Chromium Code Reviews
|
|
Chromium Code Reviews
Issue
1324213007:
vea_unittest: Fix driver doesn't suppport hole in memory mapped file (Closed)
|
Created:
5 years, 3 months ago by kcwu Modified:
5 years ago CC:
chromium-reviews, posciak+watch_chromium.org, jam, mcasas+watch_chromium.org, feature-media-reviews_chromium.org, darin-cc_chromium.org, piman+watch_chromium.org, wjia+watch_chromium.org Base URL:
https://chromium.googlesource.com/chromium/src.git@master Target Ref:
refs/pending/heads/master Project:
chromium Visibility:
Public. |
Descriptionvea_unittest: Fix driver doesn't support hole in memory mapped file
A hole in a file is an area represented by zeros, which are not stored
physically on disk (aka sparse files). Take 640x360 I420 input as
an example. The padding size for three planes are 5120, 1280, and 1280.
Note 5120 > 4k. That means for every frame, there is one page "hole" in
the file. The hole page causes troubles for some systems.
This CL fills holes with zero bytes.
BUG=chrome-os-partner:41833
TEST=video_encode_accelerator_unittest
Committed: https://crrev.com/b2e15d28fb69d5f9d94b9edbea56e1a7a399fd0d
Cr-Commit-Position: refs/heads/master@{#361089}
Patch Set 1 #
Total comments: 6
Patch Set 2 : revised #
Total comments: 2
Patch Set 3 : #Messages
Total messages: 25 (17 generated)
kcwu@chromium.org changed reviewers: + posciak@chromium.org, wuchengli@chromium.org
PTAL
Description was changed from ========== Workaround for VEA doesn't suppport hole in memory mapped file BUG=chrome-os-partner:41833 ========== to ========== vea_unittest: Fix driver doesn't suppport hole in memory mapped file BUG=chrome-os-partner:41833 ==========
Description was changed from ========== vea_unittest: Fix driver doesn't suppport hole in memory mapped file BUG=chrome-os-partner:41833 ========== to ========== vea_unittest: Fix driver doesn't support hole in memory mapped file BUG=chrome-os-partner:41833 ==========
https://codereview.chromium.org/1324213007/diff/1/content/common/gpu/media/vi... File content/common/gpu/media/video_encode_accelerator_unittest.cc (right): https://codereview.chromium.org/1324213007/diff/1/content/common/gpu/media/vi... content/common/gpu/media/video_encode_accelerator_unittest.cc:1: // Copyright 2013 The Chromium Authors. All rights reserved. There was only subject in the change description. Explain more about the issue in the change description. https://codereview.chromium.org/1324213007/diff/1/content/common/gpu/media/vi... content/common/gpu/media/video_encode_accelerator_unittest.cc:286: if (padding_sizes[i]) { We can use padding[i].size() and remove |padding_sizes| variable. https://codereview.chromium.org/1324213007/diff/1/content/common/gpu/media/vi... content/common/gpu/media/video_encode_accelerator_unittest.cc:287: CHECK(WriteFile(&dest_file, dest_offset, &padding[i][0], s/CHECK/LOG_ASSERT/
Description was changed from ========== vea_unittest: Fix driver doesn't support hole in memory mapped file BUG=chrome-os-partner:41833 ========== to ========== vea_unittest: Fix driver doesn't support hole in memory mapped file For example 640x360 I420 input, the padding size for three planes are 5120, 1280, 1280. Note 5120 > 4k, that means for every frame, there is one page "hole" in the file. The hole page cause troubles for some systems. This CL fill holes with zero bytes. BUG=chrome-os-partner:41833 TEST=video_encode_accelerator_unittest ==========
https://codereview.chromium.org/1324213007/diff/1/content/common/gpu/media/vi... File content/common/gpu/media/video_encode_accelerator_unittest.cc (right): https://codereview.chromium.org/1324213007/diff/1/content/common/gpu/media/vi... content/common/gpu/media/video_encode_accelerator_unittest.cc:1: // Copyright 2013 The Chromium Authors. All rights reserved. On 2015/11/19 03:33:40, wuchengli wrote: > There was only subject in the change description. Explain more about the issue > in the change description. Done. https://codereview.chromium.org/1324213007/diff/1/content/common/gpu/media/vi... content/common/gpu/media/video_encode_accelerator_unittest.cc:286: if (padding_sizes[i]) { On 2015/11/19 03:33:40, wuchengli wrote: > We can use padding[i].size() and remove |padding_sizes| variable. Done. https://codereview.chromium.org/1324213007/diff/1/content/common/gpu/media/vi... content/common/gpu/media/video_encode_accelerator_unittest.cc:287: CHECK(WriteFile(&dest_file, dest_offset, &padding[i][0], On 2015/11/19 03:33:40, wuchengli wrote: > s/CHECK/LOG_ASSERT/ Done.
Description was changed from ========== vea_unittest: Fix driver doesn't support hole in memory mapped file For example 640x360 I420 input, the padding size for three planes are 5120, 1280, 1280. Note 5120 > 4k, that means for every frame, there is one page "hole" in the file. The hole page cause troubles for some systems. This CL fill holes with zero bytes. BUG=chrome-os-partner:41833 TEST=video_encode_accelerator_unittest ========== to ========== vea_unittest: Fix driver doesn't support hole in memory mapped file Take 640x360 I420 input as an example. The padding size for three planes are 5120, 1280, and 1280. Note 5120 > 4k. That means for every frame, there is one page "hole" in the file. The hole page causes troubles for some systems. This CL fill holes with zero bytes. BUG=chrome-os-partner:41833 TEST=video_encode_accelerator_unittest ==========
Description was changed from ========== vea_unittest: Fix driver doesn't support hole in memory mapped file Take 640x360 I420 input as an example. The padding size for three planes are 5120, 1280, and 1280. Note 5120 > 4k. That means for every frame, there is one page "hole" in the file. The hole page causes troubles for some systems. This CL fill holes with zero bytes. BUG=chrome-os-partner:41833 TEST=video_encode_accelerator_unittest ========== to ========== vea_unittest: Fix driver doesn't support hole in memory mapped file Take 640x360 I420 input as an example. The padding size for three planes are 5120, 1280, and 1280. Note 5120 > 4k. That means for every frame, there is one page "hole" in the file. The hole page causes troubles for some systems. This CL fill holes with zero bytes. BUG=chrome-os-partner:41833 TEST=video_encode_accelerator_unittest ==========
Description was changed from ========== vea_unittest: Fix driver doesn't support hole in memory mapped file Take 640x360 I420 input as an example. The padding size for three planes are 5120, 1280, and 1280. Note 5120 > 4k. That means for every frame, there is one page "hole" in the file. The hole page causes troubles for some systems. This CL fill holes with zero bytes. BUG=chrome-os-partner:41833 TEST=video_encode_accelerator_unittest ========== to ========== vea_unittest: Fix driver doesn't support hole in memory mapped file Take 640x360 I420 input as an example. The padding size for three planes are 5120, 1280, and 1280. Note 5120 > 4k. That means for every frame, there is one page "hole" in the file. The hole page causes troubles for some systems. This CL fill holes with zero bytes. BUG=chrome-os-partner:41833 TEST=video_encode_accelerator_unittest ==========
Description was changed from ========== vea_unittest: Fix driver doesn't support hole in memory mapped file Take 640x360 I420 input as an example. The padding size for three planes are 5120, 1280, and 1280. Note 5120 > 4k. That means for every frame, there is one page "hole" in the file. The hole page causes troubles for some systems. This CL fill holes with zero bytes. BUG=chrome-os-partner:41833 TEST=video_encode_accelerator_unittest ========== to ========== vea_unittest: Fix driver doesn't support hole in memory mapped file Take 640x360 I420 input as an example. The padding size for three planes are 5120, 1280, and 1280. Note 5120 > 4k. That means for every frame, there is one page "hole" in the file. The hole page causes troubles for some systems. This CL fill holes with zero bytes. BUG=chrome-os-partner:41833 TEST=video_encode_accelerator_unittest ==========
Description was changed from ========== vea_unittest: Fix driver doesn't support hole in memory mapped file Take 640x360 I420 input as an example. The padding size for three planes are 5120, 1280, and 1280. Note 5120 > 4k. That means for every frame, there is one page "hole" in the file. The hole page causes troubles for some systems. This CL fill holes with zero bytes. BUG=chrome-os-partner:41833 TEST=video_encode_accelerator_unittest ========== to ========== vea_unittest: Fix driver doesn't support hole in memory mapped file Take 640x360 I420 input as an example. The padding size for three planes are 5120, 1280, and 1280. Note 5120 > 4k. That means for every frame, there is one page "hole" in the file. The hole page causes troubles for some systems. This CL fill holes with zero bytes. BUG=chrome-os-partner:41833 TEST=video_encode_accelerator_unittest ==========
Description was changed from ========== vea_unittest: Fix driver doesn't support hole in memory mapped file Take 640x360 I420 input as an example. The padding size for three planes are 5120, 1280, and 1280. Note 5120 > 4k. That means for every frame, there is one page "hole" in the file. The hole page causes troubles for some systems. This CL fill holes with zero bytes. BUG=chrome-os-partner:41833 TEST=video_encode_accelerator_unittest ========== to ========== vea_unittest: Fix driver doesn't support hole in memory mapped file A hole in a file is an area represented by zeros, which are not stored physically on disk (aka sparse files). Take 640x360 I420 input as an example. The padding size for three planes are 5120, 1280, and 1280. Note 5120 > 4k. That means for every frame, there is one page "hole" in the file. The hole page causes troubles for some systems. This CL fill holes with zero bytes. BUG=chrome-os-partner:41833 TEST=video_encode_accelerator_unittest ==========
Description was changed from ========== vea_unittest: Fix driver doesn't support hole in memory mapped file A hole in a file is an area represented by zeros, which are not stored physically on disk (aka sparse files). Take 640x360 I420 input as an example. The padding size for three planes are 5120, 1280, and 1280. Note 5120 > 4k. That means for every frame, there is one page "hole" in the file. The hole page causes troubles for some systems. This CL fill holes with zero bytes. BUG=chrome-os-partner:41833 TEST=video_encode_accelerator_unittest ========== to ========== vea_unittest: Fix driver doesn't support hole in memory mapped file A hole in a file is an area represented by zeros, which are not stored physically on disk (aka sparse files). Take 640x360 I420 input as an example. The padding size for three planes are 5120, 1280, and 1280. Note 5120 > 4k. That means for every frame, there is one page "hole" in the file. The hole page causes troubles for some systems. This CL fill holes with zero bytes. BUG=chrome-os-partner:41833 TEST=video_encode_accelerator_unittest ==========
Description was changed from ========== vea_unittest: Fix driver doesn't support hole in memory mapped file A hole in a file is an area represented by zeros, which are not stored physically on disk (aka sparse files). Take 640x360 I420 input as an example. The padding size for three planes are 5120, 1280, and 1280. Note 5120 > 4k. That means for every frame, there is one page "hole" in the file. The hole page causes troubles for some systems. This CL fill holes with zero bytes. BUG=chrome-os-partner:41833 TEST=video_encode_accelerator_unittest ========== to ========== vea_unittest: Fix driver doesn't support hole in memory mapped file A hole in a file is an area represented by zeros, which are not stored physically on disk (aka sparse files). Take 640x360 I420 input as an example. The padding size for three planes are 5120, 1280, and 1280. Note 5120 > 4k. That means for every frame, there is one page "hole" in the file. The hole page causes troubles for some systems. This CL fill holes with zero bytes. BUG=chrome-os-partner:41833 TEST=video_encode_accelerator_unittest ==========
Description was changed from ========== vea_unittest: Fix driver doesn't support hole in memory mapped file A hole in a file is an area represented by zeros, which are not stored physically on disk (aka sparse files). Take 640x360 I420 input as an example. The padding size for three planes are 5120, 1280, and 1280. Note 5120 > 4k. That means for every frame, there is one page "hole" in the file. The hole page causes troubles for some systems. This CL fill holes with zero bytes. BUG=chrome-os-partner:41833 TEST=video_encode_accelerator_unittest ========== to ========== vea_unittest: Fix driver doesn't support hole in memory mapped file A hole in a file is an area represented by zeros, which are not stored physically on disk (aka sparse files). Take 640x360 I420 input as an example. The padding size for three planes are 5120, 1280, and 1280. Note 5120 > 4k. That means for every frame, there is one page "hole" in the file. The hole page causes troubles for some systems. This CL fill holes with zero bytes. BUG=chrome-os-partner:41833 TEST=video_encode_accelerator_unittest ==========
Description was changed from ========== vea_unittest: Fix driver doesn't support hole in memory mapped file A hole in a file is an area represented by zeros, which are not stored physically on disk (aka sparse files). Take 640x360 I420 input as an example. The padding size for three planes are 5120, 1280, and 1280. Note 5120 > 4k. That means for every frame, there is one page "hole" in the file. The hole page causes troubles for some systems. This CL fill holes with zero bytes. BUG=chrome-os-partner:41833 TEST=video_encode_accelerator_unittest ========== to ========== vea_unittest: Fix driver doesn't support hole in memory mapped file A hole in a file is an area represented by zeros, which are not stored physically on disk (aka sparse files). Take 640x360 I420 input as an example. The padding size for three planes are 5120, 1280, and 1280. Note 5120 > 4k. That means for every frame, there is one page "hole" in the file. The hole page causes troubles for some systems. This CL fill holes with zero bytes. BUG=chrome-os-partner:41833 TEST=video_encode_accelerator_unittest ==========
Description was changed from ========== vea_unittest: Fix driver doesn't support hole in memory mapped file A hole in a file is an area represented by zeros, which are not stored physically on disk (aka sparse files). Take 640x360 I420 input as an example. The padding size for three planes are 5120, 1280, and 1280. Note 5120 > 4k. That means for every frame, there is one page "hole" in the file. The hole page causes troubles for some systems. This CL fill holes with zero bytes. BUG=chrome-os-partner:41833 TEST=video_encode_accelerator_unittest ========== to ========== vea_unittest: Fix driver doesn't support hole in memory mapped file A hole in a file is an area represented by zeros, which are not stored physically on disk (aka sparse files). Take 640x360 I420 input as an example. The padding size for three planes are 5120, 1280, and 1280. Note 5120 > 4k. That means for every frame, there is one page "hole" in the file. The hole page causes troubles for some systems. This CL fills holes with zero bytes. BUG=chrome-os-partner:41833 TEST=video_encode_accelerator_unittest ==========
LGTM https://codereview.chromium.org/1324213007/diff/20001/content/common/gpu/medi... File content/common/gpu/media/video_encode_accelerator_unittest.cc (right): https://codereview.chromium.org/1324213007/diff/20001/content/common/gpu/medi... content/common/gpu/media/video_encode_accelerator_unittest.cc:306: dest_offset += padding[i].size(); nit: this can be moved inside if (!padding[i].empty())
https://codereview.chromium.org/1324213007/diff/20001/content/common/gpu/medi... File content/common/gpu/media/video_encode_accelerator_unittest.cc (right): https://codereview.chromium.org/1324213007/diff/20001/content/common/gpu/medi... content/common/gpu/media/video_encode_accelerator_unittest.cc:306: dest_offset += padding[i].size(); On 2015/11/23 09:14:26, wuchengli wrote: > nit: this can be moved inside if (!padding[i].empty()) Done.
The CQ bit was checked by kcwu@chromium.org
The patchset sent to the CQ was uploaded after l-g-t-m from wuchengli@chromium.org Link to the patchset: https://codereview.chromium.org/1324213007/#ps40001 (title: " ")
CQ is trying da patch. Follow status at https://chromium-cq-status.appspot.com/patch-status/1324213007/40001 View timeline at https://chromium-cq-status.appspot.com/patch-timeline/1324213007/40001
Message was sent while issue was closed.
Committed patchset #3 (id:40001)
Message was sent while issue was closed.
Patchset 3 (id:??) landed as https://crrev.com/b2e15d28fb69d5f9d94b9edbea56e1a7a399fd0d Cr-Commit-Position: refs/heads/master@{#361089} |
