[Tracing] Embed V8 runtime call stats into tracing.

[Reland][Tracing] Embed V8 runtime call stats into tracing.

Currently we have V8 RuntimeCallStats that is independently from tracing when
running d8 with flag --runtime_call_stats. This patch embeds V8 runtime call
stats into tracing, by having a global table of runtime call counters each
isolate, resetting the table each time we enter a top level trace event, and
dumping the table for each top level trace event. This will make trace file more
compat, as well as enable runtime call stats in tracing system.

This patch adds ~5% overhead to V8 when the category is enabled, we measure the
overhead by running a script when category is enabled.

BUG=v8:5089
Committed: https://crrev.com/d014866173eaa2b548c566217b2c94b1d49385fa
Committed: https://crrev.com/1ca3b73bba4a7253ca8eeef39321d70e7d414331
Committed: https://crrev.com/3f936a5b17754783e92d2146eaf66c88a78ee45b
Committed: https://crrev.com/7a3631e7e12fe0e5038d80c58f3f491c61977b95
Cr-Original-Original-Original-Commit-Position: refs/heads/master@{#38270}
Cr-Original-Original-Commit-Position: refs/heads/master@{#38314}
Cr-Original-Commit-Position: refs/heads/master@{#38403}
Cr-Commit-Position: refs/heads/master@{#38510}

[lpy, 2016-07-26 21:21:12]

The CQ bit was checked by lpy@chromium.org to run a CQ dry run

[commit-bot: I haz the power, 2016-07-26 21:21:23]

Dry run: CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[lpy, 2016-07-26 21:22:03]

lpy@chromium.org changed reviewers:
+ cbruni@chromium.org, fmeawad@chromium.org

[lpy, 2016-07-26 21:22:05]

PTAL.

[commit-bot: I haz the power, 2016-07-26 21:45:02]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2016-07-26 21:45:03]

Dry run: This issue passed the CQ dry run.

[Camillo Bruni, 2016-07-27 08:29:07]

nice work!

I added some comments mainly concerning performance.
I spend quite some time figuring out how to wire the runtime-stats into the
runtime functions and builtins to pose a minimal overhead.
The moment you add a stack-allocated scope object in the function, clang doesn't
manage to optimize it away even if the scope only conditionally does something.
The only way to enforce this was to put the scopes into a separate non-inlinable
function and hoisting the checks out the default function.

To measure the overhead I just hammered one runtime-function using the natives
syntax in js.
Just modify an existing runtime function that is not used (for instance
Runtime_GetHeapUsage) and immediately return a smi:

for (var i = 0; i < 1000000; i++) {
  %GetHeapUsage();
}

https://codereview.chromium.org/2187693002/diff/1/src/arguments.h
File src/arguments.h (right):

https://codereview.chromium.org/2187693002/diff/1/src/arguments.h#newcode103
src/arguments.h:103: isolate, &tracing::TraceEventStatsTable::Name);            
          \
Could you put the the scope in the Stats_##Name macro?
I'd like to keep as many timers out of the default runtime function as possible.

https://codereview.chromium.org/2187693002/diff/1/src/builtins/builtins-utils.h
File src/builtins/builtins-utils.h (right):

https://codereview.chromium.org/2187693002/diff/1/src/builtins/builtins-utils...
src/builtins/builtins-utils.h:100: Object* result = Builtin_Impl_##name(args,
isolate);                     \
The same as for the runtime functions.
I'd like to keep all timers and stack-allocated scopes away from the default
functions.

https://codereview.chromium.org/2187693002/diff/1/src/d8.gyp
File src/d8.gyp (right):

https://codereview.chromium.org/2187693002/diff/1/src/d8.gyp#newcode47
src/d8.gyp:47: '<(DEPTH)',
We manually specify includes for each file, so I'm not sure we want this.

https://codereview.chromium.org/2187693002/diff/1/src/tracing/trace-event.cc
File src/tracing/trace-event.cc (right):

https://codereview.chromium.org/2187693002/diff/1/src/tracing/trace-event.cc#...
src/tracing/trace-event.cc:18: int kRuntimeCallsTracingEnabled = 0;
Please make this a bool to make your intentions a bit clearer, seems like you
use 0/1 state only anyway.

https://codereview.chromium.org/2187693002/diff/1/src/tracing/trace-event.cc#...
src/tracing/trace-event.cc:25: if (p_data_ && *data_.category_group_enabled) {
can you put the test in the header and delegate the rest to a function in .cc ?

https://codereview.chromium.org/2187693002/diff/1/src/tracing/trace-event.h
File src/tracing/trace-event.h (right):

https://codereview.chromium.org/2187693002/diff/1/src/tracing/trace-event.h#n...
src/tracing/trace-event.h:327: extern int kRuntimeCallsTracingEnabled;
change to bool.

https://codereview.chromium.org/2187693002/diff/1/src/v8.gyp
File src/v8.gyp (right):

https://codereview.chromium.org/2187693002/diff/1/src/v8.gyp#newcode182
src/v8.gyp:182: '<(DEPTH)',
please manually specify the new files you want to include.

[lpy, 2016-07-27 19:55:59]

The CQ bit was checked by lpy@chromium.org to run a CQ dry run

[commit-bot: I haz the power, 2016-07-27 19:56:04]

Dry run: CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[lpy, 2016-07-27 19:58:38]

Thanks, I updated the CL.

https://codereview.chromium.org/2187693002/diff/1/src/arguments.h
File src/arguments.h (right):

https://codereview.chromium.org/2187693002/diff/1/src/arguments.h#newcode103
src/arguments.h:103: isolate, &tracing::TraceEventStatsTable::Name);            
          \
On 2016/07/27 08:29:07, Camillo Bruni wrote:
> Could you put the the scope in the Stats_##Name macro?
> I'd like to keep as many timers out of the default runtime function as
possible.

I understand the concern here. The reason we are putting it here is we don't
want the scope to be behind the --runtime_call_stats flag.

What about moving the scope to everywhere that uses RUNTIME_FUNCTION macro? I
think eventually we will want to replace the RuntimeCallTimerScope, so at that
point we can move the scope in Stats_##Name and remove the flag.

WDYT?

https://codereview.chromium.org/2187693002/diff/1/src/d8.gyp
File src/d8.gyp (right):

https://codereview.chromium.org/2187693002/diff/1/src/d8.gyp#newcode47
src/d8.gyp:47: '<(DEPTH)',
On 2016/07/27 08:29:07, Camillo Bruni wrote:
> We manually specify includes for each file, so I'm not sure we want this.

This is for correctly including the trace_event_common.h when building with
Chrome, I think we use the same way in src/v8.gyp.

https://codereview.chromium.org/2187693002/diff/1/src/tracing/trace-event.cc
File src/tracing/trace-event.cc (right):

https://codereview.chromium.org/2187693002/diff/1/src/tracing/trace-event.cc#...
src/tracing/trace-event.cc:18: int kRuntimeCallsTracingEnabled = 0;
On 2016/07/27 08:29:07, Camillo Bruni wrote:
> Please make this a bool to make your intentions a bit clearer, seems like you
> use 0/1 state only anyway.

Done.

https://codereview.chromium.org/2187693002/diff/1/src/tracing/trace-event.cc#...
src/tracing/trace-event.cc:25: if (p_data_ && *data_.category_group_enabled) {
On 2016/07/27 08:29:07, Camillo Bruni wrote:
> can you put the test in the header and delegate the rest to a function in .cc
?

Done.

https://codereview.chromium.org/2187693002/diff/1/src/tracing/trace-event.h
File src/tracing/trace-event.h (right):

https://codereview.chromium.org/2187693002/diff/1/src/tracing/trace-event.h#n...
src/tracing/trace-event.h:327: extern int kRuntimeCallsTracingEnabled;
On 2016/07/27 08:29:07, Camillo Bruni wrote:
> change to bool.

Done.

https://codereview.chromium.org/2187693002/diff/1/src/v8.gyp
File src/v8.gyp (right):

https://codereview.chromium.org/2187693002/diff/1/src/v8.gyp#newcode182
src/v8.gyp:182: '<(DEPTH)',
On 2016/07/27 08:29:07, Camillo Bruni wrote:
> please manually specify the new files you want to include.

This is for correctly including the trace_event_common.h when building with
Chrome.

[commit-bot: I haz the power, 2016-07-27 20:22:13]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2016-07-27 20:22:16]

Dry run: This issue passed the CQ dry run.

[Camillo Bruni, 2016-07-28 07:15:48]

cbruni@chromium.org changed reviewers:
+ machenbach@chromium.org

[Camillo Bruni, 2016-07-28 07:15:49]

machenbach@ PTAL gyp changes

[Michael Achenbach, 2016-07-28 07:20:33]

Gyp lg. No GN changes required?

[lpy, 2016-07-28 18:20:38]

On 2016/07/28 07:20:33, Michael Achenbach (slow) wrote:
> Gyp lg. No GN changes required?

It's already there. https://cs.chromium.org/chromium/src/v8/BUILD.gn?l=806

[fmeawad, 2016-07-28 21:14:31]

https://codereview.chromium.org/2187693002/diff/20001/src/tracing/trace-event.cc
File src/tracing/trace-event.cc (right):

https://codereview.chromium.org/2187693002/diff/20001/src/tracing/trace-event...
src/tracing/trace-event.cc:128: buffer_ << "\"END\":[]}";
Given that we are trying to minimize the trace size, is there an easy way to get
rid of the END here?

[Camillo Bruni, 2016-07-29 08:56:16]

https://codereview.chromium.org/2187693002/diff/1/src/arguments.h
File src/arguments.h (right):

https://codereview.chromium.org/2187693002/diff/1/src/arguments.h#newcode103
src/arguments.h:103: isolate, &tracing::TraceEventStatsTable::Name);            
          \
Oh yes of course you don't want to be behind a flag :), but could you just OR
your checks whether the trace is enabled or not?
You will have to do this anyway to reduce the overhead when removing the
RuntimeCallTimerScope.
Since the scopes aren't optimized away and given that some runtime functions are
called really frequently I'd like to reduce the overhead to a minimum.

Pushing the uses down to the RUNTIME_FUNCTION wouldn't help in this case either,
it really has to be a guarded call to a separate non-inlined function.
Maybe use the V8_UNLIKELY() around the check as well, I forgot about that.

https://codereview.chromium.org/2187693002/diff/20001/src/tracing/trace-event.cc
File src/tracing/trace-event.cc (right):

https://codereview.chromium.org/2187693002/diff/20001/src/tracing/trace-event...
src/tracing/trace-event.cc:140: if
(V8_UNLIKELY(TRACE_EVENT_RUNTIME_CALL_STATS_TRACING_ENABLED())) {
Please put the hole function or at least the check in the header file as well.

[lpy, 2016-08-01 07:09:29]

The CQ bit was checked by lpy@chromium.org to run a CQ dry run

[commit-bot: I haz the power, 2016-08-01 07:09:39]

Dry run: CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[lpy, 2016-08-01 07:13:25]

The CQ bit was checked by lpy@chromium.org to run a CQ dry run

[commit-bot: I haz the power, 2016-08-01 07:13:29]

Dry run: CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[lpy, 2016-08-01 07:13:57]

Thanks, I updated the CL.

PTAL.

https://codereview.chromium.org/2187693002/diff/20001/src/tracing/trace-event.cc
File src/tracing/trace-event.cc (right):

https://codereview.chromium.org/2187693002/diff/20001/src/tracing/trace-event...
src/tracing/trace-event.cc:140: if
(V8_UNLIKELY(TRACE_EVENT_RUNTIME_CALL_STATS_TRACING_ENABLED())) {
On 2016/07/29 08:56:15, Camillo Bruni wrote:
> Please put the hole function or at least the check in the header file as well.

Done.

[commit-bot: I haz the power, 2016-08-01 07:39:11]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2016-08-01 07:39:12]

Dry run: This issue passed the CQ dry run.

[Camillo Bruni, 2016-08-01 07:43:55]

LGTM with nits

https://codereview.chromium.org/2187693002/diff/60001/src/builtins/builtins-u...
File src/builtins/builtins-utils.h (right):

https://codereview.chromium.org/2187693002/diff/60001/src/builtins/builtins-u...
src/builtins/builtins-utils.h:97: FLAG_runtime_call_stats)) {                   
          \
nice!

https://codereview.chromium.org/2187693002/diff/60001/src/counters.h
File src/counters.h (right):

https://codereview.chromium.org/2187693002/diff/60001/src/counters.h#newcode485
src/counters.h:485: void Dump(std::stringstream& out);
nit: Please mark as non-inlined (V8_NOINLINE void ...)
TraceEventStatsTable::Dump otherwise might end up being a huge function (we had
a report of 150KB ;)

https://codereview.chromium.org/2187693002/diff/60001/src/tracing/trace-event.h
File src/tracing/trace-event.h (right):

https://codereview.chromium.org/2187693002/diff/60001/src/tracing/trace-event...
src/tracing/trace-event.h:633: if (p_data_ && *data_.category_group_enabled) {
microtweaking nit: change to if(V8_UNLIKELY(p_data_ &&
*data_.category_group_enabled))

[lpy, 2016-08-01 18:16:29]

The CQ bit was checked by lpy@chromium.org to run a CQ dry run

[commit-bot: I haz the power, 2016-08-01 18:16:40]

Dry run: CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[lpy, 2016-08-01 18:17:55]

lpy@chromium.org changed reviewers:
+ adamk@chromium.org, bmeurer@chromium.org, rmcilroy@chromium.org

[lpy, 2016-08-01 18:17:56]

Thanks, I updated the CL.

+ bmeurer@ for full-codegen/
+ adamk@ for parsing/
+ rmcilroy@ for interpreter/

Please take a look.

https://codereview.chromium.org/2187693002/diff/60001/src/builtins/builtins-u...
File src/builtins/builtins-utils.h (right):

https://codereview.chromium.org/2187693002/diff/60001/src/builtins/builtins-u...
src/builtins/builtins-utils.h:97: FLAG_runtime_call_stats)) {                   
          \
On 2016/08/01 07:43:55, Camillo Bruni wrote:
> nice!

\o/

https://codereview.chromium.org/2187693002/diff/60001/src/counters.h
File src/counters.h (right):

https://codereview.chromium.org/2187693002/diff/60001/src/counters.h#newcode485
src/counters.h:485: void Dump(std::stringstream& out);
On 2016/08/01 07:43:55, Camillo Bruni wrote:
> nit: Please mark as non-inlined (V8_NOINLINE void ...)
> TraceEventStatsTable::Dump otherwise might end up being a huge function (we
had
> a report of 150KB ;)

Done.

https://codereview.chromium.org/2187693002/diff/60001/src/tracing/trace-event.h
File src/tracing/trace-event.h (right):

https://codereview.chromium.org/2187693002/diff/60001/src/tracing/trace-event...
src/tracing/trace-event.h:633: if (p_data_ && *data_.category_group_enabled) {
On 2016/08/01 07:43:55, Camillo Bruni wrote:
> microtweaking nit: change to if(V8_UNLIKELY(p_data_ &&
> *data_.category_group_enabled)) 

Done.

[commit-bot: I haz the power, 2016-08-01 18:41:28]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2016-08-01 18:41:29]

Dry run: This issue passed the CQ dry run.

[adamk, 2016-08-01 19:17:58]

lgtm for parsing/

[Benedikt Meurer, 2016-08-02 05:17:54]

LGTM for fullcodegen and ignition.

[lpy, 2016-08-03 05:22:12]

The CQ bit was checked by lpy@chromium.org to run a CQ dry run

[commit-bot: I haz the power, 2016-08-03 05:22:23]

Dry run: CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[commit-bot: I haz the power, 2016-08-03 05:42:19]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2016-08-03 05:42:20]

Dry run: This issue passed the CQ dry run.

[lpy, 2016-08-03 06:01:00]

The CQ bit was checked by lpy@chromium.org

[lpy, 2016-08-03 06:01:00]

The patchset sent to the CQ was uploaded after l-g-t-m from cbruni@chromium.org,
bmeurer@chromium.org, adamk@chromium.org
Link to the patchset: https://codereview.chromium.org/2187693002/#ps100001
(title: "Fix dangling pointer issue")

[lpy, 2016-08-03 06:01:05]

The CQ bit was unchecked by lpy@chromium.org

[lpy, 2016-08-03 06:09:25]

The CQ bit was checked by lpy@chromium.org

[commit-bot: I haz the power, 2016-08-03 06:09:29]

CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[commit-bot: I haz the power, 2016-08-03 06:12:27]

Committed patchset #6 (id:100001)

[commit-bot: I haz the power, 2016-08-03 06:13:41]

Description was changed from

==========
[Tracing] Embed V8 runtime call stats into tracing.

Currently we have V8 RuntimeCallStats that is independently from tracing when
running d8 with flag --runtime_call_stats. This patch embeds V8 runtime call
stats into tracing, by having a global table of runtime call counters each
isolate, resetting the table each time we enter a top level trace event, and
dumping the table for each top level trace event. This will make trace file more
compat, as well as enable runtime call stats in tracing system.

This patch adds ~5% overhead to V8 when the category is enabled, we measure the
overhead by running a script when category is enabled.

BUG=5089
==========

to

==========
[Tracing] Embed V8 runtime call stats into tracing.

Currently we have V8 RuntimeCallStats that is independently from tracing when
running d8 with flag --runtime_call_stats. This patch embeds V8 runtime call
stats into tracing, by having a global table of runtime call counters each
isolate, resetting the table each time we enter a top level trace event, and
dumping the table for each top level trace event. This will make trace file more
compat, as well as enable runtime call stats in tracing system.

This patch adds ~5% overhead to V8 when the category is enabled, we measure the
overhead by running a script when category is enabled.

BUG=5089

Committed: https://crrev.com/d014866173eaa2b548c566217b2c94b1d49385fa
Cr-Commit-Position: refs/heads/master@{#38270}
==========

[commit-bot: I haz the power, 2016-08-03 06:13:43]

Patchset 6 (id:??) landed as
https://crrev.com/d014866173eaa2b548c566217b2c94b1d49385fa
Cr-Commit-Position: refs/heads/master@{#38270}

[lpy, 2016-08-03 06:29:49]

Description was changed from

==========
[Tracing] Embed V8 runtime call stats into tracing.

Currently we have V8 RuntimeCallStats that is independently from tracing when
running d8 with flag --runtime_call_stats. This patch embeds V8 runtime call
stats into tracing, by having a global table of runtime call counters each
isolate, resetting the table each time we enter a top level trace event, and
dumping the table for each top level trace event. This will make trace file more
compat, as well as enable runtime call stats in tracing system.

This patch adds ~5% overhead to V8 when the category is enabled, we measure the
overhead by running a script when category is enabled.

BUG=5089

Committed: https://crrev.com/d014866173eaa2b548c566217b2c94b1d49385fa
Cr-Commit-Position: refs/heads/master@{#38270}
==========

to

==========
[Tracing] Embed V8 runtime call stats into tracing.

Currently we have V8 RuntimeCallStats that is independently from tracing when
running d8 with flag --runtime_call_stats. This patch embeds V8 runtime call
stats into tracing, by having a global table of runtime call counters each
isolate, resetting the table each time we enter a top level trace event, and
dumping the table for each top level trace event. This will make trace file more
compat, as well as enable runtime call stats in tracing system.

This patch adds ~5% overhead to V8 when the category is enabled, we measure the
overhead by running a script when category is enabled.

BUG=v8:5089

Committed: https://crrev.com/d014866173eaa2b548c566217b2c94b1d49385fa
Cr-Commit-Position: refs/heads/master@{#38270}
==========

[Yang, 2016-08-03 07:25:12]

A revert of this CL (patchset #6 id:100001) has been created in
https://codereview.chromium.org/2203913004/ by yangguo@chromium.org.

The reason for reverting is: Sanitizer failures:

https://build.chromium.org/p/client.v8/builders/V8%20Linux64%20TSAN/builds/10...
https://build.chromium.org/p/client.v8/builders/V8%20Mac64%20ASAN/builds/7786/.

[lpy, 2016-08-03 20:15:10]

Description was changed from

==========
[Tracing] Embed V8 runtime call stats into tracing.

Currently we have V8 RuntimeCallStats that is independently from tracing when
running d8 with flag --runtime_call_stats. This patch embeds V8 runtime call
stats into tracing, by having a global table of runtime call counters each
isolate, resetting the table each time we enter a top level trace event, and
dumping the table for each top level trace event. This will make trace file more
compat, as well as enable runtime call stats in tracing system.

This patch adds ~5% overhead to V8 when the category is enabled, we measure the
overhead by running a script when category is enabled.

BUG=v8:5089

Committed: https://crrev.com/d014866173eaa2b548c566217b2c94b1d49385fa
Cr-Commit-Position: refs/heads/master@{#38270}
==========

to

==========
[Reland][Tracing] Embed V8 runtime call stats into tracing.

Currently we have V8 RuntimeCallStats that is independently from tracing when
running d8 with flag --runtime_call_stats. This patch embeds V8 runtime call
stats into tracing, by having a global table of runtime call counters each
isolate, resetting the table each time we enter a top level trace event, and
dumping the table for each top level trace event. This will make trace file more
compat, as well as enable runtime call stats in tracing system.

This patch adds ~5% overhead to V8 when the category is enabled, we measure the
overhead by running a script when category is enabled.

BUG=v8:5089

Committed: https://crrev.com/d014866173eaa2b548c566217b2c94b1d49385fa
Cr-Commit-Position: refs/heads/master@{#38270}
==========

[lpy, 2016-08-03 20:15:32]

The CQ bit was checked by lpy@chromium.org to run a CQ dry run

[commit-bot: I haz the power, 2016-08-03 20:15:42]

Dry run: CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[commit-bot: I haz the power, 2016-08-03 20:40:12]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2016-08-03 20:40:13]

Dry run: This issue passed the CQ dry run.

[lpy, 2016-08-03 20:42:46]

The CQ bit was checked by lpy@chromium.org

[lpy, 2016-08-03 20:42:47]

The patchset sent to the CQ was uploaded after l-g-t-m from cbruni@chromium.org,
bmeurer@chromium.org, adamk@chromium.org
Link to the patchset: https://codereview.chromium.org/2187693002/#ps140001
(title: "Fix TSAN and ASAN failure")

[commit-bot: I haz the power, 2016-08-03 20:42:53]

CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[commit-bot: I haz the power, 2016-08-03 20:44:59]

Description was changed from

==========
[Reland][Tracing] Embed V8 runtime call stats into tracing.

Currently we have V8 RuntimeCallStats that is independently from tracing when
running d8 with flag --runtime_call_stats. This patch embeds V8 runtime call
stats into tracing, by having a global table of runtime call counters each
isolate, resetting the table each time we enter a top level trace event, and
dumping the table for each top level trace event. This will make trace file more
compat, as well as enable runtime call stats in tracing system.

This patch adds ~5% overhead to V8 when the category is enabled, we measure the
overhead by running a script when category is enabled.

BUG=v8:5089

Committed: https://crrev.com/d014866173eaa2b548c566217b2c94b1d49385fa
Cr-Commit-Position: refs/heads/master@{#38270}
==========

to

==========
[Reland][Tracing] Embed V8 runtime call stats into tracing.

Currently we have V8 RuntimeCallStats that is independently from tracing when
running d8 with flag --runtime_call_stats. This patch embeds V8 runtime call
stats into tracing, by having a global table of runtime call counters each
isolate, resetting the table each time we enter a top level trace event, and
dumping the table for each top level trace event. This will make trace file more
compat, as well as enable runtime call stats in tracing system.

This patch adds ~5% overhead to V8 when the category is enabled, we measure the
overhead by running a script when category is enabled.

BUG=v8:5089

Committed: https://crrev.com/d014866173eaa2b548c566217b2c94b1d49385fa
Cr-Commit-Position: refs/heads/master@{#38270}
==========

[commit-bot: I haz the power, 2016-08-03 20:45:00]

Committed patchset #8 (id:140001)

[commit-bot: I haz the power, 2016-08-03 20:48:48]

Description was changed from

==========
[Reland][Tracing] Embed V8 runtime call stats into tracing.

Currently we have V8 RuntimeCallStats that is independently from tracing when
running d8 with flag --runtime_call_stats. This patch embeds V8 runtime call
stats into tracing, by having a global table of runtime call counters each
isolate, resetting the table each time we enter a top level trace event, and
dumping the table for each top level trace event. This will make trace file more
compat, as well as enable runtime call stats in tracing system.

This patch adds ~5% overhead to V8 when the category is enabled, we measure the
overhead by running a script when category is enabled.

BUG=v8:5089

Committed: https://crrev.com/d014866173eaa2b548c566217b2c94b1d49385fa
Cr-Commit-Position: refs/heads/master@{#38270}
==========

to

==========
[Reland][Tracing] Embed V8 runtime call stats into tracing.

Currently we have V8 RuntimeCallStats that is independently from tracing when
running d8 with flag --runtime_call_stats. This patch embeds V8 runtime call
stats into tracing, by having a global table of runtime call counters each
isolate, resetting the table each time we enter a top level trace event, and
dumping the table for each top level trace event. This will make trace file more
compat, as well as enable runtime call stats in tracing system.

This patch adds ~5% overhead to V8 when the category is enabled, we measure the
overhead by running a script when category is enabled.

BUG=v8:5089

Committed: https://crrev.com/d014866173eaa2b548c566217b2c94b1d49385fa
Committed: https://crrev.com/1ca3b73bba4a7253ca8eeef39321d70e7d414331
Cr-Original-Commit-Position: refs/heads/master@{#38270}
Cr-Commit-Position: refs/heads/master@{#38314}
==========

[commit-bot: I haz the power, 2016-08-03 20:48:50]

Patchset 8 (id:??) landed as
https://crrev.com/1ca3b73bba4a7253ca8eeef39321d70e7d414331
Cr-Commit-Position: refs/heads/master@{#38314}

[lpy, 2016-08-03 21:42:39]

A revert of this CL (patchset #8 id:140001) has been created in
https://codereview.chromium.org/2200373003/ by lpy@chromium.org.

The reason for reverting is: Mac64 ASAN failure.
https://build.chromium.org/p/client.v8/builders/V8%20Mac64%20ASAN/builds/7810....

[lpy, 2016-08-04 18:24:20]

Description was changed from

==========
[Reland][Tracing] Embed V8 runtime call stats into tracing.

Currently we have V8 RuntimeCallStats that is independently from tracing when
running d8 with flag --runtime_call_stats. This patch embeds V8 runtime call
stats into tracing, by having a global table of runtime call counters each
isolate, resetting the table each time we enter a top level trace event, and
dumping the table for each top level trace event. This will make trace file more
compat, as well as enable runtime call stats in tracing system.

This patch adds ~5% overhead to V8 when the category is enabled, we measure the
overhead by running a script when category is enabled.

BUG=v8:5089

Committed: https://crrev.com/d014866173eaa2b548c566217b2c94b1d49385fa
Committed: https://crrev.com/1ca3b73bba4a7253ca8eeef39321d70e7d414331
Cr-Original-Commit-Position: refs/heads/master@{#38270}
Cr-Commit-Position: refs/heads/master@{#38314}
==========

to

==========
[Reland][Tracing] Embed V8 runtime call stats into tracing.

Currently we have V8 RuntimeCallStats that is independently from tracing when
running d8 with flag --runtime_call_stats. This patch embeds V8 runtime call
stats into tracing, by having a global table of runtime call counters each
isolate, resetting the table each time we enter a top level trace event, and
dumping the table for each top level trace event. This will make trace file more
compat, as well as enable runtime call stats in tracing system.

This patch adds ~5% overhead to V8 when the category is enabled, we measure the
overhead by running a script when category is enabled.

BUG=v8:5089

Committed: https://crrev.com/d014866173eaa2b548c566217b2c94b1d49385fa
Committed: https://crrev.com/1ca3b73bba4a7253ca8eeef39321d70e7d414331
Cr-Original-Commit-Position: refs/heads/master@{#38270}
Cr-Commit-Position: refs/heads/master@{#38314}
==========

[lpy, 2016-08-05 21:57:50]

The CQ bit was checked by lpy@chromium.org

[commit-bot: I haz the power, 2016-08-05 21:57:59]

CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[commit-bot: I haz the power, 2016-08-05 22:21:56]

Description was changed from

==========
[Reland][Tracing] Embed V8 runtime call stats into tracing.

Currently we have V8 RuntimeCallStats that is independently from tracing when
running d8 with flag --runtime_call_stats. This patch embeds V8 runtime call
stats into tracing, by having a global table of runtime call counters each
isolate, resetting the table each time we enter a top level trace event, and
dumping the table for each top level trace event. This will make trace file more
compat, as well as enable runtime call stats in tracing system.

This patch adds ~5% overhead to V8 when the category is enabled, we measure the
overhead by running a script when category is enabled.

BUG=v8:5089

Committed: https://crrev.com/d014866173eaa2b548c566217b2c94b1d49385fa
Committed: https://crrev.com/1ca3b73bba4a7253ca8eeef39321d70e7d414331
Cr-Original-Commit-Position: refs/heads/master@{#38270}
Cr-Commit-Position: refs/heads/master@{#38314}
==========

to

==========
[Reland][Tracing] Embed V8 runtime call stats into tracing.

Currently we have V8 RuntimeCallStats that is independently from tracing when
running d8 with flag --runtime_call_stats. This patch embeds V8 runtime call
stats into tracing, by having a global table of runtime call counters each
isolate, resetting the table each time we enter a top level trace event, and
dumping the table for each top level trace event. This will make trace file more
compat, as well as enable runtime call stats in tracing system.

This patch adds ~5% overhead to V8 when the category is enabled, we measure the
overhead by running a script when category is enabled.

BUG=v8:5089

Committed: https://crrev.com/d014866173eaa2b548c566217b2c94b1d49385fa
Committed: https://crrev.com/1ca3b73bba4a7253ca8eeef39321d70e7d414331
Cr-Original-Commit-Position: refs/heads/master@{#38270}
Cr-Commit-Position: refs/heads/master@{#38314}
==========

[commit-bot: I haz the power, 2016-08-05 22:21:58]

Committed patchset #8 (id:140001)

[commit-bot: I haz the power, 2016-08-05 22:24:53]

Description was changed from

==========
[Reland][Tracing] Embed V8 runtime call stats into tracing.

Currently we have V8 RuntimeCallStats that is independently from tracing when
running d8 with flag --runtime_call_stats. This patch embeds V8 runtime call
stats into tracing, by having a global table of runtime call counters each
isolate, resetting the table each time we enter a top level trace event, and
dumping the table for each top level trace event. This will make trace file more
compat, as well as enable runtime call stats in tracing system.

This patch adds ~5% overhead to V8 when the category is enabled, we measure the
overhead by running a script when category is enabled.

BUG=v8:5089

Committed: https://crrev.com/d014866173eaa2b548c566217b2c94b1d49385fa
Committed: https://crrev.com/1ca3b73bba4a7253ca8eeef39321d70e7d414331
Cr-Original-Commit-Position: refs/heads/master@{#38270}
Cr-Commit-Position: refs/heads/master@{#38314}
==========

to

==========
[Reland][Tracing] Embed V8 runtime call stats into tracing.

Currently we have V8 RuntimeCallStats that is independently from tracing when
running d8 with flag --runtime_call_stats. This patch embeds V8 runtime call
stats into tracing, by having a global table of runtime call counters each
isolate, resetting the table each time we enter a top level trace event, and
dumping the table for each top level trace event. This will make trace file more
compat, as well as enable runtime call stats in tracing system.

This patch adds ~5% overhead to V8 when the category is enabled, we measure the
overhead by running a script when category is enabled.

BUG=v8:5089

Committed: https://crrev.com/d014866173eaa2b548c566217b2c94b1d49385fa
Committed: https://crrev.com/1ca3b73bba4a7253ca8eeef39321d70e7d414331
Committed: https://crrev.com/3f936a5b17754783e92d2146eaf66c88a78ee45b
Cr-Original-Original-Commit-Position: refs/heads/master@{#38270}
Cr-Original-Commit-Position: refs/heads/master@{#38314}
Cr-Commit-Position: refs/heads/master@{#38403}
==========

[commit-bot: I haz the power, 2016-08-05 22:24:54]

Patchset 8 (id:??) landed as
https://crrev.com/3f936a5b17754783e92d2146eaf66c88a78ee45b
Cr-Commit-Position: refs/heads/master@{#38403}

[Michael Achenbach, 2016-08-08 06:45:54]

A revert of this CL (patchset #8 id:140001) has been created in
https://codereview.chromium.org/2221853002/ by machenbach@chromium.org.

The reason for reverting is: Leaks block chromium roll:
https://codereview.chromium.org/2219083003/

Example build:
https://build.chromium.org/p/tryserver.chromium.linux/builders/linux_chromium...

You can add the trybot from tryserver.chromium.linux,
linux_chromium_asan_rel_ng, on reland..

[lpy, 2016-08-08 17:01:07]

Description was changed from

==========
[Reland][Tracing] Embed V8 runtime call stats into tracing.

Currently we have V8 RuntimeCallStats that is independently from tracing when
running d8 with flag --runtime_call_stats. This patch embeds V8 runtime call
stats into tracing, by having a global table of runtime call counters each
isolate, resetting the table each time we enter a top level trace event, and
dumping the table for each top level trace event. This will make trace file more
compat, as well as enable runtime call stats in tracing system.

This patch adds ~5% overhead to V8 when the category is enabled, we measure the
overhead by running a script when category is enabled.

BUG=v8:5089

Committed: https://crrev.com/d014866173eaa2b548c566217b2c94b1d49385fa
Committed: https://crrev.com/1ca3b73bba4a7253ca8eeef39321d70e7d414331
Committed: https://crrev.com/3f936a5b17754783e92d2146eaf66c88a78ee45b
Cr-Original-Original-Commit-Position: refs/heads/master@{#38270}
Cr-Original-Commit-Position: refs/heads/master@{#38314}
Cr-Commit-Position: refs/heads/master@{#38403}
==========

to

==========
[Reland][Tracing] Embed V8 runtime call stats into tracing.

Currently we have V8 RuntimeCallStats that is independently from tracing when
running d8 with flag --runtime_call_stats. This patch embeds V8 runtime call
stats into tracing, by having a global table of runtime call counters each
isolate, resetting the table each time we enter a top level trace event, and
dumping the table for each top level trace event. This will make trace file more
compat, as well as enable runtime call stats in tracing system.

This patch adds ~5% overhead to V8 when the category is enabled, we measure the
overhead by running a script when category is enabled.

BUG=v8:5089

Committed: https://crrev.com/d014866173eaa2b548c566217b2c94b1d49385fa
Committed: https://crrev.com/1ca3b73bba4a7253ca8eeef39321d70e7d414331
Committed: https://crrev.com/3f936a5b17754783e92d2146eaf66c88a78ee45b
Cr-Original-Original-Commit-Position: refs/heads/master@{#38270}
Cr-Original-Commit-Position: refs/heads/master@{#38314}
Cr-Commit-Position: refs/heads/master@{#38403}
==========

[lpy, 2016-08-09 23:47:16]

The CQ bit was checked by lpy@chromium.org to run a CQ dry run

[commit-bot: I haz the power, 2016-08-09 23:47:24]

Dry run: CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[lpy, 2016-08-09 23:48:33]

The CQ bit was checked by lpy@chromium.org to run a CQ dry run

[commit-bot: I haz the power, 2016-08-09 23:48:44]

Dry run: CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[lpy, 2016-08-09 23:49:01]

The CQ bit was checked by lpy@chromium.org to run a CQ dry run

[commit-bot: I haz the power, 2016-08-09 23:49:04]

Dry run: CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[commit-bot: I haz the power, 2016-08-10 00:19:09]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2016-08-10 00:19:11]

Dry run: This issue passed the CQ dry run.

[lpy, 2016-08-10 01:16:30]

The CQ bit was checked by lpy@chromium.org

[lpy, 2016-08-10 01:16:30]

The patchset sent to the CQ was uploaded after l-g-t-m from cbruni@chromium.org,
bmeurer@chromium.org, adamk@chromium.org
Link to the patchset: https://codereview.chromium.org/2187693002/#ps160001
(title: "Fix memory leak")

[commit-bot: I haz the power, 2016-08-10 01:16:37]

CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[commit-bot: I haz the power, 2016-08-10 01:18:47]

Description was changed from

==========
[Reland][Tracing] Embed V8 runtime call stats into tracing.

Currently we have V8 RuntimeCallStats that is independently from tracing when
running d8 with flag --runtime_call_stats. This patch embeds V8 runtime call
stats into tracing, by having a global table of runtime call counters each
isolate, resetting the table each time we enter a top level trace event, and
dumping the table for each top level trace event. This will make trace file more
compat, as well as enable runtime call stats in tracing system.

This patch adds ~5% overhead to V8 when the category is enabled, we measure the
overhead by running a script when category is enabled.

BUG=v8:5089

Committed: https://crrev.com/d014866173eaa2b548c566217b2c94b1d49385fa
Committed: https://crrev.com/1ca3b73bba4a7253ca8eeef39321d70e7d414331
Committed: https://crrev.com/3f936a5b17754783e92d2146eaf66c88a78ee45b
Cr-Original-Original-Commit-Position: refs/heads/master@{#38270}
Cr-Original-Commit-Position: refs/heads/master@{#38314}
Cr-Commit-Position: refs/heads/master@{#38403}
==========

to

==========
[Reland][Tracing] Embed V8 runtime call stats into tracing.

Currently we have V8 RuntimeCallStats that is independently from tracing when
running d8 with flag --runtime_call_stats. This patch embeds V8 runtime call
stats into tracing, by having a global table of runtime call counters each
isolate, resetting the table each time we enter a top level trace event, and
dumping the table for each top level trace event. This will make trace file more
compat, as well as enable runtime call stats in tracing system.

This patch adds ~5% overhead to V8 when the category is enabled, we measure the
overhead by running a script when category is enabled.

BUG=v8:5089

Committed: https://crrev.com/d014866173eaa2b548c566217b2c94b1d49385fa
Committed: https://crrev.com/1ca3b73bba4a7253ca8eeef39321d70e7d414331
Committed: https://crrev.com/3f936a5b17754783e92d2146eaf66c88a78ee45b
Cr-Original-Original-Commit-Position: refs/heads/master@{#38270}
Cr-Original-Commit-Position: refs/heads/master@{#38314}
Cr-Commit-Position: refs/heads/master@{#38403}
==========

[commit-bot: I haz the power, 2016-08-10 01:18:49]

Committed patchset #9 (id:160001)

[commit-bot: I haz the power, 2016-08-10 01:19:09]

Description was changed from

==========
[Reland][Tracing] Embed V8 runtime call stats into tracing.

Currently we have V8 RuntimeCallStats that is independently from tracing when
running d8 with flag --runtime_call_stats. This patch embeds V8 runtime call
stats into tracing, by having a global table of runtime call counters each
isolate, resetting the table each time we enter a top level trace event, and
dumping the table for each top level trace event. This will make trace file more
compat, as well as enable runtime call stats in tracing system.

This patch adds ~5% overhead to V8 when the category is enabled, we measure the
overhead by running a script when category is enabled.

BUG=v8:5089

Committed: https://crrev.com/d014866173eaa2b548c566217b2c94b1d49385fa
Committed: https://crrev.com/1ca3b73bba4a7253ca8eeef39321d70e7d414331
Committed: https://crrev.com/3f936a5b17754783e92d2146eaf66c88a78ee45b
Cr-Original-Original-Commit-Position: refs/heads/master@{#38270}
Cr-Original-Commit-Position: refs/heads/master@{#38314}
Cr-Commit-Position: refs/heads/master@{#38403}
==========

to

==========
[Reland][Tracing] Embed V8 runtime call stats into tracing.

Currently we have V8 RuntimeCallStats that is independently from tracing when
running d8 with flag --runtime_call_stats. This patch embeds V8 runtime call
stats into tracing, by having a global table of runtime call counters each
isolate, resetting the table each time we enter a top level trace event, and
dumping the table for each top level trace event. This will make trace file more
compat, as well as enable runtime call stats in tracing system.

This patch adds ~5% overhead to V8 when the category is enabled, we measure the
overhead by running a script when category is enabled.

BUG=v8:5089

Committed: https://crrev.com/d014866173eaa2b548c566217b2c94b1d49385fa
Committed: https://crrev.com/1ca3b73bba4a7253ca8eeef39321d70e7d414331
Committed: https://crrev.com/3f936a5b17754783e92d2146eaf66c88a78ee45b
Committed: https://crrev.com/7a3631e7e12fe0e5038d80c58f3f491c61977b95
Cr-Original-Original-Original-Commit-Position: refs/heads/master@{#38270}
Cr-Original-Original-Commit-Position: refs/heads/master@{#38314}
Cr-Original-Commit-Position: refs/heads/master@{#38403}
Cr-Commit-Position: refs/heads/master@{#38510}
==========

[commit-bot: I haz the power, 2016-08-10 01:19:11]

Patchset 9 (id:??) landed as
https://crrev.com/7a3631e7e12fe0e5038d80c58f3f491c61977b95
Cr-Commit-Position: refs/heads/master@{#38510}