This currently requires that we are not building with Chromium.
Bug: tint:1911
Change-Id: I6c37afe6c2b8250b30d244e8cc2274253b6ce8f6
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/129560
Commit-Queue: James Price <jrprice@google.com>
Kokoro: Kokoro <noreply+kokoro@google.com>
Reviewed-by: Dan Sinclair <dsinclair@chromium.org>
Allow the resolver to understand builtin structures, like
__frexp_result_f16. This allows backend transforms to declare the types,
even if they're "untypable" by the user.
Bug: chromium:1430309
Change-Id: I392709118182a058f737ccf1b7b46fc6b0b7264d
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/129482
Kokoro: Ben Clayton <bclayton@google.com>
Reviewed-by: Dan Sinclair <dsinclair@chromium.org>
Commit-Queue: Ben Clayton <bclayton@google.com>
Contains the builtin structure creation logic moved out from
intrinsic_table.cc.
Bug: chromium:1430309
Change-Id: I2207f9ae42c6d7343c2f2ffa81effde59e6023fe
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/129481
Reviewed-by: Dan Sinclair <dsinclair@chromium.org>
Commit-Queue: Ben Clayton <bclayton@chromium.org>
Kokoro: Kokoro <noreply+kokoro@google.com>
This Cl renames the `instr` variables to `inst`.
Bug: tint:1718
Change-Id: Icf3b8c2f612c8dfe4b469d90327fef90ad813a0d
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/129460
Reviewed-by: Ben Clayton <bclayton@google.com>
Kokoro: Kokoro <noreply+kokoro@google.com>
Commit-Queue: Dan Sinclair <dsinclair@chromium.org>
This CL adds Assignment and CompoundAssignment statements to the IR
builder.
Bug: tint:1718
Change-Id: I3037da0115c7f4fe68941565b7e48866d421bbbf
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/129201
Reviewed-by: Ben Clayton <bclayton@google.com>
Kokoro: Kokoro <noreply+kokoro@google.com>
Commit-Queue: Dan Sinclair <dsinclair@chromium.org>
The const variables should all have been const-eval'd and turned into
`constants` at the usage sites. This CL updates the IR to skip constants
when emitting.
A TODO is added to validate they aren't used when identifier expressions
are supported.
Bug: tint:1718
Change-Id: I77328a0cbd3d7f6692d1d1057d6953fcf762cfd7
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/129240
Kokoro: Kokoro <noreply+kokoro@google.com>
Reviewed-by: Ben Clayton <bclayton@google.com>
Commit-Queue: Dan Sinclair <dsinclair@chromium.org>
This CL adds support for UnaryOpExpressions and converts them into Unary
instructions in the IR.
Bug: tint:1718
Change-Id: I736e29cec5e722b7c7f1b0f4f22ce55a3d3e4109
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/129221
Kokoro: Kokoro <noreply+kokoro@google.com>
Commit-Queue: Dan Sinclair <dsinclair@chromium.org>
Reviewed-by: Ben Clayton <bclayton@google.com>
This CL updates the IR builder implementation to remove the `bool`
return values and use the diagnostics as the source of truth for if the
conversion works.
This requires disabling a couple tests as they depend on identifiers
which aren't implemented. Previously the `worked` because it would just
return an ID value that would be emitted, but now they end up being an
error.
Bug: tint:1718
Change-Id: I00bc8845393cf4fae7b3eb0f5cfffb8c5fc1dec0
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/129220
Kokoro: Kokoro <noreply+kokoro@google.com>
Commit-Queue: Dan Sinclair <dsinclair@chromium.org>
Reviewed-by: Ben Clayton <bclayton@google.com>
This CL adds support for discard into the IR. The `discard` statement is
handled as an instruction in the current block. The `discard` is a
`demote_to_helper` in WGSL so control flow has to continue after the
discard, it just predicates writes. So, an instruction seems like the
most logical way to express.
Bug: tint:1718
Change-Id: I0d2fb029631523d72a7811d0be0715732427c302
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/129200
Reviewed-by: Ben Clayton <bclayton@google.com>
Commit-Queue: Dan Sinclair <dsinclair@chromium.org>
Kokoro: Kokoro <noreply+kokoro@google.com>
When an OpConstantComposite result is used by multiple instructions,
declare it as a module-scope `const` instead of inlining the constant
at each use site. This fixes an issue whereby the spirv-reader was
massively inflating the size of the WGSL it produces, which was caught
via an OOM fuzzer bug.
Bug: oss-fuzz:57795
Change-Id: Iac8c6a2147a7e2ebfddbaacae9fcb1dbe0b59e9d
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/128881
Auto-Submit: James Price <jrprice@google.com>
Kokoro: Kokoro <noreply+kokoro@google.com>
Reviewed-by: David Neto <dneto@google.com>
Commit-Queue: David Neto <dneto@google.com>
This prevents us from crashing when a module-scope private variable
has an inferred type that is a built-in struct. We instead fail to
resolve the generated code.
Bug: chromium:1430309
Change-Id: I71c6eacf28096380227ed1043821ecc3d33ffb40
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/129140
Reviewed-by: Ben Clayton <bclayton@google.com>
Kokoro: Kokoro <noreply+kokoro@google.com>
Commit-Queue: James Price <jrprice@google.com>
The declaration may be nullptr, so doing this avoids a potential
nullptr deref when the struct is an internal type (e.g. frexp
result). In these cases we will now fail to resolve the resulting
code, which is a little better than just crashing.
A future change will update the resolver to allow us to use the
internal struct names from our transforms.
Bug: chromium:1430309
Change-Id: Ic72b5105bf3159c448c20dd9228d73b25a632b69
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/129120
Kokoro: Kokoro <noreply+kokoro@google.com>
Commit-Queue: James Price <jrprice@google.com>
Reviewed-by: Ben Clayton <bclayton@google.com>
This CL renames ir::Temp to ir::Runtime to better reflect that this is a
runtime generated ID.
Bug: tint:1718
Change-Id: I24edbe0e2e19303f50355117ec7cd4b5b8f45aef
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/129100
Reviewed-by: Ben Clayton <bclayton@google.com>
Kokoro: Kokoro <noreply+kokoro@google.com>
Commit-Queue: Dan Sinclair <dsinclair@chromium.org>
Produce a meaningful error instead of just crashing.
Bug: oss-fuzz:55170
Change-Id: I09d94a910835839ce9407849446cf2928231a114
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/128540
Kokoro: Kokoro <noreply+kokoro@google.com>
Commit-Queue: James Price <jrprice@google.com>
Reviewed-by: David Neto <dneto@google.com>
When `invariant` is enabled on MSL was was incorrectly setting
`@invariant` instead of `[[invariant]]`. We test with metal1.2 which
does not have invariant, so this only showed up when using metal2.1 or
higher.
Bug: chromium:1439273
Change-Id: Iab866608195e697b0370d465f350b25277d904a3
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/128880
Commit-Queue: Ben Clayton <bclayton@google.com>
Kokoro: Kokoro <noreply+kokoro@google.com>
Reviewed-by: Ben Clayton <bclayton@google.com>
This Cl updates the lexer to only create tokens when needed, in the case
of no match an empty optional is returned.
Change-Id: Ie4ba45d72cfb263beb8d9c83e3098ff11beeecd2
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/128620
Kokoro: Kokoro <noreply+kokoro@google.com>
Reviewed-by: Ben Clayton <bclayton@google.com>
Commit-Queue: Dan Sinclair <dsinclair@chromium.org>
This CL adds a simple program which can be set to loop over specific
parts of the tint pipeline to generate profiling data.
Change-Id: I6375940619b7ef2f7e66540d4f740e6e0b9b3132
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/128541
Reviewed-by: Ben Clayton <bclayton@google.com>
Commit-Queue: Dan Sinclair <dsinclair@chromium.org>
Kokoro: Kokoro <noreply+kokoro@google.com>
When attempting to member-access a non-value expression.
GetVal() ensures the expression resolves to a value expression, and errors accordingly.
Bug: chromium:1436467
Change-Id: I77ebb44f836be3b99db4b5c26ff41db2ee3fe30a
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/128840
Auto-Submit: Ben Clayton <bclayton@google.com>
Reviewed-by: Dan Sinclair <dsinclair@chromium.org>
Kokoro: Kokoro <noreply+kokoro@google.com>
Commit-Queue: Ben Clayton <bclayton@google.com>
In the case of ASCII characters, which a lot of WGSL source is, the
decoder can early out after a checking the value.
Change-Id: Iff655565dde23b143fddb95c6c353a917e25e916
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/128120
Reviewed-by: Ben Clayton <bclayton@google.com>
Commit-Queue: Ben Clayton <bclayton@google.com>
Kokoro: Kokoro <noreply+kokoro@google.com>
This CL extends the Symbol class to store if the content of the symbol could parse as a builtin.
Change-Id: I7e14ad944c1c9c43d900f9ccf8be6539ac9ea667
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/127460
Kokoro: Kokoro <noreply+kokoro@google.com>
Commit-Queue: Dan Sinclair <dsinclair@chromium.org>
Reviewed-by: Ben Clayton <bclayton@google.com>
This CL splits the base_src set in the GN build into a symbols_src and a
utils_src.
Change-Id: I39036a4ba5163aa816233b42d9a2ebfe34ddf038
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/128361
Reviewed-by: Ben Clayton <bclayton@google.com>
Commit-Queue: Ben Clayton <bclayton@google.com>
Kokoro: Kokoro <noreply+kokoro@google.com>
Instead of a std::string. This avoids unnecessary string allocations, and follows the pattern of all other AST nodes.
Change-Id: I3faf534090a2033d671b2ef463d8b9ed3e47eecd
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/128300
Reviewed-by: Dan Sinclair <dsinclair@chromium.org>
Kokoro: Kokoro <noreply+kokoro@google.com>
Commit-Queue: Ben Clayton <bclayton@chromium.org>
Currently the `is_reserved` method in the WGSL parser uses the
`operator==` on the token to check the reserved words. This ends up
re-doing a bunch of work as it checks the token type, then gets the
`string_view` or `string` from the variant, then compares.
This CL adds a precondition that the token is an identifier (which is
true in the only case we call this method) and then extracts the string
from the token once and uses that for all the comparisons.
Change-Id: If425afea38e727169de7991a84f6fa1e47f660ef
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/128140
Reviewed-by: Ben Clayton <bclayton@google.com>
Commit-Queue: Dan Sinclair <dsinclair@chromium.org>
Kokoro: Kokoro <noreply+kokoro@google.com>
This CL consolidates the unicode code into utils. These are utility
libraries, so the extra namespace doesn't add much.
Change-Id: Id0de612b6be036392a3cb018bfe66733f2f1ebcb
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/127403
Kokoro: Kokoro <noreply+kokoro@google.com>
Commit-Queue: Ben Clayton <bclayton@google.com>
Reviewed-by: Ben Clayton <bclayton@google.com>
The Castable class is used by utils/. This Cl moves the implementation
into the utils/ folder. The `Is` and `As` methods are added into the
`tint` namespace to make usage shorter.
Change-Id: I0decedb92ebed01b6aa12d2e3efa7190742e9a33
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/127402
Commit-Queue: Dan Sinclair <dsinclair@chromium.org>
Kokoro: Kokoro <noreply+kokoro@google.com>
Reviewed-by: Ben Clayton <bclayton@google.com>
This Cl removes the boolean return values from most of the GLSL
writer methods. The diagnostics are used to determine if the generation
was successful. The writer itself just continues until complete.
Change-Id: Ia64968eaa6a0aa39a9713fa78f3e743f2de38b44
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/127020
Reviewed-by: Ben Clayton <bclayton@google.com>
Kokoro: Kokoro <noreply+kokoro@google.com>
Commit-Queue: Dan Sinclair <dsinclair@chromium.org>
Some of the benchmark shaders when run with the MSL backend are
currently failing because the robustness transform is injecting the
arrayLength method but there is no configuration for running the
arrayLength transform.
This Cl adds some default configuration to the MSL bench to make a slot
available for each possible value. It just always makes them available
such that all shaders work correctly.
Change-Id: Ie8a15abc0b7da7d2f46be11274e66302406be019
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/128060
Reviewed-by: Ben Clayton <bclayton@google.com>
Kokoro: Kokoro <noreply+kokoro@google.com>
Commit-Queue: Dan Sinclair <dsinclair@chromium.org>
This CL updates more of the SymbolTable to use string views.
Change-Id: I0d9c7817ba4e3b0d0e73cfc02406481eb796c3c2
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/127320
Kokoro: Kokoro <noreply+kokoro@google.com>
Commit-Queue: Dan Sinclair <dsinclair@chromium.org>
Reviewed-by: Ben Clayton <bclayton@google.com>
This CL moves the tint/traits file into tint/utils/traits. Traits is one
of the few items not in utils which is referred to by utils.
Change-Id: Ie955398f24e949b7618fdc868dbcb903fe20b3f1
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/127400
Reviewed-by: Ben Clayton <bclayton@google.com>
Kokoro: Kokoro <noreply+kokoro@google.com>
Commit-Queue: Dan Sinclair <dsinclair@chromium.org>
This vector was causing a lot of small allocations to happen, this CL
converts over to a utils::Vector to take advantage of the better
allocation strategy
Change-Id: Id740bd2a50a8d0bdd84f4b6e7c7ad4607436c7f4
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/127500
Kokoro: Kokoro <noreply+kokoro@google.com>
Reviewed-by: Ben Clayton <bclayton@google.com>
Commit-Queue: Dan Sinclair <dsinclair@chromium.org>
If specified, CMake will glob all the .wgsl files in this directory, and add these to the list of benchmarks to run.
Change-Id: I75b2754a6561dcd931c42bd47649c4f625e1c581
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/127980
Commit-Queue: Ben Clayton <bclayton@google.com>
Kokoro: Kokoro <noreply+kokoro@google.com>
Reviewed-by: Dan Sinclair <dsinclair@chromium.org>
Reduces hops from sem -> ast, just to know whether the variable has a binding point.
Change-Id: I5620198e6f08b73d5a0171d95874f1a2dae5d93e
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/127060
Reviewed-by: James Price <jrprice@google.com>
Auto-Submit: Ben Clayton <bclayton@google.com>
Kokoro: Kokoro <noreply+kokoro@google.com>
Commit-Queue: James Price <jrprice@google.com>
This CL removes the `NameFor` method from SymbolTable and accesses the
symbols name directly.
Change-Id: Ic4ad6eecfa78efb946d97aeaecf2d784af2e6f16
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/127301
Reviewed-by: Ben Clayton <bclayton@google.com>
Kokoro: Kokoro <noreply+kokoro@google.com>
Commit-Queue: Dan Sinclair <dsinclair@chromium.org>
This CL adds an allocator, owned by the SymbolTable, which stores the
names of all the symbols in the table. The Symbols then have a
`string_view` to their name.
Change-Id: I28e5b2aefcf9f67c1877b7ebab52416f780bd8c6
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/127300
Reviewed-by: Ben Clayton <bclayton@google.com>
Kokoro: Kokoro <noreply+kokoro@google.com>
Commit-Queue: Dan Sinclair <dsinclair@chromium.org>
This CL adds a simple allocator which will provide a chunk of memory of
the given size. It allocates out of slabs of memory.
Change-Id: I9acf59fac88cd6bef260b7ebae7d7b77fd939754
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/127302
Reviewed-by: Ben Clayton <bclayton@google.com>
Kokoro: Kokoro <noreply+kokoro@google.com>
Commit-Queue: Dan Sinclair <dsinclair@chromium.org>
Don't skip the TruncateInterstageVariables transform when
user defined interstage attribute input for fragment stage
is empty. Because builtin inputs could also cause register
mismatch for D3D12 HLSL compiler.
Add a boolean flag to Tint hlsl generator option to indicate
whether to run TruncateInterstageVariables or not.
This defaults to false in Tint, while Dawn always set
this to true for vertex stage.
Bug: dawn:1733
Change-Id: Ie4c3648b226513bf15f0e03ae4ce7f3cc09fdef4
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/127206
Reviewed-by: Ben Clayton <bclayton@google.com>
Kokoro: Kokoro <noreply+kokoro@google.com>
Commit-Queue: Shrek Shao <shrekshao@google.com>
When referencing a pointer parameter with a 'let', but not using that
let.
Also fix a bunch of places where we used the old names for the pointer
transforms.
Bug: chromium:1433499
Change-Id: I8decefeacd6150bd6f7637f80e62b9cb62936235
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/127540
Kokoro: Kokoro <noreply+kokoro@google.com>
Auto-Submit: Ben Clayton <bclayton@google.com>
Reviewed-by: Antonio Maiorano <amaiorano@google.com>
Commit-Queue: Ben Clayton <bclayton@google.com>
The demangler is no longer used by the Tint executable, and nothing else
calls it, so delete it.
Change-Id: Ic47238e4a6126e5daacd81dfc6f5986524b648ba
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/127280
Kokoro: Kokoro <noreply+kokoro@google.com>
Reviewed-by: Ben Clayton <bclayton@google.com>
Commit-Queue: Dan Sinclair <dsinclair@chromium.org>
Currently the SPIR-V headers are only included when building the
SPIRV_READER. This means the build breaks if the SPIRV_WRITER is enabled
but not the reader.
This CL fixes the conditionals to include the SPIR-V header for both the
reader and the writer.
Change-Id: I8690e3e42dbe3071dcaa3ed467d6eac8f62a7bbc
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/127160
Reviewed-by: Ben Clayton <bclayton@google.com>
Kokoro: Kokoro <noreply+kokoro@google.com>
Commit-Queue: Dan Sinclair <dsinclair@chromium.org>
This Cl removes the boolean return values from most of the SyntaxTree
writer methods. The diagnostics are used to determine if the generation
was successful. The writer itself just continues until complete.
Change-Id: I87d85e2cbe061306a4598a498ccad4510f62e8c9
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/124681
Kokoro: Kokoro <noreply+kokoro@google.com>
Reviewed-by: Ben Clayton <bclayton@google.com>
Commit-Queue: Dan Sinclair <dsinclair@chromium.org>
This Cl removes the boolean return values from most of the WGSL writer
methods. The diagnostics are used to determine if the generation was
successful. The writer itself just continues until complete.
Change-Id: Iff33128c1df286cab799f08dab8c3defc5651d76
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/124680
Commit-Queue: Dan Sinclair <dsinclair@chromium.org>
Kokoro: Kokoro <noreply+kokoro@google.com>
Reviewed-by: Ben Clayton <bclayton@google.com>
This CL removes the `error()` method from the TextGenerator in favour of
using `Diagnostics().str()`.
Change-Id: If20261ac839d11f6d29890cb17f95f2d4df5db2d
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/126923
Kokoro: Kokoro <noreply+kokoro@google.com>
Commit-Queue: Dan Sinclair <dsinclair@chromium.org>
Reviewed-by: Ben Clayton <bclayton@google.com>
Currently the multiplanar `ShouldRun` function is scanning all objects
looking for an external texture. This CL switches to using the type
manager and attempts to find an `ExternalTexture` type in the system. If
found, then the transform should run.
Change-Id: I28b8fa1835a83f2bff5d804e95bcbaa73d3d2124
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/126922
Reviewed-by: Ben Clayton <bclayton@google.com>
Kokoro: Kokoro <noreply+kokoro@google.com>
Commit-Queue: Dan Sinclair <dsinclair@chromium.org>
Currently the multiplanar transform is only run if a global variable is
seen with a `texture_external`. There are cases where a function with a
`texture_external` can exist (the fuzzers do this, the tint command line
can do this as well). These will fail when hitting the backend.
This CL makes the running of multiplanar unconditional. This causes the
logic to process the function parameters to run and the external texture
is removed.
Bug: chromium:1431610
Change-Id: I1793142e803e864c744b38cf843931521e6a449f
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/126921
Commit-Queue: Dan Sinclair <dsinclair@chromium.org>
Reviewed-by: Ben Clayton <bclayton@google.com>
Kokoro: Kokoro <noreply+kokoro@google.com>
This CL updates the common fuzzer code to enable the multiplanar
extension for the SPIR-V backend.
Bug: chromium:1427801
Change-Id: Ie03d3e10f7831ca851aa11c00cbac0aa474ccabb
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/126920
Kokoro: Kokoro <noreply+kokoro@google.com>
Commit-Queue: Dan Sinclair <dsinclair@chromium.org>
Reviewed-by: Ben Clayton <bclayton@google.com>
