Reorganize ChannelMojo related flags.

Reorganize ChannelMojo related flags.

As ChannelMojo is turned on by default, this CL do some clenaup
around the related switches:

 * It removes the about:flags entry. This no longer has to be
   user visible
 * It moves switches to mojo_channel_switches.h/cc as it does
   compute subtle decision. I make this public/ so that we can use
   it from ipc_fuzzer.
 * It adds --disable-channel-mojo for forcibly disable ChannelMojo usage.

We still leave --enable-renderer-channel-mojo because there is still a
small change that we turn ChannelMojo off again. Once ChannelMojo is
settled, we can safely remove it.

BUG=377980
R=nasko@chromium.org, sky@chromium.org

Committed: https://crrev.com/97360186f9804cc16e1134a88fdcac8aad8d18c5
Cr-Commit-Position: refs/heads/master@{#314169}

[Hajime Morrita, 2015-01-30 23:17:28]

Sky: PTAL at chrome/ ?
Nasko: PAL at content/ ?

Thanks!

[sky, 2015-01-30 23:46:43]

LGTM

[nasko, 2015-02-02 18:14:57]

LGTM

https://codereview.chromium.org/886153003/diff/1/content/public/common/mojo_c...
File content/public/common/mojo_channel_switches.h (right):

https://codereview.chromium.org/886153003/diff/1/content/public/common/mojo_c...
content/public/common/mojo_channel_switches.h:12: extern const char
kEnableRendererMojoChannel[];
Don't you need CONTENT_EXPORT for these to be accessible outside of content?
Isn't the IPC fuzzer outside of content?

[Hajime Morrita, 2015-02-02 18:29:51]

On 2015/02/02 18:14:57, nasko wrote:
> LGTM
> 
>
https://codereview.chromium.org/886153003/diff/1/content/public/common/mojo_c...
> File content/public/common/mojo_channel_switches.h (right):
> 
>
https://codereview.chromium.org/886153003/diff/1/content/public/common/mojo_c...
> content/public/common/mojo_channel_switches.h:12: extern const char
> kEnableRendererMojoChannel[];
> Don't you need CONTENT_EXPORT for these to be accessible outside of content?
> Isn't the IPC fuzzer outside of content?
Yes, it has a separate binary.

[Hajime Morrita, 2015-02-02 18:29:56]

The CQ bit was checked by morrita@chromium.org

[commit-bot: I haz the power, 2015-02-02 18:31:04]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/886153003/1

[nasko, 2015-02-02 18:33:33]

On 2015/02/02 18:29:51, morrita wrote:
> On 2015/02/02 18:14:57, nasko wrote:
> > LGTM
> > 
> >
>
https://codereview.chromium.org/886153003/diff/1/content/public/common/mojo_c...
> > File content/public/common/mojo_channel_switches.h (right):
> > 
> >
>
https://codereview.chromium.org/886153003/diff/1/content/public/common/mojo_c...
> > content/public/common/mojo_channel_switches.h:12: extern const char
> > kEnableRendererMojoChannel[];
> > Don't you need CONTENT_EXPORT for these to be accessible outside of content?
> > Isn't the IPC fuzzer outside of content?
> Yes, it has a separate binary.

So do we need CONTENT_EXPORT on the flags or do we not?

[commit-bot: I haz the power, 2015-02-02 18:35:14]

Committed patchset #1 (id:1)

[commit-bot: I haz the power, 2015-02-02 18:35:56]

Patchset 1 (id:??) landed as
https://crrev.com/97360186f9804cc16e1134a88fdcac8aad8d18c5
Cr-Commit-Position: refs/heads/master@{#314169}

[Hajime Morrita, 2015-02-02 18:39:20]

On 2015/02/02 18:33:33, nasko wrote:
> On 2015/02/02 18:29:51, morrita wrote:
> > On 2015/02/02 18:14:57, nasko wrote:
> > > LGTM
> > > 
> > >
> >
>
https://codereview.chromium.org/886153003/diff/1/content/public/common/mojo_c...
> > > File content/public/common/mojo_channel_switches.h (right):
> > > 
> > >
> >
>
https://codereview.chromium.org/886153003/diff/1/content/public/common/mojo_c...
> > > content/public/common/mojo_channel_switches.h:12: extern const char
> > > kEnableRendererMojoChannel[];
> > > Don't you need CONTENT_EXPORT for these to be accessible outside of
content?
> > > Isn't the IPC fuzzer outside of content?
> > Yes, it has a separate binary.
> 
> So do we need CONTENT_EXPORT on the flags or do we not?
Oops. I read it wrong. Sorry about that.
It needs only ShouldUseMojoChannel() so we don't need IPC_EXPORT for the flags
themselves.

[jam, 2015-03-03 21:18:24]

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

[jam, 2015-03-03 21:18:26]

https://codereview.chromium.org/886153003/diff/1/content/public/common/mojo_c...
File content/public/common/mojo_channel_switches.h (right):

https://codereview.chromium.org/886153003/diff/1/content/public/common/mojo_c...
content/public/common/mojo_channel_switches.h:19: bool CONTENT_EXPORT
ShouldUseMojoChannel();
this shouldn't be in content/public since nothing in chrome uses it

[Hajime Morrita, 2015-03-03 21:25:20]

On 2015/03/03 21:18:26, jam wrote:
>
https://codereview.chromium.org/886153003/diff/1/content/public/common/mojo_c...
> File content/public/common/mojo_channel_switches.h (right):
> 
>
https://codereview.chromium.org/886153003/diff/1/content/public/common/mojo_c...
> content/public/common/mojo_channel_switches.h:19: bool CONTENT_EXPORT
> ShouldUseMojoChannel();
> this shouldn't be in content/public since nothing in chrome uses it
ipc_fuzzer uses it but it can be taken as a hack as it is outside chrome?

[jam, 2015-03-04 17:59:42]

On 2015/03/03 21:25:20, morrita wrote:
> On 2015/03/03 21:18:26, jam wrote:
> >
>
https://codereview.chromium.org/886153003/diff/1/content/public/common/mojo_c...
> > File content/public/common/mojo_channel_switches.h (right):
> > 
> >
>
https://codereview.chromium.org/886153003/diff/1/content/public/common/mojo_c...
> > content/public/common/mojo_channel_switches.h:19: bool CONTENT_EXPORT
> > ShouldUseMojoChannel();
> > this shouldn't be in content/public since nothing in chrome uses it
> ipc_fuzzer uses it but it can be taken as a hack as it is outside chrome?

ah i missed that, ok nvm