fix windows assert

fix windows assert

Windows asserts trigger out of memory stack traces that
are confusing. This permits windows to break more
directly, stopping the IDE at the assert and providing
a less confusing stacktrace outside of the IDE.

R=reed@google.com,herb@google.com
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=2039443002

Committed: https://skia.googlesource.com/skia/+/ae92262f3ba5b3412ff468877277d817f1b61308

[caryclark, 2016-06-03 14:35:59]

Description was changed from

==========
fix windows assert

Windows asserts trigger out of memory stack traces that
are confusing. This permits windows to break more
directly, stopping the IDE at the assert and providing
a less confusing stacktrace outside of the IDE.

R=reed@google.com,herb@google.com
==========

to

==========
fix windows assert

Windows asserts trigger out of memory stack traces that
are confusing. This permits windows to break more
directly, stopping the IDE at the assert and providing
a less confusing stacktrace outside of the IDE.

R=reed@google.com,herb@google.com
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=2039443002
==========

[caryclark, 2016-06-03 14:36:21]

The CQ bit was checked by caryclark@google.com to run a CQ dry run

[commit-bot: I haz the power, 2016-06-03 14:36:29]

Dry run: CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/2039443002/1
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/2039443002/1

[herb_g, 2016-06-03 14:50:09]

lgtm

[commit-bot: I haz the power, 2016-06-03 14:55:59]

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

[commit-bot: I haz the power, 2016-06-03 14:56:00]

Dry run: This issue passed the CQ dry run.

[reed1, 2016-06-03 16:05:06]

lgtm

[caryclark, 2016-06-03 16:41:57]

The CQ bit was checked by caryclark@google.com

[commit-bot: I haz the power, 2016-06-03 16:42:00]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/2039443002/1

[commit-bot: I haz the power, 2016-06-03 16:42:55]

Description was changed from

==========
fix windows assert

Windows asserts trigger out of memory stack traces that
are confusing. This permits windows to break more
directly, stopping the IDE at the assert and providing
a less confusing stacktrace outside of the IDE.

R=reed@google.com,herb@google.com
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=2039443002
==========

to

==========
fix windows assert

Windows asserts trigger out of memory stack traces that
are confusing. This permits windows to break more
directly, stopping the IDE at the assert and providing
a less confusing stacktrace outside of the IDE.

R=reed@google.com,herb@google.com
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=2039443002

Committed:
https://skia.googlesource.com/skia/+/ae92262f3ba5b3412ff468877277d817f1b61308
==========

[commit-bot: I haz the power, 2016-06-03 16:42:57]

Committed patchset #1 (id:1) as
https://skia.googlesource.com/skia/+/ae92262f3ba5b3412ff468877277d817f1b61308

[brucedawson, 2016-06-03 22:53:40]

On 2016/06/03 16:42:57, commit-bot: I haz the power wrote:
> Committed patchset #1 (id:1) as
> https://skia.googlesource.com/skia/+/ae92262f3ba5b3412ff468877277d817f1b61308

FWIW, the usual reason for the madness is that the assert dialog runs a message
pump which then causes messages to be dispatched which leads to unexpected
recursion, possibly arbitrarily deep - the assert dialog breaks all sorts of
pre-conditions. Madness.

The fix I have used in the past (when a debugger is not attached) is to put the
assert dialog on its own thread so that it doesn't pump messages for the main
window. So, invoking the assert dialog requires some inter-thread communication.
Messier, but possible.

Anyway, just an FYI.