Sema: Allow non-final classes to satisfy properties and subscripts with covariant Self #34005
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
Collaborator
Author
|
Should I be adding some execution tests? |
…th covariant Self
AnthonyLatsis
force-pushed
the
coself
branch
from
0a8ca5b to
103a821
Compare
slavapestov
reviewed
Sep 21, 2020
Contributor
slavapestov
left a comment
slavapestov
left a comment
There was a problem hiding this comment.
Sounds good! Runtime tests would be nice to make sure the generated witness thunk is valid and recovers the correct Self type.
This ensures that we erase opened archetypes before closing an existential result
Collaborator
Author
|
@swift-ci please smoke test |
AnthonyLatsis
force-pushed
the
coself
branch
from
9ab98ba to
9f58968
Compare
AnthonyLatsis
commented
Sep 23, 2020
Comment on lines
+97
to
+108
| // CHECK: [[PCOPY_ADDR:%[0-9]+]] = open_existential_addr immutable_access [[P]] : $*P to $*@opened([[N:".*"]]) P | ||
| // CHECK: [[P_RESULT:%[0-9]+]] = alloc_stack $P | ||
| // CHECK: [[PCOPY_ADDR_1:%[0-9]+]] = alloc_stack $@opened([[N]]) P | ||
| // CHECK: copy_addr [[PCOPY_ADDR]] to [initialization] [[PCOPY_ADDR_1]] : $*@opened([[N]]) P | ||
| // CHECK: [[P_P_GETTER:%[0-9]+]] = witness_method $@opened([[N]]) P, #P.p!getter : {{.*}}, [[PCOPY_ADDR]]{{.*}} : $@convention(witness_method: P) <τ_0_0 where τ_0_0 : P> (@in_guaranteed τ_0_0) -> @out τ_0_0 | ||
| // CHECK: [[P_RESULT_ADDR2:%[0-9]+]] = init_existential_addr [[P_RESULT]] : $*P, $@opened([[N]]) P | ||
| // CHECK: apply [[P_P_GETTER]]<@opened([[N]]) P>([[P_RESULT_ADDR2]], [[PCOPY_ADDR_1]]) : $@convention(witness_method: P) <τ_0_0 where τ_0_0 : P> (@in_guaranteed τ_0_0) -> @out τ_0_0 | ||
| // CHECK: destroy_addr [[PCOPY_ADDR_1]] : $*@opened([[N]]) P | ||
| // CHECK: destroy_addr [[P_RESULT]] : $*P | ||
| // CHECK: dealloc_stack [[PCOPY_ADDR_1]] : $*@opened([[N]]) P | ||
| // CHECK: dealloc_stack [[P_RESULT]] : $*P | ||
| _ = p.p |
Collaborator
Author
There was a problem hiding this comment.
If this output happens to be sane, why do we need PCOPY_ADDR_1 in addition to PCOPY_ADDR?
Contributor
There was a problem hiding this comment.
I think it's just SILGen inserting an unnecessary copy. It does that sometimes.
slavapestov
approved these changes
Sep 24, 2020
Collaborator
Author
|
@swift-ci please smoke test |
Collaborator
Author
|
@swift-ci please smoke test Windows |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
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.
Hopefully, whatever complications were impeding this in 926e371 are long gone.
Bonus: now we can use protocols with such requirements as types (iff the requirement was the reason we previously couldn't).