linux/arch/x86/mm/tlb.c

462 lines
12 KiB
C
Raw Normal View History

#include <linux/init.h>
#include <linux/mm.h>
#include <linux/spinlock.h>
#include <linux/smp.h>
#include <linux/interrupt.h>
#include <linux/module.h>
x86: Spread tlb flush vector between nodes Currently flush tlb vector allocation is based on below equation: sender = smp_processor_id() % 8 This isn't optimal, CPUs from different node can have the same vector, this causes a lot of lock contention. Instead, we can assign the same vectors to CPUs from the same node, while different node has different vectors. This has below advantages: a. if there is lock contention, the lock contention is between CPUs from one node. This should be much cheaper than the contention between nodes. b. completely avoid lock contention between nodes. This especially benefits kswapd, which is the biggest user of tlb flush, since kswapd sets its affinity to specific node. In my test, this could reduce > 20% CPU overhead in extreme case.The test machine has 4 nodes and each node has 16 CPUs. I then bind each node's kswapd to the first CPU of the node. I run a workload with 4 sequential mmap file read thread. The files are empty sparse file. This workload will trigger a lot of page reclaim and tlbflush. The kswapd bind is to easy trigger the extreme tlb flush lock contention because otherwise kswapd keeps migrating between CPUs of a node and I can't get stable result. Sure in real workload, we can't always see so big tlb flush lock contention, but it's possible. [ hpa: folded in fix from Eric Dumazet to use this_cpu_read() ] Signed-off-by: Shaohua Li <shaohua.li@intel.com> LKML-Reference: <1287544023.4571.8.camel@sli10-conroe.sh.intel.com> Cc: Eric Dumazet <eric.dumazet@gmail.com> Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
2010-10-20 11:07:03 +08:00
#include <linux/cpu.h>
#include <asm/tlbflush.h>
#include <asm/mmu_context.h>
#include <asm/cache.h>
#include <asm/apic.h>
#include <asm/uv/uv.h>
#include <linux/debugfs.h>
/*
* Smarter SMP flushing macros.
* c/o Linus Torvalds.
*
* These mean you can really definitely utterly forget about
* writing to user space from interrupts. (Its not allowed anyway).
*
* Optimizations Manfred Spraul <manfred@colorfullife.com>
*
* More scalable flush, from Andi Kleen
*
* Implement flush IPI by CALL_FUNCTION_VECTOR, Alex Shi
*/
#ifdef CONFIG_SMP
struct flush_tlb_info {
struct mm_struct *flush_mm;
unsigned long flush_start;
unsigned long flush_end;
};
x86: Spread tlb flush vector between nodes Currently flush tlb vector allocation is based on below equation: sender = smp_processor_id() % 8 This isn't optimal, CPUs from different node can have the same vector, this causes a lot of lock contention. Instead, we can assign the same vectors to CPUs from the same node, while different node has different vectors. This has below advantages: a. if there is lock contention, the lock contention is between CPUs from one node. This should be much cheaper than the contention between nodes. b. completely avoid lock contention between nodes. This especially benefits kswapd, which is the biggest user of tlb flush, since kswapd sets its affinity to specific node. In my test, this could reduce > 20% CPU overhead in extreme case.The test machine has 4 nodes and each node has 16 CPUs. I then bind each node's kswapd to the first CPU of the node. I run a workload with 4 sequential mmap file read thread. The files are empty sparse file. This workload will trigger a lot of page reclaim and tlbflush. The kswapd bind is to easy trigger the extreme tlb flush lock contention because otherwise kswapd keeps migrating between CPUs of a node and I can't get stable result. Sure in real workload, we can't always see so big tlb flush lock contention, but it's possible. [ hpa: folded in fix from Eric Dumazet to use this_cpu_read() ] Signed-off-by: Shaohua Li <shaohua.li@intel.com> LKML-Reference: <1287544023.4571.8.camel@sli10-conroe.sh.intel.com> Cc: Eric Dumazet <eric.dumazet@gmail.com> Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
2010-10-20 11:07:03 +08:00
/*
* We cannot call mmdrop() because we are in interrupt context,
* instead update mm->cpu_vm_mask.
*/
void leave_mm(int cpu)
{
struct mm_struct *active_mm = this_cpu_read(cpu_tlbstate.active_mm);
if (this_cpu_read(cpu_tlbstate.state) == TLBSTATE_OK)
BUG();
if (cpumask_test_cpu(cpu, mm_cpumask(active_mm))) {
cpumask_clear_cpu(cpu, mm_cpumask(active_mm));
load_cr3(swapper_pg_dir);
/*
* This gets called in the idle path where RCU
* functions differently. Tracing normally
* uses RCU, so we have to call the tracepoint
* specially here.
*/
trace_tlb_flush_rcuidle(TLB_FLUSH_ON_TASK_SWITCH, TLB_FLUSH_ALL);
}
}
EXPORT_SYMBOL_GPL(leave_mm);
#endif /* CONFIG_SMP */
void switch_mm(struct mm_struct *prev, struct mm_struct *next,
struct task_struct *tsk)
{
unsigned cpu = smp_processor_id();
if (likely(prev != next)) {
#ifdef CONFIG_SMP
this_cpu_write(cpu_tlbstate.state, TLBSTATE_OK);
this_cpu_write(cpu_tlbstate.active_mm, next);
#endif
cpumask_set_cpu(cpu, mm_cpumask(next));
/*
* Re-load page tables.
*
* This logic has an ordering constraint:
*
* CPU 0: Write to a PTE for 'next'
* CPU 0: load bit 1 in mm_cpumask. if nonzero, send IPI.
* CPU 1: set bit 1 in next's mm_cpumask
* CPU 1: load from the PTE that CPU 0 writes (implicit)
*
* We need to prevent an outcome in which CPU 1 observes
* the new PTE value and CPU 0 observes bit 1 clear in
* mm_cpumask. (If that occurs, then the IPI will never
* be sent, and CPU 0's TLB will contain a stale entry.)
*
* The bad outcome can occur if either CPU's load is
* reordered before that CPU's store, so both CPUs must
* execute full barriers to prevent this from happening.
*
* Thus, switch_mm needs a full barrier between the
* store to mm_cpumask and any operation that could load
* from next->pgd. TLB fills are special and can happen
* due to instruction fetches or for no reason at all,
* and neither LOCK nor MFENCE orders them.
* Fortunately, load_cr3() is serializing and gives the
* ordering guarantee we need.
*
*/
load_cr3(next->pgd);
trace_tlb_flush(TLB_FLUSH_ON_TASK_SWITCH, TLB_FLUSH_ALL);
/* Stop flush ipis for the previous mm */
cpumask_clear_cpu(cpu, mm_cpumask(prev));
/* Load per-mm CR4 state */
load_mm_cr4(next);
#ifdef CONFIG_MODIFY_LDT_SYSCALL
/*
* Load the LDT, if the LDT is different.
*
* It's possible that prev->context.ldt doesn't match
* the LDT register. This can happen if leave_mm(prev)
* was called and then modify_ldt changed
* prev->context.ldt but suppressed an IPI to this CPU.
* In this case, prev->context.ldt != NULL, because we
* never set context.ldt to NULL while the mm still
* exists. That means that next->context.ldt !=
* prev->context.ldt, because mms never share an LDT.
*/
if (unlikely(prev->context.ldt != next->context.ldt))
load_mm_ldt(next);
#endif
}
#ifdef CONFIG_SMP
else {
this_cpu_write(cpu_tlbstate.state, TLBSTATE_OK);
BUG_ON(this_cpu_read(cpu_tlbstate.active_mm) != next);
if (!cpumask_test_cpu(cpu, mm_cpumask(next))) {
/*
* On established mms, the mm_cpumask is only changed
* from irq context, from ptep_clear_flush() while in
* lazy tlb mode, and here. Irqs are blocked during
* schedule, protecting us from simultaneous changes.
*/
cpumask_set_cpu(cpu, mm_cpumask(next));
/*
* We were in lazy tlb mode and leave_mm disabled
* tlb flush IPI delivery. We must reload CR3
* to make sure to use no freed page tables.
*
* As above, load_cr3() is serializing and orders TLB
* fills with respect to the mm_cpumask write.
*/
load_cr3(next->pgd);
trace_tlb_flush(TLB_FLUSH_ON_TASK_SWITCH, TLB_FLUSH_ALL);
load_mm_cr4(next);
load_mm_ldt(next);
}
}
#endif
}
#ifdef CONFIG_SMP
/*
* 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) 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.
* 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_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.
* 1b3) if the bit was 0: leave_mm was called, flush the tlb.
* 2) switch %%esp, ie current
*
* The interrupt must handle 2 special cases:
* - cr3 is changed before %%esp, ie. it cannot use current->{active_,}mm.
* - the cpu performs speculative tlb reads, i.e. even if the cpu only
* runs in kernel space, the cpu could load tlb entries for user space
* pages.
*
* The good news is that cpu_tlbstate is local to each cpu, no
* write/read ordering problems.
*/
/*
* 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.
*/
static void flush_tlb_func(void *info)
{
struct flush_tlb_info *f = info;
inc_irq_stat(irq_tlb_count);
if (f->flush_mm && f->flush_mm != this_cpu_read(cpu_tlbstate.active_mm))
return;
count_vm_tlb_event(NR_TLB_REMOTE_FLUSH_RECEIVED);
if (this_cpu_read(cpu_tlbstate.state) == TLBSTATE_OK) {
if (f->flush_end == TLB_FLUSH_ALL) {
local_flush_tlb();
trace_tlb_flush(TLB_REMOTE_SHOOTDOWN, TLB_FLUSH_ALL);
} else {
unsigned long addr;
unsigned long nr_pages =
(f->flush_end - f->flush_start) / PAGE_SIZE;
addr = f->flush_start;
while (addr < f->flush_end) {
__flush_tlb_single(addr);
addr += PAGE_SIZE;
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>
2012-06-28 09:02:17 +08:00
}
trace_tlb_flush(TLB_REMOTE_SHOOTDOWN, nr_pages);
}
} else
leave_mm(smp_processor_id());
}
void native_flush_tlb_others(const struct cpumask *cpumask,
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>
2012-06-28 09:02:17 +08:00
struct mm_struct *mm, unsigned long start,
unsigned long end)
{
struct flush_tlb_info info;
if (end == 0)
end = start + PAGE_SIZE;
info.flush_mm = mm;
info.flush_start = start;
info.flush_end = end;
count_vm_tlb_event(NR_TLB_REMOTE_FLUSH);
if (end == TLB_FLUSH_ALL)
trace_tlb_flush(TLB_REMOTE_SEND_IPI, TLB_FLUSH_ALL);
else
trace_tlb_flush(TLB_REMOTE_SEND_IPI,
(end - start) >> PAGE_SHIFT);
if (is_uv_system()) {
unsigned int cpu;
cpu = smp_processor_id();
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>
2012-06-28 09:02:17 +08:00
cpumask = uv_flush_tlb_others(cpumask, mm, start, end, cpu);
if (cpumask)
smp_call_function_many(cpumask, flush_tlb_func,
&info, 1);
return;
}
smp_call_function_many(cpumask, flush_tlb_func, &info, 1);
}
void flush_tlb_current_task(void)
{
struct mm_struct *mm = current->mm;
preempt_disable();
count_vm_tlb_event(NR_TLB_LOCAL_FLUSH_ALL);
/* This is an implicit full barrier that synchronizes with switch_mm. */
local_flush_tlb();
trace_tlb_flush(TLB_LOCAL_SHOOTDOWN, TLB_FLUSH_ALL);
if (cpumask_any_but(mm_cpumask(mm), smp_processor_id()) < nr_cpu_ids)
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>
2012-06-28 09:02:17 +08:00
flush_tlb_others(mm_cpumask(mm), mm, 0UL, TLB_FLUSH_ALL);
preempt_enable();
}
x86/mm: Set TLB flush tunable to sane value (33) This has been run through Intel's LKP tests across a wide range of modern sytems and workloads and it wasn't shown to make a measurable performance difference positive or negative. Now that we have some shiny new tracepoints, we can actually figure out what the heck is going on. During a kernel compile, 60% of the flush_tlb_mm_range() calls are for a single page. It breaks down like this: size percent percent<= V V V GLOBAL: 2.20% 2.20% avg cycles: 2283 1: 56.92% 59.12% avg cycles: 1276 2: 13.78% 72.90% avg cycles: 1505 3: 8.26% 81.16% avg cycles: 1880 4: 7.41% 88.58% avg cycles: 2447 5: 1.73% 90.31% avg cycles: 2358 6: 1.32% 91.63% avg cycles: 2563 7: 1.14% 92.77% avg cycles: 2862 8: 0.62% 93.39% avg cycles: 3542 9: 0.08% 93.47% avg cycles: 3289 10: 0.43% 93.90% avg cycles: 3570 11: 0.20% 94.10% avg cycles: 3767 12: 0.08% 94.18% avg cycles: 3996 13: 0.03% 94.20% avg cycles: 4077 14: 0.02% 94.23% avg cycles: 4836 15: 0.04% 94.26% avg cycles: 5699 16: 0.06% 94.32% avg cycles: 5041 17: 0.57% 94.89% avg cycles: 5473 18: 0.02% 94.91% avg cycles: 5396 19: 0.03% 94.95% avg cycles: 5296 20: 0.02% 94.96% avg cycles: 6749 21: 0.18% 95.14% avg cycles: 6225 22: 0.01% 95.15% avg cycles: 6393 23: 0.01% 95.16% avg cycles: 6861 24: 0.12% 95.28% avg cycles: 6912 25: 0.05% 95.32% avg cycles: 7190 26: 0.01% 95.33% avg cycles: 7793 27: 0.01% 95.34% avg cycles: 7833 28: 0.01% 95.35% avg cycles: 8253 29: 0.08% 95.42% avg cycles: 8024 30: 0.03% 95.45% avg cycles: 9670 31: 0.01% 95.46% avg cycles: 8949 32: 0.01% 95.46% avg cycles: 9350 33: 3.11% 98.57% avg cycles: 8534 34: 0.02% 98.60% avg cycles: 10977 35: 0.02% 98.62% avg cycles: 11400 We get in to dimishing returns pretty quickly. On pre-IvyBridge CPUs, we used to set the limit at 8 pages, and it was set at 128 on IvyBrige. That 128 number looks pretty silly considering that less than 0.5% of the flushes are that large. The previous code tried to size this number based on the size of the TLB. Good idea, but it's error-prone, needs maintenance (which it didn't get up to now), and probably would not matter in practice much. Settting it to 33 means that we cover the mallopt M_TRIM_THRESHOLD, which is the most universally common size to do flushes. That's the short version. Here's the long one for why I chose 33: 1. These numbers have a constant bias in the timestamps from the tracing. Probably counts for a couple hundred cycles in each of these tests, but it should be fairly _even_ across all of them. The smallest delta between the tracepoints I have ever seen is 335 cycles. This is one reason the cycles/page cost goes down in general as the flushes get larger. The true cost is nearer to 100 cycles. 2. A full flush is more expensive than a single invlpg, but not by much (single percentages). 3. A dtlb miss is 17.1ns (~45 cycles) and a itlb miss is 13.0ns (~34 cycles). At those rates, refilling the 512-entry dTLB takes 22,000 cycles. 4. 22,000 cycles is approximately the equivalent of doing 85 invlpg operations. But, the odds are that the TLB can actually be filled up faster than that because TLB misses that are close in time also tend to leverage the same caches. 6. ~98% of flushes are <=33 pages. There are a lot of flushes of 33 pages, probably because libc's M_TRIM_THRESHOLD is set to 128k (32 pages) 7. I've found no consistent data to support changing the IvyBridge vs. SandyBridge tunable by a factor of 16 I used the performance counters on this hardware (IvyBridge i5-3320M) to figure out the tlb miss costs: ocperf.py stat -e dtlb_load_misses.walk_duration,dtlb_load_misses.walk_completed,dtlb_store_misses.walk_duration,dtlb_store_misses.walk_completed,itlb_misses.walk_duration,itlb_misses.walk_completed,itlb.itlb_flush 7,720,030,970 dtlb_load_misses_walk_duration [57.13%] 169,856,353 dtlb_load_misses_walk_completed [57.15%] 708,832,859 dtlb_store_misses_walk_duration [57.17%] 19,346,823 dtlb_store_misses_walk_completed [57.17%] 2,779,687,402 itlb_misses_walk_duration [57.15%] 82,241,148 itlb_misses_walk_completed [57.13%] 770,717 itlb_itlb_flush [57.11%] Show that a dtlb miss is 17.1ns (~45 cycles) and a itlb miss is 13.0ns (~34 cycles). At those rates, refilling the 512-entry dTLB takes 22,000 cycles. On a SandyBridge system with more cores and larger caches, those are dtlb=13.4ns and itlb=9.5ns. cat perf.stat.txt | perl -pe 's/,//g' | awk '/itlb_misses_walk_duration/ { icyc+=$1 } /itlb_misses_walk_completed/ { imiss+=$1 } /dtlb_.*_walk_duration/ { dcyc+=$1 } /dtlb_.*.*completed/ { dmiss+=$1 } END {print "itlb cyc/miss: ", icyc/imiss, " dtlb cyc/miss: ", dcyc/dmiss, " ----- ", icyc,imiss, dcyc,dmiss } On Westmere CPUs, the counters to use are: itlb_flush,itlb_misses.walk_cycles,itlb_misses.any,dtlb_misses.walk_cycles,dtlb_misses.any The assumptions that this code went in under: https://lkml.org/lkml/2012/6/12/119 say that a flush and a refill are about 100ns. Being generous, that is over by a factor of 6 on the refill side, although it is fairly close on the cost of an invlpg. An increase of a single invlpg operation seems to lengthen the flush range operation by about 200 cycles. Here is one example of the data collected for flushing 10 and 11 pages (full data are below): 10: 0.43% 93.90% avg cycles: 3570 cycles/page: 357 samples: 4714 11: 0.20% 94.10% avg cycles: 3767 cycles/page: 342 samples: 2145 How to generate this table: echo 10000 > /sys/kernel/debug/tracing/buffer_size_kb echo x86-tsc > /sys/kernel/debug/tracing/trace_clock echo 'reason != 0' > /sys/kernel/debug/tracing/events/tlb/tlb_flush/filter echo 1 > /sys/kernel/debug/tracing/events/tlb/tlb_flush/enable Pipe the trace output in to this script: http://sr71.net/~dave/intel/201402-tlb/trace-time-diff-process.pl.txt Note that these data were gathered with the invlpg threshold set to 150 pages. Only data points with >=50 of samples were printed: Flush % of %<= in flush this pages es size ------------------------------------------------------------------------------ -1: 2.20% 2.20% avg cycles: 2283 cycles/page: xxxx samples: 23960 1: 56.92% 59.12% avg cycles: 1276 cycles/page: 1276 samples: 620895 2: 13.78% 72.90% avg cycles: 1505 cycles/page: 752 samples: 150335 3: 8.26% 81.16% avg cycles: 1880 cycles/page: 626 samples: 90131 4: 7.41% 88.58% avg cycles: 2447 cycles/page: 611 samples: 80877 5: 1.73% 90.31% avg cycles: 2358 cycles/page: 471 samples: 18885 6: 1.32% 91.63% avg cycles: 2563 cycles/page: 427 samples: 14397 7: 1.14% 92.77% avg cycles: 2862 cycles/page: 408 samples: 12441 8: 0.62% 93.39% avg cycles: 3542 cycles/page: 442 samples: 6721 9: 0.08% 93.47% avg cycles: 3289 cycles/page: 365 samples: 917 10: 0.43% 93.90% avg cycles: 3570 cycles/page: 357 samples: 4714 11: 0.20% 94.10% avg cycles: 3767 cycles/page: 342 samples: 2145 12: 0.08% 94.18% avg cycles: 3996 cycles/page: 333 samples: 864 13: 0.03% 94.20% avg cycles: 4077 cycles/page: 313 samples: 289 14: 0.02% 94.23% avg cycles: 4836 cycles/page: 345 samples: 236 15: 0.04% 94.26% avg cycles: 5699 cycles/page: 379 samples: 390 16: 0.06% 94.32% avg cycles: 5041 cycles/page: 315 samples: 643 17: 0.57% 94.89% avg cycles: 5473 cycles/page: 321 samples: 6229 18: 0.02% 94.91% avg cycles: 5396 cycles/page: 299 samples: 224 19: 0.03% 94.95% avg cycles: 5296 cycles/page: 278 samples: 367 20: 0.02% 94.96% avg cycles: 6749 cycles/page: 337 samples: 185 21: 0.18% 95.14% avg cycles: 6225 cycles/page: 296 samples: 1964 22: 0.01% 95.15% avg cycles: 6393 cycles/page: 290 samples: 83 23: 0.01% 95.16% avg cycles: 6861 cycles/page: 298 samples: 61 24: 0.12% 95.28% avg cycles: 6912 cycles/page: 288 samples: 1307 25: 0.05% 95.32% avg cycles: 7190 cycles/page: 287 samples: 533 26: 0.01% 95.33% avg cycles: 7793 cycles/page: 299 samples: 94 27: 0.01% 95.34% avg cycles: 7833 cycles/page: 290 samples: 66 28: 0.01% 95.35% avg cycles: 8253 cycles/page: 294 samples: 73 29: 0.08% 95.42% avg cycles: 8024 cycles/page: 276 samples: 846 30: 0.03% 95.45% avg cycles: 9670 cycles/page: 322 samples: 296 31: 0.01% 95.46% avg cycles: 8949 cycles/page: 288 samples: 79 32: 0.01% 95.46% avg cycles: 9350 cycles/page: 292 samples: 60 33: 3.11% 98.57% avg cycles: 8534 cycles/page: 258 samples: 33936 34: 0.02% 98.60% avg cycles: 10977 cycles/page: 322 samples: 268 35: 0.02% 98.62% avg cycles: 11400 cycles/page: 325 samples: 177 36: 0.01% 98.63% avg cycles: 11504 cycles/page: 319 samples: 161 37: 0.02% 98.65% avg cycles: 11596 cycles/page: 313 samples: 182 38: 0.02% 98.66% avg cycles: 11850 cycles/page: 311 samples: 195 39: 0.01% 98.68% avg cycles: 12158 cycles/page: 311 samples: 128 40: 0.01% 98.68% avg cycles: 11626 cycles/page: 290 samples: 78 41: 0.04% 98.73% avg cycles: 11435 cycles/page: 278 samples: 477 42: 0.01% 98.73% avg cycles: 12571 cycles/page: 299 samples: 74 43: 0.01% 98.74% avg cycles: 12562 cycles/page: 292 samples: 78 44: 0.01% 98.75% avg cycles: 12991 cycles/page: 295 samples: 108 45: 0.01% 98.76% avg cycles: 13169 cycles/page: 292 samples: 78 46: 0.02% 98.78% avg cycles: 12891 cycles/page: 280 samples: 261 47: 0.01% 98.79% avg cycles: 13099 cycles/page: 278 samples: 67 48: 0.01% 98.80% avg cycles: 13851 cycles/page: 288 samples: 77 49: 0.01% 98.80% avg cycles: 13749 cycles/page: 280 samples: 66 50: 0.01% 98.81% avg cycles: 13949 cycles/page: 278 samples: 73 52: 0.00% 98.82% avg cycles: 14243 cycles/page: 273 samples: 52 54: 0.01% 98.83% avg cycles: 15312 cycles/page: 283 samples: 87 55: 0.01% 98.84% avg cycles: 15197 cycles/page: 276 samples: 109 56: 0.02% 98.86% avg cycles: 15234 cycles/page: 272 samples: 208 57: 0.00% 98.86% avg cycles: 14888 cycles/page: 261 samples: 53 58: 0.01% 98.87% avg cycles: 15037 cycles/page: 259 samples: 59 59: 0.01% 98.87% avg cycles: 15752 cycles/page: 266 samples: 63 62: 0.00% 98.89% avg cycles: 16222 cycles/page: 261 samples: 54 64: 0.02% 98.91% avg cycles: 17179 cycles/page: 268 samples: 248 65: 0.12% 99.03% avg cycles: 18762 cycles/page: 288 samples: 1324 85: 0.00% 99.10% avg cycles: 21649 cycles/page: 254 samples: 50 127: 0.01% 99.18% avg cycles: 32397 cycles/page: 255 samples: 75 128: 0.13% 99.31% avg cycles: 31711 cycles/page: 247 samples: 1466 129: 0.18% 99.49% avg cycles: 33017 cycles/page: 255 samples: 1927 181: 0.33% 99.84% avg cycles: 2489 cycles/page: 13 samples: 3547 256: 0.05% 99.91% avg cycles: 2305 cycles/page: 9 samples: 550 512: 0.03% 99.95% avg cycles: 2133 cycles/page: 4 samples: 304 1512: 0.01% 99.99% avg cycles: 3038 cycles/page: 2 samples: 65 Here are the tlb counters during a 10-second slice of a kernel compile for a SandyBridge system. It's better than IvyBridge, but probably due to the larger caches since this was one of the 'X' extreme parts. 10,873,007,282 dtlb_load_misses_walk_duration 250,711,333 dtlb_load_misses_walk_completed 1,212,395,865 dtlb_store_misses_walk_duration 31,615,772 dtlb_store_misses_walk_completed 5,091,010,274 itlb_misses_walk_duration 163,193,511 itlb_misses_walk_completed 1,321,980 itlb_itlb_flush 10.008045158 seconds time elapsed # cat perf.stat.1392743721.txt | perl -pe 's/,//g' | awk '/itlb_misses_walk_duration/ { icyc+=$1 } /itlb_misses_walk_completed/ { imiss+=$1 } /dtlb_.*_walk_duration/ { dcyc+=$1 } /dtlb_.*.*completed/ { dmiss+=$1 } END {print "itlb cyc/miss: ", icyc/imiss/3.3, " dtlb cyc/miss: ", dcyc/dmiss/3.3, " ----- ", icyc,imiss, dcyc,dmiss }' itlb ns/miss: 9.45338 dtlb ns/miss: 12.9716 Signed-off-by: Dave Hansen <dave.hansen@linux.intel.com> Link: http://lkml.kernel.org/r/20140731154103.10C1115E@viggo.jf.intel.com Acked-by: Rik van Riel <riel@redhat.com> Acked-by: Mel Gorman <mgorman@suse.de> Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
2014-07-31 23:41:03 +08:00
/*
* See Documentation/x86/tlb.txt for details. We choose 33
* because it is large enough to cover the vast majority (at
* least 95%) of allocations, and is small enough that we are
* confident it will not cause too much overhead. Each single
* flush is about 100 ns, so this caps the maximum overhead at
* _about_ 3,000 ns.
*
* This is in units of pages.
*/
static unsigned long tlb_single_page_flush_ceiling __read_mostly = 33;
x86/mm: Rip out complicated, out-of-date, buggy TLB flushing I think the flush_tlb_mm_range() code that tries to tune the flush sizes based on the CPU needs to get ripped out for several reasons: 1. It is obviously buggy. It uses mm->total_vm to judge the task's footprint in the TLB. It should certainly be using some measure of RSS, *NOT* ->total_vm since only resident memory can populate the TLB. 2. Haswell, and several other CPUs are missing from the intel_tlb_flushall_shift_set() function. Thus, it has been demonstrated to bitrot quickly in practice. 3. It is plain wrong in my vm: [ 0.037444] Last level iTLB entries: 4KB 0, 2MB 0, 4MB 0 [ 0.037444] Last level dTLB entries: 4KB 0, 2MB 0, 4MB 0 [ 0.037444] tlb_flushall_shift: 6 Which leads to it to never use invlpg. 4. The assumptions about TLB refill costs are wrong: http://lkml.kernel.org/r/1337782555-8088-3-git-send-email-alex.shi@intel.com (more on this in later patches) 5. I can not reproduce the original data: https://lkml.org/lkml/2012/5/17/59 I believe the sample times were too short. Running the benchmark in a loop yields times that vary quite a bit. Note that this leaves us with a static ceiling of 1 page. This is a conservative, dumb setting, and will be revised in a later patch. This also removes the code which attempts to predict whether we are flushing data or instructions. We expect instruction flushes to be relatively rare and not worth tuning for explicitly. Signed-off-by: Dave Hansen <dave.hansen@linux.intel.com> Link: http://lkml.kernel.org/r/20140731154055.ABC88E89@viggo.jf.intel.com Acked-by: Rik van Riel <riel@redhat.com> Acked-by: Mel Gorman <mgorman@suse.de> Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
2014-07-31 23:40:55 +08:00
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>
2012-06-28 09:02:22 +08:00
void flush_tlb_mm_range(struct mm_struct *mm, unsigned long start,
unsigned long end, unsigned long vmflag)
{
unsigned long addr;
/* do a global flush by default */
unsigned long base_pages_to_flush = TLB_FLUSH_ALL;
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>
2012-06-28 09:02:17 +08:00
preempt_disable();
if (current->active_mm != mm) {
/* Synchronize with switch_mm. */
smp_mb();
goto out;
}
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>
2012-06-28 09:02:17 +08:00
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>
2012-06-28 09:02:22 +08:00
if (!current->mm) {
leave_mm(smp_processor_id());
/* Synchronize with switch_mm. */
smp_mb();
goto out;
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>
2012-06-28 09:02:22 +08:00
}
if ((end != TLB_FLUSH_ALL) && !(vmflag & VM_HUGETLB))
base_pages_to_flush = (end - start) >> PAGE_SHIFT;
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>
2012-06-28 09:02:17 +08:00
/*
* Both branches below are implicit full barriers (MOV to CR or
* INVLPG) that synchronize with switch_mm.
*/
if (base_pages_to_flush > tlb_single_page_flush_ceiling) {
base_pages_to_flush = TLB_FLUSH_ALL;
count_vm_tlb_event(NR_TLB_LOCAL_FLUSH_ALL);
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>
2012-06-28 09:02:22 +08:00
local_flush_tlb();
} else {
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>
2012-06-28 09:02:22 +08:00
/* flush range by one by one 'invlpg' */
for (addr = start; addr < end; addr += PAGE_SIZE) {
count_vm_tlb_event(NR_TLB_LOCAL_FLUSH_ONE);
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>
2012-06-28 09:02:22 +08:00
__flush_tlb_single(addr);
}
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>
2012-06-28 09:02:17 +08:00
}
trace_tlb_flush(TLB_LOCAL_MM_SHOOTDOWN, base_pages_to_flush);
out:
if (base_pages_to_flush == TLB_FLUSH_ALL) {
start = 0UL;
end = TLB_FLUSH_ALL;
}
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>
2012-06-28 09:02:17 +08:00
if (cpumask_any_but(mm_cpumask(mm), smp_processor_id()) < nr_cpu_ids)
flush_tlb_others(mm_cpumask(mm), mm, start, end);
preempt_enable();
}
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>
2012-06-28 09:02:17 +08:00
void flush_tlb_page(struct vm_area_struct *vma, unsigned long start)
{
struct mm_struct *mm = vma->vm_mm;
preempt_disable();
if (current->active_mm == mm) {
if (current->mm) {
/*
* Implicit full barrier (INVLPG) that synchronizes
* with switch_mm.
*/
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>
2012-06-28 09:02:17 +08:00
__flush_tlb_one(start);
} else {
leave_mm(smp_processor_id());
/* Synchronize with switch_mm. */
smp_mb();
}
}
if (cpumask_any_but(mm_cpumask(mm), smp_processor_id()) < nr_cpu_ids)
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>
2012-06-28 09:02:17 +08:00
flush_tlb_others(mm_cpumask(mm), mm, start, 0UL);
preempt_enable();
}
static void do_flush_tlb_all(void *info)
{
count_vm_tlb_event(NR_TLB_REMOTE_FLUSH_RECEIVED);
__flush_tlb_all();
if (this_cpu_read(cpu_tlbstate.state) == TLBSTATE_LAZY)
leave_mm(smp_processor_id());
}
void flush_tlb_all(void)
{
count_vm_tlb_event(NR_TLB_REMOTE_FLUSH);
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)
{
/* Balance as user space task's flush, a bit conservative */
x86/mm: Rip out complicated, out-of-date, buggy TLB flushing I think the flush_tlb_mm_range() code that tries to tune the flush sizes based on the CPU needs to get ripped out for several reasons: 1. It is obviously buggy. It uses mm->total_vm to judge the task's footprint in the TLB. It should certainly be using some measure of RSS, *NOT* ->total_vm since only resident memory can populate the TLB. 2. Haswell, and several other CPUs are missing from the intel_tlb_flushall_shift_set() function. Thus, it has been demonstrated to bitrot quickly in practice. 3. It is plain wrong in my vm: [ 0.037444] Last level iTLB entries: 4KB 0, 2MB 0, 4MB 0 [ 0.037444] Last level dTLB entries: 4KB 0, 2MB 0, 4MB 0 [ 0.037444] tlb_flushall_shift: 6 Which leads to it to never use invlpg. 4. The assumptions about TLB refill costs are wrong: http://lkml.kernel.org/r/1337782555-8088-3-git-send-email-alex.shi@intel.com (more on this in later patches) 5. I can not reproduce the original data: https://lkml.org/lkml/2012/5/17/59 I believe the sample times were too short. Running the benchmark in a loop yields times that vary quite a bit. Note that this leaves us with a static ceiling of 1 page. This is a conservative, dumb setting, and will be revised in a later patch. This also removes the code which attempts to predict whether we are flushing data or instructions. We expect instruction flushes to be relatively rare and not worth tuning for explicitly. Signed-off-by: Dave Hansen <dave.hansen@linux.intel.com> Link: http://lkml.kernel.org/r/20140731154055.ABC88E89@viggo.jf.intel.com Acked-by: Rik van Riel <riel@redhat.com> Acked-by: Mel Gorman <mgorman@suse.de> Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
2014-07-31 23:40:55 +08:00
if (end == TLB_FLUSH_ALL ||
(end - start) > tlb_single_page_flush_ceiling * PAGE_SIZE) {
on_each_cpu(do_flush_tlb_all, NULL, 1);
x86/mm: Rip out complicated, out-of-date, buggy TLB flushing I think the flush_tlb_mm_range() code that tries to tune the flush sizes based on the CPU needs to get ripped out for several reasons: 1. It is obviously buggy. It uses mm->total_vm to judge the task's footprint in the TLB. It should certainly be using some measure of RSS, *NOT* ->total_vm since only resident memory can populate the TLB. 2. Haswell, and several other CPUs are missing from the intel_tlb_flushall_shift_set() function. Thus, it has been demonstrated to bitrot quickly in practice. 3. It is plain wrong in my vm: [ 0.037444] Last level iTLB entries: 4KB 0, 2MB 0, 4MB 0 [ 0.037444] Last level dTLB entries: 4KB 0, 2MB 0, 4MB 0 [ 0.037444] tlb_flushall_shift: 6 Which leads to it to never use invlpg. 4. The assumptions about TLB refill costs are wrong: http://lkml.kernel.org/r/1337782555-8088-3-git-send-email-alex.shi@intel.com (more on this in later patches) 5. I can not reproduce the original data: https://lkml.org/lkml/2012/5/17/59 I believe the sample times were too short. Running the benchmark in a loop yields times that vary quite a bit. Note that this leaves us with a static ceiling of 1 page. This is a conservative, dumb setting, and will be revised in a later patch. This also removes the code which attempts to predict whether we are flushing data or instructions. We expect instruction flushes to be relatively rare and not worth tuning for explicitly. Signed-off-by: Dave Hansen <dave.hansen@linux.intel.com> Link: http://lkml.kernel.org/r/20140731154055.ABC88E89@viggo.jf.intel.com Acked-by: Rik van Riel <riel@redhat.com> Acked-by: Mel Gorman <mgorman@suse.de> Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
2014-07-31 23:40:55 +08:00
} else {
struct flush_tlb_info info;
info.flush_start = start;
info.flush_end = end;
on_each_cpu(do_kernel_range_flush, &info, 1);
}
}
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, "%ld\n", tlb_single_page_flush_ceiling);
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;
int ceiling;
len = min(count, sizeof(buf) - 1);
if (copy_from_user(buf, user_buf, len))
return -EFAULT;
buf[len] = '\0';
if (kstrtoint(buf, 0, &ceiling))
return -EINVAL;
if (ceiling < 0)
return -EINVAL;
tlb_single_page_flush_ceiling = ceiling;
return count;
}
static const struct file_operations fops_tlbflush = {
.read = tlbflush_read_file,
.write = tlbflush_write_file,
.llseek = default_llseek,
};
static int __init create_tlb_single_page_flush_ceiling(void)
{
debugfs_create_file("tlb_single_page_flush_ceiling", S_IRUSR | S_IWUSR,
arch_debugfs_dir, NULL, &fops_tlbflush);
return 0;
}
late_initcall(create_tlb_single_page_flush_ceiling);
#endif /* CONFIG_SMP */