diff options
Diffstat (limited to 'arch/s390')
120 files changed, 3718 insertions, 2238 deletions
diff --git a/arch/s390/Kconfig b/arch/s390/Kconfig index bb63499..f2cf1f9 100644 --- a/arch/s390/Kconfig +++ b/arch/s390/Kconfig @@ -58,6 +58,9 @@ config NO_IOPORT_MAP config PCI_QUIRKS def_bool n +config ARCH_SUPPORTS_UPROBES + def_bool 64BIT + config S390 def_bool y select ARCH_HAS_ATOMIC64_DEC_IF_POSITIVE @@ -92,10 +95,12 @@ config S390 select ARCH_INLINE_WRITE_UNLOCK_IRQ select ARCH_INLINE_WRITE_UNLOCK_IRQRESTORE select ARCH_SAVE_PAGE_KEYS if HIBERNATION + select ARCH_SUPPORTS_ATOMIC_RMW select ARCH_USE_CMPXCHG_LOCKREF select ARCH_WANT_IPC_PARSE_VERSION select BUILDTIME_EXTABLE_SORT select CLONE_BACKWARDS2 + select DYNAMIC_FTRACE if FUNCTION_TRACER select GENERIC_CLOCKEVENTS select GENERIC_CPU_DEVICES if !SMP select GENERIC_FIND_FIRST_BIT @@ -112,11 +117,11 @@ config S390 select HAVE_CMPXCHG_LOCAL select HAVE_C_RECORDMCOUNT select HAVE_DEBUG_KMEMLEAK - select HAVE_DYNAMIC_FTRACE + select HAVE_DYNAMIC_FTRACE if 64BIT + select HAVE_DYNAMIC_FTRACE_WITH_REGS if 64BIT select HAVE_FTRACE_MCOUNT_RECORD - select HAVE_FUNCTION_GRAPH_TRACER - select HAVE_FUNCTION_TRACER - select HAVE_FUNCTION_TRACE_MCOUNT_TEST + select HAVE_FUNCTION_GRAPH_TRACER if 64BIT + select HAVE_FUNCTION_TRACER if 64BIT select HAVE_FUTEX_CMPXCHG if FUTEX select HAVE_KERNEL_BZIP2 select HAVE_KERNEL_GZIP @@ -137,7 +142,6 @@ config S390 select HAVE_SYSCALL_TRACEPOINTS select HAVE_UID16 if 32BIT select HAVE_VIRT_CPU_ACCOUNTING - select KTIME_SCALAR if 32BIT select MODULES_USE_ELF_RELA select NO_BOOTMEM select OLD_SIGACTION @@ -146,6 +150,7 @@ config S390 select TTY select VIRT_CPU_ACCOUNTING select VIRT_TO_BUS + select ARCH_HAS_SG_CHAIN config SCHED_OMIT_FRAME_POINTER def_bool y @@ -496,8 +501,8 @@ config QDIO menuconfig PCI bool "PCI support" - default n depends on 64BIT + select HAVE_DMA_ATTRS select PCI_MSI help Enable PCI support. @@ -544,9 +549,6 @@ config HAS_DMA config NEED_SG_DMA_LENGTH def_bool PCI -config HAVE_DMA_ATTRS - def_bool PCI - config NEED_DMA_MAP_STATE def_bool PCI diff --git a/arch/s390/Makefile b/arch/s390/Makefile index 874e6d6..878e679 100644 --- a/arch/s390/Makefile +++ b/arch/s390/Makefile @@ -35,13 +35,16 @@ endif export LD_BFD -cflags-$(CONFIG_MARCH_G5) += -march=g5 -cflags-$(CONFIG_MARCH_Z900) += -march=z900 -cflags-$(CONFIG_MARCH_Z990) += -march=z990 -cflags-$(CONFIG_MARCH_Z9_109) += -march=z9-109 -cflags-$(CONFIG_MARCH_Z10) += -march=z10 -cflags-$(CONFIG_MARCH_Z196) += -march=z196 -cflags-$(CONFIG_MARCH_ZEC12) += -march=zEC12 +mflags-$(CONFIG_MARCH_G5) := -march=g5 +mflags-$(CONFIG_MARCH_Z900) := -march=z900 +mflags-$(CONFIG_MARCH_Z990) := -march=z990 +mflags-$(CONFIG_MARCH_Z9_109) := -march=z9-109 +mflags-$(CONFIG_MARCH_Z10) := -march=z10 +mflags-$(CONFIG_MARCH_Z196) := -march=z196 +mflags-$(CONFIG_MARCH_ZEC12) := -march=zEC12 + +aflags-y += $(mflags-y) +cflags-y += $(mflags-y) cflags-$(CONFIG_MARCH_G5_TUNE) += -mtune=g5 cflags-$(CONFIG_MARCH_Z900_TUNE) += -mtune=z900 diff --git a/arch/s390/configs/default_defconfig b/arch/s390/configs/default_defconfig index fd09a10..9d94fdd 100644 --- a/arch/s390/configs/default_defconfig +++ b/arch/s390/configs/default_defconfig @@ -63,6 +63,7 @@ CONFIG_CRASH_DUMP=y # CONFIG_CORE_DUMP_DEFAULT_ELF_HEADERS is not set CONFIG_BINFMT_MISC=m CONFIG_HIBERNATION=y +CONFIG_NET=y CONFIG_PACKET=y CONFIG_PACKET_DIAG=m CONFIG_UNIX=y @@ -366,7 +367,6 @@ CONFIG_VIRTIO_BLK=y CONFIG_ENCLOSURE_SERVICES=m CONFIG_RAID_ATTRS=m CONFIG_SCSI=y -CONFIG_SCSI_TGT=m CONFIG_BLK_DEV_SD=y CONFIG_CHR_DEV_ST=m CONFIG_CHR_DEV_OSST=m @@ -380,7 +380,6 @@ CONFIG_SCSI_LOGGING=y CONFIG_SCSI_SPI_ATTRS=m CONFIG_SCSI_SAS_LIBSAS=m CONFIG_SCSI_SRP_ATTRS=m -CONFIG_SCSI_SRP_TGT_ATTRS=y CONFIG_ISCSI_TCP=m CONFIG_LIBFCOE=m CONFIG_SCSI_DEBUG=m diff --git a/arch/s390/configs/gcov_defconfig b/arch/s390/configs/gcov_defconfig index b061180..90f514b 100644 --- a/arch/s390/configs/gcov_defconfig +++ b/arch/s390/configs/gcov_defconfig @@ -61,6 +61,7 @@ CONFIG_CRASH_DUMP=y # CONFIG_CORE_DUMP_DEFAULT_ELF_HEADERS is not set CONFIG_BINFMT_MISC=m CONFIG_HIBERNATION=y +CONFIG_NET=y CONFIG_PACKET=y CONFIG_PACKET_DIAG=m CONFIG_UNIX=y @@ -363,7 +364,6 @@ CONFIG_VIRTIO_BLK=y CONFIG_ENCLOSURE_SERVICES=m CONFIG_RAID_ATTRS=m CONFIG_SCSI=y -CONFIG_SCSI_TGT=m CONFIG_BLK_DEV_SD=y CONFIG_CHR_DEV_ST=m CONFIG_CHR_DEV_OSST=m @@ -377,7 +377,6 @@ CONFIG_SCSI_LOGGING=y CONFIG_SCSI_SPI_ATTRS=m CONFIG_SCSI_SAS_LIBSAS=m CONFIG_SCSI_SRP_ATTRS=m -CONFIG_SCSI_SRP_TGT_ATTRS=y CONFIG_ISCSI_TCP=m CONFIG_LIBFCOE=m CONFIG_SCSI_DEBUG=m diff --git a/arch/s390/configs/performance_defconfig b/arch/s390/configs/performance_defconfig index d279baa..13559d3 100644 --- a/arch/s390/configs/performance_defconfig +++ b/arch/s390/configs/performance_defconfig @@ -59,6 +59,7 @@ CONFIG_CRASH_DUMP=y # CONFIG_CORE_DUMP_DEFAULT_ELF_HEADERS is not set CONFIG_BINFMT_MISC=m CONFIG_HIBERNATION=y +CONFIG_NET=y CONFIG_PACKET=y CONFIG_PACKET_DIAG=m CONFIG_UNIX=y @@ -361,7 +362,6 @@ CONFIG_VIRTIO_BLK=y CONFIG_ENCLOSURE_SERVICES=m CONFIG_RAID_ATTRS=m CONFIG_SCSI=y -CONFIG_SCSI_TGT=m CONFIG_BLK_DEV_SD=y CONFIG_CHR_DEV_ST=m CONFIG_CHR_DEV_OSST=m @@ -375,7 +375,6 @@ CONFIG_SCSI_LOGGING=y CONFIG_SCSI_SPI_ATTRS=m CONFIG_SCSI_SAS_LIBSAS=m CONFIG_SCSI_SRP_ATTRS=m -CONFIG_SCSI_SRP_TGT_ATTRS=y CONFIG_ISCSI_TCP=m CONFIG_LIBFCOE=m CONFIG_SCSI_DEBUG=m diff --git a/arch/s390/configs/zfcpdump_defconfig b/arch/s390/configs/zfcpdump_defconfig index 948e0e0..e376789 100644 --- a/arch/s390/configs/zfcpdump_defconfig +++ b/arch/s390/configs/zfcpdump_defconfig @@ -23,6 +23,7 @@ CONFIG_CRASH_DUMP=y # CONFIG_CORE_DUMP_DEFAULT_ELF_HEADERS is not set # CONFIG_SECCOMP is not set # CONFIG_IUCV is not set +CONFIG_NET=y CONFIG_ATM=y CONFIG_ATM_LANE=y CONFIG_UEVENT_HELPER_PATH="/sbin/hotplug" diff --git a/arch/s390/defconfig b/arch/s390/defconfig index 2e56498..fab35a8 100644 --- a/arch/s390/defconfig +++ b/arch/s390/defconfig @@ -50,6 +50,7 @@ CONFIG_CMA=y CONFIG_CRASH_DUMP=y CONFIG_BINFMT_MISC=m CONFIG_HIBERNATION=y +CONFIG_NET=y CONFIG_PACKET=y CONFIG_UNIX=y CONFIG_NET_KEY=y diff --git a/arch/s390/include/asm/Kbuild b/arch/s390/include/asm/Kbuild index 57892a8..773f866 100644 --- a/arch/s390/include/asm/Kbuild +++ b/arch/s390/include/asm/Kbuild @@ -2,6 +2,8 @@ generic-y += clkdev.h generic-y += hash.h +generic-y += irq_work.h generic-y += mcs_spinlock.h generic-y += preempt.h +generic-y += scatterlist.h generic-y += trace_clock.h diff --git a/arch/s390/include/asm/barrier.h b/arch/s390/include/asm/barrier.h index 19ff956..b5dce65 100644 --- a/arch/s390/include/asm/barrier.h +++ b/arch/s390/include/asm/barrier.h @@ -15,11 +15,13 @@ #ifdef CONFIG_HAVE_MARCH_Z196_FEATURES /* Fast-BCR without checkpoint synchronization */ -#define mb() do { asm volatile("bcr 14,0" : : : "memory"); } while (0) +#define __ASM_BARRIER "bcr 14,0\n" #else -#define mb() do { asm volatile("bcr 15,0" : : : "memory"); } while (0) +#define __ASM_BARRIER "bcr 15,0\n" #endif +#define mb() do { asm volatile(__ASM_BARRIER : : : "memory"); } while (0) + #define rmb() mb() #define wmb() mb() #define read_barrier_depends() do { } while(0) diff --git a/arch/s390/include/asm/cputime.h b/arch/s390/include/asm/cputime.h index f65bd36..f8c1969 100644 --- a/arch/s390/include/asm/cputime.h +++ b/arch/s390/include/asm/cputime.h @@ -8,8 +8,6 @@ #define _S390_CPUTIME_H #include <linux/types.h> -#include <linux/percpu.h> -#include <linux/spinlock.h> #include <asm/div64.h> @@ -18,6 +16,8 @@ typedef unsigned long long __nocast cputime_t; typedef unsigned long long __nocast cputime64_t; +#define cmpxchg_cputime(ptr, old, new) cmpxchg64(ptr, old, new) + static inline unsigned long __div(unsigned long long n, unsigned long base) { #ifndef CONFIG_64BIT @@ -165,28 +165,8 @@ static inline clock_t cputime64_to_clock_t(cputime64_t cputime) return clock; } -struct s390_idle_data { - int nohz_delay; - unsigned int sequence; - unsigned long long idle_count; - unsigned long long idle_time; - unsigned long long clock_idle_enter; - unsigned long long clock_idle_exit; - unsigned long long timer_idle_enter; - unsigned long long timer_idle_exit; -}; - -DECLARE_PER_CPU(struct s390_idle_data, s390_idle); - -cputime64_t s390_get_idle_time(int cpu); - -#define arch_idle_time(cpu) s390_get_idle_time(cpu) - -static inline int s390_nohz_delay(int cpu) -{ - return __get_cpu_var(s390_idle).nohz_delay != 0; -} +cputime64_t arch_cpu_idle_time(int cpu); -#define arch_needs_cpu(cpu) s390_nohz_delay(cpu) +#define arch_idle_time(cpu) arch_cpu_idle_time(cpu) #endif /* _S390_CPUTIME_H */ diff --git a/arch/s390/include/asm/dis.h b/arch/s390/include/asm/dis.h index 04a83f5..60323c2 100644 --- a/arch/s390/include/asm/dis.h +++ b/arch/s390/include/asm/dis.h @@ -13,12 +13,13 @@ #define OPERAND_FPR 0x2 /* Operand printed as %fx */ #define OPERAND_AR 0x4 /* Operand printed as %ax */ #define OPERAND_CR 0x8 /* Operand printed as %cx */ -#define OPERAND_DISP 0x10 /* Operand printed as displacement */ -#define OPERAND_BASE 0x20 /* Operand printed as base register */ -#define OPERAND_INDEX 0x40 /* Operand printed as index register */ -#define OPERAND_PCREL 0x80 /* Operand printed as pc-relative symbol */ -#define OPERAND_SIGNED 0x100 /* Operand printed as signed value */ -#define OPERAND_LENGTH 0x200 /* Operand printed as length (+1) */ +#define OPERAND_VR 0x10 /* Operand printed as %vx */ +#define OPERAND_DISP 0x20 /* Operand printed as displacement */ +#define OPERAND_BASE 0x40 /* Operand printed as base register */ +#define OPERAND_INDEX 0x80 /* Operand printed as index register */ +#define OPERAND_PCREL 0x100 /* Operand printed as pc-relative symbol */ +#define OPERAND_SIGNED 0x200 /* Operand printed as signed value */ +#define OPERAND_LENGTH 0x400 /* Operand printed as length (+1) */ struct s390_operand { diff --git a/arch/s390/include/asm/dma-mapping.h b/arch/s390/include/asm/dma-mapping.h index 3fbc67d..709955d 100644 --- a/arch/s390/include/asm/dma-mapping.h +++ b/arch/s390/include/asm/dma-mapping.h @@ -56,24 +56,35 @@ static inline int dma_mapping_error(struct device *dev, dma_addr_t dma_addr) return dma_addr == DMA_ERROR_CODE; } -static inline void *dma_alloc_coherent(struct device *dev, size_t size, - dma_addr_t *dma_handle, gfp_t flag) +#define dma_alloc_coherent(d, s, h, f) dma_alloc_attrs(d, s, h, f, NULL) + +static inline void *dma_alloc_attrs(struct device *dev, size_t size, + dma_addr_t *dma_handle, gfp_t flags, + struct dma_attrs *attrs) { struct dma_map_ops *ops = get_dma_ops(dev); - void *ret; + void *cpu_addr; + + BUG_ON(!ops); - ret = ops->alloc(dev, size, dma_handle, flag, NULL); - debug_dma_alloc_coherent(dev, size, *dma_handle, ret); - return ret; + cpu_addr = ops->alloc(dev, size, dma_handle, flags, attrs); + debug_dma_alloc_coherent(dev, size, *dma_handle, cpu_addr); + + return cpu_addr; } -static inline void dma_free_coherent(struct device *dev, size_t size, - void *cpu_addr, dma_addr_t dma_handle) +#define dma_free_coherent(d, s, c, h) dma_free_attrs(d, s, c, h, NULL) + +static inline void dma_free_attrs(struct device *dev, size_t size, + void *cpu_addr, dma_addr_t dma_handle, + struct dma_attrs *attrs) { - struct dma_map_ops *dma_ops = get_dma_ops(dev); + struct dma_map_ops *ops = get_dma_ops(dev); + + BUG_ON(!ops); debug_dma_free_coherent(dev, size, cpu_addr, dma_handle); - dma_ops->free(dev, size, cpu_addr, dma_handle, NULL); + ops->free(dev, size, cpu_addr, dma_handle, attrs); } #endif /* _ASM_S390_DMA_MAPPING_H */ diff --git a/arch/s390/include/asm/elf.h b/arch/s390/include/asm/elf.h index 78f4f87..f6e43d3 100644 --- a/arch/s390/include/asm/elf.h +++ b/arch/s390/include/asm/elf.h @@ -102,6 +102,7 @@ #define HWCAP_S390_ETF3EH 256 #define HWCAP_S390_HIGH_GPRS 512 #define HWCAP_S390_TE 1024 +#define HWCAP_S390_VXRS 2048 /* * These are used to set parameters in the core dumps. @@ -225,6 +226,6 @@ int arch_setup_additional_pages(struct linux_binprm *, int); extern unsigned long arch_randomize_brk(struct mm_struct *mm); #define arch_randomize_brk arch_randomize_brk -void *fill_cpu_elf_notes(void *ptr, struct save_area *sa); +void *fill_cpu_elf_notes(void *ptr, struct save_area *sa, __vector128 *vxrs); #endif diff --git a/arch/s390/include/asm/ftrace.h b/arch/s390/include/asm/ftrace.h index bf246da..3aef8af 100644 --- a/arch/s390/include/asm/ftrace.h +++ b/arch/s390/include/asm/ftrace.h @@ -4,6 +4,7 @@ #ifndef __ASSEMBLY__ extern void _mcount(void); +extern char ftrace_graph_caller_end; struct dyn_arch_ftrace { }; @@ -17,10 +18,8 @@ static inline unsigned long ftrace_call_adjust(unsigned long addr) #endif /* __ASSEMBLY__ */ -#ifdef CONFIG_64BIT -#define MCOUNT_INSN_SIZE 12 -#else -#define MCOUNT_INSN_SIZE 22 -#endif +#define MCOUNT_INSN_SIZE 18 + +#define ARCH_SUPPORTS_FTRACE_OPS 1 #endif /* _ASM_S390_FTRACE_H */ diff --git a/arch/s390/include/asm/idle.h b/arch/s390/include/asm/idle.h new file mode 100644 index 0000000..6af037f --- /dev/null +++ b/arch/s390/include/asm/idle.h @@ -0,0 +1,26 @@ +/* + * Copyright IBM Corp. 2014 + * + * Author: Martin Schwidefsky <schwidefsky@de.ibm.com> + */ + +#ifndef _S390_IDLE_H +#define _S390_IDLE_H + +#include <linux/types.h> +#include <linux/device.h> + +struct s390_idle_data { + unsigned int sequence; + unsigned long long idle_count; + unsigned long long idle_time; + unsigned long long clock_idle_enter; + unsigned long long clock_idle_exit; + unsigned long long timer_idle_enter; + unsigned long long timer_idle_exit; +}; + +extern struct device_attribute dev_attr_idle_count; +extern struct device_attribute dev_attr_idle_time_us; + +#endif /* _S390_IDLE_H */ diff --git a/arch/s390/include/asm/ipl.h b/arch/s390/include/asm/ipl.h index 2fcccc0..ece606c 100644 --- a/arch/s390/include/asm/ipl.h +++ b/arch/s390/include/asm/ipl.h @@ -17,12 +17,12 @@ #define IPL_PARM_BLK_FCP_LEN (sizeof(struct ipl_list_hdr) + \ sizeof(struct ipl_block_fcp)) -#define IPL_PARM_BLK0_FCP_LEN (sizeof(struct ipl_block_fcp) + 8) +#define IPL_PARM_BLK0_FCP_LEN (sizeof(struct ipl_block_fcp) + 16) #define IPL_PARM_BLK_CCW_LEN (sizeof(struct ipl_list_hdr) + \ sizeof(struct ipl_block_ccw)) -#define IPL_PARM_BLK0_CCW_LEN (sizeof(struct ipl_block_ccw) + 8) +#define IPL_PARM_BLK0_CCW_LEN (sizeof(struct ipl_block_ccw) + 16) #define IPL_MAX_SUPPORTED_VERSION (0) @@ -38,10 +38,11 @@ struct ipl_list_hdr { u8 pbt; u8 flags; u16 reserved2; + u8 loadparm[8]; } __attribute__((packed)); struct ipl_block_fcp { - u8 reserved1[313-1]; + u8 reserved1[305-1]; u8 opt; u8 reserved2[3]; u16 reserved3; @@ -62,7 +63,6 @@ struct ipl_block_fcp { offsetof(struct ipl_block_fcp, scp_data))) struct ipl_block_ccw { - u8 load_parm[8]; u8 reserved1[84]; u8 reserved2[2]; u16 devno; @@ -89,12 +89,12 @@ extern u32 ipl_flags; extern u32 dump_prefix_page; struct dump_save_areas { - struct save_area **areas; + struct save_area_ext **areas; int count; }; extern struct dump_save_areas dump_save_areas; -struct save_area *dump_save_area_create(int cpu); +struct save_area_ext *dump_save_area_create(int cpu); extern void do_reipl(void); extern void do_halt(void); diff --git a/arch/s390/include/asm/irq.h b/arch/s390/include/asm/irq.h index c4dd400..b0d5f0a 100644 --- a/arch/s390/include/asm/irq.h +++ b/arch/s390/include/asm/irq.h @@ -51,6 +51,7 @@ enum interruption_class { IRQEXT_CMS, IRQEXT_CMC, IRQEXT_CMR, + IRQEXT_FTP, IRQIO_CIO, IRQIO_QAI, IRQIO_DAS, @@ -81,7 +82,7 @@ DECLARE_PER_CPU_SHARED_ALIGNED(struct irq_stat, irq_stat); static __always_inline void inc_irq_stat(enum interruption_class irq) { - __get_cpu_var(irq_stat).irqs[irq]++; + __this_cpu_inc(irq_stat.irqs[irq]); } struct ext_code { diff --git a/arch/s390/include/asm/kprobes.h b/arch/s390/include/asm/kprobes.h index 4176dfe..9862917 100644 --- a/arch/s390/include/asm/kprobes.h +++ b/arch/s390/include/asm/kprobes.h @@ -84,6 +84,10 @@ int kprobe_fault_handler(struct pt_regs *regs, int trapnr); int kprobe_exceptions_notify(struct notifier_block *self, unsigned long val, void *data); +int probe_is_prohibited_opcode(u16 *insn); +int probe_get_fixup_type(u16 *insn); +int probe_is_insn_relative_long(u16 *insn); + #define flush_insn_slot(p) do { } while (0) #endif /* _ASM_S390_KPROBES_H */ diff --git a/arch/s390/include/asm/kvm_host.h b/arch/s390/include/asm/kvm_host.h index 4181d7b..2175f911 100644 --- a/arch/s390/include/asm/kvm_host.h +++ b/arch/s390/include/asm/kvm_host.h @@ -13,8 +13,11 @@ #ifndef ASM_KVM_HOST_H #define ASM_KVM_HOST_H + +#include <linux/types.h> #include <linux/hrtimer.h> #include <linux/interrupt.h> +#include <linux/kvm_types.h> #include <linux/kvm_host.h> #include <linux/kvm.h> #include <asm/debug.h> @@ -154,7 +157,9 @@ struct kvm_s390_sie_block { __u8 armid; /* 0x00e3 */ __u8 reservede4[4]; /* 0x00e4 */ __u64 tecmc; /* 0x00e8 */ - __u8 reservedf0[16]; /* 0x00f0 */ + __u8 reservedf0[12]; /* 0x00f0 */ +#define CRYCB_FORMAT1 0x00000001 + __u32 crycbd; /* 0x00fc */ __u64 gcr[16]; /* 0x0100 */ __u64 gbea; /* 0x0180 */ __u8 reserved188[24]; /* 0x0188 */ @@ -187,6 +192,7 @@ struct kvm_vcpu_stat { u32 exit_stop_request; u32 exit_validity; u32 exit_instruction; + u32 halt_wakeup; u32 instruction_lctl; u32 instruction_lctlg; u32 instruction_stctl; @@ -305,7 +311,6 @@ struct kvm_s390_local_interrupt { struct list_head list; atomic_t active; struct kvm_s390_float_interrupt *float_int; - int timer_due; /* event indicator for waitqueue below */ wait_queue_head_t *wq; atomic_t *cpuflags; unsigned int action_bits; @@ -367,7 +372,6 @@ struct kvm_vcpu_arch { s390_fp_regs guest_fpregs; struct kvm_s390_local_interrupt local_int; struct hrtimer ckc_timer; - struct tasklet_struct tasklet; struct kvm_s390_pgm_info pgm; union { struct cpuid cpu_id; @@ -409,6 +413,15 @@ struct s390_io_adapter { #define MAX_S390_IO_ADAPTERS ((MAX_ISC + 1) * 8) #define MAX_S390_ADAPTER_MAPS 256 +struct kvm_s390_crypto { + struct kvm_s390_crypto_cb *crycb; + __u32 crycbd; +}; + +struct kvm_s390_crypto_cb { + __u8 reserved00[128]; /* 0x0000 */ +}; + struct kvm_arch{ struct sca_block *sca; debug_info_t *dbf; @@ -418,9 +431,11 @@ struct kvm_arch{ int css_support; int use_irqchip; int use_cmma; + int user_cpu_state_ctrl; struct s390_io_adapter *adapters[MAX_S390_IO_ADAPTERS]; wait_queue_head_t ipte_wq; spinlock_t start_stop_lock; + struct kvm_s390_crypto crypto; }; #define KVM_HVA_ERR_BAD (-1UL) @@ -432,8 +447,6 @@ static inline bool kvm_is_error_hva(unsigned long addr) } #define ASYNC_PF_PER_VCPU 64 -struct kvm_vcpu; -struct kvm_async_pf; struct kvm_arch_async_pf { unsigned long pfault_token; }; @@ -451,4 +464,18 @@ void kvm_arch_async_page_present(struct kvm_vcpu *vcpu, extern int sie64a(struct kvm_s390_sie_block *, u64 *); extern char sie_exit; + +static inline void kvm_arch_hardware_disable(void) {} +static inline void kvm_arch_check_processor_compat(void *rtn) {} +static inline void kvm_arch_exit(void) {} +static inline void kvm_arch_sync_events(struct kvm *kvm) {} +static inline void kvm_arch_vcpu_uninit(struct kvm_vcpu *vcpu) {} +static inline void kvm_arch_sched_in(struct kvm_vcpu *vcpu, int cpu) {} +static inline void kvm_arch_free_memslot(struct kvm *kvm, + struct kvm_memory_slot *free, struct kvm_memory_slot *dont) {} +static inline void kvm_arch_memslots_updated(struct kvm *kvm) {} +static inline void kvm_arch_flush_shadow_all(struct kvm *kvm) {} +static inline void kvm_arch_flush_shadow_memslot(struct kvm *kvm, + struct kvm_memory_slot *slot) {} + #endif diff --git a/arch/s390/include/asm/lowcore.h b/arch/s390/include/asm/lowcore.h index 4349197..6cc51fe 100644 --- a/arch/s390/include/asm/lowcore.h +++ b/arch/s390/include/asm/lowcore.h @@ -11,6 +11,7 @@ #include <linux/types.h> #include <asm/ptrace.h> #include <asm/cpu.h> +#include <asm/types.h> #ifdef CONFIG_32BIT @@ -31,6 +32,11 @@ struct save_area { u32 ctrl_regs[16]; } __packed; +struct save_area_ext { + struct save_area sa; + __vector128 vx_regs[32]; +}; + struct _lowcore { psw_t restart_psw; /* 0x0000 */ psw_t restart_old_psw; /* 0x0008 */ @@ -183,6 +189,11 @@ struct save_area { u64 ctrl_regs[16]; } __packed; +struct save_area_ext { + struct save_area sa; + __vector128 vx_regs[32]; +}; + struct _lowcore { __u8 pad_0x0000[0x0014-0x0000]; /* 0x0000 */ __u32 ipl_parmblock_ptr; /* 0x0014 */ @@ -310,7 +321,10 @@ struct _lowcore { /* Extended facility list */ __u64 stfle_fac_list[32]; /* 0x0f00 */ - __u8 pad_0x1000[0x11b8-0x1000]; /* 0x1000 */ + __u8 pad_0x1000[0x11b0-0x1000]; /* 0x1000 */ + + /* Pointer to vector register save area */ + __u64 vector_save_area_addr; /* 0x11b0 */ /* 64 bit extparam used for pfault/diag 250: defined by architecture */ __u64 ext_params2; /* 0x11B8 */ @@ -334,9 +348,10 @@ struct _lowcore { /* Transaction abort diagnostic block */ __u8 pgm_tdb[256]; /* 0x1800 */ + __u8 pad_0x1900[0x1c00-0x1900]; /* 0x1900 */ - /* align to the top of the prefix area */ - __u8 pad_0x1900[0x2000-0x1900]; /* 0x1900 */ + /* Software defined save area for vector registers */ + __u8 vector_save_area[1024]; /* 0x1c00 */ } __packed; #endif /* CONFIG_32BIT */ diff --git a/arch/s390/include/asm/nmi.h b/arch/s390/include/asm/nmi.h index 35f8ec1..3027a5a 100644 --- a/arch/s390/include/asm/nmi.h +++ b/arch/s390/include/asm/nmi.h @@ -38,7 +38,7 @@ struct mci { __u32 pm : 1; /* 22 psw program mask and cc validity */ __u32 ia : 1; /* 23 psw instruction address validity */ __u32 fa : 1; /* 24 failing storage address validity */ - __u32 : 1; /* 25 */ + __u32 vr : 1; /* 25 vector register validity */ __u32 ec : 1; /* 26 external damage code validity */ __u32 fp : 1; /* 27 floating point register validity */ __u32 gr : 1; /* 28 general register validity */ diff --git a/arch/s390/include/asm/page.h b/arch/s390/include/asm/page.h index 114258e..7b2ac6e 100644 --- a/arch/s390/include/asm/page.h +++ b/arch/s390/include/asm/page.h @@ -162,6 +162,4 @@ static inline int devmem_is_allowed(unsigned long pfn) #include <asm-generic/memory_model.h> #include <asm-generic/getorder.h> -#define __HAVE_ARCH_GATE_AREA 1 - #endif /* _S390_PAGE_H */ diff --git a/arch/s390/include/asm/percpu.h b/arch/s390/include/asm/percpu.h index fa91e00..933355e 100644 --- a/arch/s390/include/asm/percpu.h +++ b/arch/s390/include/asm/percpu.h @@ -31,7 +31,7 @@ pcp_op_T__ old__, new__, prev__; \ pcp_op_T__ *ptr__; \ preempt_disable(); \ - ptr__ = __this_cpu_ptr(&(pcp)); \ + ptr__ = raw_cpu_ptr(&(pcp)); \ prev__ = *ptr__; \ do { \ old__ = prev__; \ @@ -70,7 +70,7 @@ pcp_op_T__ val__ = (val); \ pcp_op_T__ old__, *ptr__; \ preempt_disable(); \ - ptr__ = __this_cpu_ptr(&(pcp)); \ + ptr__ = raw_cpu_ptr(&(pcp)); \ if (__builtin_constant_p(val__) && \ ((szcast)val__ > -129) && ((szcast)val__ < 128)) { \ asm volatile( \ @@ -97,7 +97,7 @@ pcp_op_T__ val__ = (val); \ pcp_op_T__ old__, *ptr__; \ preempt_disable(); \ - ptr__ = __this_cpu_ptr(&(pcp)); \ + ptr__ = raw_cpu_ptr(&(pcp)); \ asm volatile( \ op " %[old__],%[val__],%[ptr__]\n" \ : [old__] "=d" (old__), [ptr__] "+Q" (*ptr__) \ @@ -116,7 +116,7 @@ pcp_op_T__ val__ = (val); \ pcp_op_T__ old__, *ptr__; \ preempt_disable(); \ - ptr__ = __this_cpu_ptr(&(pcp)); \ + ptr__ = raw_cpu_ptr(&(pcp)); \ asm volatile( \ op " %[old__],%[val__],%[ptr__]\n" \ : [old__] "=d" (old__), [ptr__] "+Q" (*ptr__) \ @@ -138,7 +138,7 @@ pcp_op_T__ ret__; \ pcp_op_T__ *ptr__; \ preempt_disable(); \ - ptr__ = __this_cpu_ptr(&(pcp)); \ + ptr__ = raw_cpu_ptr(&(pcp)); \ ret__ = cmpxchg(ptr__, oval, nval); \ preempt_enable(); \ ret__; \ @@ -154,7 +154,7 @@ typeof(pcp) *ptr__; \ typeof(pcp) ret__; \ preempt_disable(); \ - ptr__ = __this_cpu_ptr(&(pcp)); \ + ptr__ = raw_cpu_ptr(&(pcp)); \ ret__ = xchg(ptr__, nval); \ preempt_enable(); \ ret__; \ @@ -173,8 +173,8 @@ typeof(pcp2) *p2__; \ int ret__; \ preempt_disable(); \ - p1__ = __this_cpu_ptr(&(pcp1)); \ - p2__ = __this_cpu_ptr(&(pcp2)); \ + p1__ = raw_cpu_ptr(&(pcp1)); \ + p2__ = raw_cpu_ptr(&(pcp2)); \ ret__ = __cmpxchg_double(p1__, p2__, o1__, o2__, n1__, n2__); \ preempt_enable(); \ ret__; \ diff --git a/arch/s390/include/asm/pgalloc.h b/arch/s390/include/asm/pgalloc.h index 9e18a61..d39a31c 100644 --- a/arch/s390/include/asm/pgalloc.h +++ b/arch/s390/include/asm/pgalloc.h @@ -18,9 +18,9 @@ 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 *, unsigned long); +unsigned long *page_table_alloc(struct mm_struct *); void page_table_free(struct mm_struct *, unsigned long *); -void page_table_free_rcu(struct mmu_gather *, unsigned long *); +void page_table_free_rcu(struct mmu_gather *, unsigned long *, unsigned long); void page_table_reset_pgste(struct mm_struct *, unsigned long, unsigned long, bool init_skey); @@ -145,8 +145,8 @@ static inline void pmd_populate(struct mm_struct *mm, /* * page table entry allocation/free routines. */ -#define pte_alloc_one_kernel(mm, vmaddr) ((pte_t *) page_table_alloc(mm, vmaddr)) -#define pte_alloc_one(mm, vmaddr) ((pte_t *) page_table_alloc(mm, vmaddr)) +#define pte_alloc_one_kernel(mm, vmaddr) ((pte_t *) page_table_alloc(mm)) +#define pte_alloc_one(mm, vmaddr) ((pte_t *) page_table_alloc(mm)) #define pte_free_kernel(mm, pte) page_table_free(mm, (unsigned long *) pte) #define pte_free(mm, pte) page_table_free(mm, (unsigned long *) pte) diff --git a/arch/s390/include/asm/pgtable.h b/arch/s390/include/asm/pgtable.h index fcba5e0..57c8827 100644 --- a/arch/s390/include/asm/pgtable.h +++ b/arch/s390/include/asm/pgtable.h @@ -30,6 +30,7 @@ #include <linux/sched.h> #include <linux/mm_types.h> #include <linux/page-flags.h> +#include <linux/radix-tree.h> #include <asm/bug.h> #include <asm/page.h> @@ -216,7 +217,6 @@ extern unsigned long MODULES_END; */ /* Hardware bits in the page table entry */ -#define _PAGE_CO 0x100 /* HW Change-bit override */ #define _PAGE_PROTECT 0x200 /* HW read-only bit */ #define _PAGE_INVALID 0x400 /* HW invalid bit */ #define _PAGE_LARGE 0x800 /* Bit to mark a large pte */ @@ -233,8 +233,8 @@ extern unsigned long MODULES_END; #define __HAVE_ARCH_PTE_SPECIAL /* Set of bits not changed in pte_modify */ -#define _PAGE_CHG_MASK (PAGE_MASK | _PAGE_SPECIAL | _PAGE_CO | \ - _PAGE_DIRTY | _PAGE_YOUNG) +#define _PAGE_CHG_MASK (PAGE_MASK | _PAGE_SPECIAL | _PAGE_DIRTY | \ + _PAGE_YOUNG) /* * handle_pte_fault uses pte_present, pte_none and pte_file to find out the @@ -287,7 +287,14 @@ extern unsigned long MODULES_END; #define _SEGMENT_ENTRY_INVALID 0x20 /* invalid segment table entry */ #define _SEGMENT_ENTRY_COMMON 0x10 /* common segment bit */ #define _SEGMENT_ENTRY_PTL 0x0f /* page table length */ -#define _SEGMENT_ENTRY_NONE _SEGMENT_ENTRY_PROTECT + +#define _SEGMENT_ENTRY_DIRTY 0 /* No sw dirty bit for 31-bit */ +#define _SEGMENT_ENTRY_YOUNG 0 /* No sw young bit for 31-bit */ +#define _SEGMENT_ENTRY_READ 0 /* No sw read bit for 31-bit */ +#define _SEGMENT_ENTRY_WRITE 0 /* No sw write bit for 31-bit */ +#define _SEGMENT_ENTRY_LARGE 0 /* No large pages for 31-bit */ +#define _SEGMENT_ENTRY_BITS_LARGE 0 +#define _SEGMENT_ENTRY_ORIGIN_LARGE 0 #define _SEGMENT_ENTRY (_SEGMENT_ENTRY_PTL) #define _SEGMENT_ENTRY_EMPTY (_SEGMENT_ENTRY_INVALID) @@ -346,11 +353,10 @@ extern unsigned long MODULES_END; #define _REGION3_ENTRY_LARGE 0x400 /* RTTE-format control, large page */ #define _REGION3_ENTRY_RO 0x200 /* page protection bit */ -#define _REGION3_ENTRY_CO 0x100 /* change-recording override */ /* Bits in the segment table entry */ #define _SEGMENT_ENTRY_BITS 0xfffffffffffffe33UL -#define _SEGMENT_ENTRY_BITS_LARGE 0xfffffffffff1ff33UL +#define _SEGMENT_ENTRY_BITS_LARGE 0xfffffffffff0ff33UL #define _SEGMENT_ENTRY_ORIGIN_LARGE ~0xfffffUL /* large page address */ #define _SEGMENT_ENTRY_ORIGIN ~0x7ffUL/* segment table origin */ #define _SEGMENT_ENTRY_PROTECT 0x200 /* page protection bit */ @@ -359,30 +365,33 @@ extern unsigned long MODULES_END; #define _SEGMENT_ENTRY (0) #define _SEGMENT_ENTRY_EMPTY (_SEGMENT_ENTRY_INVALID) -#define _SEGMENT_ENTRY_LARGE 0x400 /* STE-format control, large page */ -#define _SEGMENT_ENTRY_CO 0x100 /* change-recording override */ -#define _SEGMENT_ENTRY_SPLIT 0x001 /* THP splitting bit */ -#define _SEGMENT_ENTRY_YOUNG 0x002 /* SW segment young bit */ -#define _SEGMENT_ENTRY_NONE _SEGMENT_ENTRY_YOUNG +#define _SEGMENT_ENTRY_DIRTY 0x2000 /* SW segment dirty bit */ +#define _SEGMENT_ENTRY_YOUNG 0x1000 /* SW segment young bit */ +#define _SEGMENT_ENTRY_SPLIT 0x0800 /* THP splitting bit */ +#define _SEGMENT_ENTRY_LARGE 0x0400 /* STE-format control, large page */ +#define _SEGMENT_ENTRY_READ 0x0002 /* SW segment read bit */ +#define _SEGMENT_ENTRY_WRITE 0x0001 /* SW segment write bit */ /* * Segment table entry encoding (R = read-only, I = invalid, y = young bit): - * ..R...I...y. - * prot-none, old ..0...1...1. - * prot-none, young ..1...1...1. - * read-only, old ..1...1...0. - * read-only, young ..1...0...1. - * read-write, old ..0...1...0. - * read-write, young ..0...0...1. + * dy..R...I...wr + * prot-none, clean, old 00..1...1...00 + * prot-none, clean, young 01..1...1...00 + * prot-none, dirty, old 10..1...1...00 + * prot-none, dirty, young 11..1...1...00 + * read-only, clean, old 00..1...1...01 + * read-only, clean, young 01..1...0...01 + * read-only, dirty, old 10..1...1...01 + * read-only, dirty, young 11..1...0...01 + * read-write, clean, old 00..1...1...11 + * read-write, clean, young 01..1...0...11 + * read-write, dirty, old 10..0...1...11 + * read-write, dirty, young 11..0...0...11 * The segment table origin is used to distinguish empty (origin==0) from * read-write, old segment table entries (origin!=0) */ -#define _SEGMENT_ENTRY_SPLIT_BIT 0 /* THP splitting bit number */ - -/* Set of bits not changed in pmd_modify */ -#define _SEGMENT_CHG_MASK (_SEGMENT_ENTRY_ORIGIN | _SEGMENT_ENTRY_LARGE \ - | _SEGMENT_ENTRY_SPLIT | _SEGMENT_ENTRY_CO) +#define _SEGMENT_ENTRY_SPLIT_BIT 11 /* THP splitting bit number */ /* Page status table bits for virtualization */ #define PGSTE_ACC_BITS 0xf000000000000000UL @@ -455,10 +464,11 @@ extern unsigned long MODULES_END; * Segment entry (large page) protection definitions. */ #define SEGMENT_NONE __pgprot(_SEGMENT_ENTRY_INVALID | \ - _SEGMENT_ENTRY_NONE) -#define SEGMENT_READ __pgprot(_SEGMENT_ENTRY_INVALID | \ _SEGMENT_ENTRY_PROTECT) -#define SEGMENT_WRITE __pgprot(_SEGMENT_ENTRY_INVALID) +#define SEGMENT_READ __pgprot(_SEGMENT_ENTRY_PROTECT | \ + _SEGMENT_ENTRY_READ) +#define SEGMENT_WRITE __pgprot(_SEGMENT_ENTRY_READ | \ + _SEGMENT_ENTRY_WRITE) static inline int mm_has_pgste(struct mm_struct *mm) { @@ -569,25 +579,23 @@ static inline int pmd_none(pmd_t pmd) static inline int pmd_large(pmd_t pmd) { -#ifdef CONFIG_64BIT return (pmd_val(pmd) & _SEGMENT_ENTRY_LARGE) != 0; -#else - return 0; -#endif } -static inline int pmd_prot_none(pmd_t pmd) +static inline int pmd_pfn(pmd_t pmd) { - return (pmd_val(pmd) & _SEGMENT_ENTRY_INVALID) && - (pmd_val(pmd) & _SEGMENT_ENTRY_NONE); + unsigned long origin_mask; + + origin_mask = _SEGMENT_ENTRY_ORIGIN; + if (pmd_large(pmd)) + origin_mask = _SEGMENT_ENTRY_ORIGIN_LARGE; + return (pmd_val(pmd) & origin_mask) >> PAGE_SHIFT; } static inline int pmd_bad(pmd_t pmd) { -#ifdef CONFIG_64BIT if (pmd_large(pmd)) return (pmd_val(pmd) & ~_SEGMENT_ENTRY_BITS_LARGE) != 0; -#endif return (pmd_val(pmd) & ~_SEGMENT_ENTRY_BITS) != 0; } @@ -607,20 +615,22 @@ extern int pmdp_clear_flush_young(struct vm_area_struct *vma, #define __HAVE_ARCH_PMD_WRITE static inline int pmd_write(pmd_t pmd) { - if (pmd_prot_none(pmd)) - return 0; - return (pmd_val(pmd) & _SEGMENT_ENTRY_PROTECT) == 0; + return (pmd_val(pmd) & _SEGMENT_ENTRY_WRITE) != 0; +} + +static inline int pmd_dirty(pmd_t pmd) +{ + int dirty = 1; + if (pmd_large(pmd)) + dirty = (pmd_val(pmd) & _SEGMENT_ENTRY_DIRTY) != 0; + return dirty; } static inline int pmd_young(pmd_t pmd) { - int young = 0; -#ifdef CONFIG_64BIT - if (pmd_prot_none(pmd)) - young = (pmd_val(pmd) & _SEGMENT_ENTRY_PROTECT) != 0; - else + int young = 1; + if (pmd_large(pmd)) young = (pmd_val(pmd) & _SEGMENT_ENTRY_YOUNG) != 0; -#endif return young; } @@ -777,82 +787,67 @@ static inline pgste_t pgste_set_pte(pte_t *ptep, pgste_t pgste, pte_t entry) /** * struct gmap_struct - guest address space + * @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 * @table: pointer to the page directory * @asce: address space control element for gmap page table - * @crst_list: list of all crst tables used in the guest address space * @pfault_enabled: defines if pfaults are applicable for the guest */ struct gmap { struct list_head list; + struct list_head crst_list; struct mm_struct *mm; + struct radix_tree_root guest_to_host; + struct radix_tree_root host_to_guest; + spinlock_t guest_table_lock; unsigned long *table; unsigned long asce; + unsigned long asce_end; void *private; - struct list_head crst_list; bool pfault_enabled; }; /** - * struct gmap_rmap - reverse mapping for segment table entries - * @gmap: pointer to the gmap_struct - * @entry: pointer to a segment table entry - * @vmaddr: virtual address in the guest address space - */ -struct gmap_rmap { - struct list_head list; - struct gmap *gmap; - unsigned long *entry; - unsigned long vmaddr; -}; - -/** - * struct gmap_pgtable - gmap information attached to a page table - * @vmaddr: address of the 1MB segment in the process virtual memory - * @mapper: list of segment table entries mapping a page table - */ -struct gmap_pgtable { - unsigned long vmaddr; - struct list_head mapper; -}; - -/** * 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 address); + void (*notifier_call)(struct gmap *gmap, unsigned long gaddr); }; -struct gmap *gmap_alloc(struct mm_struct *mm); +struct gmap *gmap_alloc(struct mm_struct *mm, unsigned long limit); void gmap_free(struct gmap *gmap); void gmap_enable(struct gmap *gmap); void gmap_disable(struct gmap *gmap); 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); -unsigned long __gmap_translate(unsigned long address, struct gmap *); -unsigned long gmap_translate(unsigned long address, struct gmap *); -unsigned long __gmap_fault(unsigned long address, struct gmap *); -unsigned long gmap_fault(unsigned long address, struct gmap *); -void gmap_discard(unsigned long from, unsigned long to, struct gmap *); -void __gmap_zap(unsigned long address, struct gmap *); +unsigned long __gmap_translate(struct gmap *, unsigned long gaddr); +unsigned long gmap_translate(struct gmap *, unsigned long gaddr); +int __gmap_link(struct gmap *gmap, unsigned long gaddr, unsigned long vmaddr); +int gmap_fault(struct gmap *, unsigned long gaddr, unsigned int fault_flags); +void gmap_discard(struct gmap *, unsigned long from, unsigned long to); +void __gmap_zap(struct gmap *, unsigned long gaddr); bool gmap_test_and_clear_dirty(unsigned long address, struct gmap *); 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); -void gmap_do_ipte_notify(struct mm_struct *, pte_t *); +void gmap_do_ipte_notify(struct mm_struct *, unsigned long addr, pte_t *); static inline pgste_t pgste_ipte_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; - gmap_do_ipte_notify(mm, ptep); + gmap_do_ipte_notify(mm, addr, ptep); } #endif return pgste; @@ -875,8 +870,6 @@ static inline void set_pte_at(struct mm_struct *mm, unsigned long addr, pgste = pgste_set_pte(ptep, pgste, entry); pgste_set_unlock(ptep, pgste); } else { - if (!(pte_val(entry) & _PAGE_INVALID) && MACHINE_HAS_EDAT1) - pte_val(entry) |= _PAGE_CO; *ptep = entry; } } @@ -1046,6 +1039,22 @@ static inline void __ptep_ipte_local(unsigned long address, pte_t *ptep) : "=m" (*ptep) : "m" (*ptep), "a" (pto), "a" (address)); } +static inline void __ptep_ipte_range(unsigned long address, int nr, pte_t *ptep) +{ + unsigned long pto = (unsigned long) ptep; + +#ifndef CONFIG_64BIT + /* pto in ESA mode must point to the start of the segment table */ + pto &= 0x7ffffc00; +#endif + /* Invalidate a range of ptes + global TLB flush of the ptes */ + do { + asm volatile( + " .insn rrf,0xb2210000,%2,%0,%1,0" + : "+a" (address), "+a" (nr) : "a" (pto) : "memory"); + } while (nr != 255); +} + static inline void ptep_flush_direct(struct mm_struct *mm, unsigned long address, pte_t *ptep) { @@ -1098,7 +1107,7 @@ static inline int ptep_test_and_clear_user_dirty(struct mm_struct *mm, pgste_val(pgste) &= ~PGSTE_UC_BIT; pte = *ptep; if (dirty && (pte_val(pte) & _PAGE_PRESENT)) { - pgste = pgste_ipte_notify(mm, ptep, pgste); + pgste = pgste_ipte_notify(mm, addr, ptep, pgste); __ptep_ipte(addr, ptep); if (MACHINE_HAS_ESOP || !(pte_val(pte) & _PAGE_WRITE)) pte_val(pte) |= _PAGE_PROTECT; @@ -1115,20 +1124,21 @@ static inline int ptep_test_and_clear_young(struct vm_area_struct *vma, unsigned long addr, pte_t *ptep) { pgste_t pgste; - pte_t pte; + pte_t pte, oldpte; int young; if (mm_has_pgste(vma->vm_mm)) { pgste = pgste_get_lock(ptep); - pgste = pgste_ipte_notify(vma->vm_mm, ptep, pgste); + pgste = pgste_ipte_notify(vma->vm_mm, addr, ptep, pgste); } - pte = *ptep; + oldpte = pte = *ptep; ptep_flush_direct(vma->vm_mm, addr, ptep); young = pte_young(pte); pte = pte_mkold(pte); if (mm_has_pgste(vma->vm_mm)) { + pgste = pgste_update_all(&oldpte, pgste, vma->vm_mm); pgste = pgste_set_pte(ptep, pgste, pte); pgste_set_unlock(ptep, pgste); } else @@ -1166,7 +1176,7 @@ static inline pte_t ptep_get_and_clear(struct mm_struct *mm, if (mm_has_pgste(mm)) { pgste = pgste_get_lock(ptep); - pgste = pgste_ipte_notify(mm, ptep, pgste); + pgste = pgste_ipte_notify(mm, address, ptep, pgste); } pte = *ptep; @@ -1190,7 +1200,7 @@ static inline pte_t ptep_modify_prot_start(struct mm_struct *mm, if (mm_has_pgste(mm)) { pgste = pgste_get_lock(ptep); - pgste_ipte_notify(mm, ptep, pgste); + pgste_ipte_notify(mm, address, ptep, pgste); } pte = *ptep; @@ -1227,7 +1237,7 @@ static inline pte_t ptep_clear_flush(struct vm_area_struct *vma, if (mm_has_pgste(vma->vm_mm)) { pgste = pgste_get_lock(ptep); - pgste = pgste_ipte_notify(vma->vm_mm, ptep, pgste); + pgste = pgste_ipte_notify(vma->vm_mm, address, ptep, pgste); } pte = *ptep; @@ -1261,7 +1271,7 @@ static inline pte_t ptep_get_and_clear_full(struct mm_struct *mm, if (!full && mm_has_pgste(mm)) { pgste = pgste_get_lock(ptep); - pgste = pgste_ipte_notify(mm, ptep, pgste); + pgste = pgste_ipte_notify(mm, address, ptep, pgste); } pte = *ptep; @@ -1286,7 +1296,7 @@ static inline pte_t ptep_set_wrprotect(struct mm_struct *mm, if (pte_write(pte)) { if (mm_has_pgste(mm)) { pgste = pgste_get_lock(ptep); - pgste = pgste_ipte_notify(mm, ptep, pgste); + pgste = pgste_ipte_notify(mm, address, ptep, pgste); } ptep_flush_lazy(mm, address, ptep); @@ -1312,12 +1322,13 @@ static inline int ptep_set_access_flags(struct vm_area_struct *vma, return 0; if (mm_has_pgste(vma->vm_mm)) { pgste = pgste_get_lock(ptep); - pgste = pgste_ipte_notify(vma->vm_mm, ptep, pgste); + pgste = pgste_ipte_notify(vma->vm_mm, address, ptep, pgste); } ptep_flush_direct(vma->vm_mm, address, ptep); if (mm_has_pgste(vma->vm_mm)) { + pgste_set_key(ptep, pgste, entry, vma->vm_mm); pgste = pgste_set_pte(ptep, pgste, entry); pgste_set_unlock(ptep, pgste); } else @@ -1391,7 +1402,7 @@ static inline pmd_t *pmd_offset(pud_t *pud, unsigned long address) #define pte_pfn(x) (pte_val(x) >> PAGE_SHIFT) #define pte_page(x) pfn_to_page(pte_pfn(x)) -#define pmd_page(pmd) pfn_to_page(pmd_val(pmd) >> PAGE_SHIFT) +#define pmd_page(pmd) pfn_to_page(pmd_pfn(pmd)) /* Find an entry in the lowest level page table.. */ #define pte_offset(pmd, addr) ((pte_t *) pmd_deref(*(pmd)) + pte_index(addr)) @@ -1413,41 +1424,75 @@ static inline unsigned long massage_pgprot_pmd(pgprot_t pgprot) return pgprot_val(SEGMENT_WRITE); } -static inline pmd_t pmd_mkyoung(pmd_t pmd) +static inline pmd_t pmd_wrprotect(pmd_t pmd) { -#ifdef CONFIG_64BIT - if (pmd_prot_none(pmd)) { + pmd_val(pmd) &= ~_SEGMENT_ENTRY_WRITE; + pmd_val(pmd) |= _SEGMENT_ENTRY_PROTECT; + return pmd; +} + +static inline pmd_t pmd_mkwrite(pmd_t pmd) +{ + pmd_val(pmd) |= _SEGMENT_ENTRY_WRITE; + if (pmd_large(pmd) && !(pmd_val(pmd) & _SEGMENT_ENTRY_DIRTY)) + return pmd; + pmd_val(pmd) &= ~_SEGMENT_ENTRY_PROTECT; + return pmd; +} + +static inline pmd_t pmd_mkclean(pmd_t pmd) +{ + if (pmd_large(pmd)) { + pmd_val(pmd) &= ~_SEGMENT_ENTRY_DIRTY; pmd_val(pmd) |= _SEGMENT_ENTRY_PROTECT; - } else { + } + return pmd; +} + +static inline pmd_t pmd_mkdirty(pmd_t pmd) +{ + if (pmd_large(pmd)) { + pmd_val(pmd) |= _SEGMENT_ENTRY_DIRTY; + if (pmd_val(pmd) & _SEGMENT_ENTRY_WRITE) + pmd_val(pmd) &= ~_SEGMENT_ENTRY_PROTECT; + } + return pmd; +} + +static inline pmd_t pmd_mkyoung(pmd_t pmd) +{ + if (pmd_large(pmd)) { pmd_val(pmd) |= _SEGMENT_ENTRY_YOUNG; - pmd_val(pmd) &= ~_SEGMENT_ENTRY_INVALID; + if (pmd_val(pmd) & _SEGMENT_ENTRY_READ) + pmd_val(pmd) &= ~_SEGMENT_ENTRY_INVALID; } -#endif return pmd; } static inline pmd_t pmd_mkold(pmd_t pmd) { -#ifdef CONFIG_64BIT - if (pmd_prot_none(pmd)) { - pmd_val(pmd) &= ~_SEGMENT_ENTRY_PROTECT; - } else { + if (pmd_large(pmd)) { pmd_val(pmd) &= ~_SEGMENT_ENTRY_YOUNG; pmd_val(pmd) |= _SEGMENT_ENTRY_INVALID; } -#endif return pmd; } static inline pmd_t pmd_modify(pmd_t pmd, pgprot_t newprot) { - int young; - - young = pmd_young(pmd); - pmd_val(pmd) &= _SEGMENT_CHG_MASK; + if (pmd_large(pmd)) { + pmd_val(pmd) &= _SEGMENT_ENTRY_ORIGIN_LARGE | + _SEGMENT_ENTRY_DIRTY | _SEGMENT_ENTRY_YOUNG | + _SEGMENT_ENTRY_LARGE | _SEGMENT_ENTRY_SPLIT; + pmd_val(pmd) |= massage_pgprot_pmd(newprot); + if (!(pmd_val(pmd) & _SEGMENT_ENTRY_DIRTY)) + pmd_val(pmd) |= _SEGMENT_ENTRY_PROTECT; + if (!(pmd_val(pmd) & _SEGMENT_ENTRY_YOUNG)) + pmd_val(pmd) |= _SEGMENT_ENTRY_INVALID; + return pmd; + } + pmd_val(pmd) &= _SEGMENT_ENTRY_ORIGIN; pmd_val(pmd) |= massage_pgprot_pmd(newprot); - if (young) - pmd = pmd_mkyoung(pmd); return pmd; } @@ -1455,16 +1500,9 @@ static inline pmd_t mk_pmd_phys(unsigned long physpage, pgprot_t pgprot) { pmd_t __pmd; pmd_val(__pmd) = physpage + massage_pgprot_pmd(pgprot); - return pmd_mkyoung(__pmd); + return __pmd; } -static inline pmd_t pmd_mkwrite(pmd_t pmd) -{ - /* Do not clobber PROT_NONE segments! */ - if (!pmd_prot_none(pmd)) - pmd_val(pmd) &= ~_SEGMENT_ENTRY_PROTECT; - return pmd; -} #endif /* CONFIG_TRANSPARENT_HUGEPAGE || CONFIG_HUGETLB_PAGE */ static inline void __pmdp_csp(pmd_t *pmdp) @@ -1555,34 +1593,21 @@ extern pgtable_t pgtable_trans_huge_withdraw(struct mm_struct *mm, pmd_t *pmdp); static inline int pmd_trans_splitting(pmd_t pmd) { - return pmd_val(pmd) & _SEGMENT_ENTRY_SPLIT; + return (pmd_val(pmd) & _SEGMENT_ENTRY_LARGE) && + (pmd_val(pmd) & _SEGMENT_ENTRY_SPLIT); } static inline void set_pmd_at(struct mm_struct *mm, unsigned long addr, pmd_t *pmdp, pmd_t entry) { - if (!(pmd_val(entry) & _SEGMENT_ENTRY_INVALID) && MACHINE_HAS_EDAT1) - pmd_val(entry) |= _SEGMENT_ENTRY_CO; *pmdp = entry; } static inline pmd_t pmd_mkhuge(pmd_t pmd) { pmd_val(pmd) |= _SEGMENT_ENTRY_LARGE; - return pmd; -} - -static inline pmd_t pmd_wrprotect(pmd_t pmd) -{ - /* Do not clobber PROT_NONE segments! */ - if (!pmd_prot_none(pmd)) - pmd_val(pmd) |= _SEGMENT_ENTRY_PROTECT; - return pmd; -} - -static inline pmd_t pmd_mkdirty(pmd_t pmd) -{ - /* No dirty bit in the segment table entry. */ + pmd_val(pmd) |= _SEGMENT_ENTRY_YOUNG; + pmd_val(pmd) |= _SEGMENT_ENTRY_PROTECT; return pmd; } @@ -1647,11 +1672,6 @@ static inline int has_transparent_hugepage(void) { return MACHINE_HAS_HPAGE ? 1 : 0; } - -static inline unsigned long pmd_pfn(pmd_t pmd) -{ - return pmd_val(pmd) >> PAGE_SHIFT; -} #endif /* CONFIG_TRANSPARENT_HUGEPAGE */ /* diff --git a/arch/s390/include/asm/processor.h b/arch/s390/include/asm/processor.h index 6f02d45..d559bdb 100644 --- a/arch/s390/include/asm/processor.h +++ b/arch/s390/include/asm/processor.h @@ -13,9 +13,11 @@ #define CIF_MCCK_PENDING 0 /* machine check handling is pending */ #define CIF_ASCE 1 /* user asce needs fixup / uaccess */ +#define CIF_NOHZ_DELAY 2 /* delay HZ disable for a tick */ #define _CIF_MCCK_PENDING (1<<CIF_MCCK_PENDING) #define _CIF_ASCE (1<<CIF_ASCE) +#define _CIF_NOHZ_DELAY (1<<CIF_NOHZ_DELAY) #ifndef __ASSEMBLY__ @@ -43,6 +45,8 @@ static inline int test_cpu_flag(int flag) return !!(S390_lowcore.cpu_flags & (1U << flag)); } +#define arch_needs_cpu() test_cpu_flag(CIF_NOHZ_DELAY) + /* * Default implementation of macro that returns current * instruction pointer ("program counter"). @@ -113,6 +117,7 @@ struct thread_struct { int ri_signum; #ifdef CONFIG_64BIT unsigned char trap_tdb[256]; /* Transaction abort diagnose block */ + __vector128 *vxrs; /* Vector register save area */ #endif }; @@ -217,7 +222,7 @@ static inline void cpu_relax(void) barrier(); } -#define arch_mutex_cpu_relax() barrier() +#define cpu_relax_lowlatency() barrier() static inline void psw_set_key(unsigned int key) { @@ -285,7 +290,12 @@ static inline unsigned long __rewind_psw(psw_t psw, unsigned long ilc) return (psw.addr - ilc) & mask; #endif } - + +/* + * Function to stop a processor until the next interrupt occurs + */ +void enabled_wait(void); + /* * Function to drop a processor into disabled wait state */ diff --git a/arch/s390/include/asm/ptrace.h b/arch/s390/include/asm/ptrace.h index 55d69dd..be317fe 100644 --- a/arch/s390/include/asm/ptrace.h +++ b/arch/s390/include/asm/ptrace.h @@ -161,6 +161,12 @@ static inline long regs_return_value(struct pt_regs *regs) return regs->gprs[2]; } +static inline void instruction_pointer_set(struct pt_regs *regs, + unsigned long val) +{ + regs->psw.addr = val | PSW_ADDR_AMODE; +} + int regs_query_register_offset(const char *name); const char *regs_query_register_name(unsigned int offset); unsigned long regs_get_register(struct pt_regs *regs, unsigned int offset); diff --git a/arch/s390/include/asm/qdio.h b/arch/s390/include/asm/qdio.h index d786c63..06f3034 100644 --- a/arch/s390/include/asm/qdio.h +++ b/arch/s390/include/asm/qdio.h @@ -415,6 +415,10 @@ struct qdio_brinfo_entry_l2 { #define QDIO_FLAG_SYNC_OUTPUT 0x02 #define QDIO_FLAG_PCI_OUT 0x10 +int qdio_alloc_buffers(struct qdio_buffer **buf, unsigned int count); +void qdio_free_buffers(struct qdio_buffer **buf, unsigned int count); +void qdio_reset_buffers(struct qdio_buffer **buf, unsigned int count); + extern int qdio_allocate(struct qdio_initialize *); extern int qdio_establish(struct qdio_initialize *); extern int qdio_activate(struct ccw_device *); diff --git a/arch/s390/include/asm/scatterlist.h b/arch/s390/include/asm/scatterlist.h deleted file mode 100644 index 6d45ef6..0000000 --- a/arch/s390/include/asm/scatterlist.h +++ /dev/null @@ -1,3 +0,0 @@ -#include <asm-generic/scatterlist.h> - -#define ARCH_HAS_SG_CHAIN diff --git a/arch/s390/include/asm/setup.h b/arch/s390/include/asm/setup.h index 089a498..7736fdd 100644 --- a/arch/s390/include/asm/setup.h +++ b/arch/s390/include/asm/setup.h @@ -55,8 +55,8 @@ extern void detect_memory_memblock(void); #define MACHINE_FLAG_LPP (1UL << 13) #define MACHINE_FLAG_TOPOLOGY (1UL << 14) #define MACHINE_FLAG_TE (1UL << 15) -#define MACHINE_FLAG_RRBM (1UL << 16) #define MACHINE_FLAG_TLB_LC (1UL << 17) +#define MACHINE_FLAG_VX (1UL << 18) #define MACHINE_IS_VM (S390_lowcore.machine_flags & MACHINE_FLAG_VM) #define MACHINE_IS_KVM (S390_lowcore.machine_flags & MACHINE_FLAG_KVM) @@ -78,8 +78,8 @@ extern void detect_memory_memblock(void); #define MACHINE_HAS_LPP (0) #define MACHINE_HAS_TOPOLOGY (0) #define MACHINE_HAS_TE (0) -#define MACHINE_HAS_RRBM (0) #define MACHINE_HAS_TLB_LC (0) +#define MACHINE_HAS_VX (0) #else /* CONFIG_64BIT */ #define MACHINE_HAS_IEEE (1) #define MACHINE_HAS_CSP (1) @@ -91,8 +91,8 @@ extern void detect_memory_memblock(void); #define MACHINE_HAS_LPP (S390_lowcore.machine_flags & MACHINE_FLAG_LPP) #define MACHINE_HAS_TOPOLOGY (S390_lowcore.machine_flags & MACHINE_FLAG_TOPOLOGY) #define MACHINE_HAS_TE (S390_lowcore.machine_flags & MACHINE_FLAG_TE) -#define MACHINE_HAS_RRBM (S390_lowcore.machine_flags & MACHINE_FLAG_RRBM) #define MACHINE_HAS_TLB_LC (S390_lowcore.machine_flags & MACHINE_FLAG_TLB_LC) +#define MACHINE_HAS_VX (S390_lowcore.machine_flags & MACHINE_FLAG_VX) #endif /* CONFIG_64BIT */ /* diff --git a/arch/s390/include/asm/sigp.h b/arch/s390/include/asm/sigp.h index bf9c823..4957611 100644 --- a/arch/s390/include/asm/sigp.h +++ b/arch/s390/include/asm/sigp.h @@ -15,6 +15,7 @@ #define SIGP_SET_ARCHITECTURE 18 #define SIGP_COND_EMERGENCY_SIGNAL 19 #define SIGP_SENSE_RUNNING 21 +#define SIGP_STORE_ADDITIONAL_STATUS 23 /* SIGP condition codes */ #define SIGP_CC_ORDER_CODE_ACCEPTED 0 @@ -33,9 +34,10 @@ #ifndef __ASSEMBLY__ -static inline int __pcpu_sigp(u16 addr, u8 order, u32 parm, u32 *status) +static inline int __pcpu_sigp(u16 addr, u8 order, unsigned long parm, + u32 *status) { - register unsigned int reg1 asm ("1") = parm; + register unsigned long reg1 asm ("1") = parm; int cc; asm volatile( diff --git a/arch/s390/include/asm/smp.h b/arch/s390/include/asm/smp.h index 4f13079..762d4f8 100644 --- a/arch/s390/include/asm/smp.h +++ b/arch/s390/include/asm/smp.h @@ -29,7 +29,6 @@ extern int smp_find_processor_id(u16 address); extern int smp_store_status(int cpu); extern int smp_vcpu_scheduled(int cpu); extern void smp_yield_cpu(int cpu); -extern void smp_yield(void); extern void smp_cpu_set_polarization(int cpu, int val); extern int smp_cpu_get_polarization(int cpu); extern void smp_fill_possible_mask(void); @@ -50,7 +49,6 @@ static inline int smp_find_processor_id(u16 address) { return 0; } static inline int smp_store_status(int cpu) { return 0; } static inline int smp_vcpu_scheduled(int cpu) { return 1; } static inline void smp_yield_cpu(int cpu) { } -static inline void smp_yield(void) { } static inline void smp_fill_possible_mask(void) { } #endif /* CONFIG_SMP */ diff --git a/arch/s390/include/asm/spinlock.h b/arch/s390/include/asm/spinlock.h index 96879f7..d6bdf90 100644 --- a/arch/s390/include/asm/spinlock.h +++ b/arch/s390/include/asm/spinlock.h @@ -37,11 +37,17 @@ _raw_compare_and_swap(unsigned int *lock, unsigned int old, unsigned int new) * (the type definitions are in asm/spinlock_types.h) */ +void arch_lock_relax(unsigned int cpu); + void arch_spin_lock_wait(arch_spinlock_t *); int arch_spin_trylock_retry(arch_spinlock_t *); -void arch_spin_relax(arch_spinlock_t *); void arch_spin_lock_wait_flags(arch_spinlock_t *, unsigned long flags); +static inline void arch_spin_relax(arch_spinlock_t *lock) +{ + arch_lock_relax(lock->lock); +} + static inline u32 arch_spin_lockval(int cpu) { return ~cpu; @@ -64,11 +70,6 @@ static inline int arch_spin_trylock_once(arch_spinlock_t *lp) _raw_compare_and_swap(&lp->lock, 0, SPINLOCK_LOCKVAL)); } -static inline int arch_spin_tryrelease_once(arch_spinlock_t *lp) -{ - return _raw_compare_and_swap(&lp->lock, SPINLOCK_LOCKVAL, 0); -} - static inline void arch_spin_lock(arch_spinlock_t *lp) { if (!arch_spin_trylock_once(lp)) @@ -91,7 +92,13 @@ static inline int arch_spin_trylock(arch_spinlock_t *lp) static inline void arch_spin_unlock(arch_spinlock_t *lp) { - arch_spin_tryrelease_once(lp); + typecheck(unsigned int, lp->lock); + asm volatile( + __ASM_BARRIER + "st %1,%0\n" + : "+Q" (lp->lock) + : "d" (0) + : "cc", "memory"); } static inline void arch_spin_unlock_wait(arch_spinlock_t *lock) @@ -123,13 +130,12 @@ static inline void arch_spin_unlock_wait(arch_spinlock_t *lock) */ #define arch_write_can_lock(x) ((x)->lock == 0) -extern void _raw_read_lock_wait(arch_rwlock_t *lp); -extern void _raw_read_lock_wait_flags(arch_rwlock_t *lp, unsigned long flags); extern int _raw_read_trylock_retry(arch_rwlock_t *lp); -extern void _raw_write_lock_wait(arch_rwlock_t *lp); -extern void _raw_write_lock_wait_flags(arch_rwlock_t *lp, unsigned long flags); extern int _raw_write_trylock_retry(arch_rwlock_t *lp); +#define arch_read_lock_flags(lock, flags) arch_read_lock(lock) +#define arch_write_lock_flags(lock, flags) arch_write_lock(lock) + static inline int arch_read_trylock_once(arch_rwlock_t *rw) { unsigned int old = ACCESS_ONCE(rw->lock); @@ -144,16 +150,82 @@ static inline int arch_write_trylock_once(arch_rwlock_t *rw) _raw_compare_and_swap(&rw->lock, 0, 0x80000000)); } +#ifdef CONFIG_HAVE_MARCH_Z196_FEATURES + +#define __RAW_OP_OR "lao" +#define __RAW_OP_AND "lan" +#define __RAW_OP_ADD "laa" + +#define __RAW_LOCK(ptr, op_val, op_string) \ +({ \ + unsigned int old_val; \ + \ + typecheck(unsigned int *, ptr); \ + asm volatile( \ + op_string " %0,%2,%1\n" \ + "bcr 14,0\n" \ + : "=d" (old_val), "+Q" (*ptr) \ + : "d" (op_val) \ + : "cc", "memory"); \ + old_val; \ +}) + +#define __RAW_UNLOCK(ptr, op_val, op_string) \ +({ \ + unsigned int old_val; \ + \ + typecheck(unsigned int *, ptr); \ + asm volatile( \ + "bcr 14,0\n" \ + op_string " %0,%2,%1\n" \ + : "=d" (old_val), "+Q" (*ptr) \ + : "d" (op_val) \ + : "cc", "memory"); \ + old_val; \ +}) + +extern void _raw_read_lock_wait(arch_rwlock_t *lp); +extern void _raw_write_lock_wait(arch_rwlock_t *lp, unsigned int prev); + static inline void arch_read_lock(arch_rwlock_t *rw) { - if (!arch_read_trylock_once(rw)) + unsigned int old; + + old = __RAW_LOCK(&rw->lock, 1, __RAW_OP_ADD); + if ((int) old < 0) _raw_read_lock_wait(rw); } -static inline void arch_read_lock_flags(arch_rwlock_t *rw, unsigned long flags) +static inline void arch_read_unlock(arch_rwlock_t *rw) +{ + __RAW_UNLOCK(&rw->lock, -1, __RAW_OP_ADD); +} + +static inline void arch_write_lock(arch_rwlock_t *rw) +{ + unsigned int old; + + old = __RAW_LOCK(&rw->lock, 0x80000000, __RAW_OP_OR); + if (old != 0) + _raw_write_lock_wait(rw, old); + rw->owner = SPINLOCK_LOCKVAL; +} + +static inline void arch_write_unlock(arch_rwlock_t *rw) +{ + rw->owner = 0; + __RAW_UNLOCK(&rw->lock, 0x7fffffff, __RAW_OP_AND); +} + +#else /* CONFIG_HAVE_MARCH_Z196_FEATURES */ + +extern void _raw_read_lock_wait(arch_rwlock_t *lp); +extern void _raw_write_lock_wait(arch_rwlock_t *lp); + +static inline void arch_read_lock(arch_rwlock_t *rw) { if (!arch_read_trylock_once(rw)) - _raw_read_lock_wait_flags(rw, flags); + _raw_read_lock_wait(rw); } static inline void arch_read_unlock(arch_rwlock_t *rw) @@ -169,19 +241,24 @@ static inline void arch_write_lock(arch_rwlock_t *rw) { if (!arch_write_trylock_once(rw)) _raw_write_lock_wait(rw); -} - -static inline void arch_write_lock_flags(arch_rwlock_t *rw, unsigned long flags) -{ - if (!arch_write_trylock_once(rw)) - _raw_write_lock_wait_flags(rw, flags); + rw->owner = SPINLOCK_LOCKVAL; } static inline void arch_write_unlock(arch_rwlock_t *rw) { - _raw_compare_and_swap(&rw->lock, 0x80000000, 0); + typecheck(unsigned int, rw->lock); + + rw->owner = 0; + asm volatile( + __ASM_BARRIER + "st %1,%0\n" + : "+Q" (rw->lock) + : "d" (0) + : "cc", "memory"); } +#endif /* CONFIG_HAVE_MARCH_Z196_FEATURES */ + static inline int arch_read_trylock(arch_rwlock_t *rw) { if (!arch_read_trylock_once(rw)) @@ -191,12 +268,20 @@ static inline int arch_read_trylock(arch_rwlock_t *rw) static inline int arch_write_trylock(arch_rwlock_t *rw) { - if (!arch_write_trylock_once(rw)) - return _raw_write_trylock_retry(rw); + if (!arch_write_trylock_once(rw) && !_raw_write_trylock_retry(rw)) + return 0; + rw->owner = SPINLOCK_LOCKVAL; return 1; } -#define arch_read_relax(lock) cpu_relax() -#define arch_write_relax(lock) cpu_relax() +static inline void arch_read_relax(arch_rwlock_t *rw) +{ + arch_lock_relax(rw->owner); +} + +static inline void arch_write_relax(arch_rwlock_t *rw) +{ + arch_lock_relax(rw->owner); +} #endif /* __ASM_SPINLOCK_H */ diff --git a/arch/s390/include/asm/spinlock_types.h b/arch/s390/include/asm/spinlock_types.h index b2cd6ff..d84b693 100644 --- a/arch/s390/include/asm/spinlock_types.h +++ b/arch/s390/include/asm/spinlock_types.h @@ -13,6 +13,7 @@ typedef struct { typedef struct { unsigned int lock; + unsigned int owner; } arch_rwlock_t; #define __ARCH_RW_LOCK_UNLOCKED { 0 } diff --git a/arch/s390/include/asm/switch_to.h b/arch/s390/include/asm/switch_to.h index 18ea9e3..2542a7e 100644 --- a/arch/s390/include/asm/switch_to.h +++ b/arch/s390/include/asm/switch_to.h @@ -103,6 +103,61 @@ static inline void restore_fp_regs(freg_t *fprs) asm volatile("ld 15,%0" : : "Q" (fprs[15])); } +static inline void save_vx_regs(__vector128 *vxrs) +{ + typedef struct { __vector128 _[__NUM_VXRS]; } addrtype; + + asm volatile( + " la 1,%0\n" + " .word 0xe70f,0x1000,0x003e\n" /* vstm 0,15,0(1) */ + " .word 0xe70f,0x1100,0x0c3e\n" /* vstm 16,31,256(1) */ + : "=Q" (*(addrtype *) vxrs) : : "1"); +} + +static inline void save_vx_regs_safe(__vector128 *vxrs) +{ + unsigned long cr0, flags; + + flags = arch_local_irq_save(); + __ctl_store(cr0, 0, 0); + __ctl_set_bit(0, 17); + __ctl_set_bit(0, 18); + save_vx_regs(vxrs); + __ctl_load(cr0, 0, 0); + arch_local_irq_restore(flags); +} + +static inline void restore_vx_regs(__vector128 *vxrs) +{ + typedef struct { __vector128 _[__NUM_VXRS]; } addrtype; + + asm volatile( + " la 1,%0\n" + " .word 0xe70f,0x1000,0x0036\n" /* vlm 0,15,0(1) */ + " .word 0xe70f,0x1100,0x0c36\n" /* vlm 16,31,256(1) */ + : : "Q" (*(addrtype *) vxrs) : "1"); +} + +static inline void save_fp_vx_regs(struct task_struct *task) +{ +#ifdef CONFIG_64BIT + if (task->thread.vxrs) + save_vx_regs(task->thread.vxrs); + else +#endif + save_fp_regs(task->thread.fp_regs.fprs); +} + +static inline void restore_fp_vx_regs(struct task_struct *task) +{ +#ifdef CONFIG_64BIT + if (task->thread.vxrs) + restore_vx_regs(task->thread.vxrs); + else +#endif + restore_fp_regs(task->thread.fp_regs.fprs); +} + static inline void save_access_regs(unsigned int *acrs) { typedef struct { int _[NUM_ACRS]; } acrstype; @@ -120,16 +175,16 @@ static inline void restore_access_regs(unsigned int *acrs) #define switch_to(prev,next,last) do { \ if (prev->mm) { \ save_fp_ctl(&prev->thread.fp_regs.fpc); \ - save_fp_regs(prev->thread.fp_regs.fprs); \ + save_fp_vx_regs(prev); \ save_access_regs(&prev->thread.acrs[0]); \ save_ri_cb(prev->thread.ri_cb); \ } \ if (next->mm) { \ + update_cr_regs(next); \ restore_fp_ctl(&next->thread.fp_regs.fpc); \ - restore_fp_regs(next->thread.fp_regs.fprs); \ + restore_fp_vx_regs(next); \ restore_access_regs(&next->thread.acrs[0]); \ restore_ri_cb(next->thread.ri_cb, prev->thread.ri_cb); \ - update_cr_regs(next); \ } \ prev = __switch_to(prev,next); \ } while (0) diff --git a/arch/s390/include/asm/syscall.h b/arch/s390/include/asm/syscall.h index abad78d..5bc1259 100644 --- a/arch/s390/include/asm/syscall.h +++ b/arch/s390/include/asm/syscall.h @@ -54,7 +54,7 @@ static inline void syscall_set_return_value(struct task_struct *task, struct pt_regs *regs, int error, long val) { - regs->gprs[2] = error ? -error : val; + regs->gprs[2] = error ? error : val; } static inline void syscall_get_arguments(struct task_struct *task, diff --git a/arch/s390/include/asm/thread_info.h b/arch/s390/include/asm/thread_info.h index b833e9c..4d62fd5 100644 --- a/arch/s390/include/asm/thread_info.h +++ b/arch/s390/include/asm/thread_info.h @@ -84,11 +84,13 @@ static inline struct thread_info *current_thread_info(void) #define TIF_SYSCALL_AUDIT 4 /* syscall auditing active */ #define TIF_SECCOMP 5 /* secure computing */ #define TIF_SYSCALL_TRACEPOINT 6 /* syscall tracepoint instrumentation */ +#define TIF_UPROBE 7 /* breakpointed or single-stepping */ #define TIF_31BIT 16 /* 32bit process */ #define TIF_MEMDIE 17 /* is terminating due to OOM killer */ #define TIF_RESTORE_SIGMASK 18 /* restore signal mask in do_signal() */ #define TIF_SINGLE_STEP 19 /* This task is single stepped */ #define TIF_BLOCK_STEP 20 /* This task is block stepped */ +#define TIF_UPROBE_SINGLESTEP 21 /* This task is uprobe single stepped */ #define _TIF_NOTIFY_RESUME (1<<TIF_NOTIFY_RESUME) #define _TIF_SIGPENDING (1<<TIF_SIGPENDING) @@ -97,6 +99,7 @@ static inline struct thread_info *current_thread_info(void) #define _TIF_SYSCALL_AUDIT (1<<TIF_SYSCALL_AUDIT) #define _TIF_SECCOMP (1<<TIF_SECCOMP) #define _TIF_SYSCALL_TRACEPOINT (1<<TIF_SYSCALL_TRACEPOINT) +#define _TIF_UPROBE (1<<TIF_UPROBE) #define _TIF_31BIT (1<<TIF_31BIT) #define _TIF_SINGLE_STEP (1<<TIF_SINGLE_STEP) diff --git a/arch/s390/include/asm/tlb.h b/arch/s390/include/asm/tlb.h index a25f09f..572c599 100644 --- a/arch/s390/include/asm/tlb.h +++ b/arch/s390/include/asm/tlb.h @@ -105,7 +105,7 @@ static inline void tlb_remove_page(struct mmu_gather *tlb, struct page *page) static inline void pte_free_tlb(struct mmu_gather *tlb, pgtable_t pte, unsigned long address) { - page_table_free_rcu(tlb, (unsigned long *) pte); + page_table_free_rcu(tlb, (unsigned long *) pte, address); } /* diff --git a/arch/s390/include/asm/uprobes.h b/arch/s390/include/asm/uprobes.h new file mode 100644 index 0000000..1411dff --- /dev/null +++ b/arch/s390/include/asm/uprobes.h @@ -0,0 +1,42 @@ +/* + * User-space Probes (UProbes) for s390 + * + * Copyright IBM Corp. 2014 + * Author(s): Jan Willeke, + */ + +#ifndef _ASM_UPROBES_H +#define _ASM_UPROBES_H + +#include <linux/notifier.h> + +typedef u16 uprobe_opcode_t; + +#define UPROBE_XOL_SLOT_BYTES 256 /* cache aligned */ + +#define UPROBE_SWBP_INSN 0x0002 +#define UPROBE_SWBP_INSN_SIZE 2 + +struct arch_uprobe { + union{ + uprobe_opcode_t insn[3]; + uprobe_opcode_t ixol[3]; + }; + unsigned int saved_per : 1; + unsigned int saved_int_code; +}; + +struct arch_uprobe_task { +}; + +int arch_uprobe_analyze_insn(struct arch_uprobe *aup, struct mm_struct *mm, + unsigned long addr); +int arch_uprobe_pre_xol(struct arch_uprobe *aup, struct pt_regs *regs); +int arch_uprobe_post_xol(struct arch_uprobe *aup, struct pt_regs *regs); +bool arch_uprobe_xol_was_trapped(struct task_struct *tsk); +int arch_uprobe_exception_notify(struct notifier_block *self, unsigned long val, + void *data); +void arch_uprobe_abort_xol(struct arch_uprobe *ap, struct pt_regs *regs); +unsigned long arch_uretprobe_hijack_return_addr(unsigned long trampoline, + struct pt_regs *regs); +#endif /* _ASM_UPROBES_H */ diff --git a/arch/s390/include/asm/vdso.h b/arch/s390/include/asm/vdso.h index bc9746a..a62526d 100644 --- a/arch/s390/include/asm/vdso.h +++ b/arch/s390/include/asm/vdso.h @@ -22,13 +22,17 @@ struct vdso_data { __u64 xtime_tod_stamp; /* TOD clock for xtime 0x08 */ __u64 xtime_clock_sec; /* Kernel time 0x10 */ __u64 xtime_clock_nsec; /* 0x18 */ - __u64 wtom_clock_sec; /* Wall to monotonic clock 0x20 */ - __u64 wtom_clock_nsec; /* 0x28 */ - __u32 tz_minuteswest; /* Minutes west of Greenwich 0x30 */ - __u32 tz_dsttime; /* Type of dst correction 0x34 */ - __u32 ectg_available; /* ECTG instruction present 0x38 */ - __u32 tk_mult; /* Mult. used for xtime_nsec 0x3c */ - __u32 tk_shift; /* Shift used for xtime_nsec 0x40 */ + __u64 xtime_coarse_sec; /* Coarse kernel time 0x20 */ + __u64 xtime_coarse_nsec; /* 0x28 */ + __u64 wtom_clock_sec; /* Wall to monotonic clock 0x30 */ + __u64 wtom_clock_nsec; /* 0x38 */ + __u64 wtom_coarse_sec; /* Coarse wall to monotonic 0x40 */ + __u64 wtom_coarse_nsec; /* 0x48 */ + __u32 tz_minuteswest; /* Minutes west of Greenwich 0x50 */ + __u32 tz_dsttime; /* Type of dst correction 0x54 */ + __u32 ectg_available; /* ECTG instruction present 0x58 */ + __u32 tk_mult; /* Mult. used for xtime_nsec 0x5c */ + __u32 tk_shift; /* Shift used for xtime_nsec 0x60 */ }; struct vdso_per_cpu_data { diff --git a/arch/s390/include/asm/vtimer.h b/arch/s390/include/asm/vtimer.h index bfe25d5..10a179a 100644 --- a/arch/s390/include/asm/vtimer.h +++ b/arch/s390/include/asm/vtimer.h @@ -28,6 +28,4 @@ extern int del_virt_timer(struct vtimer_list *timer); extern void init_cpu_vtimer(void); extern void vtime_init(void); -extern void vtime_stop_cpu(void); - #endif /* _ASM_S390_TIMER_H */ diff --git a/arch/s390/include/uapi/asm/Kbuild b/arch/s390/include/uapi/asm/Kbuild index 7366373..08fe6da 100644 --- a/arch/s390/include/uapi/asm/Kbuild +++ b/arch/s390/include/uapi/asm/Kbuild @@ -16,6 +16,7 @@ header-y += ioctls.h header-y += ipcbuf.h header-y += kvm.h header-y += kvm_para.h +header-y += kvm_perf.h header-y += kvm_virtio.h header-y += mman.h header-y += monwriter.h diff --git a/arch/s390/include/uapi/asm/kvm.h b/arch/s390/include/uapi/asm/kvm.h index 0fc2643..48eda3a 100644 --- a/arch/s390/include/uapi/asm/kvm.h +++ b/arch/s390/include/uapi/asm/kvm.h @@ -111,12 +111,22 @@ struct kvm_guest_debug_arch { #define KVM_SYNC_GPRS (1UL << 1) #define KVM_SYNC_ACRS (1UL << 2) #define KVM_SYNC_CRS (1UL << 3) +#define KVM_SYNC_ARCH0 (1UL << 4) +#define KVM_SYNC_PFAULT (1UL << 5) /* definition of registers in kvm_run */ struct kvm_sync_regs { __u64 prefix; /* prefix register */ __u64 gprs[16]; /* general purpose registers */ __u32 acrs[16]; /* access registers */ __u64 crs[16]; /* control registers */ + __u64 todpr; /* tod programmable register [ARCH0] */ + __u64 cputm; /* cpu timer [ARCH0] */ + __u64 ckc; /* clock comparator [ARCH0] */ + __u64 pp; /* program parameter [ARCH0] */ + __u64 gbea; /* guest breaking-event address [ARCH0] */ + __u64 pft; /* pfault token [PFAULT] */ + __u64 pfs; /* pfault select [PFAULT] */ + __u64 pfc; /* pfault compare [PFAULT] */ }; #define KVM_REG_S390_TODPR (KVM_REG_S390 | KVM_REG_SIZE_U32 | 0x1) diff --git a/arch/s390/include/uapi/asm/kvm_perf.h b/arch/s390/include/uapi/asm/kvm_perf.h new file mode 100644 index 0000000..3972827 --- /dev/null +++ b/arch/s390/include/uapi/asm/kvm_perf.h @@ -0,0 +1,25 @@ +/* + * Definitions for perf-kvm on s390 + * + * Copyright 2014 IBM Corp. + * Author(s): Alexander Yarygin <yarygin@linux.vnet.ibm.com> + * + * 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. + */ + +#ifndef __LINUX_KVM_PERF_S390_H +#define __LINUX_KVM_PERF_S390_H + +#include <asm/sie.h> + +#define DECODE_STR_LEN 40 + +#define VCPU_ID "id" + +#define KVM_ENTRY_TRACE "kvm:kvm_s390_sie_enter" +#define KVM_EXIT_TRACE "kvm:kvm_s390_sie_exit" +#define KVM_EXIT_REASON "icptcode" + +#endif diff --git a/arch/s390/include/uapi/asm/sie.h b/arch/s390/include/uapi/asm/sie.h index 5d9cc19..d4096fd 100644 --- a/arch/s390/include/uapi/asm/sie.h +++ b/arch/s390/include/uapi/asm/sie.h @@ -108,6 +108,7 @@ exit_code_ipa0(0xB2, 0x17, "STETR"), \ exit_code_ipa0(0xB2, 0x18, "PC"), \ exit_code_ipa0(0xB2, 0x20, "SERVC"), \ + exit_code_ipa0(0xB2, 0x21, "IPTE"), \ exit_code_ipa0(0xB2, 0x28, "PT"), \ exit_code_ipa0(0xB2, 0x29, "ISKE"), \ exit_code_ipa0(0xB2, 0x2a, "RRBE"), \ diff --git a/arch/s390/include/uapi/asm/sigcontext.h b/arch/s390/include/uapi/asm/sigcontext.h index b30de9c..5f0b8d7 100644 --- a/arch/s390/include/uapi/asm/sigcontext.h +++ b/arch/s390/include/uapi/asm/sigcontext.h @@ -7,10 +7,14 @@ #define _ASM_S390_SIGCONTEXT_H #include <linux/compiler.h> +#include <linux/types.h> -#define __NUM_GPRS 16 -#define __NUM_FPRS 16 -#define __NUM_ACRS 16 +#define __NUM_GPRS 16 +#define __NUM_FPRS 16 +#define __NUM_ACRS 16 +#define __NUM_VXRS 32 +#define __NUM_VXRS_LOW 16 +#define __NUM_VXRS_HIGH 16 #ifndef __s390x__ @@ -59,6 +63,16 @@ typedef struct _s390_fp_regs fpregs; } _sigregs; +typedef struct +{ +#ifndef __s390x__ + unsigned long gprs_high[__NUM_GPRS]; +#endif + unsigned long long vxrs_low[__NUM_VXRS_LOW]; + __vector128 vxrs_high[__NUM_VXRS_HIGH]; + unsigned char __reserved[128]; +} _sigregs_ext; + struct sigcontext { unsigned long oldmask[_SIGCONTEXT_NSIG_WORDS]; diff --git a/arch/s390/include/uapi/asm/types.h b/arch/s390/include/uapi/asm/types.h index 038f2b9..3c3951e 100644 --- a/arch/s390/include/uapi/asm/types.h +++ b/arch/s390/include/uapi/asm/types.h @@ -17,6 +17,10 @@ typedef unsigned long addr_t; typedef __signed__ long saddr_t; +typedef struct { + __u32 u[4]; +} __vector128; + #endif /* __ASSEMBLY__ */ #endif /* _UAPI_S390_TYPES_H */ diff --git a/arch/s390/include/uapi/asm/ucontext.h b/arch/s390/include/uapi/asm/ucontext.h index 3e077b2..64a69aa 100644 --- a/arch/s390/include/uapi/asm/ucontext.h +++ b/arch/s390/include/uapi/asm/ucontext.h @@ -7,10 +7,15 @@ #ifndef _ASM_S390_UCONTEXT_H #define _ASM_S390_UCONTEXT_H -#define UC_EXTENDED 0x00000001 - -#ifndef __s390x__ +#define UC_GPRS_HIGH 1 /* uc_mcontext_ext has valid high gprs */ +#define UC_VXRS 2 /* uc_mcontext_ext has valid vector regs */ +/* + * The struct ucontext_extended describes how the registers are stored + * on a rt signal frame. Please note that the structure is not fixed, + * if new CPU registers are added to the user state the size of the + * struct ucontext_extended will increase. + */ struct ucontext_extended { unsigned long uc_flags; struct ucontext *uc_link; @@ -19,11 +24,9 @@ struct ucontext_extended { sigset_t uc_sigmask; /* Allow for uc_sigmask growth. Glibc uses a 1024-bit sigset_t. */ unsigned char __unused[128 - sizeof(sigset_t)]; - unsigned long uc_gprs_high[16]; + _sigregs_ext uc_mcontext_ext; }; -#endif - struct ucontext { unsigned long uc_flags; struct ucontext *uc_link; diff --git a/arch/s390/include/uapi/asm/unistd.h b/arch/s390/include/uapi/asm/unistd.h index 3802d2d..940ac49 100644 --- a/arch/s390/include/uapi/asm/unistd.h +++ b/arch/s390/include/uapi/asm/unistd.h @@ -283,7 +283,10 @@ #define __NR_sched_setattr 345 #define __NR_sched_getattr 346 #define __NR_renameat2 347 -#define NR_syscalls 348 +#define __NR_seccomp 348 +#define __NR_getrandom 349 +#define __NR_memfd_create 350 +#define NR_syscalls 351 /* * There are some system calls that are not present on 64 bit, some diff --git a/arch/s390/kernel/Makefile b/arch/s390/kernel/Makefile index a95c4ca..204c43a 100644 --- a/arch/s390/kernel/Makefile +++ b/arch/s390/kernel/Makefile @@ -28,7 +28,7 @@ CFLAGS_ptrace.o += -DUTS_MACHINE='"$(UTS_MACHINE)"' CFLAGS_sysinfo.o += -Iinclude/math-emu -Iarch/s390/math-emu -w -obj-y := traps.o time.o process.o base.o early.o setup.o vtime.o +obj-y := traps.o time.o process.o base.o early.o setup.o idle.o vtime.o obj-y += processor.o sys_s390.o ptrace.o signal.o cpcmd.o ebcdic.o nmi.o obj-y += debug.o irq.o ipl.o dis.o diag.o sclp.o vdso.o obj-y += sysinfo.o jump_label.o lgr.o os_info.o machine_kexec.o pgm_check.o @@ -52,11 +52,9 @@ obj-$(CONFIG_COMPAT) += compat_wrapper.o $(compat-obj-y) obj-$(CONFIG_STACKTRACE) += stacktrace.o obj-$(CONFIG_KPROBES) += kprobes.o -obj-$(CONFIG_FUNCTION_TRACER) += $(if $(CONFIG_64BIT),mcount64.o,mcount.o) -obj-$(CONFIG_DYNAMIC_FTRACE) += ftrace.o -obj-$(CONFIG_FUNCTION_GRAPH_TRACER) += ftrace.o -obj-$(CONFIG_FTRACE_SYSCALLS) += ftrace.o +obj-$(CONFIG_FUNCTION_TRACER) += mcount.o ftrace.o obj-$(CONFIG_CRASH_DUMP) += crash_dump.o +obj-$(CONFIG_UPROBES) += uprobes.o ifdef CONFIG_64BIT obj-$(CONFIG_PERF_EVENTS) += perf_event.o perf_cpum_cf.o perf_cpum_sf.o \ diff --git a/arch/s390/kernel/asm-offsets.c b/arch/s390/kernel/asm-offsets.c index afe1715..ef279a1 100644 --- a/arch/s390/kernel/asm-offsets.c +++ b/arch/s390/kernel/asm-offsets.c @@ -9,7 +9,7 @@ #include <linux/kbuild.h> #include <linux/kvm_host.h> #include <linux/sched.h> -#include <asm/cputime.h> +#include <asm/idle.h> #include <asm/vdso.h> #include <asm/pgtable.h> @@ -62,8 +62,12 @@ int main(void) DEFINE(__VDSO_XTIME_STAMP, offsetof(struct vdso_data, xtime_tod_stamp)); DEFINE(__VDSO_XTIME_SEC, offsetof(struct vdso_data, xtime_clock_sec)); DEFINE(__VDSO_XTIME_NSEC, offsetof(struct vdso_data, xtime_clock_nsec)); + DEFINE(__VDSO_XTIME_CRS_SEC, offsetof(struct vdso_data, xtime_coarse_sec)); + DEFINE(__VDSO_XTIME_CRS_NSEC, offsetof(struct vdso_data, xtime_coarse_nsec)); DEFINE(__VDSO_WTOM_SEC, offsetof(struct vdso_data, wtom_clock_sec)); DEFINE(__VDSO_WTOM_NSEC, offsetof(struct vdso_data, wtom_clock_nsec)); + DEFINE(__VDSO_WTOM_CRS_SEC, offsetof(struct vdso_data, wtom_coarse_sec)); + DEFINE(__VDSO_WTOM_CRS_NSEC, offsetof(struct vdso_data, wtom_coarse_nsec)); DEFINE(__VDSO_TIMEZONE, offsetof(struct vdso_data, tz_minuteswest)); DEFINE(__VDSO_ECTG_OK, offsetof(struct vdso_data, ectg_available)); DEFINE(__VDSO_TK_MULT, offsetof(struct vdso_data, tk_mult)); @@ -73,8 +77,11 @@ int main(void) /* constants used by the vdso */ DEFINE(__CLOCK_REALTIME, CLOCK_REALTIME); DEFINE(__CLOCK_MONOTONIC, CLOCK_MONOTONIC); + DEFINE(__CLOCK_REALTIME_COARSE, CLOCK_REALTIME_COARSE); + DEFINE(__CLOCK_MONOTONIC_COARSE, CLOCK_MONOTONIC_COARSE); DEFINE(__CLOCK_THREAD_CPUTIME_ID, CLOCK_THREAD_CPUTIME_ID); DEFINE(__CLOCK_REALTIME_RES, MONOTONIC_RES_NSEC); + DEFINE(__CLOCK_COARSE_RES, LOW_RES_NSEC); BLANK(); /* idle data offsets */ DEFINE(__CLOCK_IDLE_ENTER, offsetof(struct s390_idle_data, clock_idle_enter)); diff --git a/arch/s390/kernel/compat_linux.h b/arch/s390/kernel/compat_linux.h index 70d4b7c..a0a886c 100644 --- a/arch/s390/kernel/compat_linux.h +++ b/arch/s390/kernel/compat_linux.h @@ -50,6 +50,14 @@ typedef struct _s390_fp_regs32 fpregs; } _sigregs32; +typedef struct +{ + __u32 gprs_high[__NUM_GPRS]; + __u64 vxrs_low[__NUM_VXRS_LOW]; + __vector128 vxrs_high[__NUM_VXRS_HIGH]; + __u8 __reserved[128]; +} _sigregs_ext32; + #define _SIGCONTEXT_NSIG32 64 #define _SIGCONTEXT_NSIG_BPW32 32 #define __SIGNAL_FRAMESIZE32 96 @@ -72,6 +80,7 @@ struct ucontext32 { compat_sigset_t uc_sigmask; /* Allow for uc_sigmask growth. Glibc uses a 1024-bit sigset_t. */ unsigned char __unused[128 - sizeof(compat_sigset_t)]; + _sigregs_ext32 uc_mcontext_ext; }; struct stat64_emu31; diff --git a/arch/s390/kernel/compat_signal.c b/arch/s390/kernel/compat_signal.c index f204d69..009f5eb 100644 --- a/arch/s390/kernel/compat_signal.c +++ b/arch/s390/kernel/compat_signal.c @@ -36,17 +36,16 @@ typedef struct struct sigcontext32 sc; _sigregs32 sregs; int signo; - __u32 gprs_high[NUM_GPRS]; - __u8 retcode[S390_SYSCALL_SIZE]; + _sigregs_ext32 sregs_ext; + __u16 svc_insn; /* Offset of svc_insn is NOT fixed! */ } sigframe32; typedef struct { __u8 callee_used_stack[__SIGNAL_FRAMESIZE32]; - __u8 retcode[S390_SYSCALL_SIZE]; + __u16 svc_insn; compat_siginfo_t info; struct ucontext32 uc; - __u32 gprs_high[NUM_GPRS]; } rt_sigframe32; int copy_siginfo_to_user32(compat_siginfo_t __user *to, const siginfo_t *from) @@ -151,6 +150,38 @@ int copy_siginfo_from_user32(siginfo_t *to, compat_siginfo_t __user *from) return err ? -EFAULT : 0; } +/* Store registers needed to create the signal frame */ +static void store_sigregs(void) +{ + int i; + + save_access_regs(current->thread.acrs); + save_fp_ctl(¤t->thread.fp_regs.fpc); + if (current->thread.vxrs) { + save_vx_regs(current->thread.vxrs); + for (i = 0; i < __NUM_FPRS; i++) + current->thread.fp_regs.fprs[i] = + *(freg_t *)(current->thread.vxrs + i); + } else + save_fp_regs(current->thread.fp_regs.fprs); +} + +/* Load registers after signal return */ +static void load_sigregs(void) +{ + int i; + + restore_access_regs(current->thread.acrs); + /* restore_fp_ctl is done in restore_sigregs */ + if (current->thread.vxrs) { + for (i = 0; i < __NUM_FPRS; i++) + *(freg_t *)(current->thread.vxrs + i) = + current->thread.fp_regs.fprs[i]; + restore_vx_regs(current->thread.vxrs); + } else + restore_fp_regs(current->thread.fp_regs.fprs); +} + static int save_sigregs32(struct pt_regs *regs, _sigregs32 __user *sregs) { _sigregs32 user_sregs; @@ -163,11 +194,8 @@ static int save_sigregs32(struct pt_regs *regs, _sigregs32 __user *sregs) (__u32)(regs->psw.mask & PSW_MASK_BA); for (i = 0; i < NUM_GPRS; i++) user_sregs.regs.gprs[i] = (__u32) regs->gprs[i]; - save_access_regs(current->thread.acrs); memcpy(&user_sregs.regs.acrs, current->thread.acrs, sizeof(user_sregs.regs.acrs)); - save_fp_ctl(¤t->thread.fp_regs.fpc); - save_fp_regs(current->thread.fp_regs.fprs); memcpy(&user_sregs.fpregs, ¤t->thread.fp_regs, sizeof(user_sregs.fpregs)); if (__copy_to_user(sregs, &user_sregs, sizeof(_sigregs32))) @@ -207,37 +235,67 @@ static int restore_sigregs32(struct pt_regs *regs,_sigregs32 __user *sregs) regs->gprs[i] = (__u64) user_sregs.regs.gprs[i]; memcpy(¤t->thread.acrs, &user_sregs.regs.acrs, sizeof(current->thread.acrs)); - restore_access_regs(current->thread.acrs); memcpy(¤t->thread.fp_regs, &user_sregs.fpregs, sizeof(current->thread.fp_regs)); - restore_fp_regs(current->thread.fp_regs.fprs); clear_pt_regs_flag(regs, PIF_SYSCALL); /* No longer in a system call */ return 0; } -static int save_sigregs_gprs_high(struct pt_regs *regs, __u32 __user *uregs) +static int save_sigregs_ext32(struct pt_regs *regs, + _sigregs_ext32 __user *sregs_ext) { __u32 gprs_high[NUM_GPRS]; + __u64 vxrs[__NUM_VXRS_LOW]; int i; + /* Save high gprs to signal stack */ for (i = 0; i < NUM_GPRS; i++) gprs_high[i] = regs->gprs[i] >> 32; - if (__copy_to_user(uregs, &gprs_high, sizeof(gprs_high))) + if (__copy_to_user(&sregs_ext->gprs_high, &gprs_high, + sizeof(sregs_ext->gprs_high))) return -EFAULT; + + /* Save vector registers to signal stack */ + if (current->thread.vxrs) { + for (i = 0; i < __NUM_VXRS_LOW; i++) + vxrs[i] = *((__u64 *)(current->thread.vxrs + i) + 1); + if (__copy_to_user(&sregs_ext->vxrs_low, vxrs, + sizeof(sregs_ext->vxrs_low)) || + __copy_to_user(&sregs_ext->vxrs_high, + current->thread.vxrs + __NUM_VXRS_LOW, + sizeof(sregs_ext->vxrs_high))) + return -EFAULT; + } return 0; } -static int restore_sigregs_gprs_high(struct pt_regs *regs, __u32 __user *uregs) +static int restore_sigregs_ext32(struct pt_regs *regs, + _sigregs_ext32 __user *sregs_ext) { __u32 gprs_high[NUM_GPRS]; + __u64 vxrs[__NUM_VXRS_LOW]; int i; - if (__copy_from_user(&gprs_high, uregs, sizeof(gprs_high))) + /* Restore high gprs from signal stack */ + if (__copy_from_user(&gprs_high, &sregs_ext->gprs_high, + sizeof(&sregs_ext->gprs_high))) return -EFAULT; for (i = 0; i < NUM_GPRS; i++) *(__u32 *)®s->gprs[i] = gprs_high[i]; + + /* Restore vector registers from signal stack */ + if (current->thread.vxrs) { + if (__copy_from_user(vxrs, &sregs_ext->vxrs_low, + sizeof(sregs_ext->vxrs_low)) || + __copy_from_user(current->thread.vxrs + __NUM_VXRS_LOW, + &sregs_ext->vxrs_high, + sizeof(sregs_ext->vxrs_high))) + return -EFAULT; + for (i = 0; i < __NUM_VXRS_LOW; i++) + *((__u64 *)(current->thread.vxrs + i) + 1) = vxrs[i]; + } return 0; } @@ -252,8 +310,9 @@ COMPAT_SYSCALL_DEFINE0(sigreturn) set_current_blocked(&set); if (restore_sigregs32(regs, &frame->sregs)) goto badframe; - if (restore_sigregs_gprs_high(regs, frame->gprs_high)) + if (restore_sigregs_ext32(regs, &frame->sregs_ext)) goto badframe; + load_sigregs(); return regs->gprs[2]; badframe: force_sig(SIGSEGV, current); @@ -269,12 +328,13 @@ COMPAT_SYSCALL_DEFINE0(rt_sigreturn) if (__copy_from_user(&set, &frame->uc.uc_sigmask, sizeof(set))) goto badframe; set_current_blocked(&set); + if (compat_restore_altstack(&frame->uc.uc_stack)) + goto badframe; if (restore_sigregs32(regs, &frame->uc.uc_mcontext)) goto badframe; - if (restore_sigregs_gprs_high(regs, frame->gprs_high)) + if (restore_sigregs_ext32(regs, &frame->uc.uc_mcontext_ext)) goto badframe; - if (compat_restore_altstack(&frame->uc.uc_stack)) - goto badframe; + load_sigregs(); return regs->gprs[2]; badframe: force_sig(SIGSEGV, current); @@ -320,46 +380,74 @@ static inline int map_signal(int sig) return sig; } -static int setup_frame32(int sig, struct k_sigaction *ka, - sigset_t *set, struct pt_regs * regs) +static int setup_frame32(struct ksignal *ksig, sigset_t *set, + struct pt_regs *regs) { - sigframe32 __user *frame = get_sigframe(ka, regs, sizeof(sigframe32)); - + int sig = ksig->sig; + sigframe32 __user *frame; + struct sigcontext32 sc; + unsigned long restorer; + size_t frame_size; + + /* + * gprs_high are always present for 31-bit compat tasks. + * The space for vector registers is only allocated if + * the machine supports it + */ + frame_size = sizeof(*frame) - sizeof(frame->sregs_ext.__reserved); + if (!MACHINE_HAS_VX) + frame_size -= sizeof(frame->sregs_ext.vxrs_low) + + sizeof(frame->sregs_ext.vxrs_high); + frame = get_sigframe(&ksig->ka, regs, frame_size); if (frame == (void __user *) -1UL) - goto give_sigsegv; + return -EFAULT; + + /* Set up backchain. */ + if (__put_user(regs->gprs[15], (unsigned int __user *) frame)) + return -EFAULT; + + /* Create struct sigcontext32 on the signal stack */ + memcpy(&sc.oldmask, &set->sig, _SIGMASK_COPY_SIZE32); + sc.sregs = (__u32)(unsigned long __force) &frame->sregs; + if (__copy_to_user(&frame->sc, &sc, sizeof(frame->sc))) + return -EFAULT; - if (__copy_to_user(&frame->sc.oldmask, &set->sig, _SIGMASK_COPY_SIZE32)) - goto give_sigsegv; + /* Store registers needed to create the signal frame */ + store_sigregs(); + /* Create _sigregs32 on the signal stack */ if (save_sigregs32(regs, &frame->sregs)) - goto give_sigsegv; - if (save_sigregs_gprs_high(regs, frame->gprs_high)) - goto give_sigsegv; - if (__put_user((unsigned long) &frame->sregs, &frame->sc.sregs)) - goto give_sigsegv; + return -EFAULT; + + /* Place signal number on stack to allow backtrace from handler. */ + if (__put_user(regs->gprs[2], (int __force __user *) &frame->signo)) + return -EFAULT; + + /* Create _sigregs_ext32 on the signal stack */ + if (save_sigregs_ext32(regs, &frame->sregs_ext)) + return -EFAULT; /* Set up to return from userspace. If provided, use a stub already in userspace. */ - if (ka->sa.sa_flags & SA_RESTORER) { - regs->gprs[14] = (__u64 __force) ka->sa.sa_restorer | PSW32_ADDR_AMODE; + if (ksig->ka.sa.sa_flags & SA_RESTORER) { + restorer = (unsigned long __force) + ksig->ka.sa.sa_restorer | PSW32_ADDR_AMODE; } else { - regs->gprs[14] = (__u64 __force) frame->retcode | PSW32_ADDR_AMODE; - if (__put_user(S390_SYSCALL_OPCODE | __NR_sigreturn, - (u16 __force __user *)(frame->retcode))) - goto give_sigsegv; + /* Signal frames without vectors registers are short ! */ + __u16 __user *svc = (void *) frame + frame_size - 2; + if (__put_user(S390_SYSCALL_OPCODE | __NR_sigreturn, svc)) + return -EFAULT; + restorer = (unsigned long __force) svc | PSW32_ADDR_AMODE; } - /* Set up backchain. */ - if (__put_user(regs->gprs[15], (unsigned int __user *) frame)) - goto give_sigsegv; - /* Set up registers for signal handler */ + regs->gprs[14] = restorer; regs->gprs[15] = (__force __u64) frame; /* Force 31 bit amode and default user address space control. */ regs->psw.mask = PSW_MASK_BA | (PSW_USER_BITS & PSW_MASK_ASC) | (regs->psw.mask & ~PSW_MASK_ASC); - regs->psw.addr = (__force __u64) ka->sa.sa_handler; + regs->psw.addr = (__force __u64) ksig->ka.sa.sa_handler; regs->gprs[2] = map_signal(sig); regs->gprs[3] = (__force __u64) &frame->sc; @@ -374,89 +462,98 @@ static int setup_frame32(int sig, struct k_sigaction *ka, regs->gprs[6] = task_thread_info(current)->last_break; } - /* Place signal number on stack to allow backtrace from handler. */ - if (__put_user(regs->gprs[2], (int __force __user *) &frame->signo)) - goto give_sigsegv; return 0; - -give_sigsegv: - force_sigsegv(sig, current); - return -EFAULT; } -static int setup_rt_frame32(int sig, struct k_sigaction *ka, siginfo_t *info, - sigset_t *set, struct pt_regs * regs) +static int setup_rt_frame32(struct ksignal *ksig, sigset_t *set, + struct pt_regs *regs) { - int err = 0; - rt_sigframe32 __user *frame = get_sigframe(ka, regs, sizeof(rt_sigframe32)); - + rt_sigframe32 __user *frame; + unsigned long restorer; + size_t frame_size; + u32 uc_flags; + + frame_size = sizeof(*frame) - + sizeof(frame->uc.uc_mcontext_ext.__reserved); + /* + * gprs_high are always present for 31-bit compat tasks. + * The space for vector registers is only allocated if + * the machine supports it + */ + uc_flags = UC_GPRS_HIGH; + if (MACHINE_HAS_VX) { + if (current->thread.vxrs) + uc_flags |= UC_VXRS; + } else + frame_size -= sizeof(frame->uc.uc_mcontext_ext.vxrs_low) + + sizeof(frame->uc.uc_mcontext_ext.vxrs_high); + frame = get_sigframe(&ksig->ka, regs, frame_size); if (frame == (void __user *) -1UL) - goto give_sigsegv; - - if (copy_siginfo_to_user32(&frame->info, info)) - goto give_sigsegv; + return -EFAULT; - /* Create the ucontext. */ - err |= __put_user(UC_EXTENDED, &frame->uc.uc_flags); - err |= __put_user(0, &frame->uc.uc_link); - err |= __compat_save_altstack(&frame->uc.uc_stack, regs->gprs[15]); - err |= save_sigregs32(regs, &frame->uc.uc_mcontext); - err |= save_sigregs_gprs_high(regs, frame->gprs_high); - err |= __copy_to_user(&frame->uc.uc_sigmask, set, sizeof(*set)); - if (err) - goto give_sigsegv; + /* Set up backchain. */ + if (__put_user(regs->gprs[15], (unsigned int __force __user *) frame)) + return -EFAULT; /* Set up to return from userspace. If provided, use a stub already in userspace. */ - if (ka->sa.sa_flags & SA_RESTORER) { - regs->gprs[14] = (__u64 __force) ka->sa.sa_restorer | PSW32_ADDR_AMODE; + if (ksig->ka.sa.sa_flags & SA_RESTORER) { + restorer = (unsigned long __force) + ksig->ka.sa.sa_restorer | PSW32_ADDR_AMODE; } else { - regs->gprs[14] = (__u64 __force) frame->retcode | PSW32_ADDR_AMODE; - if (__put_user(S390_SYSCALL_OPCODE | __NR_rt_sigreturn, - (u16 __force __user *)(frame->retcode))) - goto give_sigsegv; + __u16 __user *svc = &frame->svc_insn; + if (__put_user(S390_SYSCALL_OPCODE | __NR_rt_sigreturn, svc)) + return -EFAULT; + restorer = (unsigned long __force) svc | PSW32_ADDR_AMODE; } - /* Set up backchain. */ - if (__put_user(regs->gprs[15], (unsigned int __force __user *) frame)) - goto give_sigsegv; + /* Create siginfo on the signal stack */ + if (copy_siginfo_to_user32(&frame->info, &ksig->info)) + return -EFAULT; + + /* Store registers needed to create the signal frame */ + store_sigregs(); + + /* Create ucontext on the signal stack. */ + if (__put_user(uc_flags, &frame->uc.uc_flags) || + __put_user(0, &frame->uc.uc_link) || + __compat_save_altstack(&frame->uc.uc_stack, regs->gprs[15]) || + save_sigregs32(regs, &frame->uc.uc_mcontext) || + __copy_to_user(&frame->uc.uc_sigmask, set, sizeof(*set)) || + save_sigregs_ext32(regs, &frame->uc.uc_mcontext_ext)) + return -EFAULT; /* Set up registers for signal handler */ + regs->gprs[14] = restorer; regs->gprs[15] = (__force __u64) frame; /* Force 31 bit amode and default user address space control. */ regs->psw.mask = PSW_MASK_BA | (PSW_USER_BITS & PSW_MASK_ASC) | (regs->psw.mask & ~PSW_MASK_ASC); - regs->psw.addr = (__u64 __force) ka->sa.sa_handler; + regs->psw.addr = (__u64 __force) ksig->ka.sa.sa_handler; - regs->gprs[2] = map_signal(sig); + regs->gprs[2] = map_signal(ksig->sig); regs->gprs[3] = (__force __u64) &frame->info; regs->gprs[4] = (__force __u64) &frame->uc; regs->gprs[5] = task_thread_info(current)->last_break; return 0; - -give_sigsegv: - force_sigsegv(sig, current); - return -EFAULT; } /* * OK, we're invoking a handler */ -void handle_signal32(unsigned long sig, struct k_sigaction *ka, - siginfo_t *info, sigset_t *oldset, struct pt_regs *regs) +void handle_signal32(struct ksignal *ksig, sigset_t *oldset, + struct pt_regs *regs) { int ret; /* Set up the stack frame */ - if (ka->sa.sa_flags & SA_SIGINFO) - ret = setup_rt_frame32(sig, ka, info, oldset, regs); + if (ksig->ka.sa.sa_flags & SA_SIGINFO) + ret = setup_rt_frame32(ksig, oldset, regs); else - ret = setup_frame32(sig, ka, oldset, regs); - if (ret) - return; - signal_delivered(sig, info, ka, regs, - test_thread_flag(TIF_SINGLE_STEP)); + ret = setup_frame32(ksig, oldset, regs); + + signal_setup_done(ret, ksig, test_thread_flag(TIF_SINGLE_STEP)); } diff --git a/arch/s390/kernel/compat_wrapper.c b/arch/s390/kernel/compat_wrapper.c index 45cdb37..faf6caa 100644 --- a/arch/s390/kernel/compat_wrapper.c +++ b/arch/s390/kernel/compat_wrapper.c @@ -214,3 +214,6 @@ COMPAT_SYSCALL_WRAP3(finit_module, int, fd, const char __user *, uargs, int, fla COMPAT_SYSCALL_WRAP3(sched_setattr, pid_t, pid, struct sched_attr __user *, attr, unsigned int, flags); COMPAT_SYSCALL_WRAP4(sched_getattr, pid_t, pid, struct sched_attr __user *, attr, unsigned int, size, unsigned int, flags); COMPAT_SYSCALL_WRAP5(renameat2, int, olddfd, const char __user *, oldname, int, newdfd, const char __user *, newname, unsigned int, flags); +COMPAT_SYSCALL_WRAP3(seccomp, unsigned int, op, unsigned int, flags, const char __user *, uargs) +COMPAT_SYSCALL_WRAP3(getrandom, char __user *, buf, size_t, count, unsigned int, flags) +COMPAT_SYSCALL_WRAP2(memfd_create, const char __user *, uname, unsigned int, flags) diff --git a/arch/s390/kernel/crash_dump.c b/arch/s390/kernel/crash_dump.c index a3b9150..9f73c80 100644 --- a/arch/s390/kernel/crash_dump.c +++ b/arch/s390/kernel/crash_dump.c @@ -46,9 +46,9 @@ struct dump_save_areas dump_save_areas; /* * Allocate and add a save area for a CPU */ -struct save_area *dump_save_area_create(int cpu) +struct save_area_ext *dump_save_area_create(int cpu) { - struct save_area **save_areas, *save_area; + struct save_area_ext **save_areas, *save_area; save_area = kmalloc(sizeof(*save_area), GFP_KERNEL); if (!save_area) @@ -386,9 +386,45 @@ static void *nt_s390_prefix(void *ptr, struct save_area *sa) } /* + * Initialize vxrs high note (full 128 bit VX registers 16-31) + */ +static void *nt_s390_vx_high(void *ptr, __vector128 *vx_regs) +{ + return nt_init(ptr, NT_S390_VXRS_HIGH, &vx_regs[16], + 16 * sizeof(__vector128), KEXEC_CORE_NOTE_NAME); +} + +/* + * Initialize vxrs low note (lower halves of VX registers 0-15) + */ +static void *nt_s390_vx_low(void *ptr, __vector128 *vx_regs) +{ + Elf64_Nhdr *note; + u64 len; + int i; + + note = (Elf64_Nhdr *)ptr; + note->n_namesz = strlen(KEXEC_CORE_NOTE_NAME) + 1; + note->n_descsz = 16 * 8; + note->n_type = NT_S390_VXRS_LOW; + len = sizeof(Elf64_Nhdr); + + memcpy(ptr + len, KEXEC_CORE_NOTE_NAME, note->n_namesz); + len = roundup(len + note->n_namesz, 4); + + ptr += len; + /* Copy lower halves of SIMD registers 0-15 */ + for (i = 0; i < 16; i++) { + memcpy(ptr, &vx_regs[i], 8); + ptr += 8; + } + return ptr; +} + +/* * Fill ELF notes for one CPU with save area registers */ -void *fill_cpu_elf_notes(void *ptr, struct save_area *sa) +void *fill_cpu_elf_notes(void *ptr, struct save_area *sa, __vector128 *vx_regs) { ptr = nt_prstatus(ptr, sa); ptr = nt_fpregset(ptr, sa); @@ -397,6 +433,10 @@ void *fill_cpu_elf_notes(void *ptr, struct save_area *sa) ptr = nt_s390_tod_preg(ptr, sa); ptr = nt_s390_ctrs(ptr, sa); ptr = nt_s390_prefix(ptr, sa); + if (MACHINE_HAS_VX && vx_regs) { + ptr = nt_s390_vx_low(ptr, vx_regs); + ptr = nt_s390_vx_high(ptr, vx_regs); + } return ptr; } @@ -484,7 +524,7 @@ static int get_cpu_cnt(void) int i, cpus = 0; for (i = 0; i < dump_save_areas.count; i++) { - if (dump_save_areas.areas[i]->pref_reg == 0) + if (dump_save_areas.areas[i]->sa.pref_reg == 0) continue; cpus++; } @@ -530,17 +570,17 @@ static void loads_init(Elf64_Phdr *phdr, u64 loads_offset) */ static void *notes_init(Elf64_Phdr *phdr, void *ptr, u64 notes_offset) { - struct save_area *sa; + struct save_area_ext *sa_ext; void *ptr_start = ptr; int i; ptr = nt_prpsinfo(ptr); for (i = 0; i < dump_save_areas.count; i++) { - sa = dump_save_areas.areas[i]; - if (sa->pref_reg == 0) + sa_ext = dump_save_areas.areas[i]; + if (sa_ext->sa.pref_reg == 0) continue; - ptr = fill_cpu_elf_notes(ptr, sa); + ptr = fill_cpu_elf_notes(ptr, &sa_ext->sa, sa_ext->vx_regs); } ptr = nt_vmcoreinfo(ptr); memset(phdr, 0, sizeof(*phdr)); @@ -581,7 +621,7 @@ int elfcorehdr_alloc(unsigned long long *addr, unsigned long long *size) mem_chunk_cnt = get_mem_chunk_cnt(); - alloc_size = 0x1000 + get_cpu_cnt() * 0x300 + + alloc_size = 0x1000 + get_cpu_cnt() * 0x4a0 + mem_chunk_cnt * sizeof(Elf64_Phdr); hdr = kzalloc_panic(alloc_size); /* Init elf header */ diff --git a/arch/s390/kernel/dis.c b/arch/s390/kernel/dis.c index 993efe6..f376293 100644 --- a/arch/s390/kernel/dis.c +++ b/arch/s390/kernel/dis.c @@ -60,6 +60,11 @@ enum { A_28, /* Access reg. starting at position 28 */ C_8, /* Control reg. starting at position 8 */ C_12, /* Control reg. starting at position 12 */ + V_8, /* Vector reg. starting at position 8, extension bit at 36 */ + V_12, /* Vector reg. starting at position 12, extension bit at 37 */ + V_16, /* Vector reg. starting at position 16, extension bit at 38 */ + V_32, /* Vector reg. starting at position 32, extension bit at 39 */ + W_12, /* Vector reg. at bit 12, extension at bit 37, used as index */ B_16, /* Base register starting at position 16 */ B_32, /* Base register starting at position 32 */ X_12, /* Index register starting at position 12 */ @@ -82,6 +87,8 @@ enum { U8_24, /* 8 bit unsigned value starting at 24 */ U8_32, /* 8 bit unsigned value starting at 32 */ I8_8, /* 8 bit signed value starting at 8 */ + I8_16, /* 8 bit signed value starting at 16 */ + I8_24, /* 8 bit signed value starting at 24 */ I8_32, /* 8 bit signed value starting at 32 */ J12_12, /* PC relative offset at 12 */ I16_16, /* 16 bit signed value starting at 16 */ @@ -96,6 +103,9 @@ enum { U32_16, /* 32 bit unsigned value starting at 16 */ M_16, /* 4 bit optional mask starting at 16 */ M_20, /* 4 bit optional mask starting at 20 */ + M_24, /* 4 bit optional mask starting at 24 */ + M_28, /* 4 bit optional mask starting at 28 */ + M_32, /* 4 bit optional mask starting at 32 */ RO_28, /* optional GPR starting at position 28 */ }; @@ -130,7 +140,7 @@ enum { INSTR_RSY_RDRM, INSTR_RS_AARD, INSTR_RS_CCRD, INSTR_RS_R0RD, INSTR_RS_RRRD, INSTR_RS_RURD, - INSTR_RXE_FRRD, INSTR_RXE_RRRD, + INSTR_RXE_FRRD, INSTR_RXE_RRRD, INSTR_RXE_RRRDM, INSTR_RXF_FRRDF, INSTR_RXY_FRRD, INSTR_RXY_RRRD, INSTR_RXY_URRD, INSTR_RX_FRRD, INSTR_RX_RRRD, INSTR_RX_URRD, @@ -143,6 +153,17 @@ enum { INSTR_SS_L0RDRD, INSTR_SS_LIRDRD, INSTR_SS_LLRDRD, INSTR_SS_RRRDRD, INSTR_SS_RRRDRD2, INSTR_SS_RRRDRD3, INSTR_S_00, INSTR_S_RD, + INSTR_VRI_V0IM, INSTR_VRI_V0I0, INSTR_VRI_V0IIM, INSTR_VRI_VVIM, + INSTR_VRI_VVV0IM, INSTR_VRI_VVV0I0, INSTR_VRI_VVIMM, + INSTR_VRR_VV00MMM, INSTR_VRR_VV000MM, INSTR_VRR_VV0000M, + INSTR_VRR_VV00000, INSTR_VRR_VVV0M0M, INSTR_VRR_VV00M0M, + INSTR_VRR_VVV000M, INSTR_VRR_VVV000V, INSTR_VRR_VVV0000, + INSTR_VRR_VVV0MMM, INSTR_VRR_VVV00MM, INSTR_VRR_VVVMM0V, + INSTR_VRR_VVVM0MV, INSTR_VRR_VVVM00V, INSTR_VRR_VRR0000, + INSTR_VRS_VVRDM, INSTR_VRS_VVRD0, INSTR_VRS_VRRDM, INSTR_VRS_VRRD0, + INSTR_VRS_RVRDM, + INSTR_VRV_VVRDM, INSTR_VRV_VWRDM, + INSTR_VRX_VRRDM, INSTR_VRX_VRRD0, }; static const struct s390_operand operands[] = @@ -168,6 +189,11 @@ static const struct s390_operand operands[] = [A_28] = { 4, 28, OPERAND_AR }, [C_8] = { 4, 8, OPERAND_CR }, [C_12] = { 4, 12, OPERAND_CR }, + [V_8] = { 4, 8, OPERAND_VR }, + [V_12] = { 4, 12, OPERAND_VR }, + [V_16] = { 4, 16, OPERAND_VR }, + [V_32] = { 4, 32, OPERAND_VR }, + [W_12] = { 4, 12, OPERAND_INDEX | OPERAND_VR }, [B_16] = { 4, 16, OPERAND_BASE | OPERAND_GPR }, [B_32] = { 4, 32, OPERAND_BASE | OPERAND_GPR }, [X_12] = { 4, 12, OPERAND_INDEX | OPERAND_GPR }, @@ -190,6 +216,11 @@ static const struct s390_operand operands[] = [U8_24] = { 8, 24, 0 }, [U8_32] = { 8, 32, 0 }, [J12_12] = { 12, 12, OPERAND_PCREL }, + [I8_8] = { 8, 8, OPERAND_SIGNED }, + [I8_16] = { 8, 16, OPERAND_SIGNED }, + [I8_24] = { 8, 24, OPERAND_SIGNED }, + [I8_32] = { 8, 32, OPERAND_SIGNED }, + [I16_32] = { 16, 32, OPERAND_SIGNED }, [I16_16] = { 16, 16, OPERAND_SIGNED }, [U16_16] = { 16, 16, 0 }, [U16_32] = { 16, 32, 0 }, @@ -202,6 +233,9 @@ static const struct s390_operand operands[] = [U32_16] = { 32, 16, 0 }, [M_16] = { 4, 16, 0 }, [M_20] = { 4, 20, 0 }, + [M_24] = { 4, 24, 0 }, + [M_28] = { 4, 28, 0 }, + [M_32] = { 4, 32, 0 }, [RO_28] = { 4, 28, OPERAND_GPR } }; @@ -283,6 +317,7 @@ static const unsigned char formats[][7] = { [INSTR_RS_RURD] = { 0xff, R_8,U4_12,D_20,B_16,0,0 }, [INSTR_RXE_FRRD] = { 0xff, F_8,D_20,X_12,B_16,0,0 }, [INSTR_RXE_RRRD] = { 0xff, R_8,D_20,X_12,B_16,0,0 }, + [INSTR_RXE_RRRDM] = { 0xff, R_8,D_20,X_12,B_16,M_32,0 }, [INSTR_RXF_FRRDF] = { 0xff, F_32,F_8,D_20,X_12,B_16,0 }, [INSTR_RXY_FRRD] = { 0xff, F_8,D20_20,X_12,B_16,0,0 }, [INSTR_RXY_RRRD] = { 0xff, R_8,D20_20,X_12,B_16,0,0 }, @@ -307,6 +342,37 @@ static const unsigned char formats[][7] = { [INSTR_SS_RRRDRD] = { 0xff, D_20,R_8,B_16,D_36,B_32,R_12 }, [INSTR_S_00] = { 0xff, 0,0,0,0,0,0 }, [INSTR_S_RD] = { 0xff, D_20,B_16,0,0,0,0 }, + [INSTR_VRI_V0IM] = { 0xff, V_8,I16_16,M_32,0,0,0 }, + [INSTR_VRI_V0I0] = { 0xff, V_8,I16_16,0,0,0,0 }, + [INSTR_VRI_V0IIM] = { 0xff, V_8,I8_16,I8_24,M_32,0,0 }, + [INSTR_VRI_VVIM] = { 0xff, V_8,I16_16,V_12,M_32,0,0 }, + [INSTR_VRI_VVV0IM]= { 0xff, V_8,V_12,V_16,I8_24,M_32,0 }, + [INSTR_VRI_VVV0I0]= { 0xff, V_8,V_12,V_16,I8_24,0,0 }, + [INSTR_VRI_VVIMM] = { 0xff, V_8,V_12,I16_16,M_32,M_28,0 }, + [INSTR_VRR_VV00MMM]={ 0xff, V_8,V_12,M_32,M_28,M_24,0 }, + [INSTR_VRR_VV000MM]={ 0xff, V_8,V_12,M_32,M_28,0,0 }, + [INSTR_VRR_VV0000M]={ 0xff, V_8,V_12,M_32,0,0,0 }, + [INSTR_VRR_VV00000]={ 0xff, V_8,V_12,0,0,0,0 }, + [INSTR_VRR_VVV0M0M]={ 0xff, V_8,V_12,V_16,M_32,M_24,0 }, + [INSTR_VRR_VV00M0M]={ 0xff, V_8,V_12,M_32,M_24,0,0 }, + [INSTR_VRR_VVV000M]={ 0xff, V_8,V_12,V_16,M_32,0,0 }, + [INSTR_VRR_VVV000V]={ 0xff, V_8,V_12,V_16,V_32,0,0 }, + [INSTR_VRR_VVV0000]={ 0xff, V_8,V_12,V_16,0,0,0 }, + [INSTR_VRR_VVV0MMM]={ 0xff, V_8,V_12,V_16,M_32,M_28,M_24 }, + [INSTR_VRR_VVV00MM]={ 0xff, V_8,V_12,V_16,M_32,M_28,0 }, + [INSTR_VRR_VVVMM0V]={ 0xff, V_8,V_12,V_16,V_32,M_20,M_24 }, + [INSTR_VRR_VVVM0MV]={ 0xff, V_8,V_12,V_16,V_32,M_28,M_20 }, + [INSTR_VRR_VVVM00V]={ 0xff, V_8,V_12,V_16,V_32,M_20,0 }, + [INSTR_VRR_VRR0000]={ 0xff, V_8,R_12,R_16,0,0,0 }, + [INSTR_VRS_VVRDM] = { 0xff, V_8,V_12,D_20,B_16,M_32,0 }, + [INSTR_VRS_VVRD0] = { 0xff, V_8,V_12,D_20,B_16,0,0 }, + [INSTR_VRS_VRRDM] = { 0xff, V_8,R_12,D_20,B_16,M_32,0 }, + [INSTR_VRS_VRRD0] = { 0xff, V_8,R_12,D_20,B_16,0,0 }, + [INSTR_VRS_RVRDM] = { 0xff, R_8,V_12,D_20,B_16,M_32,0 }, + [INSTR_VRV_VVRDM] = { 0xff, V_8,V_12,D_20,B_16,M_32,0 }, + [INSTR_VRV_VWRDM] = { 0xff, V_8,D_20,W_12,B_16,M_32,0 }, + [INSTR_VRX_VRRDM] = { 0xff, V_8,D_20,X_12,B_16,M_32,0 }, + [INSTR_VRX_VRRD0] = { 0xff, V_8,D_20,X_12,B_16,0,0 }, }; enum { @@ -381,6 +447,11 @@ enum { LONG_INSN_MPCIFC, LONG_INSN_STPCIFC, LONG_INSN_PCISTB, + LONG_INSN_VPOPCT, + LONG_INSN_VERLLV, + LONG_INSN_VESRAV, + LONG_INSN_VESRLV, + LONG_INSN_VSBCBI }; static char *long_insn_name[] = { @@ -455,6 +526,11 @@ static char *long_insn_name[] = { [LONG_INSN_MPCIFC] = "mpcifc", [LONG_INSN_STPCIFC] = "stpcifc", [LONG_INSN_PCISTB] = "pcistb", + [LONG_INSN_VPOPCT] = "vpopct", + [LONG_INSN_VERLLV] = "verllv", + [LONG_INSN_VESRAV] = "vesrav", + [LONG_INSN_VESRLV] = "vesrlv", + [LONG_INSN_VSBCBI] = "vsbcbi", }; static struct s390_insn opcode[] = { @@ -1369,6 +1445,150 @@ static struct s390_insn opcode_e5[] = { { "", 0, INSTR_INVALID } }; +static struct s390_insn opcode_e7[] = { +#ifdef CONFIG_64BIT + { "lcbb", 0x27, INSTR_RXE_RRRDM }, + { "vgef", 0x13, INSTR_VRV_VVRDM }, + { "vgeg", 0x12, INSTR_VRV_VVRDM }, + { "vgbm", 0x44, INSTR_VRI_V0I0 }, + { "vgm", 0x46, INSTR_VRI_V0IIM }, + { "vl", 0x06, INSTR_VRX_VRRD0 }, + { "vlr", 0x56, INSTR_VRR_VV00000 }, + { "vlrp", 0x05, INSTR_VRX_VRRDM }, + { "vleb", 0x00, INSTR_VRX_VRRDM }, + { "vleh", 0x01, INSTR_VRX_VRRDM }, + { "vlef", 0x03, INSTR_VRX_VRRDM }, + { "vleg", 0x02, INSTR_VRX_VRRDM }, + { "vleib", 0x40, INSTR_VRI_V0IM }, + { "vleih", 0x41, INSTR_VRI_V0IM }, + { "vleif", 0x43, INSTR_VRI_V0IM }, + { "vleig", 0x42, INSTR_VRI_V0IM }, + { "vlgv", 0x21, INSTR_VRS_RVRDM }, + { "vllez", 0x04, INSTR_VRX_VRRDM }, + { "vlm", 0x36, INSTR_VRS_VVRD0 }, + { "vlbb", 0x07, INSTR_VRX_VRRDM }, + { "vlvg", 0x22, INSTR_VRS_VRRDM }, + { "vlvgp", 0x62, INSTR_VRR_VRR0000 }, + { "vll", 0x37, INSTR_VRS_VRRD0 }, + { "vmrh", 0x61, INSTR_VRR_VVV000M }, + { "vmrl", 0x60, INSTR_VRR_VVV000M }, + { "vpk", 0x94, INSTR_VRR_VVV000M }, + { "vpks", 0x97, INSTR_VRR_VVV0M0M }, + { "vpkls", 0x95, INSTR_VRR_VVV0M0M }, + { "vperm", 0x8c, INSTR_VRR_VVV000V }, + { "vpdi", 0x84, INSTR_VRR_VVV000M }, + { "vrep", 0x4d, INSTR_VRI_VVIM }, + { "vrepi", 0x45, INSTR_VRI_V0IM }, + { "vscef", 0x1b, INSTR_VRV_VWRDM }, + { "vsceg", 0x1a, INSTR_VRV_VWRDM }, + { "vsel", 0x8d, INSTR_VRR_VVV000V }, + { "vseg", 0x5f, INSTR_VRR_VV0000M }, + { "vst", 0x0e, INSTR_VRX_VRRD0 }, + { "vsteb", 0x08, INSTR_VRX_VRRDM }, + { "vsteh", 0x09, INSTR_VRX_VRRDM }, + { "vstef", 0x0b, INSTR_VRX_VRRDM }, + { "vsteg", 0x0a, INSTR_VRX_VRRDM }, + { "vstm", 0x3e, INSTR_VRS_VVRD0 }, + { "vstl", 0x3f, INSTR_VRS_VRRD0 }, + { "vuph", 0xd7, INSTR_VRR_VV0000M }, + { "vuplh", 0xd5, INSTR_VRR_VV0000M }, + { "vupl", 0xd6, INSTR_VRR_VV0000M }, + { "vupll", 0xd4, INSTR_VRR_VV0000M }, + { "va", 0xf3, INSTR_VRR_VVV000M }, + { "vacc", 0xf1, INSTR_VRR_VVV000M }, + { "vac", 0xbb, INSTR_VRR_VVVM00V }, + { "vaccc", 0xb9, INSTR_VRR_VVVM00V }, + { "vn", 0x68, INSTR_VRR_VVV0000 }, + { "vnc", 0x69, INSTR_VRR_VVV0000 }, + { "vavg", 0xf2, INSTR_VRR_VVV000M }, + { "vavgl", 0xf0, INSTR_VRR_VVV000M }, + { "vcksm", 0x66, INSTR_VRR_VVV0000 }, + { "vec", 0xdb, INSTR_VRR_VV0000M }, + { "vecl", 0xd9, INSTR_VRR_VV0000M }, + { "vceq", 0xf8, INSTR_VRR_VVV0M0M }, + { "vch", 0xfb, INSTR_VRR_VVV0M0M }, + { "vchl", 0xf9, INSTR_VRR_VVV0M0M }, + { "vclz", 0x53, INSTR_VRR_VV0000M }, + { "vctz", 0x52, INSTR_VRR_VV0000M }, + { "vx", 0x6d, INSTR_VRR_VVV0000 }, + { "vgfm", 0xb4, INSTR_VRR_VVV000M }, + { "vgfma", 0xbc, INSTR_VRR_VVVM00V }, + { "vlc", 0xde, INSTR_VRR_VV0000M }, + { "vlp", 0xdf, INSTR_VRR_VV0000M }, + { "vmx", 0xff, INSTR_VRR_VVV000M }, + { "vmxl", 0xfd, INSTR_VRR_VVV000M }, + { "vmn", 0xfe, INSTR_VRR_VVV000M }, + { "vmnl", 0xfc, INSTR_VRR_VVV000M }, + { "vmal", 0xaa, INSTR_VRR_VVVM00V }, + { "vmae", 0xae, INSTR_VRR_VVVM00V }, + { "vmale", 0xac, INSTR_VRR_VVVM00V }, + { "vmah", 0xab, INSTR_VRR_VVVM00V }, + { "vmalh", 0xa9, INSTR_VRR_VVVM00V }, + { "vmao", 0xaf, INSTR_VRR_VVVM00V }, + { "vmalo", 0xad, INSTR_VRR_VVVM00V }, + { "vmh", 0xa3, INSTR_VRR_VVV000M }, + { "vmlh", 0xa1, INSTR_VRR_VVV000M }, + { "vml", 0xa2, INSTR_VRR_VVV000M }, + { "vme", 0xa6, INSTR_VRR_VVV000M }, + { "vmle", 0xa4, INSTR_VRR_VVV000M }, + { "vmo", 0xa7, INSTR_VRR_VVV000M }, + { "vmlo", 0xa5, INSTR_VRR_VVV000M }, + { "vno", 0x6b, INSTR_VRR_VVV0000 }, + { "vo", 0x6a, INSTR_VRR_VVV0000 }, + { { 0, LONG_INSN_VPOPCT }, 0x50, INSTR_VRR_VV0000M }, + { { 0, LONG_INSN_VERLLV }, 0x73, INSTR_VRR_VVV000M }, + { "verll", 0x33, INSTR_VRS_VVRDM }, + { "verim", 0x72, INSTR_VRI_VVV0IM }, + { "veslv", 0x70, INSTR_VRR_VVV000M }, + { "vesl", 0x30, INSTR_VRS_VVRDM }, + { { 0, LONG_INSN_VESRAV }, 0x7a, INSTR_VRR_VVV000M }, + { "vesra", 0x3a, INSTR_VRS_VVRDM }, + { { 0, LONG_INSN_VESRLV }, 0x78, INSTR_VRR_VVV000M }, + { "vesrl", 0x38, INSTR_VRS_VVRDM }, + { "vsl", 0x74, INSTR_VRR_VVV0000 }, + { "vslb", 0x75, INSTR_VRR_VVV0000 }, + { "vsldb", 0x77, INSTR_VRI_VVV0I0 }, + { "vsra", 0x7e, INSTR_VRR_VVV0000 }, + { "vsrab", 0x7f, INSTR_VRR_VVV0000 }, + { "vsrl", 0x7c, INSTR_VRR_VVV0000 }, + { "vsrlb", 0x7d, INSTR_VRR_VVV0000 }, + { "vs", 0xf7, INSTR_VRR_VVV000M }, + { "vscb", 0xf5, INSTR_VRR_VVV000M }, + { "vsb", 0xbf, INSTR_VRR_VVVM00V }, + { { 0, LONG_INSN_VSBCBI }, 0xbd, INSTR_VRR_VVVM00V }, + { "vsumg", 0x65, INSTR_VRR_VVV000M }, + { "vsumq", 0x67, INSTR_VRR_VVV000M }, + { "vsum", 0x64, INSTR_VRR_VVV000M }, + { "vtm", 0xd8, INSTR_VRR_VV00000 }, + { "vfae", 0x82, INSTR_VRR_VVV0M0M }, + { "vfee", 0x80, INSTR_VRR_VVV0M0M }, + { "vfene", 0x81, INSTR_VRR_VVV0M0M }, + { "vistr", 0x5c, INSTR_VRR_VV00M0M }, + { "vstrc", 0x8a, INSTR_VRR_VVVMM0V }, + { "vfa", 0xe3, INSTR_VRR_VVV00MM }, + { "wfc", 0xcb, INSTR_VRR_VV000MM }, + { "wfk", 0xca, INSTR_VRR_VV000MM }, + { "vfce", 0xe8, INSTR_VRR_VVV0MMM }, + { "vfch", 0xeb, INSTR_VRR_VVV0MMM }, + { "vfche", 0xea, INSTR_VRR_VVV0MMM }, + { "vcdg", 0xc3, INSTR_VRR_VV00MMM }, + { "vcdlg", 0xc1, INSTR_VRR_VV00MMM }, + { "vcgd", 0xc2, INSTR_VRR_VV00MMM }, + { "vclgd", 0xc0, INSTR_VRR_VV00MMM }, + { "vfd", 0xe5, INSTR_VRR_VVV00MM }, + { "vfi", 0xc7, INSTR_VRR_VV00MMM }, + { "vlde", 0xc4, INSTR_VRR_VV000MM }, + { "vled", 0xc5, INSTR_VRR_VV00MMM }, + { "vfm", 0xe7, INSTR_VRR_VVV00MM }, + { "vfma", 0x8f, INSTR_VRR_VVVM0MV }, + { "vfms", 0x8e, INSTR_VRR_VVVM0MV }, + { "vfpso", 0xcc, INSTR_VRR_VV00MMM }, + { "vfsq", 0xce, INSTR_VRR_VV000MM }, + { "vfs", 0xe2, INSTR_VRR_VVV00MM }, + { "vftci", 0x4a, INSTR_VRI_VVIMM }, +#endif +}; + static struct s390_insn opcode_eb[] = { #ifdef CONFIG_64BIT { "lmg", 0x04, INSTR_RSY_RRRD }, @@ -1552,16 +1772,17 @@ static struct s390_insn opcode_ed[] = { static unsigned int extract_operand(unsigned char *code, const struct s390_operand *operand) { + unsigned char *cp; unsigned int val; int bits; /* Extract fragments of the operand byte for byte. */ - code += operand->shift / 8; + cp = code + operand->shift / 8; bits = (operand->shift & 7) + operand->bits; val = 0; do { val <<= 8; - val |= (unsigned int) *code++; + val |= (unsigned int) *cp++; bits -= 8; } while (bits > 0); val >>= -bits; @@ -1571,6 +1792,18 @@ static unsigned int extract_operand(unsigned char *code, if (operand->bits == 20 && operand->shift == 20) val = (val & 0xff) << 12 | (val & 0xfff00) >> 8; + /* Check for register extensions bits for vector registers. */ + if (operand->flags & OPERAND_VR) { + if (operand->shift == 8) + val |= (code[4] & 8) << 1; + else if (operand->shift == 12) + val |= (code[4] & 4) << 2; + else if (operand->shift == 16) + val |= (code[4] & 2) << 3; + else if (operand->shift == 32) + val |= (code[4] & 1) << 4; + } + /* Sign extend value if the operand is signed or pc relative. */ if ((operand->flags & (OPERAND_SIGNED | OPERAND_PCREL)) && (val & (1U << (operand->bits - 1)))) @@ -1639,6 +1872,10 @@ struct s390_insn *find_insn(unsigned char *code) case 0xe5: table = opcode_e5; break; + case 0xe7: + table = opcode_e7; + opfrag = code[5]; + break; case 0xeb: table = opcode_eb; opfrag = code[5]; @@ -1734,6 +1971,8 @@ static int print_insn(char *buffer, unsigned char *code, unsigned long addr) ptr += sprintf(ptr, "%%a%i", value); else if (operand->flags & OPERAND_CR) ptr += sprintf(ptr, "%%c%i", value); + else if (operand->flags & OPERAND_VR) + ptr += sprintf(ptr, "%%v%i", value); else if (operand->flags & OPERAND_PCREL) ptr += sprintf(ptr, "%lx", (signed int) value + addr); diff --git a/arch/s390/kernel/early.c b/arch/s390/kernel/early.c index 0dff972..cef2879e 100644 --- a/arch/s390/kernel/early.c +++ b/arch/s390/kernel/early.c @@ -390,10 +390,10 @@ static __init void detect_machine_facilities(void) S390_lowcore.machine_flags |= MACHINE_FLAG_LPP; if (test_facility(50) && test_facility(73)) S390_lowcore.machine_flags |= MACHINE_FLAG_TE; - if (test_facility(66)) - S390_lowcore.machine_flags |= MACHINE_FLAG_RRBM; if (test_facility(51)) S390_lowcore.machine_flags |= MACHINE_FLAG_TLB_LC; + if (test_facility(129)) + S390_lowcore.machine_flags |= MACHINE_FLAG_VX; #endif } diff --git a/arch/s390/kernel/entry.h b/arch/s390/kernel/entry.h index 6ac7819..0554b97 100644 --- a/arch/s390/kernel/entry.h +++ b/arch/s390/kernel/entry.h @@ -4,7 +4,7 @@ #include <linux/types.h> #include <linux/signal.h> #include <asm/ptrace.h> -#include <asm/cputime.h> +#include <asm/idle.h> extern void *restart_stack; extern unsigned long suspend_zero_pages; @@ -21,6 +21,8 @@ void psw_idle(struct s390_idle_data *, unsigned long); asmlinkage long do_syscall_trace_enter(struct pt_regs *regs); asmlinkage void do_syscall_trace_exit(struct pt_regs *regs); +int alloc_vector_registers(struct task_struct *tsk); + void do_protection_exception(struct pt_regs *regs); void do_dat_exception(struct pt_regs *regs); @@ -43,13 +45,15 @@ void special_op_exception(struct pt_regs *regs); void specification_exception(struct pt_regs *regs); void transaction_exception(struct pt_regs *regs); void translation_exception(struct pt_regs *regs); +void vector_exception(struct pt_regs *regs); void do_per_trap(struct pt_regs *regs); +void do_report_trap(struct pt_regs *regs, int si_signo, int si_code, char *str); void syscall_trace(struct pt_regs *regs, int entryexit); void kernel_stack_overflow(struct pt_regs * regs); void do_signal(struct pt_regs *regs); -void handle_signal32(unsigned long sig, struct k_sigaction *ka, - siginfo_t *info, sigset_t *oldset, struct pt_regs *regs); +void handle_signal32(struct ksignal *ksig, sigset_t *oldset, + struct pt_regs *regs); void do_notify_resume(struct pt_regs *regs); void __init init_IRQ(void); diff --git a/arch/s390/kernel/entry64.S b/arch/s390/kernel/entry64.S index f2e674c..7b2e03a 100644 --- a/arch/s390/kernel/entry64.S +++ b/arch/s390/kernel/entry64.S @@ -42,7 +42,8 @@ STACK_SHIFT = PAGE_SHIFT + THREAD_ORDER STACK_SIZE = 1 << STACK_SHIFT STACK_INIT = STACK_SIZE - STACK_FRAME_OVERHEAD - __PT_SIZE -_TIF_WORK = (_TIF_SIGPENDING | _TIF_NOTIFY_RESUME | _TIF_NEED_RESCHED) +_TIF_WORK = (_TIF_SIGPENDING | _TIF_NOTIFY_RESUME | _TIF_NEED_RESCHED | \ + _TIF_UPROBE) _TIF_TRACE = (_TIF_SYSCALL_TRACE | _TIF_SYSCALL_AUDIT | _TIF_SECCOMP | \ _TIF_SYSCALL_TRACEPOINT) _CIF_WORK = (_CIF_MCCK_PENDING | _CIF_ASCE) @@ -265,6 +266,10 @@ sysc_work: jo sysc_mcck_pending tm __TI_flags+7(%r12),_TIF_NEED_RESCHED jo sysc_reschedule +#ifdef CONFIG_UPROBES + tm __TI_flags+7(%r12),_TIF_UPROBE + jo sysc_uprobe_notify +#endif tm __PT_FLAGS+7(%r11),_PIF_PER_TRAP jo sysc_singlestep tm __TI_flags+7(%r12),_TIF_SIGPENDING @@ -323,6 +328,16 @@ sysc_notify_resume: jg do_notify_resume # +# _TIF_UPROBE is set, call uprobe_notify_resume +# +#ifdef CONFIG_UPROBES +sysc_uprobe_notify: + lgr %r2,%r11 # pass pointer to pt_regs + larl %r14,sysc_return + jg uprobe_notify_resume +#endif + +# # _PIF_PER_TRAP is set, call do_per_trap # sysc_singlestep: diff --git a/arch/s390/kernel/ftrace.c b/arch/s390/kernel/ftrace.c index 54d6493..51d14fe5 100644 --- a/arch/s390/kernel/ftrace.c +++ b/arch/s390/kernel/ftrace.c @@ -1,7 +1,7 @@ /* * Dynamic function tracer architecture backend. * - * Copyright IBM Corp. 2009 + * Copyright IBM Corp. 2009,2014 * * Author(s): Heiko Carstens <heiko.carstens@de.ibm.com>, * Martin Schwidefsky <schwidefsky@de.ibm.com> @@ -17,100 +17,76 @@ #include <asm/asm-offsets.h> #include "entry.h" -#ifdef CONFIG_DYNAMIC_FTRACE - +void mcount_replace_code(void); void ftrace_disable_code(void); void ftrace_enable_insn(void); -#ifdef CONFIG_64BIT /* - * The 64-bit mcount code looks like this: + * The mcount code looks like this: * stg %r14,8(%r15) # offset 0 - * > larl %r1,<&counter> # offset 6 - * > brasl %r14,_mcount # offset 12 + * larl %r1,<&counter> # offset 6 + * brasl %r14,_mcount # offset 12 * lg %r14,8(%r15) # offset 18 - * Total length is 24 bytes. The middle two instructions of the mcount - * block get overwritten by ftrace_make_nop / ftrace_make_call. - * The 64-bit enabled ftrace code block looks like this: - * stg %r14,8(%r15) # offset 0 + * Total length is 24 bytes. The complete mcount block initially gets replaced + * by ftrace_make_nop. Subsequent calls to ftrace_make_call / ftrace_make_nop + * only patch the jg/lg instruction within the block. + * Note: we do not patch the first instruction to an unconditional branch, + * since that would break kprobes/jprobes. It is easier to leave the larl + * instruction in and only modify the second instruction. + * The enabled ftrace code block looks like this: + * larl %r0,.+24 # offset 0 * > lg %r1,__LC_FTRACE_FUNC # offset 6 - * > lgr %r0,%r0 # offset 12 - * > basr %r14,%r1 # offset 16 - * lg %r14,8(%15) # offset 18 - * The return points of the mcount/ftrace function have the same offset 18. - * The 64-bit disable ftrace code block looks like this: - * stg %r14,8(%r15) # offset 0 + * br %r1 # offset 12 + * brcl 0,0 # offset 14 + * brc 0,0 # offset 20 + * The ftrace function gets called with a non-standard C function call ABI + * where r0 contains the return address. It is also expected that the called + * function only clobbers r0 and r1, but restores r2-r15. + * The return point of the ftrace function has offset 24, so execution + * continues behind the mcount block. + * larl %r0,.+24 # offset 0 * > jg .+18 # offset 6 - * > lgr %r0,%r0 # offset 12 - * > basr %r14,%r1 # offset 16 - * lg %r14,8(%15) # offset 18 + * br %r1 # offset 12 + * brcl 0,0 # offset 14 + * brc 0,0 # offset 20 * The jg instruction branches to offset 24 to skip as many instructions * as possible. */ asm( " .align 4\n" + "mcount_replace_code:\n" + " larl %r0,0f\n" "ftrace_disable_code:\n" " jg 0f\n" - " lgr %r0,%r0\n" - " basr %r14,%r1\n" + " br %r1\n" + " brcl 0,0\n" + " brc 0,0\n" "0:\n" " .align 4\n" "ftrace_enable_insn:\n" " lg %r1,"__stringify(__LC_FTRACE_FUNC)"\n"); +#define MCOUNT_BLOCK_SIZE 24 +#define MCOUNT_INSN_OFFSET 6 #define FTRACE_INSN_SIZE 6 -#else /* CONFIG_64BIT */ -/* - * The 31-bit mcount code looks like this: - * st %r14,4(%r15) # offset 0 - * > bras %r1,0f # offset 4 - * > .long _mcount # offset 8 - * > .long <&counter> # offset 12 - * > 0: l %r14,0(%r1) # offset 16 - * > l %r1,4(%r1) # offset 20 - * basr %r14,%r14 # offset 24 - * l %r14,4(%r15) # offset 26 - * Total length is 30 bytes. The twenty bytes starting from offset 4 - * to offset 24 get overwritten by ftrace_make_nop / ftrace_make_call. - * The 31-bit enabled ftrace code block looks like this: - * st %r14,4(%r15) # offset 0 - * > l %r14,__LC_FTRACE_FUNC # offset 4 - * > j 0f # offset 8 - * > .fill 12,1,0x07 # offset 12 - * 0: basr %r14,%r14 # offset 24 - * l %r14,4(%r14) # offset 26 - * The return points of the mcount/ftrace function have the same offset 26. - * The 31-bit disabled ftrace code block looks like this: - * st %r14,4(%r15) # offset 0 - * > j .+26 # offset 4 - * > j 0f # offset 8 - * > .fill 12,1,0x07 # offset 12 - * 0: basr %r14,%r14 # offset 24 - * l %r14,4(%r14) # offset 26 - * The j instruction branches to offset 30 to skip as many instructions - * as possible. - */ -asm( - " .align 4\n" - "ftrace_disable_code:\n" - " j 1f\n" - " j 0f\n" - " .fill 12,1,0x07\n" - "0: basr %r14,%r14\n" - "1:\n" - " .align 4\n" - "ftrace_enable_insn:\n" - " l %r14,"__stringify(__LC_FTRACE_FUNC)"\n"); - -#define FTRACE_INSN_SIZE 4 - -#endif /* CONFIG_64BIT */ - +int ftrace_modify_call(struct dyn_ftrace *rec, unsigned long old_addr, + unsigned long addr) +{ + return 0; +} int ftrace_make_nop(struct module *mod, struct dyn_ftrace *rec, unsigned long addr) { + /* Initial replacement of the whole mcount block */ + if (addr == MCOUNT_ADDR) { + if (probe_kernel_write((void *) rec->ip - MCOUNT_INSN_OFFSET, + mcount_replace_code, + MCOUNT_BLOCK_SIZE)) + return -EPERM; + return 0; + } if (probe_kernel_write((void *) rec->ip, ftrace_disable_code, MCOUNT_INSN_SIZE)) return -EPERM; @@ -135,8 +111,6 @@ int __init ftrace_dyn_arch_init(void) return 0; } -#endif /* CONFIG_DYNAMIC_FTRACE */ - #ifdef CONFIG_FUNCTION_GRAPH_TRACER /* * Hook the return address and push it in the stack of return addresses @@ -162,31 +136,26 @@ out: return parent; } -#ifdef CONFIG_DYNAMIC_FTRACE /* * Patch the kernel code at ftrace_graph_caller location. The instruction - * there is branch relative and save to prepare_ftrace_return. To disable - * the call to prepare_ftrace_return we patch the bras offset to point - * directly after the instructions. To enable the call we calculate - * the original offset to prepare_ftrace_return and put it back. + * there is branch relative on condition. To enable the ftrace graph code + * block, we simply patch the mask field of the instruction to zero and + * turn the instruction into a nop. + * To disable the ftrace graph code the mask field will be patched to + * all ones, which turns the instruction into an unconditional branch. */ int ftrace_enable_ftrace_graph_caller(void) { - unsigned short offset; + u8 op = 0x04; /* set mask field to zero */ - offset = ((void *) prepare_ftrace_return - - (void *) ftrace_graph_caller) / 2; - return probe_kernel_write((void *) ftrace_graph_caller + 2, - &offset, sizeof(offset)); + return probe_kernel_write(__va(ftrace_graph_caller)+1, &op, sizeof(op)); } int ftrace_disable_ftrace_graph_caller(void) { - static unsigned short offset = 0x0002; + u8 op = 0xf4; /* set mask field to all ones */ - return probe_kernel_write((void *) ftrace_graph_caller + 2, - &offset, sizeof(offset)); + return probe_kernel_write(__va(ftrace_graph_caller)+1, &op, sizeof(op)); } -#endif /* CONFIG_DYNAMIC_FTRACE */ #endif /* CONFIG_FUNCTION_GRAPH_TRACER */ diff --git a/arch/s390/kernel/head.S b/arch/s390/kernel/head.S index e88d35d..d62eee1 100644 --- a/arch/s390/kernel/head.S +++ b/arch/s390/kernel/head.S @@ -398,7 +398,7 @@ ENTRY(startup_kdump) xc __LC_STFL_FAC_LIST(8),__LC_STFL_FAC_LIST #ifndef CONFIG_MARCH_G5 # check capabilities against MARCH_{G5,Z900,Z990,Z9_109,Z10} - .insn s,0xb2b10000,__LC_STFL_FAC_LIST # store facility list + .insn s,0xb2b10000,0 # store facilities @ __LC_STFL_FAC_LIST tm __LC_STFL_FAC_LIST,0x01 # stfle available ? jz 0f la %r0,1 diff --git a/arch/s390/kernel/idle.c b/arch/s390/kernel/idle.c new file mode 100644 index 0000000..7559f1b --- /dev/null +++ b/arch/s390/kernel/idle.c @@ -0,0 +1,124 @@ +/* + * Idle functions for s390. + * + * Copyright IBM Corp. 2014 + * + * Author(s): Martin Schwidefsky <schwidefsky@de.ibm.com> + */ + +#include <linux/kernel.h> +#include <linux/kernel_stat.h> +#include <linux/kprobes.h> +#include <linux/notifier.h> +#include <linux/init.h> +#include <linux/cpu.h> +#include <asm/cputime.h> +#include <asm/nmi.h> +#include <asm/smp.h> +#include "entry.h" + +static DEFINE_PER_CPU(struct s390_idle_data, s390_idle); + +void __kprobes enabled_wait(void) +{ + struct s390_idle_data *idle = this_cpu_ptr(&s390_idle); + unsigned long long idle_time; + unsigned long psw_mask; + + trace_hardirqs_on(); + + /* Wait for external, I/O or machine check interrupt. */ + psw_mask = PSW_KERNEL_BITS | PSW_MASK_WAIT | PSW_MASK_DAT | + PSW_MASK_IO | PSW_MASK_EXT | PSW_MASK_MCHECK; + clear_cpu_flag(CIF_NOHZ_DELAY); + + /* Call the assembler magic in entry.S */ + psw_idle(idle, psw_mask); + + /* Account time spent with enabled wait psw loaded as idle time. */ + idle->sequence++; + smp_wmb(); + idle_time = idle->clock_idle_exit - idle->clock_idle_enter; + idle->clock_idle_enter = idle->clock_idle_exit = 0ULL; + idle->idle_time += idle_time; + idle->idle_count++; + account_idle_time(idle_time); + smp_wmb(); + idle->sequence++; +} + +static ssize_t show_idle_count(struct device *dev, + struct device_attribute *attr, char *buf) +{ + struct s390_idle_data *idle = &per_cpu(s390_idle, dev->id); + unsigned long long idle_count; + unsigned int sequence; + + do { + sequence = ACCESS_ONCE(idle->sequence); + idle_count = ACCESS_ONCE(idle->idle_count); + if (ACCESS_ONCE(idle->clock_idle_enter)) + idle_count++; + } while ((sequence & 1) || (ACCESS_ONCE(idle->sequence) != sequence)); + return sprintf(buf, "%llu\n", idle_count); +} +DEVICE_ATTR(idle_count, 0444, show_idle_count, NULL); + +static ssize_t show_idle_time(struct device *dev, + struct device_attribute *attr, char *buf) +{ + struct s390_idle_data *idle = &per_cpu(s390_idle, dev->id); + unsigned long long now, idle_time, idle_enter, idle_exit; + unsigned int sequence; + + do { + now = get_tod_clock(); + sequence = ACCESS_ONCE(idle->sequence); + idle_time = ACCESS_ONCE(idle->idle_time); + idle_enter = ACCESS_ONCE(idle->clock_idle_enter); + idle_exit = ACCESS_ONCE(idle->clock_idle_exit); + } while ((sequence & 1) || (ACCESS_ONCE(idle->sequence) != sequence)); + idle_time += idle_enter ? ((idle_exit ? : now) - idle_enter) : 0; + return sprintf(buf, "%llu\n", idle_time >> 12); +} +DEVICE_ATTR(idle_time_us, 0444, show_idle_time, NULL); + +cputime64_t arch_cpu_idle_time(int cpu) +{ + struct s390_idle_data *idle = &per_cpu(s390_idle, cpu); + unsigned long long now, idle_enter, idle_exit; + unsigned int sequence; + + do { + now = get_tod_clock(); + sequence = ACCESS_ONCE(idle->sequence); + idle_enter = ACCESS_ONCE(idle->clock_idle_enter); + idle_exit = ACCESS_ONCE(idle->clock_idle_exit); + } while ((sequence & 1) || (ACCESS_ONCE(idle->sequence) != sequence)); + return idle_enter ? ((idle_exit ?: now) - idle_enter) : 0; +} + +void arch_cpu_idle_enter(void) +{ + local_mcck_disable(); +} + +void arch_cpu_idle(void) +{ + if (!test_cpu_flag(CIF_MCCK_PENDING)) + /* Halt the cpu and keep track of cpu time accounting. */ + enabled_wait(); + local_irq_enable(); +} + +void arch_cpu_idle_exit(void) +{ + local_mcck_enable(); + if (test_cpu_flag(CIF_MCCK_PENDING)) + s390_handle_mcck(); +} + +void arch_cpu_idle_dead(void) +{ + cpu_die(); +} diff --git a/arch/s390/kernel/ipl.c b/arch/s390/kernel/ipl.c index 633ca75..39badb9 100644 --- a/arch/s390/kernel/ipl.c +++ b/arch/s390/kernel/ipl.c @@ -455,22 +455,6 @@ DEFINE_IPL_ATTR_RO(ipl_fcp, bootprog, "%lld\n", (unsigned long long) DEFINE_IPL_ATTR_RO(ipl_fcp, br_lba, "%lld\n", (unsigned long long) IPL_PARMBLOCK_START->ipl_info.fcp.br_lba); -static struct attribute *ipl_fcp_attrs[] = { - &sys_ipl_type_attr.attr, - &sys_ipl_device_attr.attr, - &sys_ipl_fcp_wwpn_attr.attr, - &sys_ipl_fcp_lun_attr.attr, - &sys_ipl_fcp_bootprog_attr.attr, - &sys_ipl_fcp_br_lba_attr.attr, - NULL, -}; - -static struct attribute_group ipl_fcp_attr_group = { - .attrs = ipl_fcp_attrs, -}; - -/* CCW ipl device attributes */ - static ssize_t ipl_ccw_loadparm_show(struct kobject *kobj, struct kobj_attribute *attr, char *page) { @@ -487,6 +471,23 @@ static ssize_t ipl_ccw_loadparm_show(struct kobject *kobj, static struct kobj_attribute sys_ipl_ccw_loadparm_attr = __ATTR(loadparm, 0444, ipl_ccw_loadparm_show, NULL); +static struct attribute *ipl_fcp_attrs[] = { + &sys_ipl_type_attr.attr, + &sys_ipl_device_attr.attr, + &sys_ipl_fcp_wwpn_attr.attr, + &sys_ipl_fcp_lun_attr.attr, + &sys_ipl_fcp_bootprog_attr.attr, + &sys_ipl_fcp_br_lba_attr.attr, + &sys_ipl_ccw_loadparm_attr.attr, + NULL, +}; + +static struct attribute_group ipl_fcp_attr_group = { + .attrs = ipl_fcp_attrs, +}; + +/* CCW ipl device attributes */ + static struct attribute *ipl_ccw_attrs_vm[] = { &sys_ipl_type_attr.attr, &sys_ipl_device_attr.attr, @@ -765,28 +766,10 @@ DEFINE_IPL_ATTR_RW(reipl_fcp, br_lba, "%lld\n", "%lld\n", DEFINE_IPL_ATTR_RW(reipl_fcp, device, "0.0.%04llx\n", "0.0.%llx\n", reipl_block_fcp->ipl_info.fcp.devno); -static struct attribute *reipl_fcp_attrs[] = { - &sys_reipl_fcp_device_attr.attr, - &sys_reipl_fcp_wwpn_attr.attr, - &sys_reipl_fcp_lun_attr.attr, - &sys_reipl_fcp_bootprog_attr.attr, - &sys_reipl_fcp_br_lba_attr.attr, - NULL, -}; - -static struct attribute_group reipl_fcp_attr_group = { - .attrs = reipl_fcp_attrs, -}; - -/* CCW reipl device attributes */ - -DEFINE_IPL_ATTR_RW(reipl_ccw, device, "0.0.%04llx\n", "0.0.%llx\n", - reipl_block_ccw->ipl_info.ccw.devno); - static void reipl_get_ascii_loadparm(char *loadparm, struct ipl_parameter_block *ibp) { - memcpy(loadparm, ibp->ipl_info.ccw.load_parm, LOADPARM_LEN); + memcpy(loadparm, ibp->hdr.loadparm, LOADPARM_LEN); EBCASC(loadparm, LOADPARM_LEN); loadparm[LOADPARM_LEN] = 0; strim(loadparm); @@ -821,13 +804,50 @@ static ssize_t reipl_generic_loadparm_store(struct ipl_parameter_block *ipb, return -EINVAL; } /* initialize loadparm with blanks */ - memset(ipb->ipl_info.ccw.load_parm, ' ', LOADPARM_LEN); + memset(ipb->hdr.loadparm, ' ', LOADPARM_LEN); /* copy and convert to ebcdic */ - memcpy(ipb->ipl_info.ccw.load_parm, buf, lp_len); - ASCEBC(ipb->ipl_info.ccw.load_parm, LOADPARM_LEN); + memcpy(ipb->hdr.loadparm, buf, lp_len); + ASCEBC(ipb->hdr.loadparm, LOADPARM_LEN); return len; } +/* FCP wrapper */ +static ssize_t reipl_fcp_loadparm_show(struct kobject *kobj, + struct kobj_attribute *attr, char *page) +{ + return reipl_generic_loadparm_show(reipl_block_fcp, page); +} + +static ssize_t reipl_fcp_loadparm_store(struct kobject *kobj, + struct kobj_attribute *attr, + const char *buf, size_t len) +{ + return reipl_generic_loadparm_store(reipl_block_fcp, buf, len); +} + +static struct kobj_attribute sys_reipl_fcp_loadparm_attr = + __ATTR(loadparm, S_IRUGO | S_IWUSR, reipl_fcp_loadparm_show, + reipl_fcp_loadparm_store); + +static struct attribute *reipl_fcp_attrs[] = { + &sys_reipl_fcp_device_attr.attr, + &sys_reipl_fcp_wwpn_attr.attr, + &sys_reipl_fcp_lun_attr.attr, + &sys_reipl_fcp_bootprog_attr.attr, + &sys_reipl_fcp_br_lba_attr.attr, + &sys_reipl_fcp_loadparm_attr.attr, + NULL, +}; + +static struct attribute_group reipl_fcp_attr_group = { + .attrs = reipl_fcp_attrs, +}; + +/* CCW reipl device attributes */ + +DEFINE_IPL_ATTR_RW(reipl_ccw, device, "0.0.%04llx\n", "0.0.%llx\n", + reipl_block_ccw->ipl_info.ccw.devno); + /* NSS wrapper */ static ssize_t reipl_nss_loadparm_show(struct kobject *kobj, struct kobj_attribute *attr, char *page) @@ -1125,11 +1145,10 @@ static void reipl_block_ccw_fill_parms(struct ipl_parameter_block *ipb) /* LOADPARM */ /* check if read scp info worked and set loadparm */ if (sclp_ipl_info.is_valid) - memcpy(ipb->ipl_info.ccw.load_parm, - &sclp_ipl_info.loadparm, LOADPARM_LEN); + memcpy(ipb->hdr.loadparm, &sclp_ipl_info.loadparm, LOADPARM_LEN); else /* read scp info failed: set empty loadparm (EBCDIC blanks) */ - memset(ipb->ipl_info.ccw.load_parm, 0x40, LOADPARM_LEN); + memset(ipb->hdr.loadparm, 0x40, LOADPARM_LEN); ipb->hdr.flags = DIAG308_FLAGS_LP_VALID; /* VM PARM */ @@ -1251,9 +1270,16 @@ static int __init reipl_fcp_init(void) return rc; } - if (ipl_info.type == IPL_TYPE_FCP) + if (ipl_info.type == IPL_TYPE_FCP) { memcpy(reipl_block_fcp, IPL_PARMBLOCK_START, PAGE_SIZE); - else { + /* + * Fix loadparm: There are systems where the (SCSI) LOADPARM + * is invalid in the SCSI IPL parameter block, so take it + * always from sclp_ipl_info. + */ + memcpy(reipl_block_fcp->hdr.loadparm, sclp_ipl_info.loadparm, + LOADPARM_LEN); + } else { reipl_block_fcp->hdr.len = IPL_PARM_BLK_FCP_LEN; reipl_block_fcp->hdr.version = IPL_PARM_BLOCK_VERSION; reipl_block_fcp->hdr.blk0_len = IPL_PARM_BLK0_FCP_LEN; @@ -1864,7 +1890,23 @@ static void __init shutdown_actions_init(void) static int __init s390_ipl_init(void) { + char str[8] = {0x40, 0x40, 0x40, 0x40, 0x40, 0x40, 0x40, 0x40}; + sclp_get_ipl_info(&sclp_ipl_info); + /* + * Fix loadparm: There are systems where the (SCSI) LOADPARM + * returned by read SCP info is invalid (contains EBCDIC blanks) + * when the system has been booted via diag308. In that case we use + * the value from diag308, if available. + * + * There are also systems where diag308 store does not work in + * case the system is booted from HMC. Fortunately in this case + * READ SCP info provides the correct value. + */ + if (memcmp(sclp_ipl_info.loadparm, str, sizeof(str)) == 0 && + diag308_set_works) + memcpy(sclp_ipl_info.loadparm, ipl_block.hdr.loadparm, + LOADPARM_LEN); shutdown_actions_init(); shutdown_triggers_init(); return 0; @@ -2060,6 +2102,13 @@ void s390_reset_system(void (*func)(void *), void *data) S390_lowcore.program_new_psw.addr = PSW_ADDR_AMODE | (unsigned long) s390_base_pgm_handler; + /* + * Clear subchannel ID and number to signal new kernel that no CCW or + * SCSI IPL has been done (for kexec and kdump) + */ + S390_lowcore.subchannel_id = 0; + S390_lowcore.subchannel_nr = 0; + /* Store status at absolute zero */ store_status(); diff --git a/arch/s390/kernel/irq.c b/arch/s390/kernel/irq.c index 99b0b09..1b8a38a 100644 --- a/arch/s390/kernel/irq.c +++ b/arch/s390/kernel/irq.c @@ -30,6 +30,7 @@ DEFINE_PER_CPU_SHARED_ALIGNED(struct irq_stat, irq_stat); EXPORT_PER_CPU_SYMBOL_GPL(irq_stat); struct irq_class { + int irq; char *name; char *desc; }; @@ -45,9 +46,9 @@ struct irq_class { * up with having a sum which accounts each interrupt twice. */ static const struct irq_class irqclass_main_desc[NR_IRQS_BASE] = { - [EXT_INTERRUPT] = {.name = "EXT"}, - [IO_INTERRUPT] = {.name = "I/O"}, - [THIN_INTERRUPT] = {.name = "AIO"}, + {.irq = EXT_INTERRUPT, .name = "EXT"}, + {.irq = IO_INTERRUPT, .name = "I/O"}, + {.irq = THIN_INTERRUPT, .name = "AIO"}, }; /* @@ -56,38 +57,39 @@ static const struct irq_class irqclass_main_desc[NR_IRQS_BASE] = { * In addition this list contains non external / I/O events like NMIs. */ static const struct irq_class irqclass_sub_desc[NR_ARCH_IRQS] = { - [IRQEXT_CLK] = {.name = "CLK", .desc = "[EXT] Clock Comparator"}, - [IRQEXT_EXC] = {.name = "EXC", .desc = "[EXT] External Call"}, - [IRQEXT_EMS] = {.name = "EMS", .desc = "[EXT] Emergency Signal"}, - [IRQEXT_TMR] = {.name = "TMR", .desc = "[EXT] CPU Timer"}, - [IRQEXT_TLA] = {.name = "TAL", .desc = "[EXT] Timing Alert"}, - [IRQEXT_PFL] = {.name = "PFL", .desc = "[EXT] Pseudo Page Fault"}, - [IRQEXT_DSD] = {.name = "DSD", .desc = "[EXT] DASD Diag"}, - [IRQEXT_VRT] = {.name = "VRT", .desc = "[EXT] Virtio"}, - [IRQEXT_SCP] = {.name = "SCP", .desc = "[EXT] Service Call"}, - [IRQEXT_IUC] = {.name = "IUC", .desc = "[EXT] IUCV"}, - [IRQEXT_CMS] = {.name = "CMS", .desc = "[EXT] CPU-Measurement: Sampling"}, - [IRQEXT_CMC] = {.name = "CMC", .desc = "[EXT] CPU-Measurement: Counter"}, - [IRQEXT_CMR] = {.name = "CMR", .desc = "[EXT] CPU-Measurement: RI"}, - [IRQIO_CIO] = {.name = "CIO", .desc = "[I/O] Common I/O Layer Interrupt"}, - [IRQIO_QAI] = {.name = "QAI", .desc = "[I/O] QDIO Adapter Interrupt"}, - [IRQIO_DAS] = {.name = "DAS", .desc = "[I/O] DASD"}, - [IRQIO_C15] = {.name = "C15", .desc = "[I/O] 3215"}, - [IRQIO_C70] = {.name = "C70", .desc = "[I/O] 3270"}, - [IRQIO_TAP] = {.name = "TAP", .desc = "[I/O] Tape"}, - [IRQIO_VMR] = {.name = "VMR", .desc = "[I/O] Unit Record Devices"}, - [IRQIO_LCS] = {.name = "LCS", .desc = "[I/O] LCS"}, - [IRQIO_CLW] = {.name = "CLW", .desc = "[I/O] CLAW"}, - [IRQIO_CTC] = {.name = "CTC", .desc = "[I/O] CTC"}, - [IRQIO_APB] = {.name = "APB", .desc = "[I/O] AP Bus"}, - [IRQIO_ADM] = {.name = "ADM", .desc = "[I/O] EADM Subchannel"}, - [IRQIO_CSC] = {.name = "CSC", .desc = "[I/O] CHSC Subchannel"}, - [IRQIO_PCI] = {.name = "PCI", .desc = "[I/O] PCI Interrupt" }, - [IRQIO_MSI] = {.name = "MSI", .desc = "[I/O] MSI Interrupt" }, - [IRQIO_VIR] = {.name = "VIR", .desc = "[I/O] Virtual I/O Devices"}, - [IRQIO_VAI] = {.name = "VAI", .desc = "[I/O] Virtual I/O Devices AI"}, - [NMI_NMI] = {.name = "NMI", .desc = "[NMI] Machine Check"}, - [CPU_RST] = {.name = "RST", .desc = "[CPU] CPU Restart"}, + {.irq = IRQEXT_CLK, .name = "CLK", .desc = "[EXT] Clock Comparator"}, + {.irq = IRQEXT_EXC, .name = "EXC", .desc = "[EXT] External Call"}, + {.irq = IRQEXT_EMS, .name = "EMS", .desc = "[EXT] Emergency Signal"}, + {.irq = IRQEXT_TMR, .name = "TMR", .desc = "[EXT] CPU Timer"}, + {.irq = IRQEXT_TLA, .name = "TAL", .desc = "[EXT] Timing Alert"}, + {.irq = IRQEXT_PFL, .name = "PFL", .desc = "[EXT] Pseudo Page Fault"}, + {.irq = IRQEXT_DSD, .name = "DSD", .desc = "[EXT] DASD Diag"}, + {.irq = IRQEXT_VRT, .name = "VRT", .desc = "[EXT] Virtio"}, + {.irq = IRQEXT_SCP, .name = "SCP", .desc = "[EXT] Service Call"}, + {.irq = IRQEXT_IUC, .name = "IUC", .desc = "[EXT] IUCV"}, + {.irq = IRQEXT_CMS, .name = "CMS", .desc = "[EXT] CPU-Measurement: Sampling"}, + {.irq = IRQEXT_CMC, .name = "CMC", .desc = "[EXT] CPU-Measurement: Counter"}, + {.irq = IRQEXT_CMR, .name = "CMR", .desc = "[EXT] CPU-Measurement: RI"}, + {.irq = IRQEXT_FTP, .name = "FTP", .desc = "[EXT] HMC FTP Service"}, + {.irq = IRQIO_CIO, .name = "CIO", .desc = "[I/O] Common I/O Layer Interrupt"}, + {.irq = IRQIO_QAI, .name = "QAI", .desc = "[I/O] QDIO Adapter Interrupt"}, + {.irq = IRQIO_DAS, .name = "DAS", .desc = "[I/O] DASD"}, + {.irq = IRQIO_C15, .name = "C15", .desc = "[I/O] 3215"}, + {.irq = IRQIO_C70, .name = "C70", .desc = "[I/O] 3270"}, + {.irq = IRQIO_TAP, .name = "TAP", .desc = "[I/O] Tape"}, + {.irq = IRQIO_VMR, .name = "VMR", .desc = "[I/O] Unit Record Devices"}, + {.irq = IRQIO_LCS, .name = "LCS", .desc = "[I/O] LCS"}, + {.irq = IRQIO_CLW, .name = "CLW", .desc = "[I/O] CLAW"}, + {.irq = IRQIO_CTC, .name = "CTC", .desc = "[I/O] CTC"}, + {.irq = IRQIO_APB, .name = "APB", .desc = "[I/O] AP Bus"}, + {.irq = IRQIO_ADM, .name = "ADM", .desc = "[I/O] EADM Subchannel"}, + {.irq = IRQIO_CSC, .name = "CSC", .desc = "[I/O] CHSC Subchannel"}, + {.irq = IRQIO_PCI, .name = "PCI", .desc = "[I/O] PCI Interrupt" }, + {.irq = IRQIO_MSI, .name = "MSI", .desc = "[I/O] MSI Interrupt" }, + {.irq = IRQIO_VIR, .name = "VIR", .desc = "[I/O] Virtual I/O Devices"}, + {.irq = IRQIO_VAI, .name = "VAI", .desc = "[I/O] Virtual I/O Devices AI"}, + {.irq = NMI_NMI, .name = "NMI", .desc = "[NMI] Machine Check"}, + {.irq = CPU_RST, .name = "RST", .desc = "[CPU] CPU Restart"}, }; void __init init_IRQ(void) @@ -116,33 +118,37 @@ void do_IRQ(struct pt_regs *regs, int irq) */ int show_interrupts(struct seq_file *p, void *v) { - int irq = *(loff_t *) v; - int cpu; + int index = *(loff_t *) v; + int cpu, irq; get_online_cpus(); - if (irq == 0) { + if (index == 0) { seq_puts(p, " "); for_each_online_cpu(cpu) seq_printf(p, "CPU%d ", cpu); seq_putc(p, '\n'); goto out; } - if (irq < NR_IRQS) { - if (irq >= NR_IRQS_BASE) + if (index < NR_IRQS) { + if (index >= NR_IRQS_BASE) goto out; - seq_printf(p, "%s: ", irqclass_main_desc[irq].name); + /* Adjust index to process irqclass_main_desc array entries */ + index--; + seq_printf(p, "%s: ", irqclass_main_desc[index].name); + irq = irqclass_main_desc[index].irq; for_each_online_cpu(cpu) seq_printf(p, "%10u ", kstat_irqs_cpu(irq, cpu)); seq_putc(p, '\n'); goto out; } - for (irq = 0; irq < NR_ARCH_IRQS; irq++) { - seq_printf(p, "%s: ", irqclass_sub_desc[irq].name); + for (index = 0; index < NR_ARCH_IRQS; index++) { + seq_printf(p, "%s: ", irqclass_sub_desc[index].name); + irq = irqclass_sub_desc[index].irq; for_each_online_cpu(cpu) seq_printf(p, "%10u ", per_cpu(irq_stat, cpu).irqs[irq]); - if (irqclass_sub_desc[irq].desc) - seq_printf(p, " %s", irqclass_sub_desc[irq].desc); + if (irqclass_sub_desc[index].desc) + seq_printf(p, " %s", irqclass_sub_desc[index].desc); seq_putc(p, '\n'); } out: @@ -253,7 +259,7 @@ static irqreturn_t do_ext_interrupt(int irq, void *dummy) ext_code = *(struct ext_code *) ®s->int_code; if (ext_code.code != EXT_IRQ_CLK_COMP) - __get_cpu_var(s390_idle).nohz_delay = 1; + set_cpu_flag(CIF_NOHZ_DELAY); index = ext_hash(ext_code.code); rcu_read_lock(); diff --git a/arch/s390/kernel/kprobes.c b/arch/s390/kernel/kprobes.c index bc71a7b..014d472 100644 --- a/arch/s390/kernel/kprobes.c +++ b/arch/s390/kernel/kprobes.c @@ -58,161 +58,13 @@ struct kprobe_insn_cache kprobe_dmainsn_slots = { .insn_size = MAX_INSN_SIZE, }; -static int __kprobes is_prohibited_opcode(kprobe_opcode_t *insn) -{ - if (!is_known_insn((unsigned char *)insn)) - return -EINVAL; - switch (insn[0] >> 8) { - case 0x0c: /* bassm */ - case 0x0b: /* bsm */ - case 0x83: /* diag */ - case 0x44: /* ex */ - case 0xac: /* stnsm */ - case 0xad: /* stosm */ - return -EINVAL; - case 0xc6: - switch (insn[0] & 0x0f) { - case 0x00: /* exrl */ - return -EINVAL; - } - } - switch (insn[0]) { - case 0x0101: /* pr */ - case 0xb25a: /* bsa */ - case 0xb240: /* bakr */ - case 0xb258: /* bsg */ - case 0xb218: /* pc */ - case 0xb228: /* pt */ - case 0xb98d: /* epsw */ - return -EINVAL; - } - return 0; -} - -static int __kprobes get_fixup_type(kprobe_opcode_t *insn) -{ - /* default fixup method */ - int fixup = FIXUP_PSW_NORMAL; - - switch (insn[0] >> 8) { - case 0x05: /* balr */ - case 0x0d: /* basr */ - fixup = FIXUP_RETURN_REGISTER; - /* if r2 = 0, no branch will be taken */ - if ((insn[0] & 0x0f) == 0) - fixup |= FIXUP_BRANCH_NOT_TAKEN; - break; - case 0x06: /* bctr */ - case 0x07: /* bcr */ - fixup = FIXUP_BRANCH_NOT_TAKEN; - break; - case 0x45: /* bal */ - case 0x4d: /* bas */ - fixup = FIXUP_RETURN_REGISTER; - break; - case 0x47: /* bc */ - case 0x46: /* bct */ - case 0x86: /* bxh */ - case 0x87: /* bxle */ - fixup = FIXUP_BRANCH_NOT_TAKEN; - break; - case 0x82: /* lpsw */ - fixup = FIXUP_NOT_REQUIRED; - break; - case 0xb2: /* lpswe */ - if ((insn[0] & 0xff) == 0xb2) - fixup = FIXUP_NOT_REQUIRED; - break; - case 0xa7: /* bras */ - if ((insn[0] & 0x0f) == 0x05) - fixup |= FIXUP_RETURN_REGISTER; - break; - case 0xc0: - if ((insn[0] & 0x0f) == 0x05) /* brasl */ - fixup |= FIXUP_RETURN_REGISTER; - break; - case 0xeb: - switch (insn[2] & 0xff) { - case 0x44: /* bxhg */ - case 0x45: /* bxleg */ - fixup = FIXUP_BRANCH_NOT_TAKEN; - break; - } - break; - case 0xe3: /* bctg */ - if ((insn[2] & 0xff) == 0x46) - fixup = FIXUP_BRANCH_NOT_TAKEN; - break; - case 0xec: - switch (insn[2] & 0xff) { - case 0xe5: /* clgrb */ - case 0xe6: /* cgrb */ - case 0xf6: /* crb */ - case 0xf7: /* clrb */ - case 0xfc: /* cgib */ - case 0xfd: /* cglib */ - case 0xfe: /* cib */ - case 0xff: /* clib */ - fixup = FIXUP_BRANCH_NOT_TAKEN; - break; - } - break; - } - return fixup; -} - -static int __kprobes is_insn_relative_long(kprobe_opcode_t *insn) -{ - /* Check if we have a RIL-b or RIL-c format instruction which - * we need to modify in order to avoid instruction emulation. */ - switch (insn[0] >> 8) { - case 0xc0: - if ((insn[0] & 0x0f) == 0x00) /* larl */ - return true; - break; - case 0xc4: - switch (insn[0] & 0x0f) { - case 0x02: /* llhrl */ - case 0x04: /* lghrl */ - case 0x05: /* lhrl */ - case 0x06: /* llghrl */ - case 0x07: /* sthrl */ - case 0x08: /* lgrl */ - case 0x0b: /* stgrl */ - case 0x0c: /* lgfrl */ - case 0x0d: /* lrl */ - case 0x0e: /* llgfrl */ - case 0x0f: /* strl */ - return true; - } - break; - case 0xc6: - switch (insn[0] & 0x0f) { - case 0x02: /* pfdrl */ - case 0x04: /* cghrl */ - case 0x05: /* chrl */ - case 0x06: /* clghrl */ - case 0x07: /* clhrl */ - case 0x08: /* cgrl */ - case 0x0a: /* clgrl */ - case 0x0c: /* cgfrl */ - case 0x0d: /* crl */ - case 0x0e: /* clgfrl */ - case 0x0f: /* clrl */ - return true; - } - break; - } - return false; -} - static void __kprobes copy_instruction(struct kprobe *p) { s64 disp, new_disp; u64 addr, new_addr; memcpy(p->ainsn.insn, p->addr, insn_length(p->opcode >> 8)); - if (!is_insn_relative_long(p->ainsn.insn)) + if (!probe_is_insn_relative_long(p->ainsn.insn)) return; /* * For pc-relative instructions in RIL-b or RIL-c format patch the @@ -276,7 +128,7 @@ int __kprobes arch_prepare_kprobe(struct kprobe *p) if ((unsigned long) p->addr & 0x01) return -EINVAL; /* Make sure the probe isn't going on a difficult instruction */ - if (is_prohibited_opcode(p->addr)) + if (probe_is_prohibited_opcode(p->addr)) return -EINVAL; if (s390_get_insn_slot(p)) return -ENOMEM; @@ -366,9 +218,9 @@ static void __kprobes disable_singlestep(struct kprobe_ctlblk *kcb, */ static void __kprobes push_kprobe(struct kprobe_ctlblk *kcb, struct kprobe *p) { - kcb->prev_kprobe.kp = __get_cpu_var(current_kprobe); + kcb->prev_kprobe.kp = __this_cpu_read(current_kprobe); kcb->prev_kprobe.status = kcb->kprobe_status; - __get_cpu_var(current_kprobe) = p; + __this_cpu_write(current_kprobe, p); } /* @@ -378,7 +230,7 @@ static void __kprobes push_kprobe(struct kprobe_ctlblk *kcb, struct kprobe *p) */ static void __kprobes pop_kprobe(struct kprobe_ctlblk *kcb) { - __get_cpu_var(current_kprobe) = kcb->prev_kprobe.kp; + __this_cpu_write(current_kprobe, kcb->prev_kprobe.kp); kcb->kprobe_status = kcb->prev_kprobe.status; } @@ -459,7 +311,7 @@ static int __kprobes kprobe_handler(struct pt_regs *regs) enable_singlestep(kcb, regs, (unsigned long) p->ainsn.insn); return 1; } else if (kprobe_running()) { - p = __get_cpu_var(current_kprobe); + p = __this_cpu_read(current_kprobe); if (p->break_handler && p->break_handler(p, regs)) { /* * Continuation after the jprobe completed and @@ -605,7 +457,7 @@ static void __kprobes resume_execution(struct kprobe *p, struct pt_regs *regs) { struct kprobe_ctlblk *kcb = get_kprobe_ctlblk(); unsigned long ip = regs->psw.addr & PSW_ADDR_INSN; - int fixup = get_fixup_type(p->ainsn.insn); + int fixup = probe_get_fixup_type(p->ainsn.insn); if (fixup & FIXUP_PSW_NORMAL) ip += (unsigned long) p->addr - (unsigned long) p->ainsn.insn; @@ -789,11 +641,6 @@ void __kprobes jprobe_return(void) asm volatile(".word 0x0002"); } -static void __used __kprobes jprobe_return_end(void) -{ - asm volatile("bcr 0,0"); -} - int __kprobes longjmp_break_handler(struct kprobe *p, struct pt_regs *regs) { struct kprobe_ctlblk *kcb = get_kprobe_ctlblk(); diff --git a/arch/s390/kernel/machine_kexec.c b/arch/s390/kernel/machine_kexec.c index 719e27b..4685337 100644 --- a/arch/s390/kernel/machine_kexec.c +++ b/arch/s390/kernel/machine_kexec.c @@ -25,6 +25,7 @@ #include <asm/elf.h> #include <asm/asm-offsets.h> #include <asm/os_info.h> +#include <asm/switch_to.h> typedef void (*relocate_kernel_t)(kimage_entry_t *, unsigned long); @@ -43,7 +44,7 @@ static void add_elf_notes(int cpu) memcpy((void *) (4608UL + sa->pref_reg), sa, sizeof(*sa)); ptr = (u64 *) per_cpu_ptr(crash_notes, cpu); - ptr = fill_cpu_elf_notes(ptr, sa); + ptr = fill_cpu_elf_notes(ptr, sa, NULL); memset(ptr, 0, sizeof(struct elf_note)); } @@ -53,8 +54,11 @@ static void add_elf_notes(int cpu) static void setup_regs(void) { unsigned long sa = S390_lowcore.prefixreg_save_area + SAVE_AREA_BASE; + struct _lowcore *lc; int cpu, this_cpu; + /* Get lowcore pointer from store status of this CPU (absolute zero) */ + lc = (struct _lowcore *)(unsigned long)S390_lowcore.prefixreg_save_area; this_cpu = smp_find_processor_id(stap()); add_elf_notes(this_cpu); for_each_online_cpu(cpu) { @@ -64,6 +68,8 @@ static void setup_regs(void) continue; add_elf_notes(cpu); } + if (MACHINE_HAS_VX) + save_vx_regs_safe((void *) lc->vector_save_area_addr); /* Copy dump CPU store status info to absolute zero */ memcpy((void *) SAVE_AREA_BASE, (void *) sa, sizeof(struct save_area)); } diff --git a/arch/s390/kernel/mcount.S b/arch/s390/kernel/mcount.S index 08dcf21..4300ea3 100644 --- a/arch/s390/kernel/mcount.S +++ b/arch/s390/kernel/mcount.S @@ -8,66 +8,72 @@ #include <linux/linkage.h> #include <asm/asm-offsets.h> #include <asm/ftrace.h> +#include <asm/ptrace.h> .section .kprobes.text, "ax" ENTRY(ftrace_stub) br %r14 +#define STACK_FRAME_SIZE (STACK_FRAME_OVERHEAD + __PT_SIZE) +#define STACK_PTREGS (STACK_FRAME_OVERHEAD) +#define STACK_PTREGS_GPRS (STACK_PTREGS + __PT_GPRS) +#define STACK_PTREGS_PSW (STACK_PTREGS + __PT_PSW) + ENTRY(_mcount) -#ifdef CONFIG_DYNAMIC_FTRACE br %r14 ENTRY(ftrace_caller) + .globl ftrace_regs_caller + .set ftrace_regs_caller,ftrace_caller + lgr %r1,%r15 + aghi %r15,-STACK_FRAME_SIZE + stg %r1,__SF_BACKCHAIN(%r15) + stg %r1,(STACK_PTREGS_GPRS+15*8)(%r15) + stg %r0,(STACK_PTREGS_PSW+8)(%r15) + stmg %r2,%r14,(STACK_PTREGS_GPRS+2*8)(%r15) +#ifdef CONFIG_HAVE_MARCH_Z196_FEATURES + aghik %r2,%r0,-MCOUNT_INSN_SIZE + lgrl %r4,function_trace_op + lgrl %r1,ftrace_trace_function +#else + lgr %r2,%r0 + aghi %r2,-MCOUNT_INSN_SIZE + larl %r4,function_trace_op + lg %r4,0(%r4) + larl %r1,ftrace_trace_function + lg %r1,0(%r1) #endif - stm %r2,%r5,16(%r15) - bras %r1,2f -0: .long ftrace_trace_function -1: .long function_trace_stop -2: l %r2,1b-0b(%r1) - icm %r2,0xf,0(%r2) - jnz 3f - st %r14,56(%r15) - lr %r0,%r15 - ahi %r15,-96 - l %r3,100(%r15) - la %r2,0(%r14) - st %r0,__SF_BACKCHAIN(%r15) - la %r3,0(%r3) - ahi %r2,-MCOUNT_INSN_SIZE - l %r14,0b-0b(%r1) - l %r14,0(%r14) - basr %r14,%r14 + lgr %r3,%r14 + la %r5,STACK_PTREGS(%r15) + basr %r14,%r1 #ifdef CONFIG_FUNCTION_GRAPH_TRACER - l %r2,100(%r15) - l %r3,152(%r15) +# The j instruction gets runtime patched to a nop instruction. +# See ftrace_enable_ftrace_graph_caller. ENTRY(ftrace_graph_caller) -# The bras instruction gets runtime patched to call prepare_ftrace_return. -# See ftrace_enable_ftrace_graph_caller. The patched instruction is: -# bras %r14,prepare_ftrace_return - bras %r14,0f -0: st %r2,100(%r15) + j ftrace_graph_caller_end + lg %r2,(STACK_PTREGS_GPRS+14*8)(%r15) + lg %r3,(STACK_PTREGS_PSW+8)(%r15) + brasl %r14,prepare_ftrace_return + stg %r2,(STACK_PTREGS_GPRS+14*8)(%r15) +ftrace_graph_caller_end: + .globl ftrace_graph_caller_end #endif - ahi %r15,96 - l %r14,56(%r15) -3: lm %r2,%r5,16(%r15) - br %r14 + lg %r1,(STACK_PTREGS_PSW+8)(%r15) + lmg %r2,%r15,(STACK_PTREGS_GPRS+2*8)(%r15) + br %r1 #ifdef CONFIG_FUNCTION_GRAPH_TRACER ENTRY(return_to_handler) - stm %r2,%r5,16(%r15) - st %r14,56(%r15) - lr %r0,%r15 - ahi %r15,-96 - st %r0,__SF_BACKCHAIN(%r15) - bras %r1,0f - .long ftrace_return_to_handler -0: l %r2,0b-0b(%r1) - basr %r14,%r2 - lr %r14,%r2 - ahi %r15,96 - lm %r2,%r5,16(%r15) + stmg %r2,%r5,32(%r15) + lgr %r1,%r15 + aghi %r15,-STACK_FRAME_OVERHEAD + stg %r1,__SF_BACKCHAIN(%r15) + brasl %r14,ftrace_return_to_handler + aghi %r15,STACK_FRAME_OVERHEAD + lgr %r14,%r2 + lmg %r2,%r5,32(%r15) br %r14 #endif diff --git a/arch/s390/kernel/mcount64.S b/arch/s390/kernel/mcount64.S deleted file mode 100644 index 1c52eae..0000000 --- a/arch/s390/kernel/mcount64.S +++ /dev/null @@ -1,65 +0,0 @@ -/* - * Copyright IBM Corp. 2008, 2009 - * - * Author(s): Heiko Carstens <heiko.carstens@de.ibm.com>, - * - */ - -#include <linux/linkage.h> -#include <asm/asm-offsets.h> -#include <asm/ftrace.h> - - .section .kprobes.text, "ax" - -ENTRY(ftrace_stub) - br %r14 - -ENTRY(_mcount) -#ifdef CONFIG_DYNAMIC_FTRACE - br %r14 - -ENTRY(ftrace_caller) -#endif - larl %r1,function_trace_stop - icm %r1,0xf,0(%r1) - bnzr %r14 - stmg %r2,%r5,32(%r15) - stg %r14,112(%r15) - lgr %r1,%r15 - aghi %r15,-160 - stg %r1,__SF_BACKCHAIN(%r15) - lgr %r2,%r14 - lg %r3,168(%r15) - aghi %r2,-MCOUNT_INSN_SIZE - larl %r14,ftrace_trace_function - lg %r14,0(%r14) - basr %r14,%r14 -#ifdef CONFIG_FUNCTION_GRAPH_TRACER - lg %r2,168(%r15) - lg %r3,272(%r15) -ENTRY(ftrace_graph_caller) -# The bras instruction gets runtime patched to call prepare_ftrace_return. -# See ftrace_enable_ftrace_graph_caller. The patched instruction is: -# bras %r14,prepare_ftrace_return - bras %r14,0f -0: stg %r2,168(%r15) -#endif - aghi %r15,160 - lmg %r2,%r5,32(%r15) - lg %r14,112(%r15) - br %r14 - -#ifdef CONFIG_FUNCTION_GRAPH_TRACER - -ENTRY(return_to_handler) - stmg %r2,%r5,32(%r15) - lgr %r1,%r15 - aghi %r15,-160 - stg %r1,__SF_BACKCHAIN(%r15) - brasl %r14,ftrace_return_to_handler - aghi %r15,160 - lgr %r14,%r2 - lmg %r2,%r5,32(%r15) - br %r14 - -#endif diff --git a/arch/s390/kernel/nmi.c b/arch/s390/kernel/nmi.c index 210e128..dd1c24c 100644 --- a/arch/s390/kernel/nmi.c +++ b/arch/s390/kernel/nmi.c @@ -20,6 +20,7 @@ #include <asm/cputime.h> #include <asm/nmi.h> #include <asm/crw.h> +#include <asm/switch_to.h> struct mcck_struct { int kill_task; @@ -53,8 +54,12 @@ void s390_handle_mcck(void) */ local_irq_save(flags); local_mcck_disable(); - mcck = __get_cpu_var(cpu_mcck); - memset(&__get_cpu_var(cpu_mcck), 0, sizeof(struct mcck_struct)); + /* + * Ummm... Does this make sense at all? Copying the percpu struct + * and then zapping it one statement later? + */ + memcpy(&mcck, this_cpu_ptr(&cpu_mcck), sizeof(mcck)); + memset(&mcck, 0, sizeof(struct mcck_struct)); clear_cpu_flag(CIF_MCCK_PENDING); local_mcck_enable(); local_irq_restore(flags); @@ -163,6 +168,21 @@ static int notrace s390_revalidate_registers(struct mci *mci) " ld 15,120(%0)\n" : : "a" (fpt_save_area)); } + +#ifdef CONFIG_64BIT + /* Revalidate vector registers */ + if (MACHINE_HAS_VX && current->thread.vxrs) { + if (!mci->vr) { + /* + * Vector registers can't be restored and therefore + * the process needs to be terminated. + */ + kill_task = 1; + } + restore_vx_regs((__vector128 *) + S390_lowcore.vector_save_area_addr); + } +#endif /* Revalidate access registers */ asm volatile( " lam 0,15,0(%0)" @@ -253,7 +273,7 @@ void notrace s390_do_machine_check(struct pt_regs *regs) nmi_enter(); inc_irq_stat(NMI_NMI); mci = (struct mci *) &S390_lowcore.mcck_interruption_code; - mcck = &__get_cpu_var(cpu_mcck); + mcck = this_cpu_ptr(&cpu_mcck); umode = user_mode(regs); if (mci->sd) { diff --git a/arch/s390/kernel/perf_cpum_cf.c b/arch/s390/kernel/perf_cpum_cf.c index ea75d01..56fdad4 100644 --- a/arch/s390/kernel/perf_cpum_cf.c +++ b/arch/s390/kernel/perf_cpum_cf.c @@ -173,7 +173,7 @@ static int validate_ctr_auth(const struct hw_perf_event *hwc) */ static void cpumf_pmu_enable(struct pmu *pmu) { - struct cpu_hw_events *cpuhw = &__get_cpu_var(cpu_hw_events); + struct cpu_hw_events *cpuhw = this_cpu_ptr(&cpu_hw_events); int err; if (cpuhw->flags & PMU_F_ENABLED) @@ -196,7 +196,7 @@ static void cpumf_pmu_enable(struct pmu *pmu) */ static void cpumf_pmu_disable(struct pmu *pmu) { - struct cpu_hw_events *cpuhw = &__get_cpu_var(cpu_hw_events); + struct cpu_hw_events *cpuhw = this_cpu_ptr(&cpu_hw_events); int err; u64 inactive; @@ -230,7 +230,7 @@ static void cpumf_measurement_alert(struct ext_code ext_code, return; inc_irq_stat(IRQEXT_CMC); - cpuhw = &__get_cpu_var(cpu_hw_events); + cpuhw = this_cpu_ptr(&cpu_hw_events); /* Measurement alerts are shared and might happen when the PMU * is not reserved. Ignore these alerts in this case. */ @@ -250,7 +250,7 @@ static void cpumf_measurement_alert(struct ext_code ext_code, #define PMC_RELEASE 1 static void setup_pmc_cpu(void *flags) { - struct cpu_hw_events *cpuhw = &__get_cpu_var(cpu_hw_events); + struct cpu_hw_events *cpuhw = this_cpu_ptr(&cpu_hw_events); switch (*((int *) flags)) { case PMC_INIT: @@ -411,12 +411,6 @@ static int cpumf_pmu_event_init(struct perf_event *event) case PERF_TYPE_HARDWARE: case PERF_TYPE_HW_CACHE: case PERF_TYPE_RAW: - /* The CPU measurement counter facility does not have overflow - * interrupts to do sampling. Sampling must be provided by - * external means, for example, by timers. - */ - if (is_sampling_event(event)) - return -ENOENT; err = __hw_perf_event_init(event); break; default: @@ -481,7 +475,7 @@ static void cpumf_pmu_read(struct perf_event *event) static void cpumf_pmu_start(struct perf_event *event, int flags) { - struct cpu_hw_events *cpuhw = &__get_cpu_var(cpu_hw_events); + struct cpu_hw_events *cpuhw = this_cpu_ptr(&cpu_hw_events); struct hw_perf_event *hwc = &event->hw; if (WARN_ON_ONCE(!(hwc->state & PERF_HES_STOPPED))) @@ -512,7 +506,7 @@ static void cpumf_pmu_start(struct perf_event *event, int flags) static void cpumf_pmu_stop(struct perf_event *event, int flags) { - struct cpu_hw_events *cpuhw = &__get_cpu_var(cpu_hw_events); + struct cpu_hw_events *cpuhw = this_cpu_ptr(&cpu_hw_events); struct hw_perf_event *hwc = &event->hw; if (!(hwc->state & PERF_HES_STOPPED)) { @@ -533,7 +527,7 @@ static void cpumf_pmu_stop(struct perf_event *event, int flags) static int cpumf_pmu_add(struct perf_event *event, int flags) { - struct cpu_hw_events *cpuhw = &__get_cpu_var(cpu_hw_events); + struct cpu_hw_events *cpuhw = this_cpu_ptr(&cpu_hw_events); /* Check authorization for the counter set to which this * counter belongs. @@ -557,7 +551,7 @@ static int cpumf_pmu_add(struct perf_event *event, int flags) static void cpumf_pmu_del(struct perf_event *event, int flags) { - struct cpu_hw_events *cpuhw = &__get_cpu_var(cpu_hw_events); + struct cpu_hw_events *cpuhw = this_cpu_ptr(&cpu_hw_events); cpumf_pmu_stop(event, PERF_EF_UPDATE); @@ -581,7 +575,7 @@ static void cpumf_pmu_del(struct perf_event *event, int flags) */ static void cpumf_pmu_start_txn(struct pmu *pmu) { - struct cpu_hw_events *cpuhw = &__get_cpu_var(cpu_hw_events); + struct cpu_hw_events *cpuhw = this_cpu_ptr(&cpu_hw_events); perf_pmu_disable(pmu); cpuhw->flags |= PERF_EVENT_TXN; @@ -595,7 +589,7 @@ static void cpumf_pmu_start_txn(struct pmu *pmu) */ static void cpumf_pmu_cancel_txn(struct pmu *pmu) { - struct cpu_hw_events *cpuhw = &__get_cpu_var(cpu_hw_events); + struct cpu_hw_events *cpuhw = this_cpu_ptr(&cpu_hw_events); WARN_ON(cpuhw->tx_state != cpuhw->state); @@ -610,7 +604,7 @@ static void cpumf_pmu_cancel_txn(struct pmu *pmu) */ static int cpumf_pmu_commit_txn(struct pmu *pmu) { - struct cpu_hw_events *cpuhw = &__get_cpu_var(cpu_hw_events); + struct cpu_hw_events *cpuhw = this_cpu_ptr(&cpu_hw_events); u64 state; /* check if the updated state can be scheduled */ @@ -681,6 +675,12 @@ static int __init cpumf_pmu_init(void) goto out; } + /* The CPU measurement counter facility does not have overflow + * interrupts to do sampling. Sampling must be provided by + * external means, for example, by timers. + */ + cpumf_pmu.capabilities |= PERF_PMU_CAP_NO_INTERRUPT; + cpumf_pmu.attr_groups = cpumf_cf_event_group(); rc = perf_pmu_register(&cpumf_pmu, "cpum_cf", PERF_TYPE_RAW); if (rc) { diff --git a/arch/s390/kernel/perf_cpum_sf.c b/arch/s390/kernel/perf_cpum_sf.c index ea0c7b2..08e7613 100644 --- a/arch/s390/kernel/perf_cpum_sf.c +++ b/arch/s390/kernel/perf_cpum_sf.c @@ -562,7 +562,7 @@ static DEFINE_MUTEX(pmc_reserve_mutex); static void setup_pmc_cpu(void *flags) { int err; - struct cpu_hw_sf *cpusf = &__get_cpu_var(cpu_hw_sf); + struct cpu_hw_sf *cpusf = this_cpu_ptr(&cpu_hw_sf); err = 0; switch (*((int *) flags)) { @@ -849,7 +849,7 @@ static int cpumsf_pmu_event_init(struct perf_event *event) static void cpumsf_pmu_enable(struct pmu *pmu) { - struct cpu_hw_sf *cpuhw = &__get_cpu_var(cpu_hw_sf); + struct cpu_hw_sf *cpuhw = this_cpu_ptr(&cpu_hw_sf); struct hw_perf_event *hwc; int err; @@ -898,7 +898,7 @@ static void cpumsf_pmu_enable(struct pmu *pmu) static void cpumsf_pmu_disable(struct pmu *pmu) { - struct cpu_hw_sf *cpuhw = &__get_cpu_var(cpu_hw_sf); + struct cpu_hw_sf *cpuhw = this_cpu_ptr(&cpu_hw_sf); struct hws_lsctl_request_block inactive; struct hws_qsi_info_block si; int err; @@ -1306,7 +1306,7 @@ static void cpumsf_pmu_read(struct perf_event *event) */ static void cpumsf_pmu_start(struct perf_event *event, int flags) { - struct cpu_hw_sf *cpuhw = &__get_cpu_var(cpu_hw_sf); + struct cpu_hw_sf *cpuhw = this_cpu_ptr(&cpu_hw_sf); if (WARN_ON_ONCE(!(event->hw.state & PERF_HES_STOPPED))) return; @@ -1327,7 +1327,7 @@ static void cpumsf_pmu_start(struct perf_event *event, int flags) */ static void cpumsf_pmu_stop(struct perf_event *event, int flags) { - struct cpu_hw_sf *cpuhw = &__get_cpu_var(cpu_hw_sf); + struct cpu_hw_sf *cpuhw = this_cpu_ptr(&cpu_hw_sf); if (event->hw.state & PERF_HES_STOPPED) return; @@ -1346,7 +1346,7 @@ static void cpumsf_pmu_stop(struct perf_event *event, int flags) static int cpumsf_pmu_add(struct perf_event *event, int flags) { - struct cpu_hw_sf *cpuhw = &__get_cpu_var(cpu_hw_sf); + struct cpu_hw_sf *cpuhw = this_cpu_ptr(&cpu_hw_sf); int err; if (cpuhw->flags & PMU_F_IN_USE) @@ -1397,7 +1397,7 @@ out: static void cpumsf_pmu_del(struct perf_event *event, int flags) { - struct cpu_hw_sf *cpuhw = &__get_cpu_var(cpu_hw_sf); + struct cpu_hw_sf *cpuhw = this_cpu_ptr(&cpu_hw_sf); perf_pmu_disable(event->pmu); cpumsf_pmu_stop(event, PERF_EF_UPDATE); @@ -1470,7 +1470,7 @@ static void cpumf_measurement_alert(struct ext_code ext_code, if (!(alert & CPU_MF_INT_SF_MASK)) return; inc_irq_stat(IRQEXT_CMS); - cpuhw = &__get_cpu_var(cpu_hw_sf); + cpuhw = this_cpu_ptr(&cpu_hw_sf); /* Measurement alerts are shared and might happen when the PMU * is not reserved. Ignore these alerts in this case. */ diff --git a/arch/s390/kernel/pgm_check.S b/arch/s390/kernel/pgm_check.S index 813ec72..f6f8886 100644 --- a/arch/s390/kernel/pgm_check.S +++ b/arch/s390/kernel/pgm_check.S @@ -49,7 +49,7 @@ PGM_CHECK_DEFAULT /* 17 */ PGM_CHECK_64BIT(transaction_exception) /* 18 */ PGM_CHECK_DEFAULT /* 19 */ PGM_CHECK_DEFAULT /* 1a */ -PGM_CHECK_DEFAULT /* 1b */ +PGM_CHECK_64BIT(vector_exception) /* 1b */ PGM_CHECK(space_switch_exception) /* 1c */ PGM_CHECK(hfp_sqrt_exception) /* 1d */ PGM_CHECK_DEFAULT /* 1e */ diff --git a/arch/s390/kernel/process.c b/arch/s390/kernel/process.c index 93b9ca4..ed84cc2 100644 --- a/arch/s390/kernel/process.c +++ b/arch/s390/kernel/process.c @@ -61,30 +61,6 @@ unsigned long thread_saved_pc(struct task_struct *tsk) return sf->gprs[8]; } -void arch_cpu_idle(void) -{ - local_mcck_disable(); - if (test_cpu_flag(CIF_MCCK_PENDING)) { - local_mcck_enable(); - local_irq_enable(); - return; - } - /* Halt the cpu and keep track of cpu time accounting. */ - vtime_stop_cpu(); - local_irq_enable(); -} - -void arch_cpu_idle_exit(void) -{ - if (test_cpu_flag(CIF_MCCK_PENDING)) - s390_handle_mcck(); -} - -void arch_cpu_idle_dead(void) -{ - cpu_die(); -} - extern void __kprobes kernel_thread_starter(void); /* diff --git a/arch/s390/kernel/processor.c b/arch/s390/kernel/processor.c index 2461202..dbdd33e 100644 --- a/arch/s390/kernel/processor.c +++ b/arch/s390/kernel/processor.c @@ -23,15 +23,13 @@ static DEFINE_PER_CPU(struct cpuid, cpu_id); */ void cpu_init(void) { - struct s390_idle_data *idle = &__get_cpu_var(s390_idle); - struct cpuid *id = &__get_cpu_var(cpu_id); + struct cpuid *id = this_cpu_ptr(&cpu_id); get_cpu_id(id); atomic_inc(&init_mm.mm_count); current->active_mm = &init_mm; BUG_ON(current->mm); enter_lazy_tlb(&init_mm, current); - memset(idle, 0, sizeof(*idle)); } /* @@ -41,7 +39,7 @@ static int show_cpuinfo(struct seq_file *m, void *v) { static const char *hwcap_str[] = { "esan3", "zarch", "stfle", "msa", "ldisp", "eimm", "dfp", - "edat", "etf3eh", "highgprs", "te" + "edat", "etf3eh", "highgprs", "te", "vx" }; unsigned long n = (unsigned long) v - 1; int i; diff --git a/arch/s390/kernel/ptrace.c b/arch/s390/kernel/ptrace.c index 910f253..99a567b 100644 --- a/arch/s390/kernel/ptrace.c +++ b/arch/s390/kernel/ptrace.c @@ -38,15 +38,6 @@ #define CREATE_TRACE_POINTS #include <trace/events/syscalls.h> -enum s390_regset { - REGSET_GENERAL, - REGSET_FP, - REGSET_LAST_BREAK, - REGSET_TDB, - REGSET_SYSTEM_CALL, - REGSET_GENERAL_EXTENDED, -}; - void update_cr_regs(struct task_struct *task) { struct pt_regs *regs = task_pt_regs(task); @@ -55,27 +46,39 @@ void update_cr_regs(struct task_struct *task) #ifdef CONFIG_64BIT /* Take care of the enable/disable of transactional execution. */ - if (MACHINE_HAS_TE) { + if (MACHINE_HAS_TE || MACHINE_HAS_VX) { unsigned long cr, cr_new; __ctl_store(cr, 0, 0); - /* Set or clear transaction execution TXC bit 8. */ - cr_new = cr | (1UL << 55); - if (task->thread.per_flags & PER_FLAG_NO_TE) - cr_new &= ~(1UL << 55); + cr_new = cr; + if (MACHINE_HAS_TE) { + /* Set or clear transaction execution TXC bit 8. */ + cr_new |= (1UL << 55); + if (task->thread.per_flags & PER_FLAG_NO_TE) + cr_new &= ~(1UL << 55); + } + if (MACHINE_HAS_VX) { + /* Enable/disable of vector extension */ + cr_new &= ~(1UL << 17); + if (task->thread.vxrs) + cr_new |= (1UL << 17); + } if (cr_new != cr) __ctl_load(cr_new, 0, 0); - /* Set or clear transaction execution TDC bits 62 and 63. */ - __ctl_store(cr, 2, 2); - cr_new = cr & ~3UL; - if (task->thread.per_flags & PER_FLAG_TE_ABORT_RAND) { - if (task->thread.per_flags & PER_FLAG_TE_ABORT_RAND_TEND) - cr_new |= 1UL; - else - cr_new |= 2UL; + if (MACHINE_HAS_TE) { + /* Set/clear transaction execution TDC bits 62/63. */ + __ctl_store(cr, 2, 2); + cr_new = cr & ~3UL; + if (task->thread.per_flags & PER_FLAG_TE_ABORT_RAND) { + if (task->thread.per_flags & + PER_FLAG_TE_ABORT_RAND_TEND) + cr_new |= 1UL; + else + cr_new |= 2UL; + } + if (cr_new != cr) + __ctl_load(cr_new, 2, 2); } - if (cr_new != cr) - __ctl_load(cr_new, 2, 2); } #endif /* Copy user specified PER registers */ @@ -84,7 +87,8 @@ void update_cr_regs(struct task_struct *task) new.end = thread->per_user.end; /* merge TIF_SINGLE_STEP into user specified PER registers. */ - if (test_tsk_thread_flag(task, TIF_SINGLE_STEP)) { + if (test_tsk_thread_flag(task, TIF_SINGLE_STEP) || + test_tsk_thread_flag(task, TIF_UPROBE_SINGLESTEP)) { if (test_tsk_thread_flag(task, TIF_BLOCK_STEP)) new.control |= PER_EVENT_BRANCH; else @@ -93,6 +97,8 @@ void update_cr_regs(struct task_struct *task) new.control |= PER_CONTROL_SUSPENSION; new.control |= PER_EVENT_TRANSACTION_END; #endif + if (test_tsk_thread_flag(task, TIF_UPROBE_SINGLESTEP)) + new.control |= PER_EVENT_IFETCH; new.start = 0; new.end = PSW_ADDR_INSN; } @@ -803,7 +809,7 @@ asmlinkage long do_syscall_trace_enter(struct pt_regs *regs) long ret = 0; /* Do the secure computing check first. */ - if (secure_computing(regs->gprs[2])) { + if (secure_computing()) { /* seccomp failures shouldn't expose any additional code. */ ret = -1; goto out; @@ -921,7 +927,15 @@ static int s390_fpregs_get(struct task_struct *target, save_fp_ctl(&target->thread.fp_regs.fpc); save_fp_regs(target->thread.fp_regs.fprs); } +#ifdef CONFIG_64BIT + else if (target->thread.vxrs) { + int i; + for (i = 0; i < __NUM_VXRS_LOW; i++) + target->thread.fp_regs.fprs[i] = + *(freg_t *)(target->thread.vxrs + i); + } +#endif return user_regset_copyout(&pos, &count, &kbuf, &ubuf, &target->thread.fp_regs, 0, -1); } @@ -955,9 +969,20 @@ static int s390_fpregs_set(struct task_struct *target, target->thread.fp_regs.fprs, offsetof(s390_fp_regs, fprs), -1); - if (rc == 0 && target == current) { - restore_fp_ctl(&target->thread.fp_regs.fpc); - restore_fp_regs(target->thread.fp_regs.fprs); + if (rc == 0) { + if (target == current) { + restore_fp_ctl(&target->thread.fp_regs.fpc); + restore_fp_regs(target->thread.fp_regs.fprs); + } +#ifdef CONFIG_64BIT + else if (target->thread.vxrs) { + int i; + + for (i = 0; i < __NUM_VXRS_LOW; i++) + *(freg_t *)(target->thread.vxrs + i) = + target->thread.fp_regs.fprs[i]; + } +#endif } return rc; @@ -1013,6 +1038,95 @@ static int s390_tdb_set(struct task_struct *target, return 0; } +static int s390_vxrs_active(struct task_struct *target, + const struct user_regset *regset) +{ + return !!target->thread.vxrs; +} + +static int s390_vxrs_low_get(struct task_struct *target, + const struct user_regset *regset, + unsigned int pos, unsigned int count, + void *kbuf, void __user *ubuf) +{ + __u64 vxrs[__NUM_VXRS_LOW]; + int i; + + if (target->thread.vxrs) { + if (target == current) + save_vx_regs(target->thread.vxrs); + for (i = 0; i < __NUM_VXRS_LOW; i++) + vxrs[i] = *((__u64 *)(target->thread.vxrs + i) + 1); + } else + memset(vxrs, 0, sizeof(vxrs)); + return user_regset_copyout(&pos, &count, &kbuf, &ubuf, vxrs, 0, -1); +} + +static int s390_vxrs_low_set(struct task_struct *target, + const struct user_regset *regset, + unsigned int pos, unsigned int count, + const void *kbuf, const void __user *ubuf) +{ + __u64 vxrs[__NUM_VXRS_LOW]; + int i, rc; + + if (!target->thread.vxrs) { + rc = alloc_vector_registers(target); + if (rc) + return rc; + } else if (target == current) + save_vx_regs(target->thread.vxrs); + + rc = user_regset_copyin(&pos, &count, &kbuf, &ubuf, vxrs, 0, -1); + if (rc == 0) { + for (i = 0; i < __NUM_VXRS_LOW; i++) + *((__u64 *)(target->thread.vxrs + i) + 1) = vxrs[i]; + if (target == current) + restore_vx_regs(target->thread.vxrs); + } + + return rc; +} + +static int s390_vxrs_high_get(struct task_struct *target, + const struct user_regset *regset, + unsigned int pos, unsigned int count, + void *kbuf, void __user *ubuf) +{ + __vector128 vxrs[__NUM_VXRS_HIGH]; + + if (target->thread.vxrs) { + if (target == current) + save_vx_regs(target->thread.vxrs); + memcpy(vxrs, target->thread.vxrs + __NUM_VXRS_LOW, + sizeof(vxrs)); + } else + memset(vxrs, 0, sizeof(vxrs)); + return user_regset_copyout(&pos, &count, &kbuf, &ubuf, vxrs, 0, -1); +} + +static int s390_vxrs_high_set(struct task_struct *target, + const struct user_regset *regset, + unsigned int pos, unsigned int count, + const void *kbuf, const void __user *ubuf) +{ + int rc; + + if (!target->thread.vxrs) { + rc = alloc_vector_registers(target); + if (rc) + return rc; + } else if (target == current) + save_vx_regs(target->thread.vxrs); + + rc = user_regset_copyin(&pos, &count, &kbuf, &ubuf, + target->thread.vxrs + __NUM_VXRS_LOW, 0, -1); + if (rc == 0 && target == current) + restore_vx_regs(target->thread.vxrs); + + return rc; +} + #endif static int s390_system_call_get(struct task_struct *target, @@ -1036,7 +1150,7 @@ static int s390_system_call_set(struct task_struct *target, } static const struct user_regset s390_regsets[] = { - [REGSET_GENERAL] = { + { .core_note_type = NT_PRSTATUS, .n = sizeof(s390_regs) / sizeof(long), .size = sizeof(long), @@ -1044,7 +1158,7 @@ static const struct user_regset s390_regsets[] = { .get = s390_regs_get, .set = s390_regs_set, }, - [REGSET_FP] = { + { .core_note_type = NT_PRFPREG, .n = sizeof(s390_fp_regs) / sizeof(long), .size = sizeof(long), @@ -1052,8 +1166,16 @@ static const struct user_regset s390_regsets[] = { .get = s390_fpregs_get, .set = s390_fpregs_set, }, + { + .core_note_type = NT_S390_SYSTEM_CALL, + .n = 1, + .size = sizeof(unsigned int), + .align = sizeof(unsigned int), + .get = s390_system_call_get, + .set = s390_system_call_set, + }, #ifdef CONFIG_64BIT - [REGSET_LAST_BREAK] = { + { .core_note_type = NT_S390_LAST_BREAK, .n = 1, .size = sizeof(long), @@ -1061,7 +1183,7 @@ static const struct user_regset s390_regsets[] = { .get = s390_last_break_get, .set = s390_last_break_set, }, - [REGSET_TDB] = { + { .core_note_type = NT_S390_TDB, .n = 1, .size = 256, @@ -1069,15 +1191,25 @@ static const struct user_regset s390_regsets[] = { .get = s390_tdb_get, .set = s390_tdb_set, }, -#endif - [REGSET_SYSTEM_CALL] = { - .core_note_type = NT_S390_SYSTEM_CALL, - .n = 1, - .size = sizeof(unsigned int), - .align = sizeof(unsigned int), - .get = s390_system_call_get, - .set = s390_system_call_set, + { + .core_note_type = NT_S390_VXRS_LOW, + .n = __NUM_VXRS_LOW, + .size = sizeof(__u64), + .align = sizeof(__u64), + .active = s390_vxrs_active, + .get = s390_vxrs_low_get, + .set = s390_vxrs_low_set, }, + { + .core_note_type = NT_S390_VXRS_HIGH, + .n = __NUM_VXRS_HIGH, + .size = sizeof(__vector128), + .align = sizeof(__vector128), + .active = s390_vxrs_active, + .get = s390_vxrs_high_get, + .set = s390_vxrs_high_set, + }, +#endif }; static const struct user_regset_view user_s390_view = { @@ -1242,7 +1374,7 @@ static int s390_compat_last_break_set(struct task_struct *target, } static const struct user_regset s390_compat_regsets[] = { - [REGSET_GENERAL] = { + { .core_note_type = NT_PRSTATUS, .n = sizeof(s390_compat_regs) / sizeof(compat_long_t), .size = sizeof(compat_long_t), @@ -1250,7 +1382,7 @@ static const struct user_regset s390_compat_regsets[] = { .get = s390_compat_regs_get, .set = s390_compat_regs_set, }, - [REGSET_FP] = { + { .core_note_type = NT_PRFPREG, .n = sizeof(s390_fp_regs) / sizeof(compat_long_t), .size = sizeof(compat_long_t), @@ -1258,7 +1390,15 @@ static const struct user_regset s390_compat_regsets[] = { .get = s390_fpregs_get, .set = s390_fpregs_set, }, - [REGSET_LAST_BREAK] = { + { + .core_note_type = NT_S390_SYSTEM_CALL, + .n = 1, + .size = sizeof(compat_uint_t), + .align = sizeof(compat_uint_t), + .get = s390_system_call_get, + .set = s390_system_call_set, + }, + { .core_note_type = NT_S390_LAST_BREAK, .n = 1, .size = sizeof(long), @@ -1266,7 +1406,7 @@ static const struct user_regset s390_compat_regsets[] = { .get = s390_compat_last_break_get, .set = s390_compat_last_break_set, }, - [REGSET_TDB] = { + { .core_note_type = NT_S390_TDB, .n = 1, .size = 256, @@ -1274,15 +1414,25 @@ static const struct user_regset s390_compat_regsets[] = { .get = s390_tdb_get, .set = s390_tdb_set, }, - [REGSET_SYSTEM_CALL] = { - .core_note_type = NT_S390_SYSTEM_CALL, - .n = 1, - .size = sizeof(compat_uint_t), - .align = sizeof(compat_uint_t), - .get = s390_system_call_get, - .set = s390_system_call_set, + { + .core_note_type = NT_S390_VXRS_LOW, + .n = __NUM_VXRS_LOW, + .size = sizeof(__u64), + .align = sizeof(__u64), + .active = s390_vxrs_active, + .get = s390_vxrs_low_get, + .set = s390_vxrs_low_set, + }, + { + .core_note_type = NT_S390_VXRS_HIGH, + .n = __NUM_VXRS_HIGH, + .size = sizeof(__vector128), + .align = sizeof(__vector128), + .active = s390_vxrs_active, + .get = s390_vxrs_high_get, + .set = s390_vxrs_high_set, }, - [REGSET_GENERAL_EXTENDED] = { + { .core_note_type = NT_S390_HIGH_GPRS, .n = sizeof(s390_compat_regs_high) / sizeof(compat_long_t), .size = sizeof(compat_long_t), diff --git a/arch/s390/kernel/setup.c b/arch/s390/kernel/setup.c index 1e2264b..e80d9ff 100644 --- a/arch/s390/kernel/setup.c +++ b/arch/s390/kernel/setup.c @@ -24,6 +24,7 @@ #include <linux/stddef.h> #include <linux/unistd.h> #include <linux/ptrace.h> +#include <linux/random.h> #include <linux/user.h> #include <linux/tty.h> #include <linux/ioport.h> @@ -61,6 +62,7 @@ #include <asm/diag.h> #include <asm/os_info.h> #include <asm/sclp.h> +#include <asm/sysinfo.h> #include "entry.h" /* @@ -341,6 +343,9 @@ static void __init setup_lowcore(void) __ctl_set_bit(14, 29); } #else + if (MACHINE_HAS_VX) + lc->vector_save_area_addr = + (unsigned long) &lc->vector_save_area; lc->vdso_per_cpu_data = (unsigned long) &lc->paste[0]; #endif lc->sync_enter_timer = S390_lowcore.sync_enter_timer; @@ -450,8 +455,8 @@ static void __init setup_memory_end(void) #ifdef CONFIG_64BIT vmalloc_size = VMALLOC_END ?: (128UL << 30) - MODULES_LEN; tmp = (memory_end ?: max_physmem_end) / PAGE_SIZE; - tmp = tmp * (sizeof(struct page) + PAGE_SIZE) + vmalloc_size; - if (tmp <= (1UL << 42)) + tmp = tmp * (sizeof(struct page) + PAGE_SIZE); + if (tmp + vmalloc_size + MODULES_LEN <= (1UL << 42)) vmax = 1UL << 42; /* 3-level kernel page table */ else vmax = 1UL << 53; /* 4-level kernel page table */ @@ -501,6 +506,8 @@ static int kdump_mem_notifier(struct notifier_block *nb, { struct memory_notify *arg = data; + if (action != MEM_GOING_OFFLINE) + return NOTIFY_OK; if (arg->start_pfn < PFN_DOWN(resource_size(&crashk_res))) return NOTIFY_BAD; if (arg->start_pfn > PFN_DOWN(crashk_res.end)) @@ -761,9 +768,16 @@ static void __init setup_hwcaps(void) */ if (test_facility(50) && test_facility(73)) elf_hwcap |= HWCAP_S390_TE; + + /* + * Vector extension HWCAP_S390_VXRS is bit 11. + */ + if (test_facility(129)) + elf_hwcap |= HWCAP_S390_VXRS; #endif get_cpu_id(&cpu_id); + add_device_randomness(&cpu_id, sizeof(cpu_id)); switch (cpu_id.machine) { case 0x9672: #if !defined(CONFIG_64BIT) @@ -802,6 +816,19 @@ static void __init setup_hwcaps(void) } /* + * Add system information as device randomness + */ +static void __init setup_randomness(void) +{ + struct sysinfo_3_2_2 *vmms; + + vmms = (struct sysinfo_3_2_2 *) alloc_page(GFP_KERNEL); + if (vmms && stsi(vmms, 3, 2, 2) == 0 && vmms->count) + add_device_randomness(&vmms, vmms->count); + free_page((unsigned long) vmms); +} + +/* * Setup function called from init/main.c just after the banner * was printed. */ @@ -899,6 +926,9 @@ void __init setup_arch(char **cmdline_p) /* Setup zfcpdump support */ setup_zfcpdump(); + + /* Add system specific data to the random pool */ + setup_randomness(); } #ifdef CONFIG_32BIT diff --git a/arch/s390/kernel/signal.c b/arch/s390/kernel/signal.c index 42b49f9..0c1a0ff 100644 --- a/arch/s390/kernel/signal.c +++ b/arch/s390/kernel/signal.c @@ -31,30 +31,117 @@ #include <asm/switch_to.h> #include "entry.h" -typedef struct +/* + * Layout of an old-style signal-frame: + * ----------------------------------------- + * | save area (_SIGNAL_FRAMESIZE) | + * ----------------------------------------- + * | struct sigcontext | + * | oldmask | + * | _sigregs * | + * ----------------------------------------- + * | _sigregs with | + * | _s390_regs_common | + * | _s390_fp_regs | + * ----------------------------------------- + * | int signo | + * ----------------------------------------- + * | _sigregs_ext with | + * | gprs_high 64 byte (opt) | + * | vxrs_low 128 byte (opt) | + * | vxrs_high 256 byte (opt) | + * | reserved 128 byte (opt) | + * ----------------------------------------- + * | __u16 svc_insn | + * ----------------------------------------- + * The svc_insn entry with the sigreturn system call opcode does not + * have a fixed position and moves if gprs_high or vxrs exist. + * Future extensions will be added to _sigregs_ext. + */ +struct sigframe { __u8 callee_used_stack[__SIGNAL_FRAMESIZE]; struct sigcontext sc; _sigregs sregs; int signo; - __u8 retcode[S390_SYSCALL_SIZE]; -} sigframe; + _sigregs_ext sregs_ext; + __u16 svc_insn; /* Offset of svc_insn is NOT fixed! */ +}; -typedef struct +/* + * Layout of an rt signal-frame: + * ----------------------------------------- + * | save area (_SIGNAL_FRAMESIZE) | + * ----------------------------------------- + * | svc __NR_rt_sigreturn 2 byte | + * ----------------------------------------- + * | struct siginfo | + * ----------------------------------------- + * | struct ucontext_extended with | + * | unsigned long uc_flags | + * | struct ucontext *uc_link | + * | stack_t uc_stack | + * | _sigregs uc_mcontext with | + * | _s390_regs_common | + * | _s390_fp_regs | + * | sigset_t uc_sigmask | + * | _sigregs_ext uc_mcontext_ext | + * | gprs_high 64 byte (opt) | + * | vxrs_low 128 byte (opt) | + * | vxrs_high 256 byte (opt)| + * | reserved 128 byte (opt) | + * ----------------------------------------- + * Future extensions will be added to _sigregs_ext. + */ +struct rt_sigframe { __u8 callee_used_stack[__SIGNAL_FRAMESIZE]; - __u8 retcode[S390_SYSCALL_SIZE]; + __u16 svc_insn; struct siginfo info; - struct ucontext uc; -} rt_sigframe; + struct ucontext_extended uc; +}; + +/* Store registers needed to create the signal frame */ +static void store_sigregs(void) +{ + save_access_regs(current->thread.acrs); + save_fp_ctl(¤t->thread.fp_regs.fpc); +#ifdef CONFIG_64BIT + if (current->thread.vxrs) { + int i; + + save_vx_regs(current->thread.vxrs); + for (i = 0; i < __NUM_FPRS; i++) + current->thread.fp_regs.fprs[i] = + *(freg_t *)(current->thread.vxrs + i); + } else +#endif + save_fp_regs(current->thread.fp_regs.fprs); +} + +/* Load registers after signal return */ +static void load_sigregs(void) +{ + restore_access_regs(current->thread.acrs); + /* restore_fp_ctl is done in restore_sigregs */ +#ifdef CONFIG_64BIT + if (current->thread.vxrs) { + int i; + + for (i = 0; i < __NUM_FPRS; i++) + *(freg_t *)(current->thread.vxrs + i) = + current->thread.fp_regs.fprs[i]; + restore_vx_regs(current->thread.vxrs); + } else +#endif + restore_fp_regs(current->thread.fp_regs.fprs); +} /* Returns non-zero on fault. */ static int save_sigregs(struct pt_regs *regs, _sigregs __user *sregs) { _sigregs user_sregs; - save_access_regs(current->thread.acrs); - /* Copy a 'clean' PSW mask to the user to avoid leaking information about whether PER is currently on. */ user_sregs.regs.psw.mask = PSW_USER_BITS | @@ -63,12 +150,6 @@ static int save_sigregs(struct pt_regs *regs, _sigregs __user *sregs) memcpy(&user_sregs.regs.gprs, ®s->gprs, sizeof(sregs->regs.gprs)); memcpy(&user_sregs.regs.acrs, current->thread.acrs, sizeof(user_sregs.regs.acrs)); - /* - * We have to store the fp registers to current->thread.fp_regs - * to merge them with the emulated registers. - */ - save_fp_ctl(¤t->thread.fp_regs.fpc); - save_fp_regs(current->thread.fp_regs.fprs); memcpy(&user_sregs.fpregs, ¤t->thread.fp_regs, sizeof(user_sregs.fpregs)); if (__copy_to_user(sregs, &user_sregs, sizeof(_sigregs))) @@ -107,20 +188,64 @@ static int restore_sigregs(struct pt_regs *regs, _sigregs __user *sregs) memcpy(®s->gprs, &user_sregs.regs.gprs, sizeof(sregs->regs.gprs)); memcpy(¤t->thread.acrs, &user_sregs.regs.acrs, sizeof(current->thread.acrs)); - restore_access_regs(current->thread.acrs); memcpy(¤t->thread.fp_regs, &user_sregs.fpregs, sizeof(current->thread.fp_regs)); - restore_fp_regs(current->thread.fp_regs.fprs); clear_pt_regs_flag(regs, PIF_SYSCALL); /* No longer in a system call */ return 0; } +/* Returns non-zero on fault. */ +static int save_sigregs_ext(struct pt_regs *regs, + _sigregs_ext __user *sregs_ext) +{ +#ifdef CONFIG_64BIT + __u64 vxrs[__NUM_VXRS_LOW]; + int i; + + /* Save vector registers to signal stack */ + if (current->thread.vxrs) { + for (i = 0; i < __NUM_VXRS_LOW; i++) + vxrs[i] = *((__u64 *)(current->thread.vxrs + i) + 1); + if (__copy_to_user(&sregs_ext->vxrs_low, vxrs, + sizeof(sregs_ext->vxrs_low)) || + __copy_to_user(&sregs_ext->vxrs_high, + current->thread.vxrs + __NUM_VXRS_LOW, + sizeof(sregs_ext->vxrs_high))) + return -EFAULT; + } +#endif + return 0; +} + +static int restore_sigregs_ext(struct pt_regs *regs, + _sigregs_ext __user *sregs_ext) +{ +#ifdef CONFIG_64BIT + __u64 vxrs[__NUM_VXRS_LOW]; + int i; + + /* Restore vector registers from signal stack */ + if (current->thread.vxrs) { + if (__copy_from_user(vxrs, &sregs_ext->vxrs_low, + sizeof(sregs_ext->vxrs_low)) || + __copy_from_user(current->thread.vxrs + __NUM_VXRS_LOW, + &sregs_ext->vxrs_high, + sizeof(sregs_ext->vxrs_high))) + return -EFAULT; + for (i = 0; i < __NUM_VXRS_LOW; i++) + *((__u64 *)(current->thread.vxrs + i) + 1) = vxrs[i]; + } +#endif + return 0; +} + SYSCALL_DEFINE0(sigreturn) { struct pt_regs *regs = task_pt_regs(current); - sigframe __user *frame = (sigframe __user *)regs->gprs[15]; + struct sigframe __user *frame = + (struct sigframe __user *) regs->gprs[15]; sigset_t set; if (__copy_from_user(&set.sig, &frame->sc.oldmask, _SIGMASK_COPY_SIZE)) @@ -128,6 +253,9 @@ SYSCALL_DEFINE0(sigreturn) set_current_blocked(&set); if (restore_sigregs(regs, &frame->sregs)) goto badframe; + if (restore_sigregs_ext(regs, &frame->sregs_ext)) + goto badframe; + load_sigregs(); return regs->gprs[2]; badframe: force_sig(SIGSEGV, current); @@ -137,16 +265,20 @@ badframe: SYSCALL_DEFINE0(rt_sigreturn) { struct pt_regs *regs = task_pt_regs(current); - rt_sigframe __user *frame = (rt_sigframe __user *)regs->gprs[15]; + struct rt_sigframe __user *frame = + (struct rt_sigframe __user *)regs->gprs[15]; sigset_t set; if (__copy_from_user(&set.sig, &frame->uc.uc_sigmask, sizeof(set))) goto badframe; set_current_blocked(&set); + if (restore_altstack(&frame->uc.uc_stack)) + goto badframe; if (restore_sigregs(regs, &frame->uc.uc_mcontext)) goto badframe; - if (restore_altstack(&frame->uc.uc_stack)) + if (restore_sigregs_ext(regs, &frame->uc.uc_mcontext_ext)) goto badframe; + load_sigregs(); return regs->gprs[2]; badframe: force_sig(SIGSEGV, current); @@ -154,11 +286,6 @@ badframe: } /* - * Set up a signal frame. - */ - - -/* * Determine which stack to use.. */ static inline void __user * @@ -195,39 +322,63 @@ static inline int map_signal(int sig) static int setup_frame(int sig, struct k_sigaction *ka, sigset_t *set, struct pt_regs * regs) { - sigframe __user *frame; - - frame = get_sigframe(ka, regs, sizeof(sigframe)); + struct sigframe __user *frame; + struct sigcontext sc; + unsigned long restorer; + size_t frame_size; + /* + * gprs_high are only present for a 31-bit task running on + * a 64-bit kernel (see compat_signal.c) but the space for + * gprs_high need to be allocated if vector registers are + * included in the signal frame on a 31-bit system. + */ + frame_size = sizeof(*frame) - sizeof(frame->sregs_ext); + if (MACHINE_HAS_VX) + frame_size += sizeof(frame->sregs_ext); + frame = get_sigframe(ka, regs, frame_size); if (frame == (void __user *) -1UL) - goto give_sigsegv; + return -EFAULT; - if (__copy_to_user(&frame->sc.oldmask, &set->sig, _SIGMASK_COPY_SIZE)) - goto give_sigsegv; + /* Set up backchain. */ + if (__put_user(regs->gprs[15], (addr_t __user *) frame)) + return -EFAULT; + + /* Create struct sigcontext on the signal stack */ + memcpy(&sc.oldmask, &set->sig, _SIGMASK_COPY_SIZE); + sc.sregs = (_sigregs __user __force *) &frame->sregs; + if (__copy_to_user(&frame->sc, &sc, sizeof(frame->sc))) + return -EFAULT; + + /* Store registers needed to create the signal frame */ + store_sigregs(); + /* Create _sigregs on the signal stack */ if (save_sigregs(regs, &frame->sregs)) - goto give_sigsegv; - if (__put_user(&frame->sregs, &frame->sc.sregs)) - goto give_sigsegv; + return -EFAULT; + + /* Place signal number on stack to allow backtrace from handler. */ + if (__put_user(regs->gprs[2], (int __user *) &frame->signo)) + return -EFAULT; + + /* Create _sigregs_ext on the signal stack */ + if (save_sigregs_ext(regs, &frame->sregs_ext)) + return -EFAULT; /* Set up to return from userspace. If provided, use a stub already in userspace. */ if (ka->sa.sa_flags & SA_RESTORER) { - regs->gprs[14] = (unsigned long) - ka->sa.sa_restorer | PSW_ADDR_AMODE; + restorer = (unsigned long) ka->sa.sa_restorer | PSW_ADDR_AMODE; } else { - regs->gprs[14] = (unsigned long) - frame->retcode | PSW_ADDR_AMODE; - if (__put_user(S390_SYSCALL_OPCODE | __NR_sigreturn, - (u16 __user *)(frame->retcode))) - goto give_sigsegv; + /* Signal frame without vector registers are short ! */ + __u16 __user *svc = (void *) frame + frame_size - 2; + if (__put_user(S390_SYSCALL_OPCODE | __NR_sigreturn, svc)) + return -EFAULT; + restorer = (unsigned long) svc | PSW_ADDR_AMODE; } - /* Set up backchain. */ - if (__put_user(regs->gprs[15], (addr_t __user *) frame)) - goto give_sigsegv; - /* Set up registers for signal handler */ + regs->gprs[14] = restorer; regs->gprs[15] = (unsigned long) frame; /* Force default amode and default user address space control. */ regs->psw.mask = PSW_MASK_EA | PSW_MASK_BA | @@ -247,91 +398,95 @@ static int setup_frame(int sig, struct k_sigaction *ka, regs->gprs[5] = regs->int_parm_long; regs->gprs[6] = task_thread_info(current)->last_break; } - - /* Place signal number on stack to allow backtrace from handler. */ - if (__put_user(regs->gprs[2], (int __user *) &frame->signo)) - goto give_sigsegv; return 0; - -give_sigsegv: - force_sigsegv(sig, current); - return -EFAULT; } -static int setup_rt_frame(int sig, struct k_sigaction *ka, siginfo_t *info, - sigset_t *set, struct pt_regs * regs) +static int setup_rt_frame(struct ksignal *ksig, sigset_t *set, + struct pt_regs *regs) { - int err = 0; - rt_sigframe __user *frame; - - frame = get_sigframe(ka, regs, sizeof(rt_sigframe)); + struct rt_sigframe __user *frame; + unsigned long uc_flags, restorer; + size_t frame_size; + frame_size = sizeof(struct rt_sigframe) - sizeof(_sigregs_ext); + /* + * gprs_high are only present for a 31-bit task running on + * a 64-bit kernel (see compat_signal.c) but the space for + * gprs_high need to be allocated if vector registers are + * included in the signal frame on a 31-bit system. + */ + uc_flags = 0; +#ifdef CONFIG_64BIT + if (MACHINE_HAS_VX) { + frame_size += sizeof(_sigregs_ext); + if (current->thread.vxrs) + uc_flags |= UC_VXRS; + } +#endif + frame = get_sigframe(&ksig->ka, regs, frame_size); if (frame == (void __user *) -1UL) - goto give_sigsegv; - - if (copy_siginfo_to_user(&frame->info, info)) - goto give_sigsegv; + return -EFAULT; - /* Create the ucontext. */ - err |= __put_user(0, &frame->uc.uc_flags); - err |= __put_user(NULL, &frame->uc.uc_link); - err |= __save_altstack(&frame->uc.uc_stack, regs->gprs[15]); - err |= save_sigregs(regs, &frame->uc.uc_mcontext); - err |= __copy_to_user(&frame->uc.uc_sigmask, set, sizeof(*set)); - if (err) - goto give_sigsegv; + /* Set up backchain. */ + if (__put_user(regs->gprs[15], (addr_t __user *) frame)) + return -EFAULT; /* Set up to return from userspace. If provided, use a stub already in userspace. */ - if (ka->sa.sa_flags & SA_RESTORER) { - regs->gprs[14] = (unsigned long) - ka->sa.sa_restorer | PSW_ADDR_AMODE; + if (ksig->ka.sa.sa_flags & SA_RESTORER) { + restorer = (unsigned long) + ksig->ka.sa.sa_restorer | PSW_ADDR_AMODE; } else { - regs->gprs[14] = (unsigned long) - frame->retcode | PSW_ADDR_AMODE; - if (__put_user(S390_SYSCALL_OPCODE | __NR_rt_sigreturn, - (u16 __user *)(frame->retcode))) - goto give_sigsegv; + __u16 __user *svc = &frame->svc_insn; + if (__put_user(S390_SYSCALL_OPCODE | __NR_rt_sigreturn, svc)) + return -EFAULT; + restorer = (unsigned long) svc | PSW_ADDR_AMODE; } - /* Set up backchain. */ - if (__put_user(regs->gprs[15], (addr_t __user *) frame)) - goto give_sigsegv; + /* Create siginfo on the signal stack */ + if (copy_siginfo_to_user(&frame->info, &ksig->info)) + return -EFAULT; + + /* Store registers needed to create the signal frame */ + store_sigregs(); + + /* Create ucontext on the signal stack. */ + if (__put_user(uc_flags, &frame->uc.uc_flags) || + __put_user(NULL, &frame->uc.uc_link) || + __save_altstack(&frame->uc.uc_stack, regs->gprs[15]) || + save_sigregs(regs, &frame->uc.uc_mcontext) || + __copy_to_user(&frame->uc.uc_sigmask, set, sizeof(*set)) || + save_sigregs_ext(regs, &frame->uc.uc_mcontext_ext)) + return -EFAULT; /* Set up registers for signal handler */ + regs->gprs[14] = restorer; regs->gprs[15] = (unsigned long) frame; /* Force default amode and default user address space control. */ regs->psw.mask = PSW_MASK_EA | PSW_MASK_BA | (PSW_USER_BITS & PSW_MASK_ASC) | (regs->psw.mask & ~PSW_MASK_ASC); - regs->psw.addr = (unsigned long) ka->sa.sa_handler | PSW_ADDR_AMODE; + regs->psw.addr = (unsigned long) ksig->ka.sa.sa_handler | PSW_ADDR_AMODE; - regs->gprs[2] = map_signal(sig); + regs->gprs[2] = map_signal(ksig->sig); regs->gprs[3] = (unsigned long) &frame->info; regs->gprs[4] = (unsigned long) &frame->uc; regs->gprs[5] = task_thread_info(current)->last_break; return 0; - -give_sigsegv: - force_sigsegv(sig, current); - return -EFAULT; } -static void handle_signal(unsigned long sig, struct k_sigaction *ka, - siginfo_t *info, sigset_t *oldset, - struct pt_regs *regs) +static void handle_signal(struct ksignal *ksig, sigset_t *oldset, + struct pt_regs *regs) { int ret; /* Set up the stack frame */ - if (ka->sa.sa_flags & SA_SIGINFO) - ret = setup_rt_frame(sig, ka, info, oldset, regs); + if (ksig->ka.sa.sa_flags & SA_SIGINFO) + ret = setup_rt_frame(ksig, oldset, regs); else - ret = setup_frame(sig, ka, oldset, regs); - if (ret) - return; - signal_delivered(sig, info, ka, regs, - test_thread_flag(TIF_SINGLE_STEP)); + ret = setup_frame(ksig->sig, &ksig->ka, oldset, regs); + + signal_setup_done(ret, ksig, test_thread_flag(TIF_SINGLE_STEP)); } /* @@ -345,9 +500,7 @@ static void handle_signal(unsigned long sig, struct k_sigaction *ka, */ void do_signal(struct pt_regs *regs) { - siginfo_t info; - int signr; - struct k_sigaction ka; + struct ksignal ksig; sigset_t *oldset = sigmask_to_save(); /* @@ -357,9 +510,8 @@ void do_signal(struct pt_regs *regs) */ current_thread_info()->system_call = test_pt_regs_flag(regs, PIF_SYSCALL) ? regs->int_code : 0; - signr = get_signal_to_deliver(&info, &ka, regs, NULL); - if (signr > 0) { + if (get_signal(&ksig)) { /* Whee! Actually deliver the signal. */ if (current_thread_info()->system_call) { regs->int_code = current_thread_info()->system_call; @@ -370,7 +522,7 @@ void do_signal(struct pt_regs *regs) regs->gprs[2] = -EINTR; break; case -ERESTARTSYS: - if (!(ka.sa.sa_flags & SA_RESTART)) { + if (!(ksig.ka.sa.sa_flags & SA_RESTART)) { regs->gprs[2] = -EINTR; break; } @@ -387,9 +539,9 @@ void do_signal(struct pt_regs *regs) clear_pt_regs_flag(regs, PIF_SYSCALL); if (is_compat_task()) - handle_signal32(signr, &ka, &info, oldset, regs); + handle_signal32(&ksig, oldset, regs); else - handle_signal(signr, &ka, &info, oldset, regs); + handle_signal(&ksig, oldset, regs); return; } diff --git a/arch/s390/kernel/smp.c b/arch/s390/kernel/smp.c index 243c7e5..6fd9e60 100644 --- a/arch/s390/kernel/smp.c +++ b/arch/s390/kernel/smp.c @@ -45,6 +45,7 @@ #include <asm/debug.h> #include <asm/os_info.h> #include <asm/sigp.h> +#include <asm/idle.h> #include "entry.h" enum { @@ -82,7 +83,8 @@ DEFINE_MUTEX(smp_cpu_state_mutex); /* * Signal processor helper functions. */ -static inline int __pcpu_sigp_relax(u16 addr, u8 order, u32 parm, u32 *status) +static inline int __pcpu_sigp_relax(u16 addr, u8 order, unsigned long parm, + u32 *status) { int cc; @@ -178,6 +180,9 @@ static int pcpu_alloc_lowcore(struct pcpu *pcpu, int cpu) goto out; } #else + if (MACHINE_HAS_VX) + lc->vector_save_area_addr = + (unsigned long) &lc->vector_save_area; if (vdso_alloc_per_cpu(lc)) goto out; #endif @@ -333,12 +338,6 @@ int smp_vcpu_scheduled(int cpu) return pcpu_running(pcpu_devices + cpu); } -void smp_yield(void) -{ - if (MACHINE_HAS_DIAG44) - asm volatile("diag 0,0,0x44"); -} - void smp_yield_cpu(int cpu) { if (MACHINE_HAS_DIAG9C) @@ -517,35 +516,53 @@ EXPORT_SYMBOL(smp_ctl_clear_bit); static void __init smp_get_save_area(int cpu, u16 address) { void *lc = pcpu_devices[0].lowcore; - struct save_area *save_area; + struct save_area_ext *sa_ext; + unsigned long vx_sa; if (is_kdump_kernel()) return; if (!OLDMEM_BASE && (address == boot_cpu_address || ipl_info.type != IPL_TYPE_FCP_DUMP)) return; - save_area = dump_save_area_create(cpu); - if (!save_area) + sa_ext = dump_save_area_create(cpu); + if (!sa_ext) panic("could not allocate memory for save area\n"); if (address == boot_cpu_address) { /* Copy the registers of the boot cpu. */ - copy_oldmem_page(1, (void *) save_area, sizeof(*save_area), + copy_oldmem_page(1, (void *) &sa_ext->sa, sizeof(sa_ext->sa), SAVE_AREA_BASE - PAGE_SIZE, 0); + if (MACHINE_HAS_VX) + save_vx_regs_safe(sa_ext->vx_regs); return; } /* Get the registers of a non-boot cpu. */ __pcpu_sigp_relax(address, SIGP_STOP_AND_STORE_STATUS, 0, NULL); - memcpy_real(save_area, lc + SAVE_AREA_BASE, sizeof(*save_area)); + memcpy_real(&sa_ext->sa, lc + SAVE_AREA_BASE, sizeof(sa_ext->sa)); + if (!MACHINE_HAS_VX) + return; + /* Get the VX registers */ + vx_sa = __get_free_page(GFP_KERNEL); + if (!vx_sa) + panic("could not allocate memory for VX save area\n"); + __pcpu_sigp_relax(address, SIGP_STORE_ADDITIONAL_STATUS, vx_sa, NULL); + memcpy(sa_ext->vx_regs, (void *) vx_sa, sizeof(sa_ext->vx_regs)); + free_page(vx_sa); } int smp_store_status(int cpu) { + unsigned long vx_sa; struct pcpu *pcpu; pcpu = pcpu_devices + cpu; if (__pcpu_sigp_relax(pcpu->address, SIGP_STOP_AND_STORE_STATUS, 0, NULL) != SIGP_CC_ORDER_CODE_ACCEPTED) return -EIO; + if (!MACHINE_HAS_VX) + return 0; + vx_sa = __pa(pcpu->lowcore->vector_save_area_addr); + __pcpu_sigp_relax(pcpu->address, SIGP_STORE_ADDITIONAL_STATUS, + vx_sa, NULL); return 0; } @@ -667,7 +684,7 @@ static void smp_start_secondary(void *cpuvoid) cpu_init(); preempt_disable(); init_cpu_timer(); - init_cpu_vtimer(); + vtime_init(); pfault_init(); notify_cpu_starting(smp_processor_id()); set_cpu_online(smp_processor_id(), true); @@ -726,6 +743,7 @@ int __cpu_disable(void) cregs[6] &= ~0xff000000UL; /* disable all I/O interrupts */ cregs[14] &= ~0x1f000000UL; /* disable most machine checks */ __ctl_load(cregs, 0, 15); + clear_cpu_flag(CIF_NOHZ_DELAY); return 0; } @@ -898,42 +916,6 @@ static struct attribute_group cpu_common_attr_group = { .attrs = cpu_common_attrs, }; -static ssize_t show_idle_count(struct device *dev, - struct device_attribute *attr, char *buf) -{ - struct s390_idle_data *idle = &per_cpu(s390_idle, dev->id); - unsigned long long idle_count; - unsigned int sequence; - - do { - sequence = ACCESS_ONCE(idle->sequence); - idle_count = ACCESS_ONCE(idle->idle_count); - if (ACCESS_ONCE(idle->clock_idle_enter)) - idle_count++; - } while ((sequence & 1) || (ACCESS_ONCE(idle->sequence) != sequence)); - return sprintf(buf, "%llu\n", idle_count); -} -static DEVICE_ATTR(idle_count, 0444, show_idle_count, NULL); - -static ssize_t show_idle_time(struct device *dev, - struct device_attribute *attr, char *buf) -{ - struct s390_idle_data *idle = &per_cpu(s390_idle, dev->id); - unsigned long long now, idle_time, idle_enter, idle_exit; - unsigned int sequence; - - do { - now = get_tod_clock(); - sequence = ACCESS_ONCE(idle->sequence); - idle_time = ACCESS_ONCE(idle->idle_time); - idle_enter = ACCESS_ONCE(idle->clock_idle_enter); - idle_exit = ACCESS_ONCE(idle->clock_idle_exit); - } while ((sequence & 1) || (ACCESS_ONCE(idle->sequence) != sequence)); - idle_time += idle_enter ? ((idle_exit ? : now) - idle_enter) : 0; - return sprintf(buf, "%llu\n", idle_time >> 12); -} -static DEVICE_ATTR(idle_time_us, 0444, show_idle_time, NULL); - static struct attribute *cpu_online_attrs[] = { &dev_attr_idle_count.attr, &dev_attr_idle_time_us.attr, diff --git a/arch/s390/kernel/suspend.c b/arch/s390/kernel/suspend.c index a7a7537..1c4c5ac 100644 --- a/arch/s390/kernel/suspend.c +++ b/arch/s390/kernel/suspend.c @@ -13,14 +13,10 @@ #include <asm/ipl.h> #include <asm/cio.h> #include <asm/pci.h> +#include <asm/sections.h> #include "entry.h" /* - * References to section boundaries - */ -extern const void __nosave_begin, __nosave_end; - -/* * The restore of the saved pages in an hibernation image will set * the change and referenced bits in the storage key for each page. * Overindication of the referenced bits after an hibernation cycle diff --git a/arch/s390/kernel/syscalls.S b/arch/s390/kernel/syscalls.S index fe5cdf2..6fe886a 100644 --- a/arch/s390/kernel/syscalls.S +++ b/arch/s390/kernel/syscalls.S @@ -356,3 +356,6 @@ SYSCALL(sys_finit_module,sys_finit_module,compat_sys_finit_module) SYSCALL(sys_sched_setattr,sys_sched_setattr,compat_sys_sched_setattr) /* 345 */ SYSCALL(sys_sched_getattr,sys_sched_getattr,compat_sys_sched_getattr) SYSCALL(sys_renameat2,sys_renameat2,compat_sys_renameat2) +SYSCALL(sys_seccomp,sys_seccomp,compat_sys_seccomp) +SYSCALL(sys_getrandom,sys_getrandom,compat_sys_getrandom) +SYSCALL(sys_memfd_create,sys_memfd_create,compat_sys_memfd_create) /* 350 */ diff --git a/arch/s390/kernel/time.c b/arch/s390/kernel/time.c index 0931b11..005d665 100644 --- a/arch/s390/kernel/time.c +++ b/arch/s390/kernel/time.c @@ -92,7 +92,7 @@ void clock_comparator_work(void) struct clock_event_device *cd; S390_lowcore.clock_comparator = -1ULL; - cd = &__get_cpu_var(comparators); + cd = this_cpu_ptr(&comparators); cd->event_handler(cd); } @@ -214,26 +214,39 @@ void update_vsyscall(struct timekeeper *tk) { u64 nsecps; - if (tk->clock != &clocksource_tod) + if (tk->tkr.clock != &clocksource_tod) return; /* Make userspace gettimeofday spin until we're done. */ ++vdso_data->tb_update_count; smp_wmb(); - vdso_data->xtime_tod_stamp = tk->clock->cycle_last; + vdso_data->xtime_tod_stamp = tk->tkr.cycle_last; vdso_data->xtime_clock_sec = tk->xtime_sec; - vdso_data->xtime_clock_nsec = tk->xtime_nsec; + vdso_data->xtime_clock_nsec = tk->tkr.xtime_nsec; vdso_data->wtom_clock_sec = tk->xtime_sec + tk->wall_to_monotonic.tv_sec; - vdso_data->wtom_clock_nsec = tk->xtime_nsec + - + ((u64) tk->wall_to_monotonic.tv_nsec << tk->shift); - nsecps = (u64) NSEC_PER_SEC << tk->shift; + vdso_data->wtom_clock_nsec = tk->tkr.xtime_nsec + + + ((u64) tk->wall_to_monotonic.tv_nsec << tk->tkr.shift); + nsecps = (u64) NSEC_PER_SEC << tk->tkr.shift; while (vdso_data->wtom_clock_nsec >= nsecps) { vdso_data->wtom_clock_nsec -= nsecps; vdso_data->wtom_clock_sec++; } - vdso_data->tk_mult = tk->mult; - vdso_data->tk_shift = tk->shift; + + vdso_data->xtime_coarse_sec = tk->xtime_sec; + vdso_data->xtime_coarse_nsec = + (long)(tk->tkr.xtime_nsec >> tk->tkr.shift); + vdso_data->wtom_coarse_sec = + vdso_data->xtime_coarse_sec + tk->wall_to_monotonic.tv_sec; + vdso_data->wtom_coarse_nsec = + vdso_data->xtime_coarse_nsec + tk->wall_to_monotonic.tv_nsec; + while (vdso_data->wtom_coarse_nsec >= NSEC_PER_SEC) { + vdso_data->wtom_coarse_nsec -= NSEC_PER_SEC; + vdso_data->wtom_coarse_sec++; + } + + vdso_data->tk_mult = tk->tkr.mult; + vdso_data->tk_shift = tk->tkr.shift; smp_wmb(); ++vdso_data->tb_update_count; } @@ -360,7 +373,7 @@ EXPORT_SYMBOL(get_sync_clock); */ static void disable_sync_clock(void *dummy) { - atomic_t *sw_ptr = &__get_cpu_var(clock_sync_word); + atomic_t *sw_ptr = this_cpu_ptr(&clock_sync_word); /* * Clear the in-sync bit 2^31. All get_sync_clock calls will * fail until the sync bit is turned back on. In addition @@ -377,7 +390,7 @@ static void disable_sync_clock(void *dummy) */ static void enable_sync_clock(void) { - atomic_t *sw_ptr = &__get_cpu_var(clock_sync_word); + atomic_t *sw_ptr = this_cpu_ptr(&clock_sync_word); atomic_set_mask(0x80000000, sw_ptr); } diff --git a/arch/s390/kernel/topology.c b/arch/s390/kernel/topology.c index 355a16c..b93bed7 100644 --- a/arch/s390/kernel/topology.c +++ b/arch/s390/kernel/topology.c @@ -464,15 +464,17 @@ static struct sched_domain_topology_level s390_topology[] = { static int __init topology_init(void) { - if (!MACHINE_HAS_TOPOLOGY) { + if (MACHINE_HAS_TOPOLOGY) + set_topology_timer(); + else topology_update_polarization_simple(); - goto out; - } - set_topology_timer(); -out: - - set_sched_topology(s390_topology); - return device_create_file(cpu_subsys.dev_root, &dev_attr_dispatching); } device_initcall(topology_init); + +static int __init early_topology_init(void) +{ + set_sched_topology(s390_topology); + return 0; +} +early_initcall(early_topology_init); diff --git a/arch/s390/kernel/traps.c b/arch/s390/kernel/traps.c index c576232..9ff5ecb 100644 --- a/arch/s390/kernel/traps.c +++ b/arch/s390/kernel/traps.c @@ -18,6 +18,8 @@ #include <linux/ptrace.h> #include <linux/sched.h> #include <linux/mm.h> +#include <linux/slab.h> +#include <asm/switch_to.h> #include "entry.h" int show_unhandled_signals = 1; @@ -58,15 +60,10 @@ int is_valid_bugaddr(unsigned long addr) return 1; } -static void __kprobes do_trap(struct pt_regs *regs, - int si_signo, int si_code, char *str) +void do_report_trap(struct pt_regs *regs, int si_signo, int si_code, char *str) { siginfo_t info; - if (notify_die(DIE_TRAP, str, regs, 0, - regs->int_code, si_signo) == NOTIFY_STOP) - return; - if (user_mode(regs)) { info.si_signo = si_signo; info.si_errno = 0; @@ -90,6 +87,15 @@ static void __kprobes do_trap(struct pt_regs *regs, } } +static void __kprobes do_trap(struct pt_regs *regs, int si_signo, int si_code, + char *str) +{ + if (notify_die(DIE_TRAP, str, regs, 0, + regs->int_code, si_signo) == NOTIFY_STOP) + return; + do_report_trap(regs, si_signo, si_code, str); +} + void __kprobes do_per_trap(struct pt_regs *regs) { siginfo_t info; @@ -178,6 +184,7 @@ void __kprobes illegal_op(struct pt_regs *regs) siginfo_t info; __u8 opcode[6]; __u16 __user *location; + int is_uprobe_insn = 0; int signal = 0; location = get_trap_ip(regs); @@ -194,6 +201,10 @@ void __kprobes illegal_op(struct pt_regs *regs) force_sig_info(SIGTRAP, &info, current); } else signal = SIGILL; +#ifdef CONFIG_UPROBES + } else if (*((__u16 *) opcode) == UPROBE_SWBP_INSN) { + is_uprobe_insn = 1; +#endif #ifdef CONFIG_MATHEMU } else if (opcode[0] == 0xb3) { if (get_user(*((__u16 *) (opcode+2)), location+1)) @@ -219,11 +230,13 @@ void __kprobes illegal_op(struct pt_regs *regs) #endif } else signal = SIGILL; - } else { - /* - * If we get an illegal op in kernel mode, send it through the - * kprobes notifier. If kprobes doesn't pick it up, SIGILL - */ + } + /* + * We got either an illegal op in kernel mode, or user space trapped + * on a uprobes illegal instruction. See if kprobes or uprobes picks + * it up. If not, SIGILL. + */ + if (is_uprobe_insn || !user_mode(regs)) { if (notify_die(DIE_BPT, "bpt", regs, 0, 3, SIGTRAP) != NOTIFY_STOP) signal = SIGILL; @@ -292,6 +305,74 @@ DO_ERROR_INFO(specification_exception, SIGILL, ILL_ILLOPN, "specification exception"); #endif +#ifdef CONFIG_64BIT +int alloc_vector_registers(struct task_struct *tsk) +{ + __vector128 *vxrs; + int i; + + /* Allocate vector register save area. */ + vxrs = kzalloc(sizeof(__vector128) * __NUM_VXRS, + GFP_KERNEL|__GFP_REPEAT); + if (!vxrs) + return -ENOMEM; + preempt_disable(); + if (tsk == current) + save_fp_regs(tsk->thread.fp_regs.fprs); + /* Copy the 16 floating point registers */ + for (i = 0; i < 16; i++) + *(freg_t *) &vxrs[i] = tsk->thread.fp_regs.fprs[i]; + tsk->thread.vxrs = vxrs; + if (tsk == current) { + __ctl_set_bit(0, 17); + restore_vx_regs(vxrs); + } + preempt_enable(); + return 0; +} + +void vector_exception(struct pt_regs *regs) +{ + int si_code, vic; + + if (!MACHINE_HAS_VX) { + do_trap(regs, SIGILL, ILL_ILLOPN, "illegal operation"); + return; + } + + /* get vector interrupt code from fpc */ + asm volatile("stfpc %0" : "=m" (current->thread.fp_regs.fpc)); + vic = (current->thread.fp_regs.fpc & 0xf00) >> 8; + switch (vic) { + case 1: /* invalid vector operation */ + si_code = FPE_FLTINV; + break; + case 2: /* division by zero */ + si_code = FPE_FLTDIV; + break; + case 3: /* overflow */ + si_code = FPE_FLTOVF; + break; + case 4: /* underflow */ + si_code = FPE_FLTUND; + break; + case 5: /* inexact */ + si_code = FPE_FLTRES; + break; + default: /* unknown cause */ + si_code = 0; + } + do_trap(regs, SIGFPE, si_code, "vector exception"); +} + +static int __init disable_vector_extension(char *str) +{ + S390_lowcore.machine_flags &= ~MACHINE_FLAG_VX; + return 1; +} +__setup("novx", disable_vector_extension); +#endif + void data_exception(struct pt_regs *regs) { __u16 __user *location; @@ -357,6 +438,18 @@ void data_exception(struct pt_regs *regs) } } #endif +#ifdef CONFIG_64BIT + /* Check for vector register enablement */ + if (MACHINE_HAS_VX && !current->thread.vxrs && + (current->thread.fp_regs.fpc & FPC_DXC_MASK) == 0xfe00) { + alloc_vector_registers(current); + /* Vector data exception is suppressing, rewind psw. */ + regs->psw.addr = __rewind_psw(regs->psw, regs->int_code >> 16); + clear_pt_regs_flag(regs, PIF_PER_TRAP); + return; + } +#endif + if (current->thread.fp_regs.fpc & FPC_DXC_MASK) signal = SIGFPE; else diff --git a/arch/s390/kernel/uprobes.c b/arch/s390/kernel/uprobes.c new file mode 100644 index 0000000..956f4f7 --- /dev/null +++ b/arch/s390/kernel/uprobes.c @@ -0,0 +1,332 @@ +/* + * User-space Probes (UProbes) for s390 + * + * Copyright IBM Corp. 2014 + * Author(s): Jan Willeke, + */ + +#include <linux/kprobes.h> +#include <linux/uaccess.h> +#include <linux/uprobes.h> +#include <linux/compat.h> +#include <linux/kdebug.h> +#include <asm/switch_to.h> +#include <asm/facility.h> +#include <asm/dis.h> +#include "entry.h" + +#define UPROBE_TRAP_NR UINT_MAX + +int arch_uprobe_analyze_insn(struct arch_uprobe *auprobe, struct mm_struct *mm, + unsigned long addr) +{ + return probe_is_prohibited_opcode(auprobe->insn); +} + +int arch_uprobe_pre_xol(struct arch_uprobe *auprobe, struct pt_regs *regs) +{ + if (psw_bits(regs->psw).eaba == PSW_AMODE_24BIT) + return -EINVAL; + if (!is_compat_task() && psw_bits(regs->psw).eaba == PSW_AMODE_31BIT) + return -EINVAL; + clear_pt_regs_flag(regs, PIF_PER_TRAP); + auprobe->saved_per = psw_bits(regs->psw).r; + auprobe->saved_int_code = regs->int_code; + regs->int_code = UPROBE_TRAP_NR; + regs->psw.addr = current->utask->xol_vaddr; + set_tsk_thread_flag(current, TIF_UPROBE_SINGLESTEP); + update_cr_regs(current); + return 0; +} + +bool arch_uprobe_xol_was_trapped(struct task_struct *tsk) +{ + struct pt_regs *regs = task_pt_regs(tsk); + + if (regs->int_code != UPROBE_TRAP_NR) + return true; + return false; +} + +int arch_uprobe_post_xol(struct arch_uprobe *auprobe, struct pt_regs *regs) +{ + int fixup = probe_get_fixup_type(auprobe->insn); + struct uprobe_task *utask = current->utask; + + clear_tsk_thread_flag(current, TIF_UPROBE_SINGLESTEP); + update_cr_regs(current); + psw_bits(regs->psw).r = auprobe->saved_per; + regs->int_code = auprobe->saved_int_code; + + if (fixup & FIXUP_PSW_NORMAL) + regs->psw.addr += utask->vaddr - utask->xol_vaddr; + if (fixup & FIXUP_RETURN_REGISTER) { + int reg = (auprobe->insn[0] & 0xf0) >> 4; + + regs->gprs[reg] += utask->vaddr - utask->xol_vaddr; + } + if (fixup & FIXUP_BRANCH_NOT_TAKEN) { + int ilen = insn_length(auprobe->insn[0] >> 8); + + if (regs->psw.addr - utask->xol_vaddr == ilen) + regs->psw.addr = utask->vaddr + ilen; + } + /* If per tracing was active generate trap */ + if (regs->psw.mask & PSW_MASK_PER) + do_per_trap(regs); + return 0; +} + +int arch_uprobe_exception_notify(struct notifier_block *self, unsigned long val, + void *data) +{ + struct die_args *args = data; + struct pt_regs *regs = args->regs; + + if (!user_mode(regs)) + return NOTIFY_DONE; + if (regs->int_code & 0x200) /* Trap during transaction */ + return NOTIFY_DONE; + switch (val) { + case DIE_BPT: + if (uprobe_pre_sstep_notifier(regs)) + return NOTIFY_STOP; + break; + case DIE_SSTEP: + if (uprobe_post_sstep_notifier(regs)) + return NOTIFY_STOP; + default: + break; + } + return NOTIFY_DONE; +} + +void arch_uprobe_abort_xol(struct arch_uprobe *auprobe, struct pt_regs *regs) +{ + clear_thread_flag(TIF_UPROBE_SINGLESTEP); + regs->int_code = auprobe->saved_int_code; + regs->psw.addr = current->utask->vaddr; +} + +unsigned long arch_uretprobe_hijack_return_addr(unsigned long trampoline, + struct pt_regs *regs) +{ + unsigned long orig; + + orig = regs->gprs[14]; + regs->gprs[14] = trampoline; + return orig; +} + +/* Instruction Emulation */ + +static void adjust_psw_addr(psw_t *psw, unsigned long len) +{ + psw->addr = __rewind_psw(*psw, -len); +} + +#define EMU_ILLEGAL_OP 1 +#define EMU_SPECIFICATION 2 +#define EMU_ADDRESSING 3 + +#define emu_load_ril(ptr, output) \ +({ \ + unsigned int mask = sizeof(*(ptr)) - 1; \ + __typeof__(*(ptr)) input; \ + int __rc = 0; \ + \ + if (!test_facility(34)) \ + __rc = EMU_ILLEGAL_OP; \ + else if ((u64 __force)ptr & mask) \ + __rc = EMU_SPECIFICATION; \ + else if (get_user(input, ptr)) \ + __rc = EMU_ADDRESSING; \ + else \ + *(output) = input; \ + __rc; \ +}) + +#define emu_store_ril(ptr, input) \ +({ \ + unsigned int mask = sizeof(*(ptr)) - 1; \ + int __rc = 0; \ + \ + if (!test_facility(34)) \ + __rc = EMU_ILLEGAL_OP; \ + else if ((u64 __force)ptr & mask) \ + __rc = EMU_SPECIFICATION; \ + else if (put_user(*(input), ptr)) \ + __rc = EMU_ADDRESSING; \ + __rc; \ +}) + +#define emu_cmp_ril(regs, ptr, cmp) \ +({ \ + unsigned int mask = sizeof(*(ptr)) - 1; \ + __typeof__(*(ptr)) input; \ + int __rc = 0; \ + \ + if (!test_facility(34)) \ + __rc = EMU_ILLEGAL_OP; \ + else if ((u64 __force)ptr & mask) \ + __rc = EMU_SPECIFICATION; \ + else if (get_user(input, ptr)) \ + __rc = EMU_ADDRESSING; \ + else if (input > *(cmp)) \ + psw_bits((regs)->psw).cc = 1; \ + else if (input < *(cmp)) \ + psw_bits((regs)->psw).cc = 2; \ + else \ + psw_bits((regs)->psw).cc = 0; \ + __rc; \ +}) + +struct insn_ril { + u8 opc0; + u8 reg : 4; + u8 opc1 : 4; + s32 disp; +} __packed; + +union split_register { + u64 u64; + u32 u32[2]; + u16 u16[4]; + s64 s64; + s32 s32[2]; + s16 s16[4]; +}; + +/* + * pc relative instructions are emulated, since parameters may not be + * accessible from the xol area due to range limitations. + */ +static void handle_insn_ril(struct arch_uprobe *auprobe, struct pt_regs *regs) +{ + union split_register *rx; + struct insn_ril *insn; + unsigned int ilen; + void *uptr; + int rc = 0; + + insn = (struct insn_ril *) &auprobe->insn; + rx = (union split_register *) ®s->gprs[insn->reg]; + uptr = (void *)(regs->psw.addr + (insn->disp * 2)); + ilen = insn_length(insn->opc0); + + switch (insn->opc0) { + case 0xc0: + switch (insn->opc1) { + case 0x00: /* larl */ + rx->u64 = (unsigned long)uptr; + break; + } + break; + case 0xc4: + switch (insn->opc1) { + case 0x02: /* llhrl */ + rc = emu_load_ril((u16 __user *)uptr, &rx->u32[1]); + break; + case 0x04: /* lghrl */ + rc = emu_load_ril((s16 __user *)uptr, &rx->u64); + break; + case 0x05: /* lhrl */ + rc = emu_load_ril((s16 __user *)uptr, &rx->u32[1]); + break; + case 0x06: /* llghrl */ + rc = emu_load_ril((u16 __user *)uptr, &rx->u64); + break; + case 0x08: /* lgrl */ + rc = emu_load_ril((u64 __user *)uptr, &rx->u64); + break; + case 0x0c: /* lgfrl */ + rc = emu_load_ril((s32 __user *)uptr, &rx->u64); + break; + case 0x0d: /* lrl */ + rc = emu_load_ril((u32 __user *)uptr, &rx->u32[1]); + break; + case 0x0e: /* llgfrl */ + rc = emu_load_ril((u32 __user *)uptr, &rx->u64); + break; + case 0x07: /* sthrl */ + rc = emu_store_ril((u16 __user *)uptr, &rx->u16[3]); + break; + case 0x0b: /* stgrl */ + rc = emu_store_ril((u64 __user *)uptr, &rx->u64); + break; + case 0x0f: /* strl */ + rc = emu_store_ril((u32 __user *)uptr, &rx->u32[1]); + break; + } + break; + case 0xc6: + switch (insn->opc1) { + case 0x02: /* pfdrl */ + if (!test_facility(34)) + rc = EMU_ILLEGAL_OP; + break; + case 0x04: /* cghrl */ + rc = emu_cmp_ril(regs, (s16 __user *)uptr, &rx->s64); + break; + case 0x05: /* chrl */ + rc = emu_cmp_ril(regs, (s16 __user *)uptr, &rx->s32[1]); + break; + case 0x06: /* clghrl */ + rc = emu_cmp_ril(regs, (u16 __user *)uptr, &rx->u64); + break; + case 0x07: /* clhrl */ + rc = emu_cmp_ril(regs, (u16 __user *)uptr, &rx->u32[1]); + break; + case 0x08: /* cgrl */ + rc = emu_cmp_ril(regs, (s64 __user *)uptr, &rx->s64); + break; + case 0x0a: /* clgrl */ + rc = emu_cmp_ril(regs, (u64 __user *)uptr, &rx->u64); + break; + case 0x0c: /* cgfrl */ + rc = emu_cmp_ril(regs, (s32 __user *)uptr, &rx->s64); + break; + case 0x0d: /* crl */ + rc = emu_cmp_ril(regs, (s32 __user *)uptr, &rx->s32[1]); + break; + case 0x0e: /* clgfrl */ + rc = emu_cmp_ril(regs, (u32 __user *)uptr, &rx->u64); + break; + case 0x0f: /* clrl */ + rc = emu_cmp_ril(regs, (u32 __user *)uptr, &rx->u32[1]); + break; + } + break; + } + adjust_psw_addr(®s->psw, ilen); + switch (rc) { + case EMU_ILLEGAL_OP: + regs->int_code = ilen << 16 | 0x0001; + do_report_trap(regs, SIGILL, ILL_ILLOPC, NULL); + break; + case EMU_SPECIFICATION: + regs->int_code = ilen << 16 | 0x0006; + do_report_trap(regs, SIGILL, ILL_ILLOPC , NULL); + break; + case EMU_ADDRESSING: + regs->int_code = ilen << 16 | 0x0005; + do_report_trap(regs, SIGSEGV, SEGV_MAPERR, NULL); + break; + } +} + +bool arch_uprobe_skip_sstep(struct arch_uprobe *auprobe, struct pt_regs *regs) +{ + if ((psw_bits(regs->psw).eaba == PSW_AMODE_24BIT) || + ((psw_bits(regs->psw).eaba == PSW_AMODE_31BIT) && + !is_compat_task())) { + regs->psw.addr = __rewind_psw(regs->psw, UPROBE_SWBP_INSN_SIZE); + do_report_trap(regs, SIGILL, ILL_ILLADR, NULL); + return true; + } + if (probe_is_insn_relative_long(auprobe->insn)) { + handle_insn_ril(auprobe, regs); + return true; + } + return false; +} diff --git a/arch/s390/kernel/vdso.c b/arch/s390/kernel/vdso.c index 6136490..0bbb7e0 100644 --- a/arch/s390/kernel/vdso.c +++ b/arch/s390/kernel/vdso.c @@ -316,18 +316,3 @@ static int __init vdso_init(void) return 0; } early_initcall(vdso_init); - -int in_gate_area_no_mm(unsigned long addr) -{ - return 0; -} - -int in_gate_area(struct mm_struct *mm, unsigned long addr) -{ - return 0; -} - -struct vm_area_struct *get_gate_vma(struct mm_struct *mm) -{ - return NULL; -} diff --git a/arch/s390/kernel/vdso32/clock_getres.S b/arch/s390/kernel/vdso32/clock_getres.S index 36aaa25..eca3f00 100644 --- a/arch/s390/kernel/vdso32/clock_getres.S +++ b/arch/s390/kernel/vdso32/clock_getres.S @@ -19,14 +19,20 @@ .type __kernel_clock_getres,@function __kernel_clock_getres: .cfi_startproc + basr %r1,0 + la %r1,4f-.(%r1) chi %r2,__CLOCK_REALTIME je 0f chi %r2,__CLOCK_MONOTONIC + je 0f + la %r1,5f-4f(%r1) + chi %r2,__CLOCK_REALTIME_COARSE + je 0f + chi %r2,__CLOCK_MONOTONIC_COARSE jne 3f 0: ltr %r3,%r3 jz 2f /* res == NULL */ - basr %r1,0 -1: l %r0,4f-1b(%r1) +1: l %r0,0(%r1) xc 0(4,%r3),0(%r3) /* set tp->tv_sec to zero */ st %r0,4(%r3) /* store tp->tv_usec */ 2: lhi %r2,0 @@ -35,5 +41,6 @@ __kernel_clock_getres: svc 0 br %r14 4: .long __CLOCK_REALTIME_RES +5: .long __CLOCK_COARSE_RES .cfi_endproc .size __kernel_clock_getres,.-__kernel_clock_getres diff --git a/arch/s390/kernel/vdso32/clock_gettime.S b/arch/s390/kernel/vdso32/clock_gettime.S index 65fc397..48c2206 100644 --- a/arch/s390/kernel/vdso32/clock_gettime.S +++ b/arch/s390/kernel/vdso32/clock_gettime.S @@ -21,19 +21,21 @@ __kernel_clock_gettime: .cfi_startproc basr %r5,0 0: al %r5,21f-0b(%r5) /* get &_vdso_data */ - chi %r2,__CLOCK_REALTIME + chi %r2,__CLOCK_REALTIME_COARSE je 10f + chi %r2,__CLOCK_REALTIME + je 11f + chi %r2,__CLOCK_MONOTONIC_COARSE + je 9f chi %r2,__CLOCK_MONOTONIC jne 19f /* CLOCK_MONOTONIC */ - ltr %r3,%r3 - jz 9f /* tp == NULL */ 1: l %r4,__VDSO_UPD_COUNT+4(%r5) /* load update counter */ tml %r4,0x0001 /* pending update ? loop */ jnz 1b - stck 24(%r15) /* Store TOD clock */ - lm %r0,%r1,24(%r15) + stcke 24(%r15) /* Store TOD clock */ + lm %r0,%r1,25(%r15) s %r0,__VDSO_XTIME_STAMP(%r5) /* TOD - cycle_last */ sl %r1,__VDSO_XTIME_STAMP+4(%r5) brc 3,2f @@ -67,17 +69,35 @@ __kernel_clock_gettime: j 6b 8: st %r2,0(%r3) /* store tp->tv_sec */ st %r1,4(%r3) /* store tp->tv_nsec */ -9: lhi %r2,0 + lhi %r2,0 br %r14 + /* CLOCK_MONOTONIC_COARSE */ +9: l %r4,__VDSO_UPD_COUNT+4(%r5) /* load update counter */ + tml %r4,0x0001 /* pending update ? loop */ + jnz 9b + l %r2,__VDSO_WTOM_CRS_SEC+4(%r5) + l %r1,__VDSO_WTOM_CRS_NSEC+4(%r5) + cl %r4,__VDSO_UPD_COUNT+4(%r5) /* check update counter */ + jne 9b + j 8b + + /* CLOCK_REALTIME_COARSE */ +10: l %r4,__VDSO_UPD_COUNT+4(%r5) /* load update counter */ + tml %r4,0x0001 /* pending update ? loop */ + jnz 10b + l %r2,__VDSO_XTIME_CRS_SEC+4(%r5) + l %r1,__VDSO_XTIME_CRS_NSEC+4(%r5) + cl %r4,__VDSO_UPD_COUNT+4(%r5) /* check update counter */ + jne 10b + j 17f + /* CLOCK_REALTIME */ -10: ltr %r3,%r3 /* tp == NULL */ - jz 18f 11: l %r4,__VDSO_UPD_COUNT+4(%r5) /* load update counter */ tml %r4,0x0001 /* pending update ? loop */ jnz 11b - stck 24(%r15) /* Store TOD clock */ - lm %r0,%r1,24(%r15) + stcke 24(%r15) /* Store TOD clock */ + lm %r0,%r1,25(%r15) s %r0,__VDSO_XTIME_STAMP(%r5) /* TOD - cycle_last */ sl %r1,__VDSO_XTIME_STAMP+4(%r5) brc 3,12f @@ -111,7 +131,7 @@ __kernel_clock_gettime: j 15b 17: st %r2,0(%r3) /* store tp->tv_sec */ st %r1,4(%r3) /* store tp->tv_nsec */ -18: lhi %r2,0 + lhi %r2,0 br %r14 /* Fallback to system call */ diff --git a/arch/s390/kernel/vdso32/gettimeofday.S b/arch/s390/kernel/vdso32/gettimeofday.S index fd621a9..60def5f 100644 --- a/arch/s390/kernel/vdso32/gettimeofday.S +++ b/arch/s390/kernel/vdso32/gettimeofday.S @@ -29,8 +29,8 @@ __kernel_gettimeofday: l %r4,__VDSO_UPD_COUNT+4(%r5) /* load update counter */ tml %r4,0x0001 /* pending update ? loop */ jnz 1b - stck 24(%r15) /* Store TOD clock */ - lm %r0,%r1,24(%r15) + stcke 24(%r15) /* Store TOD clock */ + lm %r0,%r1,25(%r15) s %r0,__VDSO_XTIME_STAMP(%r5) /* TOD - cycle_last */ sl %r1,__VDSO_XTIME_STAMP+4(%r5) brc 3,3f diff --git a/arch/s390/kernel/vdso64/clock_getres.S b/arch/s390/kernel/vdso64/clock_getres.S index 34deba7..c8513de 100644 --- a/arch/s390/kernel/vdso64/clock_getres.S +++ b/arch/s390/kernel/vdso64/clock_getres.S @@ -19,6 +19,12 @@ .type __kernel_clock_getres,@function __kernel_clock_getres: .cfi_startproc + larl %r1,4f + cghi %r2,__CLOCK_REALTIME_COARSE + je 0f + cghi %r2,__CLOCK_MONOTONIC_COARSE + je 0f + larl %r1,3f cghi %r2,__CLOCK_REALTIME je 0f cghi %r2,__CLOCK_MONOTONIC @@ -32,7 +38,6 @@ __kernel_clock_getres: jz 2f 0: ltgr %r3,%r3 jz 1f /* res == NULL */ - larl %r1,3f lg %r0,0(%r1) xc 0(8,%r3),0(%r3) /* set tp->tv_sec to zero */ stg %r0,8(%r3) /* store tp->tv_usec */ @@ -42,5 +47,6 @@ __kernel_clock_getres: svc 0 br %r14 3: .quad __CLOCK_REALTIME_RES +4: .quad __CLOCK_COARSE_RES .cfi_endproc .size __kernel_clock_getres,.-__kernel_clock_getres diff --git a/arch/s390/kernel/vdso64/clock_gettime.S b/arch/s390/kernel/vdso64/clock_gettime.S index 91940ed..9d9761f 100644 --- a/arch/s390/kernel/vdso64/clock_gettime.S +++ b/arch/s390/kernel/vdso64/clock_gettime.S @@ -20,25 +20,27 @@ __kernel_clock_gettime: .cfi_startproc larl %r5,_vdso_data - cghi %r2,__CLOCK_REALTIME + cghi %r2,__CLOCK_REALTIME_COARSE je 4f + cghi %r2,__CLOCK_REALTIME + je 5f cghi %r2,__CLOCK_THREAD_CPUTIME_ID je 9f cghi %r2,-2 /* Per-thread CPUCLOCK with PID=0, VIRT=1 */ je 9f + cghi %r2,__CLOCK_MONOTONIC_COARSE + je 3f cghi %r2,__CLOCK_MONOTONIC jne 12f /* CLOCK_MONOTONIC */ - ltgr %r3,%r3 - jz 3f /* tp == NULL */ 0: lg %r4,__VDSO_UPD_COUNT(%r5) /* load update counter */ tmll %r4,0x0001 /* pending update ? loop */ jnz 0b - stck 48(%r15) /* Store TOD clock */ + stcke 48(%r15) /* Store TOD clock */ lgf %r2,__VDSO_TK_SHIFT(%r5) /* Timekeeper shift */ lg %r0,__VDSO_WTOM_SEC(%r5) - lg %r1,48(%r15) + lg %r1,49(%r15) sg %r1,__VDSO_XTIME_STAMP(%r5) /* TOD - cycle_last */ msgf %r1,__VDSO_TK_MULT(%r5) /* * tk->mult */ alg %r1,__VDSO_WTOM_NSEC(%r5) @@ -53,18 +55,36 @@ __kernel_clock_gettime: j 1b 2: stg %r0,0(%r3) /* store tp->tv_sec */ stg %r1,8(%r3) /* store tp->tv_nsec */ -3: lghi %r2,0 + lghi %r2,0 br %r14 + /* CLOCK_MONOTONIC_COARSE */ +3: lg %r4,__VDSO_UPD_COUNT(%r5) /* load update counter */ + tmll %r4,0x0001 /* pending update ? loop */ + jnz 3b + lg %r0,__VDSO_WTOM_CRS_SEC(%r5) + lg %r1,__VDSO_WTOM_CRS_NSEC(%r5) + clg %r4,__VDSO_UPD_COUNT(%r5) /* check update counter */ + jne 3b + j 2b + + /* CLOCK_REALTIME_COARSE */ +4: lg %r4,__VDSO_UPD_COUNT(%r5) /* load update counter */ + tmll %r4,0x0001 /* pending update ? loop */ + jnz 4b + lg %r0,__VDSO_XTIME_CRS_SEC(%r5) + lg %r1,__VDSO_XTIME_CRS_NSEC(%r5) + clg %r4,__VDSO_UPD_COUNT(%r5) /* check update counter */ + jne 4b + j 7f + /* CLOCK_REALTIME */ -4: ltr %r3,%r3 /* tp == NULL */ - jz 8f 5: lg %r4,__VDSO_UPD_COUNT(%r5) /* load update counter */ tmll %r4,0x0001 /* pending update ? loop */ jnz 5b - stck 48(%r15) /* Store TOD clock */ + stcke 48(%r15) /* Store TOD clock */ lgf %r2,__VDSO_TK_SHIFT(%r5) /* Timekeeper shift */ - lg %r1,48(%r15) + lg %r1,49(%r15) sg %r1,__VDSO_XTIME_STAMP(%r5) /* TOD - cycle_last */ msgf %r1,__VDSO_TK_MULT(%r5) /* * tk->mult */ alg %r1,__VDSO_XTIME_NSEC(%r5) /* + tk->xtime_nsec */ @@ -80,7 +100,7 @@ __kernel_clock_gettime: j 6b 7: stg %r0,0(%r3) /* store tp->tv_sec */ stg %r1,8(%r3) /* store tp->tv_nsec */ -8: lghi %r2,0 + lghi %r2,0 br %r14 /* CLOCK_THREAD_CPUTIME_ID for this thread */ diff --git a/arch/s390/kernel/vdso64/gettimeofday.S b/arch/s390/kernel/vdso64/gettimeofday.S index d0860d1..7a34499 100644 --- a/arch/s390/kernel/vdso64/gettimeofday.S +++ b/arch/s390/kernel/vdso64/gettimeofday.S @@ -28,8 +28,8 @@ __kernel_gettimeofday: lg %r4,__VDSO_UPD_COUNT(%r5) /* load update counter */ tmll %r4,0x0001 /* pending update ? loop */ jnz 0b - stck 48(%r15) /* Store TOD clock */ - lg %r1,48(%r15) + stcke 48(%r15) /* Store TOD clock */ + lg %r1,49(%r15) sg %r1,__VDSO_XTIME_STAMP(%r5) /* TOD - cycle_last */ msgf %r1,__VDSO_TK_MULT(%r5) /* * tk->mult */ alg %r1,__VDSO_XTIME_NSEC(%r5) /* + tk->xtime_nsec */ diff --git a/arch/s390/kernel/vtime.c b/arch/s390/kernel/vtime.c index 8c34363..416f2a3 100644 --- a/arch/s390/kernel/vtime.c +++ b/arch/s390/kernel/vtime.c @@ -6,27 +6,18 @@ */ #include <linux/kernel_stat.h> -#include <linux/notifier.h> -#include <linux/kprobes.h> #include <linux/export.h> #include <linux/kernel.h> #include <linux/timex.h> #include <linux/types.h> #include <linux/time.h> -#include <linux/cpu.h> -#include <linux/smp.h> -#include <asm/irq_regs.h> #include <asm/cputime.h> #include <asm/vtimer.h> #include <asm/vtime.h> -#include <asm/irq.h> -#include "entry.h" static void virt_timer_expire(void); -DEFINE_PER_CPU(struct s390_idle_data, s390_idle); - static LIST_HEAD(virt_timer_list); static DEFINE_SPINLOCK(virt_timer_lock); static atomic64_t virt_timer_current; @@ -152,49 +143,6 @@ void vtime_account_system(struct task_struct *tsk) __attribute__((alias("vtime_account_irq_enter"))); EXPORT_SYMBOL_GPL(vtime_account_system); -void __kprobes vtime_stop_cpu(void) -{ - struct s390_idle_data *idle = &__get_cpu_var(s390_idle); - unsigned long long idle_time; - unsigned long psw_mask; - - trace_hardirqs_on(); - - /* Wait for external, I/O or machine check interrupt. */ - psw_mask = PSW_KERNEL_BITS | PSW_MASK_WAIT | PSW_MASK_DAT | - PSW_MASK_IO | PSW_MASK_EXT | PSW_MASK_MCHECK; - idle->nohz_delay = 0; - - /* Call the assembler magic in entry.S */ - psw_idle(idle, psw_mask); - - /* Account time spent with enabled wait psw loaded as idle time. */ - idle->sequence++; - smp_wmb(); - idle_time = idle->clock_idle_exit - idle->clock_idle_enter; - idle->clock_idle_enter = idle->clock_idle_exit = 0ULL; - idle->idle_time += idle_time; - idle->idle_count++; - account_idle_time(idle_time); - smp_wmb(); - idle->sequence++; -} - -cputime64_t s390_get_idle_time(int cpu) -{ - struct s390_idle_data *idle = &per_cpu(s390_idle, cpu); - unsigned long long now, idle_enter, idle_exit; - unsigned int sequence; - - do { - now = get_tod_clock(); - sequence = ACCESS_ONCE(idle->sequence); - idle_enter = ACCESS_ONCE(idle->clock_idle_enter); - idle_exit = ACCESS_ONCE(idle->clock_idle_exit); - } while ((sequence & 1) || (ACCESS_ONCE(idle->sequence) != sequence)); - return idle_enter ? ((idle_exit ?: now) - idle_enter) : 0; -} - /* * Sorted add to a list. List is linear searched until first bigger * element is found. @@ -372,31 +320,8 @@ EXPORT_SYMBOL(del_virt_timer); /* * Start the virtual CPU timer on the current CPU. */ -void init_cpu_vtimer(void) +void vtime_init(void) { /* set initial cpu timer */ set_vtimer(VTIMER_MAX_SLICE); } - -static int s390_nohz_notify(struct notifier_block *self, unsigned long action, - void *hcpu) -{ - struct s390_idle_data *idle; - long cpu = (long) hcpu; - - idle = &per_cpu(s390_idle, cpu); - switch (action & ~CPU_TASKS_FROZEN) { - case CPU_DYING: - idle->nohz_delay = 0; - default: - break; - } - return NOTIFY_OK; -} - -void __init vtime_init(void) -{ - /* Enable cpu timer interrupts on the boot cpu. */ - init_cpu_vtimer(); - cpu_notifier(s390_nohz_notify, 0); -} diff --git a/arch/s390/kvm/Kconfig b/arch/s390/kvm/Kconfig index 10d529a..646db9c 100644 --- a/arch/s390/kvm/Kconfig +++ b/arch/s390/kvm/Kconfig @@ -26,6 +26,7 @@ config KVM select KVM_ASYNC_PF select KVM_ASYNC_PF_SYNC select HAVE_KVM_IRQCHIP + select HAVE_KVM_IRQFD select HAVE_KVM_IRQ_ROUTING ---help--- Support hosting paravirtualized guest machines using the SIE diff --git a/arch/s390/kvm/diag.c b/arch/s390/kvm/diag.c index 0161675..9254aff 100644 --- a/arch/s390/kvm/diag.c +++ b/arch/s390/kvm/diag.c @@ -28,22 +28,32 @@ static int diag_release_pages(struct kvm_vcpu *vcpu) start = vcpu->run->s.regs.gprs[(vcpu->arch.sie_block->ipa & 0xf0) >> 4]; end = vcpu->run->s.regs.gprs[vcpu->arch.sie_block->ipa & 0xf] + 4096; - if (start & ~PAGE_MASK || end & ~PAGE_MASK || start > end + if (start & ~PAGE_MASK || end & ~PAGE_MASK || start >= end || start < 2 * PAGE_SIZE) return kvm_s390_inject_program_int(vcpu, PGM_SPECIFICATION); VCPU_EVENT(vcpu, 5, "diag release pages %lX %lX", start, end); vcpu->stat.diagnose_10++; - /* we checked for start > end above */ - if (end < prefix || start >= prefix + 2 * PAGE_SIZE) { - gmap_discard(start, end, vcpu->arch.gmap); + /* + * We checked for start >= end above, so lets check for the + * fast path (no prefix swap page involved) + */ + if (end <= prefix || start >= prefix + 2 * PAGE_SIZE) { + gmap_discard(vcpu->arch.gmap, start, end); } else { - if (start < prefix) - gmap_discard(start, prefix, vcpu->arch.gmap); - if (end >= prefix) - gmap_discard(prefix + 2 * PAGE_SIZE, - end, vcpu->arch.gmap); + /* + * This is slow path. gmap_discard will check for start + * so lets split this into before prefix, prefix, after + * prefix and let gmap_discard make some of these calls + * NOPs. + */ + gmap_discard(vcpu->arch.gmap, start, prefix); + if (start <= prefix) + gmap_discard(vcpu->arch.gmap, 0, 4096); + if (end > prefix + 4096) + gmap_discard(vcpu->arch.gmap, 4096, 8192); + gmap_discard(vcpu->arch.gmap, prefix + 2 * PAGE_SIZE, end); } return 0; } @@ -176,7 +186,8 @@ static int __diag_ipl_functions(struct kvm_vcpu *vcpu) return -EOPNOTSUPP; } - kvm_s390_vcpu_stop(vcpu); + if (!kvm_s390_user_cpu_state_ctrl(vcpu->kvm)) + kvm_s390_vcpu_stop(vcpu); vcpu->run->s390_reset_flags |= KVM_S390_RESET_SUBSYSTEM; vcpu->run->s390_reset_flags |= KVM_S390_RESET_IPL; vcpu->run->s390_reset_flags |= KVM_S390_RESET_CPU_INIT; diff --git a/arch/s390/kvm/gaccess.c b/arch/s390/kvm/gaccess.c index 4653ac6..0f961a1 100644 --- a/arch/s390/kvm/gaccess.c +++ b/arch/s390/kvm/gaccess.c @@ -254,8 +254,7 @@ static void ipte_unlock_simple(struct kvm_vcpu *vcpu) new = old = ACCESS_ONCE(*ic); new.k = 0; } while (cmpxchg(&ic->val, old.val, new.val) != old.val); - if (!ipte_lock_count) - wake_up(&vcpu->kvm->arch.ipte_wq); + wake_up(&vcpu->kvm->arch.ipte_wq); out: mutex_unlock(&ipte_mutex); } diff --git a/arch/s390/kvm/intercept.c b/arch/s390/kvm/intercept.c index a0b586c..eaf4629 100644 --- a/arch/s390/kvm/intercept.c +++ b/arch/s390/kvm/intercept.c @@ -56,32 +56,26 @@ static int handle_noop(struct kvm_vcpu *vcpu) static int handle_stop(struct kvm_vcpu *vcpu) { int rc = 0; + unsigned int action_bits; vcpu->stat.exit_stop_request++; - spin_lock_bh(&vcpu->arch.local_int.lock); - trace_kvm_s390_stop_request(vcpu->arch.local_int.action_bits); - if (vcpu->arch.local_int.action_bits & ACTION_STOP_ON_STOP) { - kvm_s390_vcpu_stop(vcpu); - vcpu->arch.local_int.action_bits &= ~ACTION_STOP_ON_STOP; - VCPU_EVENT(vcpu, 3, "%s", "cpu stopped"); - rc = -EOPNOTSUPP; - } + action_bits = vcpu->arch.local_int.action_bits; - if (vcpu->arch.local_int.action_bits & ACTION_STORE_ON_STOP) { - vcpu->arch.local_int.action_bits &= ~ACTION_STORE_ON_STOP; - /* store status must be called unlocked. Since local_int.lock - * only protects local_int.* and not guest memory we can give - * up the lock here */ - spin_unlock_bh(&vcpu->arch.local_int.lock); + if (!(action_bits & ACTION_STOP_ON_STOP)) + return 0; + + if (action_bits & ACTION_STORE_ON_STOP) { rc = kvm_s390_vcpu_store_status(vcpu, KVM_S390_STORE_STATUS_NOADDR); - if (rc >= 0) - rc = -EOPNOTSUPP; - } else - spin_unlock_bh(&vcpu->arch.local_int.lock); - return rc; + if (rc) + return rc; + } + + if (!kvm_s390_user_cpu_state_ctrl(vcpu->kvm)) + kvm_s390_vcpu_stop(vcpu); + return -EOPNOTSUPP; } static int handle_validity(struct kvm_vcpu *vcpu) diff --git a/arch/s390/kvm/interrupt.c b/arch/s390/kvm/interrupt.c index 90c8de2..a398384 100644 --- a/arch/s390/kvm/interrupt.c +++ b/arch/s390/kvm/interrupt.c @@ -26,8 +26,9 @@ #define IOINT_SSID_MASK 0x00030000 #define IOINT_CSSID_MASK 0x03fc0000 #define IOINT_AI_MASK 0x04000000 +#define PFAULT_INIT 0x0600 -static void deliver_ckc_interrupt(struct kvm_vcpu *vcpu); +static int __must_check deliver_ckc_interrupt(struct kvm_vcpu *vcpu); static int is_ioint(u64 type) { @@ -76,7 +77,7 @@ static u64 int_word_to_isc_bits(u32 int_word) return (0x80 >> isc) << 24; } -static int __interrupt_is_deliverable(struct kvm_vcpu *vcpu, +static int __must_check __interrupt_is_deliverable(struct kvm_vcpu *vcpu, struct kvm_s390_interrupt_info *inti) { switch (inti->type) { @@ -85,6 +86,7 @@ static int __interrupt_is_deliverable(struct kvm_vcpu *vcpu, return 0; if (vcpu->arch.sie_block->gcr[0] & 0x2000ul) return 1; + return 0; case KVM_S390_INT_EMERGENCY: if (psw_extint_disabled(vcpu)) return 0; @@ -158,6 +160,9 @@ static void __reset_intercept_indicators(struct kvm_vcpu *vcpu) LCTL_CR10 | LCTL_CR11); vcpu->arch.sie_block->ictl |= (ICTL_STCTL | ICTL_PINT); } + + if (vcpu->arch.local_int.action_bits & ACTION_STOP_ON_STOP) + atomic_set_mask(CPUSTAT_STOP_INT, &vcpu->arch.sie_block->cpuflags); } static void __set_cpuflag(struct kvm_vcpu *vcpu, u32 flag) @@ -202,11 +207,30 @@ static void __set_intercept_indicator(struct kvm_vcpu *vcpu, } } -static int __deliver_prog_irq(struct kvm_vcpu *vcpu, - struct kvm_s390_pgm_info *pgm_info) +static u16 get_ilc(struct kvm_vcpu *vcpu) { const unsigned short table[] = { 2, 4, 4, 6 }; + + switch (vcpu->arch.sie_block->icptcode) { + case ICPT_INST: + case ICPT_INSTPROGI: + case ICPT_OPEREXC: + case ICPT_PARTEXEC: + case ICPT_IOINST: + /* last instruction only stored for these icptcodes */ + return table[vcpu->arch.sie_block->ipa >> 14]; + case ICPT_PROGI: + return vcpu->arch.sie_block->pgmilc; + default: + return 0; + } +} + +static int __must_check __deliver_prog_irq(struct kvm_vcpu *vcpu, + struct kvm_s390_pgm_info *pgm_info) +{ int rc = 0; + u16 ilc = get_ilc(vcpu); switch (pgm_info->code & ~PGM_PER) { case PGM_AFX_TRANSLATION: @@ -273,25 +297,7 @@ static int __deliver_prog_irq(struct kvm_vcpu *vcpu, (u8 *) __LC_PER_ACCESS_ID); } - switch (vcpu->arch.sie_block->icptcode) { - case ICPT_INST: - case ICPT_INSTPROGI: - case ICPT_OPEREXC: - case ICPT_PARTEXEC: - case ICPT_IOINST: - /* last instruction only stored for these icptcodes */ - rc |= put_guest_lc(vcpu, table[vcpu->arch.sie_block->ipa >> 14], - (u16 *) __LC_PGM_ILC); - break; - case ICPT_PROGI: - rc |= put_guest_lc(vcpu, vcpu->arch.sie_block->pgmilc, - (u16 *) __LC_PGM_ILC); - break; - default: - rc |= put_guest_lc(vcpu, 0, - (u16 *) __LC_PGM_ILC); - } - + rc |= put_guest_lc(vcpu, ilc, (u16 *) __LC_PGM_ILC); rc |= put_guest_lc(vcpu, pgm_info->code, (u16 *)__LC_PGM_INT_CODE); rc |= write_guest_lc(vcpu, __LC_PGM_OLD_PSW, @@ -302,7 +308,7 @@ static int __deliver_prog_irq(struct kvm_vcpu *vcpu, return rc; } -static void __do_deliver_interrupt(struct kvm_vcpu *vcpu, +static int __must_check __do_deliver_interrupt(struct kvm_vcpu *vcpu, struct kvm_s390_interrupt_info *inti) { const unsigned short table[] = { 2, 4, 4, 6 }; @@ -340,7 +346,7 @@ static void __do_deliver_interrupt(struct kvm_vcpu *vcpu, case KVM_S390_INT_CLOCK_COMP: trace_kvm_s390_deliver_interrupt(vcpu->vcpu_id, inti->type, inti->ext.ext_params, 0); - deliver_ckc_interrupt(vcpu); + rc = deliver_ckc_interrupt(vcpu); break; case KVM_S390_INT_CPU_TIMER: trace_kvm_s390_deliver_interrupt(vcpu->vcpu_id, inti->type, @@ -373,8 +379,9 @@ static void __do_deliver_interrupt(struct kvm_vcpu *vcpu, case KVM_S390_INT_PFAULT_INIT: trace_kvm_s390_deliver_interrupt(vcpu->vcpu_id, inti->type, 0, inti->ext.ext_params2); - rc = put_guest_lc(vcpu, 0x2603, (u16 *) __LC_EXT_INT_CODE); - rc |= put_guest_lc(vcpu, 0x0600, (u16 *) __LC_EXT_CPU_ADDR); + rc = put_guest_lc(vcpu, EXT_IRQ_CP_SERVICE, + (u16 *) __LC_EXT_INT_CODE); + rc |= put_guest_lc(vcpu, PFAULT_INIT, (u16 *) __LC_EXT_CPU_ADDR); rc |= write_guest_lc(vcpu, __LC_EXT_OLD_PSW, &vcpu->arch.sie_block->gpsw, sizeof(psw_t)); rc |= read_guest_lc(vcpu, __LC_EXT_NEW_PSW, @@ -498,14 +505,11 @@ static void __do_deliver_interrupt(struct kvm_vcpu *vcpu, default: BUG(); } - if (rc) { - printk("kvm: The guest lowcore is not mapped during interrupt " - "delivery, killing userspace\n"); - do_exit(SIGKILL); - } + + return rc; } -static void deliver_ckc_interrupt(struct kvm_vcpu *vcpu) +static int __must_check deliver_ckc_interrupt(struct kvm_vcpu *vcpu) { int rc; @@ -515,11 +519,7 @@ static void deliver_ckc_interrupt(struct kvm_vcpu *vcpu) rc |= read_guest_lc(vcpu, __LC_EXT_NEW_PSW, &vcpu->arch.sie_block->gpsw, sizeof(psw_t)); - if (rc) { - printk("kvm: The guest lowcore is not mapped during interrupt " - "delivery, killing userspace\n"); - do_exit(SIGKILL); - } + return rc; } /* Check whether SIGP interpretation facility has an external call pending */ @@ -544,13 +544,13 @@ int kvm_cpu_has_interrupt(struct kvm_vcpu *vcpu) int rc = 0; if (atomic_read(&li->active)) { - spin_lock_bh(&li->lock); + spin_lock(&li->lock); list_for_each_entry(inti, &li->list, list) if (__interrupt_is_deliverable(vcpu, inti)) { rc = 1; break; } - spin_unlock_bh(&li->lock); + spin_unlock(&li->lock); } if ((!rc) && atomic_read(&fi->active)) { @@ -585,88 +585,57 @@ int kvm_cpu_has_pending_timer(struct kvm_vcpu *vcpu) int kvm_s390_handle_wait(struct kvm_vcpu *vcpu) { u64 now, sltime; - DECLARE_WAITQUEUE(wait, current); vcpu->stat.exit_wait_state++; - if (kvm_cpu_has_interrupt(vcpu)) - return 0; - __set_cpu_idle(vcpu); - spin_lock_bh(&vcpu->arch.local_int.lock); - vcpu->arch.local_int.timer_due = 0; - spin_unlock_bh(&vcpu->arch.local_int.lock); + /* fast path */ + if (kvm_cpu_has_pending_timer(vcpu) || kvm_arch_vcpu_runnable(vcpu)) + return 0; if (psw_interrupts_disabled(vcpu)) { VCPU_EVENT(vcpu, 3, "%s", "disabled wait"); - __unset_cpu_idle(vcpu); return -EOPNOTSUPP; /* disabled wait */ } + __set_cpu_idle(vcpu); if (!ckc_interrupts_enabled(vcpu)) { VCPU_EVENT(vcpu, 3, "%s", "enabled wait w/o timer"); goto no_timer; } now = get_tod_clock_fast() + vcpu->arch.sie_block->epoch; - if (vcpu->arch.sie_block->ckc < now) { - __unset_cpu_idle(vcpu); - return 0; - } - sltime = tod_to_ns(vcpu->arch.sie_block->ckc - now); - hrtimer_start(&vcpu->arch.ckc_timer, ktime_set (0, sltime) , HRTIMER_MODE_REL); VCPU_EVENT(vcpu, 5, "enabled wait via clock comparator: %llx ns", sltime); no_timer: srcu_read_unlock(&vcpu->kvm->srcu, vcpu->srcu_idx); - spin_lock(&vcpu->arch.local_int.float_int->lock); - spin_lock_bh(&vcpu->arch.local_int.lock); - add_wait_queue(&vcpu->wq, &wait); - while (list_empty(&vcpu->arch.local_int.list) && - list_empty(&vcpu->arch.local_int.float_int->list) && - (!vcpu->arch.local_int.timer_due) && - !signal_pending(current) && - !kvm_s390_si_ext_call_pending(vcpu)) { - set_current_state(TASK_INTERRUPTIBLE); - spin_unlock_bh(&vcpu->arch.local_int.lock); - spin_unlock(&vcpu->arch.local_int.float_int->lock); - schedule(); - spin_lock(&vcpu->arch.local_int.float_int->lock); - spin_lock_bh(&vcpu->arch.local_int.lock); - } + kvm_vcpu_block(vcpu); __unset_cpu_idle(vcpu); - __set_current_state(TASK_RUNNING); - remove_wait_queue(&vcpu->wq, &wait); - spin_unlock_bh(&vcpu->arch.local_int.lock); - spin_unlock(&vcpu->arch.local_int.float_int->lock); vcpu->srcu_idx = srcu_read_lock(&vcpu->kvm->srcu); hrtimer_try_to_cancel(&vcpu->arch.ckc_timer); return 0; } -void kvm_s390_tasklet(unsigned long parm) +void kvm_s390_vcpu_wakeup(struct kvm_vcpu *vcpu) { - struct kvm_vcpu *vcpu = (struct kvm_vcpu *) parm; - - spin_lock(&vcpu->arch.local_int.lock); - vcpu->arch.local_int.timer_due = 1; - if (waitqueue_active(&vcpu->wq)) + if (waitqueue_active(&vcpu->wq)) { + /* + * The vcpu gave up the cpu voluntarily, mark it as a good + * yield-candidate. + */ + vcpu->preempted = true; wake_up_interruptible(&vcpu->wq); - spin_unlock(&vcpu->arch.local_int.lock); + vcpu->stat.halt_wakeup++; + } } -/* - * low level hrtimer wake routine. Because this runs in hardirq context - * we schedule a tasklet to do the real work. - */ enum hrtimer_restart kvm_s390_idle_wakeup(struct hrtimer *timer) { struct kvm_vcpu *vcpu; vcpu = container_of(timer, struct kvm_vcpu, arch.ckc_timer); - vcpu->preempted = true; - tasklet_schedule(&vcpu->arch.tasklet); + kvm_s390_vcpu_wakeup(vcpu); return HRTIMER_NORESTART; } @@ -676,13 +645,13 @@ void kvm_s390_clear_local_irqs(struct kvm_vcpu *vcpu) struct kvm_s390_local_interrupt *li = &vcpu->arch.local_int; struct kvm_s390_interrupt_info *n, *inti = NULL; - spin_lock_bh(&li->lock); + spin_lock(&li->lock); list_for_each_entry_safe(inti, n, &li->list, list) { list_del(&inti->list); kfree(inti); } atomic_set(&li->active, 0); - spin_unlock_bh(&li->lock); + spin_unlock(&li->lock); /* clear pending external calls set by sigp interpretation facility */ atomic_clear_mask(CPUSTAT_ECALL_PEND, &vcpu->arch.sie_block->cpuflags); @@ -690,18 +659,19 @@ void kvm_s390_clear_local_irqs(struct kvm_vcpu *vcpu) &vcpu->kvm->arch.sca->cpu[vcpu->vcpu_id].ctrl); } -void kvm_s390_deliver_pending_interrupts(struct kvm_vcpu *vcpu) +int __must_check kvm_s390_deliver_pending_interrupts(struct kvm_vcpu *vcpu) { struct kvm_s390_local_interrupt *li = &vcpu->arch.local_int; struct kvm_s390_float_interrupt *fi = vcpu->arch.local_int.float_int; struct kvm_s390_interrupt_info *n, *inti = NULL; int deliver; + int rc = 0; __reset_intercept_indicators(vcpu); if (atomic_read(&li->active)) { do { deliver = 0; - spin_lock_bh(&li->lock); + spin_lock(&li->lock); list_for_each_entry_safe(inti, n, &li->list, list) { if (__interrupt_is_deliverable(vcpu, inti)) { list_del(&inti->list); @@ -712,18 +682,18 @@ void kvm_s390_deliver_pending_interrupts(struct kvm_vcpu *vcpu) } if (list_empty(&li->list)) atomic_set(&li->active, 0); - spin_unlock_bh(&li->lock); + spin_unlock(&li->lock); if (deliver) { - __do_deliver_interrupt(vcpu, inti); + rc = __do_deliver_interrupt(vcpu, inti); kfree(inti); } - } while (deliver); + } while (!rc && deliver); } - if (kvm_cpu_has_pending_timer(vcpu)) - deliver_ckc_interrupt(vcpu); + if (!rc && kvm_cpu_has_pending_timer(vcpu)) + rc = deliver_ckc_interrupt(vcpu); - if (atomic_read(&fi->active)) { + if (!rc && atomic_read(&fi->active)) { do { deliver = 0; spin_lock(&fi->lock); @@ -740,67 +710,13 @@ void kvm_s390_deliver_pending_interrupts(struct kvm_vcpu *vcpu) atomic_set(&fi->active, 0); spin_unlock(&fi->lock); if (deliver) { - __do_deliver_interrupt(vcpu, inti); + rc = __do_deliver_interrupt(vcpu, inti); kfree(inti); } - } while (deliver); + } while (!rc && deliver); } -} -void kvm_s390_deliver_pending_machine_checks(struct kvm_vcpu *vcpu) -{ - struct kvm_s390_local_interrupt *li = &vcpu->arch.local_int; - struct kvm_s390_float_interrupt *fi = vcpu->arch.local_int.float_int; - struct kvm_s390_interrupt_info *n, *inti = NULL; - int deliver; - - __reset_intercept_indicators(vcpu); - if (atomic_read(&li->active)) { - do { - deliver = 0; - spin_lock_bh(&li->lock); - list_for_each_entry_safe(inti, n, &li->list, list) { - if ((inti->type == KVM_S390_MCHK) && - __interrupt_is_deliverable(vcpu, inti)) { - list_del(&inti->list); - deliver = 1; - break; - } - __set_intercept_indicator(vcpu, inti); - } - if (list_empty(&li->list)) - atomic_set(&li->active, 0); - spin_unlock_bh(&li->lock); - if (deliver) { - __do_deliver_interrupt(vcpu, inti); - kfree(inti); - } - } while (deliver); - } - - if (atomic_read(&fi->active)) { - do { - deliver = 0; - spin_lock(&fi->lock); - list_for_each_entry_safe(inti, n, &fi->list, list) { - if ((inti->type == KVM_S390_MCHK) && - __interrupt_is_deliverable(vcpu, inti)) { - list_del(&inti->list); - fi->irq_count--; - deliver = 1; - break; - } - __set_intercept_indicator(vcpu, inti); - } - if (list_empty(&fi->list)) - atomic_set(&fi->active, 0); - spin_unlock(&fi->lock); - if (deliver) { - __do_deliver_interrupt(vcpu, inti); - kfree(inti); - } - } while (deliver); - } + return rc; } int kvm_s390_inject_program_int(struct kvm_vcpu *vcpu, u16 code) @@ -817,11 +733,11 @@ int kvm_s390_inject_program_int(struct kvm_vcpu *vcpu, u16 code) VCPU_EVENT(vcpu, 3, "inject: program check %d (from kernel)", code); trace_kvm_s390_inject_vcpu(vcpu->vcpu_id, inti->type, code, 0, 1); - spin_lock_bh(&li->lock); + spin_lock(&li->lock); list_add(&inti->list, &li->list); atomic_set(&li->active, 1); BUG_ON(waitqueue_active(li->wq)); - spin_unlock_bh(&li->lock); + spin_unlock(&li->lock); return 0; } @@ -842,11 +758,11 @@ int kvm_s390_inject_prog_irq(struct kvm_vcpu *vcpu, inti->type = KVM_S390_PROGRAM_INT; memcpy(&inti->pgm, pgm_info, sizeof(inti->pgm)); - spin_lock_bh(&li->lock); + spin_lock(&li->lock); list_add(&inti->list, &li->list); atomic_set(&li->active, 1); BUG_ON(waitqueue_active(li->wq)); - spin_unlock_bh(&li->lock); + spin_unlock(&li->lock); return 0; } @@ -934,12 +850,10 @@ static int __inject_vm(struct kvm *kvm, struct kvm_s390_interrupt_info *inti) } dst_vcpu = kvm_get_vcpu(kvm, sigcpu); li = &dst_vcpu->arch.local_int; - spin_lock_bh(&li->lock); + spin_lock(&li->lock); atomic_set_mask(CPUSTAT_EXT_INT, li->cpuflags); - if (waitqueue_active(li->wq)) - wake_up_interruptible(li->wq); - kvm_get_vcpu(kvm, sigcpu)->preempted = true; - spin_unlock_bh(&li->lock); + spin_unlock(&li->lock); + kvm_s390_vcpu_wakeup(kvm_get_vcpu(kvm, sigcpu)); unlock_fi: spin_unlock(&fi->lock); mutex_unlock(&kvm->lock); @@ -1079,9 +993,8 @@ int kvm_s390_inject_vcpu(struct kvm_vcpu *vcpu, trace_kvm_s390_inject_vcpu(vcpu->vcpu_id, s390int->type, s390int->parm, s390int->parm64, 2); - mutex_lock(&vcpu->kvm->lock); li = &vcpu->arch.local_int; - spin_lock_bh(&li->lock); + spin_lock(&li->lock); if (inti->type == KVM_S390_PROGRAM_INT) list_add(&inti->list, &li->list); else @@ -1090,11 +1003,8 @@ int kvm_s390_inject_vcpu(struct kvm_vcpu *vcpu, if (inti->type == KVM_S390_SIGP_STOP) li->action_bits |= ACTION_STOP_ON_STOP; atomic_set_mask(CPUSTAT_EXT_INT, li->cpuflags); - if (waitqueue_active(&vcpu->wq)) - wake_up_interruptible(&vcpu->wq); - vcpu->preempted = true; - spin_unlock_bh(&li->lock); - mutex_unlock(&vcpu->kvm->lock); + spin_unlock(&li->lock); + kvm_s390_vcpu_wakeup(vcpu); return 0; } @@ -1333,7 +1243,7 @@ static int kvm_s390_adapter_map(struct kvm *kvm, unsigned int id, __u64 addr) } INIT_LIST_HEAD(&map->list); map->guest_addr = addr; - map->addr = gmap_translate(addr, kvm->arch.gmap); + map->addr = gmap_translate(kvm->arch.gmap, addr); if (map->addr == -EFAULT) { ret = -EFAULT; goto out; @@ -1443,7 +1353,6 @@ static int flic_set_attr(struct kvm_device *dev, struct kvm_device_attr *attr) r = enqueue_floating_irq(dev, attr); break; case KVM_DEV_FLIC_CLEAR_IRQS: - r = 0; kvm_s390_clear_float_irqs(dev->kvm); break; case KVM_DEV_FLIC_APF_ENABLE: @@ -1589,8 +1498,7 @@ static int set_adapter_int(struct kvm_kernel_irq_routing_entry *e, return ret; } -int kvm_set_routing_entry(struct kvm_irq_routing_table *rt, - struct kvm_kernel_irq_routing_entry *e, +int kvm_set_routing_entry(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 2f3e14f..55aade4 100644 --- a/arch/s390/kvm/kvm-s390.c +++ b/arch/s390/kvm/kvm-s390.c @@ -50,6 +50,7 @@ struct kvm_stats_debugfs_item debugfs_entries[] = { { "exit_instruction", VCPU_STAT(exit_instruction) }, { "exit_program_interruption", VCPU_STAT(exit_program_interruption) }, { "exit_instr_and_program_int", VCPU_STAT(exit_instr_and_program) }, + { "halt_wakeup", VCPU_STAT(halt_wakeup) }, { "instruction_lctlg", VCPU_STAT(instruction_lctlg) }, { "instruction_lctl", VCPU_STAT(instruction_lctl) }, { "instruction_stctl", VCPU_STAT(instruction_stctl) }, @@ -100,16 +101,12 @@ int test_vfacility(unsigned long nr) } /* Section: not file related */ -int kvm_arch_hardware_enable(void *garbage) +int kvm_arch_hardware_enable(void) { /* every s390 is virtualization enabled ;-) */ return 0; } -void kvm_arch_hardware_disable(void *garbage) -{ -} - static void kvm_gmap_notifier(struct gmap *gmap, unsigned long address); int kvm_arch_hardware_setup(void) @@ -124,17 +121,10 @@ void kvm_arch_hardware_unsetup(void) gmap_unregister_ipte_notifier(&gmap_notifier); } -void kvm_arch_check_processor_compat(void *rtn) -{ -} - int kvm_arch_init(void *opaque) { - return 0; -} - -void kvm_arch_exit(void) -{ + /* Register floating interrupt controller interface. */ + return kvm_register_device_ops(&kvm_flic_ops, KVM_DEV_TYPE_FLIC); } /* Section: device related */ @@ -146,7 +136,7 @@ long kvm_arch_dev_ioctl(struct file *filp, return -EINVAL; } -int kvm_dev_ioctl_check_extension(long ext) +int kvm_vm_ioctl_check_extension(struct kvm *kvm, long ext) { int r; @@ -166,7 +156,9 @@ int kvm_dev_ioctl_check_extension(long ext) case KVM_CAP_IOEVENTFD: case KVM_CAP_DEVICE_CTRL: case KVM_CAP_ENABLE_CAP_VM: + case KVM_CAP_S390_IRQCHIP: case KVM_CAP_VM_ATTRIBUTES: + case KVM_CAP_MP_STATE: r = 1; break; case KVM_CAP_NR_VCPUS: @@ -402,6 +394,22 @@ long kvm_arch_vm_ioctl(struct file *filp, return r; } +static int kvm_s390_crypto_init(struct kvm *kvm) +{ + if (!test_vfacility(76)) + return 0; + + kvm->arch.crypto.crycb = kzalloc(sizeof(*kvm->arch.crypto.crycb), + GFP_KERNEL | GFP_DMA); + if (!kvm->arch.crypto.crycb) + return -ENOMEM; + + kvm->arch.crypto.crycbd = (__u32) (unsigned long) kvm->arch.crypto.crycb | + CRYCB_FORMAT1; + + return 0; +} + int kvm_arch_init_vm(struct kvm *kvm, unsigned long type) { int rc; @@ -439,6 +447,9 @@ int kvm_arch_init_vm(struct kvm *kvm, unsigned long type) if (!kvm->arch.dbf) goto out_nodbf; + if (kvm_s390_crypto_init(kvm) < 0) + goto out_crypto; + spin_lock_init(&kvm->arch.float_int.lock); INIT_LIST_HEAD(&kvm->arch.float_int.list); init_waitqueue_head(&kvm->arch.ipte_wq); @@ -449,7 +460,7 @@ int kvm_arch_init_vm(struct kvm *kvm, unsigned long type) if (type & KVM_VM_S390_UCONTROL) { kvm->arch.gmap = NULL; } else { - kvm->arch.gmap = gmap_alloc(current->mm); + kvm->arch.gmap = gmap_alloc(current->mm, (1UL << 44) - 1); if (!kvm->arch.gmap) goto out_nogmap; kvm->arch.gmap->private = kvm; @@ -463,6 +474,8 @@ int kvm_arch_init_vm(struct kvm *kvm, unsigned long type) return 0; out_nogmap: + kfree(kvm->arch.crypto.crycb); +out_crypto: debug_unregister(kvm->arch.dbf); out_nodbf: free_page((unsigned long)(kvm->arch.sca)); @@ -512,15 +525,12 @@ static void kvm_free_vcpus(struct kvm *kvm) mutex_unlock(&kvm->lock); } -void kvm_arch_sync_events(struct kvm *kvm) -{ -} - void kvm_arch_destroy_vm(struct kvm *kvm) { kvm_free_vcpus(kvm); free_page((unsigned long)(kvm->arch.sca)); debug_unregister(kvm->arch.dbf); + kfree(kvm->arch.crypto.crycb); if (!kvm_is_ucontrol(kvm)) gmap_free(kvm->arch.gmap); kvm_s390_destroy_adapters(kvm); @@ -533,7 +543,7 @@ int kvm_arch_vcpu_init(struct kvm_vcpu *vcpu) vcpu->arch.pfault_token = KVM_S390_PFAULT_TOKEN_INVALID; kvm_clear_async_pf_completion_queue(vcpu); if (kvm_is_ucontrol(vcpu->kvm)) { - vcpu->arch.gmap = gmap_alloc(current->mm); + vcpu->arch.gmap = gmap_alloc(current->mm, -1UL); if (!vcpu->arch.gmap) return -ENOMEM; vcpu->arch.gmap->private = vcpu->kvm; @@ -544,15 +554,12 @@ int kvm_arch_vcpu_init(struct kvm_vcpu *vcpu) vcpu->run->kvm_valid_regs = KVM_SYNC_PREFIX | KVM_SYNC_GPRS | KVM_SYNC_ACRS | - KVM_SYNC_CRS; + KVM_SYNC_CRS | + KVM_SYNC_ARCH0 | + KVM_SYNC_PFAULT; return 0; } -void kvm_arch_vcpu_uninit(struct kvm_vcpu *vcpu) -{ - /* Nothing todo */ -} - void kvm_arch_vcpu_load(struct kvm_vcpu *vcpu, int cpu) { save_fp_ctl(&vcpu->arch.host_fpregs.fpc); @@ -595,7 +602,8 @@ static void kvm_s390_vcpu_initial_reset(struct kvm_vcpu *vcpu) vcpu->arch.sie_block->pp = 0; vcpu->arch.pfault_token = KVM_S390_PFAULT_TOKEN_INVALID; kvm_clear_async_pf_completion_queue(vcpu); - kvm_s390_vcpu_stop(vcpu); + if (!kvm_s390_user_cpu_state_ctrl(vcpu->kvm)) + kvm_s390_vcpu_stop(vcpu); kvm_s390_clear_local_irqs(vcpu); } @@ -604,6 +612,14 @@ int kvm_arch_vcpu_postcreate(struct kvm_vcpu *vcpu) return 0; } +static void kvm_s390_vcpu_crypto_setup(struct kvm_vcpu *vcpu) +{ + if (!test_vfacility(76)) + return; + + vcpu->arch.sie_block->crycbd = vcpu->kvm->arch.crypto.crycbd; +} + void kvm_s390_vcpu_unsetup_cmma(struct kvm_vcpu *vcpu) { free_page(vcpu->arch.sie_block->cbrlo); @@ -647,11 +663,12 @@ int kvm_arch_vcpu_setup(struct kvm_vcpu *vcpu) return rc; } hrtimer_init(&vcpu->arch.ckc_timer, CLOCK_REALTIME, HRTIMER_MODE_ABS); - tasklet_init(&vcpu->arch.tasklet, kvm_s390_tasklet, - (unsigned long) vcpu); vcpu->arch.ckc_timer.function = kvm_s390_idle_wakeup; get_cpu_id(&vcpu->arch.cpu_id); vcpu->arch.cpu_id.version = 0xff; + + kvm_s390_vcpu_crypto_setup(vcpu); + return rc; } @@ -926,7 +943,7 @@ static int kvm_arch_vcpu_ioctl_set_initial_psw(struct kvm_vcpu *vcpu, psw_t psw) { int rc = 0; - if (!(atomic_read(&vcpu->arch.sie_block->cpuflags) & CPUSTAT_STOPPED)) + if (!is_vcpu_stopped(vcpu)) rc = -EBUSY; else { vcpu->run->psw_mask = psw.mask; @@ -980,13 +997,34 @@ int kvm_arch_vcpu_ioctl_set_guest_debug(struct kvm_vcpu *vcpu, int kvm_arch_vcpu_ioctl_get_mpstate(struct kvm_vcpu *vcpu, struct kvm_mp_state *mp_state) { - return -EINVAL; /* not implemented yet */ + /* CHECK_STOP and LOAD are not supported yet */ + return is_vcpu_stopped(vcpu) ? KVM_MP_STATE_STOPPED : + KVM_MP_STATE_OPERATING; } int kvm_arch_vcpu_ioctl_set_mpstate(struct kvm_vcpu *vcpu, struct kvm_mp_state *mp_state) { - return -EINVAL; /* not implemented yet */ + int rc = 0; + + /* user space knows about this interface - let it control the state */ + vcpu->kvm->arch.user_cpu_state_ctrl = 1; + + switch (mp_state->mp_state) { + case KVM_MP_STATE_STOPPED: + kvm_s390_vcpu_stop(vcpu); + break; + case KVM_MP_STATE_OPERATING: + kvm_s390_vcpu_start(vcpu); + break; + case KVM_MP_STATE_LOAD: + case KVM_MP_STATE_CHECK_STOP: + /* fall through - CHECK_STOP and LOAD are not supported yet */ + default: + rc = -ENXIO; + } + + return rc; } bool kvm_s390_cmma_enabled(struct kvm *kvm) @@ -1027,6 +1065,11 @@ retry: goto retry; } + if (kvm_check_request(KVM_REQ_TLB_FLUSH, vcpu)) { + vcpu->arch.sie_block->ihcpu = 0xffff; + goto retry; + } + if (kvm_check_request(KVM_REQ_ENABLE_IBS, vcpu)) { if (!ibs_enabled(vcpu)) { trace_kvm_s390_enable_disable_ibs(vcpu->vcpu_id, 1); @@ -1045,6 +1088,9 @@ retry: goto retry; } + /* nothing to do, just clear the request */ + clear_bit(KVM_REQ_UNHALT, &vcpu->requests); + return 0; } @@ -1060,18 +1106,8 @@ retry: */ long kvm_arch_fault_in_page(struct kvm_vcpu *vcpu, gpa_t gpa, int writable) { - struct mm_struct *mm = current->mm; - hva_t hva; - long rc; - - hva = gmap_fault(gpa, vcpu->arch.gmap); - if (IS_ERR_VALUE(hva)) - return (long)hva; - down_read(&mm->mmap_sem); - rc = get_user_pages(current, mm, hva, 1, writable, 0, NULL, NULL); - up_read(&mm->mmap_sem); - - return rc < 0 ? rc : 0; + return gmap_fault(vcpu->arch.gmap, gpa, + writable ? FAULT_FLAG_WRITE : 0); } static void __kvm_inject_pfault_token(struct kvm_vcpu *vcpu, bool start_token, @@ -1166,8 +1202,11 @@ static int vcpu_pre_run(struct kvm_vcpu *vcpu) if (test_cpu_flag(CIF_MCCK_PENDING)) s390_handle_mcck(); - if (!kvm_is_ucontrol(vcpu->kvm)) - kvm_s390_deliver_pending_interrupts(vcpu); + if (!kvm_is_ucontrol(vcpu->kvm)) { + rc = kvm_s390_deliver_pending_interrupts(vcpu); + if (rc) + return rc; + } rc = kvm_s390_handle_requests(vcpu); if (rc) @@ -1271,6 +1310,48 @@ static int __vcpu_run(struct kvm_vcpu *vcpu) return rc; } +static void sync_regs(struct kvm_vcpu *vcpu, struct kvm_run *kvm_run) +{ + vcpu->arch.sie_block->gpsw.mask = kvm_run->psw_mask; + vcpu->arch.sie_block->gpsw.addr = kvm_run->psw_addr; + if (kvm_run->kvm_dirty_regs & KVM_SYNC_PREFIX) + kvm_s390_set_prefix(vcpu, kvm_run->s.regs.prefix); + if (kvm_run->kvm_dirty_regs & KVM_SYNC_CRS) { + memcpy(&vcpu->arch.sie_block->gcr, &kvm_run->s.regs.crs, 128); + /* some control register changes require a tlb flush */ + kvm_make_request(KVM_REQ_TLB_FLUSH, vcpu); + } + if (kvm_run->kvm_dirty_regs & KVM_SYNC_ARCH0) { + vcpu->arch.sie_block->cputm = kvm_run->s.regs.cputm; + vcpu->arch.sie_block->ckc = kvm_run->s.regs.ckc; + vcpu->arch.sie_block->todpr = kvm_run->s.regs.todpr; + vcpu->arch.sie_block->pp = kvm_run->s.regs.pp; + vcpu->arch.sie_block->gbea = kvm_run->s.regs.gbea; + } + if (kvm_run->kvm_dirty_regs & KVM_SYNC_PFAULT) { + vcpu->arch.pfault_token = kvm_run->s.regs.pft; + vcpu->arch.pfault_select = kvm_run->s.regs.pfs; + vcpu->arch.pfault_compare = kvm_run->s.regs.pfc; + } + kvm_run->kvm_dirty_regs = 0; +} + +static void store_regs(struct kvm_vcpu *vcpu, struct kvm_run *kvm_run) +{ + kvm_run->psw_mask = vcpu->arch.sie_block->gpsw.mask; + kvm_run->psw_addr = vcpu->arch.sie_block->gpsw.addr; + kvm_run->s.regs.prefix = kvm_s390_get_prefix(vcpu); + memcpy(&kvm_run->s.regs.crs, &vcpu->arch.sie_block->gcr, 128); + kvm_run->s.regs.cputm = vcpu->arch.sie_block->cputm; + kvm_run->s.regs.ckc = vcpu->arch.sie_block->ckc; + kvm_run->s.regs.todpr = vcpu->arch.sie_block->todpr; + kvm_run->s.regs.pp = vcpu->arch.sie_block->pp; + kvm_run->s.regs.gbea = vcpu->arch.sie_block->gbea; + kvm_run->s.regs.pft = vcpu->arch.pfault_token; + kvm_run->s.regs.pfs = vcpu->arch.pfault_select; + kvm_run->s.regs.pfc = vcpu->arch.pfault_compare; +} + int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *kvm_run) { int rc; @@ -1284,32 +1365,15 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *kvm_run) if (vcpu->sigset_active) sigprocmask(SIG_SETMASK, &vcpu->sigset, &sigsaved); - kvm_s390_vcpu_start(vcpu); - - switch (kvm_run->exit_reason) { - case KVM_EXIT_S390_SIEIC: - case KVM_EXIT_UNKNOWN: - case KVM_EXIT_INTR: - case KVM_EXIT_S390_RESET: - case KVM_EXIT_S390_UCONTROL: - case KVM_EXIT_S390_TSCH: - case KVM_EXIT_DEBUG: - break; - default: - BUG(); + if (!kvm_s390_user_cpu_state_ctrl(vcpu->kvm)) { + kvm_s390_vcpu_start(vcpu); + } else if (is_vcpu_stopped(vcpu)) { + pr_err_ratelimited("kvm-s390: can't run stopped vcpu %d\n", + vcpu->vcpu_id); + return -EINVAL; } - vcpu->arch.sie_block->gpsw.mask = kvm_run->psw_mask; - vcpu->arch.sie_block->gpsw.addr = kvm_run->psw_addr; - if (kvm_run->kvm_dirty_regs & KVM_SYNC_PREFIX) { - kvm_run->kvm_dirty_regs &= ~KVM_SYNC_PREFIX; - kvm_s390_set_prefix(vcpu, kvm_run->s.regs.prefix); - } - if (kvm_run->kvm_dirty_regs & KVM_SYNC_CRS) { - kvm_run->kvm_dirty_regs &= ~KVM_SYNC_CRS; - memcpy(&vcpu->arch.sie_block->gcr, &kvm_run->s.regs.crs, 128); - kvm_s390_set_prefix(vcpu, kvm_run->s.regs.prefix); - } + sync_regs(vcpu, kvm_run); might_fault(); rc = __vcpu_run(vcpu); @@ -1339,10 +1403,7 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *kvm_run) rc = 0; } - kvm_run->psw_mask = vcpu->arch.sie_block->gpsw.mask; - kvm_run->psw_addr = vcpu->arch.sie_block->gpsw.addr; - kvm_run->s.regs.prefix = kvm_s390_get_prefix(vcpu); - memcpy(&kvm_run->s.regs.crs, &vcpu->arch.sie_block->gcr, 128); + store_regs(vcpu, kvm_run); if (vcpu->sigset_active) sigprocmask(SIG_SETMASK, &sigsaved, NULL); @@ -1413,11 +1474,6 @@ int kvm_s390_vcpu_store_status(struct kvm_vcpu *vcpu, unsigned long addr) return kvm_s390_store_status_unloaded(vcpu, addr); } -static inline int is_vcpu_stopped(struct kvm_vcpu *vcpu) -{ - return atomic_read(&(vcpu)->arch.sie_block->cpuflags) & CPUSTAT_STOPPED; -} - static void __disable_ibs_on_vcpu(struct kvm_vcpu *vcpu) { kvm_check_request(KVM_REQ_ENABLE_IBS, vcpu); @@ -1451,7 +1507,7 @@ void kvm_s390_vcpu_start(struct kvm_vcpu *vcpu) trace_kvm_s390_vcpu_start_stop(vcpu->vcpu_id, 1); /* Only one cpu at a time may enter/leave the STOPPED state. */ - spin_lock_bh(&vcpu->kvm->arch.start_stop_lock); + spin_lock(&vcpu->kvm->arch.start_stop_lock); online_vcpus = atomic_read(&vcpu->kvm->online_vcpus); for (i = 0; i < online_vcpus; i++) { @@ -1476,8 +1532,8 @@ void kvm_s390_vcpu_start(struct kvm_vcpu *vcpu) * Another VCPU might have used IBS while we were offline. * Let's play safe and flush the VCPU at startup. */ - vcpu->arch.sie_block->ihcpu = 0xffff; - spin_unlock_bh(&vcpu->kvm->arch.start_stop_lock); + kvm_make_request(KVM_REQ_TLB_FLUSH, vcpu); + spin_unlock(&vcpu->kvm->arch.start_stop_lock); return; } @@ -1491,10 +1547,18 @@ void kvm_s390_vcpu_stop(struct kvm_vcpu *vcpu) trace_kvm_s390_vcpu_start_stop(vcpu->vcpu_id, 0); /* Only one cpu at a time may enter/leave the STOPPED state. */ - spin_lock_bh(&vcpu->kvm->arch.start_stop_lock); + spin_lock(&vcpu->kvm->arch.start_stop_lock); online_vcpus = atomic_read(&vcpu->kvm->online_vcpus); + /* Need to lock access to action_bits to avoid a SIGP race condition */ + spin_lock(&vcpu->arch.local_int.lock); atomic_set_mask(CPUSTAT_STOPPED, &vcpu->arch.sie_block->cpuflags); + + /* SIGP STOP and SIGP STOP AND STORE STATUS has been fully processed */ + vcpu->arch.local_int.action_bits &= + ~(ACTION_STOP_ON_STOP | ACTION_STORE_ON_STOP); + spin_unlock(&vcpu->arch.local_int.lock); + __disable_ibs_on_vcpu(vcpu); for (i = 0; i < online_vcpus; i++) { @@ -1512,7 +1576,7 @@ void kvm_s390_vcpu_stop(struct kvm_vcpu *vcpu) __enable_ibs_on_vcpu(started_vcpu); } - spin_unlock_bh(&vcpu->kvm->arch.start_stop_lock); + spin_unlock(&vcpu->kvm->arch.start_stop_lock); return; } @@ -1623,9 +1687,7 @@ long kvm_arch_vcpu_ioctl(struct file *filp, } #endif case KVM_S390_VCPU_FAULT: { - r = gmap_fault(arg, vcpu->arch.gmap); - if (!IS_ERR_VALUE(r)) - r = 0; + r = gmap_fault(vcpu->arch.gmap, arg, 0); break; } case KVM_ENABLE_CAP: @@ -1656,21 +1718,12 @@ int kvm_arch_vcpu_fault(struct kvm_vcpu *vcpu, struct vm_fault *vmf) return VM_FAULT_SIGBUS; } -void kvm_arch_free_memslot(struct kvm *kvm, struct kvm_memory_slot *free, - struct kvm_memory_slot *dont) -{ -} - int kvm_arch_create_memslot(struct kvm *kvm, struct kvm_memory_slot *slot, unsigned long npages) { return 0; } -void kvm_arch_memslots_updated(struct kvm *kvm) -{ -} - /* Section: memory related */ int kvm_arch_prepare_memory_region(struct kvm *kvm, struct kvm_memory_slot *memslot, @@ -1716,15 +1769,6 @@ void kvm_arch_commit_memory_region(struct kvm *kvm, return; } -void kvm_arch_flush_shadow_all(struct kvm *kvm) -{ -} - -void kvm_arch_flush_shadow_memslot(struct kvm *kvm, - struct kvm_memory_slot *slot) -{ -} - static int __init kvm_s390_init(void) { int ret; @@ -1743,7 +1787,7 @@ static int __init kvm_s390_init(void) return -ENOMEM; } memcpy(vfacilities, S390_lowcore.stfle_fac_list, 16); - vfacilities[0] &= 0xff82fff3f4fc2000UL; + vfacilities[0] &= 0xff82fffbf47c2000UL; vfacilities[1] &= 0x005c000000000000UL; return 0; } diff --git a/arch/s390/kvm/kvm-s390.h b/arch/s390/kvm/kvm-s390.h index a8655ed..244d023 100644 --- a/arch/s390/kvm/kvm-s390.h +++ b/arch/s390/kvm/kvm-s390.h @@ -45,9 +45,9 @@ do { \ d_args); \ } while (0) -static inline int __cpu_is_stopped(struct kvm_vcpu *vcpu) +static inline int is_vcpu_stopped(struct kvm_vcpu *vcpu) { - return atomic_read(&vcpu->arch.sie_block->cpuflags) & CPUSTAT_STOP_INT; + return atomic_read(&vcpu->arch.sie_block->cpuflags) & CPUSTAT_STOPPED; } static inline int kvm_is_ucontrol(struct kvm *kvm) @@ -70,7 +70,7 @@ static inline u32 kvm_s390_get_prefix(struct kvm_vcpu *vcpu) static inline void kvm_s390_set_prefix(struct kvm_vcpu *vcpu, u32 prefix) { vcpu->arch.sie_block->prefix = prefix >> GUEST_PREFIX_SHIFT; - vcpu->arch.sie_block->ihcpu = 0xffff; + kvm_make_request(KVM_REQ_TLB_FLUSH, vcpu); kvm_make_request(KVM_REQ_MMU_RELOAD, vcpu); } @@ -129,11 +129,16 @@ static inline void kvm_s390_set_psw_cc(struct kvm_vcpu *vcpu, unsigned long cc) vcpu->arch.sie_block->gpsw.mask |= cc << 44; } +/* are cpu states controlled by user space */ +static inline int kvm_s390_user_cpu_state_ctrl(struct kvm *kvm) +{ + return kvm->arch.user_cpu_state_ctrl != 0; +} + int kvm_s390_handle_wait(struct kvm_vcpu *vcpu); +void kvm_s390_vcpu_wakeup(struct kvm_vcpu *vcpu); enum hrtimer_restart kvm_s390_idle_wakeup(struct hrtimer *timer); -void kvm_s390_tasklet(unsigned long parm); -void kvm_s390_deliver_pending_interrupts(struct kvm_vcpu *vcpu); -void kvm_s390_deliver_pending_machine_checks(struct kvm_vcpu *vcpu); +int __must_check kvm_s390_deliver_pending_interrupts(struct kvm_vcpu *vcpu); void kvm_s390_clear_local_irqs(struct kvm_vcpu *vcpu); void kvm_s390_clear_float_irqs(struct kvm *kvm); int __must_check kvm_s390_inject_vm(struct kvm *kvm, @@ -222,6 +227,7 @@ int kvm_cpu_has_interrupt(struct kvm_vcpu *vcpu); int psw_extint_disabled(struct kvm_vcpu *vcpu); void kvm_s390_destroy_adapters(struct kvm *kvm); int kvm_s390_si_ext_call_pending(struct kvm_vcpu *vcpu); +extern struct kvm_device_ops kvm_flic_ops; /* implemented in guestdbg.c */ void kvm_s390_backup_guest_per_regs(struct kvm_vcpu *vcpu); diff --git a/arch/s390/kvm/priv.c b/arch/s390/kvm/priv.c index f89c1cd..72bb2dd 100644 --- a/arch/s390/kvm/priv.c +++ b/arch/s390/kvm/priv.c @@ -352,13 +352,6 @@ static int handle_stfl(struct kvm_vcpu *vcpu) return 0; } -static void handle_new_psw(struct kvm_vcpu *vcpu) -{ - /* Check whether the new psw is enabled for machine checks. */ - if (vcpu->arch.sie_block->gpsw.mask & PSW_MASK_MCHECK) - kvm_s390_deliver_pending_machine_checks(vcpu); -} - #define PSW_MASK_ADDR_MODE (PSW_MASK_EA | PSW_MASK_BA) #define PSW_MASK_UNASSIGNED 0xb80800fe7fffffffUL #define PSW_ADDR_24 0x0000000000ffffffUL @@ -405,7 +398,6 @@ int kvm_s390_handle_lpsw(struct kvm_vcpu *vcpu) gpsw->addr = new_psw.addr & ~PSW32_ADDR_AMODE; if (!is_valid_psw(gpsw)) return kvm_s390_inject_program_int(vcpu, PGM_SPECIFICATION); - handle_new_psw(vcpu); return 0; } @@ -427,7 +419,6 @@ static int handle_lpswe(struct kvm_vcpu *vcpu) vcpu->arch.sie_block->gpsw = new_psw; if (!is_valid_psw(&vcpu->arch.sie_block->gpsw)) return kvm_s390_inject_program_int(vcpu, PGM_SPECIFICATION); - handle_new_psw(vcpu); return 0; } @@ -738,7 +729,7 @@ static int handle_essa(struct kvm_vcpu *vcpu) /* invalid entry */ break; /* try to free backing */ - __gmap_zap(cbrle, gmap); + __gmap_zap(gmap, cbrle); } up_read(&gmap->mm->mmap_sem); if (i < entries) diff --git a/arch/s390/kvm/sigp.c b/arch/s390/kvm/sigp.c index 43079a4..cf243ba 100644 --- a/arch/s390/kvm/sigp.c +++ b/arch/s390/kvm/sigp.c @@ -125,8 +125,9 @@ static int __sigp_external_call(struct kvm_vcpu *vcpu, u16 cpu_addr) return rc ? rc : SIGP_CC_ORDER_CODE_ACCEPTED; } -static int __inject_sigp_stop(struct kvm_s390_local_interrupt *li, int action) +static int __inject_sigp_stop(struct kvm_vcpu *dst_vcpu, int action) { + struct kvm_s390_local_interrupt *li = &dst_vcpu->arch.local_int; struct kvm_s390_interrupt_info *inti; int rc = SIGP_CC_ORDER_CODE_ACCEPTED; @@ -135,7 +136,13 @@ static int __inject_sigp_stop(struct kvm_s390_local_interrupt *li, int action) return -ENOMEM; inti->type = KVM_S390_SIGP_STOP; - spin_lock_bh(&li->lock); + spin_lock(&li->lock); + if (li->action_bits & ACTION_STOP_ON_STOP) { + /* another SIGP STOP is pending */ + kfree(inti); + rc = SIGP_CC_BUSY; + goto out; + } if ((atomic_read(li->cpuflags) & CPUSTAT_STOPPED)) { kfree(inti); if ((action & ACTION_STORE_ON_STOP) != 0) @@ -144,19 +151,17 @@ static int __inject_sigp_stop(struct kvm_s390_local_interrupt *li, int action) } list_add_tail(&inti->list, &li->list); atomic_set(&li->active, 1); - atomic_set_mask(CPUSTAT_STOP_INT, li->cpuflags); li->action_bits |= action; - if (waitqueue_active(li->wq)) - wake_up_interruptible(li->wq); + atomic_set_mask(CPUSTAT_STOP_INT, li->cpuflags); + kvm_s390_vcpu_wakeup(dst_vcpu); out: - spin_unlock_bh(&li->lock); + spin_unlock(&li->lock); return rc; } static int __sigp_stop(struct kvm_vcpu *vcpu, u16 cpu_addr, int action) { - struct kvm_s390_local_interrupt *li; struct kvm_vcpu *dst_vcpu = NULL; int rc; @@ -166,9 +171,8 @@ static int __sigp_stop(struct kvm_vcpu *vcpu, u16 cpu_addr, int action) dst_vcpu = kvm_get_vcpu(vcpu->kvm, cpu_addr); if (!dst_vcpu) return SIGP_CC_NOT_OPERATIONAL; - li = &dst_vcpu->arch.local_int; - rc = __inject_sigp_stop(li, action); + rc = __inject_sigp_stop(dst_vcpu, action); VCPU_EVENT(vcpu, 4, "sent sigp stop to cpu %x", cpu_addr); @@ -238,7 +242,7 @@ static int __sigp_set_prefix(struct kvm_vcpu *vcpu, u16 cpu_addr, u32 address, if (!inti) return SIGP_CC_BUSY; - spin_lock_bh(&li->lock); + spin_lock(&li->lock); /* cpu must be in stopped state */ if (!(atomic_read(li->cpuflags) & CPUSTAT_STOPPED)) { *reg &= 0xffffffff00000000UL; @@ -253,13 +257,12 @@ static int __sigp_set_prefix(struct kvm_vcpu *vcpu, u16 cpu_addr, u32 address, list_add_tail(&inti->list, &li->list); atomic_set(&li->active, 1); - if (waitqueue_active(li->wq)) - wake_up_interruptible(li->wq); + kvm_s390_vcpu_wakeup(dst_vcpu); rc = SIGP_CC_ORDER_CODE_ACCEPTED; VCPU_EVENT(vcpu, 4, "set prefix of cpu %02x to %x", cpu_addr, address); out_li: - spin_unlock_bh(&li->lock); + spin_unlock(&li->lock); return rc; } @@ -275,9 +278,9 @@ static int __sigp_store_status_at_addr(struct kvm_vcpu *vcpu, u16 cpu_id, if (!dst_vcpu) return SIGP_CC_NOT_OPERATIONAL; - spin_lock_bh(&dst_vcpu->arch.local_int.lock); + spin_lock(&dst_vcpu->arch.local_int.lock); flags = atomic_read(dst_vcpu->arch.local_int.cpuflags); - spin_unlock_bh(&dst_vcpu->arch.local_int.lock); + spin_unlock(&dst_vcpu->arch.local_int.lock); if (!(flags & CPUSTAT_STOPPED)) { *reg &= 0xffffffff00000000UL; *reg |= SIGP_STATUS_INCORRECT_STATE; @@ -338,10 +341,10 @@ static int sigp_check_callable(struct kvm_vcpu *vcpu, u16 cpu_addr) if (!dst_vcpu) return SIGP_CC_NOT_OPERATIONAL; li = &dst_vcpu->arch.local_int; - spin_lock_bh(&li->lock); + spin_lock(&li->lock); if (li->action_bits & ACTION_STOP_ON_STOP) rc = SIGP_CC_BUSY; - spin_unlock_bh(&li->lock); + spin_unlock(&li->lock); return rc; } @@ -461,12 +464,7 @@ int kvm_s390_handle_sigp_pei(struct kvm_vcpu *vcpu) dest_vcpu = kvm_get_vcpu(vcpu->kvm, cpu_addr); BUG_ON(dest_vcpu == NULL); - spin_lock_bh(&dest_vcpu->arch.local_int.lock); - if (waitqueue_active(&dest_vcpu->wq)) - wake_up_interruptible(&dest_vcpu->wq); - dest_vcpu->preempted = true; - spin_unlock_bh(&dest_vcpu->arch.local_int.lock); - + kvm_s390_vcpu_wakeup(dest_vcpu); kvm_s390_set_psw_cc(vcpu, SIGP_CC_ORDER_CODE_ACCEPTED); return 0; } diff --git a/arch/s390/lib/Makefile b/arch/s390/lib/Makefile index c6d752e..a01df23 100644 --- a/arch/s390/lib/Makefile +++ b/arch/s390/lib/Makefile @@ -6,3 +6,5 @@ lib-y += delay.o string.o uaccess.o find.o obj-$(CONFIG_32BIT) += div64.o qrnnd.o ucmpdi2.o mem32.o obj-$(CONFIG_64BIT) += mem64.o lib-$(CONFIG_SMP) += spinlock.o +lib-$(CONFIG_KPROBES) += probes.o +lib-$(CONFIG_UPROBES) += probes.o diff --git a/arch/s390/lib/delay.c b/arch/s390/lib/delay.c index a9f3d00..16dc42d 100644 --- a/arch/s390/lib/delay.c +++ b/arch/s390/lib/delay.c @@ -43,7 +43,7 @@ static void __udelay_disabled(unsigned long long usecs) lockdep_off(); do { set_clock_comparator(end); - vtime_stop_cpu(); + enabled_wait(); } while (get_tod_clock_fast() < end); lockdep_on(); __ctl_load(cr0, 0, 0); @@ -62,7 +62,7 @@ static void __udelay_enabled(unsigned long long usecs) clock_saved = local_tick_disable(); set_clock_comparator(end); } - vtime_stop_cpu(); + enabled_wait(); if (clock_saved) local_tick_enable(clock_saved); } while (get_tod_clock_fast() < end); diff --git a/arch/s390/lib/probes.c b/arch/s390/lib/probes.c new file mode 100644 index 0000000..c5d64a0 --- /dev/null +++ b/arch/s390/lib/probes.c @@ -0,0 +1,159 @@ +/* + * Common helper functions for kprobes and uprobes + * + * Copyright IBM Corp. 2014 + */ + +#include <linux/kprobes.h> +#include <asm/dis.h> + +int probe_is_prohibited_opcode(u16 *insn) +{ + if (!is_known_insn((unsigned char *)insn)) + return -EINVAL; + switch (insn[0] >> 8) { + case 0x0c: /* bassm */ + case 0x0b: /* bsm */ + case 0x83: /* diag */ + case 0x44: /* ex */ + case 0xac: /* stnsm */ + case 0xad: /* stosm */ + return -EINVAL; + case 0xc6: + switch (insn[0] & 0x0f) { + case 0x00: /* exrl */ + return -EINVAL; + } + } + switch (insn[0]) { + case 0x0101: /* pr */ + case 0xb25a: /* bsa */ + case 0xb240: /* bakr */ + case 0xb258: /* bsg */ + case 0xb218: /* pc */ + case 0xb228: /* pt */ + case 0xb98d: /* epsw */ + case 0xe560: /* tbegin */ + case 0xe561: /* tbeginc */ + case 0xb2f8: /* tend */ + return -EINVAL; + } + return 0; +} + +int probe_get_fixup_type(u16 *insn) +{ + /* default fixup method */ + int fixup = FIXUP_PSW_NORMAL; + + switch (insn[0] >> 8) { + case 0x05: /* balr */ + case 0x0d: /* basr */ + fixup = FIXUP_RETURN_REGISTER; + /* if r2 = 0, no branch will be taken */ + if ((insn[0] & 0x0f) == 0) + fixup |= FIXUP_BRANCH_NOT_TAKEN; + break; + case 0x06: /* bctr */ + case 0x07: /* bcr */ + fixup = FIXUP_BRANCH_NOT_TAKEN; + break; + case 0x45: /* bal */ + case 0x4d: /* bas */ + fixup = FIXUP_RETURN_REGISTER; + break; + case 0x47: /* bc */ + case 0x46: /* bct */ + case 0x86: /* bxh */ + case 0x87: /* bxle */ + fixup = FIXUP_BRANCH_NOT_TAKEN; + break; + case 0x82: /* lpsw */ + fixup = FIXUP_NOT_REQUIRED; + break; + case 0xb2: /* lpswe */ + if ((insn[0] & 0xff) == 0xb2) + fixup = FIXUP_NOT_REQUIRED; + break; + case 0xa7: /* bras */ + if ((insn[0] & 0x0f) == 0x05) + fixup |= FIXUP_RETURN_REGISTER; + break; + case 0xc0: + if ((insn[0] & 0x0f) == 0x05) /* brasl */ + fixup |= FIXUP_RETURN_REGISTER; + break; + case 0xeb: + switch (insn[2] & 0xff) { + case 0x44: /* bxhg */ + case 0x45: /* bxleg */ + fixup = FIXUP_BRANCH_NOT_TAKEN; + break; + } + break; + case 0xe3: /* bctg */ + if ((insn[2] & 0xff) == 0x46) + fixup = FIXUP_BRANCH_NOT_TAKEN; + break; + case 0xec: + switch (insn[2] & 0xff) { + case 0xe5: /* clgrb */ + case 0xe6: /* cgrb */ + case 0xf6: /* crb */ + case 0xf7: /* clrb */ + case 0xfc: /* cgib */ + case 0xfd: /* cglib */ + case 0xfe: /* cib */ + case 0xff: /* clib */ + fixup = FIXUP_BRANCH_NOT_TAKEN; + break; + } + break; + } + return fixup; +} + +int probe_is_insn_relative_long(u16 *insn) +{ + /* Check if we have a RIL-b or RIL-c format instruction which + * we need to modify in order to avoid instruction emulation. */ + switch (insn[0] >> 8) { + case 0xc0: + if ((insn[0] & 0x0f) == 0x00) /* larl */ + return true; + break; + case 0xc4: + switch (insn[0] & 0x0f) { + case 0x02: /* llhrl */ + case 0x04: /* lghrl */ + case 0x05: /* lhrl */ + case 0x06: /* llghrl */ + case 0x07: /* sthrl */ + case 0x08: /* lgrl */ + case 0x0b: /* stgrl */ + case 0x0c: /* lgfrl */ + case 0x0d: /* lrl */ + case 0x0e: /* llgfrl */ + case 0x0f: /* strl */ + return true; + } + break; + case 0xc6: + switch (insn[0] & 0x0f) { + case 0x02: /* pfdrl */ + case 0x04: /* cghrl */ + case 0x05: /* chrl */ + case 0x06: /* clghrl */ + case 0x07: /* clhrl */ + case 0x08: /* cgrl */ + case 0x0a: /* clgrl */ + case 0x0c: /* cgfrl */ + case 0x0d: /* crl */ + case 0x0e: /* clgfrl */ + case 0x0f: /* clrl */ + return true; + } + break; + } + return false; +} diff --git a/arch/s390/lib/spinlock.c b/arch/s390/lib/spinlock.c index 5b0e445..034a35a 100644 --- a/arch/s390/lib/spinlock.c +++ b/arch/s390/lib/spinlock.c @@ -98,17 +98,6 @@ void arch_spin_lock_wait_flags(arch_spinlock_t *lp, unsigned long flags) } EXPORT_SYMBOL(arch_spin_lock_wait_flags); -void arch_spin_relax(arch_spinlock_t *lp) -{ - unsigned int cpu = lp->lock; - if (cpu != 0) { - if (MACHINE_IS_VM || MACHINE_IS_KVM || - !smp_vcpu_scheduled(~cpu)) - smp_yield_cpu(~cpu); - } -} -EXPORT_SYMBOL(arch_spin_relax); - int arch_spin_trylock_retry(arch_spinlock_t *lp) { int count; @@ -122,15 +111,21 @@ EXPORT_SYMBOL(arch_spin_trylock_retry); void _raw_read_lock_wait(arch_rwlock_t *rw) { - unsigned int old; + unsigned int owner, old; int count = spin_retry; +#ifdef CONFIG_HAVE_MARCH_Z196_FEATURES + __RAW_LOCK(&rw->lock, -1, __RAW_OP_ADD); +#endif + owner = 0; while (1) { if (count-- <= 0) { - smp_yield(); + if (owner && !smp_vcpu_scheduled(~owner)) + smp_yield_cpu(~owner); count = spin_retry; } old = ACCESS_ONCE(rw->lock); + owner = ACCESS_ONCE(rw->owner); if ((int) old < 0) continue; if (_raw_compare_and_swap(&rw->lock, old, old + 1)) @@ -139,28 +134,6 @@ void _raw_read_lock_wait(arch_rwlock_t *rw) } EXPORT_SYMBOL(_raw_read_lock_wait); -void _raw_read_lock_wait_flags(arch_rwlock_t *rw, unsigned long flags) -{ - unsigned int old; - int count = spin_retry; - - local_irq_restore(flags); - while (1) { - if (count-- <= 0) { - smp_yield(); - count = spin_retry; - } - old = ACCESS_ONCE(rw->lock); - if ((int) old < 0) - continue; - local_irq_disable(); - if (_raw_compare_and_swap(&rw->lock, old, old + 1)) - return; - local_irq_restore(flags); - } -} -EXPORT_SYMBOL(_raw_read_lock_wait_flags); - int _raw_read_trylock_retry(arch_rwlock_t *rw) { unsigned int old; @@ -177,46 +150,62 @@ int _raw_read_trylock_retry(arch_rwlock_t *rw) } EXPORT_SYMBOL(_raw_read_trylock_retry); -void _raw_write_lock_wait(arch_rwlock_t *rw) +#ifdef CONFIG_HAVE_MARCH_Z196_FEATURES + +void _raw_write_lock_wait(arch_rwlock_t *rw, unsigned int prev) { - unsigned int old; + unsigned int owner, old; int count = spin_retry; + owner = 0; while (1) { if (count-- <= 0) { - smp_yield(); + if (owner && !smp_vcpu_scheduled(~owner)) + smp_yield_cpu(~owner); count = spin_retry; } old = ACCESS_ONCE(rw->lock); - if (old) - continue; - if (_raw_compare_and_swap(&rw->lock, 0, 0x80000000)) - return; + owner = ACCESS_ONCE(rw->owner); + smp_rmb(); + if ((int) old >= 0) { + prev = __RAW_LOCK(&rw->lock, 0x80000000, __RAW_OP_OR); + old = prev; + } + if ((old & 0x7fffffff) == 0 && (int) prev >= 0) + break; } } EXPORT_SYMBOL(_raw_write_lock_wait); -void _raw_write_lock_wait_flags(arch_rwlock_t *rw, unsigned long flags) +#else /* CONFIG_HAVE_MARCH_Z196_FEATURES */ + +void _raw_write_lock_wait(arch_rwlock_t *rw) { - unsigned int old; + unsigned int owner, old, prev; int count = spin_retry; - local_irq_restore(flags); + prev = 0x80000000; + owner = 0; while (1) { if (count-- <= 0) { - smp_yield(); + if (owner && !smp_vcpu_scheduled(~owner)) + smp_yield_cpu(~owner); count = spin_retry; } old = ACCESS_ONCE(rw->lock); - if (old) - continue; - local_irq_disable(); - if (_raw_compare_and_swap(&rw->lock, 0, 0x80000000)) - return; - local_irq_restore(flags); + owner = ACCESS_ONCE(rw->owner); + if ((int) old >= 0 && + _raw_compare_and_swap(&rw->lock, old, old | 0x80000000)) + prev = old; + else + smp_rmb(); + if ((old & 0x7fffffff) == 0 && (int) prev >= 0) + break; } } -EXPORT_SYMBOL(_raw_write_lock_wait_flags); +EXPORT_SYMBOL(_raw_write_lock_wait); + +#endif /* CONFIG_HAVE_MARCH_Z196_FEATURES */ int _raw_write_trylock_retry(arch_rwlock_t *rw) { @@ -233,3 +222,13 @@ int _raw_write_trylock_retry(arch_rwlock_t *rw) return 0; } EXPORT_SYMBOL(_raw_write_trylock_retry); + +void arch_lock_relax(unsigned int cpu) +{ + if (!cpu) + return; + if (MACHINE_IS_LPAR && smp_vcpu_scheduled(~cpu)) + return; + smp_yield_cpu(~cpu); +} +EXPORT_SYMBOL(arch_lock_relax); diff --git a/arch/s390/mm/dump_pagetables.c b/arch/s390/mm/dump_pagetables.c index 46d517c..d46cade 100644 --- a/arch/s390/mm/dump_pagetables.c +++ b/arch/s390/mm/dump_pagetables.c @@ -54,7 +54,6 @@ static void print_prot(struct seq_file *m, unsigned int pr, int level) return; } seq_printf(m, "%s", pr & _PAGE_PROTECT ? "RO " : "RW "); - seq_printf(m, "%s", pr & _PAGE_CO ? "CO " : " "); seq_putc(m, '\n'); } @@ -129,7 +128,7 @@ static void walk_pte_level(struct seq_file *m, struct pg_state *st, } #ifdef CONFIG_64BIT -#define _PMD_PROT_MASK (_SEGMENT_ENTRY_PROTECT | _SEGMENT_ENTRY_CO) +#define _PMD_PROT_MASK _SEGMENT_ENTRY_PROTECT #else #define _PMD_PROT_MASK 0 #endif @@ -157,7 +156,7 @@ static void walk_pmd_level(struct seq_file *m, struct pg_state *st, } #ifdef CONFIG_64BIT -#define _PUD_PROT_MASK (_REGION3_ENTRY_RO | _REGION3_ENTRY_CO) +#define _PUD_PROT_MASK _REGION3_ENTRY_RO #else #define _PUD_PROT_MASK 0 #endif diff --git a/arch/s390/mm/fault.c b/arch/s390/mm/fault.c index 3f3b354..a2b81d6 100644 --- a/arch/s390/mm/fault.c +++ b/arch/s390/mm/fault.c @@ -442,18 +442,15 @@ static inline int do_exception(struct pt_regs *regs, int access) down_read(&mm->mmap_sem); #ifdef CONFIG_PGSTE - gmap = (struct gmap *) - ((current->flags & PF_VCPU) ? S390_lowcore.gmap : 0); + gmap = (current->flags & PF_VCPU) ? + (struct gmap *) S390_lowcore.gmap : NULL; if (gmap) { - address = __gmap_fault(address, gmap); + current->thread.gmap_addr = address; + address = __gmap_translate(gmap, address); if (address == -EFAULT) { fault = VM_FAULT_BADMAP; goto out_up; } - if (address == -ENOMEM) { - fault = VM_FAULT_OOM; - goto out_up; - } if (gmap->pfault_enabled) flags |= FAULT_FLAG_RETRY_NOWAIT; } @@ -530,6 +527,20 @@ retry: goto retry; } } +#ifdef CONFIG_PGSTE + if (gmap) { + address = __gmap_link(gmap, current->thread.gmap_addr, + address); + if (address == -EFAULT) { + fault = VM_FAULT_BADMAP; + goto out_up; + } + if (address == -ENOMEM) { + fault = VM_FAULT_OOM; + goto out_up; + } + } +#endif fault = 0; out_up: up_read(&mm->mmap_sem); diff --git a/arch/s390/mm/hugetlbpage.c b/arch/s390/mm/hugetlbpage.c index 0ff66a7..3c80d2e 100644 --- a/arch/s390/mm/hugetlbpage.c +++ b/arch/s390/mm/hugetlbpage.c @@ -10,42 +10,33 @@ static inline pmd_t __pte_to_pmd(pte_t pte) { - int none, young, prot; pmd_t pmd; /* - * Convert encoding pte bits pmd bits - * .IR...wrdytp ..R...I...y. - * empty .10...000000 -> ..0...1...0. - * prot-none, clean, old .11...000001 -> ..0...1...1. - * prot-none, clean, young .11...000101 -> ..1...1...1. - * prot-none, dirty, old .10...001001 -> ..0...1...1. - * prot-none, dirty, young .10...001101 -> ..1...1...1. - * read-only, clean, old .11...010001 -> ..1...1...0. - * read-only, clean, young .01...010101 -> ..1...0...1. - * read-only, dirty, old .11...011001 -> ..1...1...0. - * read-only, dirty, young .01...011101 -> ..1...0...1. - * read-write, clean, old .11...110001 -> ..0...1...0. - * read-write, clean, young .01...110101 -> ..0...0...1. - * read-write, dirty, old .10...111001 -> ..0...1...0. - * read-write, dirty, young .00...111101 -> ..0...0...1. - * Huge ptes are dirty by definition, a clean pte is made dirty - * by the conversion. + * Convert encoding pte bits pmd bits + * .IR...wrdytp dy..R...I...wr + * empty .10...000000 -> 00..0...1...00 + * prot-none, clean, old .11...000001 -> 00..1...1...00 + * prot-none, clean, young .11...000101 -> 01..1...1...00 + * prot-none, dirty, old .10...001001 -> 10..1...1...00 + * prot-none, dirty, young .10...001101 -> 11..1...1...00 + * read-only, clean, old .11...010001 -> 00..1...1...01 + * read-only, clean, young .01...010101 -> 01..1...0...01 + * read-only, dirty, old .11...011001 -> 10..1...1...01 + * read-only, dirty, young .01...011101 -> 11..1...0...01 + * read-write, clean, old .11...110001 -> 00..0...1...11 + * read-write, clean, young .01...110101 -> 01..0...0...11 + * read-write, dirty, old .10...111001 -> 10..0...1...11 + * read-write, dirty, young .00...111101 -> 11..0...0...11 */ if (pte_present(pte)) { pmd_val(pmd) = pte_val(pte) & PAGE_MASK; - if (pte_val(pte) & _PAGE_INVALID) - pmd_val(pmd) |= _SEGMENT_ENTRY_INVALID; - none = (pte_val(pte) & _PAGE_PRESENT) && - !(pte_val(pte) & _PAGE_READ) && - !(pte_val(pte) & _PAGE_WRITE); - prot = (pte_val(pte) & _PAGE_PROTECT) && - !(pte_val(pte) & _PAGE_WRITE); - young = pte_val(pte) & _PAGE_YOUNG; - if (none || young) - pmd_val(pmd) |= _SEGMENT_ENTRY_YOUNG; - if (prot || (none && young)) - pmd_val(pmd) |= _SEGMENT_ENTRY_PROTECT; + pmd_val(pmd) |= (pte_val(pte) & _PAGE_READ) >> 4; + pmd_val(pmd) |= (pte_val(pte) & _PAGE_WRITE) >> 4; + pmd_val(pmd) |= (pte_val(pte) & _PAGE_INVALID) >> 5; + pmd_val(pmd) |= (pte_val(pte) & _PAGE_PROTECT); + pmd_val(pmd) |= (pte_val(pte) & _PAGE_DIRTY) << 10; + pmd_val(pmd) |= (pte_val(pte) & _PAGE_YOUNG) << 10; } else pmd_val(pmd) = _SEGMENT_ENTRY_INVALID; return pmd; @@ -56,34 +47,31 @@ static inline pte_t __pmd_to_pte(pmd_t pmd) pte_t pte; /* - * Convert encoding pmd bits pte bits - * ..R...I...y. .IR...wrdytp - * empty ..0...1...0. -> .10...000000 - * prot-none, old ..0...1...1. -> .10...001001 - * prot-none, young ..1...1...1. -> .10...001101 - * read-only, old ..1...1...0. -> .11...011001 - * read-only, young ..1...0...1. -> .01...011101 - * read-write, old ..0...1...0. -> .10...111001 - * read-write, young ..0...0...1. -> .00...111101 - * Huge ptes are dirty by definition + * Convert encoding pmd bits pte bits + * dy..R...I...wr .IR...wrdytp + * empty 00..0...1...00 -> .10...001100 + * prot-none, clean, old 00..0...1...00 -> .10...000001 + * prot-none, clean, young 01..0...1...00 -> .10...000101 + * prot-none, dirty, old 10..0...1...00 -> .10...001001 + * prot-none, dirty, young 11..0...1...00 -> .10...001101 + * read-only, clean, old 00..1...1...01 -> .11...010001 + * read-only, clean, young 01..1...1...01 -> .11...010101 + * read-only, dirty, old 10..1...1...01 -> .11...011001 + * read-only, dirty, young 11..1...1...01 -> .11...011101 + * read-write, clean, old 00..0...1...11 -> .10...110001 + * read-write, clean, young 01..0...1...11 -> .10...110101 + * read-write, dirty, old 10..0...1...11 -> .10...111001 + * read-write, dirty, young 11..0...1...11 -> .10...111101 */ if (pmd_present(pmd)) { - pte_val(pte) = _PAGE_PRESENT | _PAGE_LARGE | _PAGE_DIRTY | - (pmd_val(pmd) & PAGE_MASK); - if (pmd_val(pmd) & _SEGMENT_ENTRY_INVALID) - pte_val(pte) |= _PAGE_INVALID; - if (pmd_prot_none(pmd)) { - if (pmd_val(pmd) & _SEGMENT_ENTRY_PROTECT) - pte_val(pte) |= _PAGE_YOUNG; - } else { - pte_val(pte) |= _PAGE_READ; - if (pmd_val(pmd) & _SEGMENT_ENTRY_PROTECT) - pte_val(pte) |= _PAGE_PROTECT; - else - pte_val(pte) |= _PAGE_WRITE; - if (pmd_val(pmd) & _SEGMENT_ENTRY_YOUNG) - pte_val(pte) |= _PAGE_YOUNG; - } + pte_val(pte) = pmd_val(pmd) & _SEGMENT_ENTRY_ORIGIN_LARGE; + pte_val(pte) |= _PAGE_LARGE | _PAGE_PRESENT; + pte_val(pte) |= (pmd_val(pmd) & _SEGMENT_ENTRY_READ) << 4; + pte_val(pte) |= (pmd_val(pmd) & _SEGMENT_ENTRY_WRITE) << 4; + pte_val(pte) |= (pmd_val(pmd) & _SEGMENT_ENTRY_INVALID) << 5; + pte_val(pte) |= (pmd_val(pmd) & _SEGMENT_ENTRY_PROTECT); + pmd_val(pmd) |= (pte_val(pte) & _PAGE_DIRTY) << 10; + pmd_val(pmd) |= (pte_val(pte) & _PAGE_YOUNG) << 10; } else pte_val(pte) = _PAGE_INVALID; return pte; @@ -96,10 +84,11 @@ void set_huge_pte_at(struct mm_struct *mm, unsigned long addr, pmd = __pte_to_pmd(pte); if (!MACHINE_HAS_HPAGE) { + /* Emulated huge ptes loose the dirty and young bit */ pmd_val(pmd) &= ~_SEGMENT_ENTRY_ORIGIN; pmd_val(pmd) |= pte_page(pte)[1].index; } else - pmd_val(pmd) |= _SEGMENT_ENTRY_LARGE | _SEGMENT_ENTRY_CO; + pmd_val(pmd) |= _SEGMENT_ENTRY_LARGE; *(pmd_t *) ptep = pmd; } @@ -113,6 +102,8 @@ pte_t huge_ptep_get(pte_t *ptep) origin = pmd_val(pmd) & _SEGMENT_ENTRY_ORIGIN; pmd_val(pmd) &= ~_SEGMENT_ENTRY_ORIGIN; pmd_val(pmd) |= *(unsigned long *) origin; + /* Emulated huge ptes are young and dirty by definition */ + pmd_val(pmd) |= _SEGMENT_ENTRY_YOUNG | _SEGMENT_ENTRY_DIRTY; } return __pmd_to_pte(pmd); } diff --git a/arch/s390/mm/init.c b/arch/s390/mm/init.c index 0c1073e..c7235e0 100644 --- a/arch/s390/mm/init.c +++ b/arch/s390/mm/init.c @@ -43,6 +43,7 @@ pgd_t swapper_pg_dir[PTRS_PER_PGD] __attribute__((__aligned__(PAGE_SIZE))); unsigned long empty_zero_page, zero_page_mask; EXPORT_SYMBOL(empty_zero_page); +EXPORT_SYMBOL(zero_page_mask); static void __init setup_zero_pages(void) { diff --git a/arch/s390/mm/pageattr.c b/arch/s390/mm/pageattr.c index 8400f49..3fef3b2 100644 --- a/arch/s390/mm/pageattr.c +++ b/arch/s390/mm/pageattr.c @@ -6,6 +6,7 @@ #include <linux/module.h> #include <linux/mm.h> #include <asm/cacheflush.h> +#include <asm/facility.h> #include <asm/pgtable.h> #include <asm/page.h> @@ -103,27 +104,50 @@ int set_memory_x(unsigned long addr, int numpages) } #ifdef CONFIG_DEBUG_PAGEALLOC + +static void ipte_range(pte_t *pte, unsigned long address, int nr) +{ + int i; + + if (test_facility(13) && IS_ENABLED(CONFIG_64BIT)) { + __ptep_ipte_range(address, nr - 1, pte); + return; + } + for (i = 0; i < nr; i++) { + __ptep_ipte(address, pte); + address += PAGE_SIZE; + pte++; + } +} + void kernel_map_pages(struct page *page, int numpages, int enable) { unsigned long address; + int nr, i, j; pgd_t *pgd; pud_t *pud; pmd_t *pmd; pte_t *pte; - int i; - for (i = 0; i < numpages; i++) { + for (i = 0; i < numpages;) { address = page_to_phys(page + i); pgd = pgd_offset_k(address); pud = pud_offset(pgd, address); pmd = pmd_offset(pud, address); pte = pte_offset_kernel(pmd, address); - if (!enable) { - __ptep_ipte(address, pte); - pte_val(*pte) = _PAGE_INVALID; - continue; + nr = (unsigned long)pte >> ilog2(sizeof(long)); + nr = PTRS_PER_PTE - (nr & (PTRS_PER_PTE - 1)); + nr = min(numpages - i, nr); + if (enable) { + for (j = 0; j < nr; j++) { + pte_val(*pte) = __pa(address); + address += PAGE_SIZE; + pte++; + } + } else { + ipte_range(pte, address, nr); } - pte_val(*pte) = __pa(address); + i += nr; } } diff --git a/arch/s390/mm/pgtable.c b/arch/s390/mm/pgtable.c index 37b8241..296b61a 100644 --- a/arch/s390/mm/pgtable.c +++ b/arch/s390/mm/pgtable.c @@ -145,30 +145,56 @@ void crst_table_downgrade(struct mm_struct *mm, unsigned long limit) /** * gmap_alloc - allocate 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_alloc(struct mm_struct *mm) +struct gmap *gmap_alloc(struct mm_struct *mm, unsigned long limit) { struct gmap *gmap; struct page *page; unsigned long *table; - + unsigned long etype, atype; + + if (limit < (1UL << 31)) { + limit = (1UL << 31) - 1; + atype = _ASCE_TYPE_SEGMENT; + etype = _SEGMENT_ENTRY_EMPTY; + } else if (limit < (1UL << 42)) { + limit = (1UL << 42) - 1; + atype = _ASCE_TYPE_REGION3; + etype = _REGION3_ENTRY_EMPTY; + } else if (limit < (1UL << 53)) { + limit = (1UL << 53) - 1; + atype = _ASCE_TYPE_REGION2; + etype = _REGION2_ENTRY_EMPTY; + } else { + limit = -1UL; + atype = _ASCE_TYPE_REGION1; + etype = _REGION1_ENTRY_EMPTY; + } gmap = kzalloc(sizeof(struct gmap), GFP_KERNEL); if (!gmap) goto out; INIT_LIST_HEAD(&gmap->crst_list); + INIT_RADIX_TREE(&gmap->guest_to_host, GFP_KERNEL); + INIT_RADIX_TREE(&gmap->host_to_guest, GFP_ATOMIC); + spin_lock_init(&gmap->guest_table_lock); gmap->mm = mm; page = alloc_pages(GFP_KERNEL, ALLOC_ORDER); if (!page) goto out_free; + page->index = 0; list_add(&page->lru, &gmap->crst_list); table = (unsigned long *) page_to_phys(page); - crst_table_init(table, _REGION1_ENTRY_EMPTY); + crst_table_init(table, etype); gmap->table = table; - gmap->asce = _ASCE_TYPE_REGION1 | _ASCE_TABLE_LENGTH | - _ASCE_USER_BITS | __pa(table); + 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: @@ -178,36 +204,38 @@ out: } EXPORT_SYMBOL_GPL(gmap_alloc); -static int gmap_unlink_segment(struct gmap *gmap, unsigned long *table) -{ - struct gmap_pgtable *mp; - struct gmap_rmap *rmap; - struct page *page; - - if (*table & _SEGMENT_ENTRY_INVALID) - return 0; - page = pfn_to_page(*table >> PAGE_SHIFT); - mp = (struct gmap_pgtable *) page->index; - list_for_each_entry(rmap, &mp->mapper, list) { - if (rmap->entry != table) - continue; - list_del(&rmap->list); - kfree(rmap); - break; - } - *table = mp->vmaddr | _SEGMENT_ENTRY_INVALID | _SEGMENT_ENTRY_PROTECT; - return 1; -} - static void gmap_flush_tlb(struct gmap *gmap) { if (MACHINE_HAS_IDTE) - __tlb_flush_asce(gmap->mm, (unsigned long) gmap->table | - _ASCE_TYPE_REGION1); + __tlb_flush_asce(gmap->mm, gmap->asce); else __tlb_flush_global(); } +static void gmap_radix_tree_free(struct radix_tree_root *root) +{ + 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]; + radix_tree_delete(root, index); + } + } while (nr > 0); +} + /** * gmap_free - free a guest address space * @gmap: pointer to the guest address space structure @@ -215,31 +243,21 @@ static void gmap_flush_tlb(struct gmap *gmap) void gmap_free(struct gmap *gmap) { struct page *page, *next; - unsigned long *table; - int i; - /* Flush tlb. */ if (MACHINE_HAS_IDTE) - __tlb_flush_asce(gmap->mm, (unsigned long) gmap->table | - _ASCE_TYPE_REGION1); + __tlb_flush_asce(gmap->mm, gmap->asce); else __tlb_flush_global(); /* Free all segment & region tables. */ - down_read(&gmap->mm->mmap_sem); - spin_lock(&gmap->mm->page_table_lock); - list_for_each_entry_safe(page, next, &gmap->crst_list, lru) { - table = (unsigned long *) page_to_phys(page); - if ((*table & _REGION_ENTRY_TYPE_MASK) == 0) - /* Remove gmap rmap structures for segment table. */ - for (i = 0; i < PTRS_PER_PMD; i++, table++) - gmap_unlink_segment(gmap, table); + list_for_each_entry_safe(page, next, &gmap->crst_list, lru) __free_pages(page, ALLOC_ORDER); - } - spin_unlock(&gmap->mm->page_table_lock); - up_read(&gmap->mm->mmap_sem); + 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); kfree(gmap); } EXPORT_SYMBOL_GPL(gmap_free); @@ -267,42 +285,97 @@ EXPORT_SYMBOL_GPL(gmap_disable); /* * gmap_alloc_table is assumed to be called with mmap_sem held */ -static int gmap_alloc_table(struct gmap *gmap, - unsigned long *table, unsigned long init) - __releases(&gmap->mm->page_table_lock) - __acquires(&gmap->mm->page_table_lock) +static int gmap_alloc_table(struct gmap *gmap, unsigned long *table, + unsigned long init, unsigned long gaddr) { struct page *page; unsigned long *new; /* since we dont free the gmap table until gmap_free we can unlock */ - spin_unlock(&gmap->mm->page_table_lock); page = alloc_pages(GFP_KERNEL, ALLOC_ORDER); - spin_lock(&gmap->mm->page_table_lock); if (!page) return -ENOMEM; new = (unsigned long *) page_to_phys(page); crst_table_init(new, init); + spin_lock(&gmap->mm->page_table_lock); if (*table & _REGION_ENTRY_INVALID) { list_add(&page->lru, &gmap->crst_list); *table = (unsigned long) new | _REGION_ENTRY_LENGTH | (*table & _REGION_ENTRY_TYPE_MASK); - } else + page->index = gaddr; + page = NULL; + } + spin_unlock(&gmap->mm->page_table_lock); + if (page) __free_pages(page, ALLOC_ORDER); return 0; } /** + * __gmap_segment_gaddr - find virtual address from segment pointer + * @entry: pointer to a segment table entry in the guest address space + * + * Returns the virtual address in the guest address space for the segment + */ +static unsigned long __gmap_segment_gaddr(unsigned long *entry) +{ + struct page *page; + unsigned long offset; + + offset = (unsigned long) entry / sizeof(unsigned long); + offset = (offset & (PTRS_PER_PMD - 1)) * PMD_SIZE; + page = pmd_to_page((pmd_t *) entry); + return page->index + offset; +} + +/** + * __gmap_unlink_by_vmaddr - unlink a single segment via a host address + * @gmap: pointer to the guest address space structure + * @vmaddr: address in the host process address space + * + * Returns 1 if a TLB flush is required + */ +static int __gmap_unlink_by_vmaddr(struct gmap *gmap, unsigned long vmaddr) +{ + unsigned long *entry; + int flush = 0; + + spin_lock(&gmap->guest_table_lock); + entry = radix_tree_delete(&gmap->host_to_guest, vmaddr >> PMD_SHIFT); + if (entry) { + flush = (*entry != _SEGMENT_ENTRY_INVALID); + *entry = _SEGMENT_ENTRY_INVALID; + } + spin_unlock(&gmap->guest_table_lock); + return flush; +} + +/** + * __gmap_unmap_by_gaddr - unmap a single segment via a guest address + * @gmap: pointer to the guest address space structure + * @gaddr: address in the guest address space + * + * Returns 1 if a TLB flush is required + */ +static int __gmap_unmap_by_gaddr(struct gmap *gmap, unsigned long gaddr) +{ + unsigned long vmaddr; + + vmaddr = (unsigned long) radix_tree_delete(&gmap->guest_to_host, + gaddr >> PMD_SHIFT); + return vmaddr ? __gmap_unlink_by_vmaddr(gmap, vmaddr) : 0; +} + +/** * gmap_unmap_segment - unmap segment from the guest address space * @gmap: pointer to the guest address space structure - * @addr: address in the guest address space + * @to: address in the guest address space * @len: length of the memory area to unmap * * Returns 0 if the unmap succeeded, -EINVAL if not. */ int gmap_unmap_segment(struct gmap *gmap, unsigned long to, unsigned long len) { - unsigned long *table; unsigned long off; int flush; @@ -312,31 +385,10 @@ int gmap_unmap_segment(struct gmap *gmap, unsigned long to, unsigned long len) return -EINVAL; flush = 0; - down_read(&gmap->mm->mmap_sem); - spin_lock(&gmap->mm->page_table_lock); - for (off = 0; off < len; off += PMD_SIZE) { - /* Walk the guest addr space page table */ - table = gmap->table + (((to + off) >> 53) & 0x7ff); - if (*table & _REGION_ENTRY_INVALID) - goto out; - table = (unsigned long *)(*table & _REGION_ENTRY_ORIGIN); - table = table + (((to + off) >> 42) & 0x7ff); - if (*table & _REGION_ENTRY_INVALID) - goto out; - table = (unsigned long *)(*table & _REGION_ENTRY_ORIGIN); - table = table + (((to + off) >> 31) & 0x7ff); - if (*table & _REGION_ENTRY_INVALID) - goto out; - table = (unsigned long *)(*table & _REGION_ENTRY_ORIGIN); - table = table + (((to + off) >> 20) & 0x7ff); - - /* Clear segment table entry in guest address space. */ - flush |= gmap_unlink_segment(gmap, table); - *table = _SEGMENT_ENTRY_INVALID; - } -out: - spin_unlock(&gmap->mm->page_table_lock); - up_read(&gmap->mm->mmap_sem); + down_write(&gmap->mm->mmap_sem); + for (off = 0; off < len; off += PMD_SIZE) + flush |= __gmap_unmap_by_gaddr(gmap, to + off); + up_write(&gmap->mm->mmap_sem); if (flush) gmap_flush_tlb(gmap); return 0; @@ -348,87 +400,47 @@ EXPORT_SYMBOL_GPL(gmap_unmap_segment); * @gmap: pointer to the guest address space structure * @from: source address in the parent address space * @to: target address in the guest address space + * @len: length of the memory area to map * * Returns 0 if the mmap succeeded, -EINVAL or -ENOMEM if not. */ int gmap_map_segment(struct gmap *gmap, unsigned long from, unsigned long to, unsigned long len) { - unsigned long *table; unsigned long off; int flush; if ((from | to | len) & (PMD_SIZE - 1)) return -EINVAL; - if (len == 0 || from + len > TASK_MAX_SIZE || - from + len < from || to + len < to) + if (len == 0 || from + len < from || to + len < to || + from + len > TASK_MAX_SIZE || to + len > gmap->asce_end) return -EINVAL; flush = 0; - down_read(&gmap->mm->mmap_sem); - spin_lock(&gmap->mm->page_table_lock); + down_write(&gmap->mm->mmap_sem); for (off = 0; off < len; off += PMD_SIZE) { - /* Walk the gmap address space page table */ - table = gmap->table + (((to + off) >> 53) & 0x7ff); - if ((*table & _REGION_ENTRY_INVALID) && - gmap_alloc_table(gmap, table, _REGION2_ENTRY_EMPTY)) - goto out_unmap; - table = (unsigned long *)(*table & _REGION_ENTRY_ORIGIN); - table = table + (((to + off) >> 42) & 0x7ff); - if ((*table & _REGION_ENTRY_INVALID) && - gmap_alloc_table(gmap, table, _REGION3_ENTRY_EMPTY)) - goto out_unmap; - table = (unsigned long *)(*table & _REGION_ENTRY_ORIGIN); - table = table + (((to + off) >> 31) & 0x7ff); - if ((*table & _REGION_ENTRY_INVALID) && - gmap_alloc_table(gmap, table, _SEGMENT_ENTRY_EMPTY)) - goto out_unmap; - table = (unsigned long *) (*table & _REGION_ENTRY_ORIGIN); - table = table + (((to + off) >> 20) & 0x7ff); - - /* Store 'from' address in an invalid segment table entry. */ - flush |= gmap_unlink_segment(gmap, table); - *table = (from + off) | (_SEGMENT_ENTRY_INVALID | - _SEGMENT_ENTRY_PROTECT); + /* Remove old translation */ + flush |= __gmap_unmap_by_gaddr(gmap, to + off); + /* Store new translation */ + if (radix_tree_insert(&gmap->guest_to_host, + (to + off) >> PMD_SHIFT, + (void *) from + off)) + break; } - spin_unlock(&gmap->mm->page_table_lock); - up_read(&gmap->mm->mmap_sem); + up_write(&gmap->mm->mmap_sem); if (flush) gmap_flush_tlb(gmap); - return 0; - -out_unmap: - spin_unlock(&gmap->mm->page_table_lock); - up_read(&gmap->mm->mmap_sem); + if (off >= len) + return 0; gmap_unmap_segment(gmap, to, len); return -ENOMEM; } EXPORT_SYMBOL_GPL(gmap_map_segment); -static unsigned long *gmap_table_walk(unsigned long address, struct gmap *gmap) -{ - unsigned long *table; - - table = gmap->table + ((address >> 53) & 0x7ff); - if (unlikely(*table & _REGION_ENTRY_INVALID)) - return ERR_PTR(-EFAULT); - table = (unsigned long *)(*table & _REGION_ENTRY_ORIGIN); - table = table + ((address >> 42) & 0x7ff); - if (unlikely(*table & _REGION_ENTRY_INVALID)) - return ERR_PTR(-EFAULT); - table = (unsigned long *)(*table & _REGION_ENTRY_ORIGIN); - table = table + ((address >> 31) & 0x7ff); - if (unlikely(*table & _REGION_ENTRY_INVALID)) - return ERR_PTR(-EFAULT); - table = (unsigned long *)(*table & _REGION_ENTRY_ORIGIN); - table = table + ((address >> 20) & 0x7ff); - return table; -} - /** * __gmap_translate - translate a guest address to a user space address - * @address: guest address * @gmap: pointer to guest mapping meta data structure + * @gaddr: guest address * * Returns user space address which corresponds to the guest address or * -EFAULT if no such mapping exists. @@ -436,168 +448,161 @@ static unsigned long *gmap_table_walk(unsigned long address, struct gmap *gmap) * The mmap_sem of the mm that belongs to the address space must be held * when this function gets called. */ -unsigned long __gmap_translate(unsigned long address, struct gmap *gmap) +unsigned long __gmap_translate(struct gmap *gmap, unsigned long gaddr) { - unsigned long *segment_ptr, vmaddr, segment; - struct gmap_pgtable *mp; - struct page *page; + unsigned long vmaddr; - current->thread.gmap_addr = address; - segment_ptr = gmap_table_walk(address, gmap); - if (IS_ERR(segment_ptr)) - return PTR_ERR(segment_ptr); - /* Convert the gmap address to an mm address. */ - segment = *segment_ptr; - if (!(segment & _SEGMENT_ENTRY_INVALID)) { - page = pfn_to_page(segment >> PAGE_SHIFT); - mp = (struct gmap_pgtable *) page->index; - return mp->vmaddr | (address & ~PMD_MASK); - } else if (segment & _SEGMENT_ENTRY_PROTECT) { - vmaddr = segment & _SEGMENT_ENTRY_ORIGIN; - return vmaddr | (address & ~PMD_MASK); - } - return -EFAULT; + vmaddr = (unsigned long) + radix_tree_lookup(&gmap->guest_to_host, gaddr >> PMD_SHIFT); + return vmaddr ? (vmaddr | (gaddr & ~PMD_MASK)) : -EFAULT; } EXPORT_SYMBOL_GPL(__gmap_translate); /** * gmap_translate - translate a guest address to a user space address - * @address: guest address * @gmap: pointer to guest mapping meta data structure + * @gaddr: guest address * * Returns user space address which corresponds to the guest address or * -EFAULT if no such mapping exists. * This function does not establish potentially missing page table entries. */ -unsigned long gmap_translate(unsigned long address, struct gmap *gmap) +unsigned long gmap_translate(struct gmap *gmap, unsigned long gaddr) { unsigned long rc; down_read(&gmap->mm->mmap_sem); - rc = __gmap_translate(address, gmap); + rc = __gmap_translate(gmap, gaddr); up_read(&gmap->mm->mmap_sem); return rc; } EXPORT_SYMBOL_GPL(gmap_translate); -static int gmap_connect_pgtable(unsigned long address, unsigned long segment, - unsigned long *segment_ptr, struct gmap *gmap) +/** + * gmap_unlink - disconnect a page table from the gmap shadow tables + * @gmap: pointer to guest mapping meta data structure + * @table: pointer to the host page table + * @vmaddr: vm address associated with the host page table + */ +static void gmap_unlink(struct mm_struct *mm, unsigned long *table, + unsigned long vmaddr) +{ + struct gmap *gmap; + int flush; + + list_for_each_entry(gmap, &mm->context.gmap_list, list) { + flush = __gmap_unlink_by_vmaddr(gmap, vmaddr); + if (flush) + gmap_flush_tlb(gmap); + } +} + +/** + * gmap_link - set up shadow page tables to connect a host to a guest address + * @gmap: pointer to guest mapping meta data structure + * @gaddr: guest address + * @vmaddr: vm address + * + * Returns 0 on success, -ENOMEM for out of memory conditions, and -EFAULT + * if the vm address is already mapped to a different guest segment. + * The mmap_sem of the mm that belongs to the address space must be held + * when this function gets called. + */ +int __gmap_link(struct gmap *gmap, unsigned long gaddr, unsigned long vmaddr) { - unsigned long vmaddr; - struct vm_area_struct *vma; - struct gmap_pgtable *mp; - struct gmap_rmap *rmap; struct mm_struct *mm; - struct page *page; + unsigned long *table; + spinlock_t *ptl; pgd_t *pgd; pud_t *pud; pmd_t *pmd; + int rc; - mm = gmap->mm; - vmaddr = segment & _SEGMENT_ENTRY_ORIGIN; - vma = find_vma(mm, vmaddr); - if (!vma || vma->vm_start > vmaddr) - return -EFAULT; + /* Create higher level tables in the gmap page table */ + table = gmap->table; + if ((gmap->asce & _ASCE_TYPE_MASK) >= _ASCE_TYPE_REGION1) { + table += (gaddr >> 53) & 0x7ff; + if ((*table & _REGION_ENTRY_INVALID) && + gmap_alloc_table(gmap, table, _REGION2_ENTRY_EMPTY, + gaddr & 0xffe0000000000000)) + return -ENOMEM; + table = (unsigned long *)(*table & _REGION_ENTRY_ORIGIN); + } + if ((gmap->asce & _ASCE_TYPE_MASK) >= _ASCE_TYPE_REGION2) { + table += (gaddr >> 42) & 0x7ff; + if ((*table & _REGION_ENTRY_INVALID) && + gmap_alloc_table(gmap, table, _REGION3_ENTRY_EMPTY, + gaddr & 0xfffffc0000000000)) + return -ENOMEM; + table = (unsigned long *)(*table & _REGION_ENTRY_ORIGIN); + } + if ((gmap->asce & _ASCE_TYPE_MASK) >= _ASCE_TYPE_REGION3) { + table += (gaddr >> 31) & 0x7ff; + if ((*table & _REGION_ENTRY_INVALID) && + gmap_alloc_table(gmap, table, _SEGMENT_ENTRY_EMPTY, + gaddr & 0xffffffff80000000)) + return -ENOMEM; + table = (unsigned long *)(*table & _REGION_ENTRY_ORIGIN); + } + table += (gaddr >> 20) & 0x7ff; /* Walk the parent mm page table */ + mm = gmap->mm; pgd = pgd_offset(mm, vmaddr); - pud = pud_alloc(mm, pgd, vmaddr); - if (!pud) - return -ENOMEM; - pmd = pmd_alloc(mm, pud, vmaddr); - if (!pmd) - return -ENOMEM; - if (!pmd_present(*pmd) && - __pte_alloc(mm, vma, pmd, vmaddr)) - return -ENOMEM; + VM_BUG_ON(pgd_none(*pgd)); + pud = pud_offset(pgd, vmaddr); + VM_BUG_ON(pud_none(*pud)); + pmd = pmd_offset(pud, vmaddr); + VM_BUG_ON(pmd_none(*pmd)); /* large pmds cannot yet be handled */ if (pmd_large(*pmd)) return -EFAULT; - /* pmd now points to a valid segment table entry. */ - rmap = kmalloc(sizeof(*rmap), GFP_KERNEL|__GFP_REPEAT); - if (!rmap) - return -ENOMEM; /* Link gmap segment table entry location to page table. */ - page = pmd_page(*pmd); - mp = (struct gmap_pgtable *) page->index; - rmap->gmap = gmap; - rmap->entry = segment_ptr; - rmap->vmaddr = address & PMD_MASK; - spin_lock(&mm->page_table_lock); - if (*segment_ptr == segment) { - list_add(&rmap->list, &mp->mapper); - /* Set gmap segment table entry to page table. */ - *segment_ptr = pmd_val(*pmd) & PAGE_MASK; - rmap = NULL; - } - spin_unlock(&mm->page_table_lock); - kfree(rmap); - return 0; -} - -static void gmap_disconnect_pgtable(struct mm_struct *mm, unsigned long *table) -{ - struct gmap_rmap *rmap, *next; - struct gmap_pgtable *mp; - struct page *page; - int flush; - - flush = 0; - spin_lock(&mm->page_table_lock); - page = pfn_to_page(__pa(table) >> PAGE_SHIFT); - mp = (struct gmap_pgtable *) page->index; - list_for_each_entry_safe(rmap, next, &mp->mapper, list) { - *rmap->entry = mp->vmaddr | (_SEGMENT_ENTRY_INVALID | - _SEGMENT_ENTRY_PROTECT); - list_del(&rmap->list); - kfree(rmap); - flush = 1; - } - spin_unlock(&mm->page_table_lock); - if (flush) - __tlb_flush_global(); + rc = radix_tree_preload(GFP_KERNEL); + if (rc) + return rc; + ptl = pmd_lock(mm, pmd); + spin_lock(&gmap->guest_table_lock); + if (*table == _SEGMENT_ENTRY_INVALID) { + rc = radix_tree_insert(&gmap->host_to_guest, + vmaddr >> PMD_SHIFT, table); + if (!rc) + *table = pmd_val(*pmd); + } else + rc = 0; + spin_unlock(&gmap->guest_table_lock); + spin_unlock(ptl); + radix_tree_preload_end(); + return rc; } -/* - * this function is assumed to be called with mmap_sem held +/** + * gmap_fault - resolve a fault on a guest address + * @gmap: pointer to guest mapping meta data structure + * @gaddr: guest address + * @fault_flags: flags to pass down to handle_mm_fault() + * + * Returns 0 on success, -ENOMEM for out of memory conditions, and -EFAULT + * if the vm address is already mapped to a different guest segment. */ -unsigned long __gmap_fault(unsigned long address, struct gmap *gmap) +int gmap_fault(struct gmap *gmap, unsigned long gaddr, + unsigned int fault_flags) { - unsigned long *segment_ptr, segment; - struct gmap_pgtable *mp; - struct page *page; + unsigned long vmaddr; int rc; - current->thread.gmap_addr = address; - segment_ptr = gmap_table_walk(address, gmap); - if (IS_ERR(segment_ptr)) - return -EFAULT; - /* Convert the gmap address to an mm address. */ - while (1) { - segment = *segment_ptr; - if (!(segment & _SEGMENT_ENTRY_INVALID)) { - /* Page table is present */ - page = pfn_to_page(segment >> PAGE_SHIFT); - mp = (struct gmap_pgtable *) page->index; - return mp->vmaddr | (address & ~PMD_MASK); - } - if (!(segment & _SEGMENT_ENTRY_PROTECT)) - /* Nothing mapped in the gmap address space. */ - break; - rc = gmap_connect_pgtable(address, segment, segment_ptr, gmap); - if (rc) - return rc; - } - return -EFAULT; -} - -unsigned long gmap_fault(unsigned long address, struct gmap *gmap) -{ - unsigned long rc; - down_read(&gmap->mm->mmap_sem); - rc = __gmap_fault(address, gmap); + vmaddr = __gmap_translate(gmap, gaddr); + if (IS_ERR_VALUE(vmaddr)) { + rc = vmaddr; + goto out_up; + } + if (fixup_user_fault(current, gmap->mm, vmaddr, fault_flags)) { + rc = -EFAULT; + goto out_up; + } + rc = __gmap_link(gmap, gaddr, vmaddr); +out_up: up_read(&gmap->mm->mmap_sem); - return rc; } EXPORT_SYMBOL_GPL(gmap_fault); @@ -617,17 +622,24 @@ static void gmap_zap_swap_entry(swp_entry_t entry, struct mm_struct *mm) free_swap_and_cache(entry); } -/** - * The mm->mmap_sem lock must be held +/* + * this function is assumed to be called with mmap_sem held */ -static void gmap_zap_unused(struct mm_struct *mm, unsigned long address) +void __gmap_zap(struct gmap *gmap, unsigned long gaddr) { - unsigned long ptev, pgstev; + unsigned long vmaddr, ptev, pgstev; + pte_t *ptep, pte; spinlock_t *ptl; pgste_t pgste; - pte_t *ptep, pte; - ptep = get_locked_pte(mm, address, &ptl); + /* Find the vm address for the guest address */ + vmaddr = (unsigned long) radix_tree_lookup(&gmap->guest_to_host, + gaddr >> PMD_SHIFT); + if (!vmaddr) + return; + vmaddr |= gaddr & ~PMD_MASK; + /* Get pointer to the page table entry */ + ptep = get_locked_pte(gmap->mm, vmaddr, &ptl); if (unlikely(!ptep)) return; pte = *ptep; @@ -639,87 +651,34 @@ static void gmap_zap_unused(struct mm_struct *mm, unsigned long address) ptev = pte_val(pte); if (((pgstev & _PGSTE_GPS_USAGE_MASK) == _PGSTE_GPS_USAGE_UNUSED) || ((pgstev & _PGSTE_GPS_ZERO) && (ptev & _PAGE_INVALID))) { - gmap_zap_swap_entry(pte_to_swp_entry(pte), mm); - pte_clear(mm, address, ptep); + gmap_zap_swap_entry(pte_to_swp_entry(pte), gmap->mm); + pte_clear(gmap->mm, vmaddr, ptep); } pgste_set_unlock(ptep, pgste); out_pte: pte_unmap_unlock(*ptep, ptl); } - -/* - * this function is assumed to be called with mmap_sem held - */ -void __gmap_zap(unsigned long address, struct gmap *gmap) -{ - unsigned long *table, *segment_ptr; - unsigned long segment, pgstev, ptev; - struct gmap_pgtable *mp; - struct page *page; - - segment_ptr = gmap_table_walk(address, gmap); - if (IS_ERR(segment_ptr)) - return; - segment = *segment_ptr; - if (segment & _SEGMENT_ENTRY_INVALID) - return; - page = pfn_to_page(segment >> PAGE_SHIFT); - mp = (struct gmap_pgtable *) page->index; - address = mp->vmaddr | (address & ~PMD_MASK); - /* Page table is present */ - table = (unsigned long *)(segment & _SEGMENT_ENTRY_ORIGIN); - table = table + ((address >> 12) & 0xff); - pgstev = table[PTRS_PER_PTE]; - ptev = table[0]; - /* quick check, checked again with locks held */ - if (((pgstev & _PGSTE_GPS_USAGE_MASK) == _PGSTE_GPS_USAGE_UNUSED) || - ((pgstev & _PGSTE_GPS_ZERO) && (ptev & _PAGE_INVALID))) - gmap_zap_unused(gmap->mm, address); -} EXPORT_SYMBOL_GPL(__gmap_zap); -void gmap_discard(unsigned long from, unsigned long to, struct gmap *gmap) +void gmap_discard(struct gmap *gmap, unsigned long from, unsigned long to) { - - unsigned long *table, address, size; + unsigned long gaddr, vmaddr, size; struct vm_area_struct *vma; - struct gmap_pgtable *mp; - struct page *page; down_read(&gmap->mm->mmap_sem); - address = from; - while (address < to) { - /* Walk the gmap address space page table */ - table = gmap->table + ((address >> 53) & 0x7ff); - if (unlikely(*table & _REGION_ENTRY_INVALID)) { - address = (address + PMD_SIZE) & PMD_MASK; - continue; - } - table = (unsigned long *)(*table & _REGION_ENTRY_ORIGIN); - table = table + ((address >> 42) & 0x7ff); - if (unlikely(*table & _REGION_ENTRY_INVALID)) { - address = (address + PMD_SIZE) & PMD_MASK; - continue; - } - table = (unsigned long *)(*table & _REGION_ENTRY_ORIGIN); - table = table + ((address >> 31) & 0x7ff); - if (unlikely(*table & _REGION_ENTRY_INVALID)) { - address = (address + PMD_SIZE) & PMD_MASK; + for (gaddr = from; gaddr < to; + gaddr = (gaddr + PMD_SIZE) & PMD_MASK) { + /* Find the vm address for the guest address */ + vmaddr = (unsigned long) + radix_tree_lookup(&gmap->guest_to_host, + gaddr >> PMD_SHIFT); + if (!vmaddr) continue; - } - table = (unsigned long *)(*table & _REGION_ENTRY_ORIGIN); - table = table + ((address >> 20) & 0x7ff); - if (unlikely(*table & _SEGMENT_ENTRY_INVALID)) { - address = (address + PMD_SIZE) & PMD_MASK; - continue; - } - page = pfn_to_page(*table >> PAGE_SHIFT); - mp = (struct gmap_pgtable *) page->index; - vma = find_vma(gmap->mm, mp->vmaddr); - size = min(to - address, PMD_SIZE - (address & ~PMD_MASK)); - zap_page_range(vma, mp->vmaddr | (address & ~PMD_MASK), - size, NULL); - address = (address + PMD_SIZE) & PMD_MASK; + vmaddr |= gaddr & ~PMD_MASK; + /* Find vma in the parent mm */ + vma = find_vma(gmap->mm, vmaddr); + size = min(to - gaddr, PMD_SIZE - (gaddr & ~PMD_MASK)); + zap_page_range(vma, vmaddr, size, NULL); } up_read(&gmap->mm->mmap_sem); } @@ -755,7 +714,7 @@ EXPORT_SYMBOL_GPL(gmap_unregister_ipte_notifier); /** * gmap_ipte_notify - mark a range of ptes for invalidation notification * @gmap: pointer to guest mapping meta data structure - * @start: virtual address in the guest address space + * @gaddr: virtual address in the guest address space * @len: size of area * * Returns 0 if for each page in the given range a gmap mapping exists and @@ -763,7 +722,7 @@ EXPORT_SYMBOL_GPL(gmap_unregister_ipte_notifier); * 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_ipte_notify(struct gmap *gmap, unsigned long start, unsigned long len) +int gmap_ipte_notify(struct gmap *gmap, unsigned long gaddr, unsigned long len) { unsigned long addr; spinlock_t *ptl; @@ -771,12 +730,12 @@ int gmap_ipte_notify(struct gmap *gmap, unsigned long start, unsigned long len) pgste_t pgste; int rc = 0; - if ((start & ~PAGE_MASK) || (len & ~PAGE_MASK)) + if ((gaddr & ~PAGE_MASK) || (len & ~PAGE_MASK)) return -EINVAL; down_read(&gmap->mm->mmap_sem); while (len) { /* Convert gmap address and connect the page tables */ - addr = __gmap_fault(start, gmap); + addr = __gmap_translate(gmap, gaddr); if (IS_ERR_VALUE(addr)) { rc = addr; break; @@ -786,6 +745,9 @@ int gmap_ipte_notify(struct gmap *gmap, unsigned long start, unsigned long len) rc = -EFAULT; break; } + rc = __gmap_link(gmap, gaddr, addr); + if (rc) + break; /* Walk the process page table, lock and get pte pointer */ ptep = get_locked_pte(gmap->mm, addr, &ptl); if (unlikely(!ptep)) @@ -796,7 +758,7 @@ int gmap_ipte_notify(struct gmap *gmap, unsigned long start, unsigned long len) pgste = pgste_get_lock(ptep); pgste_val(pgste) |= PGSTE_IN_BIT; pgste_set_unlock(ptep, pgste); - start += PAGE_SIZE; + gaddr += PAGE_SIZE; len -= PAGE_SIZE; } spin_unlock(ptl); @@ -809,28 +771,30 @@ EXPORT_SYMBOL_GPL(gmap_ipte_notify); /** * gmap_do_ipte_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 * * This function is assumed to be called with the page table lock held * for the pte to notify. */ -void gmap_do_ipte_notify(struct mm_struct *mm, pte_t *pte) +void gmap_do_ipte_notify(struct mm_struct *mm, unsigned long vmaddr, pte_t *pte) { - unsigned long segment_offset; + unsigned long offset, gaddr; + unsigned long *table; struct gmap_notifier *nb; - struct gmap_pgtable *mp; - struct gmap_rmap *rmap; - struct page *page; + struct gmap *gmap; - segment_offset = ((unsigned long) pte) & (255 * sizeof(pte_t)); - segment_offset = segment_offset * (4096 / sizeof(pte_t)); - page = pfn_to_page(__pa(pte) >> PAGE_SHIFT); - mp = (struct gmap_pgtable *) page->index; + offset = ((unsigned long) pte) & (255 * sizeof(pte_t)); + offset = offset * (4096 / sizeof(pte_t)); spin_lock(&gmap_notifier_lock); - list_for_each_entry(rmap, &mp->mapper, list) { + list_for_each_entry(gmap, &mm->context.gmap_list, list) { + 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(rmap->gmap, - rmap->vmaddr + segment_offset); + nb->notifier_call(gmap, gaddr); } spin_unlock(&gmap_notifier_lock); } @@ -841,29 +805,18 @@ static inline int page_table_with_pgste(struct page *page) return atomic_read(&page->_mapcount) == 0; } -static inline unsigned long *page_table_alloc_pgste(struct mm_struct *mm, - unsigned long vmaddr) +static inline unsigned long *page_table_alloc_pgste(struct mm_struct *mm) { struct page *page; unsigned long *table; - struct gmap_pgtable *mp; page = alloc_page(GFP_KERNEL|__GFP_REPEAT); if (!page) return NULL; - mp = kmalloc(sizeof(*mp), GFP_KERNEL|__GFP_REPEAT); - if (!mp) { - __free_page(page); - return NULL; - } if (!pgtable_page_ctor(page)) { - kfree(mp); __free_page(page); return NULL; } - mp->vmaddr = vmaddr & PMD_MASK; - INIT_LIST_HEAD(&mp->mapper); - page->index = (unsigned long) mp; atomic_set(&page->_mapcount, 0); table = (unsigned long *) page_to_phys(page); clear_table(table, _PAGE_INVALID, PAGE_SIZE/2); @@ -874,14 +827,10 @@ static inline unsigned long *page_table_alloc_pgste(struct mm_struct *mm, static inline void page_table_free_pgste(unsigned long *table) { struct page *page; - struct gmap_pgtable *mp; page = pfn_to_page(__pa(table) >> PAGE_SHIFT); - mp = (struct gmap_pgtable *) page->index; - BUG_ON(!list_empty(&mp->mapper)); pgtable_page_dtor(page); atomic_set(&page->_mapcount, -1); - kfree(mp); __free_page(page); } @@ -986,11 +935,21 @@ int set_guest_storage_key(struct mm_struct *mm, unsigned long addr, pte_t *ptep; down_read(&mm->mmap_sem); +retry: ptep = get_locked_pte(current->mm, addr, &ptl); if (unlikely(!ptep)) { up_read(&mm->mmap_sem); return -EFAULT; } + if (!(pte_val(*ptep) & _PAGE_INVALID) && + (pte_val(*ptep) & _PAGE_PROTECT)) { + pte_unmap_unlock(*ptep, ptl); + if (fixup_user_fault(current, mm, addr, FAULT_FLAG_WRITE)) { + up_read(&mm->mmap_sem); + return -EFAULT; + } + goto retry; + } new = old = pgste_get_lock(ptep); pgste_val(new) &= ~(PGSTE_GR_BIT | PGSTE_GC_BIT | @@ -1028,8 +987,7 @@ static inline int page_table_with_pgste(struct page *page) return 0; } -static inline unsigned long *page_table_alloc_pgste(struct mm_struct *mm, - unsigned long vmaddr) +static inline unsigned long *page_table_alloc_pgste(struct mm_struct *mm) { return NULL; } @@ -1043,8 +1001,8 @@ static inline void page_table_free_pgste(unsigned long *table) { } -static inline void gmap_disconnect_pgtable(struct mm_struct *mm, - unsigned long *table) +static inline void gmap_unlink(struct mm_struct *mm, unsigned long *table, + unsigned long vmaddr) { } @@ -1064,14 +1022,14 @@ static inline unsigned int atomic_xor_bits(atomic_t *v, unsigned int bits) /* * page table entry allocation/free routines. */ -unsigned long *page_table_alloc(struct mm_struct *mm, unsigned long vmaddr) +unsigned long *page_table_alloc(struct mm_struct *mm) { unsigned long *uninitialized_var(table); struct page *uninitialized_var(page); unsigned int mask, bit; if (mm_has_pgste(mm)) - return page_table_alloc_pgste(mm, vmaddr); + return page_table_alloc_pgste(mm); /* Allocate fragments of a 4K page as 1K/2K page table */ spin_lock_bh(&mm->context.list_lock); mask = FRAG_MASK; @@ -1113,10 +1071,8 @@ void page_table_free(struct mm_struct *mm, unsigned long *table) unsigned int bit, mask; page = pfn_to_page(__pa(table) >> PAGE_SHIFT); - if (page_table_with_pgste(page)) { - gmap_disconnect_pgtable(mm, table); + if (page_table_with_pgste(page)) return page_table_free_pgste(table); - } /* Free 1K/2K page table fragment of a 4K page */ bit = 1 << ((__pa(table) & ~PAGE_MASK)/(PTRS_PER_PTE*sizeof(pte_t))); spin_lock_bh(&mm->context.list_lock); @@ -1148,7 +1104,8 @@ static void __page_table_free_rcu(void *table, unsigned bit) } } -void page_table_free_rcu(struct mmu_gather *tlb, unsigned long *table) +void page_table_free_rcu(struct mmu_gather *tlb, unsigned long *table, + unsigned long vmaddr) { struct mm_struct *mm; struct page *page; @@ -1157,7 +1114,7 @@ void page_table_free_rcu(struct mmu_gather *tlb, unsigned long *table) mm = tlb->mm; page = pfn_to_page(__pa(table) >> PAGE_SHIFT); if (page_table_with_pgste(page)) { - gmap_disconnect_pgtable(mm, table); + gmap_unlink(mm, table, vmaddr); table = (unsigned long *) (__pa(table) | FRAG_MASK); tlb_remove_table(tlb, table); return; @@ -1279,6 +1236,7 @@ static unsigned long page_table_realloc_pmd(struct mmu_gather *tlb, { unsigned long next, *table, *new; struct page *page; + spinlock_t *ptl; pmd_t *pmd; pmd = pmd_offset(pud, addr); @@ -1292,11 +1250,11 @@ again: if (page_table_with_pgste(page)) continue; /* Allocate new page table with pgstes */ - new = page_table_alloc_pgste(mm, addr); + new = page_table_alloc_pgste(mm); if (!new) return -ENOMEM; - spin_lock(&mm->page_table_lock); + ptl = pmd_lock(mm, pmd); if (likely((unsigned long *) pmd_deref(*pmd) == table)) { /* Nuke pmd entry pointing to the "short" page table */ pmdp_flush_lazy(mm, addr, pmd); @@ -1307,10 +1265,10 @@ again: /* Establish new table */ pmd_populate(mm, pmd, (pte_t *) new); /* Free old table with rcu, there might be a walker! */ - page_table_free_rcu(tlb, table); + page_table_free_rcu(tlb, table, addr); new = NULL; } - spin_unlock(&mm->page_table_lock); + spin_unlock(ptl); if (new) { page_table_free_pgste(new); goto again; @@ -1432,6 +1390,9 @@ int pmdp_set_access_flags(struct vm_area_struct *vma, { VM_BUG_ON(address & ~HPAGE_PMD_MASK); + entry = pmd_mkyoung(entry); + if (dirty) + entry = pmd_mkdirty(entry); if (pmd_same(*pmdp, entry)) return 0; pmdp_invalidate(vma, address, pmdp); diff --git a/arch/s390/mm/vmem.c b/arch/s390/mm/vmem.c index fe9012a..b1593c2 100644 --- a/arch/s390/mm/vmem.c +++ b/arch/s390/mm/vmem.c @@ -65,7 +65,7 @@ static pte_t __ref *vmem_pte_alloc(unsigned long address) pte_t *pte; if (slab_is_available()) - pte = (pte_t *) page_table_alloc(&init_mm, address); + pte = (pte_t *) page_table_alloc(&init_mm); else pte = alloc_bootmem_align(PTRS_PER_PTE * sizeof(pte_t), PTRS_PER_PTE * sizeof(pte_t)); @@ -236,8 +236,7 @@ int __meminit vmemmap_populate(unsigned long start, unsigned long end, int node) if (!new_page) goto out; pmd_val(*pm_dir) = __pa(new_page) | - _SEGMENT_ENTRY | _SEGMENT_ENTRY_LARGE | - _SEGMENT_ENTRY_CO; + _SEGMENT_ENTRY | _SEGMENT_ENTRY_LARGE; address = (address + PMD_SIZE) & PMD_MASK; continue; } @@ -253,9 +252,9 @@ int __meminit vmemmap_populate(unsigned long start, unsigned long end, int node) pt_dir = pte_offset_kernel(pm_dir, address); if (pte_none(*pt_dir)) { - unsigned long new_page; + void *new_page; - new_page =__pa(vmem_alloc_pages(0)); + new_page = vmemmap_alloc_block(PAGE_SIZE, node); if (!new_page) goto out; pte_val(*pt_dir) = @@ -263,7 +262,6 @@ int __meminit vmemmap_populate(unsigned long start, unsigned long end, int node) } address += PAGE_SIZE; } - memset((void *)start, 0, end - start); ret = 0; out: return ret; diff --git a/arch/s390/net/bpf_jit_comp.c b/arch/s390/net/bpf_jit_comp.c index a2cbd87..c52ac77 100644 --- a/arch/s390/net/bpf_jit_comp.c +++ b/arch/s390/net/bpf_jit_comp.c @@ -5,11 +5,9 @@ * * Author(s): Martin Schwidefsky <schwidefsky@de.ibm.com> */ -#include <linux/moduleloader.h> #include <linux/netdevice.h> #include <linux/if_vlan.h> #include <linux/filter.h> -#include <linux/random.h> #include <linux/init.h> #include <asm/cacheflush.h> #include <asm/facility.h> @@ -148,6 +146,12 @@ struct bpf_jit { ret; \ }) +static void bpf_jit_fill_hole(void *area, unsigned int size) +{ + /* Fill whole space with illegal instructions */ + memset(area, 0, size); +} + static void bpf_jit_prologue(struct bpf_jit *jit) { /* Save registers and create stack frame if necessary */ @@ -223,37 +227,6 @@ static void bpf_jit_epilogue(struct bpf_jit *jit) EMIT2(0x07fe); } -/* Helper to find the offset of pkt_type in sk_buff - * Make sure its still a 3bit field starting at the MSBs within a byte. - */ -#define PKT_TYPE_MAX 0xe0 -static int pkt_type_offset; - -static int __init bpf_pkt_type_offset_init(void) -{ - struct sk_buff skb_probe = { - .pkt_type = ~0, - }; - char *ct = (char *)&skb_probe; - int off; - - pkt_type_offset = -1; - for (off = 0; off < sizeof(struct sk_buff); off++) { - if (!ct[off]) - continue; - if (ct[off] == PKT_TYPE_MAX) - pkt_type_offset = off; - else { - /* Found non matching bit pattern, fix needed. */ - WARN_ON_ONCE(1); - pkt_type_offset = -1; - return -1; - } - } - return 0; -} -device_initcall(bpf_pkt_type_offset_init); - /* * make sure we dont leak kernel information to user */ @@ -753,12 +726,10 @@ call_fn: /* lg %r1,<d(function)>(%r13) */ } break; case BPF_ANC | SKF_AD_PKTTYPE: - if (pkt_type_offset < 0) - goto out; /* lhi %r5,0 */ EMIT4(0xa7580000); /* ic %r5,<d(pkt_type_offset)>(%r2) */ - EMIT4_DISP(0x43502000, pkt_type_offset); + EMIT4_DISP(0x43502000, PKT_TYPE_OFFSET()); /* srl %r5,5 */ EMIT4_DISP(0x88500000, 5); break; @@ -780,39 +751,7 @@ out: return -1; } -/* - * Note: for security reasons, bpf code will follow a randomly - * sized amount of illegal instructions. - */ -struct bpf_binary_header { - unsigned int pages; - u8 image[]; -}; - -static struct bpf_binary_header *bpf_alloc_binary(unsigned int bpfsize, - u8 **image_ptr) -{ - struct bpf_binary_header *header; - unsigned int sz, hole; - - /* Most BPF filters are really small, but if some of them fill a page, - * allow at least 128 extra bytes for illegal instructions. - */ - sz = round_up(bpfsize + sizeof(*header) + 128, PAGE_SIZE); - header = module_alloc(sz); - if (!header) - return NULL; - memset(header, 0, sz); - header->pages = sz / PAGE_SIZE; - hole = min(sz - (bpfsize + sizeof(*header)), PAGE_SIZE - sizeof(*header)); - /* Insert random number of illegal instructions before BPF code - * and make sure the first instruction starts at an even address. - */ - *image_ptr = &header->image[(prandom_u32() % hole) & -2]; - return header; -} - -void bpf_jit_compile(struct sk_filter *fp) +void bpf_jit_compile(struct bpf_prog *fp) { struct bpf_binary_header *header = NULL; unsigned long size, prg_len, lit_len; @@ -850,7 +789,8 @@ void bpf_jit_compile(struct sk_filter *fp) size = prg_len + lit_len; if (size >= BPF_SIZE_MAX) goto out; - header = bpf_alloc_binary(size, &jit.start); + header = bpf_jit_binary_alloc(size, &jit.start, + 2, bpf_jit_fill_hole); if (!header) goto out; jit.prg = jit.mid = jit.start + prg_len; @@ -869,13 +809,13 @@ void bpf_jit_compile(struct sk_filter *fp) if (jit.start) { set_memory_ro((unsigned long)header, header->pages); fp->bpf_func = (void *) jit.start; - fp->jited = 1; + fp->jited = true; } out: kfree(addrs); } -void bpf_jit_free(struct sk_filter *fp) +void bpf_jit_free(struct bpf_prog *fp) { unsigned long addr = (unsigned long)fp->bpf_func & PAGE_MASK; struct bpf_binary_header *header = (void *)addr; @@ -884,8 +824,8 @@ void bpf_jit_free(struct sk_filter *fp) goto free_filter; set_memory_rw(addr, header->pages); - module_free(NULL, header); + bpf_jit_binary_free(header); free_filter: - kfree(fp); + bpf_prog_unlock_free(fp); } diff --git a/arch/s390/oprofile/hwsampler.c b/arch/s390/oprofile/hwsampler.c index e53c6f2..ff9b4eb 100644 --- a/arch/s390/oprofile/hwsampler.c +++ b/arch/s390/oprofile/hwsampler.c @@ -178,7 +178,7 @@ static int smp_ctl_qsi(int cpu) static void hws_ext_handler(struct ext_code ext_code, unsigned int param32, unsigned long param64) { - struct hws_cpu_buffer *cb = &__get_cpu_var(sampler_cpu_buffer); + struct hws_cpu_buffer *cb = this_cpu_ptr(&sampler_cpu_buffer); if (!(param32 & CPU_MF_INT_SF_MASK)) return; diff --git a/arch/s390/pci/pci.c b/arch/s390/pci/pci.c index 30de427..2fa7b14 100644 --- a/arch/s390/pci/pci.c +++ b/arch/s390/pci/pci.c @@ -15,8 +15,8 @@ * Thomas Klein */ -#define COMPONENT "zPCI" -#define pr_fmt(fmt) COMPONENT ": " fmt +#define KMSG_COMPONENT "zpci" +#define pr_fmt(fmt) KMSG_COMPONENT ": " fmt #include <linux/kernel.h> #include <linux/slab.h> diff --git a/arch/s390/pci/pci_clp.c b/arch/s390/pci/pci_clp.c index 96545d7..6e22a24 100644 --- a/arch/s390/pci/pci_clp.c +++ b/arch/s390/pci/pci_clp.c @@ -5,8 +5,8 @@ * Jan Glauber <jang@linux.vnet.ibm.com> */ -#define COMPONENT "zPCI" -#define pr_fmt(fmt) COMPONENT ": " fmt +#define KMSG_COMPONENT "zpci" +#define pr_fmt(fmt) KMSG_COMPONENT ": " fmt #include <linux/kernel.h> #include <linux/slab.h> diff --git a/arch/s390/pci/pci_debug.c b/arch/s390/pci/pci_debug.c index c5c6684..eec598c 100644 --- a/arch/s390/pci/pci_debug.c +++ b/arch/s390/pci/pci_debug.c @@ -5,8 +5,8 @@ * Jan Glauber <jang@linux.vnet.ibm.com> */ -#define COMPONENT "zPCI" -#define pr_fmt(fmt) COMPONENT ": " fmt +#define KMSG_COMPONENT "zpci" +#define pr_fmt(fmt) KMSG_COMPONENT ": " fmt #include <linux/kernel.h> #include <linux/seq_file.h> diff --git a/arch/s390/pci/pci_dma.c b/arch/s390/pci/pci_dma.c index f91c031..4cbb29a 100644 --- a/arch/s390/pci/pci_dma.c +++ b/arch/s390/pci/pci_dma.c @@ -16,6 +16,13 @@ static struct kmem_cache *dma_region_table_cache; static struct kmem_cache *dma_page_table_cache; +static int s390_iommu_strict; + +static int zpci_refresh_global(struct zpci_dev *zdev) +{ + return zpci_refresh_trans((u64) zdev->fh << 32, zdev->start_dma, + zdev->iommu_pages * PAGE_SIZE); +} static unsigned long *dma_alloc_cpu_table(void) { @@ -155,18 +162,15 @@ static int dma_update_trans(struct zpci_dev *zdev, unsigned long pa, } /* - * rpcit is not required to establish new translations when previously - * invalid translation-table entries are validated, however it is - * required when altering previously valid entries. + * With zdev->tlb_refresh == 0, rpcit is not required to establish new + * translations when previously invalid translation-table entries are + * validated. With lazy unmap, it also is skipped for previously valid + * entries, but a global rpcit is then required before any address can + * be re-used, i.e. after each iommu bitmap wrap-around. */ if (!zdev->tlb_refresh && - ((flags & ZPCI_PTE_VALID_MASK) == ZPCI_PTE_VALID)) - /* - * TODO: also need to check that the old entry is indeed INVALID - * and not only for one page but for the whole range... - * -> now we WARN_ON in that case but with lazy unmap that - * needs to be redone! - */ + (!s390_iommu_strict || + ((flags & ZPCI_PTE_VALID_MASK) == ZPCI_PTE_VALID))) goto no_refresh; rc = zpci_refresh_trans((u64) zdev->fh << 32, start_dma_addr, @@ -220,16 +224,21 @@ static unsigned long __dma_alloc_iommu(struct zpci_dev *zdev, static unsigned long dma_alloc_iommu(struct zpci_dev *zdev, int size) { unsigned long offset, flags; + int wrap = 0; spin_lock_irqsave(&zdev->iommu_bitmap_lock, flags); offset = __dma_alloc_iommu(zdev, zdev->next_bit, size); - if (offset == -1) + if (offset == -1) { + /* wrap-around */ offset = __dma_alloc_iommu(zdev, 0, size); + wrap = 1; + } if (offset != -1) { zdev->next_bit = offset + size; - if (zdev->next_bit >= zdev->iommu_pages) - zdev->next_bit = 0; + if (!zdev->tlb_refresh && !s390_iommu_strict && wrap) + /* global flush after wrap-around with lazy unmap */ + zpci_refresh_global(zdev); } spin_unlock_irqrestore(&zdev->iommu_bitmap_lock, flags); return offset; @@ -243,7 +252,11 @@ static void dma_free_iommu(struct zpci_dev *zdev, unsigned long offset, int size if (!zdev->iommu_bitmap) goto out; bitmap_clear(zdev->iommu_bitmap, offset, size); - if (offset >= zdev->next_bit) + /* + * Lazy flush for unmap: need to move next_bit to avoid address re-use + * until wrap-around. + */ + if (!s390_iommu_strict && offset >= zdev->next_bit) zdev->next_bit = offset + size; out: spin_unlock_irqrestore(&zdev->iommu_bitmap_lock, flags); @@ -504,3 +517,12 @@ struct dma_map_ops s390_dma_ops = { /* dma_supported is unconditionally true without a callback */ }; EXPORT_SYMBOL_GPL(s390_dma_ops); + +static int __init s390_iommu_setup(char *str) +{ + if (!strncmp(str, "strict", 6)) + s390_iommu_strict = 1; + return 0; +} + +__setup("s390_iommu=", s390_iommu_setup); diff --git a/arch/s390/pci/pci_event.c b/arch/s390/pci/pci_event.c index 6d7f5a3..460fdb2 100644 --- a/arch/s390/pci/pci_event.c +++ b/arch/s390/pci/pci_event.c @@ -5,8 +5,8 @@ * Jan Glauber <jang@linux.vnet.ibm.com> */ -#define COMPONENT "zPCI" -#define pr_fmt(fmt) COMPONENT ": " fmt +#define KMSG_COMPONENT "zpci" +#define pr_fmt(fmt) KMSG_COMPONENT ": " fmt #include <linux/kernel.h> #include <linux/pci.h> diff --git a/arch/s390/pci/pci_sysfs.c b/arch/s390/pci/pci_sysfs.c index 9190214..fa3ce89 100644 --- a/arch/s390/pci/pci_sysfs.c +++ b/arch/s390/pci/pci_sysfs.c @@ -5,8 +5,8 @@ * Jan Glauber <jang@linux.vnet.ibm.com> */ -#define COMPONENT "zPCI" -#define pr_fmt(fmt) COMPONENT ": " fmt +#define KMSG_COMPONENT "zpci" +#define pr_fmt(fmt) KMSG_COMPONENT ": " fmt #include <linux/kernel.h> #include <linux/stat.h> |