Remove implicit dependency from 'skia' on 'ui'.

Remove implicit dependency from 'skia' on 'ui'.

This is needed for the components build. Just adding an explicit dependency from skia to ui doesn't work because ui already depends on skia.

BUG=90078
TEST=No linker errors when linking libskia.dylib

Committed: http://src.chromium.org/viewvc/chrome?view=rev&revision=125009

[Robert Sesek, 2012-03-05 15:49:46]

https://chromiumcodereview.appspot.com/9578010/diff/5/skia/ext/skia_utils_mac.mm
File skia/ext/skia_utils_mac.mm (right):

https://chromiumcodereview.appspot.com/9578010/diff/5/skia/ext/skia_utils_mac...
skia/ext/skia_utils_mac.mm:114: if (!old_context &&
base::mac::IsOSLeopardOrEarlier())
Remove this condition; it's logic that would only be appropriate for the scoper
when old_context would be nil. It won't ever be so here.

[Nico, 2012-03-05 15:58:35]

Thanks for the quick review!

https://chromiumcodereview.appspot.com/9578010/diff/5/skia/ext/skia_utils_mac.mm
File skia/ext/skia_utils_mac.mm (right):

https://chromiumcodereview.appspot.com/9578010/diff/5/skia/ext/skia_utils_mac...
skia/ext/skia_utils_mac.mm:114: if (!old_context &&
base::mac::IsOSLeopardOrEarlier())
On 2012/03/05 15:49:46, rsesek wrote:
> Remove this condition; it's logic that would only be appropriate for the
scoper
> when old_context would be nil. It won't ever be so here.

Why's that? -currentContext could return nil, couldn't it?
(ScopedNSGraphicsContextSaveGState only ever sets its context_ to
-currentContext too at least.)

[Robert Sesek, 2012-03-05 17:23:58]

https://chromiumcodereview.appspot.com/9578010/diff/5/skia/ext/skia_utils_mac.mm
File skia/ext/skia_utils_mac.mm (right):

https://chromiumcodereview.appspot.com/9578010/diff/5/skia/ext/skia_utils_mac...
skia/ext/skia_utils_mac.mm:114: if (!old_context &&
base::mac::IsOSLeopardOrEarlier())
On 2012/03/05 15:58:35, Nico wrote:
> On 2012/03/05 15:49:46, rsesek wrote:
> > Remove this condition; it's logic that would only be appropriate for the
> scoper
> > when old_context would be nil. It won't ever be so here.
> 
> Why's that? -currentContext could return nil, couldn't it?
> (ScopedNSGraphicsContextSaveGState only ever sets its context_ to
> -currentContext too at least.)

Because we're in a drawing context. The scoper is designed to be versatile. You
wouldn't ever do this check if you weren't using a scoper.

[Nico, 2012-03-05 17:33:00]

https://chromiumcodereview.appspot.com/9578010/diff/5/skia/ext/skia_utils_mac.mm
File skia/ext/skia_utils_mac.mm (right):

https://chromiumcodereview.appspot.com/9578010/diff/5/skia/ext/skia_utils_mac...
skia/ext/skia_utils_mac.mm:114: if (!old_context &&
base::mac::IsOSLeopardOrEarlier())
> Because we're in a drawing context. The scoper is designed to be versatile.
You
> wouldn't ever do this check if you weren't using a scoper.

I don't understand, can you expand a bit? Why is -currentContext guaranteed to
be non-nil?

[Robert Sesek, 2012-03-05 17:40:04]

lgtm

https://chromiumcodereview.appspot.com/9578010/diff/5/skia/ext/skia_utils_mac.mm
File skia/ext/skia_utils_mac.mm (right):

https://chromiumcodereview.appspot.com/9578010/diff/5/skia/ext/skia_utils_mac...
skia/ext/skia_utils_mac.mm:114: if (!old_context &&
base::mac::IsOSLeopardOrEarlier())
On 2012/03/05 17:33:00, Nico wrote:
> > Because we're in a drawing context. The scoper is designed to be versatile.
> You
> > wouldn't ever do this check if you weren't using a scoper.
> 
> I don't understand, can you expand a bit? Why is -currentContext guaranteed to
> be non-nil?

DAMN YOU RIETEVELD. I thought this was in |-drawRect:|.

Ok, this is fine. But ugly.

[commit-bot: I haz the power, 2012-03-05 17:42:08]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/thakis@chromium.org/9578010/5

[Nico, 2012-03-05 17:43:47]

On 2012/03/05 17:40:04, rsesek wrote:
> lgtm

Thanks!

> Ok, this is fine. But ugly.

Agreed. The alternative is to move ui/gfx (or at least the bits of it that don't
depend on skia) to some other target that skia and ui can then depend on. Maybe
some other day. (I think gfx was toplevel once but then moved below ui for some
reason.)

[commit-bot: I haz the power, 2012-03-05 18:55:04]

Try job failure for 9578010-5 (retry) on win_rel for step "compile" (clobber
build).
It's a second try, previously, step "compile" failed.
http://build.chromium.org/p/tryserver.chromium/buildstatus?builder=win_rel&nu...

[commit-bot: I haz the power, 2012-03-05 20:04:29]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/thakis@chromium.org/9578010/5

[commit-bot: I haz the power, 2012-03-05 21:58:28]

Change committed as 125009