[PATCH] D38216: [C++17] Fix class template argument deduction for default constructors without an initializer

2017-11-13 Thread Zhihao Yuan via Phabricator via cfe-commits
lichray added a comment. In https://reviews.llvm.org/D38216#923580, @rsmith wrote: > That's almost right, but not all `extern` declarations are disallowed. (An > `extern` declaration is still a defining declaration if it has an > initializer.) Good to know, thanks. So gcc got this wrong? ht

r318129 - [completion] complete ObjC interface names in an expression

2017-11-13 Thread Alex Lorenz via cfe-commits
Author: arphaman Date: Mon Nov 13 17:46:24 2017 New Revision: 318129 URL: http://llvm.org/viewvc/llvm-project?rev=318129&view=rev Log: [completion] complete ObjC interface names in an expression Objective-C interfaces can be used in a class property expression. rdar://26982192 Modified: cfe

r318130 - [PM] Add a missing header that I had in the next commit but was needed

2017-11-13 Thread Chandler Carruth via cfe-commits
Author: chandlerc Date: Mon Nov 13 17:47:24 2017 New Revision: 318130 URL: http://llvm.org/viewvc/llvm-project?rev=318130&view=rev Log: [PM] Add a missing header that I had in the next commit but was needed in r318128. Should fix the build. Modified: cfe/trunk/lib/CodeGen/BackendUtil.cpp Mod

[PATCH] D39994: Loosen MSVC 2017 path requirements

2017-11-13 Thread David Major via Phabricator via cfe-commits
dmajor created this revision. Herald added subscribers: kristof.beyls, aemerson. Mozilla's build machines are currently applying this patch locally, but I thought I'd offer it upstream because it should be pretty harmless. clang-cl has some sanity checks to make sure that the cl.exe it finds is

r318131 - [PM] Wire up support for the bounds checking sanitizer with the new PM.

2017-11-13 Thread Chandler Carruth via cfe-commits
Author: chandlerc Date: Mon Nov 13 17:59:18 2017 New Revision: 318131 URL: http://llvm.org/viewvc/llvm-project?rev=318131&view=rev Log: [PM] Wire up support for the bounds checking sanitizer with the new PM. Not much interesting here. Mostly wiring things together. One thing worth noting is that

[PATCH] D39913: [ObjC] warn about availability attributes missing from a method's declaration when they're specified for a method's definition

2017-11-13 Thread Akira Hatanaka via Phabricator via cfe-commits
ahatanak added a comment. Is it not necessary to print a diagnostic when a non-member function declaration is missing an availability attribute? void foo1(); __attribute__((availability(macos, introduced=10.1))) void foo1() { } Repository: rL LLVM https://reviews.llvm.org/D39913

r318137 - [PM] Require a registered x86 target for this test which uses the x86

2017-11-13 Thread Chandler Carruth via cfe-commits
Author: chandlerc Date: Mon Nov 13 21:20:03 2017 New Revision: 318137 URL: http://llvm.org/viewvc/llvm-project?rev=318137&view=rev Log: [PM] Require a registered x86 target for this test which uses the x86 triple. Modified: cfe/trunk/test/CodeGen/bounds-checking.c Modified: cfe/trunk/test/Co

[PATCH] D39955: [Driver] Add command-line flags for the new TBAA metadata format

2017-11-13 Thread John McCall via Phabricator via cfe-commits
rjmccall added a comment. We don't want this to become a long-term feature; we just want to switch the implementation over to the new format. Is there a good reason not to do that in one patch? If so, adding an option is acceptable, but it should be a -cc1 option only, because it is purely a

[PATCH] D39505: [OpenMP] Show error if VLAs are not supported

2017-11-13 Thread John McCall via Phabricator via cfe-commits
rjmccall added inline comments. Comment at: include/clang/Basic/TargetInfo.h:944 + /// \brief Whether target supports variable-length arrays. + bool isVLASupported() const { return VLASupported; } + Hahnfeld wrote: > rjmccall wrote: > > ABataev wrote: > > > ABa

<    1   2