JIT: capture all write barrier helper addresses for SPMI #97535
Merged
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
When varying CSEs we can sometimes alter the write barrier that is needed. In particular if we CSE a heap address computation we may lose track of the fact that an indir is writing to the heap, and so change which write barrier needs to be used. See dotnet#97534. Even if that's fixed it seems like we still might change our minds for various reasons, so when running under SPMI, just collect all the possible write barrier helper addresses.
dotnet-issue-labeler
bot
added
the
area-CodeGen-coreclr
ghost
assigned 
|
Tagging subscribers to this area: @JulieLeeMSFT, @jakobbotsch Issue DetailsWhen varying CSEs we can sometimes alter the write barrier that is needed. In particular if we CSE a heap address computation we may lose track of the fact that an indir is writing to the heap, and so change which write barrier needs to be used. See #97534. Even if that's fixed it seems like we still might change our minds for various reasons, so when running under SPMI, just collect all the possible write barrier helper addresses.
|
|
@jakobbotsch PTAL |
BruceForstall
approved these changes
Jan 26, 2024
src/coreclr/jit/codegeninterface.h
Outdated
| @@ -173,6 +173,11 @@ class CodeGenInterface | |||
| bool genUseOptimizedWriteBarriers(GenTreeStoreInd* store); | |||
| CorInfoHelpFunc genWriteBarrierHelperForWriteBarrierForm(GCInfo::WriteBarrierForm wbf); | |||
|
|
|||
| #ifdef DEBUG | |||
| bool genWriteBarrierUsed; | |||
| void WriteBarrierExtraSuperPmiQueries(); | |||
There was a problem hiding this comment.
Looks like you were going to use a function, but then you didn't?
jakobbotsch
approved these changes
Jan 26, 2024
AndyAyersMS
added a commit
to AndyAyersMS/runtime
that referenced
this pull request
Jan 29, 2024
Fixes an issue introduced by dotnet#97535. AOT hosts don't support all the possible write barriers and blow things up if the JIT asks for ones they don't support.
jakobbotsch
pushed a commit
that referenced
this pull request
Jan 30, 2024
Fixes an issue introduced by #97535. AOT hosts don't support all the possible write barriers and blow things up if the JIT asks for ones they don't support. Also add back the initial semicolon to the metrics line
Sign up for free
to subscribe to this conversation on GitHub.
Already have an account?
Sign in.
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.
When varying CSEs we can sometimes alter the write barrier that is needed. In particular if we CSE a heap address computation we may lose track of the fact that an indir is writing to the heap, and so change which write barrier needs to be used.
See #97534.
Even if that's fixed it seems like we still might change our minds for various reasons, so when running under SPMI, just collect all the possible write barrier helper addresses.