From 0cab012e6205c5f1ba9848b60d177e998c6babd8 Mon Sep 17 00:00:00 2001 From: Julian Seward Date: Sat, 28 Dec 2019 20:30:39 +0100 Subject: [PATCH] Complete initial triage pass. --- docs/internals/3_15_BUGSTATUS.txt | 208 ++++++++++++++++++++++++++++++-------- 1 file changed, 166 insertions(+), 42 deletions(-) diff --git a/docs/internals/3_15_BUGSTATUS.txt b/docs/internals/3_15_BUGSTATUS.txt index 4b9542065..c0be01f69 100644 --- a/docs/internals/3_15_BUGSTATUS.txt +++ b/docs/internals/3_15_BUGSTATUS.txt @@ -5,72 +5,196 @@ Created Tues 14 May 2019. This contains bugs reported only after the release of 3.15.0. It doesn't carry over bugs from earlier versions. +=== Debuginfo reader =================================================== +=== Documentation ====================================================== +=== GDB server ========================================================= +=== KernelInterface ==================================================== +=== KernelInterface/Linux ============================================== + 407376 Update Xen support to 4.12 and add more coverage -407377 SIGSEGV if debugged application going to VMW_PORT() macro -407589 Add support for C11 aligned_alloc() and GNU reallocarray()[valgrind] -407904 Inlined member operators lose class name in logs and generated supps -407933 VgTs_Runnable syscall mach:50 (lwpid 771) -408140 Missing support for vmovdqu64 -408758 memcheck internal error: Signal 10 (SIGBUS) + ** Has patch, looks reasonable + 408858 Add new io_uring_register, setup, enter syscalls -409141 Valgrind hangs when SIGKILLed + No patch, no test case + 409206 [PATCH] Support for Linux PPS and PTP ioctls -409391 [PATCH] LSE instruction support : LDADD 32bit -409429 False positives at unexpected location due to failure to recognize - cmpeq as a dependency breaking idiom (fixed in grail? check this) -409501 amd64->IR: unhandled instruction bytes -409646 1Uto32 can be identified -409678 improvement suggestion for dhat -409934 Callgrind: Warning: L3 cache found, using its data for the LL simulation -409999 Valgrind causes SIGILL due to unrecognized instruction in rocksdb - static initialization -410102 Valgrind ir sanity check failure crash + ** Has patches, looks reasonable + 410556 [PATCH] add support for BLKIO{MIN,OPT} and BLKALIGNOFF ioctls -410562 Unrecognised instruction 'UD2' + ** Has patches, looks reasonable + 410743 shmat() calls for 32-bit programs fail when running in 64-bit valgrind + Not sure if this is important. Ask MJW. + 410757 glibc wrapper discrepancy for preadv2/pwritev2 system calls across different versions -410764 BLENDVPD, BLENDVPS, PBLENDVB not implemented in guest_x86 + Not sure if this is important. Ask MJW. + 411100 Valgrind does not provide a solution in the error message when an mmap error is detected. -411203 valgrind should have a mean to pass environment variable to target - program -411303 Illegal hardware instruction when using memcheck on lepton -411451 x86/amd64->IR of bt/btc/bts/btr with immediate clears zero flag -412338 vex x86->IR: unhandled instruction bytes: 0xF 0x3F 0x7 0xB -412377 SIGILL on cache flushes on arm64 + ** Initial load failure for very large executables? + Has discussion links. + 412408 unhandled arm-linux syscall: 124 - adjtime - on arm-linux -412711 Patch: Add a per-thread simulate flag to Callgrind -412745 Support macOS Catalina (10.15) + * trivial patch, but need to check the handler is correct + 413119 ioctl wrapper for DRM_IOCTL_I915_GEM_MMAP -413251 Compilation error using GCC 7.4.0 & OpenMPI 4.0.2 + ** plausible; contains patches + +415621 epoll_ctl reports for uninitialized padding + * maybe an inaccurate wrapper; may be easy to fix? + +=== KernelInterface/OS X =============================================== + +407933 VgTs_Runnable syscall mach:50 (lwpid 771) + +408758 memcheck internal error: Signal 10 (SIGBUS) + Memcheck OSX failure + +410562 Unrecognised instruction 'UD2' + on OSX + +412745 Support macOS Catalina (10.15) + 413369 unhandled syscall getpgid + 413410 Mac Os 10.15 valgrind 3.15.0 (== 412745 ?) -414053 vex amd64->IR: unhandled instruction bytes + +=== MPI ================================================================ + +413251 Compilation error using GCC 7.4.0 & OpenMPI 4.0.2 + May not be hard to fix, but finding the bandwidth is difficult + +=== Output ============================================================= +=== Tools ============================================================== +=== Tools/Cachegrind =================================================== +=== Tools/Callgrind ==================================================== + +409934 Callgrind: Warning: L3 cache found, using its data for the LL simulation + * Misleading warning message? + +412711 Patch: Add a per-thread simulate flag to Callgrind + Uh. Is this a good idea, even? + +415293 Incorrect call-graph tracking due + to new _dl_runtime_resolve_xsave* functions + * Potentially important, and may be easy to fix? + +=== Tools/DHAT ========================================================= + +409678 DHAT: make it able to distinguish reads from writes + Has patches, but do we want to take them? + +=== Tools/DRD ========================================================== +=== Tools/Helgrind ===================================================== +=== Tools/Lackey ======================================================= +=== Tools/Massif ======================================================= +=== Tools/Memcheck ===================================================== + +407589 Add support for C11 aligned_alloc() and GNU reallocarray() + Missing allocation intercepts? + +409429 False positives at unexpected location due to failure to recognize + cmpeq as a dependency breaking idiom (fixed in grail? check this) + In grail: 96de5118f5332ae145912ebe91b8fa143df74b8d + +415141 Possible leak with calling __libc_freeres before + all thread's tid_addresses are cleared + Not really Memcheck specific. MJW knows more. + +=== Tools/SGCheck ====================================================== +=== Uncategorised ====================================================== +=== Uncategorised/build ================================================ + +415516 Can't cross compile on openwrt + MIPS build failure + +=== Uncategorised/run ================================================== + +411203 valgrind should have a mean to pass environment variable to target + program + * I think this is invalid, but maybe I misunderstand + +=== VEX ================================================================ +=== VEX/amd64 ========================================================== + +408140 Missing support for vmovdqu64 + is an AVX512 insn + +409999 vex amd64->IR: 0x62 0xD1 0xFE 0x8 0x6F 0x84 0x24 0x8 0x0 0x0 + *** 0x62 is BOUND, but that isn't valid in 64-bit mode + == 393351 + == 414944 + == 411303 + == 414053 + 393351 has STR + +411451 x86/amd64->IR of bt/btc/bts/btr with immediate clears zero flag + * has patch and nano-test-case + +=== VEX/arm32 ========================================================== + +410102 Valgrind ir sanity check failure crash + But is on 3.13 at latest. No small test case. + +=== VEX/arm64 ========================================================== + +409391 [PATCH] LSE instruction support : LDADD 32bit + arm64 something. AH can maybe deal with it? + +409646 No case for 1Uto32 - arm64 + For 'taintgrind'. Should be trivial to fix. + +412377 SIGILL on cache flushes on arm64 + * Needs looking at. Not sure if this is valid or not. + +414270 Collection of bugs for new ARMv8.1 work and features 414268 Enable AArch64 feature detection and decoding for v8.x instructions (where x>0) -414270 Collection of bugs for new ARMv8.1 work and features +415136 ARMv8.1 Compare-and-Swap instructions are not supported + All of this 8.x stuff is WIP + +=== VEX/mips =========================================================== +=== VEX/other ========================================================== +=== VEX/ppc ============================================================ +=== VEX/s390x ========================================================== +=== VEX/x86 ============================================================ +=== zz_other =========================================================== + 414278 VG_(memcpy) used for overlapping moves + * Comment-only change + 414290 Release notes are corrupted on valgrind.org -414870 std::frexp(long double) broken under valgrind. -414944 vex amd64->IR: 0x62 0xF1 0x7D 0x48 0xEF 0xC0 0x48 0x8D 0x7D 0xD0 - (== the mysterious 393351) -415136 ARMv8.1 Compare-and-Swap instructions are not supported -415141 Possible leak with calling __libc_freeres before - all thread's tid_addresses are cleared -415293 Incorrect call-graph tracking due - to new _dl_runtime_resolve_xsave* functions 415393 Valgrind webpages and non-printable characters (== 414290) -415516 Can't cross compile on openwrt -415621 epoll_ctl reports for uninitialized padding + ** + +=== zz_other/Android =================================================== +=== zz_other/amd64 ===================================================== + +407377 SIGSEGV if debugged application going to VMW_PORT() macro + Happens when using VMWare trapdoor? + +=== zz_other/arm32 ===================================================== +=== zz_other/arm64 ===================================================== +=== zz_other/*BSD ====================================================== +=== zz_other/mips ====================================================== +=== zz_other/ppc ======================================================= +=== zz_other/s390 ====================================================== +=== zz_other/Win32 ===================================================== +=== zz_other/Win64 ===================================================== + +253657 [Contains patches to improve PDB reading] + Improve PDB* reading ? + + +=== zz_other/x86 ======================================================= + -- Sat 28 Dec 10:39:06 CET 2019 Extras (not new) 319393 bad rounding in cvtsi2ss instruction -Improve PDB* reading ? I thought I saw some patches for this .. is it 253657 ? - apply included fixes for Xen 4.6/4.7/4.8/4.9/4.10/4.11/4.12 390553 ? Can we get rid of exp-sgcheck now? -- 2.11.4.GIT