In the .mm files use the crashpad::implicit_cast .. explicitly.

In the launch.mm file use the crashpad::implicit_cast .. explicitly.

Until the base one goes away this complains of it being ambiguous.

R=mark@chromium.org, scottmg@chromium.org
BUG=529769, 472900, crashpad:51
Committed: https://chromium.googlesource.com/crashpad/crashpad/+/595803e1be18aa789ba561aa06db73cfc3c2e7a2

[danakj, 2015-09-14 21:56:20]

Maybe this?

[danakj, 2015-09-14 22:00:16]

I think we need to make sure that this can build with the implicit_cast still
present, just until we can roll once into chromium and remove it there.

[Mark Mentovai, 2015-09-14 22:01:50]

Wow, this is really weird. I don’t see that this is necessary in the
http_transport_mac file, only in the launchd file. Let me poke at it for a few
minutes.

[danakj, 2015-09-14 22:02:20]

OK sure :) Ya looks weird to me too.

[scottmg, 2015-09-14 22:04:37]

Is that only in crashpad? We could just roll crashpad to
https://codereview.chromium.org/1347483002/ instead so there's only one
implicit_cast?

[danakj, 2015-09-14 22:05:44]

On 2015/09/14 22:04:37, scottmg wrote:
> Is that only in crashpad? We could just roll crashpad to
> https://codereview.chromium.org/1347483002/ instead so there's only one
> implicit_cast?

That's the crashpad bot. But if we make it green by rolling mini_chromium, we
can't roll crashpad into chromium until implicit_cast is gone in base. But we
can't remove from base until we roll crashpad not depending on it. So.. :)

[scottmg, 2015-09-14 22:06:53]

On 2015/09/14 22:05:44, danakj wrote:
> On 2015/09/14 22:04:37, scottmg wrote:
> > Is that only in crashpad? We could just roll crashpad to
> > https://codereview.chromium.org/1347483002/ instead so there's only one
> > implicit_cast?
> 
> That's the crashpad bot. But if we make it green by rolling mini_chromium, we
> can't roll crashpad into chromium until implicit_cast is gone in base. But we
> can't remove from base until we roll crashpad not depending on it. So.. :)

The chrome base removal CL could have the crashpad deps roll I guess? But
explicit qualification seems ok to me too.

[danakj, 2015-09-14 22:08:35]

On 2015/09/14 22:06:53, scottmg wrote:
> On 2015/09/14 22:05:44, danakj wrote:
> > On 2015/09/14 22:04:37, scottmg wrote:
> > > Is that only in crashpad? We could just roll crashpad to
> > > https://codereview.chromium.org/1347483002/ instead so there's only one
> > > implicit_cast?
> > 
> > That's the crashpad bot. But if we make it green by rolling mini_chromium,
we
> > can't roll crashpad into chromium until implicit_cast is gone in base. But
we
> > can't remove from base until we roll crashpad not depending on it. So.. :)
> 
> The chrome base removal CL could have the crashpad deps roll I guess? But
> explicit qualification seems ok to me too.

Oh.. we could do both at once. I did not consider it! That's fine then, either
way.

[Mark Mentovai, 2015-09-14 22:10:01]

Must be something about the pointer types, because I’m able to compile this line
in the same file without any errors:

        int five = implicit_cast<int>(5l);

I think we can take this as a temporary workaround, but we only need it in
launchd.mm, and we still shouldn’t run past 80 columns. :)

https://codereview.chromium.org/1336413003/diff/1/util/mac/launchd.mm
File util/mac/launchd.mm (right):

https://codereview.chromium.org/1336413003/diff/1/util/mac/launchd.mm#newcode19
util/mac/launchd.mm:19: #include "base/basictypes.h"
This #include isn’t needed anymore.

[danakj, 2015-09-14 22:11:32]

PTAL

[Mark Mentovai, 2015-09-14 22:12:10]

LGTM, we’ll circle back and un-crashpad:: this once everything’s done on the
Chrome side.

[danakj, 2015-09-14 22:13:27]

Thanks. Mind landing this for me?

[danakj, 2015-09-14 22:14:38]

On 2015/09/14 22:13:27, danakj wrote:
> Thanks. Mind landing this for me?

Or, the roll landed. we can abandon ship here.

[scottmg, 2015-09-14 22:15:48]

On 2015/09/14 22:14:38, danakj wrote:
> On 2015/09/14 22:13:27, danakj wrote:
> > Thanks. Mind landing this for me?
> 
> Or, the roll landed. we can abandon ship here.

I'll land this anyway so that you can do it either way in Chrome. Maybe separate
is better.

[Mark Mentovai, 2015-09-14 22:16:28]

Let one of us know if you’d still like to take this. (Or if you’re interested in
doing more Crashpad work and want to be a committer!)

[Mark Mentovai, 2015-09-14 22:16:52]

On 2015/09/14 22:15:48, scottmg wrote:
> On 2015/09/14 22:14:38, danakj wrote:
> > On 2015/09/14 22:13:27, danakj wrote:
> > > Thanks. Mind landing this for me?
> > 
> > Or, the roll landed. we can abandon ship here.
> 
> I'll land this anyway so that you can do it either way in Chrome. Maybe
separate
> is better.

Still LGTM

[scottmg, 2015-09-14 22:17:57]

Committed patchset #3 (id:40001) manually as
595803e1be18aa789ba561aa06db73cfc3c2e7a2 (presubmit successful).

[scottmg, 2015-09-14 22:18:31]

On 2015/09/14 22:16:28, Mark Mentovai - August is over wrote:
> Let one of us know if you’d still like to take this. (Or if you’re interested
in
> doing more Crashpad work and want to be a committer!)

I tried adding Dana to the "Project committers" list in the
crashpad.googlecode.com UI, but I guess that's not enough now. Do you know where
the right place is now?

[Mark Mentovai, 2015-09-14 22:22:34]

The ACL is in Gerrit. I don’t want to post the link but if you search your
e-mail for things I sent you that say “gerrit”, you should find it.

[scottmg, 2015-09-14 22:30:09]

On 2015/09/14 22:22:34, Mark Mentovai - August is over wrote:
> The ACL is in Gerrit. I don’t want to post the link but if you search your
> e-mail for things I sent you that say “gerrit”, you should find it.

Aha! Forgot about that, thanks. Dana, you're added now (@chromium.org) if you
need to land anything in the future.

[danakj, 2015-09-14 22:31:08]

On 2015/09/14 22:30:09, scottmg wrote:
> On 2015/09/14 22:22:34, Mark Mentovai - August is over wrote:
> > The ACL is in Gerrit. I don’t want to post the link but if you search your
> > e-mail for things I sent you that say “gerrit”, you should find it.
> 
> Aha! Forgot about that, thanks. Dana, you're added now (@chromium.org) if you
> need to land anything in the future.

Cool, thank you :)