Re: [PATCH] D10802: [mips] Interrupt attribute support.

sdardis updated this revision to Diff 40382. sdardis marked 12 inline comments as done. sdardis added a comment. Updated documentation text. Switched the various hard failures to semantic warnings/failures. Strangely, the 'interrupt' attribute on a non-function defaults to a warning. Thanks.

Re: [PATCH] D14467: [MS] Fix for bug 25013 - #pragma vtordisp is unknown inside functions.

d.zobnin.bugzilla updated this revision to Diff 40380. d.zobnin.bugzilla added a comment. Only tests have been updated: enabled a test case for Sema, which pass with these changes, added corresponding test case for CodeGen. http://reviews.llvm.org/D14467 Files: include/clang/Sema/Sema.h

Re: r253269 - Make FP_CONTRACT ON the default.

Note that due to this change we're hitting an assert at lib/CodeGen/CGExprScalar.cpp:2570 in llvm::Value *tryEmitFMulAdd(const (anonymous namespace)::BinOpInfo &, const clang::CodeGen::CodeGenFunction &, clang::CodeGen::CGBuilderTy &, bool): LHSBinOp->getNumUses( ) == 0 && "Operations with

Re: r253269 - Make FP_CONTRACT ON the default.

- Original Message - > From: "Renato Golin via cfe-commits" > To: "Stephen Canon" > Cc: "Clang Commits" > Sent: Tuesday, November 17, 2015 3:51:23 AM > Subject: Re: r253269 - Make FP_CONTRACT ON the default. > >

Re: [PATCH] D14467: [MS] Fix for bug 25013 - #pragma vtordisp is unknown inside functions.

d.zobnin.bugzilla added inline comments. Comment at: test/SemaCXX/pragma-vtordisp.cpp:34-35 @@ -33,6 +33,4 @@ struct C { -// FIXME: Our implementation based on token insertion makes it impossible for -// the pragma to appear everywhere we should support it. -//#pragma

Re: [PATCH] D13731: [Analyzer] Supporting function attributes in .model files.

pgousseau added a comment. Ping! http://reviews.llvm.org/D13731 ___ cfe-commits mailing list cfe-commits@lists.llvm.org http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

Re: r253269 - Make FP_CONTRACT ON the default.

Reverted in r253337. Failing test case in commit message. On Tue, Nov 17, 2015 at 4:39 PM Manuel Klimek wrote: > Repro: > float foo(float U, float base, float cell) { return (U = 2 * base) - cell; > } > Preparing rollback of the CL. > > On Tue, Nov 17, 2015 at 2:46 PM Manuel

r253337 - Revert "Make FP_CONTRACT ON the default."

Author: klimek Date: Tue Nov 17 09:40:10 2015 New Revision: 253337 URL: http://llvm.org/viewvc/llvm-project?rev=253337=rev Log: Revert "Make FP_CONTRACT ON the default." This reverts commit r253269. This leads to assert / segfault triggering on the following reduced example: float foo(float U,

Re: r253269 - Make FP_CONTRACT ON the default.

Repro: float foo(float U, float base, float cell) { return (U = 2 * base) - cell; } Preparing rollback of the CL. On Tue, Nov 17, 2015 at 2:46 PM Manuel Klimek wrote: > Note that due to this change we're hitting an assert at > lib/CodeGen/CGExprScalar.cpp:2570 in llvm::Value

Re: [PATCH] D14134: [OpenMP] Parsing and sema support for map clause

ABataev added inline comments. Comment at: include/clang/AST/OpenMPClause.h:2660-2662 @@ +2659,5 @@ + /// + /// \param StartLoc Starting location of the clause. + /// \param EndLoc Ending location of the clause. + /// \param N Number of the variables in the clause. + ///

Re: [PATCH] D10802: [mips] Interrupt attribute support.

aaron.ballman added inline comments. Comment at: include/clang/Basic/DiagnosticSemaKinds.td:255 @@ -254,1 +254,3 @@ def err_parameter_name_omitted : Error<"parameter name omitted">; +def err_excess_arguments : Error<"function %0 has too many arguments">; +def

[PATCH] D14748: bmiintrin.h: Allow using the tzcnt intrinsics for non-BMI targets

hans created this revision. hans added reviewers: thakis, echristo. hans added a subscriber: cfe-commits. The tzcnt intrinsics are used non non-BMI targets by code (e.g. ffmpeg) that uses it as a potentially faster BSF. The TZCNT instruction is special in that it's encoded in a

r253350 - Add trivial utility to append -L arguments to linker step. NFC

Author: dougk Date: Tue Nov 17 11:41:23 2015 New Revision: 253350 URL: http://llvm.org/viewvc/llvm-project?rev=253350=rev Log: Add trivial utility to append -L arguments to linker step. NFC Modified: cfe/trunk/include/clang/Driver/ToolChain.h cfe/trunk/lib/Driver/ToolChain.cpp

Re: [PATCH] D14748: bmiintrin.h: Allow using the tzcnt intrinsics for non-BMI targets

echristo accepted this revision. echristo added a comment. This revision is now accepted and ready to land. Sounds good to me. Weird, but fine :) -eric http://reviews.llvm.org/D14748 ___ cfe-commits mailing list cfe-commits@lists.llvm.org

r253358 - bmiintrin.h: Allow using the tzcnt intrinsics for non-BMI targets

Author: hans Date: Tue Nov 17 12:46:48 2015 New Revision: 253358 URL: http://llvm.org/viewvc/llvm-project?rev=253358=rev Log: bmiintrin.h: Allow using the tzcnt intrinsics for non-BMI targets The tzcnt intrinsics are used non non-BMI targets by code (e.g. ffmpeg) that uses it as a potentially

Re: r253398 - [modules] When a #include is mapped to a module import and appears somewhere

On Tue, Nov 17, 2015 at 3:40 PM, David Blaikie wrote: > On Tue, Nov 17, 2015 at 3:32 PM, Richard Smith via cfe-commits < > cfe-commits@lists.llvm.org> wrote: > >> Author: rsmith >> Date: Tue Nov 17 17:32:01 2015 >> New Revision: 253398 >> >> URL:

r253398 - [modules] When a #include is mapped to a module import and appears somewhere

Author: rsmith Date: Tue Nov 17 17:32:01 2015 New Revision: 253398 URL: http://llvm.org/viewvc/llvm-project?rev=253398=rev Log: [modules] When a #include is mapped to a module import and appears somewhere other than the top level, we issue an error. This breaks a fair amount of C++ code wrapping

Re: [PATCH] D14744: PR10235: support for vector mode attributes + warning

DmitryPolukhin added a comment. Thank you for prompt response! Comment at: include/clang/Basic/DiagnosticSemaKinds.td:2765 @@ +2764,3 @@ +def warn_vector_mode_deprecated : Warning< + "specifying vector types with __attribute__ ((mode)) is deprecated, " + "use __attribute__

Re: [PATCH] D9600: Add scan-build python implementation

rizsotto.mailinglist added a comment. In http://reviews.llvm.org/D9600#290031, @jroelofs wrote: > Thanks for re-uploading! thanks for your comments! :) most of your comments are about to embed it more into the clang build, test infrastructure. i think these are valid points! i would like to

Re: [PATCH] D14748: bmiintrin.h: Allow using the tzcnt intrinsics for non-BMI targets

craig.topper added a subscriber: craig.topper. craig.topper added a comment. The summary of why this is ok is slightly misleading. The backend won't try to encode the TZCNT instruction when the BMI feature is not enabled. Notice this just maps the to the generic non-x86 specific __builtin_ctz.

r253410 - Fix for use-after-free which caused test failure in cuda-detect.cu.

Author: tra Date: Tue Nov 17 18:37:41 2015 New Revision: 253410 URL: http://llvm.org/viewvc/llvm-project?rev=253410=rev Log: Fix for use-after-free which caused test failure in cuda-detect.cu. Return std::string itself instead StringRef to a temporary std::string. Modified:

Re: [PATCH] D13488: [analyzer] Assume escape is possible through system functions taking void*

zaks.anna closed this revision. zaks.anna added a comment. Committed in r251449. http://reviews.llvm.org/D13488 ___ cfe-commits mailing list cfe-commits@lists.llvm.org http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

Re: r252853 - libclang: add clang_Cursor_getCXXManglings

This introduced a memory leak, which I'm testing a patch for. Aside from that, did you get Doug to review this? We probably want to be really careful about changing libclang's C API, so you should be more cautious here than you would be normally. On Wed, Nov 11, 2015 at 7:57 PM, Saleem

Re: [PATCH] D14737: Convert some ObjC msgSends to runtime calls

stephanemoore added a subscriber: stephanemoore. stephanemoore added a comment. I hope that it's not presumptuous of me to inquire but I was wondering if the intent of this patch is to optimize calls to RR methods (and alloc) in non-ARC code? Would I be correct in assuming that clang will

Re: [clang-tools-extra] r253401 - [clang-tidy] add check cppcoreguidelines-pro-bounds-constant-array-index

Hi Matthias, This is failing on the green dragon bot: http://lab.llvm.org:8080/green/job/clang-stage1-configure-RA_check/13184/ http://lab.llvm.org:8080/green/job/clang-stage1-configure-RA_check/13184/consoleFull#54616856949ba4694-19c4-4d7e-bec5-911270d8a58c Thanks, On Tue, Nov 17, 2015 at 3:43

Re: [PATCH] D14652: [analyzer] Improve modeling of static initializers.

zaks.anna added inline comments. Comment at: test/Analysis/inline.cpp:308 @@ +307,3 @@ +clang_analyzer_eval(0 != void_string); // expected-warning{{TRUE}} +clang_analyzer_eval(0 != ((char *)void_string)[1]); // expected-warning{{TRUE}} + } Why are we

[clang-tools-extra] r253428 - Revert r253401, "[clang-tidy] add check cppcoreguidelines-pro-bounds-constant-array-index"

Author: chapuni Date: Tue Nov 17 20:14:35 2015 New Revision: 253428 URL: http://llvm.org/viewvc/llvm-project?rev=253428=rev Log: Revert r253401, "[clang-tidy] add check cppcoreguidelines-pro-bounds-constant-array-index" cppcoreguidelines-pro-bounds-constant-array-index.cpp is failing in several

r253434 - BuildUniversalActions(): Prune obsolete \param(s). [-Wdocumentation]

Author: chapuni Date: Tue Nov 17 23:15:21 2015 New Revision: 253434 URL: http://llvm.org/viewvc/llvm-project?rev=253434=rev Log: BuildUniversalActions(): Prune obsolete \param(s). [-Wdocumentation] FIXME: Describe BAInputs. Modified: cfe/trunk/include/clang/Driver/Driver.h Modified:

[libcxxabi] r253435 - c++abi: use __builtin_offsetof instead of offsetof

Author: compnerd Date: Tue Nov 17 23:33:38 2015 New Revision: 253435 URL: http://llvm.org/viewvc/llvm-project?rev=253435=rev Log: c++abi: use __builtin_offsetof instead of offsetof Use `__builtin_offsetof` in place of `offsetof`. Certain environments provide a macro definition of `offsetof`

r253431 - [NFC] Change the evaluation context of a non-type default template argument from Unevaluated to ConstantEvaluated.

Author: faisalv Date: Tue Nov 17 22:29:22 2015 New Revision: 253431 URL: http://llvm.org/viewvc/llvm-project?rev=253431=rev Log: [NFC] Change the evaluation context of a non-type default template argument from Unevaluated to ConstantEvaluated. This patch emits a more appropriate (but still

Re: [PATCH] D14652: [analyzer] Improve modeling of static initializers.

ygribov added inline comments. Comment at: test/Analysis/inline.cpp:308 @@ +307,3 @@ +clang_analyzer_eval(0 != void_string); // expected-warning{{TRUE}} +clang_analyzer_eval(0 != ((char *)void_string)[1]); // expected-warning{{TRUE}} + } zaks.anna

Re: [PATCH] D14506: Porting shouldVisitImplicitCode to DataRecursiveASTVisitor.

Richard, this is still optional, right? (the AST matchers need to control visitation) On Tue, Nov 17, 2015 at 2:26 AM Argyrios Kyrtzidis via cfe-commits < cfe-commits@lists.llvm.org> wrote: > W00t! That’s awesome Richard! > > On Nov 16, 2015, at 5:10 PM, Richard Smith

Re: [PATCH] D14506: Porting shouldVisitImplicitCode to DataRecursiveASTVisitor.

LOn Nov 17, 2015 12:49 AM, "Manuel Klimek" wrote: > Richard, this is still optional, right? (the AST matchers need to control visitation) This doesn't change RAV semantics at all (or if it does, it's a bug). If any of the extension points are overridden in the derived class,

Re: [PATCH] D14506: Porting shouldVisitImplicitCode to DataRecursiveASTVisitor.

On Tue, Nov 17, 2015 at 10:14 AM Richard Smith wrote: > LOn Nov 17, 2015 12:49 AM, "Manuel Klimek" wrote: > > Richard, this is still optional, right? (the AST matchers need to > control visitation) > > This doesn't change RAV semantics at all (or if it

Re: [PATCH] D14736: [analyzer] DeadStoresChecker: Treat locals captured by reference in C++ lambdas as escaped.

xazax.hun added a comment. This looks good to me, I have one question inline. Comment at: lib/StaticAnalyzer/Checkers/DeadStoresChecker.cpp:425 @@ +424,3 @@ +llvm::DenseMap CaptureFields; +FieldDecl *ThisCaptureField; +LambdaClass->getCaptureFields(CaptureFields,

Re: [PATCH] D14170: Fix false positive warning about memory leak for QApplication::postEvent

xazax.hun added inline comments. Comment at: lib/StaticAnalyzer/Checkers/MallocChecker.cpp:2514 @@ -2513,1 +2513,3 @@ + if (FName.endswith("postEvent") && + FD->getQualifiedNameAsString() == "QCoreApplication::postEvent") { Shouldn't you use == instead of

Re: [PATCH] D14744: PR10235: support for vector mode attributes + warning

aaron.ballman added inline comments. Comment at: include/clang/Basic/DiagnosticSemaKinds.td:2765 @@ +2764,3 @@ +def warn_vector_mode_deprecated : Warning< + "specifying vector types with __attribute__ ((mode)) is deprecated, " + "use __attribute__ ((vector_size)) instead">,

[libcxx] r253382 - Fix bad variable name. project_root -> project_obj_root

Author: ericwf Date: Tue Nov 17 15:48:29 2015 New Revision: 253382 URL: http://llvm.org/viewvc/llvm-project?rev=253382=rev Log: Fix bad variable name. project_root -> project_obj_root Modified: libcxx/trunk/test/libcxx/test/config.py Modified: libcxx/trunk/test/libcxx/test/config.py URL:

[PATCH] D14756: Handle ARMv6-J as an alias, instead of fake architecture

tyomitch created this revision. tyomitch added reviewers: rengolin, bogden, compnerd. tyomitch added a subscriber: cfe-commits. Herald added subscribers: rengolin, aemerson. Clang-side update, corresponding to D14755 http://reviews.llvm.org/D14756 Files: test/Driver/arm-cortex-cpus.c Index:

r253389 - [CUDA] Make CUDA compilation usable by default.

Author: tra Date: Tue Nov 17 16:28:55 2015 New Revision: 253389 URL: http://llvm.org/viewvc/llvm-project?rev=253389=rev Log: [CUDA] Make CUDA compilation usable by default. Currently clang requires several additional command line options in order to enable new features needed during CUDA

Re: [PATCH] D13144: [CUDA] propagate to CUDA sub-compilations target triple of opposite side.

This revision was automatically updated to reflect the committed changes. Closed by commit rL253385: [CUDA] use -aux-triple to pass target triple of opposite side of compilation (authored by tra). Changed prior to commit: http://reviews.llvm.org/D13144?vs=40163=40432#toc Repository: rL LLVM

Re: [PATCH] D13170: [CUDA] Driver changes to pass flags needed to use detected CUDA installation.

This revision was automatically updated to reflect the committed changes. Closed by commit rL253386: [CUDA] added include paths for both sides of CUDA compilation. (authored by tra). Changed prior to commit: http://reviews.llvm.org/D13170?vs=39860=40433#toc Repository: rL LLVM

Re: [PATCH] D14754: [Myriad]: insert -L paths into linker cmd only when they exist.

jyknight added inline comments. Comment at: lib/Driver/ToolChains.cpp:4319 @@ +4318,3 @@ + // This contains libc, libg (a superset of libc), libm, libstdc++, libssp. + SmallString<128> LibDir(GCCInstallation.getParentLibPath()); + if (Triple.getArch() == llvm::Triple::sparcel)

r253440 - Update for llvm change.

Author: rafael Date: Wed Nov 18 00:54:13 2015 New Revision: 253440 URL: http://llvm.org/viewvc/llvm-project?rev=253440=rev Log: Update for llvm change. Modified: cfe/trunk/include/clang/Lex/HeaderSearchOptions.h cfe/trunk/lib/CodeGen/CGObjCMac.cpp Modified:

[clang-tools-extra] r253442 - Replace a dyn_cast with isa where the result was only being used as a boolean. NFC.

Author: ctopper Date: Wed Nov 18 01:08:11 2015 New Revision: 253442 URL: http://llvm.org/viewvc/llvm-project?rev=253442=rev Log: Replace a dyn_cast with isa where the result was only being used as a boolean. NFC. Modified: clang-tools-extra/trunk/clang-tidy/modernize/UseDefaultCheck.cpp