From bugzilla-daemon at llvm.org Wed Dec 1 04:11:57 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Wed, 1 Dec 2010 04:11:57 -0600 (CST) Subject: [LLVMbugs] [Bug 8712] New: value dependent member in class template Message-ID: http://llvm.org/bugs/show_bug.cgi?id=8712 Summary: value dependent member in class template Product: clang Version: trunk Platform: PC OS/Version: Linux Status: NEW Severity: normal Priority: P Component: C++ AssignedTo: unassignedclangbugs at nondot.org ReportedBy: timoskrempel at gmx.de CC: llvmbugs at cs.uiuc.edu, dgregor at apple.com The following code works an compiles in 2.8 but compilation fails in trunk (rev 120287): template class B { public: B(const unsigned char i); unsigned char value : (dim > 0 ? dim : 1); }; template inline B::B(const unsigned char i) : value(i) {} int main() {} with the following error: $ clang++ bug3.cc clang: ExprConstant.cpp:2458: ::ICEDiag CheckICE(const clang::Expr*, clang::ASTContext&): Assertion `!E->isValueDependent() && "Should not see value dependent exprs!"' failed. 0 clang 0x00000000016971cf 1 clang 0x0000000001699442 2 libpthread.so.0 0x00007fcf772f8b40 3 libc.so.6 0x00007fcf765f6ba5 gsignal + 53 4 libc.so.6 0x00007fcf765fa6b0 abort + 384 5 libc.so.6 0x00007fcf765efa71 __assert_fail + 241 6 clang 0x0000000000c8ecb9 7 clang 0x0000000000c8edef clang::Expr::isIntegerConstantExpr(llvm::APSInt&, clang::ASTContext&, clang::SourceLocation*, bool) const + 63 8 clang 0x00000000008e13cc 9 clang 0x00000000009e2cad clang::InitializationSequence::Perform(clang::Sema&, clang::InitializedEntity const&, clang::InitializationKind const&, clang::ASTMultiPtr, clang::QualType*) + 301 10 clang 0x0000000000952a68 clang::Sema::BuildMemberInitializer(clang::FieldDecl*, clang::Expr**, unsigned int, clang::SourceLocation, clang::SourceLocation, clang::SourceLocation) + 712 11 clang 0x0000000000968f59 clang::Sema::ActOnMemInitializer(clang::Decl*, clang::Scope*, clang::CXXScopeSpec&, clang::IdentifierInfo*, clang::OpaquePtr, clang::SourceLocation, clang::SourceLocation, clang::Expr**, unsigned int, clang::SourceLocation) + 697 12 clang 0x000000000089c44e clang::Parser::ParseMemInitializer(clang::Decl*) + 750 13 clang 0x000000000089c5e5 clang::Parser::ParseConstructorInitializer(clang::Decl*) + 213 14 clang 0x000000000088bd4b clang::Parser::ParseFunctionDefinition(clang::Parser::ParsingDeclarator&, clang::Parser::ParsedTemplateInfo const&) + 555 15 clang 0x00000000008b6e9e clang::Parser::ParseSingleDeclarationAfterTemplate(unsigned int, clang::Parser::ParsedTemplateInfo const&, clang::Parser::ParsingDeclRAIIObject&, clang::SourceLocation&, clang::AccessSpecifier) + 2382 16 clang 0x00000000008b9305 clang::Parser::ParseTemplateDeclarationOrSpecialization(unsigned int, clang::SourceLocation&, clang::AccessSpecifier) + 693 17 clang 0x0000000000896929 clang::Parser::ParseDeclaration(clang::ASTOwningVector&, unsigned int, clang::SourceLocation&, clang::CXX0XAttributeList) + 409 18 clang 0x000000000088ccd3 clang::Parser::ParseExternalDeclaration(clang::CXX0XAttributeList, clang::Parser::ParsingDeclSpec*) + 1091 19 clang 0x000000000088cf12 clang::Parser::ParseTopLevelDecl(clang::OpaquePtr&) + 114 20 clang 0x000000000087541b clang::ParseAST(clang::Sema&, bool) + 139 21 clang 0x0000000000747b34 clang::CodeGenAction::ExecuteAction() + 68 22 clang 0x0000000000641c75 clang::CompilerInstance::ExecuteAction(clang::FrontendAction&) + 357 23 clang 0x00000000006219ec clang::ExecuteCompilerInvocation(clang::CompilerInstance*) + 1260 24 clang 0x0000000000619775 cc1_main(char const**, char const**, char const*, void*) + 693 25 clang 0x0000000000620965 main + 4181 26 libc.so.6 0x00007fcf765e1d8e __libc_start_main + 254 27 clang 0x0000000000617e89 Stack dump: 0. Program arguments: /scratch/kram/llvm/Release+Asserts/bin/clang -cc1 -triple x86_64-unknown-linux-gnu -emit-obj -mrelax-all -disable-free -main-file-name bug3.cc -mrelocation-model static -mdisable-fp-elim -masm-verbose -mconstructor-aliases -munwind-tables -target-cpu x86-64 -target-linker-version 2.20.51 -resource-dir /scratch/kram/llvm/Release+Asserts/bin/../lib/clang/2.9 -ferror-limit 19 -fmessage-length 165 -fexceptions -fgnu-runtime -fdiagnostics-show-option -fcolor-diagnostics -o /tmp/cc-BRMnQS.o -x c++ bug3.cc 1. bug3.cc:13:1: current parser token '{' clang: error: unable to execute command: Aborted clang: error: clang frontend command failed due to signal 1 (use -v to see invocation) -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Wed Dec 1 05:16:29 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Wed, 1 Dec 2010 05:16:29 -0600 (CST) Subject: [LLVMbugs] [Bug 8713] New: "relocation R_X86_64_PC32 against undefined symbol ..." Message-ID: http://llvm.org/bugs/show_bug.cgi?id=8713 Summary: "relocation R_X86_64_PC32 against undefined symbol ..." Product: clang Version: trunk Platform: PC OS/Version: Linux Status: NEW Severity: normal Priority: P Component: -New Bugs AssignedTo: unassignedclangbugs at nondot.org ReportedBy: hans at chromium.org CC: llvmbugs at cs.uiuc.edu Trying to compile and link the following into a shared library fails: #include extern std::ostream& get_ostream(); void f() { get_ostream() << "foo"; } $ clang++ -fno-exceptions -fvisibility=hidden -fPIC -fno-rtti -fno-threadsafe-statics -fvisibility-inlines-hidden -c a.cc -o fail.o $ clang++ -shared -o fail.so fail.o /usr/bin/ld: fail.o: relocation R_X86_64_PC32 against undefined symbol `_ZStlsISt11char_traitsIcEERSt13basic_ostreamIcT_ES5_PKc@@GLIBCXX_3.4' can not be used when making a shared object; recompile with -fPIC /usr/bin/ld: final link failed: Bad value clang: error: linker command failed with exit code 1 (use -v to see invocation) If g++ is used in the first step, the link step succeeds. $ clang++ -v clang version 2.9 (trunk 120563) Target: x86_64-unknown-linux-gnu Thread model: posix gcc version 4.4.3 (Ubuntu 4.4.3-4ubuntu5) -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Wed Dec 1 05:57:44 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Wed, 1 Dec 2010 05:57:44 -0600 (CST) Subject: [LLVMbugs] [Bug 8714] New: Jump threading doesn't handle indirectbr very well. Message-ID: http://llvm.org/bugs/show_bug.cgi?id=8714 Summary: Jump threading doesn't handle indirectbr very well. Product: new-bugs Version: trunk Platform: All OS/Version: All Status: NEW Severity: normal Priority: P Component: new bugs AssignedTo: unassignedbugs at nondot.org ReportedBy: fvbommel at gmail.com CC: llvmbugs at cs.uiuc.edu Jump threading only threads through blocks ending in conditional branches or switches, not 'indirectbr'. These are probably even more profitable to thread through than switches. -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Wed Dec 1 09:22:29 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Wed, 1 Dec 2010 09:22:29 -0600 (CST) Subject: [LLVMbugs] [Bug 8715] New: incorrect .debug_line section entries in generated .s file Message-ID: http://llvm.org/bugs/show_bug.cgi?id=8715 Summary: incorrect .debug_line section entries in generated .s file Product: libraries Version: trunk Platform: Macintosh OS/Version: MacOS X Status: NEW Severity: normal Priority: P Component: Common Code Generator Code AssignedTo: unassignedbugs at nondot.org ReportedBy: dpatel at apple.com CC: llvmbugs at cs.uiuc.edu For the following input --- begin --- char c[4] = "hi\n"; --- end --- llvm emits incorrect .debug_line entries on darwin9 .section __DWARF,__debug_line,regular,debug Ltmp0: .long (Ltmp1-Ltmp0)-4 .short 2 .long (Ltmp2-Ltmp0)-10 .byte 1 .byte 1 .byte -5 .byte 14 .byte 13 .byte 0 .byte 1 .byte 1 .byte 1 .byte 1 .byte 0 .byte 0 .byte 0 .byte 1 .byte 0 .byte 0 .byte 1 .byte 0 .ascii "g1.c" .byte 0 .byte 0 .byte 0 .byte 0 .byte 0 Ltmp2: Ltmp1: This is a regression. -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Wed Dec 1 10:04:36 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Wed, 1 Dec 2010 10:04:36 -0600 (CST) Subject: [LLVMbugs] [Bug 8715] incorrect .debug_line section entries in generated .s file In-Reply-To: References: Message-ID: <20101201160436.A20062A6C12C@llvm.org> http://llvm.org/bugs/show_bug.cgi?id=8715 Rafael ?vila de Esp?ndola changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |RESOLVED Resolution| |WORKSFORME -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Wed Dec 1 10:10:52 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Wed, 1 Dec 2010 10:10:52 -0600 (CST) Subject: [LLVMbugs] [Bug 8715] incorrect .debug_line section entries in generated .s file In-Reply-To: References: Message-ID: <20101201161052.D74D22A6C12F@llvm.org> http://llvm.org/bugs/show_bug.cgi?id=8715 Devang Patel changed: What |Removed |Added ---------------------------------------------------------------------------- Status|RESOLVED |REOPENED Resolution|WORKSFORME | --- Comment #4 from Devang Patel 2010-12-01 10:10:52 CST --- I am reopening this because this is a darwin9 issue. The assembler+linker on darwin9 are not able to handle this. The linker crashes in fact. This could be a bug in linker. However, this is a new regression caused by llvm code generator changes in .debug_line output in .s file. FWIW, here is what llvm code generator used to emit before .loc related changes were introduced. .section __DWARF,__debug_line,regular,debug Lset2 = Lline_end-Lline_begin ## Length of Source Line Info .long Lset2 Lline_begin: .short 2 ## DWARF version number Lset3 = Lline_prolog_end-Lline_prolog_begin ## Prolog Length .long Lset3 Lline_prolog_begin: .byte 1 ## Minimum Instruction Length .byte 1 ## Default is_stmt_start flag .byte 246 ## Line Base Value (Special Opcodes) .byte 245 ## Line Range Value (Special Opcodes) .byte 10 ## Special Opcode Base .byte 0 ## DW_LNS_copy arg count .byte 1 ## DW_LNS_advance_pc arg count .byte 1 ## DW_LNS_advance_line arg count .byte 1 ## DW_LNS_set_file arg count .byte 1 ## DW_LNS_set_column arg count .byte 0 ## DW_LNS_negate_stmt arg count .byte 0 ## DW_LNS_set_basic_block arg count .byte 0 ## DW_LNS_const_add_pc arg count .byte 1 ## DW_LNS_fixed_advance_pc arg count .asciz "/tmp/" ## Directory .byte 0 ## End of directories .asciz "g.c" ## Source .byte 1 ## Directory # .byte 0 ## Mod date .byte 0 ## File size .byte 0 ## End of files Lline_prolog_end: .byte 0 ## Extended Op .byte 9 ## Op size .byte 2 ## DW_LNE_set_address .quad Lsection_end1 ## Section end label .byte 0 ## DW_LNE_end_sequence .byte 1 .byte 1 Lline_end: -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Wed Dec 1 10:20:54 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Wed, 1 Dec 2010 10:20:54 -0600 (CST) Subject: [LLVMbugs] [Bug 8716] New: Crash in Live Variable Analysis Message-ID: http://llvm.org/bugs/show_bug.cgi?id=8716 Summary: Crash in Live Variable Analysis Product: new-bugs Version: trunk Platform: PC OS/Version: FreeBSD Status: NEW Severity: normal Priority: P Component: new bugs AssignedTo: unassignedbugs at nondot.org ReportedBy: pawel.worach at gmail.com CC: llvmbugs at cs.uiuc.edu Created an attachment (id=5846) --> (http://llvm.org/bugs/attachment.cgi?id=5846) bugpoint-reduced-simplified.bc $ llc bugpoint-reduced-simplified.bcStack dump: 0. Program arguments: /data/buildslave/freebsd-clang-amd64/obj/obj-llvm.2/Release+Asserts/bin/llc bugpoint-reduced-simplified.bc 1. Running pass 'Function Pass Manager' on module 'bugpoint-reduced-simplified.bc'. 2. Running pass 'Live Variable Analysis' on function '@tgamma' Segmentation fault (core dumped) Reduced from FreeBSD lib/msun/bsdsrc/b_tgamma.c. -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Wed Dec 1 10:46:25 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Wed, 1 Dec 2010 10:46:25 -0600 (CST) Subject: [LLVMbugs] [Bug 8717] New: LLVM MC does not know how to handle .subsection derective Message-ID: http://llvm.org/bugs/show_bug.cgi?id=8717 Summary: LLVM MC does not know how to handle .subsection derective Product: new-bugs Version: trunk Platform: PC OS/Version: Linux Status: NEW Severity: normal Priority: P Component: new bugs AssignedTo: unassignedbugs at nondot.org ReportedBy: kaffeemonster at googlemail.com CC: llvmbugs at cs.uiuc.edu After a long time testing llvm+clang again. llvm/clang @ TOT, r120576 Given the following (silly example) code: int foo(int a) { int x; asm( "1:\n\t" "dec %0\n\t" "jnz 2f\n\t" /* * some procesessors see jumping back as likely * so jump forward, then backward */ ".subsection 2\n" "2:\n\t" "jmp 1b\n\t" ".previous" : "=r" (x) : "0" (a) ); return x; } clang gives: $ clang -O -c clang_sub.c clang_sub.c:7:12: error: expected identifier in directive "jnz 2f\n\t" ^ :4:14: note: instantiated into assembly here .subsection 2 ^ 1 error generated. Looks like the new builtin asm foo is to blame. I tried to make a patch (attached, don't shoot me, it's just a test, i have no clue about C++ or llvm), but it does not solve it. Greetings Jan -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Wed Dec 1 12:53:30 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Wed, 1 Dec 2010 12:53:30 -0600 (CST) Subject: [LLVMbugs] [Bug 8716] Crash in Live Variable Analysis In-Reply-To: References: Message-ID: <20101201185330.810D72A6C12D@llvm.org> http://llvm.org/bugs/show_bug.cgi?id=8716 Jakob Stoklund Olesen changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |RESOLVED Resolution| |WORKSFORME --- Comment #3 from Jakob Stoklund Olesen 2010-12-01 12:53:30 CST --- This no longer reproduces on TOT. Please verify. (I was able to reproduce with r120526, but not r120580) -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Wed Dec 1 12:56:31 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Wed, 1 Dec 2010 12:56:31 -0600 (CST) Subject: [LLVMbugs] [Bug 8718] New: No warning on casting between pointer and non-pointer-sized int Message-ID: http://llvm.org/bugs/show_bug.cgi?id=8718 Summary: No warning on casting between pointer and non-pointer-sized int Product: clang Version: trunk Platform: Macintosh OS/Version: MacOS X Status: NEW Severity: normal Priority: P Component: Frontend AssignedTo: unassignedclangbugs at nondot.org ReportedBy: tjw at omnigroup.com CC: llvmbugs at cs.uiuc.edu llvm-gcc-4.2 and gcc-4.2 correctly report warnings when casting between pointers and integer types of non-pointer width: llvm-gcc-4.2 -Wall -c int-ptr-cast-size-mismatch.c -o /tmp/x int-ptr-cast-size-mismatch.c: In function ?cast_from_pointer?: int-ptr-cast-size-mismatch.c:4: warning: cast from pointer to integer of different size int-ptr-cast-size-mismatch.c: In function ?cast_to_pointer?: int-ptr-cast-size-mismatch.c:9: warning: cast to pointer from integer of different size clang, as shipped in Xcode 3.2.5 and from trunk r120583, reports no warnings, however. -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Wed Dec 1 13:37:09 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Wed, 1 Dec 2010 13:37:09 -0600 (CST) Subject: [LLVMbugs] [Bug 8692] use of default argument to function that is declared later in class In-Reply-To: References: Message-ID: <20101201193709.325EE2A6C130@llvm.org> http://llvm.org/bugs/show_bug.cgi?id=8692 Douglas Gregor changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |RESOLVED Resolution| |FIXED --- Comment #3 from Douglas Gregor 2010-12-01 13:37:08 CST --- Yes, this is fixed. -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Wed Dec 1 14:53:01 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Wed, 1 Dec 2010 14:53:01 -0600 (CST) Subject: [LLVMbugs] [Bug 8603] Invalid assembly accepted on ELF In-Reply-To: References: Message-ID: <20101201205301.10D292A6C130@llvm.org> http://llvm.org/bugs/show_bug.cgi?id=8603 Rafael ?vila de Esp?ndola changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |RESOLVED Resolution| |FIXED -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Wed Dec 1 14:53:16 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Wed, 1 Dec 2010 14:53:16 -0600 (CST) Subject: [LLVMbugs] [Bug 8603] Invalid assembly accepted on ELF In-Reply-To: References: Message-ID: <20101201205316.05EF82A6C130@llvm.org> http://llvm.org/bugs/show_bug.cgi?id=8603 Rafael ?vila de Esp?ndola changed: What |Removed |Added ---------------------------------------------------------------------------- Status|RESOLVED |REOPENED Resolution|FIXED | -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Wed Dec 1 14:53:48 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Wed, 1 Dec 2010 14:53:48 -0600 (CST) Subject: [LLVMbugs] [Bug 8596] MC Temporary Symbol Name Clash In-Reply-To: References: Message-ID: <20101201205348.B33482A6C12C@llvm.org> http://llvm.org/bugs/show_bug.cgi?id=8596 Rafael ?vila de Esp?ndola changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |RESOLVED Resolution| |FIXED -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Wed Dec 1 15:57:14 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Wed, 1 Dec 2010 15:57:14 -0600 (CST) Subject: [LLVMbugs] [Bug 8719] New: clang assembler asserts Message-ID: http://llvm.org/bugs/show_bug.cgi?id=8719 Summary: clang assembler asserts Product: libraries Version: trunk Platform: PC OS/Version: Linux Status: NEW Severity: normal Priority: P Component: LLVM assembly language parser AssignedTo: unassignedbugs at nondot.org ReportedBy: tjablin at cs.princeton.edu CC: llvmbugs at cs.uiuc.edu Created an attachment (id=5853) --> (http://llvm.org/bugs/attachment.cgi?id=5853) Assembly Language testcase The clang assembler asserts on the attached two line reduced assembly-code testcase. .file 1 "cfd.cpp" .Ltmp0: -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Wed Dec 1 16:05:29 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Wed, 1 Dec 2010 16:05:29 -0600 (CST) Subject: [LLVMbugs] [Bug 8720] New: llvm-gcc testsuite regressions with symbol aliases Message-ID: http://llvm.org/bugs/show_bug.cgi?id=8720 Summary: llvm-gcc testsuite regressions with symbol aliases Product: new-bugs Version: unspecified Platform: All OS/Version: All Status: NEW Severity: normal Priority: P Component: new bugs AssignedTo: unassignedbugs at nondot.org ReportedBy: gohman at apple.com CC: llvmbugs at cs.uiuc.edu The following tests appear to have started failing with r119152 in "clang-i386-darwin10": gcc.dg/alias-5.c (test for excess errors) gcc.dg/special/alias-1.c (test for excess errors) gcc.dg/special/alias-2.c (test for excess errors) gcc.dg/special/wkali-1.c (test for excess errors) -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Wed Dec 1 16:49:19 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Wed, 1 Dec 2010 16:49:19 -0600 (CST) Subject: [LLVMbugs] [Bug 8721] New: AsmParser does not understand several ARM specific constructs Message-ID: http://llvm.org/bugs/show_bug.cgi?id=8721 Summary: AsmParser does not understand several ARM specific constructs Product: libraries Version: trunk Platform: PC OS/Version: Linux Status: NEW Severity: normal Priority: P Component: Backend: ARM AssignedTo: unassignedbugs at nondot.org ReportedBy: jasonwkim at google.com CC: llvmbugs at cs.uiuc.edu :lower16: :upper16: not supported llvm-mc -triple=armv7-linux-gnueabi -arch=arm small.s -o small.o -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Wed Dec 1 18:35:08 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Wed, 1 Dec 2010 18:35:08 -0600 (CST) Subject: [LLVMbugs] [Bug 8719] clang assembler asserts In-Reply-To: References: Message-ID: <20101202003508.121552A6C130@llvm.org> http://llvm.org/bugs/show_bug.cgi?id=8719 Thomas B. Jablin changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |RESOLVED Resolution| |DUPLICATE --- Comment #3 from Thomas B. Jablin 2010-12-01 18:35:07 CST --- *** This bug has been marked as a duplicate of bug 8596 *** -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Wed Dec 1 19:20:30 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Wed, 1 Dec 2010 19:20:30 -0600 (CST) Subject: [LLVMbugs] [Bug 8715] incorrect .debug_line section entries in generated .s file In-Reply-To: References: Message-ID: <20101202012030.7627F2A6C131@llvm.org> http://llvm.org/bugs/show_bug.cgi?id=8715 Devang Patel changed: What |Removed |Added ---------------------------------------------------------------------------- Status|REOPENED |RESOLVED Resolution| |FIXED --- Comment #10 from Devang Patel 2010-12-01 19:20:29 CST --- Thanks Nick. Instead of trying to emit dummy entry, why not just skip .debug_line section entirely if there is not any line info ? Fixed in r120637. This lets me pass this test case, but llvm-gcc bootstrap sill fails. This time there are line table entries! I'll use another bugzilla, if I need, for it though. -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Thu Dec 2 00:51:07 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Thu, 2 Dec 2010 00:51:07 -0600 (CST) Subject: [LLVMbugs] [Bug 8722] New: 'clang --analyze' crashes on casting from an object L-value to a reference Message-ID: http://llvm.org/bugs/show_bug.cgi?id=8722 Summary: 'clang --analyze' crashes on casting from an object L-value to a reference Product: clang Version: trunk Platform: All OS/Version: All Status: NEW Severity: normal Priority: P Component: Static Analyzer AssignedTo: kremenek at apple.com ReportedBy: zhanyong.wan at gmail.com CC: llvmbugs at cs.uiuc.edu $ cat test.cc class Foo {}; void Test() { int n; reinterpret_cast(n); // fine Foo foo; reinterpret_cast(foo); // crashes 'clang --analyze' } $ clang --analyze test.cc 0 clang 0x00000000012e1910 1 clang 0x00000000012e1e1a 2 libpthread.so.0 0x00002b34758db9f0 3 clang 0x00000000006fca52 clang::StoreManager::CastRegion(clang::MemRegion const*, clang::QualType) + 322 4 clang 0x0000000000700904 clang::SValuator::EvalCast(clang::SVal, clang::QualType, clang::QualType) + 1364 5 clang 0x0000000000774d7e clang::GRExprEngine::VisitCast(clang::CastExpr const*, clang::Expr const*, clang::ExplodedNode*, clang::ExplodedNodeSet&, bool) + 1438 6 clang 0x0000000000769443 clang::GRExprEngine::Visit(clang::Stmt const*, clang::ExplodedNode*, clang::ExplodedNodeSet&) + 819 7 clang 0x0000000000774949 clang::GRExprEngine::VisitCast(clang::CastExpr const*, clang::Expr const*, clang::ExplodedNode*, clang::ExplodedNodeSet&, bool) + 361 8 clang 0x0000000000769443 clang::GRExprEngine::Visit(clang::Stmt const*, clang::ExplodedNode*, clang::ExplodedNodeSet&) + 819 9 clang 0x00000000007754ae clang::GRExprEngine::ProcessStmt(clang::CFGStmt, clang::GRStmtNodeBuilder&) + 798 10 clang 0x0000000000775b19 clang::GRExprEngine::ProcessElement(clang::CFGElement, clang::GRStmtNodeBuilder&) + 121 11 clang 0x000000000074eca4 clang::GRCoreEngine::ExecuteWorkList(clang::LocationContext const*, unsigned int, clang::GRState const*) + 500 12 clang 0x00000000006af6f2 13 clang 0x00000000006b1666 14 clang 0x00000000008c84ed clang::ParseAST(clang::Sema&, bool) + 557 15 clang 0x00000000007eec9b clang::CompilerInstance::ExecuteAction(clang::FrontendAction&) + 315 16 clang 0x000000000056e18d clang::ExecuteCompilerInvocation(clang::CompilerInstance*) + 989 17 clang 0x0000000000568886 cc1_main(char const**, char const**, char const*, void*) + 982 18 clang 0x000000000056d3b7 main + 3639 19 libc.so.6 0x00002b3476033d5d __libc_start_main + 253 20 clang 0x00000000005659fe Stack dump: 0. Program arguments: clang -cc1 ... test.cc 1. parser at end of file 2. test.cc:8:3: Error evaluating statement 3. test.cc:8:3: Error evaluating statement 4. test.cc:8:9: Error evaluating statement clang: error: unable to execute command: Segmentation fault clang: error: clang frontend command failed due to signal 1 (use -v to see invocation) -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Thu Dec 2 01:07:45 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Thu, 2 Dec 2010 01:07:45 -0600 (CST) Subject: [LLVMbugs] [Bug 279] Clang initialization code generation can be improved In-Reply-To: References: Message-ID: <20101202070745.493DD2A6C12D@llvm.org> http://llvm.org/bugs/show_bug.cgi?id=279 Chris Lattner changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |RESOLVED Resolution| |FIXED --- Comment #6 from Chris Lattner 2010-12-02 01:07:44 CST --- Finished off in r120692. We now generate a memset + 1 store. -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Thu Dec 2 01:09:59 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Thu, 2 Dec 2010 01:09:59 -0600 (CST) Subject: [LLVMbugs] [Bug 792] Factor AsmWriter out of lib/VMCore In-Reply-To: References: Message-ID: <20101202070959.811C02A6C131@llvm.org> http://llvm.org/bugs/show_bug.cgi?id=792 Chris Lattner changed: What |Removed |Added ---------------------------------------------------------------------------- Status|ASSIGNED |RESOLVED Resolution| |WONTFIX --- Comment #19 from Chris Lattner 2010-12-02 01:09:57 CST --- It's fair to say that this will never happen, and if it does, that it won't matter. In practice, everything that touches IR would depend on the factored out library, because they call ->dump() in debug paths or error cases. Eliminating the vtable from Value would be a great project, but seems like it is tracked in another bug. -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Thu Dec 2 01:12:45 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Thu, 2 Dec 2010 01:12:45 -0600 (CST) Subject: [LLVMbugs] [Bug 767] Prolog Epilog Inserter should take advantage of larger aliases In-Reply-To: References: Message-ID: <20101202071245.854EC2A6C130@llvm.org> http://llvm.org/bugs/show_bug.cgi?id=767 Chris Lattner changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |RESOLVED Resolution| |LATER --- Comment #1 from Chris Lattner 2010-12-02 01:12:45 CST --- This is already tracked in the PPC readme ("cr save restore is awful"), no need for a separate bugzilla. -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Thu Dec 2 01:13:37 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Thu, 2 Dec 2010 01:13:37 -0600 (CST) Subject: [LLVMbugs] [Bug 8723] New: llvm 2.8: doesn't build on Mac OS X 10.4.11, undefined symbol ___crashreporter_info__ Message-ID: http://llvm.org/bugs/show_bug.cgi?id=8723 Summary: llvm 2.8: doesn't build on Mac OS X 10.4.11, undefined symbol ___crashreporter_info__ Product: new-bugs Version: 2.8 Platform: Macintosh OS/Version: MacOS X Status: NEW Severity: normal Priority: P Component: new bugs AssignedTo: unassignedbugs at nondot.org ReportedBy: edwintorok at gmail.com CC: llvmbugs at cs.uiuc.edu See https://wwws.clamav.net/bugzilla/show_bug.cgi?id=2402 Using gcc 4.5 on LLVM 2.8 make fails with: /usr/bin/ld: Undefined symbols: ___crashreporter_info__ collect2: ld returned 1 exit status make[2]: *** [/Users//Documents/src/llvm-2.8/Release/bin/FileCheck] Error 1 Using the (Apple provided) gcc 3.3 fails too: llvm[1]: Compiling Alarm.cpp for Release build cc1plus: error: invalid option `macosx-version-min=10.4' rm: /Users//Documents/src/llvm-2.8/lib/System/Release/Alarm.d.tmp: No such file or directory make[1]: *** [/Users//Documents/src/llvm-2.8/lib/System/Release/Alarm.o] Error 1 According to the user there is no other Apple provided gcc for this platform that he could try (well there is 4.0.1, but that is buggy). -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Thu Dec 2 01:13:41 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Thu, 2 Dec 2010 01:13:41 -0600 (CST) Subject: [LLVMbugs] [Bug 8125] switches over a 2-bit domain lowered inefficiently In-Reply-To: References: Message-ID: <20101202071341.50B982A6C132@llvm.org> http://llvm.org/bugs/show_bug.cgi?id=8125 Chris Lattner changed: What |Removed |Added ---------------------------------------------------------------------------- Status|ASSIGNED |RESOLVED Resolution| |DUPLICATE --- Comment #5 from Chris Lattner 2010-12-02 01:13:41 CST --- *** This bug has been marked as a duplicate of bug 774 *** -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Thu Dec 2 01:15:56 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Thu, 2 Dec 2010 01:15:56 -0600 (CST) Subject: [LLVMbugs] [Bug 605] [CBE] C Type layout doesn't match LLVM type layout In-Reply-To: References: Message-ID: <20101202071556.BAC822A6C131@llvm.org> http://llvm.org/bugs/show_bug.cgi?id=605 Chris Lattner changed: What |Removed |Added ---------------------------------------------------------------------------- Status|REOPENED |RESOLVED Resolution| |WONTFIX --- Comment #6 from Chris Lattner 2010-12-02 01:15:56 CST --- Noone seems interested in fixing this, it would require a substantial rewrite of the CBE. -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Thu Dec 2 01:17:38 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Thu, 2 Dec 2010 01:17:38 -0600 (CST) Subject: [LLVMbugs] [Bug 1391] [llvm-gcc] non-local gotos not supported In-Reply-To: References: Message-ID: <20101202071738.898222A6C132@llvm.org> http://llvm.org/bugs/show_bug.cgi?id=1391 Chris Lattner changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |RESOLVED Resolution| |WONTFIX --- Comment #1 from Chris Lattner 2010-12-02 01:17:38 CST --- Noone cares, and clang won't support nested functions. -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Thu Dec 2 01:17:54 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Thu, 2 Dec 2010 01:17:54 -0600 (CST) Subject: [LLVMbugs] [Bug 1150] stage2 compiles c++ files with the host c++ compiler In-Reply-To: References: Message-ID: <20101202071754.CF4CA2A6C132@llvm.org> http://llvm.org/bugs/show_bug.cgi?id=1150 Chris Lattner changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |RESOLVED Resolution| |WONTFIX --- Comment #11 from Chris Lattner 2010-12-02 01:17:53 CST --- Noone cares. -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Thu Dec 2 01:18:57 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Thu, 2 Dec 2010 01:18:57 -0600 (CST) Subject: [LLVMbugs] [Bug 1168] LLVM doesn't support "complex" type in "asm" statements In-Reply-To: References: Message-ID: <20101202071857.6FF362A6C136@llvm.org> http://llvm.org/bugs/show_bug.cgi?id=1168 Chris Lattner changed: What |Removed |Added ---------------------------------------------------------------------------- Status|REOPENED |RESOLVED Resolution| |FIXED --- Comment #14 from Chris Lattner 2010-12-02 01:18:56 CST --- I think that this is fixed in clang, it handles the case in Comment 2. -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Thu Dec 2 01:19:15 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Thu, 2 Dec 2010 01:19:15 -0600 (CST) Subject: [LLVMbugs] [Bug 1166] [meta] gcc testsuite issues In-Reply-To: References: Message-ID: <20101202071915.1DFA62A6C139@llvm.org> http://llvm.org/bugs/show_bug.cgi?id=1166 Chris Lattner changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |RESOLVED Resolution| |FIXED --- Comment #3 from Chris Lattner 2010-12-02 01:19:14 CST --- All the blocking bugs are fixed, this isn't useful anymore. -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Thu Dec 2 01:20:47 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Thu, 2 Dec 2010 01:20:47 -0600 (CST) Subject: [LLVMbugs] [Bug 1392] Need support for true tail calls In-Reply-To: References: Message-ID: <20101202072047.68E5C2A6C13A@llvm.org> http://llvm.org/bugs/show_bug.cgi?id=1392 Chris Lattner changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |RESOLVED Resolution| |FIXED --- Comment #2 from Chris Lattner 2010-12-02 01:20:47 CST --- Per comment #1, arnold implemented this a long time ago. -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Thu Dec 2 01:21:58 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Thu, 2 Dec 2010 01:21:58 -0600 (CST) Subject: [LLVMbugs] [Bug 1590] split llvm support from non-llvm support libraries In-Reply-To: References: Message-ID: <20101202072158.C05C12A6C139@llvm.org> http://llvm.org/bugs/show_bug.cgi?id=1590 Chris Lattner changed: What |Removed |Added ---------------------------------------------------------------------------- Status|REOPENED |RESOLVED Resolution| |LATER --- Comment #6 from Chris Lattner 2010-12-02 01:21:58 CST --- There is no plan to do this splitting out anymore. -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Thu Dec 2 01:22:27 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Thu, 2 Dec 2010 01:22:27 -0600 (CST) Subject: [LLVMbugs] [Bug 1623] The testsuite should test llvm-gcc codegen directly In-Reply-To: References: Message-ID: <20101202072227.D73692A6C137@llvm.org> http://llvm.org/bugs/show_bug.cgi?id=1623 Chris Lattner changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |RESOLVED Resolution| |FIXED --- Comment #2 from Chris Lattner 2010-12-02 01:22:27 CST --- I think daniels TEST=Simple fixed this. -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Thu Dec 2 01:24:58 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Thu, 2 Dec 2010 01:24:58 -0600 (CST) Subject: [LLVMbugs] [Bug 1838] x86 backend doesn't insert emms when it is needed In-Reply-To: References: Message-ID: <20101202072458.061AA2A6C139@llvm.org> http://llvm.org/bugs/show_bug.cgi?id=1838 Chris Lattner changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |RESOLVED Resolution| |FIXED --- Comment #6 from Chris Lattner 2010-12-02 01:24:57 CST --- We fixed this by having the x86 backend not implicitly insert mmx ops. -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Thu Dec 2 01:25:14 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Thu, 2 Dec 2010 01:25:14 -0600 (CST) Subject: [LLVMbugs] [Bug 1854] nightly tester is not recording users who check out from svn In-Reply-To: References: Message-ID: <20101202072514.790552A6C139@llvm.org> http://llvm.org/bugs/show_bug.cgi?id=1854 Chris Lattner changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |RESOLVED Resolution| |FIXED --- Comment #2 from Chris Lattner 2010-12-02 01:25:14 CST --- The buildbot handles this now. -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Thu Dec 2 01:25:52 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Thu, 2 Dec 2010 01:25:52 -0600 (CST) Subject: [LLVMbugs] [Bug 1917] Use address space markers to improve GC In-Reply-To: References: Message-ID: <20101202072552.CC7D02A6C139@llvm.org> http://llvm.org/bugs/show_bug.cgi?id=1917 Chris Lattner changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |RESOLVED Resolution| |WONTFIX --- Comment #3 from Chris Lattner 2010-12-02 01:25:52 CST --- This is probably not a good idea, sorry I filed it ;-) -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Thu Dec 2 01:28:15 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Thu, 2 Dec 2010 01:28:15 -0600 (CST) Subject: [LLVMbugs] [Bug 2073] CBE barfs on constant struct containing long double In-Reply-To: References: Message-ID: <20101202072815.5AE6F2A6C13A@llvm.org> http://llvm.org/bugs/show_bug.cgi?id=2073 Chris Lattner changed: What |Removed |Added ---------------------------------------------------------------------------- Status|ASSIGNED |RESOLVED Resolution| |LATER --- Comment #5 from Chris Lattner 2010-12-02 01:28:14 CST --- This isn't reasonably fixable in the current CBE. -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Thu Dec 2 01:28:49 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Thu, 2 Dec 2010 01:28:49 -0600 (CST) Subject: [LLVMbugs] [Bug 2153] simd-4.C fails at -O2 In-Reply-To: References: Message-ID: <20101202072849.64BDA2A6C142@llvm.org> http://llvm.org/bugs/show_bug.cgi?id=2153 Chris Lattner changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |RESOLVED Resolution| |LATER --- Comment #4 from Chris Lattner 2010-12-02 01:28:48 CST --- -fnon-call-exceptions is tracked elsewhere. -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Thu Dec 2 01:29:58 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Thu, 2 Dec 2010 01:29:58 -0600 (CST) Subject: [LLVMbugs] [Bug 2175] [array_ref hack] llvm-gcc abort on tricky VLA cast In-Reply-To: References: Message-ID: <20101202072958.6174B2A6C13F@llvm.org> http://llvm.org/bugs/show_bug.cgi?id=2175 Chris Lattner changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |RESOLVED Resolution| |WONTFIX --- Comment #3 from Chris Lattner 2010-12-02 01:29:57 CST --- I have no plan to fix this. -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Thu Dec 2 01:30:55 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Thu, 2 Dec 2010 01:30:55 -0600 (CST) Subject: [LLVMbugs] [Bug 2399] cbe generates illegal code for sret of array In-Reply-To: References: Message-ID: <20101202073055.E97FE2A6C13F@llvm.org> http://llvm.org/bugs/show_bug.cgi?id=2399 Chris Lattner changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |RESOLVED Resolution| |WONTFIX --- Comment #3 from Chris Lattner 2010-12-02 01:30:55 CST --- This is very very difficult to fix in the CBE, maybe if it ever gets redesigned. -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Thu Dec 2 01:32:58 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Thu, 2 Dec 2010 01:32:58 -0600 (CST) Subject: [LLVMbugs] [Bug 2905] Implement a pretty printer for LLVM assembly. In-Reply-To: References: Message-ID: <20101202073258.A342B2A6C13F@llvm.org> http://llvm.org/bugs/show_bug.cgi?id=2905 Chris Lattner changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |RESOLVED Resolution| |LATER --- Comment #4 from Chris Lattner 2010-12-02 01:32:58 CST --- I don't think anyone is interested in doing this. -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Thu Dec 2 01:33:47 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Thu, 2 Dec 2010 01:33:47 -0600 (CST) Subject: [LLVMbugs] [Bug 3370] Eliminate the vicmp and vfcmp instructions In-Reply-To: References: Message-ID: <20101202073347.9C88B2A6C13F@llvm.org> http://llvm.org/bugs/show_bug.cgi?id=3370 Chris Lattner changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |RESOLVED Resolution| |FIXED --- Comment #1 from Chris Lattner 2010-12-02 01:33:47 CST --- THis was done long ago. -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Thu Dec 2 01:34:35 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Thu, 2 Dec 2010 01:34:35 -0600 (CST) Subject: [LLVMbugs] [Bug 3373] cannot tie inline asm operands of different widths In-Reply-To: References: Message-ID: <20101202073435.BC1EC2A6C148@llvm.org> http://llvm.org/bugs/show_bug.cgi?id=3373 Chris Lattner changed: What |Removed |Added ---------------------------------------------------------------------------- Status|REOPENED |RESOLVED Resolution| |FIXED --- Comment #34 from Chris Lattner 2010-12-02 01:34:34 CST --- This is done in clang, and was reopened because llvm-gcc doesn't support it. We don't plan to fix llvm-gcc, so reclosing. If you'd like to keep a bug for this in dragonegg, please file a new one. -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Thu Dec 2 01:35:47 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Thu, 2 Dec 2010 01:35:47 -0600 (CST) Subject: [LLVMbugs] [Bug 3559] llvm-gcc emits debug info for gimple temporaries (non-user vars) In-Reply-To: References: Message-ID: <20101202073547.3019D2A6C13C@llvm.org> http://llvm.org/bugs/show_bug.cgi?id=3559 Chris Lattner changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |RESOLVED Resolution| |FIXED --- Comment #3 from Chris Lattner 2010-12-02 01:35:46 CST --- "fixed in clang" :) -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Thu Dec 2 01:36:02 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Thu, 2 Dec 2010 01:36:02 -0600 (CST) Subject: [LLVMbugs] [Bug 3565] MallocBench/gs crashes In-Reply-To: References: Message-ID: <20101202073602.9CB8D2A6C143@llvm.org> http://llvm.org/bugs/show_bug.cgi?id=3565 Chris Lattner changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |RESOLVED Resolution| |WORKSFORME --- Comment #1 from Chris Lattner 2010-12-02 01:36:02 CST --- Works for me :) -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Thu Dec 2 01:36:34 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Thu, 2 Dec 2010 01:36:34 -0600 (CST) Subject: [LLVMbugs] [Bug 3532] llvm-gcc fails to build Qt4.5-RC1 In-Reply-To: References: Message-ID: <20101202073634.5849A2A6C13F@llvm.org> http://llvm.org/bugs/show_bug.cgi?id=3532 Chris Lattner changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |RESOLVED Resolution| |FIXED --- Comment #4 from Chris Lattner 2010-12-02 01:36:33 CST --- According to the comments, this is fixed. -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Thu Dec 2 01:37:03 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Thu, 2 Dec 2010 01:37:03 -0600 (CST) Subject: [LLVMbugs] [Bug 3623] duplicate "extractvalue" not eliminated In-Reply-To: References: Message-ID: <20101202073703.F21772A6C131@llvm.org> http://llvm.org/bugs/show_bug.cgi?id=3623 Chris Lattner changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |RESOLVED Resolution| |FIXED --- Comment #5 from Chris Lattner 2010-12-02 01:37:03 CST --- GVN gets this now. -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Thu Dec 2 01:37:55 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Thu, 2 Dec 2010 01:37:55 -0600 (CST) Subject: [LLVMbugs] [Bug 3578] ConstantFold doesn't handle vector bitcast? In-Reply-To: References: Message-ID: <20101202073755.9C4C32A6C144@llvm.org> http://llvm.org/bugs/show_bug.cgi?id=3578 Chris Lattner changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |RESOLVED Resolution| |FIXED --- Comment #2 from Chris Lattner 2010-12-02 01:37:55 CST --- THis got fixed, the testcase has even been updated :) -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Thu Dec 2 01:38:40 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Thu, 2 Dec 2010 01:38:40 -0600 (CST) Subject: [LLVMbugs] [Bug 3823] llvm-gcc segmentation fault during build of libstdc++ In-Reply-To: References: Message-ID: <20101202073840.7AF772A6C145@llvm.org> http://llvm.org/bugs/show_bug.cgi?id=3823 Chris Lattner changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |RESOLVED Resolution| |WONTFIX --- Comment #3 from Chris Lattner 2010-12-02 01:38:39 CST --- This is either already fixed or not going to be. -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Thu Dec 2 01:39:35 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Thu, 2 Dec 2010 01:39:35 -0600 (CST) Subject: [LLVMbugs] [Bug 3887] llvm-gcc should not build java by default In-Reply-To: References: Message-ID: <20101202073935.21A092A6C13F@llvm.org> http://llvm.org/bugs/show_bug.cgi?id=3887 Chris Lattner changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |RESOLVED Resolution| |WONTFIX --- Comment #5 from Chris Lattner 2010-12-02 01:39:34 CST --- I don't think that anyone cares to mess with the GCC build process. "fixed in clang" :) -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Thu Dec 2 01:40:32 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Thu, 2 Dec 2010 01:40:32 -0600 (CST) Subject: [LLVMbugs] [Bug 3914] Building fail on vc++ 2008 x64 In-Reply-To: References: Message-ID: <20101202074032.B02672A6C145@llvm.org> http://llvm.org/bugs/show_bug.cgi?id=3914 Chris Lattner changed: What |Removed |Added ---------------------------------------------------------------------------- Status|ASSIGNED |RESOLVED Resolution| |FIXED --- Comment #2 from Chris Lattner 2010-12-02 01:40:31 CST --- I think this got fixed. -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Thu Dec 2 01:41:01 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Thu, 2 Dec 2010 01:41:01 -0600 (CST) Subject: [LLVMbugs] [Bug 3870] gcc testsuite not setup for LLVM (and non-Apple Platforms) In-Reply-To: References: Message-ID: <20101202074102.A3AFD2A6C145@llvm.org> http://llvm.org/bugs/show_bug.cgi?id=3870 Chris Lattner changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |RESOLVED Resolution| |LATER --- Comment #13 from Chris Lattner 2010-12-02 01:41:01 CST --- No, there is nothing actionable left here. -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Thu Dec 2 01:41:36 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Thu, 2 Dec 2010 01:41:36 -0600 (CST) Subject: [LLVMbugs] [Bug 3966] Problems with addrspace(256) = GS in x86 codegen In-Reply-To: References: Message-ID: <20101202074136.41BEE2A6C14A@llvm.org> http://llvm.org/bugs/show_bug.cgi?id=3966 Chris Lattner changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |RESOLVED Resolution| |FIXED --- Comment #3 from Chris Lattner 2010-12-02 01:41:35 CST --- I fixed this awhile ago. -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Thu Dec 2 01:42:54 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Thu, 2 Dec 2010 01:42:54 -0600 (CST) Subject: [LLVMbugs] [Bug 3984] [driver] clang foo.c -std=c++98 doesn't treat inputs as C++ In-Reply-To: References: Message-ID: <20101202074254.6ECB82A6C145@llvm.org> http://llvm.org/bugs/show_bug.cgi?id=3984 Chris Lattner changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |RESOLVED Resolution| |INVALID --- Comment #2 from Chris Lattner 2010-12-02 01:42:54 CST --- We have a c++ frontend now! Not needed :) -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Thu Dec 2 01:43:48 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Thu, 2 Dec 2010 01:43:48 -0600 (CST) Subject: [LLVMbugs] [Bug 4158] support arbitrary-width integers in the CBE In-Reply-To: References: Message-ID: <20101202074348.A143F2A6C14A@llvm.org> http://llvm.org/bugs/show_bug.cgi?id=4158 Chris Lattner changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |RESOLVED Resolution| |WONTFIX --- Comment #7 from Chris Lattner 2010-12-02 01:43:47 CST --- this depends on someone rewriting hte CBE, the related issues aren't worth tracking with individual bugs. -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Thu Dec 2 01:47:40 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Thu, 2 Dec 2010 01:47:40 -0600 (CST) Subject: [LLVMbugs] [Bug 4266] llvm-gcc segfaults on void (*f)(enum b); if b not declared In-Reply-To: References: Message-ID: <20101202074740.51AFF2A6C148@llvm.org> http://llvm.org/bugs/show_bug.cgi?id=4266 Chris Lattner changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |RESOLVED Resolution| |FIXED --- Comment #4 from Chris Lattner 2010-12-02 01:47:39 CST --- Fixed in clang, and presumably dragonegg. -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Thu Dec 2 01:50:55 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Thu, 2 Dec 2010 01:50:55 -0600 (CST) Subject: [LLVMbugs] [Bug 4447] llvm-gcc asserts on gcc extension In-Reply-To: References: Message-ID: <20101202075055.EB2882A6C147@llvm.org> http://llvm.org/bugs/show_bug.cgi?id=4447 Chris Lattner changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |RESOLVED Resolution| |FIXED --- Comment #2 from Chris Lattner 2010-12-02 01:50:55 CST --- Clang heartily rejects this code. -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Thu Dec 2 01:52:04 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Thu, 2 Dec 2010 01:52:04 -0600 (CST) Subject: [LLVMbugs] [Bug 4525] llvm-gcc does not compile In-Reply-To: References: Message-ID: <20101202075205.0C81B2A6C137@llvm.org> http://llvm.org/bugs/show_bug.cgi?id=4525 Chris Lattner changed: What |Removed |Added ---------------------------------------------------------------------------- Status|REOPENED |RESOLVED Resolution| |INVALID --- Comment #5 from Chris Lattner 2010-12-02 01:52:04 CST --- Submitter says it was a config error, should build clang anyway :-) -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Thu Dec 2 01:52:37 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Thu, 2 Dec 2010 01:52:37 -0600 (CST) Subject: [LLVMbugs] [Bug 4562] Miscompliation of GVN In-Reply-To: References: Message-ID: <20101202075237.564A02A6C14B@llvm.org> http://llvm.org/bugs/show_bug.cgi?id=4562 Chris Lattner changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |RESOLVED Resolution| |FIXED --- Comment #2 from Chris Lattner 2010-12-02 01:52:37 CST --- This has probably been fixed. -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Thu Dec 2 01:54:58 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Thu, 2 Dec 2010 01:54:58 -0600 (CST) Subject: [LLVMbugs] [Bug 6194] poor x86-64 ABI calling conv passing of struct with 3 or 4 floats In-Reply-To: References: Message-ID: <20101202075458.EFD662A6C14C@llvm.org> http://llvm.org/bugs/show_bug.cgi?id=6194 Chris Lattner changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |RESOLVED Resolution| |FIXED --- Comment #10 from Chris Lattner 2010-12-02 01:54:58 CST --- Clang generates great code for all of these testcases. -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Thu Dec 2 01:56:06 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Thu, 2 Dec 2010 01:56:06 -0600 (CST) Subject: [LLVMbugs] [Bug 1391] [llvm-gcc] non-local gotos not supported In-Reply-To: References: Message-ID: <20101202075606.2871E2A6C14A@llvm.org> http://llvm.org/bugs/show_bug.cgi?id=1391 Duncan Sands changed: What |Removed |Added ---------------------------------------------------------------------------- Status|RESOLVED |REOPENED Component|llvm-gcc |New Bugs Version|1.0 |trunk Resolution|WONTFIX | AssignedTo|unassignedbugs at nondot.org |baldrick at free.fr Product|tools |dragonegg --- Comment #2 from Duncan Sands 2010-12-02 01:56:05 CST --- Moving to dragonegg. -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Thu Dec 2 02:07:41 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Thu, 2 Dec 2010 02:07:41 -0600 (CST) Subject: [LLVMbugs] [Bug 8124] copy is only important at the start of a method name In-Reply-To: References: Message-ID: <20101202080742.025902A6C13C@llvm.org> http://llvm.org/bugs/show_bug.cgi?id=8124 Ted Kremenek changed: What |Removed |Added ---------------------------------------------------------------------------- Status|REOPENED |RESOLVED Resolution| |INVALID --- Comment #6 from Ted Kremenek 2010-12-02 02:07:41 CST --- Yes, the docs are correct. This convention has been vetted numerous times with the Cocoa guys. -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Thu Dec 2 02:59:25 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Thu, 2 Dec 2010 02:59:25 -0600 (CST) Subject: [LLVMbugs] [Bug 3373] cannot tie inline asm operands of different widths In-Reply-To: References: Message-ID: <20101202085925.EC21F2A6C13E@llvm.org> http://llvm.org/bugs/show_bug.cgi?id=3373 Duncan Sands changed: What |Removed |Added ---------------------------------------------------------------------------- Status|RESOLVED |REOPENED Component|new bugs |New Bugs Version|unspecified |trunk Resolution|FIXED | AssignedTo|unassignedbugs at nondot.org |baldrick at free.fr Product|new-bugs |dragonegg --- Comment #35 from Duncan Sands 2010-12-02 02:59:24 CST --- I plan to fix this in dragonegg. Moving it to the dragonegg component. -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Thu Dec 2 03:56:36 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Thu, 2 Dec 2010 03:56:36 -0600 (CST) Subject: [LLVMbugs] [Bug 8724] New: clang: missed optimization: copy global to local, prefetch Message-ID: http://llvm.org/bugs/show_bug.cgi?id=8724 Summary: clang: missed optimization: copy global to local, prefetch Product: new-bugs Version: 2.8 Platform: PC URL: http://gcc.gnu.org/bugzilla/show_bug.cgi?id=46763 OS/Version: Linux Status: NEW Severity: normal Priority: P Component: new bugs AssignedTo: unassignedbugs at nondot.org ReportedBy: edwintorok at gmail.com CC: llvmbugs at cs.uiuc.edu Created an attachment (id=5856) --> (http://llvm.org/bugs/attachment.cgi?id=5856) gy.i.bz2 See http://gcc.gnu.org/bugzilla/show_bug.cgi?id=46763 clang 2.8 -O2 misses that optimization too: $ clang gy.i -O2 -lm $ ./a.out default: 1.405707s ( 100.0%) glob2loc: 1.295956s ( 92.2% +- 5.219%) prefetchnta: 1.226949s ( 87.3% +- 0.9207%) prefetch: 1.292528s ( 91.9% +- 0.05732%) glob2loc+prefetchnta: 1.205552s ( 85.8% +- 0.9699%) It also misses some other optimizations, because with gcc-4.5 I get: default: 1.325195s ( 100.0%) glob2loc: 1.268875s ( 95.8% +- 1.024%) prefetchnta: 1.207342s ( 91.1% +- 0.4986%) prefetch: 1.277638s ( 96.4% +- 0.1179%) glob2loc+prefetchnta: 1.199906s ( 90.5% +- 0.3629%) CPU: AMD Phenom(tm) II X6 1090T Processor clang version 2.8 (tags/RELEASE_28/rc3) Target: x86_64-unknown-linux-gnu Thread model: posix -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Thu Dec 2 04:23:28 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Thu, 2 Dec 2010 04:23:28 -0600 (CST) Subject: [LLVMbugs] [Bug 8725] New: Regression on 2 gcc-torture test Message-ID: http://llvm.org/bugs/show_bug.cgi?id=8725 Summary: Regression on 2 gcc-torture test Product: libraries Version: trunk Platform: PC OS/Version: Linux Status: NEW Severity: normal Priority: P Component: Scalar Optimizations AssignedTo: unassignedbugs at nondot.org ReportedBy: quickslyver at free.fr CC: llvmbugs at cs.uiuc.edu Created an attachment (id=5857) --> (http://llvm.org/bugs/attachment.cgi?id=5857) first falling test the attached tests are failing, I'm sure that they passed 1 month ago, the last time I run these test. I have no time to investigate more :( They pass at -O0 but fail at -Os, so it's surely a llvm bug $ clang -v clang version 2.9 (trunk 120699) Target: i386-pc-linux-gnu Thread model: posix $ clang ~/repos/pitchoun-trunk/sim/regression/invia_regression/cpu/gcc-torture/run/gcc/20071219-1.c &&./a.out $ clang ~/repos/pitchoun-trunk/sim/regression/invia_regression/cpu/gcc-torture/run/gcc/pr35472.c && ./a.out $ clang ~/repos/pitchoun-trunk/sim/regression/invia_regression/cpu/gcc-torture/run/gcc/20071219-1.c -Os $ ./a.out Aborted $ clang ~/repos/pitchoun-trunk/sim/regression/invia_regression/cpu/gcc-torture/run/gcc/pr35472.c -Os $ ./a.out Aborted -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Thu Dec 2 05:43:54 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Thu, 2 Dec 2010 05:43:54 -0600 (CST) Subject: [LLVMbugs] [Bug 8726] New: Assertion `D && "Cannot get layout of forward declarations!"' failed. Message-ID: http://llvm.org/bugs/show_bug.cgi?id=8726 Summary: Assertion `D && "Cannot get layout of forward declarations!"' failed. Product: clang Version: trunk Platform: PC OS/Version: Linux Status: NEW Severity: normal Priority: P Component: C++ AssignedTo: unassignedclangbugs at nondot.org ReportedBy: hans at chromium.org CC: llvmbugs at cs.uiuc.edu, dgregor at apple.com Trying to compile the following with Clang built from trunk (r120702) hits an assert: class C; struct S { const C &c1; int i; const C &c2; }; void f(const C& c) { S s = {c, 42, c}; } clang: RecordLayoutBuilder.cpp:1615: const clang::ASTRecordLayout& clang::ASTContext::getASTRecordLayout(const clang::RecordDecl*): Assertion `D && "Cannot get layout of forward declarations!"' failed. -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Thu Dec 2 07:05:16 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Thu, 2 Dec 2010 07:05:16 -0600 (CST) Subject: [LLVMbugs] [Bug 3559] llvm-gcc emits debug info for gimple temporaries (non-user vars) In-Reply-To: References: Message-ID: <20101202130516.D00F02A6C12C@llvm.org> http://llvm.org/bugs/show_bug.cgi?id=3559 Duncan Sands changed: What |Removed |Added ---------------------------------------------------------------------------- Status|RESOLVED |REOPENED CC| |baldrick at free.fr Component|llvm-gcc |New Bugs Version|2.2 |trunk Resolution|FIXED | AssignedTo|unassignedbugs at nondot.org |baldrick at free.fr Product|tools |dragonegg --- Comment #4 from Duncan Sands 2010-12-02 07:05:16 CST --- Moving to dragonegg since there is no declare for A there either (on the other hand the wrong declare for the gimple temporary isn't present in dragonegg). -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Thu Dec 2 09:10:51 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Thu, 2 Dec 2010 09:10:51 -0600 (CST) Subject: [LLVMbugs] [Bug 8224] [Unit Test][ADTTests.ValueMapTest/{0, 1, 2}.LocksMutex] Failure: Mutex should already be locked. In-Reply-To: References: Message-ID: <20101202151051.7AA0E2A6C12F@llvm.org> http://llvm.org/bugs/show_bug.cgi?id=8224 NAKAMURA Takumi changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |RESOLVED Resolution| |FIXED --- Comment #3 from NAKAMURA Takumi 2010-12-02 09:10:51 CST --- It passes on msvs10. It had been suppressed on msvs<=9 since r119782. -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Thu Dec 2 09:16:38 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Thu, 2 Dec 2010 09:16:38 -0600 (CST) Subject: [LLVMbugs] [Bug 8727] New: no-operator-names doesn't work with C++ Message-ID: http://llvm.org/bugs/show_bug.cgi?id=8727 Summary: no-operator-names doesn't work with C++ Product: clang Version: trunk Platform: PC OS/Version: Linux Status: NEW Severity: normal Priority: P Component: C++ AssignedTo: unassignedclangbugs at nondot.org ReportedBy: smacvicar at gmail.com CC: llvmbugs at cs.uiuc.edu, dgregor at apple.com This is from a header file from c-client, was running it with: clang test.cpp -E -fno-operator-names g++ has no issues with this. #ifdef __cplusplus /* help out people who use C++ compilers */ extern "C" { /* If you use gcc, you may also have to use -fno-operator-names */ #define private cclientPrivate /* private to c-client */ #define and cclientAnd /* C99 doesn't realize that ISO 646 is dead */ #define or cclientOr #define not cclientNot #undef private } #endif -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Thu Dec 2 12:29:13 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Thu, 2 Dec 2010 12:29:13 -0600 (CST) Subject: [LLVMbugs] [Bug 8726] Assertion `D && "Cannot get layout of forward declarations!"' failed. In-Reply-To: References: Message-ID: <20101202182913.2C4142A6C130@llvm.org> http://llvm.org/bugs/show_bug.cgi?id=8726 Chris Lattner changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |RESOLVED Resolution| |FIXED --- Comment #2 from Chris Lattner 2010-12-02 12:29:12 CST --- Fixed in r120722, thanks for the great testcase! -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Thu Dec 2 13:22:07 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Thu, 2 Dec 2010 13:22:07 -0600 (CST) Subject: [LLVMbugs] [Bug 8728] New: likely wrong code bug Message-ID: http://llvm.org/bugs/show_bug.cgi?id=8728 Summary: likely wrong code bug Product: new-bugs Version: unspecified Platform: PC OS/Version: All Status: NEW Severity: normal Priority: P Component: new bugs AssignedTo: unassignedbugs at nondot.org ReportedBy: regehr at cs.utah.edu CC: llvmbugs at cs.uiuc.edu, chenyang at cs.utah.edu [regehr at n1 ~]$ clang -O1 foo.c -o foo [regehr at n1 ~]$ ./foo g_99.f0 = 1 [regehr at n1 ~]$ clang -O2 foo.c -o foo [regehr at n1 ~]$ ./foo g_99.f0 = 0 [regehr at n1 ~]$ cat foo.c struct S2 { unsigned f0 : 27; unsigned f1 : 17; signed f2 : 18; }; struct S2 g_99 = {0,0,0}; struct S2 *g_248 = &g_99; int printf(const char *format, ...); int main(void) { struct S2 l_438 = {1,1,1}; *g_248 = l_438; *g_248 = g_99; printf("g_99.f0 = %d\n", g_99.f0); return 0; } [regehr at n1 ~]$ clang -v clang version 2.9 (trunk 120726) Target: i386-pc-linux-gnu Thread model: posix [regehr at n1 ~]$ -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Thu Dec 2 14:06:53 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Thu, 2 Dec 2010 14:06:53 -0600 (CST) Subject: [LLVMbugs] [Bug 8124] copy is only important at the start of a method name In-Reply-To: References: Message-ID: <20101202200654.231942A6C12D@llvm.org> http://llvm.org/bugs/show_bug.cgi?id=8124 Ted Kremenek changed: What |Removed |Added ---------------------------------------------------------------------------- Status|RESOLVED |REOPENED Resolution|INVALID | --- Comment #9 from Ted Kremenek 2010-12-02 14:06:53 CST --- (In reply to comment #8) > (In reply to comment #7) > > Note, you can use the attribute 'ns_returns_not_retained' to override the > > behavior. > > This attribute is not present in the headers for any of the methods that > contain this convention. There are, in fact, no cases in Apple's Foundation or > AppKit frameworks where a method contains the word copy anywhere other than at > the start, yet returns an owned reference (many return BOOL, or some other > non-object type, some return a non-owned reference). > > If the conventions are meant to be correct, then I suggest that you remind the > Cocoa team that they have been consistently violating them for approximately 20 > years (some of the methods of this form are present in the OpenStep > specification, while some were introduced as recently as OS X 10.6). > > Surely correcting the conventions to reflect how they have been used - and the > expectations of developers, formed over many years - is a better solution than > creating lots of spurious warnings. It's all very well to say that the > conventions have been vetted, but they have clearly not been observed. A > simple grep of the headers will show that this rule is wrong. After more conversation, the Cocoa guys now agree with you. The docs will be updated, and so will the analyzer. The new rule is that any method that starts with 'copy' or 'mutableCopy' follows the copy rule. Anything else that contains 'copy' will need to use the ns_returns_retained attribute. -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Thu Dec 2 18:33:27 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Thu, 2 Dec 2010 18:33:27 -0600 (CST) Subject: [LLVMbugs] [Bug 8729] New: Allow diagnostic to be disabled: warn_indirection_through_null Message-ID: http://llvm.org/bugs/show_bug.cgi?id=8729 Summary: Allow diagnostic to be disabled: warn_indirection_through_null Product: clang Version: trunk Platform: PC OS/Version: Linux Status: NEW Severity: enhancement Priority: P Component: -New Bugs AssignedTo: unassignedclangbugs at nondot.org ReportedBy: pipping at exherbo.org CC: llvmbugs at cs.uiuc.edu Created an attachment (id=5859) --> (http://llvm.org/bugs/attachment.cgi?id=5859) Fix warn_indirection_through_null is enabled by default and can currently only be disabled through -w. 02:19 < pipping> Hi. clang gives me a warning, diag::warn_indirection_through_null in particular. I'd like to disable that, how do I find out what option I need to pass to clang? 02:20 < Bigcheese> It should tell you what flag caused it right next to the warning. 02:20 < Bigcheese> Err, what flag disables it. 02:22 < pipping> it know that clang does that for some warnings, but it didn't for this one: http://dpaste.com/282079/ 02:22 <+nlewycky> then it doesn't have one. please fix! -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Thu Dec 2 18:43:50 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Thu, 2 Dec 2010 18:43:50 -0600 (CST) Subject: [LLVMbugs] [Bug 8730] New: Missed optimization in passing bools Message-ID: http://llvm.org/bugs/show_bug.cgi?id=8730 Summary: Missed optimization in passing bools Product: clang Version: trunk Platform: PC OS/Version: Linux Status: NEW Severity: enhancement Priority: P Component: LLVM Codegen AssignedTo: unassignedclangbugs at nondot.org ReportedBy: jyasskin at google.com CC: nicholas at mxc.ca, llvmbugs at cs.uiuc.edu $ cat missed_opt.cc void called(bool b); void wrapper(bool* b) { called(*b); } $ g++ --version g++ (Ubuntu 4.4.3-4ubuntu5) 4.4.3 $ g++ -S -O2 missed_opt.cc -o - ... _Z7wrapperPb: .LFB0: .cfi_startproc .cfi_personality 0x3,__gxx_personality_v0 movzbl (%rdi), %edi jmp _Z6calledb ... $ clang++ -S -O2 missed_opt.cc -o - ... _Z7wrapperPb: # @_Z7wrapperPb .Leh_func_begin0: # BB#0: # %entry pushq %rbp .Ltmp0: movq %rsp, %rbp .Ltmp1: movb (%rdi), %al andb $1, %al movzbl %al, %edi popq %rbp jmp _Z6calledb # TAILCALL $ clang++ -S -emit-llvm -O2 missed_opt.cc -o - target datalayout = "e-p:64:64:64-i1:8:8-i8:8:8-i16:16:16-i32:32:32-i64:64:64-f32:32:32-f64:64:64-v64:64:64-v128:128:128-a0:0:64-s0:64:64-f80:128:128-n8:16:32:64" target triple = "x86_64-unknown-linux-gnu" define void @_Z7wrapperPb(i8* nocapture %b) { entry: %tmp1 = load i8* %b, align 1, !tbaa !0 %tmp = and i8 %tmp1, 1 %tobool = icmp ne i8 %tmp, 0 tail call void @_Z6calledb(i1 zeroext %tobool) ret void } declare void @_Z6calledb(i1 zeroext) $ The extra 'and' instruction seems to show up any time a bool is loaded from memory. Nick suggests fixing it by bitcasting to i1* before doing the load. -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Thu Dec 2 19:10:40 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Thu, 2 Dec 2010 19:10:40 -0600 (CST) Subject: [LLVMbugs] [Bug 8729] Allow diagnostic to be disabled: warn_indirection_through_null In-Reply-To: References: Message-ID: <20101203011040.3082B2A6C12F@llvm.org> http://llvm.org/bugs/show_bug.cgi?id=8729 Nick Lewycky changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |RESOLVED CC| |nicholas at mxc.ca Resolution| |FIXED --- Comment #2 from Nick Lewycky 2010-12-02 19:10:39 CST --- Fixed in r120771: http://lists.cs.uiuc.edu/pipermail/cfe-commits/Week-of-Mon-20101129/036973.html Thanks for the patch! -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Fri Dec 3 00:54:23 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Fri, 3 Dec 2010 00:54:23 -0600 (CST) Subject: [LLVMbugs] [Bug 8638] crash during static analysis of objective.c file In-Reply-To: References: Message-ID: <20101203065423.CBFD72A6C12D@llvm.org> http://llvm.org/bugs/show_bug.cgi?id=8638 Ted Kremenek changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |RESOLVED Resolution| |FIXED --- Comment #3 from Ted Kremenek 2010-12-03 00:54:23 CST --- Fixed: http://llvm.org/viewvc/llvm-project?view=rev&revision=120795 http://llvm.org/viewvc/llvm-project?view=rev&revision=120796 -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Fri Dec 3 01:46:09 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Fri, 3 Dec 2010 01:46:09 -0600 (CST) Subject: [LLVMbugs] [Bug 8723] llvm 2.8: doesn't build on Mac OS X 10.4.11, undefined symbol ___crashreporter_info__ In-Reply-To: References: Message-ID: <20101203074609.DC8322A6C12F@llvm.org> http://llvm.org/bugs/show_bug.cgi?id=8723 Eric Christopher changed: What |Removed |Added ---------------------------------------------------------------------------- Status|ASSIGNED |RESOLVED Resolution| |FIXED --- Comment #4 from Eric Christopher 2010-12-03 01:46:09 CST --- I think this is fixed here: [issola:~/workspaces/PR8723/llvm] echristo% svn ci Sending autoconf/configure.ac Sending configure Sending include/llvm/Config/config.h.in Sending lib/Support/PrettyStackTrace.cpp Transmitting file data .... Committed revision 120801. but I don't actually have a Tiger machine sitting around anywhere :) -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Fri Dec 3 03:30:14 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Fri, 3 Dec 2010 03:30:14 -0600 (CST) Subject: [LLVMbugs] [Bug 8731] New: Clang generates load of struct value. Message-ID: http://llvm.org/bugs/show_bug.cgi?id=8731 Summary: Clang generates load of struct value. Product: new-bugs Version: trunk Platform: PC OS/Version: Linux Status: NEW Severity: normal Priority: P Component: new bugs AssignedTo: unassignedbugs at nondot.org ReportedBy: fvbommel at gmail.com CC: llvmbugs at cs.uiuc.edu Created an attachment (id=5863) --> (http://llvm.org/bugs/attachment.cgi?id=5863) The full .ll of the C++ code. Compiling this C++ code with clang: ===== #include std::pair foo(void* a, long long b) { return std::make_pair(a, b); } ===== generates the attached .ll (on x86-64 linux). The relevant snippet: ===== %"struct.std::pair" = type { i8*, i64 } define linkonce_odr %"struct.std::pair" @_ZSt9make_pairIPvxESt4pairIT_T0_ES2_S3_(i8* %__x, i64 %__y) inlinehint { entry: %retval = alloca %"struct.std::pair", align 8 %__x.addr = alloca i8*, align 8 %__y.addr = alloca i64, align 8 store i8* %__x, i8** %__x.addr, align 8 store i64 %__y, i64* %__y.addr, align 8 call void @_ZNSt4pairIPvxEC1ERKS0_RKx(%"struct.std::pair"* %retval, i8** %__x.addr, i64* %__y.addr) %0 = load %"struct.std::pair"* %retval ret %"struct.std::pair" %0 } ===== Notice that it generates a load of a struct value. At -O3, after inlining, a later transformation (-scalarrepl?) can insert an extractvalue of such loads. (this happens in LLVM's lib/CodeGen/PreAllocSplitting.cpp). According to Dan Gohman (reference: http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20101129/112708.html) this merits a bug report. (The problem apparently being that loads of aggregates are compiled less efficiently than loads of their elements followed by 'insertvalue' instructions.) -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Fri Dec 3 03:39:01 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Fri, 3 Dec 2010 03:39:01 -0600 (CST) Subject: [LLVMbugs] [Bug 8732] New: Report function/variable/types usage Message-ID: http://llvm.org/bugs/show_bug.cgi?id=8732 Summary: Report function/variable/types usage Product: clang Version: unspecified Platform: All OS/Version: All Status: NEW Severity: enhancement Priority: P Component: Static Analyzer AssignedTo: kremenek at apple.com ReportedBy: ranma42 at gmail.com CC: llvmbugs at cs.uiuc.edu Sometimes (expecially when refactoring code) it is useful to know callers and callees of a function, functions accessing global variables and/or where a type is used. I believe that Clang Static Analyzer already has most, if not all, the infrastructure needed to compute this (obviously ignoring cases like function pointers). Would it be possible/hard/a good idea to have this kind of report in addition to the "bugreport" for the code? -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Fri Dec 3 09:25:23 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Fri, 3 Dec 2010 09:25:23 -0600 (CST) Subject: [LLVMbugs] [Bug 8733] New: AV when reporting diagnostics in some code-completion results Message-ID: http://llvm.org/bugs/show_bug.cgi?id=8733 Summary: AV when reporting diagnostics in some code-completion results Product: clang Version: trunk Platform: All OS/Version: Windows NT Status: NEW Severity: normal Priority: P Component: Frontend AssignedTo: unassignedclangbugs at nondot.org ReportedBy: plc at rowley.co.uk CC: llvmbugs at cs.uiuc.edu Created an attachment (id=5865) --> (http://llvm.org/bugs/attachment.cgi?id=5865) Contains repro case Code completion generates an exception when reporting errors in #included macro expansions with precompiled headers enabled. > more pl230.h #define __C(X) (0xFFFFFFFFU >> (31-(X))) enum { V = __C(31), }; > more lm3s_mac.c #include "pl230.h" static void foo(void) { / } >set CINDEXTEST_EDITING=1 >c-index-test -code-completion-at=lm3s_mac.c:6:2 lm3s_mac.c -nostdinc -I. [snip] 599B66BE (0x00000004 0x00EEDAC8 0x00BDF760 0x599B6796), clang::SourceManager::ge tSLocEntry()+0142 bytes(s), e:\llvm\llvm\tools\clang\include\clang\basic\sourcem anager.h, line 818+0022 byte(s) 599B660A (0x00000004 0x00BDFC20 0x00000004 0x00E6B620), clang::SourceManager::ge tSLocEntry()+0026 bytes(s), e:\llvm\llvm\tools\clang\include\clang\basic\sourcem anager.h, line 824 ...and so on. Actually, this bug is triggered *ONLY* by CXTranslationUnit_PrecompiledPreamble as far as I can tell. Turning precompiled preamble off and we're good. You change getDefaultParsingOptions() in c-index-test, you provoke a crash or not, which is what I did. static unsigned getDefaultParsingOptions() { unsigned options = 0; // Nuke this one.. // CXTranslationUnit_DetailedPreprocessingRecord; if (getenv("CINDEXTEST_EDITING")) { // Figure out which option triggers fault: break them out... options |= CXTranslationUnit_PrecompiledPreamble; // options |= CXTranslationUnit_CacheCompletionResults; // options |= CXTranslationUnit_CXXPrecompiledPreamble; // was: options |= clang_defaultEditingTranslationUnitOptions(); } if (getenv("CINDEXTEST_COMPLETION_CACHING")) options |= CXTranslationUnit_CacheCompletionResults; return options; } This may well affect released versions of clang, I have no idea. I'm just trunking at present. -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Fri Dec 3 16:59:23 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Fri, 3 Dec 2010 16:59:23 -0600 (CST) Subject: [LLVMbugs] [Bug 8734] New: [MC assembler] .float not supported Message-ID: http://llvm.org/bugs/show_bug.cgi?id=8734 Summary: [MC assembler] .float not supported Product: new-bugs Version: trunk Platform: PC OS/Version: FreeBSD Status: NEW Severity: normal Priority: P Component: new bugs AssignedTo: unassignedbugs at nondot.org ReportedBy: dimitry at andric.com CC: llvmbugs at cs.uiuc.edu It looks like integrated-as supports the .single directive, but not the synonym .float. This should be easy to implement. :) http://sourceware.org/binutils/docs/as/Float.html says: 7.53 .float flonums This directive assembles zero or more flonums, separated by commas. It has the same effect as .single. The exact kind of floating point numbers emitted depends on how as is configured. See [Machine Dependencies]. -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Fri Dec 3 17:56:35 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Fri, 3 Dec 2010 17:56:35 -0600 (CST) Subject: [LLVMbugs] [Bug 8735] New: c++ miscompile at -O2 Message-ID: http://llvm.org/bugs/show_bug.cgi?id=8735 Summary: c++ miscompile at -O2 Product: new-bugs Version: trunk Platform: PC OS/Version: Linux Status: NEW Severity: normal Priority: P Component: new bugs AssignedTo: unassignedbugs at nondot.org ReportedBy: nlewycky at google.com CC: chandlerc at gmail.com, llvmbugs at cs.uiuc.edu I have a testcase which fails when used with __versa_string. For those not familiar, versa_string is an alternate string type in libstdc++ (#include ) that implements the small string optimization. This header (and its dependencies) are missing lots of "this->" so I'm posting only the preprocessed source. You can extract the original testcase from the bottom if you're willing to change your system headers. Steps to repro: nlewycky at ducttape:~$ llvm/Debug+Asserts/bin/clang++ -O2 -fno-exceptions rr.ii -o rr nlewycky at ducttape:~$ ./rr r1 removeremove r2 removeremove i insertBar The two 'removeremove' strings should be 'removeBar'. The problem is that string::operator+ is getting the wrong 'this' pointer (but a valid one! it's just one for a different object) after opt -O2. -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Sat Dec 4 13:44:59 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Sat, 4 Dec 2010 13:44:59 -0600 (CST) Subject: [LLVMbugs] [Bug 8736] New: LLVM-GCC-4.2 recursive global blocks are nil at runtime Message-ID: http://llvm.org/bugs/show_bug.cgi?id=8736 Summary: LLVM-GCC-4.2 recursive global blocks are nil at runtime Product: new-bugs Version: unspecified Platform: Macintosh OS/Version: MacOS X Status: NEW Severity: normal Priority: P Component: new bugs AssignedTo: unassignedbugs at nondot.org ReportedBy: keith at 33software.com CC: llvmbugs at cs.uiuc.edu Created an attachment (id=5869) --> (http://llvm.org/bugs/attachment.cgi?id=5869) Demonstrates issue. Summary: When compiling a global block variable that calls itself, using LLVM GCC 4.2, the block isn't created - but compilation succeeds and the symbol is still exported. This is also filed as rdar://problem/8639147 Steps to Reproduce: Compile and run the attached test case using: @@@ llvm-gcc-4.2 RecursiveGlobalBlocks.m -framework Foundation ./a.out @@@ This will exit with EXIT_FAILURE. The output of `llvm-gcc-4.2 -v` is: @@@ Using built-in specs. Target: i686-apple-darwin10 Configured with: /var/tmp/llvmgcc42/llvmgcc42-2333.4~7/src/configure --disable-checking --enable-werror --prefix=/Developer/usr/llvm-gcc-4.2 --mandir=/share/man --enable-languages=c,objc,c++,obj-c++ --program-prefix=llvm- --program-transform-name=/^[cg][^.-]*$/s/$/-4.2/ --with-slibdir=/usr/lib --build=i686-apple-darwin10 --enable-llvm=/var/tmp/llvmgcc42/llvmgcc42-2333.4~7/dst-llvmCore/Developer/usr/local --program-prefix=i686-apple-darwin10- --host=x86_64-apple-darwin10 --target=i686-apple-darwin10 --with-gxx-include-dir=/usr/include/c++/4.2.1 Thread model: posix gcc version 4.2.1 (Based on Apple Inc. build 5658) (LLVM build 2333.4) @@@ Recompile and run the test case using: @@@ clang RecursiveGlobalBlocks.m -framework Foundation ./a.out @@@ Compiled with clang the program will output 120 (5!) and exit with EXIT_SUCCESS The output of `clang -v` is: @@@ Apple clang version 1.6 (tags/Apple/clang-70) Target: x86_64-apple-darwin10 Thread model: posix @@@ Expected Results: Fail to compile, or ensure global blocks can call themselves. Actual Results: Global block variables which call themselves, compiled with llvm-gcc-4.2, are nil at runtime. Invoking them results in a crash when de-referencing to retrieve the function pointer. -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Sat Dec 4 16:51:23 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Sat, 4 Dec 2010 16:51:23 -0600 (CST) Subject: [LLVMbugs] [Bug 8737] New: [mc assembler] .extern directive ignored Message-ID: http://llvm.org/bugs/show_bug.cgi?id=8737 Summary: [mc assembler] .extern directive ignored Product: new-bugs Version: trunk Platform: PC OS/Version: FreeBSD Status: NEW Severity: normal Priority: P Component: new bugs AssignedTo: unassignedbugs at nondot.org ReportedBy: kwm at FreeBSD.org CC: llvmbugs at cs.uiuc.edu While building libGL with clang ToT. The following warning happens. > clang -v clang version 2.9 (trunk 120744) Target: x86_64-unknown-freebsd8.1 Thread model: posix clang -c -I. -I../../../include -I../../../include/GL/internal -I../../../src/mesa -I../../../src/mesa/glapi -I/usr/local/include -I/usr/local/include/drm -I/usr/local/include -D_THREAD_SAFE -I/usr/local/include -I/usr/local/include -O2 -pipe -fno-strict-aliasing -Wall -Wmissing-prototypes -std=c99 -ffast-math -fno-strict-aliasing -fPIC -DUSE_X86_64_ASM -DHAVE_POSIX_MEMALIGN -DUSE_XCB -DPTHREADS -DUSE_EXTERNAL_DXTN_LIB=1 -DIN_DRI_DRIVER -DHAVE_ALIAS -DGLX_INDIRECT_RENDERING -DGLX_DIRECT_RENDERING -DXF86VIDMODE -D_REENTRANT -UIN_DRI_DRIVER -DDEFAULT_DRIVER_DIR=\"/usr/local/lib/dri\" ../../../src/mesa/x86-64/glapi_x86-64.S -o ../../../src/mesa/x86-64/glapi_x86-64.o /tmp/cc-uLf4Jd.s:14:2: warning: ignoring directive for now .extern _glapi_Dispatch ^ /tmp/cc-uLf4Jd.s:15:2: warning: ignoring directive for now .extern _gl_DispatchTSD ^ /tmp/cc-uLf4Jd.s:16:2: warning: ignoring directive for now .extern pthread_getspecific ^ -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Sat Dec 4 17:33:07 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Sat, 4 Dec 2010 17:33:07 -0600 (CST) Subject: [LLVMbugs] [Bug 8736] LLVM-GCC-4.2 recursive global blocks are nil at runtime In-Reply-To: References: Message-ID: <20101204233307.4A8352A6C124@llvm.org> http://llvm.org/bugs/show_bug.cgi?id=8736 Chris Lattner changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |RESOLVED Resolution| |WONTFIX --- Comment #1 from Chris Lattner 2010-12-04 17:33:06 CST --- Sorry, but we're not fixing frontend bugs in llvm-gcc anymore, please try clang. -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Sat Dec 4 17:33:58 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Sat, 4 Dec 2010 17:33:58 -0600 (CST) Subject: [LLVMbugs] [Bug 8732] Report function/variable/types usage In-Reply-To: References: Message-ID: <20101204233358.4C3282A6C12E@llvm.org> http://llvm.org/bugs/show_bug.cgi?id=8732 Chris Lattner changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |RESOLVED Resolution| |FIXED --- Comment #1 from Chris Lattner 2010-12-04 17:33:57 CST --- libclang provides a programatic interface that provides this information. Please see Doug's talk on the devmtg 2010 page for some more info on it. -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Sat Dec 4 17:35:32 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Sat, 4 Dec 2010 17:35:32 -0600 (CST) Subject: [LLVMbugs] [Bug 8731] Clang generates load of struct value. In-Reply-To: References: Message-ID: <20101204233532.9006D2A6C12C@llvm.org> http://llvm.org/bugs/show_bug.cgi?id=8731 Chris Lattner changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |RESOLVED Resolution| |INVALID --- Comment #1 from Chris Lattner 2010-12-04 17:35:32 CST --- Dan's point is that first class values in general disable fastisel. However, there is nothing that we can do about it in this case, as the ABI requires us to return them. Using insertvalue instead of load won't help us, it will cause the same fastisel bail out. When optimizations are enabled, fast isel isn't used, so there is no problem at -O. -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Sun Dec 5 01:03:46 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Sun, 5 Dec 2010 01:03:46 -0600 (CST) Subject: [LLVMbugs] [Bug 8738] New: Failure to type infer literal integer Message-ID: http://llvm.org/bugs/show_bug.cgi?id=8738 Summary: Failure to type infer literal integer Product: clang Version: trunk Platform: PC OS/Version: Linux Status: NEW Severity: normal Priority: P Component: C++0x AssignedTo: unassignedclangbugs at nondot.org ReportedBy: sashan at zenskg.net CC: llvmbugs at cs.uiuc.edu, dgregor at apple.com Hi Bumped into this clang++ failure: sashan at cyclops build $ clang++ -c -std=c++0x ~/code/cpp/auto.cpp /home/sashan/code/cpp/auto.cpp:1:6: error: cannot initialize a variable of type 'auto' with an rvalue of type 'int' auto x = 1; ^ ~ 1 error generated. sashan at cyclops build 1 $ Where auto.cpp is: auto x = 1; This compiles with g++ -c -std=c++0x. -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Sun Dec 5 12:20:43 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Sun, 5 Dec 2010 12:20:43 -0600 (CST) Subject: [LLVMbugs] [Bug 8739] New: Assertion `DerivationOkay && "Can only be used with a derived-to-base conversion"' Message-ID: http://llvm.org/bugs/show_bug.cgi?id=8739 Summary: Assertion `DerivationOkay && "Can only be used with a derived-to-base conversion"' Product: clang Version: trunk Platform: PC OS/Version: Linux Status: NEW Severity: normal Priority: P Component: C++ AssignedTo: unassignedclangbugs at nondot.org ReportedBy: grzegorz.dabrowski at gmail.com CC: llvmbugs at cs.uiuc.edu, dgregor at apple.com Assertion fails in trunk, but works in 2.8: class BSplitView {}; class BView {}; template class Base; template class Split : public Base { public: typedef Split ThisBuilder; ThisBuilder& GetView(BView** _view); private: BSplitView* fView; }; template typename Split::ThisBuilder& Split::GetView(BView** _view) { *_view = fView; return *this; } -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Sun Dec 5 14:53:00 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Sun, 5 Dec 2010 14:53:00 -0600 (CST) Subject: [LLVMbugs] [Bug 4170] Verifier should catch illegal extractvalue instruction In-Reply-To: References: Message-ID: <20101205205300.4847A2A6C12D@llvm.org> http://llvm.org/bugs/show_bug.cgi?id=4170 Frits van Bommel changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |RESOLVED CC| |fvbommel at gmail.com Resolution| |FIXED --- Comment #6 from Frits van Bommel 2010-12-05 14:52:59 CST --- Fixed by r120956. http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20101129/113151.html -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Sun Dec 5 18:37:17 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Sun, 5 Dec 2010 18:37:17 -0600 (CST) Subject: [LLVMbugs] [Bug 8725] Regression on 2 gcc-torture test In-Reply-To: References: Message-ID: <20101206003717.0DF602A6C124@llvm.org> http://llvm.org/bugs/show_bug.cgi?id=8725 Chris Lattner changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |RESOLVED Resolution| |DUPLICATE --- Comment #4 from Chris Lattner 2010-12-05 18:37:16 CST --- Ok, I'm happy to dupe this, will take a look at your patch next. Thanks Jakub! *** This bug has been marked as a duplicate of bug 8728 *** -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Sun Dec 5 18:53:55 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Sun, 5 Dec 2010 18:53:55 -0600 (CST) Subject: [LLVMbugs] [Bug 8681] clang ToT can't parse boost::addressof any more In-Reply-To: References: Message-ID: <20101206005355.4091D2A6C12E@llvm.org> http://llvm.org/bugs/show_bug.cgi?id=8681 Richard Smith changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |RESOLVED Resolution| |FIXED --- Comment #11 from Richard Smith 2010-12-05 18:53:54 CST --- The code in comment#0 and comment#5 both work with ToT. Thanks! -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Sun Dec 5 19:48:54 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Sun, 5 Dec 2010 19:48:54 -0600 (CST) Subject: [LLVMbugs] [Bug 8728] likely wrong code bug In-Reply-To: References: Message-ID: <20101206014854.DAA172A6C12E@llvm.org> http://llvm.org/bugs/show_bug.cgi?id=8728 Chris Lattner changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |RESOLVED Resolution| |FIXED --- Comment #6 from Chris Lattner 2010-12-05 19:48:54 CST --- Fixed in r120974. Also, it turns out that eliminating the memset in: tail call void @llvm.memset.i64(i8* %P, i8 42, i64 8, i32 1) tail call void @llvm.memcpy.i64(i8* %P, i8* %Q, i64 12, i32 1) isn't safe if P/Q might alias. -Chris -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Sun Dec 5 22:36:05 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Sun, 5 Dec 2010 22:36:05 -0600 (CST) Subject: [LLVMbugs] [Bug 8740] New: dse can be too eager Message-ID: http://llvm.org/bugs/show_bug.cgi?id=8740 Summary: dse can be too eager Product: libraries Version: trunk Platform: PC OS/Version: Linux Status: NEW Severity: normal Priority: P Component: Scalar Optimizations AssignedTo: unassignedbugs at nondot.org ReportedBy: tom.prince at ualberta.net CC: llvmbugs at cs.uiuc.edu Created an attachment (id=5873) --> (http://llvm.org/bugs/attachment.cgi?id=5873) llvm IR that breaks dse. Running the attached llvm ir with lli gives -10 -10 -30 with opt -basicaa -dse | lli gives -10 -10 -10 -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Sun Dec 5 23:27:12 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Sun, 5 Dec 2010 23:27:12 -0600 (CST) Subject: [LLVMbugs] [Bug 8739] Assertion `DerivationOkay && "Can only be used with a derived-to-base conversion"' In-Reply-To: References: Message-ID: <20101206052712.1223A2A6C12E@llvm.org> http://llvm.org/bugs/show_bug.cgi?id=8739 John McCall changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |RESOLVED Resolution| |FIXED --- Comment #1 from John McCall 2010-12-05 23:27:11 CST --- Fixed in r120990. -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Mon Dec 6 00:12:04 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Mon, 6 Dec 2010 00:12:04 -0600 (CST) Subject: [LLVMbugs] [Bug 8741] New: [MC-COFF] Cannot handle globals @Lxxx (begins with 'L') on Win64. Message-ID: http://llvm.org/bugs/show_bug.cgi?id=8741 Summary: [MC-COFF] Cannot handle globals @Lxxx (begins with 'L') on Win64. Product: new-bugs Version: unspecified Platform: PC OS/Version: Windows XP Status: NEW Severity: normal Priority: P Component: new bugs AssignedTo: unassignedbugs at nondot.org ReportedBy: geek4civic at gmail.com CC: llvmbugs at cs.uiuc.edu, bigcheesegs at gmail.com ; llc -mtriple=x86_64-win32 -filetype=obj declare void @L() define void @foo() { entry: call void @L() ret void } It barfs. It seems globals @Lxxx is marked as 'IsTemporary' on MCSymbol. -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Mon Dec 6 01:52:54 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Mon, 6 Dec 2010 01:52:54 -0600 (CST) Subject: [LLVMbugs] [Bug 8735] c++ miscompile at -O2 In-Reply-To: References: Message-ID: <20101206075254.CBC752A6C12F@llvm.org> http://llvm.org/bugs/show_bug.cgi?id=8735 Chris Lattner changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |RESOLVED Resolution| |FIXED --- Comment #13 from Chris Lattner 2010-12-06 01:52:54 CST --- Fixed in r120995 -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Mon Dec 6 03:58:48 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Mon, 6 Dec 2010 03:58:48 -0600 (CST) Subject: [LLVMbugs] [Bug 6290] "Subtype of ScopedDecl not handled" in C++ mode In-Reply-To: References: Message-ID: <20101206095848.087CF2A6C12D@llvm.org> http://llvm.org/bugs/show_bug.cgi?id=6290 Dmitry G. Dyachenko changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |RESOLVED Resolution| |INVALID --- Comment #1 from Dmitry G. Dyachenko 2010-12-06 03:58:46 CST --- current C++ static checker has no relation to old bug -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Mon Dec 6 04:28:05 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Mon, 6 Dec 2010 04:28:05 -0600 (CST) Subject: [LLVMbugs] [Bug 8742] New: assert in clang after commit r120890 when compiling some valid obj-c code. Message-ID: http://llvm.org/bugs/show_bug.cgi?id=8742 Summary: assert in clang after commit r120890 when compiling some valid obj-c code. Product: clang Version: trunk Platform: Macintosh OS/Version: MacOS X Status: NEW Severity: release blocker Priority: P Component: -New Bugs AssignedTo: unassignedclangbugs at nondot.org ReportedBy: devlists at shadowlab.org CC: llvmbugs at cs.uiuc.edu Created an attachment (id=5874) --> (http://llvm.org/bugs/attachment.cgi?id=5874) File that clang cannot compile r120890 by rjmccall, 04:47 ?Although we currently have explicit lvalue-to-rvalue conversions, they're not actually frequently used, because ImpCastExprToType only creates a node if the types differ?? After this commit, clang asserts while compiling the following code. ----------------------- extern int printf(const char *, ...); @interface WBGradientStep {} @property float start; @end void wb_checkLocation() { WBGradientStep *step = 0; printf("%.2f", step.start); } ---------------------- > clang --version clang version 2.9 (trunk 120890) Target: x86_64-apple-darwin10 Thread model: posix > clang -c test.m Assertion failed: (E->getObjectKind() == OK_Ordinary && "reached property reference without lvalue-to-rvalue"), function VisitObjCPropertyRefExpr, file /Volumes/MacPro/Projects/OpenSource/llvm/tools/clang/lib/CodeGen/CGExprScalar.cpp, line 241. 0 clang 0x0000000100e68a62 PrintStackTrace(void*) + 34 1 clang 0x0000000100e69009 SignalHandler(int) + 857 2 libSystem.B.dylib 0x00007fff8326667a _sigtramp + 26 3 libSystem.B.dylib 000000000000000000 _sigtramp + 2094635424 4 clang 0x0000000100011a76 abort + 22 5 clang 0x0000000100011a38 __assert_rtn + 56 6 clang 0x000000010019d9e9 clang::StmtVisitor<(anonymous namespace)::ScalarExprEmitter, llvm::Value*>::Visit(clang::Stmt*) + 41449 7 clang 0x000000010019f7e6 (anonymous namespace)::ScalarExprEmitter::EmitCastExpr(clang::CastExpr*) + 4550 ? -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Mon Dec 6 04:59:04 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Mon, 6 Dec 2010 04:59:04 -0600 (CST) Subject: [LLVMbugs] [Bug 8743] New: Register allocation is invalid with Win64 tailcall jump Message-ID: http://llvm.org/bugs/show_bug.cgi?id=8743 Summary: Register allocation is invalid with Win64 tailcall jump Product: libraries Version: trunk Platform: PC OS/Version: Windows XP Status: NEW Severity: normal Priority: P Component: Backend: X86 AssignedTo: unassignedbugs at nondot.org ReportedBy: geek4civic at gmail.com CC: llvmbugs at cs.uiuc.edu ; llc -O1 -mtriple=x86_64-mingw64 %class = type { [8 x i8] } %vt = type { i32 (...)** } define %vt* @_ZN4llvm9UnsetInit20convertInitializerToEPNS_5RecTyE(%class* %this, %vt* %Ty) align 2 { entry: %0 = bitcast %vt* %Ty to %vt* (%vt*, %class*)*** %vtable = load %vt* (%vt*, %class*)*** %0, align 8 %vfn = getelementptr inbounds %vt* (%vt*, %class*)** %vtable, i64 4 %1 = load %vt* (%vt*, %class*)** %vfn, align 8 %call = tail call %vt* %1(%vt* %Ty, %class* %this) ret %vt* %call } ---- _ZN4llvm9UnsetInit20convertInitializerToEPNS_5RecTyE: # BB#0: # %entry subq $216, %rsp Ltmp0: movaps %xmm15, 32(%rsp) # 16-byte Spill movaps %xmm14, 48(%rsp) # 16-byte Spill movaps %xmm13, 64(%rsp) # 16-byte Spill movaps %xmm12, 80(%rsp) # 16-byte Spill movaps %xmm11, 96(%rsp) # 16-byte Spill movaps %xmm10, 112(%rsp) # 16-byte Spill movaps %xmm9, 128(%rsp) # 16-byte Spill movaps %xmm8, 144(%rsp) # 16-byte Spill movaps %xmm7, 160(%rsp) # 16-byte Spill movaps %xmm6, 176(%rsp) # 16-byte Spill movq %rsi, 200(%rsp) # 8-byte Spill movq %rdi, 208(%rsp) # 8-byte Spill movq %rcx, %rax movq (%rdx), %rcx movq 32(%rcx), %rsi movq %rdx, %rcx movq %rax, %rdx movq 208(%rsp), %rdi # 8-byte Reload movq 200(%rsp), %rsi # 8-byte Reload movaps 176(%rsp), %xmm6 # 16-byte Reload movaps 160(%rsp), %xmm7 # 16-byte Reload movaps 144(%rsp), %xmm8 # 16-byte Reload movaps 128(%rsp), %xmm9 # 16-byte Reload movaps 112(%rsp), %xmm10 # 16-byte Reload movaps 96(%rsp), %xmm11 # 16-byte Reload movaps 80(%rsp), %xmm12 # 16-byte Reload movaps 64(%rsp), %xmm13 # 16-byte Reload movaps 48(%rsp), %xmm14 # 16-byte Reload movaps 32(%rsp), %xmm15 # 16-byte Reload addq $216, %rsp jmpq *%rsi # TAILCALL ---- %RSI should not be used for tail jump. FYI, I don't meet this issue with -regalloc=pbqp. -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Mon Dec 6 05:01:40 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Mon, 6 Dec 2010 05:01:40 -0600 (CST) Subject: [LLVMbugs] [Bug 8744] New: Assert when code completing a #include Message-ID: http://llvm.org/bugs/show_bug.cgi?id=8744 Summary: Assert when code completing a #include Product: clang Version: unspecified Platform: All OS/Version: Windows NT Status: NEW Severity: normal Priority: P Component: Frontend AssignedTo: unassignedclangbugs at nondot.org ReportedBy: plc at rowley.co.uk CC: llvmbugs at cs.uiuc.edu clang will AV when trying to code complete at the opening < of an #include directive that has no closing >. (Not sure about whether additional tokens will do this either...) Repro: > c-index-test -code-completion-at=foo1.c:1:11 foo1.c foo1.c: #include < void foo(void) { } Also attached foo1.c. It seems as if clang is not expecting the code completion token in this case. Can't say I blame it, but the code completion token can probably pop up anywhere. There is no documented API restriction on *where* the token can appear, but just some words on where you'd expect to call code completion. I can guess at a fix; however, my clang skills are still being sharpened, so time to let the pros have a bash. -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Mon Dec 6 11:11:44 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Mon, 6 Dec 2010 11:11:44 -0600 (CST) Subject: [LLVMbugs] [Bug 8731] Clang generates load of struct value. In-Reply-To: References: Message-ID: <20101206171145.2690B2A6C12C@llvm.org> http://llvm.org/bugs/show_bug.cgi?id=8731 Dan Gohman changed: What |Removed |Added ---------------------------------------------------------------------------- Status|RESOLVED |REOPENED Resolution|INVALID | --- Comment #2 from Dan Gohman 2010-12-06 11:11:44 CST --- (In reply to comment #1) > Dan's point is that first class values in general disable fastisel. No it wasn't. The concern I expressed was about optimization. Aggregate loads are never better than individual scalar loads, and can potentially be a lot worse. Fast-isel's present limitations are another reason for preferring scalar loads though. > However, > there is nothing that we can do about it in this case, as the ABI requires us > to return them. To return them yes, but there is an alternative to loading them: call void @_ZNSt4pairIPvxEC1ERKS0_RKx(%"struct.std::pair"* %retval, i8** %__x.addr, i64* %__y.addr) %0 = getelementptr %"struct.std::pair"* %retval, 0, 0 %1 = load i8** %0 %2 = getelementptr %"struct.std::pair"* %retval, 0, 1 %3 = load i64* %2 %4 = insertvalue %"struct.std::pair" undef, %1, 0 %5 = insertvalue %"struct.std::pair" %4, %3, 1 ret %"struct.std::pair" %5 I'm contemplating whether to consider it a bug in clang for not doing this, or to say that instcombine should expand aggregate loads into this form. -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Mon Dec 6 12:37:10 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Mon, 6 Dec 2010 12:37:10 -0600 (CST) Subject: [LLVMbugs] [Bug 8742] assert in clang after commit r120890 when compiling some valid obj-c code. In-Reply-To: References: Message-ID: <20101206183710.410312A6C12E@llvm.org> http://llvm.org/bugs/show_bug.cgi?id=8742 John McCall changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |RESOLVED Resolution| |FIXED --- Comment #2 from John McCall 2010-12-06 12:37:09 CST --- Ha, yes, this is an amusingly well-targeted bug, thank you. Fixed in r121022. -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Mon Dec 6 13:17:12 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Mon, 6 Dec 2010 13:17:12 -0600 (CST) Subject: [LLVMbugs] [Bug 8745] New: [mc] Try and reenable the inst to data lowering optimization Message-ID: http://llvm.org/bugs/show_bug.cgi?id=8745 Summary: [mc] Try and reenable the inst to data lowering optimization Product: libraries Version: trunk Platform: PC OS/Version: All Status: NEW Severity: normal Priority: P Component: Common Code Generator Code AssignedTo: unassignedbugs at nondot.org ReportedBy: rafael.espindola at gmail.com CC: llvmbugs at cs.uiuc.edu The Inst to Data lowering was causing reads after free since we deleted the old inst fragment, but symbols would still point to it. It has been disable for now, but it could be reenabled if each fragment knew its symbols. That way we would be able to move the symbols to the new fragment. We would also be able to do this during layout instead of after it. Yet another way to do it is make sure that symbols never point to instruction fragments by creating dummy data fragments when necessary. -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Mon Dec 6 15:11:13 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Mon, 6 Dec 2010 15:11:13 -0600 (CST) Subject: [LLVMbugs] [Bug 8740] dse can be too eager In-Reply-To: References: Message-ID: <20101206211113.6C56A2A6C12D@llvm.org> http://llvm.org/bugs/show_bug.cgi?id=8740 Chris Lattner changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |RESOLVED Resolution| |INVALID --- Comment #1 from Chris Lattner 2010-12-06 15:11:13 CST --- DSE is doing the correct thing here, your IR is invalid: %11 = tail call fastcc {} %10(i8* %9) nounwind This call is loading from the %0 alloca, which is illegal for calls marked tail. See LangRef.html for more information. If you remove "tail" from the call, DSE doesn't change the behavior of the program. -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Mon Dec 6 15:17:27 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Mon, 6 Dec 2010 15:17:27 -0600 (CST) Subject: [LLVMbugs] [Bug 8746] New: Scalar evolution loops forever (after 2h I got impatient :) on input file Message-ID: http://llvm.org/bugs/show_bug.cgi?id=8746 Summary: Scalar evolution loops forever (after 2h I got impatient :) on input file Product: tools Version: trunk Platform: PC OS/Version: All Status: NEW Severity: normal Priority: P Component: opt AssignedTo: unassignedbugs at nondot.org ReportedBy: arnold.schwaighofer at gmail.com CC: llvmbugs at cs.uiuc.edu Running 'opt -O2 IndVarFailure.red.ll' will loop probably forever (if one can ever make that claim). The C-file that generated this IR is also attached: IndVarFailure.red.c. This C file is the result of unrolling the HPC Challenge random access benchmark 16 times. I tried to figure out what is going and could track it down to the scalar evolution analysis that runs before 'indvars'. I tried to debug it, but I got lost in deep recursions of scalar evolution. I have also attached the IR at that point: 'opt -indvars indvarsbug.2.bp.ll' will exhibit the same behavior. Last tested with revision 121016. -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Mon Dec 6 15:21:23 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Mon, 6 Dec 2010 15:21:23 -0600 (CST) Subject: [LLVMbugs] [Bug 5335] [DSE] slow compile time on this testcase In-Reply-To: References: Message-ID: <20101206212123.64EEB2A6C12D@llvm.org> http://llvm.org/bugs/show_bug.cgi?id=5335 Chris Lattner changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |RESOLVED Resolution| |FIXED --- Comment #3 from Chris Lattner 2010-12-06 15:21:22 CST --- I believe that the root problem (initializer lowering) has already been fixed in clang. If you have the original C testcase that this came from, please verify or attach to the PR and I'll do it. -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Mon Dec 6 15:46:13 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Mon, 6 Dec 2010 15:46:13 -0600 (CST) Subject: [LLVMbugs] [Bug 8747] New: [x86] error in definition of SSE 4.1 round_ps intrinsic? Message-ID: http://llvm.org/bugs/show_bug.cgi?id=8747 Summary: [x86] error in definition of SSE 4.1 round_ps intrinsic? Product: new-bugs Version: trunk Platform: PC OS/Version: All Status: NEW Severity: normal Priority: P Component: new bugs AssignedTo: unassignedbugs at nondot.org ReportedBy: matt at pharr.org CC: llvmbugs at cs.uiuc.edu The round_ps intrinsic seems to be defined to take 2 float vector parameters in addition to the desired rounding mode. This doesn't seem to match the definition of the instruction, which does take two SSE registers, but the first operand is the register where the result is returned. As a repro case, when I run the attached test program, I get as output: % ./a.out <4 x float> (<4 x float>, <4 x float>, i32) % whereas I'd expect it to give "<4 x float> (<4 x float>, i32)" I'm not enough of an llvm internals expert to provide a suggested patch, but I assume this is an easy fix (assuming my bug report is correct.) Thanks, -matt -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Mon Dec 6 16:34:46 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Mon, 6 Dec 2010 16:34:46 -0600 (CST) Subject: [LLVMbugs] [Bug 8748] New: Default template arguments in out-of-line member definitions not rejected. Message-ID: http://llvm.org/bugs/show_bug.cgi?id=8748 Summary: Default template arguments in out-of-line member definitions not rejected. Product: clang Version: trunk Platform: PC OS/Version: Linux Status: NEW Severity: normal Priority: P Component: C++ AssignedTo: unassignedclangbugs at nondot.org ReportedBy: schaub-johannes at web.de CC: llvmbugs at cs.uiuc.edu, dgregor at apple.com Clang does accept both of the following // Testcase 1 struct A { template struct B; }; template struct A::B { }; // Testcase 2 template struct A { template struct B; }; template template struct A::B { }; The draft standard n3225 however says at [temp.param]p9 (and C++03 uses wording too, to that effect there): "A default template-argument shall not be specified in the template-parameter-lists of the definition of a member of a class template that appears outside of the member's class." So Testcase 2 seems to be ill-formed. Incidentally, Clang doesn't compile Testcase 2 if we want to make use of the default argument, as in "A::B<>". I'm not exactly sure how much of it is QoI and how much is required to be diagnosed. If we do *not* make use of the default argument, then clang doesn't complain at all if we say "A::B". However here the ill-formeness doesn't occur in the definition of the default argument, but with its mere presence in the definition, so if we require the definition of "A::B" but not the definition of the default argument, I can't clearly cut the border of where required diagnostics end and QoI starts. It would seem to be easy to diagnose the default argument nontheless to me, already when its definition is parsed. -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Mon Dec 6 16:36:46 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Mon, 6 Dec 2010 16:36:46 -0600 (CST) Subject: [LLVMbugs] [Bug 8747] [x86] error in definition of SSE 4.1 round_ss intrinsic? In-Reply-To: References: Message-ID: <20101206223646.08BC82A6C12D@llvm.org> http://llvm.org/bugs/show_bug.cgi?id=8747 Eric Christopher changed: What |Removed |Added ---------------------------------------------------------------------------- Status|ASSIGNED |RESOLVED Resolution| |INVALID --- Comment #8 from Eric Christopher 2010-12-06 16:36:45 CST --- Pretty sure this is correct. The llvm intrinsic is designed to be similar to the _mm_round_ss api intrinsic which is: __m128 _mm_round_ss (__m128 a, __m128 b, const int c) >From some documentation: A 128-bit parameter. The lowest 32 bits are the result of the rounding function on b0. The higher order 96 bits are copied directly from input parameter a. The return value is described by the following equations: r0 = RND(b0) r1 = a1 r2 = a2 r3 = a3 Yeah, it's weird. No, it really doesn't make any sense. Apparently it's intentional though. -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Mon Dec 6 17:36:39 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Mon, 6 Dec 2010 17:36:39 -0600 (CST) Subject: [LLVMbugs] [Bug 8749] New: PPC stack references broken Message-ID: http://llvm.org/bugs/show_bug.cgi?id=8749 Summary: PPC stack references broken Product: new-bugs Version: trunk Platform: Macintosh OS/Version: MacOS X Status: NEW Severity: normal Priority: P Component: new bugs AssignedTo: unassignedbugs at nondot.org ReportedBy: dalej at apple.com CC: llvmbugs at cs.uiuc.edu Run the following through llc -disable-fp-elim ; ModuleID = 'main.c' target datalayout = "E-p:32:32:32-i1:8:8-i8:8:8-i16:16:16-i32:32:32-i64:32:64-f32:32:32-f64:32:64-v64:64:64-v128:128:128-a0:0:64-f128:64:128-n32" target triple = "powerpc-apple-darwin9.8" define i32 @main() nounwind { entry: %retval = alloca i32 %0 = alloca i32 %"alloca point" = bitcast i32 0 to i32 store i32 0, i32* %0, align 4 %1 = load i32* %0, align 4 store i32 %1, i32* %retval, align 4 br label %return return: ; preds = %entry %retval1 = load i32* %retval ret i32 %retval1 } You will see stw r3, -12(r31) but R31 is not set anywhere. I suspect this was introduced in Anton's rewrite of this area, but haven't analyzed further. -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Mon Dec 6 17:41:41 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Mon, 6 Dec 2010 17:41:41 -0600 (CST) Subject: [LLVMbugs] [Bug 8714] Jump threading doesn't handle indirectbr very well. In-Reply-To: References: Message-ID: <20101206234141.C5F3A2A6C12C@llvm.org> http://llvm.org/bugs/show_bug.cgi?id=8714 Frits van Bommel changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |RESOLVED Resolution| |FIXED --- Comment #5 from Frits van Bommel 2010-12-06 17:41:41 CST --- Fixed in r121066. -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Mon Dec 6 18:58:08 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Mon, 6 Dec 2010 18:58:08 -0600 (CST) Subject: [LLVMbugs] [Bug 8723] llvm 2.8: doesn't build on Mac OS X 10.4.11, undefined symbol ___crashreporter_info__ In-Reply-To: References: Message-ID: <20101207005808.58D1B2A6C124@llvm.org> http://llvm.org/bugs/show_bug.cgi?id=8723 margali at imapmail.org changed: What |Removed |Added ---------------------------------------------------------------------------- Status|RESOLVED |REOPENED CC| |margali at imapmail.org Resolution|FIXED | --- Comment #5 from margali at imapmail.org 2010-12-06 18:58:07 CST --- (In reply to comment #4) > I think this is fixed here: > > [issola:~/workspaces/PR8723/llvm] echristo% svn ci > Sending autoconf/configure.ac > Sending configure > Sending include/llvm/Config/config.h.in > Sending lib/Support/PrettyStackTrace.cpp > Transmitting file data .... > Committed revision 120801. > > but I don't actually have a Tiger machine sitting around anywhere :) I am still getting the same error. I checked out llvm from svn (2010-12-06). configured with: ./configure \ CC=gcc-4.5 \ CXX=g++-4.5 \ CPP=cpp-4.5 \ GCJ=gcj-4.5 \ --enable-targets=host-only make fails with the following error: llvm[2]: Compiling FileCheck.cpp for Debug+Asserts build llvm[2]: Linking Debug+Asserts executable FileCheck /usr/bin/ld: Undefined symbols: ___crashreporter_info__ collect2: ld returned 1 exit status make[2]: *** [/Users//Documents/src/llvm-svn-2010-12-06/Debug+Asserts/bin/FileCheck] Error 1 make[1]: *** [FileCheck/.makeall] Error 2 make: *** [all] Error 1 I do not understand these things at all but should config.log contain: #define HAVE_CRASHREPORTER_INFO 1 when ___crashreporter_info__ is apparently unavailable? Mac OS X 10.4.11: Darwin Kernel Version 8.11.0: Wed Oct 10 18:26:00 PDT 2007; root:xnu-792.24.17~1/RELEASE_PPC Power Macintosh powerpc gcc: Target: powerpc-apple-darwin8.11.0 Configured with: ../gcc-4.5.1/configure --disable-libjava-multilib --program-suffix=-4.5 Thread model: posix gcc version 4.5.1 (GCC) config.log attached -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Mon Dec 6 19:45:48 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Mon, 6 Dec 2010 19:45:48 -0600 (CST) Subject: [LLVMbugs] [Bug 8223] [Unit Test] All death tests fail when run from lit on Windows 7. In-Reply-To: References: Message-ID: <20101207014548.9ED162A6C12C@llvm.org> http://llvm.org/bugs/show_bug.cgi?id=8223 Michael Spencer changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |RESOLVED Resolution| |FIXED --- Comment #1 from Michael Spencer 2010-12-06 19:45:48 CST --- Fixed by r121095. -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Mon Dec 6 19:59:46 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Mon, 6 Dec 2010 19:59:46 -0600 (CST) Subject: [LLVMbugs] [Bug 8710] Optimizer moves a @__cxx_global_var_init into a constant expression, causing as to fail In-Reply-To: References: Message-ID: <20101207015946.3EAE42A6C12D@llvm.org> http://llvm.org/bugs/show_bug.cgi?id=8710 Chris Lattner changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |RESOLVED Resolution| |FIXED --- Comment #7 from Chris Lattner 2010-12-06 19:59:45 CST --- Fixed in r121100, thanks! -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Mon Dec 6 20:06:27 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Mon, 6 Dec 2010 20:06:27 -0600 (CST) Subject: [LLVMbugs] [Bug 8723] llvm 2.8: doesn't build on Mac OS X 10.4.11, undefined symbol ___crashreporter_info__ In-Reply-To: References: Message-ID: <20101207020627.4D23E2A6C12E@llvm.org> http://llvm.org/bugs/show_bug.cgi?id=8723 Eric Christopher changed: What |Removed |Added ---------------------------------------------------------------------------- Status|ASSIGNED |RESOLVED Resolution| |FIXED --- Comment #9 from Eric Christopher 2010-12-06 20:06:26 CST --- OK. I've committed something, please give that a shot. Thanks! -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Mon Dec 6 21:24:36 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Mon, 6 Dec 2010 21:24:36 -0600 (CST) Subject: [LLVMbugs] [Bug 8741] [MC-COFF] Cannot handle globals @Lxxx (begins with 'L') on Win64. In-Reply-To: References: Message-ID: <20101207032436.5A6262A6C12D@llvm.org> http://llvm.org/bugs/show_bug.cgi?id=8741 NAKAMURA Takumi changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |RESOLVED Resolution| |FIXED --- Comment #2 from NAKAMURA Takumi 2010-12-06 21:24:35 CST --- Applied in r121103. There are still many issues to build clang-x64 selfhost. -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Mon Dec 6 23:25:14 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Mon, 6 Dec 2010 23:25:14 -0600 (CST) Subject: [LLVMbugs] [Bug 8750] New: Incorrect OR of conversion function templates against non-template conversion functions Message-ID: http://llvm.org/bugs/show_bug.cgi?id=8750 Summary: Incorrect OR of conversion function templates against non-template conversion functions Product: clang Version: trunk Platform: PC OS/Version: Linux Status: NEW Severity: normal Priority: P Component: C++ AssignedTo: unassignedclangbugs at nondot.org ReportedBy: schaub.johannes at googlemail.com CC: llvmbugs at cs.uiuc.edu, dgregor at apple.com Clang accepts the following code void f(int); void f(long); struct A { template operator T(); operator long(); }; int main() { f(A()); } However it's supposed to reject it: For the first f, you end up using the "operator T" (since int -> int is better than long -> int), and for the second f, you end up using the non-template operator long (since it's a non-template, it's preferred over the function template specialization). The two user defined conversion sequences created aren't comparable because they don't use the same conversion function. So the call is ambiguous. -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Mon Dec 6 23:28:41 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Mon, 6 Dec 2010 23:28:41 -0600 (CST) Subject: [LLVMbugs] [Bug 8699] Store of 3 constant bytes should be combined into one i24 store (regression) In-Reply-To: References: Message-ID: <20101207052841.458282A6C12F@llvm.org> http://llvm.org/bugs/show_bug.cgi?id=8699 Chris Lattner changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |RESOLVED Resolution| |DUPLICATE --- Comment #4 from Chris Lattner 2010-12-06 23:28:40 CST --- Agreed, those should be merged. However, they should be merged at codegen time, not in the IR. Doing it at the IR level will obscure the type relation and confuse things. Merging at codegen time is tracked by PR3473 *** This bug has been marked as a duplicate of bug 3473 *** -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Tue Dec 7 05:33:26 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Tue, 7 Dec 2010 05:33:26 -0600 (CST) Subject: [LLVMbugs] [Bug 8751] New: r120890 breaks property in Obj-C++ Message-ID: http://llvm.org/bugs/show_bug.cgi?id=8751 Summary: r120890 breaks property in Obj-C++ Product: clang Version: trunk Platform: PC OS/Version: All Status: NEW Severity: normal Priority: P Component: C++ AssignedTo: unassignedclangbugs at nondot.org ReportedBy: devlists at shadowlab.org CC: llvmbugs at cs.uiuc.edu, dgregor at apple.com r120890 by rjmccall, 04:47 ?Although we currently have explicit lvalue-to-rvalue conversions, they're not actually frequently used, because ImpCastExprToType only creates a node if the types differ?? After this commit, clang asserts while compiling the following code in Obj-C++ (it works when compiling in Obj-C) ---------------------- @interface Test {} @property int foo; @end int test(Test *t) { return t.foo; } --------------------- -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Tue Dec 7 06:52:06 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Tue, 7 Dec 2010 06:52:06 -0600 (CST) Subject: [LLVMbugs] [Bug 8752] New: Clang -Wunused should warn on unusable (member) functions. Message-ID: http://llvm.org/bugs/show_bug.cgi?id=8752 Summary: Clang -Wunused should warn on unusable (member) functions. Product: new-bugs Version: trunk Platform: All OS/Version: All Status: NEW Severity: normal Priority: P Component: new bugs AssignedTo: unassignedbugs at nondot.org ReportedBy: fvbommel at gmail.com CC: llvmbugs at cs.uiuc.edu -Wunused (or -Wunused-function) should produce warnings for (member) functions that can't possibly be used, for instance because their return value or one of their parameters (including 'this') contains a class type declared in an anonymous namespace. (extern "C" functions should probably be exempt) Some examples: ===== // From some .h file. class Bar { public: virtual void frob(); }; // From some .cpp file. namespace { // Because Foo is in an anonymous namespace, newly-introduced members // can't be used from different translation units. class Foo : public Bar { virtual void frob() { // May be used externally. } int quux() { // Unused, but doesn't produce warning. return 1; } static int baz() { // Unused, but doesn't produce warning. return 2; } }; // Unused, produces warning. Bar* getAnonNSBar() { return new Foo; } } // Unused, produces warning. static Bar* getStaticBar() { return new Foo; } // Unused, but doesn't produce warning. Foo* getFoo() { return new Foo; } // May be used externally. Bar* getBar() { return new Foo; } ===== -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Tue Dec 7 07:50:08 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Tue, 7 Dec 2010 07:50:08 -0600 (CST) Subject: [LLVMbugs] [Bug 8753] New: clang compiling net/atm/mpc.c in linux runs indefinetely Message-ID: http://llvm.org/bugs/show_bug.cgi?id=8753 Summary: clang compiling net/atm/mpc.c in linux runs indefinetely Product: clang Version: trunk Platform: PC OS/Version: Linux Status: NEW Severity: normal Priority: P Component: LLVM Codegen AssignedTo: unassignedclangbugs at nondot.org ReportedBy: dave at jikos.cz CC: llvmbugs at cs.uiuc.edu Created an attachment (id=5881) --> (http://llvm.org/bugs/attachment.cgi?id=5881) net/atm/mpc.c file set to reproduce Tested clang up to 121127, and same problem with llvm-gcc frontend -- compilation runs indefinitely (at least one whole night ...). The file is preprocessed without problems (attached with compilation commands). -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Tue Dec 7 09:25:44 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Tue, 7 Dec 2010 09:25:44 -0600 (CST) Subject: [LLVMbugs] [Bug 8720] llvm-gcc testsuite regressions with symbol aliases In-Reply-To: References: Message-ID: <20101207152544.2E54B2A6C12D@llvm.org> http://llvm.org/bugs/show_bug.cgi?id=8720 Rafael ?vila de Esp?ndola changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |RESOLVED Resolution| |FIXED --- Comment #4 from Rafael ?vila de Esp?ndola 2010-12-07 09:25:43 CST --- Fixed in 121137. -- Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug. From bugzilla-daemon at llvm.org Tue Dec 7 14:00:34 2010 From: bugzilla-daemon at llvm.org (bugzilla-daemon at llvm.org) Date: Tue, 7 Dec 2010 14:00:34 -0600 (CST) Subject: [LLVMbugs] [Bug 8723] llvm 2.8: doesn't build on Mac OS X 10.4.11, undefined symbol ___crashreporter_info__ In-Reply-To: