Move LayoutTestMessageFilter-related IPCs to their own message file.

Move LayoutTestMessageFilter-related IPCs to their own message file.

This does not yet contain all layout-test only IPC messages, but moves those
used in LayoutTestMessageFilter to a separate message file. As refactoring
continues, more IPCs will move there from the ShellMessageFilter.

BUG=420994
Committed: https://crrev.com/e0420359b53ecc20f1475bf64eb7c96d05ed9c23
Cr-Commit-Position: refs/heads/master@{#301392}

[Peter Beverloo, 2014-10-24 11:52:35]

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

[Peter Beverloo, 2014-10-24 11:52:35]

PTAL

[Mike West, 2014-10-24 12:05:42]

The bots do not like your patch.

[Mike West, 2014-10-24 12:08:44]

Otherwise LGTM, assuming the bots are happy. :)

[Peter Beverloo, 2014-10-24 12:18:31]

On 2014/10/24 12:08:44, Mike West wrote:
> Otherwise LGTM, assuming the bots are happy. :)

Ah, forgot to mention that it depends on
https://codereview.chromium.org/659043003/. Let me run try-jobs based on HEAD to
verify.

[Peter Beverloo, 2014-10-27 12:16:30]

On 2014/10/24 12:18:31, Peter Beverloo wrote:
> On 2014/10/24 12:08:44, Mike West wrote:
> > Otherwise LGTM, assuming the bots are happy. :)
> 
> Ah, forgot to mention that it depends on
> https://codereview.chromium.org/659043003/. Let me run try-jobs based on HEAD
to
> verify.

There's a linking error on win_chromium_compile_dbg here that I don't
understand. Have you seen this before, Mike?

[Mike West, 2014-10-27 12:39:55]

On 2014/10/27 12:16:30, Peter Beverloo wrote:
> On 2014/10/24 12:18:31, Peter Beverloo wrote:
> > On 2014/10/24 12:08:44, Mike West wrote:
> > > Otherwise LGTM, assuming the bots are happy. :)
> > 
> > Ah, forgot to mention that it depends on
> > https://codereview.chromium.org/659043003/. Let me run try-jobs based on
HEAD
> to
> > verify.
> 
> There's a linking error on win_chromium_compile_dbg here that I don't
> understand. Have you seen this before, Mike?

Nope. It should be linking into the same DLL as shell_messages... Let's throw it
at the bot again and see what happens. :)

[Peter Beverloo, 2014-10-27 13:51:51]

On 2014/10/27 12:39:55, Mike West wrote:
> On 2014/10/27 12:16:30, Peter Beverloo wrote:
> > On 2014/10/24 12:18:31, Peter Beverloo wrote:
> > > On 2014/10/24 12:08:44, Mike West wrote:
> > > > Otherwise LGTM, assuming the bots are happy. :)
> > > 
> > > Ah, forgot to mention that it depends on
> > > https://codereview.chromium.org/659043003/. Let me run try-jobs based on
> HEAD
> > to
> > > verify.
> > 
> > There's a linking error on win_chromium_compile_dbg here that I don't
> > understand. Have you seen this before, Mike?
> 
> Nope. It should be linking into the same DLL as shell_messages... Let's throw
it
> at the bot again and see what happens. :)

I ran the bot a couple of times over the weekend, to no avail. It doesn't seem
to be flake.

The confusing part is that it shouldn't be linked in a DLL at all --
content_shell_lib is a static library (.lib). The content_browsertests target
later on could be a shared library (for Android), but that wouldn't show up in
this phase of the build process.

A further strange thing is that the duplicate definition failures being thrown
are different for each of the messages defined in the file. For example,
LayoutTestHostMsg_ClearWebNotificationPermissions only has ::Log() as a failure.

[Peter Beverloo, 2014-10-27 16:25:52]

On 2014/10/27 13:51:51, Peter Beverloo wrote:
> On 2014/10/27 12:39:55, Mike West wrote:
> > On 2014/10/27 12:16:30, Peter Beverloo wrote:
> > > On 2014/10/24 12:18:31, Peter Beverloo wrote:
> > > > On 2014/10/24 12:08:44, Mike West wrote:
> > > > > Otherwise LGTM, assuming the bots are happy. :)
> > > > 
> > > > Ah, forgot to mention that it depends on
> > > > https://codereview.chromium.org/659043003/. Let me run try-jobs based on
> > HEAD
> > > to
> > > > verify.
> > > 
> > > There's a linking error on win_chromium_compile_dbg here that I don't
> > > understand. Have you seen this before, Mike?
> > 
> > Nope. It should be linking into the same DLL as shell_messages... Let's
throw
> it
> > at the bot again and see what happens. :)
> 
> I ran the bot a couple of times over the weekend, to no avail. It doesn't seem
> to be flake.
> 
> The confusing part is that it shouldn't be linked in a DLL at all --
> content_shell_lib is a static library (.lib). The content_browsertests target
> later on could be a shared library (for Android), but that wouldn't show up in
> this phase of the build process.
> 
> A further strange thing is that the duplicate definition failures being thrown
> are different for each of the messages defined in the file. For example,
> LayoutTestHostMsg_ClearWebNotificationPermissions only has ::Log() as a
failure.

Including ipc_message_macros.h seems to have fixed it, so putting this on the
queue. My theory is that redefining IPC_MESSAGE_EXPORT as a void-macro must've
helped.

[Peter Beverloo, 2014-10-27 16:26:00]

The CQ bit was checked by peter@chromium.org

[commit-bot: I haz the power, 2014-10-27 16:28:27]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/677673002/20001

[Mike West, 2014-10-27 16:29:26]

On 2014/10/27 16:28:27, I haz the power (commit-bot) wrote:
> CQ is trying da patch. Follow status at
>  https://chromium-cq-status.appspot.com/patch-status/677673002/20001

Funky. Good job finding the bug. :)

[commit-bot: I haz the power, 2014-10-27 17:57:43]

Committed patchset #2 (id:20001)

[commit-bot: I haz the power, 2014-10-27 17:58:52]

Patchset 2 (id:??) landed as
https://crrev.com/e0420359b53ecc20f1475bf64eb7c96d05ed9c23
Cr-Commit-Position: refs/heads/master@{#301392}