From e7b52ffd45a6d834473f43b349e7d86593d763c7 Mon Sep 17 00:00:00 2001
From: Alex Shi <alex.shi@intel.com>
Date: Thu, 28 Jun 2012 09:02:17 +0800
Subject: x86/flush_tlb: try flush_tlb_single one by one in flush_tlb_range

x86 has no flush_tlb_range support in instruction level. Currently the
flush_tlb_range just implemented by flushing all page table. That is not
the best solution for all scenarios. In fact, if we just use 'invlpg' to
flush few lines from TLB, we can get the performance gain from later
remain TLB lines accessing.

But the 'invlpg' instruction costs much of time. Its execution time can
compete with cr3 rewriting, and even a bit more on SNB CPU.

So, on a 512 4KB TLB entries CPU, the balance points is at:
	(512 - X) * 100ns(assumed TLB refill cost) =
		X(TLB flush entries) * 100ns(assumed invlpg cost)

Here, X is 256, that is 1/2 of 512 entries.

But with the mysterious CPU pre-fetcher and page miss handler Unit, the
assumed TLB refill cost is far lower then 100ns in sequential access. And
2 HT siblings in one core makes the memory access more faster if they are
accessing the same memory. So, in the patch, I just do the change when
the target entries is less than 1/16 of whole active tlb entries.
Actually, I have no data support for the percentage '1/16', so any
suggestions are welcomed.

As to hugetlb, guess due to smaller page table, and smaller active TLB
entries, I didn't see benefit via my benchmark, so no optimizing now.

My micro benchmark show in ideal scenarios, the performance improves 70
percent in reading. And in worst scenario, the reading/writing
performance is similar with unpatched 3.4-rc4 kernel.

Here is the reading data on my 2P * 4cores *HT NHM EP machine, with THP
'always':

multi thread testing, '-t' paramter is thread number:
	       	        with patch   unpatched 3.4-rc4
./mprotect -t 1           14ns		24ns
./mprotect -t 2           13ns		22ns
./mprotect -t 4           12ns		19ns
./mprotect -t 8           14ns		16ns
./mprotect -t 16          28ns		26ns
./mprotect -t 32          54ns		51ns
./mprotect -t 128         200ns		199ns

Single process with sequencial flushing and memory accessing:

		       	with patch   unpatched 3.4-rc4
./mprotect		    7ns			11ns
./mprotect -p 4096  -l 8 -n 10240
			    21ns		21ns

[ hpa: http://lkml.kernel.org/r/1B4B44D9196EFF41AE41FDA404FC0A100BFF94@SHSMSX101.ccr.corp.intel.com
  has additional performance numbers. ]

Signed-off-by: Alex Shi <alex.shi@intel.com>
Link: http://lkml.kernel.org/r/1340845344-27557-3-git-send-email-alex.shi@intel.com
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
---
 arch/x86/mm/tlb.c | 97 ++++++++++++++++++++++++++++++++++++++++++++++---------
 1 file changed, 81 insertions(+), 16 deletions(-)

(limited to 'arch/x86/mm')

diff --git a/arch/x86/mm/tlb.c b/arch/x86/mm/tlb.c
index 5e57e113b72..3b91c981a27 100644
--- a/arch/x86/mm/tlb.c
+++ b/arch/x86/mm/tlb.c
@@ -41,7 +41,8 @@ DEFINE_PER_CPU_SHARED_ALIGNED(struct tlb_state, cpu_tlbstate)
 union smp_flush_state {
 	struct {
 		struct mm_struct *flush_mm;
-		unsigned long flush_va;
+		unsigned long flush_start;
+		unsigned long flush_end;
 		raw_spinlock_t tlbstate_lock;
 		DECLARE_BITMAP(flush_cpumask, NR_CPUS);
 	};
@@ -156,10 +157,19 @@ void smp_invalidate_interrupt(struct pt_regs *regs)
 
 	if (f->flush_mm == this_cpu_read(cpu_tlbstate.active_mm)) {
 		if (this_cpu_read(cpu_tlbstate.state) == TLBSTATE_OK) {
-			if (f->flush_va == TLB_FLUSH_ALL)
+			if (f->flush_end == TLB_FLUSH_ALL
+					|| !cpu_has_invlpg)
 				local_flush_tlb();
-			else
-				__flush_tlb_one(f->flush_va);
+			else if (!f->flush_end)
+				__flush_tlb_single(f->flush_start);
+			else {
+				unsigned long addr;
+				addr = f->flush_start;
+				while (addr < f->flush_end) {
+					__flush_tlb_single(addr);
+					addr += PAGE_SIZE;
+				}
+			}
 		} else
 			leave_mm(cpu);
 	}
@@ -172,7 +182,8 @@ out:
 }
 
 static void flush_tlb_others_ipi(const struct cpumask *cpumask,
-				 struct mm_struct *mm, unsigned long va)
+				 struct mm_struct *mm, unsigned long start,
+				 unsigned long end)
 {
 	unsigned int sender;
 	union smp_flush_state *f;
@@ -185,7 +196,8 @@ static void flush_tlb_others_ipi(const struct cpumask *cpumask,
 		raw_spin_lock(&f->tlbstate_lock);
 
 	f->flush_mm = mm;
-	f->flush_va = va;
+	f->flush_start = start;
+	f->flush_end = end;
 	if (cpumask_andnot(to_cpumask(f->flush_cpumask), cpumask, cpumask_of(smp_processor_id()))) {
 		/*
 		 * We have to send the IPI only to
@@ -199,24 +211,26 @@ static void flush_tlb_others_ipi(const struct cpumask *cpumask,
 	}
 
 	f->flush_mm = NULL;
-	f->flush_va = 0;
+	f->flush_start = 0;
+	f->flush_end = 0;
 	if (nr_cpu_ids > NUM_INVALIDATE_TLB_VECTORS)
 		raw_spin_unlock(&f->tlbstate_lock);
 }
 
 void native_flush_tlb_others(const struct cpumask *cpumask,
-			     struct mm_struct *mm, unsigned long va)
+				 struct mm_struct *mm, unsigned long start,
+				 unsigned long end)
 {
 	if (is_uv_system()) {
 		unsigned int cpu;
 
 		cpu = smp_processor_id();
-		cpumask = uv_flush_tlb_others(cpumask, mm, va, cpu);
+		cpumask = uv_flush_tlb_others(cpumask, mm, start, end, cpu);
 		if (cpumask)
-			flush_tlb_others_ipi(cpumask, mm, va);
+			flush_tlb_others_ipi(cpumask, mm, start, end);
 		return;
 	}
-	flush_tlb_others_ipi(cpumask, mm, va);
+	flush_tlb_others_ipi(cpumask, mm, start, end);
 }
 
 static void __cpuinit calculate_tlb_offset(void)
@@ -282,7 +296,7 @@ void flush_tlb_current_task(void)
 
 	local_flush_tlb();
 	if (cpumask_any_but(mm_cpumask(mm), smp_processor_id()) < nr_cpu_ids)
-		flush_tlb_others(mm_cpumask(mm), mm, TLB_FLUSH_ALL);
+		flush_tlb_others(mm_cpumask(mm), mm, 0UL, TLB_FLUSH_ALL);
 	preempt_enable();
 }
 
@@ -297,12 +311,63 @@ void flush_tlb_mm(struct mm_struct *mm)
 			leave_mm(smp_processor_id());
 	}
 	if (cpumask_any_but(mm_cpumask(mm), smp_processor_id()) < nr_cpu_ids)
-		flush_tlb_others(mm_cpumask(mm), mm, TLB_FLUSH_ALL);
+		flush_tlb_others(mm_cpumask(mm), mm, 0UL, TLB_FLUSH_ALL);
+
+	preempt_enable();
+}
+
+#define FLUSHALL_BAR	16
+
+void flush_tlb_range(struct vm_area_struct *vma,
+				   unsigned long start, unsigned long end)
+{
+	struct mm_struct *mm;
+
+	if (!cpu_has_invlpg || vma->vm_flags & VM_HUGETLB) {
+		flush_tlb_mm(vma->vm_mm);
+		return;
+	}
+
+	preempt_disable();
+	mm = vma->vm_mm;
+	if (current->active_mm == mm) {
+		if (current->mm) {
+			unsigned long addr, vmflag = vma->vm_flags;
+			unsigned act_entries, tlb_entries = 0;
+
+			if (vmflag & VM_EXEC)
+				tlb_entries = tlb_lli_4k[ENTRIES];
+			else
+				tlb_entries = tlb_lld_4k[ENTRIES];
+
+			act_entries = tlb_entries > mm->total_vm ?
+					mm->total_vm : tlb_entries;
 
+			if ((end - start)/PAGE_SIZE > act_entries/FLUSHALL_BAR)
+				local_flush_tlb();
+			else {
+				for (addr = start; addr < end;
+						addr += PAGE_SIZE)
+					__flush_tlb_single(addr);
+
+				if (cpumask_any_but(mm_cpumask(mm),
+					smp_processor_id()) < nr_cpu_ids)
+					flush_tlb_others(mm_cpumask(mm), mm,
+								start, end);
+				preempt_enable();
+				return;
+			}
+		} else {
+			leave_mm(smp_processor_id());
+		}
+	}
+	if (cpumask_any_but(mm_cpumask(mm), smp_processor_id()) < nr_cpu_ids)
+		flush_tlb_others(mm_cpumask(mm), mm, 0UL, TLB_FLUSH_ALL);
 	preempt_enable();
 }
 
-void flush_tlb_page(struct vm_area_struct *vma, unsigned long va)
+
+void flush_tlb_page(struct vm_area_struct *vma, unsigned long start)
 {
 	struct mm_struct *mm = vma->vm_mm;
 
@@ -310,13 +375,13 @@ void flush_tlb_page(struct vm_area_struct *vma, unsigned long va)
 
 	if (current->active_mm == mm) {
 		if (current->mm)
-			__flush_tlb_one(va);
+			__flush_tlb_one(start);
 		else
 			leave_mm(smp_processor_id());
 	}
 
 	if (cpumask_any_but(mm_cpumask(mm), smp_processor_id()) < nr_cpu_ids)
-		flush_tlb_others(mm_cpumask(mm), mm, va);
+		flush_tlb_others(mm_cpumask(mm), mm, start, 0UL);
 
 	preempt_enable();
 }
-- 
cgit v1.2.3-70-g09d2


From d8dfe60d6dcad5989c4558b753b98d657e2813c0 Mon Sep 17 00:00:00 2001
From: Alex Shi <alex.shi@intel.com>
Date: Thu, 28 Jun 2012 09:02:18 +0800
Subject: x86/tlb: fall back to flush all when meet a THP large page

We don't need to flush large pages by PAGE_SIZE step, that just waste
time. and actually, large page don't need 'invlpg' optimizing according
to our micro benchmark. So, just flush whole TLB is enough for them.

The following result is tested on a 2CPU * 4cores * 2HT NHM EP machine,
with THP 'always' setting.

Multi-thread testing, '-t' paramter is thread number:
                       without this patch 	with this patch
./mprotect -t 1         14ns                       13ns
./mprotect -t 2         13ns                       13ns
./mprotect -t 4         12ns                       11ns
./mprotect -t 8         14ns                       10ns
./mprotect -t 16        28ns                       28ns
./mprotect -t 32        54ns                       52ns
./mprotect -t 128       200ns                      200ns

Signed-off-by: Alex Shi <alex.shi@intel.com>
Link: http://lkml.kernel.org/r/1340845344-27557-4-git-send-email-alex.shi@intel.com
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
---
 arch/x86/mm/tlb.c | 34 ++++++++++++++++++++++++++++++++++
 1 file changed, 34 insertions(+)

(limited to 'arch/x86/mm')

diff --git a/arch/x86/mm/tlb.c b/arch/x86/mm/tlb.c
index 3b91c981a27..184a02a4d87 100644
--- a/arch/x86/mm/tlb.c
+++ b/arch/x86/mm/tlb.c
@@ -318,12 +318,42 @@ void flush_tlb_mm(struct mm_struct *mm)
 
 #define FLUSHALL_BAR	16
 
+#ifdef CONFIG_TRANSPARENT_HUGEPAGE
+static inline unsigned long has_large_page(struct mm_struct *mm,
+				 unsigned long start, unsigned long end)
+{
+	pgd_t *pgd;
+	pud_t *pud;
+	pmd_t *pmd;
+	unsigned long addr = ALIGN(start, HPAGE_SIZE);
+	for (; addr < end; addr += HPAGE_SIZE) {
+		pgd = pgd_offset(mm, addr);
+		if (likely(!pgd_none(*pgd))) {
+			pud = pud_offset(pgd, addr);
+			if (likely(!pud_none(*pud))) {
+				pmd = pmd_offset(pud, addr);
+				if (likely(!pmd_none(*pmd)))
+					if (pmd_large(*pmd))
+						return addr;
+			}
+		}
+	}
+	return 0;
+}
+#else
+static inline unsigned long has_large_page(struct mm_struct *mm,
+				 unsigned long start, unsigned long end)
+{
+	return 0;
+}
+#endif
 void flush_tlb_range(struct vm_area_struct *vma,
 				   unsigned long start, unsigned long end)
 {
 	struct mm_struct *mm;
 
 	if (!cpu_has_invlpg || vma->vm_flags & VM_HUGETLB) {
+flush_all:
 		flush_tlb_mm(vma->vm_mm);
 		return;
 	}
@@ -346,6 +376,10 @@ void flush_tlb_range(struct vm_area_struct *vma,
 			if ((end - start)/PAGE_SIZE > act_entries/FLUSHALL_BAR)
 				local_flush_tlb();
 			else {
+				if (has_large_page(mm, start, end)) {
+					preempt_enable();
+					goto flush_all;
+				}
 				for (addr = start; addr < end;
 						addr += PAGE_SIZE)
 					__flush_tlb_single(addr);
-- 
cgit v1.2.3-70-g09d2


From c4211f42d3e66875298a5e26a75109878c80f15b Mon Sep 17 00:00:00 2001
From: Alex Shi <alex.shi@intel.com>
Date: Thu, 28 Jun 2012 09:02:19 +0800
Subject: x86/tlb: add tlb_flushall_shift for specific CPU

Testing show different CPU type(micro architectures and NUMA mode) has
different balance points between the TLB flush all and multiple invlpg.
And there also has cases the tlb flush change has no any help.

This patch give a interface to let x86 vendor developers have a chance
to set different shift for different CPU type.

like some machine in my hands, balance points is 16 entries on
Romely-EP; while it is at 8 entries on Bloomfield NHM-EP; and is 256 on
IVB mobile CPU. but on model 15 core2 Xeon using invlpg has nothing
help.

For untested machine, do a conservative optimization, same as NHM CPU.

Signed-off-by: Alex Shi <alex.shi@intel.com>
Link: http://lkml.kernel.org/r/1340845344-27557-5-git-send-email-alex.shi@intel.com
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
---
 arch/x86/include/asm/processor.h |  2 ++
 arch/x86/kernel/cpu/common.c     | 14 ++++++++++++--
 arch/x86/kernel/cpu/intel.c      | 34 ++++++++++++++++++++++++++++++++++
 arch/x86/mm/tlb.c                |  7 +++----
 include/asm-generic/tlb.h        |  3 ++-
 5 files changed, 53 insertions(+), 7 deletions(-)

(limited to 'arch/x86/mm')

diff --git a/arch/x86/include/asm/processor.h b/arch/x86/include/asm/processor.h
index 39b2bd48dfb..d048cad9bca 100644
--- a/arch/x86/include/asm/processor.h
+++ b/arch/x86/include/asm/processor.h
@@ -72,6 +72,8 @@ extern u16 __read_mostly tlb_lli_4m[NR_INFO];
 extern u16 __read_mostly tlb_lld_4k[NR_INFO];
 extern u16 __read_mostly tlb_lld_2m[NR_INFO];
 extern u16 __read_mostly tlb_lld_4m[NR_INFO];
+extern s8  __read_mostly tlb_flushall_shift;
+
 /*
  *  CPU type and hardware bug flags. Kept separately for each CPU.
  *  Members of this structure are referenced in head.S, so think twice
diff --git a/arch/x86/kernel/cpu/common.c b/arch/x86/kernel/cpu/common.c
index b2016df0081..7595552600b 100644
--- a/arch/x86/kernel/cpu/common.c
+++ b/arch/x86/kernel/cpu/common.c
@@ -459,16 +459,26 @@ u16 __read_mostly tlb_lld_4k[NR_INFO];
 u16 __read_mostly tlb_lld_2m[NR_INFO];
 u16 __read_mostly tlb_lld_4m[NR_INFO];
 
+/*
+ * tlb_flushall_shift shows the balance point in replacing cr3 write
+ * with multiple 'invlpg'. It will do this replacement when
+ *   flush_tlb_lines <= active_lines/2^tlb_flushall_shift.
+ * If tlb_flushall_shift is -1, means the replacement will be disabled.
+ */
+s8  __read_mostly tlb_flushall_shift = -1;
+
 void __cpuinit cpu_detect_tlb(struct cpuinfo_x86 *c)
 {
 	if (this_cpu->c_detect_tlb)
 		this_cpu->c_detect_tlb(c);
 
 	printk(KERN_INFO "Last level iTLB entries: 4KB %d, 2MB %d, 4MB %d\n" \
-		"Last level dTLB entries: 4KB %d, 2MB %d, 4MB %d\n",
+		"Last level dTLB entries: 4KB %d, 2MB %d, 4MB %d\n"	     \
+		"tlb_flushall_shift is 0x%x\n",
 		tlb_lli_4k[ENTRIES], tlb_lli_2m[ENTRIES],
 		tlb_lli_4m[ENTRIES], tlb_lld_4k[ENTRIES],
-		tlb_lld_2m[ENTRIES], tlb_lld_4m[ENTRIES]);
+		tlb_lld_2m[ENTRIES], tlb_lld_4m[ENTRIES],
+		tlb_flushall_shift);
 }
 
 void __cpuinit detect_ht(struct cpuinfo_x86 *c)
diff --git a/arch/x86/kernel/cpu/intel.c b/arch/x86/kernel/cpu/intel.c
index ed0d512cf51..0a4ce2980a5 100644
--- a/arch/x86/kernel/cpu/intel.c
+++ b/arch/x86/kernel/cpu/intel.c
@@ -610,6 +610,39 @@ static void __cpuinit intel_tlb_lookup(const unsigned char desc)
 	}
 }
 
+static void __cpuinit intel_tlb_flushall_shift_set(struct cpuinfo_x86 *c)
+{
+	if (!cpu_has_invlpg) {
+		tlb_flushall_shift = -1;
+		return;
+	}
+	switch ((c->x86 << 8) + c->x86_model) {
+	case 0x60f: /* original 65 nm celeron/pentium/core2/xeon, "Merom"/"Conroe" */
+	case 0x616: /* single-core 65 nm celeron/core2solo "Merom-L"/"Conroe-L" */
+	case 0x617: /* current 45 nm celeron/core2/xeon "Penryn"/"Wolfdale" */
+	case 0x61d: /* six-core 45 nm xeon "Dunnington" */
+		tlb_flushall_shift = -1;
+		break;
+	case 0x61a: /* 45 nm nehalem, "Bloomfield" */
+	case 0x61e: /* 45 nm nehalem, "Lynnfield" */
+	case 0x625: /* 32 nm nehalem, "Clarkdale" */
+	case 0x62c: /* 32 nm nehalem, "Gulftown" */
+	case 0x62e: /* 45 nm nehalem-ex, "Beckton" */
+	case 0x62f: /* 32 nm Xeon E7 */
+		tlb_flushall_shift = 6;
+		break;
+	case 0x62a: /* SandyBridge */
+	case 0x62d: /* SandyBridge, "Romely-EP" */
+		tlb_flushall_shift = 5;
+		break;
+	case 0x63a: /* Ivybridge */
+		tlb_flushall_shift = 1;
+		break;
+	default:
+		tlb_flushall_shift = 6;
+	}
+}
+
 static void __cpuinit intel_detect_tlb(struct cpuinfo_x86 *c)
 {
 	int i, j, n;
@@ -630,6 +663,7 @@ static void __cpuinit intel_detect_tlb(struct cpuinfo_x86 *c)
 		for (j = 1 ; j < 16 ; j++)
 			intel_tlb_lookup(desc[j]);
 	}
+	intel_tlb_flushall_shift_set(c);
 }
 
 static const struct cpu_dev __cpuinitconst intel_cpu_dev = {
diff --git a/arch/x86/mm/tlb.c b/arch/x86/mm/tlb.c
index 184a02a4d87..2939f2f9edb 100644
--- a/arch/x86/mm/tlb.c
+++ b/arch/x86/mm/tlb.c
@@ -316,8 +316,6 @@ void flush_tlb_mm(struct mm_struct *mm)
 	preempt_enable();
 }
 
-#define FLUSHALL_BAR	16
-
 #ifdef CONFIG_TRANSPARENT_HUGEPAGE
 static inline unsigned long has_large_page(struct mm_struct *mm,
 				 unsigned long start, unsigned long end)
@@ -352,7 +350,7 @@ void flush_tlb_range(struct vm_area_struct *vma,
 {
 	struct mm_struct *mm;
 
-	if (!cpu_has_invlpg || vma->vm_flags & VM_HUGETLB) {
+	if (vma->vm_flags & VM_HUGETLB || tlb_flushall_shift == -1) {
 flush_all:
 		flush_tlb_mm(vma->vm_mm);
 		return;
@@ -373,7 +371,8 @@ flush_all:
 			act_entries = tlb_entries > mm->total_vm ?
 					mm->total_vm : tlb_entries;
 
-			if ((end - start)/PAGE_SIZE > act_entries/FLUSHALL_BAR)
+			if ((end - start) >> PAGE_SHIFT >
+					act_entries >> tlb_flushall_shift)
 				local_flush_tlb();
 			else {
 				if (has_large_page(mm, start, end)) {
diff --git a/include/asm-generic/tlb.h b/include/asm-generic/tlb.h
index f96a5b58a97..75e888b3cfd 100644
--- a/include/asm-generic/tlb.h
+++ b/include/asm-generic/tlb.h
@@ -113,7 +113,8 @@ static inline int tlb_fast_mode(struct mmu_gather *tlb)
 
 void tlb_gather_mmu(struct mmu_gather *tlb, struct mm_struct *mm, bool fullmm);
 void tlb_flush_mmu(struct mmu_gather *tlb);
-void tlb_finish_mmu(struct mmu_gather *tlb, unsigned long start, unsigned long end);
+void tlb_finish_mmu(struct mmu_gather *tlb, unsigned long start,
+							unsigned long end);
 int __tlb_remove_page(struct mmu_gather *tlb, struct page *page);
 
 /* tlb_remove_page
-- 
cgit v1.2.3-70-g09d2


From 3df3212f9722c7e45c723b9ea231a04ba4dbc47c Mon Sep 17 00:00:00 2001
From: Alex Shi <alex.shi@intel.com>
Date: Thu, 28 Jun 2012 09:02:20 +0800
Subject: x86/tlb: add tlb_flushall_shift knob into debugfs

kernel will replace cr3 rewrite with invlpg when
  tlb_flush_entries <= active_tlb_entries / 2^tlb_flushall_factor
if tlb_flushall_factor is -1, kernel won't do this replacement.

User can modify its value according to specific CPU/applications.

Thanks for Borislav providing the help message of
CONFIG_DEBUG_TLBFLUSH.

Signed-off-by: Alex Shi <alex.shi@intel.com>
Link: http://lkml.kernel.org/r/1340845344-27557-6-git-send-email-alex.shi@intel.com
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
---
 arch/x86/Kconfig.debug | 19 +++++++++++++++++++
 arch/x86/mm/tlb.c      | 51 ++++++++++++++++++++++++++++++++++++++++++++++++++
 2 files changed, 70 insertions(+)

(limited to 'arch/x86/mm')

diff --git a/arch/x86/Kconfig.debug b/arch/x86/Kconfig.debug
index e46c2147397..b322f124ee3 100644
--- a/arch/x86/Kconfig.debug
+++ b/arch/x86/Kconfig.debug
@@ -129,6 +129,25 @@ config DOUBLEFAULT
 	  option saves about 4k and might cause you much additional grey
 	  hair.
 
+config DEBUG_TLBFLUSH
+	bool "Set upper limit of TLB entries to flush one-by-one"
+	depends on DEBUG_KERNEL && (X86_64 || X86_INVLPG)
+	---help---
+
+	X86-only for now.
+
+	This option allows the user to tune the amount of TLB entries the
+	kernel flushes one-by-one instead of doing a full TLB flush. In
+	certain situations, the former is cheaper. This is controlled by the
+	tlb_flushall_shift knob under /sys/kernel/debug/x86. If you set it
+	to -1, the code flushes the whole TLB unconditionally. Otherwise,
+	for positive values of it, the kernel will use single TLB entry
+	invalidating instructions according to the following formula:
+
+	flush_entries <= active_tlb_entries / 2^tlb_flushall_shift
+
+	If in doubt, say "N".
+
 config IOMMU_DEBUG
 	bool "Enable IOMMU debugging"
 	depends on GART_IOMMU && DEBUG_KERNEL
diff --git a/arch/x86/mm/tlb.c b/arch/x86/mm/tlb.c
index 2939f2f9edb..5911f61e300 100644
--- a/arch/x86/mm/tlb.c
+++ b/arch/x86/mm/tlb.c
@@ -12,6 +12,7 @@
 #include <asm/cache.h>
 #include <asm/apic.h>
 #include <asm/uv/uv.h>
+#include <linux/debugfs.h>
 
 DEFINE_PER_CPU_SHARED_ALIGNED(struct tlb_state, cpu_tlbstate)
 			= { &init_mm, 0, };
@@ -430,3 +431,53 @@ void flush_tlb_all(void)
 {
 	on_each_cpu(do_flush_tlb_all, NULL, 1);
 }
+
+#ifdef CONFIG_DEBUG_TLBFLUSH
+static ssize_t tlbflush_read_file(struct file *file, char __user *user_buf,
+			     size_t count, loff_t *ppos)
+{
+	char buf[32];
+	unsigned int len;
+
+	len = sprintf(buf, "%hd\n", tlb_flushall_shift);
+	return simple_read_from_buffer(user_buf, count, ppos, buf, len);
+}
+
+static ssize_t tlbflush_write_file(struct file *file,
+		 const char __user *user_buf, size_t count, loff_t *ppos)
+{
+	char buf[32];
+	ssize_t len;
+	s8 shift;
+
+	len = min(count, sizeof(buf) - 1);
+	if (copy_from_user(buf, user_buf, len))
+		return -EFAULT;
+
+	buf[len] = '\0';
+	if (kstrtos8(buf, 0, &shift))
+		return -EINVAL;
+
+	if (shift > 64)
+		return -EINVAL;
+
+	tlb_flushall_shift = shift;
+	return count;
+}
+
+static const struct file_operations fops_tlbflush = {
+	.read = tlbflush_read_file,
+	.write = tlbflush_write_file,
+	.llseek = default_llseek,
+};
+
+static int __cpuinit create_tlb_flushall_shift(void)
+{
+	if (cpu_has_invlpg) {
+		debugfs_create_file("tlb_flushall_shift", S_IRUSR | S_IWUSR,
+			arch_debugfs_dir, NULL, &fops_tlbflush);
+	}
+	return 0;
+}
+late_initcall(create_tlb_flushall_shift);
+#endif
-- 
cgit v1.2.3-70-g09d2


From 611ae8e3f5204f7480b3b405993b3352cfa16662 Mon Sep 17 00:00:00 2001
From: Alex Shi <alex.shi@intel.com>
Date: Thu, 28 Jun 2012 09:02:22 +0800
Subject: x86/tlb: enable tlb flush range support for x86

Not every tlb_flush execution moment is really need to evacuate all
TLB entries, like in munmap, just few 'invlpg' is better for whole
process performance, since it leaves most of TLB entries for later
accessing.

This patch also rewrite flush_tlb_range for 2 purposes:
1, split it out to get flush_blt_mm_range function.
2, clean up to reduce line breaking, thanks for Borislav's input.

My micro benchmark 'mummap' http://lkml.org/lkml/2012/5/17/59
show that the random memory access on other CPU has 0~50% speed up
on a 2P * 4cores * HT NHM EP while do 'munmap'.

Thanks Yongjie's testing on this patch:
-------------
I used Linux 3.4-RC6 w/ and w/o his patches as Xen dom0 and guest
kernel.
After running two benchmarks in Xen HVM guest, I found his patches
brought about 1%~3% performance gain in 'kernel build' and 'netperf'
testing, though the performance gain was not very stable in 'kernel
build' testing.

Some detailed testing results are below.

Testing Environment:
	Hardware: Romley-EP platform
	Xen version: latest upstream
	Linux kernel: 3.4-RC6
	Guest vCPU number: 8
	NIC: Intel 82599 (10GB bandwidth)

In 'kernel build' testing in guest:
	Command line  |  performance gain
    make -j 4      |    3.81%
    make -j 8      |    0.37%
    make -j 16     |    -0.52%

In 'netperf' testing, we tested TCP_STREAM with default socket size
16384 byte as large packet and 64 byte as small packet.
I used several clients to add networking pressure, then 'netperf' server
automatically generated several threads to response them.
I also used large-size packet and small-size packet in the testing.
	Packet size  |  Thread number | performance gain
	16384 bytes  |      4       |   0.02%
	16384 bytes  |      8       |   2.21%
	16384 bytes  |      16      |   2.04%
	64 bytes     |      4       |   1.07%
	64 bytes     |      8       |   3.31%
	64 bytes     |      16      |   0.71%

Signed-off-by: Alex Shi <alex.shi@intel.com>
Link: http://lkml.kernel.org/r/1340845344-27557-8-git-send-email-alex.shi@intel.com
Tested-by: Ren, Yongjie <yongjie.ren@intel.com>
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
---
 arch/x86/include/asm/tlb.h      |   9 +++-
 arch/x86/include/asm/tlbflush.h |  17 ++++--
 arch/x86/mm/tlb.c               | 112 +++++++++++++++++-----------------------
 3 files changed, 68 insertions(+), 70 deletions(-)

(limited to 'arch/x86/mm')

diff --git a/arch/x86/include/asm/tlb.h b/arch/x86/include/asm/tlb.h
index 829215fef9e..4fef20773b8 100644
--- a/arch/x86/include/asm/tlb.h
+++ b/arch/x86/include/asm/tlb.h
@@ -4,7 +4,14 @@
 #define tlb_start_vma(tlb, vma) do { } while (0)
 #define tlb_end_vma(tlb, vma) do { } while (0)
 #define __tlb_remove_tlb_entry(tlb, ptep, address) do { } while (0)
-#define tlb_flush(tlb) flush_tlb_mm((tlb)->mm)
+
+#define tlb_flush(tlb)							\
+{									\
+	if (tlb->fullmm == 0)						\
+		flush_tlb_mm_range(tlb->mm, tlb->start, tlb->end, 0UL);	\
+	else								\
+		flush_tlb_mm_range(tlb->mm, 0UL, TLB_FLUSH_ALL, 0UL);	\
+}
 
 #include <asm-generic/tlb.h>
 
diff --git a/arch/x86/include/asm/tlbflush.h b/arch/x86/include/asm/tlbflush.h
index 33608d96d68..621b959e1db 100644
--- a/arch/x86/include/asm/tlbflush.h
+++ b/arch/x86/include/asm/tlbflush.h
@@ -105,6 +105,13 @@ static inline void flush_tlb_range(struct vm_area_struct *vma,
 		__flush_tlb();
 }
 
+static inline void flush_tlb_mm_range(struct vm_area_struct *vma,
+	   unsigned long start, unsigned long end, unsigned long vmflag)
+{
+	if (vma->vm_mm == current->active_mm)
+		__flush_tlb();
+}
+
 static inline void native_flush_tlb_others(const struct cpumask *cpumask,
 					   struct mm_struct *mm,
 					   unsigned long start,
@@ -122,12 +129,16 @@ static inline void reset_lazy_tlbstate(void)
 
 #define local_flush_tlb() __flush_tlb()
 
+#define flush_tlb_mm(mm)	flush_tlb_mm_range(mm, 0UL, TLB_FLUSH_ALL, 0UL)
+
+#define flush_tlb_range(vma, start, end)	\
+		flush_tlb_mm_range(vma->vm_mm, start, end, vma->vm_flags)
+
 extern void flush_tlb_all(void);
 extern void flush_tlb_current_task(void);
-extern void flush_tlb_mm(struct mm_struct *);
 extern void flush_tlb_page(struct vm_area_struct *, unsigned long);
-extern void flush_tlb_range(struct vm_area_struct *vma,
-				   unsigned long start, unsigned long end);
+extern void flush_tlb_mm_range(struct mm_struct *mm, unsigned long start,
+				unsigned long end, unsigned long vmflag);
 
 #define flush_tlb()	flush_tlb_current_task()
 
diff --git a/arch/x86/mm/tlb.c b/arch/x86/mm/tlb.c
index 5911f61e300..481737def84 100644
--- a/arch/x86/mm/tlb.c
+++ b/arch/x86/mm/tlb.c
@@ -301,23 +301,10 @@ void flush_tlb_current_task(void)
 	preempt_enable();
 }
 
-void flush_tlb_mm(struct mm_struct *mm)
-{
-	preempt_disable();
-
-	if (current->active_mm == mm) {
-		if (current->mm)
-			local_flush_tlb();
-		else
-			leave_mm(smp_processor_id());
-	}
-	if (cpumask_any_but(mm_cpumask(mm), smp_processor_id()) < nr_cpu_ids)
-		flush_tlb_others(mm_cpumask(mm), mm, 0UL, TLB_FLUSH_ALL);
-
-	preempt_enable();
-}
-
-#ifdef CONFIG_TRANSPARENT_HUGEPAGE
+/*
+ * It can find out the THP large page, or
+ * HUGETLB page in tlb_flush when THP disabled
+ */
 static inline unsigned long has_large_page(struct mm_struct *mm,
 				 unsigned long start, unsigned long end)
 {
@@ -339,68 +326,61 @@ static inline unsigned long has_large_page(struct mm_struct *mm,
 	}
 	return 0;
 }
-#else
-static inline unsigned long has_large_page(struct mm_struct *mm,
-				 unsigned long start, unsigned long end)
-{
-	return 0;
-}
-#endif
-void flush_tlb_range(struct vm_area_struct *vma,
-				   unsigned long start, unsigned long end)
-{
-	struct mm_struct *mm;
 
-	if (vma->vm_flags & VM_HUGETLB || tlb_flushall_shift == -1) {
-flush_all:
-		flush_tlb_mm(vma->vm_mm);
-		return;
-	}
+void flush_tlb_mm_range(struct mm_struct *mm, unsigned long start,
+				unsigned long end, unsigned long vmflag)
+{
+	unsigned long addr;
+	unsigned act_entries, tlb_entries = 0;
 
 	preempt_disable();
-	mm = vma->vm_mm;
-	if (current->active_mm == mm) {
-		if (current->mm) {
-			unsigned long addr, vmflag = vma->vm_flags;
-			unsigned act_entries, tlb_entries = 0;
+	if (current->active_mm != mm)
+		goto flush_all;
 
-			if (vmflag & VM_EXEC)
-				tlb_entries = tlb_lli_4k[ENTRIES];
-			else
-				tlb_entries = tlb_lld_4k[ENTRIES];
-
-			act_entries = tlb_entries > mm->total_vm ?
-					mm->total_vm : tlb_entries;
+	if (!current->mm) {
+		leave_mm(smp_processor_id());
+		goto flush_all;
+	}
 
-			if ((end - start) >> PAGE_SHIFT >
-					act_entries >> tlb_flushall_shift)
-				local_flush_tlb();
-			else {
-				if (has_large_page(mm, start, end)) {
-					preempt_enable();
-					goto flush_all;
-				}
-				for (addr = start; addr < end;
-						addr += PAGE_SIZE)
-					__flush_tlb_single(addr);
+	if (end == TLB_FLUSH_ALL || tlb_flushall_shift == -1
+					|| vmflag == VM_HUGETLB) {
+		local_flush_tlb();
+		goto flush_all;
+	}
 
-				if (cpumask_any_but(mm_cpumask(mm),
-					smp_processor_id()) < nr_cpu_ids)
-					flush_tlb_others(mm_cpumask(mm), mm,
-								start, end);
-				preempt_enable();
-				return;
-			}
-		} else {
-			leave_mm(smp_processor_id());
+	/* In modern CPU, last level tlb used for both data/ins */
+	if (vmflag & VM_EXEC)
+		tlb_entries = tlb_lli_4k[ENTRIES];
+	else
+		tlb_entries = tlb_lld_4k[ENTRIES];
+	/* Assume all of TLB entries was occupied by this task */
+	act_entries = mm->total_vm > tlb_entries ? tlb_entries : mm->total_vm;
+
+	/* tlb_flushall_shift is on balance point, details in commit log */
+	if ((end - start) >> PAGE_SHIFT > act_entries >> tlb_flushall_shift)
+		local_flush_tlb();
+	else {
+		if (has_large_page(mm, start, end)) {
+			local_flush_tlb();
+			goto flush_all;
 		}
+		/* flush range by one by one 'invlpg' */
+		for (addr = start; addr < end;	addr += PAGE_SIZE)
+			__flush_tlb_single(addr);
+
+		if (cpumask_any_but(mm_cpumask(mm),
+				smp_processor_id()) < nr_cpu_ids)
+			flush_tlb_others(mm_cpumask(mm), mm, start, end);
+		preempt_enable();
+		return;
 	}
+
+flush_all:
 	if (cpumask_any_but(mm_cpumask(mm), smp_processor_id()) < nr_cpu_ids)
 		flush_tlb_others(mm_cpumask(mm), mm, 0UL, TLB_FLUSH_ALL);
 	preempt_enable();
 }
 
-
 void flush_tlb_page(struct vm_area_struct *vma, unsigned long start)
 {
 	struct mm_struct *mm = vma->vm_mm;
-- 
cgit v1.2.3-70-g09d2


From 52aec3308db85f4e9f5c8b9f5dc4fbd0138c6fa4 Mon Sep 17 00:00:00 2001
From: Alex Shi <alex.shi@intel.com>
Date: Thu, 28 Jun 2012 09:02:23 +0800
Subject: x86/tlb: replace INVALIDATE_TLB_VECTOR by CALL_FUNCTION_VECTOR

There are 32 INVALIDATE_TLB_VECTOR now in kernel. That is quite big
amount of vector in IDT. But it is still not enough, since modern x86
sever has more cpu number. That still causes heavy lock contention
in TLB flushing.

The patch using generic smp call function to replace it. That saved 32
vector number in IDT, and resolved the lock contention in TLB
flushing on large system.

In the NHM EX machine 4P * 8cores * HT = 64 CPUs, hackbench pthread
has 3% performance increase.

Signed-off-by: Alex Shi <alex.shi@intel.com>
Link: http://lkml.kernel.org/r/1340845344-27557-9-git-send-email-alex.shi@intel.com
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
---
 arch/x86/include/asm/entry_arch.h  |   9 --
 arch/x86/include/asm/irq_vectors.h |  11 --
 arch/x86/kernel/entry_64.S         |  18 ---
 arch/x86/kernel/irqinit.c          |  73 -----------
 arch/x86/mm/tlb.c                  | 242 +++++++------------------------------
 5 files changed, 47 insertions(+), 306 deletions(-)

(limited to 'arch/x86/mm')

diff --git a/arch/x86/include/asm/entry_arch.h b/arch/x86/include/asm/entry_arch.h
index 0baa628e330..40afa0005c6 100644
--- a/arch/x86/include/asm/entry_arch.h
+++ b/arch/x86/include/asm/entry_arch.h
@@ -15,15 +15,6 @@ BUILD_INTERRUPT(call_function_interrupt,CALL_FUNCTION_VECTOR)
 BUILD_INTERRUPT(call_function_single_interrupt,CALL_FUNCTION_SINGLE_VECTOR)
 BUILD_INTERRUPT(irq_move_cleanup_interrupt,IRQ_MOVE_CLEANUP_VECTOR)
 BUILD_INTERRUPT(reboot_interrupt,REBOOT_VECTOR)
-
-.irp idx,0,1,2,3,4,5,6,7,8,9,10,11,12,13,14,15, \
-	16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31
-.if NUM_INVALIDATE_TLB_VECTORS > \idx
-BUILD_INTERRUPT3(invalidate_interrupt\idx,
-		 (INVALIDATE_TLB_VECTOR_START)+\idx,
-		 smp_invalidate_interrupt)
-.endif
-.endr
 #endif
 
 BUILD_INTERRUPT(x86_platform_ipi, X86_PLATFORM_IPI_VECTOR)
diff --git a/arch/x86/include/asm/irq_vectors.h b/arch/x86/include/asm/irq_vectors.h
index 4b4448761e8..1508e518c7e 100644
--- a/arch/x86/include/asm/irq_vectors.h
+++ b/arch/x86/include/asm/irq_vectors.h
@@ -119,17 +119,6 @@
  */
 #define LOCAL_TIMER_VECTOR		0xef
 
-/* up to 32 vectors used for spreading out TLB flushes: */
-#if NR_CPUS <= 32
-# define NUM_INVALIDATE_TLB_VECTORS	(NR_CPUS)
-#else
-# define NUM_INVALIDATE_TLB_VECTORS	(32)
-#endif
-
-#define INVALIDATE_TLB_VECTOR_END	(0xee)
-#define INVALIDATE_TLB_VECTOR_START	\
-	(INVALIDATE_TLB_VECTOR_END-NUM_INVALIDATE_TLB_VECTORS+1)
-
 #define NR_VECTORS			 256
 
 #define FPU_IRQ				  13
diff --git a/arch/x86/kernel/entry_64.S b/arch/x86/kernel/entry_64.S
index 7d65133b51b..bcf28e1ce1a 100644
--- a/arch/x86/kernel/entry_64.S
+++ b/arch/x86/kernel/entry_64.S
@@ -1048,24 +1048,6 @@ apicinterrupt LOCAL_TIMER_VECTOR \
 apicinterrupt X86_PLATFORM_IPI_VECTOR \
 	x86_platform_ipi smp_x86_platform_ipi
 
-#ifdef CONFIG_SMP
-	ALIGN
-	INTR_FRAME
-.irp idx,1,2,3,4,5,6,7,8,9,10,11,12,13,14,15, \
-	16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31
-.if NUM_INVALIDATE_TLB_VECTORS > \idx
-ENTRY(invalidate_interrupt\idx)
-	pushq_cfi $~(INVALIDATE_TLB_VECTOR_START+\idx)
-	jmp .Lcommon_invalidate_interrupt0
-	CFI_ADJUST_CFA_OFFSET -8
-END(invalidate_interrupt\idx)
-.endif
-.endr
-	CFI_ENDPROC
-apicinterrupt INVALIDATE_TLB_VECTOR_START, \
-	invalidate_interrupt0, smp_invalidate_interrupt
-#endif
-
 apicinterrupt THRESHOLD_APIC_VECTOR \
 	threshold_interrupt smp_threshold_interrupt
 apicinterrupt THERMAL_APIC_VECTOR \
diff --git a/arch/x86/kernel/irqinit.c b/arch/x86/kernel/irqinit.c
index 252981afd6c..6e03b0d6913 100644
--- a/arch/x86/kernel/irqinit.c
+++ b/arch/x86/kernel/irqinit.c
@@ -171,79 +171,6 @@ static void __init smp_intr_init(void)
 	 */
 	alloc_intr_gate(RESCHEDULE_VECTOR, reschedule_interrupt);
 
-	/* IPIs for invalidation */
-#define ALLOC_INVTLB_VEC(NR) \
-	alloc_intr_gate(INVALIDATE_TLB_VECTOR_START+NR, \
-		invalidate_interrupt##NR)
-
-	switch (NUM_INVALIDATE_TLB_VECTORS) {
-	default:
-		ALLOC_INVTLB_VEC(31);
-	case 31:
-		ALLOC_INVTLB_VEC(30);
-	case 30:
-		ALLOC_INVTLB_VEC(29);
-	case 29:
-		ALLOC_INVTLB_VEC(28);
-	case 28:
-		ALLOC_INVTLB_VEC(27);
-	case 27:
-		ALLOC_INVTLB_VEC(26);
-	case 26:
-		ALLOC_INVTLB_VEC(25);
-	case 25:
-		ALLOC_INVTLB_VEC(24);
-	case 24:
-		ALLOC_INVTLB_VEC(23);
-	case 23:
-		ALLOC_INVTLB_VEC(22);
-	case 22:
-		ALLOC_INVTLB_VEC(21);
-	case 21:
-		ALLOC_INVTLB_VEC(20);
-	case 20:
-		ALLOC_INVTLB_VEC(19);
-	case 19:
-		ALLOC_INVTLB_VEC(18);
-	case 18:
-		ALLOC_INVTLB_VEC(17);
-	case 17:
-		ALLOC_INVTLB_VEC(16);
-	case 16:
-		ALLOC_INVTLB_VEC(15);
-	case 15:
-		ALLOC_INVTLB_VEC(14);
-	case 14:
-		ALLOC_INVTLB_VEC(13);
-	case 13:
-		ALLOC_INVTLB_VEC(12);
-	case 12:
-		ALLOC_INVTLB_VEC(11);
-	case 11:
-		ALLOC_INVTLB_VEC(10);
-	case 10:
-		ALLOC_INVTLB_VEC(9);
-	case 9:
-		ALLOC_INVTLB_VEC(8);
-	case 8:
-		ALLOC_INVTLB_VEC(7);
-	case 7:
-		ALLOC_INVTLB_VEC(6);
-	case 6:
-		ALLOC_INVTLB_VEC(5);
-	case 5:
-		ALLOC_INVTLB_VEC(4);
-	case 4:
-		ALLOC_INVTLB_VEC(3);
-	case 3:
-		ALLOC_INVTLB_VEC(2);
-	case 2:
-		ALLOC_INVTLB_VEC(1);
-	case 1:
-		ALLOC_INVTLB_VEC(0);
-		break;
-	}
-
 	/* IPI for generic function call */
 	alloc_intr_gate(CALL_FUNCTION_VECTOR, call_function_interrupt);
 
diff --git a/arch/x86/mm/tlb.c b/arch/x86/mm/tlb.c
index 481737def84..2b5f506a765 100644
--- a/arch/x86/mm/tlb.c
+++ b/arch/x86/mm/tlb.c
@@ -28,34 +28,14 @@ DEFINE_PER_CPU_SHARED_ALIGNED(struct tlb_state, cpu_tlbstate)
  *
  *	More scalable flush, from Andi Kleen
  *
- *	To avoid global state use 8 different call vectors.
- *	Each CPU uses a specific vector to trigger flushes on other
- *	CPUs. Depending on the received vector the target CPUs look into
- *	the right array slot for the flush data.
- *
- *	With more than 8 CPUs they are hashed to the 8 available
- *	vectors. The limited global vector space forces us to this right now.
- *	In future when interrupts are split into per CPU domains this could be
- *	fixed, at the cost of triggering multiple IPIs in some cases.
+ *	Implement flush IPI by CALL_FUNCTION_VECTOR, Alex Shi
  */
 
-union smp_flush_state {
-	struct {
-		struct mm_struct *flush_mm;
-		unsigned long flush_start;
-		unsigned long flush_end;
-		raw_spinlock_t tlbstate_lock;
-		DECLARE_BITMAP(flush_cpumask, NR_CPUS);
-	};
-	char pad[INTERNODE_CACHE_BYTES];
-} ____cacheline_internodealigned_in_smp;
-
-/* State is put into the per CPU data section, but padded
-   to a full cache line because other CPUs can access it and we don't
-   want false sharing in the per cpu data segment. */
-static union smp_flush_state flush_state[NUM_INVALIDATE_TLB_VECTORS];
-
-static DEFINE_PER_CPU_READ_MOSTLY(int, tlb_vector_offset);
+struct flush_tlb_info {
+	struct mm_struct *flush_mm;
+	unsigned long flush_start;
+	unsigned long flush_end;
+};
 
 /*
  * We cannot call mmdrop() because we are in interrupt context,
@@ -74,28 +54,25 @@ void leave_mm(int cpu)
 EXPORT_SYMBOL_GPL(leave_mm);
 
 /*
- *
  * The flush IPI assumes that a thread switch happens in this order:
  * [cpu0: the cpu that switches]
  * 1) switch_mm() either 1a) or 1b)
  * 1a) thread switch to a different mm
- * 1a1) cpu_clear(cpu, old_mm->cpu_vm_mask);
- *	Stop ipi delivery for the old mm. This is not synchronized with
- *	the other cpus, but smp_invalidate_interrupt ignore flush ipis
- *	for the wrong mm, and in the worst case we perform a superfluous
- *	tlb flush.
- * 1a2) set cpu mmu_state to TLBSTATE_OK
- *	Now the smp_invalidate_interrupt won't call leave_mm if cpu0
- *	was in lazy tlb mode.
- * 1a3) update cpu active_mm
+ * 1a1) set cpu_tlbstate to TLBSTATE_OK
+ *	Now the tlb flush NMI handler flush_tlb_func won't call leave_mm
+ *	if cpu0 was in lazy tlb mode.
+ * 1a2) update cpu active_mm
  *	Now cpu0 accepts tlb flushes for the new mm.
- * 1a4) cpu_set(cpu, new_mm->cpu_vm_mask);
+ * 1a3) cpu_set(cpu, new_mm->cpu_vm_mask);
  *	Now the other cpus will send tlb flush ipis.
  * 1a4) change cr3.
+ * 1a5) cpu_clear(cpu, old_mm->cpu_vm_mask);
+ *	Stop ipi delivery for the old mm. This is not synchronized with
+ *	the other cpus, but flush_tlb_func ignore flush ipis for the wrong
+ *	mm, and in the worst case we perform a superfluous tlb flush.
  * 1b) thread switch without mm change
- *	cpu active_mm is correct, cpu0 already handles
- *	flush ipis.
- * 1b1) set cpu mmu_state to TLBSTATE_OK
+ *	cpu active_mm is correct, cpu0 already handles flush ipis.
+ * 1b1) set cpu_tlbstate to TLBSTATE_OK
  * 1b2) test_and_set the cpu bit in cpu_vm_mask.
  *	Atomically set the bit [other cpus will start sending flush ipis],
  *	and test the bit.
@@ -108,186 +85,61 @@ EXPORT_SYMBOL_GPL(leave_mm);
  *   runs in kernel space, the cpu could load tlb entries for user space
  *   pages.
  *
- * The good news is that cpu mmu_state is local to each cpu, no
+ * The good news is that cpu_tlbstate is local to each cpu, no
  * write/read ordering problems.
  */
 
 /*
- * TLB flush IPI:
- *
+ * TLB flush funcation:
  * 1) Flush the tlb entries if the cpu uses the mm that's being flushed.
  * 2) Leave the mm if we are in the lazy tlb mode.
- *
- * Interrupts are disabled.
- */
-
-/*
- * FIXME: use of asmlinkage is not consistent.  On x86_64 it's noop
- * but still used for documentation purpose but the usage is slightly
- * inconsistent.  On x86_32, asmlinkage is regparm(0) but interrupt
- * entry calls in with the first parameter in %eax.  Maybe define
- * intrlinkage?
  */
-#ifdef CONFIG_X86_64
-asmlinkage
-#endif
-void smp_invalidate_interrupt(struct pt_regs *regs)
+static void flush_tlb_func(void *info)
 {
-	unsigned int cpu;
-	unsigned int sender;
-	union smp_flush_state *f;
-
-	cpu = smp_processor_id();
-	/*
-	 * orig_rax contains the negated interrupt vector.
-	 * Use that to determine where the sender put the data.
-	 */
-	sender = ~regs->orig_ax - INVALIDATE_TLB_VECTOR_START;
-	f = &flush_state[sender];
-
-	if (!cpumask_test_cpu(cpu, to_cpumask(f->flush_cpumask)))
-		goto out;
-		/*
-		 * This was a BUG() but until someone can quote me the
-		 * line from the intel manual that guarantees an IPI to
-		 * multiple CPUs is retried _only_ on the erroring CPUs
-		 * its staying as a return
-		 *
-		 * BUG();
-		 */
-
-	if (f->flush_mm == this_cpu_read(cpu_tlbstate.active_mm)) {
-		if (this_cpu_read(cpu_tlbstate.state) == TLBSTATE_OK) {
-			if (f->flush_end == TLB_FLUSH_ALL
-					|| !cpu_has_invlpg)
-				local_flush_tlb();
-			else if (!f->flush_end)
-				__flush_tlb_single(f->flush_start);
-			else {
-				unsigned long addr;
-				addr = f->flush_start;
-				while (addr < f->flush_end) {
-					__flush_tlb_single(addr);
-					addr += PAGE_SIZE;
-				}
-			}
-		} else
-			leave_mm(cpu);
-	}
-out:
-	ack_APIC_irq();
-	smp_mb__before_clear_bit();
-	cpumask_clear_cpu(cpu, to_cpumask(f->flush_cpumask));
-	smp_mb__after_clear_bit();
-	inc_irq_stat(irq_tlb_count);
-}
+	struct flush_tlb_info *f = info;
 
-static void flush_tlb_others_ipi(const struct cpumask *cpumask,
-				 struct mm_struct *mm, unsigned long start,
-				 unsigned long end)
-{
-	unsigned int sender;
-	union smp_flush_state *f;
-
-	/* Caller has disabled preemption */
-	sender = this_cpu_read(tlb_vector_offset);
-	f = &flush_state[sender];
-
-	if (nr_cpu_ids > NUM_INVALIDATE_TLB_VECTORS)
-		raw_spin_lock(&f->tlbstate_lock);
-
-	f->flush_mm = mm;
-	f->flush_start = start;
-	f->flush_end = end;
-	if (cpumask_andnot(to_cpumask(f->flush_cpumask), cpumask, cpumask_of(smp_processor_id()))) {
-		/*
-		 * We have to send the IPI only to
-		 * CPUs affected.
-		 */
-		apic->send_IPI_mask(to_cpumask(f->flush_cpumask),
-			      INVALIDATE_TLB_VECTOR_START + sender);
-
-		while (!cpumask_empty(to_cpumask(f->flush_cpumask)))
-			cpu_relax();
-	}
+	if (f->flush_mm != this_cpu_read(cpu_tlbstate.active_mm))
+		return;
+
+	if (this_cpu_read(cpu_tlbstate.state) == TLBSTATE_OK) {
+		if (f->flush_end == TLB_FLUSH_ALL || !cpu_has_invlpg)
+			local_flush_tlb();
+		else if (!f->flush_end)
+			__flush_tlb_single(f->flush_start);
+		else {
+			unsigned long addr;
+			addr = f->flush_start;
+			while (addr < f->flush_end) {
+				__flush_tlb_single(addr);
+				addr += PAGE_SIZE;
+			}
+		}
+	} else
+		leave_mm(smp_processor_id());
 
-	f->flush_mm = NULL;
-	f->flush_start = 0;
-	f->flush_end = 0;
-	if (nr_cpu_ids > NUM_INVALIDATE_TLB_VECTORS)
-		raw_spin_unlock(&f->tlbstate_lock);
 }
 
 void native_flush_tlb_others(const struct cpumask *cpumask,
 				 struct mm_struct *mm, unsigned long start,
 				 unsigned long end)
 {
+	struct flush_tlb_info info;
+	info.flush_mm = mm;
+	info.flush_start = start;
+	info.flush_end = end;
+
 	if (is_uv_system()) {
 		unsigned int cpu;
 
 		cpu = smp_processor_id();
 		cpumask = uv_flush_tlb_others(cpumask, mm, start, end, cpu);
 		if (cpumask)
-			flush_tlb_others_ipi(cpumask, mm, start, end);
+			smp_call_function_many(cpumask, flush_tlb_func,
+								&info, 1);
 		return;
 	}
-	flush_tlb_others_ipi(cpumask, mm, start, end);
-}
-
-static void __cpuinit calculate_tlb_offset(void)
-{
-	int cpu, node, nr_node_vecs, idx = 0;
-	/*
-	 * we are changing tlb_vector_offset for each CPU in runtime, but this
-	 * will not cause inconsistency, as the write is atomic under X86. we
-	 * might see more lock contentions in a short time, but after all CPU's
-	 * tlb_vector_offset are changed, everything should go normal
-	 *
-	 * Note: if NUM_INVALIDATE_TLB_VECTORS % nr_online_nodes !=0, we might
-	 * waste some vectors.
-	 **/
-	if (nr_online_nodes > NUM_INVALIDATE_TLB_VECTORS)
-		nr_node_vecs = 1;
-	else
-		nr_node_vecs = NUM_INVALIDATE_TLB_VECTORS/nr_online_nodes;
-
-	for_each_online_node(node) {
-		int node_offset = (idx % NUM_INVALIDATE_TLB_VECTORS) *
-			nr_node_vecs;
-		int cpu_offset = 0;
-		for_each_cpu(cpu, cpumask_of_node(node)) {
-			per_cpu(tlb_vector_offset, cpu) = node_offset +
-				cpu_offset;
-			cpu_offset++;
-			cpu_offset = cpu_offset % nr_node_vecs;
-		}
-		idx++;
-	}
-}
-
-static int __cpuinit tlb_cpuhp_notify(struct notifier_block *n,
-		unsigned long action, void *hcpu)
-{
-	switch (action & 0xf) {
-	case CPU_ONLINE:
-	case CPU_DEAD:
-		calculate_tlb_offset();
-	}
-	return NOTIFY_OK;
-}
-
-static int __cpuinit init_smp_flush(void)
-{
-	int i;
-
-	for (i = 0; i < ARRAY_SIZE(flush_state); i++)
-		raw_spin_lock_init(&flush_state[i].tlbstate_lock);
-
-	calculate_tlb_offset();
-	hotcpu_notifier(tlb_cpuhp_notify, 0);
-	return 0;
+	smp_call_function_many(cpumask, flush_tlb_func, &info, 1);
 }
-core_initcall(init_smp_flush);
 
 void flush_tlb_current_task(void)
 {
-- 
cgit v1.2.3-70-g09d2


From effee4b9b3b0aa5770bcd98de5f672b05b27703c Mon Sep 17 00:00:00 2001
From: Alex Shi <alex.shi@intel.com>
Date: Thu, 28 Jun 2012 09:02:24 +0800
Subject: x86/tlb: do flush_tlb_kernel_range by 'invlpg'

This patch do flush_tlb_kernel_range by 'invlpg'. The performance pay
and gain was analyzed in previous patch
(x86/flush_tlb: try flush_tlb_single one by one in flush_tlb_range).

In the testing: http://lkml.org/lkml/2012/6/21/10

The pay is mostly covered by long kernel path, but the gain is still
quite clear, memory access in user APP can increase 30+% when kernel
execute this funtion.

Signed-off-by: Alex Shi <alex.shi@intel.com>
Link: http://lkml.kernel.org/r/1340845344-27557-10-git-send-email-alex.shi@intel.com
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
---
 arch/x86/include/asm/tlbflush.h | 13 +++++++------
 arch/x86/mm/tlb.c               | 30 ++++++++++++++++++++++++++++++
 2 files changed, 37 insertions(+), 6 deletions(-)

(limited to 'arch/x86/mm')

diff --git a/arch/x86/include/asm/tlbflush.h b/arch/x86/include/asm/tlbflush.h
index 621b959e1db..b5a27bd7766 100644
--- a/arch/x86/include/asm/tlbflush.h
+++ b/arch/x86/include/asm/tlbflush.h
@@ -123,6 +123,12 @@ static inline void reset_lazy_tlbstate(void)
 {
 }
 
+static inline void flush_tlb_kernel_range(unsigned long start,
+					  unsigned long end)
+{
+	flush_tlb_all();
+}
+
 #else  /* SMP */
 
 #include <asm/smp.h>
@@ -139,6 +145,7 @@ extern void flush_tlb_current_task(void);
 extern void flush_tlb_page(struct vm_area_struct *, unsigned long);
 extern void flush_tlb_mm_range(struct mm_struct *mm, unsigned long start,
 				unsigned long end, unsigned long vmflag);
+extern void flush_tlb_kernel_range(unsigned long start, unsigned long end);
 
 #define flush_tlb()	flush_tlb_current_task()
 
@@ -168,10 +175,4 @@ static inline void reset_lazy_tlbstate(void)
 	native_flush_tlb_others(mask, mm, start, end)
 #endif
 
-static inline void flush_tlb_kernel_range(unsigned long start,
-					  unsigned long end)
-{
-	flush_tlb_all();
-}
-
 #endif /* _ASM_X86_TLBFLUSH_H */
diff --git a/arch/x86/mm/tlb.c b/arch/x86/mm/tlb.c
index 2b5f506a765..613cd83e8c0 100644
--- a/arch/x86/mm/tlb.c
+++ b/arch/x86/mm/tlb.c
@@ -264,6 +264,36 @@ void flush_tlb_all(void)
 	on_each_cpu(do_flush_tlb_all, NULL, 1);
 }
 
+static void do_kernel_range_flush(void *info)
+{
+	struct flush_tlb_info *f = info;
+	unsigned long addr;
+
+	/* flush range by one by one 'invlpg' */
+	for (addr = f->flush_start; addr < f->flush_end; addr += PAGE_SIZE)
+		__flush_tlb_single(addr);
+}
+
+void flush_tlb_kernel_range(unsigned long start, unsigned long end)
+{
+	unsigned act_entries;
+	struct flush_tlb_info info;
+
+	/* In modern CPU, last level tlb used for both data/ins */
+	act_entries = tlb_lld_4k[ENTRIES];
+
+	/* Balance as user space task's flush, a bit conservative */
+	if (end == TLB_FLUSH_ALL || tlb_flushall_shift == -1 ||
+		(end - start) >> PAGE_SHIFT > act_entries >> tlb_flushall_shift)
+
+		on_each_cpu(do_flush_tlb_all, NULL, 1);
+	else {
+		info.flush_start = start;
+		info.flush_end = end;
+		on_each_cpu(do_kernel_range_flush, &info, 1);
+	}
+}
+
 #ifdef CONFIG_DEBUG_TLBFLUSH
 static ssize_t tlbflush_read_file(struct file *file, char __user *user_buf,
 			     size_t count, loff_t *ppos)
-- 
cgit v1.2.3-70-g09d2