Revert of Add isnan() workaround for Intel Mac

Revert of Add isnan() workaround for Intel Mac (patchset #4 id:60001 of https://codereview.chromium.org/2376033004/ )

Reason for revert:
Seems to have broken NavigatingExtensionPopupBrowserTest.DownloadViaPost on Mac 10.10:
https://build.chromium.org/p/chromium.mac/builders/Mac10.10%20Tests/builds/8270

Original issue's description:
> Add isnan() workaround for Intel Mac
>
> On some Intel drivers (skylake Windows and Mac), calling
> function isnan() on highp float will get wrong answer.
> This patch work arounds this bug by using an expression
> to emulate this GLSL function. Skylake device ids are
> referenced from:
> https://cgit.freedesktop.org/vaapi/intel-driver/tree/src/i965_pciids.h
>
> BUG=650547
> CQ_INCLUDE_TRYBOTS=master.tryserver.chromium.linux:linux_optional_gpu_tests_rel;master.tryserver.chromium.mac:mac_optional_gpu_tests_rel;master.tryserver.chromium.win:win_optional_gpu_tests_rel
>
> Committed: https://crrev.com/9358866a937bc8cb9facb7feb9d087fab355e459
> Cr-Commit-Position: refs/heads/master@{#424101}

TBR=qiankun.miao@intel.com,kbr@chromium.org,zmo@chromium.org,jiawei.shao@intel.com
# Skipping CQ checks because original CL landed less than 1 days ago.
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=650547
Committed: https://crrev.com/7026791422f0046a47bb849b2f2dc3212aab8415
Cr-Commit-Position: refs/heads/master@{#424121}

[Marc Treib, 2016-10-10 08:42:49]

The CQ bit was checked by treib@chromium.org

[Marc Treib, 2016-10-10 08:42:50]

Created Revert of Add isnan() workaround for Intel Mac

[commit-bot: I haz the power, 2016-10-10 08:43:01]

CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[commit-bot: I haz the power, 2016-10-10 08:43:55]

Description was changed from

==========
Revert of Add isnan() workaround for Intel Mac (patchset #4 id:60001 of
https://codereview.chromium.org/2376033004/ )

Reason for revert:
Seems to have broken NavigatingExtensionPopupBrowserTest.DownloadViaPost on Mac
10.10:
https://build.chromium.org/p/chromium.mac/builders/Mac10.10%20Tests/builds/8270

Original issue's description:
> Add isnan() workaround for Intel Mac
> 
> On some Intel drivers (skylake Windows and Mac), calling
> function isnan() on highp float will get wrong answer.
> This patch work arounds this bug by using an expression
> to emulate this GLSL function. Skylake device ids are
> referenced from:
> https://cgit.freedesktop.org/vaapi/intel-driver/tree/src/i965_pciids.h
> 
> BUG=650547
>
CQ_INCLUDE_TRYBOTS=master.tryserver.chromium.linux:linux_optional_gpu_tests_rel;master.tryserver.chromium.mac:mac_optional_gpu_tests_rel;master.tryserver.chromium.win:win_optional_gpu_tests_rel
> 
> Committed: https://crrev.com/9358866a937bc8cb9facb7feb9d087fab355e459
> Cr-Commit-Position: refs/heads/master@{#424101}

TBR=qiankun.miao@intel.com,kbr@chromium.org,zmo@chromium.org,jiawei.shao@inte...
# Skipping CQ checks because original CL landed less than 1 days ago.
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=650547
==========

to

==========
Revert of Add isnan() workaround for Intel Mac (patchset #4 id:60001 of
https://codereview.chromium.org/2376033004/ )

Reason for revert:
Seems to have broken NavigatingExtensionPopupBrowserTest.DownloadViaPost on Mac
10.10:
https://build.chromium.org/p/chromium.mac/builders/Mac10.10%20Tests/builds/8270

Original issue's description:
> Add isnan() workaround for Intel Mac
> 
> On some Intel drivers (skylake Windows and Mac), calling
> function isnan() on highp float will get wrong answer.
> This patch work arounds this bug by using an expression
> to emulate this GLSL function. Skylake device ids are
> referenced from:
> https://cgit.freedesktop.org/vaapi/intel-driver/tree/src/i965_pciids.h
> 
> BUG=650547
>
CQ_INCLUDE_TRYBOTS=master.tryserver.chromium.linux:linux_optional_gpu_tests_rel;master.tryserver.chromium.mac:mac_optional_gpu_tests_rel;master.tryserver.chromium.win:win_optional_gpu_tests_rel
> 
> Committed: https://crrev.com/9358866a937bc8cb9facb7feb9d087fab355e459
> Cr-Commit-Position: refs/heads/master@{#424101}

TBR=qiankun.miao@intel.com,kbr@chromium.org,zmo@chromium.org,jiawei.shao@inte...
# Skipping CQ checks because original CL landed less than 1 days ago.
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=650547
==========

[commit-bot: I haz the power, 2016-10-10 08:43:56]

Committed patchset #1 (id:1)

[commit-bot: I haz the power, 2016-10-10 08:46:05]

Description was changed from

==========
Revert of Add isnan() workaround for Intel Mac (patchset #4 id:60001 of
https://codereview.chromium.org/2376033004/ )

Reason for revert:
Seems to have broken NavigatingExtensionPopupBrowserTest.DownloadViaPost on Mac
10.10:
https://build.chromium.org/p/chromium.mac/builders/Mac10.10%20Tests/builds/8270

Original issue's description:
> Add isnan() workaround for Intel Mac
> 
> On some Intel drivers (skylake Windows and Mac), calling
> function isnan() on highp float will get wrong answer.
> This patch work arounds this bug by using an expression
> to emulate this GLSL function. Skylake device ids are
> referenced from:
> https://cgit.freedesktop.org/vaapi/intel-driver/tree/src/i965_pciids.h
> 
> BUG=650547
>
CQ_INCLUDE_TRYBOTS=master.tryserver.chromium.linux:linux_optional_gpu_tests_rel;master.tryserver.chromium.mac:mac_optional_gpu_tests_rel;master.tryserver.chromium.win:win_optional_gpu_tests_rel
> 
> Committed: https://crrev.com/9358866a937bc8cb9facb7feb9d087fab355e459
> Cr-Commit-Position: refs/heads/master@{#424101}

TBR=qiankun.miao@intel.com,kbr@chromium.org,zmo@chromium.org,jiawei.shao@inte...
# Skipping CQ checks because original CL landed less than 1 days ago.
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=650547
==========

to

==========
Revert of Add isnan() workaround for Intel Mac (patchset #4 id:60001 of
https://codereview.chromium.org/2376033004/ )

Reason for revert:
Seems to have broken NavigatingExtensionPopupBrowserTest.DownloadViaPost on Mac
10.10:
https://build.chromium.org/p/chromium.mac/builders/Mac10.10%20Tests/builds/8270

Original issue's description:
> Add isnan() workaround for Intel Mac
>
> On some Intel drivers (skylake Windows and Mac), calling
> function isnan() on highp float will get wrong answer.
> This patch work arounds this bug by using an expression
> to emulate this GLSL function. Skylake device ids are
> referenced from:
> https://cgit.freedesktop.org/vaapi/intel-driver/tree/src/i965_pciids.h
>
> BUG=650547
>
CQ_INCLUDE_TRYBOTS=master.tryserver.chromium.linux:linux_optional_gpu_tests_rel;master.tryserver.chromium.mac:mac_optional_gpu_tests_rel;master.tryserver.chromium.win:win_optional_gpu_tests_rel
>
> Committed: https://crrev.com/9358866a937bc8cb9facb7feb9d087fab355e459
> Cr-Commit-Position: refs/heads/master@{#424101}

TBR=qiankun.miao@intel.com,kbr@chromium.org,zmo@chromium.org,jiawei.shao@inte...
# Skipping CQ checks because original CL landed less than 1 days ago.
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=650547

Committed: https://crrev.com/7026791422f0046a47bb849b2f2dc3212aab8415
Cr-Commit-Position: refs/heads/master@{#424121}
==========

[commit-bot: I haz the power, 2016-10-10 08:46:06]

Patchset 1 (id:??) landed as
https://crrev.com/7026791422f0046a47bb849b2f2dc3212aab8415
Cr-Commit-Position: refs/heads/master@{#424121}

[Jiawei, 2016-10-10 12:58:16]

Hi Treib,

Could you tell me how to run this test for me to fix this regression? 

Thanks!

[Marc Treib, 2016-10-10 13:05:15]

On 2016/10/10 12:58:16, Jiawei wrote:
> Hi Treib,
> 
> Could you tell me how to run this test for me to fix this regression? 
> 
> Thanks!

First: I'm not actually sure that this CL caused the regression (it's possible
there's something wrong with the bot or whatever). The corresponding try run is
still running:
https://build.chromium.org/p/chromium.mac/builders/Mac10.10%20Tests/builds/8273

To run the test: (replace "out/Release" with your output folder)
ninja -C out/Release browser_tests
out/Release/browser_tests
--gtest_filter=NavigatingExtensionPopupBrowserTest.DownloadViaPost

Note that the failure seems specific to Mac 10.10, or possibly to some
particular GPU? Anyway I couldn't reproduce it locally on my Mac 10.11 machine.

[Marc Treib, 2016-10-10 13:09:17]

On 2016/10/10 13:05:15, Marc Treib wrote:
> On 2016/10/10 12:58:16, Jiawei wrote:
> > Hi Treib,
> > 
> > Could you tell me how to run this test for me to fix this regression? 
> > 
> > Thanks!
> 
> First: I'm not actually sure that this CL caused the regression (it's possible
> there's something wrong with the bot or whatever). The corresponding try run
is
> still running:
>
https://build.chromium.org/p/chromium.mac/builders/Mac10.10%20Tests/builds/8273
> 
> To run the test: (replace "out/Release" with your output folder)
> ninja -C out/Release browser_tests
> out/Release/browser_tests
> --gtest_filter=NavigatingExtensionPopupBrowserTest.DownloadViaPost
> 
> Note that the failure seems specific to Mac 10.10, or possibly to some
> particular GPU? Anyway I couldn't reproduce it locally on my Mac 10.11
machine.

Looking at the output from browser_tests on that bot
(https://build.chromium.org/p/chromium.mac/builders/Mac10.10%20Tests/builds/82...),
the test failed again even with the revert. That means your CL is not the cause,
and I'll reland it. Sorry for the trouble!

[Marc Treib, 2016-10-10 13:10:24]

A revert of this CL (patchset #1 id:1) has been created in
https://codereview.chromium.org/2404863002/ by treib@chromium.org.

The reason for reverting is: Relanding the original CL - the test failed again,
so this isn't the cause..