A recent CL (http://review.skia.org/464121) made it an error to coerce a
literal value to a type that cannot hold the value. The fuzzer found a
case where we assumed type-coercion of a literal would always succeed,
and failed to null-check the result. We now null-check the result.
Change-Id: Id97c6016e56c20ef724028f71bbf4688dde3c064
Bug: oss-fuzz:40428
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/464919
Commit-Queue: Brian Osman <brianosman@google.com>
Reviewed-by: Brian Osman <brianosman@google.com>
Auto-Submit: John Stiles <johnstiles@google.com>
On low memory machines a common cause of crashes is when a draw
operation requires a large layer (or many large layers) because an
implementation of SkSurface_Base::onCopyOnWrite does not actually have
the resources available to do the copy when required by a draw. Allow
this method to fail and percolate up the call chain so that the draw
simply does not happen instead of crashing.
It appears some users are creating their own subclasses of
SkSurface_Base for their tests. Add SK_SURFACE_COPY_ON_WRITE_CRASHES to
keep the old crashy behavior until they can be updated.
Bug: chromium:1116362
Revert: 5b19ebe0c5.
Revert-Change-Id: I2873589f996ded9c9fd6d27b19155ca18d5b5326
Revert-Reviewed-on: https://skia-review.googlesource.com/c/skia/+/463956
Change-Id: Id06d1e6d3aacb409a3b00b9a862bd8ddd1aaa22f
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/464456
Reviewed-by: Greg Daniel <egdaniel@google.com>
Commit-Queue: Ben Wagner <bungeman@google.com>
Yesterday's negation-related changes (http://review.skia.org/464123)
exposed a flaw that the fuzzer was able to exploit. We were previously
able to assume that `simplify_negation` would always return a non-null
expression; in some cases, that is no longer true.
Change-Id: Ia585232b0e35fafe0c642384a59ef94ce743ffd5
Bug: oss-fuzz:40427
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/464916
Commit-Queue: John Stiles <johnstiles@google.com>
Commit-Queue: Brian Osman <brianosman@google.com>
Auto-Submit: John Stiles <johnstiles@google.com>
Reviewed-by: Brian Osman <brianosman@google.com>
SkTypes.h makes sure SK_SUPPORT_GPU is defined.
I was tempted to move these files to include/gpu and
src/gpu, but that might be best handled in a follow-up
CL in case any clients depend on them.
Change-Id: Ia4d6717567fe6b1842bed2c7fc0439e85e1795b2
Bug: skia:12584
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/464638
Reviewed-by: Brian Osman <brianosman@google.com>
Adds Attributes and supporting types to RenderPipelineDesc so they can
be created for the RenderPipeline.
Bug: skia:12466
Change-Id: I7ed920ea6d44f27f7dace81d35cd967a8dea55de
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/464377
Reviewed-by: Greg Daniel <egdaniel@google.com>
Commit-Queue: Jim Van Verth <jvanverth@google.com>
This reverts commit c77620c308.
Reason for revert: Android failing LayerTypeAndRenderTypeTransactionTest#SetAlphaClamped
Original change's description:
> Move alpha modulation into paint conversion (Step 1)
>
> Skia has a rule that says (roughly), any SkShader is modulated by the
> paint alpha. The CPU backend implements this precisely - the entire
> shader sees the opaque paint color, and the alpha is applied once at
> the root. The GPU backend has previously made each SkShader responsible
> for doing this modulation. That led to two kinds of problems:
>
> 1) Some shaders forgot to do the modulation at all.
> 2) In shader trees, it's possible for the alpha to get applied
> multiple times, or to child effects where it isn't desired.
>
> A consequence of #2 is that the GPU implementation of things like
> blend-shaders have to jump through hoops to invoke their children with
> opaque versions of their input colors.
>
> This CL allows us to remove the explicit alpha modulation from a variety
> of shaders, and should also cleanup of the blend shader logic. However,
> to get the CL past chrome's layout tests, we need to guard the change.
> So this version just makes the major change so paint conversion, in a
> way that will work universally. Once chrome switches to the new method,
> it will be safe to clean up the various shader implementations.
>
> Bug: skia:11942
> Change-Id: I518534becdd5d10cbd78cf3ff7d4a46dd1faabf9
> Reviewed-on: https://skia-review.googlesource.com/c/skia/+/464116
> Reviewed-by: Brian Salomon <bsalomon@google.com>
> Commit-Queue: Brian Osman <brianosman@google.com>
Bug: skia:11942
Change-Id: I382945cfe441412fe440cf2a658b5caa2e3ff696
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/464697
Auto-Submit: Brian Osman <brianosman@google.com>
Commit-Queue: Rubber Stamper <rubber-stamper@appspot.gserviceaccount.com>
Bot-Commit: Rubber Stamper <rubber-stamper@appspot.gserviceaccount.com>
With this change, we no longer have any SkSL tests which are able to
make a Literal integer that overflows its type. Literal::MakeInt now
asserts that its value is within bounds. I look forward to the fuzzer's
inevitable attempts to trigger these assertions.
Change-Id: I7b15e862caaf65984d33f5d72d2c1de816d1d292
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/464124
Auto-Submit: John Stiles <johnstiles@google.com>
Commit-Queue: Brian Osman <brianosman@google.com>
Reviewed-by: Brian Osman <brianosman@google.com>
This aligns the order of the two checks that we do here. It's rare for a
device to advertise both extensions, but we've now seen one that does,
and it requires that we use the image format from the EXT extension.
Change-Id: I8c7af72d852b915cd6645aaa1ec4e0db0f7d0d31
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/464696
Reviewed-by: Brian Salomon <bsalomon@google.com>
Commit-Queue: Brian Osman <brianosman@google.com>
Change-Id: Icf6f45069b078f7936cfa08224fd8796d8c283b4
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/464122
Commit-Queue: Kevin Lubick <kjlubick@google.com>
Reviewed-by: Brian Osman <brianosman@google.com>
This reverts commit 2e228bb72c.
Reason for revert: breaks android roll
Original change's description:
> Avoid crash when surface CoW allocation fails
>
> On low memory machines a common cause of crashes is when a draw
> operation requires a large layer (or many large layers) because an
> implementation of SkSurface_Base::onCopyOnWrite does not actually have
> the resources available to do the copy when required by a draw. Allow
> this method to fail and percolate up the call chain so that the draw
> simply does not happen instead of crashing.
>
> Bug: chromium:1116362
> Change-Id: I2873589f996ded9c9fd6d27b19155ca18d5b5326
> Reviewed-on: https://skia-review.googlesource.com/c/skia/+/463956
> Reviewed-by: Greg Daniel <egdaniel@google.com>
> Reviewed-by: Herb Derby <herb@google.com>
> Commit-Queue: Ben Wagner <bungeman@google.com>
Bug: chromium:1116362
Change-Id: I5ab590b6fc14bcb6712c00dda75d1e7cdc931447
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/464077
Auto-Submit: Greg Daniel <egdaniel@google.com>
Commit-Queue: Rubber Stamper <rubber-stamper@appspot.gserviceaccount.com>
Bot-Commit: Rubber Stamper <rubber-stamper@appspot.gserviceaccount.com>
SkImage is supposed to be immutable, convert remaining methods to const.
Change-Id: Icf673204474f09992a57c10f29703ae7b33e3904
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/464256
Commit-Queue: Florin Malita <fmalita@google.com>
Reviewed-by: Brian Salomon <bsalomon@google.com>
This was mistakenly using dFdx in some portions (copy-paste error).
Change-Id: Ifb159b3c44185d9166c10725b24002a28a0895b2
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/464381
Commit-Queue: John Stiles <johnstiles@google.com>
Commit-Queue: Brian Osman <brianosman@google.com>
Auto-Submit: John Stiles <johnstiles@google.com>
Reviewed-by: Brian Osman <brianosman@google.com>
On low memory machines a common cause of crashes is when a draw
operation requires a large layer (or many large layers) because an
implementation of SkSurface_Base::onCopyOnWrite does not actually have
the resources available to do the copy when required by a draw. Allow
this method to fail and percolate up the call chain so that the draw
simply does not happen instead of crashing.
Bug: chromium:1116362
Change-Id: I2873589f996ded9c9fd6d27b19155ca18d5b5326
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/463956
Reviewed-by: Greg Daniel <egdaniel@google.com>
Reviewed-by: Herb Derby <herb@google.com>
Commit-Queue: Ben Wagner <bungeman@google.com>
This class will manage and suballocate all buffers used for draws. This
includes vertex, index, and uniform buffers.
Bug: skia:12466
Change-Id: I30f498fe3465e4cc693b974019dcef27efaa8073
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/462097
Commit-Queue: Greg Daniel <egdaniel@google.com>
Reviewed-by: Michael Ludwig <michaelludwig@google.com>
This is a reland of 8a2a020ef4
Original change's description:
> Create a common Tessellation.h header
>
> This header and corresponding implementation will contain common
> definitions and subroutines for tessellation code.
>
> Bug: skia:12524
> Change-Id: Ib29b444177f284acb88a3d5644936674c48c0b89
> Reviewed-on: https://skia-review.googlesource.com/c/skia/+/463437
> Commit-Queue: Chris Dalton <csmartdalton@google.com>
> Reviewed-by: Greg Daniel <egdaniel@google.com>
Bug: skia:12524
Change-Id: I874fed63cdab50df841e8a5d25d8c822690b5af8
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/464294
Reviewed-by: Greg Daniel <egdaniel@google.com>
Commit-Queue: Chris Dalton <csmartdalton@google.com>
Bug: skia:12466
Change-Id: If633ce39c8f45b1ee3c042b5b72d7e0f95ca5c19
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/459597
Reviewed-by: Jim Van Verth <jvanverth@google.com>
Commit-Queue: Robert Phillips <robertphillips@google.com>
Previously, we would create a Literal with the negated value even if it
was outside the type's minimum/maximum values. Error reporting would
happen elsewhere, if at all (e.g. during assignment or coercion).
Change-Id: I020a93daf2b0f5741fb805a58a690489d7578dab
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/464123
Commit-Queue: John Stiles <johnstiles@google.com>
Auto-Submit: John Stiles <johnstiles@google.com>
Reviewed-by: Brian Osman <brianosman@google.com>
Previously, we would create SkSL literals of ints that didn't fit into
an int. This change causes a few errors to report differently. (In
particular, we no longer create global variables containing values that
wouldn't fit in that variable, so those symbols are invalid later.)
Change-Id: I29d219e853126ea78dd2d2a6d8a69b23ef2b06b8
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/464121
Auto-Submit: John Stiles <johnstiles@google.com>
Reviewed-by: Brian Osman <brianosman@google.com>
Commit-Queue: John Stiles <johnstiles@google.com>
This reverts commit 8a2a020ef4.
Reason for revert: breaking chrome roll. Needs change in chrome to land first
Original change's description:
> Create a common Tessellation.h header
>
> This header and corresponding implementation will contain common
> definitions and subroutines for tessellation code.
>
> Bug: skia:12524
> Change-Id: Ib29b444177f284acb88a3d5644936674c48c0b89
> Reviewed-on: https://skia-review.googlesource.com/c/skia/+/463437
> Commit-Queue: Chris Dalton <csmartdalton@google.com>
> Reviewed-by: Greg Daniel <egdaniel@google.com>
Bug: skia:12524
Change-Id: Ifd7e81eed8c337506d67ecd0fd2501eedf2e4408
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/464297
Auto-Submit: Greg Daniel <egdaniel@google.com>
Commit-Queue: Rubber Stamper <rubber-stamper@appspot.gserviceaccount.com>
Bot-Commit: Rubber Stamper <rubber-stamper@appspot.gserviceaccount.com>
This reverts commit 75e1f4c028.
Reason for revert: need to revert parent CL
Original change's description:
> Lift recursive curve culling out of tessellators
>
> We need recursive chopping/culling logic in order to draw
> astronomically large paths. But rather than do that at the same time
> the tessellators chop curves, this CL moves that logic into an
> SkPath -> SkPath transformation that runs ahead of time (and only if
> the path is extremely large to begin with). This will enable us to
> remove recursion from the tessellators and quickly determine ahead of
> time the size of buffers they need.
>
> Bug: skia:12524
> Change-Id: Ib2800fb23054f1548501811203173e58273fbc83
> Reviewed-on: https://skia-review.googlesource.com/c/skia/+/463936
> Commit-Queue: Chris Dalton <csmartdalton@google.com>
> Reviewed-by: Greg Daniel <egdaniel@google.com>
Bug: skia:12524
Change-Id: Ib075e8398767fd7b9f49e50ecee98ce47a2ad6b1
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/464296
Auto-Submit: Greg Daniel <egdaniel@google.com>
Bot-Commit: Rubber Stamper <rubber-stamper@appspot.gserviceaccount.com>
Commit-Queue: Greg Daniel <egdaniel@google.com>
This was causing errors in UBSAN when compiling some of our existing
SkSL tests.
Change-Id: I66f22607094df77d47ff70948a139c77feae8624
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/464118
Auto-Submit: John Stiles <johnstiles@google.com>
Commit-Queue: Brian Osman <brianosman@google.com>
Reviewed-by: Brian Osman <brianosman@google.com>
Check number of palette entries from FreeType and wrap palette in SkSpan
to pass that around, do size checks when accessing it.
Bug: skia:12579
Cq-Include-Trybots: luci.skia.skia.primary:Test-Android-Clang-GalaxyS20-GPU-MaliG77-arm64-Release-All-Android_NativeFonts
Change-Id: I1f3705a8c88225d95fb11fd42a03f220eac9334b
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/463657
Commit-Queue: Dominik Röttsches <drott@google.com>
Reviewed-by: Ben Wagner <bungeman@google.com>
We need recursive chopping/culling logic in order to draw
astronomically large paths. But rather than do that at the same time
the tessellators chop curves, this CL moves that logic into an
SkPath -> SkPath transformation that runs ahead of time (and only if
the path is extremely large to begin with). This will enable us to
remove recursion from the tessellators and quickly determine ahead of
time the size of buffers they need.
Bug: skia:12524
Change-Id: Ib2800fb23054f1548501811203173e58273fbc83
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/463936
Commit-Queue: Chris Dalton <csmartdalton@google.com>
Reviewed-by: Greg Daniel <egdaniel@google.com>
Skia has a rule that says (roughly), any SkShader is modulated by the
paint alpha. The CPU backend implements this precisely - the entire
shader sees the opaque paint color, and the alpha is applied once at
the root. The GPU backend has previously made each SkShader responsible
for doing this modulation. That led to two kinds of problems:
1) Some shaders forgot to do the modulation at all.
2) In shader trees, it's possible for the alpha to get applied
multiple times, or to child effects where it isn't desired.
A consequence of #2 is that the GPU implementation of things like
blend-shaders have to jump through hoops to invoke their children with
opaque versions of their input colors.
This CL allows us to remove the explicit alpha modulation from a variety
of shaders, and should also cleanup of the blend shader logic. However,
to get the CL past chrome's layout tests, we need to guard the change.
So this version just makes the major change so paint conversion, in a
way that will work universally. Once chrome switches to the new method,
it will be safe to clean up the various shader implementations.
Bug: skia:11942
Change-Id: I518534becdd5d10cbd78cf3ff7d4a46dd1faabf9
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/464116
Reviewed-by: Brian Salomon <bsalomon@google.com>
Commit-Queue: Brian Osman <brianosman@google.com>
This header and corresponding implementation will contain common
definitions and subroutines for tessellation code.
Bug: skia:12524
Change-Id: Ib29b444177f284acb88a3d5644936674c48c0b89
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/463437
Commit-Queue: Chris Dalton <csmartdalton@google.com>
Reviewed-by: Greg Daniel <egdaniel@google.com>
Bug: skia:12466
Change-Id: I3f29f403a3ac91e81f54b0ec3bad19df8dd58739
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/462885
Commit-Queue: Jim Van Verth <jvanverth@google.com>
Reviewed-by: Michael Ludwig <michaelludwig@google.com>
Reviewed-by: Greg Daniel <egdaniel@google.com>
This also adds a base class BufferWriter which Vertex, Index, and
Uniform Writers inherit from
Bug: skia:12466
Change-Id: Icbac1210fbbd07321f9d88728ddde1e761fe4bb0
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/463496
Commit-Queue: Greg Daniel <egdaniel@google.com>
Reviewed-by: Chris Dalton <csmartdalton@google.com>
The original values assumed that `half` types would be sized the same as
a `float`. This is still the case (as skia:12339 is not yet fully
implemented) but this CL now contains the expected half values as well.
Change-Id: I94ff73446d29318f127600fb681c222a759936cc
Bug: skia:12339
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/463736
Auto-Submit: John Stiles <johnstiles@google.com>
Reviewed-by: Greg Daniel <egdaniel@google.com>
Commit-Queue: Greg Daniel <egdaniel@google.com>
warning: use of bitwise '|' with boolean operands [-Wbitwise-instead-of-logical]
inherit_if_needed(currentNode->fX , attrs->fX) |
^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
note: cast one or both operands to int to silence this warning
Change-Id: I6d59a5c33826af29560c0e30ddf9f4e16581882c
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/463896
Commit-Queue: John Stiles <johnstiles@google.com>
Commit-Queue: Florin Malita <fmalita@google.com>
Auto-Submit: John Stiles <johnstiles@google.com>
Reviewed-by: Florin Malita <fmalita@google.com>
- Drop the following dimensions:
reserved_for_xcode_version:*
gpu:none
- Loosen the cpu dimension.
- Restrict to cores:12 to avoid using the VMs.
The failing iPhone6 bots are failing on the main waterfall today
so they are WAI.
See also: https://bugs.chromium.org/p/chromium/issues/detail?id=1257420
Change-Id: I9f0eef7c6904e652a011937ab99c465e9d483172
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/463141
Reviewed-by: Eric Boren <borenet@google.com>
Commit-Queue: Joe Gregorio <jcgregorio@google.com>
This makes use of Bazel's pre-defined platforms
https://github.com/bazelbuild/platforms
and some of our own defined values (see
//bazel/common_config_settings/BUILD.bazel) to customize
the build rules.
I verified this by building bazel_test locally for
linux x64 as well as using the third_party deps for
a WASM build (using build files not seen in this CL).
Suggested Review Order:
- https://docs.bazel.build/versions/main/platforms.html if not
already familiar with Bazel Platforms
- third_party/BUILD.bazel to see that 1) all globs have
been removed and 2) select() targets various
platform constants or groups of constants to control
sources, headers, and local_defines.
- common_config_settings/ to see the groups of constraints
created, as well as new constraint_settings defined
(skdebug_impl)
- supported_combinations/ to see how we can define supported
sets of the constraint values (aka Bazel platforms).
I imagine expanding this more, so we might have platforms
named "linux_x64_emptyfontmgr_vulkan" or such.
- //BUILD.bazel and bazel_test.cpp to see use of SkDebugf.
- Everything else.
Change-Id: I49e4abdbcf7b76f0674efdbb1f53dc8823d110ee
Bug: skia:12541
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/463517
Reviewed-by: Ben Wagner <bungeman@google.com>
Reviewed-by: Leandro Lovisolo <lovisolo@google.com>
Owners-Override: Kevin Lubick <kjlubick@google.com>
On linux we need to support drm format modifiers on vulkan images. When
we have a modifier it can restrict the allowed usages/features of the
VkImage.
The current use case we have for images coming into skia that have
these modifiers are from video decoders. All these images are only used
as sampled images in draws and they have no filtering applied.
Therefore, for now, instead of tracking all format and modifier pairs
to know what is supported, we internally set these images to be external
which already restricts their use to basic in shader sampling.
Additionally in chrome, all these images are coming in ycbcr conversions
already which we treat as external.
Bug: skia:12336
Change-Id: I59a564f937f49a6d906efe954b24cebe5c7470ff
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/441298
Reviewed-by: Jim Van Verth <jvanverth@google.com>
Commit-Queue: Greg Daniel <egdaniel@google.com>
Bug: skia:12466
Change-Id: I1e7812e3600cacc8fe678e924d48c27fb8b88138
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/463716
Commit-Queue: Jim Van Verth <jvanverth@google.com>
Commit-Queue: Greg Daniel <egdaniel@google.com>
Auto-Submit: Jim Van Verth <jvanverth@google.com>
Reviewed-by: Greg Daniel <egdaniel@google.com>
Ideally we could move fPtr on VertexWriter to be private to force all
users of VertexWriter to go through the write functions. This CL at
least removes all users that were using fPtr for validity checking of
the VertexWriter.
I also move fPtr to private and added a temporary public getter. This
will allow us to make a base Writer class where fPtr is private and
we'll only expose the getter on the VertexWriter.
Change-Id: Ib389778d0f530fb31cef85460f00bd9f687c5219
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/463456
Commit-Queue: Greg Daniel <egdaniel@google.com>
Auto-Submit: Greg Daniel <egdaniel@google.com>
Reviewed-by: Chris Dalton <csmartdalton@google.com>
Gerrit didn't do so well on this CL. What's changed in the UniformManager is:
a lot of unused stuff was pared out of the UniformDataManager
the Write methods now return the number of bytes written in the call (incl. padding)
all the "#ifdef SK_DEBUG" stuff is new to validate against the prior way of writing uniforms
Bug: skia:12466
Change-Id: I172a74273c774257f751750aef37f4bd4c4e7e7b
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/463458
Reviewed-by: Greg Daniel <egdaniel@google.com>
Reviewed-by: Jim Van Verth <jvanverth@google.com>
Commit-Queue: Robert Phillips <robertphillips@google.com>
