- User Since
- May 27 2014, 6:39 AM (243 w, 1 d)
- Add a test that __kindof works with type sugar such as typedef.
- Use early return to reduce nesting. Also it's consistent with the previous early returns.
Update diff to ignore parent patch.
Tue, Jan 22
Wed, Jan 16
Thanks for the review.
Tue, Jan 15
Thanks for the review.
Fri, Jan 11
Thu, Jan 10
Mon, Jan 7
- Test initializing vector of pointers with array of pointers of convertible type.
- Test initializing vector of unsigned int with array of int.
Fri, Jan 4
- Tweak the test.
- Use for __construct_range_forward approach recommended by @ldionne (it works with C++03).
- Use __is_default_allocator for __construct_forward and __construct_backward.
Dec 20 2018
The change itself looks correct. Cannot really tell if you need to make changes in other places. For that I rely on Richard's opinion.
Thanks for the review, John. Committed the change and will watch the buildbots on non-Darwin platform.
Dec 19 2018
- [ObjC++] Verify there are no unexpected calls.
Dec 18 2018
- [ObjC++] Pass struct with __strong field as a value, not as a reference.
Dec 17 2018
- Test order of destroying parameters to inlined inherited constructor.
- [ObjC++] Add test using struct with __strong field as a parameter for inlined inherited constructor.
Dec 14 2018
Dec 12 2018
Dec 11 2018
Regarding the tests. I've moved most of new tests for custom allocators to test/libcxx/*. And in destroy.pass.cpp I'm just checking _LIBCPP_VERSION to avoid copying the test to a different file.
- Update __has_construct, __has_destroy according to review comments.
- Tighten tests: custom allocators aren't mandated by C++03 but libc++ has the support.
Dec 10 2018
Dec 7 2018
Thanks for the review, Erik.
Thanks for the feedback. Answered one simple question, the rest needs more time.
Dec 6 2018
- Remove now unnecessary #include "llvm/ADT/DenseSet.h"
- Different approach: don't enter already processed files after reaching max include depth.
- Keep indexing after include cycle.
Dec 4 2018
Nov 29 2018
Nov 16 2018
Have you considered the same approach as typo correction? I.e. for the max allowed edit distance use percentage of the input size. For example, something similar to TypoCorrectionConsumer::addName.
Nov 15 2018
Nov 13 2018
The test case I've promised is
Nov 9 2018
Nov 8 2018
How do you want it to work with symlinks? Would getRealPath be sufficient for your purpose?
Nov 7 2018
Depending on isLocal usage, it can be worth implementing it for RedirectingFileSystem.
Nov 6 2018
OK. Good to know you are still working on it.
Nov 5 2018
Sorry, you've decided to abandon the patch, it took a lot of good work. Xing, are you sure you don't want to see this change finished? I agree that delays in code review can be frustrating and I think it is something we can improve.
Disabling test in D54132.
Nov 1 2018
Thanks for the review, John. I'll update the comment with TODO and commit.
Oct 31 2018
Can you please check https://reviews.llvm.org/D53674 ? It builds on top of this change and I plan to commit them together.
Exclude commits tracked in a different review.
- Rename EmitConstant to emitScalarConstant.
- Tweak comment to be explicitly about intended IR code, not about Obj-C++ code.
- Switch to camelCase.
Looks good to me.
Oct 30 2018
- Rename EmitConstant to EmitScalarConstant.