[PATCH] D26546: [PPC] Add vec_insert4b/vec_extract4b to altivec.h

echristo closed this revision. echristo added a comment. This was committed: commit d65cd1f9424369c4ae7f945fac7fd9e4357451b2 Author: Sean Fertile Date: Thu Jan 5 21:43:30 2017 + Add vec_insert4b and vec_extract4b functions to altivec.h Add builtins for the

[PATCH] D29077: [lsan] Enable LSan for x86 Linux.

This revision was automatically updated to reflect the committed changes. Closed by commit rL293609: [lsan] Enable LSan for x86 Linux (authored by chefmax). Changed prior to commit: https://reviews.llvm.org/D29077?vs=85756=86387#toc Repository: rL LLVM https://reviews.llvm.org/D29077

r293604 - In VirtualCallChecker, handle indirect calls

Author: sammccall Date: Mon Jan 30 23:23:20 2017 New Revision: 293604 URL: http://llvm.org/viewvc/llvm-project?rev=293604=rev Log: In VirtualCallChecker, handle indirect calls Summary: In VirtualCallChecker, handle indirect calls. getDirectCallee() can be nullptr, and dyn_cast(nullptr) is UB

[PATCH] D29303: In VirtualCallChecker, handle indirect calls

This revision was automatically updated to reflect the committed changes. Closed by commit rL293604: In VirtualCallChecker, handle indirect calls (authored by sammccall). Changed prior to commit: https://reviews.llvm.org/D29303?vs=86342=86385#toc Repository: rL LLVM

[PATCH] D28001: [X86] Teach Clang about -mfentry flag

dberris accepted this revision. dberris added a comment. This revision is now accepted and ready to land. LGTM https://reviews.llvm.org/D28001 ___ cfe-commits mailing list cfe-commits@lists.llvm.org

r293600 - AMDGPU: Add builtin for fmed3 intrinsic

Author: arsenm Date: Mon Jan 30 21:42:07 2017 New Revision: 293600 URL: http://llvm.org/viewvc/llvm-project?rev=293600=rev Log: AMDGPU: Add builtin for fmed3 intrinsic Modified: cfe/trunk/include/clang/Basic/BuiltinsAMDGPU.def cfe/trunk/lib/CodeGen/CGBuiltin.cpp

[libcxx] r293599 - Fix PR#31779: basic_string::operator= isn't exception safe.

Author: marshall Date: Mon Jan 30 21:40:52 2017 New Revision: 293599 URL: http://llvm.org/viewvc/llvm-project?rev=293599=rev Log: Fix PR#31779: basic_string::operator= isn't exception safe. Modified: libcxx/trunk/include/string

[PATCH] D24933: Enable configuration files in clang

hfinkel added inline comments. Comment at: lib/Driver/Driver.cpp:172 +NumConfigArgs = static_cast(NumCfgArgs); + } + sepavloff wrote: > bruno wrote: > > If `NumCfgArgs == 0` it would be nice to warn that the config file is empty? > Not sure if it makes

Re: r293596 - Handle ObjCEncodeExpr in extractStringLiteralCharacter.

Hans, can this be merged to 4.0? This fixes a regression from 3.9 (the part I touched was committed a few years ago, but didn’t cause an assert until recently). > On Jan 30, 2017, at 6:31 PM, Akira Hatanaka via cfe-commits > wrote: > > Author: ahatanak > Date:

Re: [libcxx] r293581 - Revert "Adorn __call_once_proxy with `inline` and `_LIBCPP_INLINE_VISIBILITY`"

Can you file a PR for changing this to internal visibility please? I think that we could probably do that in the unstable ABI versioning in fact. On Mon, Jan 30, 2017 at 5:26 PM, Justin Bogner via cfe-commits < cfe-commits@lists.llvm.org> wrote: > Author: bogner > Date: Mon Jan 30 19:26:09 2017

Re: Add warning for c++ member variable shadowing

I think that the patch is starting to look pretty good! Can you add some test cases for the particular cases to diagnose in a separate test set to ensure that we have proper coverage of the various cases rather than relying on the existing test cases? Something to make sure that we get the

r293596 - Handle ObjCEncodeExpr in extractStringLiteralCharacter.

Author: ahatanak Date: Mon Jan 30 20:31:39 2017 New Revision: 293596 URL: http://llvm.org/viewvc/llvm-project?rev=293596=rev Log: Handle ObjCEncodeExpr in extractStringLiteralCharacter. This fixes an assertion failure that occurs later in the function when an ObjCEncodeExpr is cast to

r293595 - Improve fix for PR28739

Author: rsmith Date: Mon Jan 30 20:23:02 2017 New Revision: 293595 URL: http://llvm.org/viewvc/llvm-project?rev=293595=rev Log: Improve fix for PR28739 Don't try to map an APSInt addend to an int64_t in pointer arithmetic before bounds-checking it. This gives more consistent behavior (outside

[PATCH] D29304: [cmake] Hint find_package() to prefer LLVM installed alongside clang

rnk accepted this revision. rnk added a comment. This revision is now accepted and ready to land. lgtm Repository: rL LLVM https://reviews.llvm.org/D29304 ___ cfe-commits mailing list cfe-commits@lists.llvm.org

Re: Add warning for c++ member variable shadowing

Hi Richard Sorry for the late reply. Thank you for giving the feedback! The updated version is attached. Please let me know if there is anything improper. Thanks James From: on behalf of Richard Smith Date: Friday, January 27, 2017 at 3:03 PM To:

[PATCH] D24333: [CleanupInfo] Use cleanupsHaveSideEffects instead of exprNeedsCleanups in assertions

timshen updated this revision to Diff 86377. timshen added a comment. ActOnFinishFullExpr after exiting the expression evaluation context. https://reviews.llvm.org/D24333 Files: clang/lib/Sema/SemaTemplateInstantiateDecl.cpp clang/lib/Sema/TreeTransform.h clang/test/Sema/pr30306.cpp

[PATCH] D21675: New ODR checker for modules

This revision was automatically updated to reflect the committed changes. Closed by commit rL293585: Add better ODR checking for modules. (authored by rtrieu). Changed prior to commit: https://reviews.llvm.org/D21675?vs=86142=86376#toc Repository: rL LLVM https://reviews.llvm.org/D21675

r293585 - Add better ODR checking for modules.

Author: rtrieu Date: Mon Jan 30 19:44:15 2017 New Revision: 293585 URL: http://llvm.org/viewvc/llvm-project?rev=293585=rev Log: Add better ODR checking for modules. When objects are imported for modules, there is a chance that a name collision will cause an ODR violation. Previously, only a

[PATCH] D24333: [CleanupInfo] Use cleanupsHaveSideEffects instead of exprNeedsCleanups in assertions

timshen updated this revision to Diff 86375. timshen added a comment. Fix in the right way as rsmith pointed out. https://reviews.llvm.org/D24333 Files: clang/lib/Sema/SemaTemplateInstantiateDecl.cpp clang/lib/Sema/TreeTransform.h clang/test/Sema/pr30306.cpp Index:

[libcxx] r293581 - Revert "Adorn __call_once_proxy with `inline` and `_LIBCPP_INLINE_VISIBILITY`"

Author: bogner Date: Mon Jan 30 19:26:09 2017 New Revision: 293581 URL: http://llvm.org/viewvc/llvm-project?rev=293581=rev Log: Revert "Adorn __call_once_proxy with `inline` and `_LIBCPP_INLINE_VISIBILITY`" While this change didn't really hurt, it does lead to spurious warnings about not being

Re: r289787 - Re-commit r289252 and r289285, and fix PR31374

On 15 December 2016 at 00:09, Yaxun Liu via cfe-commits < cfe-commits@lists.llvm.org> wrote: > Author: yaxunl > Date: Thu Dec 15 02:09:08 2016 > New Revision: 289787 > > URL: http://llvm.org/viewvc/llvm-project?rev=289787=rev > Log: > Re-commit r289252 and r289285, and fix PR31374 > > Added: >

[PATCH] D16135: Macro Debug Info support in Clang

mkuper resigned from this revision. mkuper added a comment. I admit it would be nice if someone reviewed this, but I'm really the wrong person for this code area. :-) https://reviews.llvm.org/D16135 ___ cfe-commits mailing list

[PATCH] D28365: [Driver] Updated for Visual Studio 2017

rnk accepted this revision. rnk added a comment. This revision is now accepted and ready to land. lgtm Comment at: lib/Driver/MSVCToolChain.cpp:34 + #if 0 +#define USE_VS_SETUP_CONFIG + #endif hamzasood wrote: > rnk wrote: > > What are the outstanding

Re: r293475 - Revert "r293343 - [ubsan] Sanity-check shift amounts before truncation

Sorry about that! I took another shot at it in r293572. vedant > On Jan 30, 2017, at 3:37 AM, Alex Lorenz via cfe-commits > wrote: > > Author: arphaman > Date: Mon Jan 30 05:37:18 2017 > New Revision: 293475 > > URL:

r293572 - Re-apply "[ubsan] Sanity-check shift amounts before truncation"

Author: vedantk Date: Mon Jan 30 17:38:54 2017 New Revision: 293572 URL: http://llvm.org/viewvc/llvm-project?rev=293572=rev Log: Re-apply "[ubsan] Sanity-check shift amounts before truncation" This re-applies r293343 (reverts commit r293475) with a fix for an assertion failure caused by a

r293568 - PR28739: Check that integer values fit into 64 bits before extracting them as 64 bit values for pointer arithmetic.

Author: rsmith Date: Mon Jan 30 17:30:26 2017 New Revision: 293568 URL: http://llvm.org/viewvc/llvm-project?rev=293568=rev Log: PR28739: Check that integer values fit into 64 bits before extracting them as 64 bit values for pointer arithmetic. This fixes various ways to tickle an assertion in

r293556 - Serialization: use range based for loop (NFC)

Author: compnerd Date: Mon Jan 30 16:25:28 2017 New Revision: 293556 URL: http://llvm.org/viewvc/llvm-project?rev=293556=rev Log: Serialization: use range based for loop (NFC) Just a small clean up noticed when doing post-commit review of Duncan's previous change for ModuleFile memory ownership

[PATCH] D29303: In VirtualCallChecker, handle indirect calls

bkramer accepted this revision. bkramer added a comment. This revision is now accepted and ready to land. lg https://reviews.llvm.org/D29303 ___ cfe-commits mailing list cfe-commits@lists.llvm.org

[PATCH] D29303: In VirtualCallChecker, handle indirect calls

sammccall updated this revision to Diff 86342. sammccall added a comment. Add regression test. https://reviews.llvm.org/D29303 Files: lib/StaticAnalyzer/Checkers/VirtualCallChecker.cpp test/Analysis/virtualcall.cpp Index: test/Analysis/virtualcall.cpp

[PATCH] D29304: [cmake] Hint find_package() to prefer LLVM installed alongside clang

mgorny created this revision. Include a path hint for find_package() in ClangConfig.cmake to ensure that CMake prefers LLVM installed alongside clang over the default search path. If two versions of LLVM are installed in the system, and one of them is in PATH, CMake's find_package() magic

[PATCH] D29303: In VirtualCallChecker, handle indirect calls

bkramer added a comment. Your test case is fine, it crashes with assertions enabled. https://reviews.llvm.org/D29303 ___ cfe-commits mailing list cfe-commits@lists.llvm.org http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[PATCH] D24333: [CleanupInfo] Use cleanupsHaveSideEffects instead of exprNeedsCleanups in assertions

rsmith added inline comments. Comment at: clang/test/SemaCXX/pr30306.cpp:5 +template +void g(T) { int a[f(3)]; } // expected-no-diagnostics + Shouldn't we be (somehow) handling the cleanups from the array bound expression here -- either discarding them or

[PATCH] D29303: In VirtualCallChecker, handle indirect calls

sammccall added a comment. I couldn't work out how to add a test for this, advice appreciated. Closest I could get was adding something like this to test/Analysis/virtualcall.cpp: class F { public: F(); void foo(); }; F::F() { void (F::* ptr) = ::foo; (this->*ptr)(); } which crashes, but

[PATCH] D29303: In VirtualCallChecker, handle indirect calls

sammccall updated this revision to Diff 86337. sammccall added a comment. Oops, reverted noise :( https://reviews.llvm.org/D29303 Files: lib/StaticAnalyzer/Checkers/VirtualCallChecker.cpp Index: lib/StaticAnalyzer/Checkers/VirtualCallChecker.cpp

[PATCH] D29303: In VirtualCallChecker, handle indirect calls

sammccall created this revision. In VirtualCallChecker, handle indirect calls. getDirectCallee() can be nullptr, and dyn_cast(nullptr) is UB https://reviews.llvm.org/D29303 Files: lib/StaticAnalyzer/Checkers/VirtualCallChecker.cpp Index: lib/StaticAnalyzer/Checkers/VirtualCallChecker.cpp

[PATCH] D29298: [clang-format] Fix regression that breaks comments without a comment prefix

This revision was automatically updated to reflect the committed changes. Closed by commit rL293548: [clang-format] Fix regression that breaks comments without a comment prefix (authored by krasimir). Changed prior to commit: https://reviews.llvm.org/D29298?vs=86328=86331#toc Repository: rL

r293548 - [clang-format] Fix regression that breaks comments without a comment prefix

Author: krasimir Date: Mon Jan 30 15:00:01 2017 New Revision: 293548 URL: http://llvm.org/viewvc/llvm-project?rev=293548=rev Log: [clang-format] Fix regression that breaks comments without a comment prefix Summary: Consider formatting the following code fragment with column limit 20: ``` { //

[PATCH] D29300: [clang-format] Refactor WhitespaceManager and friends

djasper created this revision. The main motivation behind this is to cleanup the WhitespaceManager and make it more extensible for future alignment etc. features. Specifically, WhitespaceManager has started to copy more and more code that is already present in FormatToken. Instead, I think it

Re: r293473 - Revert r293455, which breaks v8 with a spurious error. Testcase added.

Thanks for the revert; fixed and re-committed as r293544. On 30 January 2017 at 02:44, Sam McCall via cfe-commits < cfe-commits@lists.llvm.org> wrote: > Author: sammccall > Date: Mon Jan 30 04:44:11 2017 > New Revision: 293473 > > URL: http://llvm.org/viewvc/llvm-project?rev=293473=rev > Log: >

r293544 - Towards P0091R3: parsing support for class template argument deduction in typename-specifiers.

Author: rsmith Date: Mon Jan 30 14:39:26 2017 New Revision: 293544 URL: http://llvm.org/viewvc/llvm-project?rev=293544=rev Log: Towards P0091R3: parsing support for class template argument deduction in typename-specifiers. This reinstates r293455, reverted in r293455, with a fix for

[libcxx] r293543 - experimental: avoid using raw _WIN32 in filesystem

Author: compnerd Date: Mon Jan 30 13:57:27 2017 New Revision: 293543 URL: http://llvm.org/viewvc/llvm-project?rev=293543=rev Log: experimental: avoid using raw _WIN32 in filesystem Use the _LIBCPP_WIN32API macro instead of _WIN32 checks. Fix a missed renaming for style conformance. Modified:

[PATCH] D29291: [clang-format] Separate line comment sections after a right brace from comment sections in the scope.

This revision was automatically updated to reflect the committed changes. Closed by commit rL293539: [clang-format] Separate line comment sections after a right brace from comment… (authored by krasimir). Changed prior to commit: https://reviews.llvm.org/D29291?vs=86316=86317#toc Repository:

r293539 - [clang-format] Separate line comment sections after a right brace from comment sections in the scope.

Author: krasimir Date: Mon Jan 30 13:18:55 2017 New Revision: 293539 URL: http://llvm.org/viewvc/llvm-project?rev=293539=rev Log: [clang-format] Separate line comment sections after a right brace from comment sections in the scope. Summary: The following two comment lines form a single comment

[PATCH] D29291: [clang-format] Separate line comment sections after a right brace from comment sections in the scope.

krasimir updated this revision to Diff 86316. krasimir added a comment. - Remove scope https://reviews.llvm.org/D29291 Files: lib/Format/UnwrappedLineParser.cpp unittests/Format/FormatTest.cpp Index: unittests/Format/FormatTest.cpp

[PATCH] D29291: [clang-format] Separate line comment sections after a right brace from comment sections in the scope.

djasper accepted this revision. djasper added inline comments. This revision is now accepted and ready to land. Comment at: lib/Format/UnwrappedLineParser.cpp:2095 + const FormatToken *MinColumnToken = Line.Tokens.front().Tok; + { +// Scan for '{//'. If found, use the

[libcxx] r293531 - experimental: port directory_iterator to Windows

Author: compnerd Date: Mon Jan 30 12:50:34 2017 New Revision: 293531 URL: http://llvm.org/viewvc/llvm-project?rev=293531=rev Log: experimental: port directory_iterator to Windows This adds a basic first cut implementation for directory_iterator on Windows. It uses the FindFirstFile/FindNextFile

[PATCH] D28050: [Clang][Driver] Clean up Clang::ConstructJob a little bit, NFC

ahatanak added a comment. I think this patch is an improvement, but Clang::ConstructJob is still one giant function. Do you have ideas to improve readability of this function or plans to further reduce its size? Repository: rL LLVM https://reviews.llvm.org/D28050

[libcxx] r293530 - experimental: add missing file header

Author: compnerd Date: Mon Jan 30 12:50:32 2017 New Revision: 293530 URL: http://llvm.org/viewvc/llvm-project?rev=293530=rev Log: experimental: add missing file header The directory_iterator implementation file was missing the file header. Add one. NFC. Modified:

[PATCH] D22955: [MSVC] Improved late parsing of template functions.

ABataev added a comment. Ping https://reviews.llvm.org/D22955 ___ cfe-commits mailing list cfe-commits@lists.llvm.org http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

r293523 - [AST] Give TemplateArgumentLoc a constexpr ctor.

Author: d0k Date: Mon Jan 30 12:32:46 2017 New Revision: 293523 URL: http://llvm.org/viewvc/llvm-project?rev=293523=rev Log: [AST] Give TemplateArgumentLoc a constexpr ctor. This removes the thread-safe static from clang::TemplateTemplateParmDecl::getDefaultArgument() const::None Modified:

r293518 - [ASTMatchers] Sprinkle some constexpr on the global matcher constructors.

Author: d0k Date: Mon Jan 30 12:20:00 2017 New Revision: 293518 URL: http://llvm.org/viewvc/llvm-project?rev=293518=rev Log: [ASTMatchers] Sprinkle some constexpr on the global matcher constructors. This dramatically reduces the size of the global constructors we emit for those variables in

[PATCH] D28835: [coroutines] NFC: Refactor Sema::CoroutineBodyStmt construction.

GorNishanov added a comment. LGTM? Pretty please :) https://reviews.llvm.org/D28835 ___ cfe-commits mailing list cfe-commits@lists.llvm.org http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

RE: r293457 - Tidy up codegen modules test & make it x86 specific since it relies on Itanium name manglings

Okay as is, then. Thanks for the explanation. --paulr From: David Blaikie [mailto:dblai...@gmail.com] Sent: Monday, January 30, 2017 9:32 AM To: Robinson, Paul Cc: cfe-commits (cfe-commits@lists.llvm.org) Subject: Re: r293457 - Tidy up codegen modules test & make it x86 specific since it relies

Re: r293457 - Tidy up codegen modules test & make it x86 specific since it relies on Itanium name manglings

On reflection itanium wouldn't be sufficient/my comment wasn't sufficiently descriptive (I realized after I made this change that it would also fix the MacOS buildbot failure I was seeing & hadn't understood) - Darwin doesn't use comdats, for example, but is still an itanium ABI. I could remove

RE: r293457 - Tidy up codegen modules test & make it x86 specific since it relies on Itanium name manglings

Use %itanium_abi_triple instead? > -Original Message- > From: cfe-commits [mailto:cfe-commits-boun...@lists.llvm.org] On Behalf Of > David Blaikie via cfe-commits > Sent: Sunday, January 29, 2017 9:34 PM > To: cfe-commits@lists.llvm.org > Subject: r293457 - Tidy up codegen modules test &

[PATCH] D27985: Add demangling support for C++11 thread_local variables

mehdi_amini added a comment. In https://reviews.llvm.org/D27985#660354, @davidb wrote: > In https://reviews.llvm.org/D27985#660029, @mehdi_amini wrote: > > > LGTM. > > > > Can you apply the patch in LLVM (libDemangle) as well please? > > > Thank you Mehdi. I have an LLVM (libDemangle) with an

[PATCH] D29262: Fixes to modernize-use-using

krystyna updated this revision to Diff 86298. krystyna edited the summary of this revision. krystyna added a comment. Style fixes. https://reviews.llvm.org/D29262 Files: clang-tidy/modernize/UseUsingCheck.cpp clang-tidy/modernize/UseUsingCheck.h

[PATCH] D29152: Drop 'dllimport' when redeclaring inline function template without the attribute (PR31695)

rnk accepted this revision. rnk added a comment. This revision is now accepted and ready to land. lgtm https://reviews.llvm.org/D29152 ___ cfe-commits mailing list cfe-commits@lists.llvm.org

[PATCH] D29152: Drop 'dllimport' when redeclaring inline function template without the attribute (PR31695)

hans added a comment. Ping? https://reviews.llvm.org/D29152 ___ cfe-commits mailing list cfe-commits@lists.llvm.org http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

r293493 - Adjust tests after folding inlining analysis into missed remarks

Author: anemet Date: Mon Jan 30 10:22:50 2017 New Revision: 293493 URL: http://llvm.org/viewvc/llvm-project?rev=293493=rev Log: Adjust tests after folding inlining analysis into missed remarks Modified: cfe/trunk/test/Frontend/optimization-remark-with-hotness.c

[PATCH] D29205: Change debug-info-for-profiling from a TargetOption to a function attribute.

danielcdh updated this revision to Diff 86292. danielcdh added a comment. update https://reviews.llvm.org/D29205 Files: lib/CodeGen/BackendUtil.cpp lib/CodeGen/CGDebugInfo.cpp Index: lib/CodeGen/CGDebugInfo.cpp === ---

Re: [PATCH] D28007: Switch TableGen to emit calls to ASTRecordReader for AttrPCHRead.

This review thread is missing on-list approval. I assume it happened on Phab & just wasn't reflected here (if the Phab approval has no text it doesn't send an email, which is unfortunate - not sure if anyone's planning to fix that/change the settings, but until then it's helpful to include "LGTM"

Re: r293416 - [c-index-test] Provide capability for 'c-index-test core' to dump symbol information from a PCH/module file.

Hi Hans, Could this go into the stable branch, along with the follow-ups: r293461 r293463 r293466 > On Jan 28, 2017, at 8:50 PM, Argyrios Kyrtzidis via cfe-commits > wrote: > > Author: akirtzidis > Date: Sat Jan 28 22:50:35 2017 > New Revision: 293416 > > URL:

r293488 - [AST] Make header standalone.

Author: d0k Date: Mon Jan 30 10:00:57 2017 New Revision: 293488 URL: http://llvm.org/viewvc/llvm-project?rev=293488=rev Log: [AST] Make header standalone. Most implementations get ptrdiff_t transitively, some don't. Explicitly include cstddef. Modified:

[PATCH] D29267: [clang-tidy] safety-no-assembler

jbcoe updated this revision to Diff 86289. jbcoe added a comment. Add link to HIC++ website and fix release notes. https://reviews.llvm.org/D29267 Files: clang-tools-extra/clang-tidy/CMakeLists.txt clang-tools-extra/clang-tidy/safety/CMakeLists.txt

Re: r293395 - Modules: Clarify ownership of ModuleFile instances in ModuleManager, NFC

Thanks! Always love to see cleanup like this :) On Sat, Jan 28, 2017 at 2:35 PM Duncan P. N. Exon Smith via cfe-commits < cfe-commits@lists.llvm.org> wrote: > Author: dexonsmith > Date: Sat Jan 28 16:24:01 2017 > New Revision: 293395 > > URL: http://llvm.org/viewvc/llvm-project?rev=293395=rev >

r293485 - [IRGen] Make header standalone.

Author: d0k Date: Mon Jan 30 09:39:18 2017 New Revision: 293485 URL: http://llvm.org/viewvc/llvm-project?rev=293485=rev Log: [IRGen] Make header standalone. Modified: cfe/trunk/lib/CodeGen/ABIInfo.h Modified: cfe/trunk/lib/CodeGen/ABIInfo.h URL:

[PATCH] D27985: Add demangling support for C++11 thread_local variables

davidb added a comment. In https://reviews.llvm.org/D27985#660029, @mehdi_amini wrote: > LGTM. > > Can you apply the patch in LLVM (libDemangle) as well please? Thank you Mehdi. I have an LLVM (libDemangle) with an identical change (minus the tests) waiting. Would you like my to submit it as

r293473 - Revert r293455, which breaks v8 with a spurious error. Testcase added.

Author: sammccall Date: Mon Jan 30 04:44:11 2017 New Revision: 293473 URL: http://llvm.org/viewvc/llvm-project?rev=293473=rev Log: Revert r293455, which breaks v8 with a spurious error. Testcase added. Summary: Revert r293455, which breaks v8 with a spurious error. Testcase added. Reviewers:

Re: r293343 - [ubsan] Sanity-check shift amounts before truncation (fixes PR27271)

Another example + possible fix (two WidthMinusOne, one per (possibly different) bitwidth): (not fully tested) int f() { return 0 << 0L; } diff --git a/lib/CodeGen/CGExprScalar.cpp b/lib/CodeGen/CGExprScalar.cpp index 40d949dece..5c9055d49a 100644 --- a/lib/CodeGen/CGExprScalar.cpp +++

Re: r293343 - [ubsan] Sanity-check shift amounts before truncation (fixes PR27271)

Hi Vedant, This commit has caused a compiler crash in our stage 2 green dragon ASAN+Ubsan bot ( http://lab.llvm.org:8080/green/job/clang-stage2-cmake-RgSan_build/). I have reverted it in r293475. The following example reproduces the crash with -fsanitize=undefined : typedef unsigned long long

r293475 - Revert "r293343 - [ubsan] Sanity-check shift amounts before truncation

Author: arphaman Date: Mon Jan 30 05:37:18 2017 New Revision: 293475 URL: http://llvm.org/viewvc/llvm-project?rev=293475=rev Log: Revert "r293343 - [ubsan] Sanity-check shift amounts before truncation (fixes PR27271)" After r293343 clang fails to compile itself with -fsanitize=undefined (

[PATCH] D29271: Revert r293455, which breaks v8 with a spurious error. Testcase added.

sammccall added a comment. In https://reviews.llvm.org/D29271#660169, @sberg wrote: > Btw, ran into that with the even simpler test case Thanks, I stole it :-) https://reviews.llvm.org/D29271 ___ cfe-commits mailing list

[PATCH] D29271: Revert r293455, which breaks v8 with a spurious error. Testcase added.

This revision was automatically updated to reflect the committed changes. Closed by commit rL293473: Revert r293455, which breaks v8 with a spurious error. Testcase added. (authored by sammccall). Changed prior to commit: https://reviews.llvm.org/D29271?vs=86257=86259#toc Repository: rL

[PATCH] D29271: Revert r293455, which breaks v8 with a spurious error. Testcase added.

klimek accepted this revision. klimek added a comment. This revision is now accepted and ready to land. lg https://reviews.llvm.org/D29271 ___ cfe-commits mailing list cfe-commits@lists.llvm.org

[PATCH] D29271: Revert r293455, which breaks v8 with a spurious error. Testcase added.

sberg added a comment. Btw, ran into that with the even simpler test case template struct S { static typename T::t const n = 0; }; struct U { using t = int; }; int main() { return S::n; } https://reviews.llvm.org/D29271 ___

[PATCH] D29271: Revert r293455, which breaks v8 with a spurious error. Testcase added.

sammccall created this revision. Revert r293455, which breaks v8 with a spurious error. Testcase added. https://reviews.llvm.org/D29271 Files: include/clang/AST/DeclTemplate.h include/clang/Basic/DiagnosticSemaKinds.td include/clang/Sema/Sema.h lib/Sema/SemaDecl.cpp