- User Since
- Dec 9 2012, 11:41 PM (411 w, 3 h)
Wed, Oct 21
Tue, Oct 20
Is there a reason to not use the existing -mlinker-version= option and expanding that beyond just Darwin targets? I feel like having the same option would be much nicer.
Mon, Oct 19
Thanks for doing this!
I know that many people find these tools useful, but they are meant as testing tools. Marking them as utilities seems reasonable.
Fri, Oct 16
Thanks for the feedback!
Thu, Oct 15
Passes locally now
Given that Darwin has fully moved to DWARF at this point, why STABS?
Tue, Oct 13
Rebase; there are a 4 more test failures to go through
Sure, I think I can do that instead. Thanks for pointing me to that!
Mon, Oct 12
Personally, I think that switch is nicer for the ExceptionCode check, but thats not your doing nor a real concern.
Sat, Oct 10
I've not taken a thorough look at the changes, but I am definitely in favour of this change (conceptually).
Fri, Oct 9
Remove unnecessary include, fix an incorrect wait (verified that unit tests now pass!)
Thu, Oct 8
Wed, Oct 7
This seems pretty reasonable to me, it just makes the DLL storage handling more similar to Microsoft's which is what the Windows Itanium target is intending to do.
The reason that I added the overlapped event was specifically for the cancellation and overlooked the fact that it will be used by the kernel side as well, thanks for catching that!
Mon, Oct 5
Sun, Oct 4
There already is testing coverage for this - I just missed the CMake changes.
Fri, Oct 2
Address feedback from @rsmith
Update text based on feedback.
Thu, Oct 1
@aaron.ballman - I completely agree with you about the testing. The interfaces are tested via https://github.com/llvm/llvm-project/blob/master/clang/unittests/DirectoryWatcher/CMakeLists.txt, which now that I look at again, seems to need an additional case for system name.
Wed, Sep 30
Tue, Sep 29
If we require LLVM for any of the projects, perhaps we should hoist the HandleCompilerRT.cmake into LLVM and avoid having to maintain the copy everywhere.
Mon, Sep 28
Sep 25 2020
Sep 24 2020
lld changes seem unrelated (include changes, whitespace); LGTM with the copyright header for lit.cfg
- address all feedback, add more test cases
add more tests to cover the confusing case of attribute merge handling
Sep 23 2020
Correct, this is just upstreaming the original ArchSpec change that we found on Swift.
Emphatic +1 to @smeenai's comment. This really should be done via lit substitutions.
Sep 22 2020
Update for feedback from @aaron.ballman
The ParseAttrCommonArgs refactoring still needs to be looked at, but this should address the feedback, and doesn't need to be predicated on the refactoring.
Sep 21 2020
Sep 17 2020
Sep 16 2020
Change diagnostics for conflicting swift_name, add a test case
Address everything but warning
Address feedback from @aaron.ballman
I think that tracking the segment length should be okay. It’s a total overhead of 8 bytes, which is only allocated once, however, the naming here is egregiously bad. This is the segment length and not the DSO length nor the DSO mapped length (which may actually be larger or smaller based on the load segment layout).