[llvm-bugs] [Bug 30714] New: [MIPS] Emitted instructions do not meet their predicates
https://llvm.org/bugs/show_bug.cgi?id=30714 Bug ID: 30714 Summary: [MIPS] Emitted instructions do not meet their predicates Product: new-bugs Version: trunk Hardware: PC OS: All Status: NEW Severity: normal Priority: P Component: new bugs Assignee: unassignedb...@nondot.org Reporter: simon.dar...@imgtec.com CC: llvm-bugs@lists.llvm.org Classification: Unclassified Certain instructions can be emitted in cases where they do not meet their predicates. See: https://reviews.llvm.org/D25618 and https://reviews.llvm.org/D25622 Quoting D25622: The following tests fail when this patch is applied. Failing Tests (26): LLVM :: CodeGen/Mips/cconv/callee-saved-float.ll LLVM :: CodeGen/Mips/micromips-atomic1.ll LLVM :: CodeGen/Mips/tailcall/tailcall-wrong-isa.ll LLVM :: MC/Mips/cprestore-noreorder.s LLVM :: MC/Mips/cprestore-reorder.s LLVM :: MC/Mips/elf_eflags_mips16.s LLVM :: MC/Mips/expansion-jal-sym-pic.s LLVM :: MC/Mips/expr1.s LLVM :: MC/Mips/micromips-16-bit-instructions.s LLVM :: MC/Mips/micromips-alu-instructions.s LLVM :: MC/Mips/micromips-branch-fixup.s LLVM :: MC/Mips/micromips-branch-instructions.s LLVM :: MC/Mips/micromips-control-instructions.s LLVM :: MC/Mips/micromips-dsp/valid-micromips32r3.s LLVM :: MC/Mips/micromips-el-fixup-data.s LLVM :: MC/Mips/micromips-expansions.s LLVM :: MC/Mips/micromips-fpu-instructions.s LLVM :: MC/Mips/micromips-jump-instructions.s LLVM :: MC/Mips/micromips-jump26.s LLVM :: MC/Mips/micromips-tailr.s LLVM :: MC/Mips/micromips/valid.s LLVM :: MC/Mips/micromips32r6/valid.s LLVM :: MC/Mips/mips2/valid.s LLVM :: MC/Mips/mips3/valid.s LLVM :: MC/Mips/mips4/valid.s LLVM :: MC/Mips/mips5/valid.s Some look like bugs but others look like they are deliberate (e.g. SYNC is using a loophole to implement the MIPS-II version as an alias of the MIPS32 version). -- You are receiving this mail because: You are on the CC list for the bug. ___ llvm-bugs mailing list llvm-bugs@lists.llvm.org http://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-bugs
[llvm-bugs] [Bug 30715] New: Legalization code assumes only integer typed ops may need expanding
https://llvm.org/bugs/show_bug.cgi?id=30715 Bug ID: 30715 Summary: Legalization code assumes only integer typed ops may need expanding Product: libraries Version: 3.9 Hardware: PC OS: Linux Status: NEW Severity: normal Priority: P Component: Common Code Generator Code Assignee: unassignedb...@nondot.org Reporter: nemanja.i@gmail.com CC: llvm-bugs@lists.llvm.org Classification: Unclassified Created attachment 17456 --> https://llvm.org/bugs/attachment.cgi?id=17456&action=edit Test case that causes the failure When a floating point type needs to be legalized by expanding, the code in DAGTypeLegalizer::ExpandOp_BITCAST causes an assert in DAGTypeLegalizer::IntegerToVector to trip because it assumes that the input operand is integral. But there is a type that may need to be legalized by expanding (namely ppcf128 is a pair of doubles). To reproduce: llc < legalize-bug.ll .text .abiversion 2 .file"" llc: /home/nemanjai/llvm/llvm-clean/lib/CodeGen/SelectionDAG/LegalizeTypesGeneric.cpp:321: void llvm::DAGTypeLegalizer::IntegerToVector(llvm::SDValue, unsigned int, llvm::SmallVectorImpl&, llvm::EVT): Assertion `Op.getValueType().isInteger()' failed. <... rest of stack trace omitted for brevity ...> -- You are receiving this mail because: You are on the CC list for the bug. ___ llvm-bugs mailing list llvm-bugs@lists.llvm.org http://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-bugs
[llvm-bugs] Question about debug symbol
Hello LLVM admins, I'm jinmoo seok, live in seout, Korea. i have some problem about debug symbol. i want to how to solve this problem, so i email you. source code is : #includeint main() { std::cout << "hello world" << std::endl; return 0; } this is simple source. but when i build it width debug symbol using clang, binary size is 100kb. when using gcc, binary size is 19kb i tryed "readelf -wi " readelf: dwarf_loclist_form_expr_b: Invalid argument [dwarf_loclist_from_expr_b(279)] <237> DW_AT_data_member_location: 0 byte block: () <4><23a>: Abbrev Number: 9 (DW_TAG_member) <23b> DW_AT_name: (indirect string) __precision_ <23f> DW_AT_type: <0x799> <243> DW_AT_decl_file : 2 <244> DW_AT_decl_line : 363this message is repeatedly output. what does message mean? and i tryed "readelf -s " 0 NOTYPE LOCAL DEFAULT 31 this message is output too many. why debug_str size is too big? i test both clang34 and clang38 clang38's binary size is nearly as twice as clang34's binary. i wan't to why this problem happen. My environment: i386 freebsd11 ___ llvm-bugs mailing list llvm-bugs@lists.llvm.org http://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-bugs
[llvm-bugs] [Bug 30696] lld: implement -nopie
https://llvm.org/bugs/show_bug.cgi?id=30696 George Rimar changed: What|Removed |Added Status|NEW |RESOLVED Resolution|--- |FIXED --- Comment #4 from George Rimar --- r284388 -- You are receiving this mail because: You are on the CC list for the bug. ___ llvm-bugs mailing list llvm-bugs@lists.llvm.org http://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-bugs
[llvm-bugs] [Bug 30712] clang ignores constexpr for static members of objects with reference template type
https://llvm.org/bugs/show_bug.cgi?id=30712 Richard Smith changed: What|Removed |Added Status|NEW |RESOLVED CC||richard-l...@metafoo.co.uk Resolution|--- |INVALID --- Comment #1 from Richard Smith --- Clang is correct. In order for a call to a constexpr function to be evaluated at compile time, it must be possible to (symbolically) evaluate all of its arguments. In the failing case, it is not possible to evaluate the value of the implicit 'this' parameter. In the successful case, it evaluates symbolically to the address of the caller's parameter. -- You are receiving this mail because: You are on the CC list for the bug. ___ llvm-bugs mailing list llvm-bugs@lists.llvm.org http://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-bugs
[llvm-bugs] [Bug 30680] lld + LTO: symbol __start_($SECTION_NAME) is missing type
https://llvm.org/bugs/show_bug.cgi?id=30680 Peter Collingbourne changed: What|Removed |Added Status|NEW |RESOLVED Resolution|--- |FIXED --- Comment #7 from Peter Collingbourne --- Yes, the fix landed in r284168. -- You are receiving this mail because: You are on the CC list for the bug. ___ llvm-bugs mailing list llvm-bugs@lists.llvm.org http://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-bugs
[llvm-bugs] [Bug 30716] New: Invalid suffix 'l' on instructions accessing 16-bit registers
https://llvm.org/bugs/show_bug.cgi?id=30716 Bug ID: 30716 Summary: Invalid suffix 'l' on instructions accessing 16-bit registers Product: libraries Version: 3.9 Hardware: PC OS: Linux Status: NEW Severity: normal Priority: P Component: Backend: X86 Assignee: unassignedb...@nondot.org Reporter: nholc...@wisc.edu CC: llvm-bugs@lists.llvm.org Classification: Unclassified The below are two examples of decodings produced by LLVM where the instruction is suffixed with 'l', but accesses a 16-bit register. Bytes: 66 d1 d2 Output: rcll %dx Bytes: 66 b8 09 35 Output: movl $13577, %sp I suspect this is a result of ignoring the 0x66 prefix when choosing opcode suffix. -- You are receiving this mail because: You are on the CC list for the bug. ___ llvm-bugs mailing list llvm-bugs@lists.llvm.org http://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-bugs
[llvm-bugs] [Bug 30717] New: Exhange instruction uses two different sized registers
https://llvm.org/bugs/show_bug.cgi?id=30717 Bug ID: 30717 Summary: Exhange instruction uses two different sized registers Product: libraries Version: 3.9 Hardware: PC OS: Linux Status: NEW Severity: normal Priority: P Component: Backend: X86 Assignee: unassignedb...@nondot.org Reporter: nholc...@wisc.edu CC: llvm-bugs@lists.llvm.org Classification: Unclassified When the 0x66 prefix is present, two differently sized registers are used by the decoded xchg instruction. Here's an example decoded with LLVM: Bytes: 66 94 Output: xchgl %sp, %eax The second operand should be %ax. -- You are receiving this mail because: You are on the CC list for the bug. ___ llvm-bugs mailing list llvm-bugs@lists.llvm.org http://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-bugs
[llvm-bugs] [Bug 30718] New: Failure to disassemble certain MOV with rex.W
https://llvm.org/bugs/show_bug.cgi?id=30718 Bug ID: 30718 Summary: Failure to disassemble certain MOV with rex.W Product: libraries Version: 3.9 Hardware: PC OS: Linux Status: NEW Severity: normal Priority: P Component: Backend: X86 Assignee: unassignedb...@nondot.org Reporter: nholc...@wisc.edu CC: llvm-bugs@lists.llvm.org Classification: Unclassified The LLVM disassembler returns an error when given the following bytes: 4c 8e ca Looking at the output of XED and other disassemblers, it appears that the instruction should be: mov %rdx, %cs -- You are receiving this mail because: You are on the CC list for the bug. ___ llvm-bugs mailing list llvm-bugs@lists.llvm.org http://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-bugs
[llvm-bugs] [Bug 30719] New: Failure to disassemble variant encoding of test instruction
https://llvm.org/bugs/show_bug.cgi?id=30719 Bug ID: 30719 Summary: Failure to disassemble variant encoding of test instruction Product: libraries Version: 3.9 Hardware: PC OS: Linux Status: NEW Severity: normal Priority: P Component: Backend: X86 Assignee: unassignedb...@nondot.org Reporter: nholc...@wisc.edu CC: llvm-bugs@lists.llvm.org Classification: Unclassified The following are the bytes of a test instruction and the expected output. LLVM returns an error when disassembling these bytes. Bytes: f6 0b 50 Expected output: test $0x50,(%rbx) The other possible encoding of this instruction is f6 03 50. -- You are receiving this mail because: You are on the CC list for the bug. ___ llvm-bugs mailing list llvm-bugs@lists.llvm.org http://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-bugs
[llvm-bugs] [Bug 30702] ICE when compiling with --coverage but without -g
https://llvm.org/bugs/show_bug.cgi?id=30702 Davide Italiano changed: What|Removed |Added Status|ASSIGNED|RESOLVED Resolution|--- |FIXED --- Comment #2 from Davide Italiano --- r284418 -- You are receiving this mail because: You are on the CC list for the bug. ___ llvm-bugs mailing list llvm-bugs@lists.llvm.org http://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-bugs
[llvm-bugs] [Bug 30720] New: [LTO] Try to be option compatible with the gold plugin
https://llvm.org/bugs/show_bug.cgi?id=30720 Bug ID: 30720 Summary: [LTO] Try to be option compatible with the gold plugin Product: lld Version: unspecified Hardware: PC OS: All Status: NEW Severity: normal Priority: P Component: ELF Assignee: unassignedb...@nondot.org Reporter: dav...@freebsd.org CC: llvm-bugs@lists.llvm.org Classification: Unclassified ThinLTO landed, and Sean pointed out that we currently ignore --plugin-opt and we could instead change LLD to DTRT (i.e. parse plugin-opt arguments and alias to existing lld options). There's an agreement this is a good thing to do, so let's keep track of this here. -- You are receiving this mail because: You are on the CC list for the bug. ___ llvm-bugs mailing list llvm-bugs@lists.llvm.org http://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-bugs
[llvm-bugs] [Bug 30721] New: SAL instruction decoded as bad bytes
https://llvm.org/bugs/show_bug.cgi?id=30721 Bug ID: 30721 Summary: SAL instruction decoded as bad bytes Product: libraries Version: 3.9 Hardware: PC OS: Linux Status: NEW Severity: normal Priority: P Component: Backend: X86 Assignee: unassignedb...@nondot.org Reporter: nholc...@wisc.edu CC: llvm-bugs@lists.llvm.org Classification: Unclassified The following bytes correspond to a 2-byte sal instruction, but LLVM returns an error instead. Bytes: d2 f0 Expected output: salb %cl, %al -- You are receiving this mail because: You are on the CC list for the bug. ___ llvm-bugs mailing list llvm-bugs@lists.llvm.org http://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-bugs
[llvm-bugs] [Bug 30722] New: Warn if -coverage is passed (and debug info are not present)
https://llvm.org/bugs/show_bug.cgi?id=30722 Bug ID: 30722 Summary: Warn if -coverage is passed (and debug info are not present) Product: clang Version: unspecified Hardware: PC OS: All Status: NEW Severity: normal Priority: P Component: Frontend Assignee: unassignedclangb...@nondot.org Reporter: dav...@freebsd.org CC: llvm-bugs@lists.llvm.org Classification: Unclassified Gcov data is useless without debug info so we should error/warn if -coverage is passed but -g is not. -- You are receiving this mail because: You are on the CC list for the bug. ___ llvm-bugs mailing list llvm-bugs@lists.llvm.org http://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-bugs
[llvm-bugs] [Bug 30723] New: false positive for "-Wzero-length-array"
https://llvm.org/bugs/show_bug.cgi?id=30723
Bug ID: 30723
Summary: false positive for "-Wzero-length-array"
Product: clang
Version: trunk
Hardware: PC
OS: Windows NT
Status: NEW
Severity: normal
Priority: P
Component: C++
Assignee: unassignedclangb...@nondot.org
Reporter: ca...@carter.net
CC: dgre...@apple.com, llvm-bugs@lists.llvm.org
Classification: Unclassified
This program:
template
auto f() -> decltype(new int[N]{});
using F = decltype(f());
compiled with -Wzero-length-array (which happens to be under -pedantic) results
in:
prog.cc:2:32: warning: zero size arrays are an extension [-Wzero-length-array]
auto f() -> decltype(new int[N]{});
^
--
You are receiving this mail because:
You are on the CC list for the bug.
___
llvm-bugs mailing list
llvm-bugs@lists.llvm.org
http://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-bugs
[llvm-bugs] [Bug 30709] fatal error: llvm/IR/Attributes.inc: No such file or directory
https://llvm.org/bugs/show_bug.cgi?id=30709 Chris Bieneman changed: What|Removed |Added Status|NEW |RESOLVED CC||chris.biene...@me.com Resolution|--- |FIXED --- Comment #1 from Chris Bieneman --- Fixed in r28. We should probably flag this for a 3.9.x update. I'll file the merge request bug shortly. -- You are receiving this mail because: You are on the CC list for the bug. ___ llvm-bugs mailing list llvm-bugs@lists.llvm.org http://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-bugs
[llvm-bugs] [Bug 30724] New: Merge r284444 to 3.9.x
https://llvm.org/bugs/show_bug.cgi?id=30724 Bug ID: 30724 Summary: Merge r28 to 3.9.x Product: lld Version: unspecified Hardware: PC OS: All Status: NEW Severity: normal Priority: P Component: All Bugs Assignee: unassignedb...@nondot.org Reporter: chris.biene...@me.com CC: llvm-bugs@lists.llvm.org Classification: Unclassified This is a minor fix for a missing dependency in the build system. -- You are receiving this mail because: You are on the CC list for the bug. ___ llvm-bugs mailing list llvm-bugs@lists.llvm.org http://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-bugs
[llvm-bugs] [Bug 30725] New: Merge r281650 to 3.9
https://llvm.org/bugs/show_bug.cgi?id=30725 Bug ID: 30725 Summary: Merge r281650 to 3.9 Product: libraries Version: 3.9 Hardware: All OS: All Status: NEW Severity: normal Priority: P Component: Scalar Optimizations Assignee: unassignedb...@nondot.org Reporter: b...@comstyle.com CC: llvm-bugs@lists.llvm.org Classification: Unclassified Please merge r281650 to 3.9. This fixes PR30402. https://github.com/rust-lang/rust/issues/36474 -- You are receiving this mail because: You are on the CC list for the bug. ___ llvm-bugs mailing list llvm-bugs@lists.llvm.org http://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-bugs
