Fix OverwritesInput metadata for ops that allocate new values#580
Open
eric wants to merge 1 commit intogorgonia:masterfrom
Open
Fix OverwritesInput metadata for ops that allocate new values#580eric wants to merge 1 commit intogorgonia:masterfrom
eric wants to merge 1 commit intogorgonia:masterfrom
Conversation
transposeOp, reshapeOp, maxOp, and sumOp all declared OverwritesInput() == 0 and ReturnsPtr() == true, telling the register allocator they modify their input in-place. In reality, their Do() methods allocate new values: - transposeOp.Do() calls tensor.Transpose() which allocates via SafeT - reshapeOp.Do() calls ShallowClone() creating a new tensor header - maxOp/sumOp.Do() call reductionDo() which returns new tensors This mismatch caused the register allocator to share input/output registers for these ops. While not currently causing failures (the liveness analysis prevents reuse of live registers), it violates the OverwritesInput contract and could amplify future liveness bugs. Ops that correctly declare OverwritesInput >= 0 are unchanged: - elemBinOp: has UsePreallocDo that writes into preallocated memory - elemUnaryOp: has UnsafeDo that modifies in-place - repeatOp: has UsePreallocDo + CallsExtern=true - sliceIncrOp: has UsePreallocDo + CallsExtern=true
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
transposeOp, reshapeOp, maxOp, and sumOp all declared OverwritesInput() == 0 and ReturnsPtr() == true, telling the register allocator they modify their input in-place. In reality, their Do() methods allocate new values:
This mismatch caused the register allocator to share input/output registers for these ops. While not currently causing failures (the liveness analysis prevents reuse of live registers), it violates the OverwritesInput contract and could amplify future liveness bugs.
Ops that correctly declare OverwritesInput >= 0 are unchanged: