[wasm] Generate bailouts for switch opcodes in jiterpreter instead of aborting #107323
Conversation
This file contains 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
…es instead of aborting the trace Better diagnostic output when jiterp hits an invalid opcode
|
Tagging subscribers to this area: @BrzVlad, @kotlarmilos |
pavelsavara
approved these changes
Sep 4, 2024
This was referenced Sep 4, 2024
|
should we backport this? |
|
We can, but it's simultaneously not-zero-risk and probably not high-value |
radekdoulik
pushed a commit
to radekdoulik/runtime
that referenced
this pull request
Sep 6, 2024
…es instead of aborting the trace (dotnet#107323) Better diagnostic output when jiterp hits an invalid opcode
jtschuster
pushed a commit
to jtschuster/runtime
that referenced
this pull request
Sep 17, 2024
…es instead of aborting the trace (dotnet#107323) Better diagnostic output when jiterp hits an invalid opcode
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Outdated suggestions cannot be applied.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.
Right now none of the jiterpreter support code knows how to decode switch opcodes since they're variable length - hitting one causes codegen to just halt. This is undesirable since the switch might be in an unlikely-to-execute block, i.e. an error handler.
As a first step this PR implements the size calculation algorithm for interpreter switches and turns them into an unconditional bailout, so codegen can continue past them.
In the future to implement switches we'd need to update the logic that finds branch targets so it can see 'into' switches and enumerate all their potential branch targets, and then pair that with jiterp codegen that produces a wasm jump table with the right displacements. We'd probably want to limit that to switches of a certain size and only do it after the trace size limit has been raised.