This transform converts compound assignment statements into regular
assignments, hoisting LHS expressions and converting for-loops and
else-if statements if necessary.
The vector-component case needs particular care, as we cannot take the
address of a vector component. We need to capture a pointer to the
whole vector and also the component index expression:
// Before
vector_array[foo()][bar()] *= 2.0;
// After:
let _vec = &vector_array[foo()];
let _idx = bar();
(*_vec)[_idx] = (*_vec)[_idx] * 2.0;
Bug: tint:1325
Change-Id: I8b9b31fc9ac4b3697f954100ceb4be24d063bca6
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/85282
Reviewed-by: Antonio Maiorano <amaiorano@google.com>
Kokoro: Kokoro <noreply+kokoro@google.com>
Reuse the logic for resolving binary operator result types that was
implemented for binary expressions. This validates that the LHS and
RHS are compatible for the target operator. We then try to match the
resolved result type against the LHS store type.
Bug: tint:1325
Change-Id: If80a883079bb71fa6c4eb5545654279fefffacb4
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/74362
Reviewed-by: Antonio Maiorano <amaiorano@google.com>
Reviewed-by: Ben Clayton <bclayton@google.com>
Kokoro: Kokoro <noreply+kokoro@google.com>
Added a new transform::RemoveContinueInSwitch that replaces continue
statements in switch cases with setting a bool variable, and checking if
the variable is set after the switch to continue.
Bug: tint:1080
Change-Id: I3c0a6c790e1bb612fac3f927a4bd5beb2d0d4ed1
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/84960
Reviewed-by: James Price <jrprice@google.com>
Kokoro: Kokoro <noreply+kokoro@google.com>
Commit-Queue: Antonio Maiorano <amaiorano@google.com>
Implement new transform UnwindDiscardFunctions that replaces discard
statements with setting a module-level bool, adds a check and return for
this bool after every function call that may discard, and finally
invokes a single function that executes a discard from top-level
functions.
Regenerated tests and remove HLSL ones that used to fail FXC because it
had difficulty with discard.
Bug: tint:1478
Bug: chromium:1118
Change-Id: I09d680f59e2d5d0cad907bfbbdd426aae76d4bf3
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/84221
Reviewed-by: James Price <jrprice@google.com>
Reviewed-by: Ben Clayton <bclayton@google.com>
Kokoro: Kokoro <noreply+kokoro@google.com>
Commit-Queue: Antonio Maiorano <amaiorano@google.com>
When enabled, for each test, FXC is run first, and if it succeeds, DXC
is run. If both succeed, the test passes, otherwise it fails. This
option allows us to get meaningful feedback from running this script
against HLSL files, as well as taking advantage of the feature to delete
skips for tests that now pass (both FXC and DXC in this case).
Change-Id: Iae2ebfda7bd92f1b94893e648e2d1fb1f6979b39
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/84680
Reviewed-by: Ben Clayton <bclayton@google.com>
Kokoro: Kokoro <noreply+kokoro@google.com>
Commit-Queue: Antonio Maiorano <amaiorano@google.com>
We still use the StrideAttribute AST node in the SPIR-V reader for
strided arrays and matrices, which are then removed by transforms.
The WGSL parser no longer has to handle attributes on types.
Bug: tint:1381
Change-Id: Ifa39575ce207d3fdfcbef7125fe6a3686fad5f20
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/83963
Reviewed-by: Ben Clayton <bclayton@google.com>
Kokoro: Kokoro <noreply+kokoro@google.com>
Since this was the only attribute allowed on structures, we can also
remove the parsing code for them. However, we still need to have
attributes on the struct AST node, since the AddSpirvBlockAttribute
transform adds one.
Fixed: tint:1324
Change-Id: I7966237765b1d8a58c59908b59e1f1152a8a0439
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/83740
Kokoro: Kokoro <noreply+kokoro@google.com>
Reviewed-by: Ben Clayton <bclayton@google.com>
The WGSL spec says that line comments are terminated by any blankspace
other than a space or a horizontal tab.
Also rename is_whitespace to is_blankspace and tighten up the
definition to only include the characters listed in the WGSL spec.
Change-Id: I4fee0175980ab70e9baf107a6e79ab5c2e4f906d
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/83920
Reviewed-by: Ben Clayton <bclayton@google.com>
Kokoro: Kokoro <noreply+kokoro@google.com>
Commit-Queue: James Price <jrprice@google.com>
Invalid UTF-8 was being best-effort consumed, which given the right sequence of brokenness, could end up with diagnostic locations referring to bytes beyond the end of a line.
Improve the UTF-8 decoding so that it can detect when multi-byte codepoints are missing the high-bit being set.
Actually detect this in a lexer, and parser and produce errors.
Bug: tint:1437
Bug: chromium:1305648
Change-Id: I459f0df840b4ce8c4f5f82363f93602bf8326984
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/83540
Kokoro: Kokoro <noreply+kokoro@google.com>
Reviewed-by: Antonio Maiorano <amaiorano@google.com>
Commit-Queue: Ben Clayton <bclayton@google.com>
This transform ensures that expressions are evaluated in the order
defined in the WGSL spec. It does this by making sure to hoist
expressions that have side-effects (calls) along with variables that may
receive these side-effects to lets in the correct order.
Bug: tint:1300
Change-Id: Ic027dc4e0d894beff626a68b5837bd2eed26d8a5
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/78620
Reviewed-by: James Price <jrprice@google.com>
Kokoro: Kokoro <noreply+kokoro@google.com>
Commit-Queue: Antonio Maiorano <amaiorano@google.com>
Fails to compile this reinterpret_cast with:
error : cast from 'FARPROC' (aka 'long long (*)()') to 'pD3DCompile' (aka 'long (*)(const void *, unsigned long long, const char *, const _D3D_SHADER_MACRO *, ID3DInclude *, const char *, const char *, unsigned int, unsigned int, ID3D10Blob **, ID3D10Blob **)') converts to incompatible function type [-Werror,-Wcast-function-type]
pD3DCompile d3dCompile = reinterpret_cast<pD3DCompile>(
^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Casting the result of GetProcAddress to void* fixes this. Note that this
is the same thing Dawn does.
Change-Id: Ib185a4fe96c60163cb66cd9591679856ae95d7f2
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/83360
Kokoro: Kokoro <noreply+kokoro@google.com>
Reviewed-by: Ben Clayton <bclayton@google.com>
Commit-Queue: Antonio Maiorano <amaiorano@google.com>
On MSVC debug builds (VS 2022), running Tint against the
"unicode\identifiers.wgsl" test was triggering the following assert:
```
Program: C:\src\tint\out\build\x64-Debug\tint.exe
File: minkernel\crts\ucrt\src\appcrt\convert\isctype.cpp
Line: 36
Expression: c >= -1 && c <= 255
```
std::isdigit, isxdigit, and isspace specify that the behaviour is
undefined if the argument's value is neither representable as unsigned
char nor equal to EOF. For example, see:
https://en.cppreference.com/w/cpp/string/byte/isdigit. As suggested on
cppreference, to safely use these functions, we should first convert the
char argument to unsigned char.
Bug: tint:1437
Change-Id: I80e061820cfd87aca51758ae2e3b59306b157b04
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/83180
Reviewed-by: Ben Clayton <bclayton@google.com>
Kokoro: Kokoro <noreply+kokoro@google.com>
Commit-Queue: Antonio Maiorano <amaiorano@google.com>
Add a Hash() and Equals() methods to sem::Type.
Have sem::Manager (which should be called sem::TypeManager) derive from
utils::UniqueAllocator. This now uses the Hash() and Equals() for uniquely
constructing semantic types instead of building strings and comparing those.
Bug: tint:1383
Change-Id: I5e3229bd087391ac594d333a0ab4232cfcddf54d
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/82743
Kokoro: Kokoro <noreply+kokoro@google.com>
Reviewed-by: David Neto <dneto@google.com>
