[compiler] Initialize abstract values of places in InferReferenceEffects #30973
+77
−30
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
Test Plan: In this diff, we now populate the abstractValue field of places during the InferReferenceEffects pass. The value we populate it with is the same value that we use internally in this pass, but it now will remain accessible to downstream phases as part of the Place. For phis specifically, we need to do a bit of extra work to compute the appropriate value for the phi, since InferReferenceEffects currently doesn't need to infer a value for phis themselves, only values downstream of them. However, the value we compute should correspond to the value available downstream of the phi. This changes the error message for one todo test case, because we now are querying for the valueKind earlier in the pass, and hitting an invariant violation as a result of that rather than a later invariant violation. [ghstack-poisoned]
|
The latest updates on your projects. Learn more about Vercel for Git ↗︎
|
This was referenced Sep 16, 2024
…ferenceEffects" Test Plan: In this diff, we now populate the abstractValue field of places during the InferReferenceEffects pass. The value we populate it with is the same value that we use internally in this pass, but it now will remain accessible to downstream phases as part of the Place. For phis specifically, we need to do a bit of extra work to compute the appropriate value for the phi, since InferReferenceEffects currently doesn't need to infer a value for phis themselves, only values downstream of them. However, the value we compute should correspond to the value available downstream of the phi. This changes the error message for one todo test case, because we now are querying for the valueKind earlier in the pass, and hitting an invariant violation as a result of that rather than a later invariant violation. [ghstack-poisoned]
mofeiZ
reviewed
Sep 16, 2024
| reason: `[hoisting] Expected value kind to be initialized`, | ||
| description: `${printPlace(place)}`, | ||
| reason: `[hoisting] Expected value kind to be initialized in call to kind()`, | ||
| description: `${place.identifier.id}`, |
There was a problem hiding this comment.
Nit: printing the entire identifier / place might be more descriptive (or is this change intentional?)
josephsavona
approved these changes
Sep 16, 2024
…ferenceEffects" Test Plan: In this diff, we now populate the abstractValue field of places during the InferReferenceEffects pass. The value we populate it with is the same value that we use internally in this pass, but it now will remain accessible to downstream phases as part of the Place. For phis specifically, we need to do a bit of extra work to compute the appropriate value for the phi, since InferReferenceEffects currently doesn't need to infer a value for phis themselves, only values downstream of them. However, the value we compute should correspond to the value available downstream of the phi. This changes the error message for one todo test case, because we now are querying for the valueKind earlier in the pass, and hitting an invariant violation as a result of that rather than a later invariant violation. [ghstack-poisoned]
…ferenceEffects" Test Plan: In this diff, we now populate the abstractValue field of places during the InferReferenceEffects pass. The value we populate it with is the same value that we use internally in this pass, but it now will remain accessible to downstream phases as part of the Place. For phis specifically, we need to do a bit of extra work to compute the appropriate value for the phi, since InferReferenceEffects currently doesn't need to infer a value for phis themselves, only values downstream of them. However, the value we compute should correspond to the value available downstream of the phi. This changes the error message for one todo test case, because we now are querying for the valueKind earlier in the pass, and hitting an invariant violation as a result of that rather than a later invariant violation. [ghstack-poisoned]
…ferenceEffects" Test Plan: In this diff, we now populate the abstractValue field of places during the InferReferenceEffects pass. The value we populate it with is the same value that we use internally in this pass, but it now will remain accessible to downstream phases as part of the Place. For phis specifically, we need to do a bit of extra work to compute the appropriate value for the phi, since InferReferenceEffects currently doesn't need to infer a value for phis themselves, only values downstream of them. However, the value we compute should correspond to the value available downstream of the phi. This changes the error message for one todo test case, because we now are querying for the valueKind earlier in the pass, and hitting an invariant violation as a result of that rather than a later invariant violation. [ghstack-poisoned]
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.
Stack from ghstack (oldest at bottom):
Test Plan:
In this diff, we now populate the abstractValue field of places during the InferReferenceEffects pass. The value we populate it with is the same value that we use internally in this pass, but it now will remain accessible to downstream phases as part of the Place.
For phis specifically, we need to do a bit of extra work to compute the appropriate value for the phi, since InferReferenceEffects currently doesn't need to infer a value for phis themselves, only values downstream of them. However, the value we compute should correspond to the value available downstream of the phi.
This changes the error message for one todo test case, because we now are querying for the valueKind earlier in the pass, and hitting an invariant violation as a result of that rather than a later invariant violation.