My favorites | Sign in
Project Home Downloads Wiki Issues Code Search
New issue   Search
for
  Advanced search   Search tips   Subscriptions
Issue 133828: Enable HW video decode by default on mac
59 people starred this issue and may be notified of changes. Back to list
 
Project Member Reported by fischman@chromium.org, Jun 20, 2012
Tracking bug for --enable-accelerated-video-decode by default on mac.

Jun 20, 2012
#6 albe...@chromium.org
(No comment was entered for this change.)
Cc: jlut...@chromium.org albe...@chromium.org
Jun 20, 2012
#7 jeffr...@google.com
PMs are responsible for shepherding Type-Launch bugs through the labels (as appropriate). They are familiar with the details of the process.
Cc: wiltz...@chromium.org
Labels: Launch-DesignDocReview-No Launch-AccessibilityReview-No Launch-QAReview-No Launch-UIReview-No Launch-StringsReviewed-No Launch-SecurityReview-No Launch-PrivacyReview-No Launch-LegalReview-No Launch-MarketingReview-No Launch-ConopsReview-No Launch-SREReview-No
Jun 20, 2012
#8 jeffr...@google.com
OK, I am told that jluther@ is the PM owner for this.

The best practice is to  set the label to "NA" if the label doesn't apply. E.g., "Launch-SecurityReview-NA".

thanks,
Jeff
Jun 22, 2012
#9 jlut...@chromium.org
(No comment was entered for this change.)
Owner: jlut...@chromium.org
Labels: -Launch-DesignDocReview-No -Launch-AccessibilityReview-No -Launch-QAReview-No -Launch-UIReview-No -Launch-StringsReviewed-No -Launch-SecurityReview-No -Launch-PrivacyReview-No -Launch-LegalReview-No -Launch-MarketingReview-No -Launch-ConopsReview-No -Launch-SREReview-No Launch-DesignDocReview-NA Launch-AccessibilityReview-NA Launch-QAReview-NA Launch-UIReview-NA Launch-StringsReviewed-NA Launch-SecurityReview-NA Launch-PrivacyReview-NA Launch-LegalReview-NA Launch-MarketingReview-NA Launch-ConopsReview-NA Launch-SREReview-NA
Jun 22, 2012
#10 jlut...@chromium.org
(No comment was entered for this change.)
Status: Assigned
Cc: -jlut...@chromium.org
Jun 22, 2012
#11 jlut...@chromium.org
(No comment was entered for this change.)
Owner: fischman@chromium.org
Jun 22, 2012
#12 jlut...@chromium.org
(No comment was entered for this change.)
Cc: jlut...@chromium.org
Jul 20, 2012
#13 jlut...@chromium.org
fischman@ is implementing the feature, as PM I assigned myself ownership.
Owner: jlut...@chromium.org
Cc: -jlut...@chromium.org fischman@chromium.org
Jul 25, 2012
#14 jeffr...@google.com
heya, is this enabled-by-default on the Dev channel?
Jul 25, 2012
#15 fischman@chromium.org
Not yet; currently blocked on http://codereview.chromium.org/10407070/
Jul 25, 2012
#16 bugdro...@chromium.org
The following revision refers to this bug:
    http://src.chromium.org/viewvc/chrome?view=rev&revision=148452

------------------------------------------------------------------------
r148452 | fischman@chromium.org | 2012-07-26T00:49:40.558649Z

Changed paths:
   M http://src.chromium.org/viewvc/chrome/trunk/src/content/public/common/content_switches.cc?r1=148452&r2=148451&pathrev=148452
   M http://src.chromium.org/viewvc/chrome/trunk/src/content/public/common/content_switches.h?r1=148452&r2=148451&pathrev=148452
   M http://src.chromium.org/viewvc/chrome/trunk/src/chrome/browser/about_flags.cc?r1=148452&r2=148451&pathrev=148452
   M http://src.chromium.org/viewvc/chrome/trunk/src/content/renderer/render_view_impl.cc?r1=148452&r2=148451&pathrev=148452
   M http://src.chromium.org/viewvc/chrome/trunk/src/content/browser/renderer_host/render_process_host_impl.cc?r1=148452&r2=148451&pathrev=148452
   M http://src.chromium.org/viewvc/chrome/trunk/src/chrome/app/generated_resources.grd?r1=148452&r2=148451&pathrev=148452
   M http://src.chromium.org/viewvc/chrome/trunk/src/chrome/browser/chromeos/login/login_utils.cc?r1=148452&r2=148451&pathrev=148452

Flip HW video decode from being behind an "enable" flag to being behind a "disable" flag
with per-platform default settings (currently off for all but cros).

This is in preparation for turning on HW decode by default one platform at a time.

BUG=133827,133828


Review URL: https://chromiumcodereview.appspot.com/10805070
------------------------------------------------------------------------
Jul 25, 2012
#17 pavanv@chromium.org
(No comment was entered for this change.)
Cc: vivi...@chromium.org
Jul 25, 2012
#18 jeffr...@google.com
K, sounds like M23 is the right target then. (M22 branches in less than 2 weeks)
Labels: -Mstone-22 Mstone-23
Jul 27, 2012
#19 fischman@chromium.org
It appears most h264 streams are unsupported by the OS/X API even on Mountain Lion.  sail@ is going to ask Apple about planned format support, but for now this launch seems unlikely (no point pushing complexity to users if it'll almost never be used).
Cc: sail@chromium.org
Aug 7, 2012
#20 jeffr...@google.com
Just making sure that this is disabled on the M22 branch, since you are targeting M23.
Aug 7, 2012
#21 fischman@chromium.org
Yep, this was never enabled by default on trunk or any branch.
Aug 22, 2012
#22 jlut...@chromium.org
(No comment was entered for this change.)
Labels: -Mstone-23 Mstone-24
Aug 29, 2012
#23 joshwood...@chromium.org
Bulk update:

Sumit and I have continued investigating how type=launch bugs might be visible to non-Googlers. As a part of this investigation, we found a series of bugs (like this one and 35 others) that did not have the Restrict-View-Google label on them. 

It appears these bugs started off as non-type=launch bugs, were converted, but the Restrict-View-Google label didn't stick or was not added. It's unclear to us how this happened.

I'm adding the Restrict-View-Google label now (without checking 'Send email') as a precaution.
Labels: Restrict-View-Google
Oct 11, 2012
#24 jlut...@chromium.org
(No comment was entered for this change.)
Labels: -Mstone-24 Mstone-25
Nov 8, 2012
#25 jlut...@chromium.org
(No comment was entered for this change.)
Labels: -Mstone-25 Mstone-26
Nov 12, 2012
#26 scherkus@chromium.org
Aside from the bugs listed as blockers, there are the following two major issues:

1) Hardware decoder won't fail until data starts streaming in
One fix would be to change the decoder initialize function to take a SPS and PPS header (see http://codereview.chromium.org/10408003 for ideas)

2) Most videos on the web seem to fail HW decode using the MacVDA despite apparently working in QT/apple's players
This can be observed by running the same clip in QT versus the VDA sample application or something like VLC and comparing CPU usage.
Nov 17, 2012
#27 bugdro...@chromium.org
The following revision refers to this bug:
    http://src.chromium.org/viewvc/chrome?view=rev&revision=168441

------------------------------------------------------------------------
r168441 | fischman@chromium.org | 2012-11-17T15:48:57.232578Z

Changed paths:
   M http://src.chromium.org/viewvc/chrome/trunk/src/chrome/test/functional/PYAUTO_TESTS?r1=168441&r2=168440&pathrev=168441
   M http://src.chromium.org/viewvc/chrome/trunk/src/content/common/gpu/media/mac_video_decode_accelerator.mm?r1=168441&r2=168440&pathrev=168441

Forcibly disable MacVDA, even for users who ignore the GPU blacklist.

BUG=125424,133828,157394,160090

Review URL: https://chromiumcodereview.appspot.com/11415036
------------------------------------------------------------------------
Dec 7, 2012
#28 meh@chromium.org
(No comment was entered for this change.)
Cc: owp-laun...@chromium.org
Jan 16, 2013
#29 scherkus@chromium.org
(No comment was entered for this change.)
Cc: cduviv...@chromium.org
Jan 16, 2013
#30 jlut...@chromium.org
Welcome, Christian! Is this realistic for M26 or should we move to 27?
Jan 16, 2013
#31 cduviv...@chromium.org
Don't know yet, will let you know as soon as I do.

What are the deadlines for M26 and M27?

--cd
Jan 16, 2013
#32 jlut...@chromium.org
For M26, branch is 11 February but features must be enabled-by-default on trunk by January 25. For M27 will probably be ~March 8.
Jan 17, 2013
#33 scherkus@chromium.org
(No comment was entered for this change.)
Labels: -Feature-Media-Video Feature-Media hardware
Jan 23, 2013
#34 scherkus@chromium.org
(No comment was entered for this change.)
Labels: -hardware Feature-Media-Hardware
Jan 24, 2013
#35 jlut...@chromium.org
(No comment was entered for this change.)
Labels: -Mstone-26 Mstone-27
Mar 4, 2013
#36 scherkus@chromium.org
Punting for now. cduvivier@ investigated and determined the current public API isn't well suited to supporting the variety of web H264 content.

Waiting to hear back from contacts whether there's a more appropriate API to use.
Labels: -Mstone-27 Mstone-28
Mar 8, 2013
#37 bugdro...@chromium.org
(No comment was entered for this change.)
Labels: -Launch-DesignDocReview-NA -Launch-AccessibilityReview-NA -Launch-QAReview-NA -Launch-UIReview-NA -Launch-StringsReviewed-NA -Launch-SecurityReview-NA -Launch-PrivacyReview-NA -Launch-LegalReview-NA -Launch-MarketingReview-NA -Launch-ConopsReview-NA -Launch-SREReview-NA Launch-QA-NA Launch-Accessibility-NA Launch-Security-NA Launch-SRE-NA Launch-Conops-NA Launch-UI-NA Launch-DesignDoc-NA Launch-Privacy-NA Launch-Marketing-NA Launch-Legal-NA Launch-Strings-NA
Mar 10, 2013
#38 bugdro...@chromium.org
(No comment was entered for this change.)
Labels: -Feature-Media -Area-WebKit -Mstone-28 -Feature-Media-Hardware Cr-Content Cr-Internals-Media Cr-Internals-Media-Hardware M-28
Apr 5, 2013
#39 bugdro...@chromium.org
(No comment was entered for this change.)
Labels: -Cr-Content Cr-Blink
Apr 17, 2013
#40 jeffr...@google.com
Hi all,

There are only a few days left to enable your features by default if you're targeting M28.

At the M28 go/no-go meeting during the week of April 29 (with Linus and the QA + TPM teams), we will be enforcing the criteria that features must be enabled-by-default in the April 22 Dev channel push (at the latest).

Please adjust the milestone label as appropriate. I'll be pinging the Chrome PMs next week to ask for the info needed at the go/no-go meeting.

thanks,
Jeff
Apr 23, 2013
#41 jeffr...@google.com
I see that you're targeting M28 for this feature, which branches in less than 2 weeks.

Has this been enabled-by-default on trunk yet?
Apr 23, 2013
#42 vivi...@chromium.org
(No comment was entered for this change.)
Cc: yiho...@chromium.org dbbro...@chromium.org
Apr 23, 2013
#43 jlut...@chromium.org
(No comment was entered for this change.)
Labels: -M-28 M-29
Apr 24, 2013
#44 scherkus@chromium.org
jluther: should we remove the M-## label until someone is actually working on this?
Apr 24, 2013
#45 jlut...@chromium.org
(No comment was entered for this change.)
Labels: -M-29
Apr 25, 2013
#46 fischman@chromium.org
Nobody is working on this or scheduled to work on this.  I propose to delete the MacVDA code since it confuses developers (I field a question about this every 2 weeks or so) and close this bug until we decide to re-prioritize this work.
Apr 25, 2013
#47 scherkus@chromium.org
SGTM

it's always possible to resurrect the code from the revision history
Apr 25, 2013
#48 bugdro...@chromium.org
------------------------------------------------------------------------
r196621 | fischman@chromium.org | 2013-04-26T05:14:19.821906Z

Changed paths:
   D http://src.chromium.org/viewvc/chrome/trunk/src/content/common/gpu/media/avc_config_record_builder.h?r1=196621&r2=196620&pathrev=196621
   D http://src.chromium.org/viewvc/chrome/trunk/src/ui/gfx/video_decode_acceleration_support_mac_unittest.mm?r1=196621&r2=196620&pathrev=196621
   D http://src.chromium.org/viewvc/chrome/trunk/src/ui/gfx/video_decode_acceleration_support_mac.mm?r1=196621&r2=196620&pathrev=196621
   M http://src.chromium.org/viewvc/chrome/trunk/src/chrome/browser/about_flags.cc?r1=196621&r2=196620&pathrev=196621
   M http://src.chromium.org/viewvc/chrome/trunk/src/content/common/gpu/media/video_decode_accelerator_unittest.cc?r1=196621&r2=196620&pathrev=196621
   M http://src.chromium.org/viewvc/chrome/trunk/src/content/content_tests.gypi?r1=196621&r2=196620&pathrev=196621
   D http://src.chromium.org/viewvc/chrome/trunk/src/content/common/gpu/media/rendering_helper_mac.mm?r1=196621&r2=196620&pathrev=196621
   D http://src.chromium.org/viewvc/chrome/trunk/src/content/common/gpu/media/mac_video_decode_accelerator.mm?r1=196621&r2=196620&pathrev=196621
   M http://src.chromium.org/viewvc/chrome/trunk/src/ui/ui_unittests.gypi?r1=196621&r2=196620&pathrev=196621
   D http://src.chromium.org/viewvc/chrome/trunk/src/ui/gfx/video_decode_acceleration_support_mac.h?r1=196621&r2=196620&pathrev=196621
   M http://src.chromium.org/viewvc/chrome/trunk/src/content/common/gpu/media/gpu_video_decode_accelerator.cc?r1=196621&r2=196620&pathrev=196621
   M http://src.chromium.org/viewvc/chrome/trunk/src/ui/ui.gyp?r1=196621&r2=196620&pathrev=196621
   D http://src.chromium.org/viewvc/chrome/trunk/src/content/common/gpu/media/mac_video_decode_accelerator.h?r1=196621&r2=196620&pathrev=196621
   D http://src.chromium.org/viewvc/chrome/trunk/src/content/common/gpu/media/avc_config_record_builder_unittest.cc?r1=196621&r2=196620&pathrev=196621
   D http://src.chromium.org/viewvc/chrome/trunk/src/content/common/gpu/media/avc_config_record_builder.cc?r1=196621&r2=196620&pathrev=196621
   M http://src.chromium.org/viewvc/chrome/trunk/src/content/content_common.gypi?r1=196621&r2=196620&pathrev=196621

Delete MacVDA from the codebase since it is dead code and not being worked on.

http://crbug.com/133828#c36 summarizes why this wasn't turned on.
Also related: http://openradar.appspot.com/radar?id=1845403 

BUG=133828

Review URL: https://chromiumcodereview.appspot.com/14401013
------------------------------------------------------------------------
Apr 28, 2013
#49 fischman@chromium.org
WontFix'ing pending finding the right person to work on this (currently unprioritized).
Status: WontFix
May 10, 2013
#50 evan@chromium.org
This bug is linked from chrome://gpu but is marked restrict-view-google -- intentional?
Dec 20, 2013
#53 fischman@chromium.org
Removing R-V-G since it's referred to by about:gpu, there's nothing confidential here, and the Launchiness of the bug has been aborted for many months.
Dec 20, 2013
#54 fischman@chromium.org
(No comment was entered for this change.)
Labels: -Restrict-View-Google
Dec 20, 2013
#55 fischman@chromium.org
Switching bug Type to see if that opens up the bug (even though I removed R-V-G the bug is still Restricted)
Labels: -Type-Launch Type-Feature
Feb 28, 2014
#56 dongseon...@intel.com
(No comment was entered for this change.)
Cc: dongseon...@intel.com
Jun 10, 2014
#57 eri...@chromium.org
reactivating and assigning to sandersd@, who will be working on this.
Status: Assigned
Owner: sande...@chromium.org
Jun 10, 2014
#58 smori...@adobe.com
(No comment was entered for this change.)
Cc: adobe-fl...@chromium.org
Jun 10, 2014
#59 fischman@chromium.org
(No comment was entered for this change.)
Cc: -fischman@chromium.org
Jun 11, 2014
#60 bugdro...@chromium.org
The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/c78d8192e1159aab87b03bed005e776f4dcee173

commit c78d8192e1159aab87b03bed005e776f4dcee173
Author: sandersd@chromium.org <sandersd@chromium.org@0039d316-1c4b-4281-b951-d872f2087c98>
Date: Wed Jun 11 07:21:56 2014

Create VTVideoDecodeAccelerator.

This is a stub for what will eventually become a video decode accelerator for Mac.

BUG=133828

Review URL: https://codereview.chromium.org/322433004

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@276290 0039d316-1c4b-4281-b951-d872f2087c98


Jun 19, 2014
#62 sande...@chromium.org
Status update: My VideoToolbox-based implementation is now operational enough to decode (but not render) frames. The next step is to check in the VideoToolbox dynamic linking glue code, after which the rest of the work can be submitted in small pieces.

The good news is that so far it appears that VideoToolbox does not have the same video size restrictions that VDA had, so it is more likely that this can be enabled for all video (including MediaSource).

The bad news is that I've taken some implementation shortcuts that will restrict the implementation to OS X 10.9+. These are fixable, but I don't have test machines handy for older versions of OS X.
Status: Started
Jun 24, 2014
#63 bugdro...@chromium.org
The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/c2dac535dea5647736dec19201d34ea339fcd109

commit c2dac535dea5647736dec19201d34ea339fcd109
Author: sandersd@chromium.org <sandersd@chromium.org@0039d316-1c4b-4281-b951-d872f2087c98>
Date: Wed Jun 25 04:56:04 2014

Add VTSupport for dynamically linking VideoToolbox.

BUG=133828

Review URL: https://codereview.chromium.org/340933002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@279593 0039d316-1c4b-4281-b951-d872f2087c98


Jul 10, 2014
#65 bugdro...@chromium.org
The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/fa18b3c8505b45caea43f1997af81154003d2446

commit fa18b3c8505b45caea43f1997af81154003d2446
Author: sandersd@chromium.org <sandersd@chromium.org@0039d316-1c4b-4281-b951-d872f2087c98>
Date: Fri Jul 11 04:56:06 2014

Use PPS/SPS NALUs to initialize a VTDecompressionSession.

BUG=133828

Review URL: https://codereview.chromium.org/374153003

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@282549 0039d316-1c4b-4281-b951-d872f2087c98


Jul 10, 2014
#66 bugdro...@chromium.org
------------------------------------------------------------------
r282549 | sandersd@chromium.org | 2014-07-11T04:56:06.564426Z

Changed paths:
   M http://src.chromium.org/viewvc/chrome/trunk/src/content/content_common.gypi?r1=282549&r2=282548&pathrev=282549
   M http://src.chromium.org/viewvc/chrome/trunk/src/content/common/gpu/media/vt_video_decode_accelerator.cc?r1=282549&r2=282548&pathrev=282549
   M http://src.chromium.org/viewvc/chrome/trunk/src/content/common/gpu/media/vt_video_decode_accelerator.h?r1=282549&r2=282548&pathrev=282549

Use PPS/SPS NALUs to initialize a VTDecompressionSession.

BUG=133828

Review URL: https://codereview.chromium.org/374153003
-----------------------------------------------------------------
Aug 4, 2014
#67 bugdro...@chromium.org
The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/e68e3228ca7cd2988142ff4c90428bb0b35ab2f3

commit e68e3228ca7cd2988142ff4c90428bb0b35ab2f3
Author: sandersd@chromium.org <sandersd@chromium.org@0039d316-1c4b-4281-b951-d872f2087c98>
Date: Tue Aug 05 04:17:44 2014

Implement actually decoding frames in VTVideoDecodeAccelerator.

This adds translation from Annex B to AVCC format along with decoding
frames and binding them to textures.

It seems that kVTDecodeFrame_EnableTemporalProcessing is just a
suggestion to VideoToolbox, and one that it ignores. That means that,
for now, this code only outputs frames in the correct order for I-frame
only video.

BUG=133828

Review URL: https://codereview.chromium.org/397883002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@287451 0039d316-1c4b-4281-b951-d872f2087c98


Aug 4, 2014
#68 bugdro...@chromium.org
------------------------------------------------------------------
r287451 | sandersd@chromium.org | 2014-08-05T04:17:44.652934Z

Changed paths:
   M http://src.chromium.org/viewvc/chrome/trunk/src/content/common/gpu/media/vt_video_decode_accelerator.cc?r1=287451&r2=287450&pathrev=287451
   M http://src.chromium.org/viewvc/chrome/trunk/src/content/common/gpu/media/gpu_video_decode_accelerator.cc?r1=287451&r2=287450&pathrev=287451
   M http://src.chromium.org/viewvc/chrome/trunk/src/content/common/gpu/media/vt_video_decode_accelerator.h?r1=287451&r2=287450&pathrev=287451
   M http://src.chromium.org/viewvc/chrome/trunk/src/content/common/gpu/media/vt_stubs_header.fragment?r1=287451&r2=287450&pathrev=287451

Implement actually decoding frames in VTVideoDecodeAccelerator.

This adds translation from Annex B to AVCC format along with decoding
frames and binding them to textures.

It seems that kVTDecodeFrame_EnableTemporalProcessing is just a
suggestion to VideoToolbox, and one that it ignores. That means that,
for now, this code only outputs frames in the correct order for I-frame
only video.

BUG=133828

Review URL: https://codereview.chromium.org/397883002
-----------------------------------------------------------------
Aug 19, 2014
#69 i...@chromium.org
(No comment was entered for this change.)
Blocking: chromium:335836
Sep 26 (5 days ago)
#70 pinkerton@chromium.org
(No comment was entered for this change.)
Cc: pinkerton@chromium.org
Today (5 hours ago)
#72 pinkerton@chromium.org
(No comment was entered for this change.)
Cc: ccameron@chromium.org
Sign in to add a comment

Powered by Google Project Hosting