Revert of Accelerate small canvases if Ganesh is on.

Revert of Accelerate small canvases if Ganesh is on. (https://codereview.chromium.org/326283002/)

Reason for revert:
Need to revert the blink side and this (incorrectly) adds an OVERRIDE for a cross-repository override.

Original issue's description:
> Accelerate small canvases if Ganesh is on.
> 
> cc side of change.
> 
> TBR=phajdan.jr@chromium.org
> BUG=382738
> 
> Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=276753

TBR=ajuma@chromium.org,kbr@chromium.org,phajdan.jr@chromium.org,chrishtr@chromium.org
NOTREECHECKS=true
NOTRY=true
BUG=382738
Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=278842

[jamesr, 2014-06-20 22:31:08]

Created Revert of Accelerate small canvases if Ganesh is on.

[chrishtr, 2014-06-20 22:33:04]

Why do you want to revert?

[jamesr, 2014-06-20 22:33:39]

On 2014/06/20 22:33:04, chrishtr wrote:
> Why do you want to revert?

See https://codereview.chromium.org/343273002

[jamesr, 2014-06-20 22:34:19]

On 2014/06/20 22:33:39, jamesr wrote:
> On 2014/06/20 22:33:04, chrishtr wrote:
> > Why do you want to revert?
> 
> See https://codereview.chromium.org/343273002

FYI this is crashing on the bots and blocking the Blink roll.

[jamesr, 2014-06-20 22:37:32]

Committed patchset #1 manually as r278842 (presubmit successful).

[commit-bot: I haz the power, 2014-06-20 22:37:33]

CQ is trying da patch. Follow status at
 https://chromium-status.appspot.com/cq/jamesr@chromium.org/349733002/1

[jamesr, 2014-06-20 22:38:26]

The reason this needed to be reverted instead of simply reverting the Blink side
is that you used OVERRIDE on a cross-repository override, which means simply
reverting the blink side patch and not this one breaks the compile.  Don't do
that!

[chrishtr, 2014-06-20 22:39:04]

Please link to the failing bots. Also, why is it failing now and not
earlier?


On Fri, Jun 20, 2014 at 3:38 PM, <jamesr@chromium.org> wrote:

> The reason this needed to be reverted instead of simply reverting the
> Blink side
> is that you used OVERRIDE on a cross-repository override, which means
> simply
> reverting the blink side patch and not this one breaks the compile.  Don't
> do
> that!
>
> https://codereview.chromium.org/349733002/
>

To unsubscribe from this group and stop receiving emails from it, send an email
to chromium-reviews+unsubscribe@chromium.org.

[chromium-reviews, 2014-06-20 22:42:31]

http://build.chromium.org/p/chromium.webkit/builders/Win7%20%28dbg%29
http://build.chromium.org/p/tryserver.chromium/builders/linux_chromium_chrome...

crash stack is this:

Received signal 11 SEGV_MAPERR 000000000000
#0 0x0000027d1c1e base::debug::StackTrace::StackTrace()
#1 0x0000027d1757 base::debug::(anonymous
namespace)::StackDumpSignalHandler()
#2 0x7f2bddcf4cb0 <unknown>
#3 0x000006231909 blink::ChromeClientImpl::usesGpuRasterization()
#4 0x000006517fa1 WebCore::HTMLCanvasElement::shouldAccelerate()
#5 0x00000651812b WebCore::HTMLCanvasElement::createImageBufferSurface()
#6 0x0000065184fa WebCore::HTMLCanvasElement::createImageBufferInternal()
#7 0x0000065182d9 WebCore::HTMLCanvasElement::createImageBuffer()
#8 0x0000065177bb WebCore::HTMLCanvasElement::buffer()
#9 0x000006518999 WebCore::HTMLCanvasElement::drawingContext()
#10 0x0000065b27e7 WebCore::CanvasRenderingContext2D::drawingContext()
#11 0x0000065b7156 WebCore::CanvasRenderingContext2D::clearRect()
#12 0x00000772a415
WebCore::CanvasRenderingContext2DV8Internal::clearRectMethod()
#13 0x0000076f7e14
WebCore::CanvasRenderingContext2DV8Internal::clearRectMethodCallback()
#14 0x00000619b413 v8::internal::FunctionCallbackArguments::Call()
#15 0x000005d0d331 v8::internal::Builtin_HandleApiCall()
#16 0x3ff9ab6066ee <unknown>
  r8: 0000000000000001  r9: 0000000000000000 r10: 00000162a2aa3490 r11:
00007f2bdb45be2e
 r12: 000000001287d23b r13: 0000000000000000 r14: 00000000076f7df0 r15:
00000162a2aa0020
  di: 000013ba62654010  si: 000037069b248010  bp: 00007fffbededa50  bx:
0000000000000000
  dx: 0000000000000001  ax: 0000000000000000  cx: 000013ba62654301  sp:
00007fffbededa40
  ip: 0000000006231909 efl: 0000000000010202 cgf: ffff000000000033 erf:
0000000000000004


it's showing up now because your patch depends on very subtle timing which
happened to show up now.  But it's definitely a broken approach.



On Fri, Jun 20, 2014 at 3:38 PM, Chris Harrelson <chrishtr@chromium.org>
wrote:

> Please link to the failing bots. Also, why is it failing now and not
> earlier?
>
>
> On Fri, Jun 20, 2014 at 3:38 PM, <jamesr@chromium.org> wrote:
>
>> The reason this needed to be reverted instead of simply reverting the
>> Blink side
>> is that you used OVERRIDE on a cross-repository override, which means
>> simply
>> reverting the blink side patch and not this one breaks the compile.
>>  Don't do
>> that!
>>
>> https://codereview.chromium.org/349733002/
>>
>
>

To unsubscribe from this group and stop receiving emails from it, send an email
to chromium-reviews+unsubscribe@chromium.org.

[chrishtr, 2014-06-20 22:48:34]

I believe you, but I can't find that stack anywhere in the links you
provided?


On Fri, Jun 20, 2014 at 3:42 PM, James Robinson <jamesr@google.com> wrote:

> http://build.chromium.org/p/chromium.webkit/builders/Win7%20%28dbg%29
>
>
http://build.chromium.org/p/tryserver.chromium/builders/linux_chromium_chrome...
>
> crash stack is this:
>
> Received signal 11 SEGV_MAPERR 000000000000
> #0 0x0000027d1c1e base::debug::StackTrace::StackTrace()
> #1 0x0000027d1757 base::debug::(anonymous
> namespace)::StackDumpSignalHandler()
> #2 0x7f2bddcf4cb0 <unknown>
> #3 0x000006231909 blink::ChromeClientImpl::usesGpuRasterization()
> #4 0x000006517fa1 WebCore::HTMLCanvasElement::shouldAccelerate()
> #5 0x00000651812b WebCore::HTMLCanvasElement::createImageBufferSurface()
> #6 0x0000065184fa WebCore::HTMLCanvasElement::createImageBufferInternal()
> #7 0x0000065182d9 WebCore::HTMLCanvasElement::createImageBuffer()
> #8 0x0000065177bb WebCore::HTMLCanvasElement::buffer()
> #9 0x000006518999 WebCore::HTMLCanvasElement::drawingContext()
> #10 0x0000065b27e7 WebCore::CanvasRenderingContext2D::drawingContext()
> #11 0x0000065b7156 WebCore::CanvasRenderingContext2D::clearRect()
> #12 0x00000772a415
> WebCore::CanvasRenderingContext2DV8Internal::clearRectMethod()
> #13 0x0000076f7e14
> WebCore::CanvasRenderingContext2DV8Internal::clearRectMethodCallback()
> #14 0x00000619b413 v8::internal::FunctionCallbackArguments::Call()
> #15 0x000005d0d331 v8::internal::Builtin_HandleApiCall()
> #16 0x3ff9ab6066ee <unknown>
>   r8: 0000000000000001  r9: 0000000000000000 r10: 00000162a2aa3490 r11:
> 00007f2bdb45be2e
>  r12: 000000001287d23b r13: 0000000000000000 r14: 00000000076f7df0 r15:
> 00000162a2aa0020
>   di: 000013ba62654010  si: 000037069b248010  bp: 00007fffbededa50  bx:
> 0000000000000000
>   dx: 0000000000000001  ax: 0000000000000000  cx: 000013ba62654301  sp:
> 00007fffbededa40
>   ip: 0000000006231909 efl: 0000000000010202 cgf: ffff000000000033 erf:
> 0000000000000004
>
>
> it's showing up now because your patch depends on very subtle timing which
> happened to show up now.  But it's definitely a broken approach.
>
>
>
> On Fri, Jun 20, 2014 at 3:38 PM, Chris Harrelson <chrishtr@chromium.org>
> wrote:
>
>> Please link to the failing bots. Also, why is it failing now and not
>> earlier?
>>
>>
>> On Fri, Jun 20, 2014 at 3:38 PM, <jamesr@chromium.org> wrote:
>>
>>> The reason this needed to be reverted instead of simply reverting the
>>> Blink side
>>> is that you used OVERRIDE on a cross-repository override, which means
>>> simply
>>> reverting the blink side patch and not this one breaks the compile.
>>>  Don't do
>>> that!
>>>
>>> https://codereview.chromium.org/349733002/
>>>
>>
>>
>

To unsubscribe from this group and stop receiving emails from it, send an email
to chromium-reviews+unsubscribe@chromium.org.

[chromium-reviews, 2014-06-20 22:51:24]

I know - the release and windows bots don't produce stacks, so that stack
is from my box.  If the bots produced stacktraces this would have been a
lot easier to track down...


On Fri, Jun 20, 2014 at 3:48 PM, Chris Harrelson <chrishtr@chromium.org>
wrote:

> I believe you, but I can't find that stack anywhere in the links you
> provided?
>
>
> On Fri, Jun 20, 2014 at 3:42 PM, James Robinson <jamesr@google.com> wrote:
>
>> http://build.chromium.org/p/chromium.webkit/builders/Win7%20%28dbg%29
>>
>>
http://build.chromium.org/p/tryserver.chromium/builders/linux_chromium_chrome...
>>
>> crash stack is this:
>>
>> Received signal 11 SEGV_MAPERR 000000000000
>> #0 0x0000027d1c1e base::debug::StackTrace::StackTrace()
>> #1 0x0000027d1757 base::debug::(anonymous
>> namespace)::StackDumpSignalHandler()
>> #2 0x7f2bddcf4cb0 <unknown>
>> #3 0x000006231909 blink::ChromeClientImpl::usesGpuRasterization()
>> #4 0x000006517fa1 WebCore::HTMLCanvasElement::shouldAccelerate()
>> #5 0x00000651812b WebCore::HTMLCanvasElement::createImageBufferSurface()
>> #6 0x0000065184fa WebCore::HTMLCanvasElement::createImageBufferInternal()
>> #7 0x0000065182d9 WebCore::HTMLCanvasElement::createImageBuffer()
>> #8 0x0000065177bb WebCore::HTMLCanvasElement::buffer()
>> #9 0x000006518999 WebCore::HTMLCanvasElement::drawingContext()
>> #10 0x0000065b27e7 WebCore::CanvasRenderingContext2D::drawingContext()
>> #11 0x0000065b7156 WebCore::CanvasRenderingContext2D::clearRect()
>> #12 0x00000772a415
>> WebCore::CanvasRenderingContext2DV8Internal::clearRectMethod()
>> #13 0x0000076f7e14
>> WebCore::CanvasRenderingContext2DV8Internal::clearRectMethodCallback()
>> #14 0x00000619b413 v8::internal::FunctionCallbackArguments::Call()
>> #15 0x000005d0d331 v8::internal::Builtin_HandleApiCall()
>> #16 0x3ff9ab6066ee <unknown>
>>   r8: 0000000000000001  r9: 0000000000000000 r10: 00000162a2aa3490 r11:
>> 00007f2bdb45be2e
>>  r12: 000000001287d23b r13: 0000000000000000 r14: 00000000076f7df0 r15:
>> 00000162a2aa0020
>>   di: 000013ba62654010  si: 000037069b248010  bp: 00007fffbededa50  bx:
>> 0000000000000000
>>   dx: 0000000000000001  ax: 0000000000000000  cx: 000013ba62654301  sp:
>> 00007fffbededa40
>>   ip: 0000000006231909 efl: 0000000000010202 cgf: ffff000000000033 erf:
>> 0000000000000004
>>
>>
>> it's showing up now because your patch depends on very subtle timing
>> which happened to show up now.  But it's definitely a broken approach.
>>
>>
>>
>> On Fri, Jun 20, 2014 at 3:38 PM, Chris Harrelson <chrishtr@chromium.org>
>> wrote:
>>
>>> Please link to the failing bots. Also, why is it failing now and not
>>> earlier?
>>>
>>>
>>> On Fri, Jun 20, 2014 at 3:38 PM, <jamesr@chromium.org> wrote:
>>>
>>>> The reason this needed to be reverted instead of simply reverting the
>>>> Blink side
>>>> is that you used OVERRIDE on a cross-repository override, which means
>>>> simply
>>>> reverting the blink side patch and not this one breaks the compile.
>>>>  Don't do
>>>> that!
>>>>
>>>> https://codereview.chromium.org/349733002/
>>>>
>>>
>>>
>>
>

To unsubscribe from this group and stop receiving emails from it, send an email
to chromium-reviews+unsubscribe@chromium.org.

[chrishtr, 2014-06-20 22:52:44]

I see, ok, thanks.


On Fri, Jun 20, 2014 at 3:51 PM, James Robinson <jamesr@google.com> wrote:

> I know - the release and windows bots don't produce stacks, so that stack
> is from my box.  If the bots produced stacktraces this would have been a
> lot easier to track down...
>
>
> On Fri, Jun 20, 2014 at 3:48 PM, Chris Harrelson <chrishtr@chromium.org>
> wrote:
>
>> I believe you, but I can't find that stack anywhere in the links you
>> provided?
>>
>>
>> On Fri, Jun 20, 2014 at 3:42 PM, James Robinson <jamesr@google.com>
>> wrote:
>>
>>> http://build.chromium.org/p/chromium.webkit/builders/Win7%20%28dbg%29
>>>
>>>
http://build.chromium.org/p/tryserver.chromium/builders/linux_chromium_chrome...
>>>
>>> crash stack is this:
>>>
>>> Received signal 11 SEGV_MAPERR 000000000000
>>> #0 0x0000027d1c1e base::debug::StackTrace::StackTrace()
>>> #1 0x0000027d1757 base::debug::(anonymous
>>> namespace)::StackDumpSignalHandler()
>>> #2 0x7f2bddcf4cb0 <unknown>
>>> #3 0x000006231909 blink::ChromeClientImpl::usesGpuRasterization()
>>> #4 0x000006517fa1 WebCore::HTMLCanvasElement::shouldAccelerate()
>>> #5 0x00000651812b WebCore::HTMLCanvasElement::createImageBufferSurface()
>>> #6 0x0000065184fa WebCore::HTMLCanvasElement::createImageBufferInternal()
>>> #7 0x0000065182d9 WebCore::HTMLCanvasElement::createImageBuffer()
>>> #8 0x0000065177bb WebCore::HTMLCanvasElement::buffer()
>>> #9 0x000006518999 WebCore::HTMLCanvasElement::drawingContext()
>>> #10 0x0000065b27e7 WebCore::CanvasRenderingContext2D::drawingContext()
>>> #11 0x0000065b7156 WebCore::CanvasRenderingContext2D::clearRect()
>>> #12 0x00000772a415
>>> WebCore::CanvasRenderingContext2DV8Internal::clearRectMethod()
>>> #13 0x0000076f7e14
>>> WebCore::CanvasRenderingContext2DV8Internal::clearRectMethodCallback()
>>> #14 0x00000619b413 v8::internal::FunctionCallbackArguments::Call()
>>> #15 0x000005d0d331 v8::internal::Builtin_HandleApiCall()
>>> #16 0x3ff9ab6066ee <unknown>
>>>   r8: 0000000000000001  r9: 0000000000000000 r10: 00000162a2aa3490 r11:
>>> 00007f2bdb45be2e
>>>  r12: 000000001287d23b r13: 0000000000000000 r14: 00000000076f7df0 r15:
>>> 00000162a2aa0020
>>>   di: 000013ba62654010  si: 000037069b248010  bp: 00007fffbededa50  bx:
>>> 0000000000000000
>>>   dx: 0000000000000001  ax: 0000000000000000  cx: 000013ba62654301  sp:
>>> 00007fffbededa40
>>>   ip: 0000000006231909 efl: 0000000000010202 cgf: ffff000000000033 erf:
>>> 0000000000000004
>>>
>>>
>>> it's showing up now because your patch depends on very subtle timing
>>> which happened to show up now.  But it's definitely a broken approach.
>>>
>>>
>>>
>>> On Fri, Jun 20, 2014 at 3:38 PM, Chris Harrelson <chrishtr@chromium.org>
>>> wrote:
>>>
>>>> Please link to the failing bots. Also, why is it failing now and not
>>>> earlier?
>>>>
>>>>
>>>> On Fri, Jun 20, 2014 at 3:38 PM, <jamesr@chromium.org> wrote:
>>>>
>>>>> The reason this needed to be reverted instead of simply reverting the
>>>>> Blink side
>>>>> is that you used OVERRIDE on a cross-repository override, which means
>>>>> simply
>>>>> reverting the blink side patch and not this one breaks the compile.
>>>>>  Don't do
>>>>> that!
>>>>>
>>>>> https://codereview.chromium.org/349733002/
>>>>>
>>>>
>>>>
>>>
>>
>

To unsubscribe from this group and stop receiving emails from it, send an email
to chromium-reviews+unsubscribe@chromium.org.