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>
In order to preserve padding properly for MSL, we need to use its
packed_vec type for all vec3 types in storage buffers, not just struct
members. This commit includes a complete rewrite of the PackedVec3
transform to achieve this. The key details are:
* An internal `__packed_vec3<>` type was added, which corresponds to a
`type::Vector` with an additional flag to indicate that it will be
emitted as packed vector.
* The `PackedVec3` transform replaces all vec3 types used in
host-shareable address spaces with the internal `__packed_vec3`
type. This includes vec3 types that appear as the store type of a
pointer.
* When used as an array element, these `__packed_vec3` types are
wrapped in a struct that contains a single `__packed_vec3`
member. This allows us to add an `@align()` attribute that ensures
that `array<vec3<T>>` still has the correct array element stride.
* When the `vec3<T>` appears as a struct member in the input program,
we apply the `@align()` to that member to ensure that we do not
change its offset.
* Matrix types with three rows that are used in memory are replaced
with an array of columns, where each column uses a `__packed_vec3`
inside an aligned wrapper structure as above.
* Accesses to host-shareable memory that involve any of these types
invoke a "pack" or "unpack" helper function to convert them to the
equivalent type that uses `__packed_vec3` or a regular `vec3` as
required.
* The `chromium_internal_relaxed_uniform_layout` extension is used to
avoid issues where modifying a type in the uniform address space
triggers stricter layout validation rules.
Bug: tint:1571
Fixed: tint:1837
Change-Id: Idaf2da2f5bcb2be00c85ec657edfb614186476bb
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/121200
Reviewed-by: Ben Clayton <bclayton@google.com>
Commit-Queue: James Price <jrprice@google.com>
Kokoro: Kokoro <noreply+kokoro@google.com>
The resolver now wraps sem::Expression objects with a sem::Load object
anywhere that the load rule is invoked. sem::Expression provides an
`UnwrapLoad()` method that returns the inner expression (or
passthrough, if no load is present), which is analaguous to
Type::UnwrapRef().
The logic for alias analysis in `RegisterLoadIfNeeded` has been folded
into the new `Resolver::Load` method.
Fixed up many transforms and tests. The only difference in output is
for a single SPIR-V backend test, where some IDs have changed due to
slight re-ordering of when expressions are generated.
There may be further clean-ups possible (e.g. removing unnecessary
calls to `UnwrapRef`, and simplifying places in the SPIR-V writer or
transforms that deal with memory accesses), but these can be addressed
in future patches.
Fixed: tint:1654
Change-Id: I69adecfe9251faae46546b64d0cdc29eea26cd4e
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/99706
Commit-Queue: James Price <jrprice@google.com>
Kokoro: Kokoro <noreply+kokoro@google.com>
Reviewed-by: Antonio Maiorano <amaiorano@google.com>
Reviewed-by: Ben Clayton <bclayton@google.com>
Change tint's `--fxc` flag to take the path of the FXC compiler DLL.
Have tint attempt to validate with both FXC and DXC if `--validate` is
passed.
Fix the 'dirsWithNoPassExpectations' logic which looks like it got
broken with the tint -> dawn merge. It also incorrectly applied
filepath.FromSlash() on windows.
Change-Id: I0f46aa5c21bc48a2abc48402c41f846aff4a8633
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/96800
Reviewed-by: Antonio Maiorano <amaiorano@google.com>
Commit-Queue: Ben Clayton <bclayton@chromium.org>
Enable the parsing of 'const'.
Warn on use of module-scope 'let', and automatically replace with 'const'.
Fixed: tint:1580
Change-Id: I214aabca80686dc6b60ae21a7a57fbfb4898ea83
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/93786
Commit-Queue: Ben Clayton <bclayton@google.com>
Reviewed-by: Dan Sinclair <dsinclair@chromium.org>
Kokoro: Kokoro <noreply+kokoro@google.com>
Once 'const' is introduced, let will no longer resolve to a
creation-time constant value.
Add temporary code into each of the writers to prevent constants that
originate from a 'let' from being inlined. This will reduce the amount
of noise in later CLs.
Bug: tint:1580
Change-Id: Id3493a43ac09fe9f042ff2d517d04b2ae854d43e
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/94541
Commit-Queue: Ben Clayton <bclayton@google.com>
Reviewed-by: David Neto <dneto@google.com>
Kokoro: Kokoro <noreply+kokoro@google.com>
Commit-Queue: Ben Clayton <bclayton@chromium.org>
If the object and index are both constant expressions, resolve the
index as a constant.
Note: Expectations have been updated for indexing on 'let's. Once
'const' is introduced, 'let' will no longer be treated as a constant
expression, and these expectations will be reverted.
Bug: tint:1580
Change-Id: I42793d36c1a5f82890ccaa5dbbb71b4346493bef
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/94329
Kokoro: Kokoro <noreply+kokoro@google.com>
Reviewed-by: Dan Sinclair <dsinclair@chromium.org>
Commit-Queue: Ben Clayton <bclayton@chromium.org>
This CL updates all of the Tint unittests to the new @stage shorter
syntax. This also updates the WGSL writer to emit the new short forms
instead of using the deprecated form.
Bug: tint:1503
Change-Id: I8c49e5319a19cccb5b4b5078f3ab39c50f31a9a8
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/92483
Reviewed-by: Ben Clayton <bclayton@google.com>
Commit-Queue: Dan Sinclair <dsinclair@chromium.org>
Kokoro: Kokoro <noreply+kokoro@google.com>
This is required to handle materialized values, and for constant
expressions.
Bug: tint:1504
Change-Id: I79ad567954de2d1cfea09dda255894e4e2aa678e
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/92081
Kokoro: Kokoro <noreply+kokoro@google.com>
Reviewed-by: Antonio Maiorano <amaiorano@google.com>
Commit-Queue: Ben Clayton <bclayton@chromium.org>
This is required to handle materialized values, and for constant
expressions.
Bug: tint:1504
Change-Id: Ic3ac62317241fa6f7009360128f222aeb56f62e4
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/92083
Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Commit-Queue: Ben Clayton <bclayton@google.com>
This is required to handle materialized values, and for constant
expressions.
Bug: tint:1504
Change-Id: Ie0177f148e08a0e1a3f4d7e06e283f121655804b
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/92080
Commit-Queue: Ben Clayton <bclayton@google.com>
Kokoro: Kokoro <noreply+kokoro@google.com>
Reviewed-by: Antonio Maiorano <amaiorano@google.com>
If the literal was constructed with an 'f', make sure we print it.
Bug: tint:1504
Change-Id: I6f04e31a166919c07574db56b0a2063ce5b8ca5c
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/91965
Reviewed-by: Antonio Maiorano <amaiorano@google.com>
Commit-Queue: Ben Clayton <bclayton@google.com>
Unsuffixed integer literals are currently treated as i32,
but will shortly become AbstractInteger. To keep tests behaving
identically to how they are currently, change all test literals
to using either 'i' or 'u' suffixes.
Bug: tint:1504
Change-Id: Ic373d18ce1c718a16b6905568aec89da3641d36b
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/88845
Reviewed-by: Dan Sinclair <dsinclair@chromium.org>
Commit-Queue: Ben Clayton <bclayton@google.com>
For all i32 literal values.
Reduces risk of the SPIR-V reader producing WGSL that behaves
differently, when abstract-integers are fully implemented.
Bug: tint:1504
Change-Id: Ieaf8afec5b09c7978c75a38c6ed144633ddc017e
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/88843
Commit-Queue: Ben Clayton <bclayton@google.com>
Reviewed-by: Dan Sinclair <dsinclair@chromium.org>
James Price
Ben Clayton
dan sinclair
Ryan Harrison