[llvm-bugs] [Bug 51711] New: [postmisched] wrong code generated after post-mi scheduling
https://bugs.llvm.org/show_bug.cgi?id=51711 Bug ID: 51711 Summary: [postmisched] wrong code generated after post-mi scheduling Product: libraries Version: trunk Hardware: PC OS: Linux Status: NEW Severity: enhancement Priority: P Component: Common Code Generator Code Assignee: unassignedb...@nondot.org Reporter: wweiand...@gmail.com CC: llvm-bugs@lists.llvm.org before post-ra scheduling: ... renamable $w17 = ANDSWri renamable $w6, 15, implicit-def dead $nzcv $w18 = SUBSWrs renamable $w13, renamable $w14, 0, implicit-def $nzcv, implicit-def dead $nzcv ... renamable $w11 = ANDSWri killed renamable $w2, 0, implicit-def $nzcv renamable $w9 = CSINCWr killed renamable $w8, renamable $w8, 0, implicit killed $nzcv ... after post-ra scheduling: ... $w18 = SUBSWrs renamable $w13, renamable $w14, 0, implicit-def $nzcv, implicit-def dead $nzcv ... renamable $w11 = ANDSWri killed renamable $w2, 0, implicit-def $nzcv renamable $w17 = ANDSWri renamable $w6, 15, implicit-def dead $nzcv renamable $w9 = CSINCWr killed renamable $w8, renamable $w8, 0, implicit killed $nzcv ... "renamable $w17 = ANDSWri renamable $w6, 15, implicit-def dead $nzcv" was wrong scheduled, because "implicit-def dead $nzcv" in this inst will break the true dependence between "renamable $w11 = ..." and "renamable $w9 = ..." -- You are receiving this mail because: You are on the CC list for the bug.___ llvm-bugs mailing list llvm-bugs@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-bugs
[llvm-bugs] [Bug 51712] New: Clang runs OOM when checking for constant initialization of array
https://bugs.llvm.org/show_bug.cgi?id=51712 Bug ID: 51712 Summary: Clang runs OOM when checking for constant initialization of array Product: clang Version: trunk Hardware: PC OS: All Status: NEW Severity: enhancement Priority: P Component: Frontend Assignee: unassignedclangb...@nondot.org Reporter: kadircetinkaya.06...@gmail.com CC: llvm-bugs@lists.llvm.org, neeil...@live.com, richard-l...@metafoo.co.uk ``` cat > a.cc <(void) () at /usr/local/google/home/kadircet/repos/llvm/llvm/include/llvm/ADT/STLExtras.h:185 #27 0x02870041 in operator() () at /usr/local/google/home/kadircet/repos/llvm/llvm/include/llvm/ADT/STLExtras.h:209 #28 RunSafely () at /usr/local/google/home/kadircet/repos/llvm/llvm/lib/Support/CrashRecoveryContext.cpp:424 #29 0x02f5c1f7 in Execute () at /usr/local/google/home/kadircet/repos/llvm/clang/lib/Driver/Job.cpp:404 #30 0x02f272c8 in ExecuteCommand () at /usr/local/google/home/kadircet/repos/llvm/clang/lib/Driver/Compilation.cpp:195 #31 0x02f27817 in ExecuteJobs () at /usr/local/google/home/kadircet/repos/llvm/clang/lib/Driver/Compilation.cpp:246 #32 0x02f41c88 in ExecuteCompilation () at /usr/local/google/home/kadircet/repos/llvm/clang/lib/Driver/Driver.cpp:1510 #33 0x009e5df4 in main () at /usr/local/google/home/kadircet/repos/llvm/clang/tools/driver/driver.cpp:502 ``` All of the backtrace samples I grabbed had `checkForConstantInitialization` as LCA. -- You are receiving this mail because: You are on the CC list for the bug.___ llvm-bugs mailing list llvm-bugs@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-bugs
[llvm-bugs] [Bug 51711] [postmisched] wrong code generated after post-mi scheduling
https://bugs.llvm.org/show_bug.cgi?id=51711 Florian Hahn changed: What|Removed |Added CC||florian_h...@apple.com Resolution|--- |INVALID Status|NEW |RESOLVED --- Comment #1 from Florian Hahn --- Thanks for the report. Could you attach a MIR file to reproduce the issue with instructions on how to reproduce it and re-open the issue? -- You are receiving this mail because: You are on the CC list for the bug.___ llvm-bugs mailing list llvm-bugs@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-bugs
[llvm-bugs] [Bug 51711] [postmisched] wrong code generated after post-mi scheduling
https://bugs.llvm.org/show_bug.cgi?id=51711 Dave Green changed: What|Removed |Added Status|RESOLVED|REOPENED Resolution|INVALID |--- --- Comment #2 from Dave Green --- This is https://reviews.llvm.org/D109069? Or a separate issue? -- You are receiving this mail because: You are on the CC list for the bug.___ llvm-bugs mailing list llvm-bugs@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-bugs
[llvm-bugs] [Bug 51713] New: __builtin_frame_address(0) returning RSP for x86_64-windows (instead of RBP)
https://bugs.llvm.org/show_bug.cgi?id=51713
Bug ID: 51713
Summary: __builtin_frame_address(0) returning RSP for
x86_64-windows (instead of RBP)
Product: new-bugs
Version: trunk
Hardware: PC
OS: Linux
Status: NEW
Severity: enhancement
Priority: P
Component: new bugs
Assignee: unassignedb...@nondot.org
Reporter: xtk...@gmail.com
CC: htmldevelo...@gmail.com, llvm-bugs@lists.llvm.org
Created attachment 25219
--> https://bugs.llvm.org/attachment.cgi?id=25219&action=edit
Repro
$ cat frameaddr.c
void bar(int *n);
void foo(int **p, int n)
{
__asm__ __volatile__ ("nop");
*p = __builtin_frame_address(0);
__asm__ __volatile__ ("nop");
bar(&n);
}
$ clang --target=x86_64-windows frameaddr.c -c -o - | llvm-objdump -d -
:file format coff-x86-64
Disassembly of section .text:
:
0: 55pushq %rbp
1: 48 83 ec 30 subq$48, %rsp
5: 48 8d 6c 24 30leaq48(%rsp), %rbp
a: 89 55 fc movl%edx, -4(%rbp)
d: 48 89 4d f0 movq%rcx, -16(%rbp)
11: 90nop
12: 48 8d 4d d0 leaq-48(%rbp), %rcx
16: 48 8b 45 f0 movq-16(%rbp), %rax
1a: 48 89 08 movq%rcx, (%rax)
1d: 90nop
1e: 48 8d 4d fc leaq-4(%rbp), %rcx
22: e8 00 00 00 00callq 0x27
27: 90nop
28: 48 83 c4 30 addq$48, %rsp
2c: 5dpopq%rbp
2d: c3retq
The instruction "leaq -48(%rbp), %rcx" at 0x12 should instead be something like
"movq %rbp, %rcx".
I don't know whether this is expected, but this behavior breaks
__builtin_longjmp for x86_64-windows (PR50142).
--
You are receiving this mail because:
You are on the CC list for the bug.___
llvm-bugs mailing list
llvm-bugs@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-bugs
[llvm-bugs] [Bug 51714] New: liblzma compress/uncompress mismatch when built with llvm >=12 on FreeBSD powerpc64 (BE)
https://bugs.llvm.org/show_bug.cgi?id=51714 Bug ID: 51714 Summary: liblzma compress/uncompress mismatch when built with llvm >=12 on FreeBSD powerpc64 (BE) Product: new-bugs Version: unspecified Hardware: PC OS: FreeBSD Status: NEW Severity: enhancement Priority: P Component: new bugs Assignee: unassignedb...@nondot.org Reporter: alfredo.jun...@eldorado.org.br CC: htmldevelo...@gmail.com, llvm-bugs@lists.llvm.org On FreeBSD 13/powerpc64 (BE, ELFv2 ABI), a xz (liblzma.so) compiled with llvm12 or later isn't able to uncompress files produced by other machines or downloaded from the Internet (data corruption). It can uncompress files compressed by itself (the same library binary), but other machines won't recognize the .xz as well (data corruption). Bisect points to this commit[1] as source of regression. Reverting it makes liblzma.so work correctly again, but it's not clear to me how it's interefering. FreeBSD 13/powerpc64le and amd64 are not affected. How to reproduce: 1 - install FreeBSD 13/powerpc64 2 - install llvm12 or latest devel (pkg install llvm12 llvm-devel) 3 - take source from http://git.tukaani.org/xz.git and compile branch v5.0 using clang12 or later as CC 4 - download an .xz from the internet and try to uncompress it [1] https://github.com/llvm/llvm-project/commit/42eaf4fe0adef3344adfd9fbccd49f325cb549ef -- You are receiving this mail because: You are on the CC list for the bug.___ llvm-bugs mailing list llvm-bugs@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-bugs
[llvm-bugs] [Bug 51715] New: [AMDGPU][MC][GFX10] Incorrect FLAT GLOBAL fcmpswap data size
https://bugs.llvm.org/show_bug.cgi?id=51715 Bug ID: 51715 Summary: [AMDGPU][MC][GFX10] Incorrect FLAT GLOBAL fcmpswap data size Product: libraries Version: trunk Hardware: PC OS: Windows NT Status: NEW Severity: enhancement Priority: P Component: Backend: AMDGPU Assignee: unassignedb...@nondot.org Reporter: d-...@mail.ru CC: llvm-bugs@lists.llvm.org global_atomic_fcmpswap and global_atomic_fcmpswap_x2 are similar to cmpswap opcodes but operate with f32 data. They have 2-element data operand: - global_atomic_fcmpswap: data operand type is f32x2 - global_atomic_fcmpswap: data operand type is f64x2 However llvm assembler expects one-element data operand. An example of a failed test: global_atomic_fcmpswap v[1:2], v[2:3], off The assembler expects that data operand size is 32 bits which is incorrect: global_atomic_fcmpswap v[1:2], v2, off // invalid opcode but assembled ok Compare this with cmpswap test which is accepted by the assembler: global_atomic_cmpswap v[1:2], v[2:3], off -- You are receiving this mail because: You are on the CC list for the bug.___ llvm-bugs mailing list llvm-bugs@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-bugs
[llvm-bugs] [Bug 51716] New: Please remove autoconf in the "build scripts" components
https://bugs.llvm.org/show_bug.cgi?id=51716 Bug ID: 51716 Summary: Please remove autoconf in the "build scripts" components Product: Bugzilla Admin Version: unspecified Hardware: PC OS: All Status: NEW Severity: enhancement Priority: P Component: Products Assignee: unassignedb...@nondot.org Reporter: sylves...@debian.org CC: kristof.be...@gmail.com, llvm-bugs@lists.llvm.org, r...@google.com, t.p.northo...@gmail.com it has been a while since we don't have autoconf around: https://bugs.llvm.org/enter_bug.cgi?product=Build%20scripts -- You are receiving this mail because: You are on the CC list for the bug.___ llvm-bugs mailing list llvm-bugs@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-bugs
[llvm-bugs] [Bug 51717] New: [AMDGPU][MC][GFX10] GLOBAL and FLAT atomics discard dlc modifier
https://bugs.llvm.org/show_bug.cgi?id=51717 Bug ID: 51717 Summary: [AMDGPU][MC][GFX10] GLOBAL and FLAT atomics discard dlc modifier Product: libraries Version: trunk Hardware: PC OS: Windows NT Status: NEW Severity: enhancement Priority: P Component: Backend: AMDGPU Assignee: unassignedb...@nondot.org Reporter: d-...@mail.ru CC: llvm-bugs@lists.llvm.org GLOBAL and FLAT atomics accept dlc modifier but it is silently discarded. An example: flat_atomic_add v[1:2], v2 flat_atomic_add v[1:2], v2 dlc Both instructions are assembled into the same code with zero dlc bit: 0x00,0x00,0xc8,0xdc,0x01,0x02,0x7d,0x00 AMD documentation does not describe any limitations on use of dlc with atomics so the questions are: 1. Is dlc discarded on purpose? 2. If yes, should validator be corrected to trigger an error for atomics with dlc? -- You are receiving this mail because: You are on the CC list for the bug.___ llvm-bugs mailing list llvm-bugs@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-bugs
[llvm-bugs] [Bug 51718] New: -DLLVM_ENABLE_RUNTIMES="libcxx; libcxxabi" is showing projects as disabled
https://bugs.llvm.org/show_bug.cgi?id=51718 Bug ID: 51718 Summary: -DLLVM_ENABLE_RUNTIMES="libcxx;libcxxabi" is showing projects as disabled Product: Build scripts Version: trunk Hardware: PC OS: All Status: NEW Severity: enhancement Priority: P Component: cmake Assignee: unassignedb...@nondot.org Reporter: sylves...@debian.org CC: llvm-bugs@lists.llvm.org Following https://libcxx.llvm.org/BuildingLibcxx.html#bootstrapping-build cmake -G Ninja -S llvm -B build -DLLVM_ENABLE_PROJECTS="clang" -DLLVM_ENABLE_RUNTIMES="libcxx;libcxxabi" [...] -- libcxx project is disabled -- libcxxabi project is disabled I get that runtimes are different from projects but we should probably have -- libcxx runtime is enabled (or -- libcxx project is enabled) -- You are receiving this mail because: You are on the CC list for the bug.___ llvm-bugs mailing list llvm-bugs@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-bugs
[llvm-bugs] [Bug 51719] New: [AArch64][SVE] Merge 14e1a4a6eef2fb95ec852c9ddfc597f80bba3226 into LLVM 13.0.0
https://bugs.llvm.org/show_bug.cgi?id=51719 Bug ID: 51719 Summary: [AArch64][SVE] Merge 14e1a4a6eef2fb95ec852c9ddfc597f80bba3226 into LLVM 13.0.0 Product: libraries Version: trunk Hardware: All OS: All Status: NEW Severity: release blocker Priority: P Component: Backend: AArch64 Assignee: unassignedb...@nondot.org Reporter: bradley.sm...@arm.com CC: arnaud.degrandmai...@arm.com, llvm-bugs@lists.llvm.org, smithp...@googlemail.com, ties.st...@arm.com Blocks: 51236 Please merge 14e1a4a6eef2fb95ec852c9ddfc597f80bba3226 (https://reviews.llvm.org/D109145) into the LLVM 13 release branch. This fixes another codegen fault when lowering fixed length SVE gather/scatter nodes when using a specific addressing mode and datatype. Referenced Bugs: https://bugs.llvm.org/show_bug.cgi?id=51236 [Bug 51236] [meta] 13.0.0 Release Blockers -- You are receiving this mail because: You are on the CC list for the bug.___ llvm-bugs mailing list llvm-bugs@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-bugs
[llvm-bugs] [Bug 51720] New: [AMDGPU][GFX10] MIMG opcodes with bias and a16 have incorrect address size
https://bugs.llvm.org/show_bug.cgi?id=51720 Bug ID: 51720 Summary: [AMDGPU][GFX10] MIMG opcodes with bias and a16 have incorrect address size Product: libraries Version: trunk Hardware: PC OS: Windows NT Status: NEW Severity: enhancement Priority: P Component: Backend: AMDGPU Assignee: unassignedb...@nondot.org Reporter: d-...@mail.ru CC: llvm-bugs@lists.llvm.org According with AMD documentation, bias size is affected by a16. However AMDGPU assembler does not correct bias size which results in test failures for GATHER_B* and SAMPLER_B* opcodes. For example, the following instruction may be assembled using sp3 but not llvm AMDGPU assembler: image_gather4_b v[5:8], v[1:2], s[8:15], s[12:15] dmask:0x8 dim:SQ_RSRC_IMG_CUBE a16 AMDGPU assembler expects 96-bit address size: first dword:bias second dword: t, s third dword:0, face Actual address in this case should be as follows: first dword:s, bias second dword: face, t Current implementation computes address size as shown here (see https://github.com/llvm/llvm-project/blob/e5438f386854136d848989315f53788808afa37a/llvm/lib/Target/AMDGPU/Utils/AMDGPUBaseInfo.cpp#L158): unsigned AddrWords = BaseOpcode->NumExtraArgs; unsigned AddrComponents = ... if (IsA16) AddrWords += divideCeil(AddrComponents, 2); else AddrWords += AddrComponents; In other words, NumExtraArgs accounts for offset, bias and z-compare (but not derivatives which are handled elsewhere). So currently bias size is not affected by a16. Note that bias size may affect address size under the following conditions: - a16=1. - opcode has no z-compare (it follows bias and always 32-bit aligned). - opcode has no derivatives (it follows z-comp and also 32-bit aligned). It looks like NumExtraArgs should be split into 3 separate fields (offset, bias, zcomp) to account for bias size variations. -- You are receiving this mail because: You are on the CC list for the bug.___ llvm-bugs mailing list llvm-bugs@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-bugs
[llvm-bugs] [Bug 51721] New: clang_rt_asan_dynamic_x86_64 crash running ASan build of Firefox on Windows 11
https://bugs.llvm.org/show_bug.cgi?id=51721 Bug ID: 51721 Summary: clang_rt_asan_dynamic_x86_64 crash running ASan build of Firefox on Windows 11 Product: compiler-rt Version: unspecified Hardware: PC OS: Windows NT Status: NEW Severity: enhancement Priority: P Component: asan Assignee: unassignedb...@nondot.org Reporter: cpeter...@mozilla.com CC: llvm-bugs@lists.llvm.org When I try to run an ASan build of Firefox on Windows 11 Beta (version 10.0.22000.1), it crashes with a 0xc142 (STATUS_DLL_INIT_FAILED) error. I filed Firefox bug https://bugzilla.mozilla.org/show_bug.cgi?id=1723593 The crash happens in Clang's runtime clang_rt.asan_dynamic-x86_64.dll here: 0:000> r rax=0005 rbx=7ffcc085c200 rcx=00d28548 rdx= rsi= rdi=7ffcc11b7bd0 rip=7ffcc0831ac6 rsp=00b44f5fe1b8 rbp=0001 r8=74d28548 r9=8548 r10=8548 r11=8080808080808080 r12=7ffcf50d8200 r13=7ffcc0873508 r14=0002 r15=7ffcc11b7bd0 iopl=0 nv up ei pl nz na po nc cs=0033 ss=002b ds=002b es=002b fs=0053 gs=002b efl=0204 clang_rt_asan_dynamic_x86_64+0x1ac6: 7ffc`c0831ac6 cc int 3 0:000> knL # Child-SP RetAddr Call Site 00 00b4`4f5fe1b8 7ffc`c0831b13 clang_rt_asan_dynamic_x86_64+0x1ac6 01 00b4`4f5fe1c0 7ffc`c08320db clang_rt_asan_dynamic_x86_64+0x1b13 02 00b4`4f5fe260 7ffc`c0861ebd clang_rt_asan_dynamic_x86_64+0x20db 03 00b4`4f5fe2c0 7ffc`c086b88f clang_rt_asan_dynamic_x86_64!_asan_wrap_atol+0x2fd 04 00b4`4f5fe2f0 7ffc`c084d4f9 clang_rt_asan_dynamic_x86_64!_asan_storeN_noabort+0x57f 05 00b4`4f5fe360 7ffc`c084dcb8 clang_rt_asan_dynamic_x86_64!_ubsan_get_current_report_data+0x1679 06 00b4`4f5fe400 7ffc`c0863106 clang_rt_asan_dynamic_x86_64!_ubsan_get_current_report_data+0x1e38 07 00b4`4f5fe450 7ffc`c08a0d16 clang_rt_asan_dynamic_x86_64!_asan_memmove+0x5d6 08 00b4`4f5fece0 7ffc`c08a0d9c clang_rt_asan_dynamic_x86_64!_ubsan_handle_function_type_mismatch_v1_abort+0x2ea76 09 00b4`4f5fed10 7ffc`c08ab7af clang_rt_asan_dynamic_x86_64!_ubsan_handle_function_type_mismatch_v1_abort+0x2eafc 0a 00b4`4f5fed40 7ffc`c0872f1c clang_rt_asan_dynamic_x86_64!_ubsan_handle_function_type_mismatch_v1_abort+0x3950f 0b 00b4`4f5fed70 7ffc`c0873226 clang_rt_asan_dynamic_x86_64!_ubsan_handle_function_type_mismatch_v1_abort+0xc7c 0c 00b4`4f5feda0 7ffc`c0873424 clang_rt_asan_dynamic_x86_64!_ubsan_handle_function_type_mismatch_v1_abort+0xf86 0d 00b4`4f5fedd0 7ffc`f507fb97 clang_rt_asan_dynamic_x86_64!_ubsan_handle_function_type_mismatch_v1_abort+0x1184 0e 00b4`4f5fee30 7ffc`f50b2a2e ntdll!LdrpCallInitRoutine+0x6b 0f 00b4`4f5feea0 7ffc`f50b27de ntdll!LdrpInitializeNode+0x1ca 10 00b4`4f5fefe0 7ffc`f50b2850 ntdll!LdrpInitializeGraphRecurse+0x42 11 00b4`4f5ff020 7ffc`f511f5b2 ntdll!LdrpInitializeGraphRecurse+0xb4 12 00b4`4f5ff060 7ffc`f510cee2 ntdll!LdrpInitializeProcess+0x1c86 13 00b4`4f5ff420 7ffc`f50ba7a3 ntdll!_LdrpInitialize+0x52706 14 00b4`4f5ff4a0 7ffc`f50ba6ce ntdll!LdrpInitializeInternal+0x6b 15 00b4`4f5ff720 ` ntdll!LdrInitializeThunk+0xe Firefox ASan builds had a similar crash on Windows 10 four years ago: GetInstructionSize bug 32895 -- You are receiving this mail because: You are on the CC list for the bug.___ llvm-bugs mailing list llvm-bugs@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-bugs
[llvm-bugs] [Bug 48622] "make check" in glibc doesn't generate test summary on Linux/x86-64
https://bugs.llvm.org/show_bug.cgi?id=48622 Fangrui Song changed: What|Removed |Added Resolution|--- |FIXED Status|NEW |RESOLVED --- Comment #1 from Fangrui Song --- Fixed in LLD 13.0.0 with glibc 2.35 -- You are receiving this mail because: You are on the CC list for the bug.___ llvm-bugs mailing list llvm-bugs@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-bugs
[llvm-bugs] [Bug 51722] New: LoopIdiom regression: ctlz no longer recognized
https://bugs.llvm.org/show_bug.cgi?id=51722
Bug ID: 51722
Summary: LoopIdiom regression: ctlz no longer recognized
Product: libraries
Version: trunk
Hardware: PC
OS: Linux
Status: NEW
Severity: enhancement
Priority: P
Component: Scalar Optimizations
Assignee: unassignedb...@nondot.org
Reporter: david.bolvan...@gmail.com
CC: llvm-bugs@lists.llvm.org
unsigned long long FIO_highbit64(unsigned long long v)
{
unsigned count = 0;
v >>= 1;
while (v) { v >>= 1; count++; }
return count;
}
Since LLVM 13, we lost recognized "@llvm.ctlz.i64".
https://godbolt.org/z/nodKdfab4
--
You are receiving this mail because:
You are on the CC list for the bug.___
llvm-bugs mailing list
llvm-bugs@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-bugs
[llvm-bugs] [Bug 51723] New: [AMDGPU][MC][GFX10] MIMG gather4 opcodes failed to disassemble when d16 is specified
https://bugs.llvm.org/show_bug.cgi?id=51723 Bug ID: 51723 Summary: [AMDGPU][MC][GFX10] MIMG gather4 opcodes failed to disassemble when d16 is specified Product: libraries Version: trunk Hardware: PC OS: Windows NT Status: NEW Severity: enhancement Priority: P Component: Backend: AMDGPU Assignee: unassignedb...@nondot.org Reporter: d-...@mail.ru CC: llvm-bugs@lists.llvm.org An example of failed test: 0x08,0x08,0x00,0xf1,0x01,0xfe,0x62,0x80 Expected output: image_gather4 v[254:255], v[1:2], s[8:15], s[12:15] dmask:0x8 dim:SQ_RSRC_IMG_2D d16 Actual result: s_add_u32 s98, s1, src_lds_direct ; encoding: [0x01,0xfe,0x62,0x80] Note that instruction image_gather4 shown above is assembled fine with expected encoding but this encoding cannot be disassembled. -- You are receiving this mail because: You are on the CC list for the bug.___ llvm-bugs mailing list llvm-bugs@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-bugs
[llvm-bugs] [Bug 51724] New: Constant ValueID not recognized
https://bugs.llvm.org/show_bug.cgi?id=51724 Bug ID: 51724 Summary: Constant ValueID not recognized Product: new-bugs Version: trunk Hardware: PC OS: All Status: NEW Severity: enhancement Priority: P Component: new bugs Assignee: unassignedb...@nondot.org Reporter: pho...@chromium.org CC: htmldevelo...@gmail.com, llvm-bugs@lists.llvm.org Created attachment 25221 --> https://bugs.llvm.org/attachment.cgi?id=25221&action=edit Clang crash reproducer We ran into this failure when trying to use -fmerge-functions. ``` Constant ValueID not recognized. UNREACHABLE executed at /src/clang-llvm/llvm-project/llvm/lib/Transforms/Utils/FunctionComparator.cpp:399! PLEASE submit a bug report to https://bugs.llvm.org/ and include the crash backtrace, preprocessed source, and associated run script. Stack dump: 0. Program arguments: ../bin/clang++ -cc1 -triple x86_64-unknown-fuchsia -emit-obj -massembler-fatal-warnings --mrelax-relocations -disable-free -disable-llvm-verifier -discard-value-names -main-file-name zircon_platform_buffer.cc -mrelocation-model pic -pic-level 2 -pic-is-pie -mframe-pointer=none -fno-rounding-math -mconstructor-aliases -munwind-tables -target-cpu x86-64-v2 -mllvm -x86-branches-within-32B-boundaries -debug-info-kind=constructor -dwarf-version=5 -debugger-tuning=gdb -mllvm -crash-diagnostics-dir=clang-crashreports -ffunction-sections -fdata-sections -fcoverage-compilation-dir=. -sys-header-deps -D TOOLCHAIN_VERSION=dO8igrHyLDfgq8txd8Qx0mI9oJqFLswAMLcIBKLF6TYC -D _LIBCPP_DISABLE_VISIBILITY_ANNOTATIONS -D NDEBUG=1 -D _LIBCPP_ENABLE_THREAD_SAFETY_ANNOTATIONS=1 -D MAGMA_DEBUG_INTERNAL_USE_ONLY=0 -D MAGMA_ENABLE_TRACING -O3 -Wall -Wextra -Wnewline-eof -Wconversion -Wimplicit-fallthrough -Wno-unused-parameter -Wno-sign-conversion -Wno-c99-designator -Werror -Wno-error=deprecated-declarations -Wthread-safety -Wno-deprecated-copy -Wno-suggest-override -std=c++17 -fdeprecated-macro -fdebug-compilation-dir=. -ferror-limit 19 -fvisibility hidden -fvisibility-inlines-hidden -fsanitize=safe-stack -stack-protector 2 -ftrivial-auto-var-init=pattern -fno-rtti -fgnuc-version=4.2.1 -fno-legacy-pass-manager -fcolor-diagnostics -vectorize-loops -vectorize-slp -debug-info-kind=constructor -fmerge-functions -faddrsig -D__GCC_HAVE_DWARF2_CFI_ASM=1 -x c++ zircon_platform_buffer-84a54b.cpp 1. parser at end of file 2. Optimizer #0 0x092b3a5a llvm::sys::PrintStackTrace(llvm::raw_ostream&, int) /src/clang-llvm/llvm-project/llvm/lib/Support/Unix/Signals.inc:565:11 #1 0x092b3b6b PrintStackTraceSignalHandler(void*) /src/clang-llvm/llvm-project/llvm/lib/Support/Unix/Signals.inc:632:1 #2 0x092b2423 llvm::sys::RunSignalHandlers() /src/clang-llvm/llvm-project/llvm/lib/Support/Signals.cpp:97:5 #3 0x092b41a5 SignalHandler(int) /src/clang-llvm/llvm-project/llvm/lib/Support/Unix/Signals.inc:407:1 #4 0x7feff6aac140 __restore_rt (/lib/x86_64-linux-gnu/libpthread.so.0+0x14140) #5 0x7feff67b9ce1 raise (/lib/x86_64-linux-gnu/libc.so.6+0x3bce1) #6 0x7feff67a3537 abort (/lib/x86_64-linux-gnu/libc.so.6+0x25537) #7 0x091daff0 llvm::install_out_of_memory_new_handler() /src/clang-llvm/llvm-project/llvm/lib/Support/ErrorHandling.cpp:194:0 #8 0x09332127 llvm::FunctionComparator::cmpConstants(llvm::Constant const*, llvm::Constant const*) const /src/clang-llvm/llvm-project/llvm/lib/Transforms/Utils/FunctionComparator.cpp:402:1 #9 0x09331d87 llvm::FunctionComparator::cmpConstants(llvm::Constant const*, llvm::Constant const*) const /src/clang-llvm/llvm-project/llvm/lib/Transforms/Utils/FunctionComparator.cpp:357:15 #10 0x093322a7 llvm::FunctionComparator::cmpValues(llvm::Value const*, llvm::Value const*) const /src/clang-llvm/llvm-project/llvm/lib/Transforms/Utils/FunctionComparator.cpp:767:5 #11 0x09333da1 llvm::FunctionComparator::cmpBasicBlocks(llvm::BasicBlock const*, llvm::BasicBlock const*) const /src/clang-llvm/llvm-project/llvm/lib/Transforms/Utils/FunctionComparator.cpp:807:17 #12 0x09334565 llvm::FunctionComparator::compare() /src/clang-llvm/llvm-project/llvm/lib/Transforms/Utils/FunctionComparator.cpp:896:13 #13 0x08894caa (anonymous namespace)::MergeFunctions::FunctionNodeCmp::operator()((anonymous namespace)::FunctionNode const&, (anonymous namespace)::FunctionNode const&) const /src/clang-llvm/llvm-project/llvm/lib/Transforms/IPO/MergeFunctions.cpp:220:29 #14 0x088947ef std::__2::__tree_node_base*& std::__2::__tree<(anonymous namespace)::FunctionNode, (anonymous namespace)::MergeFunctions::FunctionNodeCmp, std::__2::allocator<(anonymous namespace)::FunctionNode> >::__find_equal<(anonymous namespace)::FunctionNode>(std::__2::__tree_end_node*>*&, (anonymous namespace)::FunctionNode const&) /usr/local/google/home/phosek/fuchsia/prebuilt/third_party/clang/linux-x64
[llvm-bugs] [Bug 51725] New: [AMDGPU][MC][GFX10] image_atomic_fcmpswap has incorrect dst size with dmask:0xf
https://bugs.llvm.org/show_bug.cgi?id=51725 Bug ID: 51725 Summary: [AMDGPU][MC][GFX10] image_atomic_fcmpswap has incorrect dst size with dmask:0xf Product: libraries Version: trunk Hardware: PC OS: Windows NT Status: NEW Severity: enhancement Priority: P Component: Backend: AMDGPU Assignee: unassignedb...@nondot.org Reporter: d-...@mail.ru CC: llvm-bugs@lists.llvm.org An example of failed test: image_atomic_fcmpswap v[1:4], v2, s[12:19] dmask:0xf dim:SQ_RSRC_IMG_1D unorm sp3 assembles this instruction fine and produces the correct code: 0x00,0x1f,0x74,0xf0,0x02,0x01,0x03,0x00 This code may be decoded by llvm AMDGPU disassembler but with incorrect result: image_atomic_fcmpswap v1, v2, s[12:19] dmask:0xf dim:SQ_RSRC_IMG_1D unorm -- You are receiving this mail because: You are on the CC list for the bug.___ llvm-bugs mailing list llvm-bugs@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-bugs
[llvm-bugs] [Bug 51706] clang-cl crashes when using _setbv without /arch:AVX
https://bugs.llvm.org/show_bug.cgi?id=51706 Nico Weber changed: What|Removed |Added Fixed By Commit(s)||https://reviews.llvm.org/rG ||3e89cc5cda22bdc35b1d0a21e6d ||8eae08a4f61dc Resolution|--- |FIXED Status|NEW |RESOLVED -- You are receiving this mail because: You are on the CC list for the bug.___ llvm-bugs mailing list llvm-bugs@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-bugs
[llvm-bugs] [Bug 51726] New: Building with gcc 10 fails with "unrecognized command-line option ‘-mno-omit-leaf-frame-pointer’"
https://bugs.llvm.org/show_bug.cgi?id=51726 Bug ID: 51726 Summary: Building with gcc 10 fails with "unrecognized command-line option ‘-mno-omit-leaf-frame-pointer’" Product: compiler-rt Version: unspecified Hardware: PC OS: Linux Status: NEW Severity: release blocker Priority: P Component: scudo Assignee: unassignedb...@nondot.org Reporter: sylves...@debian.org CC: llvm-bugs@lists.llvm.org Building 13.0.0 rc2 with g++ 10 fails with: cd "/<>/build-llvm/projects/compiler-rt/lib/scudo/standalone" && /usr/bin/g++-10 -D_FILE_OFFSET_BITS=64 -D_GNU_SOURCE -D_LARGEFILE_SOURCE -D__STDC_CONSTANT_MACROS -D__STDC_FORMAT_MACROS -D__STDC_LIMIT_MACROS -I"/<>/build-llvm/projects/compiler-rt/lib/scudo/standalone" -I"/<>/compiler-rt/lib/scudo/standalone" -I"/<>/build-llvm/include" -I"/<>/llvm/include" -I"/<>/compiler-rt/lib/scudo/standalone/../.." -I"/<>/compiler-rt/lib/scudo/standalone/include" -g -O2 -fdebug-prefix-map=/<>=. -fstack-protector-strong -Wformat -Werror=format-security -fPIC -fno-semantic-interposition -fvisibility-inlines-hidden -Werror=date-time -Wall -Wextra -Wno-unused-parameter -Wwrite-strings -Wcast-qual -Wno-missing-field-initializers -pedantic -Wno-long-long -Wimplicit-fallthrough -Wno-maybe-uninitialized -Wno-class-memaccess -Wno-redundant-move -Wno-pessimizing-move -Wno-noexcept-type -Wdelete-non-virtual-dtor -Wsuggest-override -Wno-comment -Wmisleading-indentation -ffunction-sections -fdata-sections -Wall -std=c++14 -Wno-unused-parameter -O2 -DNDEBUG -g1 -march=armv7-a -mfloat-abi=hard -fno-lto -Werror=conversion -Wall -g -nostdinc++ -fvisibility=hidden -fno-exceptions -Wno-pedantic -fno-lto -O3 -DGWP_ASAN_HOOKS -fno-omit-frame-pointer -mno-omit-leaf-frame-pointer -std=c++14 -o CMakeFiles/clang_rt.scudo_standalone-armhf.dir/checksum.cpp.o -c "/<>/compiler-rt/lib/scudo/standalone/checksum.cpp" g++-10: error: unrecognized command-line option ‘-mno-omit-leaf-frame-pointer’; did you mean ‘-fno-omit-frame-pointer’? make[5]: *** [projects/compiler-rt/lib/scudo/standalone/CMakeFiles/clang_rt.scudo_standalone-armhf.dir/build.make:85: projects/compiler-rt/lib/scudo/standalone/CMakeFiles/clang_rt.scudo_standalone-armhf.dir/checksum.cpp.o] Error 1 make[5]: Leaving directory '/<>/build-llvm' make[4]: *** [CMakeFiles/Makefile2:47492: projects/compiler-rt/lib/scudo/standalone/CMakeFiles/clang_rt.scudo_standalone-armhf.dir/all] Error 2 -- You are receiving this mail because: You are on the CC list for the bug.___ llvm-bugs mailing list llvm-bugs@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-bugs
[llvm-bugs] [Bug 51188] lib/clang/13.0.0/include/intrin.h is missing __readfsdword
https://bugs.llvm.org/show_bug.cgi?id=51188 Nico Weber changed: What|Removed |Added Status|NEW |RESOLVED Resolution|--- |FIXED Fixed By Commit(s)||https://reviews.llvm.org/rG ||e5438f386854136d848989315f5 ||3788808afa37a -- You are receiving this mail because: You are on the CC list for the bug.___ llvm-bugs mailing list llvm-bugs@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-bugs
[llvm-bugs] [Bug 51447] Cyclic symlink between libclang1-13 and libclang-13-dev
https://bugs.llvm.org/show_bug.cgi?id=51447 Sylvestre Ledru changed: What|Removed |Added Status|NEW |RESOLVED Assignee|unassignedb...@nondot.org |sylves...@debian.org CC||sylves...@debian.org Resolution|--- |FIXED --- Comment #1 from Sylvestre Ledru --- Yeah, this is fixed now, sorry about that :) -- You are receiving this mail because: You are on the CC list for the bug.___ llvm-bugs mailing list llvm-bugs@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-bugs
[llvm-bugs] [Bug 40750] Better optimizations for logical ops with bit fields
https://bugs.llvm.org/show_bug.cgi?id=40750 David Bolvansky changed: What|Removed |Added Status|NEW |RESOLVED Resolution|--- |FIXED --- Comment #2 from David Bolvansky --- Resolved in LLVM 10+. Testcase added in llvmorg-14-init-3006-g00f8aecf6ef4 -- You are receiving this mail because: You are on the CC list for the bug.___ llvm-bugs mailing list llvm-bugs@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-bugs
[llvm-bugs] [Bug 51292] New ExecutionEngine/OrcLazy tests segfault on 32-bit x86
https://bugs.llvm.org/show_bug.cgi?id=51292 Tom Stellard changed: What|Removed |Added Resolution|--- |FIXED Fixed By Commit(s)|c5ab55f5331c9da3c352b61d10d |c5ab55f5331c9da3c352b61d10d |2f8a470a08b5b |2f8a470a08b5b c8905f1bb304 Status|NEW |RESOLVED --- Comment #14 from Tom Stellard --- Merged: c8905f1bb304 -- You are receiving this mail because: You are on the CC list for the bug.___ llvm-bugs mailing list llvm-bugs@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-bugs
[llvm-bugs] [Bug 51236] [meta] 13.0.0 Release Blockers
https://bugs.llvm.org/show_bug.cgi?id=51236 Bug 51236 depends on bug 51292, which changed state. Bug 51292 Summary: New ExecutionEngine/OrcLazy tests segfault on 32-bit x86 https://bugs.llvm.org/show_bug.cgi?id=51292 What|Removed |Added Status|NEW |RESOLVED Resolution|--- |FIXED -- You are receiving this mail because: You are on the CC list for the bug.___ llvm-bugs mailing list llvm-bugs@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-bugs
[llvm-bugs] [Bug 51698] [Linker] Merge 92f54e1c752253b5a17eb9ec7942f580d878f64d into 13.0.0
https://bugs.llvm.org/show_bug.cgi?id=51698 Tom Stellard changed: What|Removed |Added Resolution|--- |FIXED Status|NEW |RESOLVED --- Comment #1 from Tom Stellard --- Merged: 7a4a5d413624bd8bd54c79e2519ebd80a9ef9acb -- You are receiving this mail because: You are on the CC list for the bug.___ llvm-bugs mailing list llvm-bugs@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-bugs
[llvm-bugs] [Bug 51236] [meta] 13.0.0 Release Blockers
https://bugs.llvm.org/show_bug.cgi?id=51236 Bug 51236 depends on bug 51698, which changed state. Bug 51698 Summary: [Linker] Merge 92f54e1c752253b5a17eb9ec7942f580d878f64d into 13.0.0 https://bugs.llvm.org/show_bug.cgi?id=51698 What|Removed |Added Status|NEW |RESOLVED Resolution|--- |FIXED -- You are receiving this mail because: You are on the CC list for the bug.___ llvm-bugs mailing list llvm-bugs@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-bugs
[llvm-bugs] [Bug 51555] WebAssembly .s format can omit .globaltype for __stack_pointer
https://bugs.llvm.org/show_bug.cgi?id=51555 Wouter van Oortmerssen changed: What|Removed |Added Status|CONFIRMED |RESOLVED Resolution|--- |FIXED --- Comment #3 from Wouter van Oortmerssen --- Thanks for reporting, and figuring out the problem! The point of MachineSymbolsUsed is that they are indeed all collected before emitExternalDecls ever gets called, but that wasn't happening, because the pass that does so, WebAssemblyMCLowerPrePass, was a FunctionPass which can get scheduled interleaved with the AsmPrinter. This patch changes it to a ModulePass to fix that: https://reviews.llvm.org/D109202 -- You are receiving this mail because: You are on the CC list for the bug.___ llvm-bugs mailing list llvm-bugs@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-bugs
[llvm-bugs] [Bug 51728] New: Ability to customize indentation of designated initializers
https://bugs.llvm.org/show_bug.cgi?id=51728
Bug ID: 51728
Summary: Ability to customize indentation of designated
initializers
Product: clang
Version: unspecified
Hardware: PC
OS: Linux
Status: NEW
Severity: enhancement
Priority: P
Component: Formatter
Assignee: unassignedclangb...@nondot.org
Reporter: vlov...@gmail.com
CC: djas...@google.com, kli...@google.com,
llvm-bugs@lists.llvm.org
When a designated initializer needs to spilled onto a new line, it uses
`ContinuationIndentWidth`. However in the KJ style guide
(https://github.com/capnproto/capnproto/blob/master/kjdoc/style-guide.md), such
a case should have `IndentWidth` applied.
Currently (undesirable):
IndentWidth: 2
ContinuationIndentWidth: 2
return SomeStruct {
.x = 5,
.y = 6,
.z = 7
};
Desired:
IndentWidth: 2
ContinuationIndentWidth: 2
return SomeStruct {
.x = 5,
.y = 6,
.z = 7
};
--
You are receiving this mail because:
You are on the CC list for the bug.___
llvm-bugs mailing list
llvm-bugs@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-bugs
[llvm-bugs] [Bug 51729] New: LLVM 13 Regression: __builtin_longjmp not working with -Oz
https://bugs.llvm.org/show_bug.cgi?id=51729
Bug ID: 51729
Summary: LLVM 13 Regression: __builtin_longjmp not working with
-Oz
Product: new-bugs
Version: trunk
Hardware: PC
OS: Linux
Status: NEW
Severity: enhancement
Priority: P
Component: new bugs
Assignee: unassignedb...@nondot.org
Reporter: xtk...@gmail.com
CC: htmldevelo...@gmail.com, llvm-bugs@lists.llvm.org
Created attachment 25224
--> https://bugs.llvm.org/attachment.cgi?id=25224&action=edit
Repro
Tested with 13.0.0-rc2 and 12.0.1. The repro could be made shorter.
Note that this issue is not necessarily specific to X86_64. A segfault is also
observed for ARM.
$ cat bsjlj.c
void foo(void **jb, int n);
void bar(int n);
void
bsj(void **jb, volatile int n)
{
if (__builtin_setjmp(jb) == 0) {
foo(jb, n);
} else {
bar(n);
}
}
void
blj(void **jb)
{
__builtin_longjmp(jb, 1);
}
int printf (const char *__restrict __format, ...);
void
foo(void **jb, int n)
{
printf("foo: n = %d\n", n);
blj(jb);
}
void
bar(int n)
{
printf("bar: n = %d\n", n);
}
int
main(int argc, char *argv[] __attribute__((unused)))
{
void *jb[5];
bsj(jb, argc);
bsj(jb, argc+1);
return 0;
}
$ /usr/lib/llvm/12/bin/clang --target=x86_64-linux-gnu bsjlj.c -Oz -Wall
-Wextra -Wundef
$ ./a.out
foo: n = 1
bar: n = 1
foo: n = 2
bar: n = 2
$ /usr/lib/llvm/13/bin/clang --target=x86_64-linux-gnu bsjlj.c -Oz -Wall
-Wextra -Wundef
$ ./a.out
foo: n = 1
bar: n = 32764
Segmentation fault
--
You are receiving this mail because:
You are on the CC list for the bug.___
llvm-bugs mailing list
llvm-bugs@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-bugs
[llvm-bugs] [Bug 51223] AVR driver can't find avr-libc in /opt/local/avr/
https://bugs.llvm.org/show_bug.cgi?id=51223 Ben Shi changed: What|Removed |Added Resolution|--- |FIXED CC||powerman...@163.com Status|NEW |RESOLVED --- Comment #1 from Ben Shi --- As discussed in https://reviews.llvm.org/D107682, this issue has been fixed in https://reviews.llvm.org/rGb31199bab4865deef4e778d7a028c8ec64285654 -- You are receiving this mail because: You are on the CC list for the bug.___ llvm-bugs mailing list llvm-bugs@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-bugs
