* [PATCH] trace: adjust code layout in get_recursion_context
@ 2017-08-22 14:40 Jesper Dangaard Brouer
2017-08-22 15:14 ` Peter Zijlstra
0 siblings, 1 reply; 9+ messages in thread
From: Jesper Dangaard Brouer @ 2017-08-22 14:40 UTC (permalink / raw)
To: Arnaldo Carvalho de Melo
Cc: Peter Zijlstra, Jesper Dangaard Brouer, linux-kernel, Jiri Olsa,
Ingo Molnar
In an XDP redirect applications using tracepoint xdp:xdp_redirect to
diagnose TX overrun, I noticed perf_swevent_get_recursion_context()
was consuming 2% CPU. This was reduced to 1.6% with this simple
change.
Looking at the annotated asm code, it was clear that the unlikely case
in_nmi() test was chosen (by the compiler) as the most likely
event/branch. This small adjustment makes the compiler (gcc version
7.1.1 20170622 (Red Hat 7.1.1-3)) put in_nmi() as an unlikely branch.
Signed-off-by: Jesper Dangaard Brouer <brouer@redhat.com>
---
kernel/events/internal.h | 8 ++++----
1 file changed, 4 insertions(+), 4 deletions(-)
diff --git a/kernel/events/internal.h b/kernel/events/internal.h
index 486fd78eb8d5..56aa462760fa 100644
--- a/kernel/events/internal.h
+++ b/kernel/events/internal.h
@@ -208,12 +208,12 @@ static inline int get_recursion_context(int *recursion)
{
int rctx;
- if (in_nmi())
- rctx = 3;
+ if (in_softirq())
+ rctx = 1;
else if (in_irq())
rctx = 2;
- else if (in_softirq())
- rctx = 1;
+ else if (in_nmi())
+ rctx = 3;
else
rctx = 0;
^ permalink raw reply related [flat|nested] 9+ messages in thread
* Re: [PATCH] trace: adjust code layout in get_recursion_context
2017-08-22 14:40 [PATCH] trace: adjust code layout in get_recursion_context Jesper Dangaard Brouer
@ 2017-08-22 15:14 ` Peter Zijlstra
2017-08-22 15:20 ` Peter Zijlstra
0 siblings, 1 reply; 9+ messages in thread
From: Peter Zijlstra @ 2017-08-22 15:14 UTC (permalink / raw)
To: Jesper Dangaard Brouer
Cc: Arnaldo Carvalho de Melo, linux-kernel, Jiri Olsa, Ingo Molnar
On Tue, Aug 22, 2017 at 04:40:24PM +0200, Jesper Dangaard Brouer wrote:
> In an XDP redirect applications using tracepoint xdp:xdp_redirect to
> diagnose TX overrun, I noticed perf_swevent_get_recursion_context()
> was consuming 2% CPU. This was reduced to 1.6% with this simple
> change.
It is also incorrect. What do you suppose it now returns when the NMI
hits a hard IRQ which hit during a Soft IRQ?
> @@ -208,12 +208,12 @@ static inline int get_recursion_context(int *recursion)
> {
> int rctx;
>
> + if (in_softirq())
> + rctx = 1;
> else if (in_irq())
> rctx = 2;
> + else if (in_nmi())
> + rctx = 3;
> else
> rctx = 0;
>
>
^ permalink raw reply [flat|nested] 9+ messages in thread
* Re: [PATCH] trace: adjust code layout in get_recursion_context
2017-08-22 15:14 ` Peter Zijlstra
@ 2017-08-22 15:20 ` Peter Zijlstra
2017-08-22 17:00 ` Jesper Dangaard Brouer
0 siblings, 1 reply; 9+ messages in thread
From: Peter Zijlstra @ 2017-08-22 15:20 UTC (permalink / raw)
To: Jesper Dangaard Brouer
Cc: Arnaldo Carvalho de Melo, linux-kernel, Jiri Olsa, Ingo Molnar
On Tue, Aug 22, 2017 at 05:14:10PM +0200, Peter Zijlstra wrote:
> On Tue, Aug 22, 2017 at 04:40:24PM +0200, Jesper Dangaard Brouer wrote:
> > In an XDP redirect applications using tracepoint xdp:xdp_redirect to
> > diagnose TX overrun, I noticed perf_swevent_get_recursion_context()
> > was consuming 2% CPU. This was reduced to 1.6% with this simple
> > change.
>
> It is also incorrect. What do you suppose it now returns when the NMI
> hits a hard IRQ which hit during a Soft IRQ?
Does this help any? I can imagine the compiler could struggle to CSE
preempt_count() seeing how its an asm thing.
---
kernel/events/internal.h | 7 ++++---
1 file changed, 4 insertions(+), 3 deletions(-)
diff --git a/kernel/events/internal.h b/kernel/events/internal.h
index 486fd78eb8d5..e0b5b8fa83a2 100644
--- a/kernel/events/internal.h
+++ b/kernel/events/internal.h
@@ -206,13 +206,14 @@ perf_callchain(struct perf_event *event, struct pt_regs *regs);
static inline int get_recursion_context(int *recursion)
{
+ unsigned int pc = preempt_count();
int rctx;
- if (in_nmi())
+ if (pc & NMI_MASK)
rctx = 3;
- else if (in_irq())
+ else if (pc & HARDIRQ_MASK)
rctx = 2;
- else if (in_softirq())
+ else if (pc & SOFTIRQ_OFFSET)
rctx = 1;
else
rctx = 0;
^ permalink raw reply related [flat|nested] 9+ messages in thread
* Re: [PATCH] trace: adjust code layout in get_recursion_context
2017-08-22 15:20 ` Peter Zijlstra
@ 2017-08-22 17:00 ` Jesper Dangaard Brouer
2017-08-22 17:10 ` Jesper Dangaard Brouer
2017-08-22 17:55 ` [PATCH] trace: adjust code layout in get_recursion_context Peter Zijlstra
0 siblings, 2 replies; 9+ messages in thread
From: Jesper Dangaard Brouer @ 2017-08-22 17:00 UTC (permalink / raw)
To: Peter Zijlstra
Cc: Arnaldo Carvalho de Melo, linux-kernel, Jiri Olsa, Ingo Molnar, brouer
On Tue, 22 Aug 2017 17:20:25 +0200
Peter Zijlstra <peterz@infradead.org> wrote:
> On Tue, Aug 22, 2017 at 05:14:10PM +0200, Peter Zijlstra wrote:
> > On Tue, Aug 22, 2017 at 04:40:24PM +0200, Jesper Dangaard Brouer wrote:
> > > In an XDP redirect applications using tracepoint xdp:xdp_redirect to
> > > diagnose TX overrun, I noticed perf_swevent_get_recursion_context()
> > > was consuming 2% CPU. This was reduced to 1.6% with this simple
> > > change.
> >
> > It is also incorrect. What do you suppose it now returns when the NMI
> > hits a hard IRQ which hit during a Soft IRQ?
>
> Does this help any? I can imagine the compiler could struggle to CSE
> preempt_count() seeing how its an asm thing.
Nope, it does not help (see assembly below, with perf percentages).
But I think I can achieve that I want by a simple unlikely(in_nmi()) annotation.
> ---
> kernel/events/internal.h | 7 ++++---
> 1 file changed, 4 insertions(+), 3 deletions(-)
>
> diff --git a/kernel/events/internal.h b/kernel/events/internal.h
> index 486fd78eb8d5..e0b5b8fa83a2 100644
> --- a/kernel/events/internal.h
> +++ b/kernel/events/internal.h
> @@ -206,13 +206,14 @@ perf_callchain(struct perf_event *event, struct pt_regs *regs);
>
> static inline int get_recursion_context(int *recursion)
> {
> + unsigned int pc = preempt_count();
> int rctx;
>
> - if (in_nmi())
> + if (pc & NMI_MASK)
> rctx = 3;
> - else if (in_irq())
> + else if (pc & HARDIRQ_MASK)
> rctx = 2;
> - else if (in_softirq())
> + else if (pc & SOFTIRQ_OFFSET)
Hmmm... shouldn't this be SOFTIRQ_MASK?
> rctx = 1;
> else
> rctx = 0;
perf_swevent_get_recursion_context /proc/kcore
│
│
│ Disassembly of section load0:
│
│ ffffffff811465c0 <load0>:
13.32 │ push %rbp
1.43 │ mov $0x14d20,%rax
5.12 │ mov %rsp,%rbp
6.56 │ add %gs:0x7eec3b5d(%rip),%rax
0.72 │ lea 0x34(%rax),%rdx
0.31 │ mov %gs:0x7eec5db2(%rip),%eax
2.46 │ mov %eax,%ecx
6.86 │ and $0x7fffffff,%ecx
0.72 │ test $0x100000,%eax
│ ↓ jne 40
│ test $0xf0000,%eax
0.41 │ ↓ je 5b
│ mov $0x8,%ecx
│ mov $0x2,%eax
│ ↓ jmp 4a
│40: mov $0xc,%ecx
│ mov $0x3,%eax
2.05 │4a: add %rcx,%rdx
16.60 │ mov (%rdx),%ecx
2.66 │ test %ecx,%ecx
│ ↓ jne 6d
1.33 │ movl $0x1,(%rdx)
1.54 │ pop %rbp
4.51 │ ← retq
3.89 │5b: shr $0x8,%ecx
9.53 │ and $0x1,%ecx
0.61 │ movzbl %cl,%eax
0.92 │ movzbl %cl,%ecx
4.30 │ shl $0x2,%rcx
14.14 │ ↑ jmp 4a
│6d: mov $0xffffffff,%eax
│ pop %rbp
│ ← retq
│ xchg %ax,%ax
--
Best regards,
Jesper Dangaard Brouer
MSc.CS, Principal Kernel Engineer at Red Hat
LinkedIn: http://www.linkedin.com/in/brouer
^ permalink raw reply [flat|nested] 9+ messages in thread
* Re: [PATCH] trace: adjust code layout in get_recursion_context
2017-08-22 17:00 ` Jesper Dangaard Brouer
@ 2017-08-22 17:10 ` Jesper Dangaard Brouer
2017-08-22 17:22 ` [PATCH V2] " Jesper Dangaard Brouer
2017-08-22 17:55 ` [PATCH] trace: adjust code layout in get_recursion_context Peter Zijlstra
1 sibling, 1 reply; 9+ messages in thread
From: Jesper Dangaard Brouer @ 2017-08-22 17:10 UTC (permalink / raw)
To: Peter Zijlstra
Cc: Arnaldo Carvalho de Melo, linux-kernel, Jiri Olsa, Ingo Molnar, brouer
On Tue, 22 Aug 2017 19:00:39 +0200
Jesper Dangaard Brouer <brouer@redhat.com> wrote:
> On Tue, 22 Aug 2017 17:20:25 +0200
> Peter Zijlstra <peterz@infradead.org> wrote:
>
> > On Tue, Aug 22, 2017 at 05:14:10PM +0200, Peter Zijlstra wrote:
> > > On Tue, Aug 22, 2017 at 04:40:24PM +0200, Jesper Dangaard Brouer wrote:
> > > > In an XDP redirect applications using tracepoint xdp:xdp_redirect to
> > > > diagnose TX overrun, I noticed perf_swevent_get_recursion_context()
> > > > was consuming 2% CPU. This was reduced to 1.6% with this simple
> > > > change.
> > >
> > > It is also incorrect. What do you suppose it now returns when the NMI
> > > hits a hard IRQ which hit during a Soft IRQ?
> >
> > Does this help any? I can imagine the compiler could struggle to CSE
> > preempt_count() seeing how its an asm thing.
>
> Nope, it does not help (see assembly below, with perf percentages).
>
> But I think I can achieve that I want by a simple unlikely(in_nmi()) annotation.
Like:
diff --git a/kernel/events/internal.h b/kernel/events/internal.h
index 486fd78eb8d5..e1a7ac7bd686 100644
--- a/kernel/events/internal.h
+++ b/kernel/events/internal.h
@@ -208,7 +208,7 @@ static inline int get_recursion_context(int *recursion)
{
int rctx;
- if (in_nmi())
+ if (unlikely(in_nmi()))
rctx = 3;
else if (in_irq())
rctx = 2;
Testing this show I get the expected result. Although, the 2% is
reduced to 1.85% (and not 1.6% as before).
perf_swevent_get_recursion_context /proc/kcore
│
│ Disassembly of section load0:
│
│ ffffffff811465c0 <load0>:
4.94 │ push %rbp
2.56 │ mov $0x14d20,%rax
14.81 │ mov %rsp,%rbp
3.47 │ add %gs:0x7eec3b5d(%rip),%rax
0.91 │ lea 0x34(%rax),%rdx
1.46 │ mov %gs:0x7eec5db2(%rip),%eax
8.04 │ test $0x100000,%eax
│ ↓ jne 59
3.11 │ test $0xf0000,%eax
│ ↓ jne 4d
0.37 │ test $0xff,%ah
1.83 │ setne %cl
9.87 │ movzbl %cl,%eax
2.01 │ movzbl %cl,%ecx
1.65 │ shl $0x2,%rcx
4.39 │3c: add %rcx,%rdx
29.62 │ mov (%rdx),%ecx
2.93 │ test %ecx,%ecx
│ ↓ jne 65
0.55 │ movl $0x1,(%rdx)
2.56 │ pop %rbp
4.94 │ ← retq
│4d: mov $0x8,%ecx
│ mov $0x2,%eax
│ ↑ jmp 3c
│59: mov $0xc,%ecx
│ mov $0x3,%eax
│ ↑ jmp 3c
│65: mov $0xffffffff,%eax
│ pop %rbp
│ ← retq
> > ---
> > kernel/events/internal.h | 7 ++++---
> > 1 file changed, 4 insertions(+), 3 deletions(-)
> >
> > diff --git a/kernel/events/internal.h b/kernel/events/internal.h
> > index 486fd78eb8d5..e0b5b8fa83a2 100644
> > --- a/kernel/events/internal.h
> > +++ b/kernel/events/internal.h
> > @@ -206,13 +206,14 @@ perf_callchain(struct perf_event *event, struct pt_regs *regs);
> >
> > static inline int get_recursion_context(int *recursion)
> > {
> > + unsigned int pc = preempt_count();
> > int rctx;
> >
> > - if (in_nmi())
> > + if (pc & NMI_MASK)
> > rctx = 3;
> > - else if (in_irq())
> > + else if (pc & HARDIRQ_MASK)
> > rctx = 2;
> > - else if (in_softirq())
> > + else if (pc & SOFTIRQ_OFFSET)
>
> Hmmm... shouldn't this be SOFTIRQ_MASK?
>
> > rctx = 1;
> > else
> > rctx = 0;
>
> perf_swevent_get_recursion_context /proc/kcore
> │
> │
> │ Disassembly of section load0:
> │
> │ ffffffff811465c0 <load0>:
> 13.32 │ push %rbp
> 1.43 │ mov $0x14d20,%rax
> 5.12 │ mov %rsp,%rbp
> 6.56 │ add %gs:0x7eec3b5d(%rip),%rax
> 0.72 │ lea 0x34(%rax),%rdx
> 0.31 │ mov %gs:0x7eec5db2(%rip),%eax
> 2.46 │ mov %eax,%ecx
> 6.86 │ and $0x7fffffff,%ecx
> 0.72 │ test $0x100000,%eax
> │ ↓ jne 40
> │ test $0xf0000,%eax
> 0.41 │ ↓ je 5b
> │ mov $0x8,%ecx
> │ mov $0x2,%eax
> │ ↓ jmp 4a
> │40: mov $0xc,%ecx
> │ mov $0x3,%eax
> 2.05 │4a: add %rcx,%rdx
> 16.60 │ mov (%rdx),%ecx
> 2.66 │ test %ecx,%ecx
> │ ↓ jne 6d
> 1.33 │ movl $0x1,(%rdx)
> 1.54 │ pop %rbp
> 4.51 │ ← retq
> 3.89 │5b: shr $0x8,%ecx
> 9.53 │ and $0x1,%ecx
> 0.61 │ movzbl %cl,%eax
> 0.92 │ movzbl %cl,%ecx
> 4.30 │ shl $0x2,%rcx
> 14.14 │ ↑ jmp 4a
> │6d: mov $0xffffffff,%eax
> │ pop %rbp
> │ ← retq
> │ xchg %ax,%ax
>
>
>
--
Best regards,
Jesper Dangaard Brouer
MSc.CS, Principal Kernel Engineer at Red Hat
LinkedIn: http://www.linkedin.com/in/brouer
^ permalink raw reply related [flat|nested] 9+ messages in thread
* [PATCH V2] trace: adjust code layout in get_recursion_context
2017-08-22 17:10 ` Jesper Dangaard Brouer
@ 2017-08-22 17:22 ` Jesper Dangaard Brouer
2017-08-25 11:54 ` [tip:perf/core] tracing, perf: Adjust code layout in get_recursion_context() tip-bot for Jesper Dangaard Brouer
0 siblings, 1 reply; 9+ messages in thread
From: Jesper Dangaard Brouer @ 2017-08-22 17:22 UTC (permalink / raw)
To: Arnaldo Carvalho de Melo
Cc: Peter Zijlstra, Jesper Dangaard Brouer, linux-kernel, Jiri Olsa,
Ingo Molnar
In an XDP redirect applications using tracepoint xdp:xdp_redirect to
diagnose TX overrun, I noticed perf_swevent_get_recursion_context()
was consuming 2% CPU. This was reduced to 1.85% with this simple
change.
Looking at the annotated asm code, it was clear that the unlikely case
in_nmi() test was chosen (by the compiler) as the most likely
event/branch. This small adjustment makes the compiler (gcc version
7.1.1 20170622 (Red Hat 7.1.1-3)) put in_nmi() as an unlikely branch.
Signed-off-by: Jesper Dangaard Brouer <brouer@redhat.com>
---
kernel/events/internal.h | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/kernel/events/internal.h b/kernel/events/internal.h
index 486fd78eb8d5..e1a7ac7bd686 100644
--- a/kernel/events/internal.h
+++ b/kernel/events/internal.h
@@ -208,7 +208,7 @@ static inline int get_recursion_context(int *recursion)
{
int rctx;
- if (in_nmi())
+ if (unlikely(in_nmi()))
rctx = 3;
else if (in_irq())
rctx = 2;
^ permalink raw reply related [flat|nested] 9+ messages in thread
* Re: [PATCH] trace: adjust code layout in get_recursion_context
2017-08-22 17:00 ` Jesper Dangaard Brouer
2017-08-22 17:10 ` Jesper Dangaard Brouer
@ 2017-08-22 17:55 ` Peter Zijlstra
2017-08-23 8:12 ` Ingo Molnar
1 sibling, 1 reply; 9+ messages in thread
From: Peter Zijlstra @ 2017-08-22 17:55 UTC (permalink / raw)
To: Jesper Dangaard Brouer
Cc: Arnaldo Carvalho de Melo, linux-kernel, Jiri Olsa, Ingo Molnar
On Tue, Aug 22, 2017 at 07:00:39PM +0200, Jesper Dangaard Brouer wrote:
> > static inline int get_recursion_context(int *recursion)
> > {
> > + unsigned int pc = preempt_count();
> > int rctx;
> >
> > - if (in_nmi())
> > + if (pc & NMI_MASK)
> > rctx = 3;
> > - else if (in_irq())
> > + else if (pc & HARDIRQ_MASK)
> > rctx = 2;
> > - else if (in_softirq())
> > + else if (pc & SOFTIRQ_OFFSET)
>
> Hmmm... shouldn't this be SOFTIRQ_MASK?
No, that was actually broken, this is correct. See the comment near
__local_bh_disable_ip(). We use the low SOFTIRQ bit to indicate if we're
in softirq and the rest of the bits to 'disable' softirq.
The code should've been using in_serving_softirq().
> perf_swevent_get_recursion_context /proc/kcore
> │
> │
> │ Disassembly of section load0:
> │
> │ ffffffff811465c0 <load0>:
> 13.32 │ push %rbp
> 1.43 │ mov $0x14d20,%rax
> 5.12 │ mov %rsp,%rbp
> 6.56 │ add %gs:0x7eec3b5d(%rip),%rax
> 0.72 │ lea 0x34(%rax),%rdx
> 0.31 │ mov %gs:0x7eec5db2(%rip),%eax
> 2.46 │ mov %eax,%ecx
> 6.86 │ and $0x7fffffff,%ecx
> 0.72 │ test $0x100000,%eax
> │ ↓ jne 40
> │ test $0xf0000,%eax
> 0.41 │ ↓ je 5b
> │ mov $0x8,%ecx
> │ mov $0x2,%eax
> │ ↓ jmp 4a
> │40: mov $0xc,%ecx
> │ mov $0x3,%eax
> 2.05 │4a: add %rcx,%rdx
> 16.60 │ mov (%rdx),%ecx
> 2.66 │ test %ecx,%ecx
> │ ↓ jne 6d
> 1.33 │ movl $0x1,(%rdx)
> 1.54 │ pop %rbp
> 4.51 │ ← retq
> 3.89 │5b: shr $0x8,%ecx
> 9.53 │ and $0x1,%ecx
> 0.61 │ movzbl %cl,%eax
> 0.92 │ movzbl %cl,%ecx
> 4.30 │ shl $0x2,%rcx
> 14.14 │ ↑ jmp 4a
> │6d: mov $0xffffffff,%eax
> │ pop %rbp
> │ ← retq
> │ xchg %ax,%ax
>
Bah, that's fairly disgusting code but I'm failing to come up with
anything significantly better with the current rules.
However, if we change the rules on what *_{enter,exit} do, such that we
always have all the bits set, we can. The below patch (100% untested)
gives me the following asm, which looks better (of course, given how
things go today, it'll both explode and be slower once you get it to
work).
Also, its not strictly correct in that while we _should_ not have nested
hardirqs, they _could_ happen and push the SOFTIRQ count out of the 2
bit 'serving' range.
In any case, something to play with...
$ objdump -dr defconfig-build/kernel/events/core.o | awk '/<[^>]*>:$/ {p=0} /<perf_swevent_get_recursion_context>:/ {p=1} {if (p) print $0}'
0000000000004450 <perf_swevent_get_recursion_context>:
4450: 55 push %rbp
4451: 48 c7 c6 00 00 00 00 mov $0x0,%rsi
4454: R_X86_64_32S .data..percpu+0x20
4458: 48 89 e5 mov %rsp,%rbp
445b: 65 48 03 35 00 00 00 add %gs:0x0(%rip),%rsi # 4463 <perf_swevent_get_recursion_context+0x13>
4462: 00
445f: R_X86_64_PC32 this_cpu_off-0x4
4463: 65 8b 15 00 00 00 00 mov %gs:0x0(%rip),%edx # 446a <perf_swevent_get_recursion_context+0x1a>
4466: R_X86_64_PC32 __preempt_count-0x4
446a: 89 d0 mov %edx,%eax
446c: c1 e8 08 shr $0x8,%eax
446f: 83 e0 01 and $0x1,%eax
4472: 89 c1 mov %eax,%ecx
4474: 83 c0 01 add $0x1,%eax
4477: f7 c2 00 00 0f 00 test $0xf0000,%edx
447d: 0f 44 c1 cmove %ecx,%eax
4480: 81 e2 00 00 10 00 and $0x100000,%edx
4486: 83 fa 01 cmp $0x1,%edx
4489: 83 d8 ff sbb $0xffffffff,%eax
448c: 48 63 d0 movslq %eax,%rdx
448f: 48 8d 54 96 2c lea 0x2c(%rsi,%rdx,4),%rdx
4494: 8b 0a mov (%rdx),%ecx
4496: 85 c9 test %ecx,%ecx
4498: 75 08 jne 44a2 <perf_swevent_get_recursion_context+0x52>
449a: c7 02 01 00 00 00 movl $0x1,(%rdx)
44a0: 5d pop %rbp
44a1: c3 retq
44a2: b8 ff ff ff ff mov $0xffffffff,%eax
44a7: 5d pop %rbp
44a8: c3 retq
44a9: 0f 1f 80 00 00 00 00 nopl 0x0(%rax)
---
diff --git a/include/linux/hardirq.h b/include/linux/hardirq.h
index c683996110b1..480babbbc2a5 100644
--- a/include/linux/hardirq.h
+++ b/include/linux/hardirq.h
@@ -35,7 +35,7 @@ static inline void rcu_nmi_exit(void)
#define __irq_enter() \
do { \
account_irq_enter_time(current); \
- preempt_count_add(HARDIRQ_OFFSET); \
+ preempt_count_add(HARDIRQ_OFFSET+SOFTIRQ_OFFSET); \
trace_hardirq_enter(); \
} while (0)
@@ -51,7 +51,7 @@ static inline void rcu_nmi_exit(void)
do { \
trace_hardirq_exit(); \
account_irq_exit_time(current); \
- preempt_count_sub(HARDIRQ_OFFSET); \
+ preempt_count_sub(HARDIRQ_OFFSET+SOFTIRQ_OFFSET); \
} while (0)
/*
@@ -65,7 +65,7 @@ static inline void rcu_nmi_exit(void)
lockdep_off(); \
ftrace_nmi_enter(); \
BUG_ON(in_nmi()); \
- preempt_count_add(NMI_OFFSET + HARDIRQ_OFFSET); \
+ preempt_count_add(NMI_OFFSET + HARDIRQ_OFFSET + SOFTIRQ_OFFSET); \
rcu_nmi_enter(); \
trace_hardirq_enter(); \
} while (0)
@@ -75,7 +75,7 @@ static inline void rcu_nmi_exit(void)
trace_hardirq_exit(); \
rcu_nmi_exit(); \
BUG_ON(!in_nmi()); \
- preempt_count_sub(NMI_OFFSET + HARDIRQ_OFFSET); \
+ preempt_count_sub(NMI_OFFSET + HARDIRQ_OFFSET + SOFTIRQ_OFFSET); \
ftrace_nmi_exit(); \
lockdep_on(); \
printk_nmi_exit(); \
diff --git a/include/linux/preempt.h b/include/linux/preempt.h
index cae461224948..e8e7b759b196 100644
--- a/include/linux/preempt.h
+++ b/include/linux/preempt.h
@@ -50,7 +50,7 @@
#define HARDIRQ_OFFSET (1UL << HARDIRQ_SHIFT)
#define NMI_OFFSET (1UL << NMI_SHIFT)
-#define SOFTIRQ_DISABLE_OFFSET (2 * SOFTIRQ_OFFSET)
+#define SOFTIRQ_DISABLE_OFFSET (4 * SOFTIRQ_OFFSET)
/* We use the MSB mostly because its available */
#define PREEMPT_NEED_RESCHED 0x80000000
diff --git a/kernel/events/internal.h b/kernel/events/internal.h
index 486fd78eb8d5..981e4163e16c 100644
--- a/kernel/events/internal.h
+++ b/kernel/events/internal.h
@@ -206,16 +206,15 @@ static inline unsigned long perf_aux_size(struct ring_buffer *rb)
static inline int get_recursion_context(int *recursion)
{
- int rctx;
-
- if (in_nmi())
- rctx = 3;
- else if (in_irq())
- rctx = 2;
- else if (in_softirq())
- rctx = 1;
- else
- rctx = 0;
+ unsigned int pc = preempt_count();
+ int rctx = 0;
+
+ if (pc & SOFTIRQ_OFFSET)
+ rctx++;
+ if (pc & HARDIRQ_MASK)
+ rctx++;
+ if (pc & NMI_MASK)
+ rctx++;
if (recursion[rctx])
return -1;
^ permalink raw reply related [flat|nested] 9+ messages in thread
* Re: [PATCH] trace: adjust code layout in get_recursion_context
2017-08-22 17:55 ` [PATCH] trace: adjust code layout in get_recursion_context Peter Zijlstra
@ 2017-08-23 8:12 ` Ingo Molnar
0 siblings, 0 replies; 9+ messages in thread
From: Ingo Molnar @ 2017-08-23 8:12 UTC (permalink / raw)
To: Peter Zijlstra
Cc: Jesper Dangaard Brouer, Arnaldo Carvalho de Melo, linux-kernel,
Jiri Olsa
* Peter Zijlstra <peterz@infradead.org> wrote:
> diff --git a/kernel/events/internal.h b/kernel/events/internal.h
> index 486fd78eb8d5..981e4163e16c 100644
> --- a/kernel/events/internal.h
> +++ b/kernel/events/internal.h
> @@ -206,16 +206,15 @@ static inline unsigned long perf_aux_size(struct ring_buffer *rb)
>
> static inline int get_recursion_context(int *recursion)
> {
> - int rctx;
> -
> - if (in_nmi())
> - rctx = 3;
> - else if (in_irq())
> - rctx = 2;
> - else if (in_softirq())
> - rctx = 1;
> - else
> - rctx = 0;
> + unsigned int pc = preempt_count();
> + int rctx = 0;
> +
> + if (pc & SOFTIRQ_OFFSET)
> + rctx++;
> + if (pc & HARDIRQ_MASK)
> + rctx++;
> + if (pc & NMI_MASK)
> + rctx++;
Just a nit: if this ever gets beyond the proof of concent stage please rename 'pc'
to something like 'count', because 'pc' stands for so many other things (program
counter, etc.) which makes it all look a bit weird ...
Thanks,
Ingo
^ permalink raw reply [flat|nested] 9+ messages in thread
* [tip:perf/core] tracing, perf: Adjust code layout in get_recursion_context()
2017-08-22 17:22 ` [PATCH V2] " Jesper Dangaard Brouer
@ 2017-08-25 11:54 ` tip-bot for Jesper Dangaard Brouer
0 siblings, 0 replies; 9+ messages in thread
From: tip-bot for Jesper Dangaard Brouer @ 2017-08-25 11:54 UTC (permalink / raw)
To: linux-tip-commits
Cc: peterz, jolsa, brouer, tglx, linux-kernel, mingo, acme, torvalds, hpa
Commit-ID: d0618410eced4eb092295fad10312a4545fcdfaf
Gitweb: http://git.kernel.org/tip/d0618410eced4eb092295fad10312a4545fcdfaf
Author: Jesper Dangaard Brouer <brouer@redhat.com>
AuthorDate: Tue, 22 Aug 2017 19:22:43 +0200
Committer: Ingo Molnar <mingo@kernel.org>
CommitDate: Fri, 25 Aug 2017 11:04:18 +0200
tracing, perf: Adjust code layout in get_recursion_context()
In an XDP redirect applications using tracepoint xdp:xdp_redirect to
diagnose TX overrun, I noticed perf_swevent_get_recursion_context()
was consuming 2% CPU. This was reduced to 1.85% with this simple
change.
Looking at the annotated asm code, it was clear that the unlikely case
in_nmi() test was chosen (by the compiler) as the most likely
event/branch. This small adjustment makes the compiler (GCC version
7.1.1 20170622 (Red Hat 7.1.1-3)) put in_nmi() as an unlikely branch.
Signed-off-by: Jesper Dangaard Brouer <brouer@redhat.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Arnaldo Carvalho de Melo <acme@kernel.org>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Link: http://lkml.kernel.org/r/150342256382.16595.986861478681783732.stgit@firesoul
Signed-off-by: Ingo Molnar <mingo@kernel.org>
---
kernel/events/internal.h | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/kernel/events/internal.h b/kernel/events/internal.h
index 5377c59..843e970 100644
--- a/kernel/events/internal.h
+++ b/kernel/events/internal.h
@@ -208,7 +208,7 @@ static inline int get_recursion_context(int *recursion)
{
int rctx;
- if (in_nmi())
+ if (unlikely(in_nmi()))
rctx = 3;
else if (in_irq())
rctx = 2;
^ permalink raw reply related [flat|nested] 9+ messages in thread
end of thread, other threads:[~2017-08-25 11:57 UTC | newest]
Thread overview: 9+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2017-08-22 14:40 [PATCH] trace: adjust code layout in get_recursion_context Jesper Dangaard Brouer
2017-08-22 15:14 ` Peter Zijlstra
2017-08-22 15:20 ` Peter Zijlstra
2017-08-22 17:00 ` Jesper Dangaard Brouer
2017-08-22 17:10 ` Jesper Dangaard Brouer
2017-08-22 17:22 ` [PATCH V2] " Jesper Dangaard Brouer
2017-08-25 11:54 ` [tip:perf/core] tracing, perf: Adjust code layout in get_recursion_context() tip-bot for Jesper Dangaard Brouer
2017-08-22 17:55 ` [PATCH] trace: adjust code layout in get_recursion_context Peter Zijlstra
2017-08-23 8:12 ` Ingo Molnar
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).