Skip to content

Slava Pestov
slavapestov

Organizations

@factor @apple
slavapestov commented on pull request apple/swift#1229
@slavapestov

I vaguely recall we used to have this and it got taken out. @lattner ?

slavapestov commented on pull request apple/swift#1229
@slavapestov

Formatting -- line up with ( on previous line

@slavapestov
slavapestov merged pull request apple/swift#1075
@slavapestov
[Sema] implement better type circularity check
3 commits with 292 additions and 74 deletions
slavapestov commented on commit apple/swift@6a254042ee
@slavapestov

@nadavrot It is not really a fix because default implementations are not fully plumbed through yet. There's no regression here.

@slavapestov
slavapestov commented on pull request apple/swift#1075
@slavapestov

@dduan No worries! If you find some time to rip out the remaining dead code from SIL too, that would be great.

slavapestov commented on pull request apple/swift#1075
@slavapestov

@swift-ci Please test

slavapestov commented on pull request apple/swift#1075
@slavapestov

Looks great, thanks for sticking with this!

@slavapestov
slavapestov commented on pull request apple/swift#1075
@slavapestov

Ah, I see now, I was just misunderstanding the code. It might be clearer if you always pass baseType in, and just set it to getDeclaredTypeOfContex…

slavapestov commented on pull request apple/swift#1075
@slavapestov

This parameter should be named baseType for consistency with the previous one

slavapestov commented on pull request apple/swift#1075
@slavapestov

I think so, yeah. Thanks!

slavapestov commented on pull request apple/swift#1075
@slavapestov

This isn't quite the right approach. So say we have a struct decl S<T>, with a member of type H<T>. If the base type is S<Int>, then getTypeOfMembe…

slavapestov commented on pull request apple/swift#1075
@slavapestov

You should always have a baseType

slavapestov commented on pull request apple/swift#1075
@slavapestov

You should be doing 'baseType->getTypeOfMember()' here -- the idea is to apply the baseType substitutions to the member's type

slavapestov commented on pull request apple/swift#1075
@slavapestov

Line up with opening (

slavapestov commented on pull request apple/swift#1075
@slavapestov

We don't like using gcc extensions like ?:, please write this out with the standard trinary operator

slavapestov commented on pull request apple/swift#1075
@slavapestov

Is this necessary?

slavapestov commented on pull request apple/swift#1075
@slavapestov

Typo: unwrap

slavapestov commented on pull request apple/swift#1075
@slavapestov

Don't optionals desugar to enums?

slavapestov commented on pull request apple/swift#1075
@slavapestov

Typo: "a circularity"

slavapestov commented on pull request apple/swift#1075
@slavapestov

Line these up with the opening (

slavapestov commented on commit apple/swift@34904f9edb
@slavapestov

Michael, it appears that this merge reverted a bunch of recent changes to master. Look at this diff: git diff 11978a6..ee7f011

@slavapestov
  • @harlanhaskins b4c7678
    [coverage] Documented modes for SWIFT_ANALYZE_CODE_COVERAGE and added…
  • @harlanhaskins bb8160a
    [coverage] Declared SWIFT_ANALYZE_CODE_COVERAGE in CMakeLists and doc…
  • 314 more commits »
@slavapestov
  • @slavapestov 96b542e
    SILGen: Clean up witness thunk re-abstraction
@slavapestov
@slavapestov
Something went wrong with that request. Please try again.