[hannk] Clean up aliasing #6321
Closed
Conversation
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
Minor optimization to avoid identical overrides for OpGroup in multiple places (when all we want to process are the leaf ops)
The code for aliasing tensors was janky. This cleans it up and makes a clear distinction between aliasing done to overlay buffers with crop-and-translate, vs the aliasing done when we reshape tensors. We no longer allow a given tensor to do both of these, and we give preference to Reshape aliasing first. Also fixed the bounds calculation in lower_tflite_fullyconnected() to preserve the invariants expected, and added a testcase that previously failed.
|
Gentle review ping |
Recent updates to Chrome/FF/etc require serving pages with certain headers if you want to allow SharedArrayBuffer (and thus, multithreading); update the README and add a helper script to make this documented and easier to manage. Also add $(CXXFLAGS) to the benchmark targets so they enable C++11 or more (which HalideBuffer.h now requires).
|
Gentle review ping |
steven-johnson
added a commit
that referenced
this pull request
Oct 28, 2021
The code for aliasing tensors was janky. This cleans it up and makes a clear distinction between aliasing done to overlay buffers with crop-and-translate, vs the aliasing done when we reshape tensors. We no longer allow a given tensor to do both of these, and we give preference to Reshape aliasing first. (Cherry-picked from #6321)
|
(Cherry-picking the pieces I need from this; it won't land as-is but I don't want to close it until all the pieces are done) |
steven-johnson
added a commit
that referenced
this pull request
Nov 1, 2021
* wip * [hannk] Clean up aliasing (v2) The code for aliasing tensors was janky. This cleans it up and makes a clear distinction between aliasing done to overlay buffers with crop-and-translate, vs the aliasing done when we reshape tensors. We no longer allow a given tensor to do both of these, and we give preference to Reshape aliasing first. (Cherry-picked from #6321) * Move alias_type into shared ptr
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
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.
The code for aliasing tensors was janky. This cleans it up and makes a clear distinction between aliasing done to overlay buffers with crop-and-translate, vs the aliasing done when we reshape tensors. We no longer allow a given tensor to do both of these, and we give preference to Reshape aliasing first.
Also fixed the bounds calculation in lower_tflite_fullyconnected() to preserve the invariants expected, and added a testcase that previously failed.