Performs a module-scope (global) declaration dependency analysis, so
that out-of-order global declarations can be re-ordered into dependency
order for consumption by the resolver.
The WGSL working group are currently debating whether out-of-order
declarations should be included in WebGPU V1, so this implementation
currently errors if module-scope declarations are declared out-of-order,
and the resolver does not currently use this sorted global list.
The analysis does however provide significantly better error diagnostics
when cyclic dependencies are formed, and when globals are declared
out-of-order.
The DependencyGraph also correctly now detects symbol collisions between
functions and types (tint:1308).
With this change, validation is duplicated between the DependencyGraph
and the Resolver. The now-unreachable validation will be removed from
the Resolver with a followup change.
Fixed: tint:1308
Bug: tint:1266
Change-Id: I809c23a069a86cf429f5ec8ef3ad9a98246766ab
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/69381
Kokoro: Kokoro <noreply+kokoro@google.com>
Commit-Queue: Ben Clayton <bclayton@chromium.org>
Reviewed-by: David Neto <dneto@google.com>
This is no longer used.
Fixed: tint:1225
Change-Id: I0cfe9955687a2b7ded3e645c573f3bffbc2f1f84
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/66380
Kokoro: Kokoro <noreply+kokoro@google.com>
Reviewed-by: David Neto <dneto@google.com>
Reviewed-by: James Price <jrprice@google.com>
Commit-Queue: Ben Clayton <bclayton@google.com>
The intrinsics that did anything useful with this were deprecated
several releases ago.
Change-Id: I79e3c901b6a78583853a067ec46cfa98e346517c
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/66262
Kokoro: Kokoro <noreply+kokoro@google.com>
Commit-Queue: Ben Clayton <bclayton@google.com>
Reviewed-by: James Price <jrprice@google.com>
- function scope variable store type must be constructible
- add IsConstructible() to sem::atomic
Bug: tint:1069
Change-Id: Ib0616b486ecf278dbdd99640dc4ede7f3007feb8
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/60120
Kokoro: Kokoro <noreply+kokoro@google.com>
Reviewed-by: Ben Clayton <bclayton@google.com>
Commit-Queue: Sarah Mashayekhi <sarahmashay@google.com>
Auto-Submit: Sarah Mashayekhi <sarahmashay@google.com>
Adds a new single-function API for the generators, which applies the
sanitizing transform and performs the generation in one step, and
returns a result object which contains the generated code and success
status/diagnostics.
The new APIs take an `Option` structure to control backend-specific
generation details (e.g. MSL fixed sample mask). The result objects
also provide backend-specific feedback (e.g. whether a UBO of buffer
lengths was generated).
HLSL needs a list of entry points to validate, and it's the HLSL
sanitizer that generates an entry point for programs that do not have
one. This change makes the HLSL generator return the list of
post-sanitize entry points so that the Tint executable can forward
them to the validation code.
Change-Id: I2d5aa27fda95d7c50c5bef41e206aee38f2fd2eb
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/57101
Auto-Submit: James Price <jrprice@google.com>
Kokoro: Kokoro <noreply+kokoro@google.com>
Commit-Queue: James Price <jrprice@google.com>
Reviewed-by: Ben Clayton <bclayton@chromium.org>
Also split out validation tests from call_test.cc into call_validation_test.cc.
Bug: tint:886
Change-Id: I1e1dee9b7c348363e89080cdecd3119cc004658f
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/54063
Kokoro: Kokoro <noreply+kokoro@google.com>
Reviewed-by: David Neto <dneto@google.com>
Add a Resolver::IsPlain() method to check for plain types, which is
then used instead of IsStorable() for validating array and struct
subtypes.
Remove validation of assignment and constructor RHS types, instead
validating the type of the variable declaration. This catches
additional errors that were previously missed, such as using a pointer
for a var declaration with no constructor.
Change-Id: I5786a262159d2a42cc05b44743c6c26f6b5647c0
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/53960
Auto-Submit: James Price <jrprice@google.com>
Commit-Queue: Ben Clayton <bclayton@google.com>
Kokoro: Kokoro <noreply+kokoro@google.com>
Reviewed-by: Ben Clayton <bclayton@google.com>
Handle access control on var declarations instead of via [[access]]
decorations. This change does the minimal work to migrate the WGSL
parser over to the new syntax. Additional changes will be needed
to correctly generate defaulted access qualifiers, as well as
validating access usage.
The [[access]] decorations are still supported by the WGSL parser,
with new deprecated warnings, but not for aliases. Example:
var x : [[access(x)]] alias_to_struct;
Making this work is far more effort than I want to dedicate to backwards
compatibility, and I do not beleive any real-world usage will be doing
this.
Still TODO:
* Adding access control as the optional, third parameter to ptr<>.
* Calculating default accesses for the various storage types.
* Validating usage of variables against the different accesses.
Bug: tint:846
Change-Id: If8ca82e5d16ec319ecd01f9a2cafffd930963bde
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/53088
Commit-Queue: Ben Clayton <bclayton@google.com>
Reviewed-by: James Price <jrprice@google.com>
Reviewed-by: David Neto <dneto@google.com>
Kokoro: Kokoro <noreply+kokoro@google.com>
Validate that:
* That resource variables have resource bindings
* Only resource variables have resource bindings
* That a [[binding]] decoration is paired with a [[group]]
* That binding points are not reused in the same entry point
Fixed: tint:235
Fixed: tint:645
Bug: tint:645
Change-Id: I2542934b4c6a2b4bbde48242932c04c796033a90
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/50500
Commit-Queue: Ben Clayton <bclayton@google.com>
Reviewed-by: David Neto <dneto@google.com>
Helps catch brokenness before asking the developer to stare at AST dumps.
Change-Id: Id39602225ad2d4a1b72dcdf998b164feda870cf1
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/49751
Kokoro: Kokoro <noreply+kokoro@google.com>
Commit-Queue: Ben Clayton <bclayton@chromium.org>
Reviewed-by: Antonio Maiorano <amaiorano@google.com>
Produces ugly code for the common case.
Change-Id: I8e5b1215e19fc6461dc40b8a91922db25f9cbd76
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/47764
Commit-Queue: Ben Clayton <bclayton@google.com>
Reviewed-by: Ryan Harrison <rharrison@chromium.org>
Reviewed-by: James Price <jrprice@google.com>
https://gpuweb.github.io/gpuweb/wgsl/#variable-declaration
Variables in the storage storage class and variables with a storage
texture type must have an access attribute applied to the store type.
https://gpuweb.github.io/gpuweb/wgsl/#module-scope-variables
A variable in the storage storage class is a storage buffer variable. Its
store type must be a host-shareable structure type with block attribute,
satisfying the storage class constraints.
Fixup tests, including those that were producing warnings about `var <in>`
The WGSL writer seems to want to put a newline after every decoration block, leading to some ugly output. I'll fix this as a separate change.
Fixes: tint:531
Fixes: tint:692
Change-Id: If09d987477247ab4a7c635f6ee6e616a06061515
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/47763
Commit-Queue: Ben Clayton <bclayton@google.com>
Kokoro: Kokoro <noreply+kokoro@google.com>
Reviewed-by: James Price <jrprice@google.com>
Reviewed-by: Ryan Harrison <rharrison@chromium.org>
* Moved global variable resolving logic to new function Resolver::GlobalVariable, and moved validation logic there.
* Moved global variable-related tests to resolver tests.
* Fixed many tests that started failing after this change, mainly because many globals were declared with no storage class. I set most of these to "Input".
Bug: tint:642
Change-Id: I0f8ea2091ed2bb3faa358f9497cd884b2994a40f
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/46940
Kokoro: Kokoro <noreply+kokoro@google.com>
Reviewed-by: Ben Clayton <bclayton@google.com>
Commit-Queue: Antonio Maiorano <amaiorano@google.com>
* Formerly, we reported the same error message if we detected no default
clause or more than one. I made it so that we output a different error
message for each. This makes it more clear, and in the case of more than
one, the error source location points at the second default clause,
rather than at the switch statement.
* Add functions to ProgramBuilder to more easily define switch and case
statements.
* Fix broken tests as a result of this change.
Bug: tint:642
Change-Id: Iab4e610a563165862d9bc190772d32a4dd24ac45
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/45880
Kokoro: Kokoro <noreply+kokoro@google.com>
Reviewed-by: Ben Clayton <bclayton@google.com>
Commit-Queue: Antonio Maiorano <amaiorano@google.com>
Added enforcement for vector constructor type rules according to the
table in https://gpuweb.github.io/gpuweb/wgsl.html#type-constructor-expr.
This surfaced a number of existing tests that violated some of these
rules or had a type-declaration related bug, so this CL fixes those as
well (these tests either passed the incorrect number of arguments to a
vector constructor or relied on implicit conversions between numeric
types).
Fixed: tint:632
Fixed: tint:476
Change-Id: I8279be3eeae50b64db486ee7a91a43bd94fdff62
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/44480
Commit-Queue: Arman Uguray <armansito@chromium.org>
Reviewed-by: Ben Clayton <bclayton@google.com>
Implements https://github.com/gpuweb/gpuweb/pull/1447
SPIR-V Reader is still TODO, but continues to function as the offset
decoration is still supported.
Bug: tint:626
Bug: tint:629
Change-Id: Id574eb3a5c6729559382812de37b23f0c68fd406
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/43640
Commit-Queue: Ben Clayton <bclayton@chromium.org>
Reviewed-by: David Neto <dneto@google.com>
All includes from .cc to .h are preserved, even when transitively included.
It's clear that there are far too many includes in header files, and we should be more aggressive with forward declarations. tint:532 will continue to track this work.
There are, however, plenty of includes that have accumulated over time which are no longer required directly or transitively, so this change starts with a clean slate of *required* includes.
Bug: tint:532
Change-Id: Ie1718dad565f8309fa180ef91bcf3920e76dba18
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/44042
Commit-Queue: Ben Clayton <bclayton@google.com>
Reviewed-by: Antonio Maiorano <amaiorano@google.com>
Statements like `var u : u32 = 0;` should fail because '0' is a signed
integer being used to initialize an unsigned variable.
Added test.
Bug: tint:79
Change-Id: I34f6d21b4355167f49f9a8b5d4ed34a808823185
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/42702
Reviewed-by: Ben Clayton <bclayton@google.com>
Reviewed-by: Antonio Maiorano <amaiorano@google.com>
Commit-Queue: Antonio Maiorano <amaiorano@google.com>
Remove the Parser classes from the wgsl and spirv namespaces.
These have been replaced with a Parse() method.
Remove the TypeDeterminer::Run() method, this was not called by tint and
the TypeDeterminer is now non-public API.
Change-Id: I5ddb82768da04398ab3958d1647be44f9fe30c21
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/41840
Commit-Queue: Ben Clayton <bclayton@google.com>
Reviewed-by: dan sinclair <dsinclair@chromium.org>
Change 41302 correctly fixed up Module::Clone(), but this wasn't actually called by the CloneContext, as Module::Clone() returns a new Module, where as the CloneContext needs to clone into an existing Module.
Refactor the code so that this duplicated logic is moved into a single Module::Copy() method.
Fixed: 1177275
Change-Id: Ia8c45ef05e03b2891b5785ee6f425dd01cb989c6
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/41542
Reviewed-by: James Price <jrprice@google.com>
Reviewed-by: dan sinclair <dsinclair@chromium.org>
Commit-Queue: dan sinclair <dsinclair@chromium.org>
Automatically set the storage class to UniformConstant.
Producing an error if an explicit storage class is given is deferred
until downstream users have caught up.
Bug: tint:332
Change-Id: I70e7390dc95d6f578a0fdeb675ca63a8b5b4fa26
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/40160
Commit-Queue: dan sinclair <dsinclair@chromium.org>
Reviewed-by: dan sinclair <dsinclair@chromium.org>
Auto-Submit: James Price <jrprice@google.com>
Add helpers on Program and ProgramBuilder that significantly simplify
usage.
Also demangle - this also reduces a bunch of copy-pasta code.
Change-Id: I6215c346e7f6e49c20aced058a6150603253ed93
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/39342
Commit-Queue: Ben Clayton <bclayton@google.com>
Reviewed-by: dan sinclair <dsinclair@chromium.org>
Will hold the mutable fields that currently reside in the otherwise immutable-AST.
Change the AST string methods to accept a `const semantic::Info&`. This is required as some nodes include type-resolved information in their output strings.
Bug: tint:390
Change-Id: Iba494a9c5645ce2096da0a8cfe63a4309a9d9c3c
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/39003
Commit-Queue: Ben Clayton <bclayton@google.com>
Reviewed-by: dan sinclair <dsinclair@chromium.org>
By putting diagnostics into the program, we can hold all the diagnostic messages for parsing and type determination in one place.
This also means that we can simplify the public WGSL and SPIR-V Parser interfaces to a single function.
Change-Id: Ib6ab5fa180addd45c4aaf0c6b192d47182ffb50a
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/38920
Reviewed-by: dan sinclair <dsinclair@chromium.org>
Commit-Queue: Ben Clayton <bclayton@google.com>
Program is now immutable*, and remains part of the public Tint
interface.
ProgramBuilder is the mutable builder for Programs, and is not part of
the public Tint interface. ast::Builder has been folded into
ProgramBuilder.
Immutable Programs can be cloned into a mutable ProgramBuilder with
Program::CloneAsBuilder().
Mutable ProgramBuilders can be moved into immutable Programs.
* - mostly immutable. It still has a move constructor and move
assignment operator - required for practical usage - and the
semantic information on AST nodes is still mutable.
Bug: tint:390
Change-Id: Ia856c50b1880c2f95c91467a9eef5024cbc380c6
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/38240
Reviewed-by: dan sinclair <dsinclair@chromium.org>
Have this take a SymbolTable instead of a Program.
Program will be split into Program (immutable) and ProgramBuilder (mutable). We'll need Demangler to support both.
Bug: tint:390
Change-Id: I6447dd9674919d4867ed8ba126880cdfd9bf7128
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/38550
Reviewed-by: dan sinclair <dsinclair@chromium.org>
Enforce all places where Dawn passes in or returns a ast::Module, now takes a `const Program* ` or returns a `Program`.
As the end goal of all this is to have immutable Programs, all Program inputs take a pointer instead of moving the actual object.
As consumers of a Program are now all const, we have to const_cast to work around all the places we've been incorrectly mutating a ast::Module.
These const_casts are temporary, and will be fixed in the next set of changes.
Depends on https://dawn-review.googlesource.com/c/dawn/+/38522
Bug: tint:390
Change-Id: Ie05b112b16134937d1b601e9b713ea4ec4e1c677
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/38541
Reviewed-by: dan sinclair <dsinclair@chromium.org>
Internally use BlockAllocator to allocate the types.
When we optimize the allocation patterns of BlockAllocator, this will now benefit both AST nodes and types.
Remove Reset(). It was not used.
Remove type::Manager::Get(std::unique_ptr<type::Type>) - this was used (via Module::unique_type) in one place, which has easily been migrated to using the standard Module::create<>.
Replace all remaining uses of std::unique_ptr<> of types in tests with the standard create<> so we can guarantee uniqueness of the types.
Change-Id: Ib0e1fe94e492b31816450df5de0c839a0aefcb9e
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/38362
Commit-Queue: Ben Clayton <bclayton@google.com>
Reviewed-by: dan sinclair <dsinclair@chromium.org>
Despite `tint::ast::type::Type` being in the AST namespace, these classes are clearly not AST nodes:
* They don't derive from ast::Node
* They're deduplicated by the type manager
* None of the types have an Source - they have no lexical declaration point
* The fact we have `ast::Struct` and `ast::type::Struct` clearly demonstrates what is an AST node, and what is a type.
* We have code scattered in the codebase (TypeDeterminer, writers, etc) that create new types after parsing - so clearly not part of the original syntax tree.
Types in tint are closer to being semantic info, but due to the parse-time generation of types, and tight dependency of ast::Nodes to types, I'd be reluctant to class these as semantic info. Instead, put these into a separate root level `tint::type` namespace and `src/tint` directory.
The fact that types exist in the ast::Module has already caused bugs (https://dawn-review.googlesource.com/c/tint/+/37261). This is a first step in separating out types from the ast::Module.
Bug: tint:390
Change-Id: I8349bbbd1b19597b8e6d51d5cda0890de46ecaec
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/38002
Commit-Queue: Ben Clayton <bclayton@google.com>
Reviewed-by: dan sinclair <dsinclair@chromium.org>
A container and allocator of objects of (or deriving from) the template type `T`.
Objects are allocated by calling Create(), and are owned by the BlockAllocator.
When the BlockAllocator is destructed, all constructed objects are automatically destructed and freed.
Objects held by the BlockAllocator can be iterated over using a View or ConstView.
Use this to hold the ast::Nodes in the ast::Module
This is called BlockAllocator as it can be optimized to hold objects in contiguous memory blocks, which will improve cache coherencey. Currently BlockAllocator is a straight port of the vector-of-unique-ptr, taken from ast::Module.
Change-Id: I4bf4d298aec3c70d2ddf833e2f168416cbb024c0
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/38001
Reviewed-by: dan sinclair <dsinclair@chromium.org>
Commit-Queue: Ben Clayton <bclayton@google.com>
This Cl updates the `storage_buffer` storage class to just be `storage`.
Change-Id: Ibfaecbb0862bd60d39665eb937c0b6300899e177
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/38161
Auto-Submit: dan sinclair <dsinclair@chromium.org>
Reviewed-by: David Neto <dneto@google.com>
Commit-Queue: dan sinclair <dsinclair@chromium.org>
This CL changes the set decoration to a group decoration. We still parse
`set` until downstream users can be updated.
Bug: tint:338
Change-Id: I610d1ed769b3a26c117ad6d875f8a99a3d5b7754
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/37920
Auto-Submit: dan sinclair <dsinclair@chromium.org>
Commit-Queue: Ben Clayton <bclayton@google.com>
Reviewed-by: Ben Clayton <bclayton@google.com>
This CL adds support for the decorated texture storage tokens. The old
_ro_ and _wo_ tokens still exist until downstream users are updated.
Bug: tint:286
Change-Id: I4ddc50be3b22bd3feeab41b3b4fe4ded63e6e59a
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/37780
Auto-Submit: dan sinclair <dsinclair@chromium.org>
Commit-Queue: Ben Clayton <bclayton@google.com>
Reviewed-by: Ben Clayton <bclayton@google.com>
This CL reverts the changes to use the top level namer in the various
backends. This is causing issues when rolling into Dawn in the case
where the tint generator is used to create SPIR-V which is sent to
SPIRV-Cross but then generator and inspector aren't used. The entry
points end up being incorrect as SPIRV-Cross gets the renamed entry
points.
Change-Id: I4749e1d773f2bd9edcce83e63555f07a443d5ca5
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/37342
Reviewed-by: dan sinclair <dsinclair@chromium.org>
Commit-Queue: dan sinclair <dsinclair@chromium.org>
This CL changes the generators so we don't have to move the module into
the generator. This will then allow calling the demangler at a later
point and still having access to the module.
Change-Id: Icad16ddb2b89921cbb174cf7fec520c410139285
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/36942
Reviewed-by: Ben Clayton <bclayton@google.com>
The interface for cloning a module was made significantly more complex in https://dawn-review.googlesource.com/c/tint/+/35502/ as some of the transforms required constructing symbols before the clone. This was temporary solution in 35502 was to copy the symbol table, then construct the new types, then perform the clone. This lead to a really messy callback interface, that was extremely error prone (e.g. lamda-capturing stack variables from the initializer callback that had been unwound).
Instead, clone the symbols as they're encountered. This may produce an entirely different set of identifiers, but no longer ever-grows the symbol list, and keeps the interface clean.
Bug: tint:396
Bug: tint:390
Change-Id: I54affd68ac3b730b649af9b47eba685c8a1d784a
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/35663
Commit-Queue: Ben Clayton <bclayton@google.com>
Reviewed-by: dan sinclair <dsinclair@chromium.org>
This CL strips the context object out of Tint.
Change-Id: Id0dcb9c557b217c03a8d9ac08fc9fe1c799f3fdc
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/34742
Commit-Queue: dan sinclair <dsinclair@chromium.org>
Reviewed-by: Ben Clayton <bclayton@google.com>
Reviewed-by: Ryan Harrison <rharrison@chromium.org>
Deep-clones all `Node`s and `Type`s into a new module.
Instead of writing a million standalone tests that'll only ever test the
existing fields of each type, I've opted to write the tests using
wgsl<->ast<->wgsl conversion. This means the tests require the enabling
of TINT_BUILD_WGSL_READER and TINT_BUILD_WGSL_WRITER, but I believe this
is much easier to maintain.
I'm aware there are probably gaps in the tests, and that even full
coverage is likely to rapidly rot, so I've also added
fuzzers/tint_ast_clone_fuzzer.cc - a fuzzer based test that ensures that
all AST modules can be cloned with identical reproduction.
I've run this across 100 cores of a 3990x for 4 hours, fixing the
single issue it detected.
Note: Expressions do not currently clone their `TypeManager` determined
types. This is for two reasons:
(a) This initial CL is mahoosive enough.
(b) I'm uncertain whether we actually want to clone this info, or to
re-run the `TypeDeterminer` after each AST transform. Maybe it should
be optional. Time will tell.
Fixed: tint:307
Change-Id: Id90fab06aaa740c805d12b66f3f11d1f452c6805
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/33300
Commit-Queue: Ben Clayton <bclayton@google.com>
Reviewed-by: dan sinclair <dsinclair@chromium.org>
Reviewed-by: David Neto <dneto@google.com>
