I want to land this so we can start testing color space aware
decoding on Android. In particular, it will be interesting to
see how linear premultiplication will affect existing content.
This will only modify BitmapRegionDecoder behavior. I'll
follow up with a similar change to BitmapFactory.cpp in Android.
This will cause image diffs on Gold.
BUG=skia:
Change-Id: Iffda5f035447f2608ce26945570b503f8971b735
Reviewed-on: https://skia-review.googlesource.com/5698
Commit-Queue: Matt Sarett <msarett@google.com>
Reviewed-by: Mike Reed <reed@google.com>
Reviewed-by: Leon Scroggins <scroggo@google.com>
To avoid undefined int underflow behavior.
BUG=skia:6017
R=mtklein@google.com,reed@google.com
Change-Id: Ib707ad5e1d87eda70525c56db14849c4164c5640
Reviewed-on: https://skia-review.googlesource.com/5861
Reviewed-by: Mike Klein <mtklein@chromium.org>
Commit-Queue: Florin Malita <fmalita@chromium.org>
Rather than auto-adding the Job, assert that it's listed. This enforces
that the JOBS list is accurate. Add all missing compile Jobs to the list.
BUG=skia:
Change-Id: Ic7a90165ccac36baa52a4674798977021d6812d7
Reviewed-on: https://skia-review.googlesource.com/5848
Reviewed-by: Mike Klein <mtklein@chromium.org>
Commit-Queue: Eric Boren <borenet@google.com>
Don't walk off the end if the loop doesn't
contain the expected value.
R=kjlubick@google.com
BUG=skia:6047
Change-Id: I96815180dc7c92b45691037ae6c4b40beedc009a
Reviewed-on: https://skia-review.googlesource.com/5845
Reviewed-by: Kevin Lubick <kjlubick@google.com>
Commit-Queue: Cary Clark <caryclark@google.com>
Prior to this CL, if a GIF file was truncated before reading the local
color map of a frame, incremental decode would do the wrong thing. In
onStartIncrementalDecode, we would either create a color table based on
the global color map, or we would create a dummy one with only one
color (transparent). The dummy color table is correct if there is
neither a global nor a local color map, and allows us to fill the frame
with transparent. But if more data is provided, and it includes an
actual color map and image data, one of the following can happen:
- If the created color table is smaller than the actual one, the
decoded data may include indices outside of the range of the created
color table, resulting in a crash.
- If we get lucky, and the created color table is large enough, it may
still be the wrong colors (and most likely is).
To solve this, make onStartIncrementalDecode fail if there is a local
color map that has not been read yet. A future call may read more data
and read the correct color map.
This is done by returning kIncompleteInput in
SkGifCodec::prepareToDecode if there is a local color map that has not
yet been read. (It is possible that there is no color map at all, in
which case we still need to support decoding that frame. Skip
attempting to decode in that case.)
In onGetPixels, if prepareToDecode returned kIncompleteInput, return
kInvalidInput. Although the input is technically incomplete, no future
call will provide more data (unlike in incremental decoding), and there
is nothing interesting for the client to draw. This also prevents
SkCodec from attempting to fill the data with an SkSwizzler, which has
not been created. (An alternative solution would be create the dummy
color table and an SkSwizzler, which would keep the current behavior.
But I think the new behavior of returning kInvalidInput makes more
sense.)
Add tests to verify the intended behavior:
- getPixels fails.
- startIncrementalDecode fails, but after providing more data it will
succeed and incremental decoding matches the image decoded from the
full stream.
- Both succeed if there is no color table at all.
Change-Id: Ifb52fe7f723673406a28e80c8805a552f0ac33b6
Reviewed-on: https://skia-review.googlesource.com/5758
Reviewed-by: Matt Sarett <msarett@google.com>
Commit-Queue: Leon Scroggins <scroggo@google.com>
This reverts commit ce33f10677.
Reason for revert: Breaking many gpu bots
Change-Id: I94c813ed6a9311458c872f74bb1b0792f46ff414
Reviewed-on: https://skia-review.googlesource.com/5737
Commit-Queue: Greg Daniel <egdaniel@google.com>
Reviewed-by: Greg Daniel <egdaniel@google.com>
Tool will now check for and output all unsuccessfully parsed ICC
profiles in input sksp images if --testColorCorrectionSupported is set
as a flag. All ICC-aware codecs had to be slightly modified in order to
expose this information, as the logic for accessing the ICC profiles is
all within the codecs. If --writeFailedImages is set, it will also
output all images whoses ICC profiles were not supported.
TBR=reed@google.com
BUG=skia:
Change-Id: Ic310d82bdebf92f8d3bc0ad3dcc688136b6de377
Reviewed-on: https://skia-review.googlesource.com/5355
Reviewed-by: Leon Scroggins <scroggo@google.com>
Reviewed-by: Matt Sarett <msarett@google.com>
Commit-Queue: Robert Aftias <raftias@google.com>
BUG=skia:6048
Change-Id: I86abfc0059c6523db08e552df16e88ae71379ed7
Reviewed-on: https://skia-review.googlesource.com/5733
Reviewed-by: Mike Klein <mtklein@chromium.org>
Commit-Queue: Eric Boren <borenet@google.com>
Instead of relying on cpu-features.c, just do what it does.
Good reading: http://man7.org/linux/man-pages/man3/getauxval.3.html
While it's nice to use the headers when possible, should either of these headers not be available, we can fall back to doing it all manually:
extern "C" uint32_t getauxval(uint32_t)
static const int AT_HWCAP = 16;
static const int HWCAP_CRC32 = (1<<7);
To keep things simple I've slimmed cpu feature detection down to just the features we actually make use of. This removes all runtime feature detection for ARMv7... we expect NEON to be globally available, and so far we haven't used the other FMA/FP16 bits on ARMv7. ARMv8 feature dection remains the same, CRC32 before, CRC32 after. x86 (cpuid-based detection) and MIPS (nothing) are untouched.
We need to keep //third_party/cpu-features for //third_party/libwebp.
Change-Id: I6c96df9a09ae68c8c0e54c1152aa177ba9bafc83
Reviewed-on: https://skia-review.googlesource.com/5800
Reviewed-by: Derek Sollenberger <djsollen@google.com>
Commit-Queue: Mike Klein <mtklein@chromium.org>
This is a precursor to using mask load, mask store, and gather instructions for f16. This is a slight performance win too, through slightly simpler code generation. Having done this, it now makes sense to give a name to f16->f32 conversion, from_f16().
Finally, while we're at this, also send store_f32 through store(), so that now all formats use load, gather, and store uniformly.
CQ_INCLUDE_TRYBOTS=skia.primary:Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Release-SKNX_NO_SIMD
Change-Id: I403f16f712936e2bcf3294e72c863cb6c6fbcf0c
Reviewed-on: https://skia-review.googlesource.com/5731
Reviewed-by: Herb Derby <herb@google.com>
Commit-Queue: Mike Klein <mtklein@chromium.org>
in response to 5784
Change-Id: I3ad34a30743e7ffbd04767668c288a4f884eb19c
Reviewed-on: https://skia-review.googlesource.com/5732
Reviewed-by: Leon Scroggins <scroggo@google.com>
Commit-Queue: Mike Klein <mtklein@chromium.org>
We were always already supplying this, makes it simpler
BUG=skia:
Change-Id: I36ac35205df5ab2a0fb7ec26e83ddb1547154816
Reviewed-on: https://skia-review.googlesource.com/5778
Reviewed-by: Brian Salomon <bsalomon@google.com>
Commit-Queue: Brian Osman <brianosman@google.com>
The storage cost is the same, so might as well do this when building the pipeline instead of when running it. This also avoids the awkward cvtsi2ss instruction that screws with register renaming.
CQ_INCLUDE_TRYBOTS=skia.primary:Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Release-SKNX_NO_SIMD
Change-Id: I1c7d5bad558870256a31e3da969eee5d80fb93a8
Reviewed-on: https://skia-review.googlesource.com/5782
Reviewed-by: Herb Derby <herb@google.com>
Commit-Queue: Mike Klein <mtklein@chromium.org>
Change-Id: Ie3c41aa0910599c9413b4943fbe63000226e526f
Reviewed-on: https://skia-review.googlesource.com/5776
Reviewed-by: Brian Osman <brianosman@google.com>
Commit-Queue: Brian Salomon <bsalomon@google.com>
Change-Id: I35efd4ad2b7132145c1e477f0b1f283276e9fad5
Reviewed-on: https://skia-review.googlesource.com/5704
Reviewed-by: Brian Osman <brianosman@google.com>
Commit-Queue: Brian Salomon <bsalomon@google.com>
- explicitly separate bilinear_ stages in x and y too
BUG=skia:
CQ_INCLUDE_TRYBOTS=skia.primary:Test-Ubuntu-GCC-GCE-CPU-AVX2-x86_64-Release-SKNX_NO_SIMD
Change-Id: Ib7b4f9d26ea6abe9171068e92424479d811ee606
Reviewed-on: https://skia-review.googlesource.com/5636
Reviewed-by: Brian Osman <brianosman@google.com>
Commit-Queue: Mike Klein <mtklein@chromium.org>
Change-Id: I1b3aee2c16075bc481d96052a82f3b3da82061fe
Reviewed-on: https://skia-review.googlesource.com/5699
Commit-Queue: Brian Salomon <bsalomon@google.com>
Reviewed-by: Brian Osman <brianosman@google.com>
Fixes a bad merge.
This reverts commit 073285c059.
Change-Id: I5e92339d9b33d3a6dc58b9fcd2a1b3a5684e8f8a
Reviewed-on: https://skia-review.googlesource.com/5774
Reviewed-by: Brian Salomon <bsalomon@google.com>
Commit-Queue: Brian Salomon <bsalomon@google.com>
It's mostly "How to Effectively Fight with GYP", which is no longer interesting.
BUG=skia:6043
NOTRY=true
DOCS_PREVIEW= https://skia.org/?cl=5761
Change-Id: I760e8d9c6577846b34d8f1f0d2e37b5a3c553ccf
Reviewed-on: https://skia-review.googlesource.com/5761
Reviewed-by: Heather Miller <hcm@google.com>
Commit-Queue: Mike Klein <mtklein@chromium.org>
This is less to type in most cases, and gives us more information
(for things like picture-backed images, where we need to know all
about the destination surface).
Additionally, strip out the plumbing entirely for bitmap sources,
where we don't need to know anything.
BUG=skia:
Change-Id: I4deff6c7c345fcf62eb08b2aff0560adae4313da
Reviewed-on: https://skia-review.googlesource.com/5748
Reviewed-by: Mike Klein <mtklein@chromium.org>
Reviewed-by: Brian Salomon <bsalomon@google.com>
Commit-Queue: Brian Osman <brianosman@google.com>
This reverts commit 3944484020.
Reason for revert: Merges badly with a recent change. Will rebase and reland.
Original change's description:
> Reland "Remove antialiasing control from GrPaint."
>
> This contains fixes for GLPrograms test and mixed samples rendering.
>
> This reverts commit 419d81eed4.
>
> BUG=skia:
>
> Change-Id: If8f002fbfaaaab6d1607403f2b15ccc7f1e17e87
> Reviewed-on: https://skia-review.googlesource.com/5763
> Commit-Queue: Brian Salomon <bsalomon@google.com>
> Reviewed-by: Robert Phillips <robertphillips@google.com>
>
TBR=bsalomon@google.com,robertphillips@google.com,reviews@skia.org
BUG=skia:
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
Change-Id: Iff9657041e28604a845bc5a9acec7c9b248c53bd
Reviewed-on: https://skia-review.googlesource.com/5772
Commit-Queue: Brian Salomon <bsalomon@google.com>
Reviewed-by: Brian Salomon <bsalomon@google.com>
This contains fixes for GLPrograms test and mixed samples rendering.
This reverts commit 419d81eed4.
BUG=skia:
Change-Id: If8f002fbfaaaab6d1607403f2b15ccc7f1e17e87
Reviewed-on: https://skia-review.googlesource.com/5763
Commit-Queue: Brian Salomon <bsalomon@google.com>
Reviewed-by: Robert Phillips <robertphillips@google.com>
No one (other than test code) was using this API, and it lacks the context
to do the right thing. Specifically, if this forces a decode of an encoded
image, we don't know the intended use (re: color spaces) to determine how
we should decode.
BUG=skia:
Change-Id: I6ff700b3a5adce8257f35c5e3dd5ba557b2a3219
Reviewed-on: https://skia-review.googlesource.com/5614
Reviewed-by: Brian Salomon <bsalomon@google.com>
Commit-Queue: Brian Osman <brianosman@google.com>
This CL reverts https://skia-review.googlesource.com/c/5148/ (Fix gpu blurring on platforms that "useDrawInsteadOfClear") (all the worstCaseWidth/Height stuff) and adds a new GrRenderTargetContext entry point (absClear) to specify clears that can't be discarded or altered.
BUG=skia:
Change-Id: I18b1373ecf4a153ca8c0f290ab8b1d00770426da
Reviewed-on: https://skia-review.googlesource.com/5484
Commit-Queue: Robert Phillips <robertphillips@google.com>
Reviewed-by: Brian Salomon <bsalomon@google.com>
We have much more capacity on the Pixel Cs.
BUG=skia:
Change-Id: I65e136d5e2f44ec4bef9c168a86e706537878532
Reviewed-on: https://skia-review.googlesource.com/5755
Reviewed-by: Mike Klein <mtklein@chromium.org>
Reviewed-by: Eric Boren <borenet@google.com>
Commit-Queue: Kevin Lubick <kjlubick@google.com>
BUG=chrome:655247
Change-Id: I1c987939ac224075cdd385c879dd091fb0efd82a
Reviewed-on: https://skia-review.googlesource.com/5753
Commit-Queue: Brian Osman <brianosman@google.com>
Reviewed-by: Brian Salomon <bsalomon@google.com>
Seems like the bots don't run hooks. That's okay, even ideal.
They'll keep getting GN via recipes.
DOCS_PREVIEW= https://skia.org/?cl=5725
Change-Id: I000bad3390dddaeb4548972f29c96b8b3288ea6c
Reviewed-on: https://skia-review.googlesource.com/5725
Commit-Queue: Mike Klein <mtklein@chromium.org>
Reviewed-by: Eric Boren <borenet@google.com>
BUG=skia:
Change-Id: I523357617135678d0be43a1b41150be58e89f780
Reviewed-on: https://skia-review.googlesource.com/5754
Reviewed-by: Mike Klein <mtklein@chromium.org>
Commit-Queue: Hal Canary <halcanary@google.com>
ClipPaths can be clipped too, e.g.:
<clipPath id="clip1" clip-path="url(#clip2)">...</clipPath>
Since we're not really drawing clips but resolving their geometry,
asPath() needs to take composed clipping into account (and intersect as
needed).
R=reed@google.com,robertphillips@google.com,stephana@google.com
Change-Id: I25959e22fe50f72042147cfe6b416b6b9ac20cd4
Reviewed-on: https://skia-review.googlesource.com/5720
Reviewed-by: Robert Phillips <robertphillips@google.com>
Commit-Queue: Florin Malita <fmalita@chromium.org>
The new GN doesn't like "ar = ar + ...", etc.
CQ_INCLUDE_TRYBOTS=skia.primary:Build-Win-Clang-arm64-Release-Android
Change-Id: Ib131ee367c4af144f8ffb8562fc26b67675e4f45
Reviewed-on: https://skia-review.googlesource.com/5726
Reviewed-by: Herb Derby <herb@google.com>
Commit-Queue: Mike Klein <mtklein@chromium.org>
If there is enough data in the stream to read the reported canvas size,
but not enough to read the first image's header, we do not know the
true canvas size, since we may expand it to fit the first frame. In
that case, return nullptr from NewFromStream.
Add a test.
SkGifCodec.cpp:
Correct a comment - parse returns false if there is a fatal error.
parse() returning true does not guarantee that the size was found.
Instead of checking the width and height, check to see whether the
first frame exists and has its header defined. If not, we do not yet
know the true canvas size. Assert that the canvas size is non-zero,
which is a fatal error from parse.
SkGifImageReader.cpp:
Move the code to set the header defined before the SkGIFSizeQuery exit
condition. This allows SkGifCodec to check the first frame's header to
determine whether the size is known.
GifTest.cpp:
Add a test which truncates the file just before the image header (and
after the global header). Prior to the other changes, this would create
an SkCodec. For an image that needs its canvas size expanded, the
SkCodec would have an incorrect size.
CodecPartialTest.cpp:
randPixels.gif now needs more than half of its data to create an
SkCodec, so set a minimum for test_partial.
Change-Id: I40482f524128b2f1fe59b8f27dd64c7cbe793079
Reviewed-on: https://skia-review.googlesource.com/5701
Reviewed-by: Derek Sollenberger <djsollen@google.com>
Commit-Queue: Leon Scroggins <scroggo@google.com>
This reverts commit 45aac57ac6.
Reason for revert:
Pixel test suppression has landed in Chrome:
https://codereview.chromium.org/2559213002 (Add suppression for GPU Pixel test in preparation for Skia CL)
Original change's description:
> Revert "Add antialiasing to SkDropShadowImageFilter's shadow draw"
>
> This reverts commit 78e8165ec3.
>
> Reason for revert: Chrome pixel test
>
> Original change's description:
> > Add antialiasing to SkDropShadowImageFilter's shadow draw
> >
> > A BW draw in the SkDropShadowImageFilter can lead to unexpected popping of the shadow when the dest rect lies on half pixel boundaries.
> >
> > Change-Id: Ibc59dacc79bca8955981ec2889e79facd7d2de83
> > Reviewed-on: https://skia-review.googlesource.com/5669
> > Reviewed-by: Brian Salomon <bsalomon@google.com>
> > Commit-Queue: Robert Phillips <robertphillips@google.com>
> >
>
> TBR=bsalomon@google.com,robertphillips@google.com
> NOPRESUBMIT=true
> NOTREECHECKS=true
> NOTRY=true
>
> Change-Id: I0b0073cd8d9f1502daaa452c2153af029c11e52a
> Reviewed-on: https://skia-review.googlesource.com/5652
> Commit-Queue: Robert Phillips <robertphillips@google.com>
> Reviewed-by: Robert Phillips <robertphillips@google.com>
>
TBR=bsalomon@google.com,robertphillips@google.com
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
Change-Id: I7d844c486fd6db57d3df60a32c12699f2d85ad16
Reviewed-on: https://skia-review.googlesource.com/5747
Commit-Queue: Robert Phillips <robertphillips@google.com>
Reviewed-by: Robert Phillips <robertphillips@google.com>
Mostly to get
5d7f4ce [GN] Fix generate Xcode project for macOS to work with Xcode 8.
After this lands, gclient sync && python bin/fetch-gn to get the new GN.
Change-Id: I3e1a186996b1db69ef767922a5b9efff173b950c
Reviewed-on: https://skia-review.googlesource.com/5723
Reviewed-by: Mike Reed <reed@google.com>
Commit-Queue: Mike Klein <mtklein@chromium.org>
