The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/6c1dc894125c68f12d3f14065d057d29c0229860

commit 6c1dc894125c68f12d3f14065d057d29c0229860
Author: ssid <ssid@chromium.org>
Date: Mon Mar 28 19:56:49 2016

Move TRACE_EVENT in IPC message dispatch to add message name

The IPC message names are not displayed unless logging is enabled
(debug builds) in ChannelProxy::Context::OnDispatchMessage. But the
name is actually available in release build binary. Specific message
names are required for memory-infra heap profiler to understand the
memory usage in Chrome, since the task name just shows DispatchMessage
for all ipc messages. More context: https://goo.gl/hcqY5p.

This CL adds the existing names in the chrome binary to tracing when ipc
category is enabled.

BUG=594803

Review URL: https://codereview.chromium.org/1833573002

Cr-Commit-Position: refs/heads/master@{#383546}

[modify] https://crrev.com/6c1dc894125c68f12d3f14065d057d29c0229860/chrome/common/trace_event_args_whitelist.cc
[modify] https://crrev.com/6c1dc894125c68f12d3f14065d057d29c0229860/ipc/ipc_channel_proxy.cc
[modify] https://crrev.com/6c1dc894125c68f12d3f14065d057d29c0229860/ipc/ipc_message_templates.h

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/b3fabc66203904dd110597b11ebb89b89697f1a9

commit b3fabc66203904dd110597b11ebb89b89697f1a9
Author: ssid <ssid@chromium.org>
Date: Fri Apr 01 04:39:25 2016

[tracing] Add thread name to the pseudo stack of heap profiler

The thread name that allocated the memory is very useful to understand
the memory usage of the process. This CL adds it to the top of the
pseudo stack trace.
The name is set from ThreadIdNameManager since it leaks the thread name.
Since getting the name from AllocationContextTracker can cause rentrancy
into ThreadIdNameManager, the name is passed from ThreadIdNameManager.

BUG=594803

Review URL: https://codereview.chromium.org/1814083002

Cr-Commit-Position: refs/heads/master@{#384484}

[modify] https://crrev.com/b3fabc66203904dd110597b11ebb89b89697f1a9/base/threading/thread_id_name_manager.cc
[modify] https://crrev.com/b3fabc66203904dd110597b11ebb89b89697f1a9/base/trace_event/heap_profiler_allocation_context_tracker.cc
[modify] https://crrev.com/b3fabc66203904dd110597b11ebb89b89697f1a9/base/trace_event/heap_profiler_allocation_context_tracker.h
[modify] https://crrev.com/b3fabc66203904dd110597b11ebb89b89697f1a9/base/trace_event/heap_profiler_allocation_context_tracker_unittest.cc
[modify] https://crrev.com/b3fabc66203904dd110597b11ebb89b89697f1a9/content/browser/browser_main_runner.cc

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/533c402f8b21e503ebcaf3f0de427f41076b4de1

commit 533c402f8b21e503ebcaf3f0de427f41076b4de1
Author: ssid <ssid@chromium.org>
Date: Fri Apr 01 08:41:51 2016

[tracing] Adding task information to heap profiler

This CL adds more information to the malloc heap profiler added in
crrev.com/1675183006. The heap profiler only shows trace events and all
the tasks show up as "MessageLoop::Runtask".

This CL does:
 - Moves TRACE_TASK_EXECUTION macro from task_annotator to
   trace_event.h.
 - Adds extra scoped trace event to track the task context in heap
   profiler.
 - The folder where the task was posted from is used to categorize
   allocations.
 - Uses the type_name for context temporarily till we define a new
   context dimension for allocations.

BUG=594803

Review URL: https://codereview.chromium.org/1784133002

Cr-Commit-Position: refs/heads/master@{#384521}

[modify] https://crrev.com/533c402f8b21e503ebcaf3f0de427f41076b4de1/base/debug/task_annotator.h
[modify] https://crrev.com/533c402f8b21e503ebcaf3f0de427f41076b4de1/base/threading/sequenced_worker_pool.cc
[modify] https://crrev.com/533c402f8b21e503ebcaf3f0de427f41076b4de1/base/threading/worker_pool_posix.cc
[modify] https://crrev.com/533c402f8b21e503ebcaf3f0de427f41076b4de1/base/threading/worker_pool_win.cc
[modify] https://crrev.com/533c402f8b21e503ebcaf3f0de427f41076b4de1/base/trace_event/common/trace_event_common.h
[modify] https://crrev.com/533c402f8b21e503ebcaf3f0de427f41076b4de1/base/trace_event/heap_profiler_allocation_context_tracker.cc
[modify] https://crrev.com/533c402f8b21e503ebcaf3f0de427f41076b4de1/base/trace_event/heap_profiler_allocation_context_tracker.h
[modify] https://crrev.com/533c402f8b21e503ebcaf3f0de427f41076b4de1/base/trace_event/heap_profiler_allocation_context_tracker_unittest.cc
[modify] https://crrev.com/533c402f8b21e503ebcaf3f0de427f41076b4de1/base/trace_event/heap_profiler_type_name_deduplicator.cc
[modify] https://crrev.com/533c402f8b21e503ebcaf3f0de427f41076b4de1/base/trace_event/heap_profiler_type_name_deduplicator_unittest.cc
[modify] https://crrev.com/533c402f8b21e503ebcaf3f0de427f41076b4de1/base/trace_event/trace_event.h
[modify] https://crrev.com/533c402f8b21e503ebcaf3f0de427f41076b4de1/components/scheduler/base/task_queue_manager.cc

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/7a3a6793b5a471280e64842060d565e6f52e939f

commit 7a3a6793b5a471280e64842060d565e6f52e939f
Author: kbr <kbr@chromium.org>
Date: Fri Apr 01 21:45:46 2016

Revert of [tracing] Adding task information to heap profiler (patchset #18 id:630001 of https://codereview.chromium.org/1784133002/ )

Reason for revert:
Caused flakiness of trace_test; see  Issue 599794 .

Original issue's description:
> [tracing] Adding task information to heap profiler
>
> This CL adds more information to the malloc heap profiler added in
> crrev.com/1675183006. The heap profiler only shows trace events and all
> the tasks show up as "MessageLoop::Runtask".
>
> This CL does:
>  - Moves TRACE_TASK_EXECUTION macro from task_annotator to
>    trace_event.h.
>  - Adds extra scoped trace event to track the task context in heap
>    profiler.
>  - The folder where the task was posted from is used to categorize
>    allocations.
>  - Uses the type_name for context temporarily till we define a new
>    context dimension for allocations.
>
> BUG=594803
>
> Committed: https://crrev.com/533c402f8b21e503ebcaf3f0de427f41076b4de1
> Cr-Commit-Position: refs/heads/master@{#384521}

TBR=jochen@chromium.org,dskiba@google.com,oysteine@chromium.org,petrcermak@chromium.org,primiano@chromium.org,skyostil@chromium.org,thakis@chromium.org,ssid@chromium.org
# Skipping CQ checks because original CL landed less than 1 days ago.
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=594803

Review URL: https://codereview.chromium.org/1846333002

Cr-Commit-Position: refs/heads/master@{#384695}

[modify] https://crrev.com/7a3a6793b5a471280e64842060d565e6f52e939f/base/debug/task_annotator.h
[modify] https://crrev.com/7a3a6793b5a471280e64842060d565e6f52e939f/base/threading/sequenced_worker_pool.cc
[modify] https://crrev.com/7a3a6793b5a471280e64842060d565e6f52e939f/base/threading/worker_pool_posix.cc
[modify] https://crrev.com/7a3a6793b5a471280e64842060d565e6f52e939f/base/threading/worker_pool_win.cc
[modify] https://crrev.com/7a3a6793b5a471280e64842060d565e6f52e939f/base/trace_event/common/trace_event_common.h
[modify] https://crrev.com/7a3a6793b5a471280e64842060d565e6f52e939f/base/trace_event/heap_profiler_allocation_context_tracker.cc
[modify] https://crrev.com/7a3a6793b5a471280e64842060d565e6f52e939f/base/trace_event/heap_profiler_allocation_context_tracker.h
[modify] https://crrev.com/7a3a6793b5a471280e64842060d565e6f52e939f/base/trace_event/heap_profiler_allocation_context_tracker_unittest.cc
[modify] https://crrev.com/7a3a6793b5a471280e64842060d565e6f52e939f/base/trace_event/heap_profiler_type_name_deduplicator.cc
[modify] https://crrev.com/7a3a6793b5a471280e64842060d565e6f52e939f/base/trace_event/heap_profiler_type_name_deduplicator_unittest.cc
[modify] https://crrev.com/7a3a6793b5a471280e64842060d565e6f52e939f/base/trace_event/trace_event.h
[modify] https://crrev.com/7a3a6793b5a471280e64842060d565e6f52e939f/components/scheduler/base/task_queue_manager.cc

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/092c384b364d376d8ef4971cf43a5b11aa2884cf

commit 092c384b364d376d8ef4971cf43a5b11aa2884cf
Author: ssid <ssid@chromium.org>
Date: Sat Apr 02 04:22:07 2016

Reland of [tracing] Adding task information to heap profiler (patchset #1 id:1 of https://codereview.chromium.org/1846333002/ )

Reason for revert:
Fixing the flakiness in trace_test by removing the event from
sequenced_worker_pool.cc. 2 trace events together somehow causes multiple
calls to UpdateDuration. I couldn't reproduce yet.
Also because 2 trace events there is not necessary, I will figure out a
way to add WorkerPool task information in next CL.

Original issue's description:
> Revert of [tracing] Adding task information to heap profiler (patchset #18 id:630001 of https://codereview.chromium.org/1784133002/ )
>
> Reason for revert:
> Caused flakiness of trace_test; see  Issue 599794 .
>
> Original issue's description:
> > [tracing] Adding task information to heap profiler
> >
> > This CL adds more information to the malloc heap profiler added in
> > crrev.com/1675183006. The heap profiler only shows trace events and all
> > the tasks show up as "MessageLoop::Runtask".
> >
> > This CL does:
> >  - Moves TRACE_TASK_EXECUTION macro from task_annotator to
> >    trace_event.h.
> >  - Adds extra scoped trace event to track the task context in heap
> >    profiler.
> >  - The folder where the task was posted from is used to categorize
> >    allocations.
> >  - Uses the type_name for context temporarily till we define a new
> >    context dimension for allocations.
> >
> > BUG=594803
> >
> > Committed: https://crrev.com/533c402f8b21e503ebcaf3f0de427f41076b4de1
> > Cr-Commit-Position: refs/heads/master@{#384521}
>
> TBR=jochen@chromium.org,dskiba@google.com,oysteine@chromium.org,petrcermak@chromium.org,primiano@chromium.org,skyostil@chromium.org,thakis@chromium.org,ssid@chromium.org
> # Skipping CQ checks because original CL landed less than 1 days ago.
> NOPRESUBMIT=true
> NOTREECHECKS=true
> NOTRY=true
> BUG=594803
>
> Committed: https://crrev.com/7a3a6793b5a471280e64842060d565e6f52e939f
> Cr-Commit-Position: refs/heads/master@{#384695}

TBR=jochen@chromium.org,dskiba@google.com,oysteine@chromium.org,petrcermak@chromium.org,primiano@chromium.org,skyostil@chromium.org,thakis@chromium.org,kbr@chromium.org
# Skipping CQ checks because original CL landed less than 1 days ago.
NOPRESUBMIT=true
NOTREECHECKS=true
BUG=594803

Review URL: https://codereview.chromium.org/1851233002

Cr-Commit-Position: refs/heads/master@{#384795}

[modify] https://crrev.com/092c384b364d376d8ef4971cf43a5b11aa2884cf/base/debug/task_annotator.h
[modify] https://crrev.com/092c384b364d376d8ef4971cf43a5b11aa2884cf/base/threading/worker_pool_posix.cc
[modify] https://crrev.com/092c384b364d376d8ef4971cf43a5b11aa2884cf/base/threading/worker_pool_win.cc
[modify] https://crrev.com/092c384b364d376d8ef4971cf43a5b11aa2884cf/base/trace_event/common/trace_event_common.h
[modify] https://crrev.com/092c384b364d376d8ef4971cf43a5b11aa2884cf/base/trace_event/heap_profiler_allocation_context_tracker.cc
[modify] https://crrev.com/092c384b364d376d8ef4971cf43a5b11aa2884cf/base/trace_event/heap_profiler_allocation_context_tracker.h
[modify] https://crrev.com/092c384b364d376d8ef4971cf43a5b11aa2884cf/base/trace_event/heap_profiler_allocation_context_tracker_unittest.cc
[modify] https://crrev.com/092c384b364d376d8ef4971cf43a5b11aa2884cf/base/trace_event/heap_profiler_type_name_deduplicator.cc
[modify] https://crrev.com/092c384b364d376d8ef4971cf43a5b11aa2884cf/base/trace_event/heap_profiler_type_name_deduplicator_unittest.cc
[modify] https://crrev.com/092c384b364d376d8ef4971cf43a5b11aa2884cf/base/trace_event/trace_event.h
[modify] https://crrev.com/092c384b364d376d8ef4971cf43a5b11aa2884cf/components/scheduler/base/task_queue_manager.cc

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/ee345086542ba00bae935fcfb13780ac5bdb63f1

commit ee345086542ba00bae935fcfb13780ac5bdb63f1
Author: ssid <ssid@chromium.org>
Date: Wed Sep 07 07:28:51 2016

Heap Profiler: Add trace category group names as type names for allocations

The category for allocations are decided using the file that posted
current task. But, this does not work always since sometimes the code
is invoked by event listeners and tasks are not posted. To work around
this issue, the category name from trace events is marked as the
category for allocations as fallback.

BUG=594803

Review-Url: https://codereview.chromium.org/2272843002
Cr-Commit-Position: refs/heads/master@{#416873}

[modify] https://crrev.com/ee345086542ba00bae935fcfb13780ac5bdb63f1/base/trace_event/heap_profiler_allocation_context_tracker.cc
[modify] https://crrev.com/ee345086542ba00bae935fcfb13780ac5bdb63f1/base/trace_event/heap_profiler_allocation_context_tracker.h
[modify] https://crrev.com/ee345086542ba00bae935fcfb13780ac5bdb63f1/base/trace_event/heap_profiler_allocation_context_tracker_unittest.cc
[modify] https://crrev.com/ee345086542ba00bae935fcfb13780ac5bdb63f1/base/trace_event/heap_profiler_type_name_deduplicator.cc
[modify] https://crrev.com/ee345086542ba00bae935fcfb13780ac5bdb63f1/base/trace_event/trace_log.cc

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/ee345086542ba00bae935fcfb13780ac5bdb63f1

commit ee345086542ba00bae935fcfb13780ac5bdb63f1
Author: ssid <ssid@chromium.org>
Date: Wed Sep 07 07:28:51 2016

Heap Profiler: Add trace category group names as type names for allocations

The category for allocations are decided using the file that posted
current task. But, this does not work always since sometimes the code
is invoked by event listeners and tasks are not posted. To work around
this issue, the category name from trace events is marked as the
category for allocations as fallback.

BUG=594803

Review-Url: https://codereview.chromium.org/2272843002
Cr-Commit-Position: refs/heads/master@{#416873}

[modify] https://crrev.com/ee345086542ba00bae935fcfb13780ac5bdb63f1/base/trace_event/heap_profiler_allocation_context_tracker.cc
[modify] https://crrev.com/ee345086542ba00bae935fcfb13780ac5bdb63f1/base/trace_event/heap_profiler_allocation_context_tracker.h
[modify] https://crrev.com/ee345086542ba00bae935fcfb13780ac5bdb63f1/base/trace_event/heap_profiler_allocation_context_tracker_unittest.cc
[modify] https://crrev.com/ee345086542ba00bae935fcfb13780ac5bdb63f1/base/trace_event/heap_profiler_type_name_deduplicator.cc
[modify] https://crrev.com/ee345086542ba00bae935fcfb13780ac5bdb63f1/base/trace_event/trace_log.cc

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/e6b9ddc6d7301c74c8e82789e78081a7d4c520fb

commit e6b9ddc6d7301c74c8e82789e78081a7d4c520fb
Author: ssid <ssid@chromium.org>
Date: Thu Feb 09 21:53:35 2017

Add context for tasks run by TaskGraphRunner

Other task runners use TRACE_TASK_EXECUTION macro which records context
for heap profiler. TaskGraphRunner does not have posted from
information. So, record the context for the tasks as "cc".

BUG=594803
CQ_INCLUDE_TRYBOTS=master.tryserver.blink:linux_trusty_blink_rel

Review-Url: https://codereview.chromium.org/2687553004
Cr-Commit-Position: refs/heads/master@{#449428}

[modify] https://crrev.com/e6b9ddc6d7301c74c8e82789e78081a7d4c520fb/cc/raster/synchronous_task_graph_runner.cc

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/27cd72d3ccb96dc69f03b968cb47aae4801a8db1

commit 27cd72d3ccb96dc69f03b968cb47aae4801a8db1
Author: ssid <ssid@chromium.org>
Date: Thu Feb 16 06:39:46 2017

Use TRACK_RUN_IN_THIS_SCOPED_REGION for heap profiler

Currently the heap profiler records the current task context using the
posted from location. But for ipc messages all the tasks are posted
from ipc.h. TRACK_RUN_IN_THIS_SCOPED_REGION solves the same problem for
showing the posted from info in about:profiler. Use this macro for heap
profiling as well.

BUG=594803

Review-Url: https://codereview.chromium.org/2695833004
Cr-Commit-Position: refs/heads/master@{#450880}

[modify] https://crrev.com/27cd72d3ccb96dc69f03b968cb47aae4801a8db1/base/profiler/scoped_profile.h

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/36b5511e4f35e3063317921e3b623cec5dbe0caa

commit 36b5511e4f35e3063317921e3b623cec5dbe0caa
Author: ssid <ssid@chromium.org>
Date: Thu Feb 16 08:59:42 2017

Uncomment tracking macro in IPC_MESSAGE_HANDLER_GENERIC

This was commented by codereview.chromium.org/8463033 because of
failures in ASAN bots. The reason says some hint about some message not
supplying |code| argument. Currently we do not have any code that calls
this macro with empty |code|. So, I am guessing that this issue is
fixed.

If this causes ASAN problems with browser tests, then
I'll revert it.

BUG=594803

Review-Url: https://codereview.chromium.org/2692123004
Cr-Commit-Position: refs/heads/master@{#450905}

[modify] https://crrev.com/36b5511e4f35e3063317921e3b623cec5dbe0caa/ipc/ipc_message_macros.h

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/5dd4fb35920fffa84d53cb908b83668b2c7db1b8

commit 5dd4fb35920fffa84d53cb908b83668b2c7db1b8
Author: ssid <ssid@chromium.org>
Date: Thu Feb 16 23:57:17 2017

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}

[modify] https://crrev.com/5dd4fb35920fffa84d53cb908b83668b2c7db1b8/base/files/file_descriptor_watcher_posix.cc
[modify] https://crrev.com/5dd4fb35920fffa84d53cb908b83668b2c7db1b8/base/message_loop/message_loop_unittest.cc
[modify] https://crrev.com/5dd4fb35920fffa84d53cb908b83668b2c7db1b8/base/message_loop/message_pump_io_ios.cc
[modify] https://crrev.com/5dd4fb35920fffa84d53cb908b83668b2c7db1b8/base/message_loop/message_pump_io_ios.h
[modify] https://crrev.com/5dd4fb35920fffa84d53cb908b83668b2c7db1b8/base/message_loop/message_pump_io_ios_unittest.cc
[modify] https://crrev.com/5dd4fb35920fffa84d53cb908b83668b2c7db1b8/base/message_loop/message_pump_libevent.cc
[modify] https://crrev.com/5dd4fb35920fffa84d53cb908b83668b2c7db1b8/base/message_loop/message_pump_libevent.h
[modify] https://crrev.com/5dd4fb35920fffa84d53cb908b83668b2c7db1b8/base/message_loop/message_pump_libevent_unittest.cc
[modify] https://crrev.com/5dd4fb35920fffa84d53cb908b83668b2c7db1b8/chrome/browser/apps/app_shim/unix_domain_socket_acceptor.cc
[modify] https://crrev.com/5dd4fb35920fffa84d53cb908b83668b2c7db1b8/chrome/browser/chrome_browser_main_extra_parts_exo.cc
[modify] https://crrev.com/5dd4fb35920fffa84d53cb908b83668b2c7db1b8/chrome/common/service_process_util_posix.cc
[modify] https://crrev.com/5dd4fb35920fffa84d53cb908b83668b2c7db1b8/chromeos/binder/ipc_thread.cc
[modify] https://crrev.com/5dd4fb35920fffa84d53cb908b83668b2c7db1b8/components/crash/content/browser/crash_handler_host_linux.cc
[modify] https://crrev.com/5dd4fb35920fffa84d53cb908b83668b2c7db1b8/dbus/bus.cc
[modify] https://crrev.com/5dd4fb35920fffa84d53cb908b83668b2c7db1b8/mojo/edk/system/channel_posix.cc
[modify] https://crrev.com/5dd4fb35920fffa84d53cb908b83668b2c7db1b8/net/base/address_tracker_linux.cc
[modify] https://crrev.com/5dd4fb35920fffa84d53cb908b83668b2c7db1b8/net/dns/notify_watcher_mac.cc
[modify] https://crrev.com/5dd4fb35920fffa84d53cb908b83668b2c7db1b8/net/proxy/proxy_config_service_linux.cc
[modify] https://crrev.com/5dd4fb35920fffa84d53cb908b83668b2c7db1b8/net/socket/socket_posix.cc
[modify] https://crrev.com/5dd4fb35920fffa84d53cb908b83668b2c7db1b8/net/socket/udp_socket_posix.cc
[modify] https://crrev.com/5dd4fb35920fffa84d53cb908b83668b2c7db1b8/ui/events/ozone/device/udev/device_manager_udev.cc
[modify] https://crrev.com/5dd4fb35920fffa84d53cb908b83668b2c7db1b8/ui/events/ozone/evdev/event_converter_evdev.cc
[modify] https://crrev.com/5dd4fb35920fffa84d53cb908b83668b2c7db1b8/ui/events/ozone/evdev/event_converter_evdev_impl.cc
[modify] https://crrev.com/5dd4fb35920fffa84d53cb908b83668b2c7db1b8/ui/events/ozone/evdev/tablet_event_converter_evdev.cc
[modify] https://crrev.com/5dd4fb35920fffa84d53cb908b83668b2c7db1b8/ui/events/platform/x11/x11_event_source_libevent.cc
[modify] https://crrev.com/5dd4fb35920fffa84d53cb908b83668b2c7db1b8/ui/ozone/platform/drm/gpu/drm_device.cc
[modify] https://crrev.com/5dd4fb35920fffa84d53cb908b83668b2c7db1b8/ui/ozone/platform/wayland/fake_server.cc
[modify] https://crrev.com/5dd4fb35920fffa84d53cb908b83668b2c7db1b8/ui/ozone/platform/wayland/wayland_connection.cc

Currently we categorize 90% of the allocations on Android. The 10% is calls from jni methods which are not tracked by FROM_HERE. On Linux we do better than 90% in most of the test runs. So, lowering priority and probably won't get to this isue for sometime.

I'll soon run an experiment to check how accurately our (pseudo) categorization is.