Just like for divide, FXC fails with the exact same error when
performing a modulo on a value that FXC determines to be zero. We
address it in the same way as we do for divide.
This also fixes a couple of the vk-gl-cts tests for which I manually
generated expectation files for.
Bug: tint:1083
Change-Id: Ia388bf002112afded542adb791d37e88e35a77ff
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/74220
Reviewed-by: James Price <jrprice@google.com>
Kokoro: Kokoro <noreply+kokoro@google.com>
Commit-Queue: Antonio Maiorano <amaiorano@google.com>
Also ensure correct paramter ordering: bias is always after offset.
Bug: tint:1351
Change-Id: I41ee66b86cd9d912f3857e5377b660c50d035c6e
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/73720
Kokoro: Kokoro <noreply+kokoro@google.com>
Reviewed-by: James Price <jrprice@google.com>
Commit-Queue: Stephen White <senorblanco@chromium.org>
FXC fails to compile when it determines that the rhs of an integral
division is zero with "error X4010: Unsigned integer divide by zero".
bclayton's fix (https://dawn-review.googlesource.com/c/tint/+/60500)
addressed cases for division by an integer constant 0. This CL adds the
missing support for division by integral vectors with 0 components.
FXC also fails on division by integral expressions that it can fold to
0. To handle these cases, we now emit a runtime check for 0 and replace
by 1. In the cases I've tested, FXC seems able to optimize these checks
away.
Bug: tint:1083
Change-Id: I02f08e9077882f03c1e42b62dacb742a48fa48ba
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/73580
Kokoro: Kokoro <noreply+kokoro@google.com>
Reviewed-by: David Neto <dneto@google.com>
Reviewed-by: James Price <jrprice@google.com>
Commit-Queue: Antonio Maiorano <amaiorano@google.com>
There are a few places where a branch or return is created,
conditionally on whether a terminator was the last thing seen.
The goal is to generate a SPIR-V basic block terminator exactly
when needed, and to avoid generating a branch or return immediately
after a prior terminator.
Previously, the decision was based on the last thing seen in the AST.
But we should instead check the emitted SPIR-V instead.
This fixes cases such as a break or return inside an else-if.
That's because an if/elseif is actually a selection inside a selection.
Looking at the AST only works when trying to terminate the *inside*
selection. In the outer recursive call, the last AST node is
no longer a terminator, and we would skip generating the branch
to the merge block.
Fixed: tint:1315
Change-Id: I6b886ce85d1d681f2063997e469e0c1b4e5973a2
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/73480
Kokoro: Kokoro <noreply+kokoro@google.com>
Auto-Submit: David Neto <dneto@google.com>
Reviewed-by: James Price <jrprice@google.com>
Commit-Queue: James Price <jrprice@google.com>
SPIR-V validation fails when an object has multiple ArrayStride
decorations. Remove the SPIR-V reader tests that did this.
Change-Id: Iaac0f0f0c9230ccbc7dde5414d6aaf6f0e972534
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/73240
Auto-Submit: David Neto <dneto@google.com>
Reviewed-by: James Price <jrprice@google.com>
Kokoro: David Neto <dneto@google.com>
Commit-Queue: David Neto <dneto@google.com>
* Rename GenerateNonReferenceExpression to
GenerateExpressionWithLoadIfNeeded.
This version takes an ast::Expression
* Add a variant that takes a sem::Expression, because the sem
expression already knows the resolved type, and so we can save
a lookup.
* Replace most uses of GenerateExpression ... GenerateLoadIfNeeded
with a call to one of the above.
This is a non-functional change.
Followup to the fix in tint:1343.
Bug: tint:1343
Change-Id: If19a1bc7670edd2badc1533861d8b42f0825c7b8
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/72720
Auto-Submit: David Neto <dneto@google.com>
Kokoro: Kokoro <noreply+kokoro@google.com>
Reviewed-by: Antonio Maiorano <amaiorano@google.com>
Commit-Queue: David Neto <dneto@google.com>
A one letter typo would lead to invalid memory access in the very
specific case of outputting the layout for a struct within a struct with
field alignment padding, and the inner struct has more members than the
outer.
Bug: tint:1344
Bug: oss-fuzz:72642
Change-Id: I749e3fb172e78a20ece68b40be1a0a57dc5746f4
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/72642
Reviewed-by: Ben Clayton <bclayton@google.com>
Reviewed-by: David Neto <dneto@google.com>
Kokoro: Kokoro <noreply+kokoro@google.com>
Commit-Queue: Antonio Maiorano <amaiorano@google.com>
Vulkan allows them, but WGSL does not.
A duplicate decoration is purely redundant.
A parameterized decoratio with different parameterization is
an inconsistency and a semantic error, at least for currently defined
SPIR-V decorations.
So for each target, only take the first decoration of each kind.
Fixed: tint:1337
Change-Id: I6ed5c39cf2e213c695cb8217ed1b97814da3db56
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/72500
Reviewed-by: Ben Clayton <bclayton@google.com>
Auto-Submit: David Neto <dneto@google.com>
Kokoro: Kokoro <noreply+kokoro@google.com>
Reviewed-by: James Price <jrprice@google.com>
Commit-Queue: James Price <jrprice@google.com>
FXC fails to compile code that assigns to dynamically-indexed fixed-size
arrays in structs on internal shader variables with:
error X3500: array reference cannot be used as an l-value; not natively
addressable
This CL detects this case, and transforms such assignments into copying
out the array to a local variable, assigning to that local, and then
copying the array back.
Also manually regenerate SKIPs for HLSL/FXC after this change, which
fixes 30 tests. Also exposes some "compilation aborted unexpectedly" now
that "array reference cannot be used as an l-value" has been fixed. For
tests that fail for both DXC and FXC, updating SKIPs to the DXC one to
help distinguish actual FXC bugs from valid errors.
Bug: tint:998
Bug: tint:1206
Change-Id: I09204d8d81ab27d1c257538ad702414ccc386543
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/71620
Reviewed-by: Ben Clayton <bclayton@google.com>
Kokoro: Kokoro <noreply+kokoro@google.com>
Commit-Queue: Antonio Maiorano <amaiorano@google.com>
These are errors:
let a = 0x;
let b = -0x;
Fixes: tint:1338
Change-Id: I9d26ad66e32deb954550c0ecfbda0a9005bcd380
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/72380
Reviewed-by: Ben Clayton <bclayton@google.com>
Commit-Queue: David Neto <dneto@google.com>
Auto-Submit: David Neto <dneto@google.com>
Kokoro: Kokoro <noreply+kokoro@google.com>
The GLSL emitted for these was invalid, and we don't need these
structs since they're only used as the store types of buffers, which
are handled elsewhere.
Change-Id: I17c15e408b5c36e9b895e5950528a6d02d1802a6
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/72381
Reviewed-by: Stephen White <senorblanco@chromium.org>
Kokoro: Kokoro <noreply+kokoro@google.com>
These are no longer necessary and will soon be deprecated and removed.
Bug: tint:1324
Change-Id: I3fa076e7ce5eb36466d24c80fd1c83658c28c5ab
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/72086
Reviewed-by: David Neto <dneto@google.com>
Reviewed-by: Ben Clayton <bclayton@google.com>
Kokoro: Kokoro <noreply+kokoro@google.com>
These are no longer necessary, and the attribute will soon be
deprecated and removed.
Bug: tint:1324
Change-Id: Ia3cdd4e9d9bd7bca45d734a06466993199b3e838
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/72085
Reviewed-by: Ben Clayton <bclayton@google.com>
Kokoro: Kokoro <noreply+kokoro@google.com>
No code should rely on the presence of the block attribute, which will
soon be deprecated and removed.
Bug: tint:1324
Change-Id: I868d5e795e66a93bdf99b94389c07dec98cb0ec2
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/72084
Reviewed-by: Ben Clayton <bclayton@google.com>
Kokoro: Kokoro <noreply+kokoro@google.com>
Skipping those that are block-decorated is not correct, as
block-decorated structures can also have non-buffer usages. This is
even clearer now that WGSL has removed the block attribute.
Bug: tint:1324
Change-Id: I6484766a5c541d39e2dc08beb3ae7b889759a3fb
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/72083
Kokoro: Kokoro <noreply+kokoro@google.com>
Reviewed-by: Ben Clayton <bclayton@google.com>
Any struct which is used as the store type of a buffer variable needs
to have a block decoration. If that struct is nested inside an array
or another struct, we wrap it inside another struct first.
This removes the SPIR-V backend's reliance on the [[block]] attribute,
which will soon be deprecated and removed.
Bug: tint:1324
Change-Id: Ib6ad54f24a3e4a090da9faeed699f266abcb66ba
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/72082
Kokoro: Kokoro <noreply+kokoro@google.com>
Reviewed-by: Ben Clayton <bclayton@google.com>
Replace all validation rules that rely on the block attribute with the
new rules based on fixed-footprint types.
Bug: tint:1324
Change-Id: I02656537bee66e6e1af95875e503a37bf23d4a6b
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/72081
Kokoro: Kokoro <noreply+kokoro@google.com>
Reviewed-by: Ben Clayton <bclayton@google.com>
This transform does not use the top-level ctx.Clone() method, so we
need to manually copy the applied transforms over to the destination
program.
This fixes the GLSL backend, where this transform was dropping the
applied transforms and breaking a required transform dependency later
in the chain.
Change-Id: I3db688fe35dfbe67468b9941f739cb2ecf3a843b
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/72220
Kokoro: Kokoro <noreply+kokoro@google.com>
Reviewed-by: Ben Clayton <bclayton@google.com>
Commit-Queue: Ben Clayton <bclayton@google.com>
This fixes errors like "error X3500: array reference cannot be used as
an l-value; not natively addressable". Note that FXC treats matrices
like arrays. We still get this error for dynamically indexed arrays in
structs.
Also improved HLSL assign tests, and add missing ones for vector
indexing.
Manually removed 20 e2e skip hlsl SKIP files that are now passing with
this change.
Bug: tint:1333
Change-Id: If23881a667857a4d4ec6881e72666af0a666ef10
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/71982
Reviewed-by: Ben Clayton <bclayton@google.com>
Kokoro: Kokoro <noreply+kokoro@google.com>
Commit-Queue: Antonio Maiorano <amaiorano@google.com>
All writers implemented, along with resolving and validation.
TODO: SPIR-V Reader.
Bug: tint:1330
Change-Id: I8ba2f6023749474f80efb8a5422ac187e6c73a69
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/71820
Reviewed-by: David Neto <dneto@google.com>
Commit-Queue: Ben Clayton <bclayton@google.com>
Kokoro: Ben Clayton <bclayton@google.com>
If TINT_SYMBOL_STORE_DEBUG_NAME is 1, Symbol instances store a
`debug_name_` member initialized with the name of the identifier they
represent. This member is not exposed, but is useful for debugging
purposes.
Bug: tint:1331
Change-Id: Ia98e266aefc1ca26bbf30c6ece73d9eac8afdbd7
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/71780
Kokoro: Kokoro <noreply+kokoro@google.com>
Commit-Queue: Antonio Maiorano <amaiorano@google.com>
Reviewed-by: Ben Clayton <bclayton@google.com>
FXC is buggy, and I recently landed changes in Dawn to run with "/O0"
rather than /"O2" because of these bugs. Let's make sure Tint end-to-end
tests do the same. Also do the same when running against DXC.
Bug: dawn:1203
Change-Id: I1a30f16dee8306bd645d87b3ccb0cc87691c5972
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/71800
Reviewed-by: Ben Clayton <bclayton@google.com>
Kokoro: Kokoro <noreply+kokoro@google.com>
Commit-Queue: Antonio Maiorano <amaiorano@google.com>
Fix dependency graph traversal for bitcasts. These were not being traversed, leading to an ICE if the bitcast type was an alias, as the symbol was not resolved for later use by the resolver.
Add missing validation for bitcasts. We were permitting any bitcast that wasn't a being cast to a pointer type, when the spec only allows:
* numeric_scalar to numeric_scalar
* vecN<numeric_scalar> to vecN<numeric_scalar>
Add lots of tests.
Fixed: chromium:1276320
Change-Id: I9e5487ec7649ac543f73fc878e7e282bf932d8cb
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/71681
Kokoro: Kokoro <noreply+kokoro@google.com>
Commit-Queue: Ben Clayton <bclayton@google.com>
Reviewed-by: Antonio Maiorano <amaiorano@google.com>
Migrate some of the validation logic over to use the results of behavior
analysis.
The most significant changes are:
* Unreachable-statements now consider merge-points of control flow. For
example, if all branches of a if-statement or switch-statement either
return or discard, the next statement will be considered unreachable.
* Unreachable statements are no longer an error, but a warning. See
https://github.com/gpuweb/gpuweb/issues/2378.
* Statements that follow a loops that does not break, or have a
conditional will now be considered unreachable.
* Unreachable statements produced by the SPIR-V reader are now removed
using the new RemoveUnreachableStatements transform.
Some other new changes include additional validation for the continuing
block for for-loops, to match the rules of a loop continuing block.
The new cases this validation is testing for are not expressible in
WGSL, but some transforms may produce complex continuing statements that
might violate these rules. All the writers are able to decay these
complex for-loop continuing statements to regular loops.
Bug: tint:1302
Change-Id: I0d8a48c73d5d5c30a1cddf92cc3383a692a58e61
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/71500
Kokoro: Kokoro <noreply+kokoro@google.com>
Reviewed-by: David Neto <dneto@google.com>
writer::spirv::Function attempted to append a trailing OpReturn if the
function does not end with a terminating instruction. This wasn't
considering functions that have a non-void return type.
This has now been moved to spirv::Builder::GenerateFunction(),
where we can actually examine the function return type, and generate a
zero-expression to return if we need to.
Note: this was masked by WGSL validation that required all functions to
end with a return statement.
Bug: tint:1302
Change-Id: Iddfeda25a956622c318b8235dc6fc093a2a5c26d
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/71604
Kokoro: Kokoro <noreply+kokoro@google.com>
Reviewed-by: David Neto <dneto@google.com>
This change implements the behavior analysis for expressions and
statements as described in:
https://www.w3.org/TR/WGSL/#behaviors
This CL makes no changes to the validation rules. This will be done as a
followup change.
Bug: tint:1302
Change-Id: If0a251a7982ea15ff5d93b54a5cc5ed03ba60608
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/68408
Kokoro: Kokoro <noreply+kokoro@google.com>
Commit-Queue: Ben Clayton <bclayton@google.com>
Reviewed-by: David Neto <dneto@google.com>
The SwitchCaseBlockStatement was bound to the BlockStatement of an ast::CaseStatement, but we had nothing that mapped to the actual ast::CaseStatement.
sem::CaseStatement replaces sem::SwitchCaseBlockStatement, and has a Block() accessor, providing a superset of the old behavior.
With this, we can now easily validate the `fallthrough` rules directly, instead of scanning the switch case. This keeps the validation more tigtly coupled to the ast / sem nodes.
Change-Id: I0f22eba37bb164b9e071a6166c7a41fc1a5ac532
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/71460
Kokoro: Kokoro <noreply+kokoro@google.com>
Reviewed-by: Antonio Maiorano <amaiorano@google.com>
Commit-Queue: Ben Clayton <bclayton@google.com>
Produces a direct SEM -> SEM pointer, reducing AST <-> SEM hopping.
Change-Id: I233b4c47d4e55b5f2c6e14ed08699a302b8fb64d
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/71321
Kokoro: Kokoro <noreply+kokoro@google.com>
Commit-Queue: Ben Clayton <bclayton@chromium.org>
Reviewed-by: James Price <jrprice@google.com>
Adds functionality and tests for performing the multiplanar external
texture transform on texture_external parameters in user-defined
functions.
Bug: dawn:1082
Change-Id: I15f23e639a6cd24ed3428055420908f05b69c0c6
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/70840
Kokoro: Kokoro <noreply+kokoro@google.com>
Reviewed-by: Ben Clayton <bclayton@google.com>
Commit-Queue: Ben Clayton <bclayton@google.com>
Some versions of clang warn as error without this:
error: definition of implicit copy constructor for 'EnumSet<tint::ast::PipelineStage>' is deprecated because it has a user-declared copy assignment operator [-Werror,-Wdeprecated-copy]
Change-Id: I02cee0b1f7b70d74f29da0067e888b50acf27ee1
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/71600
Commit-Queue: Ben Clayton <bclayton@google.com>
Kokoro: Kokoro <noreply+kokoro@google.com>
Reviewed-by: Corentin Wallez <cwallez@chromium.org>
Add assignment operators and methods for adding, removing and operators for performing set arithmatic.
Change-Id: I13d30734354f503180f75480866b54034f647c2a
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/71320
Commit-Queue: Ben Clayton <bclayton@google.com>
Reviewed-by: Antonio Maiorano <amaiorano@google.com>
Kokoro: Kokoro <noreply+kokoro@google.com>
Nothing currently does the analysis to calculate these.
Change-Id: Ia2103102fbf36109f357aebf32cdfda24d6d8155
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/68407
Kokoro: Kokoro <noreply+kokoro@google.com>
Reviewed-by: David Neto <dneto@google.com>
Commit-Queue: Ben Clayton <bclayton@chromium.org>
Break up Resolver::Statement() into multiple resolver functions.
Move simple statement validation out to resolver_validation.cc
Change-Id: Ifa29433af0a9afa39a66ac3e4f7ca376351adfbf
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/71102
Kokoro: Kokoro <noreply+kokoro@google.com>
Commit-Queue: Ben Clayton <bclayton@google.com>
Reviewed-by: Antonio Maiorano <amaiorano@google.com>
Now that I've landed this change to Dawn to disable FXC optimizations:
https://dawn-review.googlesource.com/c/dawn/+/70700,, we can reland this
change. The Tint-into-Dawn roll was failing because FXC would miscompile
certain loops into infinite loops when not unrolled.
Also reland
test/bug/fxc/gradient_in_varying_loop/1112.wgsl.expected.hlsl
Bug: tint:1112
Bug: dawn:1203
Change-Id: I641d68864b833e0fbe3b117d397b89ae96482536
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/71000
Kokoro: Kokoro <noreply+kokoro@google.com>
Reviewed-by: Ben Clayton <bclayton@google.com>
Commit-Queue: Antonio Maiorano <amaiorano@google.com>
If we encounter a global twice, the stack vector was not popped, which could trigger false cyclic-dependency errors. Because the cyclic dependency did not actually exist, later logic could break, expecting to find a global->global edge which did not exist.
Fixed: chromium:1273451
Change-Id: I9d99f1aeeaea042d9ed847a878c4717803122240
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/70980
Reviewed-by: David Neto <dneto@google.com>
Kokoro: Kokoro <noreply+kokoro@google.com>
Commit-Queue: Ben Clayton <bclayton@google.com>
RemovePhonies was transforming:
_ = f32(1)
_ = vec2<f32>(1.0, 2.0)
into:
f32(1)
vec2<f32>(1.0, 2.0)
Which the resolver gets grumpy about, as these are expressions, not statements.
Fixed: chromium:1273230
Change-Id: Ie85d3cee705fa3f792db686c021d76331e241f17
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/70960
Auto-Submit: Ben Clayton <bclayton@google.com>
Commit-Queue: James Price <jrprice@google.com>
Kokoro: Ben Clayton <bclayton@chromium.org>
Reviewed-by: James Price <jrprice@google.com>
David Neto
Antonio Maiorano
Stephen White
James Price
Ben Clayton
Brandon Jones