The only way that this can happen is if we get through the
ShouldTreatAsCompleteDueToSyncDecode check returning true in the case of
the image being errored.
https://hg.mozilla.org/integration/autoland/rev/645a4d6461ca was
supposed to deal with this, but my guess is that there is a slight race
condition in which the error status isn't there at the beginning, but is
there after the StartDecoding call.
It seems returning BAD_IMAGE rather than painting transparent if we hit
a broken image is a better thing to do than what we're doing now, and
should fix the intermittent issue.
Differential Revision: https://phabricator.services.mozilla.com/D101361
From what I can see in source history, pal8v4.bmp started out with a fuzzy of (3, 6376) on all platforms. Then one day aarch64-windows started producing a result of (1, 899) and so that platform's expectations were adjusted.
In the upcoming clang 12, the behavior of this test gets "fixed" in that we go back to those 6376 differing pixels like on other platforms. The max difference rises to 4 though. In light of the fact that aarch64-windows is seeing less priority these days, I can't justify digging into the exact code reason for the change, so this patch just updates the fuzzy setting to allow these values.
Differential Revision: https://phabricator.services.mozilla.com/D100823
This patch implements the transparency support for AVIF image files.
To convert the decoded YCbCr and Alpha data to RGBA, a function named
`ConvertYCbCrAToARGB` is created to do this job in the following
procedure:
ConvertYCbCrAToARGB:
If the layout of the YCbCr is I420
Calling libyuv::I420AlphaToARGB
Else
Fill RGB data converted by ConvertYCbCrToRGB in ARGB buffer first
Insert the alpha data to ARGB buffer
On the other hand, this patch refactors the nsAVIFDecoder a bit to make
the lifetime of the parsed data and decoded image data clearer. They
won't live longer than Parser object and {Dav1d, AOM}Decoder object.
This should improve the code readability.
This patch also adds a transparent image test (TransparentAVIFTestCase)
to check the FLAG_HAS_TRANSPARENCY is posted or not. The test image file
`transparent.avif` is from Bug 1654462.
Differential Revision: https://phabricator.services.mozilla.com/D98951
Import the improvements made in mp4parse-rust repo. The changes would
save some redundant copy when calling avif related APIs and provide the
ability to get the alpha data of the parsed avif image.
Differential Revision: https://phabricator.services.mozilla.com/D98950
Import the improvements made in mp4parse-rust repo. The changes would
save some redundant copy when calling avif related APIs and provide the
ability to get the alpha data of the parsed avif image.
Differential Revision: https://phabricator.services.mozilla.com/D98950
I was running into issues where these names would conflict with the type's own Get/Set methods
and these names have the added benefit of indicating a bit more that atomic stuff is going on.
Differential Revision: https://phabricator.services.mozilla.com/D99268
When loading a progressive jpegs with non-interleaved DC scans,
we add a check to see if we have already seen data from all DC
channels and only start rendering when this is the case.
Differential Revision: https://phabricator.services.mozilla.com/D95683
In https://hg.mozilla.org/integration/autoland/rev/471ad96ddc3e we changed this from calling |Surface().IsFullyDecoded()| to calling Seek because IsFullyDecoded always returns false for large enough animations because we don't keep around all of their frames.
In the old model, where mIsCurrentlyDecoded tracks if the image is fully decoded, we don't want to set mCompositedFrameInvalid to true if simply if the image isn't fully decoded because the current frame could be available.
In the new model, where mIsCurrentlyDecoded tracks if the current frame is available, we definitely do want to mark mCompositedFrameInvalid if the current frame is not available.
Depends on D96944
Differential Revision: https://phabricator.services.mozilla.com/D96945
In https://hg.mozilla.org/integration/autoland/rev/471ad96ddc3e we changed this from calling |Surface().IsFullyDecoded()| to calling Seek because IsFullyDecoded always returns false for large enough animations because we don't keep around all of their frames.
However RequestRefresh/AdvanceFrame call GetFrame. Seek is intended to be used for display purposes, GetFrame is intended to be used for advancing the frame purposes. And Seek and GetFrame can actually returns different results if it is the first frame that we are asking for.
Differential Revision: https://phabricator.services.mozilla.com/D96944
Add a telemetry probe tracking the errors in dav1d or aom decoder. We
only care the errors from `aom_codec_decode` and `dav1d_get_picture` so
other error code won't be counted.
Differential Revision: https://phabricator.services.mozilla.com/D94489
Allow-list all Python code in tree for use with the black linter, and re-format all code in-tree accordingly.
To produce this patch I did all of the following:
1. Make changes to tools/lint/black.yml to remove include: stanza and update list of source extensions.
2. Run ./mach lint --linter black --fix
3. Make some ad-hoc manual updates to python/mozbuild/mozbuild/test/configure/test_configure.py -- it has some hard-coded line numbers that the reformat breaks.
4. Make some ad-hoc manual updates to `testing/marionette/client/setup.py`, `testing/marionette/harness/setup.py`, and `testing/firefox-ui/harness/setup.py`, which have hard-coded regexes that break after the reformat.
5. Add a set of exclusions to black.yml. These will be deleted in a follow-up bug (1672023).
# ignore-this-changeset
Differential Revision: https://phabricator.services.mozilla.com/D94045
Allow-list all Python code in tree for use with the black linter, and re-format all code in-tree accordingly.
To produce this patch I did all of the following:
1. Make changes to tools/lint/black.yml to remove include: stanza and update list of source extensions.
2. Run ./mach lint --linter black --fix
3. Make some ad-hoc manual updates to python/mozbuild/mozbuild/test/configure/test_configure.py -- it has some hard-coded line numbers that the reformat breaks.
4. Make some ad-hoc manual updates to `testing/marionette/client/setup.py`, `testing/marionette/harness/setup.py`, and `testing/firefox-ui/harness/setup.py`, which have hard-coded regexes that break after the reformat.
5. Add a set of exclusions to black.yml. These will be deleted in a follow-up bug (1672023).
# ignore-this-changeset
Differential Revision: https://phabricator.services.mozilla.com/D94045
sizemode/displaymode media queries only affect a given browsing context
tree so there's no need to propagate the change to images in that case.
Differential Revision: https://phabricator.services.mozilla.com/D94422
Allow-list all Python code in tree for use with the black linter, and re-format all code in-tree accordingly.
To produce this patch I did all of the following:
1. Make changes to tools/lint/black.yml to remove include: stanza and update list of source extensions.
2. Run ./mach lint --linter black --fix
3. Make some ad-hoc manual updates to python/mozbuild/mozbuild/test/configure/test_configure.py -- it has some hard-coded line numbers that the reformat breaks.
4. Add a set of exclusions to black.yml. These will be deleted in a follow-up bug (1672023).
# ignore-this-changeset
Differential Revision: https://phabricator.services.mozilla.com/D94045
This changes the way we deal with page use counters so that we can
handle out of process iframes.
Currently, when a parent document is being destroyed, we poke into all
of the sub-documents to merge their use counters into the parent's page
use counters, which we then report via Telemetry. With Fission enabled,
the sub-documents may be out of process. We can't simply turn these
into async IPC calls, since the parent document will be destroyed
shortly, as might the content processes holding the sub-documents.
So instead, each document during its initialization identifies which
ancestor document it will contribute its page use counters to, and
stores its WindowContext id to identify that ancestor. A message is
sent to the parent process to notify it that page use counter data will
be sent at some later point. That later point is when the document
loses its window. It doesn't matter if the ancestor document has
already been destroyed at this point, since all we need is its
WindowContext id to uniquely identify it. Once the parent process has
received all of the use counters it expects to accumulate to a given
WindowContext Id, it reports them via Telemetry.
Reporting of document use counters remains unchanged and is done by each
document in their content process.
While we're here, we also:
* Limit use counters to be reported for a pre-defined set of document
URL schemes, rather than be based on the document principal.
* Add proper MOZ_LOG logging for use counters instead of printfs.
Differential Revision: https://phabricator.services.mozilla.com/D87188
Emilio Cobos Álvarez
Jon Bauman
David Major
Andi-Bogdan Postelnicu
Masatoshi Kimura
Paul Zuehlcke
Chun-Min Chang
Kartik Gautam
Dorel Luca
Csoregi Natalia
Alexis Beingessner
Timothy Nikkel
Moritz Firsching
Simon Giesecke
Sylvestre Ledru
Matt Woodrow
Razvan Maries
Lee Salzman
Butkovits Atila
Randell Jesup
Bas Schouten
Ricky Stewart
Bogdan Tara
Dimi Lee
Cameron McCormack