1067377 - Removed compatibility layer of libvpx

Status: RESOLVED FIXED

(Core :: Audio/Video, defect)

Description

[Steffen Michalke]

Attachment: libvpx_removed_compat-VP8TrackEncoder.patch

User Agent: Mozilla/5.0 (X11; Linux x86_64; rv:35.0) Gecko/20100101 Firefox/35.0
Build ID: 2014091513

Steps to reproduce:

I compiled the latest mozilla-central tree, --with-system-libvpx enabled.


Actual results:

Compilation fails, because the libvpx devs removed the compat layer which mapped the VPX_PLANE_* and VPX_IMG_FMT_* constants introduced in libvpx v0.9.1 to their older counterparts w/o the VPX_ prefix.


Expected results:

There are a few places where Mozilla still uses the old constants. I attached two patches which renames the constants.

Comment 1

[Steffen Michalke]

Attachment: The 2nd patch.

2nd patch.

Comment 2

[Ralph Giles (:rillian)]

Comment on attachment 8489368 [details] [diff] [review]
libvpx_removed_compat-VP8TrackEncoder.patch

Review of attachment 8489368 [details] [diff] [review]:
-----------------------------------------------------------------

Hi! Thanks for the patch submission. Normally you should set the r=? flag with one of the suggested reviewers to get prompt attention.

Comment 3

[Ralph Giles (:rillian)]

Comment on attachment 8489369 [details] [diff] [review]
The 2nd patch.

Passing review of the webrtc bit to jesup.

Steffan, can you take care of submitting this one upstream as well? This part of the code comes from webrtc.org.

Comment 4

[Ralph Giles (:rillian)]

Attachment: Part 3: Use prefixed VPX_IMG_FMT macros in WebMReader.

Comment 5

[Steffen Michalke]

I've submitted a bug report @ webrtc.org (#3839). The first two patches should therefore be obsolete as far as I can see, one could wait for a fixed upstream version of webrtc.

Comment 6

[Ralph Giles (:rillian)]

No, these should land. We want to update our in-tree copy of libvpx as well, and that will likely happen sooner than a webrtc update. Doing so requires these patches, so thanks for the timely fix!

Comment 7

[Ralph Giles (:rillian)]

I also wonder how chromium handles this, since they're using an equally new libvpx.

Comment 8

[Ralph Giles (:rillian)]

https://hg.mozilla.org/integration/mozilla-inbound/rev/fdf547a07616
https://hg.mozilla.org/integration/mozilla-inbound/rev/6b34adefbcac
https://hg.mozilla.org/integration/mozilla-inbound/rev/cb9c67b221bb

Comment 9

[Carsten Book [:Tomcat]]

https://hg.mozilla.org/mozilla-central/rev/fdf547a07616
https://hg.mozilla.org/mozilla-central/rev/6b34adefbcac
https://hg.mozilla.org/mozilla-central/rev/cb9c67b221bb

Comment 10

[Steffen Michalke]

(In reply to Ralph Giles (:rillian) from comment #7)
> I also wonder how chromium handles this, since they're using an equally new
> libvpx.

They are fixing it on their own before it goes upstream with webrtc. Their pritority.