6 Commits

Author SHA1 Message Date
Mehdi Amini
41f65666f6
[MLIR] Revamp RegionBranchOpInterface (#165429)
This is still somehow a WIP, we have some issues with this interface
that are not trivial to solve. This patch tries to make the concepts of
RegionBranchPoint and RegionSuccessor more robust and aligned with their
definition:
- A `RegionBranchPoint` is either the parent (`RegionBranchOpInterface`)
op or a `RegionBranchTerminatorOpInterface` operation in a nested
region.
- A `RegionSuccessor` is either one of the nested region or the parent
`RegionBranchOpInterface`

Some new methods with reasonnable default implementation are added to
help resolving the flow of values across the RegionBranchOpInterface.

It is still not trivial in the current state to walk the def-use chain
backward with this interface. For example when you have the 3rd block
argument in the entry block of a for-loop, finding the matching operands
requires to know about the hidden loop iterator block argument and where
the iterargs start. The API is designed around forward-tracking of the
chain unfortunately.

Try to reland #161575 ; I suspect a buildbot incremental build issue.
2025-10-28 09:53:56 -07:00
Mehdi Amini
e3c547179f
Revert " [MLIR] Revamp RegionBranchOpInterface " (#165356)
Reverts llvm/llvm-project#161575

Broke Windows on ARM buildbot build, needs investigations.
2025-10-28 01:06:14 -07:00
Mehdi Amini
ab1fd21b54
[MLIR] Revamp RegionBranchOpInterface (#161575)
This is still somehow a WIP, we have some issues with this interface
that are not trivial to solve. This patch tries to make the concepts of
RegionBranchPoint and RegionSuccessor more robust and aligned with their
definition:
- A `RegionBranchPoint` is either the parent (`RegionBranchOpInterface`)
op or a `RegionBranchTerminatorOpInterface` operation in a nested
region.
- A `RegionSuccessor` is either one of the nested region or the parent
`RegionBranchOpInterface`

Some new methods with reasonnable default implementation are added to
help resolving the flow of values across the RegionBranchOpInterface.

It is still not trivial in the current state to walk the def-use chain
backward with this interface. For example when you have the 3rd block
argument in the entry block of a for-loop, finding the matching operands
requires to know about the hidden loop iterator block argument and where
the iterargs start. The API is designed around forward-tracking of the
chain unfortunately.
2025-10-28 07:47:26 +00:00
Adrian Kuegel
b3a2208c56 [mlir] Apply ClangTidy fixes.
- Prefer to check empty() instead of size() == 0.
- Remove unused using declarations.
2024-09-17 11:02:20 +00:00
srcarroll
6b4c122847
[mlir][loops] Add getters for multi dim loop variables in LoopLikeOpInterface (#94516)
This patch adds `getLoopInductionVars`, `getLoopLowerBounds`,
`getLoopBounds`, `getLoopSteps` interface methods to
`LoopLIkeOpInterface`. The corresponding single value versions have been
moved to shared class declaration and have been implemented based on the
new interface methods.
2024-06-07 18:25:43 -05:00
Jorn Tuyls
286bd42a7a
[mlir] Extract forall_to_for logic into reusable function and add pass (#89636)
This PR extracts the existing `scf.forall` to `scf.for` conversion logic
inside a transform op (https://github.com/llvm/llvm-project/pull/65474)
into a standalone function which can be used in other transformations
and adds a `scf-forall-to-for` pass.
2024-04-24 09:57:48 -07:00