Use the location where FileDescriptorWatcher is created to track the libevent handlers

Use the location where FileDescriptorWatcher is created to track the libevent handlers

Currently the trace events are without the location information and it
is of less use without the location for heap profiler. The libevent
handlers are now annotated with the location where the watcher
controller is created for tracing and heap profiler.

This only handles the posix version of watcher events. Windows will be
fixed in different CL.

BUG=609912
Review-Url: https://codereview.chromium.org/2695593009
Cr-Commit-Position: refs/heads/master@{#451142}
Committed: https://chromium.googlesource.com/chromium/src/+/5dd4fb35920fffa84d53cb908b83668b2c7db1b8

[ssid, 2017-02-16 03:01:02]

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

[commit-bot: I haz the power, 2017-02-16 03:03:56]

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

[ssid, 2017-02-16 03:04:29]

Description was changed from

==========
Use the location where FileDescriptorWatcher is created to track the libevent
handlers

Currently the events are without the location information and it is of
less use without the location for heap profiler. The libevent handlers
are annotated with the location where the watcher controller is created
for tracing and heap profiler.

BUG=609912
==========

to

==========
Use the location where FileDescriptorWatcher is created to track the libevent
handlers

Currently the trace events are without the location information and it
is of less use without the location for heap profiler. The libevent
handlers are now annotated with the location where the watcher
controller is created for tracing and heap profiler.

This only handles the posix version of watcher events. Windows will be
fixed in different CL.

BUG=609912
==========

[commit-bot: I haz the power, 2017-02-16 03:16:11]

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

[commit-bot: I haz the power, 2017-02-16 03:16:12]

Dry run: Try jobs failed on following builders:
  ios-device on master.tryserver.chromium.mac (JOB_FAILED,
http://build.chromium.org/p/tryserver.chromium.mac/builders/ios-device/builds...)
  ios-simulator-xcode-clang on master.tryserver.chromium.mac (JOB_FAILED,
http://build.chromium.org/p/tryserver.chromium.mac/builders/ios-simulator-xco...)

[ssid, 2017-02-16 05:19:53]

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

[ssid, 2017-02-16 05:20:36]

Patchset #1 (id:1) has been deleted

[commit-bot: I haz the power, 2017-02-16 05:20:51]

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, 2017-02-16 05:53:17]

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

[commit-bot: I haz the power, 2017-02-16 05:53:21]

Dry run: Try jobs failed on following builders:
  mac_chromium_compile_dbg_ng on master.tryserver.chromium.mac (JOB_FAILED,
http://build.chromium.org/p/tryserver.chromium.mac/builders/mac_chromium_comp...)

[ssid, 2017-02-16 18:00:02]

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

[commit-bot: I haz the power, 2017-02-16 18:00:41]

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

[ssid, 2017-02-16 18:00:50]

ssid@chromium.org changed reviewers:
+ thakis@chromium.org

[ssid, 2017-02-16 18:00:51]

Nico, wdyt?

[Nico, 2017-02-16 18:37:40]

seems fine to me (lgtm), but it'd be cool if someone closer to tracing / perf
stuff could have a short look too (one of erikchen primiano dskiba & co)

[ssid, 2017-02-16 18:40:07]

ssid@chromium.org changed reviewers:
+ oysteine@chromium.org

[ssid, 2017-02-16 18:40:07]

+oysteine for the tracing change wdyt?

[commit-bot: I haz the power, 2017-02-16 18:43:32]

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

[commit-bot: I haz the power, 2017-02-16 18:43:33]

Dry run: Try jobs failed on following builders:
  mac_chromium_compile_dbg_ng on master.tryserver.chromium.mac (JOB_FAILED,
http://build.chromium.org/p/tryserver.chromium.mac/builders/mac_chromium_comp...)

[ssid, 2017-02-16 18:46:54]

ssid@chromium.org changed reviewers:
+ dskiba@chromium.org, rockot@chromium.org, rsesek@chromium.org,
stevenjb@chromium.org, xunjieli@chromium.org

[ssid, 2017-02-16 18:46:54]

Small changes for tracing, ptal. Thanks.

+rsesek for components/crash

+rockot for mojo/

+xunjieli for net/

+stevenjb for dbus/ and binder/

+dskiba This fixes the unknown categories for native mode :)

[Ken Rockot(use gerrit already), 2017-02-16 18:48:16]

mojo rs lgtm

[Robert Sesek, 2017-02-16 18:49:14]

lgtm

[xunjieli, 2017-02-16 18:56:03]

net/ lgtm.

[ssid, 2017-02-16 19:26:38]

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

[commit-bot: I haz the power, 2017-02-16 19:27:40]

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

[stevenjb, 2017-02-16 19:38:26]

owner lgtm

[oystein (OOO til 10th of July), 2017-02-16 21:02:31]

On 2017/02/16 at 18:40:07, ssid wrote:
> +oysteine for the tracing change wdyt?

I'm not too familiar with how FileDescriptorWatcher is used, assuming these
aren't super-frequent events and blow up the trace size, then lgtm.

[commit-bot: I haz the power, 2017-02-16 21:11:52]

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

[commit-bot: I haz the power, 2017-02-16 21:11:53]

Dry run: This issue passed the CQ dry run.

[ssid, 2017-02-16 21:41:53]

On 2017/02/16 21:02:31, oystein wrote:
> On 2017/02/16 at 18:40:07, ssid wrote:
> > +oysteine for the tracing change wdyt?
> 
> I'm not too familiar with how FileDescriptorWatcher is used, assuming these
> aren't super-frequent events and blow up the trace size, then lgtm.

On a simple tracing session they accounted for 5-8% of the tasks. And we record
the location for all tasks already. So, this must not cause a big difference.

[ssid, 2017-02-16 22:34:31]

For memory-infra folks:

The heap profiler categories will change this way on Linux:
View with this CL:
Ⓣnet/socket	2,126.5 KiB
Ⓣcontent/browser/frame_host	1,369.8 KiB
Ⓣnet/disk_cache/simple	899.0 KiB
Ⓣnet/extras/sqlite	812.7 KiB
Ⓣnet/url_request	674.6 KiB
Ⓣ[unknown]	509.5 KiB
Ⓣui/views/corewm	503.6 KiB
Ⓣcc/trees	496.9 KiB
Ⓣcontent/common/service_manager	390.2 KiB
Ⓣmojo/edk/system	328.1 KiB
Ⓣcontent/browser/dom_storage	275.4 KiB
Ⓣmedia::mojom::RemoterFactory	237.4 KiB
Ⓣcontent/browser/fileapi	237.1 KiB
Ⓣnet/dns	231.3 KiB

Without this CL:
Ⓣ[unknown]	3,173.1 KiB
Ⓣchrome/browser	1,621.8 KiB
Ⓣcontent/browser/frame_host	1,315.4 KiB
Ⓣnet/extras/sqlite	807.0 KiB
Ⓣnet/url_request	674.3 KiB
Ⓣcc/trees	506.8 KiB
Ⓣui/views/corewm	487.4 KiB
Ⓣnet/disk_cache/simple	446.1 KiB
Ⓣcontent/common/service_manager	385.2 KiB
Ⓣnet/socket	318.8 KiB
Ⓣcontent/browser/dom_storage	291.6 KiB
Ⓣcontent/browser/service_worker	236.8 KiB
Ⓣmedia::mojom::RemoterFactory	231.9 KiB
Ⓣnet/dns	224.0 KiB


Observable changes are: Unknown changes to net/socket and mojo/edk/system

On Android:
With this CL:
Ⓣcc/trees	1,274.7 KiB
Ⓣnet/socket	325.6 KiB
Ⓣcontent/browser/devtools	310.8 KiB
Ⓣchrome/browser/android/thumbnail	233.1 KiB
Ⓣ[unknown]	155.4 KiB
Ⓣnet/cert	95.6 KiB
Ⓣmojo/edk/system	80.5 KiB
Ⓣnet/disk_cache/simple	71.4 KiB
Ⓣcc/scheduler	46.5 KiB


Without CL:
Ⓣcc/trees	1,466.9 KiB
Ⓣ[unknown]	330.2 KiB
Ⓣcontent/browser/devtools	310.9 KiB
Ⓣnet/socket	298.2 KiB
Ⓣnet/disk_cache/simple	53.6 KiB
Ⓣcc/scheduler	46.3 KiB
Ⓣnet/http	43.3 KiB
Ⓣnet/cert	42.8 KiB
Ⓣcc/surfaces	40.5 KiB


Only observable changes are unknown decreases and net/socket increases.

[ssid, 2017-02-16 22:34:42]

The CQ bit was checked by ssid@chromium.org

[ssid, 2017-02-16 22:34:43]

The patchset sent to the CQ was uploaded after l-g-t-m from rockot@chromium.org,
thakis@chromium.org, rsesek@chromium.org
Link to the patchset: https://codereview.chromium.org/2695593009/#ps60001
(title: "fix mac.")

[commit-bot: I haz the power, 2017-02-16 22:37:20]

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, 2017-02-16 23:57:22]

CQ is committing da patch.

Bot data: {"patchset_id": 60001, "attempt_start_ts": 1487284482448610,
"parent_rev": "9caf6364ab67c04029f037ffecb96169b681a469", "commit_rev":
"5dd4fb35920fffa84d53cb908b83668b2c7db1b8"}

[commit-bot: I haz the power, 2017-02-16 23:58:21]

Description was changed from

==========
Use the location where FileDescriptorWatcher is created to track the libevent
handlers

Currently the trace events are without the location information and it
is of less use without the location for heap profiler. The libevent
handlers are now annotated with the location where the watcher
controller is created for tracing and heap profiler.

This only handles the posix version of watcher events. Windows will be
fixed in different CL.

BUG=609912
==========

to

==========
Use the location where FileDescriptorWatcher is created to track the libevent
handlers

Currently the trace events are without the location information and it
is of less use without the location for heap profiler. The libevent
handlers are now annotated with the location where the watcher
controller is created for tracing and heap profiler.

This only handles the posix version of watcher events. Windows will be
fixed in different CL.

BUG=609912

Review-Url: https://codereview.chromium.org/2695593009
Cr-Commit-Position: refs/heads/master@{#451142}
Committed:
https://chromium.googlesource.com/chromium/src/+/5dd4fb35920fffa84d53cb908b83...
==========

[commit-bot: I haz the power, 2017-02-16 23:58:24]

Committed patchset #3 (id:60001) as
https://chromium.googlesource.com/chromium/src/+/5dd4fb35920fffa84d53cb908b83...