Migrate all uses to use this and the new `unique_type<T>()` and `types()` methods.
Remove the `type_mgr()` accessor. `TypeManager` is now an implementation detail of the module, allowing us to unify the allocation of types and nodes (if we so wish).
Fixes: tint:337
Bug: tint:307
Change-Id: I233fa9dc73d60515dd721f02ea7ba089ef7d374f
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/33667
Commit-Queue: Ben Clayton <bclayton@google.com>
Reviewed-by: David Neto <dneto@google.com>
First step to moving this to the `ast::Module`.
Also remove a bunch of redundant includes to `type_manager.h` as this is already included in `context.h`
Bug: tint:307
Bug: tint:337
Change-Id: Ic4baffa7b76ddefa29f56f758c25b1003ef40888
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/33665
Commit-Queue: Ben Clayton <bclayton@google.com>
Reviewed-by: David Neto <dneto@google.com>
Synthesize Tint types for handle variables and function parameters.
Bug: tint:109
Change-Id: I3c155e03b154c5ebf46e79c96a6e2b054dbca9b4
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/33341
Reviewed-by: David Neto <dneto@google.com>
Reviewed-by: dan sinclair <dsinclair@chromium.org>
Commit-Queue: David Neto <dneto@google.com>
Auto-Submit: David Neto <dneto@google.com>
The reason being that some tests called parse() twice, which will silently destruct the first parser.
Once the `Module` owns the AST nodes, the second call will end up deleting all the AST nodes. Tests would then perform use-after-free for the AST nodes belonging to the first parser / module.
There's no reason why the unique_ptr can't be returned, which is cleaner overall.
Bug: tint:335
Change-Id: I7ff2e9777a7ebeb76702f806294fe4c2c49bd7c9
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/33241
Auto-Submit: Ben Clayton <bclayton@google.com>
Reviewed-by: dan sinclair <dsinclair@chromium.org>
Commit-Queue: dan sinclair <dsinclair@chromium.org>
Change-Id: I2fed4cf2a052686e2121066588686f05907a169f
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/33200
Commit-Queue: Ben Clayton <bclayton@google.com>
Commit-Queue: dan sinclair <dsinclair@chromium.org>
Auto-Submit: Ben Clayton <bclayton@google.com>
Reviewed-by: dan sinclair <dsinclair@chromium.org>
It's valid to look for but not find an underlying memory object
declaration.
Bug: tint:109
Change-Id: I7296d79550a50050d2438996dc3e0c8d09a6babd
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/33140
Auto-Submit: David Neto <dneto@google.com>
Reviewed-by: dan sinclair <dsinclair@chromium.org>
Commit-Queue: dan sinclair <dsinclair@chromium.org>
We convert types of samplers, images, and sampled images entirely
differently, but still find it useful to generalize ParserImpl::ConvertType
to cover them.
Fake it: Make ConvertType return void for them.
Bug: tint:109
Change-Id: I0982eb987d0914db8227bc0fce552989831129b9
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/33020
Commit-Queue: David Neto <dneto@google.com>
Commit-Queue: dan sinclair <dsinclair@chromium.org>
Auto-Submit: David Neto <dneto@google.com>
Reviewed-by: dan sinclair <dsinclair@chromium.org>
Change-Id: I56a314d0c4eb1f1dfdb8804024a5824e5456356c
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/32940
Commit-Queue: David Neto <dneto@google.com>
Commit-Queue: dan sinclair <dsinclair@chromium.org>
Auto-Submit: David Neto <dneto@google.com>
Reviewed-by: dan sinclair <dsinclair@chromium.org>
Remove all redundant std::move()s. I've also removed calls to
std::move() in tests, even if they act as an optimization. This is for
two reasons:
(a) Performance is not important for testing, and this helps with
readability.
(b) A whole bunch tests were relying on std::move() clearing vectors so
they can be repopulated and used again. This is undefined behavior:
> Objects of types defined in the C++ standard library may be moved from
> (12.8). Move operations may be explicitly specified or implicitly
> generated. Unless otherwise specified, such moved-from objects shall
> be placed in a valid but unspecified state.
All of these UB cases have been fixed.
Removed all duplicate variables left over from:
`auto* foo_ptr = foo.get()`
which became:
`auto* foo_ptr = foo`
Bug: tint:322
Change-Id: Ibd08a2379671382320fd4d8da296ccc6a378b8af
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/32900
Commit-Queue: Ben Clayton <bclayton@google.com>
Reviewed-by: dan sinclair <dsinclair@chromium.org>
This is a minimal effort to fix up the code. There's substantial code
cleanup which can now be done, which is done in the next change.
Bug: tint:322
Change-Id: Iafcf5e814837d9534889e8c21333de4931a19cfa
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/32864
Commit-Queue: dan sinclair <dsinclair@chromium.org>
Reviewed-by: dan sinclair <dsinclair@chromium.org>
create() is currently just a simple forwarder to std::make_unique<>, but
will be later replaced with a function that returns a raw pointer,
and owned by the context.
Bug: tint:322
Change-Id: I281fe91864a98365db5ccd40e264d042e6476172
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/32861
Reviewed-by: dan sinclair <dsinclair@chromium.org>
Commit-Queue: Ben Clayton <bclayton@google.com>
create() is currently just a simple forwarder to std::make_unique<>, but
will be later replaced with a function that returns a raw pointer,
and owned by the context.
Bug: tint:322
Change-Id: I72558482c4b6aff7a655087ee010b3e16b006192
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/32860
Commit-Queue: Ben Clayton <bclayton@google.com>
Reviewed-by: dan sinclair <dsinclair@chromium.org>
In a near-future change, AST nodes, such as ast::BlockStatement will no longer
be std::unique_ptrs, and will have to be constructed and owned by an external
class. This means AST nodes can no longer allocate default child nodes.
Bug: tint:322
Change-Id: I36a1cf55c31a1dabccde272b2be415f98c16b18f
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/32677
Commit-Queue: Ben Clayton <bclayton@google.com>
Commit-Queue: David Neto <dneto@google.com>
Reviewed-by: David Neto <dneto@google.com>
Reviewed-by: dan sinclair <dsinclair@chromium.org>
This CL adds the result_type type_name into the AST dump if available.
Bug: tint:310, tint:308
Change-Id: Iea678fd4f7a2dadbfca86f29043c75459c421cb3
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/32780
Reviewed-by: Sarah Mashayekhi <sarahmashay@google.com>
Commit-Queue: dan sinclair <dsinclair@chromium.org>
In a near-future change, AST nodes, such as ast::BlockStatement will no longer
be std::unique_ptrs, and will have to be constructed and owned by an external
class. This means AST nodes can no longer allocate default child nodes.
Bug: tint:322
Change-Id: I2a571d0a4727d6dc3d6c38e8b6602e131292f49c
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/32676
Reviewed-by: dan sinclair <dsinclair@chromium.org>
Commit-Queue: dan sinclair <dsinclair@chromium.org>
In a near-future change, AST nodes, such as ast::BlockStatement will no longer
be std::unique_ptrs, and will have to be constructed and owned by an external
class. This means AST nodes can no longer allocate default child nodes.
Bug: tint:322
Change-Id: I3db9b3c037896f07b84b14b7b8d4da0f066b69b0
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/32679
Commit-Queue: Ben Clayton <bclayton@google.com>
Reviewed-by: dan sinclair <dsinclair@chromium.org>
In a near-future change, AST nodes, such as ast::BlockStatement will no longer
be std::unique_ptrs, and will have to be constructed and owned by an external
class. This means AST nodes can no longer allocate default child nodes.
Bug: tint:322
Change-Id: Iddb5605b9bc0de80ad2710ced0e429f89410af2f
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/32675
Commit-Queue: dan sinclair <dsinclair@chromium.org>
Reviewed-by: dan sinclair <dsinclair@chromium.org>
A common base class is required to move from std::unique_ptr<> to raw pointers for AST types.
Also unifies a bunch of similar APIs.
Bug: tint:322
Change-Id: If829f8c3f22069adf62751365f1f1eeb646aba08
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/32660
Commit-Queue: Ben Clayton <bclayton@google.com>
Commit-Queue: dan sinclair <dsinclair@chromium.org>
Reviewed-by: dan sinclair <dsinclair@chromium.org>
Use synchronization tokens to ensure the parser can resynchronize on error.
Bug: tint:282
Change-Id: I8bb033f8a723eb8f2bc029e1ffc8350174c964e2
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/32284
Commit-Queue: dan sinclair <dsinclair@chromium.org>
Reviewed-by: dan sinclair <dsinclair@chromium.org>
I feel this reads better.
Bug: tint:282
Change-Id: I04fe42a0347ea050e93e5cf5ccae7259bc79bb58
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/32283
Commit-Queue: dan sinclair <dsinclair@chromium.org>
Reviewed-by: dan sinclair <dsinclair@chromium.org>
A number of tests check the first encountered error is correct.
As the parser currently aborts after the first error, later errors are ignored.
Once the parser supports resynchronization, we'll emit multiple error messages, and these tests will start to fail. Fix them now.
Bug: tint:282
Change-Id: If8d0c41f030c652500b2e3b7284297b7a448d23e
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/32282
Commit-Queue: dan sinclair <dsinclair@chromium.org>
Reviewed-by: dan sinclair <dsinclair@chromium.org>
Split out statments that are non-block (non-loops, etc) into a separate function.
These all end with a semi-colon, which is important for resynchronization on errors (coming up in another change).
Bug: tint:282
Change-Id: I0e58c4938f2bbe859dc6ffb8dcd45c8cf26101da
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/32281
Commit-Queue: dan sinclair <dsinclair@chromium.org>
Reviewed-by: dan sinclair <dsinclair@chromium.org>
Reduces code, paves the way for multiple errors with resynchronization points.
Bug: tint:282
Change-Id: I68018ea8cabe4ec347afa21d1220126d6348d3d1
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/32280
Commit-Queue: dan sinclair <dsinclair@chromium.org>
Reviewed-by: dan sinclair <dsinclair@chromium.org>
For values of type T* and std::unique_ptr<T>.
This allows us to replace all occurances of `res.value->member` with: `res->member`, which also asserts that `res` is not in an error state.
Brings the verbosity back down to pre-expect and pre-maybe levels.
Bug: tint:282
Change-Id: Ib00018affca53ac5e71ee2140e7e0cd607b83715
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/32141
Commit-Queue: Ben Clayton <bclayton@google.com>
Reviewed-by: dan sinclair <dsinclair@chromium.org>
And use it for the non-ParserImpl::expect_xxx() methods.
Another step towards supporting multiple error messages, as the caller can now test to see if the specific call errored, or didn't match, instead of using a global error state.
Makes reading the control flow conditionals a bit easier too.
Change-Id: Ie8627b8499ec9079167965da2a566401cd6bd903
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/32102
Commit-Queue: Ben Clayton <bclayton@google.com>
Reviewed-by: dan sinclair <dsinclair@chromium.org>
And use it for the ParserImpl::expect_xxx() methods.
This is the first step towards supporting multiple error messages, as
the caller can now test to see if the specific call errored, instead of
using a global error state.
Also cleans up a bunch of code.
Bug: tint:282
Change-Id: I5e39fc33bd1e16620cee80d27fa728bc2af3387e
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/32101
Commit-Queue: Ben Clayton <bclayton@google.com>
Reviewed-by: dan sinclair <dsinclair@chromium.org>
Reduces code. Keeps things more consistent.
Bug: tint:282
Change-Id: Iff280880eb033fbcee4c6095c2da2d4af06835b5
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/32103
Commit-Queue: Ben Clayton <bclayton@google.com>
Reviewed-by: dan sinclair <dsinclair@chromium.org>
All the call sites of `storage_class()` add their own error handling, so transform this into `expect_storage_class()`.
Also makes error messages more consistent.
Bug: tint:282
Change-Id: I5131acd84f91fc7494ed6b90965853b7d0fc37f0
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/32104
Commit-Queue: Ben Clayton <bclayton@google.com>
Reviewed-by: dan sinclair <dsinclair@chromium.org>
All the call sites of `variable_ident_decl()` add their own error handling, so transform this into `expect_variable_ident_decl()`.
Also makes error messages more consistent.
Bug: tint:282
Change-Id: I0b5ac984018ba78896ddec0320636f5b5c4ad0b2
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/32100
Reviewed-by: dan sinclair <dsinclair@chromium.org>
Commit-Queue: dan sinclair <dsinclair@chromium.org>
The `expect_` prefixes now clearly indicate when a method will internally error, or produce a valid AST object.
Verified by code coverage.
Bug: tint:282
Change-Id: Icbdae9db02bd48c69aec010a4f8fdc5a496125f8
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/32002
Reviewed-by: dan sinclair <dsinclair@chromium.org>
Commit-Queue: Ben Clayton <bclayton@google.com>
... for those that will internally error if the grammar does not match,
otherwise will always return a valid AST object.
This helps identify whether the caller is expected to error or not.
Bug: tint:282
Change-Id: Ied94f717526a63033f2e6c9e94fca43dbf0b8f05
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/32001
Commit-Queue: Ben Clayton <bclayton@google.com>
Commit-Queue: dan sinclair <dsinclair@chromium.org>
Reviewed-by: dan sinclair <dsinclair@chromium.org>
This Cl updates the storage textures to have `storage` in the name. So,
`texture_ro_1d` -> `texture_storage_ro_1d`
Bug: tint:286
Change-Id: I0a9ea02f15de2681d64e272cb42be51a940b6a13
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/31840
Commit-Queue: dan sinclair <dsinclair@chromium.org>
Commit-Queue: Ryan Harrison <rharrison@chromium.org>
Reviewed-by: Ryan Harrison <rharrison@chromium.org>
Parse all decorations with the same function, and create a
`ast::DecorationList`. Once the parser has progressed to the consumer of
the decorations, we attempt to downcast these to the required type,
erroring if they're the wrong kind.
While the error message could be improved further, this greatly reduces
the headscratching around crbug.com/tint/291.
Also knocks another 223 lines off parser_impl.cc.
Bug: tint:291
Bug: tint:282
Change-Id: I7506faeb56d876e5446d900c7c134669a9db6409
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/31660
Commit-Queue: Ben Clayton <bclayton@google.com>
Commit-Queue: dan sinclair <dsinclair@chromium.org>
Reviewed-by: dan sinclair <dsinclair@chromium.org>
`expect_block()` takes a start and end token, along with a function to parse a lexical block body.
This reduces code, keeps error messages consistent, and also gives us a future place to try resynchronising the parser so we can have more than one error emitted.
`expect_paren_block()` and `expect_brace_block()` are convenience helpers for providing the start and end tokens for common block types.
Bug: tint:282
Change-Id: I432a0301727b131a6fce875687b952dfc6889a4b
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/31736
Commit-Queue: Ben Clayton <bclayton@google.com>
Reviewed-by: dan sinclair <dsinclair@chromium.org>
The only place that calls `pipeline_stage()` expects a stage to exist, so follow the `expect_` pattern.
Bug: tint:282
Change-Id: Ie18d24ed25a5f882e66a8e553e53b4fb52dcf6fa
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/31734
Commit-Queue: Ben Clayton <bclayton@google.com>
Reviewed-by: dan sinclair <dsinclair@chromium.org>
