Send the decoded size when response completed and stop summing in ResourceLoader::DidReceiveData()

Send the decoded size when response completed and stop summing in ResourceLoader::DidReceiveData()

When the size properties of PerformanceResourceTiming were introduced by
https://crrev.com/2105713002, both encodedBodySize and decodedBodySize were
summed up in ResourceLoader::DidReceiveData().
The CL https://crrev.com/2510333002 changed this to send encodedBodySize when
response completed. And currently only decodedBodySize is summed up in
ResourceLoader::DidReceiveData().

We are planning to directly pass the mojo data pipe to FetchResponseData when
fetch() API is used. If we do so, ResourceLoader::DidReceiveData() will not be
called. And it will not be guaranteed that the renderer process read the all
data before the RequestComplete IPC. This will be a problem because the size
properties are passed to the ResourceTimingInfo structure when the renderer
receives the RequestComplete IPC message.

So this cl change it to send the decodedBodySize when response completed same as
encodedBodySize.

And this is also useful to support the resource timing info for service worker
navigation preload response where we don't use ResourceLoader.

BUG=712809
Review-Url: https://codereview.chromium.org/2835123005
Cr-Commit-Position: refs/heads/master@{#467744}
Committed: https://chromium.googlesource.com/chromium/src/+/e8442e6b8cab49e5e56f24d2779ee48cce71b2d5

[horo, 2017-04-25 08:44:47]

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

[commit-bot: I haz the power, 2017-04-25 08:44:58]

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, 2017-04-25 10:46:40]

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

[commit-bot: I haz the power, 2017-04-25 10:46:40]

Dry run: This issue passed the CQ dry run.

[horo, 2017-04-25 11:38:11]

Description was changed from

==========
Send the decoded body length on DidFinishLoading

BUG=
==========

to

==========
Stop summing the decoded size in ResourceLoader::DidReceiveData and send it when
response completed

When the size properties of PerformanceResourceTiming were introduced by
https://crrev.com/2105713002, both encodedBodySize and decodedBodySize were
summed up in ResourceLoader::DidReceiveData().
The CL https://crrev.com/2510333002 changed this to send encodedBodySize when
response completed. And currently only decodedBodySize is summed up in
ResourceLoader::DidReceiveData().

We are planning to directly pass the mojo data pipe to FetchResponseData when
fetch() API is used. If we do so, ResourceLoader::DidReceiveData() will not be
called. And it will not be guaranteed that the renderer process read the all
data before the RequestComplete IPC. This will be a problem because the size
properties are passed to the ResourceTimingInfo structure when the renderer
receives the RequestComplete IPC message.

So this cl change it to send the decodedBodySize when response completed same as
encodedBodySize.

And this is also useful to support the resource timing info for service worker
navigation preload response.

BUG=712809
==========

[horo, 2017-04-25 11:39:23]

Description was changed from

==========
Stop summing the decoded size in ResourceLoader::DidReceiveData and send it when
response completed

When the size properties of PerformanceResourceTiming were introduced by
https://crrev.com/2105713002, both encodedBodySize and decodedBodySize were
summed up in ResourceLoader::DidReceiveData().
The CL https://crrev.com/2510333002 changed this to send encodedBodySize when
response completed. And currently only decodedBodySize is summed up in
ResourceLoader::DidReceiveData().

We are planning to directly pass the mojo data pipe to FetchResponseData when
fetch() API is used. If we do so, ResourceLoader::DidReceiveData() will not be
called. And it will not be guaranteed that the renderer process read the all
data before the RequestComplete IPC. This will be a problem because the size
properties are passed to the ResourceTimingInfo structure when the renderer
receives the RequestComplete IPC message.

So this cl change it to send the decodedBodySize when response completed same as
encodedBodySize.

And this is also useful to support the resource timing info for service worker
navigation preload response.

BUG=712809
==========

to

==========
Stop summing the decoded size in ResourceLoader::DidReceiveData and send it when
response completed

When the size properties of PerformanceResourceTiming were introduced by
https://crrev.com/2105713002, both encodedBodySize and decodedBodySize were
summed up in ResourceLoader::DidReceiveData().
The CL https://crrev.com/2510333002 changed this to send encodedBodySize when
response completed. And currently only decodedBodySize is summed up in
ResourceLoader::DidReceiveData().

We are planning to directly pass the mojo data pipe to FetchResponseData when
fetch() API is used. If we do so, ResourceLoader::DidReceiveData() will not be
called. And it will not be guaranteed that the renderer process read the all
data before the RequestComplete IPC. This will be a problem because the size
properties are passed to the ResourceTimingInfo structure when the renderer
receives the RequestComplete IPC message.

So this cl change it to send the decodedBodySize when response completed same as
encodedBodySize.

And this is also useful to support the resource timing info for service worker
navigation preload response where we don't use ResourceLoader.

BUG=712809
==========

[horo, 2017-04-25 14:23:57]

Patchset #1 (id:1) has been deleted

[horo, 2017-04-25 14:24:02]

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

[commit-bot: I haz the power, 2017-04-25 14:24:14]

Dry run: CQ is trying da patch.

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

[horo, 2017-04-25 14:29:57]

Description was changed from

==========
Stop summing the decoded size in ResourceLoader::DidReceiveData and send it when
response completed

When the size properties of PerformanceResourceTiming were introduced by
https://crrev.com/2105713002, both encodedBodySize and decodedBodySize were
summed up in ResourceLoader::DidReceiveData().
The CL https://crrev.com/2510333002 changed this to send encodedBodySize when
response completed. And currently only decodedBodySize is summed up in
ResourceLoader::DidReceiveData().

We are planning to directly pass the mojo data pipe to FetchResponseData when
fetch() API is used. If we do so, ResourceLoader::DidReceiveData() will not be
called. And it will not be guaranteed that the renderer process read the all
data before the RequestComplete IPC. This will be a problem because the size
properties are passed to the ResourceTimingInfo structure when the renderer
receives the RequestComplete IPC message.

So this cl change it to send the decodedBodySize when response completed same as
encodedBodySize.

And this is also useful to support the resource timing info for service worker
navigation preload response where we don't use ResourceLoader.

BUG=712809
==========

to

==========
Send the decoded size when response completed and stop summing in
ResourceLoader::DidReceiveData()

When the size properties of PerformanceResourceTiming were introduced by
https://crrev.com/2105713002, both encodedBodySize and decodedBodySize were
summed up in ResourceLoader::DidReceiveData().
The CL https://crrev.com/2510333002 changed this to send encodedBodySize when
response completed. And currently only decodedBodySize is summed up in
ResourceLoader::DidReceiveData().

We are planning to directly pass the mojo data pipe to FetchResponseData when
fetch() API is used. If we do so, ResourceLoader::DidReceiveData() will not be
called. And it will not be guaranteed that the renderer process read the all
data before the RequestComplete IPC. This will be a problem because the size
properties are passed to the ResourceTimingInfo structure when the renderer
receives the RequestComplete IPC message.

So this cl change it to send the decodedBodySize when response completed same as
encodedBodySize.

And this is also useful to support the resource timing info for service worker
navigation preload response where we don't use ResourceLoader.

BUG=712809
==========

[horo, 2017-04-25 14:32:20]

horo@chromium.org changed reviewers:
+ yhirano@chromium.org

[horo, 2017-04-25 14:32:21]

yhirano@
Could you please review this?

[commit-bot: I haz the power, 2017-04-25 16:21:18]

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

[commit-bot: I haz the power, 2017-04-25 16:21:19]

Dry run: This issue passed the CQ dry run.

[yhirano, 2017-04-26 08:17:28]

lgtm

[horo, 2017-04-26 08:33:39]

horo@chromium.org changed reviewers:
+ jam@chromium.org

[horo, 2017-04-26 08:33:40]

jam@
Could you please review this CL?

[jam, 2017-04-26 15:36:44]

content& chrome lgtm

you need someone for the various webkit files you're not owners for (I'm not an
owner in webkit)

[horo, 2017-04-26 16:02:37]

horo@chromium.org changed reviewers:
+ mkwst@chromium.org

[horo, 2017-04-26 16:02:37]

mkwst@
Please review webkit files and resource_messages.h

Thank you.

[Mike West, 2017-04-27 11:20:59]

//third_party/WebKit and IPC LGTM.

[horo, 2017-04-27 15:33:20]

The CQ bit was checked by horo@chromium.org

[commit-bot: I haz the power, 2017-04-27 15:33:50]

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, 2017-04-27 19:10:51]

CQ is committing da patch.

Bot data: {"patchset_id": 20001, "attempt_start_ts": 1493307200691090,
"parent_rev": "859a37371104d1cbea1afe33cdc59a31d335ce02", "commit_rev":
"e8442e6b8cab49e5e56f24d2779ee48cce71b2d5"}

[commit-bot: I haz the power, 2017-04-27 19:10:58]

Description was changed from

==========
Send the decoded size when response completed and stop summing in
ResourceLoader::DidReceiveData()

When the size properties of PerformanceResourceTiming were introduced by
https://crrev.com/2105713002, both encodedBodySize and decodedBodySize were
summed up in ResourceLoader::DidReceiveData().
The CL https://crrev.com/2510333002 changed this to send encodedBodySize when
response completed. And currently only decodedBodySize is summed up in
ResourceLoader::DidReceiveData().

We are planning to directly pass the mojo data pipe to FetchResponseData when
fetch() API is used. If we do so, ResourceLoader::DidReceiveData() will not be
called. And it will not be guaranteed that the renderer process read the all
data before the RequestComplete IPC. This will be a problem because the size
properties are passed to the ResourceTimingInfo structure when the renderer
receives the RequestComplete IPC message.

So this cl change it to send the decodedBodySize when response completed same as
encodedBodySize.

And this is also useful to support the resource timing info for service worker
navigation preload response where we don't use ResourceLoader.

BUG=712809
==========

to

==========
Send the decoded size when response completed and stop summing in
ResourceLoader::DidReceiveData()

When the size properties of PerformanceResourceTiming were introduced by
https://crrev.com/2105713002, both encodedBodySize and decodedBodySize were
summed up in ResourceLoader::DidReceiveData().
The CL https://crrev.com/2510333002 changed this to send encodedBodySize when
response completed. And currently only decodedBodySize is summed up in
ResourceLoader::DidReceiveData().

We are planning to directly pass the mojo data pipe to FetchResponseData when
fetch() API is used. If we do so, ResourceLoader::DidReceiveData() will not be
called. And it will not be guaranteed that the renderer process read the all
data before the RequestComplete IPC. This will be a problem because the size
properties are passed to the ResourceTimingInfo structure when the renderer
receives the RequestComplete IPC message.

So this cl change it to send the decodedBodySize when response completed same as
encodedBodySize.

And this is also useful to support the resource timing info for service worker
navigation preload response where we don't use ResourceLoader.

BUG=712809

Review-Url: https://codereview.chromium.org/2835123005
Cr-Commit-Position: refs/heads/master@{#467744}
Committed:
https://chromium.googlesource.com/chromium/src/+/e8442e6b8cab49e5e56f24d2779e...
==========

[commit-bot: I haz the power, 2017-04-27 19:11:01]

Committed patchset #1 (id:20001) as
https://chromium.googlesource.com/chromium/src/+/e8442e6b8cab49e5e56f24d2779e...