We now have two functions `writeOpLoad` and `writeOpStore` which are
in charge of writing SpvOpLoad and SpvOpStore instructions.
`writeOpStore` also keeps track of pointer stores in a "store cache."
Subsequent loads from that same pointer will be found in the cache and
will return the value stored in that pointer instead.
Such a cache definitely cannot work in the face of control flow, so we
make the following concessions:
- `pruneReachableOps` is now `pruneConditionalOps`. Any pointers that
are altered inside a potentially-unreachable block are cleared from
the cache entirely.
- The entire store cache is cleared at all OpLabels within a loop.
The cache also cannot work in the presence of swizzled stores, so we
make another significant concession:
- The entire store cache is cleared whenever we store into a non-memory
pointer (e.g., assigning into a swizzled LValue, such as `foo.xz`).
Despite these significant limitations, this manages to dramatically
shrink many real-world examples.
Change-Id: I0981a0cf7b45b064e153e9ada271494c8e00cad5
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/530054
Reviewed-by: Brian Osman <brianosman@google.com>
Commit-Queue: John Stiles <johnstiles@google.com>
Auto-Submit: John Stiles <johnstiles@google.com>
Previously, we only handled the simple case of extracting from an
OpConstantComposite. Now we also handle the complex case of extracting
from an OpCompositeConstruct, where vectors can be composed of other
vectors.
This is particularly challenging because OpCompositeConstruct can
contain SpvIds from almost any other instruction, so we need to be
able to decode those instructions and figure out their type. For
instance:
%5 = OpFAdd Vec2 %1 %2
%6 = OpFAdd Scalar %3 %4
%7 = OpCompositeConstruct FloatVec3 %5 %6
%8 = OpCompositeExtract %7 2
The %8 (OpCompositeExtract) could be replaced with %6 but we need to
peek at the type in *both* OpFAdd instructions to decode this. It
only works when the affected instructions are in-cache, so many
opportunities are currently not optimized because their code still
uses the original, uncached form of writeInstruction.
Change-Id: I5719ae6284f32e1d6f2c898eca282c22b94fc764
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/529743
Reviewed-by: Brian Osman <brianosman@google.com>
Auto-Submit: John Stiles <johnstiles@google.com>
Commit-Queue: John Stiles <johnstiles@google.com>
The output changes here are almost entirely a wash, because we already
had support for caching scalars and vectors. Almost all changes are just
inconsequential reorderings of IDs, and the removal of RelaxedPrecision
decorators on constants (which were not meaningful).
Change-Id: I45340c4a240cb504b7c4a934b3db178d2f39ec99
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/528709
Reviewed-by: Arman Uguray <armansito@google.com>
Commit-Queue: John Stiles <johnstiles@google.com>
The additional tests from http://review.skia.org/441238 uncovered a gap
in the constant folder's abilities; it was not able to fold away
boolean vector comparisons even when they were constant. These are ES2
constant-expressions, so folding them properly is a requirement.
Change-Id: Ia0b4d5d1215c5fc2b247ac3f0dec4c8747d2153e
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/441579
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>
This was another place where we needed to use
`getConstantSubexpression` to rebuild vectors/matrices; it is a more
robust approach than trying handle each ctor type individually. The
fuzzer found an edge case with double-casting matrices to vectors that
fell through the cracks with the original approach.
In adding additional tests, I also found a case that the constant-folder
seems to ignore, `bool4(x,x,x,x) == bool4(x)`. This does fold for ints
and floats, so this ought to be fixable in a followup, but it's not a
big deal either way; this is very unlikely to occer in real code.
Change-Id: I4d577c87ef7049306685ca95250ecdf93b1dbc06
Bug: oss-fuzz:37464
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/441238
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>
The tests now check bool4-mat2 conversions, which fortunately do work,
and the vector-to-matrix tests include int and bool conversions as well.
Change-Id: I971271838a93081b9258deb7c1d13b7732fb2440
Bug: skia:12067
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/426757
Commit-Queue: John Stiles <johnstiles@google.com>
Auto-Submit: John Stiles <johnstiles@google.com>
Reviewed-by: Brian Osman <brianosman@google.com>
Reviewed-by: Ethan Nicholas <ethannicholas@google.com>
The fuzzer quickly discovered that the newly introduced mat2-to-vec4
conversion code did not account for integer vectors. We now handle
`ivec4(mat2)` casts properly. This required some non-trivial
restructuring of the logic, but in the vast majority of cases, the types
will match and the end result will be identical.
Change-Id: If07c2fe4b4345bd767384b1802374910f65cd3f0
Bug: oss-fuzz:35998
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/426756
Auto-Submit: John Stiles <johnstiles@google.com>
Reviewed-by: Brian Osman <brianosman@google.com>
Commit-Queue: John Stiles <johnstiles@google.com>
GLSL supports casting vec4 into mat2 and vice versa, so SkSL should have
equivalent support. This CL allows the Compound constructor to take a
matrix as input, and fixes up backends to do the right thing when a
matrix shows up in the compound-constructor path.
Change-Id: I13289ad0a27ba59bddc3706093820594efebc693
Bug: skia:12067
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/426003
Auto-Submit: John Stiles <johnstiles@google.com>
Reviewed-by: Brian Osman <brianosman@google.com>
Commit-Queue: John Stiles <johnstiles@google.com>
This reverts commit a89781215a.
Reason for revert: breakage on Windows
Original change's description:
> Add tests for matrix-vector conversions.
>
> GLSL supports casting vec4 into mat2 and vice versa, so SkSL should have
> equivalent support. Adding tests as a starting point.
>
> Change-Id: If8bcbf99afcec94d948d5da9e6205cb4a232af18
> Bug: skia:12067
> Reviewed-on: https://skia-review.googlesource.com/c/skia/+/425837
> Auto-Submit: John Stiles <johnstiles@google.com>
> Commit-Queue: Brian Osman <brianosman@google.com>
> Reviewed-by: Brian Osman <brianosman@google.com>
TBR=brianosman@google.com,ethannicholas@google.com,johnstiles@google.com
Change-Id: I2563041f538b1b20074385f1b61af5fc506ffad5
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Bug: skia:12067
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/426057
Reviewed-by: John Stiles <johnstiles@google.com>
Commit-Queue: John Stiles <johnstiles@google.com>
GLSL supports casting vec4 into mat2 and vice versa, so SkSL should have
equivalent support. Adding tests as a starting point.
Change-Id: If8bcbf99afcec94d948d5da9e6205cb4a232af18
Bug: skia:12067
Reviewed-on: https://skia-review.googlesource.com/c/skia/+/425837
Auto-Submit: John Stiles <johnstiles@google.com>
Commit-Queue: Brian Osman <brianosman@google.com>
Reviewed-by: Brian Osman <brianosman@google.com>
