From: Ingo Molnar <mingo@kernel.org>
To: Linus Torvalds <torvalds@linux-foundation.org>
Cc: linux-kernel@vger.kernel.org,
Thomas Gleixner <tglx@linutronix.de>,
"H. Peter Anvin" <hpa@zytor.com>,
Peter Zijlstra <a.p.zijlstra@chello.nl>,
Andrew Morton <akpm@linux-foundation.org>
Subject: [GIT PULL] x86 fixes
Date: Sat, 14 Oct 2017 18:16:37 +0200 [thread overview]
Message-ID: <20171014161636.tufvzolmzst64jaa@gmail.com> (raw)
Linus,
Please pull the latest x86-urgent-for-linus git tree from:
git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip.git x86-urgent-for-linus
# HEAD: 1f161f67a272cc4f29f27934dd3f74cb657eb5c4 x86/microcode: Do the family check first
A landry list of fixes:
- fix reboot breakage on some PCID-enabled system
- fix crashes/hangs on some PCID-enabled systems
- fix microcode loading on certain older CPUs
- various unwinder fixes
- extend an APIC quirk to more hardware systems and disable APIC related warning
on virtualized systems
- various Hyper-V fixes
- a macro definition robustness fix
- remove jprobes IRQ disabling
- various mem-encryption fixes
Thanks,
Ingo
------------------>
Andy Lutomirski (2):
x86/mm/64: Fix reboot interaction with CR4.PCIDE
x86/mm: Flush more aggressively in lazy TLB mode
Borislav Petkov (1):
x86/microcode: Do the family check first
Josh Poimboeuf (5):
kprobes/x86: Set up frame pointer in kprobe trampoline
x86/unwind: Fix dereference of untrusted pointer
x86/unwind: Use MSB for frame pointer encoding on 32-bit
x86/unwind: Align stack pointer in unwinder dump
x86/unwind: Disable unwinder warnings on 32-bit
Len Brown (1):
x86/apic: Update TSC_DEADLINE quirk with additional SKX stepping
Marcelo Henrique Cerri (1):
x86/hyperv: Fix hypercalls with extended CPU ranges for TLB flushing
Masami Hiramatsu (1):
kprobes/x86: Remove IRQ disabling from jprobe handlers
Mathias Krause (1):
x86/alternatives: Fix alt_max_short macro to really be a max()
Paolo Bonzini (1):
x86/apic: Silence "FW_BUG TSC_DEADLINE disabled due to Errata" on hypervisors
Tom Lendacky (1):
x86/mm: Disable various instrumentations of mm/mem_encrypt.c and mm/tlb.c
Vitaly Kuznetsov (2):
x86/hyperv: Clear vCPU banks between calls to avoid flushing unneeded vCPUs
x86/hyperv: Don't use percpu areas for pcpu_flush/pcpu_flush_ex structures
arch/x86/entry/entry_32.S | 4 +-
arch/x86/hyperv/hv_init.c | 5 ++
arch/x86/hyperv/mmu.c | 57 +++++++++---
arch/x86/include/asm/alternative-asm.h | 4 +-
arch/x86/include/asm/alternative.h | 6 +-
arch/x86/include/asm/mmu_context.h | 8 +-
arch/x86/include/asm/mshyperv.h | 1 +
arch/x86/include/asm/tlbflush.h | 24 ++++++
arch/x86/kernel/apic/apic.c | 15 +++-
arch/x86/kernel/cpu/microcode/core.c | 27 ++++--
arch/x86/kernel/kprobes/common.h | 13 ++-
arch/x86/kernel/kprobes/core.c | 2 -
arch/x86/kernel/reboot.c | 4 +
arch/x86/kernel/unwind_frame.c | 38 +++++++-
arch/x86/mm/Makefile | 11 ++-
arch/x86/mm/tlb.c | 153 ++++++++++++++++++++++++---------
16 files changed, 284 insertions(+), 88 deletions(-)
diff --git a/arch/x86/entry/entry_32.S b/arch/x86/entry/entry_32.S
index 8a13d468635a..50e0d2bc4528 100644
--- a/arch/x86/entry/entry_32.S
+++ b/arch/x86/entry/entry_32.S
@@ -176,7 +176,7 @@
/*
* This is a sneaky trick to help the unwinder find pt_regs on the stack. The
* frame pointer is replaced with an encoded pointer to pt_regs. The encoding
- * is just setting the LSB, which makes it an invalid stack address and is also
+ * is just clearing the MSB, which makes it an invalid stack address and is also
* a signal to the unwinder that it's a pt_regs pointer in disguise.
*
* NOTE: This macro must be used *after* SAVE_ALL because it corrupts the
@@ -185,7 +185,7 @@
.macro ENCODE_FRAME_POINTER
#ifdef CONFIG_FRAME_POINTER
mov %esp, %ebp
- orl $0x1, %ebp
+ andl $0x7fffffff, %ebp
#endif
.endm
diff --git a/arch/x86/hyperv/hv_init.c b/arch/x86/hyperv/hv_init.c
index 1a8eb550c40f..a5db63f728a2 100644
--- a/arch/x86/hyperv/hv_init.c
+++ b/arch/x86/hyperv/hv_init.c
@@ -85,6 +85,8 @@ EXPORT_SYMBOL_GPL(hyperv_cs);
u32 *hv_vp_index;
EXPORT_SYMBOL_GPL(hv_vp_index);
+u32 hv_max_vp_index;
+
static int hv_cpu_init(unsigned int cpu)
{
u64 msr_vp_index;
@@ -93,6 +95,9 @@ static int hv_cpu_init(unsigned int cpu)
hv_vp_index[smp_processor_id()] = msr_vp_index;
+ if (msr_vp_index > hv_max_vp_index)
+ hv_max_vp_index = msr_vp_index;
+
return 0;
}
diff --git a/arch/x86/hyperv/mmu.c b/arch/x86/hyperv/mmu.c
index 39e7f6e50919..9cc9e1c1e2db 100644
--- a/arch/x86/hyperv/mmu.c
+++ b/arch/x86/hyperv/mmu.c
@@ -36,9 +36,9 @@ struct hv_flush_pcpu_ex {
/* Each gva in gva_list encodes up to 4096 pages to flush */
#define HV_TLB_FLUSH_UNIT (4096 * PAGE_SIZE)
-static struct hv_flush_pcpu __percpu *pcpu_flush;
+static struct hv_flush_pcpu __percpu **pcpu_flush;
-static struct hv_flush_pcpu_ex __percpu *pcpu_flush_ex;
+static struct hv_flush_pcpu_ex __percpu **pcpu_flush_ex;
/*
* Fills in gva_list starting from offset. Returns the number of items added.
@@ -76,6 +76,18 @@ static inline int cpumask_to_vp_set(struct hv_flush_pcpu_ex *flush,
{
int cpu, vcpu, vcpu_bank, vcpu_offset, nr_bank = 1;
+ /* valid_bank_mask can represent up to 64 banks */
+ if (hv_max_vp_index / 64 >= 64)
+ return 0;
+
+ /*
+ * Clear all banks up to the maximum possible bank as hv_flush_pcpu_ex
+ * structs are not cleared between calls, we risk flushing unneeded
+ * vCPUs otherwise.
+ */
+ for (vcpu_bank = 0; vcpu_bank <= hv_max_vp_index / 64; vcpu_bank++)
+ flush->hv_vp_set.bank_contents[vcpu_bank] = 0;
+
/*
* Some banks may end up being empty but this is acceptable.
*/
@@ -83,11 +95,6 @@ static inline int cpumask_to_vp_set(struct hv_flush_pcpu_ex *flush,
vcpu = hv_cpu_number_to_vp_number(cpu);
vcpu_bank = vcpu / 64;
vcpu_offset = vcpu % 64;
-
- /* valid_bank_mask can represent up to 64 banks */
- if (vcpu_bank >= 64)
- return 0;
-
__set_bit(vcpu_offset, (unsigned long *)
&flush->hv_vp_set.bank_contents[vcpu_bank]);
if (vcpu_bank >= nr_bank)
@@ -102,6 +109,7 @@ static void hyperv_flush_tlb_others(const struct cpumask *cpus,
const struct flush_tlb_info *info)
{
int cpu, vcpu, gva_n, max_gvas;
+ struct hv_flush_pcpu **flush_pcpu;
struct hv_flush_pcpu *flush;
u64 status = U64_MAX;
unsigned long flags;
@@ -116,7 +124,17 @@ static void hyperv_flush_tlb_others(const struct cpumask *cpus,
local_irq_save(flags);
- flush = this_cpu_ptr(pcpu_flush);
+ flush_pcpu = this_cpu_ptr(pcpu_flush);
+
+ if (unlikely(!*flush_pcpu))
+ *flush_pcpu = page_address(alloc_page(GFP_ATOMIC));
+
+ flush = *flush_pcpu;
+
+ if (unlikely(!flush)) {
+ local_irq_restore(flags);
+ goto do_native;
+ }
if (info->mm) {
flush->address_space = virt_to_phys(info->mm->pgd);
@@ -173,6 +191,7 @@ static void hyperv_flush_tlb_others_ex(const struct cpumask *cpus,
const struct flush_tlb_info *info)
{
int nr_bank = 0, max_gvas, gva_n;
+ struct hv_flush_pcpu_ex **flush_pcpu;
struct hv_flush_pcpu_ex *flush;
u64 status = U64_MAX;
unsigned long flags;
@@ -187,7 +206,17 @@ static void hyperv_flush_tlb_others_ex(const struct cpumask *cpus,
local_irq_save(flags);
- flush = this_cpu_ptr(pcpu_flush_ex);
+ flush_pcpu = this_cpu_ptr(pcpu_flush_ex);
+
+ if (unlikely(!*flush_pcpu))
+ *flush_pcpu = page_address(alloc_page(GFP_ATOMIC));
+
+ flush = *flush_pcpu;
+
+ if (unlikely(!flush)) {
+ local_irq_restore(flags);
+ goto do_native;
+ }
if (info->mm) {
flush->address_space = virt_to_phys(info->mm->pgd);
@@ -222,18 +251,18 @@ static void hyperv_flush_tlb_others_ex(const struct cpumask *cpus,
flush->flags |= HV_FLUSH_NON_GLOBAL_MAPPINGS_ONLY;
status = hv_do_rep_hypercall(
HVCALL_FLUSH_VIRTUAL_ADDRESS_SPACE_EX,
- 0, nr_bank + 2, flush, NULL);
+ 0, nr_bank, flush, NULL);
} else if (info->end &&
((info->end - info->start)/HV_TLB_FLUSH_UNIT) > max_gvas) {
status = hv_do_rep_hypercall(
HVCALL_FLUSH_VIRTUAL_ADDRESS_SPACE_EX,
- 0, nr_bank + 2, flush, NULL);
+ 0, nr_bank, flush, NULL);
} else {
gva_n = fill_gva_list(flush->gva_list, nr_bank,
info->start, info->end);
status = hv_do_rep_hypercall(
HVCALL_FLUSH_VIRTUAL_ADDRESS_LIST_EX,
- gva_n, nr_bank + 2, flush, NULL);
+ gva_n, nr_bank, flush, NULL);
}
local_irq_restore(flags);
@@ -266,7 +295,7 @@ void hyper_alloc_mmu(void)
return;
if (!(ms_hyperv.hints & HV_X64_EX_PROCESSOR_MASKS_RECOMMENDED))
- pcpu_flush = __alloc_percpu(PAGE_SIZE, PAGE_SIZE);
+ pcpu_flush = alloc_percpu(struct hv_flush_pcpu *);
else
- pcpu_flush_ex = __alloc_percpu(PAGE_SIZE, PAGE_SIZE);
+ pcpu_flush_ex = alloc_percpu(struct hv_flush_pcpu_ex *);
}
diff --git a/arch/x86/include/asm/alternative-asm.h b/arch/x86/include/asm/alternative-asm.h
index e7636bac7372..6c98821fef5e 100644
--- a/arch/x86/include/asm/alternative-asm.h
+++ b/arch/x86/include/asm/alternative-asm.h
@@ -62,8 +62,10 @@
#define new_len2 145f-144f
/*
- * max without conditionals. Idea adapted from:
+ * gas compatible max based on the idea from:
* http://graphics.stanford.edu/~seander/bithacks.html#IntegerMinOrMax
+ *
+ * The additional "-" is needed because gas uses a "true" value of -1.
*/
#define alt_max_short(a, b) ((a) ^ (((a) ^ (b)) & -(-((a) < (b)))))
diff --git a/arch/x86/include/asm/alternative.h b/arch/x86/include/asm/alternative.h
index c096624137ae..ccbe24e697c4 100644
--- a/arch/x86/include/asm/alternative.h
+++ b/arch/x86/include/asm/alternative.h
@@ -103,12 +103,12 @@ static inline int alternatives_text_reserved(void *start, void *end)
alt_end_marker ":\n"
/*
- * max without conditionals. Idea adapted from:
+ * gas compatible max based on the idea from:
* http://graphics.stanford.edu/~seander/bithacks.html#IntegerMinOrMax
*
- * The additional "-" is needed because gas works with s32s.
+ * The additional "-" is needed because gas uses a "true" value of -1.
*/
-#define alt_max_short(a, b) "((" a ") ^ (((" a ") ^ (" b ")) & -(-((" a ") - (" b ")))))"
+#define alt_max_short(a, b) "((" a ") ^ (((" a ") ^ (" b ")) & -(-((" a ") < (" b ")))))"
/*
* Pad the second replacement alternative with additional NOPs if it is
diff --git a/arch/x86/include/asm/mmu_context.h b/arch/x86/include/asm/mmu_context.h
index c120b5db178a..3c856a15b98e 100644
--- a/arch/x86/include/asm/mmu_context.h
+++ b/arch/x86/include/asm/mmu_context.h
@@ -126,13 +126,7 @@ static inline void switch_ldt(struct mm_struct *prev, struct mm_struct *next)
DEBUG_LOCKS_WARN_ON(preemptible());
}
-static inline void enter_lazy_tlb(struct mm_struct *mm, struct task_struct *tsk)
-{
- int cpu = smp_processor_id();
-
- if (cpumask_test_cpu(cpu, mm_cpumask(mm)))
- cpumask_clear_cpu(cpu, mm_cpumask(mm));
-}
+void enter_lazy_tlb(struct mm_struct *mm, struct task_struct *tsk);
static inline int init_new_context(struct task_struct *tsk,
struct mm_struct *mm)
diff --git a/arch/x86/include/asm/mshyperv.h b/arch/x86/include/asm/mshyperv.h
index 738503e1f80c..530f448fddaf 100644
--- a/arch/x86/include/asm/mshyperv.h
+++ b/arch/x86/include/asm/mshyperv.h
@@ -289,6 +289,7 @@ static inline u64 hv_do_rep_hypercall(u16 code, u16 rep_count, u16 varhead_size,
* to this information.
*/
extern u32 *hv_vp_index;
+extern u32 hv_max_vp_index;
/**
* hv_cpu_number_to_vp_number() - Map CPU to VP.
diff --git a/arch/x86/include/asm/tlbflush.h b/arch/x86/include/asm/tlbflush.h
index 4893abf7f74f..d362161d3291 100644
--- a/arch/x86/include/asm/tlbflush.h
+++ b/arch/x86/include/asm/tlbflush.h
@@ -83,6 +83,13 @@ static inline u64 inc_mm_tlb_gen(struct mm_struct *mm)
#endif
/*
+ * If tlb_use_lazy_mode is true, then we try to avoid switching CR3 to point
+ * to init_mm when we switch to a kernel thread (e.g. the idle thread). If
+ * it's false, then we immediately switch CR3 when entering a kernel thread.
+ */
+DECLARE_STATIC_KEY_TRUE(tlb_use_lazy_mode);
+
+/*
* 6 because 6 should be plenty and struct tlb_state will fit in
* two cache lines.
*/
@@ -105,6 +112,23 @@ struct tlb_state {
u16 next_asid;
/*
+ * We can be in one of several states:
+ *
+ * - Actively using an mm. Our CPU's bit will be set in
+ * mm_cpumask(loaded_mm) and is_lazy == false;
+ *
+ * - Not using a real mm. loaded_mm == &init_mm. Our CPU's bit
+ * will not be set in mm_cpumask(&init_mm) and is_lazy == false.
+ *
+ * - Lazily using a real mm. loaded_mm != &init_mm, our bit
+ * is set in mm_cpumask(loaded_mm), but is_lazy == true.
+ * We're heuristically guessing that the CR3 load we
+ * skipped more than makes up for the overhead added by
+ * lazy mode.
+ */
+ bool is_lazy;
+
+ /*
* Access to this CR4 shadow and to H/W CR4 is protected by
* disabling interrupts when modifying either one.
*/
diff --git a/arch/x86/kernel/apic/apic.c b/arch/x86/kernel/apic/apic.c
index d705c769f77d..ff891772c9f8 100644
--- a/arch/x86/kernel/apic/apic.c
+++ b/arch/x86/kernel/apic/apic.c
@@ -573,11 +573,21 @@ static u32 bdx_deadline_rev(void)
return ~0U;
}
+static u32 skx_deadline_rev(void)
+{
+ switch (boot_cpu_data.x86_mask) {
+ case 0x03: return 0x01000136;
+ case 0x04: return 0x02000014;
+ }
+
+ return ~0U;
+}
+
static const struct x86_cpu_id deadline_match[] = {
DEADLINE_MODEL_MATCH_FUNC( INTEL_FAM6_HASWELL_X, hsx_deadline_rev),
DEADLINE_MODEL_MATCH_REV ( INTEL_FAM6_BROADWELL_X, 0x0b000020),
DEADLINE_MODEL_MATCH_FUNC( INTEL_FAM6_BROADWELL_XEON_D, bdx_deadline_rev),
- DEADLINE_MODEL_MATCH_REV ( INTEL_FAM6_SKYLAKE_X, 0x02000014),
+ DEADLINE_MODEL_MATCH_FUNC( INTEL_FAM6_SKYLAKE_X, skx_deadline_rev),
DEADLINE_MODEL_MATCH_REV ( INTEL_FAM6_HASWELL_CORE, 0x22),
DEADLINE_MODEL_MATCH_REV ( INTEL_FAM6_HASWELL_ULT, 0x20),
@@ -600,7 +610,8 @@ static void apic_check_deadline_errata(void)
const struct x86_cpu_id *m;
u32 rev;
- if (!boot_cpu_has(X86_FEATURE_TSC_DEADLINE_TIMER))
+ if (!boot_cpu_has(X86_FEATURE_TSC_DEADLINE_TIMER) ||
+ boot_cpu_has(X86_FEATURE_HYPERVISOR))
return;
m = x86_match_cpu(deadline_match);
diff --git a/arch/x86/kernel/cpu/microcode/core.c b/arch/x86/kernel/cpu/microcode/core.c
index 86e8f0b2537b..c4fa4a85d4cb 100644
--- a/arch/x86/kernel/cpu/microcode/core.c
+++ b/arch/x86/kernel/cpu/microcode/core.c
@@ -122,9 +122,6 @@ static bool __init check_loader_disabled_bsp(void)
bool *res = &dis_ucode_ldr;
#endif
- if (!have_cpuid_p())
- return *res;
-
/*
* CPUID(1).ECX[31]: reserved for hypervisor use. This is still not
* completely accurate as xen pv guests don't see that CPUID bit set but
@@ -166,24 +163,36 @@ bool get_builtin_firmware(struct cpio_data *cd, const char *name)
void __init load_ucode_bsp(void)
{
unsigned int cpuid_1_eax;
+ bool intel = true;
- if (check_loader_disabled_bsp())
+ if (!have_cpuid_p())
return;
cpuid_1_eax = native_cpuid_eax(1);
switch (x86_cpuid_vendor()) {
case X86_VENDOR_INTEL:
- if (x86_family(cpuid_1_eax) >= 6)
- load_ucode_intel_bsp();
+ if (x86_family(cpuid_1_eax) < 6)
+ return;
break;
+
case X86_VENDOR_AMD:
- if (x86_family(cpuid_1_eax) >= 0x10)
- load_ucode_amd_bsp(cpuid_1_eax);
+ if (x86_family(cpuid_1_eax) < 0x10)
+ return;
+ intel = false;
break;
+
default:
- break;
+ return;
}
+
+ if (check_loader_disabled_bsp())
+ return;
+
+ if (intel)
+ load_ucode_intel_bsp();
+ else
+ load_ucode_amd_bsp(cpuid_1_eax);
}
static bool check_loader_disabled_ap(void)
diff --git a/arch/x86/kernel/kprobes/common.h b/arch/x86/kernel/kprobes/common.h
index db2182d63ed0..3fc0f9a794cb 100644
--- a/arch/x86/kernel/kprobes/common.h
+++ b/arch/x86/kernel/kprobes/common.h
@@ -3,6 +3,15 @@
/* Kprobes and Optprobes common header */
+#include <asm/asm.h>
+
+#ifdef CONFIG_FRAME_POINTER
+# define SAVE_RBP_STRING " push %" _ASM_BP "\n" \
+ " mov %" _ASM_SP ", %" _ASM_BP "\n"
+#else
+# define SAVE_RBP_STRING " push %" _ASM_BP "\n"
+#endif
+
#ifdef CONFIG_X86_64
#define SAVE_REGS_STRING \
/* Skip cs, ip, orig_ax. */ \
@@ -17,7 +26,7 @@
" pushq %r10\n" \
" pushq %r11\n" \
" pushq %rbx\n" \
- " pushq %rbp\n" \
+ SAVE_RBP_STRING \
" pushq %r12\n" \
" pushq %r13\n" \
" pushq %r14\n" \
@@ -48,7 +57,7 @@
" pushl %es\n" \
" pushl %ds\n" \
" pushl %eax\n" \
- " pushl %ebp\n" \
+ SAVE_RBP_STRING \
" pushl %edi\n" \
" pushl %esi\n" \
" pushl %edx\n" \
diff --git a/arch/x86/kernel/kprobes/core.c b/arch/x86/kernel/kprobes/core.c
index f0153714ddac..0742491cbb73 100644
--- a/arch/x86/kernel/kprobes/core.c
+++ b/arch/x86/kernel/kprobes/core.c
@@ -1080,8 +1080,6 @@ int setjmp_pre_handler(struct kprobe *p, struct pt_regs *regs)
* raw stack chunk with redzones:
*/
__memcpy(kcb->jprobes_stack, (kprobe_opcode_t *)addr, MIN_STACK_SIZE(addr));
- regs->flags &= ~X86_EFLAGS_IF;
- trace_hardirqs_off();
regs->ip = (unsigned long)(jp->entry);
/*
diff --git a/arch/x86/kernel/reboot.c b/arch/x86/kernel/reboot.c
index 54180fa6f66f..add33f600531 100644
--- a/arch/x86/kernel/reboot.c
+++ b/arch/x86/kernel/reboot.c
@@ -105,6 +105,10 @@ void __noreturn machine_real_restart(unsigned int type)
load_cr3(initial_page_table);
#else
write_cr3(real_mode_header->trampoline_pgd);
+
+ /* Exiting long mode will fail if CR4.PCIDE is set. */
+ if (static_cpu_has(X86_FEATURE_PCID))
+ cr4_clear_bits(X86_CR4_PCIDE);
#endif
/* Jump to the identity-mapped low memory code */
diff --git a/arch/x86/kernel/unwind_frame.c b/arch/x86/kernel/unwind_frame.c
index d145a0b1f529..3dc26f95d46e 100644
--- a/arch/x86/kernel/unwind_frame.c
+++ b/arch/x86/kernel/unwind_frame.c
@@ -44,7 +44,8 @@ static void unwind_dump(struct unwind_state *state)
state->stack_info.type, state->stack_info.next_sp,
state->stack_mask, state->graph_idx);
- for (sp = state->orig_sp; sp; sp = PTR_ALIGN(stack_info.next_sp, sizeof(long))) {
+ for (sp = PTR_ALIGN(state->orig_sp, sizeof(long)); sp;
+ sp = PTR_ALIGN(stack_info.next_sp, sizeof(long))) {
if (get_stack_info(sp, state->task, &stack_info, &visit_mask))
break;
@@ -174,6 +175,7 @@ static bool is_last_task_frame(struct unwind_state *state)
* This determines if the frame pointer actually contains an encoded pointer to
* pt_regs on the stack. See ENCODE_FRAME_POINTER.
*/
+#ifdef CONFIG_X86_64
static struct pt_regs *decode_frame_pointer(unsigned long *bp)
{
unsigned long regs = (unsigned long)bp;
@@ -183,6 +185,23 @@ static struct pt_regs *decode_frame_pointer(unsigned long *bp)
return (struct pt_regs *)(regs & ~0x1);
}
+#else
+static struct pt_regs *decode_frame_pointer(unsigned long *bp)
+{
+ unsigned long regs = (unsigned long)bp;
+
+ if (regs & 0x80000000)
+ return NULL;
+
+ return (struct pt_regs *)(regs | 0x80000000);
+}
+#endif
+
+#ifdef CONFIG_X86_32
+#define KERNEL_REGS_SIZE (sizeof(struct pt_regs) - 2*sizeof(long))
+#else
+#define KERNEL_REGS_SIZE (sizeof(struct pt_regs))
+#endif
static bool update_stack_state(struct unwind_state *state,
unsigned long *next_bp)
@@ -202,7 +221,7 @@ static bool update_stack_state(struct unwind_state *state,
regs = decode_frame_pointer(next_bp);
if (regs) {
frame = (unsigned long *)regs;
- len = regs_size(regs);
+ len = KERNEL_REGS_SIZE;
state->got_irq = true;
} else {
frame = next_bp;
@@ -226,6 +245,14 @@ static bool update_stack_state(struct unwind_state *state,
frame < prev_frame_end)
return false;
+ /*
+ * On 32-bit with user mode regs, make sure the last two regs are safe
+ * to access:
+ */
+ if (IS_ENABLED(CONFIG_X86_32) && regs && user_mode(regs) &&
+ !on_stack(info, frame, len + 2*sizeof(long)))
+ return false;
+
/* Move state to the next frame: */
if (regs) {
state->regs = regs;
@@ -328,6 +355,13 @@ bool unwind_next_frame(struct unwind_state *state)
state->regs->sp < (unsigned long)task_pt_regs(state->task))
goto the_end;
+ /*
+ * There are some known frame pointer issues on 32-bit. Disable
+ * unwinder warnings on 32-bit until it gets objtool support.
+ */
+ if (IS_ENABLED(CONFIG_X86_32))
+ goto the_end;
+
if (state->regs) {
printk_deferred_once(KERN_WARNING
"WARNING: kernel stack regs at %p in %s:%d has bad 'bp' value %p\n",
diff --git a/arch/x86/mm/Makefile b/arch/x86/mm/Makefile
index 72bf8c01c6e3..e1f095884386 100644
--- a/arch/x86/mm/Makefile
+++ b/arch/x86/mm/Makefile
@@ -1,5 +1,12 @@
-# Kernel does not boot with instrumentation of tlb.c.
-KCOV_INSTRUMENT_tlb.o := n
+# Kernel does not boot with instrumentation of tlb.c and mem_encrypt.c
+KCOV_INSTRUMENT_tlb.o := n
+KCOV_INSTRUMENT_mem_encrypt.o := n
+
+KASAN_SANITIZE_mem_encrypt.o := n
+
+ifdef CONFIG_FUNCTION_TRACER
+CFLAGS_REMOVE_mem_encrypt.o = -pg
+endif
obj-y := init.o init_$(BITS).o fault.o ioremap.o extable.o pageattr.o mmap.o \
pat.o pgtable.o physaddr.o setup_nx.o tlb.o
diff --git a/arch/x86/mm/tlb.c b/arch/x86/mm/tlb.c
index 49d9778376d7..658bf0090565 100644
--- a/arch/x86/mm/tlb.c
+++ b/arch/x86/mm/tlb.c
@@ -30,6 +30,8 @@
atomic64_t last_mm_ctx_id = ATOMIC64_INIT(1);
+DEFINE_STATIC_KEY_TRUE(tlb_use_lazy_mode);
+
static void choose_new_asid(struct mm_struct *next, u64 next_tlb_gen,
u16 *new_asid, bool *need_flush)
{
@@ -80,7 +82,7 @@ void leave_mm(int cpu)
return;
/* Warn if we're not lazy. */
- WARN_ON(cpumask_test_cpu(smp_processor_id(), mm_cpumask(loaded_mm)));
+ WARN_ON(!this_cpu_read(cpu_tlbstate.is_lazy));
switch_mm(NULL, &init_mm, NULL);
}
@@ -142,45 +144,24 @@ void switch_mm_irqs_off(struct mm_struct *prev, struct mm_struct *next,
__flush_tlb_all();
}
#endif
+ this_cpu_write(cpu_tlbstate.is_lazy, false);
if (real_prev == next) {
VM_BUG_ON(this_cpu_read(cpu_tlbstate.ctxs[prev_asid].ctx_id) !=
next->context.ctx_id);
- if (cpumask_test_cpu(cpu, mm_cpumask(next))) {
- /*
- * There's nothing to do: we weren't lazy, and we
- * aren't changing our mm. We don't need to flush
- * anything, nor do we need to update CR3, CR4, or
- * LDTR.
- */
- return;
- }
-
- /* Resume remote flushes and then read tlb_gen. */
- cpumask_set_cpu(cpu, mm_cpumask(next));
- next_tlb_gen = atomic64_read(&next->context.tlb_gen);
-
- if (this_cpu_read(cpu_tlbstate.ctxs[prev_asid].tlb_gen) <
- next_tlb_gen) {
- /*
- * Ideally, we'd have a flush_tlb() variant that
- * takes the known CR3 value as input. This would
- * be faster on Xen PV and on hypothetical CPUs
- * on which INVPCID is fast.
- */
- this_cpu_write(cpu_tlbstate.ctxs[prev_asid].tlb_gen,
- next_tlb_gen);
- write_cr3(build_cr3(next, prev_asid));
- trace_tlb_flush(TLB_FLUSH_ON_TASK_SWITCH,
- TLB_FLUSH_ALL);
- }
-
/*
- * We just exited lazy mode, which means that CR4 and/or LDTR
- * may be stale. (Changes to the required CR4 and LDTR states
- * are not reflected in tlb_gen.)
+ * We don't currently support having a real mm loaded without
+ * our cpu set in mm_cpumask(). We have all the bookkeeping
+ * in place to figure out whether we would need to flush
+ * if our cpu were cleared in mm_cpumask(), but we don't
+ * currently use it.
*/
+ if (WARN_ON_ONCE(real_prev != &init_mm &&
+ !cpumask_test_cpu(cpu, mm_cpumask(next))))
+ cpumask_set_cpu(cpu, mm_cpumask(next));
+
+ return;
} else {
u16 new_asid;
bool need_flush;
@@ -199,10 +180,9 @@ void switch_mm_irqs_off(struct mm_struct *prev, struct mm_struct *next,
}
/* Stop remote flushes for the previous mm */
- if (cpumask_test_cpu(cpu, mm_cpumask(real_prev)))
- cpumask_clear_cpu(cpu, mm_cpumask(real_prev));
-
- VM_WARN_ON_ONCE(cpumask_test_cpu(cpu, mm_cpumask(next)));
+ VM_WARN_ON_ONCE(!cpumask_test_cpu(cpu, mm_cpumask(real_prev)) &&
+ real_prev != &init_mm);
+ cpumask_clear_cpu(cpu, mm_cpumask(real_prev));
/*
* Start remote flushes and then read tlb_gen.
@@ -233,6 +213,37 @@ void switch_mm_irqs_off(struct mm_struct *prev, struct mm_struct *next,
}
/*
+ * enter_lazy_tlb() is a hint from the scheduler that we are entering a
+ * kernel thread or other context without an mm. Acceptable implementations
+ * include doing nothing whatsoever, switching to init_mm, or various clever
+ * lazy tricks to try to minimize TLB flushes.
+ *
+ * The scheduler reserves the right to call enter_lazy_tlb() several times
+ * in a row. It will notify us that we're going back to a real mm by
+ * calling switch_mm_irqs_off().
+ */
+void enter_lazy_tlb(struct mm_struct *mm, struct task_struct *tsk)
+{
+ if (this_cpu_read(cpu_tlbstate.loaded_mm) == &init_mm)
+ return;
+
+ if (static_branch_unlikely(&tlb_use_lazy_mode)) {
+ /*
+ * There's a significant optimization that may be possible
+ * here. We have accurate enough TLB flush tracking that we
+ * don't need to maintain coherence of TLB per se when we're
+ * lazy. We do, however, need to maintain coherence of
+ * paging-structure caches. We could, in principle, leave our
+ * old mm loaded and only switch to init_mm when
+ * tlb_remove_page() happens.
+ */
+ this_cpu_write(cpu_tlbstate.is_lazy, true);
+ } else {
+ switch_mm(NULL, &init_mm, NULL);
+ }
+}
+
+/*
* Call this when reinitializing a CPU. It fixes the following potential
* problems:
*
@@ -303,16 +314,20 @@ static void flush_tlb_func_common(const struct flush_tlb_info *f,
/* This code cannot presently handle being reentered. */
VM_WARN_ON(!irqs_disabled());
+ if (unlikely(loaded_mm == &init_mm))
+ return;
+
VM_WARN_ON(this_cpu_read(cpu_tlbstate.ctxs[loaded_mm_asid].ctx_id) !=
loaded_mm->context.ctx_id);
- if (!cpumask_test_cpu(smp_processor_id(), mm_cpumask(loaded_mm))) {
+ if (this_cpu_read(cpu_tlbstate.is_lazy)) {
/*
- * We're in lazy mode -- don't flush. We can get here on
- * remote flushes due to races and on local flushes if a
- * kernel thread coincidentally flushes the mm it's lazily
- * still using.
+ * We're in lazy mode. We need to at least flush our
+ * paging-structure cache to avoid speculatively reading
+ * garbage into our TLB. Since switching to init_mm is barely
+ * slower than a minimal flush, just switch to init_mm.
*/
+ switch_mm_irqs_off(NULL, &init_mm, NULL);
return;
}
@@ -611,3 +626,57 @@ static int __init create_tlb_single_page_flush_ceiling(void)
return 0;
}
late_initcall(create_tlb_single_page_flush_ceiling);
+
+static ssize_t tlblazy_read_file(struct file *file, char __user *user_buf,
+ size_t count, loff_t *ppos)
+{
+ char buf[2];
+
+ buf[0] = static_branch_likely(&tlb_use_lazy_mode) ? '1' : '0';
+ buf[1] = '\n';
+
+ return simple_read_from_buffer(user_buf, count, ppos, buf, 2);
+}
+
+static ssize_t tlblazy_write_file(struct file *file,
+ const char __user *user_buf, size_t count, loff_t *ppos)
+{
+ bool val;
+
+ if (kstrtobool_from_user(user_buf, count, &val))
+ return -EINVAL;
+
+ if (val)
+ static_branch_enable(&tlb_use_lazy_mode);
+ else
+ static_branch_disable(&tlb_use_lazy_mode);
+
+ return count;
+}
+
+static const struct file_operations fops_tlblazy = {
+ .read = tlblazy_read_file,
+ .write = tlblazy_write_file,
+ .llseek = default_llseek,
+};
+
+static int __init init_tlb_use_lazy_mode(void)
+{
+ if (boot_cpu_has(X86_FEATURE_PCID)) {
+ /*
+ * Heuristic: with PCID on, switching to and from
+ * init_mm is reasonably fast, but remote flush IPIs
+ * as expensive as ever, so turn off lazy TLB mode.
+ *
+ * We can't do this in setup_pcid() because static keys
+ * haven't been initialized yet, and it would blow up
+ * badly.
+ */
+ static_branch_disable(&tlb_use_lazy_mode);
+ }
+
+ debugfs_create_file("tlb_use_lazy_mode", S_IRUSR | S_IWUSR,
+ arch_debugfs_dir, NULL, &fops_tlblazy);
+ return 0;
+}
+late_initcall(init_tlb_use_lazy_mode);
next reply other threads:[~2017-10-14 16:16 UTC|newest]
Thread overview: 567+ messages / expand[flat|nested] mbox.gz Atom feed top
2017-10-14 16:16 Ingo Molnar [this message]
-- strict thread matches above, loose matches on Subject: below --
2024-05-05 9:14 [GIT PULL] x86 fixes Ingo Molnar
2024-05-05 19:43 ` pr-tracker-bot
2024-04-28 8:34 Ingo Molnar
2024-04-28 19:28 ` pr-tracker-bot
2024-04-14 9:08 Ingo Molnar
2024-04-14 18:48 ` pr-tracker-bot
2023-12-23 14:34 Ingo Molnar
2023-12-23 20:21 ` pr-tracker-bot
2024-04-07 8:15 ` Ingo Molnar
2024-04-07 16:44 ` pr-tracker-bot
2023-11-26 9:51 Ingo Molnar
2023-11-26 17:16 ` pr-tracker-bot
2023-10-28 10:50 Ingo Molnar
2023-10-28 18:17 ` pr-tracker-bot
2023-10-14 22:08 Ingo Molnar
2023-10-14 22:49 ` pr-tracker-bot
2023-10-08 9:41 Ingo Molnar
2023-10-08 18:06 ` pr-tracker-bot
2023-10-01 9:04 Ingo Molnar
2023-10-01 17:08 ` pr-tracker-bot
2023-09-22 10:33 Ingo Molnar
2023-09-22 20:19 ` pr-tracker-bot
2023-09-17 17:44 Ingo Molnar
2023-09-17 18:24 ` pr-tracker-bot
2023-09-02 10:24 [GIT PULL] x86 fix Ingo Molnar
2023-09-10 16:26 ` [GIT PULL] x86 fixes Ingo Molnar
2023-09-10 18:08 ` pr-tracker-bot
2023-08-26 17:54 Ingo Molnar
2023-08-26 18:08 ` pr-tracker-bot
2023-02-11 8:59 Ingo Molnar
2023-02-11 19:24 ` pr-tracker-bot
2022-08-28 14:57 Ingo Molnar
2022-08-28 18:18 ` pr-tracker-bot
2022-08-06 19:29 Ingo Molnar
2022-08-07 0:50 ` pr-tracker-bot
2021-03-28 10:44 Ingo Molnar
2021-03-28 19:22 ` pr-tracker-bot
2020-10-11 8:08 Ingo Molnar
2020-10-11 18:00 ` Linus Torvalds
2020-10-11 20:00 ` Thomas Gleixner
2020-10-11 18:23 ` pr-tracker-bot
2020-09-06 8:15 Ingo Molnar
2020-09-06 19:14 ` pr-tracker-bot
2020-08-15 11:45 Ingo Molnar
2020-08-16 1:55 ` pr-tracker-bot
2020-07-25 11:46 Ingo Molnar
2020-07-25 22:30 ` pr-tracker-bot
2020-03-02 8:49 Ingo Molnar
2020-03-03 23:35 ` pr-tracker-bot
2020-01-31 11:52 Ingo Molnar
2020-01-31 19:35 ` pr-tracker-bot
2020-01-18 18:52 Ingo Molnar
2020-01-18 21:05 ` pr-tracker-bot
2019-12-01 22:22 Ingo Molnar
2019-12-02 4:40 ` pr-tracker-bot
2019-11-16 21:42 Ingo Molnar
2019-11-17 0:35 ` pr-tracker-bot
2019-10-12 13:19 Ingo Molnar
2019-10-12 22:35 ` pr-tracker-bot
2019-09-12 12:57 Ingo Molnar
2019-09-12 14:05 ` pr-tracker-bot
2019-09-05 8:07 Ingo Molnar
2019-09-05 21:15 ` pr-tracker-bot
2019-06-29 9:14 Ingo Molnar
2019-06-29 11:45 ` pr-tracker-bot
2019-06-02 17:44 Ingo Molnar
2019-06-02 18:15 ` pr-tracker-bot
2019-05-16 16:26 Ingo Molnar
2019-05-16 18:20 ` pr-tracker-bot
2019-04-27 14:42 Ingo Molnar
2019-04-27 18:45 ` pr-tracker-bot
2019-04-20 7:38 Ingo Molnar
2019-04-20 19:25 ` pr-tracker-bot
2019-04-12 13:10 Ingo Molnar
2019-04-13 4:05 ` pr-tracker-bot
2019-02-17 10:19 Ingo Molnar
2019-02-17 16:50 ` pr-tracker-bot
2019-02-10 9:13 Ingo Molnar
2019-02-10 18:30 ` pr-tracker-bot
2019-01-11 7:14 Ingo Molnar
2019-01-11 18:00 ` pr-tracker-bot
2018-12-21 12:25 Ingo Molnar
2018-12-21 19:30 ` pr-tracker-bot
2018-12-09 22:06 Ingo Molnar
2018-12-09 23:45 ` pr-tracker-bot
2018-11-30 6:29 Ingo Molnar
2018-11-30 21:00 ` pr-tracker-bot
2018-11-03 23:09 Ingo Molnar
2018-11-04 1:27 ` Linus Torvalds
2018-10-20 8:54 Ingo Molnar
2018-10-20 13:28 ` Greg Kroah-Hartman
2018-10-11 9:14 Ingo Molnar
2018-10-11 12:32 ` Greg Kroah-Hartman
2018-10-05 9:53 Ingo Molnar
2018-10-05 23:06 ` Greg Kroah-Hartman
2018-09-15 13:24 Ingo Molnar
2018-07-30 17:59 Ingo Molnar
2018-06-30 8:49 Ingo Molnar
2018-06-30 19:01 ` Linus Torvalds
2018-07-02 18:47 ` Andy Lutomirski
2018-07-02 18:53 ` Linus Torvalds
2018-07-03 7:56 ` Ingo Molnar
2018-03-31 10:36 Ingo Molnar
2018-02-15 0:45 Ingo Molnar
2018-01-17 15:41 Ingo Molnar
2018-01-17 20:35 ` Linus Torvalds
2018-01-18 0:24 ` Ingo Molnar
2018-01-18 0:29 ` Andrew Morton
2018-01-12 13:56 Ingo Molnar
2017-12-15 15:43 Ingo Molnar
2017-12-15 15:50 ` Andy Lutomirski
2017-12-15 16:07 ` Ingo Molnar
2017-12-17 3:25 ` Andy Lutomirski
2017-12-17 8:32 ` Ingo Molnar
2017-12-17 11:41 ` Thomas Gleixner
2017-12-17 15:15 ` Borislav Petkov
2017-12-06 22:36 Ingo Molnar
2017-11-26 12:48 Ingo Molnar
2017-11-05 14:46 Ingo Molnar
2017-10-27 19:24 Ingo Molnar
2017-09-24 11:28 Ingo Molnar
2017-09-13 17:54 Ingo Molnar
2017-09-12 15:38 Ingo Molnar
2017-08-26 7:26 Ingo Molnar
2017-07-21 10:26 Ingo Molnar
2017-06-10 9:03 Ingo Molnar
2017-06-02 6:54 Ingo Molnar
2017-05-12 7:39 Ingo Molnar
2017-03-07 20:40 Ingo Molnar
2017-02-28 8:08 Ingo Molnar
2017-02-11 18:18 Ingo Molnar
2017-02-02 21:04 Ingo Molnar
2017-01-15 10:06 Ingo Molnar
2016-12-23 22:57 Ingo Molnar
2016-12-07 18:53 Ingo Molnar
2016-11-22 15:41 Ingo Molnar
2016-11-14 8:03 Ingo Molnar
2016-10-28 8:41 Ingo Molnar
2016-10-22 11:16 Ingo Molnar
2016-10-18 11:22 Ingo Molnar
2016-09-13 18:20 Ingo Molnar
2016-08-18 20:49 Ingo Molnar
2016-08-12 19:46 Ingo Molnar
2016-08-06 6:13 Ingo Molnar
2016-07-13 12:54 Ingo Molnar
2016-07-08 14:00 Ingo Molnar
2016-06-10 14:43 Ingo Molnar
2016-05-25 22:00 Ingo Molnar
2016-05-10 12:01 Ingo Molnar
2016-05-06 19:20 Ingo Molnar
2016-04-28 18:00 Ingo Molnar
2016-04-23 11:38 Ingo Molnar
2016-04-14 14:13 Ingo Molnar
2016-03-24 8:01 Ingo Molnar
2016-03-12 19:06 Ingo Molnar
2016-02-20 11:30 Ingo Molnar
2016-01-14 10:16 Ingo Molnar
2016-01-08 12:57 Ingo Molnar
2015-10-23 11:45 Ingo Molnar
2015-10-03 10:24 Ingo Molnar
2015-10-03 10:57 ` Ingo Molnar
2015-10-03 19:40 ` Thomas Gleixner
2015-09-17 8:28 Ingo Molnar
2015-08-22 12:21 Ingo Molnar
2015-08-14 7:15 Ingo Molnar
2015-08-14 18:25 ` Linus Torvalds
2015-08-14 18:46 ` Andy Lutomirski
2015-08-14 18:57 ` Linus Torvalds
2015-08-14 19:06 ` Linus Torvalds
2015-08-14 19:18 ` Andy Lutomirski
2015-08-14 19:37 ` Linus Torvalds
2015-08-14 19:14 ` Andy Lutomirski
2015-08-17 8:01 ` Ingo Molnar
2015-08-17 10:59 ` Denys Vlasenko
2015-08-17 16:57 ` Linus Torvalds
2015-08-18 7:57 ` Ingo Molnar
2015-08-17 16:47 ` Linus Torvalds
2015-08-17 16:58 ` H. Peter Anvin
2015-08-17 17:17 ` Linus Torvalds
2015-08-17 22:17 ` H. Peter Anvin
2015-08-19 5:59 ` Ingo Molnar
2015-08-19 6:15 ` Ingo Molnar
2015-08-19 6:50 ` Ingo Molnar
2015-08-19 10:00 ` H. Peter Anvin
2015-08-19 22:33 ` Linus Torvalds
2015-08-20 6:54 ` H. Peter Anvin
2015-08-19 21:53 ` H. Peter Anvin
2015-08-21 10:17 ` Denys Vlasenko
2015-08-17 23:47 ` Bryan O'Donoghue
2015-08-17 21:03 ` H. Peter Anvin
2015-08-17 23:59 ` Andy Lutomirski
2015-08-18 0:01 ` H. Peter Anvin
2015-08-18 0:06 ` H. Peter Anvin
2015-08-18 0:19 ` Andy Lutomirski
2015-08-18 5:56 ` H. Peter Anvin
2015-08-18 5:59 ` H. Peter Anvin
2015-08-18 7:55 ` Ingo Molnar
2015-08-01 8:44 Ingo Molnar
2015-07-18 3:18 Ingo Molnar
2015-07-20 7:20 ` Heiko Carstens
2015-07-04 11:29 Ingo Molnar
2015-06-05 8:40 Ingo Molnar
2015-05-27 12:54 Ingo Molnar
2015-05-06 12:58 Ingo Molnar
2015-05-06 18:14 ` Linus Torvalds
2015-04-18 15:26 Ingo Molnar
2015-04-03 13:16 Ingo Molnar
2015-03-17 16:54 Ingo Molnar
2015-03-05 17:02 Ingo Molnar
2015-03-01 17:14 Ingo Molnar
2015-02-20 13:47 Ingo Molnar
2015-01-11 8:51 Ingo Molnar
2014-12-14 19:46 Ingo Molnar
2014-11-20 8:02 Ingo Molnar
2014-11-16 9:07 Ingo Molnar
2014-11-17 7:42 ` Markus Trippelsdorf
2014-11-17 8:27 ` Markus Trippelsdorf
2014-11-17 13:58 ` Ingo Molnar
2014-11-17 21:02 ` Kees Cook
2014-11-17 21:05 ` Markus Trippelsdorf
2014-11-17 21:21 ` Markus Trippelsdorf
2014-11-17 23:09 ` Kees Cook
2014-10-31 11:26 Ingo Molnar
2014-09-27 6:02 Ingo Molnar
2014-09-19 10:40 Ingo Molnar
2014-09-23 5:22 ` Linus Torvalds
2014-09-23 5:35 ` Ingo Molnar
2014-09-23 5:37 ` Ingo Molnar
2014-09-23 5:44 ` H. Peter Anvin
2014-09-23 5:59 ` Linus Torvalds
2014-09-23 6:07 ` Linus Torvalds
2014-09-23 6:56 ` Matt Fleming
[not found] ` <CA+55aFz+2tf7zEGjVmkVuncZssiDdVRKJ=OUfgnDFf2TYN-KvA@mail.gmail.com>
2014-09-23 7:35 ` Matt Fleming
2014-09-23 12:18 ` Josh Boyer
2014-09-23 5:58 ` Ingo Molnar
2014-09-23 7:20 ` Matt Fleming
2014-09-23 8:18 ` Ard Biesheuvel
2014-09-23 13:18 ` Matt Fleming
2014-09-23 13:59 ` Leif Lindholm
2014-09-23 14:25 ` Maarten Lankhorst
2014-09-23 14:37 ` Matt Fleming
2014-09-23 16:01 ` Linus Torvalds
2014-09-24 7:26 ` Ingo Molnar
2014-09-24 11:42 ` Matt Fleming
2014-09-24 13:08 ` Ingo Molnar
2014-09-24 13:18 ` Matt Fleming
2014-09-24 13:18 ` Ingo Molnar
2014-09-23 16:05 ` Linus Torvalds
2014-09-23 16:11 ` Matt Fleming
2014-09-23 16:17 ` Josh Boyer
2014-09-23 17:21 ` Josh Boyer
2014-09-23 20:43 ` Matt Fleming
2014-08-24 20:28 Ingo Molnar
2014-04-16 13:21 Ingo Molnar
2013-11-13 20:47 Ingo Molnar
2013-10-18 19:11 Ingo Molnar
2013-10-12 17:15 Ingo Molnar
2013-10-12 18:05 ` Linus Torvalds
2013-10-12 18:18 ` H. Peter Anvin
2013-10-12 18:49 ` Ingo Molnar
2013-10-15 7:15 ` Ingo Molnar
2013-10-15 10:57 ` Borislav Petkov
2013-10-12 19:28 ` Matthew Garrett
2013-10-12 19:41 ` Linus Torvalds
2013-10-12 20:35 ` H. Peter Anvin
2013-10-04 7:57 Ingo Molnar
2013-09-25 18:16 Ingo Molnar
2013-09-18 16:24 Ingo Molnar
2013-09-05 11:03 Ingo Molnar
2013-08-19 11:23 Ingo Molnar
2013-04-14 15:55 Ingo Molnar
2013-02-26 12:10 Ingo Molnar
2013-02-04 18:31 Ingo Molnar
2012-10-26 14:52 Ingo Molnar
2012-09-21 19:15 Ingo Molnar
2012-08-23 10:54 Ingo Molnar
2012-08-20 9:21 Ingo Molnar
2012-08-21 8:00 ` Ingo Molnar
2012-08-03 16:51 Ingo Molnar
2012-06-29 15:33 Ingo Molnar
2012-06-15 18:53 Ingo Molnar
2012-06-08 14:46 Ingo Molnar
2012-05-17 8:24 Ingo Molnar
2012-04-27 6:57 Ingo Molnar
2012-04-03 22:45 Ingo Molnar
2012-04-03 23:47 ` Konrad Rzeszutek Wilk
2012-04-04 6:56 ` Ingo Molnar
2012-04-04 13:03 ` Konrad Rzeszutek Wilk
2012-02-27 10:32 Ingo Molnar
2012-02-02 10:10 Ingo Molnar
2012-01-26 20:15 Ingo Molnar
2012-01-15 13:40 Ingo Molnar
2011-12-13 23:00 Ingo Molnar
2011-12-05 19:18 Ingo Molnar
2011-07-07 18:24 Ingo Molnar
2011-06-19 9:09 Ingo Molnar
2011-06-13 9:49 Ingo Molnar
2011-06-07 18:44 Ingo Molnar
2011-05-31 16:30 Ingo Molnar
2011-05-31 16:35 ` Joe Perches
2011-05-31 18:16 ` Borislav Petkov
2011-05-31 19:04 ` Ingo Molnar
2011-05-31 19:51 ` Borislav Petkov
2011-05-31 21:35 ` Linus Torvalds
2011-06-01 6:00 ` Ingo Molnar
2011-06-01 6:08 ` Joe Perches
2011-06-01 6:18 ` Borislav Petkov
2011-06-01 6:24 ` Ingo Molnar
2011-05-23 10:19 Ingo Molnar
2011-05-17 21:43 Ingo Molnar
2011-05-03 11:44 Ingo Molnar
2011-04-29 18:02 Ingo Molnar
2011-04-21 16:06 Ingo Molnar
2011-04-16 10:11 Ingo Molnar
2011-04-07 17:36 Ingo Molnar
2011-04-02 10:52 Ingo Molnar
2011-03-25 13:35 Ingo Molnar
2011-03-22 10:20 Ingo Molnar
2011-03-18 13:54 Ingo Molnar
2011-03-16 16:21 Ingo Molnar
2011-03-10 8:10 Ingo Molnar
2011-02-28 17:37 Ingo Molnar
2011-02-25 19:58 Ingo Molnar
2011-02-15 16:36 Ingo Molnar
2011-02-06 11:18 Ingo Molnar
2011-01-27 17:28 Ingo Molnar
2011-01-24 13:01 Ingo Molnar
2011-01-19 19:01 Ingo Molnar
2011-01-18 19:05 Ingo Molnar
2011-01-15 15:17 Ingo Molnar
2010-12-19 15:30 Ingo Molnar
2010-11-26 13:27 Ingo Molnar
2010-11-11 11:03 Ingo Molnar
2010-10-27 16:05 Ingo Molnar
2010-10-27 16:07 ` Ingo Molnar
2010-09-26 8:50 Ingo Molnar
2010-09-08 13:08 Ingo Molnar
2010-06-02 11:49 Ingo Molnar
2010-03-30 12:30 Ingo Molnar
2010-03-26 15:43 Ingo Molnar
2010-03-29 15:47 ` Linus Torvalds
2010-03-29 16:47 ` Ingo Molnar
2010-03-13 16:39 Ingo Molnar
2010-01-31 17:19 Ingo Molnar
2010-01-16 17:03 Ingo Molnar
2010-01-16 20:34 ` Linus Torvalds
2010-01-16 20:53 ` Cyrill Gorcunov
2010-01-16 21:16 ` Ian Campbell
2010-01-16 22:12 ` Cyrill Gorcunov
2010-01-17 0:50 ` H. Peter Anvin
2010-01-16 21:06 ` H. Peter Anvin
2010-01-16 21:09 ` H. Peter Anvin
2010-01-17 0:18 ` Brian Gerst
2010-01-17 6:00 ` Ian Campbell
2009-12-31 12:03 Ingo Molnar
2009-12-31 12:56 ` Borislav Petkov
2009-12-18 18:56 Ingo Molnar
2009-12-15 20:36 Ingo Molnar
2009-12-14 19:06 Ingo Molnar
2009-12-10 19:42 Ingo Molnar
2009-11-10 17:40 Ingo Molnar
2009-11-04 15:48 Ingo Molnar
2009-11-01 15:24 Ingo Molnar
2009-10-23 14:40 Ingo Molnar
2009-10-15 10:55 Ingo Molnar
2009-10-13 18:15 Ingo Molnar
2009-10-08 18:57 Ingo Molnar
2009-10-02 12:36 Ingo Molnar
2009-09-26 12:21 Ingo Molnar
2009-09-21 12:59 Ingo Molnar
2009-08-28 10:40 Ingo Molnar
2009-08-25 18:00 Ingo Molnar
2009-08-17 21:37 Ingo Molnar
2009-08-13 18:49 Ingo Molnar
2009-08-09 16:01 Ingo Molnar
2009-08-04 18:55 Ingo Molnar
2009-06-26 19:07 Ingo Molnar
2009-06-12 10:47 Ingo Molnar
2009-05-18 14:38 Ingo Molnar
2009-05-08 18:46 Ingo Molnar
2009-05-05 9:26 Ingo Molnar
2009-04-26 17:18 Ingo Molnar
2009-04-17 1:32 Ingo Molnar
2009-04-13 17:36 Ingo Molnar
2009-04-09 15:47 Ingo Molnar
2009-04-03 22:46 Ingo Molnar
2009-03-06 18:36 [git pull] " Ingo Molnar
2009-03-03 20:59 Ingo Molnar
2009-03-02 8:47 Ingo Molnar
2009-02-27 16:28 Ingo Molnar
2009-02-21 17:08 Ingo Molnar
2009-02-20 14:18 Ingo Molnar
2009-02-19 17:10 Ingo Molnar
2009-02-21 2:13 ` Linus Torvalds
2009-02-21 6:56 ` H. Peter Anvin
2009-02-21 8:32 ` Ingo Molnar
2009-02-21 8:39 ` Ingo Molnar
2009-02-21 8:42 ` H. Peter Anvin
2009-02-21 9:18 ` Sam Ravnborg
2009-02-21 9:46 ` Ingo Molnar
2009-02-17 16:36 Ingo Molnar
2009-02-11 14:31 Ingo Molnar
2009-02-04 19:22 Ingo Molnar
2009-01-30 23:00 Ingo Molnar
2009-01-26 17:17 Ingo Molnar
2009-01-26 19:05 ` Andrew Morton
2009-01-26 19:20 ` Ingo Molnar
2009-01-26 19:40 ` Andrew Morton
2009-01-26 19:59 ` Ingo Molnar
2009-01-26 20:14 ` Andrew Morton
2009-01-26 20:28 ` Ingo Molnar
2009-01-19 23:23 Ingo Molnar
2009-01-12 18:28 Ingo Molnar
2009-01-11 14:39 Ingo Molnar
2009-01-11 16:45 ` Torsten Kaiser
2009-01-11 18:18 ` Ingo Molnar
2009-01-12 18:17 ` Pallipadi, Venkatesh
2009-01-12 19:01 ` Torsten Kaiser
2009-01-12 19:19 ` Pallipadi, Venkatesh
2009-01-12 19:29 ` Pallipadi, Venkatesh
2009-01-12 19:47 ` Linus Torvalds
2009-01-12 19:54 ` Pallipadi, Venkatesh
2009-01-12 20:38 ` Ingo Molnar
2009-01-12 20:52 ` Ingo Molnar
2009-01-12 21:03 ` Harvey Harrison
2009-01-12 21:12 ` Ingo Molnar
2009-01-12 21:55 ` Torsten Kaiser
2009-01-12 22:03 ` Ingo Molnar
2009-01-12 20:05 ` Torsten Kaiser
2009-01-12 20:40 ` Ingo Molnar
2009-01-12 21:50 ` Torsten Kaiser
2009-01-12 22:13 ` Ingo Molnar
2009-01-13 19:20 ` Torsten Kaiser
2009-01-12 22:16 ` Ingo Molnar
2009-01-02 21:48 Ingo Molnar
2008-12-20 13:43 Ingo Molnar
2008-12-20 19:16 ` Linus Torvalds
2008-12-20 19:31 ` Ingo Molnar
2008-12-20 22:11 ` Linus Torvalds
2008-12-20 20:58 ` Joerg Roedel
2008-12-08 18:26 Ingo Molnar
2008-12-04 19:46 Ingo Molnar
2008-11-29 19:31 Ingo Molnar
2008-11-20 11:22 Ingo Molnar
2008-11-18 20:35 Ingo Molnar
2008-11-06 21:29 Ingo Molnar
2008-11-01 17:06 Ingo Molnar
2008-10-30 23:34 Ingo Molnar
2008-10-28 10:49 Ingo Molnar
2008-10-23 19:33 Ingo Molnar
2008-10-17 17:27 Ingo Molnar
2008-10-15 16:32 Ingo Molnar
2008-10-01 18:05 Ingo Molnar
2008-09-27 21:02 Ingo Molnar
2008-09-23 19:34 Ingo Molnar
2008-09-17 9:58 Ingo Molnar
2008-09-09 19:03 H. Peter Anvin
2008-09-08 19:32 H. Peter Anvin
2008-09-08 20:34 ` David Sanders
2008-09-08 21:20 ` H. Peter Anvin
2008-09-08 21:22 ` H. Peter Anvin
2008-09-08 21:43 ` H. Peter Anvin
2008-09-08 22:16 ` David Sanders
2008-09-09 6:05 ` Ingo Molnar
2008-09-09 7:19 ` Ingo Molnar
2008-09-09 19:18 ` David Sanders
2008-09-09 19:56 ` Linus Torvalds
2008-09-09 20:37 ` David Sanders
2008-09-09 20:45 ` Linus Torvalds
2008-09-09 20:46 ` Linus Torvalds
2008-09-09 20:49 ` Ingo Molnar
2008-09-09 20:53 ` David Sanders
2008-09-08 17:52 H. Peter Anvin
2008-09-08 18:04 ` Linus Torvalds
2008-09-08 18:17 ` Linus Torvalds
2008-09-08 22:42 ` Andi Kleen
2008-09-08 18:22 ` H. Peter Anvin
2008-09-08 18:46 ` Arjan van de Ven
2008-09-08 18:51 ` H. Peter Anvin
2008-09-08 19:02 ` Ingo Molnar
2008-09-08 19:30 ` Linus Torvalds
2008-09-08 19:55 ` Arjan van de Ven
2008-09-08 20:14 ` H. Peter Anvin
2008-09-08 23:17 ` Krzysztof Halasa
2008-09-08 18:42 ` Arjan van de Ven
2008-09-09 10:24 ` Andi Kleen
2008-09-09 14:54 ` Linus Torvalds
2008-09-09 17:01 ` H. Peter Anvin
2008-09-09 17:17 ` Mark Lord
2008-09-09 17:19 ` H. Peter Anvin
2008-09-09 17:48 ` Mark Lord
2008-09-09 18:40 ` Andi Kleen
2008-09-09 16:05 ` Adrian Bunk
2008-09-09 16:15 ` Linus Torvalds
2008-09-08 20:25 ` Valdis.Kletnieks
2008-09-09 7:27 ` Ingo Molnar
2008-09-08 22:43 ` Andi Kleen
2008-09-09 16:57 ` Adrian Bunk
2008-09-09 17:03 ` H. Peter Anvin
2008-09-09 17:43 ` Adrian Bunk
2008-09-09 18:12 ` H. Peter Anvin
2008-09-06 19:01 Ingo Molnar
2008-09-05 18:51 Ingo Molnar
2008-08-28 11:41 Ingo Molnar
2008-08-25 17:50 Ingo Molnar
2008-08-22 12:23 Ingo Molnar
2008-08-18 18:36 Ingo Molnar
2008-07-31 21:42 Ingo Molnar
2008-07-29 15:53 Ingo Molnar
2008-07-26 19:15 Ingo Molnar
2008-07-24 15:12 Ingo Molnar
2008-07-24 19:36 ` Linus Torvalds
2008-07-24 20:38 ` H. Peter Anvin
2008-07-22 14:03 Ingo Molnar
2008-07-22 14:35 ` Johannes Weiner
2008-07-22 15:08 ` Jeremy Fitzhardinge
2008-07-22 15:23 ` Johannes Weiner
2008-07-17 17:32 Ingo Molnar
2008-07-15 15:01 Ingo Molnar
2008-07-15 15:13 ` Ingo Molnar
2008-07-15 16:03 ` Linus Torvalds
2008-07-05 19:29 Ingo Molnar
2008-07-04 16:48 Ingo Molnar
2008-06-30 15:30 Ingo Molnar
2008-06-19 15:13 Ingo Molnar
2008-06-19 21:29 ` Simon Holm Thøgersen
2008-06-19 23:34 ` Suresh Siddha
2008-06-12 19:51 Ingo Molnar
2008-05-13 19:27 Ingo Molnar
2008-05-13 19:40 ` Adrian Bunk
2008-05-13 20:02 ` Adrian Bunk
2008-05-13 20:38 ` Adrian Bunk
2008-05-13 21:01 ` H. Peter Anvin
2008-05-13 20:20 ` Linus Torvalds
2008-05-04 19:35 Ingo Molnar
2008-05-05 15:12 ` Adrian Bunk
2008-05-05 15:29 ` Andres Salomon
2008-05-06 12:49 ` Thomas Gleixner
2008-05-07 15:41 ` Andres Salomon
2008-05-07 19:08 ` Thomas Gleixner
2008-05-07 19:48 ` Andres Salomon
2008-05-07 20:07 ` Andrew Morton
2008-05-09 10:28 ` Ingo Molnar
2008-04-30 21:24 Ingo Molnar
2008-04-24 21:37 Ingo Molnar
2008-04-07 19:38 Ingo Molnar
2008-03-27 20:03 Ingo Molnar
2008-03-27 20:31 ` Linus Torvalds
2008-03-27 20:48 ` Harvey Harrison
2008-03-27 20:55 ` Ingo Molnar
2008-03-27 21:01 ` Ingo Molnar
2008-03-27 21:08 ` Harvey Harrison
2008-03-27 20:50 ` Ingo Molnar
2008-03-27 21:24 ` Ingo Molnar
2008-03-26 21:41 Ingo Molnar
2008-03-21 16:20 Ingo Molnar
2008-03-11 16:12 Ingo Molnar
2008-03-07 15:50 Ingo Molnar
2008-03-04 16:59 Ingo Molnar
2008-03-03 13:18 Ingo Molnar
2008-01-01 17:21 Ingo Molnar
2007-12-21 0:46 Ingo Molnar
2007-12-19 23:04 Ingo Molnar
2007-12-04 16:41 Ingo Molnar
2007-12-02 19:12 Ingo Molnar
2007-12-03 16:23 ` Linus Torvalds
2007-12-03 16:38 ` Ingo Molnar
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20171014161636.tufvzolmzst64jaa@gmail.com \
--to=mingo@kernel.org \
--cc=a.p.zijlstra@chello.nl \
--cc=akpm@linux-foundation.org \
--cc=hpa@zytor.com \
--cc=linux-kernel@vger.kernel.org \
--cc=tglx@linutronix.de \
--cc=torvalds@linux-foundation.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).