1907121 - VideoFrame copyTo crashes when converting I420 frame with rgb matrix to rgb buffer

Status: RESOLVED FIXED

(Core :: Audio/Video: Web Codecs, defect)

Description

[C.M.Chang[:chunmin] | PTO]

Attachment: test.html

VideoFrame copyTo crashes when converting I420 frame with a colorSpace with rgb matrix.

Comment 1

[C.M.Chang[:chunmin] | PTO]

Attachment: Bug 1907121 - Add crashtest for VideoFrame copyTo with I420 and RGB colorspace matrix

VideoFrame copyTo will crash when converting a I420 VideoFrame
in colorSpace with rgb matrix.

Comment 2

[C.M.Chang[:chunmin] | PTO]

Attachment: Bug 1907121 - Alphabetize include headers in yuv_convert.cpp

Depends on D216184

Comment 3

[C.M.Chang[:chunmin] | PTO]

Attachment: Bug 1907121 - Modify yuv_convert APIs to return libyuv execution results

Depends on D216185

Comment 4

[C.M.Chang[:chunmin] | PTO]

Attachment: Bug 1907121 - Sort include headers in YCbCrUtils.* per coding style

Depends on D216186

Comment 5

[C.M.Chang[:chunmin] | PTO]

Attachment: Bug 1907121 - Modify YCbCrUtils APIs to return yuv_convert execution results

Depends on D216187

Comment 6

[C.M.Chang[:chunmin] | PTO]

Attachment: Bug 1907121 - Add validation for 16-bit to 8-bit color conversion in YCbCrUtils APIs

Depends on D216189

Comment 7

[C.M.Chang[:chunmin] | PTO]

Attachment: Bug 1907121 - Move YCbCrUtils API assertions to caller side

Previously, the YCbCrUtils APIs assert the success inside the APIs
themselves. With the changes made in previous patches, those assertions
are removed. This patch addes those assertion back, but on callers side
instead.

Depends on D216190

Comment 8

[C.M.Chang[:chunmin] | PTO]

Attachment: Bug 1907121 - Return error for non-I444 Identity color space in ConvertYCbCrToRGB32

Depends on D216191

Comment 9

[Donal Meehan [:dmeehan]]

:chumin release tracking requests are automatically added when the S1 severity is used.
Checking if you intended to set this as an S1?
https://firefox-source-docs.mozilla.org/bug-mgmt/guides/severity.html

Comment 10

[C.M.Chang[:chunmin] | PTO]

Ah, my bad. It's shouldn't be S1. I am setting it to S4 right now.

Comment 11

[C.M.Chang[:chunmin] | PTO]

Attachment: Bug 1907121 - Correct type for stride

Depends on D216190

Comment 12

[Pulsebot]

Pushed by [e-mail]:
https://hg.mozilla.org/integration/autoland/rev/ba07896445cf
Add crashtest for VideoFrame copyTo with I420 and RGB colorspace matrix r=media-playback-reviewers,alwu
https://hg.mozilla.org/integration/autoland/rev/c9962b7dabd2
Alphabetize include headers in yuv_convert.cpp r=gfx-reviewers,nical
https://hg.mozilla.org/integration/autoland/rev/91d6c77b02d4
Modify yuv_convert APIs to return libyuv execution results r=gfx-reviewers,nical
https://hg.mozilla.org/integration/autoland/rev/c5ae9c49d02b
Sort include headers in YCbCrUtils.* per coding style r=gfx-reviewers,nical
https://hg.mozilla.org/integration/autoland/rev/497fa8c01852
Modify YCbCrUtils APIs to return yuv_convert execution results r=gfx-reviewers,nical
https://hg.mozilla.org/integration/autoland/rev/03cf7c930405
Add validation for 16-bit to 8-bit color conversion in YCbCrUtils APIs r=gfx-reviewers,nical
https://hg.mozilla.org/integration/autoland/rev/1ae7f5736844
Correct type for stride r=gfx-reviewers,bradwerth
https://hg.mozilla.org/integration/autoland/rev/c844e41d16c9
Move YCbCrUtils API assertions to caller side r=gfx-reviewers,nical
https://hg.mozilla.org/integration/autoland/rev/227c24230d4a
Return error for non-I444 Identity color space in ConvertYCbCrToRGB32 r=gfx-reviewers,nical

Comment 13

[pstanciu]

https://hg.mozilla.org/mozilla-central/rev/ba07896445cf
https://hg.mozilla.org/mozilla-central/rev/c9962b7dabd2
https://hg.mozilla.org/mozilla-central/rev/91d6c77b02d4
https://hg.mozilla.org/mozilla-central/rev/c5ae9c49d02b
https://hg.mozilla.org/mozilla-central/rev/497fa8c01852
https://hg.mozilla.org/mozilla-central/rev/03cf7c930405
https://hg.mozilla.org/mozilla-central/rev/1ae7f5736844
https://hg.mozilla.org/mozilla-central/rev/c844e41d16c9
https://hg.mozilla.org/mozilla-central/rev/227c24230d4a