summaryrefslogtreecommitdiff
path: root/arch/x86/kernel
diff options
context:
space:
mode:
Diffstat (limited to 'arch/x86/kernel')
-rw-r--r--arch/x86/kernel/acpi/boot.c15
-rw-r--r--arch/x86/kernel/apic/apic.c26
-rw-r--r--arch/x86/kernel/apic/io_apic.c2
-rw-r--r--arch/x86/kernel/cpu/mcheck/mce_amd.c3
-rw-r--r--arch/x86/kernel/cpu/microcode/intel.c19
-rw-r--r--arch/x86/kernel/fpu/regset.c9
-rw-r--r--arch/x86/kernel/fpu/signal.c4
-rw-r--r--arch/x86/kernel/kvm.c9
-rw-r--r--arch/x86/kernel/process_64.c236
-rw-r--r--arch/x86/kernel/setup.c7
-rw-r--r--arch/x86/kernel/smpboot.c11
-rw-r--r--arch/x86/kernel/tsc.c9
12 files changed, 209 insertions, 141 deletions
diff --git a/arch/x86/kernel/acpi/boot.c b/arch/x86/kernel/acpi/boot.c
index 931ced8..b89bef9 100644
--- a/arch/x86/kernel/acpi/boot.c
+++ b/arch/x86/kernel/acpi/boot.c
@@ -176,10 +176,15 @@ static int acpi_register_lapic(int id, u32 acpiid, u8 enabled)
return -EINVAL;
}
+ if (!enabled) {
+ ++disabled_cpus;
+ return -EINVAL;
+ }
+
if (boot_cpu_physical_apicid != -1U)
ver = boot_cpu_apic_version;
- cpu = __generic_processor_info(id, ver, enabled);
+ cpu = generic_processor_info(id, ver);
if (cpu >= 0)
early_per_cpu(x86_cpu_to_acpiid, cpu) = acpiid;
@@ -338,6 +343,14 @@ static void __init mp_override_legacy_irq(u8 bus_irq, u8 polarity, u8 trigger,
struct mpc_intsrc mp_irq;
/*
+ * Check bus_irq boundary.
+ */
+ if (bus_irq >= NR_IRQS_LEGACY) {
+ pr_warn("Invalid bus_irq %u for legacy override\n", bus_irq);
+ return;
+ }
+
+ /*
* Convert 'gsi' to 'ioapic.pin'.
*/
ioapic = mp_find_ioapic(gsi);
diff --git a/arch/x86/kernel/apic/apic.c b/arch/x86/kernel/apic/apic.c
index f223491..e2ead34 100644
--- a/arch/x86/kernel/apic/apic.c
+++ b/arch/x86/kernel/apic/apic.c
@@ -2070,7 +2070,7 @@ static int allocate_logical_cpuid(int apicid)
return nr_logical_cpuids++;
}
-int __generic_processor_info(int apicid, int version, bool enabled)
+int generic_processor_info(int apicid, int version)
{
int cpu, max = nr_cpu_ids;
bool boot_cpu_detected = physid_isset(boot_cpu_physical_apicid,
@@ -2128,11 +2128,9 @@ int __generic_processor_info(int apicid, int version, bool enabled)
if (num_processors >= nr_cpu_ids) {
int thiscpu = max + disabled_cpus;
- if (enabled) {
- pr_warning("APIC: NR_CPUS/possible_cpus limit of %i "
- "reached. Processor %d/0x%x ignored.\n",
- max, thiscpu, apicid);
- }
+ pr_warning("APIC: NR_CPUS/possible_cpus limit of %i "
+ "reached. Processor %d/0x%x ignored.\n",
+ max, thiscpu, apicid);
disabled_cpus++;
return -EINVAL;
@@ -2184,23 +2182,13 @@ int __generic_processor_info(int apicid, int version, bool enabled)
apic->x86_32_early_logical_apicid(cpu);
#endif
set_cpu_possible(cpu, true);
-
- if (enabled) {
- num_processors++;
- physid_set(apicid, phys_cpu_present_map);
- set_cpu_present(cpu, true);
- } else {
- disabled_cpus++;
- }
+ physid_set(apicid, phys_cpu_present_map);
+ set_cpu_present(cpu, true);
+ num_processors++;
return cpu;
}
-int generic_processor_info(int apicid, int version)
-{
- return __generic_processor_info(apicid, version, true);
-}
-
int hard_smp_processor_id(void)
{
return read_apic_id();
diff --git a/arch/x86/kernel/apic/io_apic.c b/arch/x86/kernel/apic/io_apic.c
index 7249f15..cf89928 100644
--- a/arch/x86/kernel/apic/io_apic.c
+++ b/arch/x86/kernel/apic/io_apic.c
@@ -2116,7 +2116,7 @@ static inline void __init check_timer(void)
int idx;
idx = find_irq_entry(apic1, pin1, mp_INT);
if (idx != -1 && irq_trigger(idx))
- unmask_ioapic_irq(irq_get_chip_data(0));
+ unmask_ioapic_irq(irq_get_irq_data(0));
}
irq_domain_deactivate_irq(irq_data);
irq_domain_activate_irq(irq_data);
diff --git a/arch/x86/kernel/cpu/mcheck/mce_amd.c b/arch/x86/kernel/cpu/mcheck/mce_amd.c
index 3dfca7b..a5b47c1 100644
--- a/arch/x86/kernel/cpu/mcheck/mce_amd.c
+++ b/arch/x86/kernel/cpu/mcheck/mce_amd.c
@@ -955,6 +955,9 @@ static int threshold_create_bank(unsigned int cpu, unsigned int bank)
const char *name = get_name(bank, NULL);
int err = 0;
+ if (!dev)
+ return -ENODEV;
+
if (is_shared_bank(bank)) {
nb = node_to_amd_nb(amd_get_nb_id(cpu));
diff --git a/arch/x86/kernel/cpu/microcode/intel.c b/arch/x86/kernel/cpu/microcode/intel.c
index cdc0dea..13dbcc0 100644
--- a/arch/x86/kernel/cpu/microcode/intel.c
+++ b/arch/x86/kernel/cpu/microcode/intel.c
@@ -34,6 +34,7 @@
#include <linux/mm.h>
#include <asm/microcode_intel.h>
+#include <asm/intel-family.h>
#include <asm/processor.h>
#include <asm/tlbflush.h>
#include <asm/setup.h>
@@ -1046,6 +1047,18 @@ static int get_ucode_fw(void *to, const void *from, size_t n)
return 0;
}
+static bool is_blacklisted(unsigned int cpu)
+{
+ struct cpuinfo_x86 *c = &cpu_data(cpu);
+
+ if (c->x86 == 6 && c->x86_model == INTEL_FAM6_BROADWELL_X) {
+ pr_err_once("late loading on model 79 is disabled.\n");
+ return true;
+ }
+
+ return false;
+}
+
static enum ucode_state request_microcode_fw(int cpu, struct device *device,
bool refresh_fw)
{
@@ -1054,6 +1067,9 @@ static enum ucode_state request_microcode_fw(int cpu, struct device *device,
const struct firmware *firmware;
enum ucode_state ret;
+ if (is_blacklisted(cpu))
+ return UCODE_NFOUND;
+
sprintf(name, "intel-ucode/%02x-%02x-%02x",
c->x86, c->x86_model, c->x86_mask);
@@ -1078,6 +1094,9 @@ static int get_ucode_user(void *to, const void *from, size_t n)
static enum ucode_state
request_microcode_user(int cpu, const void __user *buf, size_t size)
{
+ if (is_blacklisted(cpu))
+ return UCODE_NFOUND;
+
return generic_load_microcode(cpu, (void *)buf, size, &get_ucode_user);
}
diff --git a/arch/x86/kernel/fpu/regset.c b/arch/x86/kernel/fpu/regset.c
index c114b13..7052d9a 100644
--- a/arch/x86/kernel/fpu/regset.c
+++ b/arch/x86/kernel/fpu/regset.c
@@ -130,11 +130,16 @@ int xstateregs_set(struct task_struct *target, const struct user_regset *regset,
fpu__activate_fpstate_write(fpu);
- if (boot_cpu_has(X86_FEATURE_XSAVES))
+ if (boot_cpu_has(X86_FEATURE_XSAVES)) {
ret = copyin_to_xsaves(kbuf, ubuf, xsave);
- else
+ } else {
ret = user_regset_copyin(&pos, &count, &kbuf, &ubuf, xsave, 0, -1);
+ /* xcomp_bv must be 0 when using uncompacted format */
+ if (!ret && xsave->header.xcomp_bv)
+ ret = -EINVAL;
+ }
+
/*
* In case of failure, mark all states as init:
*/
diff --git a/arch/x86/kernel/fpu/signal.c b/arch/x86/kernel/fpu/signal.c
index a184c21..3ec0d2d 100644
--- a/arch/x86/kernel/fpu/signal.c
+++ b/arch/x86/kernel/fpu/signal.c
@@ -329,6 +329,10 @@ static int __fpu__restore_sig(void __user *buf, void __user *buf_fx, int size)
} else {
err = __copy_from_user(&fpu->state.xsave,
buf_fx, state_size);
+
+ /* xcomp_bv must be 0 when using uncompacted format */
+ if (!err && state_size > offsetof(struct xregs_state, header) && fpu->state.xsave.header.xcomp_bv)
+ err = -EINVAL;
}
if (err || __copy_from_user(&env, buf, sizeof(env))) {
diff --git a/arch/x86/kernel/kvm.c b/arch/x86/kernel/kvm.c
index 9cf697c..77f17cb 100644
--- a/arch/x86/kernel/kvm.c
+++ b/arch/x86/kernel/kvm.c
@@ -141,7 +141,8 @@ void kvm_async_pf_task_wait(u32 token)
n.token = token;
n.cpu = smp_processor_id();
- n.halted = is_idle_task(current) || preempt_count() > 1;
+ n.halted = is_idle_task(current) || preempt_count() > 1 ||
+ rcu_preempt_depth();
init_swait_queue_head(&n.wq);
hlist_add_head(&n.link, &b->list);
raw_spin_unlock(&b->lock);
@@ -152,6 +153,8 @@ void kvm_async_pf_task_wait(u32 token)
if (hlist_unhashed(&n.link))
break;
+ rcu_irq_exit();
+
if (!n.halted) {
local_irq_enable();
schedule();
@@ -160,11 +163,11 @@ void kvm_async_pf_task_wait(u32 token)
/*
* We cannot reschedule. So halt.
*/
- rcu_irq_exit();
native_safe_halt();
local_irq_disable();
- rcu_irq_enter();
}
+
+ rcu_irq_enter();
}
if (!n.halted)
finish_swait(&n.wq, &wait);
diff --git a/arch/x86/kernel/process_64.c b/arch/x86/kernel/process_64.c
index b3760b3..0887d2a 100644
--- a/arch/x86/kernel/process_64.c
+++ b/arch/x86/kernel/process_64.c
@@ -136,6 +136,123 @@ void release_thread(struct task_struct *dead_task)
}
}
+enum which_selector {
+ FS,
+ GS
+};
+
+/*
+ * Saves the FS or GS base for an outgoing thread if FSGSBASE extensions are
+ * not available. The goal is to be reasonably fast on non-FSGSBASE systems.
+ * It's forcibly inlined because it'll generate better code and this function
+ * is hot.
+ */
+static __always_inline void save_base_legacy(struct task_struct *prev_p,
+ unsigned short selector,
+ enum which_selector which)
+{
+ if (likely(selector == 0)) {
+ /*
+ * On Intel (without X86_BUG_NULL_SEG), the segment base could
+ * be the pre-existing saved base or it could be zero. On AMD
+ * (with X86_BUG_NULL_SEG), the segment base could be almost
+ * anything.
+ *
+ * This branch is very hot (it's hit twice on almost every
+ * context switch between 64-bit programs), and avoiding
+ * the RDMSR helps a lot, so we just assume that whatever
+ * value is already saved is correct. This matches historical
+ * Linux behavior, so it won't break existing applications.
+ *
+ * To avoid leaking state, on non-X86_BUG_NULL_SEG CPUs, if we
+ * report that the base is zero, it needs to actually be zero:
+ * see the corresponding logic in load_seg_legacy.
+ */
+ } else {
+ /*
+ * If the selector is 1, 2, or 3, then the base is zero on
+ * !X86_BUG_NULL_SEG CPUs and could be anything on
+ * X86_BUG_NULL_SEG CPUs. In the latter case, Linux
+ * has never attempted to preserve the base across context
+ * switches.
+ *
+ * If selector > 3, then it refers to a real segment, and
+ * saving the base isn't necessary.
+ */
+ if (which == FS)
+ prev_p->thread.fsbase = 0;
+ else
+ prev_p->thread.gsbase = 0;
+ }
+}
+
+static __always_inline void save_fsgs(struct task_struct *task)
+{
+ savesegment(fs, task->thread.fsindex);
+ savesegment(gs, task->thread.gsindex);
+ save_base_legacy(task, task->thread.fsindex, FS);
+ save_base_legacy(task, task->thread.gsindex, GS);
+}
+
+static __always_inline void loadseg(enum which_selector which,
+ unsigned short sel)
+{
+ if (which == FS)
+ loadsegment(fs, sel);
+ else
+ load_gs_index(sel);
+}
+
+static __always_inline void load_seg_legacy(unsigned short prev_index,
+ unsigned long prev_base,
+ unsigned short next_index,
+ unsigned long next_base,
+ enum which_selector which)
+{
+ if (likely(next_index <= 3)) {
+ /*
+ * The next task is using 64-bit TLS, is not using this
+ * segment at all, or is having fun with arcane CPU features.
+ */
+ if (next_base == 0) {
+ /*
+ * Nasty case: on AMD CPUs, we need to forcibly zero
+ * the base.
+ */
+ if (static_cpu_has_bug(X86_BUG_NULL_SEG)) {
+ loadseg(which, __USER_DS);
+ loadseg(which, next_index);
+ } else {
+ /*
+ * We could try to exhaustively detect cases
+ * under which we can skip the segment load,
+ * but there's really only one case that matters
+ * for performance: if both the previous and
+ * next states are fully zeroed, we can skip
+ * the load.
+ *
+ * (This assumes that prev_base == 0 has no
+ * false positives. This is the case on
+ * Intel-style CPUs.)
+ */
+ if (likely(prev_index | next_index | prev_base))
+ loadseg(which, next_index);
+ }
+ } else {
+ if (prev_index != next_index)
+ loadseg(which, next_index);
+ wrmsrl(which == FS ? MSR_FS_BASE : MSR_KERNEL_GS_BASE,
+ next_base);
+ }
+ } else {
+ /*
+ * The next task is using a real segment. Loading the selector
+ * is sufficient.
+ */
+ loadseg(which, next_index);
+ }
+}
+
int copy_thread_tls(unsigned long clone_flags, unsigned long sp,
unsigned long arg, struct task_struct *p, unsigned long tls)
{
@@ -216,10 +333,19 @@ start_thread_common(struct pt_regs *regs, unsigned long new_ip,
unsigned long new_sp,
unsigned int _cs, unsigned int _ss, unsigned int _ds)
{
+ WARN_ON_ONCE(regs != current_pt_regs());
+
+ if (static_cpu_has(X86_BUG_NULL_SEG)) {
+ /* Loading zero below won't clear the base. */
+ loadsegment(fs, __USER_DS);
+ load_gs_index(__USER_DS);
+ }
+
loadsegment(fs, 0);
loadsegment(es, _ds);
loadsegment(ds, _ds);
load_gs_index(0);
+
regs->ip = new_ip;
regs->sp = new_sp;
regs->cs = _cs;
@@ -264,7 +390,6 @@ __switch_to(struct task_struct *prev_p, struct task_struct *next_p)
struct fpu *next_fpu = &next->fpu;
int cpu = smp_processor_id();
struct tss_struct *tss = &per_cpu(cpu_tss, cpu);
- unsigned prev_fsindex, prev_gsindex;
fpu_switch_t fpu_switch;
fpu_switch = switch_fpu_prepare(prev_fpu, next_fpu, cpu);
@@ -274,8 +399,7 @@ __switch_to(struct task_struct *prev_p, struct task_struct *next_p)
*
* (e.g. xen_load_tls())
*/
- savesegment(fs, prev_fsindex);
- savesegment(gs, prev_gsindex);
+ save_fsgs(prev_p);
/*
* Load TLS before restoring any segments so that segment loads
@@ -314,108 +438,10 @@ __switch_to(struct task_struct *prev_p, struct task_struct *next_p)
if (unlikely(next->ds | prev->ds))
loadsegment(ds, next->ds);
- /*
- * Switch FS and GS.
- *
- * These are even more complicated than DS and ES: they have
- * 64-bit bases are that controlled by arch_prctl. The bases
- * don't necessarily match the selectors, as user code can do
- * any number of things to cause them to be inconsistent.
- *
- * We don't promise to preserve the bases if the selectors are
- * nonzero. We also don't promise to preserve the base if the
- * selector is zero and the base doesn't match whatever was
- * most recently passed to ARCH_SET_FS/GS. (If/when the
- * FSGSBASE instructions are enabled, we'll need to offer
- * stronger guarantees.)
- *
- * As an invariant,
- * (fsbase != 0 && fsindex != 0) || (gsbase != 0 && gsindex != 0) is
- * impossible.
- */
- if (next->fsindex) {
- /* Loading a nonzero value into FS sets the index and base. */
- loadsegment(fs, next->fsindex);
- } else {
- if (next->fsbase) {
- /* Next index is zero but next base is nonzero. */
- if (prev_fsindex)
- loadsegment(fs, 0);
- wrmsrl(MSR_FS_BASE, next->fsbase);
- } else {
- /* Next base and index are both zero. */
- if (static_cpu_has_bug(X86_BUG_NULL_SEG)) {
- /*
- * We don't know the previous base and can't
- * find out without RDMSR. Forcibly clear it.
- */
- loadsegment(fs, __USER_DS);
- loadsegment(fs, 0);
- } else {
- /*
- * If the previous index is zero and ARCH_SET_FS
- * didn't change the base, then the base is
- * also zero and we don't need to do anything.
- */
- if (prev->fsbase || prev_fsindex)
- loadsegment(fs, 0);
- }
- }
- }
- /*
- * Save the old state and preserve the invariant.
- * NB: if prev_fsindex == 0, then we can't reliably learn the base
- * without RDMSR because Intel user code can zero it without telling
- * us and AMD user code can program any 32-bit value without telling
- * us.
- */
- if (prev_fsindex)
- prev->fsbase = 0;
- prev->fsindex = prev_fsindex;
-
- if (next->gsindex) {
- /* Loading a nonzero value into GS sets the index and base. */
- load_gs_index(next->gsindex);
- } else {
- if (next->gsbase) {
- /* Next index is zero but next base is nonzero. */
- if (prev_gsindex)
- load_gs_index(0);
- wrmsrl(MSR_KERNEL_GS_BASE, next->gsbase);
- } else {
- /* Next base and index are both zero. */
- if (static_cpu_has_bug(X86_BUG_NULL_SEG)) {
- /*
- * We don't know the previous base and can't
- * find out without RDMSR. Forcibly clear it.
- *
- * This contains a pointless SWAPGS pair.
- * Fixing it would involve an explicit check
- * for Xen or a new pvop.
- */
- load_gs_index(__USER_DS);
- load_gs_index(0);
- } else {
- /*
- * If the previous index is zero and ARCH_SET_GS
- * didn't change the base, then the base is
- * also zero and we don't need to do anything.
- */
- if (prev->gsbase || prev_gsindex)
- load_gs_index(0);
- }
- }
- }
- /*
- * Save the old state and preserve the invariant.
- * NB: if prev_gsindex == 0, then we can't reliably learn the base
- * without RDMSR because Intel user code can zero it without telling
- * us and AMD user code can program any 32-bit value without telling
- * us.
- */
- if (prev_gsindex)
- prev->gsbase = 0;
- prev->gsindex = prev_gsindex;
+ load_seg_legacy(prev->fsindex, prev->fsbase,
+ next->fsindex, next->fsbase, FS);
+ load_seg_legacy(prev->gsindex, prev->gsbase,
+ next->gsindex, next->gsbase, GS);
switch_fpu_finish(next_fpu, fpu_switch);
diff --git a/arch/x86/kernel/setup.c b/arch/x86/kernel/setup.c
index 9c337b0..feaab07 100644
--- a/arch/x86/kernel/setup.c
+++ b/arch/x86/kernel/setup.c
@@ -1054,6 +1054,13 @@ void __init setup_arch(char **cmdline_p)
max_possible_pfn = max_pfn;
/*
+ * This call is required when the CPU does not support PAT. If
+ * mtrr_bp_init() invoked it already via pat_init() the call has no
+ * effect.
+ */
+ init_cache_modes();
+
+ /*
* Define random base addresses for memory sections after max_pfn is
* defined and before each memory section base is used.
*/
diff --git a/arch/x86/kernel/smpboot.c b/arch/x86/kernel/smpboot.c
index 36171bc..9fe7b9e 100644
--- a/arch/x86/kernel/smpboot.c
+++ b/arch/x86/kernel/smpboot.c
@@ -181,6 +181,12 @@ static void smp_callin(void)
smp_store_cpu_info(cpuid);
/*
+ * The topology information must be up to date before
+ * calibrate_delay() and notify_cpu_starting().
+ */
+ set_cpu_sibling_map(raw_smp_processor_id());
+
+ /*
* Get our bogomips.
* Update loops_per_jiffy in cpu_data. Previous call to
* smp_store_cpu_info() stored a value that is close but not as
@@ -190,11 +196,6 @@ static void smp_callin(void)
cpu_data(cpuid).loops_per_jiffy = loops_per_jiffy;
pr_debug("Stack at about %p\n", &cpuid);
- /*
- * This must be done before setting cpu_online_mask
- * or calling notify_cpu_starting.
- */
- set_cpu_sibling_map(raw_smp_processor_id());
wmb();
notify_cpu_starting(cpuid);
diff --git a/arch/x86/kernel/tsc.c b/arch/x86/kernel/tsc.c
index eea88fe..44bf5cf 100644
--- a/arch/x86/kernel/tsc.c
+++ b/arch/x86/kernel/tsc.c
@@ -694,6 +694,7 @@ unsigned long native_calibrate_tsc(void)
crystal_khz = 24000; /* 24.0 MHz */
break;
case INTEL_FAM6_SKYLAKE_X:
+ case INTEL_FAM6_ATOM_DENVERTON:
crystal_khz = 25000; /* 25.0 MHz */
break;
case INTEL_FAM6_ATOM_GOLDMONT:
@@ -1381,12 +1382,10 @@ void __init tsc_init(void)
unsigned long calibrate_delay_is_known(void)
{
int sibling, cpu = smp_processor_id();
- struct cpumask *mask = topology_core_cpumask(cpu);
+ int constant_tsc = cpu_has(&cpu_data(cpu), X86_FEATURE_CONSTANT_TSC);
+ const struct cpumask *mask = topology_core_cpumask(cpu);
- if (!tsc_disabled && !cpu_has(&cpu_data(cpu), X86_FEATURE_CONSTANT_TSC))
- return 0;
-
- if (!mask)
+ if (tsc_disabled || !constant_tsc || !mask)
return 0;
sibling = cpumask_any_but(mask, cpu);
generated by cgit v0.10.2 at 2024-06-02 04:44:16 (GMT)