo->name, sizeof(sctns->par.infppnam)); + + sctns->par.infpval1 |= PAR_ID_VLD; +} + +static void fill_stsi(struct sthyi_sctns *sctns) +{ + void *sysinfo; + + /* Errors are handled through the validity bits in the response. */ + sysinfo = (void *)__get_free_page(GFP_KERNEL); + if (!sysinfo) + return; + + fill_stsi_mac(sctns, sysinfo); + fill_stsi_par(sctns, sysinfo); + + free_pages((unsigned long)sysinfo, 0); +} + +static void fill_diag_mac(struct sthyi_sctns *sctns, + struct diag204_x_phys_block *block, + void *diag224_buf) +{ + int i; + + for (i = 0; i < block->hdr.cpus; i++) { + switch (cpu_id(block->cpus[i].ctidx, diag224_buf)) { + case CP: + if (block->cpus[i].weight == DED_WEIGHT) + sctns->mac.infmdcps++; + else + sctns->mac.infmscps++; + break; + case IFL: + if (block->cpus[i].weight == DED_WEIGHT) + sctns->mac.infmdifl++; + else + sctns->mac.infmsifl++; + break; + } + } + sctns->mac.infmval1 |= MAC_CNT_VLD; +} + +/* Returns a pointer to the the next partition block. */ +static struct diag204_x_part_block *lpar_cpu_inf(struct lpar_cpu_inf *part_inf, + bool this_lpar, + void *diag224_buf, + struct diag204_x_part_block *block) +{ + int i, capped = 0, weight_cp = 0, weight_ifl = 0; + struct cpu_inf *cpu_inf; + + for (i = 0; i < block->hdr.rcpus; i++) { + if (!(block->cpus[i].cflag & DIAG204_CPU_ONLINE)) + continue; + + switch (cpu_id(block->cpus[i].ctidx, diag224_buf)) { + case CP: + cpu_inf = &part_inf->cp; + if (block->cpus[i].cur_weight < DED_WEIGHT) + weight_cp |= block->cpus[i].cur_weight; + break; + case IFL: + cpu_inf = &part_inf->ifl; + if (block->cpus[i].cur_weight < DED_WEIGHT) + weight_ifl |= block->cpus[i].cur_weight; + break; + default: + continue; + } + + if (!this_lpar) + continue; + + capped |= block->cpus[i].cflag & DIAG204_CPU_CAPPED; + cpu_inf->lpar_cap |= block->cpus[i].cpu_type_cap; + cpu_inf->lpar_grp_cap |= block->cpus[i].group_cpu_type_cap; + + if (block->cpus[i].weight == DED_WEIGHT) + cpu_inf->cpu_num_ded += 1; + else + cpu_inf->cpu_num_shd += 1; + } + + if (this_lpar && capped) { + part_inf->cp.lpar_weight = weight_cp; + part_inf->ifl.lpar_weight = weight_ifl; + } + part_inf->cp.all_weight += weight_cp; + part_inf->ifl.all_weight += weight_ifl; + return (struct diag204_x_part_block *)&block->cpus[i]; +} + +static void fill_diag(struct sthyi_sctns *sctns) +{ + int i, r, pages; + bool this_lpar; + void *diag204_buf; + void *diag224_buf = NULL; + struct diag204_x_info_blk_hdr *ti_hdr; + struct diag204_x_part_block *part_block; + struct diag204_x_phys_block *phys_block; + struct lpar_cpu_inf lpar_inf = {}; + + /* Errors are handled through the validity bits in the response. */ + pages = diag204((unsigned long)DIAG204_SUBC_RSI | + (unsigned long)DIAG204_INFO_EXT, 0, NULL); + if (pages <= 0) + return; + + diag204_buf = vmalloc(PAGE_SIZE * pages); + if (!diag204_buf) + return; + + r = diag204((unsigned long)DIAG204_SUBC_STIB7 | + (unsigned long)DIAG204_INFO_EXT, pages, diag204_buf); + if (r < 0) + goto out; + + diag224_buf = kmalloc(PAGE_SIZE, GFP_KERNEL | GFP_DMA); + if (!diag224_buf || diag224(diag224_buf)) + goto out; + + ti_hdr = diag204_buf; + part_block = diag204_buf + sizeof(*ti_hdr); + + for (i = 0; i < ti_hdr->npar; i++) { + /* + * For the calling lpar we also need to get the cpu + * caps and weights. The time information block header + * specifies the offset to the partition block of the + * caller lpar, so we know when we process its data. + */ + this_lpar = (void *)part_block - diag204_buf == ti_hdr->this_part; + part_block = lpar_cpu_inf(&lpar_inf, this_lpar, diag224_buf, + part_block); + } + + phys_block = (struct diag204_x_phys_block *)part_block; + part_block = diag204_buf + ti_hdr->this_part; + if (part_block->hdr.mtid) + sctns->par.infpflg1 = PAR_MT_EN; + + sctns->par.infpval1 |= PAR_GRP_VLD; + sctns->par.infplgcp = scale_cap(lpar_inf.cp.lpar_grp_cap); + sctns->par.infplgif = scale_cap(lpar_inf.ifl.lpar_grp_cap); + memcpy(sctns->par.infplgnm, part_block->hdr.hardware_group_name, + sizeof(sctns->par.infplgnm)); + + sctns->par.infpscps = lpar_inf.cp.cpu_num_shd; + sctns->par.infpdcps = lpar_inf.cp.cpu_num_ded; + sctns->par.infpsifl = lpar_inf.ifl.cpu_num_shd; + sctns->par.infpdifl = lpar_inf.ifl.cpu_num_ded; + sctns->par.infpval1 |= PAR_PCNT_VLD; + + sctns->par.infpabcp = scale_cap(lpar_inf.cp.lpar_cap); + sctns->par.infpabif = scale_cap(lpar_inf.ifl.lpar_cap); + sctns->par.infpval1 |= PAR_ABS_VLD; + + /* + * Everything below needs global performance data to be + * meaningful. + */ + if (!(ti_hdr->flags & DIAG204_LPAR_PHYS_FLG)) { + sctns->hdr.infhflg1 |= HDR_PERF_UNAV; + goto out; + } + + fill_diag_mac(sctns, phys_block, diag224_buf); + + if (lpar_inf.cp.lpar_weight) { + sctns->par.infpwbcp = sctns->mac.infmscps * 0x10000 * + lpar_inf.cp.lpar_weight / lpar_inf.cp.all_weight; + } + + if (lpar_inf.ifl.lpar_weight) { + sctns->par.infpwbif = sctns->mac.infmsifl * 0x10000 * + lpar_inf.ifl.lpar_weight / lpar_inf.ifl.all_weight; + } + sctns->par.infpval1 |= PAR_WGHT_VLD; + +out: + kfree(diag224_buf); + vfree(diag204_buf); +} + +static int sthyi(u64 vaddr) +{ + register u64 code asm("0") = 0; + register u64 addr asm("2") = vaddr; + int cc; + + asm volatile( + ".insn rre,0xB2560000,%[code],%[addr]\n" + "ipm %[cc]\n" + "srl %[cc],28\n" + : [cc] "=d" (cc) + : [code] "d" (code), [addr] "a" (addr) + : "memory", "cc"); + return cc; +} + +int handle_sthyi(struct kvm_vcpu *vcpu) +{ + int reg1, reg2, r = 0; + u64 code, addr, cc = 0; + struct sthyi_sctns *sctns = NULL; + + /* + * STHYI requires extensive locking in the higher hypervisors + * and is very computational/memory expensive. Therefore we + * ratelimit the executions per VM. + */ + if (!__ratelimit(&vcpu->kvm->arch.sthyi_limit)) { + kvm_s390_retry_instr(vcpu); + return 0; + } + + kvm_s390_get_regs_rre(vcpu, ®1, ®2); + code = vcpu->run->s.regs.gprs[reg1]; + addr = vcpu->run->s.regs.gprs[reg2]; + + vcpu->stat.instruction_sthyi++; + VCPU_EVENT(vcpu, 3, "STHYI: fc: %llu addr: 0x%016llx", code, addr); + trace_kvm_s390_handle_sthyi(vcpu, code, addr); + + if (reg1 == reg2 || reg1 & 1 || reg2 & 1 || addr & ~PAGE_MASK) + return kvm_s390_inject_program_int(vcpu, PGM_SPECIFICATION); + + if (code & 0xffff) { + cc = 3; + goto out; + } + + /* + * If the page has not yet been faulted in, we want to do that + * now and not after all the expensive calculations. + */ + r = write_guest(vcpu, addr, reg2, &cc, 1); + if (r) + return kvm_s390_inject_prog_cond(vcpu, r); + + sctns = (void *)get_zeroed_page(GFP_KERNEL); + if (!sctns) + return -ENOMEM; + + /* + * If we are a guest, we don't want to emulate an emulated + * instruction. We ask the hypervisor to provide the data. + */ + if (test_facility(74)) { + cc = sthyi((u64)sctns); + goto out; + } + + fill_hdr(sctns); + fill_stsi(sctns); + fill_diag(sctns); + +out: + if (!cc) { + r = write_guest(vcpu, addr, reg2, sctns, PAGE_SIZE); + if (r) { + free_page((unsigned long)sctns); + return kvm_s390_inject_prog_cond(vcpu, r); + } + } + + free_page((unsigned long)sctns); + vcpu->run->s.regs.gprs[reg2 + 1] = cc ? 4 : 0; + kvm_s390_set_psw_cc(vcpu, cc); + return r; +} diff --git a/arch/s390/kvm/trace.h b/arch/s390/kvm/trace.h index 916834d..4fc9d4e 100644 --- a/arch/s390/kvm/trace.h +++ b/arch/s390/kvm/trace.h @@ -41,7 +41,7 @@ TRACE_EVENT(kvm_s390_skey_related_inst, TP_fast_assign( VCPU_ASSIGN_COMMON ), - VCPU_TP_PRINTK("%s", "first instruction related to skeys on vcpu") + VCPU_TP_PRINTK("%s", "storage key related instruction") ); TRACE_EVENT(kvm_s390_major_guest_pfault, @@ -185,8 +185,10 @@ TRACE_EVENT(kvm_s390_intercept_prog, __entry->code = code; ), - VCPU_TP_PRINTK("intercepted program interruption %04x", - __entry->code) + VCPU_TP_PRINTK("intercepted program interruption %04x (%s)", + __entry->code, + __print_symbolic(__entry->code, + icpt_prog_codes)) ); /* @@ -412,6 +414,47 @@ TRACE_EVENT(kvm_s390_handle_stsi, __entry->addr) ); +TRACE_EVENT(kvm_s390_handle_operexc, + TP_PROTO(VCPU_PROTO_COMMON, __u16 ipa, __u32 ipb), + TP_ARGS(VCPU_ARGS_COMMON, ipa, ipb), + + TP_STRUCT__entry( + VCPU_FIELD_COMMON + __field(__u64, instruction) + ), + + TP_fast_assign( + VCPU_ASSIGN_COMMON + __entry->instruction = ((__u64)ipa << 48) | + ((__u64)ipb << 16); + ), + + VCPU_TP_PRINTK("operation exception on instruction %016llx (%s)", + __entry->instruction, + __print_symbolic(icpt_insn_decoder(__entry->instruction), + icpt_insn_codes)) + ); + +TRACE_EVENT(kvm_s390_handle_sthyi, + TP_PROTO(VCPU_PROTO_COMMON, u64 code, u64 addr), + TP_ARGS(VCPU_ARGS_COMMON, code, addr), + + TP_STRUCT__entry( + VCPU_FIELD_COMMON + __field(u64, code) + __field(u64, addr) + ), + + TP_fast_assign( + VCPU_ASSIGN_COMMON + __entry->code = code; + __entry->addr = addr; + ), + + VCPU_TP_PRINTK("STHYI fc: %llu addr: %016llx", + __entry->code, __entry->addr) + ); + #endif /* _TRACE_KVM_H */ /* This part must be outside protection */ diff --git a/arch/s390/kvm/vsie.c b/arch/s390/kvm/vsie.c new file mode 100644 index 0000000..c106488 --- /dev/null +++ b/arch/s390/kvm/vsie.c @@ -0,0 +1,1091 @@ +/* + * kvm nested virtualization support for s390x + * + * Copyright IBM Corp. 2016 + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License (version 2 only) + * as published by the Free Software Foundation. + * + * Author(s): David Hildenbrand <dahi@linux.vnet.ibm.com> + */ +#include <linux/vmalloc.h> +#include <linux/kvm_host.h> +#include <linux/bug.h> +#include <linux/list.h> +#include <linux/bitmap.h> +#include <asm/gmap.h> +#include <asm/mmu_context.h> +#include <asm/sclp.h> +#include <asm/nmi.h> +#include <asm/dis.h> +#include "kvm-s390.h" +#include "gaccess.h" + +struct vsie_page { + struct kvm_s390_sie_block scb_s; /* 0x0000 */ + /* the pinned originial scb */ + struct kvm_s390_sie_block *scb_o; /* 0x0200 */ + /* the shadow gmap in use by the vsie_page */ + struct gmap *gmap; /* 0x0208 */ + /* address of the last reported fault to guest2 */ + unsigned long fault_addr; /* 0x0210 */ + __u8 reserved[0x0700 - 0x0218]; /* 0x0218 */ + struct kvm_s390_crypto_cb crycb; /* 0x0700 */ + __u8 fac[S390_ARCH_FAC_LIST_SIZE_BYTE]; /* 0x0800 */ +} __packed; + +/* trigger a validity icpt for the given scb */ +static int set_validity_icpt(struct kvm_s390_sie_block *scb, + __u16 reason_code) +{ + scb->ipa = 0x1000; + scb->ipb = ((__u32) reason_code) << 16; + scb->icptcode = ICPT_VALIDITY; + return 1; +} + +/* mark the prefix as unmapped, this will block the VSIE */ +static void prefix_unmapped(struct vsie_page *vsie_page) +{ + atomic_or(PROG_REQUEST, &vsie_page->scb_s.prog20); +} + +/* mark the prefix as unmapped and wait until the VSIE has been left */ +static void prefix_unmapped_sync(struct vsie_page *vsie_page) +{ + prefix_unmapped(vsie_page); + if (vsie_page->scb_s.prog0c & PROG_IN_SIE) + atomic_or(CPUSTAT_STOP_INT, &vsie_page->scb_s.cpuflags); + while (vsie_page->scb_s.prog0c & PROG_IN_SIE) + cpu_relax(); +} + +/* mark the prefix as mapped, this will allow the VSIE to run */ +static void prefix_mapped(struct vsie_page *vsie_page) +{ + atomic_andnot(PROG_REQUEST, &vsie_page->scb_s.prog20); +} + +/* test if the prefix is mapped into the gmap shadow */ +static int prefix_is_mapped(struct vsie_page *vsie_page) +{ + return !(atomic_read(&vsie_page->scb_s.prog20) & PROG_REQUEST); +} + +/* copy the updated intervention request bits into the shadow scb */ +static void update_intervention_requests(struct vsie_page *vsie_page) +{ + const int bits = CPUSTAT_STOP_INT | CPUSTAT_IO_INT | CPUSTAT_EXT_INT; + int cpuflags; + + cpuflags = atomic_read(&vsie_page->scb_o->cpuflags); + atomic_andnot(bits, &vsie_page->scb_s.cpuflags); + atomic_or(cpuflags & bits, &vsie_page->scb_s.cpuflags); +} + +/* shadow (filter and validate) the cpuflags */ +static int prepare_cpuflags(struct kvm_vcpu *vcpu, struct vsie_page *vsie_page) +{ + struct kvm_s390_sie_block *scb_s = &vsie_page->scb_s; + struct kvm_s390_sie_block *scb_o = vsie_page->scb_o; + int newflags, cpuflags = atomic_read(&scb_o->cpuflags); + + /* we don't allow ESA/390 guests */ + if (!(cpuflags & CPUSTAT_ZARCH)) + return set_validity_icpt(scb_s, 0x0001U); + + if (cpuflags & (CPUSTAT_RRF | CPUSTAT_MCDS)) + return set_validity_icpt(scb_s, 0x0001U); + else if (cpuflags & (CPUSTAT_SLSV | CPUSTAT_SLSR)) + return set_validity_icpt(scb_s, 0x0007U); + + /* intervention requests will be set later */ + newflags = CPUSTAT_ZARCH; + if (cpuflags & CPUSTAT_GED && test_kvm_facility(vcpu->kvm, 8)) + newflags |= CPUSTAT_GED; + if (cpuflags & CPUSTAT_GED2 && test_kvm_facility(vcpu->kvm, 78)) { + if (cpuflags & CPUSTAT_GED) + return set_validity_icpt(scb_s, 0x0001U); + newflags |= CPUSTAT_GED2; + } + if (test_kvm_cpu_feat(vcpu->kvm, KVM_S390_VM_CPU_FEAT_GPERE)) + newflags |= cpuflags & CPUSTAT_P; + if (test_kvm_cpu_feat(vcpu->kvm, KVM_S390_VM_CPU_FEAT_GSLS)) + newflags |= cpuflags & CPUSTAT_SM; + if (test_kvm_cpu_feat(vcpu->kvm, KVM_S390_VM_CPU_FEAT_IBS)) + newflags |= cpuflags & CPUSTAT_IBS; + + atomic_set(&scb_s->cpuflags, newflags); + return 0; +} + +/* + * Create a shadow copy of the crycb block and setup key wrapping, if + * requested for guest 3 and enabled for guest 2. + * + * We only accept format-1 (no AP in g2), but convert it into format-2 + * There is nothing to do for format-0. + * + * Returns: - 0 if shadowed or nothing to do + * - > 0 if control has to be given to guest 2 + */ +static int shadow_crycb(struct kvm_vcpu *vcpu, struct vsie_page *vsie_page) +{ + struct kvm_s390_sie_block *scb_s = &vsie_page->scb_s; + struct kvm_s390_sie_block *scb_o = vsie_page->scb_o; + u32 crycb_addr = scb_o->crycbd & 0x7ffffff8U; + unsigned long *b1, *b2; + u8 ecb3_flags; + + scb_s->crycbd = 0; + if (!(scb_o->crycbd & vcpu->arch.sie_block->crycbd & CRYCB_FORMAT1)) + return 0; + /* format-1 is supported with message-security-assist extension 3 */ + if (!test_kvm_facility(vcpu->kvm, 76)) + return 0; + /* we may only allow it if enabled for guest 2 */ + ecb3_flags = scb_o->ecb3 & vcpu->arch.sie_block->ecb3 & + (ECB3_AES | ECB3_DEA); + if (!ecb3_flags) + return 0; + + if ((crycb_addr & PAGE_MASK) != ((crycb_addr + 128) & PAGE_MASK)) + return set_validity_icpt(scb_s, 0x003CU); + else if (!crycb_addr) + return set_validity_icpt(scb_s, 0x0039U); + + /* copy only the wrapping keys */ + if (read_guest_real(vcpu, crycb_addr + 72, &vsie_page->crycb, 56)) + return set_validity_icpt(scb_s, 0x0035U); + + scb_s->ecb3 |= ecb3_flags; + scb_s->crycbd = ((__u32)(__u64) &vsie_page->crycb) | CRYCB_FORMAT1 | + CRYCB_FORMAT2; + + /* xor both blocks in one run */ + b1 = (unsigned long *) vsie_page->crycb.dea_wrapping_key_mask; + b2 = (unsigned long *) + vcpu->kvm->arch.crypto.crycb->dea_wrapping_key_mask; + /* as 56%8 == 0, bitmap_xor won't overwrite any data */ + bitmap_xor(b1, b1, b2, BITS_PER_BYTE * 56); + return 0; +} + +/* shadow (round up/down) the ibc to avoid validity icpt */ +static void prepare_ibc(struct kvm_vcpu *vcpu, struct vsie_page *vsie_page) +{ + struct kvm_s390_sie_block *scb_s = &vsie_page->scb_s; + struct kvm_s390_sie_block *scb_o = vsie_page->scb_o; + __u64 min_ibc = (sclp.ibc >> 16) & 0x0fffU; + + scb_s->ibc = 0; + /* ibc installed in g2 and requested for g3 */ + if (vcpu->kvm->arch.model.ibc && (scb_o->ibc & 0x0fffU)) { + scb_s->ibc = scb_o->ibc & 0x0fffU; + /* takte care of the minimum ibc level of the machine */ + if (scb_s->ibc < min_ibc) + scb_s->ibc = min_ibc; + /* take care of the maximum ibc level set for the guest */ + if (scb_s->ibc > vcpu->kvm->arch.model.ibc) + scb_s->ibc = vcpu->kvm->arch.model.ibc; + } +} + +/* unshadow the scb, copying parameters back to the real scb */ +static void unshadow_scb(struct kvm_vcpu *vcpu, struct vsie_page *vsie_page) +{ + struct kvm_s390_sie_block *scb_s = &vsie_page->scb_s; + struct kvm_s390_sie_block *scb_o = vsie_page->scb_o; + + /* interception */ + scb_o->icptcode = scb_s->icptcode; + scb_o->icptstatus = scb_s->icptstatus; + scb_o->ipa = scb_s->ipa; + scb_o->ipb = scb_s->ipb; + scb_o->gbea = scb_s->gbea; + + /* timer */ + scb_o->cputm = scb_s->cputm; + scb_o->ckc = scb_s->ckc; + scb_o->todpr = scb_s->todpr; + + /* guest state */ + scb_o->gpsw = scb_s->gpsw; + scb_o->gg14 = scb_s->gg14; + scb_o->gg15 = scb_s->gg15; + memcpy(scb_o->gcr, scb_s->gcr, 128); + scb_o->pp = scb_s->pp; + + /* interrupt intercept */ + switch (scb_s->icptcode) { + case ICPT_PROGI: + case ICPT_INSTPROGI: + case ICPT_EXTINT: + memcpy((void *)((u64)scb_o + 0xc0), + (void *)((u64)scb_s + 0xc0), 0xf0 - 0xc0); + break; + case ICPT_PARTEXEC: + /* MVPG only */ + memcpy((void *)((u64)scb_o + 0xc0), + (void *)((u64)scb_s + 0xc0), 0xd0 - 0xc0); + break; + } + + if (scb_s->ihcpu != 0xffffU) + scb_o->ihcpu = scb_s->ihcpu; +} + +/* + * Setup the shadow scb by copying and checking the relevant parts of the g2 + * provided scb. + * + * Returns: - 0 if the scb has been shadowed + * - > 0 if control has to be given to guest 2 + */ +static int shadow_scb(struct kvm_vcpu *vcpu, struct vsie_page *vsie_page) +{ + struct kvm_s390_sie_block *scb_o = vsie_page->scb_o; + struct kvm_s390_sie_block *scb_s = &vsie_page->scb_s; + bool had_tx = scb_s->ecb & 0x10U; + unsigned long new_mso = 0; + int rc; + + /* make sure we don't have any leftovers when reusing the scb */ + scb_s->icptcode = 0; + scb_s->eca = 0; + scb_s->ecb = 0; + scb_s->ecb2 = 0; + scb_s->ecb3 = 0; + scb_s->ecd = 0; + scb_s->fac = 0; + + rc = prepare_cpuflags(vcpu, vsie_page); + if (rc) + goto out; + + /* timer */ + scb_s->cputm = scb_o->cputm; + scb_s->ckc = scb_o->ckc; + scb_s->todpr = scb_o->todpr; + scb_s->epoch = scb_o->epoch; + + /* guest state */ + scb_s->gpsw = scb_o->gpsw; + scb_s->gg14 = scb_o->gg14; + scb_s->gg15 = scb_o->gg15; + memcpy(scb_s->gcr, scb_o->gcr, 128); + scb_s->pp = scb_o->pp; + + /* interception / execution handling */ + scb_s->gbea = scb_o->gbea; + scb_s->lctl = scb_o->lctl; + scb_s->svcc = scb_o->svcc; + scb_s->ictl = scb_o->ictl; + /* + * SKEY handling functions can't deal with false setting of PTE invalid + * bits. Therefore we cannot provide interpretation and would later + * have to provide own emulation handlers. + */ + scb_s->ictl |= ICTL_ISKE | ICTL_SSKE | ICTL_RRBE; + scb_s->icpua = scb_o->icpua; + + if (!(atomic_read(&scb_s->cpuflags) & CPUSTAT_SM)) + new_mso = scb_o->mso & 0xfffffffffff00000UL; + /* if the hva of the prefix changes, we have to remap the prefix */ + if (scb_s->mso != new_mso || scb_s->prefix != scb_o->prefix) + prefix_unmapped(vsie_page); + /* SIE will do mso/msl validity and exception checks for us */ + scb_s->msl = scb_o->msl & 0xfffffffffff00000UL; + scb_s->mso = new_mso; + scb_s->prefix = scb_o->prefix; + + /* We have to definetly flush the tlb if this scb never ran */ + if (scb_s->ihcpu != 0xffffU) + scb_s->ihcpu = scb_o->ihcpu; + + /* MVPG and Protection Exception Interpretation are always available */ + scb_s->eca |= scb_o->eca & 0x01002000U; + /* Host-protection-interruption introduced with ESOP */ + if (test_kvm_cpu_feat(vcpu->kvm, KVM_S390_VM_CPU_FEAT_ESOP)) + scb_s->ecb |= scb_o->ecb & 0x02U; + /* transactional execution */ + if (test_kvm_facility(vcpu->kvm, 73)) { + /* remap the prefix is tx is toggled on */ + if ((scb_o->ecb & 0x10U) && !had_tx) + prefix_unmapped(vsie_page); + scb_s->ecb |= scb_o->ecb & 0x10U; + } + /* SIMD */ + if (test_kvm_facility(vcpu->kvm, 129)) { + scb_s->eca |= scb_o->eca & 0x00020000U; + scb_s->ecd |= scb_o->ecd & 0x20000000U; + } + /* Run-time-Instrumentation */ + if (test_kvm_facility(vcpu->kvm, 64)) + scb_s->ecb3 |= scb_o->ecb3 & 0x01U; + if (test_kvm_cpu_feat(vcpu->kvm, KVM_S390_VM_CPU_FEAT_SIIF)) + scb_s->eca |= scb_o->eca & 0x00000001U; + if (test_kvm_cpu_feat(vcpu->kvm, KVM_S390_VM_CPU_FEAT_IB)) + scb_s->eca |= scb_o->eca & 0x40000000U; + if (test_kvm_cpu_feat(vcpu->kvm, KVM_S390_VM_CPU_FEAT_CEI)) + scb_s->eca |= scb_o->eca & 0x80000000U; + + prepare_ibc(vcpu, vsie_page); + rc = shadow_crycb(vcpu, vsie_page); +out: + if (rc) + unshadow_scb(vcpu, vsie_page); + return rc; +} + +void kvm_s390_vsie_gmap_notifier(struct gmap *gmap, unsigned long start, + unsigned long end) +{ + struct kvm *kvm = gmap->private; + struct vsie_page *cur; + unsigned long prefix; + struct page *page; + int i; + + if (!gmap_is_shadow(gmap)) + return; + if (start >= 1UL << 31) + /* We are only interested in prefix pages */ + return; + + /* + * Only new shadow blocks are added to the list during runtime, + * therefore we can safely reference them all the time. + */ + for (i = 0; i < kvm->arch.vsie.page_count; i++) { + page = READ_ONCE(kvm->arch.vsie.pages[i]); + if (!page) + continue; + cur = page_to_virt(page); + if (READ_ONCE(cur->gmap) != gmap) + continue; + prefix = cur->scb_s.prefix << GUEST_PREFIX_SHIFT; + /* with mso/msl, the prefix lies at an offset */ + prefix += cur->scb_s.mso; + if (prefix <= end && start <= prefix + 2 * PAGE_SIZE - 1) + prefix_unmapped_sync(cur); + } +} + +/* + * Map the first prefix page and if tx is enabled also the second prefix page. + * + * The prefix will be protected, a gmap notifier will inform about unmaps. + * The shadow scb must not be executed until the prefix is remapped, this is + * guaranteed by properly handling PROG_REQUEST. + * + * Returns: - 0 on if successfully mapped or already mapped + * - > 0 if control has to be given to guest 2 + * - -EAGAIN if the caller can retry immediately + * - -ENOMEM if out of memory + */ +static int map_prefix(struct kvm_vcpu *vcpu, struct vsie_page *vsie_page) +{ + struct kvm_s390_sie_block *scb_s = &vsie_page->scb_s; + u64 prefix = scb_s->prefix << GUEST_PREFIX_SHIFT; + int rc; + + if (prefix_is_mapped(vsie_page)) + return 0; + + /* mark it as mapped so we can catch any concurrent unmappers */ + prefix_mapped(vsie_page); + + /* with mso/msl, the prefix lies at offset *mso* */ + prefix += scb_s->mso; + + rc = kvm_s390_shadow_fault(vcpu, vsie_page->gmap, prefix); + if (!rc && (scb_s->ecb & 0x10U)) + rc = kvm_s390_shadow_fault(vcpu, vsie_page->gmap, + prefix + PAGE_SIZE); + /* + * We don't have to mprotect, we will be called for all unshadows. + * SIE will detect if protection applies and trigger a validity. + */ + if (rc) + prefix_unmapped(vsie_page); + if (rc > 0 || rc == -EFAULT) + rc = set_validity_icpt(scb_s, 0x0037U); + return rc; +} + +/* + * Pin the guest page given by gpa and set hpa to the pinned host address. + * Will always be pinned writable. + * + * Returns: - 0 on success + * - -EINVAL if the gpa is not valid guest storage + * - -ENOMEM if out of memory + */ +static int pin_guest_page(struct kvm *kvm, gpa_t gpa, hpa_t *hpa) +{ + struct page *page; + hva_t hva; + int rc; + + hva = gfn_to_hva(kvm, gpa_to_gfn(gpa)); + if (kvm_is_error_hva(hva)) + return -EINVAL; + rc = get_user_pages_fast(hva, 1, 1, &page); + if (rc < 0) + return rc; + else if (rc != 1) + return -ENOMEM; + *hpa = (hpa_t) page_to_virt(page) + (gpa & ~PAGE_MASK); + return 0; +} + +/* Unpins a page previously pinned via pin_guest_page, marking it as dirty. */ +static void unpin_guest_page(struct kvm *kvm, gpa_t gpa, hpa_t hpa) +{ + struct page *page; + + page = virt_to_page(hpa); + set_page_dirty_lock(page); + put_page(page); + /* mark the page always as dirty for migration */ + mark_page_dirty(kvm, gpa_to_gfn(gpa)); +} + +/* unpin all blocks previously pinned by pin_blocks(), marking them dirty */ +static void unpin_blocks(struct kvm_vcpu *vcpu, struct vsie_page *vsie_page) +{ + struct kvm_s390_sie_block *scb_o = vsie_page->scb_o; + struct kvm_s390_sie_block *scb_s = &vsie_page->scb_s; + hpa_t hpa; + gpa_t gpa; + + hpa = (u64) scb_s->scaoh << 32 | scb_s->scaol; + if (hpa) { + gpa = scb_o->scaol & ~0xfUL; + if (test_kvm_cpu_feat(vcpu->kvm, KVM_S390_VM_CPU_FEAT_64BSCAO)) + gpa |= (u64) scb_o->scaoh << 32; + unpin_guest_page(vcpu->kvm, gpa, hpa); + scb_s->scaol = 0; + scb_s->scaoh = 0; + } + + hpa = scb_s->itdba; + if (hpa) { + gpa = scb_o->itdba & ~0xffUL; + unpin_guest_page(vcpu->kvm, gpa, hpa); + scb_s->itdba = 0; + } + + hpa = scb_s->gvrd; + if (hpa) { + gpa = scb_o->gvrd & ~0x1ffUL; + unpin_guest_page(vcpu->kvm, gpa, hpa); + scb_s->gvrd = 0; + } + + hpa = scb_s->riccbd; + if (hpa) { + gpa = scb_o->riccbd & ~0x3fUL; + unpin_guest_page(vcpu->kvm, gpa, hpa); + scb_s->riccbd = 0; + } +} + +/* + * Instead of shadowing some blocks, we can simply forward them because the + * addresses in the scb are 64 bit long. + * + * This works as long as the data lies in one page. If blocks ever exceed one + * page, we have to fall back to shadowing. + * + * As we reuse the sca, the vcpu pointers contained in it are invalid. We must + * therefore not enable any facilities that access these pointers (e.g. SIGPIF). + * + * Returns: - 0 if all blocks were pinned. + * - > 0 if control has to be given to guest 2 + * - -ENOMEM if out of memory + */ +static int pin_blocks(struct kvm_vcpu *vcpu, struct vsie_page *vsie_page) +{ + struct kvm_s390_sie_block *scb_o = vsie_page->scb_o; + struct kvm_s390_sie_block *scb_s = &vsie_page->scb_s; + hpa_t hpa; + gpa_t gpa; + int rc = 0; + + gpa = scb_o->scaol & ~0xfUL; + if (test_kvm_cpu_feat(vcpu->kvm, KVM_S390_VM_CPU_FEAT_64BSCAO)) + gpa |= (u64) scb_o->scaoh << 32; + if (gpa) { + if (!(gpa & ~0x1fffUL)) + rc = set_validity_icpt(scb_s, 0x0038U); + else if ((gpa & ~0x1fffUL) == kvm_s390_get_prefix(vcpu)) + rc = set_validity_icpt(scb_s, 0x0011U); + else if ((gpa & PAGE_MASK) != + ((gpa + sizeof(struct bsca_block) - 1) & PAGE_MASK)) + rc = set_validity_icpt(scb_s, 0x003bU); + if (!rc) { + rc = pin_guest_page(vcpu->kvm, gpa, &hpa); + if (rc == -EINVAL) + rc = set_validity_icpt(scb_s, 0x0034U); + } + if (rc) + goto unpin; + scb_s->scaoh = (u32)((u64)hpa >> 32); + scb_s->scaol = (u32)(u64)hpa; + } + + gpa = scb_o->itdba & ~0xffUL; + if (gpa && (scb_s->ecb & 0x10U)) { + if (!(gpa & ~0x1fffU)) { + rc = set_validity_icpt(scb_s, 0x0080U); + goto unpin; + } + /* 256 bytes cannot cross page boundaries */ + rc = pin_guest_page(vcpu->kvm, gpa, &hpa); + if (rc == -EINVAL) + rc = set_validity_icpt(scb_s, 0x0080U); + if (rc) + goto unpin; + scb_s->itdba = hpa; + } + + gpa = scb_o->gvrd & ~0x1ffUL; + if (gpa && (scb_s->eca & 0x00020000U) && + !(scb_s->ecd & 0x20000000U)) { + if (!(gpa & ~0x1fffUL)) { + rc = set_validity_icpt(scb_s, 0x1310U); + goto unpin; + } + /* + * 512 bytes vector registers cannot cross page boundaries + * if this block gets bigger, we have to shadow it. + */ + rc = pin_guest_page(vcpu->kvm, gpa, &hpa); + if (rc == -EINVAL) + rc = set_validity_icpt(scb_s, 0x1310U); + if (rc) + goto unpin; + scb_s->gvrd = hpa; + } + + gpa = scb_o->riccbd & ~0x3fUL; + if (gpa && (scb_s->ecb3 & 0x01U)) { + if (!(gpa & ~0x1fffUL)) { + rc = set_validity_icpt(scb_s, 0x0043U); + goto unpin; + } + /* 64 bytes cannot cross page boundaries */ + rc = pin_guest_page(vcpu->kvm, gpa, &hpa); + if (rc == -EINVAL) + rc = set_validity_icpt(scb_s, 0x0043U); + /* Validity 0x0044 will be checked by SIE */ + if (rc) + goto unpin; + scb_s->gvrd = hpa; + } + return 0; +unpin: + unpin_blocks(vcpu, vsie_page); + return rc; +} + +/* unpin the scb provided by guest 2, marking it as dirty */ +static void unpin_scb(struct kvm_vcpu *vcpu, struct vsie_page *vsie_page, + gpa_t gpa) +{ + hpa_t hpa = (hpa_t) vsie_page->scb_o; + + if (hpa) + unpin_guest_page(vcpu->kvm, gpa, hpa); + vsie_page->scb_o = NULL; +} + +/* + * Pin the scb at gpa provided by guest 2 at vsie_page->scb_o. + * + * Returns: - 0 if the scb was pinned. + * - > 0 if control has to be given to guest 2 + * - -ENOMEM if out of memory + */ +static int pin_scb(struct kvm_vcpu *vcpu, struct vsie_page *vsie_page, + gpa_t gpa) +{ + hpa_t hpa; + int rc; + + rc = pin_guest_page(vcpu->kvm, gpa, &hpa); + if (rc == -EINVAL) { + rc = kvm_s390_inject_program_int(vcpu, PGM_ADDRESSING); + if (!rc) + rc = 1; + } + if (!rc) + vsie_page->scb_o = (struct kvm_s390_sie_block *) hpa; + return rc; +} + +/* + * Inject a fault into guest 2. + * + * Returns: - > 0 if control has to be given to guest 2 + * < 0 if an error occurred during injection. + */ +static int inject_fault(struct kvm_vcpu *vcpu, __u16 code, __u64 vaddr, + bool write_flag) +{ + struct kvm_s390_pgm_info pgm = { + .code = code, + .trans_exc_code = + /* 0-51: virtual address */ + (vaddr & 0xfffffffffffff000UL) | + /* 52-53: store / fetch */ + (((unsigned int) !write_flag) + 1) << 10, + /* 62-63: asce id (alway primary == 0) */ + .exc_access_id = 0, /* always primary */ + .op_access_id = 0, /* not MVPG */ + }; + int rc; + + if (code == PGM_PROTECTION) + pgm.trans_exc_code |= 0x4UL; + + rc = kvm_s390_inject_prog_irq(vcpu, &pgm); + return rc ? rc : 1; +} + +/* + * Handle a fault during vsie execution on a gmap shadow. + * + * Returns: - 0 if the fault was resolved + * - > 0 if control has to be given to guest 2 + * - < 0 if an error occurred + */ +static int handle_fault(struct kvm_vcpu *vcpu, struct vsie_page *vsie_page) +{ + int rc; + + if (current->thread.gmap_int_code == PGM_PROTECTION) + /* we can directly forward all protection exceptions */ + return inject_fault(vcpu, PGM_PROTECTION, + current->thread.gmap_addr, 1); + + rc = kvm_s390_shadow_fault(vcpu, vsie_page->gmap, + current->thread.gmap_addr); + if (rc > 0) { + rc = inject_fault(vcpu, rc, + current->thread.gmap_addr, + current->thread.gmap_write_flag); + if (rc >= 0) + vsie_page->fault_addr = current->thread.gmap_addr; + } + return rc; +} + +/* + * Retry the previous fault that required guest 2 intervention. This avoids + * one superfluous SIE re-entry and direct exit. + * + * Will ignore any errors. The next SIE fault will do proper fault handling. + */ +static void handle_last_fault(struct kvm_vcpu *vcpu, + struct vsie_page *vsie_page) +{ + if (vsie_page->fault_addr) + kvm_s390_shadow_fault(vcpu, vsie_page->gmap, + vsie_page->fault_addr); + vsie_page->fault_addr = 0; +} + +static inline void clear_vsie_icpt(struct vsie_page *vsie_page) +{ + vsie_page->scb_s.icptcode = 0; +} + +/* rewind the psw and clear the vsie icpt, so we can retry execution */ +static void retry_vsie_icpt(struct vsie_page *vsie_page) +{ + struct kvm_s390_sie_block *scb_s = &vsie_page->scb_s; + int ilen = insn_length(scb_s->ipa >> 8); + + /* take care of EXECUTE instructions */ + if (scb_s->icptstatus & 1) { + ilen = (scb_s->icptstatus >> 4) & 0x6; + if (!ilen) + ilen = 4; + } + scb_s->gpsw.addr = __rewind_psw(scb_s->gpsw, ilen); + clear_vsie_icpt(vsie_page); +} + +/* + * Try to shadow + enable the guest 2 provided facility list. + * Retry instruction execution if enabled for and provided by guest 2. + * + * Returns: - 0 if handled (retry or guest 2 icpt) + * - > 0 if control has to be given to guest 2 + */ +static int handle_stfle(struct kvm_vcpu *vcpu, struct vsie_page *vsie_page) +{ + struct kvm_s390_sie_block *scb_s = &vsie_page->scb_s; + __u32 fac = vsie_page->scb_o->fac & 0x7ffffff8U; + + if (fac && test_kvm_facility(vcpu->kvm, 7)) { + retry_vsie_icpt(vsie_page); + if (read_guest_real(vcpu, fac, &vsie_page->fac, + sizeof(vsie_page->fac))) + return set_validity_icpt(scb_s, 0x1090U); + scb_s->fac = (__u32)(__u64) &vsie_page->fac; + } + return 0; +} + +/* + * Run the vsie on a shadow scb and a shadow gmap, without any further + * sanity checks, handling SIE faults. + * + * Returns: - 0 everything went fine + * - > 0 if control has to be given to guest 2 + * - < 0 if an error occurred + */ +static int do_vsie_run(struct kvm_vcpu *vcpu, struct vsie_page *vsie_page) +{ + struct kvm_s390_sie_block *scb_s = &vsie_page->scb_s; + struct kvm_s390_sie_block *scb_o = vsie_page->scb_o; + int rc; + + handle_last_fault(vcpu, vsie_page); + + if (need_resched()) + schedule(); + if (test_cpu_flag(CIF_MCCK_PENDING)) + s390_handle_mcck(); + + srcu_read_unlock(&vcpu->kvm->srcu, vcpu->srcu_idx); + local_irq_disable(); + guest_enter_irqoff(); + local_irq_enable(); + + rc = sie64a(scb_s, vcpu->run->s.regs.gprs); + + local_irq_disable(); + guest_exit_irqoff(); + local_irq_enable(); + vcpu->srcu_idx = srcu_read_lock(&vcpu->kvm->srcu); + + if (rc > 0) + rc = 0; /* we could still have an icpt */ + else if (rc == -EFAULT) + return handle_fault(vcpu, vsie_page); + + switch (scb_s->icptcode) { + case ICPT_INST: + if (scb_s->ipa == 0xb2b0) + rc = handle_stfle(vcpu, vsie_page); + break; + case ICPT_STOP: + /* stop not requested by g2 - must have been a kick */ + if (!(atomic_read(&scb_o->cpuflags) & CPUSTAT_STOP_INT)) + clear_vsie_icpt(vsie_page); + break; + case ICPT_VALIDITY: + if ((scb_s->ipa & 0xf000) != 0xf000) + scb_s->ipa += 0x1000; + break; + } + return rc; +} + +static void release_gmap_shadow(struct vsie_page *vsie_page) +{ + if (vsie_page->gmap) + gmap_put(vsie_page->gmap); + WRITE_ONCE(vsie_page->gmap, NULL); + prefix_unmapped(vsie_page); +} + +static int acquire_gmap_shadow(struct kvm_vcpu *vcpu, + struct vsie_page *vsie_page) +{ + unsigned long asce; + union ctlreg0 cr0; + struct gmap *gmap; + int edat; + + asce = vcpu->arch.sie_block->gcr[1]; + cr0.val = vcpu->arch.sie_block->gcr[0]; + edat = cr0.edat && test_kvm_facility(vcpu->kvm, 8); + edat += edat && test_kvm_facility(vcpu->kvm, 78); + + /* + * ASCE or EDAT could have changed since last icpt, or the gmap + * we're holding has been unshadowed. If the gmap is still valid, + * we can safely reuse it. + */ + if (vsie_page->gmap && gmap_shadow_valid(vsie_page->gmap, asce, edat)) + return 0; + + /* release the old shadow - if any, and mark the prefix as unmapped */ + release_gmap_shadow(vsie_page); + gmap = gmap_shadow(vcpu->arch.gmap, asce, edat); + if (IS_ERR(gmap)) + return PTR_ERR(gmap); + gmap->private = vcpu->kvm; + WRITE_ONCE(vsie_page->gmap, gmap); + return 0; +} + +/* + * Register the shadow scb at the VCPU, e.g. for kicking out of vsie. + */ +static void register_shadow_scb(struct kvm_vcpu *vcpu, + struct vsie_page *vsie_page) +{ + struct kvm_s390_sie_block *scb_s = &vsie_page->scb_s; + + WRITE_ONCE(vcpu->arch.vsie_block, &vsie_page->scb_s); + /* + * External calls have to lead to a kick of the vcpu and + * therefore the vsie -> Simulate Wait state. + */ + atomic_or(CPUSTAT_WAIT, &vcpu->arch.sie_block->cpuflags); + /* + * We have to adjust the g3 epoch by the g2 epoch. The epoch will + * automatically be adjusted on tod clock changes via kvm_sync_clock. + */ + preempt_disable(); + scb_s->epoch += vcpu->kvm->arch.epoch; + preempt_enable(); +} + +/* + * Unregister a shadow scb from a VCPU. + */ +static void unregister_shadow_scb(struct kvm_vcpu *vcpu) +{ + atomic_andnot(CPUSTAT_WAIT, &vcpu->arch.sie_block->cpuflags); + WRITE_ONCE(vcpu->arch.vsie_block, NULL); +} + +/* + * Run the vsie on a shadowed scb, managing the gmap shadow, handling + * prefix pages and faults. + * + * Returns: - 0 if no errors occurred + * - > 0 if control has to be given to guest 2 + * - -ENOMEM if out of memory + */ +static int vsie_run(struct kvm_vcpu *vcpu, struct vsie_page *vsie_page) +{ + struct kvm_s390_sie_block *scb_s = &vsie_page->scb_s; + int rc = 0; + + while (1) { + rc = acquire_gmap_shadow(vcpu, vsie_page); + if (!rc) + rc = map_prefix(vcpu, vsie_page); + if (!rc) { + gmap_enable(vsie_page->gmap); + update_intervention_requests(vsie_page); + rc = do_vsie_run(vcpu, vsie_page); + gmap_enable(vcpu->arch.gmap); + } + atomic_andnot(PROG_BLOCK_SIE, &scb_s->prog20); + + if (rc == -EAGAIN) + rc = 0; + if (rc || scb_s->icptcode || signal_pending(current) || + kvm_s390_vcpu_has_irq(vcpu, 0)) + break; + }; + + if (rc == -EFAULT) { + /* + * Addressing exceptions are always presentes as intercepts. + * As addressing exceptions are suppressing and our guest 3 PSW + * points at the responsible instruction, we have to + * forward the PSW and set the ilc. If we can't read guest 3 + * instruction, we can use an arbitrary ilc. Let's always use + * ilen = 4 for now, so we can avoid reading in guest 3 virtual + * memory. (we could also fake the shadow so the hardware + * handles it). + */ + scb_s->icptcode = ICPT_PROGI; + scb_s->iprcc = PGM_ADDRESSING; + scb_s->pgmilc = 4; + scb_s->gpsw.addr = __rewind_psw(scb_s->gpsw, 4); + } + return rc; +} + +/* + * Get or create a vsie page for a scb address. + * + * Returns: - address of a vsie page (cached or new one) + * - NULL if the same scb address is already used by another VCPU + * - ERR_PTR(-ENOMEM) if out of memory + */ +static struct vsie_page *get_vsie_page(struct kvm *kvm, unsigned long addr) +{ + struct vsie_page *vsie_page; + struct page *page; + int nr_vcpus; + + rcu_read_lock(); + page = radix_tree_lookup(&kvm->arch.vsie.addr_to_page, addr >> 9); + rcu_read_unlock(); + if (page) { + if (page_ref_inc_return(page) == 2) + return page_to_virt(page); + page_ref_dec(page); + } + + /* + * We want at least #online_vcpus shadows, so every VCPU can execute + * the VSIE in parallel. + */ + nr_vcpus = atomic_read(&kvm->online_vcpus); + + mutex_lock(&kvm->arch.vsie.mutex); + if (kvm->arch.vsie.page_count < nr_vcpus) { + page = alloc_page(GFP_KERNEL | __GFP_ZERO | GFP_DMA); + if (!page) { + mutex_unlock(&kvm->arch.vsie.mutex); + return ERR_PTR(-ENOMEM); + } + page_ref_inc(page); + kvm->arch.vsie.pages[kvm->arch.vsie.page_count] = page; + kvm->arch.vsie.page_count++; + } else { + /* reuse an existing entry that belongs to nobody */ + while (true) { + page = kvm->arch.vsie.pages[kvm->arch.vsie.next]; + if (page_ref_inc_return(page) == 2) + break; + page_ref_dec(page); + kvm->arch.vsie.next++; + kvm->arch.vsie.next %= nr_vcpus; + } + radix_tree_delete(&kvm->arch.vsie.addr_to_page, page->index >> 9); + } + page->index = addr; + /* double use of the same address */ + if (radix_tree_insert(&kvm->arch.vsie.addr_to_page, addr >> 9, page)) { + page_ref_dec(page); + mutex_unlock(&kvm->arch.vsie.mutex); + return NULL; + } + mutex_unlock(&kvm->arch.vsie.mutex); + + vsie_page = page_to_virt(page); + memset(&vsie_page->scb_s, 0, sizeof(struct kvm_s390_sie_block)); + release_gmap_shadow(vsie_page); + vsie_page->fault_addr = 0; + vsie_page->scb_s.ihcpu = 0xffffU; + return vsie_page; +} + +/* put a vsie page acquired via get_vsie_page */ +static void put_vsie_page(struct kvm *kvm, struct vsie_page *vsie_page) +{ + struct page *page = pfn_to_page(__pa(vsie_page) >> PAGE_SHIFT); + + page_ref_dec(page); +} + +int kvm_s390_handle_vsie(struct kvm_vcpu *vcpu) +{ + struct vsie_page *vsie_page; + unsigned long scb_addr; + int rc; + + vcpu->stat.instruction_sie++; + if (!test_kvm_cpu_feat(vcpu->kvm, KVM_S390_VM_CPU_FEAT_SIEF2)) + return -EOPNOTSUPP; + if (vcpu->arch.sie_block->gpsw.mask & PSW_MASK_PSTATE) + return kvm_s390_inject_program_int(vcpu, PGM_PRIVILEGED_OP); + + BUILD_BUG_ON(sizeof(struct vsie_page) != 4096); + scb_addr = kvm_s390_get_base_disp_s(vcpu, NULL); + + /* 512 byte alignment */ + if (unlikely(scb_addr & 0x1ffUL)) + return kvm_s390_inject_program_int(vcpu, PGM_SPECIFICATION); + + if (signal_pending(current) || kvm_s390_vcpu_has_irq(vcpu, 0)) + return 0; + + vsie_page = get_vsie_page(vcpu->kvm, scb_addr); + if (IS_ERR(vsie_page)) + return PTR_ERR(vsie_page); + else if (!vsie_page) + /* double use of sie control block - simply do nothing */ + return 0; + + rc = pin_scb(vcpu, vsie_page, scb_addr); + if (rc) + goto out_put; + rc = shadow_scb(vcpu, vsie_page); + if (rc) + goto out_unpin_scb; + rc = pin_blocks(vcpu, vsie_page); + if (rc) + goto out_unshadow; + register_shadow_scb(vcpu, vsie_page); + rc = vsie_run(vcpu, vsie_page); + unregister_shadow_scb(vcpu); + unpin_blocks(vcpu, vsie_page); +out_unshadow: + unshadow_scb(vcpu, vsie_page); +out_unpin_scb: + unpin_scb(vcpu, vsie_page, scb_addr); +out_put: + put_vsie_page(vcpu->kvm, vsie_page); + + return rc < 0 ? rc : 0; +} + +/* Init the vsie data structures. To be called when a vm is initialized. */ +void kvm_s390_vsie_init(struct kvm *kvm) +{ + mutex_init(&kvm->arch.vsie.mutex); + INIT_RADIX_TREE(&kvm->arch.vsie.addr_to_page, GFP_KERNEL); +} + +/* Destroy the vsie data structures. To be called when a vm is destroyed. */ +void kvm_s390_vsie_destroy(struct kvm *kvm) +{ + struct vsie_page *vsie_page; + struct page *page; + int i; + + mutex_lock(&kvm->arch.vsie.mutex); + for (i = 0; i < kvm->arch.vsie.page_count; i++) { + page = kvm->arch.vsie.pages[i]; + kvm->arch.vsie.pages[i] = NULL; + vsie_page = page_to_virt(page); + release_gmap_shadow(vsie_page); + /* free the radix tree entry */ + radix_tree_delete(&kvm->arch.vsie.addr_to_page, page->index >> 9); + __free_page(page); + } + kvm->arch.vsie.page_count = 0; + mutex_unlock(&kvm->arch.vsie.mutex); +} + +void kvm_s390_vsie_kick(struct kvm_vcpu *vcpu) +{ + struct kvm_s390_sie_block *scb = READ_ONCE(vcpu->arch.vsie_block); + + /* + * Even if the VCPU lets go of the shadow sie block reference, it is + * still valid in the cache. So we can safely kick it. + */ + if (scb) { + atomic_or(PROG_BLOCK_SIE, &scb->prog20); + if (scb->prog0c & PROG_IN_SIE) + atomic_or(CPUSTAT_STOP_INT, &scb->cpuflags); + } +} diff --git a/arch/s390/mm/fault.c b/arch/s390/mm/fault.c index 25783dc..a58bca6 100644 --- a/arch/s390/mm/fault.c +++ b/arch/s390/mm/fault.c @@ -418,6 +418,8 @@ static inline int do_exception(struct pt_regs *regs, int access) (struct gmap *) S390_lowcore.gmap : NULL; if (gmap) { current->thread.gmap_addr = address; + current->thread.gmap_write_flag = !!(flags & FAULT_FLAG_WRITE); + current->thread.gmap_int_code = regs->int_code & 0xffff; address = __gmap_translate(gmap, address); if (address == -EFAULT) { fault = VM_FAULT_BADMAP; diff --git a/arch/s390/mm/gmap.c b/arch/s390/mm/gmap.c index 063c721..2ce6bb3 100644 --- a/arch/s390/mm/gmap.c +++ b/arch/s390/mm/gmap.c @@ -20,14 +20,16 @@ #include <asm/gmap.h> #include <asm/tlb.h> +#define GMAP_SHADOW_FAKE_TABLE 1ULL + /** - * gmap_alloc - allocate a guest address space + * gmap_alloc - allocate and initialize a guest address space * @mm: pointer to the parent mm_struct * @limit: maximum address of the gmap address space * * Returns a guest address space structure. */ -struct gmap *gmap_alloc(struct mm_struct *mm, unsigned long limit) +static struct gmap *gmap_alloc(unsigned long limit) { struct gmap *gmap; struct page *page; @@ -55,10 +57,14 @@ struct gmap *gmap_alloc(struct mm_struct *mm, unsigned long limit) if (!gmap) goto out; INIT_LIST_HEAD(&gmap->crst_list); + INIT_LIST_HEAD(&gmap->children); + INIT_LIST_HEAD(&gmap->pt_list); INIT_RADIX_TREE(&gmap->guest_to_host, GFP_KERNEL); INIT_RADIX_TREE(&gmap->host_to_guest, GFP_ATOMIC); + INIT_RADIX_TREE(&gmap->host_to_rmap, GFP_ATOMIC); spin_lock_init(&gmap->guest_table_lock); - gmap->mm = mm; + spin_lock_init(&gmap->shadow_lock); + atomic_set(&gmap->ref_count, 1); page = alloc_pages(GFP_KERNEL, 2); if (!page) goto out_free; @@ -70,9 +76,6 @@ struct gmap *gmap_alloc(struct mm_struct *mm, unsigned long limit) gmap->asce = atype | _ASCE_TABLE_LENGTH | _ASCE_USER_BITS | __pa(table); gmap->asce_end = limit; - down_write(&mm->mmap_sem); - list_add(&gmap->list, &mm->context.gmap_list); - up_write(&mm->mmap_sem); return gmap; out_free: @@ -80,7 +83,28 @@ out_free: out: return NULL; } -EXPORT_SYMBOL_GPL(gmap_alloc); + +/** + * gmap_create - create a guest address space + * @mm: pointer to the parent mm_struct + * @limit: maximum size of the gmap address space + * + * Returns a guest address space structure. + */ +struct gmap *gmap_create(struct mm_struct *mm, unsigned long limit) +{ + struct gmap *gmap; + + gmap = gmap_alloc(limit); + if (!gmap) + return NULL; + gmap->mm = mm; + spin_lock(&mm->context.gmap_lock); + list_add_rcu(&gmap->list, &mm->context.gmap_list); + spin_unlock(&mm->context.gmap_lock); + return gmap; +} +EXPORT_SYMBOL_GPL(gmap_create); static void gmap_flush_tlb(struct gmap *gmap) { @@ -114,31 +138,117 @@ static void gmap_radix_tree_free(struct radix_tree_root *root) } while (nr > 0); } +static void gmap_rmap_radix_tree_free(struct radix_tree_root *root) +{ + struct gmap_rmap *rmap, *rnext, *head; + struct radix_tree_iter iter; + unsigned long indices[16]; + unsigned long index; + void **slot; + int i, nr; + + /* A radix tree is freed by deleting all of its entries */ + index = 0; + do { + nr = 0; + radix_tree_for_each_slot(slot, root, &iter, index) { + indices[nr] = iter.index; + if (++nr == 16) + break; + } + for (i = 0; i < nr; i++) { + index = indices[i]; + head = radix_tree_delete(root, index); + gmap_for_each_rmap_safe(rmap, rnext, head) + kfree(rmap); + } + } while (nr > 0); +} + /** * gmap_free - free a guest address space * @gmap: pointer to the guest address space structure + * + * No locks required. There are no references to this gmap anymore. */ -void gmap_free(struct gmap *gmap) +static void gmap_free(struct gmap *gmap) { struct page *page, *next; - /* Flush tlb. */ - if (MACHINE_HAS_IDTE) - __tlb_flush_idte(gmap->asce); - else - __tlb_flush_global(); - + /* Flush tlb of all gmaps (if not already done for shadows) */ + if (!(gmap_is_shadow(gmap) && gmap->removed)) + gmap_flush_tlb(gmap); /* Free all segment & region tables. */ list_for_each_entry_safe(page, next, &gmap->crst_list, lru) __free_pages(page, 2); gmap_radix_tree_free(&gmap->guest_to_host); gmap_radix_tree_free(&gmap->host_to_guest); - down_write(&gmap->mm->mmap_sem); - list_del(&gmap->list); - up_write(&gmap->mm->mmap_sem); + + /* Free additional data for a shadow gmap */ + if (gmap_is_shadow(gmap)) { + /* Free all page tables. */ + list_for_each_entry_safe(page, next, &gmap->pt_list, lru) + page_table_free_pgste(page); + gmap_rmap_radix_tree_free(&gmap->host_to_rmap); + /* Release reference to the parent */ + gmap_put(gmap->parent); + } + kfree(gmap); } -EXPORT_SYMBOL_GPL(gmap_free); + +/** + * gmap_get - increase reference counter for guest address space + * @gmap: pointer to the guest address space structure + * + * Returns the gmap pointer + */ +struct gmap *gmap_get(struct gmap *gmap) +{ + atomic_inc(&gmap->ref_count); + return gmap; +} +EXPORT_SYMBOL_GPL(gmap_get); + +/** + * gmap_put - decrease reference counter for guest address space + * @gmap: pointer to the guest address space structure + * + * If the reference counter reaches zero the guest address space is freed. + */ +void gmap_put(struct gmap *gmap) +{ + if (atomic_dec_return(&gmap->ref_count) == 0) + gmap_free(gmap); +} +EXPORT_SYMBOL_GPL(gmap_put); + +/** + * gmap_remove - remove a guest address space but do not free it yet + * @gmap: pointer to the guest address space structure + */ +void gmap_remove(struct gmap *gmap) +{ + struct gmap *sg, *next; + + /* Remove all shadow gmaps linked to this gmap */ + if (!list_empty(&gmap->children)) { + spin_lock(&gmap->shadow_lock); + list_for_each_entry_safe(sg, next, &gmap->children, list) { + list_del(&sg->list); + gmap_put(sg); + } + spin_unlock(&gmap->shadow_lock); + } + /* Remove gmap from the pre-mm list */ + spin_lock(&gmap->mm->context.gmap_lock); + list_del_rcu(&gmap->list); + spin_unlock(&gmap->mm->context.gmap_lock); + synchronize_rcu(); + /* Put reference */ + gmap_put(gmap); +} +EXPORT_SYMBOL_GPL(gmap_remove); /** * gmap_enable - switch primary space to the guest address space @@ -160,6 +270,17 @@ void gmap_disable(struct gmap *gmap) } EXPORT_SYMBOL_GPL(gmap_disable); +/** + * gmap_get_enabled - get a pointer to the currently enabled gmap + * + * Returns a pointer to the currently enabled gmap. 0 if none is enabled. + */ +struct gmap *gmap_get_enabled(void) +{ + return (struct gmap *) S390_lowcore.gmap; +} +EXPORT_SYMBOL_GPL(gmap_get_enabled); + /* * gmap_alloc_table is assumed to be called with mmap_sem held */ @@ -175,7 +296,7 @@ static int gmap_alloc_table(struct gmap *gmap, unsigned long *table, return -ENOMEM; new = (unsigned long *) page_to_phys(page); crst_table_init(new, init); - spin_lock(&gmap->mm->page_table_lock); + spin_lock(&gmap->guest_table_lock); if (*table & _REGION_ENTRY_INVALID) { list_add(&page->lru, &gmap->crst_list); *table = (unsigned long) new | _REGION_ENTRY_LENGTH | @@ -183,7 +304,7 @@ static int gmap_alloc_table(struct gmap *gmap, unsigned long *table, page->index = gaddr; page = NULL; } - spin_unlock(&gmap->mm->page_table_lock); + spin_unlock(&gmap->guest_table_lock); if (page) __free_pages(page, 2); return 0; @@ -219,6 +340,7 @@ static int __gmap_unlink_by_vmaddr(struct gmap *gmap, unsigned long vmaddr) unsigned long *entry; int flush = 0; + BUG_ON(gmap_is_shadow(gmap)); spin_lock(&gmap->guest_table_lock); entry = radix_tree_delete(&gmap->host_to_guest, vmaddr >> PMD_SHIFT); if (entry) { @@ -258,6 +380,7 @@ int gmap_unmap_segment(struct gmap *gmap, unsigned long to, unsigned long len) unsigned long off; int flush; + BUG_ON(gmap_is_shadow(gmap)); if ((to | len) & (PMD_SIZE - 1)) return -EINVAL; if (len == 0 || to + len < to) @@ -289,6 +412,7 @@ int gmap_map_segment(struct gmap *gmap, unsigned long from, unsigned long off; int flush; + BUG_ON(gmap_is_shadow(gmap)); if ((from | to | len) & (PMD_SIZE - 1)) return -EINVAL; if (len == 0 || from + len < from || to + len < to || @@ -326,6 +450,8 @@ EXPORT_SYMBOL_GPL(gmap_map_segment); * This function does not establish potentially missing page table entries. * The mmap_sem of the mm that belongs to the address space must be held * when this function gets called. + * + * Note: Can also be called for shadow gmaps. */ unsigned long __gmap_translate(struct gmap *gmap, unsigned long gaddr) { @@ -333,6 +459,7 @@ unsigned long __gmap_translate(struct gmap *gmap, unsigned long gaddr) vmaddr = (unsigned long) radix_tree_lookup(&gmap->guest_to_host, gaddr >> PMD_SHIFT); + /* Note: guest_to_host is empty for a shadow gmap */ return vmaddr ? (vmaddr | (gaddr & ~PMD_MASK)) : -EFAULT; } EXPORT_SYMBOL_GPL(__gmap_translate); @@ -369,11 +496,13 @@ void gmap_unlink(struct mm_struct *mm, unsigned long *table, struct gmap *gmap; int flush; - list_for_each_entry(gmap, &mm->context.gmap_list, list) { + rcu_read_lock(); + list_for_each_entry_rcu(gmap, &mm->context.gmap_list, list) { flush = __gmap_unlink_by_vmaddr(gmap, vmaddr); if (flush) gmap_flush_tlb(gmap); } + rcu_read_unlock(); } /** @@ -397,6 +526,7 @@ int __gmap_link(struct gmap *gmap, unsigned long gaddr, unsigned long vmaddr) pmd_t *pmd; int rc; + BUG_ON(gmap_is_shadow(gmap)); /* Create higher level tables in the gmap page table */ table = gmap->table; if ((gmap->asce & _ASCE_TYPE_MASK) >= _ASCE_TYPE_REGION1) { @@ -552,116 +682,1412 @@ static LIST_HEAD(gmap_notifier_list); static DEFINE_SPINLOCK(gmap_notifier_lock); /** - * gmap_register_ipte_notifier - register a pte invalidation callback + * gmap_register_pte_notifier - register a pte invalidation callback * @nb: pointer to the gmap notifier block */ -void gmap_register_ipte_notifier(struct gmap_notifier *nb) +void gmap_register_pte_notifier(struct gmap_notifier *nb) { spin_lock(&gmap_notifier_lock); - list_add(&nb->list, &gmap_notifier_list); + list_add_rcu(&nb->list, &gmap_notifier_list); spin_unlock(&gmap_notifier_lock); } -EXPORT_SYMBOL_GPL(gmap_register_ipte_notifier); +EXPORT_SYMBOL_GPL(gmap_register_pte_notifier); /** - * gmap_unregister_ipte_notifier - remove a pte invalidation callback + * gmap_unregister_pte_notifier - remove a pte invalidation callback * @nb: pointer to the gmap notifier block */ -void gmap_unregister_ipte_notifier(struct gmap_notifier *nb) +void gmap_unregister_pte_notifier(struct gmap_notifier *nb) { spin_lock(&gmap_notifier_lock); - list_del_init(&nb->list); + list_del_rcu(&nb->list); spin_unlock(&gmap_notifier_lock); + synchronize_rcu(); +} +EXPORT_SYMBOL_GPL(gmap_unregister_pte_notifier); + +/** + * gmap_call_notifier - call all registered invalidation callbacks + * @gmap: pointer to guest mapping meta data structure + * @start: start virtual address in the guest address space + * @end: end virtual address in the guest address space + */ +static void gmap_call_notifier(struct gmap *gmap, unsigned long start, + unsigned long end) +{ + struct gmap_notifier *nb; + + list_for_each_entry(nb, &gmap_notifier_list, list) + nb->notifier_call(gmap, start, end); +} + +/** + * gmap_table_walk - walk the gmap page tables + * @gmap: pointer to guest mapping meta data structure + * @gaddr: virtual address in the guest address space + * @level: page table level to stop at + * + * Returns a table entry pointer for the given guest address and @level + * @level=0 : returns a pointer to a page table table entry (or NULL) + * @level=1 : returns a pointer to a segment table entry (or NULL) + * @level=2 : returns a pointer to a region-3 table entry (or NULL) + * @level=3 : returns a pointer to a region-2 table entry (or NULL) + * @level=4 : returns a pointer to a region-1 table entry (or NULL) + * + * Returns NULL if the gmap page tables could not be walked to the + * requested level. + * + * Note: Can also be called for shadow gmaps. + */ +static inline unsigned long *gmap_table_walk(struct gmap *gmap, + unsigned long gaddr, int level) +{ + unsigned long *table; + + if ((gmap->asce & _ASCE_TYPE_MASK) + 4 < (level * 4)) + return NULL; + if (gmap_is_shadow(gmap) && gmap->removed) + return NULL; + if (gaddr & (-1UL << (31 + ((gmap->asce & _ASCE_TYPE_MASK) >> 2)*11))) + return NULL; + table = gmap->table; + switch (gmap->asce & _ASCE_TYPE_MASK) { + case _ASCE_TYPE_REGION1: + table += (gaddr >> 53) & 0x7ff; + if (level == 4) + break; + if (*table & _REGION_ENTRY_INVALID) + return NULL; + table = (unsigned long *)(*table & _REGION_ENTRY_ORIGIN); + /* Fallthrough */ + case _ASCE_TYPE_REGION2: + table += (gaddr >> 42) & 0x7ff; + if (level == 3) + break; + if (*table & _REGION_ENTRY_INVALID) + return NULL; + table = (unsigned long *)(*table & _REGION_ENTRY_ORIGIN); + /* Fallthrough */ + case _ASCE_TYPE_REGION3: + table += (gaddr >> 31) & 0x7ff; + if (level == 2) + break; + if (*table & _REGION_ENTRY_INVALID) + return NULL; + table = (unsigned long *)(*table & _REGION_ENTRY_ORIGIN); + /* Fallthrough */ + case _ASCE_TYPE_SEGMENT: + table += (gaddr >> 20) & 0x7ff; + if (level == 1) + break; + if (*table & _REGION_ENTRY_INVALID) + return NULL; + table = (unsigned long *)(*table & _SEGMENT_ENTRY_ORIGIN); + table += (gaddr >> 12) & 0xff; + } + return table; +} + +/** + * gmap_pte_op_walk - walk the gmap page table, get the page table lock + * and return the pte pointer + * @gmap: pointer to guest mapping meta data structure + * @gaddr: virtual address in the guest address space + * @ptl: pointer to the spinlock pointer + * + * Returns a pointer to the locked pte for a guest address, or NULL + * + * Note: Can also be called for shadow gmaps. + */ +static pte_t *gmap_pte_op_walk(struct gmap *gmap, unsigned long gaddr, + spinlock_t **ptl) +{ + unsigned long *table; + + if (gmap_is_shadow(gmap)) + spin_lock(&gmap->guest_table_lock); + /* Walk the gmap page table, lock and get pte pointer */ + table = gmap_table_walk(gmap, gaddr, 1); /* get segment pointer */ + if (!table || *table & _SEGMENT_ENTRY_INVALID) { + if (gmap_is_shadow(gmap)) + spin_unlock(&gmap->guest_table_lock); + return NULL; + } + if (gmap_is_shadow(gmap)) { + *ptl = &gmap->guest_table_lock; + return pte_offset_map((pmd_t *) table, gaddr); + } + return pte_alloc_map_lock(gmap->mm, (pmd_t *) table, gaddr, ptl); +} + +/** + * gmap_pte_op_fixup - force a page in and connect the gmap page table + * @gmap: pointer to guest mapping meta data structure + * @gaddr: virtual address in the guest address space + * @vmaddr: address in the host process address space + * @prot: indicates access rights: PROT_NONE, PROT_READ or PROT_WRITE + * + * Returns 0 if the caller can retry __gmap_translate (might fail again), + * -ENOMEM if out of memory and -EFAULT if anything goes wrong while fixing + * up or connecting the gmap page table. + */ +static int gmap_pte_op_fixup(struct gmap *gmap, unsigned long gaddr, + unsigned long vmaddr, int prot) +{ + struct mm_struct *mm = gmap->mm; + unsigned int fault_flags; + bool unlocked = false; + + BUG_ON(gmap_is_shadow(gmap)); + fault_flags = (prot == PROT_WRITE) ? FAULT_FLAG_WRITE : 0; + if (fixup_user_fault(current, mm, vmaddr, fault_flags, &unlocked)) + return -EFAULT; + if (unlocked) + /* lost mmap_sem, caller has to retry __gmap_translate */ + return 0; + /* Connect the page tables */ + return __gmap_link(gmap, gaddr, vmaddr); } -EXPORT_SYMBOL_GPL(gmap_unregister_ipte_notifier); /** - * gmap_ipte_notify - mark a range of ptes for invalidation notification + * gmap_pte_op_end - release the page table lock + * @ptl: pointer to the spinlock pointer + */ +static void gmap_pte_op_end(spinlock_t *ptl) +{ + spin_unlock(ptl); +} + +/* + * gmap_protect_range - remove access rights to memory and set pgste bits * @gmap: pointer to guest mapping meta data structure * @gaddr: virtual address in the guest address space * @len: size of area + * @prot: indicates access rights: PROT_NONE, PROT_READ or PROT_WRITE + * @bits: pgste notification bits to set * - * Returns 0 if for each page in the given range a gmap mapping exists and - * the invalidation notification could be set. If the gmap mapping is missing - * for one or more pages -EFAULT is returned. If no memory could be allocated - * -ENOMEM is returned. This function establishes missing page table entries. + * Returns 0 if successfully protected, -ENOMEM if out of memory and + * -EFAULT if gaddr is invalid (or mapping for shadows is missing). + * + * Called with sg->mm->mmap_sem in read. + * + * Note: Can also be called for shadow gmaps. */ -int gmap_ipte_notify(struct gmap *gmap, unsigned long gaddr, unsigned long len) +static int gmap_protect_range(struct gmap *gmap, unsigned long gaddr, + unsigned long len, int prot, unsigned long bits) { - unsigned long addr; + unsigned long vmaddr; spinlock_t *ptl; pte_t *ptep; - bool unlocked; - int rc = 0; + int rc; + + while (len) { + rc = -EAGAIN; + ptep = gmap_pte_op_walk(gmap, gaddr, &ptl); + if (ptep) { + rc = ptep_force_prot(gmap->mm, gaddr, ptep, prot, bits); + gmap_pte_op_end(ptl); + } + if (rc) { + vmaddr = __gmap_translate(gmap, gaddr); + if (IS_ERR_VALUE(vmaddr)) + return vmaddr; + rc = gmap_pte_op_fixup(gmap, gaddr, vmaddr, prot); + if (rc) + return rc; + continue; + } + gaddr += PAGE_SIZE; + len -= PAGE_SIZE; + } + return 0; +} + +/** + * gmap_mprotect_notify - change access rights for a range of ptes and + * call the notifier if any pte changes again + * @gmap: pointer to guest mapping meta data structure + * @gaddr: virtual address in the guest address space + * @len: size of area + * @prot: indicates access rights: PROT_NONE, PROT_READ or PROT_WRITE + * + * Returns 0 if for each page in the given range a gmap mapping exists, + * the new access rights could be set and the notifier could be armed. + * If the gmap mapping is missing for one or more pages -EFAULT is + * returned. If no memory could be allocated -ENOMEM is returned. + * This function establishes missing page table entries. + */ +int gmap_mprotect_notify(struct gmap *gmap, unsigned long gaddr, + unsigned long len, int prot) +{ + int rc; - if ((gaddr & ~PAGE_MASK) || (len & ~PAGE_MASK)) + if ((gaddr & ~PAGE_MASK) || (len & ~PAGE_MASK) || gmap_is_shadow(gmap)) + return -EINVAL; + if (!MACHINE_HAS_ESOP && prot == PROT_READ) return -EINVAL; down_read(&gmap->mm->mmap_sem); - while (len) { - unlocked = false; - /* Convert gmap address and connect the page tables */ - addr = __gmap_translate(gmap, gaddr); - if (IS_ERR_VALUE(addr)) { - rc = addr; + rc = gmap_protect_range(gmap, gaddr, len, prot, PGSTE_IN_BIT); + up_read(&gmap->mm->mmap_sem); + return rc; +} +EXPORT_SYMBOL_GPL(gmap_mprotect_notify); + +/** + * gmap_read_table - get an unsigned long value from a guest page table using + * absolute addressing, without marking the page referenced. + * @gmap: pointer to guest mapping meta data structure + * @gaddr: virtual address in the guest address space + * @val: pointer to the unsigned long value to return + * + * Returns 0 if the value was read, -ENOMEM if out of memory and -EFAULT + * if reading using the virtual address failed. + * + * Called with gmap->mm->mmap_sem in read. + */ +int gmap_read_table(struct gmap *gmap, unsigned long gaddr, unsigned long *val) +{ + unsigned long address, vmaddr; + spinlock_t *ptl; + pte_t *ptep, pte; + int rc; + + while (1) { + rc = -EAGAIN; + ptep = gmap_pte_op_walk(gmap, gaddr, &ptl); + if (ptep) { + pte = *ptep; + if (pte_present(pte) && (pte_val(pte) & _PAGE_READ)) { + address = pte_val(pte) & PAGE_MASK; + address += gaddr & ~PAGE_MASK; + *val = *(unsigned long *) address; + pte_val(*ptep) |= _PAGE_YOUNG; + /* Do *NOT* clear the _PAGE_INVALID bit! */ + rc = 0; + } + gmap_pte_op_end(ptl); + } + if (!rc) + break; + vmaddr = __gmap_translate(gmap, gaddr); + if (IS_ERR_VALUE(vmaddr)) { + rc = vmaddr; break; } - /* Get the page mapped */ - if (fixup_user_fault(current, gmap->mm, addr, FAULT_FLAG_WRITE, - &unlocked)) { - rc = -EFAULT; + rc = gmap_pte_op_fixup(gmap, gaddr, vmaddr, PROT_READ); + if (rc) break; + } + return rc; +} +EXPORT_SYMBOL_GPL(gmap_read_table); + +/** + * gmap_insert_rmap - add a rmap to the host_to_rmap radix tree + * @sg: pointer to the shadow guest address space structure + * @vmaddr: vm address associated with the rmap + * @rmap: pointer to the rmap structure + * + * Called with the sg->guest_table_lock + */ +static inline void gmap_insert_rmap(struct gmap *sg, unsigned long vmaddr, + struct gmap_rmap *rmap) +{ + void **slot; + + BUG_ON(!gmap_is_shadow(sg)); + slot = radix_tree_lookup_slot(&sg->host_to_rmap, vmaddr >> PAGE_SHIFT); + if (slot) { + rmap->next = radix_tree_deref_slot_protected(slot, + &sg->guest_table_lock); + radix_tree_replace_slot(slot, rmap); + } else { + rmap->next = NULL; + radix_tree_insert(&sg->host_to_rmap, vmaddr >> PAGE_SHIFT, + rmap); + } +} + +/** + * gmap_protect_rmap - modify access rights to memory and create an rmap + * @sg: pointer to the shadow guest address space structure + * @raddr: rmap address in the shadow gmap + * @paddr: address in the parent guest address space + * @len: length of the memory area to protect + * @prot: indicates access rights: none, read-only or read-write + * + * Returns 0 if successfully protected and the rmap was created, -ENOMEM + * if out of memory and -EFAULT if paddr is invalid. + */ +static int gmap_protect_rmap(struct gmap *sg, unsigned long raddr, + unsigned long paddr, unsigned long len, int prot) +{ + struct gmap *parent; + struct gmap_rmap *rmap; + unsigned long vmaddr; + spinlock_t *ptl; + pte_t *ptep; + int rc; + + BUG_ON(!gmap_is_shadow(sg)); + parent = sg->parent; + while (len) { + vmaddr = __gmap_translate(parent, paddr); + if (IS_ERR_VALUE(vmaddr)) + return vmaddr; + rmap = kzalloc(sizeof(*rmap), GFP_KERNEL); + if (!rmap) + return -ENOMEM; + rmap->raddr = raddr; + rc = radix_tree_preload(GFP_KERNEL); + if (rc) { + kfree(rmap); + return rc; + } + rc = -EAGAIN; + ptep = gmap_pte_op_walk(parent, paddr, &ptl); + if (ptep) { + spin_lock(&sg->guest_table_lock); + rc = ptep_force_prot(parent->mm, paddr, ptep, prot, + PGSTE_VSIE_BIT); + if (!rc) + gmap_insert_rmap(sg, vmaddr, rmap); + spin_unlock(&sg->guest_table_lock); + gmap_pte_op_end(ptl); } - /* While trying to map mmap_sem got unlocked. Let us retry */ - if (unlocked) + radix_tree_preload_end(); + if (rc) { + kfree(rmap); + rc = gmap_pte_op_fixup(parent, paddr, vmaddr, prot); + if (rc) + return rc; continue; - rc = __gmap_link(gmap, gaddr, addr); + } + paddr += PAGE_SIZE; + len -= PAGE_SIZE; + } + return 0; +} + +#define _SHADOW_RMAP_MASK 0x7 +#define _SHADOW_RMAP_REGION1 0x5 +#define _SHADOW_RMAP_REGION2 0x4 +#define _SHADOW_RMAP_REGION3 0x3 +#define _SHADOW_RMAP_SEGMENT 0x2 +#define _SHADOW_RMAP_PGTABLE 0x1 + +/** + * gmap_idte_one - invalidate a single region or segment table entry + * @asce: region or segment table *origin* + table-type bits + * @vaddr: virtual address to identify the table entry to flush + * + * The invalid bit of a single region or segment table entry is set + * and the associated TLB entries depending on the entry are flushed. + * The table-type of the @asce identifies the portion of the @vaddr + * that is used as the invalidation index. + */ +static inline void gmap_idte_one(unsigned long asce, unsigned long vaddr) +{ + asm volatile( + " .insn rrf,0xb98e0000,%0,%1,0,0" + : : "a" (asce), "a" (vaddr) : "cc", "memory"); +} + +/** + * gmap_unshadow_page - remove a page from a shadow page table + * @sg: pointer to the shadow guest address space structure + * @raddr: rmap address in the shadow guest address space + * + * Called with the sg->guest_table_lock + */ +static void gmap_unshadow_page(struct gmap *sg, unsigned long raddr) +{ + unsigned long *table; + + BUG_ON(!gmap_is_shadow(sg)); + table = gmap_table_walk(sg, raddr, 0); /* get page table pointer */ + if (!table || *table & _PAGE_INVALID) + return; + gmap_call_notifier(sg, raddr, raddr + (1UL << 12) - 1); + ptep_unshadow_pte(sg->mm, raddr, (pte_t *) table); +} + +/** + * __gmap_unshadow_pgt - remove all entries from a shadow page table + * @sg: pointer to the shadow guest address space structure + * @raddr: rmap address in the shadow guest address space + * @pgt: pointer to the start of a shadow page table + * + * Called with the sg->guest_table_lock + */ +static void __gmap_unshadow_pgt(struct gmap *sg, unsigned long raddr, + unsigned long *pgt) +{ + int i; + + BUG_ON(!gmap_is_shadow(sg)); + for (i = 0; i < 256; i++, raddr += 1UL << 12) + pgt[i] = _PAGE_INVALID; +} + +/** + * gmap_unshadow_pgt - remove a shadow page table from a segment entry + * @sg: pointer to the shadow guest address space structure + * @raddr: address in the shadow guest address space + * + * Called with the sg->guest_table_lock + */ +static void gmap_unshadow_pgt(struct gmap *sg, unsigned long raddr) +{ + unsigned long sto, *ste, *pgt; + struct page *page; + + BUG_ON(!gmap_is_shadow(sg)); + ste = gmap_table_walk(sg, raddr, 1); /* get segment pointer */ + if (!ste || !(*ste & _SEGMENT_ENTRY_ORIGIN)) + return; + gmap_call_notifier(sg, raddr, raddr + (1UL << 20) - 1); + sto = (unsigned long) (ste - ((raddr >> 20) & 0x7ff)); + gmap_idte_one(sto | _ASCE_TYPE_SEGMENT, raddr); + pgt = (unsigned long *)(*ste & _SEGMENT_ENTRY_ORIGIN); + *ste = _SEGMENT_ENTRY_EMPTY; + __gmap_unshadow_pgt(sg, raddr, pgt); + /* Free page table */ + page = pfn_to_page(__pa(pgt) >> PAGE_SHIFT); + list_del(&page->lru); + page_table_free_pgste(page); +} + +/** + * __gmap_unshadow_sgt - remove all entries from a shadow segment table + * @sg: pointer to the shadow guest address space structure + * @raddr: rmap address in the shadow guest address space + * @sgt: pointer to the start of a shadow segment table + * + * Called with the sg->guest_table_lock + */ +static void __gmap_unshadow_sgt(struct gmap *sg, unsigned long raddr, + unsigned long *sgt) +{ + unsigned long asce, *pgt; + struct page *page; + int i; + + BUG_ON(!gmap_is_shadow(sg)); + asce = (unsigned long) sgt | _ASCE_TYPE_SEGMENT; + for (i = 0; i < 2048; i++, raddr += 1UL << 20) { + if (!(sgt[i] & _SEGMENT_ENTRY_ORIGIN)) + continue; + pgt = (unsigned long *)(sgt[i] & _REGION_ENTRY_ORIGIN); + sgt[i] = _SEGMENT_ENTRY_EMPTY; + __gmap_unshadow_pgt(sg, raddr, pgt); + /* Free page table */ + page = pfn_to_page(__pa(pgt) >> PAGE_SHIFT); + list_del(&page->lru); + page_table_free_pgste(page); + } +} + +/** + * gmap_unshadow_sgt - remove a shadow segment table from a region-3 entry + * @sg: pointer to the shadow guest address space structure + * @raddr: rmap address in the shadow guest address space + * + * Called with the shadow->guest_table_lock + */ +static void gmap_unshadow_sgt(struct gmap *sg, unsigned long raddr) +{ + unsigned long r3o, *r3e, *sgt; + struct page *page; + + BUG_ON(!gmap_is_shadow(sg)); + r3e = gmap_table_walk(sg, raddr, 2); /* get region-3 pointer */ + if (!r3e || !(*r3e & _REGION_ENTRY_ORIGIN)) + return; + gmap_call_notifier(sg, raddr, raddr + (1UL << 31) - 1); + r3o = (unsigned long) (r3e - ((raddr >> 31) & 0x7ff)); + gmap_idte_one(r3o | _ASCE_TYPE_REGION3, raddr); + sgt = (unsigned long *)(*r3e & _REGION_ENTRY_ORIGIN); + *r3e = _REGION3_ENTRY_EMPTY; + __gmap_unshadow_sgt(sg, raddr, sgt); + /* Free segment table */ + page = pfn_to_page(__pa(sgt) >> PAGE_SHIFT); + list_del(&page->lru); + __free_pages(page, 2); +} + +/** + * __gmap_unshadow_r3t - remove all entries from a shadow region-3 table + * @sg: pointer to the shadow guest address space structure + * @raddr: address in the shadow guest address space + * @r3t: pointer to the start of a shadow region-3 table + * + * Called with the sg->guest_table_lock + */ +static void __gmap_unshadow_r3t(struct gmap *sg, unsigned long raddr, + unsigned long *r3t) +{ + unsigned long asce, *sgt; + struct page *page; + int i; + + BUG_ON(!gmap_is_shadow(sg)); + asce = (unsigned long) r3t | _ASCE_TYPE_REGION3; + for (i = 0; i < 2048; i++, raddr += 1UL << 31) { + if (!(r3t[i] & _REGION_ENTRY_ORIGIN)) + continue; + sgt = (unsigned long *)(r3t[i] & _REGION_ENTRY_ORIGIN); + r3t[i] = _REGION3_ENTRY_EMPTY; + __gmap_unshadow_sgt(sg, raddr, sgt); + /* Free segment table */ + page = pfn_to_page(__pa(sgt) >> PAGE_SHIFT); + list_del(&page->lru); + __free_pages(page, 2); + } +} + +/** + * gmap_unshadow_r3t - remove a shadow region-3 table from a region-2 entry + * @sg: pointer to the shadow guest address space structure + * @raddr: rmap address in the shadow guest address space + * + * Called with the sg->guest_table_lock + */ +static void gmap_unshadow_r3t(struct gmap *sg, unsigned long raddr) +{ + unsigned long r2o, *r2e, *r3t; + struct page *page; + + BUG_ON(!gmap_is_shadow(sg)); + r2e = gmap_table_walk(sg, raddr, 3); /* get region-2 pointer */ + if (!r2e || !(*r2e & _REGION_ENTRY_ORIGIN)) + return; + gmap_call_notifier(sg, raddr, raddr + (1UL << 42) - 1); + r2o = (unsigned long) (r2e - ((raddr >> 42) & 0x7ff)); + gmap_idte_one(r2o | _ASCE_TYPE_REGION2, raddr); + r3t = (unsigned long *)(*r2e & _REGION_ENTRY_ORIGIN); + *r2e = _REGION2_ENTRY_EMPTY; + __gmap_unshadow_r3t(sg, raddr, r3t); + /* Free region 3 table */ + page = pfn_to_page(__pa(r3t) >> PAGE_SHIFT); + list_del(&page->lru); + __free_pages(page, 2); +} + +/** + * __gmap_unshadow_r2t - remove all entries from a shadow region-2 table + * @sg: pointer to the shadow guest address space structure + * @raddr: rmap address in the shadow guest address space + * @r2t: pointer to the start of a shadow region-2 table + * + * Called with the sg->guest_table_lock + */ +static void __gmap_unshadow_r2t(struct gmap *sg, unsigned long raddr, + unsigned long *r2t) +{ + unsigned long asce, *r3t; + struct page *page; + int i; + + BUG_ON(!gmap_is_shadow(sg)); + asce = (unsigned long) r2t | _ASCE_TYPE_REGION2; + for (i = 0; i < 2048; i++, raddr += 1UL << 42) { + if (!(r2t[i] & _REGION_ENTRY_ORIGIN)) + continue; + r3t = (unsigned long *)(r2t[i] & _REGION_ENTRY_ORIGIN); + r2t[i] = _REGION2_ENTRY_EMPTY; + __gmap_unshadow_r3t(sg, raddr, r3t); + /* Free region 3 table */ + page = pfn_to_page(__pa(r3t) >> PAGE_SHIFT); + list_del(&page->lru); + __free_pages(page, 2); + } +} + +/** + * gmap_unshadow_r2t - remove a shadow region-2 table from a region-1 entry + * @sg: pointer to the shadow guest address space structure + * @raddr: rmap address in the shadow guest address space + * + * Called with the sg->guest_table_lock + */ +static void gmap_unshadow_r2t(struct gmap *sg, unsigned long raddr) +{ + unsigned long r1o, *r1e, *r2t; + struct page *page; + + BUG_ON(!gmap_is_shadow(sg)); + r1e = gmap_table_walk(sg, raddr, 4); /* get region-1 pointer */ + if (!r1e || !(*r1e & _REGION_ENTRY_ORIGIN)) + return; + gmap_call_notifier(sg, raddr, raddr + (1UL << 53) - 1); + r1o = (unsigned long) (r1e - ((raddr >> 53) & 0x7ff)); + gmap_idte_one(r1o | _ASCE_TYPE_REGION1, raddr); + r2t = (unsigned long *)(*r1e & _REGION_ENTRY_ORIGIN); + *r1e = _REGION1_ENTRY_EMPTY; + __gmap_unshadow_r2t(sg, raddr, r2t); + /* Free region 2 table */ + page = pfn_to_page(__pa(r2t) >> PAGE_SHIFT); + list_del(&page->lru); + __free_pages(page, 2); +} + +/** + * __gmap_unshadow_r1t - remove all entries from a shadow region-1 table + * @sg: pointer to the shadow guest address space structure + * @raddr: rmap address in the shadow guest address space + * @r1t: pointer to the start of a shadow region-1 table + * + * Called with the shadow->guest_table_lock + */ +static void __gmap_unshadow_r1t(struct gmap *sg, unsigned long raddr, + unsigned long *r1t) +{ + unsigned long asce, *r2t; + struct page *page; + int i; + + BUG_ON(!gmap_is_shadow(sg)); + asce = (unsigned long) r1t | _ASCE_TYPE_REGION1; + for (i = 0; i < 2048; i++, raddr += 1UL << 53) { + if (!(r1t[i] & _REGION_ENTRY_ORIGIN)) + continue; + r2t = (unsigned long *)(r1t[i] & _REGION_ENTRY_ORIGIN); + __gmap_unshadow_r2t(sg, raddr, r2t); + /* Clear entry and flush translation r1t -> r2t */ + gmap_idte_one(asce, raddr); + r1t[i] = _REGION1_ENTRY_EMPTY; + /* Free region 2 table */ + page = pfn_to_page(__pa(r2t) >> PAGE_SHIFT); + list_del(&page->lru); + __free_pages(page, 2); + } +} + +/** + * gmap_unshadow - remove a shadow page table completely + * @sg: pointer to the shadow guest address space structure + * + * Called with sg->guest_table_lock + */ +static void gmap_unshadow(struct gmap *sg) +{ + unsigned long *table; + + BUG_ON(!gmap_is_shadow(sg)); + if (sg->removed) + return; + sg->removed = 1; + gmap_call_notifier(sg, 0, -1UL); + gmap_flush_tlb(sg); + table = (unsigned long *)(sg->asce & _ASCE_ORIGIN); + switch (sg->asce & _ASCE_TYPE_MASK) { + case _ASCE_TYPE_REGION1: + __gmap_unshadow_r1t(sg, 0, table); + break; + case _ASCE_TYPE_REGION2: + __gmap_unshadow_r2t(sg, 0, table); + break; + case _ASCE_TYPE_REGION3: + __gmap_unshadow_r3t(sg, 0, table); + break; + case _ASCE_TYPE_SEGMENT: + __gmap_unshadow_sgt(sg, 0, table); + break; + } +} + +/** + * gmap_find_shadow - find a specific asce in the list of shadow tables + * @parent: pointer to the parent gmap + * @asce: ASCE for which the shadow table is created + * @edat_level: edat level to be used for the shadow translation + * + * Returns the pointer to a gmap if a shadow table with the given asce is + * already available, ERR_PTR(-EAGAIN) if another one is just being created, + * otherwise NULL + */ +static struct gmap *gmap_find_shadow(struct gmap *parent, unsigned long asce, + int edat_level) +{ + struct gmap *sg; + + list_for_each_entry(sg, &parent->children, list) { + if (sg->orig_asce != asce || sg->edat_level != edat_level || + sg->removed) + continue; + if (!sg->initialized) + return ERR_PTR(-EAGAIN); + atomic_inc(&sg->ref_count); + return sg; + } + return NULL; +} + +/** + * gmap_shadow_valid - check if a shadow guest address space matches the + * given properties and is still valid + * @sg: pointer to the shadow guest address space structure + * @asce: ASCE for which the shadow table is requested + * @edat_level: edat level to be used for the shadow translation + * + * Returns 1 if the gmap shadow is still valid and matches the given + * properties, the caller can continue using it. Returns 0 otherwise, the + * caller has to request a new shadow gmap in this case. + * + */ +int gmap_shadow_valid(struct gmap *sg, unsigned long asce, int edat_level) +{ + if (sg->removed) + return 0; + return sg->orig_asce == asce && sg->edat_level == edat_level; +} +EXPORT_SYMBOL_GPL(gmap_shadow_valid); + +/** + * gmap_shadow - create/find a shadow guest address space + * @parent: pointer to the parent gmap + * @asce: ASCE for which the shadow table is created + * @edat_level: edat level to be used for the shadow translation + * + * The pages of the top level page table referred by the asce parameter + * will be set to read-only and marked in the PGSTEs of the kvm process. + * The shadow table will be removed automatically on any change to the + * PTE mapping for the source table. + * + * Returns a guest address space structure, ERR_PTR(-ENOMEM) if out of memory, + * ERR_PTR(-EAGAIN) if the caller has to retry and ERR_PTR(-EFAULT) if the + * parent gmap table could not be protected. + */ +struct gmap *gmap_shadow(struct gmap *parent, unsigned long asce, + int edat_level) +{ + struct gmap *sg, *new; + unsigned long limit; + int rc; + + BUG_ON(gmap_is_shadow(parent)); + spin_lock(&parent->shadow_lock); + sg = gmap_find_shadow(parent, asce, edat_level); + spin_unlock(&parent->shadow_lock); + if (sg) + return sg; + /* Create a new shadow gmap */ + limit = -1UL >> (33 - (((asce & _ASCE_TYPE_MASK) >> 2) * 11)); + if (asce & _ASCE_REAL_SPACE) + limit = -1UL; + new = gmap_alloc(limit); + if (!new) + return ERR_PTR(-ENOMEM); + new->mm = parent->mm; + new->parent = gmap_get(parent); + new->orig_asce = asce; + new->edat_level = edat_level; + new->initialized = false; + spin_lock(&parent->shadow_lock); + /* Recheck if another CPU created the same shadow */ + sg = gmap_find_shadow(parent, asce, edat_level); + if (sg) { + spin_unlock(&parent->shadow_lock); + gmap_free(new); + return sg; + } + if (asce & _ASCE_REAL_SPACE) { + /* only allow one real-space gmap shadow */ + list_for_each_entry(sg, &parent->children, list) { + if (sg->orig_asce & _ASCE_REAL_SPACE) { + spin_lock(&sg->guest_table_lock); + gmap_unshadow(sg); + spin_unlock(&sg->guest_table_lock); + list_del(&sg->list); + gmap_put(sg); + break; + } + } + } + atomic_set(&new->ref_count, 2); + list_add(&new->list, &parent->children); + if (asce & _ASCE_REAL_SPACE) { + /* nothing to protect, return right away */ + new->initialized = true; + spin_unlock(&parent->shadow_lock); + return new; + } + spin_unlock(&parent->shadow_lock); + /* protect after insertion, so it will get properly invalidated */ + down_read(&parent->mm->mmap_sem); + rc = gmap_protect_range(parent, asce & _ASCE_ORIGIN, + ((asce & _ASCE_TABLE_LENGTH) + 1) * 4096, + PROT_READ, PGSTE_VSIE_BIT); + up_read(&parent->mm->mmap_sem); + spin_lock(&parent->shadow_lock); + new->initialized = true; + if (rc) { + list_del(&new->list); + gmap_free(new); + new = ERR_PTR(rc); + } + spin_unlock(&parent->shadow_lock); + return new; +} +EXPORT_SYMBOL_GPL(gmap_shadow); + +/** + * gmap_shadow_r2t - create an empty shadow region 2 table + * @sg: pointer to the shadow guest address space structure + * @saddr: faulting address in the shadow gmap + * @r2t: parent gmap address of the region 2 table to get shadowed + * @fake: r2t references contiguous guest memory block, not a r2t + * + * The r2t parameter specifies the address of the source table. The + * four pages of the source table are made read-only in the parent gmap + * address space. A write to the source table area @r2t will automatically + * remove the shadow r2 table and all of its decendents. + * + * Returns 0 if successfully shadowed or already shadowed, -EAGAIN if the + * shadow table structure is incomplete, -ENOMEM if out of memory and + * -EFAULT if an address in the parent gmap could not be resolved. + * + * Called with sg->mm->mmap_sem in read. + */ +int gmap_shadow_r2t(struct gmap *sg, unsigned long saddr, unsigned long r2t, + int fake) +{ + unsigned long raddr, origin, offset, len; + unsigned long *s_r2t, *table; + struct page *page; + int rc; + + BUG_ON(!gmap_is_shadow(sg)); + /* Allocate a shadow region second table */ + page = alloc_pages(GFP_KERNEL, 2); + if (!page) + return -ENOMEM; + page->index = r2t & _REGION_ENTRY_ORIGIN; + if (fake) + page->index |= GMAP_SHADOW_FAKE_TABLE; + s_r2t = (unsigned long *) page_to_phys(page); + /* Install shadow region second table */ + spin_lock(&sg->guest_table_lock); + table = gmap_table_walk(sg, saddr, 4); /* get region-1 pointer */ + if (!table) { + rc = -EAGAIN; /* Race with unshadow */ + goto out_free; + } + if (!(*table & _REGION_ENTRY_INVALID)) { + rc = 0; /* Already established */ + goto out_free; + } else if (*table & _REGION_ENTRY_ORIGIN) { + rc = -EAGAIN; /* Race with shadow */ + goto out_free; + } + crst_table_init(s_r2t, _REGION2_ENTRY_EMPTY); + /* mark as invalid as long as the parent table is not protected */ + *table = (unsigned long) s_r2t | _REGION_ENTRY_LENGTH | + _REGION_ENTRY_TYPE_R1 | _REGION_ENTRY_INVALID; + if (sg->edat_level >= 1) + *table |= (r2t & _REGION_ENTRY_PROTECT); + list_add(&page->lru, &sg->crst_list); + if (fake) { + /* nothing to protect for fake tables */ + *table &= ~_REGION_ENTRY_INVALID; + spin_unlock(&sg->guest_table_lock); + return 0; + } + spin_unlock(&sg->guest_table_lock); + /* Make r2t read-only in parent gmap page table */ + raddr = (saddr & 0xffe0000000000000UL) | _SHADOW_RMAP_REGION1; + origin = r2t & _REGION_ENTRY_ORIGIN; + offset = ((r2t & _REGION_ENTRY_OFFSET) >> 6) * 4096; + len = ((r2t & _REGION_ENTRY_LENGTH) + 1) * 4096 - offset; + rc = gmap_protect_rmap(sg, raddr, origin + offset, len, PROT_READ); + spin_lock(&sg->guest_table_lock); + if (!rc) { + table = gmap_table_walk(sg, saddr, 4); + if (!table || (*table & _REGION_ENTRY_ORIGIN) != + (unsigned long) s_r2t) + rc = -EAGAIN; /* Race with unshadow */ + else + *table &= ~_REGION_ENTRY_INVALID; + } else { + gmap_unshadow_r2t(sg, raddr); + } + spin_unlock(&sg->guest_table_lock); + return rc; +out_free: + spin_unlock(&sg->guest_table_lock); + __free_pages(page, 2); + return rc; +} +EXPORT_SYMBOL_GPL(gmap_shadow_r2t); + +/** + * gmap_shadow_r3t - create a shadow region 3 table + * @sg: pointer to the shadow guest address space structure + * @saddr: faulting address in the shadow gmap + * @r3t: parent gmap address of the region 3 table to get shadowed + * @fake: r3t references contiguous guest memory block, not a r3t + * + * Returns 0 if successfully shadowed or already shadowed, -EAGAIN if the + * shadow table structure is incomplete, -ENOMEM if out of memory and + * -EFAULT if an address in the parent gmap could not be resolved. + * + * Called with sg->mm->mmap_sem in read. + */ +int gmap_shadow_r3t(struct gmap *sg, unsigned long saddr, unsigned long r3t, + int fake) +{ + unsigned long raddr, origin, offset, len; + unsigned long *s_r3t, *table; + struct page *page; + int rc; + + BUG_ON(!gmap_is_shadow(sg)); + /* Allocate a shadow region second table */ + page = alloc_pages(GFP_KERNEL, 2); + if (!page) + return -ENOMEM; + page->index = r3t & _REGION_ENTRY_ORIGIN; + if (fake) + page->index |= GMAP_SHADOW_FAKE_TABLE; + s_r3t = (unsigned long *) page_to_phys(page); + /* Install shadow region second table */ + spin_lock(&sg->guest_table_lock); + table = gmap_table_walk(sg, saddr, 3); /* get region-2 pointer */ + if (!table) { + rc = -EAGAIN; /* Race with unshadow */ + goto out_free; + } + if (!(*table & _REGION_ENTRY_INVALID)) { + rc = 0; /* Already established */ + goto out_free; + } else if (*table & _REGION_ENTRY_ORIGIN) { + rc = -EAGAIN; /* Race with shadow */ + } + crst_table_init(s_r3t, _REGION3_ENTRY_EMPTY); + /* mark as invalid as long as the parent table is not protected */ + *table = (unsigned long) s_r3t | _REGION_ENTRY_LENGTH | + _REGION_ENTRY_TYPE_R2 | _REGION_ENTRY_INVALID; + if (sg->edat_level >= 1) + *table |= (r3t & _REGION_ENTRY_PROTECT); + list_add(&page->lru, &sg->crst_list); + if (fake) { + /* nothing to protect for fake tables */ + *table &= ~_REGION_ENTRY_INVALID; + spin_unlock(&sg->guest_table_lock); + return 0; + } + spin_unlock(&sg->guest_table_lock); + /* Make r3t read-only in parent gmap page table */ + raddr = (saddr & 0xfffffc0000000000UL) | _SHADOW_RMAP_REGION2; + origin = r3t & _REGION_ENTRY_ORIGIN; + offset = ((r3t & _REGION_ENTRY_OFFSET) >> 6) * 4096; + len = ((r3t & _REGION_ENTRY_LENGTH) + 1) * 4096 - offset; + rc = gmap_protect_rmap(sg, raddr, origin + offset, len, PROT_READ); + spin_lock(&sg->guest_table_lock); + if (!rc) { + table = gmap_table_walk(sg, saddr, 3); + if (!table || (*table & _REGION_ENTRY_ORIGIN) != + (unsigned long) s_r3t) + rc = -EAGAIN; /* Race with unshadow */ + else + *table &= ~_REGION_ENTRY_INVALID; + } else { + gmap_unshadow_r3t(sg, raddr); + } + spin_unlock(&sg->guest_table_lock); + return rc; +out_free: + spin_unlock(&sg->guest_table_lock); + __free_pages(page, 2); + return rc; +} +EXPORT_SYMBOL_GPL(gmap_shadow_r3t); + +/** + * gmap_shadow_sgt - create a shadow segment table + * @sg: pointer to the shadow guest address space structure + * @saddr: faulting address in the shadow gmap + * @sgt: parent gmap address of the segment table to get shadowed + * @fake: sgt references contiguous guest memory block, not a sgt + * + * Returns: 0 if successfully shadowed or already shadowed, -EAGAIN if the + * shadow table structure is incomplete, -ENOMEM if out of memory and + * -EFAULT if an address in the parent gmap could not be resolved. + * + * Called with sg->mm->mmap_sem in read. + */ +int gmap_shadow_sgt(struct gmap *sg, unsigned long saddr, unsigned long sgt, + int fake) +{ + unsigned long raddr, origin, offset, len; + unsigned long *s_sgt, *table; + struct page *page; + int rc; + + BUG_ON(!gmap_is_shadow(sg) || (sgt & _REGION3_ENTRY_LARGE)); + /* Allocate a shadow segment table */ + page = alloc_pages(GFP_KERNEL, 2); + if (!page) + return -ENOMEM; + page->index = sgt & _REGION_ENTRY_ORIGIN; + if (fake) + page->index |= GMAP_SHADOW_FAKE_TABLE; + s_sgt = (unsigned long *) page_to_phys(page); + /* Install shadow region second table */ + spin_lock(&sg->guest_table_lock); + table = gmap_table_walk(sg, saddr, 2); /* get region-3 pointer */ + if (!table) { + rc = -EAGAIN; /* Race with unshadow */ + goto out_free; + } + if (!(*table & _REGION_ENTRY_INVALID)) { + rc = 0; /* Already established */ + goto out_free; + } else if (*table & _REGION_ENTRY_ORIGIN) { + rc = -EAGAIN; /* Race with shadow */ + goto out_free; + } + crst_table_init(s_sgt, _SEGMENT_ENTRY_EMPTY); + /* mark as invalid as long as the parent table is not protected */ + *table = (unsigned long) s_sgt | _REGION_ENTRY_LENGTH | + _REGION_ENTRY_TYPE_R3 | _REGION_ENTRY_INVALID; + if (sg->edat_level >= 1) + *table |= sgt & _REGION_ENTRY_PROTECT; + list_add(&page->lru, &sg->crst_list); + if (fake) { + /* nothing to protect for fake tables */ + *table &= ~_REGION_ENTRY_INVALID; + spin_unlock(&sg->guest_table_lock); + return 0; + } + spin_unlock(&sg->guest_table_lock); + /* Make sgt read-only in parent gmap page table */ + raddr = (saddr & 0xffffffff80000000UL) | _SHADOW_RMAP_REGION3; + origin = sgt & _REGION_ENTRY_ORIGIN; + offset = ((sgt & _REGION_ENTRY_OFFSET) >> 6) * 4096; + len = ((sgt & _REGION_ENTRY_LENGTH) + 1) * 4096 - offset; + rc = gmap_protect_rmap(sg, raddr, origin + offset, len, PROT_READ); + spin_lock(&sg->guest_table_lock); + if (!rc) { + table = gmap_table_walk(sg, saddr, 2); + if (!table || (*table & _REGION_ENTRY_ORIGIN) != + (unsigned long) s_sgt) + rc = -EAGAIN; /* Race with unshadow */ + else + *table &= ~_REGION_ENTRY_INVALID; + } else { + gmap_unshadow_sgt(sg, raddr); + } + spin_unlock(&sg->guest_table_lock); + return rc; +out_free: + spin_unlock(&sg->guest_table_lock); + __free_pages(page, 2); + return rc; +} +EXPORT_SYMBOL_GPL(gmap_shadow_sgt); + +/** + * gmap_shadow_lookup_pgtable - find a shadow page table + * @sg: pointer to the shadow guest address space structure + * @saddr: the address in the shadow aguest address space + * @pgt: parent gmap address of the page table to get shadowed + * @dat_protection: if the pgtable is marked as protected by dat + * @fake: pgt references contiguous guest memory block, not a pgtable + * + * Returns 0 if the shadow page table was found and -EAGAIN if the page + * table was not found. + * + * Called with sg->mm->mmap_sem in read. + */ +int gmap_shadow_pgt_lookup(struct gmap *sg, unsigned long saddr, + unsigned long *pgt, int *dat_protection, + int *fake) +{ + unsigned long *table; + struct page *page; + int rc; + + BUG_ON(!gmap_is_shadow(sg)); + spin_lock(&sg->guest_table_lock); + table = gmap_table_walk(sg, saddr, 1); /* get segment pointer */ + if (table && !(*table & _SEGMENT_ENTRY_INVALID)) { + /* Shadow page tables are full pages (pte+pgste) */ + page = pfn_to_page(*table >> PAGE_SHIFT); + *pgt = page->index & ~GMAP_SHADOW_FAKE_TABLE; + *dat_protection = !!(*table & _SEGMENT_ENTRY_PROTECT); + *fake = !!(page->index & GMAP_SHADOW_FAKE_TABLE); + rc = 0; + } else { + rc = -EAGAIN; + } + spin_unlock(&sg->guest_table_lock); + return rc; + +} +EXPORT_SYMBOL_GPL(gmap_shadow_pgt_lookup); + +/** + * gmap_shadow_pgt - instantiate a shadow page table + * @sg: pointer to the shadow guest address space structure + * @saddr: faulting address in the shadow gmap + * @pgt: parent gmap address of the page table to get shadowed + * @fake: pgt references contiguous guest memory block, not a pgtable + * + * Returns 0 if successfully shadowed or already shadowed, -EAGAIN if the + * shadow table structure is incomplete, -ENOMEM if out of memory, + * -EFAULT if an address in the parent gmap could not be resolved and + * + * Called with gmap->mm->mmap_sem in read + */ +int gmap_shadow_pgt(struct gmap *sg, unsigned long saddr, unsigned long pgt, + int fake) +{ + unsigned long raddr, origin; + unsigned long *s_pgt, *table; + struct page *page; + int rc; + + BUG_ON(!gmap_is_shadow(sg) || (pgt & _SEGMENT_ENTRY_LARGE)); + /* Allocate a shadow page table */ + page = page_table_alloc_pgste(sg->mm); + if (!page) + return -ENOMEM; + page->index = pgt & _SEGMENT_ENTRY_ORIGIN; + if (fake) + page->index |= GMAP_SHADOW_FAKE_TABLE; + s_pgt = (unsigned long *) page_to_phys(page); + /* Install shadow page table */ + spin_lock(&sg->guest_table_lock); + table = gmap_table_walk(sg, saddr, 1); /* get segment pointer */ + if (!table) { + rc = -EAGAIN; /* Race with unshadow */ + goto out_free; + } + if (!(*table & _SEGMENT_ENTRY_INVALID)) { + rc = 0; /* Already established */ + goto out_free; + } else if (*table & _SEGMENT_ENTRY_ORIGIN) { + rc = -EAGAIN; /* Race with shadow */ + goto out_free; + } + /* mark as invalid as long as the parent table is not protected */ + *table = (unsigned long) s_pgt | _SEGMENT_ENTRY | + (pgt & _SEGMENT_ENTRY_PROTECT) | _SEGMENT_ENTRY_INVALID; + list_add(&page->lru, &sg->pt_list); + if (fake) { + /* nothing to protect for fake tables */ + *table &= ~_SEGMENT_ENTRY_INVALID; + spin_unlock(&sg->guest_table_lock); + return 0; + } + spin_unlock(&sg->guest_table_lock); + /* Make pgt read-only in parent gmap page table (not the pgste) */ + raddr = (saddr & 0xfffffffffff00000UL) | _SHADOW_RMAP_SEGMENT; + origin = pgt & _SEGMENT_ENTRY_ORIGIN & PAGE_MASK; + rc = gmap_protect_rmap(sg, raddr, origin, PAGE_SIZE, PROT_READ); + spin_lock(&sg->guest_table_lock); + if (!rc) { + table = gmap_table_walk(sg, saddr, 1); + if (!table || (*table & _SEGMENT_ENTRY_ORIGIN) != + (unsigned long) s_pgt) + rc = -EAGAIN; /* Race with unshadow */ + else + *table &= ~_SEGMENT_ENTRY_INVALID; + } else { + gmap_unshadow_pgt(sg, raddr); + } + spin_unlock(&sg->guest_table_lock); + return rc; +out_free: + spin_unlock(&sg->guest_table_lock); + page_table_free_pgste(page); + return rc; + +} +EXPORT_SYMBOL_GPL(gmap_shadow_pgt); + +/** + * gmap_shadow_page - create a shadow page mapping + * @sg: pointer to the shadow guest address space structure + * @saddr: faulting address in the shadow gmap + * @pte: pte in parent gmap address space to get shadowed + * + * Returns 0 if successfully shadowed or already shadowed, -EAGAIN if the + * shadow table structure is incomplete, -ENOMEM if out of memory and + * -EFAULT if an address in the parent gmap could not be resolved. + * + * Called with sg->mm->mmap_sem in read. + */ +int gmap_shadow_page(struct gmap *sg, unsigned long saddr, pte_t pte) +{ + struct gmap *parent; + struct gmap_rmap *rmap; + unsigned long vmaddr, paddr; + spinlock_t *ptl; + pte_t *sptep, *tptep; + int prot; + int rc; + + BUG_ON(!gmap_is_shadow(sg)); + parent = sg->parent; + prot = (pte_val(pte) & _PAGE_PROTECT) ? PROT_READ : PROT_WRITE; + + rmap = kzalloc(sizeof(*rmap), GFP_KERNEL); + if (!rmap) + return -ENOMEM; + rmap->raddr = (saddr & PAGE_MASK) | _SHADOW_RMAP_PGTABLE; + + while (1) { + paddr = pte_val(pte) & PAGE_MASK; + vmaddr = __gmap_translate(parent, paddr); + if (IS_ERR_VALUE(vmaddr)) { + rc = vmaddr; + break; + } + rc = radix_tree_preload(GFP_KERNEL); if (rc) break; - /* Walk the process page table, lock and get pte pointer */ - ptep = get_locked_pte(gmap->mm, addr, &ptl); - VM_BUG_ON(!ptep); - /* Set notification bit in the pgste of the pte */ - if ((pte_val(*ptep) & (_PAGE_INVALID | _PAGE_PROTECT)) == 0) { - ptep_set_notify(gmap->mm, addr, ptep); - gaddr += PAGE_SIZE; - len -= PAGE_SIZE; + rc = -EAGAIN; + sptep = gmap_pte_op_walk(parent, paddr, &ptl); + if (sptep) { + spin_lock(&sg->guest_table_lock); + /* Get page table pointer */ + tptep = (pte_t *) gmap_table_walk(sg, saddr, 0); + if (!tptep) { + spin_unlock(&sg->guest_table_lock); + gmap_pte_op_end(ptl); + radix_tree_preload_end(); + break; + } + rc = ptep_shadow_pte(sg->mm, saddr, sptep, tptep, pte); + if (rc > 0) { + /* Success and a new mapping */ + gmap_insert_rmap(sg, vmaddr, rmap); + rmap = NULL; + rc = 0; + } + gmap_pte_op_end(ptl); + spin_unlock(&sg->guest_table_lock); } - pte_unmap_unlock(ptep, ptl); + radix_tree_preload_end(); + if (!rc) + break; + rc = gmap_pte_op_fixup(parent, paddr, vmaddr, prot); + if (rc) + break; } - up_read(&gmap->mm->mmap_sem); + kfree(rmap); return rc; } -EXPORT_SYMBOL_GPL(gmap_ipte_notify); +EXPORT_SYMBOL_GPL(gmap_shadow_page); + +/** + * gmap_shadow_notify - handle notifications for shadow gmap + * + * Called with sg->parent->shadow_lock. + */ +static void gmap_shadow_notify(struct gmap *sg, unsigned long vmaddr, + unsigned long offset, pte_t *pte) +{ + struct gmap_rmap *rmap, *rnext, *head; + unsigned long gaddr, start, end, bits, raddr; + unsigned long *table; + + BUG_ON(!gmap_is_shadow(sg)); + spin_lock(&sg->parent->guest_table_lock); + table = radix_tree_lookup(&sg->parent->host_to_guest, + vmaddr >> PMD_SHIFT); + gaddr = table ? __gmap_segment_gaddr(table) + offset : 0; + spin_unlock(&sg->parent->guest_table_lock); + if (!table) + return; + + spin_lock(&sg->guest_table_lock); + if (sg->removed) { + spin_unlock(&sg->guest_table_lock); + return; + } + /* Check for top level table */ + start = sg->orig_asce & _ASCE_ORIGIN; + end = start + ((sg->orig_asce & _ASCE_TABLE_LENGTH) + 1) * 4096; + if (!(sg->orig_asce & _ASCE_REAL_SPACE) && gaddr >= start && + gaddr < end) { + /* The complete shadow table has to go */ + gmap_unshadow(sg); + spin_unlock(&sg->guest_table_lock); + list_del(&sg->list); + gmap_put(sg); + return; + } + /* Remove the page table tree from on specific entry */ + head = radix_tree_delete(&sg->host_to_rmap, vmaddr >> 12); + gmap_for_each_rmap_safe(rmap, rnext, head) { + bits = rmap->raddr & _SHADOW_RMAP_MASK; + raddr = rmap->raddr ^ bits; + switch (bits) { + case _SHADOW_RMAP_REGION1: + gmap_unshadow_r2t(sg, raddr); + break; + case _SHADOW_RMAP_REGION2: + gmap_unshadow_r3t(sg, raddr); + break; + case _SHADOW_RMAP_REGION3: + gmap_unshadow_sgt(sg, raddr); + break; + case _SHADOW_RMAP_SEGMENT: + gmap_unshadow_pgt(sg, raddr); + break; + case _SHADOW_RMAP_PGTABLE: + gmap_unshadow_page(sg, raddr); + break; + } + kfree(rmap); + } + spin_unlock(&sg->guest_table_lock); +} /** * ptep_notify - call all invalidation callbacks for a specific pte. * @mm: pointer to the process mm_struct * @addr: virtual address in the process address space * @pte: pointer to the page table entry + * @bits: bits from the pgste that caused the notify call * * This function is assumed to be called with the page table lock held * for the pte to notify. */ -void ptep_notify(struct mm_struct *mm, unsigned long vmaddr, pte_t *pte) +void ptep_notify(struct mm_struct *mm, unsigned long vmaddr, + pte_t *pte, unsigned long bits) { unsigned long offset, gaddr; unsigned long *table; - struct gmap_notifier *nb; - struct gmap *gmap; + struct gmap *gmap, *sg, *next; offset = ((unsigned long) pte) & (255 * sizeof(pte_t)); offset = offset * (4096 / sizeof(pte_t)); - spin_lock(&gmap_notifier_lock); - list_for_each_entry(gmap, &mm->context.gmap_list, list) { + rcu_read_lock(); + list_for_each_entry_rcu(gmap, &mm->context.gmap_list, list) { + if (!list_empty(&gmap->children) && (bits & PGSTE_VSIE_BIT)) { + spin_lock(&gmap->shadow_lock); + list_for_each_entry_safe(sg, next, + &gmap->children, list) + gmap_shadow_notify(sg, vmaddr, offset, pte); + spin_unlock(&gmap->shadow_lock); + } + if (!(bits & PGSTE_IN_BIT)) + continue; + spin_lock(&gmap->guest_table_lock); table = radix_tree_lookup(&gmap->host_to_guest, vmaddr >> PMD_SHIFT); - if (!table) - continue; - gaddr = __gmap_segment_gaddr(table) + offset; - list_for_each_entry(nb, &gmap_notifier_list, list) - nb->notifier_call(gmap, gaddr); + if (table) + gaddr = __gmap_segment_gaddr(table) + offset; + spin_unlock(&gmap->guest_table_lock); + if (table) + gmap_call_notifier(gmap, gaddr, gaddr + PAGE_SIZE - 1); } - spin_unlock(&gmap_notifier_lock); + rcu_read_unlock(); } EXPORT_SYMBOL_GPL(ptep_notify); diff --git a/arch/s390/mm/pgalloc.c b/arch/s390/mm/pgalloc.c index e2565d2..995f785 100644 --- a/arch/s390/mm/pgalloc.c +++ b/arch/s390/mm/pgalloc.c @@ -137,6 +137,29 @@ static inline unsigned int atomic_xor_bits(atomic_t *v, unsigned int bits) return new; } +#ifdef CONFIG_PGSTE + +struct page *page_table_alloc_pgste(struct mm_struct *mm) +{ + struct page *page; + unsigned long *table; + + page = alloc_page(GFP_KERNEL|__GFP_REPEAT); + if (page) { + table = (unsigned long *) page_to_phys(page); + clear_table(table, _PAGE_INVALID, PAGE_SIZE/2); + clear_table(table + PTRS_PER_PTE, 0, PAGE_SIZE/2); + } + return page; +} + +void page_table_free_pgste(struct page *page) +{ + __free_page(page); +} + +#endif /* CONFIG_PGSTE */ + /* * page table entry allocation/free routines. */ @@ -149,7 +172,7 @@ unsigned long *page_table_alloc(struct mm_struct *mm) /* Try to get a fragment of a 4K page as a 2K page table */ if (!mm_alloc_pgste(mm)) { table = NULL; - spin_lock_bh(&mm->context.list_lock); + spin_lock_bh(&mm->context.pgtable_lock); if (!list_empty(&mm->context.pgtable_list)) { page = list_first_entry(&mm->context.pgtable_list, struct page, lru); @@ -164,7 +187,7 @@ unsigned long *page_table_alloc(struct mm_struct *mm) list_del(&page->lru); } } - spin_unlock_bh(&mm->context.list_lock); + spin_unlock_bh(&mm->context.pgtable_lock); if (table) return table; } @@ -187,9 +210,9 @@ unsigned long *page_table_alloc(struct mm_struct *mm) /* Return the first 2K fragment of the page */ atomic_set(&page->_mapcount, 1); clear_table(table, _PAGE_INVALID, PAGE_SIZE); - spin_lock_bh(&mm->context.list_lock); + spin_lock_bh(&mm->context.pgtable_lock); list_add(&page->lru, &mm->context.pgtable_list); - spin_unlock_bh(&mm->context.list_lock); + spin_unlock_bh(&mm->context.pgtable_lock); } return table; } @@ -203,13 +226,13 @@ void page_table_free(struct mm_struct *mm, unsigned long *table) if (!mm_alloc_pgste(mm)) { /* Free 2K page table fragment of a 4K page */ bit = (__pa(table) & ~PAGE_MASK)/(PTRS_PER_PTE*sizeof(pte_t)); - spin_lock_bh(&mm->context.list_lock); + spin_lock_bh(&mm->context.pgtable_lock); mask = atomic_xor_bits(&page->_mapcount, 1U << bit); if (mask & 3) list_add(&page->lru, &mm->context.pgtable_list); else list_del(&page->lru); - spin_unlock_bh(&mm->context.list_lock); + spin_unlock_bh(&mm->context.pgtable_lock); if (mask != 0) return; } @@ -235,13 +258,13 @@ void page_table_free_rcu(struct mmu_gather *tlb, unsigned long *table, return; } bit = (__pa(table) & ~PAGE_MASK) / (PTRS_PER_PTE*sizeof(pte_t)); - spin_lock_bh(&mm->context.list_lock); + spin_lock_bh(&mm->context.pgtable_lock); mask = atomic_xor_bits(&page->_mapcount, 0x11U << bit); if (mask & 3) list_add_tail(&page->lru, &mm->context.pgtable_list); else list_del(&page->lru); - spin_unlock_bh(&mm->context.list_lock); + spin_unlock_bh(&mm->context.pgtable_lock); table = (unsigned long *) (__pa(table) | (1U << bit)); tlb_remove_table(tlb, table); } diff --git a/arch/s390/mm/pgtable.c b/arch/s390/mm/pgtable.c index b98d1a1..5f09201 100644 --- a/arch/s390/mm/pgtable.c +++ b/arch/s390/mm/pgtable.c @@ -174,14 +174,17 @@ static inline pgste_t pgste_set_pte(pte_t *ptep, pgste_t pgste, pte_t entry) return pgste; } -static inline pgste_t pgste_ipte_notify(struct mm_struct *mm, - unsigned long addr, - pte_t *ptep, pgste_t pgste) +static inline pgste_t pgste_pte_notify(struct mm_struct *mm, + unsigned long addr, + pte_t *ptep, pgste_t pgste) { #ifdef CONFIG_PGSTE - if (pgste_val(pgste) & PGSTE_IN_BIT) { - pgste_val(pgste) &= ~PGSTE_IN_BIT; - ptep_notify(mm, addr, ptep); + unsigned long bits; + + bits = pgste_val(pgste) & (PGSTE_IN_BIT | PGSTE_VSIE_BIT); + if (bits) { + pgste_val(pgste) ^= bits; + ptep_notify(mm, addr, ptep, bits); } #endif return pgste; @@ -194,7 +197,7 @@ static inline pgste_t ptep_xchg_start(struct mm_struct *mm, if (mm_has_pgste(mm)) { pgste = pgste_get_lock(ptep); - pgste = pgste_ipte_notify(mm, addr, ptep, pgste); + pgste = pgste_pte_notify(mm, addr, ptep, pgste); } return pgste; } @@ -459,6 +462,90 @@ void ptep_set_notify(struct mm_struct *mm, unsigned long addr, pte_t *ptep) preempt_enable(); } +/** + * ptep_force_prot - change access rights of a locked pte + * @mm: pointer to the process mm_struct + * @addr: virtual address in the guest address space + * @ptep: pointer to the page table entry + * @prot: indicates guest access rights: PROT_NONE, PROT_READ or PROT_WRITE + * @bit: pgste bit to set (e.g. for notification) + * + * Returns 0 if the access rights were changed and -EAGAIN if the current + * and requested access rights are incompatible. + */ +int ptep_force_prot(struct mm_struct *mm, unsigned long addr, + pte_t *ptep, int prot, unsigned long bit) +{ + pte_t entry; + pgste_t pgste; + int pte_i, pte_p; + + pgste = pgste_get_lock(ptep); + entry = *ptep; + /* Check pte entry after all locks have been acquired */ + pte_i = pte_val(entry) & _PAGE_INVALID; + pte_p = pte_val(entry) & _PAGE_PROTECT; + if ((pte_i && (prot != PROT_NONE)) || + (pte_p && (prot & PROT_WRITE))) { + pgste_set_unlock(ptep, pgste); + return -EAGAIN; + } + /* Change access rights and set pgste bit */ + if (prot == PROT_NONE && !pte_i) { + ptep_flush_direct(mm, addr, ptep); + pgste = pgste_update_all(entry, pgste, mm); + pte_val(entry) |= _PAGE_INVALID; + } + if (prot == PROT_READ && !pte_p) { + ptep_flush_direct(mm, addr, ptep); + pte_val(entry) &= ~_PAGE_INVALID; + pte_val(entry) |= _PAGE_PROTECT; + } + pgste_val(pgste) |= bit; + pgste = pgste_set_pte(ptep, pgste, entry); + pgste_set_unlock(ptep, pgste); + return 0; +} + +int ptep_shadow_pte(struct mm_struct *mm, unsigned long saddr, + pte_t *sptep, pte_t *tptep, pte_t pte) +{ + pgste_t spgste, tpgste; + pte_t spte, tpte; + int rc = -EAGAIN; + + if (!(pte_val(*tptep) & _PAGE_INVALID)) + return 0; /* already shadowed */ + spgste = pgste_get_lock(sptep); + spte = *sptep; + if (!(pte_val(spte) & _PAGE_INVALID) && + !((pte_val(spte) & _PAGE_PROTECT) && + !(pte_val(pte) & _PAGE_PROTECT))) { + pgste_val(spgste) |= PGSTE_VSIE_BIT; + tpgste = pgste_get_lock(tptep); + pte_val(tpte) = (pte_val(spte) & PAGE_MASK) | + (pte_val(pte) & _PAGE_PROTECT); + /* don't touch the storage key - it belongs to parent pgste */ + tpgste = pgste_set_pte(tptep, tpgste, tpte); + pgste_set_unlock(tptep, tpgste); + rc = 1; + } + pgste_set_unlock(sptep, spgste); + return rc; +} + +void ptep_unshadow_pte(struct mm_struct *mm, unsigned long saddr, pte_t *ptep) +{ + pgste_t pgste; + + pgste = pgste_get_lock(ptep); + /* notifier is called by the caller */ + ptep_flush_direct(mm, saddr, ptep); + /* don't touch the storage key - it belongs to parent pgste */ + pgste = pgste_set_pte(ptep, pgste, __pte(_PAGE_INVALID)); + pgste_set_unlock(ptep, pgste); +} + static void ptep_zap_swap_entry(struct mm_struct *mm, swp_entry_t entry) { if (!non_swap_entry(entry)) @@ -532,7 +619,7 @@ bool test_and_clear_guest_dirty(struct mm_struct *mm, unsigned long addr) pgste_val(pgste) &= ~PGSTE_UC_BIT; pte = *ptep; if (dirty && (pte_val(pte) & _PAGE_PRESENT)) { - pgste = pgste_ipte_notify(mm, addr, ptep, pgste); + pgste = pgste_pte_notify(mm, addr, ptep, pgste); __ptep_ipte(addr, ptep); if (MACHINE_HAS_ESOP || !(pte_val(pte) & _PAGE_WRITE)) pte_val(pte) |= _PAGE_PROTECT; @@ -555,12 +642,9 @@ int set_guest_storage_key(struct mm_struct *mm, unsigned long addr, pgste_t old, new; pte_t *ptep; - down_read(&mm->mmap_sem); ptep = get_locked_pte(mm, addr, &ptl); - if (unlikely(!ptep)) { - up_read(&mm->mmap_sem); + if (unlikely(!ptep)) return -EFAULT; - } new = old = pgste_get_lock(ptep); pgste_val(new) &= ~(PGSTE_GR_BIT | PGSTE_GC_BIT | @@ -587,45 +671,100 @@ int set_guest_storage_key(struct mm_struct *mm, unsigned long addr, pgste_set_unlock(ptep, new); pte_unmap_unlock(ptep, ptl); - up_read(&mm->mmap_sem); return 0; } EXPORT_SYMBOL(set_guest_storage_key); -unsigned char get_guest_storage_key(struct mm_struct *mm, unsigned long addr) +/** + * Conditionally set a guest storage key (handling csske). + * oldkey will be updated when either mr or mc is set and a pointer is given. + * + * Returns 0 if a guests storage key update wasn't necessary, 1 if the guest + * storage key was updated and -EFAULT on access errors. + */ +int cond_set_guest_storage_key(struct mm_struct *mm, unsigned long addr, + unsigned char key, unsigned char *oldkey, + bool nq, bool mr, bool mc) +{ + unsigned char tmp, mask = _PAGE_ACC_BITS | _PAGE_FP_BIT; + int rc; + + /* we can drop the pgste lock between getting and setting the key */ + if (mr | mc) { + rc = get_guest_storage_key(current->mm, addr, &tmp); + if (rc) + return rc; + if (oldkey) + *oldkey = tmp; + if (!mr) + mask |= _PAGE_REFERENCED; + if (!mc) + mask |= _PAGE_CHANGED; + if (!((tmp ^ key) & mask)) + return 0; + } + rc = set_guest_storage_key(current->mm, addr, key, nq); + return rc < 0 ? rc : 1; +} +EXPORT_SYMBOL(cond_set_guest_storage_key); + +/** + * Reset a guest reference bit (rrbe), returning the reference and changed bit. + * + * Returns < 0 in case of error, otherwise the cc to be reported to the guest. + */ +int reset_guest_reference_bit(struct mm_struct *mm, unsigned long addr) { - unsigned char key; spinlock_t *ptl; - pgste_t pgste; + pgste_t old, new; pte_t *ptep; + int cc = 0; - down_read(&mm->mmap_sem); ptep = get_locked_pte(mm, addr, &ptl); - if (unlikely(!ptep)) { - up_read(&mm->mmap_sem); + if (unlikely(!ptep)) return -EFAULT; - } - pgste = pgste_get_lock(ptep); - if (pte_val(*ptep) & _PAGE_INVALID) { - key = (pgste_val(pgste) & PGSTE_ACC_BITS) >> 56; - key |= (pgste_val(pgste) & PGSTE_FP_BIT) >> 56; - key |= (pgste_val(pgste) & PGSTE_GR_BIT) >> 48; - key |= (pgste_val(pgste) & PGSTE_GC_BIT) >> 48; - } else { - key = page_get_storage_key(pte_val(*ptep) & PAGE_MASK); + new = old = pgste_get_lock(ptep); + /* Reset guest reference bit only */ + pgste_val(new) &= ~PGSTE_GR_BIT; - /* Reflect guest's logical view, not physical */ - if (pgste_val(pgste) & PGSTE_GR_BIT) - key |= _PAGE_REFERENCED; - if (pgste_val(pgste) & PGSTE_GC_BIT) - key |= _PAGE_CHANGED; + if (!(pte_val(*ptep) & _PAGE_INVALID)) { + cc = page_reset_referenced(pte_val(*ptep) & PAGE_MASK); + /* Merge real referenced bit into host-set */ + pgste_val(new) |= ((unsigned long) cc << 53) & PGSTE_HR_BIT; } + /* Reflect guest's logical view, not physical */ + cc |= (pgste_val(old) & (PGSTE_GR_BIT | PGSTE_GC_BIT)) >> 49; + /* Changing the guest storage key is considered a change of the page */ + if ((pgste_val(new) ^ pgste_val(old)) & PGSTE_GR_BIT) + pgste_val(new) |= PGSTE_UC_BIT; + + pgste_set_unlock(ptep, new); + pte_unmap_unlock(ptep, ptl); + return 0; +} +EXPORT_SYMBOL(reset_guest_reference_bit); + +int get_guest_storage_key(struct mm_struct *mm, unsigned long addr, + unsigned char *key) +{ + spinlock_t *ptl; + pgste_t pgste; + pte_t *ptep; + ptep = get_locked_pte(mm, addr, &ptl); + if (unlikely(!ptep)) + return -EFAULT; + + pgste = pgste_get_lock(ptep); + *key = (pgste_val(pgste) & (PGSTE_ACC_BITS | PGSTE_FP_BIT)) >> 56; + if (!(pte_val(*ptep) & _PAGE_INVALID)) + *key = page_get_storage_key(pte_val(*ptep) & PAGE_MASK); + /* Reflect guest's logical view, not physical */ + *key |= (pgste_val(pgste) & (PGSTE_GR_BIT | PGSTE_GC_BIT)) >> 48; pgste_set_unlock(ptep, pgste); pte_unmap_unlock(ptep, ptl); - up_read(&mm->mmap_sem); - return key; + return 0; } EXPORT_SYMBOL(get_guest_storage_key); #endif diff --git a/arch/x86/include/asm/kvm_host.h b/arch/x86/include/asm/kvm_host.h index 69e62862..33ae3a4 100644 --- a/arch/x86/include/asm/kvm_host.h +++ b/arch/x86/include/asm/kvm_host.h @@ -35,8 +35,9 @@ #include <asm/asm.h> #include <asm/kvm_page_track.h> -#define KVM_MAX_VCPUS 255 -#define KVM_SOFT_MAX_VCPUS 160 +#define KVM_MAX_VCPUS 288 +#define KVM_SOFT_MAX_VCPUS 240 +#define KVM_MAX_VCPU_ID 1023 #define KVM_USER_MEM_SLOTS 509 /* memory slots that are not exposed to userspace */ #define KVM_PRIVATE_MEM_SLOTS 3 @@ -599,6 +600,7 @@ struct kvm_vcpu_arch { u64 mcg_cap; u64 mcg_status; u64 mcg_ctl; + u64 mcg_ext_ctl; u64 *mce_banks; /* Cache MMIO info */ @@ -682,9 +684,12 @@ struct kvm_arch_memory_slot { struct kvm_apic_map { struct rcu_head rcu; u8 mode; - struct kvm_lapic *phys_map[256]; - /* first index is cluster id second is cpu id in a cluster */ - struct kvm_lapic *logical_map[16][16]; + u32 max_apic_id; + union { + struct kvm_lapic *xapic_flat_map[8]; + struct kvm_lapic *xapic_cluster_map[16][4]; + }; + struct kvm_lapic *phys_map[]; }; /* Hyper-V emulation context */ @@ -779,6 +784,9 @@ struct kvm_arch { u32 ldr_mode; struct page *avic_logical_id_table_page; struct page *avic_physical_id_table_page; + + bool x2apic_format; + bool x2apic_broadcast_quirk_disabled; }; struct kvm_vm_stat { @@ -1006,6 +1014,11 @@ struct kvm_x86_ops { int (*update_pi_irte)(struct kvm *kvm, unsigned int host_irq, uint32_t guest_irq, bool set); void (*apicv_post_state_restore)(struct kvm_vcpu *vcpu); + + int (*set_hv_timer)(struct kvm_vcpu *vcpu, u64 guest_deadline_tsc); + void (*cancel_hv_timer)(struct kvm_vcpu *vcpu); + + void (*setup_mce)(struct kvm_vcpu *vcpu); }; struct kvm_arch_async_pf { @@ -1026,7 +1039,7 @@ void kvm_mmu_setup(struct kvm_vcpu *vcpu); void kvm_mmu_init_vm(struct kvm *kvm); void kvm_mmu_uninit_vm(struct kvm *kvm); void kvm_mmu_set_mask_ptes(u64 user_mask, u64 accessed_mask, - u64 dirty_mask, u64 nx_mask, u64 x_mask); + u64 dirty_mask, u64 nx_mask, u64 x_mask, u64 p_mask); void kvm_mmu_reset_context(struct kvm_vcpu *vcpu); void kvm_mmu_slot_remove_write_access(struct kvm *kvm, @@ -1077,6 +1090,10 @@ extern u32 kvm_max_guest_tsc_khz; extern u8 kvm_tsc_scaling_ratio_frac_bits; /* maximum allowed value of TSC scaling ratio */ extern u64 kvm_max_tsc_scaling_ratio; +/* 1ull << kvm_tsc_scaling_ratio_frac_bits */ +extern u64 kvm_default_tsc_scaling_ratio; + +extern u64 kvm_mce_cap_supported; enum emulation_result { EMULATE_DONE, /* no further processing */ @@ -1352,7 +1369,7 @@ bool kvm_vcpu_is_bsp(struct kvm_vcpu *vcpu); bool kvm_intr_is_single_vcpu(struct kvm *kvm, struct kvm_lapic_irq *irq, struct kvm_vcpu **dest_vcpu); -void kvm_set_msi_irq(struct kvm_kernel_irq_routing_entry *e, +void kvm_set_msi_irq(struct kvm *kvm, struct kvm_kernel_irq_routing_entry *e, struct kvm_lapic_irq *irq); static inline void kvm_arch_vcpu_blocking(struct kvm_vcpu *vcpu) diff --git a/arch/x86/include/asm/svm.h b/arch/x86/include/asm/svm.h index d0fe23e..14824fc 100644 --- a/arch/x86/include/asm/svm.h +++ b/arch/x86/include/asm/svm.h @@ -193,7 +193,6 @@ struct __attribute__ ((__packed__)) vmcb { struct vmcb_save_area save; }; -#define SVM_CPUID_FEATURE_SHIFT 2 #define SVM_CPUID_FUNC 0x8000000a #define SVM_VM_CR_SVM_DISABLE 4 diff --git a/arch/x86/include/asm/virtext.h b/arch/x86/include/asm/virtext.h index cce9ee6..0116b2e 100644 --- a/arch/x86/include/asm/virtext.h +++ b/arch/x86/include/asm/virtext.h @@ -83,23 +83,19 @@ static inline void cpu_emergency_vmxoff(void) */ static inline int cpu_has_svm(const char **msg) { - uint32_t eax, ebx, ecx, edx; - if (boot_cpu_data.x86_vendor != X86_VENDOR_AMD) { if (msg) *msg = "not amd"; return 0; } - cpuid(0x80000000, &eax, &ebx, &ecx, &edx); - if (eax < SVM_CPUID_FUNC) { + if (boot_cpu_data.extended_cpuid_level < SVM_CPUID_FUNC) { if (msg) *msg = "can't execute cpuid_8000000a"; return 0; } - cpuid(0x80000001, &eax, &ebx, &ecx, &edx); - if (!(ecx & (1 << SVM_CPUID_FEATURE_SHIFT))) { + if (!boot_cpu_has(X86_FEATURE_SVM)) { if (msg) *msg = "svm not available"; return 0; diff --git a/arch/x86/kvm/Kconfig b/arch/x86/kvm/Kconfig index 639a6e3..ab8e32f 100644 --- a/arch/x86/kvm/Kconfig +++ b/arch/x86/kvm/Kconfig @@ -32,7 +32,6 @@ config KVM select HAVE_KVM_IRQ_BYPASS select HAVE_KVM_IRQ_ROUTING select HAVE_KVM_EVENTFD - select KVM_APIC_ARCHITECTURE select KVM_ASYNC_PF select USER_RETURN_NOTIFIER select KVM_MMIO diff --git a/arch/x86/kvm/i8254.c b/arch/x86/kvm/i8254.c index a4bf5b4..5fb6c62 100644 --- a/arch/x86/kvm/i8254.c +++ b/arch/x86/kvm/i8254.c @@ -645,7 +645,6 @@ static const struct kvm_io_device_ops speaker_dev_ops = { .write = speaker_ioport_write, }; -/* Caller must hold slots_lock */ struct kvm_pit *kvm_create_pit(struct kvm *kvm, u32 flags) { struct kvm_pit *pit; @@ -690,6 +689,7 @@ struct kvm_pit *kvm_create_pit(struct kvm *kvm, u32 flags) kvm_pit_set_reinject(pit, true); + mutex_lock(&kvm->slots_lock); kvm_iodevice_init(&pit->dev, &pit_dev_ops); ret = kvm_io_bus_register_dev(kvm, KVM_PIO_BUS, KVM_PIT_BASE_ADDRESS, KVM_PIT_MEM_LENGTH, &pit->dev); @@ -704,12 +704,14 @@ struct kvm_pit *kvm_create_pit(struct kvm *kvm, u32 flags) if (ret < 0) goto fail_register_speaker; } + mutex_unlock(&kvm->slots_lock); return pit; fail_register_speaker: kvm_io_bus_unregister_dev(kvm, KVM_PIO_BUS, &pit->dev); fail_register_pit: + mutex_unlock(&kvm->slots_lock); kvm_pit_set_reinject(pit, false); kthread_stop(pit->worker_task); fail_kthread: diff --git a/arch/x86/kvm/iommu.c b/arch/x86/kvm/iommu.c index 95e0e64..b181426 100644 --- a/arch/x86/kvm/iommu.c +++ b/arch/x86/kvm/iommu.c @@ -28,9 +28,7 @@ #include <linux/moduleparam.h> #include <linux/pci.h> #include <linux/stat.h> -#include <linux/dmar.h> #include <linux/iommu.h> -#include <linux/intel-iommu.h> #include "assigned-dev.h" static bool allow_unsafe_assigned_interrupts; diff --git a/arch/x86/kvm/irq_comm.c b/arch/x86/kvm/irq_comm.c index dfb4c64..25810b1 100644 --- a/arch/x86/kvm/irq_comm.c +++ b/arch/x86/kvm/irq_comm.c @@ -110,13 +110,17 @@ int kvm_irq_delivery_to_apic(struct kvm *kvm, struct kvm_lapic *src, return r; } -void kvm_set_msi_irq(struct kvm_kernel_irq_routing_entry *e, +void kvm_set_msi_irq(struct kvm *kvm, struct kvm_kernel_irq_routing_entry *e, struct kvm_lapic_irq *irq) { - trace_kvm_msi_set_irq(e->msi.address_lo, e->msi.data); + trace_kvm_msi_set_irq(e->msi.address_lo | (kvm->arch.x2apic_format ? + (u64)e->msi.address_hi << 32 : 0), + e->msi.data); irq->dest_id = (e->msi.address_lo & MSI_ADDR_DEST_ID_MASK) >> MSI_ADDR_DEST_ID_SHIFT; + if (kvm->arch.x2apic_format) + irq->dest_id |= MSI_ADDR_EXT_DEST_ID(e->msi.address_hi); irq->vector = (e->msi.data & MSI_DATA_VECTOR_MASK) >> MSI_DATA_VECTOR_SHIFT; irq->dest_mode = (1 << MSI_ADDR_DEST_MODE_SHIFT) & e->msi.address_lo; @@ -129,15 +133,24 @@ void kvm_set_msi_irq(struct kvm_kernel_irq_routing_entry *e, } EXPORT_SYMBOL_GPL(kvm_set_msi_irq); +static inline bool kvm_msi_route_invalid(struct kvm *kvm, + struct kvm_kernel_irq_routing_entry *e) +{ + return kvm->arch.x2apic_format && (e->msi.address_hi & 0xff); +} + int kvm_set_msi(struct kvm_kernel_irq_routing_entry *e, struct kvm *kvm, int irq_source_id, int level, bool line_status) { struct kvm_lapic_irq irq; + if (kvm_msi_route_invalid(kvm, e)) + return -EINVAL; + if (!level) return -1; - kvm_set_msi_irq(e, &irq); + kvm_set_msi_irq(kvm, e, &irq); return kvm_irq_delivery_to_apic(kvm, NULL, &irq, NULL); } @@ -153,7 +166,10 @@ int kvm_arch_set_irq_inatomic(struct kvm_kernel_irq_routing_entry *e, if (unlikely(e->type != KVM_IRQ_ROUTING_MSI)) return -EWOULDBLOCK; - kvm_set_msi_irq(e, &irq); + if (kvm_msi_route_invalid(kvm, e)) + return -EINVAL; + + kvm_set_msi_irq(kvm, e, &irq); if (kvm_irq_delivery_to_apic_fast(kvm, NULL, &irq, &r, NULL)) return r; @@ -248,7 +264,8 @@ static int kvm_hv_set_sint(struct kvm_kernel_irq_routing_entry *e, return kvm_hv_synic_set_irq(kvm, e->hv_sint.vcpu, e->hv_sint.sint); } -int kvm_set_routing_entry(struct kvm_kernel_irq_routing_entry *e, +int kvm_set_routing_entry(struct kvm *kvm, + struct kvm_kernel_irq_routing_entry *e, const struct kvm_irq_routing_entry *ue) { int r = -EINVAL; @@ -285,6 +302,9 @@ int kvm_set_routing_entry(struct kvm_kernel_irq_routing_entry *e, e->msi.address_lo = ue->u.msi.address_lo; e->msi.address_hi = ue->u.msi.address_hi; e->msi.data = ue->u.msi.data; + + if (kvm_msi_route_invalid(kvm, e)) + goto out; break; case KVM_IRQ_ROUTING_HV_SINT: e->set = kvm_hv_set_sint; @@ -388,21 +408,16 @@ void kvm_scan_ioapic_routes(struct kvm_vcpu *vcpu, kvm->arch.nr_reserved_ioapic_pins); for (i = 0; i < nr_ioapic_pins; ++i) { hlist_for_each_entry(entry, &table->map[i], link) { - u32 dest_id, dest_mode; - bool level; + struct kvm_lapic_irq irq; if (entry->type != KVM_IRQ_ROUTING_MSI) continue; - dest_id = (entry->msi.address_lo >> 12) & 0xff; - dest_mode = (entry->msi.address_lo >> 2) & 0x1; - level = entry->msi.data & MSI_DATA_TRIGGER_LEVEL; - if (level && kvm_apic_match_dest(vcpu, NULL, 0, - dest_id, dest_mode)) { - u32 vector = entry->msi.data & 0xff; - - __set_bit(vector, - ioapic_handled_vectors); - } + + kvm_set_msi_irq(vcpu->kvm, entry, &irq); + + if (irq.level && kvm_apic_match_dest(vcpu, NULL, 0, + irq.dest_id, irq.dest_mode)) + __set_bit(irq.vector, ioapic_handled_vectors); } } srcu_read_unlock(&kvm->irq_srcu, idx); diff --git a/arch/x86/kvm/lapic.c b/arch/x86/kvm/lapic.c index 57549ed..730cf17 100644 --- a/arch/x86/kvm/lapic.c +++ b/arch/x86/kvm/lapic.c @@ -115,26 +115,43 @@ static inline int apic_enabled(struct kvm_lapic *apic) (LVT_MASK | APIC_MODE_MASK | APIC_INPUT_POLARITY | \ APIC_LVT_REMOTE_IRR | APIC_LVT_LEVEL_TRIGGER) -/* The logical map is definitely wrong if we have multiple - * modes at the same time. (Physical map is always right.) - */ -static inline bool kvm_apic_logical_map_valid(struct kvm_apic_map *map) -{ - return !(map->mode & (map->mode - 1)); +static inline bool kvm_apic_map_get_logical_dest(struct kvm_apic_map *map, + u32 dest_id, struct kvm_lapic ***cluster, u16 *mask) { + switch (map->mode) { + case KVM_APIC_MODE_X2APIC: { + u32 offset = (dest_id >> 16) * 16; + u32 max_apic_id = map->max_apic_id; + + if (offset <= max_apic_id) { + u8 cluster_size = min(max_apic_id - offset + 1, 16U); + + *cluster = &map->phys_map[offset]; + *mask = dest_id & (0xffff >> (16 - cluster_size)); + } else { + *mask = 0; + } + + return true; + } + case KVM_APIC_MODE_XAPIC_FLAT: + *cluster = map->xapic_flat_map; + *mask = dest_id & 0xff; + return true; + case KVM_APIC_MODE_XAPIC_CLUSTER: + *cluster = map->xapic_cluster_map[dest_id >> 4]; + *mask = dest_id & 0xf; + return true; + default: + /* Not optimized. */ + return false; + } } -static inline void -apic_logical_id(struct kvm_apic_map *map, u32 dest_id, u16 *cid, u16 *lid) +static void kvm_apic_map_free(struct rcu_head *rcu) { - unsigned lid_bits; + struct kvm_apic_map *map = container_of(rcu, struct kvm_apic_map, rcu); - BUILD_BUG_ON(KVM_APIC_MODE_XAPIC_CLUSTER != 4); - BUILD_BUG_ON(KVM_APIC_MODE_XAPIC_FLAT != 8); - BUILD_BUG_ON(KVM_APIC_MODE_X2APIC != 16); - lid_bits = map->mode; - - *cid = dest_id >> lid_bits; - *lid = dest_id & ((1 << lid_bits) - 1); + kvfree(map); } static void recalculate_apic_map(struct kvm *kvm) @@ -142,17 +159,26 @@ static void recalculate_apic_map(struct kvm *kvm) struct kvm_apic_map *new, *old = NULL; struct kvm_vcpu *vcpu; int i; - - new = kzalloc(sizeof(struct kvm_apic_map), GFP_KERNEL); + u32 max_id = 255; mutex_lock(&kvm->arch.apic_map_lock); + kvm_for_each_vcpu(i, vcpu, kvm) + if (kvm_apic_present(vcpu)) + max_id = max(max_id, kvm_apic_id(vcpu->arch.apic)); + + new = kvm_kvzalloc(sizeof(struct kvm_apic_map) + + sizeof(struct kvm_lapic *) * ((u64)max_id + 1)); + if (!new) goto out; + new->max_apic_id = max_id; + kvm_for_each_vcpu(i, vcpu, kvm) { struct kvm_lapic *apic = vcpu->arch.apic; - u16 cid, lid; + struct kvm_lapic **cluster; + u16 mask; u32 ldr, aid; if (!kvm_apic_present(vcpu)) @@ -161,7 +187,7 @@ static void recalculate_apic_map(struct kvm *kvm) aid = kvm_apic_id(apic); ldr = kvm_lapic_get_reg(apic, APIC_LDR); - if (aid < ARRAY_SIZE(new->phys_map)) + if (aid <= new->max_apic_id) new->phys_map[aid] = apic; if (apic_x2apic_mode(apic)) { @@ -174,13 +200,11 @@ static void recalculate_apic_map(struct kvm *kvm) new->mode |= KVM_APIC_MODE_XAPIC_CLUSTER; } - if (!kvm_apic_logical_map_valid(new)) + if (!kvm_apic_map_get_logical_dest(new, ldr, &cluster, &mask)) continue; - apic_logical_id(new, ldr, &cid, &lid); - - if (lid && cid < ARRAY_SIZE(new->logical_map)) - new->logical_map[cid][ffs(lid) - 1] = apic; + if (mask) + cluster[ffs(mask) - 1] = apic; } out: old = rcu_dereference_protected(kvm->arch.apic_map, @@ -189,7 +213,7 @@ out: mutex_unlock(&kvm->arch.apic_map_lock); if (old) - kfree_rcu(old, rcu); + call_rcu(&old->rcu, kvm_apic_map_free); kvm_make_scan_ioapic_request(kvm); } @@ -210,7 +234,7 @@ static inline void apic_set_spiv(struct kvm_lapic *apic, u32 val) } } -static inline void kvm_apic_set_id(struct kvm_lapic *apic, u8 id) +static inline void kvm_apic_set_xapic_id(struct kvm_lapic *apic, u8 id) { kvm_lapic_set_reg(apic, APIC_ID, id << 24); recalculate_apic_map(apic->vcpu->kvm); @@ -222,11 +246,11 @@ static inline void kvm_apic_set_ldr(struct kvm_lapic *apic, u32 id) recalculate_apic_map(apic->vcpu->kvm); } -static inline void kvm_apic_set_x2apic_id(struct kvm_lapic *apic, u8 id) +static inline void kvm_apic_set_x2apic_id(struct kvm_lapic *apic, u32 id) { u32 ldr = ((id >> 4) << 16) | (1 << (id & 0xf)); - kvm_lapic_set_reg(apic, APIC_ID, id << 24); + kvm_lapic_set_reg(apic, APIC_ID, id); kvm_lapic_set_reg(apic, APIC_LDR, ldr); recalculate_apic_map(apic->vcpu->kvm); } @@ -599,17 +623,30 @@ static bool kvm_apic_match_logical_addr(struct kvm_lapic *apic, u32 mda) } } -/* KVM APIC implementation has two quirks - * - dest always begins at 0 while xAPIC MDA has offset 24, - * - IOxAPIC messages have to be delivered (directly) to x2APIC. +/* The KVM local APIC implementation has two quirks: + * + * - the xAPIC MDA stores the destination at bits 24-31, while this + * is not true of struct kvm_lapic_irq's dest_id field. This is + * just a quirk in the API and is not problematic. + * + * - in-kernel IOAPIC messages have to be delivered directly to + * x2APIC, because the kernel does not support interrupt remapping. + * In order to support broadcast without interrupt remapping, x2APIC + * rewrites the destination of non-IPI messages from APIC_BROADCAST + * to X2APIC_BROADCAST. + * + * The broadcast quirk can be disabled with KVM_CAP_X2APIC_API. This is + * important when userspace wants to use x2APIC-format MSIs, because + * APIC_BROADCAST (0xff) is a legal route for "cluster 0, CPUs 0-7". */ -static u32 kvm_apic_mda(unsigned int dest_id, struct kvm_lapic *source, - struct kvm_lapic *target) +static u32 kvm_apic_mda(struct kvm_vcpu *vcpu, unsigned int dest_id, + struct kvm_lapic *source, struct kvm_lapic *target) { bool ipi = source != NULL; bool x2apic_mda = apic_x2apic_mode(ipi ? source : target); - if (!ipi && dest_id == APIC_BROADCAST && x2apic_mda) + if (!vcpu->kvm->arch.x2apic_broadcast_quirk_disabled && + !ipi && dest_id == APIC_BROADCAST && x2apic_mda) return X2APIC_BROADCAST; return x2apic_mda ? dest_id : SET_APIC_DEST_FIELD(dest_id); @@ -619,7 +656,7 @@ bool kvm_apic_match_dest(struct kvm_vcpu *vcpu, struct kvm_lapic *source, int short_hand, unsigned int dest, int dest_mode) { struct kvm_lapic *target = vcpu->arch.apic; - u32 mda = kvm_apic_mda(dest, source, target); + u32 mda = kvm_apic_mda(vcpu, dest, source, target); apic_debug("target %p, source %p, dest 0x%x, " "dest_mode 0x%x, short_hand 0x%x\n", @@ -671,102 +708,126 @@ static void kvm_apic_disabled_lapic_found(struct kvm *kvm) } } -bool kvm_irq_delivery_to_apic_fast(struct kvm *kvm, struct kvm_lapic *src, - struct kvm_lapic_irq *irq, int *r, struct dest_map *dest_map) +static bool kvm_apic_is_broadcast_dest(struct kvm *kvm, struct kvm_lapic **src, + struct kvm_lapic_irq *irq, struct kvm_apic_map *map) { - struct kvm_apic_map *map; - unsigned long bitmap = 1; - struct kvm_lapic **dst; - int i; - bool ret, x2apic_ipi; + if (kvm->arch.x2apic_broadcast_quirk_disabled) { + if ((irq->dest_id == APIC_BROADCAST && + map->mode != KVM_APIC_MODE_X2APIC)) + return true; + if (irq->dest_id == X2APIC_BROADCAST) + return true; + } else { + bool x2apic_ipi = src && *src && apic_x2apic_mode(*src); + if (irq->dest_id == (x2apic_ipi ? + X2APIC_BROADCAST : APIC_BROADCAST)) + return true; + } - *r = -1; + return false; +} - if (irq->shorthand == APIC_DEST_SELF) { - *r = kvm_apic_set_irq(src->vcpu, irq, dest_map); - return true; - } +/* Return true if the interrupt can be handled by using *bitmap as index mask + * for valid destinations in *dst array. + * Return false if kvm_apic_map_get_dest_lapic did nothing useful. + * Note: we may have zero kvm_lapic destinations when we return true, which + * means that the interrupt should be dropped. In this case, *bitmap would be + * zero and *dst undefined. + */ +static inline bool kvm_apic_map_get_dest_lapic(struct kvm *kvm, + struct kvm_lapic **src, struct kvm_lapic_irq *irq, + struct kvm_apic_map *map, struct kvm_lapic ***dst, + unsigned long *bitmap) +{ + int i, lowest; - if (irq->shorthand) + if (irq->shorthand == APIC_DEST_SELF && src) { + *dst = src; + *bitmap = 1; + return true; + } else if (irq->shorthand) return false; - x2apic_ipi = src && apic_x2apic_mode(src); - if (irq->dest_id == (x2apic_ipi ? X2APIC_BROADCAST : APIC_BROADCAST)) + if (!map || kvm_apic_is_broadcast_dest(kvm, src, irq, map)) return false; - ret = true; - rcu_read_lock(); - map = rcu_dereference(kvm->arch.apic_map); - - if (!map) { - ret = false; - goto out; + if (irq->dest_mode == APIC_DEST_PHYSICAL) { + if (irq->dest_id > map->max_apic_id) { + *bitmap = 0; + } else { + *dst = &map->phys_map[irq->dest_id]; + *bitmap = 1; + } + return true; } - if (irq->dest_mode == APIC_DEST_PHYSICAL) { - if (irq->dest_id >= ARRAY_SIZE(map->phys_map)) - goto out; + *bitmap = 0; + if (!kvm_apic_map_get_logical_dest(map, irq->dest_id, dst, + (u16 *)bitmap)) + return false; - dst = &map->phys_map[irq->dest_id]; - } else { - u16 cid; + if (!kvm_lowest_prio_delivery(irq)) + return true; - if (!kvm_apic_logical_map_valid(map)) { - ret = false; - goto out; + if (!kvm_vector_hashing_enabled()) { + lowest = -1; + for_each_set_bit(i, bitmap, 16) { + if (!(*dst)[i]) + continue; + if (lowest < 0) + lowest = i; + else if (kvm_apic_compare_prio((*dst)[i]->vcpu, + (*dst)[lowest]->vcpu) < 0) + lowest = i; } + } else { + if (!*bitmap) + return true; - apic_logical_id(map, irq->dest_id, &cid, (u16 *)&bitmap); + lowest = kvm_vector_to_index(irq->vector, hweight16(*bitmap), + bitmap, 16); - if (cid >= ARRAY_SIZE(map->logical_map)) - goto out; + if (!(*dst)[lowest]) { + kvm_apic_disabled_lapic_found(kvm); + *bitmap = 0; + return true; + } + } - dst = map->logical_map[cid]; + *bitmap = (lowest >= 0) ? 1 << lowest : 0; - if (!kvm_lowest_prio_delivery(irq)) - goto set_irq; + return true; +} - if (!kvm_vector_hashing_enabled()) { - int l = -1; - for_each_set_bit(i, &bitmap, 16) { - if (!dst[i]) - continue; - if (l < 0) - l = i; - else if (kvm_apic_compare_prio(dst[i]->vcpu, - dst[l]->vcpu) < 0) - l = i; - } - bitmap = (l >= 0) ? 1 << l : 0; - } else { - int idx; - unsigned int dest_vcpus; +bool kvm_irq_delivery_to_apic_fast(struct kvm *kvm, struct kvm_lapic *src, + struct kvm_lapic_irq *irq, int *r, struct dest_map *dest_map) +{ + struct kvm_apic_map *map; + unsigned long bitmap; + struct kvm_lapic **dst = NULL; + int i; + bool ret; - dest_vcpus = hweight16(bitmap); - if (dest_vcpus == 0) - goto out; + *r = -1; - idx = kvm_vector_to_index(irq->vector, - dest_vcpus, &bitmap, 16); + if (irq->shorthand == APIC_DEST_SELF) { + *r = kvm_apic_set_irq(src->vcpu, irq, dest_map); + return true; + } - if (!dst[idx]) { - kvm_apic_disabled_lapic_found(kvm); - goto out; - } + rcu_read_lock(); + map = rcu_dereference(kvm->arch.apic_map); - bitmap = (idx >= 0) ? 1 << idx : 0; + ret = kvm_apic_map_get_dest_lapic(kvm, &src, irq, map, &dst, &bitmap); + if (ret) + for_each_set_bit(i, &bitmap, 16) { + if (!dst[i]) + continue; + if (*r < 0) + *r = 0; + *r += kvm_apic_set_irq(dst[i]->vcpu, irq, dest_map); } - } -set_irq: - for_each_set_bit(i, &bitmap, 16) { - if (!dst[i]) - continue; - if (*r < 0) - *r = 0; - *r += kvm_apic_set_irq(dst[i]->vcpu, irq, dest_map); - } -out: rcu_read_unlock(); return ret; } @@ -789,8 +850,9 @@ bool kvm_intr_is_single_vcpu_fast(struct kvm *kvm, struct kvm_lapic_irq *irq, struct kvm_vcpu **dest_vcpu) { struct kvm_apic_map *map; + unsigned long bitmap; + struct kvm_lapic **dst = NULL; bool ret = false; - struct kvm_lapic *dst = NULL; if (irq->shorthand) return false; @@ -798,69 +860,16 @@ bool kvm_intr_is_single_vcpu_fast(struct kvm *kvm, struct kvm_lapic_irq *irq, rcu_read_lock(); map = rcu_dereference(kvm->arch.apic_map); - if (!map) - goto out; - - if (irq->dest_mode == APIC_DEST_PHYSICAL) { - if (irq->dest_id == 0xFF) - goto out; - - if (irq->dest_id >= ARRAY_SIZE(map->phys_map)) - goto out; - - dst = map->phys_map[irq->dest_id]; - if (dst && kvm_apic_present(dst->vcpu)) - *dest_vcpu = dst->vcpu; - else - goto out; - } else { - u16 cid; - unsigned long bitmap = 1; - int i, r = 0; - - if (!kvm_apic_logical_map_valid(map)) - goto out; - - apic_logical_id(map, irq->dest_id, &cid, (u16 *)&bitmap); - - if (cid >= ARRAY_SIZE(map->logical_map)) - goto out; - - if (kvm_vector_hashing_enabled() && - kvm_lowest_prio_delivery(irq)) { - int idx; - unsigned int dest_vcpus; + if (kvm_apic_map_get_dest_lapic(kvm, NULL, irq, map, &dst, &bitmap) && + hweight16(bitmap) == 1) { + unsigned long i = find_first_bit(&bitmap, 16); - dest_vcpus = hweight16(bitmap); - if (dest_vcpus == 0) - goto out; - - idx = kvm_vector_to_index(irq->vector, dest_vcpus, - &bitmap, 16); - - dst = map->logical_map[cid][idx]; - if (!dst) { - kvm_apic_disabled_lapic_found(kvm); - goto out; - } - - *dest_vcpu = dst->vcpu; - } else { - for_each_set_bit(i, &bitmap, 16) { - dst = map->logical_map[cid][i]; - if (++r == 2) - goto out; - } - - if (dst && kvm_apic_present(dst->vcpu)) - *dest_vcpu = dst->vcpu; - else - goto out; + if (dst[i]) { + *dest_vcpu = dst[i]->vcpu; + ret = true; } } - ret = true; -out: rcu_read_unlock(); return ret; } @@ -1127,12 +1136,6 @@ static u32 __apic_read(struct kvm_lapic *apic, unsigned int offset) return 0; switch (offset) { - case APIC_ID: - if (apic_x2apic_mode(apic)) - val = kvm_apic_id(apic); - else - val = kvm_apic_id(apic) << 24; - break; case APIC_ARBPRI: apic_debug("Access APIC ARBPRI register which is for P6\n"); break; @@ -1314,6 +1317,108 @@ void wait_lapic_expire(struct kvm_vcpu *vcpu) nsec_to_cycles(vcpu, lapic_timer_advance_ns))); } +static void start_sw_tscdeadline(struct kvm_lapic *apic) +{ + u64 guest_tsc, tscdeadline = apic->lapic_timer.tscdeadline; + u64 ns = 0; + ktime_t expire; + struct kvm_vcpu *vcpu = apic->vcpu; + unsigned long this_tsc_khz = vcpu->arch.virtual_tsc_khz; + unsigned long flags; + ktime_t now; + + if (unlikely(!tscdeadline || !this_tsc_khz)) + return; + + local_irq_save(flags); + + now = apic->lapic_timer.timer.base->get_time(); + guest_tsc = kvm_read_l1_tsc(vcpu, rdtsc()); + if (likely(tscdeadline > guest_tsc)) { + ns = (tscdeadline - guest_tsc) * 1000000ULL; + do_div(ns, this_tsc_khz); + expire = ktime_add_ns(now, ns); + expire = ktime_sub_ns(expire, lapic_timer_advance_ns); + hrtimer_start(&apic->lapic_timer.timer, + expire, HRTIMER_MODE_ABS_PINNED); + } else + apic_timer_expired(apic); + + local_irq_restore(flags); +} + +bool kvm_lapic_hv_timer_in_use(struct kvm_vcpu *vcpu) +{ + return vcpu->arch.apic->lapic_timer.hv_timer_in_use; +} +EXPORT_SYMBOL_GPL(kvm_lapic_hv_timer_in_use); + +static void cancel_hv_tscdeadline(struct kvm_lapic *apic) +{ + kvm_x86_ops->cancel_hv_timer(apic->vcpu); + apic->lapic_timer.hv_timer_in_use = false; +} + +void kvm_lapic_expired_hv_timer(struct kvm_vcpu *vcpu) +{ + struct kvm_lapic *apic = vcpu->arch.apic; + + WARN_ON(!apic->lapic_timer.hv_timer_in_use); + WARN_ON(swait_active(&vcpu->wq)); + cancel_hv_tscdeadline(apic); + apic_timer_expired(apic); +} +EXPORT_SYMBOL_GPL(kvm_lapic_expired_hv_timer); + +static bool start_hv_tscdeadline(struct kvm_lapic *apic) +{ + u64 tscdeadline = apic->lapic_timer.tscdeadline; + + if (atomic_read(&apic->lapic_timer.pending) || + kvm_x86_ops->set_hv_timer(apic->vcpu, tscdeadline)) { + if (apic->lapic_timer.hv_timer_in_use) + cancel_hv_tscdeadline(apic); + } else { + apic->lapic_timer.hv_timer_in_use = true; + hrtimer_cancel(&apic->lapic_timer.timer); + + /* In case the sw timer triggered in the window */ + if (atomic_read(&apic->lapic_timer.pending)) + cancel_hv_tscdeadline(apic); + } + trace_kvm_hv_timer_state(apic->vcpu->vcpu_id, + apic->lapic_timer.hv_timer_in_use); + return apic->lapic_timer.hv_timer_in_use; +} + +void kvm_lapic_switch_to_hv_timer(struct kvm_vcpu *vcpu) +{ + struct kvm_lapic *apic = vcpu->arch.apic; + + WARN_ON(apic->lapic_timer.hv_timer_in_use); + + if (apic_lvtt_tscdeadline(apic)) + start_hv_tscdeadline(apic); +} +EXPORT_SYMBOL_GPL(kvm_lapic_switch_to_hv_timer); + +void kvm_lapic_switch_to_sw_timer(struct kvm_vcpu *vcpu) +{ + struct kvm_lapic *apic = vcpu->arch.apic; + + /* Possibly the TSC deadline timer is not enabled yet */ + if (!apic->lapic_timer.hv_timer_in_use) + return; + + cancel_hv_tscdeadline(apic); + + if (atomic_read(&apic->lapic_timer.pending)) + return; + + start_sw_tscdeadline(apic); +} +EXPORT_SYMBOL_GPL(kvm_lapic_switch_to_sw_timer); + static void start_apic_timer(struct kvm_lapic *apic) { ktime_t now; @@ -1360,32 +1465,8 @@ static void start_apic_timer(struct kvm_lapic *apic) ktime_to_ns(ktime_add_ns(now, apic->lapic_timer.period))); } else if (apic_lvtt_tscdeadline(apic)) { - /* lapic timer in tsc deadline mode */ - u64 guest_tsc, tscdeadline = apic->lapic_timer.tscdeadline; - u64 ns = 0; - ktime_t expire; - struct kvm_vcpu *vcpu = apic->vcpu; - unsigned long this_tsc_khz = vcpu->arch.virtual_tsc_khz; - unsigned long flags; - - if (unlikely(!tscdeadline || !this_tsc_khz)) - return; - - local_irq_save(flags); - - now = apic->lapic_timer.timer.base->get_time(); - guest_tsc = kvm_read_l1_tsc(vcpu, rdtsc()); - if (likely(tscdeadline > guest_tsc)) { - ns = (tscdeadline - guest_tsc) * 1000000ULL; - do_div(ns, this_tsc_khz); - expire = ktime_add_ns(now, ns); - expire = ktime_sub_ns(expire, lapic_timer_advance_ns); - hrtimer_start(&apic->lapic_timer.timer, - expire, HRTIMER_MODE_ABS_PINNED); - } else - apic_timer_expired(apic); - - local_irq_restore(flags); + if (!(kvm_x86_ops->set_hv_timer && start_hv_tscdeadline(apic))) + start_sw_tscdeadline(apic); } } @@ -1413,7 +1494,7 @@ int kvm_lapic_reg_write(struct kvm_lapic *apic, u32 reg, u32 val) switch (reg) { case APIC_ID: /* Local APIC ID */ if (!apic_x2apic_mode(apic)) - kvm_apic_set_id(apic, val >> 24); + kvm_apic_set_xapic_id(apic, val >> 24); else ret = 1; break; @@ -1674,9 +1755,10 @@ void kvm_lapic_set_base(struct kvm_vcpu *vcpu, u64 value) /* update jump label if enable bit changes */ if ((old_value ^ value) & MSR_IA32_APICBASE_ENABLE) { - if (value & MSR_IA32_APICBASE_ENABLE) + if (value & MSR_IA32_APICBASE_ENABLE) { + kvm_apic_set_xapic_id(apic, vcpu->vcpu_id); static_key_slow_dec_deferred(&apic_hw_disabled); - else + } else static_key_slow_inc(&apic_hw_disabled.key); recalculate_apic_map(vcpu->kvm); } @@ -1716,8 +1798,11 @@ void kvm_lapic_reset(struct kvm_vcpu *vcpu, bool init_event) /* Stop the timer in case it's a reset to an active apic */ hrtimer_cancel(&apic->lapic_timer.timer); - if (!init_event) - kvm_apic_set_id(apic, vcpu->vcpu_id); + if (!init_event) { + kvm_lapic_set_base(vcpu, APIC_DEFAULT_PHYS_BASE | + MSR_IA32_APICBASE_ENABLE); + kvm_apic_set_xapic_id(apic, vcpu->vcpu_id); + } kvm_apic_set_version(apic->vcpu); for (i = 0; i < KVM_APIC_LVT_NUM; i++) @@ -1856,9 +1941,6 @@ int kvm_create_lapic(struct kvm_vcpu *vcpu) * thinking that APIC satet has changed. */ vcpu->arch.apic_base = MSR_IA32_APICBASE_ENABLE; - kvm_lapic_set_base(vcpu, - APIC_DEFAULT_PHYS_BASE | MSR_IA32_APICBASE_ENABLE); - static_key_slow_inc(&apic_sw_disabled.key); /* sw disabled at reset */ kvm_lapic_reset(vcpu, false); kvm_iodevice_init(&apic->dev, &apic_mmio_ops); @@ -1938,17 +2020,48 @@ int kvm_get_apic_interrupt(struct kvm_vcpu *vcpu) return vector; } -void kvm_apic_post_state_restore(struct kvm_vcpu *vcpu, - struct kvm_lapic_state *s) +static int kvm_apic_state_fixup(struct kvm_vcpu *vcpu, + struct kvm_lapic_state *s, bool set) +{ + if (apic_x2apic_mode(vcpu->arch.apic)) { + u32 *id = (u32 *)(s->regs + APIC_ID); + + if (vcpu->kvm->arch.x2apic_format) { + if (*id != vcpu->vcpu_id) + return -EINVAL; + } else { + if (set) + *id >>= 24; + else + *id <<= 24; + } + } + + return 0; +} + +int kvm_apic_get_state(struct kvm_vcpu *vcpu, struct kvm_lapic_state *s) +{ + memcpy(s->regs, vcpu->arch.apic->regs, sizeof(*s)); + return kvm_apic_state_fixup(vcpu, s, false); +} + +int kvm_apic_set_state(struct kvm_vcpu *vcpu, struct kvm_lapic_state *s) { struct kvm_lapic *apic = vcpu->arch.apic; + int r; + kvm_lapic_set_base(vcpu, vcpu->arch.apic_base); /* set SPIV separately to get count of SW disabled APICs right */ apic_set_spiv(apic, *((u32 *)(s->regs + APIC_SPIV))); + + r = kvm_apic_state_fixup(vcpu, s, true); + if (r) + return r; memcpy(vcpu->arch.apic->regs, s->regs, sizeof *s); - /* call kvm_apic_set_id() to put apic into apic_map */ - kvm_apic_set_id(apic, kvm_apic_id(apic)); + + recalculate_apic_map(vcpu->kvm); kvm_apic_set_version(vcpu); apic_update_ppr(apic); @@ -1974,6 +2087,8 @@ void kvm_apic_post_state_restore(struct kvm_vcpu *vcpu, kvm_rtc_eoi_tracking_restore_one(vcpu); vcpu->arch.apic_arb_prio = 0; + + return 0; } void __kvm_migrate_apic_timer(struct kvm_vcpu *vcpu) diff --git a/arch/x86/kvm/lapic.h b/arch/x86/kvm/lapic.h index 891c6da..f60d01c 100644 --- a/arch/x86/kvm/lapic.h +++ b/arch/x86/kvm/lapic.h @@ -20,6 +20,7 @@ struct kvm_timer { u64 tscdeadline; u64 expired_tscdeadline; atomic_t pending; /* accumulated triggered timers */ + bool hv_timer_in_use; }; struct kvm_lapic { @@ -80,8 +81,8 @@ bool kvm_irq_delivery_to_apic_fast(struct kvm *kvm, struct kvm_lapic *src, u64 kvm_get_apic_base(struct kvm_vcpu *vcpu); int kvm_set_apic_base(struct kvm_vcpu *vcpu, struct msr_data *msr_info); -void kvm_apic_post_state_restore(struct kvm_vcpu *vcpu, - struct kvm_lapic_state *s); +int kvm_apic_get_state(struct kvm_vcpu *vcpu, struct kvm_lapic_state *s); +int kvm_apic_set_state(struct kvm_vcpu *vcpu, struct kvm_lapic_state *s); int kvm_lapic_find_highest_irr(struct kvm_vcpu *vcpu); u64 kvm_get_lapic_tscdeadline_msr(struct kvm_vcpu *vcpu); @@ -199,9 +200,15 @@ static inline int kvm_lapic_latched_init(struct kvm_vcpu *vcpu) return lapic_in_kernel(vcpu) && test_bit(KVM_APIC_INIT, &vcpu->arch.apic->pending_events); } -static inline int kvm_apic_id(struct kvm_lapic *apic) +static inline u32 kvm_apic_id(struct kvm_lapic *apic) { - return (kvm_lapic_get_reg(apic, APIC_ID) >> 24) & 0xff; + /* To avoid a race between apic_base and following APIC_ID update when + * switching to x2apic_mode, the x2apic mode returns initial x2apic id. + */ + if (apic_x2apic_mode(apic)) + return apic->vcpu->vcpu_id; + + return kvm_lapic_get_reg(apic, APIC_ID) >> 24; } bool kvm_apic_pending_eoi(struct kvm_vcpu *vcpu, int vector); @@ -212,4 +219,8 @@ bool kvm_intr_is_single_vcpu_fast(struct kvm *kvm, struct kvm_lapic_irq *irq, struct kvm_vcpu **dest_vcpu); int kvm_vector_to_index(u32 vector, u32 dest_vcpus, const unsigned long *bitmap, u32 bitmap_size); +void kvm_lapic_switch_to_sw_timer(struct kvm_vcpu *vcpu); +void kvm_lapic_switch_to_hv_timer(struct kvm_vcpu *vcpu); +void kvm_lapic_expired_hv_timer(struct kvm_vcpu *vcpu); +bool kvm_lapic_hv_timer_in_use(struct kvm_vcpu *vcpu); #endif diff --git a/arch/x86/kvm/mmu.c b/arch/x86/kvm/mmu.c index 745a5f4..3d4cc8cc 100644 --- a/arch/x86/kvm/mmu.c +++ b/arch/x86/kvm/mmu.c @@ -176,6 +176,7 @@ static u64 __read_mostly shadow_user_mask; static u64 __read_mostly shadow_accessed_mask; static u64 __read_mostly shadow_dirty_mask; static u64 __read_mostly shadow_mmio_mask; +static u64 __read_mostly shadow_present_mask; static void mmu_spte_set(u64 *sptep, u64 spte); static void mmu_free_roots(struct kvm_vcpu *vcpu); @@ -283,13 +284,14 @@ static bool check_mmio_spte(struct kvm_vcpu *vcpu, u64 spte) } void kvm_mmu_set_mask_ptes(u64 user_mask, u64 accessed_mask, - u64 dirty_mask, u64 nx_mask, u64 x_mask) + u64 dirty_mask, u64 nx_mask, u64 x_mask, u64 p_mask) { shadow_user_mask = user_mask; shadow_accessed_mask = accessed_mask; shadow_dirty_mask = dirty_mask; shadow_nx_mask = nx_mask; shadow_x_mask = x_mask; + shadow_present_mask = p_mask; } EXPORT_SYMBOL_GPL(kvm_mmu_set_mask_ptes); @@ -305,7 +307,7 @@ static int is_nx(struct kvm_vcpu *vcpu) static int is_shadow_present_pte(u64 pte) { - return pte & PT_PRESENT_MASK && !is_mmio_spte(pte); + return (pte & 0xFFFFFFFFull) && !is_mmio_spte(pte); } static int is_large_pte(u64 pte) @@ -524,7 +526,7 @@ static void mmu_spte_set(u64 *sptep, u64 new_spte) } /* Rules for using mmu_spte_update: - * Update the state bits, it means the mapped pfn is not changged. + * Update the state bits, it means the mapped pfn is not changed. * * Whenever we overwrite a writable spte with a read-only one we * should flush remote TLBs. Otherwise rmap_write_protect @@ -2246,10 +2248,9 @@ static void link_shadow_page(struct kvm_vcpu *vcpu, u64 *sptep, { u64 spte; - BUILD_BUG_ON(VMX_EPT_READABLE_MASK != PT_PRESENT_MASK || - VMX_EPT_WRITABLE_MASK != PT_WRITABLE_MASK); + BUILD_BUG_ON(VMX_EPT_WRITABLE_MASK != PT_WRITABLE_MASK); - spte = __pa(sp->spt) | PT_PRESENT_MASK | PT_WRITABLE_MASK | + spte = __pa(sp->spt) | shadow_present_mask | PT_WRITABLE_MASK | shadow_user_mask | shadow_x_mask | shadow_accessed_mask; mmu_spte_set(sptep, spte); @@ -2516,13 +2517,19 @@ static int set_spte(struct kvm_vcpu *vcpu, u64 *sptep, gfn_t gfn, kvm_pfn_t pfn, bool speculative, bool can_unsync, bool host_writable) { - u64 spte; + u64 spte = 0; int ret = 0; if (set_mmio_spte(vcpu, sptep, gfn, pfn, pte_access)) return 0; - spte = PT_PRESENT_MASK; + /* + * For the EPT case, shadow_present_mask is 0 if hardware + * supports exec-only page table entries. In that case, + * ACC_USER_MASK and shadow_user_mask are used to represent + * read access. See FNAME(gpte_access) in paging_tmpl.h. + */ + spte |= shadow_present_mask; if (!speculative) spte |= shadow_accessed_mask; @@ -3190,7 +3197,7 @@ static int mmu_alloc_shadow_roots(struct kvm_vcpu *vcpu) MMU_WARN_ON(VALID_PAGE(root)); if (vcpu->arch.mmu.root_level == PT32E_ROOT_LEVEL) { pdptr = vcpu->arch.mmu.get_pdptr(vcpu, i); - if (!is_present_gpte(pdptr)) { + if (!(pdptr & PT_PRESENT_MASK)) { vcpu->arch.mmu.pae_root[i] = 0; continue; } @@ -3915,9 +3922,7 @@ static void update_permission_bitmask(struct kvm_vcpu *vcpu, * clearer. */ smap = cr4_smap && u && !uf && !ff; - } else - /* Not really needed: no U/S accesses on ept */ - u = 1; + } fault = (ff && !x) || (uf && !u) || (wf && !w) || (smapf && smap); diff --git a/arch/x86/kvm/mmu.h b/arch/x86/kvm/mmu.h index 66b33b9..ddc56e9 100644 --- a/arch/x86/kvm/mmu.h +++ b/arch/x86/kvm/mmu.h @@ -93,11 +93,6 @@ static inline int kvm_mmu_reload(struct kvm_vcpu *vcpu) return kvm_mmu_load(vcpu); } -static inline int is_present_gpte(unsigned long pte) -{ - return pte & PT_PRESENT_MASK; -} - /* * Currently, we have two sorts of write-protection, a) the first one * write-protects guest page to sync the guest modification, b) another one is diff --git a/arch/x86/kvm/paging_tmpl.h b/arch/x86/kvm/paging_tmpl.h index bc019f7..a011054 100644 --- a/arch/x86/kvm/paging_tmpl.h +++ b/arch/x86/kvm/paging_tmpl.h @@ -131,7 +131,7 @@ static inline void FNAME(protect_clean_gpte)(unsigned *access, unsigned gpte) static inline int FNAME(is_present_gpte)(unsigned long pte) { #if PTTYPE != PTTYPE_EPT - return is_present_gpte(pte); + return pte & PT_PRESENT_MASK; #else return pte & 7; #endif @@ -181,13 +181,19 @@ no_present: return true; } +/* + * For PTTYPE_EPT, a page table can be executable but not readable + * on supported processors. Therefore, set_spte does not automatically + * set bit 0 if execute only is supported. Here, we repurpose ACC_USER_MASK + * to signify readability since it isn't used in the EPT case + */ static inline unsigned FNAME(gpte_access)(struct kvm_vcpu *vcpu, u64 gpte) { unsigned access; #if PTTYPE == PTTYPE_EPT access = ((gpte & VMX_EPT_WRITABLE_MASK) ? ACC_WRITE_MASK : 0) | ((gpte & VMX_EPT_EXECUTABLE_MASK) ? ACC_EXEC_MASK : 0) | - ACC_USER_MASK; + ((gpte & VMX_EPT_READABLE_MASK) ? ACC_USER_MASK : 0); #else BUILD_BUG_ON(ACC_EXEC_MASK != PT_PRESENT_MASK); BUILD_BUG_ON(ACC_EXEC_MASK != 1); diff --git a/arch/x86/kvm/pmu_intel.c b/arch/x86/kvm/pmu_intel.c index ab38af4..9d4a850 100644 --- a/arch/x86/kvm/pmu_intel.c +++ b/arch/x86/kvm/pmu_intel.c @@ -93,7 +93,7 @@ static unsigned intel_find_fixed_event(int idx) return intel_arch_events[fixed_pmc_events[idx]].event_type; } -/* check if a PMC is enabled by comparising it with globl_ctrl bits. */ +/* check if a PMC is enabled by comparing it with globl_ctrl bits. */ static bool intel_pmc_is_enabled(struct kvm_pmc *pmc) { struct kvm_pmu *pmu = pmc_to_pmu(pmc); diff --git a/arch/x86/kvm/svm.c b/arch/x86/kvm/svm.c index 16ef31b..af523d8 100644 --- a/arch/x86/kvm/svm.c +++ b/arch/x86/kvm/svm.c @@ -1577,7 +1577,7 @@ static unsigned long svm_get_rflags(struct kvm_vcpu *vcpu) static void svm_set_rflags(struct kvm_vcpu *vcpu, unsigned long rflags) { /* - * Any change of EFLAGS.VM is accompained by a reload of SS + * Any change of EFLAGS.VM is accompanied by a reload of SS * (caused by either a task switch or an inter-privilege IRET), * so we do not need to update the CPL here. */ @@ -4940,6 +4940,12 @@ out: static void svm_handle_external_intr(struct kvm_vcpu *vcpu) { local_irq_enable(); + /* + * We must have an instruction with interrupts enabled, so + * the timer interrupt isn't delayed by the interrupt shadow. + */ + asm("nop"); + local_irq_disable(); } static void svm_sched_in(struct kvm_vcpu *vcpu, int cpu) diff --git a/arch/x86/kvm/trace.h b/arch/x86/kvm/trace.h index 8de9250..0a6cc67 100644 --- a/arch/x86/kvm/trace.h +++ b/arch/x86/kvm/trace.h @@ -1348,6 +1348,21 @@ TRACE_EVENT(kvm_avic_unaccelerated_access, __entry->vec) ); +TRACE_EVENT(kvm_hv_timer_state, + TP_PROTO(unsigned int vcpu_id, unsigned int hv_timer_in_use), + TP_ARGS(vcpu_id, hv_timer_in_use), + TP_STRUCT__entry( + __field(unsigned int, vcpu_id) + __field(unsigned int, hv_timer_in_use) + ), + TP_fast_assign( + __entry->vcpu_id = vcpu_id; + __entry->hv_timer_in_use = hv_timer_in_use; + ), + TP_printk("vcpu_id %x hv_timer %x\n", + __entry->vcpu_id, + __entry->hv_timer_in_use) +); #endif /* _TRACE_KVM_H */ #undef TRACE_INCLUDE_PATH diff --git a/arch/x86/kvm/vmx.c b/arch/x86/kvm/vmx.c index df07a0a..bc354f0 100644 --- a/arch/x86/kvm/vmx.c +++ b/arch/x86/kvm/vmx.c @@ -110,6 +110,13 @@ module_param_named(pml, enable_pml, bool, S_IRUGO); #define KVM_VMX_TSC_MULTIPLIER_MAX 0xffffffffffffffffULL +/* Guest_tsc -> host_tsc conversion requires 64-bit division. */ +static int __read_mostly cpu_preemption_timer_multi; +static bool __read_mostly enable_preemption_timer = 1; +#ifdef CONFIG_X86_64 +module_param_named(preemption_timer, enable_preemption_timer, bool, S_IRUGO); +#endif + #define KVM_GUEST_CR0_MASK (X86_CR0_NW | X86_CR0_CD) #define KVM_VM_CR0_ALWAYS_ON_UNRESTRICTED_GUEST (X86_CR0_WP | X86_CR0_NE) #define KVM_VM_CR0_ALWAYS_ON \ @@ -398,6 +405,12 @@ struct nested_vmx { /* The host-usable pointer to the above */ struct page *current_vmcs12_page; struct vmcs12 *current_vmcs12; + /* + * Cache of the guest's VMCS, existing outside of guest memory. + * Loaded from guest memory during VMPTRLD. Flushed to guest + * memory during VMXOFF, VMCLEAR, VMPTRLD. + */ + struct vmcs12 *cached_vmcs12; struct vmcs *current_shadow_vmcs; /* * Indicates if the shadow vmcs must be updated with the @@ -421,7 +434,6 @@ struct nested_vmx { struct pi_desc *pi_desc; bool pi_pending; u16 posted_intr_nv; - u64 msr_ia32_feature_control; struct hrtimer preemption_timer; bool preemption_timer_expired; @@ -597,11 +609,22 @@ struct vcpu_vmx { #define PML_ENTITY_NUM 512 struct page *pml_pg; + /* apic deadline value in host tsc */ + u64 hv_deadline_tsc; + u64 current_tsc_ratio; bool guest_pkru_valid; u32 guest_pkru; u32 host_pkru; + + /* + * Only bits masked by msr_ia32_feature_control_valid_bits can be set in + * msr_ia32_feature_control. FEATURE_CONTROL_LOCKED is always included + * in msr_ia32_feature_control_valid_bits. + */ + u64 msr_ia32_feature_control; + u64 msr_ia32_feature_control_valid_bits; }; enum segment_cache_field { @@ -841,7 +864,7 @@ static inline short vmcs_field_to_offset(unsigned long field) static inline struct vmcs12 *get_vmcs12(struct kvm_vcpu *vcpu) { - return to_vmx(vcpu)->nested.current_vmcs12; + return to_vmx(vcpu)->nested.cached_vmcs12; } static struct page *nested_get_page(struct kvm_vcpu *vcpu, gpa_t addr) @@ -1056,6 +1079,58 @@ static inline bool cpu_has_vmx_virtual_intr_delivery(void) SECONDARY_EXEC_VIRTUAL_INTR_DELIVERY; } +/* + * Comment's format: document - errata name - stepping - processor name. + * Refer from + * https://www.virtualbox.org/svn/vbox/trunk/src/VBox/VMM/VMMR0/HMR0.cpp + */ +static u32 vmx_preemption_cpu_tfms[] = { +/* 323344.pdf - BA86 - D0 - Xeon 7500 Series */ +0x000206E6, +/* 323056.pdf - AAX65 - C2 - Xeon L3406 */ +/* 322814.pdf - AAT59 - C2 - i7-600, i5-500, i5-400 and i3-300 Mobile */ +/* 322911.pdf - AAU65 - C2 - i5-600, i3-500 Desktop and Pentium G6950 */ +0x00020652, +/* 322911.pdf - AAU65 - K0 - i5-600, i3-500 Desktop and Pentium G6950 */ +0x00020655, +/* 322373.pdf - AAO95 - B1 - Xeon 3400 Series */ +/* 322166.pdf - AAN92 - B1 - i7-800 and i5-700 Desktop */ +/* + * 320767.pdf - AAP86 - B1 - + * i7-900 Mobile Extreme, i7-800 and i7-700 Mobile + */ +0x000106E5, +/* 321333.pdf - AAM126 - C0 - Xeon 3500 */ +0x000106A0, +/* 321333.pdf - AAM126 - C1 - Xeon 3500 */ +0x000106A1, +/* 320836.pdf - AAJ124 - C0 - i7-900 Desktop Extreme and i7-900 Desktop */ +0x000106A4, + /* 321333.pdf - AAM126 - D0 - Xeon 3500 */ + /* 321324.pdf - AAK139 - D0 - Xeon 5500 */ + /* 320836.pdf - AAJ124 - D0 - i7-900 Extreme and i7-900 Desktop */ +0x000106A5, +}; + +static inline bool cpu_has_broken_vmx_preemption_timer(void) +{ + u32 eax = cpuid_eax(0x00000001), i; + + /* Clear the reserved bits */ + eax &= ~(0x3U << 14 | 0xfU << 28); + for (i = 0; i < ARRAY_SIZE(vmx_preemption_cpu_tfms); i++) + if (eax == vmx_preemption_cpu_tfms[i]) + return true; + + return false; +} + +static inline bool cpu_has_vmx_preemption_timer(void) +{ + return vmcs_config.pin_based_exec_ctrl & + PIN_BASED_VMX_PREEMPTION_TIMER; +} + static inline bool cpu_has_vmx_posted_intr(void) { return IS_ENABLED(CONFIG_X86_LOCAL_APIC) && @@ -1603,6 +1678,11 @@ static __always_inline void vmcs_set_bits(unsigned long field, u32 mask) __vmcs_writel(field, __vmcs_readl(field) | mask); } +static inline void vm_entry_controls_reset_shadow(struct vcpu_vmx *vmx) +{ + vmx->vm_entry_controls_shadow = vmcs_read32(VM_ENTRY_CONTROLS); +} + static inline void vm_entry_controls_init(struct vcpu_vmx *vmx, u32 val) { vmcs_write32(VM_ENTRY_CONTROLS, val); @@ -1631,6 +1711,11 @@ static inline void vm_entry_controls_clearbit(struct vcpu_vmx *vmx, u32 val) vm_entry_controls_set(vmx, vm_entry_controls_get(vmx) & ~val); } +static inline void vm_exit_controls_reset_shadow(struct vcpu_vmx *vmx) +{ + vmx->vm_exit_controls_shadow = vmcs_read32(VM_EXIT_CONTROLS); +} + static inline void vm_exit_controls_init(struct vcpu_vmx *vmx, u32 val) { vmcs_write32(VM_EXIT_CONTROLS, val); @@ -2121,22 +2206,14 @@ static void vmx_vcpu_load(struct kvm_vcpu *vcpu, int cpu) { struct vcpu_vmx *vmx = to_vmx(vcpu); u64 phys_addr = __pa(per_cpu(vmxarea, cpu)); + bool already_loaded = vmx->loaded_vmcs->cpu == cpu; if (!vmm_exclusive) kvm_cpu_vmxon(phys_addr); - else if (vmx->loaded_vmcs->cpu != cpu) + else if (!already_loaded) loaded_vmcs_clear(vmx->loaded_vmcs); - if (per_cpu(current_vmcs, cpu) != vmx->loaded_vmcs->vmcs) { - per_cpu(current_vmcs, cpu) = vmx->loaded_vmcs->vmcs; - vmcs_load(vmx->loaded_vmcs->vmcs); - } - - if (vmx->loaded_vmcs->cpu != cpu) { - struct desc_ptr *gdt = this_cpu_ptr(&host_gdt); - unsigned long sysenter_esp; - - kvm_make_request(KVM_REQ_TLB_FLUSH, vcpu); + if (!already_loaded) { local_irq_disable(); crash_disable_local_vmclear(cpu); @@ -2151,6 +2228,18 @@ static void vmx_vcpu_load(struct kvm_vcpu *vcpu, int cpu) &per_cpu(loaded_vmcss_on_cpu, cpu)); crash_enable_local_vmclear(cpu); local_irq_enable(); + } + + if (per_cpu(current_vmcs, cpu) != vmx->loaded_vmcs->vmcs) { + per_cpu(current_vmcs, cpu) = vmx->loaded_vmcs->vmcs; + vmcs_load(vmx->loaded_vmcs->vmcs); + } + + if (!already_loaded) { + struct desc_ptr *gdt = this_cpu_ptr(&host_gdt); + unsigned long sysenter_esp; + + kvm_make_request(KVM_REQ_TLB_FLUSH, vcpu); /* * Linux uses per-cpu TSS and GDT, so set these when switching @@ -2716,6 +2805,9 @@ static void nested_vmx_setup_ctls_msrs(struct vcpu_vmx *vmx) vmx->nested.nested_vmx_ept_caps = VMX_EPT_PAGE_WALK_4_BIT | VMX_EPTP_WB_BIT | VMX_EPT_2MB_PAGE_BIT | VMX_EPT_INVEPT_BIT; + if (cpu_has_vmx_ept_execute_only()) + vmx->nested.nested_vmx_ept_caps |= + VMX_EPT_EXECUTE_ONLY_BIT; vmx->nested.nested_vmx_ept_caps &= vmx_capability.ept; /* * For nested guests, we don't do anything specific @@ -2864,6 +2956,14 @@ static int vmx_get_vmx_msr(struct kvm_vcpu *vcpu, u32 msr_index, u64 *pdata) return 0; } +static inline bool vmx_feature_control_msr_valid(struct kvm_vcpu *vcpu, + uint64_t val) +{ + uint64_t valid_bits = to_vmx(vcpu)->msr_ia32_feature_control_valid_bits; + + return !(val & ~valid_bits); +} + /* * Reads an msr value (of 'msr_index') into 'pdata'. * Returns 0 on success, non-0 otherwise. @@ -2905,10 +3005,15 @@ static int vmx_get_msr(struct kvm_vcpu *vcpu, struct msr_data *msr_info) return 1; msr_info->data = vmcs_read64(GUEST_BNDCFGS); break; - case MSR_IA32_FEATURE_CONTROL: - if (!nested_vmx_allowed(vcpu)) + case MSR_IA32_MCG_EXT_CTL: + if (!msr_info->host_initiated && + !(to_vmx(vcpu)->msr_ia32_feature_control & + FEATURE_CONTROL_LMCE)) return 1; - msr_info->data = to_vmx(vcpu)->nested.msr_ia32_feature_control; + msr_info->data = vcpu->arch.mcg_ext_ctl; + break; + case MSR_IA32_FEATURE_CONTROL: + msr_info->data = to_vmx(vcpu)->msr_ia32_feature_control; break; case MSR_IA32_VMX_BASIC ... MSR_IA32_VMX_VMFUNC: if (!nested_vmx_allowed(vcpu)) @@ -2998,12 +3103,20 @@ static int vmx_set_msr(struct kvm_vcpu *vcpu, struct msr_data *msr_info) case MSR_IA32_TSC_ADJUST: ret = kvm_set_msr_common(vcpu, msr_info); break; + case MSR_IA32_MCG_EXT_CTL: + if ((!msr_info->host_initiated && + !(to_vmx(vcpu)->msr_ia32_feature_control & + FEATURE_CONTROL_LMCE)) || + (data & ~MCG_EXT_CTL_LMCE_EN)) + return 1; + vcpu->arch.mcg_ext_ctl = data; + break; case MSR_IA32_FEATURE_CONTROL: - if (!nested_vmx_allowed(vcpu) || - (to_vmx(vcpu)->nested.msr_ia32_feature_control & + if (!vmx_feature_control_msr_valid(vcpu, data) || + (to_vmx(vcpu)->msr_ia32_feature_control & FEATURE_CONTROL_LOCKED && !msr_info->host_initiated)) return 1; - vmx->nested.msr_ia32_feature_control = data; + vmx->msr_ia32_feature_control = data; if (msr_info->host_initiated && data == 0) vmx_leave_nested(vcpu); break; @@ -3297,25 +3410,27 @@ static __init int setup_vmcs_config(struct vmcs_config *vmcs_conf) vmx_capability.ept, vmx_capability.vpid); } - min = VM_EXIT_SAVE_DEBUG_CONTROLS; + min = VM_EXIT_SAVE_DEBUG_CONTROLS | VM_EXIT_ACK_INTR_ON_EXIT; #ifdef CONFIG_X86_64 min |= VM_EXIT_HOST_ADDR_SPACE_SIZE; #endif opt = VM_EXIT_SAVE_IA32_PAT | VM_EXIT_LOAD_IA32_PAT | - VM_EXIT_ACK_INTR_ON_EXIT | VM_EXIT_CLEAR_BNDCFGS; + VM_EXIT_CLEAR_BNDCFGS; if (adjust_vmx_controls(min, opt, MSR_IA32_VMX_EXIT_CTLS, &_vmexit_control) < 0) return -EIO; min = PIN_BASED_EXT_INTR_MASK | PIN_BASED_NMI_EXITING; - opt = PIN_BASED_VIRTUAL_NMIS | PIN_BASED_POSTED_INTR; + opt = PIN_BASED_VIRTUAL_NMIS | PIN_BASED_POSTED_INTR | + PIN_BASED_VMX_PREEMPTION_TIMER; if (adjust_vmx_controls(min, opt, MSR_IA32_VMX_PINBASED_CTLS, &_pin_based_exec_control) < 0) return -EIO; + if (cpu_has_broken_vmx_preemption_timer()) + _pin_based_exec_control &= ~PIN_BASED_VMX_PREEMPTION_TIMER; if (!(_cpu_based_2nd_exec_control & - SECONDARY_EXEC_VIRTUAL_INTR_DELIVERY) || - !(_vmexit_control & VM_EXIT_ACK_INTR_ON_EXIT)) + SECONDARY_EXEC_VIRTUAL_INTR_DELIVERY)) _pin_based_exec_control &= ~PIN_BASED_POSTED_INTR; min = VM_ENTRY_LOAD_DEBUG_CONTROLS; @@ -3364,7 +3479,7 @@ static __init int setup_vmcs_config(struct vmcs_config *vmcs_conf) /* * Some cpus support VM_ENTRY_(LOAD|SAVE)_IA32_PERF_GLOBAL_CTRL - * but due to arrata below it can't be used. Workaround is to use + * but due to errata below it can't be used. Workaround is to use * msr load mechanism to switch IA32_PERF_GLOBAL_CTRL. * * VM Exit May Incorrectly Clear IA32_PERF_GLOBAL_CTRL [34:32] @@ -4781,6 +4896,8 @@ static u32 vmx_pin_based_exec_ctrl(struct vcpu_vmx *vmx) if (!kvm_vcpu_apicv_active(&vmx->vcpu)) pin_based_exec_ctrl &= ~PIN_BASED_POSTED_INTR; + /* Enable the preemption timer dynamically */ + pin_based_exec_ctrl &= ~PIN_BASED_VMX_PREEMPTION_TIMER; return pin_based_exec_ctrl; } @@ -4896,6 +5013,7 @@ static int vmx_vcpu_setup(struct vcpu_vmx *vmx) /* Control */ vmcs_write32(PIN_BASED_VM_EXEC_CONTROL, vmx_pin_based_exec_ctrl(vmx)); + vmx->hv_deadline_tsc = -1; vmcs_write32(CPU_BASED_VM_EXEC_CONTROL, vmx_exec_control(vmx)); @@ -6016,12 +6134,14 @@ static int handle_ept_violation(struct kvm_vcpu *vcpu) gpa = vmcs_read64(GUEST_PHYSICAL_ADDRESS); trace_kvm_page_fault(gpa, exit_qualification); - /* It is a write fault? */ - error_code = exit_qualification & PFERR_WRITE_MASK; + /* it is a read fault? */ + error_code = (exit_qualification << 2) & PFERR_USER_MASK; + /* it is a write fault? */ + error_code |= exit_qualification & PFERR_WRITE_MASK; /* It is a fetch fault? */ error_code |= (exit_qualification << 2) & PFERR_FETCH_MASK; /* ept page table is present? */ - error_code |= (exit_qualification >> 3) & PFERR_PRESENT_MASK; + error_code |= (exit_qualification & 0x38) != 0; vcpu->arch.exit_qualification = exit_qualification; @@ -6355,9 +6475,6 @@ static __init int hardware_setup(void) for (msr = 0x800; msr <= 0x8ff; msr++) vmx_disable_intercept_msr_read_x2apic(msr); - /* According SDM, in x2apic mode, the whole id reg is used. But in - * KVM, it only use the highest eight bits. Need to intercept it */ - vmx_enable_intercept_msr_read_x2apic(0x802); /* TMCCT */ vmx_enable_intercept_msr_read_x2apic(0x839); /* TPR */ @@ -6368,10 +6485,12 @@ static __init int hardware_setup(void) vmx_disable_intercept_msr_write_x2apic(0x83f); if (enable_ept) { - kvm_mmu_set_mask_ptes(0ull, + kvm_mmu_set_mask_ptes(VMX_EPT_READABLE_MASK, (enable_ept_ad_bits) ? VMX_EPT_ACCESS_BIT : 0ull, (enable_ept_ad_bits) ? VMX_EPT_DIRTY_BIT : 0ull, - 0ull, VMX_EPT_EXECUTABLE_MASK); + 0ull, VMX_EPT_EXECUTABLE_MASK, + cpu_has_vmx_ept_execute_only() ? + 0ull : VMX_EPT_READABLE_MASK); ept_set_mmio_spte_mask(); kvm_enable_tdp(); } else @@ -6393,8 +6512,21 @@ static __init int hardware_setup(void) kvm_x86_ops->enable_log_dirty_pt_masked = NULL; } + if (cpu_has_vmx_preemption_timer() && enable_preemption_timer) { + u64 vmx_msr; + + rdmsrl(MSR_IA32_VMX_MISC, vmx_msr); + cpu_preemption_timer_multi = + vmx_msr & VMX_MISC_PREEMPTION_TIMER_RATE_MASK; + } else { + kvm_x86_ops->set_hv_timer = NULL; + kvm_x86_ops->cancel_hv_timer = NULL; + } + kvm_set_posted_intr_wakeup_handler(wakeup_handler); + kvm_mce_cap_supported |= MCG_LMCE_P; + return alloc_kvm_area(); out8: @@ -6862,16 +6994,22 @@ static int handle_vmon(struct kvm_vcpu *vcpu) return 1; } - if ((vmx->nested.msr_ia32_feature_control & VMXON_NEEDED_FEATURES) + if ((vmx->msr_ia32_feature_control & VMXON_NEEDED_FEATURES) != VMXON_NEEDED_FEATURES) { kvm_inject_gp(vcpu, 0); return 1; } + vmx->nested.cached_vmcs12 = kmalloc(VMCS12_SIZE, GFP_KERNEL); + if (!vmx->nested.cached_vmcs12) + return -ENOMEM; + if (enable_shadow_vmcs) { shadow_vmcs = alloc_vmcs(); - if (!shadow_vmcs) + if (!shadow_vmcs) { + kfree(vmx->nested.cached_vmcs12); return -ENOMEM; + } /* mark vmcs as shadow */ shadow_vmcs->revision_id |= (1u << 31); /* init shadow vmcs */ @@ -6942,6 +7080,11 @@ static inline void nested_release_vmcs12(struct vcpu_vmx *vmx) vmcs_write64(VMCS_LINK_POINTER, -1ull); } vmx->nested.posted_intr_nv = -1; + + /* Flush VMCS12 to guest memory */ + memcpy(vmx->nested.current_vmcs12, vmx->nested.cached_vmcs12, + VMCS12_SIZE); + kunmap(vmx->nested.current_vmcs12_page); nested_release_page(vmx->nested.current_vmcs12_page); vmx->nested.current_vmptr = -1ull; @@ -6962,6 +7105,7 @@ static void free_nested(struct vcpu_vmx *vmx) nested_release_vmcs12(vmx); if (enable_shadow_vmcs) free_vmcs(vmx->nested.current_shadow_vmcs); + kfree(vmx->nested.cached_vmcs12); /* Unpin physical memory we referred to in current vmcs02 */ if (vmx->nested.apic_access_page) { nested_release_page(vmx->nested.apic_access_page); @@ -7365,6 +7509,13 @@ static int handle_vmptrld(struct kvm_vcpu *vcpu) vmx->nested.current_vmptr = vmptr; vmx->nested.current_vmcs12 = new_vmcs12; vmx->nested.current_vmcs12_page = page; + /* + * Load VMCS12 from guest memory since it is not already + * cached. + */ + memcpy(vmx->nested.cached_vmcs12, + vmx->nested.current_vmcs12, VMCS12_SIZE); + if (enable_shadow_vmcs) { vmcs_set_bits(SECONDARY_VM_EXEC_CONTROL, SECONDARY_EXEC_SHADOW_VMCS); @@ -7560,6 +7711,12 @@ static int handle_pml_full(struct kvm_vcpu *vcpu) return 1; } +static int handle_preemption_timer(struct kvm_vcpu *vcpu) +{ + kvm_lapic_expired_hv_timer(vcpu); + return 1; +} + /* * The exit handlers return 1 if the exit was handled fully and guest execution * may resume. Otherwise they set the kvm_run parameter to indicate what needs @@ -7610,6 +7767,7 @@ static int (*const kvm_vmx_exit_handlers[])(struct kvm_vcpu *vcpu) = { [EXIT_REASON_XSAVES] = handle_xsaves, [EXIT_REASON_XRSTORS] = handle_xrstors, [EXIT_REASON_PML_FULL] = handle_pml_full, + [EXIT_REASON_PREEMPTION_TIMER] = handle_preemption_timer, }; static const int kvm_vmx_max_exit_handlers = @@ -7918,6 +8076,8 @@ static bool nested_vmx_exit_handled(struct kvm_vcpu *vcpu) * the XSS exit bitmap in vmcs12. */ return nested_cpu_has2(vmcs12, SECONDARY_EXEC_XSAVES); + case EXIT_REASON_PREEMPTION_TIMER: + return false; default: return true; } @@ -8303,7 +8463,7 @@ static void vmx_set_apic_access_page_addr(struct kvm_vcpu *vcpu, hpa_t hpa) * the next L2->L1 exit. */ if (!is_guest_mode(vcpu) || - !nested_cpu_has2(vmx->nested.current_vmcs12, + !nested_cpu_has2(get_vmcs12(&vmx->vcpu), SECONDARY_EXEC_VIRTUALIZE_APIC_ACCESSES)) vmcs_write64(APIC_ACCESS_ADDR, hpa); } @@ -8436,7 +8596,6 @@ static void vmx_handle_external_intr(struct kvm_vcpu *vcpu) "push %[sp]\n\t" #endif "pushf\n\t" - "orl $0x200, (%%" _ASM_SP ")\n\t" __ASM_SIZE(push) " $%c[cs]\n\t" "call *%[entry]\n\t" : @@ -8449,8 +8608,7 @@ static void vmx_handle_external_intr(struct kvm_vcpu *vcpu) [ss]"i"(__KERNEL_DS), [cs]"i"(__KERNEL_CS) ); - } else - local_irq_enable(); + } } static bool vmx_has_high_real_mode_segbase(void) @@ -8601,6 +8759,26 @@ static void atomic_switch_perf_msrs(struct vcpu_vmx *vmx) msrs[i].host); } +void vmx_arm_hv_timer(struct kvm_vcpu *vcpu) +{ + struct vcpu_vmx *vmx = to_vmx(vcpu); + u64 tscl; + u32 delta_tsc; + + if (vmx->hv_deadline_tsc == -1) + return; + + tscl = rdtsc(); + if (vmx->hv_deadline_tsc > tscl) + /* sure to be 32 bit only because checked on set_hv_timer */ + delta_tsc = (u32)((vmx->hv_deadline_tsc - tscl) >> + cpu_preemption_timer_multi); + else + delta_tsc = 0; + + vmcs_write32(VMX_PREEMPTION_TIMER_VALUE, delta_tsc); +} + static void __noclone vmx_vcpu_run(struct kvm_vcpu *vcpu) { struct vcpu_vmx *vmx = to_vmx(vcpu); @@ -8650,6 +8828,8 @@ static void __noclone vmx_vcpu_run(struct kvm_vcpu *vcpu) atomic_switch_perf_msrs(vmx); debugctlmsr = get_debugctlmsr(); + vmx_arm_hv_timer(vcpu); + vmx->__launched = vmx->loaded_vmcs->launched; asm( /* Store host registers */ @@ -8940,6 +9120,8 @@ static struct kvm_vcpu *vmx_create_vcpu(struct kvm *kvm, unsigned int id) vmx->nested.current_vmptr = -1ull; vmx->nested.current_vmcs12 = NULL; + vmx->msr_ia32_feature_control_valid_bits = FEATURE_CONTROL_LOCKED; + return &vmx->vcpu; free_vmcs: @@ -9080,6 +9262,13 @@ static void vmx_cpuid_update(struct kvm_vcpu *vcpu) if (cpu_has_secondary_exec_ctrls()) vmcs_set_secondary_exec_control(secondary_exec_ctl); + + if (nested_vmx_allowed(vcpu)) + to_vmx(vcpu)->msr_ia32_feature_control_valid_bits |= + FEATURE_CONTROL_VMXON_ENABLED_OUTSIDE_SMX; + else + to_vmx(vcpu)->msr_ia32_feature_control_valid_bits &= + ~FEATURE_CONTROL_VMXON_ENABLED_OUTSIDE_SMX; } static void vmx_set_supported_cpuid(u32 func, struct kvm_cpuid_entry2 *entry) @@ -9636,9 +9825,14 @@ static void prepare_vmcs02(struct kvm_vcpu *vcpu, struct vmcs12 *vmcs12) vmcs_write64(VMCS_LINK_POINTER, -1ull); exec_control = vmcs12->pin_based_vm_exec_control; - exec_control |= vmcs_config.pin_based_exec_ctrl; + + /* Preemption timer setting is only taken from vmcs01. */ exec_control &= ~PIN_BASED_VMX_PREEMPTION_TIMER; + exec_control |= vmcs_config.pin_based_exec_ctrl; + if (vmx->hv_deadline_tsc == -1) + exec_control &= ~PIN_BASED_VMX_PREEMPTION_TIMER; + /* Posted interrupts setting is only taken from vmcs12. */ if (nested_cpu_has_posted_intr(vmcs12)) { /* * Note that we use L0's vector here and in @@ -10556,8 +10750,8 @@ static void nested_vmx_vmexit(struct kvm_vcpu *vcpu, u32 exit_reason, vmcs12->vm_exit_intr_error_code, KVM_ISA_VMX); - vm_entry_controls_init(vmx, vmcs_read32(VM_ENTRY_CONTROLS)); - vm_exit_controls_init(vmx, vmcs_read32(VM_EXIT_CONTROLS)); + vm_entry_controls_reset_shadow(vmx); + vm_exit_controls_reset_shadow(vmx); vmx_segment_cache_clear(vmx); /* if no vmcs02 cache requested, remove the one we used */ @@ -10566,8 +10760,14 @@ static void nested_vmx_vmexit(struct kvm_vcpu *vcpu, u32 exit_reason, load_vmcs12_host_state(vcpu, vmcs12); - /* Update TSC_OFFSET if TSC was changed while L2 ran */ + /* Update any VMCS fields that might have changed while L2 ran */ vmcs_write64(TSC_OFFSET, vmx->nested.vmcs01_tsc_offset); + if (vmx->hv_deadline_tsc == -1) + vmcs_clear_bits(PIN_BASED_VM_EXEC_CONTROL, + PIN_BASED_VMX_PREEMPTION_TIMER); + else + vmcs_set_bits(PIN_BASED_VM_EXEC_CONTROL, + PIN_BASED_VMX_PREEMPTION_TIMER); /* This is needed for same reason as it was needed in prepare_vmcs02 */ vmx->host_rsp = 0; @@ -10647,6 +10847,64 @@ static int vmx_check_intercept(struct kvm_vcpu *vcpu, return X86EMUL_CONTINUE; } +#ifdef CONFIG_X86_64 +/* (a << shift) / divisor, return 1 if overflow otherwise 0 */ +static inline int u64_shl_div_u64(u64 a, unsigned int shift, + u64 divisor, u64 *result) +{ + u64 low = a << shift, high = a >> (64 - shift); + + /* To avoid the overflow on divq */ + if (high >= divisor) + return 1; + + /* Low hold the result, high hold rem which is discarded */ + asm("divq %2\n\t" : "=a" (low), "=d" (high) : + "rm" (divisor), "0" (low), "1" (high)); + *result = low; + + return 0; +} + +static int vmx_set_hv_timer(struct kvm_vcpu *vcpu, u64 guest_deadline_tsc) +{ + struct vcpu_vmx *vmx = to_vmx(vcpu); + u64 tscl = rdtsc(); + u64 guest_tscl = kvm_read_l1_tsc(vcpu, tscl); + u64 delta_tsc = max(guest_deadline_tsc, guest_tscl) - guest_tscl; + + /* Convert to host delta tsc if tsc scaling is enabled */ + if (vcpu->arch.tsc_scaling_ratio != kvm_default_tsc_scaling_ratio && + u64_shl_div_u64(delta_tsc, + kvm_tsc_scaling_ratio_frac_bits, + vcpu->arch.tsc_scaling_ratio, + &delta_tsc)) + return -ERANGE; + + /* + * If the delta tsc can't fit in the 32 bit after the multi shift, + * we can't use the preemption timer. + * It's possible that it fits on later vmentries, but checking + * on every vmentry is costly so we just use an hrtimer. + */ + if (delta_tsc >> (cpu_preemption_timer_multi + 32)) + return -ERANGE; + + vmx->hv_deadline_tsc = tscl + delta_tsc; + vmcs_set_bits(PIN_BASED_VM_EXEC_CONTROL, + PIN_BASED_VMX_PREEMPTION_TIMER); + return 0; +} + +static void vmx_cancel_hv_timer(struct kvm_vcpu *vcpu) +{ + struct vcpu_vmx *vmx = to_vmx(vcpu); + vmx->hv_deadline_tsc = -1; + vmcs_clear_bits(PIN_BASED_VM_EXEC_CONTROL, + PIN_BASED_VMX_PREEMPTION_TIMER); +} +#endif + static void vmx_sched_in(struct kvm_vcpu *vcpu, int cpu) { if (ple_gap) @@ -10691,7 +10949,7 @@ static void vmx_enable_log_dirty_pt_masked(struct kvm *kvm, * this case, return 1, otherwise, return 0. * */ -static int vmx_pre_block(struct kvm_vcpu *vcpu) +static int pi_pre_block(struct kvm_vcpu *vcpu) { unsigned long flags; unsigned int dest; @@ -10758,7 +11016,18 @@ static int vmx_pre_block(struct kvm_vcpu *vcpu) return 0; } -static void vmx_post_block(struct kvm_vcpu *vcpu) +static int vmx_pre_block(struct kvm_vcpu *vcpu) +{ + if (pi_pre_block(vcpu)) + return 1; + + if (kvm_lapic_hv_timer_in_use(vcpu)) + kvm_lapic_switch_to_sw_timer(vcpu); + + return 0; +} + +static void pi_post_block(struct kvm_vcpu *vcpu) { struct pi_desc *pi_desc = vcpu_to_pi_desc(vcpu); struct pi_desc old, new; @@ -10800,6 +11069,14 @@ static void vmx_post_block(struct kvm_vcpu *vcpu) } } +static void vmx_post_block(struct kvm_vcpu *vcpu) +{ + if (kvm_x86_ops->set_hv_timer) + kvm_lapic_switch_to_hv_timer(vcpu); + + pi_post_block(vcpu); +} + /* * vmx_update_pi_irte - set IRTE for Posted-Interrupts * @@ -10844,7 +11121,7 @@ static int vmx_update_pi_irte(struct kvm *kvm, unsigned int host_irq, * We will support full lowest-priority interrupt later. */ - kvm_set_msi_irq(e, &irq); + kvm_set_msi_irq(kvm, e, &irq); if (!kvm_intr_is_single_vcpu(kvm, &irq, &vcpu)) { /* * Make sure the IRTE is in remapped mode if @@ -10889,6 +11166,16 @@ out: return ret; } +static void vmx_setup_mce(struct kvm_vcpu *vcpu) +{ + if (vcpu->arch.mcg_cap & MCG_LMCE_P) + to_vmx(vcpu)->msr_ia32_feature_control_valid_bits |= + FEATURE_CONTROL_LMCE; + else + to_vmx(vcpu)->msr_ia32_feature_control_valid_bits &= + ~FEATURE_CONTROL_LMCE; +} + static struct kvm_x86_ops vmx_x86_ops = { .cpu_has_kvm_support = cpu_has_kvm_support, .disabled_by_bios = vmx_disabled_by_bios, @@ -11013,6 +11300,13 @@ static struct kvm_x86_ops vmx_x86_ops = { .pmu_ops = &intel_pmu_ops, .update_pi_irte = vmx_update_pi_irte, + +#ifdef CONFIG_X86_64 + .set_hv_timer = vmx_set_hv_timer, + .cancel_hv_timer = vmx_cancel_hv_timer, +#endif + + .setup_mce = vmx_setup_mce, }; static int __init vmx_init(void) diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c index 9c496c7..19f9f9e 100644 --- a/arch/x86/kvm/x86.c +++ b/arch/x86/kvm/x86.c @@ -71,7 +71,8 @@ #define MAX_IO_MSRS 256 #define KVM_MAX_MCE_BANKS 32 -#define KVM_MCE_CAP_SUPPORTED (MCG_CTL_P | MCG_SER_P) +u64 __read_mostly kvm_mce_cap_supported = MCG_CTL_P | MCG_SER_P; +EXPORT_SYMBOL_GPL(kvm_mce_cap_supported); #define emul_to_vcpu(ctxt) \ container_of(ctxt, struct kvm_vcpu, arch.emulate_ctxt) @@ -90,8 +91,12 @@ static u64 __read_mostly efer_reserved_bits = ~((u64)EFER_SCE); #define VM_STAT(x) offsetof(struct kvm, stat.x), KVM_STAT_VM #define VCPU_STAT(x) offsetof(struct kvm_vcpu, stat.x), KVM_STAT_VCPU +#define KVM_X2APIC_API_VALID_FLAGS (KVM_X2APIC_API_USE_32BIT_IDS | \ + KVM_X2APIC_API_DISABLE_BROADCAST_QUIRK) + static void update_cr8_intercept(struct kvm_vcpu *vcpu); static void process_nmi(struct kvm_vcpu *vcpu); +static void enter_smm(struct kvm_vcpu *vcpu); static void __kvm_set_rflags(struct kvm_vcpu *vcpu, unsigned long rflags); struct kvm_x86_ops *kvm_x86_ops __read_mostly; @@ -114,7 +119,8 @@ u8 __read_mostly kvm_tsc_scaling_ratio_frac_bits; EXPORT_SYMBOL_GPL(kvm_tsc_scaling_ratio_frac_bits); u64 __read_mostly kvm_max_tsc_scaling_ratio; EXPORT_SYMBOL_GPL(kvm_max_tsc_scaling_ratio); -static u64 __read_mostly kvm_default_tsc_scaling_ratio; +u64 __read_mostly kvm_default_tsc_scaling_ratio; +EXPORT_SYMBOL_GPL(kvm_default_tsc_scaling_ratio); /* tsc tolerance in parts per million - default to 1/2 of the NTP threshold */ static u32 __read_mostly tsc_tolerance_ppm = 250; @@ -538,7 +544,7 @@ int load_pdptrs(struct kvm_vcpu *vcpu, struct kvm_mmu *mmu, unsigned long cr3) goto out; } for (i = 0; i < ARRAY_SIZE(pdpte); ++i) { - if (is_present_gpte(pdpte[i]) && + if ((pdpte[i] & PT_PRESENT_MASK) && (pdpte[i] & vcpu->arch.mmu.guest_rsvd_check.rsvd_bits_mask[0][2])) { ret = 0; @@ -983,6 +989,7 @@ static u32 emulated_msrs[] = { MSR_IA32_MISC_ENABLE, MSR_IA32_MCG_STATUS, MSR_IA32_MCG_CTL, + MSR_IA32_MCG_EXT_CTL, MSR_IA32_SMBASE, }; @@ -1162,7 +1169,7 @@ static void kvm_write_wall_clock(struct kvm *kvm, gpa_t wall_clock) int version; int r; struct pvclock_wall_clock wc; - struct timespec boot; + struct timespec64 boot; if (!wall_clock) return; @@ -1185,13 +1192,13 @@ static void kvm_write_wall_clock(struct kvm *kvm, gpa_t wall_clock) * wall clock specified here. guest system time equals host * system time for us, thus we must fill in host boot time here. */ - getboottime(&boot); + getboottime64(&boot); if (kvm->arch.kvmclock_offset) { - struct timespec ts = ns_to_timespec(kvm->arch.kvmclock_offset); - boot = timespec_sub(boot, ts); + struct timespec64 ts = ns_to_timespec64(kvm->arch.kvmclock_offset); + boot = timespec64_sub(boot, ts); } - wc.sec = boot.tv_sec; + wc.sec = (u32)boot.tv_sec; /* overflow in 2106 guest time */ wc.nsec = boot.tv_nsec; wc.version = version; @@ -2616,6 +2623,9 @@ int kvm_vm_ioctl_check_extension(struct kvm *kvm, long ext) case KVM_CAP_TSC_CONTROL: r = kvm_has_tsc_control; break; + case KVM_CAP_X2APIC_API: + r = KVM_X2APIC_API_VALID_FLAGS; + break; default: r = 0; break; @@ -2678,11 +2688,9 @@ long kvm_arch_dev_ioctl(struct file *filp, break; } case KVM_X86_GET_MCE_CAP_SUPPORTED: { - u64 mce_cap; - - mce_cap = KVM_MCE_CAP_SUPPORTED; r = -EFAULT; - if (copy_to_user(argp, &mce_cap, sizeof mce_cap)) + if (copy_to_user(argp, &kvm_mce_cap_supported, + sizeof(kvm_mce_cap_supported))) goto out; r = 0; break; @@ -2734,6 +2742,11 @@ void kvm_arch_vcpu_load(struct kvm_vcpu *vcpu, int cpu) rdtsc() - vcpu->arch.last_host_tsc; if (tsc_delta < 0) mark_tsc_unstable("KVM discovered backwards TSC"); + + if (kvm_lapic_hv_timer_in_use(vcpu) && + kvm_x86_ops->set_hv_timer(vcpu, + kvm_get_lapic_tscdeadline_msr(vcpu))) + kvm_lapic_switch_to_sw_timer(vcpu); if (check_tsc_unstable()) { u64 offset = kvm_compute_tsc_offset(vcpu, vcpu->arch.last_guest_tsc); @@ -2767,15 +2780,17 @@ static int kvm_vcpu_ioctl_get_lapic(struct kvm_vcpu *vcpu, if (vcpu->arch.apicv_active) kvm_x86_ops->sync_pir_to_irr(vcpu); - memcpy(s->regs, vcpu->arch.apic->regs, sizeof *s); - - return 0; + return kvm_apic_get_state(vcpu, s); } static int kvm_vcpu_ioctl_set_lapic(struct kvm_vcpu *vcpu, struct kvm_lapic_state *s) { - kvm_apic_post_state_restore(vcpu, s); + int r; + + r = kvm_apic_set_state(vcpu, s); + if (r) + return r; update_cr8_intercept(vcpu); return 0; @@ -2860,7 +2875,7 @@ static int kvm_vcpu_ioctl_x86_setup_mce(struct kvm_vcpu *vcpu, r = -EINVAL; if (!bank_num || bank_num >= KVM_MAX_MCE_BANKS) goto out; - if (mcg_cap & ~(KVM_MCE_CAP_SUPPORTED | 0xff | 0xff0000)) + if (mcg_cap & ~(kvm_mce_cap_supported | 0xff | 0xff0000)) goto out; r = 0; vcpu->arch.mcg_cap = mcg_cap; @@ -2870,6 +2885,9 @@ static int kvm_vcpu_ioctl_x86_setup_mce(struct kvm_vcpu *vcpu, /* Init IA32_MCi_CTL to all 1s */ for (bank = 0; bank < bank_num; bank++) vcpu->arch.mce_banks[bank*4] = ~(u64)0; + + if (kvm_x86_ops->setup_mce) + kvm_x86_ops->setup_mce(vcpu); out: return r; } @@ -3768,7 +3786,7 @@ static int kvm_vm_ioctl_enable_cap(struct kvm *kvm, r = -EEXIST; if (irqchip_in_kernel(kvm)) goto split_irqchip_unlock; - if (atomic_read(&kvm->online_vcpus)) + if (kvm->created_vcpus) goto split_irqchip_unlock; r = kvm_setup_empty_irq_routing(kvm); if (r) @@ -3782,6 +3800,18 @@ split_irqchip_unlock: mutex_unlock(&kvm->lock); break; } + case KVM_CAP_X2APIC_API: + r = -EINVAL; + if (cap->args[0] & ~KVM_X2APIC_API_VALID_FLAGS) + break; + + if (cap->args[0] & KVM_X2APIC_API_USE_32BIT_IDS) + kvm->arch.x2apic_format = true; + if (cap->args[0] & KVM_X2APIC_API_DISABLE_BROADCAST_QUIRK) + kvm->arch.x2apic_broadcast_quirk_disabled = true; + + r = 0; + break; default: r = -EINVAL; break; @@ -3833,7 +3863,7 @@ long kvm_arch_vm_ioctl(struct file *filp, if (kvm->arch.vpic) goto create_irqchip_unlock; r = -EINVAL; - if (atomic_read(&kvm->online_vcpus)) + if (kvm->created_vcpus) goto create_irqchip_unlock; r = -ENOMEM; vpic = kvm_create_pic(kvm); @@ -3873,7 +3903,7 @@ long kvm_arch_vm_ioctl(struct file *filp, sizeof(struct kvm_pit_config))) goto out; create_pit: - mutex_lock(&kvm->slots_lock); + mutex_lock(&kvm->lock); r = -EEXIST; if (kvm->arch.vpit) goto create_pit_unlock; @@ -3882,7 +3912,7 @@ long kvm_arch_vm_ioctl(struct file *filp, if (kvm->arch.vpit) r = 0; create_pit_unlock: - mutex_unlock(&kvm->slots_lock); + mutex_unlock(&kvm->lock); break; case KVM_GET_IRQCHIP: { /* 0: PIC master, 1: PIC slave, 2: IOAPIC */ @@ -3989,7 +4019,7 @@ long kvm_arch_vm_ioctl(struct file *filp, case KVM_SET_BOOT_CPU_ID: r = 0; mutex_lock(&kvm->lock); - if (atomic_read(&kvm->online_vcpus) != 0) + if (kvm->created_vcpus) r = -EBUSY; else kvm->arch.bsp_vcpu_id = arg; @@ -5297,13 +5327,8 @@ static void kvm_smm_changed(struct kvm_vcpu *vcpu) /* This is a good place to trace that we are exiting SMM. */ trace_kvm_enter_smm(vcpu->vcpu_id, vcpu->arch.smbase, false); - if (unlikely(vcpu->arch.smi_pending)) { - kvm_make_request(KVM_REQ_SMI, vcpu); - vcpu->arch.smi_pending = 0; - } else { - /* Process a latched INIT, if any. */ - kvm_make_request(KVM_REQ_EVENT, vcpu); - } + /* Process a latched INIT or SMI, if any. */ + kvm_make_request(KVM_REQ_EVENT, vcpu); } kvm_mmu_reset_context(vcpu); @@ -5849,8 +5874,8 @@ int kvm_arch_init(void *opaque) kvm_x86_ops = ops; kvm_mmu_set_mask_ptes(PT_USER_MASK, PT_ACCESSED_MASK, - PT_DIRTY_MASK, PT64_NX_MASK, 0); - + PT_DIRTY_MASK, PT64_NX_MASK, 0, + PT_PRESENT_MASK); kvm_timer_init(); perf_register_guest_info_callbacks(&kvm_guest_cbs); @@ -6084,7 +6109,10 @@ static int inject_pending_event(struct kvm_vcpu *vcpu, bool req_int_win) } /* try to inject new event if pending */ - if (vcpu->arch.nmi_pending && kvm_x86_ops->nmi_allowed(vcpu)) { + if (vcpu->arch.smi_pending && !is_smm(vcpu)) { + vcpu->arch.smi_pending = false; + enter_smm(vcpu); + } else if (vcpu->arch.nmi_pending && kvm_x86_ops->nmi_allowed(vcpu)) { --vcpu->arch.nmi_pending; vcpu->arch.nmi_injected = true; kvm_x86_ops->set_nmi(vcpu); @@ -6107,6 +6135,7 @@ static int inject_pending_event(struct kvm_vcpu *vcpu, bool req_int_win) kvm_x86_ops->set_irq(vcpu); } } + return 0; } @@ -6130,7 +6159,7 @@ static void process_nmi(struct kvm_vcpu *vcpu) #define put_smstate(type, buf, offset, val) \ *(type *)((buf) + (offset) - 0x7e00) = val -static u32 process_smi_get_segment_flags(struct kvm_segment *seg) +static u32 enter_smm_get_segment_flags(struct kvm_segment *seg) { u32 flags = 0; flags |= seg->g << 23; @@ -6144,7 +6173,7 @@ static u32 process_smi_get_segment_flags(struct kvm_segment *seg) return flags; } -static void process_smi_save_seg_32(struct kvm_vcpu *vcpu, char *buf, int n) +static void enter_smm_save_seg_32(struct kvm_vcpu *vcpu, char *buf, int n) { struct kvm_segment seg; int offset; @@ -6159,11 +6188,11 @@ static void process_smi_save_seg_32(struct kvm_vcpu *vcpu, char *buf, int n) put_smstate(u32, buf, offset + 8, seg.base); put_smstate(u32, buf, offset + 4, seg.limit); - put_smstate(u32, buf, offset, process_smi_get_segment_flags(&seg)); + put_smstate(u32, buf, offset, enter_smm_get_segment_flags(&seg)); } #ifdef CONFIG_X86_64 -static void process_smi_save_seg_64(struct kvm_vcpu *vcpu, char *buf, int n) +static void enter_smm_save_seg_64(struct kvm_vcpu *vcpu, char *buf, int n) { struct kvm_segment seg; int offset; @@ -6172,7 +6201,7 @@ static void process_smi_save_seg_64(struct kvm_vcpu *vcpu, char *buf, int n) kvm_get_segment(vcpu, &seg, n); offset = 0x7e00 + n * 16; - flags = process_smi_get_segment_flags(&seg) >> 8; + flags = enter_smm_get_segment_flags(&seg) >> 8; put_smstate(u16, buf, offset, seg.selector); put_smstate(u16, buf, offset + 2, flags); put_smstate(u32, buf, offset + 4, seg.limit); @@ -6180,7 +6209,7 @@ static void process_smi_save_seg_64(struct kvm_vcpu *vcpu, char *buf, int n) } #endif -static void process_smi_save_state_32(struct kvm_vcpu *vcpu, char *buf) +static void enter_smm_save_state_32(struct kvm_vcpu *vcpu, char *buf) { struct desc_ptr dt; struct kvm_segment seg; @@ -6204,13 +6233,13 @@ static void process_smi_save_state_32(struct kvm_vcpu *vcpu, char *buf) put_smstate(u32, buf, 0x7fc4, seg.selector); put_smstate(u32, buf, 0x7f64, seg.base); put_smstate(u32, buf, 0x7f60, seg.limit); - put_smstate(u32, buf, 0x7f5c, process_smi_get_segment_flags(&seg)); + put_smstate(u32, buf, 0x7f5c, enter_smm_get_segment_flags(&seg)); kvm_get_segment(vcpu, &seg, VCPU_SREG_LDTR); put_smstate(u32, buf, 0x7fc0, seg.selector); put_smstate(u32, buf, 0x7f80, seg.base); put_smstate(u32, buf, 0x7f7c, seg.limit); - put_smstate(u32, buf, 0x7f78, process_smi_get_segment_flags(&seg)); + put_smstate(u32, buf, 0x7f78, enter_smm_get_segment_flags(&seg)); kvm_x86_ops->get_gdt(vcpu, &dt); put_smstate(u32, buf, 0x7f74, dt.address); @@ -6221,7 +6250,7 @@ static void process_smi_save_state_32(struct kvm_vcpu *vcpu, char *buf) put_smstate(u32, buf, 0x7f54, dt.size); for (i = 0; i < 6; i++) - process_smi_save_seg_32(vcpu, buf, i); + enter_smm_save_seg_32(vcpu, buf, i); put_smstate(u32, buf, 0x7f14, kvm_read_cr4(vcpu)); @@ -6230,7 +6259,7 @@ static void process_smi_save_state_32(struct kvm_vcpu *vcpu, char *buf) put_smstate(u32, buf, 0x7ef8, vcpu->arch.smbase); } -static void process_smi_save_state_64(struct kvm_vcpu *vcpu, char *buf) +static void enter_smm_save_state_64(struct kvm_vcpu *vcpu, char *buf) { #ifdef CONFIG_X86_64 struct desc_ptr dt; @@ -6262,7 +6291,7 @@ static void process_smi_save_state_64(struct kvm_vcpu *vcpu, char *buf) kvm_get_segment(vcpu, &seg, VCPU_SREG_TR); put_smstate(u16, buf, 0x7e90, seg.selector); - put_smstate(u16, buf, 0x7e92, process_smi_get_segment_flags(&seg) >> 8); + put_smstate(u16, buf, 0x7e92, enter_smm_get_segment_flags(&seg) >> 8); put_smstate(u32, buf, 0x7e94, seg.limit); put_smstate(u64, buf, 0x7e98, seg.base); @@ -6272,7 +6301,7 @@ static void process_smi_save_state_64(struct kvm_vcpu *vcpu, char *buf) kvm_get_segment(vcpu, &seg, VCPU_SREG_LDTR); put_smstate(u16, buf, 0x7e70, seg.selector); - put_smstate(u16, buf, 0x7e72, process_smi_get_segment_flags(&seg) >> 8); + put_smstate(u16, buf, 0x7e72, enter_smm_get_segment_flags(&seg) >> 8); put_smstate(u32, buf, 0x7e74, seg.limit); put_smstate(u64, buf, 0x7e78, seg.base); @@ -6281,31 +6310,26 @@ static void process_smi_save_state_64(struct kvm_vcpu *vcpu, char *buf) put_smstate(u64, buf, 0x7e68, dt.address); for (i = 0; i < 6; i++) - process_smi_save_seg_64(vcpu, buf, i); + enter_smm_save_seg_64(vcpu, buf, i); #else WARN_ON_ONCE(1); #endif } -static void process_smi(struct kvm_vcpu *vcpu) +static void enter_smm(struct kvm_vcpu *vcpu) { struct kvm_segment cs, ds; struct desc_ptr dt; char buf[512]; u32 cr0; - if (is_smm(vcpu)) { - vcpu->arch.smi_pending = true; - return; - } - trace_kvm_enter_smm(vcpu->vcpu_id, vcpu->arch.smbase, true); vcpu->arch.hflags |= HF_SMM_MASK; memset(buf, 0, 512); if (guest_cpuid_has_longmode(vcpu)) - process_smi_save_state_64(vcpu, buf); + enter_smm_save_state_64(vcpu, buf); else - process_smi_save_state_32(vcpu, buf); + enter_smm_save_state_32(vcpu, buf); kvm_vcpu_write_guest(vcpu, vcpu->arch.smbase + 0xfe00, buf, sizeof(buf)); @@ -6361,6 +6385,12 @@ static void process_smi(struct kvm_vcpu *vcpu) kvm_mmu_reset_context(vcpu); } +static void process_smi(struct kvm_vcpu *vcpu) +{ + vcpu->arch.smi_pending = true; + kvm_make_request(KVM_REQ_EVENT, vcpu); +} + void kvm_make_scan_ioapic_request(struct kvm *kvm) { kvm_make_all_cpus_request(kvm, KVM_REQ_SCAN_IOAPIC); @@ -6555,8 +6585,18 @@ static int vcpu_enter_guest(struct kvm_vcpu *vcpu) if (inject_pending_event(vcpu, req_int_win) != 0) req_immediate_exit = true; - /* enable NMI/IRQ window open exits if needed */ else { + /* Enable NMI/IRQ window open exits if needed. + * + * SMIs have two cases: 1) they can be nested, and + * then there is nothing to do here because RSM will + * cause a vmexit anyway; 2) or the SMI can be pending + * because inject_pending_event has completed the + * injection of an IRQ or NMI from the previous vmexit, + * and then we request an immediate exit to inject the SMI. + */ + if (vcpu->arch.smi_pending && !is_smm(vcpu)) + req_immediate_exit = true; if (vcpu->arch.nmi_pending) kvm_x86_ops->enable_nmi_window(vcpu); if (kvm_cpu_has_injectable_intr(vcpu) || req_int_win) @@ -6607,12 +6647,14 @@ static int vcpu_enter_guest(struct kvm_vcpu *vcpu) kvm_load_guest_xcr0(vcpu); - if (req_immediate_exit) + if (req_immediate_exit) { + kvm_make_request(KVM_REQ_EVENT, vcpu); smp_send_reschedule(vcpu->cpu); + } trace_kvm_entry(vcpu->vcpu_id); wait_lapic_expire(vcpu); - __kvm_guest_enter(); + guest_enter_irqoff(); if (unlikely(vcpu->arch.switch_db_regs)) { set_debugreg(0, 7); @@ -6663,16 +6705,9 @@ static int vcpu_enter_guest(struct kvm_vcpu *vcpu) ++vcpu->stat.exits; - /* - * We must have an instruction between local_irq_enable() and - * kvm_guest_exit(), so the timer interrupt isn't delayed by - * the interrupt shadow. The stat.exits increment will do nicely. - * But we need to prevent reordering, hence this barrier(): - */ - barrier(); - - kvm_guest_exit(); + guest_exit_irqoff(); + local_irq_enable(); preempt_enable(); vcpu->srcu_idx = srcu_read_lock(&vcpu->kvm->srcu); @@ -7409,6 +7444,7 @@ void kvm_vcpu_reset(struct kvm_vcpu *vcpu, bool init_event) { vcpu->arch.hflags = 0; + vcpu->arch.smi_pending = 0; atomic_set(&vcpu->arch.nmi_queued, 0); vcpu->arch.nmi_pending = 0; vcpu->arch.nmi_injected = false; @@ -7601,11 +7637,6 @@ bool kvm_vcpu_is_bsp(struct kvm_vcpu *vcpu) return (vcpu->arch.apic_base & MSR_IA32_APICBASE_BSP) != 0; } -bool kvm_vcpu_compatible(struct kvm_vcpu *vcpu) -{ - return irqchip_in_kernel(vcpu->kvm) == lapic_in_kernel(vcpu); -} - struct static_key kvm_no_apic_vcpu __read_mostly; EXPORT_SYMBOL_GPL(kvm_no_apic_vcpu); @@ -7872,7 +7903,7 @@ void kvm_arch_destroy_vm(struct kvm *kvm) kfree(kvm->arch.vpic); kfree(kvm->arch.vioapic); kvm_free_vcpus(kvm); - kfree(rcu_dereference_check(kvm->arch.apic_map, 1)); + kvfree(rcu_dereference_check(kvm->arch.apic_map, 1)); kvm_mmu_uninit_vm(kvm); } @@ -8380,7 +8411,7 @@ void kvm_arch_irq_bypass_del_producer(struct irq_bypass_consumer *cons, /* * When producer of consumer is unregistered, we change back to * remapped mode, so we can re-use the current implementation - * when the irq is masked/disabed or the consumer side (KVM + * when the irq is masked/disabled or the consumer side (KVM * int this case doesn't want to receive the interrupts. */ ret = kvm_x86_ops->update_pi_irte(irqfd->kvm, prod->irq, irqfd->gsi, 0); |