tracing: Make synthetic events a separate option

PATCH In review Tom Zanussi 3 participants 5 messages May 28, 2020 – May 29, 2020

Subsystems (2)

Kselftest Tracing

lore

Share & tools

lore

Tom Zanussi

author

2020-05-28 19:32 UTC #0 lore raw

v2 cover tracing: Make synthetic events a separate option

Hi Steve,

These are the same patches as v1, but after posting v1 I realized that if you applied them on top of the other patchset I posted awhile back ("[PATCH 0/3] tracing: histogram internals doc, hist_debug, and misc") [1], you wouldn't be able to apply these.

So assuming you apply the other patchset first, these should apply cleanly on top of those.

Thanks,

Tom

[1] https://lore.kernel.org/lkml/[email protected]/

v1 text:

Since synthetic events can now be used separately from hist triggers, it makes sense to actually separate them into different options. It also makes sense in terms of file bloat - trace_events_hist.c was getting very large and this makes it smaller and easier to follow.

Thanks,

Tom

The following changes since commit 77914c22b0ba493d9783c53bbfbc6087d6a7e1b1:

tracing: Add hist_debug trace event files for histogram debugging (2020-04-03 13:46:05 -0500)

are available in the Git repository at:

git://git.kernel.org/pub/scm/linux/kernel/git/zanussi/linux-trace.git ftrace/separate-synth-v2

Tom Zanussi (2):
tracing: Move synthetic events to a separate file selftests/ftrace: Distinguish between hist and synthetic event checks

 kernel/trace/Kconfig                          |   20 +-
 kernel/trace/Makefile                         |    1 +
 kernel/trace/trace_events_hist.c              | 2071 ++---------------
 kernel/trace/trace_events_synth.c             | 1789 ++++++++++++++
 kernel/trace/trace_synth.h                    |   36 +
 .../trigger-field-variable-support.tc         |    5 +
 .../trigger-inter-event-combined-hist.tc      |    5 +
 .../trigger-multi-actions-accept.tc           |    5 +
 .../trigger-onmatch-action-hist.tc            |    5 +
 .../trigger-onmatch-onmax-action-hist.tc      |    5 +
 .../inter-event/trigger-onmax-action-hist.tc  |    5 +
 .../trigger-snapshot-action-hist.tc           |    5 +
 .../inter-event/trigger-trace-action-hist.tc  |    5 +
 13 files changed, 2029 insertions(+), 1928 deletions(-)
 create mode 100644 kernel/trace/trace_events_synth.c
 create mode 100644 kernel/trace/trace_synth.h

Signature · 1 line

2.17.1

Tom Zanussi

author

2020-05-28 19:32 UTC #2 lore raw

PATCH v2 2/2 selftests/ftrace: Distinguish between hist and synthetic event checks

With synthetic events now a separate config item as a result of 'tracing: Move synthetic events to a separate file', tests that use both need to explicitly check for hist trigger support rather than relying on hist triggers to pull in synthetic events.

Add an additional hist trigger check to all the trigger tests that now require it, otherwise they'll fail if synthetic events but not hist triggers are enabled.

Signed-off-by Tom Zanussi <[email protected]>

---
 .../trigger/inter-event/trigger-field-variable-support.tc    | 5 +++++
 .../trigger/inter-event/trigger-inter-event-combined-hist.tc | 5 +++++
 .../trigger/inter-event/trigger-multi-actions-accept.tc      | 5 +++++
 .../trigger/inter-event/trigger-onmatch-action-hist.tc       | 5 +++++
 .../trigger/inter-event/trigger-onmatch-onmax-action-hist.tc | 5 +++++
 .../test.d/trigger/inter-event/trigger-onmax-action-hist.tc  | 5 +++++
 .../trigger/inter-event/trigger-snapshot-action-hist.tc      | 5 +++++
 .../test.d/trigger/inter-event/trigger-trace-action-hist.tc  | 5 +++++
 8 files changed, 40 insertions(+)

tools/testing/selftests/ftrace/test.d/trigger/inter-event/trigger-field-variable-support.tc +5 -0

index 77be6e1f6e7b..e232059a8ab2 100644
@@ -17,6 +17,11 @@ if [ ! -f synthetic_events ]; then
     exit_unsupported
 fi
 
+if [ ! -f events/sched/sched_process_fork/hist ]; then
+    echo "hist trigger is not supported"
+    exit_unsupported
+fi
+
 echo "Test field variable support"
 
 echo 'wakeup_latency u64 lat; pid_t pid; int prio; char comm[16]' > synthetic_events

tools/testing/selftests/ftrace/test.d/trigger/inter-event/trigger-inter-event-combined-hist.tc +5 -0

index f3eb8aacec0e..07cfcb8157b6 100644
@@ -17,6 +17,11 @@ if [ ! -f synthetic_events ]; then
     exit_unsupported
 fi
 
+if [ ! -f events/sched/sched_process_fork/hist ]; then
+    echo "hist trigger is not supported"
+    exit_unsupported
+fi
+
 echo "Test create synthetic event"
 
 echo 'waking_latency  u64 lat pid_t pid' > synthetic_events

tools/testing/selftests/ftrace/test.d/trigger/inter-event/trigger-multi-actions-accept.tc +5 -0

index d281f056f980..73e413c2ca26 100644
@@ -17,6 +17,11 @@ if [ ! -f synthetic_events ]; then
     exit_unsupported
 fi
 
+if [ ! -f events/sched/sched_process_fork/hist ]; then
+    echo "hist trigger is not supported"
+    exit_unsupported
+fi
+
 echo "Test multiple actions on hist trigger"
 echo 'wakeup_latency u64 lat; pid_t pid' >> synthetic_events
 TRIGGER1=events/sched/sched_wakeup/trigger

tools/testing/selftests/ftrace/test.d/trigger/inter-event/trigger-onmatch-action-hist.tc +5 -0

index a708f0e7858a..ebe0ad827f9f 100644
@@ -17,6 +17,11 @@ if [ ! -f synthetic_events ]; then
     exit_unsupported
 fi
 
+if [ ! -f events/sched/sched_process_fork/hist ]; then
+    echo "hist trigger is not supported"
+    exit_unsupported
+fi
+
 echo "Test create synthetic event"
 
 echo 'wakeup_latency  u64 lat pid_t pid char comm[16]' > synthetic_events

tools/testing/selftests/ftrace/test.d/trigger/inter-event/trigger-onmatch-onmax-action-hist.tc +5 -0

index dfce6932d8be..2a2ef767249e 100644
@@ -17,6 +17,11 @@ if [ ! -f synthetic_events ]; then
     exit_unsupported
 fi
 
+if [ ! -f events/sched/sched_process_fork/hist ]; then
+    echo "hist trigger is not supported"
+    exit_unsupported
+fi
+
 echo "Test create synthetic event"
 
 echo 'wakeup_latency  u64 lat pid_t pid char comm[16]' > synthetic_events

tools/testing/selftests/ftrace/test.d/trigger/inter-event/trigger-onmax-action-hist.tc +5 -0

index 0035995c2194..98d73bfb0296 100644
@@ -17,6 +17,11 @@ if [ ! -f synthetic_events ]; then
     exit_unsupported
 fi
 
+if [ ! -f events/sched/sched_process_fork/hist ]; then
+    echo "hist trigger is not supported"
+    exit_unsupported
+fi
+
 echo "Test create synthetic event"
 
 echo 'wakeup_latency  u64 lat pid_t pid char comm[16]' > synthetic_events

tools/testing/selftests/ftrace/test.d/trigger/inter-event/trigger-snapshot-action-hist.tc +5 -0

index f546c1b66a9b..01b01b9c4e07 100644
@@ -12,6 +12,11 @@ if [ ! -f set_event ]; then
     exit_unsupported
 fi
 
+if [ ! -f events/sched/sched_process_fork/hist ]; then
+    echo "hist trigger is not supported"
+    exit_unsupported
+fi
+
 if [ ! -f snapshot ]; then
     echo "snapshot is not supported"
     exit_unsupported

tools/testing/selftests/ftrace/test.d/trigger/inter-event/trigger-trace-action-hist.tc +5 -0

index 8021d60aafec..c3baa486aeb4 100644
@@ -17,6 +17,11 @@ if [ ! -f synthetic_events ]; then
     exit_unsupported
 fi
 
+if [ ! -f events/sched/sched_process_fork/hist ]; then
+    echo "hist trigger is not supported"
+    exit_unsupported
+fi
+
 grep -q "trace(<synthetic_event>" README || exit_unsupported # version issue
 
 echo "Test create synthetic event"

Signature · 1 line

2.17.1

Tom Zanussi

author

2020-05-28 19:32 UTC #1 lore raw

PATCH v2 1/2 tracing: Move synthetic events to a separate file

In reply to Tom Zanussi

With the addition of the in-kernel synthetic event API, synthetic events are no longer specifically tied to the histogram triggers.

The synthetic event code is also making trace_event_hist.c very bloated, so for those reasons, move it to a separate file, trace_events_synth.c, along with a new trace_synth.h header file.

Because synthetic events are now independent from hist triggers, add a new CONFIG_SYNTH_EVENTS config option, and have CONFIG_HIST_TRIGGERS select it, and have CONFIG_SYNTH_EVENT_GEN_TEST depend on it.

Signed-off-by Tom Zanussi <[email protected]>

---
 kernel/trace/Kconfig              |   20 +-
 kernel/trace/Makefile             |    1 +
 kernel/trace/trace_events_hist.c  | 2071 ++---------------------------
 kernel/trace/trace_events_synth.c | 1789 +++++++++++++++++++++++++
 kernel/trace/trace_synth.h        |   36 +
 5 files changed, 1989 insertions(+), 1928 deletions(-)
 create mode 100644 kernel/trace/trace_events_synth.c
 create mode 100644 kernel/trace/trace_synth.h

kernel/trace/Kconfig +19 -1

index 357f6304e86a..adfcfd02c868 100644
@@ -624,12 +624,30 @@ config TRACING_MAP
 	  generally used outside of that context, and is normally
 	  selected by tracers that use it.
 
+config SYNTH_EVENTS
+	bool "Synthetic trace events"
+	select TRACING
+	select DYNAMIC_EVENTS
+	default n
+	help
+	  Synthetic events are user-defined trace events that can be
+	  used to combine data from other trace events or in fact any
+	  data source.  Synthetic events can be generated indirectly
+	  via the trace() action of histogram triggers or directly
+	  by way of an in-kernel API.
+
+	  See Documentation/trace/events.rst or
+	  Documentation/trace/histogram.rst for details and examples.
+
+	  If in doubt, say N.
+
 config HIST_TRIGGERS
 	bool "Histogram triggers"
 	depends on ARCH_HAVE_NMI_SAFE_CMPXCHG
 	select TRACING_MAP
 	select TRACING
 	select DYNAMIC_EVENTS
+	select SYNTH_EVENTS
 	default n
 	help
 	  Hist triggers allow one or more arbitrary trace event fields

@@ -825,7 +843,7 @@ config PREEMPTIRQ_DELAY_TEST
 
 config SYNTH_EVENT_GEN_TEST
 	tristate "Test module for in-kernel synthetic event generation"
-	depends on HIST_TRIGGERS
+	depends on SYNTH_EVENTS
 	help
           This option creates a test module to check the base
           functionality of in-kernel synthetic event definition and

kernel/trace/Makefile +1 -0

index f9dcd19165fa..1d8aaa546412 100644
@@ -72,6 +72,7 @@ endif
 obj-$(CONFIG_EVENT_TRACING) += trace_events_filter.o
 obj-$(CONFIG_EVENT_TRACING) += trace_events_trigger.o
 obj-$(CONFIG_TRACE_EVENT_INJECT) += trace_events_inject.o
+obj-$(CONFIG_SYNTH_EVENTS) += trace_events_synth.o
 obj-$(CONFIG_HIST_TRIGGERS) += trace_events_hist.o
 obj-$(CONFIG_BPF_EVENTS) += bpf_trace.o
 obj-$(CONFIG_KPROBE_EVENTS) += trace_kprobe.o

kernel/trace/trace_events_hist.c +144 ~~-1927~~

2,145 lines not on this page · raw message

kernel/trace/trace_events_synth.c new file +1789 -0

810 lines not on this page · raw message

kernel/trace/trace_synth.h new file +36 -0

new file mode 100644
index 000000000000..ac35c45207c4

Lines of this file are not shown. View the raw message.

The rest of this diff is not shown. View the raw message.

Signature · 1 line

2.17.1

Part of the diff is not shown here. View the raw message.

Masami Hiramatsu

maintainer Acked-by on v2 2/2

2020-05-29 14:38 UTC #3 lore raw

Re: [PATCH v2 2/2] selftests/ftrace: Distinguish between hist and synthetic event checks

In reply to Tom Zanussi

Hi Tom,

Quoted 8 lines · On Thu, 28 May 2020 14:32:38 -0500 Tom Zanussi <[email protected]> wrote:

With synthetic events now a separate config item as a result of 'tracing: Move synthetic events to a separate file', tests that use both need to explicitly check for hist trigger support rather than relying on hist triggers to pull in synthetic events.
Add an additional hist trigger check to all the trigger tests that now require it, otherwise they'll fail if synthetic events but not hist triggers are enabled.

OK, this looks good to me. And if you don't want to repeat it, you can also put the check function into the test.d/functions.

Acked-by Masami Hiramatsu <[email protected]>

Thank you,

Quoted 144 lines

Signed-off-by Tom Zanussi <[email protected]>

---
 .../trigger/inter-event/trigger-field-variable-support.tc    | 5 +++++
 .../trigger/inter-event/trigger-inter-event-combined-hist.tc | 5 +++++
 .../trigger/inter-event/trigger-multi-actions-accept.tc      | 5 +++++
 .../trigger/inter-event/trigger-onmatch-action-hist.tc       | 5 +++++
 .../trigger/inter-event/trigger-onmatch-onmax-action-hist.tc | 5 +++++
 .../test.d/trigger/inter-event/trigger-onmax-action-hist.tc  | 5 +++++
 .../trigger/inter-event/trigger-snapshot-action-hist.tc      | 5 +++++
 .../test.d/trigger/inter-event/trigger-trace-action-hist.tc  | 5 +++++
 8 files changed, 40 insertions(+)

tools/testing/selftests/ftrace/test.d/trigger/inter-event/trigger-field-variable-support.tc +5 -0

index 77be6e1f6e7b..e232059a8ab2 100644
@@ -17,6 +17,11 @@ if [ ! -f synthetic_events ]; then
     exit_unsupported
 fi
 
+if [ ! -f events/sched/sched_process_fork/hist ]; then
+    echo "hist trigger is not supported"
+    exit_unsupported
+fi
+
 echo "Test field variable support"
 
 echo 'wakeup_latency u64 lat; pid_t pid; int prio; char comm[16]' > synthetic_events

tools/testing/selftests/ftrace/test.d/trigger/inter-event/trigger-inter-event-combined-hist.tc +5 -0

index f3eb8aacec0e..07cfcb8157b6 100644
@@ -17,6 +17,11 @@ if [ ! -f synthetic_events ]; then
     exit_unsupported
 fi
 
+if [ ! -f events/sched/sched_process_fork/hist ]; then
+    echo "hist trigger is not supported"
+    exit_unsupported
+fi
+
 echo "Test create synthetic event"
 
 echo 'waking_latency  u64 lat pid_t pid' > synthetic_events

tools/testing/selftests/ftrace/test.d/trigger/inter-event/trigger-multi-actions-accept.tc +5 -0

index d281f056f980..73e413c2ca26 100644
@@ -17,6 +17,11 @@ if [ ! -f synthetic_events ]; then
     exit_unsupported
 fi
 
+if [ ! -f events/sched/sched_process_fork/hist ]; then
+    echo "hist trigger is not supported"
+    exit_unsupported
+fi
+
 echo "Test multiple actions on hist trigger"
 echo 'wakeup_latency u64 lat; pid_t pid' >> synthetic_events
 TRIGGER1=events/sched/sched_wakeup/trigger

tools/testing/selftests/ftrace/test.d/trigger/inter-event/trigger-onmatch-action-hist.tc +5 -0

index a708f0e7858a..ebe0ad827f9f 100644
@@ -17,6 +17,11 @@ if [ ! -f synthetic_events ]; then
     exit_unsupported
 fi
 
+if [ ! -f events/sched/sched_process_fork/hist ]; then
+    echo "hist trigger is not supported"
+    exit_unsupported
+fi
+
 echo "Test create synthetic event"
 
 echo 'wakeup_latency  u64 lat pid_t pid char comm[16]' > synthetic_events

tools/testing/selftests/ftrace/test.d/trigger/inter-event/trigger-onmatch-onmax-action-hist.tc +5 -0

index dfce6932d8be..2a2ef767249e 100644
@@ -17,6 +17,11 @@ if [ ! -f synthetic_events ]; then
     exit_unsupported
 fi
 
+if [ ! -f events/sched/sched_process_fork/hist ]; then
+    echo "hist trigger is not supported"
+    exit_unsupported
+fi
+
 echo "Test create synthetic event"
 
 echo 'wakeup_latency  u64 lat pid_t pid char comm[16]' > synthetic_events

tools/testing/selftests/ftrace/test.d/trigger/inter-event/trigger-onmax-action-hist.tc +5 -0

index 0035995c2194..98d73bfb0296 100644
@@ -17,6 +17,11 @@ if [ ! -f synthetic_events ]; then
     exit_unsupported
 fi
 
+if [ ! -f events/sched/sched_process_fork/hist ]; then
+    echo "hist trigger is not supported"
+    exit_unsupported
+fi
+
 echo "Test create synthetic event"
 
 echo 'wakeup_latency  u64 lat pid_t pid char comm[16]' > synthetic_events

tools/testing/selftests/ftrace/test.d/trigger/inter-event/trigger-snapshot-action-hist.tc +5 -0

index f546c1b66a9b..01b01b9c4e07 100644
@@ -12,6 +12,11 @@ if [ ! -f set_event ]; then
     exit_unsupported
 fi
 
+if [ ! -f events/sched/sched_process_fork/hist ]; then
+    echo "hist trigger is not supported"
+    exit_unsupported
+fi
+
 if [ ! -f snapshot ]; then
     echo "snapshot is not supported"
     exit_unsupported

tools/testing/selftests/ftrace/test.d/trigger/inter-event/trigger-trace-action-hist.tc +5 -0

index 8021d60aafec..c3baa486aeb4 100644
@@ -17,6 +17,11 @@ if [ ! -f synthetic_events ]; then
     exit_unsupported
 fi
 
+if [ ! -f events/sched/sched_process_fork/hist ]; then
+    echo "hist trigger is not supported"
+    exit_unsupported
+fi
+
 grep -q "trace(<synthetic_event>" README || exit_unsupported # version issue
 
 echo "Test create synthetic event"

Signature · 1 line

2.17.1

Signature · 1 line

Masami Hiramatsu <[email protected]>

Steven Rostedt

maintainer

2020-05-29 14:53 UTC #4 lore raw

Re: [PATCH v2 2/2] selftests/ftrace: Distinguish between hist and synthetic event checks

Quoted 10 lines · On Fri, 29 May 2020 23:38:45 +0900 Masami Hiramatsu <[email protected]> wrote:

Hi Tom,
Quoted 5 lines · On Thu, 28 May 2020 14:32:38 -0500 Tom Zanussi <[email protected]> wrote:
With synthetic events now a separate config item as a result of 'tracing: Move synthetic events to a separate file', tests that use both need to explicitly check for hist trigger support rather than relying on hist triggers to pull in synthetic events.

Quoted 8 lines

Quoted 3 lines
Add an additional hist trigger check to all the trigger tests that now require it, otherwise they'll fail if synthetic events but not hist triggers are enabled.
OK, this looks good to me. And if you don't want to repeat it, you can also put the check function into the test.d/functions.
Acked-by Masami Hiramatsu <[email protected]>

Thanks Tom and Masami. I'm running tests on these now.

-- Steve