Track DisplayItemClient aliveness for each PaintController

Track DisplayItemClient aliveness for each PaintController

After CHECK_DISPLAY_ITEM_CLIENT_ALIVENESS was enabled, we didn't get
crashes in DisplayItemClient::~DisplayItemClient() when the
DisplayItemClient should be alive. We still get crashes in
PaintController::updateCacheGeneration() [1]
which indicates that we still have short-lived DisplayItemClient but
not catched by aliveness checking. This might happen if there are
multiple PaintControllers (e.g. one created by SkPictureBuilder and
another by GraphicsLayer).

Separately track DisplayItemClient aliveness for each PaintController
in order to catch the issue.

[1] https://crash.corp.google.com/browse?q=custom_data.ChromeCrashProto.ptype%3D%27renderer%27%20AND%20custom_data.ChromeCrashProto.magic_signature_1.name%20like%20%27blink%3A%3APaintController%3A%3A%25%27%20AND%20product.version%3D%2753.0.2751.0%27&ignore_case=false&enable_rewrite=true&omit_field_name=&omit_field_value=&omit_field_opt=%3D&stbtiq=&reportid=20bc8a5c00000000&index=2#4

BUG=609218
Committed: https://crrev.com/53559da8fd5faa99d30a00eadbc6da3aebbde2b1
Cr-Commit-Position: refs/heads/master@{#397243}

[Xianzhu, 2016-06-01 17:15:11]

Description was changed from

==========
Track DisplayItemClient aliveness for each PaintController

After CHECK_DISPLAY_ITEM_CLIENT_ALIVENESS was enabled, we didn't get
crashes in DisplayItemClient::~DisplayItemClient() when the
DisplayItemClient should be alive. We still get crashes in
PaintController::updateCacheGeneration() [1]
which indicates that we still have short-lived DisplayItemClient but
not catched by aliveness checking. This might happen if there are
multiple PaintControllers (e.g. one created by SkPictureBuilder and
another by GraphicsLayer).

Separately track DisplayItemClient aliveness for each PaintController
in order to catch the issue.

[1]
https://crash.corp.google.com/browse?q=custom_data.ChromeCrashProto.ptype%3D%...

BUG=609218

x
==========

to

==========
Track DisplayItemClient aliveness for each PaintController

After CHECK_DISPLAY_ITEM_CLIENT_ALIVENESS was enabled, we didn't get
crashes in DisplayItemClient::~DisplayItemClient() when the
DisplayItemClient should be alive. We still get crashes in
PaintController::updateCacheGeneration() [1]
which indicates that we still have short-lived DisplayItemClient but
not catched by aliveness checking. This might happen if there are
multiple PaintControllers (e.g. one created by SkPictureBuilder and
another by GraphicsLayer).

Separately track DisplayItemClient aliveness for each PaintController
in order to catch the issue.

[1]
https://crash.corp.google.com/browse?q=custom_data.ChromeCrashProto.ptype%3D%...

BUG=609218

x
==========

[Xianzhu, 2016-06-01 17:15:11]

wangxianzhu@chromium.org changed reviewers:
+ chrishtr@chromium.org

[Xianzhu, 2016-06-01 17:15:28]

Description was changed from

==========
Track DisplayItemClient aliveness for each PaintController

After CHECK_DISPLAY_ITEM_CLIENT_ALIVENESS was enabled, we didn't get
crashes in DisplayItemClient::~DisplayItemClient() when the
DisplayItemClient should be alive. We still get crashes in
PaintController::updateCacheGeneration() [1]
which indicates that we still have short-lived DisplayItemClient but
not catched by aliveness checking. This might happen if there are
multiple PaintControllers (e.g. one created by SkPictureBuilder and
another by GraphicsLayer).

Separately track DisplayItemClient aliveness for each PaintController
in order to catch the issue.

[1]
https://crash.corp.google.com/browse?q=custom_data.ChromeCrashProto.ptype%3D%...

BUG=609218

x
==========

to

==========
Track DisplayItemClient aliveness for each PaintController

After CHECK_DISPLAY_ITEM_CLIENT_ALIVENESS was enabled, we didn't get
crashes in DisplayItemClient::~DisplayItemClient() when the
DisplayItemClient should be alive. We still get crashes in
PaintController::updateCacheGeneration() [1]
which indicates that we still have short-lived DisplayItemClient but
not catched by aliveness checking. This might happen if there are
multiple PaintControllers (e.g. one created by SkPictureBuilder and
another by GraphicsLayer).

Separately track DisplayItemClient aliveness for each PaintController
in order to catch the issue.

[1]
https://crash.corp.google.com/browse?q=custom_data.ChromeCrashProto.ptype%3D%...

BUG=609218
==========

[chrishtr, 2016-06-01 17:34:11]

lgtm

[Xianzhu, 2016-06-01 20:23:14]

The CQ bit was checked by wangxianzhu@chromium.org

[commit-bot: I haz the power, 2016-06-01 20:24:02]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/2031623002/1
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/2031623002/1

[commit-bot: I haz the power, 2016-06-01 21:14:51]

Description was changed from

==========
Track DisplayItemClient aliveness for each PaintController

After CHECK_DISPLAY_ITEM_CLIENT_ALIVENESS was enabled, we didn't get
crashes in DisplayItemClient::~DisplayItemClient() when the
DisplayItemClient should be alive. We still get crashes in
PaintController::updateCacheGeneration() [1]
which indicates that we still have short-lived DisplayItemClient but
not catched by aliveness checking. This might happen if there are
multiple PaintControllers (e.g. one created by SkPictureBuilder and
another by GraphicsLayer).

Separately track DisplayItemClient aliveness for each PaintController
in order to catch the issue.

[1]
https://crash.corp.google.com/browse?q=custom_data.ChromeCrashProto.ptype%3D%...

BUG=609218
==========

to

==========
Track DisplayItemClient aliveness for each PaintController

After CHECK_DISPLAY_ITEM_CLIENT_ALIVENESS was enabled, we didn't get
crashes in DisplayItemClient::~DisplayItemClient() when the
DisplayItemClient should be alive. We still get crashes in
PaintController::updateCacheGeneration() [1]
which indicates that we still have short-lived DisplayItemClient but
not catched by aliveness checking. This might happen if there are
multiple PaintControllers (e.g. one created by SkPictureBuilder and
another by GraphicsLayer).

Separately track DisplayItemClient aliveness for each PaintController
in order to catch the issue.

[1]
https://crash.corp.google.com/browse?q=custom_data.ChromeCrashProto.ptype%3D%...

BUG=609218
==========

[commit-bot: I haz the power, 2016-06-01 21:14:52]

Committed patchset #1 (id:1)

[commit-bot: I haz the power, 2016-06-01 21:17:07]

Description was changed from

==========
Track DisplayItemClient aliveness for each PaintController

After CHECK_DISPLAY_ITEM_CLIENT_ALIVENESS was enabled, we didn't get
crashes in DisplayItemClient::~DisplayItemClient() when the
DisplayItemClient should be alive. We still get crashes in
PaintController::updateCacheGeneration() [1]
which indicates that we still have short-lived DisplayItemClient but
not catched by aliveness checking. This might happen if there are
multiple PaintControllers (e.g. one created by SkPictureBuilder and
another by GraphicsLayer).

Separately track DisplayItemClient aliveness for each PaintController
in order to catch the issue.

[1]
https://crash.corp.google.com/browse?q=custom_data.ChromeCrashProto.ptype%3D%...

BUG=609218
==========

to

==========
Track DisplayItemClient aliveness for each PaintController

After CHECK_DISPLAY_ITEM_CLIENT_ALIVENESS was enabled, we didn't get
crashes in DisplayItemClient::~DisplayItemClient() when the
DisplayItemClient should be alive. We still get crashes in
PaintController::updateCacheGeneration() [1]
which indicates that we still have short-lived DisplayItemClient but
not catched by aliveness checking. This might happen if there are
multiple PaintControllers (e.g. one created by SkPictureBuilder and
another by GraphicsLayer).

Separately track DisplayItemClient aliveness for each PaintController
in order to catch the issue.

[1]
https://crash.corp.google.com/browse?q=custom_data.ChromeCrashProto.ptype%3D%...

BUG=609218

Committed: https://crrev.com/53559da8fd5faa99d30a00eadbc6da3aebbde2b1
Cr-Commit-Position: refs/heads/master@{#397243}
==========

[commit-bot: I haz the power, 2016-06-01 21:17:08]

Patchset 1 (id:??) landed as
https://crrev.com/53559da8fd5faa99d30a00eadbc6da3aebbde2b1
Cr-Commit-Position: refs/heads/master@{#397243}