Moved the MatrixFoldingES2.sksl test case for matrix construction with
side-effects into a new PreserveSideEffects.sksl test and added new test
cases for various vector and matrix types and constructors. The new test
is written such that none of its contents should be folded away.
Note: This test does not pass on NVIDIA GPUs when using OpenGL as
discussed in skia:13035. Notably, NONE of the increments are executed on
those GPUs as ALL increment expressions seemingly get subjected to
constant-folding. The test is disabled on NVIDIA GPU bots.
This also means that the remaining MatrixFoldingES2.sksl tests now work
on NVIDIA GPUs when using OpenGL (with the exception of Tegra3 + OpenGL
ES).
Bug: skia:13035, skia:11919
Change-Id: I561bb62fe2b6b814ba80fbc492d3885bbcd6b65b
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/518278
Reviewed-by: John Stiles <johnstiles@google.com>
Commit-Queue: Arman Uguray <armansito@google.com>
This lets us differentiate SkQP from other testing harnesses (like DM or
Viewer) at runtime.
This allows us to change strictness or deactivate tests when SkQP is
running. Previously we had a macro SK_BUILD_FOR_SKQP for this, but this
did not work on a local skqp binary; it only activated when compiling
for Android.
Change-Id: I7334e04ea1eddda783a5d2f26699edd20828f81a
Bug: skia:13037
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/518939
Reviewed-by: Derek Sollenberger <djsollen@google.com>
Reviewed-by: Brian Osman <brianosman@google.com>
Commit-Queue: John Stiles <johnstiles@google.com>
The Android builds of SkQP will conditionally exclude any tests that are
not flagged with `SkQP`.
Note that local builds of the C++ skqp binary do not actually set
SK_BUILD_FOR_SKQP. This flag is set by gn_to_bp. So local skqp binaries
will continue to run these tests. A possible fix for this limitation
has been written in the followup CL: http://review.skia.org/518939
Change-Id: I879650d165c5931693b14102586bae2c9e45dd43
Bug: skia:13037
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/518936
Auto-Submit: John Stiles <johnstiles@google.com>
Reviewed-by: Derek Sollenberger <djsollen@google.com>
Commit-Queue: Derek Sollenberger <djsollen@google.com>
At this CL, we should run exactly the same tests as before; the only
difference is that we enable and disable tests via SkSLTestFlags instead
of using different SKSL_TEST_xxxxxx macros for different test types.
There is an inert flag for SkQP; this will be implemented in a followup
CL.
The SKSL_TEST block now extends past 100 characters in some spots
because it was, overall, a lot easier to read and edit tabular data in
a tabular format than with inconsistent formatting and line wrapping.
Change-Id: Ic90f88b10c6e59357275b471e6e86e53e611f101
Bug: skia:13037
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/518138
Auto-Submit: John Stiles <johnstiles@google.com>
Reviewed-by: Arman Uguray <armansito@google.com>
Commit-Queue: John Stiles <johnstiles@google.com>
Change-Id: I1fa160941b18392616708f4e08a4ebbf7398480e
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/518521
Reviewed-by: John Stiles <johnstiles@google.com>
Reviewed-by: Brian Osman <brianosman@google.com>
Commit-Queue: Ethan Nicholas <ethannicholas@google.com>
4 byte alignment was being used, when it should be 256 on Intel Macs.
Bug: skia:13033
Change-Id: If902885f62e019b10c316dcfd7f4a336e0982db7
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/518522
Reviewed-by: Robert Phillips <robertphillips@google.com>
Commit-Queue: Jim Van Verth <jvanverth@google.com>
Reviewed-by: John Stiles <johnstiles@google.com>
ExtractPaintData currently returns:
std::tuple<SkUniquePaintParamsID, std::unique_ptr<SkPipelineData>>
SkShaderCodeDictionary::findOrCreate is going to change in an upcoming CL and this sets up to better isolate those changes.
Bug: skia:12701
Change-Id: I8d077b3d342e6ea16ac99227eb145d92e8247ea4
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/517736
Reviewed-by: Jim Van Verth <jvanverth@google.com>
Commit-Queue: Robert Phillips <robertphillips@google.com>
PS1 regenerates BUILD.bazel files
I suggest reviewing the deltas between PS1 and the latest
PS to focus on the interesting bits.
The changes here allow for a Vulkan-only build of HelloWorld
based on sk_app. The toughest change was properly fetching
the VisualID after removing the gl calls that used to
fill that in.
There are a few changes that fix resolution of Dawn
header files, but those won't actually be built until
a follow-on CL.
Change-Id: I54fb58b5dd7ecd4313562aed401759b3eaed53c0
Bug: skia:12541
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/516999
Reviewed-by: Greg Daniel <egdaniel@google.com>
Commit-Queue: Kevin Lubick <kjlubick@google.com>
This was motivated by SkScalerContextFlags putting `kNone` into the
global namespace. Instead of adding a suffix, change these internal
enums to be scoped and update the users.
Because SkScalerContextFlags is used like a bitfield,
`SK_MAKE_BITFIELD_OPS` is used to provide the appropriate operators. The
macro implementation is updated to handle scoped enums, remove the very
dangerous, unwanted, and unused templated operators, and add operator~
since that is needed by one existing user.
Also, for some reason this allows the compiler to find an unused field
which is removed.
Change-Id: Id31c80b381bffaf9b623f6a8ec7d86751d3d99f0
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/517616
Reviewed-by: Brian Osman <brianosman@google.com>
Commit-Queue: Ben Wagner <bungeman@google.com>
SkPipelineData is evolving into a big deal - it deserves it owns files.
Bug: skia:12701
Change-Id: I78b4100f1b90fa10f2b264e6d13fca6f15bba39b
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/516157
Reviewed-by: Jim Van Verth <jvanverth@google.com>
Commit-Queue: Robert Phillips <robertphillips@google.com>
Additionally make the enums enum-classes.
We will need this to specify fixed function blend modes in Graphite.
This CL is mainly mechanical (i.e., global replace) but some fix-ups were required for our internal bit twiddling with the enum-classes.
Bug: skia:12701
Change-Id: I67abcf8274d14e691974cf253bf65dcb4697bd9a
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/515318
Reviewed-by: Jim Van Verth <jvanverth@google.com>
Commit-Queue: Robert Phillips <robertphillips@google.com>
DEF_CONDITIONAL_TEST and DEF_CONDITIONAL_GPUTEST_FOR_xxxx_CONTEXTS will
allow unit tests to be considered or ignored based on a passed-in
boolean condition. DEF_TEST_DISABLED is now implemented as a conditional
test hard-coded with a "false" condition.
This allows macros (such as SKSL_TEST) to add unit tests based on the
result of a comparison.
Change-Id: Ic5325569009cfce768ba8942907c1d3e7d69ca1c
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/517516
Auto-Submit: John Stiles <johnstiles@google.com>
Reviewed-by: Brian Osman <brianosman@google.com>
Commit-Queue: John Stiles <johnstiles@google.com>
ES2 disallows opaque types in expressions (other than passing them to
their associated builtin functions). We now enforce a similar
restriction on SkSL opaque types.
While I was there, I added several other cases to the invalid-shader
test to make sure that they were all caught.
I needed to reorder some code to make sure that ternary expression error
messages didn't change. Ternary expressions now check for opaque types
before checking that the left-side type and right-side type are
compatible. This is because we check for "compatible" ternary
expressions by checking if `leftSide == rightSide` would be accepted.
`shader1 == shader2` used to be considered a valid expression for the
purposes of this test, but not anymore.
Change-Id: I62a0a31feca9dadd428da7d1b48d7693c4b6434d
Bug: skia:13026
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/516802
Commit-Queue: John Stiles <johnstiles@google.com>
Auto-Submit: John Stiles <johnstiles@google.com>
Reviewed-by: Brian Osman <brianosman@google.com>
Commit-Queue: Brian Osman <brianosman@google.com>
The fuzzer discovered that we allow == on void types (confusing the SkVM
backend).
Change-Id: Ia9494642faf67f3f86e3a365807be8bd4a7062e4
Bug: skia:13026
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/516796
Auto-Submit: John Stiles <johnstiles@google.com>
Reviewed-by: Brian Osman <brianosman@google.com>
Commit-Queue: John Stiles <johnstiles@google.com>
Previously, we would take the vector-folding path for all types. This
didn't cause any problem for scalars, but failed for "zero-size" types
like void. It isn't valid to compare zero-size values, but we currently
don't reject such code (see skia:13026), and the fuzzer noticed this.
It's safest to only run the vector-folding code when we actually have
multiple slots that need to be folded into one result.
Change-Id: I0bc88043d9a4aeea38ae24dc1a6d1a7430d3d7b0
Bug: oss-fuzz:45279, skia:13026
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/516676
Auto-Submit: John Stiles <johnstiles@google.com>
Reviewed-by: Brian Osman <brianosman@google.com>
Commit-Queue: Brian Osman <brianosman@google.com>
Bug: skia:12754
Change-Id: Iec046dedb088f226e5928712222fa786426b2689
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/515517
Reviewed-by: Michael Ludwig <michaelludwig@google.com>
Commit-Queue: Greg Daniel <egdaniel@google.com>
This will allow us to easily replace unordered_set/unordered_map in more
places.
Change-Id: Ic0b5245e8cdce2918d7b043bfc370eca95caf889
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/515319
Auto-Submit: John Stiles <johnstiles@google.com>
Reviewed-by: Ben Wagner <bungeman@google.com>
Commit-Queue: John Stiles <johnstiles@google.com>
The slot-assignment logic has been changed to associate slots with
function calls, instead of function definitions. In our test case, you
can now see that the calls to `get` are now mapped to $15, $17 and $18.
This change also jiggles some existing tests and improves their
register allocation slightly (!).
One minor hitch here is that there's no FunctionCall node associated
with main() (it's never explicitly called). However, our slot map key
can be any IRNode, and we know main() can't be called by anyone else,
so it's harmless to use the function definition as the key in this case.
(This entry could probably stay out the map entirely if it made a
difference, but I don't think it matters.)
Change-Id: I68a6ff24cbd3a2db77f24126502bd3d11e8c0962
Bug: skia:13011
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/514578
Reviewed-by: Brian Osman <brianosman@google.com>
Commit-Queue: John Stiles <johnstiles@google.com>
Auto-Submit: John Stiles <johnstiles@google.com>
If a function is called multiple times on one line, stepping over that
line does not show all of the function-call results. It only shows the
last result.
e.g. in this example, I have just stepped over the first line which
calls "get" three times. We should see three results, but we only see
one: http://screen/3WfJoZWm77cSexM
In this test you can see that all three calls to `get` are assigned to
the same slot, $15.
Change-Id: Id0c486ef349a1e527001efbcee2ed2b836f56e83
Bug: skia:13011
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/514577
Auto-Submit: John Stiles <johnstiles@google.com>
Reviewed-by: Brian Osman <brianosman@google.com>
Navigating away from the Debugger in Viewer could lead to a null
dereference.
Change-Id: I2686edd3716e7a375d394a4dc1a296395b5f35b8
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/514579
Reviewed-by: Arman Uguray <armansito@google.com>
Auto-Submit: John Stiles <johnstiles@google.com>
Commit-Queue: John Stiles <johnstiles@google.com>
This leverages our new support for standard strings in SkGoodHash, added
in prior CL http://review.skia.org/515356.
These tests already included checks for SkString, so it was easy to add
a template which tested std::string and std::string_view with the same
checks.
Change-Id: Ia2d9b889e66a50556a25e40493d14a5b692214a8
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/515358
Auto-Submit: John Stiles <johnstiles@google.com>
Reviewed-by: Ben Wagner <bungeman@google.com>
Commit-Queue: John Stiles <johnstiles@google.com>
This will allow these types to be used as the key in an SkTHashSet/Map
without providing a custom Traits class.
Change-Id: I41bfc1d43f3909d3653b4f12ac39e2a288235052
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/515356
Auto-Submit: John Stiles <johnstiles@google.com>
Reviewed-by: Ben Wagner <bungeman@google.com>
I added a comment and didn't rebuild; this shifted line numbers around,
which is reflected in the debug trace opcodes.
Change-Id: I1b8e00ff65557a03483e8d7ff0c4dbec9852866f
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/514777
Commit-Queue: John Stiles <johnstiles@google.com>
Auto-Submit: John Stiles <johnstiles@google.com>
Reviewed-by: Kevin Lubick <kjlubick@google.com>
Commit-Queue: Kevin Lubick <kjlubick@google.com>
The `break_loop` test causes LLVM to get confused and crash when
compiled on some GPUs. The crash goes away if we pass a literal 5
instead of a 5 that is computed at runtime. This also results in a
simpler test for SkVM, for better or worse, but we still have
coverage for dynamic loop exits in other tests.
LLVM crash: https://paste.googleplex.com/4718583155261440
Dangerous shader: https://paste.googleplex.com/4776089520963584
Change-Id: Ic6cbd55a36d2de58e5dd3459d4dfd74acdbc9f91
Bug: skia:13005
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/514538
Reviewed-by: Arman Uguray <armansito@google.com>
Commit-Queue: John Stiles <johnstiles@google.com>
Auto-Submit: John Stiles <johnstiles@google.com>
MSL does not support the unary "-" operator on matrix types. Similarly
the SPIR-V OpFNegate/OpSNegate operations only work on scalar and vector
type.
* An expression such as "-<mat>" is now transformed to "-1.0 * <mat>" when
generating MSL.
* The same expression now generates a component-wise negation in SPIR-V,
matching what glslang outputs for GLSL.
* A unary "+" is now treated as NOP for MSL, matching the SPIR-V backend.
An expression such as "+<expr>" is now evaluated as "<expr>".
* The shared/Negation.sksl has been moved to folding/ as much of its
contents exercise constant-folding of comparison expressions.
* The shared/UnaryPositiveNegative.sksl test has been extended to
exercise scalar and matrix types.
NOTE: The SPIR-V backend changes have caused a minor re-ordering of SSA
IDs generated when writing out a prefix-expression. The affected gold
files have been updated.
Bug: skia:12627, skia:12992
Change-Id: Iec5cdafc591aed7e49b3b52bda42a02661380bab
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/513976
Auto-Submit: Arman Uguray <armansito@google.com>
Reviewed-by: John Stiles <johnstiles@google.com>
Commit-Queue: Arman Uguray <armansito@google.com>
We previously had no way to signal a parse error from arraySize,
resulting in a cascade of additional errors downstream. This tightens
up the behavior and allows us to fail more gracefully.
Bug: skia:12416
Change-Id: I83d3d5bc1dc63395edb325297375a6eb52415817
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/512952
Reviewed-by: John Stiles <johnstiles@google.com>
Commit-Queue: Ethan Nicholas <ethannicholas@google.com>
dFdy wasn't working due to the DSL never having been started.
Bug: skia:12985
Change-Id: Ic58018e3dec3e9df15c2e784f8732ee99e793353
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/513938
Reviewed-by: John Stiles <johnstiles@google.com>
Commit-Queue: Ethan Nicholas <ethannicholas@google.com>
These tests were not being run for ES3-only GPU and non-strict-ES2 CPU
tests. CLONE and REHYDRATE variants are now generated for all test
declarations.
Change-Id: I8ce14bd50f175d0bae66d9593b68a42fab701f93
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/513056
Reviewed-by: John Stiles <johnstiles@google.com>
Commit-Queue: Arman Uguray <armansito@google.com>
On modern hardware, this will give the correct result for `NaN != x`
(true).
Change-Id: I9683f74756da5da5f34ccacec02c1f2449791f26
Bug: skia:12977
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/513317
Commit-Queue: John Stiles <johnstiles@google.com>
Auto-Submit: John Stiles <johnstiles@google.com>
Reviewed-by: Brian Osman <brianosman@google.com>
Reviewed-by: Arman Uguray <armansito@google.com>
Commit-Queue: Arman Uguray <armansito@google.com>
The SKSL_TEST_CPU_ES3 has been misinterpreted to mean that the tests
should run on "SkVM and OpenGL" while it was originally meant for tests
that were explicitly disabled on GPU backends.
* This macro is now called "SKSL_TEST_CPU_NO_STRICT_ES2" and the comments
reflect the fact that the tests are meant for CPU-only.
* Removed the infra bot exceptions for the matrix constructor tests as
they are now restricted to run only on CPU.
Bug: skia:12970
Change-Id: I46dcec51ef6998f6a8a7b4610c39560da1e59057
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/512578
Reviewed-by: John Stiles <johnstiles@google.com>
Commit-Queue: Arman Uguray <armansito@google.com>
PS1 is the automatic regenerated changes.
PS2-3 adds an #ifdef guard to the addToKey method on shaders.
The SkShaderCodeDictionary class helps generate SkSL and is
only necessary when we are building with SkSL (gpu builds and
cpu builds with SkVM).
Suggested Review order:
- Use Gerrit to diff PS 1 and the last PS
- src/core/BUILD.bazel adds some sources to the "only
necessary if sksl is enabled" bucket
- All the .cpp and .h files to see the #ifdef is added
correctly.
Change-Id: I4d4ce61a4957ef1e0840204acff08ce7e616f9cb
Bug: skia:12541
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/512157
Reviewed-by: Michael Ludwig <michaelludwig@google.com>
Commit-Queue: Kevin Lubick <kjlubick@google.com>
This should reduce any possibility of keys being built w/ mixed backend snippets. The decision of which backend the key is being created for is made once, when the builder is created.
Bug: skia:12701
Change-Id: I070330627450501eb93ed98271659e1441aaa935
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/511804
Reviewed-by: Jim Van Verth <jvanverth@google.com>
Commit-Queue: Robert Phillips <robertphillips@google.com>
Right now this just checks the creation of the key in debug. In follow up CLs we will pass the data payload to the glue code.
The additional key structure also obviates the need for the block-specific Dump methods.
Bug: skia:12701
Change-Id: Ieddea596de92a4f6b3cf6003f33d7f01d3d28c20
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/508678
Reviewed-by: Michael Ludwig <michaelludwig@google.com>
Commit-Queue: Robert Phillips <robertphillips@google.com>
Switch to using the public API similar to SkTextBlob for
serialize/Deserialize. Add deserializeSlug to SkStrikeClient.
Bug: chromium:1278340
Change-Id: I91b93487859c662e3bfdfba49ba4758f13529cd9
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/511601
Reviewed-by: Robert Phillips <robertphillips@google.com>
Commit-Queue: Herb Derby <herb@google.com>
Use the SkStrikeClient to do the TypefaceID translation if present
when creating a GrGlyphVector.
With this change, Chromium should be able to use the
flatten/MakeFromBuffer API. Currently, this will only
serialize DirectMask (the most common type) drawing.
Other SubRun type will follow soon.
Bug: chromium:1278340
Change-Id: I08a46c7e4e13f7bd899abfdad89c5b3db2548d6a
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/511416
Reviewed-by: Robert Phillips <robertphillips@google.com>
Commit-Queue: Herb Derby <herb@google.com>
This failure happens when using the SkRemoteGlyphCache system
*and* serializing slugs. The problem is that the origin from the
blob conversion was applied twice causing all the sub-pixel
positions to be off. This often pass because this position often
is {0, 0}.
Change-Id: I0384cfc1de5c32851381390661d1f4a8a6595ab2
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/510724
Reviewed-by: Robert Phillips <robertphillips@google.com>
Commit-Queue: Herb Derby <herb@google.com>
This updates the key builder system so the builder can be used in a tight loop with the memory backing the key being reused.
Bug: skia:12701
Change-Id: I79a72ca26570dcfea9aa45b0fbad8e598688ee98
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/510016
Reviewed-by: Michael Ludwig <michaelludwig@google.com>
Commit-Queue: Robert Phillips <robertphillips@google.com>
This will leave the PaintParamsKeys more clearly immutable and manage more of the cruft of key creation (e.g., beginBlock/endBlock pairing).
Bug: skia:12701
Change-Id: I1944397465f845e974e950f47cedb90c3f1895b1
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/506881
Reviewed-by: Michael Ludwig <michaelludwig@google.com>
Commit-Queue: Robert Phillips <robertphillips@google.com>
This reverts commit f436cf2343.
Reason for revert: May need to be behind flag or more
suppressions. Breaking linux-rel vulkan_swiftshader_blink_web_tests
css3/filters/effect-blur-hw.html .
Original change's description:
> Preserve base device origin on saveLayer and image filters
>
> SaveLayerOriginTest taken from https://skia-review.googlesource.com/c/skia/+/277977
>
> Bug: skia:12732
> Change-Id: I5ce75355bb16237043c229e1cbc7a106eb636d18
> Reviewed-on: https://skia-review.googlesource.com/c/skia/+/508919
> Reviewed-by: Greg Daniel <egdaniel@google.com>
> Commit-Queue: Michael Ludwig <michaelludwig@google.com>
Bug: skia:12732
Change-Id: I74cc8dc279d22c4fbd313ae3caeb4d0748daf003
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/510196
Auto-Submit: Ben Wagner <bungeman@google.com>
Commit-Queue: Rubber Stamper <rubber-stamper@appspot.gserviceaccount.com>
Bot-Commit: Rubber Stamper <rubber-stamper@appspot.gserviceaccount.com>
