Tracepoint: fix documentation of RCU lockdep checks
Subsystems (2)
Mathieu Desnoyers
reviewerPATCH 1/3 Tracepoint: fix documentation of RCU lockdep checks
The documentation on top of __DECLARE_TRACE() does not match its implementation since the condition check has been added to the RCU lockdep checks. Update the documentation to match its implementation.
--- include/linux/tracepoint.h | 9 +++++---- 1 file changed, 5 insertions(+), 4 deletions(-)
include/linux/tracepoint.h +5 -4
@@ -167,10 +167,11 @@ extern void syscall_unregfunc(void);* structure. Force alignment to the same alignment as the section start. * * When lockdep is enabled, we make sure to always do the RCU portions of -* the tracepoint code, regardless of whether tracing is on or we match the -* condition. This lets us find RCU issues triggered with tracepoints even -* when this tracepoint is off. This code has no purpose other than poking -* RCU a bit. +* the tracepoint code, regardless of whether tracing is on. However, +* don't check if the condition is false, due to interaction with idle +* instrumentation. This lets us find RCU issues triggered with tracepoints +* even when this tracepoint is off. This code has no purpose other than +* poking RCU a bit. */ #define __DECLARE_TRACE (name, proto, args, cond, data_proto, data_args) \ extern struct tracepoint __tracepoint_##name; \
Signature · 1 line
2.1.4
Mathieu Desnoyers
reviewerPATCH 2/3 ftrace: add ftrace-buffer option
In order to use ftrace tracers to generate tracepoints without doing tracing to its own hardcoded ring buffers, add a ftrace-buffer option (default: 1). When set to 0, it disables tracing into the ftrace hardcoded buffers.
--- kernel/trace/trace.c | 4 +++- kernel/trace/trace.h | 1 + kernel/trace/trace_irqsoff.c | 19 +++++++++++-------- 3 files changed, 15 insertions(+), 9 deletions(-)
kernel/trace/trace.c +3 -1
@@ -472,7 +472,8 @@ static inline void trace_access_lock_init(void)unsigned long trace_flags =TRACE_ITER_PRINT_PARENT |TRACE_ITER_PRINTK |TRACE_ITER_ANNOTATE |TRACE_ITER_CONTEXT_INFO |TRACE_ITER_SLEEP_TIME |TRACE_ITER_GRAPH_TIME |TRACE_ITER_RECORD_CMD |TRACE_ITER_OVERWRITE | -TRACE_ITER_IRQ_INFO |TRACE_ITER_MARKERS |TRACE_ITER_FUNCTION ; +TRACE_ITER_IRQ_INFO |TRACE_ITER_MARKERS |TRACE_ITER_FUNCTION | +TRACE_ITER_FTRACE_BUFFER ;static void tracer_tracing_on (struct trace_array *tr) {
@@ -862,6 +863,7 @@ static const char *trace_options[] = {"irq-info" ,"markers" ,"function-trace" , +"ftrace-buffer" , NULL };
kernel/trace/trace.h +1 -0
@@ -930,6 +930,7 @@ enum trace_iterator_flags {TRACE_ITER_IRQ_INFO =0x800000 ,TRACE_ITER_MARKERS =0x1000000 ,TRACE_ITER_FUNCTION =0x2000000 , + TRACE_ITER_FTRACE_BUFFER = 0x4000000, };/*
kernel/trace/trace_irqsoff.c +11 -8
@@ -334,9 +334,11 @@ check_critical_timing(struct trace_array *tr,if (!report_latency (tr, delta)) goto out_unlock; -__trace_function (tr,CALLER_ADDR0 , parent_ip, flags, pc); -/* Skip 5 functions to get to the irq/preempt enable function */ -__trace_stack (tr, flags,5 , pc); +if (trace_flags &TRACE_ITER_FTRACE_BUFFER ) { +__trace_function (tr,CALLER_ADDR0 , parent_ip, flags, pc); +/* Skip 5 functions to get to the irq/preempt enable function */ +__trace_stack (tr, flags,5 , pc); + }if (data->critical_sequence != max_sequence) goto out_unlock;
@@ -356,7 +358,8 @@ out_unlock: out: data->critical_sequence = max_sequence; data->preempt_timestamp =ftrace_now (cpu); -__trace_function (tr,CALLER_ADDR0 , parent_ip, flags, pc); +if (trace_flags &TRACE_ITER_FTRACE_BUFFER ) +__trace_function (tr,CALLER_ADDR0 , parent_ip, flags, pc); } static inline void
@@ -387,9 +390,8 @@ start_critical_timing(unsigned long ip, unsigned long parent_ip) data->critical_start = parent_ip ? : ip;local_save_flags (flags); - -__trace_function (tr, ip, parent_ip, flags,preempt_count ()); - +if (trace_flags &TRACE_ITER_FTRACE_BUFFER ) +__trace_function (tr, ip, parent_ip, flags,preempt_count ());per_cpu (tracing_cpu, cpu) =1 ;atomic_dec (&data->disabled);
@@ -422,7 +424,8 @@ stop_critical_timing(unsigned long ip, unsigned long parent_ip)atomic_inc (&data->disabled);local_save_flags (flags); -__trace_function (tr, ip, parent_ip, flags,preempt_count ()); +if (trace_flags &TRACE_ITER_FTRACE_BUFFER ) +__trace_function (tr, ip, parent_ip, flags,preempt_count ());check_critical_timing (tr, data, parent_ip ? : ip, cpu); data->critical_start =0 ;atomic_dec (&data->disabled);
Signature · 1 line
2.1.4
Steven Rostedt
maintainerRe: [PATCH 2/3] ftrace: add ftrace-buffer option
Quoted 5 lines · On Mon, 2 Nov 2015 17:42:43 -0500 Mathieu Desnoyers…
In order to use ftrace tracers to generate tracepoints without doing tracing to its own hardcoded ring buffers, add a ftrace-buffer option (default: 1). When set to 0, it disables tracing into the ftrace hardcoded buffers.
This needs a more in depth change log. I have no idea why this is needed.
Also, the trace_options code have been redesigned, and this won't apply to it (see linux-next). Also, if this is only for irqsoff latency tracers, it should be a tracer option, not a global one.
-- Steve
Quoted 96 lines
Signed-off-by Mathieu Desnoyers <[email protected]>CC Thomas Gleixner <[email protected]>CC Steven Rostedt <[email protected]>CC Ingo Molnar <[email protected]>CC Peter Zijlstra <[email protected]>--- kernel/trace/trace.c | 4 +++- kernel/trace/trace.h | 1 + kernel/trace/trace_irqsoff.c | 19 +++++++++++-------- 3 files changed, 15 insertions(+), 9 deletions(-)kernel/trace/trace.c +3
-1@@ -472,7 +472,8 @@ static inline void trace_access_lock_init(void)unsigned long trace_flags =TRACE_ITER_PRINT_PARENT |TRACE_ITER_PRINTK |TRACE_ITER_ANNOTATE |TRACE_ITER_CONTEXT_INFO |TRACE_ITER_SLEEP_TIME |TRACE_ITER_GRAPH_TIME |TRACE_ITER_RECORD_CMD |TRACE_ITER_OVERWRITE | -TRACE_ITER_IRQ_INFO |TRACE_ITER_MARKERS |TRACE_ITER_FUNCTION ; +TRACE_ITER_IRQ_INFO |TRACE_ITER_MARKERS |TRACE_ITER_FUNCTION | +TRACE_ITER_FTRACE_BUFFER ;static void tracer_tracing_on (struct trace_array *tr) {@@ -862,6 +863,7 @@ static const char *trace_options[] = {"irq-info" ,"markers" ,"function-trace" , +"ftrace-buffer" , NULL };kernel/trace/trace.h +1
-0@@ -930,6 +930,7 @@ enum trace_iterator_flags {TRACE_ITER_IRQ_INFO =0x800000 ,TRACE_ITER_MARKERS =0x1000000 ,TRACE_ITER_FUNCTION =0x2000000 , + TRACE_ITER_FTRACE_BUFFER = 0x4000000, };/* kernel/trace/trace_irqsoff.c +11
-8@@ -334,9 +334,11 @@ check_critical_timing(struct trace_array *tr,if (!report_latency (tr, delta)) goto out_unlock; -__trace_function (tr,CALLER_ADDR0 , parent_ip, flags, pc); -/* Skip 5 functions to get to the irq/preempt enable function */ -__trace_stack (tr, flags,5 , pc); +if (trace_flags &TRACE_ITER_FTRACE_BUFFER ) { +__trace_function (tr,CALLER_ADDR0 , parent_ip, flags, pc); +/* Skip 5 functions to get to the irq/preempt enable function */ +__trace_stack (tr, flags,5 , pc); + }if (data->critical_sequence != max_sequence) goto out_unlock;@@ -356,7 +358,8 @@ out_unlock: out: data->critical_sequence = max_sequence; data->preempt_timestamp =ftrace_now (cpu); -__trace_function (tr,CALLER_ADDR0 , parent_ip, flags, pc); +if (trace_flags &TRACE_ITER_FTRACE_BUFFER ) +__trace_function (tr,CALLER_ADDR0 , parent_ip, flags, pc); } static inline void@@ -387,9 +390,8 @@ start_critical_timing(unsigned long ip, unsigned long parent_ip) data->critical_start = parent_ip ? : ip;local_save_flags (flags); - -__trace_function (tr, ip, parent_ip, flags,preempt_count ()); - +if (trace_flags &TRACE_ITER_FTRACE_BUFFER ) +__trace_function (tr, ip, parent_ip, flags,preempt_count ());per_cpu (tracing_cpu, cpu) =1 ;atomic_dec (&data->disabled);@@ -422,7 +424,8 @@ stop_critical_timing(unsigned long ip, unsigned long parent_ip)atomic_inc (&data->disabled);local_save_flags (flags); -__trace_function (tr, ip, parent_ip, flags,preempt_count ()); +if (trace_flags &TRACE_ITER_FTRACE_BUFFER ) +__trace_function (tr, ip, parent_ip, flags,preempt_count ());check_critical_timing (tr, data, parent_ip ? : ip, cpu); data->critical_start =0 ;atomic_dec (&data->disabled);
Mathieu Desnoyers
reviewerRe: [PATCH 2/3] ftrace: add ftrace-buffer option
Quoted 11 lines · ----- On Nov 2, 2015, at 5:56 PM, rostedt [email protected] wrote:
Quoted 5 lines · On Mon, 2 Nov 2015 17:42:43 -0500 Mathieu Desnoyers…
In order to use ftrace tracers to generate tracepoints without doing tracing to its own hardcoded ring buffers, add a ftrace-buffer option (default: 1). When set to 0, it disables tracing into the ftrace hardcoded buffers.
This needs a more in depth change log. I have no idea why this is needed.
I can do an updated change log. This is needed for the next patch in this series, which adds tracepoints in the ftrace irqs and preempt off tracer, so other tracers such as Perf and LTTng can use them.
Quoted 3 lines
Also, the
trace_optionscode have been redesigned, and this won't apply to it (see linux-next).
I can rebase my work on top of linux-next if you are OK with the general idea.
Quoted 2 lines
Also, if this is only for irqsoff latency tracers, it should be a tracer option, not a global one.
I've currently only done the irqsoff latency tracer part, but I'm wondering whether:
1- we should make it a irqsoff latency tracer option, 2- we should keep it as a global ftrace option, and apply it to
other ftrace tracers as well,
Thoughts ?
Thanks,
Mathieu
Quoted 102 lines
-- Steve
Quoted 98 lines
Signed-off-by Mathieu Desnoyers <[email protected]>CC Thomas Gleixner <[email protected]>CC Steven Rostedt <[email protected]>CC Ingo Molnar <[email protected]>CC Peter Zijlstra <[email protected]>--- kernel/trace/trace.c | 4 +++- kernel/trace/trace.h | 1 + kernel/trace/trace_irqsoff.c | 19 +++++++++++-------- 3 files changed, 15 insertions(+), 9 deletions(-)kernel/trace/trace.c +3
-1@@ -472,7 +472,8 @@ static inline void trace_access_lock_init(void)unsigned long trace_flags =TRACE_ITER_PRINT_PARENT |TRACE_ITER_PRINTK |TRACE_ITER_ANNOTATE |TRACE_ITER_CONTEXT_INFO |TRACE_ITER_SLEEP_TIME |TRACE_ITER_GRAPH_TIME |TRACE_ITER_RECORD_CMD |TRACE_ITER_OVERWRITE | -TRACE_ITER_IRQ_INFO |TRACE_ITER_MARKERS |TRACE_ITER_FUNCTION ; +TRACE_ITER_IRQ_INFO |TRACE_ITER_MARKERS |TRACE_ITER_FUNCTION | +TRACE_ITER_FTRACE_BUFFER ;static void tracer_tracing_on (struct trace_array *tr) {@@ -862,6 +863,7 @@ static const char *trace_options[] = {"irq-info" ,"markers" ,"function-trace" , +"ftrace-buffer" , NULL };kernel/trace/trace.h +1
-0@@ -930,6 +930,7 @@ enum trace_iterator_flags {TRACE_ITER_IRQ_INFO =0x800000 ,TRACE_ITER_MARKERS =0x1000000 ,TRACE_ITER_FUNCTION =0x2000000 , + TRACE_ITER_FTRACE_BUFFER = 0x4000000, };/* kernel/trace/trace_irqsoff.c +9
-7@@ -334,9 +334,11 @@ check_critical_timing(struct trace_array *tr,if (!report_latency (tr, delta)) goto out_unlock; -__trace_function (tr,CALLER_ADDR0 , parent_ip, flags, pc); -/* Skip 5 functions to get to the irq/preempt enable function */ -__trace_stack (tr, flags,5 , pc); +if (trace_flags &TRACE_ITER_FTRACE_BUFFER ) { +__trace_function (tr,CALLER_ADDR0 , parent_ip, flags, pc); +/* Skip 5 functions to get to the irq/preempt enable function */ +__trace_stack (tr, flags,5 , pc); + }if (data->critical_sequence != max_sequence) goto out_unlock;@@ -356,7 +358,8 @@ out_unlock: out: data->critical_sequence = max_sequence; data->preempt_timestamp =ftrace_now (cpu); -__trace_function (tr,CALLER_ADDR0 , parent_ip, flags, pc); +if (trace_flags &TRACE_ITER_FTRACE_BUFFER ) +__trace_function (tr,CALLER_ADDR0 , parent_ip, flags, pc); } static inline void@@ -387,9 +390,8 @@ start_critical_timing(unsigned long ip, unsigned long parent_ip) data->critical_start = parent_ip ? : ip;local_save_flags (flags); - -__trace_function (tr, ip, parent_ip, flags,preempt_count ()); - +if (trace_flags &TRACE_ITER_FTRACE_BUFFER ) +__trace_function (tr, ip, parent_ip, flags,preempt_count ());per_cpu (tracing_cpu, cpu) =1 ;kernel/trace/trace_irqsoff.c +2
-1atomic_dec (&data->disabled);@@ -422,7 +424,8 @@ stop_critical_timing(unsigned long ip, unsigned long parent_ip) atomic_inc(&data->disabled);local_save_flags (flags); -__trace_function (tr, ip, parent_ip, flags,preempt_count ()); +if (trace_flags &TRACE_ITER_FTRACE_BUFFER ) +__trace_function (tr, ip, parent_ip, flags,preempt_count ());check_critical_timing (tr, data, parent_ip ? : ip, cpu); data->critical_start =0 ;
atomic_dec(&data->disabled);
Signature · 3 lines
Mathieu Desnoyers EfficiOS Inc. http://www.efficios.com
Steven Rostedt
maintainerRe: [PATCH 2/3] ftrace: add ftrace-buffer option
Quoted 28 lines · On Mon, 2 Nov 2015 23:01:08 +0000 (UTC) Mathieu Desnoyers…
Quoted 11 lines · ----- On Nov 2, 2015, at 5:56 PM, rostedt [email protected] wrote:
Quoted 5 lines · On Mon, 2 Nov 2015 17:42:43 -0500 Mathieu Desnoyers…
In order to use ftrace tracers to generate tracepoints without doing tracing to its own hardcoded ring buffers, add a ftrace-buffer option (default: 1). When set to 0, it disables tracing into the ftrace hardcoded buffers.
This needs a more in depth change log. I have no idea why this is needed.
I can do an updated change log. This is needed for the next patch in this series, which adds tracepoints in the ftrace irqs and preempt off tracer, so other tracers such as Perf and LTTng can use them.
Quoted 3 lines
Also, the
trace_optionscode have been redesigned, and this won't apply to it (see linux-next).I can rebase my work on top of linux-next if you are OK with the general idea.
Quoted 2 lines
Also, if this is only for irqsoff latency tracers, it should be a tracer option, not a global one.
Quoted 10 lines
I've currently only done the irqsoff latency tracer part, but I'm wondering whether:
1- we should make it a irqsoff latency tracer option, 2- we should keep it as a global ftrace option, and apply it to
other ftrace tracers as well,Thoughts ?
Tomorrow I'll take a deeper look at patch 3 to try to understand.
-- Steve
Steven Rostedt
maintainerRe: [PATCH 2/3] ftrace: add ftrace-buffer option
In reply to Mathieu DesnoyersQuoted 5 lines · On Mon, 2 Nov 2015 17:42:43 -0500 Mathieu Desnoyers…
In order to use ftrace tracers to generate tracepoints without doing tracing to its own hardcoded ring buffers, add a ftrace-buffer option (default: 1). When set to 0, it disables tracing into the ftrace hardcoded buffers.
This should be a tracer specific option. And it shouldn't be called "ftrace-buffer" as that is a very confusing name.
Perhaps I can add an option that will prevent the snapshots as well. Perhaps an option called "quiet", and a "no_snapshot" one too. I can create the no snapshot one. But this should be converted to "quiet" and added as a tracer specific option.
-- Steve
Quoted 10 lines
Signed-off-by Mathieu Desnoyers <[email protected]>CC Thomas Gleixner <[email protected]>CC Steven Rostedt <[email protected]>CC Ingo Molnar <[email protected]>CC Peter Zijlstra <[email protected]>--- kernel/trace/trace.c | 4 +++- kernel/trace/trace.h | 1 + kernel/trace/trace_irqsoff.c | 19 +++++++++++-------- 3 files changed, 15 insertions(+), 9 deletions(-)
Mathieu Desnoyers
reviewerRe: [PATCH 2/3] ftrace: add ftrace-buffer option
Quoted 9 lines · ----- On Nov 23, 2015, at 4:46 PM, rostedt [email protected] wrote:
Quoted 5 lines · On Mon, 2 Nov 2015 17:42:43 -0500 Mathieu Desnoyers…
In order to use ftrace tracers to generate tracepoints without doing tracing to its own hardcoded ring buffers, add a ftrace-buffer option (default: 1). When set to 0, it disables tracing into the ftrace hardcoded buffers.
Quoted 7 lines
This should be a tracer specific option. And it shouldn't be called "ftrace-buffer" as that is a very confusing name.
Perhaps I can add an option that will prevent the snapshots as well. Perhaps an option called "quiet", and a "
no_snapshot" one too. I can create the no snapshot one. But this should be converted to "quiet" and added as a tracer specific option.
Sure, I can add the "quiet" option to the irqsoff/irqspreemptoff tracers if that's the route you recommend,
Thanks!
Mathieu
Quoted 14 lines
-- Steve
Quoted 10 lines
Signed-off-by Mathieu Desnoyers <[email protected]>CC Thomas Gleixner <[email protected]>CC Steven Rostedt <[email protected]>CC Ingo Molnar <[email protected]>CC Peter Zijlstra <[email protected]>--- kernel/trace/trace.c | 4 +++- kernel/trace/trace.h | 1 + kernel/trace/trace_irqsoff.c | 19 +++++++++++-------- 3 files changed, 15 insertions(+), 9 deletions(-)
Signature · 3 lines
Mathieu Desnoyers EfficiOS Inc. http://www.efficios.com
Mathieu Desnoyers
reviewerPATCH v2 3/3 trace: emit tracepoint in preempt and irqs off tracer
In reply to Mathieu DesnoyersAllow Ftrace, Perf, and LTTng to hook on the irqsoff tracer and trace irqsoff and preemptoff critical sections.
How to setup the irqsoff tracer for this use-case (needed for use with Ftrace, Perf, or LTTng): if the tracing_thresh value is set, all durations beyond the specified threshold will emit a "core_critical_timing_hit" tracepoint. If unset, then the max duration will be tracked. The current max duration can be reset using the tracing_max_latency file.
echo 0 > /sys/kernel/debug/tracing/options/function-trace # don't need function tracing echo 0 > /sys/kernel/debug/tracing/options/ftrace-buffer
echo preemptirqsoff > /sys/kernel/debug/tracing/current_tracer
echo 2 > /sys/kernel/debug/tracing/tracing_thresh # set threshold (e.g. 2 µs)
echo 1 > /sys/kernel/debug/tracing/tracing_on
The following two commands can also be useful to tweak the irqsoff tracer:
echo 0 > /sys/kernel/debug/tracing/tracing_max_latency # reset max latency
echo 0 > /sys/kernel/debug/tracing/tracing_thresh # disable threshold
*** An example usage of Ftrace hooking on this tracepoint: ***
echo 1 > /sys/kernel/debug/tracing/events/core/core_critical_timing_hit/enable cat /sys/kernel/debug/tracing/trace_pipe
[...]
bash-2380 19d..3 52794942us : core_critical_timing_hit: ip=0xffffffff811021de parent_ip=0xffffffff818d1b17 delta_ns=3402 irqs_disabled=0 preempt_disabled=1 in_softirq=0 in_irq=0 in_nmi=0
*** An example usage of Perf hooking on this tracepoint: ***
perf record -g -e core:core_critical_timing_hit -c 1 sleep 30 perf report -g
+ 100.00% 100.00% sleep [kernel.vmlinux] [k] check_critical_timing + 92.75% 0.00% sleep [kernel.vmlinux] [k] preempt_count_sub + 92.75% 0.00% sleep [kernel.vmlinux] [k] trace_preempt_on + 65.22% 0.00% sleep [kernel.vmlinux] [k] page_fault + 65.22% 0.00% sleep [kernel.vmlinux] [k] do_page_fault + 63.77% 0.00% sleep [kernel.vmlinux] [k] __do_page_fault + 63.77% 0.00% sleep [kernel.vmlinux] [k] handle_mm_fault + 56.52% 0.00% sleep [kernel.vmlinux] [k] _raw_spin_unlock + 30.43% 0.00% sleep ld-2.19.so [.] _dl_sysdep_start
[...]
*** An example usage of LTTng hooking on this tracepoint: ***
(as root)
lttng-sessiond -d
(then, as root or tracing group)
lttng create --live
lttng enable-event -k core_critical_timing_hit
lttng start
lttng view
[...]
[17:34:17.090669588] (+1.569689755) compudjdev core_critical_timing_hit: { cpu_id = 3 }, { ip = 0xFFFFFFFF81102135, parent_ip = 0xFFFFFFFFA0798E13, delta_ns = 252994, irqs_disabled = 0, preempt_disabled = 1, in_softirq = 0, in_irq = 0, in_nmi = 0 }
Tracepoints core_critical_timing_start and core_critical_timing_stop are also available. They trigger at a much higher rate than core_critical_timing_hit.
---
Changes since v1:
- Remove unneeded tracepoint jump label bypass.
- Enhance TP_printk() formatting.
- Use new ftrace-buffer option in changelog.
- Add core_critical_timing_start/stop tracepoints.
--- include/trace/events/core.h | 96 ++++++++++++++++++++++++++++++++++++++++++++ kernel/trace/trace_irqsoff.c | 17 ++++++-- 2 files changed, 109 insertions(+), 4 deletions(-) create mode 100644 include/trace/events/core.h
include/trace/events/core.h new file +96 -0
@@ -0,0 +1,96 @@ +#undef TRACE_SYSTEM +#define TRACE_SYSTEM core + +#if !defined(_TRACE_CORE_H) || defined(TRACE_HEADER_MULTI_READ ) +#define _TRACE_CORE_H + +#include <linux/tracepoint.h> +#include <linux/irqflags.h> +#include <linux/preempt.h> + +/* +* Tracepoint for critical timings max/threshold hit. +*/ +TRACE_EVENT (core_critical_timing_hit, + + TP_PROTO(unsigned long ip, unsigned long parent_ip, + unsigned long flags, int preempt_cnt, + cycles_t delta_ns), + +TP_ARGS (ip, parent_ip, flags, preempt_cnt, delta_ns), + +TP_STRUCT__entry ( + __field( unsigned long, ip ) + __field( unsigned long, parent_ip ) + __field( cycles_t, delta_ns ) + __field( unsigned long, flags ) +__field ( int, preempt_cnt ) + ), + +TP_fast_assign ( + __entry->ip = ip; + __entry->parent_ip = parent_ip; + __entry->delta_ns = delta_ns; + __entry->flags = flags; + __entry->preempt_cnt = preempt_cnt; + ), + +TP_printk ("ip=0x%lx parent_ip=0x%lx delta_ns=%llu irqs_disabled=%u preempt_disabled=%u in_softirq=%u in_irq=%u in_nmi=%u" , + __entry->ip, __entry->parent_ip, + (unsigned long long ) __entry->delta_ns, + !!raw_irqs_disabled_flags (__entry->flags), + !!(__entry->preempt_cnt &PREEMPT_MASK ), + !!(__entry->preempt_cnt &SOFTIRQ_MASK ), + !!(__entry->preempt_cnt &HARDIRQ_MASK ), + !!(__entry->preempt_cnt &NMI_MASK )) +); + +/* +* Tracepoint for critical timings start/stop. +*/ +DECLARE_EVENT_CLASS (core_critical_timing, + + TP_PROTO(unsigned long ip, unsigned long parent_ip, + unsigned long flags, int preempt_cnt), + +TP_ARGS (ip, parent_ip, flags, preempt_cnt), + +TP_STRUCT__entry ( + __field( unsigned long, ip ) + __field( unsigned long, parent_ip ) + __field( unsigned long, flags ) +__field ( int, preempt_cnt ) + ), + +TP_fast_assign ( + __entry->ip = ip; + __entry->parent_ip = parent_ip; + __entry->flags = flags; + __entry->preempt_cnt = preempt_cnt; + ), + +TP_printk ("ip=0x%lx parent_ip=0x%lx irqs_disabled=%u preempt_disabled=%u in_softirq=%u in_irq=%u in_nmi=%u" , + __entry->ip, __entry->parent_ip, + !!raw_irqs_disabled_flags (__entry->flags), + !!(__entry->preempt_cnt &PREEMPT_MASK ), + !!(__entry->preempt_cnt &SOFTIRQ_MASK ), + !!(__entry->preempt_cnt &HARDIRQ_MASK ), + !!(__entry->preempt_cnt &NMI_MASK )) +); + +DEFINE_EVENT (core_critical_timing, core_critical_timing_start, + TP_PROTO(unsigned long ip, unsigned long parent_ip, + unsigned long flags, int preempt_cnt), +TP_ARGS (ip, parent_ip, flags, preempt_cnt) +); + +DEFINE_EVENT (core_critical_timing, core_critical_timing_stop, + TP_PROTO(unsigned long ip, unsigned long parent_ip, + unsigned long flags, int preempt_cnt), +TP_ARGS (ip, parent_ip, flags, preempt_cnt) +); + +#endif /* _TRACE_CORE_H */ + +/* This part must be outside protection */ +#include <trace/define_trace.h>
kernel/trace/trace_irqsoff.c +13 -4
@@ -14,6 +14,9 @@#include <linux/module.h> #include <linux/ftrace.h> +#define CREATE_TRACE_POINTS +#include <trace/events/core.h> +#include "trace.h" static struct trace_array *irqsoff_trace__read_mostly ;
@@ -339,6 +342,8 @@ check_critical_timing(struct trace_array *tr,/* Skip 5 functions to get to the irq/preempt enable function */ __trace_stack (tr, flags,5 , pc); } +trace_core_critical_timing_hit (CALLER_ADDR0 , parent_ip, flags, pc, + delta);if (data->critical_sequence != max_sequence) goto out_unlock;
@@ -365,7 +370,7 @@ out:static inline void start_critical_timing (unsigned long ip,unsigned long parent_ip) { -int cpu; +int cpu, pc;struct trace_array *tr = irqsoff_trace;struct trace_array_cpu *data;unsigned long flags;
@@ -390,8 +395,10 @@ start_critical_timing(unsigned long ip, unsigned long parent_ip) data->critical_start = parent_ip ? : ip;local_save_flags (flags); + pc =preempt_count ();if (trace_flags &TRACE_ITER_FTRACE_BUFFER ) -__trace_function (tr, ip, parent_ip, flags,preempt_count ()); +__trace_function (tr, ip, parent_ip, flags, pc); +trace_core_critical_timing_start (ip, parent_ip, flags, pc);per_cpu (tracing_cpu, cpu) =1 ;atomic_dec (&data->disabled);
@@ -400,7 +407,7 @@ start_critical_timing(unsigned long ip, unsigned long parent_ip)static inline void stop_critical_timing (unsigned long ip,unsigned long parent_ip) { -int cpu; +int cpu, pc;struct trace_array *tr = irqsoff_trace;struct trace_array_cpu *data;unsigned long flags;
@@ -424,8 +431,10 @@ stop_critical_timing(unsigned long ip, unsigned long parent_ip)atomic_inc (&data->disabled);local_save_flags (flags); + pc =preempt_count ();if (trace_flags &TRACE_ITER_FTRACE_BUFFER ) -__trace_function (tr, ip, parent_ip, flags,preempt_count ()); +__trace_function (tr, ip, parent_ip, flags, pc); +trace_core_critical_timing_stop (ip, parent_ip, flags, pc);check_critical_timing (tr, data, parent_ip ? : ip, cpu); data->critical_start =0 ;atomic_dec (&data->disabled);
Signature · 1 line
2.1.4
Mathieu Desnoyers
reviewerRe: [PATCH v2 3/3] trace: emit tracepoint in preempt and irqs off tracer
Quoted 2 lines · ----- On Nov 2, 2015, at 5:42 PM, Mathieu Desnoyers…
Allow Ftrace, Perf, and LTTng to hook on the irqsoff tracer and trace irqsoff and preemptoff critical sections.
Hi Steven,
I guess you ended up being busy lately. Your feedback on this patchset is still very welcome.
Thanks,
Mathieu
Quoted 257 lines
How to setup the irqsoff tracer for this use-case (needed for use with Ftrace, Perf, or LTTng): if the
tracing_threshvalue is set, all durations beyond the specified threshold will emit a "core_critical_timing_hit" tracepoint. If unset, then the max duration will be tracked. The current max duration can be reset using thetracing_max_latencyfile.echo 0 >
/sys/kernel/debug/tracing/options/function-trace# don't need function tracing
echo 0 >/sys/kernel/debug/tracing/options/ftrace-buffer
echo preemptirqsoff >/sys/kernel/debug/tracing/current_tracer
echo 2 >/sys/kernel/debug/tracing/tracing_thresh# set threshold (e.g. 2
µs)
echo 1 >/sys/kernel/debug/tracing/tracing_onThe following two commands can also be useful to tweak the irqsoff tracer:
echo 0 >/sys/kernel/debug/tracing/tracing_max_latency# reset max latency
echo 0 >/sys/kernel/debug/tracing/tracing_thresh# disable threshold*** An example usage of Ftrace hooking on this tracepoint: ***
echo 1 >
/sys/kernel/debug/tracing/events/core/core_critical_timing_hit/enablecat/sys/kernel/debug/tracing/trace_pipe
[...]
bash-2380 19d..3 52794942us :core_critical_timing_hit: ip=0xffffffff811021de
parent_ip=0xffffffff818d1b17 delta_ns=3402 irqs_disabled=0 preempt_disabled=1 in_softirq=0 in_irq=0 in_nmi=0*** An example usage of Perf hooking on this tracepoint: ***
perf record -g -e core:core_critical_timing_hit -c 1 sleep 30 perf report -g
(quoted diff) +9
-0+ 100.00% 100.00% sleep [kernel.vmlinux] [k] check_critical_timing + 92.75% 0.00% sleep [kernel.vmlinux] [k] preempt_count_sub + 92.75% 0.00% sleep [kernel.vmlinux] [k] trace_preempt_on + 65.22% 0.00% sleep [kernel.vmlinux] [k] page_fault + 65.22% 0.00% sleep [kernel.vmlinux] [k] do_page_fault + 63.77% 0.00% sleep [kernel.vmlinux] [k] __do_page_fault + 63.77% 0.00% sleep [kernel.vmlinux] [k] handle_mm_fault + 56.52% 0.00% sleep [kernel.vmlinux] [k] _raw_spin_unlock + 30.43% 0.00% sleep ld-2.19.so [.] _dl_sysdep_start*** An example usage of LTTng hooking on this tracepoint: ***
(as root)
lttng-sessiond -d(then, as root or tracing group)
lttng create --live
lttng enable-event -kcore_critical_timing_hit
lttng start
lttng view
[...]
[17:34:17.090669588] (+1.569689755) compudjdevcore_critical_timing_hit: {cpu_id= 3 }, { ip =0xFFFFFFFF81102135,parent_ip=0xFFFFFFFFA0798E13,delta_ns= 252994,irqs_disabled= 0,preempt_disabled= 1,in_softirq= 0,in_irq= 0,in_nmi= 0 }Tracepoints
core_critical_timing_startandcore_critical_timing_stopare also available. They trigger at a much higher rate thancore_critical_timing_hit.Signed-off-by Mathieu Desnoyers <[email protected]>CC Thomas Gleixner <[email protected]>CC Steven Rostedt <[email protected]>CC Ingo Molnar <[email protected]>CC Peter Zijlstra <[email protected]>---
Changes since v1:
- Remove unneeded tracepoint jump label bypass.
- EnhanceTP_printk()formatting.
- Use new ftrace-buffer option in changelog.
- Add core_critical_timing_start/stop tracepoints.
---include/trace/events/core.h| 96 ++++++++++++++++++++++++++++++++++++++++++++kernel/trace/trace_irqsoff.c| 17 ++++++--2 files changed, 109 insertions(+), 4 deletions(-)create mode 100644
include/trace/events/core.hinclude/trace/events/core.h new file +38
-0@@ -0,0 +1,96 @@ +#undef TRACE_SYSTEM +#define TRACE_SYSTEM core + +#if !defined(_TRACE_CORE_H) || defined(TRACE_HEADER_MULTI_READ ) +#define _TRACE_CORE_H + +#include <linux/tracepoint.h> +#include <linux/irqflags.h> +#include <linux/preempt.h> + +/* +* Tracepoint for critical timings max/threshold hit. +*/ +TRACE_EVENT(core_critical_timing_hit, + + TP_PROTO(unsigned long ip, unsigned long parent_ip, + unsigned long flags, int preempt_cnt, + cycles_t delta_ns), + + TP_ARGS(ip, parent_ip, flags, preempt_cnt, delta_ns), + + TP_STRUCT__entry( + __field( unsigned long, ip ) + __field( unsigned long, parent_ip ) + __field( cycles_t, delta_ns ) + __field( unsigned long, flags ) + __field( int, preempt_cnt ) + ), + + TP_fast_assign( + __entry->ip = ip; + __entry->parent_ip = parent_ip; + __entry->delta_ns = delta_ns; + __entry->flags = flags; + __entry->preempt_cnt = preempt_cnt; + ), + + TP_printk("ip=0x%lx parent_ip=0x%lx delta_ns=%llu irqs_disabled=%upreempt_disabled=%u in_softirq=%u in_irq=%u in_nmi=%u",
include/trace/events/core.h new file +34
-0+ __entry->ip, __entry->parent_ip, + (unsigned long long ) __entry->delta_ns, + !!raw_irqs_disabled_flags (__entry->flags), + !!(__entry->preempt_cnt &PREEMPT_MASK ), + !!(__entry->preempt_cnt &SOFTIRQ_MASK ), + !!(__entry->preempt_cnt &HARDIRQ_MASK ), + !!(__entry->preempt_cnt & NMI_MASK)) +); + +/* +* Tracepoint for critical timings start/stop. +*/ +DECLARE_EVENT_CLASS(core_critical_timing, + + TP_PROTO(unsigned long ip, unsigned long parent_ip, + unsigned long flags, int preempt_cnt), + + TP_ARGS(ip, parent_ip, flags, preempt_cnt), + + TP_STRUCT__entry( + __field( unsigned long, ip ) + __field( unsigned long, parent_ip ) + __field( unsigned long, flags ) + __field( int, preempt_cnt ) + ), + + TP_fast_assign( + __entry->ip = ip; + __entry->parent_ip = parent_ip; + __entry->flags = flags; + __entry->preempt_cnt = preempt_cnt; + ), + + TP_printk("ip=0x%lx parent_ip=0x%lx irqs_disabled=%u preempt_disabled=%uin_softirq=%u in_irq=%u in_nmi=%u",
include/trace/events/core.h new file +24
-0+ __entry->ip, __entry->parent_ip, + !!raw_irqs_disabled_flags (__entry->flags), + !!(__entry->preempt_cnt &PREEMPT_MASK ), + !!(__entry->preempt_cnt &SOFTIRQ_MASK ), + !!(__entry->preempt_cnt &HARDIRQ_MASK ), + !!(__entry->preempt_cnt & NMI_MASK)) +); + +DEFINE_EVENT (core_critical_timing, core_critical_timing_start, + TP_PROTO(unsigned long ip, unsigned long parent_ip, + unsigned long flags, int preempt_cnt), +TP_ARGS (ip, parent_ip, flags, preempt_cnt) +); + +DEFINE_EVENT (core_critical_timing, core_critical_timing_stop, + TP_PROTO(unsigned long ip, unsigned long parent_ip, + unsigned long flags, int preempt_cnt), +TP_ARGS (ip, parent_ip, flags, preempt_cnt) +); + +#endif /* _TRACE_CORE_H */ + +/* This part must be outside protection */ +#include <trace/define_trace.h> kernel/trace/trace_irqsoff.c +5
-0@@ -14,6 +14,9 @@#include <linux/module.h> #include <linux/ftrace.h> +#define CREATE_TRACE_POINTS +#include <trace/events/core.h> +#include "trace.h" static struct trace_array *irqsoff_trace__read_mostly ;@@ -339,6 +342,8 @@ check_critical_timing(struct trace_array *tr,/* Skip 5 functions to get to the irq/preempt enable function */ __trace_stack (tr, flags,5 , pc); } +trace_core_critical_timing_hit (CALLER_ADDR0 , parent_ip, flags, pc, + delta);if (data->critical_sequence != max_sequence) goto out_unlock;@@ -365,7 +370,7 @@ out:static inline void
start_critical_timing(unsigned long ip, unsigned long parent_ip){kernel/trace/trace_irqsoff.c +1
-1-int cpu; +int cpu, pc;struct trace_array *tr = irqsoff_trace; struct trace_array_cpu *data; unsigned long flags;kernel/trace/trace_irqsoff.c +3
-1@@ -390,8 +395,10 @@ start_critical_timing(unsigned long ip, unsigned long parent_ip) data->critical_start = parent_ip ? : ip;local_save_flags (flags); + pc =preempt_count ();if (trace_flags &TRACE_ITER_FTRACE_BUFFER ) -__trace_function (tr, ip, parent_ip, flags,preempt_count ()); +__trace_function (tr, ip, parent_ip, flags, pc); +trace_core_critical_timing_start (ip, parent_ip, flags, pc);per_cpu (tracing_cpu, cpu) =1 ;atomic_dec(&data->disabled);kernel/trace/trace_irqsoff.c +0
-0@@ -400,7 +407,7 @@ start_critical_timing(unsigned long ip, unsigned long parent_ip)static inline void
stop_critical_timing(unsigned long ip, unsigned long parent_ip){kernel/trace/trace_irqsoff.c +1
-1-int cpu; +int cpu, pc;struct trace_array *tr = irqsoff_trace; struct trace_array_cpu *data; unsigned long flags;kernel/trace/trace_irqsoff.c +3
-1@@ -424,8 +431,10 @@ stop_critical_timing(unsigned long ip, unsigned long parent_ip) atomic_inc(&data->disabled);local_save_flags (flags); + pc =preempt_count ();if (trace_flags &TRACE_ITER_FTRACE_BUFFER ) -__trace_function (tr, ip, parent_ip, flags,preempt_count ()); +__trace_function (tr, ip, parent_ip, flags, pc); +trace_core_critical_timing_stop (ip, parent_ip, flags, pc);check_critical_timing (tr, data, parent_ip ? : ip, cpu); data->critical_start =0 ;atomic_dec(&data->disabled);--
2.1.4
Signature · 3 lines
Mathieu Desnoyers EfficiOS Inc. http://www.efficios.com
Steven Rostedt
maintainerRe: [PATCH v2 3/3] trace: emit tracepoint in preempt and irqs off tracer
Quoted 9 lines · On Wed, 18 Nov 2015 14:39:31 +0000 (UTC) Mathieu Desnoyers…
Quoted 2 lines · ----- On Nov 2, 2015, at 5:42 PM, Mathieu Desnoyers…
Allow Ftrace, Perf, and LTTng to hook on the irqsoff tracer and trace irqsoff and preemptoff critical sections.
Hi Steven,
I guess you ended up being busy lately. Your feedback on this patchset is still very welcome.
Bah, I started looking at this three times and was pulled away each time. I still want to finish my review. I'm going to be traveling today so I don't want to guarantee anything. I'm driving this time so there's not going to be a "review it on the plane" possibility.
-- Steve
Quoted 3 lines
Thanks,
Re: [PATCH v2 3/3] trace: emit tracepoint in preempt and irqs off tracer
Quoted 4 lines · Steven Rostedt
Bah, I started looking at this three times and was pulled away each time. I still want to finish my review. I'm going to be traveling today so I don't want to guarantee anything. I'm driving this time so there's not going to be a "review it on the plane" possibility.
Get one of those self driving cars already!
Steven Rostedt
maintainerQuoted 5 lines · On Mon, 2 Nov 2015 17:42:42 -0500 Mathieu Desnoyers…
The documentation on top of
__DECLARE_TRACE()does not match its implementation since the condition check has been added to the RCU lockdep checks. Update the documentation to match its implementation.
Thanks, I can pull this now for 4.4.
-- Steve
Quoted 28 lines
Fixes a05d59a56733 "tracing: Add condition check to RCU lockdep checks"Signed-off-by Mathieu Desnoyers <[email protected]>CC Dave Hansen <[email protected]>CC Steven Rostedt <[email protected]>--- include/linux/tracepoint.h | 9 +++++---- 1 file changed, 5 insertions(+), 4 deletions(-)include/linux/tracepoint.h +5
-4@@ -167,10 +167,11 @@ extern void syscall_unregfunc(void);* structure. Force alignment to the same alignment as the section start. * * When lockdep is enabled, we make sure to always do the RCU portions of -* the tracepoint code, regardless of whether tracing is on or we match the -* condition. This lets us find RCU issues triggered with tracepoints even -* when this tracepoint is off. This code has no purpose other than poking -* RCU a bit. +* the tracepoint code, regardless of whether tracing is on. However, +* don't check if the condition is false, due to interaction with idle +* instrumentation. This lets us find RCU issues triggered with tracepoints +* even when this tracepoint is off. This code has no purpose other than +* poking RCU a bit. */ #define __DECLARE_TRACE (name, proto, args, cond, data_proto, data_args) \ extern struct tracepoint __tracepoint_##name; \