View file File name : dist.news.html Content :<html> <head> <meta http-equiv="Content-Type" content="text/html; charset=UTF-8"> <title>2. NEWS</title> <link rel="stylesheet" type="text/css" href="vg_basic.css"> <meta name="generator" content="DocBook XSL Stylesheets Vsnapshot"> <link rel="home" href="index.html" title="Valgrind Documentation"> <link rel="up" href="dist.html" title="Valgrind Distribution Documents"> <link rel="prev" href="dist.authors.html" title="1. AUTHORS"> <link rel="next" href="dist.news.old.html" title="3. OLDER NEWS"> </head> <body bgcolor="white" text="black" link="#0000FF" vlink="#840084" alink="#0000FF"> <div><table class="nav" width="100%" cellspacing="3" cellpadding="3" border="0" summary="Navigation header"><tr> <td width="22px" align="center" valign="middle"><a accesskey="p" href="dist.authors.html"><img src="images/prev.png" width="18" height="21" border="0" alt="Prev"></a></td> <td width="25px" align="center" valign="middle"><a accesskey="u" href="dist.html"><img src="images/up.png" width="21" height="18" border="0" alt="Up"></a></td> <td width="31px" align="center" valign="middle"><a accesskey="h" href="index.html"><img src="images/home.png" width="27" height="20" border="0" alt="Up"></a></td> <th align="center" valign="middle">Valgrind Distribution Documents</th> <td width="22px" align="center" valign="middle"><a accesskey="n" href="dist.news.old.html"><img src="images/next.png" width="18" height="21" border="0" alt="Next"></a></td> </tr></table></div> <div class="chapter"> <div class="titlepage"><div><div><h1 class="title"> <a name="dist.news"></a>2. NEWS</h1></div></div></div> <div class="literallayout"><p><br> Release 3.22.0 (31 Oct 2023)<br> ~~~~~~~~~~~~~~~~~~~~~~~~~~~~<br> <br> This release supports X86/Linux, AMD64/Linux, ARM32/Linux, ARM64/Linux,<br> PPC32/Linux, PPC64BE/Linux, PPC64LE/Linux, S390X/Linux, MIPS32/Linux,<br> MIPS64/Linux, ARM/Android, ARM64/Android, MIPS32/Android, X86/Android,<br> X86/Solaris, AMD64/Solaris, AMD64/MacOSX 10.12, X86/FreeBSD and<br> AMD64/FreeBSD. There is also preliminary support for X86/macOS 10.13,<br> AMD64/macOS 10.13 and nanoMIPS/Linux.<br> <br> * ==================== CORE CHANGES ===================<br> <br> * A new configure option --with-gdbscripts-dir lets you install<br> the gdb valgrind python monitor scripts in a specific location.<br> For example a distro could use it to install the scripts in a<br> safe load location --with-gdbscripts-dir=%{_datadir}/gdb/auto-load<br> It is also possible to configure --without-gdb-scripts-dir so no<br> .debug_gdb_scripts section is added to the vgpreload library and<br> no valgrind-monitor python scripts are installed at all.<br> <br> * ================== PLATFORM CHANGES =================<br> <br> * Support has been added for FreeBSD 14 and FreeBSD 15.<br> * Add support for the folllowing FreeBSD system calls:<br> close_range, kqueuex, membarrier, timerfd_create,<br> timerfd_settime and timerfd_gettime (all added in FreeBSD 15).<br> <br> * ==================== TOOL CHANGES ===================<br> <br> * Memcheck now tests and warns about the values used for<br> alignment and size. These apply to various functions: memalign,<br> posix_memalign and aligned_alloc in C and various overloads<br> of operators new and delete in C++. The kinds of error that can<br> be detected are<br> - invalid alignment, for instance the alignment is usually required<br> to be a power of 2<br> - mismatched alignment between aligned allocation and aligned<br> deallocation<br> - mismatched size when sized delete is used<br> - bad size for functions that have implementation defined behaviour<br> when the requested size is zero<br> <br> * Cachegrind:<br> - You can now profile part of a program's execution using the new<br> `CACHEGRIND_START_INSTRUMENTATION` and `CACHEGRIND_STOP_INSTRUMENTATION`<br> client requests, along with the new `--instr-at-start` option. The<br> behaviour is the same as Callgrind's equivalent functionality.<br> <br> * ==================== FIXED BUGS ====================<br> <br> The following bugs have been fixed or resolved. Note that "n-i-bz"<br> stands for "not in bugzilla" -- that is, a bug that was reported to us<br> but never got a bugzilla entry. We encourage you to file bugs in<br> bugzilla (https://bugs.kde.org/enter_bug.cgi?product=valgrind) rather<br> than mailing the developers (or mailing lists) directly -- bugs that<br> are not entered into bugzilla tend to get forgotten about or ignored.<br> <br> 390871 ELF debug info reader confused with multiple .rodata* sections<br> 417993 vbit-test fail on s390x with Iop_Add32: spurious dependency on uninit<br> 426751 Valgrind reports "still reachable" memory using musl<br> (alpine running inside docker)<br> 432801 Valgrind 3.16.1 reports a jump based on uninitialized memory somehow<br> related to clang and signals<br> 433857 Add validation to C++17 aligned new/delete alignment size<br> 433859 Add mismatched detection to C++ 17 aligned new/delete<br> 460192 Add epoll_pwait2<br> 461074 DWARF2 CFI reader: unhandled DW_OP_ 0x11 (consts) DW_OP_ 0x92 (bregx)<br> 465782 s390x: Valgrind doesn't compile with Clang on s390x<br> 466105 aligned_alloc problems, part 2<br> 467441 Add mismatched detection to C++ 14 sized delete<br> 469049 link failure on ppc64 (big endian) valgrind 3.20<br> 469146 massif --ignore-fn does not ignore inlined functions<br> 469768 Make it possible to install gdb scripts in a different location<br> 470121 Can't run callgrind_control with valgrind 3.21.0 because of perl errors<br> 470132 s390x: Assertion failure on VGM instruction<br> 470520 Multiple realloc zero errors crash in MC_(eq_Error)<br> 470713 Failure on the Yosys project: valgrind: m_libcfile.c:1802<br> (Bool vgPlain_realpath(const HChar *, HChar *)):<br> Assertion 'resolved' failed<br> 470830 Don't print actions vgdb me ... continue for vgdb --multi mode<br> 470978 s390x: Valgrind cannot start qemu-kvm when "sysctl vm.allocate_pgste=0"<br> 471311 gdb --multi mode stdout redirecting to stderr<br> 471807 Add support for lazy reading and downloading of DWARF debuginfo<br> 472219 Syscall param ppoll(ufds.events) points to uninitialised byte(s)<br> 472875 none/tests/s390x/dfp-1 failure <br> 472963 Broken regular expression in configure.ac<br> 473604 Fix bug472219.c compile failure with Clang 16<br> 473677 make check compile failure with Clang 16 based on GCC 13.x<br> 473745 must-be-redirected function - strlen<br> 473870 FreeBSD 14 applications fail early at startup<br> 473944 Handle mold linker split RW PT_LOAD segments correctly<br> 474332 aligned_alloc under Valgrind returns nullptr when alignment is not a multiple of sizeof(void *)<br> 475650 DRD does not work with C11 threads<br> 475652 Missing suppression for __wcsncpy_avx2 (strncpy-avx2.S:308)?<br> 476108 vg_replace_malloc DELETE checks size<br> n-i-bz Allow arguments with spaces in .valgrindrc files<br> n-i-bz FreeBSD fixed reading of Valgrind tools own debuginfo<br> <br> To see details of a given bug, visit<br> https://bugs.kde.org/show_bug.cgi?id=XXXXXX<br> where XXXXXX is the bug number as listed above.<br> <br> (3.22.0.RC1: 17 Oct 2023)<br> (3.22.0.RC2: 26 Oct 2023)<br> <br> Release 3.21.0 (28 Apr 2023)<br> ~~~~~~~~~~~~~~~~~~~~~~~~~~~~<br> <br> This release supports X86/Linux, AMD64/Linux, ARM32/Linux, ARM64/Linux,<br> PPC32/Linux, PPC64BE/Linux, PPC64LE/Linux, S390X/Linux, MIPS32/Linux,<br> MIPS64/Linux, ARM/Android, ARM64/Android, MIPS32/Android, X86/Android,<br> X86/Solaris, AMD64/Solaris, AMD64/MacOSX 10.12, X86/FreeBSD and<br> AMD64/FreeBSD. There is also preliminary support for X86/macOS 10.13,<br> AMD64/macOS 10.13 and nanoMIPS/Linux.<br> <br> * ==================== CORE CHANGES ===================<br> <br> * When GDB is used to debug a program running under valgrind using<br> the valgrind gdbserver, GDB will automatically load some<br> python code provided in valgrind defining GDB front end commands<br> corresponding to the valgrind monitor commands.<br> These GDB front end commands accept the same format as<br> the monitor commands directly sent to the Valgrind gdbserver.<br> These GDB front end commands provide a better integration<br> in the GDB command line interface, so as to use for example<br> GDB auto-completion, command specific help, searching for<br> a command or command help matching a regexp, ...<br> For relevant monitor commands, GDB will evaluate arguments<br> to make the use of monitor commands easier.<br> For example, instead of having to print the address of a variable<br> to pass it to a subsequent monitor command, the GDB front end<br> command will evaluate the address argument. It is for example<br> possible to do:<br> (gdb) memcheck who_points_at &some_struct sizeof(some_struct)<br> instead of:<br> (gdb) p &some_struct<br> $2 = (some_struct_type *) 0x1130a0 <some_struct><br> (gdb) p sizeof(some_struct)<br> $3 = 40<br> (gdb) monitor who_point_at 0x1130a0 40<br> <br> * The vgdb utility now supports extended-remote protocol when<br> invoked with --multi. In this mode the GDB run command is<br> supported. Which means you don't need to run gdb and valgrind<br> from different terminals. So for example to start your program<br> in gdb and run it under valgrind you can do:<br> $ gdb prog<br> (gdb) set remote exec-file prog<br> (gdb) set sysroot /<br> (gdb) target extended-remote | vgdb --multi<br> (gdb) start<br> <br> * The behaviour of realloc with a size of zero can now<br> be changed for tools that intercept malloc. Those<br> tools are memcheck, helgrind, drd, massif and dhat.<br> Realloc implementations generally do one of two things<br> - free the memory like free() and return NULL<br> (GNU libc and ptmalloc).<br> - either free the memory and then allocate a<br> minimum sized block or just return the<br> original pointer. Return NULL if the<br> allocation of the minimum sized block fails<br> (jemalloc, musl, snmalloc, Solaris, macOS).<br> When Valgrind is configured and built it will<br> try to match the OS and libc behaviour. However<br> if you are using a non-default library to replace<br> malloc and family (e.g., musl on a glibc Linux or<br> tcmalloc on FreeBSD) then you can use a command line<br> option to change the behaviour of Valgrind:<br> --realloc-zero-bytes-frees=yes|no [yes on Linux glibc, no otherwise]<br> <br> * ================== PLATFORM CHANGES =================<br> <br> * Make the address space limit on FreeBSD amd64 128Gbytes<br> (the same as Linux and Solaris, it was 32Gbytes)<br> <br> * ==================== TOOL CHANGES ===================<br> <br> * Memcheck:<br> - When doing a delta leak_search, it is now possible to only<br> output the new loss records compared to the previous leak search.<br> This is available in the memcheck monitor command 'leak_search'<br> by specifying the "new" keyword or in your program by using<br> the client request VALGRIND_DO_NEW_LEAK_CHECK.<br> Whenever a "delta" leak search is done (i.e. when specifying<br> "new" or "increased" or "changed" in the monitor command),<br> the new loss records have a "new" marker.<br> - Valgrind now contains python code that defines GDB memcheck<br> front end monitor commands. See CORE CHANGES.<br> - Performs checks for the use of realloc with a size of zero.<br> This is non-portable and a source of errors. If memcheck<br> detects such a usage it will generate an error<br> realloc() with size 0<br> followed by the usual callstacks.<br> A switch has been added to allow this to be turned off:<br> --show-realloc-size-zero=yes|no [yes]<br> <br> * Helgrind:<br> - The option ---history-backtrace-size=<number> allows to configure<br> the number of entries to record in the stack traces of "old"<br> accesses. Previously, this number was hardcoded to 8.<br> - Valgrind now contains python code that defines GDB helgrind<br> front end monitor commands. See CORE CHANGES.<br> <br> * Cachegrind:<br> - `--cache-sim=no` is now the default. The cache simulation is old and<br> unlikely to match any real modern machine. This means only the `Ir`<br> event are gathered by default, but that is by far the most useful<br> event.<br> - `cg_annotate`, `cg_diff`, and `cg_merge` have been rewritten in<br> Python. As a result, they all have more flexible command line<br> argument handling, e.g. supporting `--show-percs` and<br> `--no-show-percs` forms as well as the existing `--show-percs=yes`<br> and `--show-percs=no`.<br> - `cg_annotate` has some functional changes.<br> - It's much faster, e.g. 3-4x on common cases.<br> - It now supports diffing (with `--diff`, `--mod-filename`, and<br> `--mod-funcname`) and merging (by passing multiple data files).<br> - It now provides more information at the file and function level.<br> There are now "File:function" and "Function:file" sections. These<br> are very useful for programs that use inlining a lot.<br> - Support for user-annotated files and the `-I`/`--include` option<br> has been removed, because it was of little use and blocked other<br> improvements.<br> - The `--auto` option is renamed `--annotate`, though the old<br> `--auto=yes`/`--auto=no` forms are still supported.<br> - `cg_diff` and `cg_merge` are now deprecated, because `cg_annotate`<br> now does a better job of diffing and merging.<br> - The Cachegrind output file format has changed very slightly, but in<br> ways nobody is likely to notice.<br> <br> * Callgrind:<br> - Valgrind now contains python code that defines GDB callgrind<br> front end monitor commands. See CORE CHANGES.<br> <br> * Massif:<br> - Valgrind now contains python code that defines GDB massif<br> front end monitor commands. See CORE CHANGES.<br> <br> * DHAT:<br> - A new kind of user request has been added which allows you to<br> override the 1024 byte limit on access count histograms for blocks<br> of memory. The client request is DHAT_HISTOGRAM_MEMORY.<br> <br> * ==================== FIXED BUGS ====================<br> <br> The following bugs have been fixed or resolved. Note that "n-i-bz"<br> stands for "not in bugzilla" -- that is, a bug that was reported to us<br> but never got a bugzilla entry. We encourage you to file bugs in<br> bugzilla (https://bugs.kde.org/enter_bug.cgi?product=valgrind) rather<br> than mailing the developers (or mailing lists) directly -- bugs that<br> are not entered into bugzilla tend to get forgotten about or ignored.<br> <br> 170510 Don't warn about ioctl of size 0 without direction hint<br> 241072 List tools in --help output<br> 327548 false positive while destroying mutex<br> 382034 Testcases build fixes for musl<br> 351857 confusing error message about valid command line option<br> 374596 inconsistent RDTSCP support on x86_64<br> 392331 Spurious lock not held error from inside pthread_cond_timedwait<br> 397083 Likely false positive "uninitialised value(s)" for __wmemchr_avx2 and __wmemcmp_avx2_movbe<br> 400793 pthread_rwlock_timedwrlock false positive<br> 419054 Unhandled syscall getcpu on arm32<br> 433873 openat2 syscall unimplemented on Linux<br> 434057 Add stdio mode to valgrind's gdbserver<br> 435441 valgrind fails to interpose malloc on musl 1.2.2 due to weak symbol name and no libc soname<br> 436413 Warn about realloc of size zero<br> 439685 compiler warning in callgrind/main.c<br> 444110 priv/guest_ppc_toIR.c:36198:31: warning: duplicated 'if' condition.<br> 444487 hginfo test detects an extra lock inside data symbol "_rtld_local"<br> 444488 Use glibc.pthread.stack_cache_size tunable<br> 444568 drd/tests/pth_barrier_thr_cr fails on Fedora 38<br> 445743 "The impossible happened: mutex is locked simultaneously by two threads"<br> while using mutexes with priority inheritance and signals<br> 449309 Missing loopback device ioctl(s) <br> 459476 vgdb: allow address reuse to avoid "address already in use" errorsuse" errors<br> 460356 s390: Sqrt32Fx4 -- cannot reduce tree<br> 462830 WARNING: unhandled amd64-freebsd syscall: 474<br> 463027 broken check for MPX instruction support in assembler<br> 464103 Enhancement: add a client request to DHAT to mark memory to be histogrammed<br> 464476 Firefox fails to start under Valgrind<br> 464609 Valgrind memcheck should support Linux pidfd_open<br> 464680 Show issues caused by memory policies like selinux deny_execmem<br> 464859 Build failures with GCC-13 (drd tsan_unittest)<br> 464969 D language demangling<br> 465435 m_libcfile.c:66 (vgPlain_safe_fd): Assertion 'newfd >= VG_(fd_hard_limit)' failed.<br> 466104 aligned_alloc problems, part 1<br> 467036 Add time cost statistics for Regtest<br> 467482 Build failure on aarch64 Alpine<br> 467714 fdleak_* and rlimit tests fail when parent process has more than<br> 64 descriptors opened<br> 467839 Gdbserver: Improve compatibility of library directory name<br> 468401 [PATCH] Add a style file for clang-format<br> 468556 Build failure for vgdb<br> 468606 build: remove "Valgrind relies on GCC" check/output<br> 469097 ppc64(be) doesn't support SCV syscall instruction<br> n-i-bz FreeBSD rfork syscall fail with EINVAL or ENOSYS rather than VG_(unimplemented)<br> <br> To see details of a given bug, visit<br> https://bugs.kde.org/show_bug.cgi?id=XXXXXX<br> where XXXXXX is the bug number as listed above.<br> <br> * ==================== KNOWN ISSUES ===================<br> <br> * configure --enable-lto=yes is know to not work in all setups.<br> See bug 469049. Workaround: Build without LTO.<br> <br> (3.21.0.RC1: 14 Apr 2023)<br> (3.21.0.RC2: 21 Apr 2023)<br> <br> Release 3.20.0 (24 Oct 2022)<br> ~~~~~~~~~~~~~~~~~~~~~~~~~~~~<br> <br> This release supports X86/Linux, AMD64/Linux, ARM32/Linux, ARM64/Linux,<br> PPC32/Linux, PPC64BE/Linux, PPC64LE/Linux, S390X/Linux, MIPS32/Linux,<br> MIPS64/Linux, ARM/Android, ARM64/Android, MIPS32/Android, X86/Android,<br> X86/Solaris, AMD64/Solaris, AMD64/MacOSX 10.12, X86/FreeBSD and<br> AMD64/FreeBSD. There is also preliminary support for X86/macOS 10.13,<br> AMD64/macOS 10.13 and nanoMIPS/Linux.<br> <br> * ==================== CORE CHANGES ===================<br> <br> * The option "--vgdb-stop-at=event1,event2,..." accepts the new value abexit.<br> This indicates to invoke gdbserver when your program exits abnormally<br> (i.e. with a non zero exit code).<br> * Fix Rust v0 name demangling.<br> * The Linux rseq syscall is now implemented as (silently) returning ENOSYS.<br> * Add FreeBSD syscall wrappers for __specialfd and __realpathat.<br> * Remove FreeBSD dependencies on COMPAT10, which fixes compatibility with<br> HardenedBSD<br> * The option --enable-debuginfod=<no|yes> [default: yes] has been added on<br> Linux.<br> * More DWARF5 support as generated by clang14.<br> <br> * ==================== FIXED BUGS ====================<br> <br> The following bugs have been fixed or resolved. Note that "n-i-bz"<br> stands for "not in bugzilla" -- that is, a bug that was reported to us<br> but never got a bugzilla entry. We encourage you to file bugs in<br> bugzilla (https://bugs.kde.org/enter_bug.cgi?product=valgrind) rather<br> than mailing the developers (or mailing lists) directly -- bugs that<br> are not entered into bugzilla tend to get forgotten about or ignored.<br> <br> 131186 writev reports error in (vector[...])<br> 434764 iconv_open causes ld.so v2.28+ to use optimised strncmp<br> 446754 Improve error codes from alloc functions under memcheck<br> 452274 memcheck crashes with Assertion 'sci->status.what == SsIdle' failed<br> 452779 Valgrind fails to build on FreeBSD 13.0 with llvm-devel (15.0.0)<br> 453055 shared_timed_mutex drd test fails with "Lock shared failed" message<br> 453602 Missing command line option to enable/disable debuginfod<br> 452802 Handle lld 9+ split RW PT_LOAD segments correctly<br> 454040 s390x: False-positive memcheck:cond in memmem on arch13 systems <br> 456171 [PATCH] FreeBSD: Don't record address errors when accessing the 'kern.ps_strings' sysctl struct<br> n-i-bz Implement vgdb invoker on FreeBSD<br> 458845 PowerPC: The L field for the dcbf and sync instruction should be<br> 3 bits in ISA 3.1.<br> 458915 Remove register cache to fix 458915 gdbserver causes wrong syscall return<br> 459031 Documentation on --error-exitcode incomplete<br> 459477 XERROR messages lacks ending '\n' in vgdb<br> 462007 Implicit int in none/tests/faultstatus.c<br> <br> To see details of a given bug, visit<br> https://bugs.kde.org/show_bug.cgi?id=XXXXXX<br> where XXXXXX is the bug number as listed above.<br> <br> (3.20.0.RC1: 20 Oct 2022)<br> <br> <br> Release 3.19.0 (11 Apr 2022)<br> ~~~~~~~~~~~~~~~~~~~~~~~~~~~~<br> <br> This release supports X86/Linux, AMD64/Linux, ARM32/Linux, ARM64/Linux,<br> PPC32/Linux, PPC64BE/Linux, PPC64LE/Linux, S390X/Linux, MIPS32/Linux,<br> MIPS64/Linux, ARM/Android, ARM64/Android, MIPS32/Android, X86/Android,<br> X86/Solaris, AMD64/Solaris, AMD64/MacOSX 10.12, X86/FreeBSD and<br> AMD64/FreeBSD. There is also preliminary support for X86/macOS 10.13,<br> AMD64/macOS 10.13 and nanoMIPS/Linux.<br> <br> * ==================== CORE CHANGES ===================<br> <br> * Fix Rust v0 name demangling.<br> * The Linux rseq syscall is now implemented as (silently) returning ENOSYS.<br> * Add FreeBSD syscall wrappers for __specialfd and __realpathat.<br> * Remove FreeBSD dependencies on COMPAT10, which fixes compatibility with HardenedBSD<br> <br> * ================== PLATFORM CHANGES =================<br> <br> * arm64:<br> - ignore the "v8.x" architecture levels, only look at actual CPU features<br> present. Fixes mismatch detected between RDMA and atomics features<br> preventing startup on some QEMU configurations.<br> - Implement LD{,A}XP and ST{,L}XP<br> - Fix incorrect code emitted for doubleword CAS.<br> <br> * s390:<br> - Fix sys_ipc semtimedop syscall<br> - Fix VFLRX and WFLRX instructions<br> - Fix EXRL instruction with negative offset<br> <br> * ppc64:<br> - Reimplement the vbpermq instruction support to generate less Iops and<br> avoid overflowing internal buffers.<br> - Fix checking for scv support to avoid "Facility 'SCV' unavailable (12),<br> exception" messages in dmsg.<br> - Fix setting condition code for Vector Compare quad word instructions.<br> - Fix fix lxsibzx, lxsihzx and lxsihzx instructions so they only load<br> their respective sized data.<br> - Fix the prefixed stq instruction in PC relative mode.<br> <br> * ==================== TOOL CHANGES ===================<br> <br> * Memcheck:<br> - Speed up --track-origins=yes for large (in the range of hundreds to<br> thousands of megabytes) mmap/munmaps.<br> * DRD/Helgrind:<br> - Several fixes for new versions of libstd++ using new posix try_lock<br> functions<br> <br> * ==================== FIXED BUGS ====================<br> <br> The following bugs have been fixed or resolved. Note that "n-i-bz"<br> stands for "not in bugzilla" -- that is, a bug that was reported to us<br> but never got a bugzilla entry. We encourage you to file bugs in<br> bugzilla (https://bugs.kde.org/enter_bug.cgi?product=valgrind) rather<br> than mailing the developers (or mailing lists) directly -- bugs that<br> are not entered into bugzilla tend to get forgotten about or ignored.<br> <br> 403802 leak_cpp_interior fails with some reachable blocks different than expected<br> 435732 memcheck/tests/leak_cpp_interior fails with gcc11<br> 444242 s390x: Valgrind crashes on EXRL with negative offset<br> 444399 arm64: unhandled instruction 0xC87F2D89 (LD{,A}XP and ST{,L}XP).<br> == 434283<br> 444481 gdb_server test failures on s390x<br> 444495 dhat/tests/copy fails on s390x<br> 444552 memcheck/tests/sem fails on s390x with glibc 2.34<br> 444571 PPC, fix the lxsibzx and lxsihzx so they only load their respective<br> sized data.<br> 444836 PPC, pstq instruction for R=1 is not storing to the correct address.<br> 444925 fexecve syscall wrapper not properly implemented<br> 445032 valgrind/memcheck crash with SIGSEGV when SIGVTALRM timer used and<br> libthr.so associated<br> 445211 Fix out of tree builds<br> 445300 [PATCH] Fix building tests with Musl<br> 445011 SIGCHLD is sent when valgrind uses debuginfod-find<br> 445354 arm64 backend: incorrect code emitted for doubleword CAS<br> 445415 arm64 front end: alignment checks missing for atomic instructions<br> 445504 Using C++ condition_variable results in bogus "mutex is locked simultaneously by two threads" warning<br> 445607 Unhandled amd64-freebsd syscall: 247<br> 445668 Inline stack frame generation is broken for Rust binaries<br> 445916 Demangle Rust v0 symbols with .llvm suffix<br> 446139 DRD/Helgrind with std::shared_timed_mutex::try_lock_until and try_lock_shared_until false positives<br> 446138 DRD/Helgrind with std::timed_mutex::try_lock_until false positives<br> 446281 Add a DRD suppression for fwrite<br> 446103 Memcheck: `--track-origins=yes` causes extreme slowdowns for large mmap/munmap<br> 446139 DRD/Helgrind with std::shared_timed_mutex::try_lock_until and try_lock_shared_until false<br> 446251 TARGET_SIGNAL_THR added to enum target_signal<br> 446823 FreeBSD - missing syscalls when using libzm4<br> 447991 s390x: Valgrind indicates illegal instruction on wflrx<br> 447995 Valgrind segfault on power10 due to hwcap checking code<br> 449483 Powerpc: vcmpgtsq., vcmpgtuq,, vcmpequq. instructions not setting the<br> condition code correctly.<br> 449672 ppc64 --track-origins=yes failures because of bad cmov addHRegUse<br> 449838 sigsegv liburing the 'impossible' happened for io_uring_setup<br> 450025 Powerc: ACC file not implemented as a logical overlay of the VSR<br> registers.<br> 450437 Warn for execve syscall with argv or argv[0] being NULL<br> 450536 Powerpc: valgrind throws 'facility scv unavailable exception'<br> 451626 Syscall param bpf(attr->raw_tracepoint.name) points to unaddressable byte(s)<br> 451827 [ppc64le] VEX temporary storage exhausted with several vbpermq instructions<br> 451843 valgrind fails to start on a FreeBSD system which enforces W^X<br> <br> To see details of a given bug, visit<br> https://bugs.kde.org/show_bug.cgi?id=XXXXXX<br> where XXXXXX is the bug number as listed above.<br> <br> (3.19.0.RC1: 02 Apr 2022)<br> (3.19.0.RC2: 08 Apr 2022)<br> <br> <br> Release 3.18.0 (15 Oct 2021)<br> ~~~~~~~~~~~~~~~~~~~~~~~~~~~~<br> <br> This release supports X86/Linux, AMD64/Linux, ARM32/Linux, ARM64/Linux,<br> PPC32/Linux, PPC64BE/Linux, PPC64LE/Linux, S390X/Linux, MIPS32/Linux,<br> MIPS64/Linux, ARM/Android, ARM64/Android, MIPS32/Android, X86/Android,<br> X86/Solaris, AMD64/Solaris, AMD64/MacOSX 10.12, X86/FreeBSD and<br> AMD64/FreeBSD. There is also preliminary support for X86/macOS 10.13,<br> AMD64/macOS 10.13 and nanoMIPS/Linux.<br> <br> * ==================== CORE CHANGES ===================<br> <br> * The libiberty demangler has been updated, which brings support for<br> Rust v0 name demangling. [Update: alas, due to a bug, this support<br> isn't working in 3.18.0.]<br> <br> * __libc_freeres isn't called anymore after the program recieves a<br> fatal signal. Causing some internal glibc resources to hang around,<br> but preventing any crashes after the program has ended.<br> <br> * The DWARF reader is now very much faster at startup when just<br> --read-inline-info=yes (the default in most cases) is given.<br> <br> * glibc 2.34, which moved various functions from libpthread.so into<br> libc.so, is now supported.<br> <br> * ================== PLATFORM CHANGES =================<br> <br> * arm64:<br> <br> - v8.2 scalar and vector FABD, FACGE, FACGT and FADD.<br> - v8.2 FP compare & conditional compare instructions.<br> - Zero variants of v8.2 FP compare instructions.<br> <br> * s390:<br> <br> - Support the miscellaneous-instruction-extensions facility 3 and<br> the vector-enhancements facility 2. This enables programs<br> compiled with "-march=arch13" or "-march=z15" to be executed<br> under Valgrind.<br> <br> * ppc64:<br> <br> - ISA 3.1 support is now complete<br> - ISA 3.0 support for the darn instruction added.<br> - ISA 3.0 support for the vector system call instruction scv added.<br> - ISA 3.0 support for the copy, paste and cpabort instructions added.<br> <br> * Support for X86/FreeBSD and AMD64/FreeBSD has been added.<br> <br> * ==================== OTHER CHANGES ====================<br> <br> * Memcheck on amd64: minor fixes to remove some false positive<br> undef-value errors<br> <br> * ==================== FIXED BUGS ====================<br> <br> The following bugs have been fixed or resolved. Note that "n-i-bz"<br> stands for "not in bugzilla" -- that is, a bug that was reported to us<br> but never got a bugzilla entry. We encourage you to file bugs in<br> bugzilla (https://bugs.kde.org/enter_bug.cgi?product=valgrind) rather<br> than mailing the developers (or mailing lists) directly -- bugs that<br> are not entered into bugzilla tend to get forgotten about or ignored.<br> <br> 208531 [PATCH]: FreeBSD support for valgrind<br> 368960 WARNING: unhandled amd64-linux syscall: 163 (acct)<br> 407589 [Linux] Add support for C11 aligned_alloc() and GNU reallocarray()<br> 423963 Error in child thread when CLONE_PIDFD is used<br> 426148 crash with "impossible happened" when running BPF CO-RE programs<br> 429375 PPC ISA 3.1 support is missing, part 9<br> 431157 PPC_FEATURE2_SCV needs to be masked in AT_HWCAP2<br> 431306 Update demangler to support Rust v0 name mangling<br> 432387 s390x: z15 instructions support<br> 433437 FreeBSD support, part 1<br> 433438 FreeBSD support, part 2<br> 433439 FreeBSD support, part 3<br> 433469 FreeBSD support, part 4<br> 433473 FreeBSD support, part 5<br> 433477 FreeBSD support, part 6<br> 433479 FreeBSD support, part 7<br> 433504 FreeBSD support, part 8<br> 433506 FreeBSD support, part 9<br> 433507 FreeBSD support, part 10<br> 433508 FreeBSD support, part 11<br> 433510 FreeBSD support, part 12<br> 433801 PPC ISA 3.1 support is missing, part 10 (ISA 3.1 support complete)<br> 433863 s390x: memcheck/tests/s390x/{cds,cs,csg} failures<br> 434296 s390x: False-positive memcheck diagnostics from vector string<br> instructions<br> 434840 PPC64 darn instruction not supported<br> 435665 PPC ISA 3.0 copy, paste, cpabort instructions are not supported <br> 435908 valgrind tries to fetch from deubginfod for files which already<br> have debug information<br> 438871 unhandled instruction bytes: 0xF3 0x49 0xF 0x6F 0x9C 0x24 0x60 0x2<br> 439046 valgrind is unusably large when linked with lld<br> 439090 Implement close_range(2)<br> 439326 Valgrind 3.17.0 won't compile with Intel 2021 oneAPI compilers<br> 439590 glibc-2.34 breaks suppressions against obj:*/lib*/libc-2.*so*<br> 440670 unhandled ppc64le-linux syscall: 252 statfs64 and 253 fstatfs64<br> 440906 Fix impossible constraint issue in P10 testcase.<br> 441512 Remove a unneeded / unnecessary prefix check.<br> 441534 Update the expected output for test_isa_3_1_VRT.<br> 442061 very slow execution under Fedora 34 (readdwarf3)<br> 443031 Gcc -many change requires explicit .machine directives<br> 443033 Add support for the ISA 3.0 mcrxrx instruction<br> 443034 Sraw, srawi, srad, sradi, mfs<br> 443178 Powerpc, test jm-mfspr expected output needs to be updated.<br> 443179 Need new test for the lxvx and stxvx instructions on ISA 2.07 and<br> ISA 3.0 systems.<br> 443180 The subnormal test and the ISA 3.0 test generate compiler warnings<br> 443314 In the latest GIT version, Valgrind with "--trace-flags" crashes<br> at "al" register<br> 443605 Don't call final_tidyup (__libc_freeres) on FatalSignal<br> <br> To see details of a given bug, visit<br> https://bugs.kde.org/show_bug.cgi?id=XXXXXX<br> where XXXXXX is the bug number as listed below.<br> <br> (3.18.0.RC1: 12 Oct 2021)<br> (3.18.0: 15 Oct 2021)<br> <br> <br> <br> Release 3.17.0 (19 Mar 2021)<br> ~~~~~~~~~~~~~~~~~~~~~~~~~~~~<br> <br> 3.17.0 fixes a number of bugs and adds some functional changes: support for<br> GCC 11, Clang 11, DWARF5 debuginfo, the 'debuginfod' debuginfo server, and<br> some new instructions for Arm64, S390 and POWER. There are also some tool<br> updates.<br> <br> This release supports X86/Linux, AMD64/Linux, ARM32/Linux, ARM64/Linux,<br> PPC32/Linux, PPC64BE/Linux, PPC64LE/Linux, S390X/Linux, MIPS32/Linux,<br> MIPS64/Linux, ARM/Android, ARM64/Android, MIPS32/Android, X86/Android,<br> X86/Solaris, AMD64/Solaris and AMD64/MacOSX 10.12. There is also preliminary<br> support for X86/macOS 10.13, AMD64/macOS 10.13 and nanoMIPS/Linux.<br> <br> * ==================== CORE CHANGES ===================<br> <br> * DWARF version 5 support. Valgrind can now read DWARF version 5 debuginfo as<br> produced by GCC 11.<br> <br> * Valgrind now supports debuginfod, an HTTP server for distributing ELF/DWARF<br> debugging information. When a debuginfo file cannot be found locally,<br> Valgrind is able to query debuginfod servers for the file using its<br> build-id. See the user manual for more information about debuginfod support.<br> <br> * ================== PLATFORM CHANGES =================<br> <br> * arm64:<br> <br> - Inaccuracies resulting from double-rounding in the simulation of<br> floating-point multiply-add/subtract instructions have been fixed. These<br> should now behave exactly as the hardware does.<br> <br> - Partial support for the ARM v8.2 instruction set. v8.2 support work is<br> ongoing. Support for the half-word variants of at least the following<br> instructions has been added:<br> FABS <Hd>, <Hn><br> FABS <Vd>.<T>, <Vn>.<T><br> FNEG <Hd>, <Hn><br> FNEG <Vd>.<T>, <Vn>.<T><br> FSQRT <Hd>, <Hn><br> FSQRT <Vd>.<T>, <Vn>.<T><br> FADDP<br> <br> * s390:<br> <br> - Implement the new instructions/features that were added to z/Architecture<br> with the vector-enhancements facility 1. Also cover the instructions from<br> the vector-packed-decimal facility that are defined outside the chapter<br> "Vector Decimal Instructions", but not the ones from that chapter itself.<br> <br> For a detailed list of newly supported instructions see the updates to<br> `docs/internals/s390-opcodes.csv'.<br> <br> Since the miscellaneous instruction extensions facility 2 was already<br> added in Valgrind 3.16.0, this completes the support necessary to run<br> general programs built with `--march=z14' under Valgrind. The<br> vector-packed-decimal facility is currently not exploited by the standard<br> toolchain and libraries.<br> <br> * ppc64:<br> <br> - Various bug fixes. Fix for the sync field to limit setting just two of<br> the two bits in the L-field. Fix the write size for the stxsibx and<br> stxsihx instructions. Fix the modsw and modsd instructions.<br> <br> - Partial support for ISA 3.1 has been added. Support for the VSX PCV mask<br> instructions, bfloat16 GER instructions, and bfloat16 to/from float 32-bit<br> conversion instructions are still missing.<br> <br> * ==================== TOOL CHANGES ====================<br> <br> * General tool changes<br> <br> - All the tools and their vgpreload libraries are now installed under<br> libexec because they cannot be executed directly and should be run through<br> the valgrind executable. This should be an internal, not user visible,<br> change, but might impact valgrind packagers.<br> <br> - The --track-fds option now respects -q, --quiet and won't output anything<br> if no file descriptors are leaked. It also won't report the standard stdin<br> (0), stdout (1) or stderr (2) descriptors as being leaked with<br> --trace-fds=yes anymore. To track whether the standard file descriptors<br> are still open at the end of the program run use --trace-fds=all.<br> <br> * DHAT:<br> <br> - DHAT has been extended, with two new modes of operation. The new<br> --mode=copy flag triggers copy profiling, which records calls to memcpy,<br> strcpy, and similar functions. The new --mode=ad-hoc flag triggers ad hoc<br> profiling, which records calls to the DHAT_AD_HOC_EVENT client request in<br> the new dhat/dhat.h file. This is useful for learning more about hot code<br> paths. See the user manual for more information about the new modes.<br> <br> - Because of these changes, DHAT's file format has changed. DHAT output<br> files produced with earlier versions of DHAT will not work with this<br> version of DHAT's viewer, and DHAT output files produced with this version<br> of DHAT will not work with earlier versions of DHAT's viewer.<br> <br> * ==================== FIXED BUGS ====================<br> <br> The following bugs have been fixed or resolved. Note that "n-i-bz"<br> stands for "not in bugzilla" -- that is, a bug that was reported to us<br> but never got a bugzilla entry. We encourage you to file bugs in<br> bugzilla (https://bugs.kde.org/enter_bug.cgi?product=valgrind) rather<br> than mailing the developers (or mailing lists) directly -- bugs that<br> are not entered into bugzilla tend to get forgotten about or ignored.<br> <br> To see details of a given bug, visit<br> https://bugs.kde.org/show_bug.cgi?id=XXXXXX<br> where XXXXXX is the bug number as listed below.<br> <br> 140178 open("/proc/self/exe", ...); doesn't quite work<br> 140939 --track-fds reports leakage of stdout/in/err and doesn't respect -q<br> 217695 malloc/calloc/realloc/memalign failure doesn't set errno to ENOMEM<br> 338633 gdbserver_tests/nlcontrolc.vgtest hangs on arm64<br> 345077 linux syscall execveat support (linux 3.19)<br> 361770 Missing F_ADD_SEALS<br> 369029 handle linux syscalls sched_getattr and sched_setattr<br> 384729 __libc_freeres inhibits cross-platform valgrind<br> 388787 Support for C++17 new/delete<br> 391853 Makefile.all.am:L247 and @SOLARIS_UNDEF_LARGESOURCE@ being empty<br> 396656 Warnings while reading debug info<br> 397605 ioctl FICLONE mishandled<br> 401416 Compile failure with openmpi 4.0<br> 408663 Suppression file for musl libc<br> 404076 s390x: z14 vector instructions not implemented<br> 410743 shmat() calls for 32-bit programs fail when running in 64-bit valgrind<br> (actually affected all x86 and nanomips regardless of host bitness)<br> 413547 regression test does not check for Arm 64 features.<br> 414268 Enable AArch64 feature detection and decoding for v8.x instructions<br> 415293 Incorrect call-graph tracking due to new _dl_runtime_resolve_xsave*<br> 422174 unhandled instruction bytes: 0x48 0xE9 (REX prefixed JMP instruction)<br> 422261 platform selection fails for unqualified client name<br> 422623 epoll_ctl warns for uninitialized padding on non-amd64 64bit arches<br> 423021 PPC: Add missing ISA 3.0 documentation link and HWCAPS test.<br> 423195 PPC ISA 3.1 support is missing, part 1<br> 423361 Adds io_uring support on arm64/aarch64 (and all other arches)<br> 424012 crash with readv/writev having invalid but not NULL arg2 iovec<br> 424298 amd64: Implement RDSEED<br> 425232 PPC ISA 3.1 support is missing, part 2<br> 425820 Failure to recognize vpcmpeqq as a dependency breaking idiom.<br> 426014 arm64: implement fmadd and fmsub as Iop_MAdd/Sub<br> 426123 PPC ISA 3.1 support is missing, part 3<br> 426144 Fix "condition variable has not been initialized" on Fedora 33.<br> 427400 PPC ISA 3.1 support is missing, part 4<br> 427401 PPC ISA 3.1 support is missing, part 5<br> 427404 PPC ISA 3.1 support is missing, part 6<br> 427870 lmw, lswi and related PowerPC insns aren't allowed on ppc64le<br> 427787 Support new faccessat2 linux syscall (439)<br> 427969 debuginfo section duplicates a section in the main ELF file<br> 428035 drd: Unbreak the musl build<br> 428648 s390_emit_load_mem panics due to 20-bit offset for vector load<br> 428716 cppcheck detects potential leak in VEX/useful/smchash.c<br> 428909 helgrind: need to intercept duplicate libc definitions for Fedora 33<br> 429352 PPC ISA 3.1 support is missing, part 7<br> 429354 PPC ISA 3.1 support is missing, part 8<br> 429692 unhandled ppc64le-linux syscall: 147 (getsid)<br> 429864 s390x: C++ atomic test_and_set yields false-positive memcheck<br> diagnostics<br> 429952 Errors when building regtest with clang<br> 430354 ppc stxsibx and stxsihx instructions write too much data<br> 430429 valgrind.h doesn't compile on s390x with clang<br> 430485 expr_is_guardable doesn't handle Iex_Qop<br> 431556 Complete arm64 FADDP v8.2 instruction support<br> 432102 Add support for DWARF5 as produced by GCC11<br> 432161 Addition of arm64 v8.2 FADDP, FNEG and FSQRT<br> 432381 drd: Process STACK_REGISTER client requests<br> 432552 [AArch64] invalid error emitted for pre-decremented byte/hword addresses<br> 432672 vg_regtest: test-specific environment variables not reset between tests<br> 432809 VEX should support REX.W + POPF<br> 432861 PPC modsw and modsd give incorrect results for 1 mod 12<br> 432870 gdbserver_tests:nlcontrolc hangs with newest glibc2.33 x86-64<br> 432215 Add debuginfod functionality<br> 433323 Use pkglibexecdir as vglibdir<br> 433500 DRD regtest faulures when libstdc++ and libgcc debuginfo are installed<br> 433629 valgrind/README has type "abd" instead of "and"<br> 433641 Rust std::sys::unix::fs::try_statx Syscall param fstatat(file_name)<br> 433898 arm64: Handle sp, lr, fp as DwReg in CfiExpr<br> 434193 GCC 9+ inlined strcmp causes "Conditional jump or move [..] value" report<br> n-i-bz helgrind: If hg_cli__realloc fails, return NULL.<br> n-i-bz arm64 front end: avoid Memcheck false positives relating to CPUID<br> <br> (3.17.0.RC1: 13 Mar 2021)<br> (3.17.0.RC2: 17 Mar 2021)<br> (3.17.0: 19 Mar 2021)<br> <br> <br> <br> Release 3.16.1 (22 June 2020)<br> ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~<br> <br> 3.16.1 fixes two critical bugs discovered after 3.16.0 was frozen. It also<br> fixes character encoding problems in the documentation HTML.<br> <br> 422677 PPC sync instruction L field should only be 2 bits in ISA 3.0<br> 422715 32-bit x86: vex: the `impossible' happened: expr_is_guardable: unhandled expr<br> <br> (3.16.1, 22 June 2020, 36d6727e1d768333a536f274491e5879cab2c2f7)<br> <br> <br> <br> Release 3.16.0 (27 May 2020)<br> ~~~~~~~~~~~~~~~~~~~~~~~~~~~~<br> <br> 3.16.0 is a feature release with many improvements and the usual collection of<br> bug fixes.<br> <br> This release supports X86/Linux, AMD64/Linux, ARM32/Linux, ARM64/Linux,<br> PPC32/Linux, PPC64BE/Linux, PPC64LE/Linux, S390X/Linux, MIPS32/Linux,<br> MIPS64/Linux, ARM/Android, ARM64/Android, MIPS32/Android, X86/Android,<br> X86/Solaris, AMD64/Solaris and AMD64/MacOSX 10.12. There is also preliminary<br> support for X86/macOS 10.13, AMD64/macOS 10.13 and nanoMIPS/Linux.<br> <br> * ==================== CORE CHANGES ===================<br> <br> * It is now possible to dynamically change the value of many command line<br> options while your program (or its children) are running under Valgrind.<br> <br> To see the list of dynamically changeable options, run<br> "valgrind --help-dyn-options".<br> <br> You can change the options from the shell by using vgdb to launch<br> the monitor command "v.clo <clo option>...".<br> The same monitor command can be used from a gdb connected<br> to the valgrind gdbserver.<br> Your program can also change the dynamically changeable options using<br> the client request VALGRIND_CLO_CHANGE(option).<br> <br> * ================== PLATFORM CHANGES =================<br> <br> * MIPS: preliminary support for nanoMIPS instruction set has been added.<br> <br> * ==================== TOOL CHANGES ====================<br> <br> * DHAT:<br> <br> - The implicit memcpy done by each call to realloc now counts towards the<br> read and write counts of resized heap blocks, making those counts higher<br> and more accurate.<br> <br> * Cachegrind:<br> <br> - cg_annotate's --auto and --show-percs options now default to 'yes', because<br> they are usually wanted.<br> <br> * Callgrind:<br> <br> - callgrind_annotate's --auto and --show-percs options now default to 'yes',<br> because they are usually wanted.<br> <br> - The command option --collect-systime has been enhanced to specify<br> the unit used to record the elapsed time spent during system calls.<br> The command option now accepts the values no|yes|msec|usec|nsec,<br> where yes is a synonym of msec. When giving the value nsec, the<br> system cpu time of system calls is also recorded.<br> <br> * Memcheck:<br> <br> - Several memcheck options are now dynamically changeable.<br> Use valgrind --help-dyn-options to list them.<br> <br> - The release 3.15 introduced a backward incompatible change for<br> some suppression entries related to preadv and pwritev syscalls.<br> When reading a suppression entry using the unsupported 3.14 format,<br> valgrind will now produce a warning to say the suppression entry will not<br> work, and suggest the needed change.<br> <br> - Significantly fewer false positive errors on optimised code generated by<br> Clang and GCC. In particular, Memcheck now deals better with the<br> situation where the compiler will transform C-level "A && B" into "B && A"<br> under certain circumstances (in which the transformation is valid).<br> Handling of integer equality/non-equality checks on partially defined<br> values is also improved on some architectures.<br> <br> * exp-sgcheck:<br> <br> - The exprimental Stack and Global Array Checking tool has been removed.<br> It only ever worked on x86 and amd64, and even on those it had a<br> high false positive rate and was slow. An alternative for detecting<br> stack and global array overruns is using the AddressSanitizer (ASAN)<br> facility of the GCC and Clang compilers, which require you to rebuild<br> your code with -fsanitize=address.<br> <br> * ==================== OTHER CHANGES ====================<br> <br> * New and modified GDB server monitor features:<br> <br> - Option -T tells vgdb to output a timestamp in the vgdb information messages.<br> <br> - The gdbserver monitor commands that require an address and an optional<br> length argument now accepts the alternate 'C like' syntax "address[length]".<br> For example, the memcheck command "monitor who_points_at 0x12345678 120"<br> can now also be given as "monitor who_points_at 0x12345678[120]".<br> <br> * ==================== FIXED BUGS ====================<br> <br> The following bugs have been fixed or resolved. Note that "n-i-bz"<br> stands for "not in bugzilla" -- that is, a bug that was reported to us<br> but never got a bugzilla entry. We encourage you to file bugs in<br> bugzilla (https://bugs.kde.org/enter_bug.cgi?product=valgrind) rather<br> than mailing the developers (or mailing lists) directly -- bugs that<br> are not entered into bugzilla tend to get forgotten about or ignored.<br> <br> To see details of a given bug, visit<br> https://bugs.kde.org/show_bug.cgi?id=XXXXXX<br> where XXXXXX is the bug number as listed below.<br> <br> 343099 Linux setns syscall wrapper missing, unhandled syscall: 308<br> == 368923 WARNING: unhandled arm64-linux syscall: 268 (setns)<br> == 369031 WARNING: unhandled amd64-linux syscall: 308 (setns)<br> 385386 Assertion failed "szB >= CACHE_ENTRY_SIZE" at m_debuginfo/image.c:517<br> 400162 Patch: Guard against __GLIBC_PREREQ for musl libc<br> 400593 In Coregrind, use statx for some internal syscalls if [f]stat[64] fail<br> 400872 Add nanoMIPS support to Valgrind<br> 403212 drd/tests/trylock hangs on FreeBSD<br> 404406 s390x: z14 miscellaneous instructions not implemented<br> 405201 Incorrect size of struct vki_siginfo on 64-bit Linux architectures<br> 406561 mcinfcallWSRU gdbserver_test fails on ppc64<br> 406824 Unsupported baseline<br> 407218 Add support for the copy_file_range syscall<br> 407307 Intercept stpcpy also in ld.so for arm64<br> 407376 Update Xen support to 4.12 (4.13, actually) and add more coverage<br> == 390553<br> 407764 drd cond_post_wait gets wrong (?) condition on s390x z13 system<br> 408009 Expose rdrand and f16c even on avx if host cpu supports them<br> 408091 Missing pkey syscalls<br> 408414 Add support for missing for preadv2 and pwritev2 syscalls<br> 409141 Valgrind hangs when SIGKILLed<br> 409206 Support for Linux PPS and PTP ioctls<br> 409367 exit_group() after signal to thread waiting in futex() causes hangs<br> 409429 amd64: recognize 'cmpeq' variants as a dependency breaking idiom<br> 409780 References to non-existent configure.in<br> 410556 Add support for BLKIO{MIN,OPT} and BLKALIGNOFF ioctls<br> 410599 Non-deterministic behaviour of pth_self_kill_15_other test<br> 410757 discrepancy for preadv2/pwritev2 syscalls across different versions<br> 411134 Allow the user to change a set of command line options during execution<br> 411451 amd64->IR of bt/btc/bts/btr with immediate clears zero flag<br> 412344 Problem setting mips flags with specific paths<br> 412408 unhandled arm-linux syscall: 124 - adjtime - on arm-linux<br> 413119 Ioctl wrapper for DRM_IOCTL_I915_GEM_MMAP<br> 413330 avx-1 test fails on AMD EPYC 7401P 24-Core Processor<br> 413603 callgrind_annotate/cg_annotate truncate function names at '#'<br> 414565 Specific use case bug found in SysRes VG_(do_sys_sigprocmask)<br> 415136 ARMv8.1 Compare-and-Swap instructions are not supported<br> 415757 vex x86->IR: 0x66 0xF 0xCE 0x4F (bswapw)<br> 416239 valgrind crashes when handling clock_adjtime<br> 416285 Use prlimit64 in VG_(getrlimit) and VG_(setrlimit)<br> 416286 DRD reports "conflicting load" error on std::mutex::lock()<br> 416301 s390x: "compare and signal" not supported<br> 416387 finit_module and bpf syscalls are unhandled on arm64<br> 416464 Fix false reports for uninitialized memory for PR_CAPBSET_READ/DROP<br> 416667 gcc10 ppc64le impossible constraint in 'asm' in test_isa.<br> 416753 new 32bit time syscalls for 2038+<br> 417075 pwritev(vector[...]) suppression ignored<br> 417075 is not fixed, but incompatible supp entries are detected<br> and a warning is produced for these.<br> 417187 [MIPS] Conditional branch problem since 'grail' changes<br> 417238 Test memcheck/tests/vbit-test fails on mips64 BE<br> 417266 Make memcheck/tests/linux/sigqueue usable with musl<br> 417281 s390x: /bin/true segfaults with "grail" enabled<br> 417427 commit to fix vki_siginfo_t definition created numerous regression<br> errors on ppc64<br> 417452 s390_insn_store_emit: dst->tag for HRcVec128<br> 417578 Add suppressions for glibc DTV leaks<br> 417906 clone with CLONE_VFORK and no CLONE_VM fails<br> 418004 Grail code additions break ppc64.<br> 418435 s390x: spurious "Conditional jump or move depends on uninitialised [..]"<br> 418997 s390x: Support Iex_ITE for float and vector types<br> 419503 s390x: Avoid modifying registers returned from isel functions<br> 421321 gcc10 arm64 build needs __getauxval for linking with libgcc<br> 421570 std_mutex fails on Arm v8.1 h/w<br> 434035 vgdb might crash if valgrind is killed<br> n-i-bz Fix minor one time leaks in dhat.<br> n-i-bz Add --run-cxx-freeres=no in outer args to avoid inner crashes.<br> n-i-bz Add support for the Linux io_uring system calls<br> n-i-bz sys_statx: don't complain if both |filename| and |buf| are NULL.<br> n-i-bz Fix non-glibc build of test suite with s390x_features<br> n-i-bz MinGW, include/valgrind.h: Fix detection of 64-bit mode<br> 423195 PPC ISA 3.1 support is missing, part 1<br> <br> (3.16.0.RC1: 18 May 2020, git 6052ee66a0cf5234e8e2a2b49a8760226bc13b92)<br> (3.16.0.RC2: 19 May 2020, git 940ec1ca69a09f7fdae3e800b7359f85c13c4b37)<br> (3.16.0: 27 May 2020, git bf5e647edb9e96cbd5c57cc944984402eeee296d)<br> <br> <br> <br> Release 3.15.0 (12 April 2019)<br> ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~<br> <br> 3.15.0 is a feature release with many improvements and the usual collection of<br> bug fixes.<br> <br> This release supports X86/Linux, AMD64/Linux, ARM32/Linux, ARM64/Linux,<br> PPC32/Linux, PPC64BE/Linux, PPC64LE/Linux, S390X/Linux, MIPS32/Linux,<br> MIPS64/Linux, ARM/Android, ARM64/Android, MIPS32/Android, X86/Android,<br> X86/Solaris, AMD64/Solaris and AMD64/MacOSX 10.12. There is also preliminary<br> support for X86/macOS 10.13 and AMD64/macOS 10.13.<br> <br> * ==================== CORE CHANGES ===================<br> <br> * The XTree Massif output format now makes use of the information obtained<br> when specifying --read-inline-info=yes.<br> <br> * amd64 (x86_64): the RDRAND and F16C insn set extensions are now supported.<br> <br> * ==================== TOOL CHANGES ====================<br> <br> * DHAT: <br> <br> - DHAT been thoroughly overhauled, improved, and given a GUI. As a result,<br> it has been promoted from an experimental tool to a regular tool. Run it<br> with --tool=dhat instead of --tool=exp-dhat.<br> <br> - DHAT now prints only minimal data when the program ends, instead writing<br> the bulk of the profiling data to a file. As a result, the --show-top-n<br> and --sort-by options have been removed.<br> <br> - Profile results can be viewed with the new viewer, dh_view.html. When<br> a run ends, a short message is printed, explaining how to view the result.<br> <br> - See the documentation for more details.<br> <br> * Cachegrind:<br> <br> - cg_annotate has a new option, --show-percs, which prints percentages next<br> to all event counts.<br> <br> * Callgrind:<br> <br> - callgrind_annotate has a new option, --show-percs, which prints percentages<br> next to all event counts.<br> <br> - callgrind_annotate now inserts commas in call counts, and<br> sort the caller/callee lists in the call tree.<br> <br> * Massif:<br> <br> - The default value for --read-inline-info is now "yes" on<br> Linux/Android/Solaris. It is still "no" on other OS.<br> <br> * Memcheck:<br> <br> - The option --xtree-leak=yes (to output leak result in xtree format)<br> automatically activates the option --show-leak-kinds=all, as xtree<br> visualisation tools such as kcachegrind can in any case select what kind<br> of leak to visualise.<br> <br> - There has been further work to avoid false positives. In particular,<br> integer equality on partially defined inputs (C == and !=) is now handled<br> better.<br> <br> * ==================== OTHER CHANGES ====================<br> <br> * The new option --show-error-list=no|yes displays, at the end of the run, the<br> list of detected errors and the used suppressions. Prior to this change,<br> showing this information could only be done by specifying "-v -v", but that<br> also produced a lot of other possibly-non-useful messages. The option -s is<br> equivalent to --show-error-list=yes.<br> <br> * ==================== FIXED BUGS ====================<br> <br> The following bugs have been fixed or resolved. Note that "n-i-bz"<br> stands for "not in bugzilla" -- that is, a bug that was reported to us<br> but never got a bugzilla entry. We encourage you to file bugs in<br> bugzilla (https://bugs.kde.org/enter_bug.cgi?product=valgrind) rather<br> than mailing the developers (or mailing lists) directly -- bugs that<br> are not entered into bugzilla tend to get forgotten about or ignored.<br> <br> To see details of a given bug, visit<br> https://bugs.kde.org/show_bug.cgi?id=XXXXXX<br> where XXXXXX is the bug number as listed below.<br> <br> 385411 s390x: z13 vector floating-point instructions not implemented<br> 397187 z13 vector register support for vgdb gdbserver<br> 398183 Vex errors with _mm256_shuffle_epi8/vpshufb<br> 398870 Please add support for instruction vcvtps2ph<br> 399287 amd64 front end: Illegal Instruction vcmptrueps<br> 399301 Use inlined frames in Massif XTree output. <br> 399322 Improve callgrind_annotate output<br> 399444 VEX/priv/guest_s390_toIR.c:17407: (style) Mismatching assignment [..]<br> 400164 helgrind test encounters mips x-compiler warnings and assembler error<br> 400490 s390x: VRs allocated as if separate from FPRs<br> 400491 s390x: Operand of LOCH treated as unsigned integer<br> 400975 Compile error: error: '-mips64r2' conflicts with the other architecture<br> options, which specify a mips64 processor<br> 401112 LLVM 5.0 generates comparison against partially initialized data<br> 401277 More bugs in z13 support<br> 401454 Add a --show-percs option to cg_annotate and callgrind_annotate.<br> 401578 drd: crashes sometimes on fork()<br> 401627 memcheck errors with glibc avx2 optimized wcsncmp<br> 401822 none/tests/ppc64/jm-vmx fails and produces assembler warnings<br> 401827 none/tests/ppc64/test_isa_2_06_part3 failure on ppc64le (xvrsqrtesp)<br> 401828 none/tests/ppc64/test_isa_2_06_part1 failure on ppc64le (fcfids and<br> fcfidus)<br> 402006 mark helper regs defined in final_tidyup before freeres_wrapper call<br> 402048 WARNING: unhandled ppc64[be|le]-linux syscall: 26 (ptrace)<br> 402123 invalid assembler opcodes for mips32r2<br> 402134 assertion fail in mc_translate.c (noteTmpUsesIn) Iex_VECRET on arm64<br> 402327 Warning: DWARF2 CFI reader: unhandled DW_OP_ opcode 0x13 (DW_OP_drop)<br> 402341 drd/tests/tsan_thread_wrappers_pthread.h:369: suspicious code ?<br> 402351 mips64 libvexmultiarch_test fails on s390x<br> 402369 Overhaul DHAT<br> 402395 coregrind/vgdb-invoker-solaris.c: 2 * poor error checking<br> 402480 Do not use %rsp in clobber list<br> 402481 vbit-test fails on x86 for Iop_CmpEQ64 iselInt64Expr Sar64<br> 402515 Implement new option --show-error-list=no|yes / -s<br> 402519 POWER 3.0 addex instruction incorrectly implemented<br> 402781 Redo the cache used to process indirect branch targets<br> 403123 vex amd64->IR:0xF3 0x48 0xF 0xAE 0xD3 (wrfsbase)<br> 403552 s390x: wrong facility bit checked for vector facility<br> 404054 memcheck powerpc subfe x, x, x initializes x to 0 or -1 based on CA<br> 404638 Add VG_(replaceIndexXA)<br> 404843 s390x: backtrace sometimes ends prematurely<br> 404888 autotools cleanup series<br> 405079 unhandled ppc64le-linux syscall: 131 (quotactl)<br> 405182 Valgrind fails to build with Clang<br> 405205 filter_libc: remove the line holding the futex syscall error entirely<br> 405356 PPC64, xvcvsxdsp, xvcvuxdsp are supposed to write the 32-bit result to<br> the upper and lower 32-bits of the 64-bit result<br> 405362 PPC64, vmsummbm instruction doesn't handle overflow case correctly<br> 405363 PPC64, xvcvdpsxws, xvcvdpuxws, do not handle NaN arguments correctly.<br> 405365 PPC64, function _get_maxmin_fp_NaN() doesn't handle QNaN, SNaN case<br> correctly.<br> 405403 s390x disassembler cannot be used on x86<br> 405430 Use gcc -Wimplicit-fallthrough=2 by default if available<br> 405458 MIPS mkFormVEC arguments swapped?<br> 405716 drd: Fix an integer overflow in the stack margin calculation<br> 405722 Support arm64 core dump<br> 405733 PPC64, xvcvdpsp should write 32-bit result to upper and lower 32-bits<br> of the 64-bit destination field.<br> 405734 PPC64, vrlwnm, vrlwmi, vrldrm, vrldmi do not work properly when me < mb<br> 405782 "VEX temporary storage exhausted" when attempting to debug slic3r-pe<br> 406198 none/tests/ppc64/test_isa_3_0_other test sporadically including CA<br> bit in output.<br> 406256 PPC64, vector floating point instructions don't handle subnormal<br> according to VSCR[NJ] bit setting.<br> 406352 cachegrind/callgrind fails ann tests because of missing a.c<br> 406354 dhat is broken on x86 (32bit)<br> 406355 mcsignopass, mcsigpass, mcbreak fail due to difference in gdb output<br> 406357 gdbserver_tests fails because of gdb output change<br> 406360 memcheck/tests/libstdc++.supp needs more supression variants<br> 406422 none/tests/amd64-linux/map_32bits.vgtest fails too easily<br> 406465 arm64 insn selector fails on "t0 = <expr>" where <expr> has type Ity_F16<br> 407340 PPC64, does not support the vlogefp, vexptefp instructions.<br> n-i-bz add syswrap for PTRACE_GET|SET_THREAD_AREA on amd64.<br> n-i-bz Fix callgrind_annotate non deterministic order for equal total<br> n-i-bz callgrind_annotate --threshold=100 does not print all functions.<br> n-i-bz callgrind_annotate Use of uninitialized value in numeric gt (>)<br> n-i-bz amd64 (x86_64): RDRAND and F16C insn set extensions are supported<br> <br> (3.15.0.RC1: 8 April 2019, git ce94d674de5b99df173aad4c3ee48fc2a92e5d9c)<br> (3.15.0.RC2: 11 April 2019, git 0c8be9bbede189ec580ec270521811766429595f)<br> (3.15.0: 14 April 2019, git 270037da8b508954f0f7d703a0bebf5364eec548)<br> <br> <br> <br> Release 3.14.0 (9 October 2018)<br> ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~<br> <br> 3.14.0 is a feature release with many improvements and the usual collection of<br> bug fixes.<br> <br> This release supports X86/Linux, AMD64/Linux, ARM32/Linux, ARM64/Linux,<br> PPC32/Linux, PPC64BE/Linux, PPC64LE/Linux, S390X/Linux, MIPS32/Linux,<br> MIPS64/Linux, ARM/Android, ARM64/Android, MIPS32/Android, X86/Android,<br> X86/Solaris, AMD64/Solaris and AMD64/MacOSX 10.12. There is also preliminary<br> support for X86/macOS 10.13, AMD64/macOS 10.13.<br> <br> * ==================== CORE CHANGES ===================<br> <br> * The new option --keep-debuginfo=no|yes (default no) can be used to retain<br> debug info for unloaded code. This allows saved stack traces (e.g. for<br> memory leaks) to include file/line info for code that has been dlclose'd (or<br> similar). See the user manual for more information and known limitations.<br> <br> * Ability to specify suppressions based on source file name and line number.<br> <br> * Majorly overhauled register allocator. No end-user changes, but the JIT<br> generates code a bit more quickly now.<br> <br> * ================== PLATFORM CHANGES =================<br> <br> * Preliminary support for macOS 10.13 has been added.<br> <br> * mips: support for MIPS32/MIPS64 Revision 6 has been added.<br> <br> * mips: support for MIPS SIMD architecture (MSA) has been added.<br> <br> * mips: support for MIPS N32 ABI has been added.<br> <br> * s390: partial support for vector instructions (integer and string) has been<br> added.<br> <br> * ==================== TOOL CHANGES ====================<br> <br> * Helgrind: Addition of a flag<br> --delta-stacktrace=no|yes [yes on linux amd64/x86]<br> which specifies how full history stack traces should be computed.<br> Setting this to =yes can speed up Helgrind by 25% when using<br> --history-level=full.<br> <br> * Memcheck: reduced false positive rate for optimised code created by Clang 6<br> / LLVM 6 on x86, amd64 and arm64. In particular, Memcheck analyses code<br> blocks more carefully to determine where it can avoid expensive definedness<br> checks without loss of precision. This is controlled by the flag<br> --expensive-definedness-checks=no|auto|yes [auto].<br> <br> * ==================== OTHER CHANGES ====================<br> <br> * Valgrind is now buildable with link-time optimisation (LTO). A new<br> configure option --enable-lto=yes allows building Valgrind with LTO. If the<br> toolchain supports it, this produces a smaller/faster Valgrind (up to 10%).<br> Note that if you are doing Valgrind development, --enable-lto=yes massively<br> slows down the build process.<br> <br> * ==================== FIXED BUGS ====================<br> <br> The following bugs have been fixed or resolved. Note that "n-i-bz"<br> stands for "not in bugzilla" -- that is, a bug that was reported to us<br> but never got a bugzilla entry. We encourage you to file bugs in<br> bugzilla (https://bugs.kde.org/enter_bug.cgi?product=valgrind) rather<br> than mailing the developers (or mailing lists) directly -- bugs that<br> are not entered into bugzilla tend to get forgotten about or ignored.<br> <br> To see details of a given bug, visit<br> https://bugs.kde.org/show_bug.cgi?id=XXXXXX<br> where XXXXXX is the bug number as listed below.<br> <br> 79362 Debug info is lost for .so files when they are dlclose'd<br> 208052 strlcpy error when n = 0<br> 255603 exp-sgcheck Assertion '!already_present' failed<br> 338252 building valgrind with -flto (link time optimisation) fails<br> 345763 MIPS N32 ABI support<br> 368913 WARNING: unhandled arm64-linux syscall: 117 (ptrace)<br> == 388664 unhandled arm64-linux syscall: 117 (ptrace)<br> 372347 Replacement problem of the additional c++14/c++17 new/delete operators<br> 373069 memcheck/tests/leak_cpp_interior fails with GCC 5.1+<br> 376257 helgrind history full speed up using a cached stack<br> 379373 Fix syscall param msg->desc.port.name points to uninitialised byte(s)<br> on macOS 10.12<br> 379748 Fix missing pselect syscall (OS X 10.11)<br> 379754 Fix missing syscall ulock_wait (OS X 10.12)<br> 380397 s390x: __GI_strcspn() replacemenet needed<br> 381162 possible array overrun in VEX register allocator<br> 381272 ppc64 doesn't compile test_isa_2_06_partx.c without VSX support<br> 381274 powerpc too chatty even with --sigill-diagnostics=no<br> 381289 epoll_pwait can have a NULL sigmask<br> 381553 VEX register allocator v3<br> 381556 arm64: Handle feature registers access on 4.11 Linux kernel or later<br> 381769 Use ucontext_t instead of struct ucontext<br> 381805 arm32 needs ld.so index hardwire for new glibc security fixes<br> 382256 gz compiler flag test doesn't work for gold<br> 382407 vg_perf needs "--terse" command line option<br> 382515 "Assertion 'di->have_dinfo' failed." on wine's dlls/mscoree/tests/[..]<br> 382563 MIPS MSA ASE support<br> 382998 xml-socket doesn't work<br> 383275 massif: m_xarray.c:162 (ensureSpaceXA): Assertion '!xa->arr' failed<br> 383723 Fix missing kevent_qos syscall (macOS 10.11)<br> == 385604 illegal hardware instruction (OpenCV cv::namedWindow)<br> 384096 Mention AddrCheck at Memcheck's command line option [..]<br> 384230 vex x86->IR: 0x67 0xE8 0xAB 0x68<br> == 384156 vex x86->IR: 0x67 0xE8 0x6B 0x6A<br> == 386115 vex x86->IR: 0x67 0xE8 0xD3 0x8B any program<br> == 388407 vex x86->IR: 0x67 0xE8 0xAB 0x29<br> == 394903 vex x86->IR: 0x67 0xE8 0x1B 0xDA<br> 384337 performance improvements to VEX register allocator v2 and v3<br> 384526 reduce number of spill insns generated by VEX register allocator v3<br> 384584 Callee saved regs listed first for AMD64, X86, and PPC architectures<br> 384631 Sanitise client args as printed with -v<br> 384633 Add a simple progress-reporting facility<br> 384987 VEX regalloc: allocate caller-save registers for short lived vregs<br> 385055 PPC VEX temporary storage exhausted<br> 385182 PPC64 is missing support for the DSCR<br> 385183 PPC64, Add support for xscmpeqdp, xscmpgtdp, xscmpgedp, xsmincdp<br> 385207 PPC64, generate_store_FPRF() generates too many Iops<br> 385208 PPC64, xxperm instruction exhausts temporary memory<br> 385210 PPC64, vpermr instruction could exhaust temporary memory<br> 385279 unhandled syscall: mach:43 (mach_generate_activity_id)<br> == 395136 valgrind: m_syswrap/syswrap-main.c:438 (Bool eq_Syscall[..]<br> == 387045 Valgrind crashing on High Sierra when testing any newly [..]<br> 385334 PPC64, fix vpermr, xxperm, xxpermr mask value.<br> 385408 s390x: z13 vector "support" instructions not implemented<br> 385409 s390x: z13 vector integer instructions not implemented<br> 385410 s390x: z13 vector string instructions not implemented<br> 385412 s390x: new non-vector z13 instructions not implemented<br> 385868 glibc ld.so _dl_runtime_resolve_avx_slow conditional jump warning.<br> 385912 none/tests/rlimit_nofile fails on newer glibc/kernel.<br> 385939 Optionally exit on the first error<br> 386318 valgrind.org/info/tools.html is missing SGCheck<br> 386425 running valgrind + wine on armv7l gives illegal opcode<br> 386397 PPC64, valgrind truncates powerpc timebase to 32-bits.<br> 387410 MIPSr6 support<br> 387664 Memcheck: make expensive-definedness-checks be the default<br> 387712 s390x cgijnl reports Conditional jump depends on uninitialised value<br> 387766 asm shifts cause false positive "Conditional jump or move depends<br> on uninitialised value"<br> 387773 .gnu_debugaltlink paths resolve relative to .debug file, not symlink<br> 388174 valgrind with Wine quits with "Assertion 'cfsi_fits' failed"<br> 388786 Support bpf syscall in amd64 Linux<br> 388862 Add replacements for wmemchr and wcsnlen on Linux<br> 389065 valgrind meets gcc flag -Wlogical-op<br> 389373 exp-sgcheck the 'impossible' happened as Ist_LoadG is not instrumented<br> 390471 suppression by specification of source-file line number<br> 390723 make xtree dump files world wide readable, similar to log files<br> 391164 constraint bug in tests/ppc64/test_isa_2_07_part1.c for mtfprwa<br> 391861 Massif Assertion 'n_ips >= 1 && n_ips <= VG_(clo_backtrace_size)'<br> 392118 unhandled amd64-linux syscall: 332 (statx)<br> 392449 callgrind not clearing the number of calls properly<br> 393017 Add missing support for xsmaxcdp instruction, bug fixes for xsmincdp,<br> lxssp, stxssp and stxvl instructions.<br> 393023 callgrind_control risks using the wrong vgdb<br> 393062 build-id ELF phdrs read causes "debuginfo reader: ensure_valid failed"<br> 393099 posix_memalign() invalid write if alignment == 0<br> 393146 failing assert "is_DebugInfo_active(di)"<br> 395709 PPC64 is missing support for the xvnegsp instruction<br> 395682 Accept read-only PT_LOAD segments and .rodata by ld -z separate-code<br> == 384727<br> 396475 valgrind OS-X build: config.h not found (out-of-tree macOS builds)<br> 395991 arm-linux: wine's unit tests enter a signal delivery loop [..]<br> 396839 s390x: Trap instructions not implemented<br> 396887 arch_prctl should return EINVAL on unknown option<br> == 397286 crash before launching binary (Unsupported arch_prctl option)<br> == 397393 valgrind: the 'impossible' happened: (Archlinux)<br> == 397521 valgrind: the 'impossible' happened: Unsupported [..]<br> 396906 compile tests failure on mips32-linux: broken inline asm in tests on<br> mips32-linux<br> 397012 glibc ld.so uses arch_prctl on i386<br> 397089 amd64: Incorrect decoding of three-register vmovss/vmovsd opcode 11h<br> 397354 utimensat should ignore timespec tv_sec if tv_nsec is UTIME_NOW/OMIT<br> 397424 glibc 2.27 and gdb_server tests<br> 398028 Assertion `cfsi_fits` failing in simple C program<br> 398066 s390x: cgijl dep1, 0 reports false unitialised values warning<br> <br> n-i-bz Fix missing workq_ops operations (macOS)<br> n-i-bz fix bug in strspn replacement<br> n-i-bz Add support for the Linux BLKFLSBUF ioctl<br> n-i-bz Add support for the Linux BLKREPORTZONE and BLKRESETZONE ioctls<br> n-i-bz Fix possible stack trashing by semctl syscall wrapping<br> n-i-bz Add support for the Linux membarrier() system call<br> n-i-bz x86 front end: recognise and handle UD2 correctly<br> n-i-bz Signal delivery for x86-linux: ensure that the stack pointer is<br> correctly aligned before entering the handler.<br> <br> (3.14.0.RC1: 30 September 2018, git c2aeea2d28acb0639bcc8cc1e4ab115067db1eae)<br> (3.14.0.RC2: 3 October 2018, git 3e214c4858a6fdd5697e767543a0c19e30505582)<br> (3.14.0: 9 October 2018, git 353a3587bb0e2757411f9138f5e936728ed6cc4f)<br> <br> <br> <br> Release 3.13.0 (15 June 2017)<br> ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~<br> <br> 3.13.0 is a feature release with many improvements and the usual collection of<br> bug fixes.<br> <br> This release supports X86/Linux, AMD64/Linux, ARM32/Linux, ARM64/Linux,<br> PPC32/Linux, PPC64BE/Linux, PPC64LE/Linux, S390X/Linux, MIPS32/Linux,<br> MIPS64/Linux, ARM/Android, ARM64/Android, MIPS32/Android, X86/Android,<br> X86/Solaris, AMD64/Solaris and AMD64/MacOSX 10.12.<br> <br> * ==================== CORE CHANGES ===================<br> <br> * The translation cache size has been increased to keep up with the demands of<br> large applications. The maximum number of sectors has increased from 24 to<br> 48. The default number of sectors has increased from 16 to 32 on all<br> targets except Android, where the increase is from 6 to 12.<br> <br> * The amount of memory that Valgrind can use has been increased from 64GB to<br> 128GB. In particular this means your application can allocate up to about<br> 60GB when running on Memcheck.<br> <br> * Valgrind's default load address has been changed from 0x3800'0000 to<br> 0x5800'0000, so as to make it possible to load larger executables. This<br> should make it possible to load executables of size at least 1200MB.<br> <br> * A massive spaceleak caused by reading compressed debuginfo files has been<br> fixed. Valgrind should now be entirely usable with gcc-7.0 "-gz" created<br> debuginfo.<br> <br> * The C++ demangler has been updated.<br> <br> * Support for demangling Rust symbols has been added.<br> <br> * A new representation of stack traces, the "XTree", has been added. An XTree<br> is a tree of stacktraces with data associated with the stacktraces. This is<br> used by various tools (Memcheck, Helgrind, Massif) to report on the heap<br> consumption of your program. Reporting is controlled by the new options<br> --xtree-memory=none|allocs|full and --xtree-memory-file=<file>.<br> <br> A report can also be produced on demand using the gdbserver monitor command<br> 'xtmemory [<filename>]>'. The XTree can be output in 2 formats: 'callgrind<br> format' and 'massif format. The existing visualisers for these formats (e.g.<br> callgrind_annotate, KCachegrind, ms_print) can be used to visualise and<br> analyse these reports.<br> <br> Memcheck can also produce XTree leak reports using the Callgrind file<br> format. For more details, see the user manual.<br> <br> * ================== PLATFORM CHANGES =================<br> <br> * ppc64: support for ISA 3.0B and various fixes for existing 3.0 support<br> <br> * amd64: fixes for JIT failure problems on long AVX2 code blocks<br> <br> * amd64 and x86: support for CET prefixes has been added<br> <br> * arm32: a few missing ARMv8 instructions have been implemented<br> <br> * arm64, mips64, mips32: an alternative implementation of Load-Linked and<br> Store-Conditional instructions has been added. This is to deal with<br> processor implementations that implement the LL/SC specifications strictly<br> and as a result cause Valgrind to hang in certain situations. The<br> alternative implementation is automatically enabled at startup, as required.<br> You can use the option --sim-hints=fallback-llsc to force-enable it if you<br> want.<br> <br> * Support for OSX 10.12 has been improved.<br> <br> * On Linux, clone handling has been improved to honour CLONE_VFORK that<br> involves a child stack. Note however that CLONE_VFORK | CLONE_VM is handled<br> like CLONE_VFORK (by removing CLONE_VM), so applications that depend on<br> CLONE_VM exact semantics will (still) not work.<br> <br> * The TileGX/Linux port has been removed because it appears to be both unused<br> and unsupported.<br> <br> * ==================== TOOL CHANGES ====================<br> <br> * Memcheck:<br> <br> - Memcheck should give fewer false positives when running optimised<br> Clang/LLVM generated code.<br> <br> - Support for --xtree-memory and 'xtmemory [<filename>]>'.<br> <br> - New command line options --xtree-leak=no|yes and --xtree-leak-file=<file><br> to produce the end of execution leak report in a xtree callgrind format<br> file.<br> <br> - New option 'xtleak' in the memcheck leak_check monitor command, to produce<br> the leak report in an xtree file.<br> <br> * Massif:<br> <br> - Support for --xtree-memory and 'xtmemory [<filename>]>'.<br> <br> - For some workloads (typically, for big applications), Massif memory<br> consumption and CPU consumption has decreased significantly.<br> <br> * Helgrind:<br> <br> - Support for --xtree-memory and 'xtmemory [<filename>]>'.<br> <br> - addition of client request VALGRIND_HG_GNAT_DEPENDENT_MASTER_JOIN, useful<br> for Ada gnat compiled applications.<br> <br> * ==================== OTHER CHANGES ====================<br> <br> * For Valgrind developers: in an outer/inner setup, the outer Valgrind will<br> append the inner guest stacktrace to the inner host stacktrace. This helps<br> to investigate the errors reported by the outer, when they are caused by the<br> inner guest program (such as an inner regtest). See README_DEVELOPERS for<br> more info.<br> <br> * To allow fast detection of callgrind files by desktop environments and file<br> managers, the format was extended to have an optional first line that<br> uniquely identifies the format ("# callgrind format"). Callgrind creates<br> this line now, as does the new xtree functionality.<br> <br> * File name template arguments (such as --log-file, --xtree-memory-file, ...)<br> have a new %n format letter that is replaced by a sequence number.<br> <br> * "--version -v" now shows the SVN revision numbers from which Valgrind was<br> built.<br> <br> * ==================== FIXED BUGS ====================<br> <br> The following bugs have been fixed or resolved. Note that "n-i-bz"<br> stands for "not in bugzilla" -- that is, a bug that was reported to us<br> but never got a bugzilla entry. We encourage you to file bugs in<br> bugzilla (https://bugs.kde.org/enter_bug.cgi?product=valgrind) rather<br> than mailing the developers (or mailing lists) directly -- bugs that<br> are not entered into bugzilla tend to get forgotten about or ignored.<br> <br> To see details of a given bug, visit<br> https://bugs.kde.org/show_bug.cgi?id=XXXXXX<br> where XXXXXX is the bug number as listed below.<br> <br> 162848 --log-file output isn't split when a program forks<br> 340777 Illegal instruction on mips (ar71xx)<br> 341481 MIPS64: Iop_CmpNE32 triggers false warning on MIPS64 platforms<br> 342040 Valgrind mishandles clone with CLONE_VFORK | CLONE_VM that clones<br> to a different stack.<br> 344139 x86 stack-seg overrides, needed by the Wine people<br> 344524 store conditional of guest applications always fail - observed on<br> Octeon3(MIPS)<br> 348616 Wine/valgrind: noted but unhandled ioctl 0x5390 [..] (DVD_READ_STRUCT)<br> 352395 Please provide SVN revision info in --version -v<br> 352767 Wine/valgrind: noted but unhandled ioctl 0x5307 [..] (CDROMSTOP)<br> 356374 Assertion 'DRD_(g_threadinfo)[tid].pt_threadid !=<br> INVALID_POSIX_THREADID' failed<br> 358213 helgrind/drd bar_bad testcase hangs or crashes with new glibc pthread<br> barrier implementation<br> 358697 valgrind.h: Some code remains even when defining NVALGRIND<br> 359202 Add musl libc configure/compile<br> 360415 amd64 instructions ADCX and ADOX are not implemented in VEX<br> == 372828 (vex amd64->IR: 0x66 0xF 0x3A 0x62 0x4A 0x10)<br> 360429 unhandled ioctl 0x530d with no size/direction hints (CDROMREADMODE1)<br> 362223 assertion failed when .valgrindrc is a directory instead of a file<br> 367543 bt/btc/btr/bts x86/x86_64 instructions are poorly-handled wrt flags<br> 367942 Segfault vgPlain_do_sys_sigaction (m_signals.c:1138)<br> 368507 can't malloc chunks larger than about 34GB<br> 368529 Android arm target link error, missing atexit and pthread_atfork<br> 368863 WARNING: unhandled arm64-linux syscall: 100 (get_robust_list)<br> 368865 WARNING: unhandled arm64-linux syscall: 272 (kcmp)<br> 368868 disInstr(arm64): unhandled instruction 0xD53BE000 = cntfrq_el0 (ARMv8)<br> 368917 WARNING: unhandled arm64-linux syscall: 218 (request_key)<br> 368918 WARNING: unhandled arm64-linux syscall: 127 (sched_rr_get_interval)<br> 368922 WARNING: unhandled arm64-linux syscall: 161 (sethostname)<br> 368924 WARNING: unhandled arm64-linux syscall: 84 (sync_file_range)<br> 368925 WARNING: unhandled arm64-linux syscall: 130 (tkill)<br> 368926 WARNING: unhandled arm64-linux syscall: 97 (unshare)<br> 369459 valgrind on arm64 violates the ARMv8 spec (ldxr/stxr)<br> 370028 Reduce the number of compiler warnings on MIPS platforms<br> 370635 arm64 missing syscall getcpu<br> 371225 Fix order of timer_{gettime,getoverrun,settime} syscalls on arm64<br> 371227 Clean AArch64 syscall table<br> 371412 Rename wrap_sys_shmat to sys_shmat like other wrappers<br> 371471 Valgrind complains about non legit memory leaks on placement new (C++)<br> 371491 handleAddrOverrides() is [incorrect] when ASO prefix is used<br> 371503 disInstr(arm64): unhandled instruction 0xF89F0000<br> 371869 support '%' in symbol Z-encoding<br> 371916 execution tree xtree concept<br> 372120 c++ demangler demangles symbols which are not c++<br> 372185 Support of valgrind on ARMv8 with 32 bit executable<br> 372188 vex amd64->IR: 0x66 0xF 0x3A 0x62 0x4A 0x10 0x10 0x48 (PCMPxSTRx $0x10)<br> 372195 Power PC, xxsel instruction is not always recognized.<br> 372504 Hanging on exit_group<br> 372600 process loops forever when fatal signals are arriving quickly<br> 372794 LibVEX (arm32 front end): 'Assertion szBlg2 <= 3' failed<br> 373046 Stacks registered by core are never deregistered<br> 373069 memcheck/tests/leak_cpp_interior fails with GCC 5.1+<br> 373086 Implement additional Xen hypercalls<br> 373192 Calling posix_spawn in glibc 2.24 completely broken<br> 373488 Support for fanotify API on ARM64 architecture<br> == 368864 WARNING: unhandled arm64-linux syscall: 262 (fanotify_init)<br> 373555 Rename BBPTR to GSPTR as it denotes guest state pointer only<br> 373938 const IRExpr arguments for matchIRExpr()<br> 374719 some spelling fixes<br> 374963 increase valgrind's load address to prevent mmap failure<br> 375514 valgrind_get_tls_addr() does not work in case of static TLS<br> 375772 +1 error in get_elf_symbol_info() when computing value of 'hi' address<br> for ML_(find_rx_mapping)()<br> 375806 Test helgrind/tests/tc22_exit_w_lock fails with glibc 2.24<br> 375839 Temporary storage exhausted, with long sequence of vfmadd231ps insns<br> == 377159 "vex: the `impossible' happened" still present<br> == 375150 Assertion 'tres.status == VexTransOK' failed<br> == 378068 valgrind crashes on AVX2 function in FFmpeg<br> 376142 Segfaults on MIPS Cavium Octeon boards<br> 376279 disInstr(arm64): unhandled instruction 0xD50320FF<br> 376455 Solaris: unhandled syscall lgrpsys(180)<br> 376518 Solaris: unhandled fast trap getlgrp(6)<br> 376611 ppc64 and arm64 don't know about prlimit64 syscall<br> 376729 PPC64, remove R2 from the clobber list<br> == 371668<br> 376956 syswrap of SNDDRV and DRM_IOCTL_VERSION causing some addresses<br> to be wrongly marked as addressable<br> 377066 Some Valgrind unit tests fail to compile on Ubuntu 16.10 with<br> PIE enabled by default<br> 377376 memcheck/tests/linux/getregset fails with glibc2.24<br> 377427 PPC64, lxv instruction failing on odd destination register <br> 377478 PPC64: ISA 3.0 setup fixes<br> 377698 Missing memory check for futex() uaddr arg for FUTEX_WAKE<br> and FUTEX_WAKE_BITSET, check only 4 args for FUTEX_WAKE_BITSET,<br> and 2 args for FUTEX_TRYLOCK_PI<br> 377717 Fix massive space leak when reading compressed debuginfo sections<br> 377891 Update Xen 4.6 domctl wrappers<br> 377930 fcntl syscall wrapper is missing flock structure check<br> 378524 libvexmultiarch_test regression on s390x and ppc64<br> 378535 Valgrind reports INTERNAL ERROR in execve syscall wrapper<br> 378673 Update libiberty demangler<br> 378931 Add ISA 3.0B additional isnstructions, add OV32, CA32 setting support<br> 379039 syscall wrapper for prctl(PR_SET_NAME) must not check more than 16 bytes<br> 379094 Valgrind reports INTERNAL ERROR in rt_sigsuspend syscall wrapper<br> 379371 UNKNOWN task message [id 3444, to mach_task_self(), reply 0x603]<br> (task_register_dyld_image_infos)<br> 379372 UNKNOWN task message [id 3447, to mach_task_self(), reply 0x603]<br> (task_register_dyld_shared_cache_image_info)<br> 379390 unhandled syscall: mach:70 (host_create_mach_voucher_trap)<br> 379473 MIPS: add support for rdhwr cycle counter register<br> 379504 remove TileGX/Linux port<br> 379525 Support more x86 nop opcodes<br> 379838 disAMode(x86): not an addr!<br> 379703 PC ISA 3.0 fixes: stxvx, stxv, xscmpexpdp instructions<br> 379890 arm: unhandled instruction: 0xEBAD 0x1B05 (sub.w fp, sp, r5, lsl #4)<br> 379895 clock_gettime does not execute POST syscall wrapper<br> 379925 PPC64, mtffs does not set the FPCC and C bits in the FPSCR correctly<br> 379966 WARNING: unhandled amd64-linux syscall: 313 (finit_module)<br> 380200 xtree generated callgrind files refer to files without directory name<br> 380202 Assertion failure for cache line size (cls == 64) on aarch64.<br> 380397 s390x: __GI_strcspn() replacement needed<br> n-i-bz Fix pub_tool_basics.h build issue with g++ 4.4.7.<br> <br> (3.13.0.RC1: 2 June 2017, vex r3386, valgrind r16434)<br> (3.13.0.RC2: 9 June 2017, vex r3389, valgrind r16443)<br> (3.13.0: 14 June 2017, vex r3396, valgrind r16446)<br> <br> <br> <br> Release 3.12.0 (20 October 2016)<br> ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~<br> <br> 3.12.0 is a feature release with many improvements and the usual<br> collection of bug fixes.<br> <br> This release supports X86/Linux, AMD64/Linux, ARM32/Linux,<br> ARM64/Linux, PPC32/Linux, PPC64BE/Linux, PPC64LE/Linux, S390X/Linux,<br> MIPS32/Linux, MIPS64/Linux, ARM/Android, ARM64/Android,<br> MIPS32/Android, X86/Android, X86/Solaris, AMD64/Solaris, X86/MacOSX<br> 10.10 and AMD64/MacOSX 10.10. There is also preliminary support for<br> X86/MacOSX 10.11/12, AMD64/MacOSX 10.11/12 and TILEGX/Linux.<br> <br> * ================== PLATFORM CHANGES =================<br> <br> * POWER: Support for ISA 3.0 has been added<br> <br> * mips: support for O32 FPXX ABI has been added.<br> * mips: improved recognition of different processors<br> * mips: determination of page size now done at run time<br> <br> * amd64: Partial support for AMD FMA4 instructions.<br> <br> * arm, arm64: Support for v8 crypto and CRC instructions.<br> <br> * Improvements and robustification of the Solaris port.<br> <br> * Preliminary support for MacOS 10.12 (Sierra) has been added.<br> <br> Whilst 3.12.0 continues to support the 32-bit x86 instruction set, we<br> would prefer users to migrate to 64-bit x86 (a.k.a amd64 or x86_64)<br> where possible. Valgrind's support for 32-bit x86 has stagnated in<br> recent years and has fallen far behind that for 64-bit x86<br> instructions. By contrast 64-bit x86 is well supported, up to and<br> including AVX2.<br> <br> * ==================== TOOL CHANGES ====================<br> <br> * Memcheck:<br> <br> - Added meta mempool support for describing a custom allocator which:<br> - Auto-frees all chunks assuming that destroying a pool destroys all<br> objects in the pool<br> - Uses itself to allocate other memory blocks<br> <br> - New flag --ignore-range-below-sp to ignore memory accesses below<br> the stack pointer, if you really have to. The related flag<br> --workaround-gcc296-bugs=yes is now deprecated. Use<br> --ignore-range-below-sp=1024-1 as a replacement.<br> <br> * DRD:<br> <br> - Improved thread startup time significantly on non-Linux platforms.<br> <br> * DHAT<br> <br> - Added collection of the metric "tot-blocks-allocd"<br> <br> * ==================== OTHER CHANGES ====================<br> <br> * Replacement/wrapping of malloc/new related functions is now done not just<br> for system libraries by default, but for any globally defined malloc/new<br> related function (both in shared libraries and statically linked alternative<br> malloc implementations). The dynamic (runtime) linker is excluded, though.<br> To only intercept malloc/new related functions in<br> system libraries use --soname-synonyms=somalloc=nouserintercepts (where<br> "nouserintercepts" can be any non-existing library name).<br> This new functionality is not implemented for MacOS X.<br> <br> * The maximum number of callers in a suppression entry is now equal to<br> the maximum size for --num-callers (500).<br> Note that --gen-suppressions=yes|all similarly generates suppressions<br> containing up to --num-callers frames.<br> <br> * New and modified GDB server monitor features:<br> <br> - Valgrind's gdbserver now accepts the command 'catch syscall'.<br> Note that you must have GDB >= 7.11 to use 'catch syscall' with<br> gdbserver.<br> <br> * New option --run-cxx-freeres=<yes|no> can be used to change whether<br> __gnu_cxx::__freeres() cleanup function is called or not. Default is<br> 'yes'.<br> <br> * Valgrind is able to read compressed debuginfo sections in two formats:<br> - zlib ELF gABI format with SHF_COMPRESSED flag (gcc option -gz=zlib)<br> - zlib GNU format with .zdebug sections (gcc option -gz=zlib-gnu)<br> <br> * Modest JIT-cost improvements: the cost of instrumenting code blocks<br> for the most common use case (x86_64-linux, Memcheck) has been<br> reduced by 10%-15%.<br> <br> * Improved performance for programs that do a lot of discarding of<br> instruction address ranges of 8KB or less.<br> <br> * The C++ symbol demangler has been updated.<br> <br> * More robustness against invalid syscall parameters on Linux.<br> <br> * ==================== FIXED BUGS ====================<br> <br> The following bugs have been fixed or resolved. Note that "n-i-bz"<br> stands for "not in bugzilla" -- that is, a bug that was reported to us<br> but never got a bugzilla entry. We encourage you to file bugs in<br> bugzilla (https://bugs.kde.org/enter_bug.cgi?product=valgrind) rather<br> than mailing the developers (or mailing lists) directly -- bugs that<br> are not entered into bugzilla tend to get forgotten about or ignored.<br> <br> To see details of a given bug, visit<br> https://bugs.kde.org/show_bug.cgi?id=XXXXXX<br> where XXXXXX is the bug number as listed below.<br> <br> 191069 Exiting due to signal not reported in XML output<br> 199468 Suppressions: stack size limited to 25<br> while --num-callers allows more frames<br> 212352 vex amd64 unhandled opc_aux = 0x 2, first_opcode == 0xDC (FCOM)<br> 278744 cvtps2pd with redundant RexW<br> 303877 valgrind doesn't support compressed debuginfo sections.<br> 345307 Warning about "still reachable" memory when using libstdc++ from gcc 5<br> 348345 Assertion fails for negative lineno<br> 348924 MIPS: Load doubles through memory so the code compiles with the FPXX ABI<br> 351282 V 3.10.1 MIPS softfloat build broken with GCC 4.9.3 / binutils 2.25.1<br> 351692 Dumps created by valgrind are not readable by gdb (mips32 specific)<br> 351804 Crash on generating suppressions for "printf" call on OS X 10.10<br> 352197 mips: mmap2() not wrapped correctly for page size > 4096<br> 353083 arm64 doesn't implement various xattr system calls<br> 353084 arm64 doesn't support sigpending system call<br> 353137 www: update info for Supported Platforms<br> 353138 www: update "The Valgrind Developers" page<br> 353370 don't advertise RDRAND in cpuid for Core-i7-4910-like avx2 machine<br> == 365325<br> == 357873<br> 353384 amd64->IR: 0x66 0xF 0x3A 0x62 0xD1 0x62 (pcmpXstrX $0x62)<br> 353398 WARNING: unhandled amd64-solaris syscall: 207<br> 353660 XML in auxwhat tag not escaping reserved symbols properly<br> 353680 s390x: Crash with certain glibc versions due to non-implemented TBEGIN<br> 353727 amd64->IR: 0x66 0xF 0x3A 0x62 0xD1 0x72 (pcmpXstrX $0x72)<br> 353802 ELF debug info reader confused with multiple .rodata sections<br> 353891 Assert 'bad_scanned_addr < VG_ROUNDDN(start+len, sizeof(Addr))' failed<br> 353917 unhandled amd64-solaris syscall fchdir(120)<br> 353920 unhandled amd64-solaris syscall: 170<br> 354274 arm: unhandled instruction: 0xEBAD 0x0AC1 (sub.w sl, sp, r1, lsl #3)<br> 354392 unhandled amd64-solaris syscall: 171<br> 354797 Vbit test does not include Iops for Power 8 instruction support<br> 354883 tst->os_state.pthread - magic_delta assertion failure on OSX 10.11<br> == 361351<br> == 362920<br> == 366222<br> 354933 Fix documentation of --kernel-variant=android-no-hw-tls option<br> 355188 valgrind should intercept all malloc related global functions<br> 355454 do not intercept malloc related symbols from the runtime linker<br> 355455 stderr.exp of test cases wrapmalloc and wrapmallocstatic overconstrained<br> 356044 Dwarf line info reader misinterprets is_stmt register<br> 356112 mips: replace addi with addiu<br> 356393 valgrind (vex) crashes because isZeroU happened<br> == 363497<br> == 364497<br> 356676 arm64-linux: unhandled syscalls 125, 126 (sched_get_priority_max/min)<br> 356678 arm64-linux: unhandled syscall 232 (mincore)<br> 356817 valgrind.h triggers compiler errors on MSVC when defining NVALGRIND<br> 356823 Unsupported ARM instruction: stlex<br> 357059 x86/amd64: SSE cvtpi2ps with memory source does transition to MMX state<br> 357338 Unhandled instruction for SHA instructions libcrypto Boring SSL<br> 357673 crash if I try to run valgrind with a binary link with libcurl<br> 357833 Setting RLIMIT_DATA to zero breaks with linux 4.5+<br> 357871 pthread_spin_destroy not properly wrapped<br> 357887 Calls to VG_(fclose) do not close the file descriptor<br> 357932 amd64->IR: accept redundant REX prefixes for {minsd,maxsd} m128, xmm.<br> 358030 support direct socket calls on x86 32bit (new in linux 4.3)<br> 358478 drd/tests/std_thread.cpp doesn't build with GCC6<br> 359133 Assertion 'eltSzB <= ddpa->poolSzB' failed<br> 359181 Buffer Overflow during Demangling<br> 359201 futex syscall "skips" argument 5 if op is FUTEX_WAIT_BITSET<br> 359289 s390x: popcnt (B9E1) not implemented<br> 359472 The Power PC vsubuqm instruction doesn't always give the correct result<br> 359503 Add missing syscalls for aarch64 (arm64)<br> 359645 "You need libc6-dbg" help message could be more helpful<br> 359703 s390: wire up separate socketcalls system calls<br> 359724 getsockname might crash - deref_UInt should call safe_to_deref<br> 359733 amd64 implement ld.so strchr/index override like x86<br> 359767 Valgrind does not support the IBM POWER ISA 3.0 instructions, part 1/5<br> 359829 Power PC test suite none/tests/ppc64/test_isa_2_07.c uses<br> uninitialized data<br> 359838 arm64: Unhandled instruction 0xD5033F5F (clrex)<br> 359871 Incorrect mask handling in ppoll<br> 359952 Unrecognised PCMPESTRM variants (0x70, 0x19)<br> 360008 Contents of Power vr registers contents is not printed correctly when<br> the --vgdb-shadow-registers=yes option is used<br> 360035 POWER PC instruction bcdadd and bcdsubtract generate result with<br> non-zero shadow bits<br> 360378 arm64: Unhandled instruction 0x5E280844 (sha1h s4, s2)<br> 360425 arm64 unsupported instruction ldpsw<br> == 364435<br> 360519 none/tests/arm64/memory.vgtest might fail with newer gcc<br> 360571 Error about the Android Runtime reading below the stack pointer on ARM<br> 360574 Wrong parameter type for an ashmem ioctl() call on Android and ARM64<br> 360749 kludge for multiple .rodata sections on Solaris no longer needed<br> 360752 raise the number of reserved fds in m_main.c from 10 to 12<br> 361207 Valgrind does not support the IBM POWER ISA 3.0 instructions, part 2/5<br> 361226 s390x: risbgn (EC59) not implemented<br> 361253 [s390x] ex_clone.c:42: undefined reference to `pthread_create'<br> 361354 ppc64[le]: wire up separate socketcalls system calls<br> 361615 Inconsistent termination for multithreaded process terminated by signal<br> 361926 Unhandled Solaris syscall: sysfs(84)<br> 362009 V dumps core on unimplemented functionality before threads are created<br> 362329 Valgrind does not support the IBM POWER ISA 3.0 instructions, part 3/5<br> 362894 missing (broken) support for wbit field on mtfsfi instruction (ppc64)<br> 362935 [AsusWRT] Assertion 'sizeof(TTEntryC) <= 88' failed<br> 362953 Request for an update to the Valgrind Developers page<br> 363680 add renameat2() support<br> 363705 arm64 missing syscall name_to_handle_at and open_by_handle_at<br> 363714 ppc64 missing syscalls sync, waitid and name_to/open_by_handle_at<br> 363858 Valgrind does not support the IBM POWER ISA 3.0 instructions, part 4/5<br> 364058 clarify in manual limitations of array overruns detections<br> 364413 pselect sycallwrapper mishandles NULL sigmask<br> 364728 Power PC, missing support for several HW registers in<br> get_otrack_shadow_offset_wrk()<br> 364948 Valgrind does not support the IBM POWER ISA 3.0 instructions, part 5/5<br> 365273 Invalid write to stack location reported after signal handler runs<br> 365912 ppc64BE segfault during jm-insns test (RELRO)<br> 366079 FPXX Support for MIPS32 Valgrind<br> 366138 Fix configure errors out when using Xcode 8 (clang 8.0.0)<br> 366344 Multiple unhandled instruction for Aarch64<br> (0x0EE0E020, 0x1AC15800, 0x4E284801, 0x5E040023, 0x5E056060)<br> 367995 Integration of memcheck with custom memory allocator<br> 368120 x86_linux asm _start functions do not keep 16-byte aligned stack pointer<br> 368412 False positive result for altivec capability check<br> 368416 Add tc06_two_races_xml.exp output for ppc64<br> 368419 Perf Events ioctls not implemented<br> 368461 mmapunmap test fails on ppc64<br> 368823 run_a_thread_NORETURN assembly code typo for VGP_arm64_linux target<br> 369000 AMD64 fma4 instructions unsupported.<br> 369169 ppc64 fails jm_int_isa_2_07 test<br> 369175 jm_vec_isa_2_07 test crashes on ppc64<br> 369209 valgrind loops and eats up all memory if cwd doesn't exist.<br> 369356 pre_mem_read_sockaddr syscall wrapper can crash with bad sockaddr<br> 369359 msghdr_foreachfield can crash when handling bad iovec<br> 369360 Bad sigprocmask old or new sets can crash valgrind<br> 369361 vmsplice syscall wrapper crashes on bad iovec<br> 369362 Bad sigaction arguments crash valgrind<br> 369383 x86 sys_modify_ldt wrapper crashes on bad ptr<br> 369402 Bad set/get_thread_area pointer crashes valgrind<br> 369441 bad lvec argument crashes process_vm_readv/writev syscall wrappers<br> 369446 valgrind crashes on unknown fcntl command<br> 369439 S390x: Unhandled insns RISBLG/RISBHG and LDE/LDER <br> 369468 Remove quadratic metapool algorithm using VG_(HT_remove_at_Iter)<br> 370265 ISA 3.0 HW cap stuff needs updating<br> 371128 BCD add and subtract instructions on Power BE in 32-bit mode do not work<br> 372195 Power PC, xxsel instruction is not always recognized<br> <br> n-i-bz Fix incorrect (or infinite loop) unwind on RHEL7 x86 and amd64<br> n-i-bz massif --pages-as-heap=yes does not report peak caused by mmap+munmap<br> n-i-bz false positive leaks due to aspacemgr merging heap & non heap segments<br> n-i-bz Fix ppoll_alarm exclusion on OS X<br> n-i-bz Document brk segment limitation, reference manual in limit reached msg.<br> n-i-bz Fix clobber list in none/tests/amd64/xacq_xrel.c [valgrind r15737]<br> n-i-bz Bump allowed shift value for "add.w reg, sp, reg, lsl #N" [vex r3206]<br> n-i-bz amd64: memcheck false positive with shr %edx<br> n-i-bz arm3: Allow early writeback of SP base register in "strd rD, [sp, #-16]"<br> n-i-bz ppc: Fix two cases of PPCAvFpOp vs PPCFpOp enum confusion<br> n-i-bz arm: Fix incorrect register-number constraint check for LDAEX{,B,H,D}<br> n-i-bz DHAT: added collection of the metric "tot-blocks-allocd" <br> <br> (3.12.0.RC1: 20 October 2016, vex r3282, valgrind r16094)<br> (3.12.0.RC2: 20 October 2016, vex r3282, valgrind r16096)<br> (3.12.0: 21 October 2016, vex r3282, valgrind r16098)<br> <br> <br> <br> Release 3.11.0 (22 September 2015)<br> ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~<br> <br> 3.11.0 is a feature release with many improvements and the usual<br> collection of bug fixes.<br> <br> This release supports X86/Linux, AMD64/Linux, ARM32/Linux,<br> ARM64/Linux, PPC32/Linux, PPC64BE/Linux, PPC64LE/Linux, S390X/Linux,<br> MIPS32/Linux, MIPS64/Linux, ARM/Android, ARM64/Android,<br> MIPS32/Android, X86/Android, X86/Solaris, AMD64/Solaris, X86/MacOSX<br> 10.10 and AMD64/MacOSX 10.10. There is also preliminary support for<br> X86/MacOSX 10.11, AMD64/MacOSX 10.11 and TILEGX/Linux.<br> <br> * ================== PLATFORM CHANGES =================<br> <br> * Support for Solaris/x86 and Solaris/amd64 has been added.<br> <br> * Preliminary support for Mac OS X 10.11 (El Capitan) has been added.<br> <br> * Preliminary support for the Tilera TileGX architecture has been added.<br> <br> * s390x: It is now required for the host to have the "long displacement"<br> facility. The oldest supported machine model is z990.<br> <br> * x86: on an SSE2 only host, Valgrind in 32 bit mode now claims to be a<br> Pentium 4. 3.10.1 wrongly claimed to be a Core 2, which is SSSE3.<br> <br> * The JIT's register allocator is significantly faster, making the JIT<br> as a whole somewhat faster, so JIT-intensive activities, for example<br> program startup, are modestly faster, around 5%.<br> <br> * There have been changes to the default settings of several command<br> line flags, as detailed below.<br> <br> * Intel AVX2 support is more complete (64 bit targets only). On AVX2<br> capable hosts, the simulated CPUID will now indicate AVX2 support.<br> <br> * ==================== TOOL CHANGES ====================<br> <br> * Memcheck:<br> <br> - The default value for --leak-check-heuristics has been changed from<br> "none" to "all". This helps to reduce the number of possibly<br> lost blocks, in particular for C++ applications.<br> <br> - The default value for --keep-stacktraces has been changed from<br> "malloc-then-free" to "malloc-and-free". This has a small cost in<br> memory (one word per malloc-ed block) but allows Memcheck to show the<br> 3 stacktraces of a dangling reference: where the block was allocated,<br> where it was freed, and where it is acccessed after being freed.<br> <br> - The default value for --partial-loads-ok has been changed from "no" to <br> "yes", so as to avoid false positive errors resulting from some kinds<br> of vectorised loops.<br> <br> - A new monitor command 'xb <addr> <len>' shows the validity bits of<br> <len> bytes at <addr>. The monitor command 'xb' is easier to use<br> than get_vbits when you need to associate byte data value with<br> their corresponding validity bits.<br> <br> - The 'block_list' monitor command has been enhanced:<br> o it can print a range of loss records<br> o it now accepts an optional argument 'limited <max_blocks>'<br> to control the number of blocks printed.<br> o if a block has been found using a heuristic, then<br> 'block_list' now shows the heuristic after the block size.<br> o the loss records/blocks to print can be limited to the blocks<br> found via specified heuristics.<br> <br> - The C helper functions used to instrument loads on<br> x86-{linux,solaris} and arm-linux (both 32-bit only) have been<br> replaced by handwritten assembly sequences. This gives speedups<br> in the region of 0% to 7% for those targets only.<br> <br> - A new command line option, --expensive-definedness-checks=yes|no,<br> has been added. This is useful for avoiding occasional invalid<br> uninitialised-value errors in optimised code. Watch out for<br> runtime degradation, as this can be up to 25%. As always, though,<br> the slowdown is highly application specific. The default setting<br> is "no".<br> <br> * Massif:<br> <br> - A new monitor command 'all_snapshots <filename>' dumps all<br> snapshots taken so far.<br> <br> * Helgrind:<br> <br> - Significant memory reduction and moderate speedups for<br> --history-level=full for applications accessing a lot of memory<br> with many different stacktraces.<br> <br> - The default value for --conflict-cache-size=N has been doubled to<br> 2000000. Users that were not using the default value should<br> preferably also double the value they give.<br> <br> The default was changed due to the changes in the "full history"<br> implementation. Doubling the value gives on average a slightly more<br> complete history and uses similar memory (or significantly less memory<br> in the worst case) than the previous implementation.<br> <br> - The Helgrind monitor command 'info locks' now accepts an optional<br> argument 'lock_addr', which shows information about the lock at the<br> given address only.<br> <br> - When using --history-level=full, the new Helgrind monitor command<br> 'accesshistory <addr> [<len>]' will show the recorded accesses for<br> <len> (or 1) bytes at <addr>.<br> <br> * ==================== OTHER CHANGES ====================<br> <br> * The default value for the --smc-check option has been changed from<br> "stack" to "all-non-file" on targets that provide automatic D-I<br> cache coherence (x86, amd64 and s390x). The result is to provide,<br> by default, transparent support for JIT generated and self-modifying<br> code on all targets.<br> <br> * Mac OS X only: the default value for the --dsymutil option has been<br> changed from "no" to "yes", since any serious usage on Mac OS X<br> always required it to be "yes".<br> <br> * The command line options --db-attach and --db-command have been removed.<br> They were deprecated in 3.10.0.<br> <br> * When a process dies due to a signal, Valgrind now shows the signal<br> and the stacktrace at default verbosity (i.e. verbosity 1).<br> <br> * The address description logic used by Memcheck and Helgrind now<br> describes addresses in anonymous segments, file mmap-ed segments,<br> shared memory segments and the brk data segment.<br> <br> * The new option --error-markers=<begin>,<end> can be used to mark the<br> begin/end of errors in textual output mode, to facilitate<br> searching/extracting errors in output files that mix valgrind errors<br> with program output.<br> <br> * The new option --max-threads=<number> can be used to change the number<br> of threads valgrind can handle. The default is 500 threads which<br> should be more than enough for most applications.<br> <br> * The new option --valgrind-stacksize=<number> can be used to change the<br> size of the private thread stacks used by Valgrind. This is useful<br> for reducing memory use or increasing the stack size if Valgrind<br> segfaults due to stack overflow.<br> <br> * The new option --avg-transtab-entry-size=<number> can be used to specify<br> the expected instrumented block size, either to reduce memory use or<br> to avoid excessive retranslation.<br> <br> * Valgrind can be built with Intel's ICC compiler, version 14.0 or later.<br> <br> * New and modified GDB server monitor features:<br> <br> - When a signal is reported in GDB, you can now use the GDB convenience<br> variable $_siginfo to examine detailed signal information.<br> <br> - Valgrind's gdbserver now allows the user to change the signal<br> to deliver to the process. So, use 'signal SIGNAL' to continue execution<br> with SIGNAL instead of the signal reported to GDB. Use 'signal 0' to<br> continue without passing the signal to the process.<br> <br> - With GDB >= 7.10, the command 'target remote'<br> will automatically load the executable file of the process running<br> under Valgrind. This means you do not need to specify the executable<br> file yourself, GDB will discover it itself. See GDB documentation about<br> 'qXfer:exec-file:read' packet for more info.<br> <br> * ==================== FIXED BUGS ====================<br> <br> The following bugs have been fixed or resolved. Note that "n-i-bz"<br> stands for "not in bugzilla" -- that is, a bug that was reported to us<br> but never got a bugzilla entry. We encourage you to file bugs in<br> bugzilla (https://bugs.kde.org/enter_bug.cgi?product=valgrind) rather<br> than mailing the developers (or mailing lists) directly -- bugs that<br> are not entered into bugzilla tend to get forgotten about or ignored.<br> <br> To see details of a given bug, visit<br> https://bugs.kde.org/show_bug.cgi?id=XXXXXX<br> where XXXXXX is the bug number as listed below.<br> <br> 116002 VG_(printf): Problems with justification of strings and integers<br> 155125 avoid cutting away file:lineno after long function name<br> 197259 Unsupported arch_prtctl PR_SET_GS option<br> 201152 ppc64: Assertion in ppc32g_dirtyhelper_MFSPR_268_269<br> 201216 Fix Valgrind does not support pthread_sigmask() on OS X<br> 201435 Fix Darwin: -v does not show kernel version<br> 208217 "Warning: noted but unhandled ioctl 0x2000747b" on Mac OS X<br> 211256 Fixed an outdated comment regarding the default platform.<br> 211529 Incomplete call stacks for code compiled by newer versions of MSVC<br> 211926 Avoid compilation warnings in valgrind.h with -pedantic<br> 212291 Fix unhandled syscall: unix:132 (mkfifo) on OS X<br> == 263119<br> 226609 Crediting upstream authors in man page<br> 231257 Valgrind omits path when executing script from shebang line<br> 254164 OS X task_info: UNKNOWN task message [id 3405, to mach_task_self() [..]<br> 294065 Improve the pdb file reader by avoiding hardwired absolute pathnames<br> 269360 s390x: Fix addressing mode selection for compare-and-swap<br> 302630 Memcheck: Assertion failed: 'sizeof(UWord) == sizeof(UInt)'<br> == 326797<br> 312989 ioctl handling needs to do POST handling on generic ioctls and [..]<br> 319274 Fix unhandled syscall: unix:410 (sigsuspend_nocancel) on OS X<br> 324181 mmap does not handle MAP_32BIT (handle it now, rather than fail it)<br> 327745 Fix valgrind 3.9.0 build fails on Mac OS X 10.6.8<br> 330147 libmpiwrap PMPI_Get_count returns undefined value<br> 333051 mmap of huge pages fails due to incorrect alignment<br> == 339163<br> 334802 valgrind does not always explain why a given option is bad<br> 335618 mov.w rN, pc/sp (ARM32)<br> 335785 amd64->IR 0xC4 0xE2 0x75 0x2F (vmaskmovpd)<br> == 307399<br> == 343175<br> == 342740<br> == 346912<br> 335907 segfault when running wine's ddrawex/tests/surface.c under valgrind<br> 338602 AVX2 bit in CPUID missing<br> 338606 Strange message for scripts with invalid interpreter<br> 338731 ppc: Fix testuite build for toolchains not supporting -maltivec<br> 338995 shmat with hugepages (SHM_HUGETLB) fails with EINVAL<br> 339045 Getting valgrind to compile and run on OS X Yosemite (10.10)<br> == 340252<br> 339156 gdbsrv not called for fatal signal<br> 339215 Valgrind 3.10.0 contain 2013 in copyrights notice<br> 339288 support Cavium Octeon MIPS specific BBIT*32 instructions<br> 339636 Use fxsave64 and fxrstor64 mnemonics instead of old-school rex64 prefix<br> 339442 Fix testsuite build failure on OS X 10.9<br> 339542 Enable compilation with Intel's ICC compiler<br> 339563 The DVB demux DMX_STOP ioctl doesn't have a wrapper<br> 339688 Mac-specific ASM does not support .version directive (cpuid,<br> tronical and pushfpopf tests)<br> 339745 Valgrind crash when check Marmalade app (partial fix)<br> 339755 Fix known deliberate memory leak in setenv() on Mac OS X 10.9<br> 339778 Linux/TileGx platform support for Valgrind<br> 339780 Fix known uninitialised read in pthread_rwlock_init() on Mac OS X 10.9 <br> 339789 Fix none/tests/execve test on Mac OS X 10.9<br> 339808 Fix none/tests/rlimit64_nofile test on Mac OS X 10.9<br> 339820 vex amd64->IR: 0x66 0xF 0x3A 0x63 0xA 0x42 0x74 0x9 (pcmpistri $0x42)<br> 340115 Fix none/tests/cmdline[1|2] tests on systems which define TMPDIR<br> 340392 Allow user to select more accurate definedness checking in memcheck<br> to avoid invalid complaints on optimised code<br> 340430 Fix some grammatical weirdness in the manual.<br> 341238 Recognize GCC5/DWARFv5 DW_LANG constants (Go, C11, C++11, C++14)<br> 341419 Signal handler ucontext_t not filled out correctly on OS X<br> 341539 VG_(describe_addr) should not describe address as belonging to client<br> segment if it is past the heap end<br> 341613 Enable building of manythreads and thread-exits tests on Mac OS X<br> 341615 Fix none/tests/darwin/access_extended test on Mac OS X<br> 341698 Valgrind's AESKEYGENASSIST gives wrong result in words 0 and 2 [..]<br> 341789 aarch64: shmat fails with valgrind on ARMv8<br> 341997 MIPS64: Cavium OCTEON insns - immediate operand handled incorrectly<br> 342008 valgrind.h needs type cast [..] for clang/llvm in 64-bit mode<br> 342038 Unhandled syscalls on aarch64 (mbind/get/set_mempolicy)<br> 342063 wrong format specifier for test mcblocklistsearch in gdbserver_tests<br> 342117 Hang when loading PDB file for MSVC compiled Firefox under Wine<br> 342221 socket connect false positive uninit memory for unknown af family<br> 342353 Allow dumping full massif output while valgrind is still running<br> 342571 Valgrind chokes on AVX compare intrinsic with _CMP_GE_QS<br> == 346476<br> == 348387<br> == 350593<br> 342603 Add I2C_SMBUS ioctl support<br> 342635 OS X 10.10 (Yosemite) - missing system calls and fcntl code<br> 342683 Mark memory past the initial brk limit as unaddressable<br> 342783 arm: unhandled instruction 0xEEFE1ACA = "vcvt.s32.f32 s3, s3, #12"<br> 342795 Internal glibc __GI_mempcpy call should be intercepted<br> 342841 s390x: Support instructions fiebr(a) and fidbr(a)<br> 343012 Unhandled syscall 319 (memfd_create)<br> 343069 Patch updating v4l2 API support<br> 343173 helgrind crash during stack unwind<br> 343219 fix GET_STARTREGS for arm<br> 343303 Fix known deliberate memory leak in setenv() on Mac OS X 10.10<br> 343306 OS X 10.10: UNKNOWN mach_msg unhandled MACH_SEND_TRAILER option<br> 343332 Unhandled instruction 0x9E310021 (fcvtmu) on aarch64<br> 343335 unhandled instruction 0x1E638400 (fccmp) aarch64<br> 343523 OS X mach_ports_register: UNKNOWN task message [id 3403, to [..]<br> 343525 OS X host_get_special_port: UNKNOWN host message [id 412, to [..]<br> 343597 ppc64le: incorrect use of offseof macro<br> 343649 OS X host_create_mach_voucher: UNKNOWN host message [id 222, to [..]<br> 343663 OS X 10.10 Memchecj always reports a leak regardless of [..]<br> 343732 Unhandled syscall 144 (setgid) on aarch64<br> 343733 Unhandled syscall 187 (msgctl and related) on aarch64<br> 343802 s390x: False positive "conditional jump or move depends on [..]<br> 343902 --vgdb=yes doesn't break when --xml=yes is used<br> 343967 Don't warn about setuid/setgid/setcap executable for directories<br> 343978 Recognize DWARF5/GCC5 DW_LANG_Fortran 2003 and 2008 constants<br> 344007 accept4 syscall unhandled on arm64 (242) and ppc64 (344)<br> 344033 Helgrind on ARM32 loses track of mutex state in pthread_cond_wait<br> 344054 www - update info for Solaris/illumos<br> 344416 'make regtest' does not work cleanly on OS X<br> 344235 Remove duplicate include of pub_core_aspacemgr.h<br> 344279 syscall sendmmsg on arm64 (269) and ppc32/64 (349) unhandled<br> 344295 syscall recvmmsg on arm64 (243) and ppc32/64 (343) unhandled<br> 344307 2 unhandled syscalls on aarch64/arm64: umount2(39), mount (40)<br> 344314 callgrind_annotate ... warnings about commands containing newlines<br> 344318 socketcall should wrap recvmmsg and sendmmsg<br> 344337 Fix unhandled syscall: mach:41 (_kernelrpc_mach_port_guard_trap)<br> 344416 Fix 'make regtest' does not work cleanly on OS X<br> 344499 Fix compilation for Linux kernel >= 4.0.0<br> 344512 OS X: unhandled syscall: unix:348 (__pthread_chdir), <br> unix:349 (__pthread_fchdir)<br> 344559 Garbage collection of unused segment names in address space manager<br> 344560 Fix stack traces missing penultimate frame on OS X<br> 344621 Fix memcheck/tests/err_disable4 test on OS X<br> 344686 Fix suppression for pthread_rwlock_init on OS X 10.10<br> 344702 Fix missing libobjc suppressions on OS X 10.10<br> == 344543<br> 344936 Fix unhandled syscall: unix:473 (readlinkat) on OS X 10.10<br> 344939 Fix memcheck/tests/xml1 on OS X 10.10<br> 345016 helgrind/tests/locked_vs_unlocked2 is failing sometimes<br> 345079 Fix build problems in VEX/useful/test_main.c<br> 345126 Incorrect handling of VIDIOC_G_AUDIO and G_AUDOUT<br> 345177 arm64: prfm (reg) not implemented<br> 345215 Performance improvements for the register allocator<br> 345248 add support for Solaris OS in valgrind<br> 345338 TIOCGSERIAL and TIOCSSERIAL ioctl support on Linux<br> 345394 Fix memcheck/tests/strchr on OS X<br> 345637 Fix memcheck/tests/sendmsg on OS X<br> 345695 Add POWERPC support for AT_DCACHESIZE and HWCAP2<br> 345824 Fix aspacem segment mismatch: seen with none/tests/bigcode<br> 345887 Fix an assertion in the address space manager<br> 345928 amd64: callstack only contains current function for small stacks<br> 345984 disInstr(arm): unhandled instruction: 0xEE193F1E<br> 345987 MIPS64: Implement cavium LHX instruction<br> 346031 MIPS: Implement support for the CvmCount register (rhwr %0, 31)<br> 346185 Fix typo saving altivec register v24<br> 346267 Compiler warnings for PPC64 code on call to LibVEX_GuestPPC64_get_XER()<br> and LibVEX_GuestPPC64_get_CR()<br> 346270 Regression tests none/tests/jm_vec/isa_2_07 and<br> none/tests/test_isa_2_07_part2 have failures on PPC64 little endian<br> 346307 fuse filesystem syscall deadlocks<br> 346324 PPC64 missing support for lbarx, lharx, stbcx and sthcx instructions<br> 346411 MIPS: SysRes::_valEx handling is incorrect<br> 346416 Add support for LL_IOC_PATH2FID and LL_IOC_GETPARENT Lustre ioctls<br> 346474 PPC64 Power 8, spr TEXASRU register not supported<br> 346487 Compiler generates "note" about a future ABI change for PPC64<br> 346562 MIPS64: lwl/lwr instructions are performing 64bit loads<br> and causing spurious "invalid read of size 8" warnings<br> 346801 Fix link error on OS X: _vgModuleLocal_sf_maybe_extend_stack<br> 347151 Fix suppression for pthread_rwlock_init on OS X 10.8<br> 347233 Fix memcheck/tests/strchr on OS X 10.10 (Haswell) <br> 347322 Power PC regression test cleanup<br> 347379 valgrind --leak-check=full leak errors from system libs on OS X 10.8<br> == 217236<br> 347389 unhandled syscall: 373 (Linux ARM syncfs)<br> 347686 Patch set to cleanup PPC64 regtests<br> 347978 Remove bash dependencies where not needed<br> 347982 OS X: undefined symbols for architecture x86_64: "_global" [..]<br> 347988 Memcheck: the 'impossible' happened: unexpected size for Addr (OSX/wine)<br> == 345929<br> 348102 Patch updating v4l2 API support<br> 348247 amd64 front end: jno jumps wrongly when overflow is not set<br> 348269 Improve mmap MAP_HUGETLB support.<br> 348334 (ppc) valgrind does not simulate dcbfl - then my program terminates<br> 348345 Assertion fails for negative lineno<br> 348377 Unsupported ARM instruction: yield<br> 348565 Fix detection of command line option availability for clang<br> 348574 vex amd64->IR pcmpistri SSE4.2 unsupported (pcmpistri $0x18)<br> 348728 Fix broken check for VIDIOC_G_ENC_INDEX<br> 348748 Fix redundant condition<br> 348890 Fix clang warning about unsupported --param inline-unit-growth=900<br> 348949 Bogus "ERROR: --ignore-ranges: suspiciously large range"<br> 349034 Add Lustre ioctls LL_IOC_GROUP_LOCK and LL_IOC_GROUP_UNLOCK<br> 349086 Fix UNKNOWN task message [id 3406, to mach_task_self(), [..]<br> 349087 Fix UNKNOWN task message [id 3410, to mach_task_self(), [..]<br> 349626 Implemented additional Xen hypercalls<br> 349769 Clang/osx: ld: warning: -read_only_relocs cannot be used with x86_64<br> 349790 Clean up of the hardware capability checking utilities.<br> 349828 memcpy intercepts memmove causing src/dst overlap error (ppc64 ld.so)<br> 349874 Fix typos in source code<br> 349879 memcheck: add handwritten assembly for helperc_LOADV*<br> 349941 di_notify_mmap might create wrong start/size DebugInfoMapping<br> 350062 vex x86->IR: 0x66 0xF 0x3A 0xB (ROUNDSD) on OS X<br> 350202 Add limited param to 'monitor block_list'<br> 350290 s390x: Support instructions fixbr(a)<br> 350359 memcheck/tests/x86/fxsave hangs indefinetely on OS X<br> 350809 Fix none/tests/async-sigs for Solaris<br> 350811 Remove reference to --db-attach which has been removed.<br> 350813 Memcheck/x86: enable handwritten assembly helpers for x86/Solaris too<br> 350854 hard-to-understand code in VG_(load_ELF)()<br> 351140 arm64 syscalls setuid (146) and setresgid (149) not implemented<br> 351386 Solaris: Cannot run ld.so.1 under Valgrind<br> 351474 Fix VG_(iseqsigset) as obvious<br> 351531 Typo in /include/vki/vki-xen-physdev.h header guard<br> 351756 Intercept platform_memchr$VARIANT$Haswell on OS X<br> 351858 ldsoexec support on Solaris<br> 351873 Newer gcc doesn't allow __builtin_tabortdc[i] in ppc32 mode<br> 352130 helgrind reports false races for printfs using mempcpy on FILE* state<br> 352284 s390: Conditional jump depends on uninitialised value(s) in vfprintf <br> 352320 arm64 crash on none/tests/nestedfs<br> 352765 Vbit test fails on Power 6<br> 352768 The mbar instruction is missing from the Power PC support<br> 352769 Power PC program priority register (PPR) is not supported<br> n-i-bz Provide implementations of certain compiler builtins to support<br> compilers that may not provide those<br> n-i-bz Old STABS code is still being compiled, but never used. Remove it.<br> n-i-bz Fix compilation on distros with glibc < 2.5<br> n-i-bz (vex 3098) Avoid generation of Neon insns on non-Neon hosts<br> n-i-bz Enable rt_sigpending syscall on ppc64 linux.<br> n-i-bz mremap did not work properly on shared memory<br> n-i-bz Fix incorrect sizeof expression in syswrap-xen.c reported by Coverity<br> n-i-bz In VALGRIND_PRINTF write out thread name, if any, to xml<br> <br> (3.11.0.TEST1: 8 September 2015, vex r3187, valgrind r15646)<br> (3.11.0.TEST2: 21 September 2015, vex r3193, valgrind r15667)<br> (3.11.0: 22 September 2015, vex r3195, valgrind r15674)<br> <br> <br> <br> Release 3.10.1 (25 November 2014)<br> ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~<br> 3.10.1 is a bug fix release. It fixes various bugs reported in 3.10.0<br> and backports fixes for all reported missing AArch64 ARMv8 instructions<br> and syscalls from the trunk. If you package or deliver 3.10.0 for others<br> to use, you might want to consider upgrading to 3.10.1 instead.<br> <br> The following bugs have been fixed or resolved. Note that "n-i-bz"<br> stands for "not in bugzilla" -- that is, a bug that was reported to us<br> but never got a bugzilla entry. We encourage you to file bugs in<br> bugzilla (https://bugs.kde.org/enter_bug.cgi?product=valgrind) rather<br> than mailing the developers (or mailing lists) directly -- bugs that<br> are not entered into bugzilla tend to get forgotten about or ignored.<br> <br> To see details of a given bug, visit<br> https://bugs.kde.org/show_bug.cgi?id=XXXXXX<br> where XXXXXX is the bug number as listed below.<br> <br> 335440 arm64: ld1 (single structure) is not implemented<br> 335713 arm64: unhanded instruction: prfm (immediate)<br> 339020 ppc64: memcheck/tests/ppc64/power_ISA2_05 failing in nightly build<br> 339182 ppc64: AvSplat ought to load destination vector register with [..]<br> 339336 PPC64 store quad instruction (stq) is not supposed to change [..]<br> 339433 ppc64 lxvw4x instruction uses four 32-byte loads<br> 339645 Use correct tag names in sys_getdents/64 wrappers<br> 339706 Fix false positive for ioctl(TIOCSIG) on linux<br> 339721 assertion 'check_sibling == sibling' failed in readdwarf3.c ...<br> 339853 arm64 times syscall unknown<br> 339855 arm64 unhandled getsid/setsid syscalls<br> 339858 arm64 dmb sy not implemented<br> 339926 Unhandled instruction 0x1E674001 (frintx) on aarm64<br> 339927 Unhandled instruction 0x9E7100C6 (fcvtmu) on aarch64<br> 339938 disInstr(arm64): unhandled instruction 0x4F8010A4 (fmla)<br> == 339950<br> 339940 arm64: unhandled syscall: 83 (sys_fdatasync) + patch<br> 340033 arm64: unhandled insn dmb ishld and some other isb-dmb-dsb variants<br> 340028 unhandled syscalls for arm64 (msync, pread64, setreuid and setregid)<br> 340036 arm64: Unhandled instruction ld4 (multiple structures, no offset)<br> 340236 arm64: unhandled syscalls: mknodat, fchdir, chroot, fchownat<br> 340509 arm64: unhandled instruction fcvtas<br> 340630 arm64: fchmod (52) and fchown (55) syscalls not recognized<br> 340632 arm64: unhandled instruction fcvtas<br> 340722 Resolve "UNKNOWN attrlist flags 0:0x10000000"<br> 340725 AVX2: Incorrect decoding of vpbroadcast{b,w} reg,reg forms<br> 340788 warning: unhandled syscall: 318 (getrandom)<br> 340807 disInstr(arm): unhandled instruction: 0xEE989B20<br> 340856 disInstr(arm64): unhandled instruction 0x1E634C45 (fcsel)<br> 340922 arm64: unhandled getgroups/setgroups syscalls<br> 350251 Fix typo in VEX utility program (test_main.c).<br> 350407 arm64: unhandled instruction ucvtf (vector, integer)<br> 350809 none/tests/async-sigs breaks when run under cron on Solaris<br> 350811 update README.solaris after r15445<br> 350813 Use handwritten memcheck assembly helpers on x86/Solaris [..]<br> 350854 strange code in VG_(load_ELF)()<br> 351140 arm64 syscalls setuid (146) and setresgid (149) not implemented<br> n-i-bz DRD and Helgrind: Handle Imbe_CancelReservation (clrex on ARM)<br> n-i-bz Add missing ]] to terminate CDATA.<br> n-i-bz Glibc versions prior to 2.5 do not define PTRACE_GETSIGINFO<br> n-i-bz Enable sys_fadvise64_64 on arm32.<br> n-i-bz Add test cases for all remaining AArch64 SIMD, FP and memory insns.<br> n-i-bz Add test cases for all known arm64 load/store instructions.<br> n-i-bz PRE(sys_openat): when checking whether ARG1 == VKI_AT_FDCWD [..]<br> n-i-bz Add detection of old ppc32 magic instructions from bug 278808.<br> n-i-bz exp-dhat: Implement missing function "dh_malloc_usable_size".<br> n-i-bz arm64: Implement "fcvtpu w, s".<br> n-i-bz arm64: implement ADDP and various others<br> n-i-bz arm64: Implement {S,U}CVTF (scalar, fixedpt).<br> n-i-bz arm64: enable FCVT{A,N}S X,S.<br> <br> (3.10.1: 25 November 2014, vex r3026, valgrind r14785)<br> <br> <br> <br> Release 3.10.0 (10 September 2014)<br> ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~<br> <br> 3.10.0 is a feature release with many improvements and the usual<br> collection of bug fixes.<br> <br> This release supports X86/Linux, AMD64/Linux, ARM32/Linux, ARM64/Linux,<br> PPC32/Linux, PPC64BE/Linux, PPC64LE/Linux, S390X/Linux, MIPS32/Linux,<br> MIPS64/Linux, ARM/Android, MIPS32/Android, X86/Android, X86/MacOSX 10.9<br> and AMD64/MacOSX 10.9. Support for MacOSX 10.8 and 10.9 is<br> significantly improved relative to the 3.9.0 release.<br> <br> * ================== PLATFORM CHANGES =================<br> <br> * Support for the 64-bit ARM Architecture (AArch64 ARMv8). This port<br> is mostly complete, and is usable, but some SIMD instructions are as<br> yet unsupported.<br> <br> * Support for little-endian variant of the 64-bit POWER architecture.<br> <br> * Support for Android on MIPS32.<br> <br> * Support for 64bit FPU on MIPS32 platforms.<br> <br> * Both 32- and 64-bit executables are supported on MacOSX 10.8 and 10.9.<br> <br> * Configuration for and running on Android targets has changed.<br> See README.android in the source tree for details.<br> <br> * ================== DEPRECATED FEATURES =================<br> <br> * --db-attach is now deprecated and will be removed in the next<br> valgrind feature release. The built-in GDB server capabilities are<br> superior and should be used instead. Learn more here:<br> http://valgrind.org/docs/manual/manual-core-adv.html#manual-core-adv.gdbserver<br> <br> * ==================== TOOL CHANGES ====================<br> <br> * Memcheck:<br> <br> - Client code can now selectively disable and re-enable reporting of<br> invalid address errors in specific ranges using the new client<br> requests VALGRIND_DISABLE_ADDR_ERROR_REPORTING_IN_RANGE and<br> VALGRIND_ENABLE_ADDR_ERROR_REPORTING_IN_RANGE.<br> <br> - Leak checker: there is a new leak check heuristic called<br> "length64". This is used to detect interior pointers pointing 8<br> bytes inside a block, on the assumption that the first 8 bytes<br> holds the value "block size - 8". This is used by<br> sqlite3MemMalloc, for example.<br> <br> - Checking of system call parameters: if a syscall parameter<br> (e.g. bind struct sockaddr, sendmsg struct msghdr, ...) has<br> several fields not initialised, an error is now reported for each<br> field. Previously, an error was reported only for the first<br> uninitialised field.<br> <br> - Mismatched alloc/free checking: a new flag<br> --show-mismatched-frees=no|yes [yes] makes it possible to turn off<br> such checks if necessary.<br> <br> * Helgrind:<br> <br> - Improvements to error messages:<br> <br> o Race condition error message involving heap allocated blocks also<br> show the thread number that allocated the raced-on block.<br> <br> o All locks referenced by an error message are now announced.<br> Previously, some error messages only showed the lock addresses.<br> <br> o The message indicating where a lock was first observed now also<br> describes the address/location of the lock.<br> <br> - Helgrind now understands the Ada task termination rules and<br> creates a happens-before relationship between a terminated task<br> and its master. This avoids some false positives and avoids a big<br> memory leak when a lot of Ada tasks are created and terminated.<br> The interceptions are only activated with forthcoming releases of<br> gnatpro >= 7.3.0w-20140611 and gcc >= 5.0.<br> <br> - A new GDB server monitor command "info locks" giving the list of<br> locks, their location, and their status.<br> <br> * Callgrind:<br> <br> - callgrind_control now supports the --vgdb-prefix argument,<br> which is needed if valgrind was started with this same argument.<br> <br> * ==================== OTHER CHANGES ====================<br> <br> * Unwinding through inlined function calls. Stack unwinding can now<br> make use of Dwarf3 inlined-unwind information if it is available.<br> The practical effect is that inlined calls become visible in stack<br> traces. The suppression matching machinery has been adjusted<br> accordingly. This is controlled by the new option<br> --read-inline-info=yes|no. Currently this is enabled by default<br> only on Linux and Android targets and only for the tools Memcheck,<br> Helgrind and DRD.<br> <br> * Valgrind can now read EXIDX unwind information on 32-bit ARM<br> targets. If an object contains both CFI and EXIDX unwind<br> information, Valgrind will prefer the CFI over the EXIDX. This<br> facilitates unwinding through system libraries on arm-android<br> targets.<br> <br> * Address description logic has been improved and is now common<br> between Memcheck and Helgrind, resulting in better address<br> descriptions for some kinds of error messages.<br> <br> * Error messages about dubious arguments (eg, to malloc or calloc) are<br> output like other errors. This means that they can be suppressed<br> and they have a stack trace.<br> <br> * The C++ demangler has been updated for better C++11 support.<br> <br> * New and modified GDB server monitor features:<br> <br> - Thread local variables/storage (__thread) can now be displayed.<br> <br> - The GDB server monitor command "v.info location <address>"<br> displays information about an address. The information produced<br> depends on the tool and on the options given to valgrind.<br> Possibly, the following are described: global variables, local<br> (stack) variables, allocated or freed blocks, ...<br> <br> - The option "--vgdb-stop-at=event1,event2,..." allows the user to<br> ask the GDB server to stop at the start of program execution, at<br> the end of the program execution and on Valgrind internal errors.<br> <br> - A new monitor command "v.info stats" shows various Valgrind core<br> and tool statistics.<br> <br> - A new monitor command "v.set hostvisibility" allows the GDB server<br> to provide access to Valgrind internal host status/memory.<br> <br> * A new option "--aspace-minaddr=<address>" can in some situations<br> allow the use of more memory by decreasing the address above which<br> Valgrind maps memory. It can also be used to solve address<br> conflicts with system libraries by increasing the default value.<br> See user manual for details.<br> <br> * The amount of memory used by Valgrind to store debug info (unwind<br> info, line number information and symbol data) has been<br> significantly reduced, even though Valgrind now reads more<br> information in order to support unwinding of inlined function calls.<br> <br> * Dwarf3 handling with --read-var-info=yes has been improved:<br> <br> - Ada and C struct containing VLAs no longer cause a "bad DIE" error<br> <br> - Code compiled with<br> -ffunction-sections -fdata-sections -Wl,--gc-sections<br> no longer causes assertion failures.<br> <br> * Improved checking for the --sim-hints= and --kernel-variant=<br> options. Unknown strings are now detected and reported to the user<br> as a usage error.<br> <br> * The semantics of stack start/end boundaries in the valgrind.h<br> VALGRIND_STACK_REGISTER client request has been clarified and<br> documented. The convention is that start and end are respectively<br> the lowest and highest addressable bytes of the stack.<br> <br> * ==================== FIXED BUGS ====================<br> <br> The following bugs have been fixed or resolved. Note that "n-i-bz"<br> stands for "not in bugzilla" -- that is, a bug that was reported to us<br> but never got a bugzilla entry. We encourage you to file bugs in<br> bugzilla (https://bugs.kde.org/enter_bug.cgi?product=valgrind) rather<br> than mailing the developers (or mailing lists) directly -- bugs that<br> are not entered into bugzilla tend to get forgotten about or ignored.<br> <br> To see details of a given bug, visit<br> https://bugs.kde.org/show_bug.cgi?id=XXXXXX<br> where XXXXXX is the bug number as listed below.<br> <br> 175819 Support for ipv6 socket reporting with --track-fds<br> 232510 make distcheck fails<br> 249435 Analyzing wine programs with callgrind triggers a crash<br> 278972 support for inlined function calls in stacktraces and suppression<br> == 199144<br> 291310 FXSAVE instruction marks memory as undefined on amd64<br> 303536 ioctl for SIOCETHTOOL (ethtool(8)) isn't wrapped<br> 308729 vex x86->IR: unhandled instruction bytes 0xf 0x5 (syscall) <br> 315199 vgcore file for threaded app does not show which thread crashed<br> 315952 tun/tap ioctls are not supported<br> 323178 Unhandled instruction: PLDW register (ARM) <br> 323179 Unhandled instruction: PLDW immediate (ARM)<br> 324050 Helgrind: SEGV because of unaligned stack when using movdqa<br> 325110 Add test-cases for Power ISA 2.06 insns: divdo/divdo. and divduo/divduo.<br> 325124 [MIPSEL] Compilation error<br> 325477 Phase 4 support for IBM Power ISA 2.07<br> 325538 cavium octeon mips64, valgrind reported "dumping core" [...]<br> 325628 Phase 5 support for IBM Power ISA 2.07<br> 325714 Empty vgcore but RLIMIT_CORE is big enough (too big) <br> 325751 Missing the two privileged Power PC Transactional Memory Instructions<br> 325816 Phase 6 support for IBM Power ISA 2.07<br> 325856 Make SGCheck fail gracefully on unsupported platforms<br> 326026 Iop names for count leading zeros/sign bits incorrectly imply [..]<br> 326436 DRD: False positive in libstdc++ std::list::push_back<br> 326444 Cavium MIPS Octeon Specific Load Indexed Instructions<br> 326462 Refactor vgdb to isolate invoker stuff into separate module<br> 326469 amd64->IR: 0x66 0xF 0x3A 0x63 0xC1 0xE (pcmpistri 0x0E)<br> 326623 DRD: false positive conflict report in a field assignment<br> 326724 Valgrind does not compile on OSX 1.9 Mavericks<br> 326816 Intercept for __strncpy_sse2_unaligned missing?<br> 326921 coregrind fails to compile m_trampoline.S with MIPS/Linux port of V<br> 326983 Clear direction flag after tests on amd64.<br> 327212 Do not prepend the current directory to absolute path names.<br> 327223 Support for Cavium MIPS Octeon Atomic and Count Instructions<br> 327238 Callgrind Assertion 'passed <= last_bb->cjmp_count' failed<br> 327284 s390x: Fix translation of the risbg instruction<br> 327639 vex amd64->IR pcmpestri SSE4.2 instruction is unsupported 0x34<br> 327837 dwz compressed alternate .debug_info and .debug_str not read correctly<br> 327916 DW_TAG_typedef may have no name<br> 327943 s390x: add a redirection for the 'index' function<br> 328100 XABORT not implemented<br> 328205 Implement additional Xen hypercalls<br> 328454 add support Backtraces with ARM unwind tables (EXIDX)<br> 328455 s390x: SIGILL after emitting wrong register pair for ldxbr<br> 328711 valgrind.1 manpage "memcheck options" section is badly generated<br> 328878 vex amd64->IR pcmpestri SSE4.2 instruction is unsupported 0x14<br> 329612 Incorrect handling of AT_BASE for image execution <br> 329694 clang warns about using uninitialized variable <br> 329956 valgrind crashes when lmw/stmw instructions are used on ppc64<br> 330228 mmap must align to VKI_SHMLBA on mips32<br> 330257 LLVM does not support `-mno-dynamic-no-pic` option<br> 330319 amd64->IR: unhandled instruction bytes: 0xF 0x1 0xD5 (xend)<br> 330459 --track-fds=yes doesn't track eventfds<br> 330469 Add clock_adjtime syscall support<br> 330594 Missing sysalls on PowerPC / uClibc<br> 330622 Add test to regression suite for POWER instruction: dcbzl<br> 330939 Support for AMD's syscall instruction on x86<br> == 308729<br> 330941 Typo in PRE(poll) syscall wrapper<br> 331057 unhandled instruction: 0xEEE01B20 (vfma.f64) (has patch)<br> 331254 Fix expected output for memcheck/tests/dw4<br> 331255 Fix race condition in test none/tests/coolo_sigaction<br> 331257 Fix type of jump buffer in test none/tests/faultstatus<br> 331305 configure uses bash specific syntax<br> 331337 s390x WARNING: unhandled syscall: 326 (dup3)<br> 331380 Syscall param timer_create(evp) points to uninitialised byte(s)<br> 331476 Patch to handle ioctl 0x5422 on Linux (x86 and amd64)<br> 331829 Unexpected ioctl opcode sign extension<br> 331830 ppc64: WARNING: unhandled syscall: 96/97<br> 331839 drd/tests/sem_open specifies invalid semaphore name <br> 331847 outcome of drd/tests/thread_name is nondeterministic<br> 332037 Valgrind cannot handle Thumb "add pc, reg"<br> 332055 drd asserts on platforms with VG_STACK_REDZONE_SZB == 0 and<br> consistency checks enabled<br> 332263 intercepts for pthread_rwlock_timedrdlock and<br> pthread_rwlock_timedwrlock are incorrect<br> 332265 drd could do with post-rwlock_init and pre-rwlock_destroy<br> client requests<br> 332276 Implement additional Xen hypercalls<br> 332658 ldrd.w r1, r2, [PC, #imm] does not adjust for 32bit alignment<br> 332765 Fix ms_print to create temporary files in a proper directory<br> 333072 drd: Add semaphore annotations<br> 333145 Tests for missaligned PC+#imm access for arm<br> 333228 AAarch64 Missing instruction encoding: mrs %[reg], ctr_el0<br> 333230 AAarch64 missing instruction encodings: dc, ic, dsb.<br> 333248 WARNING: unhandled syscall: unix:443<br> 333428 ldr.w pc [rD, #imm] instruction leads to assertion<br> 333501 cachegrind: assertion: Cache set count is not a power of two.<br> == 336577<br> == 292281<br> 333666 Recognize MPX instructions and bnd prefix.<br> 333788 Valgrind does not support the CDROM_DISC_STATUS ioctl (has patch)<br> 333817 Valgrind reports the memory areas written to by the SG_IO<br> ioctl as untouched<br> 334049 lzcnt fails silently (x86_32)<br> 334384 Valgrind does not have support Little Endian support for<br> IBM POWER PPC 64<br> 334585 recvmmsg unhandled (+patch) (arm)<br> 334705 sendmsg and recvmsg should guard against bogus msghdr fields.<br> 334727 Build fails with -Werror=format-security<br> 334788 clarify doc about --log-file initial program directory<br> 334834 PPC64 Little Endian support, patch 2<br> 334836 PPC64 Little Endian support, patch 3 testcase fixes<br> 334936 patch to fix false positives on alsa SNDRV_CTL_* ioctls<br> 335034 Unhandled ioctl: HCIGETDEVLIST<br> 335155 vgdb, fix error print statement.<br> 335262 arm64: movi 8bit version is not supported<br> 335263 arm64: dmb instruction is not implemented<br> 335441 unhandled ioctl 0x8905 (SIOCATMARK) when running wine under valgrind<br> 335496 arm64: sbc/abc instructions are not implemented<br> 335554 arm64: unhandled instruction: abs<br> 335564 arm64: unhandled instruction: fcvtpu Xn, Sn<br> 335735 arm64: unhandled instruction: cnt<br> 335736 arm64: unhandled instruction: uaddlv<br> 335848 arm64: unhandled instruction: {s,u}cvtf<br> 335902 arm64: unhandled instruction: sli<br> 335903 arm64: unhandled instruction: umull (vector)<br> 336055 arm64: unhandled instruction: mov (element)<br> 336062 arm64: unhandled instruction: shrn{,2}<br> 336139 mip64: [...] valgrind hangs and spins on a single core [...]<br> 336189 arm64: unhandled Instruction: mvn<br> 336435 Valgrind hangs in pthread_spin_lock consuming 100% CPU<br> 336619 valgrind --read-var-info=yes doesn't handle DW_TAG_restrict_type<br> 336772 Make moans about unknown ioctls more informative<br> 336957 Add a section about the Solaris/illumos port on the webpage<br> 337094 ifunc wrapper is broken on ppc64<br> 337285 fcntl commands F_OFD_SETLK, F_OFD_SETLKW, and F_OFD_GETLK not supported<br> 337528 leak check heuristic for block prefixed by length as 64bit number<br> 337740 Implement additional Xen hypercalls<br> 337762 guest_arm64_toIR.c:4166 (dis_ARM64_load_store): Assertion `0' failed.<br> 337766 arm64-linux: unhandled syscalls mlock (228) and mlockall (230)<br> 337871 deprecate --db-attach<br> 338023 Add support for all V4L2/media ioctls<br> 338024 inlined functions are not shown if DW_AT_ranges is used<br> 338106 Add support for 'kcmp' syscall<br> 338115 DRD: computed conflict set differs from actual after fork<br> 338160 implement display of thread local storage in gdbsrv<br> 338205 configure.ac and check for -Wno-tautological-compare<br> 338300 coredumps are missing one byte of every segment<br> 338445 amd64 vbit-test fails with unknown opcodes used by arm64 VEX<br> 338499 --sim-hints parsing broken due to wrong order in tokens<br> 338615 suppress glibc 2.20 optimized strcmp implementation for ARMv7<br> 338681 Unable to unwind through clone thread created on i386-linux<br> 338698 race condition between gdbsrv and vgdb on startup<br> 338703 helgrind on arm-linux gets false positives in dynamic loader<br> 338791 alt dwz files can be relative of debug/main file<br> 338878 on MacOS: assertion 'VG_IS_PAGE_ALIGNED(clstack_end+1)' failed<br> 338932 build V-trunk with gcc-trunk<br> 338974 glibc 2.20 changed size of struct sigaction sa_flags field on s390<br> 345079 Fix build problems in VEX/useful/test_main.c<br> n-i-bz Fix KVM_CREATE_IRQCHIP ioctl handling<br> n-i-bz s390x: Fix memory corruption for multithreaded applications<br> n-i-bz vex arm->IR: allow PC as basereg in some LDRD cases<br> n-i-bz internal error in Valgrind if vgdb transmit signals when ptrace invoked<br> n-i-bz Fix mingw64 support in valgrind.h (dev@, 9 May 2014)<br> n-i-bz drd manual: Document how to C++11 programs that use class "std::thread"<br> n-i-bz Add command-line option --default-suppressions<br> n-i-bz Add support for BLKDISCARDZEROES ioctl<br> n-i-bz ppc32/64: fix a regression with the mtfsb0/mtfsb1 instructions<br> n-i-bz Add support for sys_pivot_root and sys_unshare<br> <br> (3.10.0.BETA1: 2 September 2014, vex r2940, valgrind r14428)<br> (3.10.0.BETA2: 8 September 2014, vex r2950, valgrind r14503)<br> (3.10.0: 10 September 2014, vex r2950, valgrind r14514)<br> <br> <br> <br> Release 3.9.0 (31 October 2013)<br> ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~<br> 3.9.0 is a feature release with many improvements and the usual<br> collection of bug fixes.<br> <br> This release supports X86/Linux, AMD64/Linux, ARM/Linux, PPC32/Linux,<br> PPC64/Linux, S390X/Linux, MIPS32/Linux, MIPS64/Linux, ARM/Android,<br> X86/Android, X86/MacOSX 10.7 and AMD64/MacOSX 10.7. Support for<br> MacOSX 10.8 is significantly improved relative to the 3.8.0 release.<br> <br> * ================== PLATFORM CHANGES =================<br> <br> * Support for MIPS64 LE and BE running Linux. Valgrind has been<br> tested on MIPS64 Debian Squeeze and Debian Wheezy distributions.<br> <br> * Support for MIPS DSP ASE on MIPS32 platforms.<br> <br> * Support for s390x Decimal Floating Point instructions on hosts that<br> have the DFP facility installed.<br> <br> * Support for POWER8 (Power ISA 2.07) instructions<br> <br> * Support for Intel AVX2 instructions. This is available only on 64<br> bit code.<br> <br> * Initial support for Intel Transactional Synchronization Extensions,<br> both RTM and HLE.<br> <br> * Initial support for Hardware Transactional Memory on POWER.<br> <br> * Improved support for MacOSX 10.8 (64-bit only). Memcheck can now<br> run large GUI apps tolerably well.<br> <br> * ==================== TOOL CHANGES ====================<br> <br> * Memcheck:<br> <br> - Improvements in handling of vectorised code, leading to<br> significantly fewer false error reports. You need to use the flag<br> --partial-loads-ok=yes to get the benefits of these changes.<br> <br> - Better control over the leak checker. It is now possible to<br> specify which leak kinds (definite/indirect/possible/reachable)<br> should be displayed, which should be regarded as errors, and which<br> should be suppressed by a given leak suppression. This is done<br> using the options --show-leak-kinds=kind1,kind2,..,<br> --errors-for-leak-kinds=kind1,kind2,.. and an optional<br> "match-leak-kinds:" line in suppression entries, respectively.<br> <br> Note that generated leak suppressions contain this new line and<br> are therefore more specific than in previous releases. To get the<br> same behaviour as previous releases, remove the "match-leak-kinds:"<br> line from generated suppressions before using them.<br> <br> - Reduced "possible leak" reports from the leak checker by the use<br> of better heuristics. The available heuristics provide detection<br> of valid interior pointers to std::stdstring, to new[] allocated<br> arrays with elements having destructors and to interior pointers<br> pointing to an inner part of a C++ object using multiple<br> inheritance. They can be selected individually using the<br> option --leak-check-heuristics=heur1,heur2,...<br> <br> - Better control of stacktrace acquisition for heap-allocated<br> blocks. Using the --keep-stacktraces option, it is possible to<br> control independently whether a stack trace is acquired for each<br> allocation and deallocation. This can be used to create better<br> "use after free" errors or to decrease Valgrind's resource<br> consumption by recording less information.<br> <br> - Better reporting of leak suppression usage. The list of used<br> suppressions (shown when the -v option is given) now shows, for<br> each leak suppressions, how many blocks and bytes it suppressed<br> during the last leak search.<br> <br> * Helgrind:<br> <br> - False errors resulting from the use of statically initialised<br> mutexes and condition variables (PTHREAD_MUTEX_INITIALISER, etc)<br> have been removed.<br> <br> - False errors resulting from the use of pthread_cond_waits that<br> timeout, have been removed.<br> <br> * ==================== OTHER CHANGES ====================<br> <br> * Some attempt to tune Valgrind's space requirements to the expected<br> capabilities of the target:<br> <br> - The default size of the translation cache has been reduced from 8<br> sectors to 6 on Android platforms, since each sector occupies<br> about 40MB when using Memcheck.<br> <br> - The default size of the translation cache has been increased to 16<br> sectors on all other platforms, reflecting the fact that large<br> applications require instrumentation and storage of huge amounts<br> of code. For similar reasons, the number of memory mapped<br> segments that can be tracked has been increased by a factor of 6.<br> <br> - In all cases, the maximum number of sectors in the translation<br> cache can be controlled by the new flag --num-transtab-sectors.<br> <br> * Changes in how debug info (line numbers, etc) is read:<br> <br> - Valgrind no longer temporarily mmaps the entire object to read<br> from it. Instead, reading is done through a small fixed sized<br> buffer. This avoids virtual memory usage spikes when Valgrind<br> reads debuginfo from large shared objects.<br> <br> - A new experimental remote debug info server. Valgrind can read<br> debug info from a different machine (typically, a build host)<br> where debuginfo objects are stored. This can save a lot of time<br> and hassle when running Valgrind on resource-constrained targets<br> (phones, tablets) when the full debuginfo objects are stored<br> somewhere else. This is enabled by the --debuginfo-server=<br> option.<br> <br> - Consistency checking between main and debug objects can be<br> disabled using the --allow-mismatched-debuginfo option.<br> <br> * Stack unwinding by stack scanning, on ARM. Unwinding by stack<br> scanning can recover stack traces in some cases when the normal<br> unwind mechanisms fail. Stack scanning is best described as "a<br> nasty, dangerous and misleading hack" and so is disabled by default.<br> Use --unw-stack-scan-thresh and --unw-stack-scan-frames to enable<br> and control it.<br> <br> * Detection and merging of recursive stack frame cycles. When your<br> program has recursive algorithms, this limits the memory used by<br> Valgrind for recorded stack traces and avoids recording<br> uninteresting repeated calls. This is controlled by the command<br> line option --merge-recursive-frame and by the monitor command<br> "v.set merge-recursive-frames".<br> <br> * File name and line numbers for used suppressions. The list of used<br> suppressions (shown when the -v option is given) now shows, for each<br> used suppression, the file name and line number where the suppression<br> is defined.<br> <br> * New and modified GDB server monitor features:<br> <br> - valgrind.h has a new client request, VALGRIND_MONITOR_COMMAND,<br> that can be used to execute gdbserver monitor commands from the<br> client program.<br> <br> - A new monitor command, "v.info open_fds", that gives the list of<br> open file descriptors and additional details.<br> <br> - An optional message in the "v.info n_errs_found" monitor command,<br> for example "v.info n_errs_found test 1234 finished", allowing a<br> comment string to be added to the process output, perhaps for the<br> purpose of separating errors of different tests or test phases.<br> <br> - A new monitor command "v.info execontext" that shows information<br> about the stack traces recorded by Valgrind.<br> <br> - A new monitor command "v.do expensive_sanity_check_general" to run<br> some internal consistency checks.<br> <br> * New flag --sigill-diagnostics to control whether a diagnostic<br> message is printed when the JIT encounters an instruction it can't<br> translate. The actual behavior -- delivery of SIGILL to the<br> application -- is unchanged.<br> <br> * The maximum amount of memory that Valgrind can use on 64 bit targets<br> has been increased from 32GB to 64GB. This should make it possible<br> to run applications on Memcheck that natively require up to about 35GB.<br> <br> * ==================== FIXED BUGS ====================<br> <br> The following bugs have been fixed or resolved. Note that "n-i-bz"<br> stands for "not in bugzilla" -- that is, a bug that was reported to us<br> but never got a bugzilla entry. We encourage you to file bugs in<br> bugzilla (https://bugs.kde.org/enter_bug.cgi?product=valgrind) rather<br> than mailing the developers (or mailing lists) directly -- bugs that<br> are not entered into bugzilla tend to get forgotten about or ignored.<br> <br> To see details of a given bug, visit<br> https://bugs.kde.org/show_bug.cgi?id=XXXXXX<br> where XXXXXX is the bug number as listed below.<br> <br> 123837 system call: 4th argument is optional, depending on cmd<br> 135425 memcheck should tell you where Freed blocks were Mallocd<br> 164485 VG_N_SEGNAMES and VG_N_SEGMENTS are (still) too small<br> 207815 Adds some of the drm ioctls to syswrap-linux.c <br> 251569 vex amd64->IR: 0xF 0x1 0xF9 0xBF 0x90 0xD0 0x3 0x0 (RDTSCP)<br> 252955 Impossible to compile with ccache<br> 253519 Memcheck reports auxv pointer accesses as invalid reads.<br> 263034 Crash when loading some PPC64 binaries<br> 269599 Increase deepest backtrace<br> 274695 s390x: Support "compare to/from logical" instructions (z196)<br> 275800 s390x: Autodetect cache info (part 2)<br> 280271 Valgrind reports possible memory leaks on still-reachable std::string<br> 284540 Memcheck shouldn't count suppressions matching still-reachable [..]<br> 289578 Backtraces with ARM unwind tables (stack scan flags)<br> 296311 Wrong stack traces due to -fomit-frame-pointer (x86) <br> 304832 ppc32: build failure<br> 305431 Use find_buildid shdr fallback for separate .debug files<br> 305728 Add support for AVX2 instructions<br> 305948 ppc64: code generation for ShlD64 / ShrD64 asserts<br> 306035 s390x: Fix IR generation for LAAG and friends<br> 306054 s390x: Condition code computation for convert-to-int/logical<br> 306098 s390x: alternate opcode form for convert to/from fixed<br> 306587 Fix cache line detection from auxiliary vector for PPC.<br> 306783 Mips unhandled syscall : 4025 / 4079 / 4182<br> 307038 DWARF2 CFI reader: unhandled DW_OP_ opcode 0x8 (DW_OP_const1u et al)<br> 307082 HG false positive: pthread_cond_destroy: destruction of unknown CV<br> 307101 sys_capget second argument can be NULL<br> 307103 sys_openat: If pathname is absolute, then dirfd is ignored.<br> 307106 amd64->IR: f0 0f c0 02 (lock xadd byte)<br> 307113 s390x: DFP support<br> 307141 valgrind does't work in mips-linux system<br> 307155 filter_gdb should filter out syscall-template.S T_PSEUDO<br> 307285 x86_amd64 feature test for avx in test suite is wrong<br> 307290 memcheck overlap testcase needs memcpy version filter<br> 307463 Please add "&limit=0" to the "all open bugs" link<br> 307465 --show-possibly-lost=no should reduce the error count / exit code<br> 307557 Leaks on Mac OS X 10.7.5 libraries at ImageLoader::recursiveInit[..]<br> 307729 pkgconfig support broken valgrind.pc<br> 307828 Memcheck false errors SSE optimized wcscpy, wcscmp, wcsrchr, wcschr<br> 307955 Building valgrind 3.7.0-r4 fails in Gentoo AMD64 when using clang<br> 308089 Unhandled syscall on ppc64: prctl<br> 308135 PPC32 MPC8xx has 16 bytes cache size<br> 308321 testsuite memcheck filter interferes with gdb_filter <br> 308333 == 307106<br> 308341 vgdb should report process exit (or fatal signal)<br> 308427 s390 memcheck reports tsearch cjump/cmove depends on uninit<br> 308495 Remove build dependency on installed Xen headers<br> 308573 Internal error on 64-bit instruction executed in 32-bit mode<br> 308626 == 308627<br> 308627 pmovmskb validity bit propagation is imprecise<br> 308644 vgdb command for having the info for the track-fds option<br> 308711 give more info about aspacemgr and arenas in out_of_memory<br> 308717 ARM: implement fixed-point VCVT.F64.[SU]32<br> 308718 ARM implement SMLALBB family of instructions<br> 308886 Missing support for PTRACE_SET/GETREGSET <br> 308930 syscall name_to_handle_at (303 on amd64) not handled<br> 309229 V-bit tester does not report number of tests generated<br> 309323 print unrecognized instuction on MIPS<br> 309425 Provide a --sigill-diagnostics flag to suppress illegal [..]<br> 309427 SSE optimized stpncpy trigger uninitialised value [..] errors<br> 309430 Self hosting ppc64 encounters a vassert error on operand type<br> 309600 valgrind is a bit confused about 0-sized sections<br> 309823 Generate errors for still reachable blocks<br> 309921 PCMPISTRI validity bit propagation is imprecise<br> 309922 none/tests/ppc64/test_dfp5 sometimes fails<br> 310169 The Iop_CmpORD class of Iops is not supported by the vbit checker.<br> 310424 --read-var-info does not properly describe static variables <br> 310792 search additional path for debug symbols<br> 310931 s390x: Message-security assist (MSA) instruction extension [..]<br> 311100 PPC DFP implementation of the integer operands is inconsistent [..]<br> 311318 ARM: "128-bit constant is not implemented" error message<br> 311407 ssse3 bcopy (actually converted memcpy) causes invalid read [..]<br> 311690 V crashes because it redirects branches inside of a redirected function<br> 311880 x86_64: make regtest hangs at shell_valid1<br> 311922 WARNING: unhandled syscall: 170<br> 311933 == 251569<br> 312171 ppc: insn selection for DFP<br> 312571 Rounding mode call wrong for the DFP Iops [..]<br> 312620 Change to Iop_D32toD64 [..] for s390 DFP support broke ppc [..]<br> 312913 Dangling pointers error should also report the alloc stack trace<br> 312980 Building on Mountain Lion generates some compiler warnings<br> 313267 Adding MIPS64/Linux port to Valgrind<br> 313348 == 251569<br> 313354 == 251569<br> 313811 Buffer overflow in assert_fail<br> 314099 coverity pointed out error in VEX guest_ppc_toIR.c insn_suffix<br> 314269 ppc: dead code in insn selection<br> 314718 ARM: implement integer divide instruction (sdiv and udiv)<br> 315345 cl-format.xml and callgrind/dump.c don't agree on using cfl= or cfi=<br> 315441 sendmsg syscall should ignore unset msghdr msg_flags<br> 315534 msgrcv inside a thread causes valgrind to hang (block)<br> 315545 Assertion '(UChar*)sec->tt[tteNo].tcptr <= (UChar*)hcode' failed<br> 315689 disInstr(thumb): unhandled instruction: 0xF852 0x0E10 (LDRT)<br> 315738 disInstr(arm): unhandled instruction: 0xEEBE0BEE (vcvt.s32.f64)<br> 315959 valgrind man page has bogus SGCHECK (and no BBV) OPTIONS section<br> 316144 valgrind.1 manpage contains unknown ??? strings [..]<br> 316145 callgrind command line options in manpage reference (unknown) [..]<br> 316145 callgrind command line options in manpage reference [..]<br> 316181 drd: Fixed a 4x slowdown for certain applications<br> 316503 Valgrind does not support SSE4 "movntdqa" instruction<br> 316535 Use of |signed int| instead of |size_t| in valgrind messages<br> 316696 fluidanimate program of parsec 2.1 stuck <br> 316761 syscall open_by_handle_at (304 on amd64, 342 on x86) not handled<br> 317091 Use -Wl,-Ttext-segment when static linking if possible [..]<br> 317186 "Impossible happens" when occurs VCVT instruction on ARM<br> 317318 Support for Threading Building Blocks "scalable_malloc"<br> 317444 amd64->IR: 0xC4 0x41 0x2C 0xC2 0xD2 0x8 (vcmpeq_uqps)<br> 317461 Fix BMI assembler configure check and avx2/bmi/fma vgtest prereqs<br> 317463 bmi testcase IR SANITY CHECK FAILURE<br> 317506 memcheck/tests/vbit-test fails with unknown opcode after [..]<br> 318050 libmpiwrap fails to compile with out-of-source build<br> 318203 setsockopt handling needs to handle SOL_SOCKET/SO_ATTACH_FILTER<br> 318643 annotate_trace_memory tests infinite loop on arm and ppc [..]<br> 318773 amd64->IR: 0xF3 0x48 0x0F 0xBC 0xC2 0xC3 0x66 0x0F<br> 318929 Crash with: disInstr(thumb): 0xF321 0x0001 (ssat16)<br> 318932 Add missing PPC64 and PPC32 system call support<br> 319235 --db-attach=yes is broken with Yama (ptrace scoping) enabled<br> 319395 Crash with unhandled instruction on STRT (Thumb) instructions<br> 319494 VEX Makefile-gcc standalone build update after r2702<br> 319505 [MIPSEL] Crash: unhandled UNRAY operator.<br> 319858 disInstr(thumb): unhandled instruction on instruction STRBT<br> 319932 disInstr(thumb): unhandled instruction on instruction STRHT<br> 320057 Problems when we try to mmap more than 12 memory pages on MIPS32<br> 320063 Memory from PTRACE_GET_THREAD_AREA is reported uninitialised<br> 320083 disInstr(thumb): unhandled instruction on instruction LDRBT<br> 320116 bind on AF_BLUETOOTH produces warnings because of sockaddr_rc padding<br> 320131 WARNING: unhandled syscall: 369 on ARM (prlimit64)<br> 320211 Stack buffer overflow in ./coregrind/m_main.c with huge TMPDIR<br> 320661 vgModuleLocal_read_elf_debug_info(): "Assertion '!di->soname'<br> 320895 add fanotify support (patch included)<br> 320998 vex amd64->IR pcmpestri and pcmpestrm SSE4.2 instruction<br> 321065 Valgrind updates for Xen 4.3<br> 321148 Unhandled instruction: PLI (Thumb 1, 2, 3)<br> 321363 Unhandled instruction: SSAX (ARM + Thumb)<br> 321364 Unhandled instruction: SXTAB16 (ARM + Thumb)<br> 321466 Unhandled instruction: SHASX (ARM + Thumb)<br> 321467 Unhandled instruction: SHSAX (ARM + Thumb)<br> 321468 Unhandled instruction: SHSUB16 (ARM + Thumb)<br> 321619 Unhandled instruction: SHSUB8 (ARM + Thumb)<br> 321620 Unhandled instruction: UASX (ARM + Thumb)<br> 321621 Unhandled instruction: USAX (ARM + Thumb)<br> 321692 Unhandled instruction: UQADD16 (ARM + Thumb)<br> 321693 Unhandled instruction: LDRSBT (Thumb)<br> 321694 Unhandled instruction: UQASX (ARM + Thumb)<br> 321696 Unhandled instruction: UQSAX (Thumb + ARM)<br> 321697 Unhandled instruction: UHASX (ARM + Thumb)<br> 321703 Unhandled instruction: UHSAX (ARM + Thumb)<br> 321704 Unhandled instruction: REVSH (ARM + Thumb)<br> 321730 Add cg_diff and cg_merge man pages<br> 321738 Add vgdb and valgrind-listener man pages<br> 321814 == 315545<br> 321891 Unhandled instruction: LDRHT (Thumb)<br> 321960 pthread_create() then alloca() causing invalid stack write errors<br> 321969 ppc32 and ppc64 don't support [lf]setxattr<br> 322254 Show threadname together with tid if set by application<br> 322294 Add initial support for IBM Power ISA 2.07<br> 322368 Assertion failure in wqthread_hijack under OS X 10.8<br> 322563 vex mips->IR: 0x70 0x83 0xF0 0x3A<br> 322807 VALGRIND_PRINTF_BACKTRACE writes callstack to xml and text to stderr<br> 322851 0bXXX binary literal syntax is not standard <br> 323035 Unhandled instruction: LDRSHT(Thumb)<br> 323036 Unhandled instruction: SMMLS (ARM and Thumb)<br> 323116 The memcheck/tests/ppc64/power_ISA2_05.c fails to build [..]<br> 323175 Unhandled instruction: SMLALD (ARM + Thumb)<br> 323177 Unhandled instruction: SMLSLD (ARM + Thumb)<br> 323432 Calling pthread_cond_destroy() or pthread_mutex_destroy() [..]<br> 323437 Phase 2 support for IBM Power ISA 2.07<br> 323713 Support mmxext (integer sse) subset on i386 (athlon)<br> 323803 Transactional memory instructions are not supported for Power<br> 323893 SSE3 not available on amd cpus in valgrind<br> 323905 Probable false positive from Valgrind/drd on close()<br> 323912 valgrind.h header isn't compatible for mingw64<br> 324047 Valgrind doesn't support [LDR,ST]{S}[B,H]T ARM instructions<br> 324149 helgrind: When pthread_cond_timedwait returns ETIMEDOUT [..]<br> 324181 mmap does not handle MAP_32BIT<br> 324227 memcheck false positive leak when a thread calls exit+block [..]<br> 324421 Support for fanotify API on ARM architecture<br> 324514 gdbserver monitor cmd output behaviour consistency [..]<br> 324518 ppc64: Emulation of dcbt instructions does not handle [..]<br> 324546 none/tests/ppc32 test_isa_2_07_part2 requests -m64<br> 324582 When access is made to freed memory, report both allocation [..]<br> 324594 Fix overflow computation for Power ISA 2.06 insns: mulldo/mulldo.<br> 324765 ppc64: illegal instruction when executing none/tests/ppc64/jm-misc<br> 324816 Incorrect VEX implementation for xscvspdp/xvcvspdp for SNaN inputs<br> 324834 Unhandled instructions in Microsoft C run-time for x86_64<br> 324894 Phase 3 support for IBM Power ISA 2.07<br> 326091 drd: Avoid false race reports from optimized strlen() impls<br> 326113 valgrind libvex hwcaps error on AMD64 <br> n-i-bz Some wrong command line options could be ignored<br> n-i-bz patch to allow fair-sched on android<br> n-i-bz report error for vgdb snapshot requested before execution<br> n-i-bz same as 303624 (fixed in 3.8.0), but for x86 android<br> <br> (3.9.0: 31 October 2013, vex r2796, valgrind r13708)<br> <br> <br> <br> Release 3.8.1 (19 September 2012)<br> ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~<br> 3.8.1 is a bug fix release. It fixes some assertion failures in 3.8.0<br> that occur moderately frequently in real use cases, adds support for<br> some missing instructions on ARM, and fixes a deadlock condition on<br> MacOSX. If you package or deliver 3.8.0 for others to use, you might<br> want to consider upgrading to 3.8.1 instead.<br> <br> The following bugs have been fixed or resolved. Note that "n-i-bz"<br> stands for "not in bugzilla" -- that is, a bug that was reported to us<br> but never got a bugzilla entry. We encourage you to file bugs in<br> bugzilla (https://bugs.kde.org/enter_bug.cgi?product=valgrind) rather<br> than mailing the developers (or mailing lists) directly -- bugs that<br> are not entered into bugzilla tend to get forgotten about or ignored.<br> <br> To see details of a given bug, visit<br> https://bugs.kde.org/show_bug.cgi?id=XXXXXX<br> where XXXXXX is the bug number as listed below.<br> <br> 284004 == 301281<br> 289584 Unhandled instruction: 0xF 0x29 0xE5 (MOVAPS)<br> 295808 amd64->IR: 0xF3 0xF 0xBC 0xC0 (TZCNT)<br> 298281 wcslen causes false(?) uninitialised value warnings<br> 301281 valgrind hangs on OS X when the process calls system()<br> 304035 disInstr(arm): unhandled instruction 0xE1023053<br> 304867 implement MOVBE instruction in x86 mode<br> 304980 Assertion 'lo <= hi' failed in vgModuleLocal_find_rx_mapping<br> 305042 amd64: implement 0F 7F encoding of movq between two registers<br> 305199 ARM: implement QDADD and QDSUB<br> 305321 amd64->IR: 0xF 0xD 0xC (prefetchw)<br> 305513 killed by fatal signal: SIGSEGV<br> 305690 DRD reporting invalid semaphore when sem_trywait fails<br> 305926 Invalid alignment checks for some AVX instructions<br> 306297 disInstr(thumb): unhandled instruction 0xE883 0x000C<br> 306310 3.8.0 release tarball missing some files<br> 306612 RHEL 6 glibc-2.X default suppressions need /lib*/libc-*patterns<br> 306664 vex amd64->IR: 0x66 0xF 0x3A 0x62 0xD1 0x46 0x66 0xF<br> n-i-bz shmat of a segment > 4Gb does not work <br> n-i-bz simulate_control_c script wrong USR1 signal number on mips<br> n-i-bz vgdb ptrace calls wrong on mips [...]<br> n-i-bz Fixes for more MPI false positives<br> n-i-bz exp-sgcheck's memcpy causes programs to segfault<br> n-i-bz OSX build w/ clang: asserts at startup<br> n-i-bz Incorrect undef'dness prop for Iop_DPBtoBCD and Iop_BCDtoDPB<br> n-i-bz fix a couple of union tag-vs-field mixups<br> n-i-bz OSX: use __NR_poll_nocancel rather than __NR_poll<br> <br> The following bugs were fixed in 3.8.0 but not listed in this NEWS<br> file at the time:<br> <br> 254088 Valgrind should know about UD2 instruction<br> 301280 == 254088<br> 301902 == 254088<br> 304754 NEWS blows TeX's little mind<br> <br> (3.8.1: 19 September 2012, vex r2537, valgrind r12996)<br> <br> <br> <br> Release 3.8.0 (10 August 2012)<br> ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~<br> 3.8.0 is a feature release with many improvements and the usual<br> collection of bug fixes.<br> <br> This release supports X86/Linux, AMD64/Linux, ARM/Linux, PPC32/Linux,<br> PPC64/Linux, S390X/Linux, MIPS/Linux, ARM/Android, X86/Android,<br> X86/MacOSX 10.6/10.7 and AMD64/MacOSX 10.6/10.7. Support for recent<br> distros and toolchain components (glibc 2.16, gcc 4.7) has been added.<br> There is initial support for MacOSX 10.8, but it is not usable for<br> serious work at present.<br> <br> * ================== PLATFORM CHANGES =================<br> <br> * Support for MIPS32 platforms running Linux. Valgrind has been<br> tested on MIPS32 and MIPS32r2 platforms running different Debian<br> Squeeze and MeeGo distributions. Both little-endian and big-endian<br> cores are supported. The tools Memcheck, Massif and Lackey have<br> been tested and are known to work. See README.mips for more details.<br> <br> * Preliminary support for Android running on x86.<br> <br> * Preliminary (as-yet largely unusable) support for MacOSX 10.8.<br> <br> * Support for Intel AVX instructions and for AES instructions. This<br> support is available only for 64 bit code.<br> <br> * Support for POWER Decimal Floating Point instructions.<br> <br> * ==================== TOOL CHANGES ====================<br> <br> * Non-libc malloc implementations are now supported. This is useful<br> for tools that replace malloc (Memcheck, Massif, DRD, Helgrind).<br> Using the new option --soname-synonyms, such tools can be informed<br> that the malloc implementation is either linked statically into the<br> executable, or is present in some other shared library different<br> from libc.so. This makes it possible to process statically linked<br> programs, and programs using other malloc libraries, for example<br> TCMalloc or JEMalloc.<br> <br> * For tools that provide their own replacement for malloc et al, the<br> option --redzone-size=<number> allows users to specify the size of<br> the padding blocks (redzones) added before and after each client<br> allocated block. Smaller redzones decrease the memory needed by<br> Valgrind. Bigger redzones increase the chance to detect blocks<br> overrun or underrun. Prior to this change, the redzone size was<br> hardwired to 16 bytes in Memcheck.<br> <br> * Memcheck:<br> <br> - The leak_check GDB server monitor command now can<br> control the maximum nr of loss records to output.<br> <br> - Reduction of memory use for applications allocating<br> many blocks and/or having many partially defined bytes.<br> <br> - Addition of GDB server monitor command 'block_list' that lists<br> the addresses/sizes of the blocks of a leak search loss record.<br> <br> - Addition of GDB server monitor command 'who_points_at' that lists<br> the locations pointing at a block.<br> <br> - If a redzone size > 0 is given, VALGRIND_MALLOCLIKE_BLOCK now will<br> detect an invalid access of these redzones, by marking them<br> noaccess. Similarly, if a redzone size is given for a memory<br> pool, VALGRIND_MEMPOOL_ALLOC will mark the redzones no access.<br> This still allows to find some bugs if the user has forgotten to<br> mark the pool superblock noaccess.<br> <br> - Performance of memory leak check has been improved, especially in<br> cases where there are many leaked blocks and/or many suppression<br> rules used to suppress leak reports.<br> <br> - Reduced noise (false positive) level on MacOSX 10.6/10.7, due to<br> more precise analysis, which is important for LLVM/Clang<br> generated code. This is at the cost of somewhat reduced<br> performance. Note there is no change to analysis precision or<br> costs on Linux targets.<br> <br> * DRD:<br> <br> - Added even more facilities that can help finding the cause of a data<br> race, namely the command-line option --ptrace-addr and the macro<br> DRD_STOP_TRACING_VAR(x). More information can be found in the manual.<br> <br> - Fixed a subtle bug that could cause false positive data race reports.<br> <br> * ==================== OTHER CHANGES ====================<br> <br> * The C++ demangler has been updated so as to work well with C++ <br> compiled by up to at least g++ 4.6.<br> <br> * Tool developers can make replacement/wrapping more flexible thanks<br> to the new option --soname-synonyms. This was reported above, but<br> in fact is very general and applies to all function<br> replacement/wrapping, not just to malloc-family functions.<br> <br> * Round-robin scheduling of threads can be selected, using the new<br> option --fair-sched= yes. Prior to this change, the pipe-based<br> thread serialisation mechanism (which is still the default) could<br> give very unfair scheduling. --fair-sched=yes improves<br> responsiveness of interactive multithreaded applications, and<br> improves repeatability of results from the thread checkers Helgrind<br> and DRD.<br> <br> * For tool developers: support to run Valgrind on Valgrind has been<br> improved. We can now routinely Valgrind on Helgrind or Memcheck.<br> <br> * gdbserver now shows the float shadow registers as integer<br> rather than float values, as the shadow values are mostly<br> used as bit patterns.<br> <br> * Increased limit for the --num-callers command line flag to 500.<br> <br> * Performance improvements for error matching when there are many<br> suppression records in use.<br> <br> * Improved support for DWARF4 debugging information (bug 284184).<br> <br> * Initial support for DWZ compressed Dwarf debug info.<br> <br> * Improved control over the IR optimiser's handling of the tradeoff<br> between performance and precision of exceptions. Specifically,<br> --vex-iropt-precise-memory-exns has been removed and replaced by<br> --vex-iropt-register-updates, with extended functionality. This<br> allows the Valgrind gdbserver to always show up to date register<br> values to GDB.<br> <br> * Modest performance gains through the use of translation chaining for<br> JIT-generated code.<br> <br> * ==================== FIXED BUGS ====================<br> <br> The following bugs have been fixed or resolved. Note that "n-i-bz"<br> stands for "not in bugzilla" -- that is, a bug that was reported to us<br> but never got a bugzilla entry. We encourage you to file bugs in<br> bugzilla (https://bugs.kde.org/enter_bug.cgi?product=valgrind) rather<br> than mailing the developers (or mailing lists) directly -- bugs that<br> are not entered into bugzilla tend to get forgotten about or ignored.<br> <br> To see details of a given bug, visit<br> https://bugs.kde.org/show_bug.cgi?id=XXXXXX<br> where XXXXXX is the bug number as listed below.<br> <br> 197914 Building valgrind from svn now requires automake-1.10<br> 203877 increase to 16Mb maximum allowed alignment for memalign et al<br> 219156 Handle statically linked malloc or other malloc lib (e.g. tcmalloc) <br> 247386 make perf does not run all performance tests<br> 270006 Valgrind scheduler unfair <br> 270777 Adding MIPS/Linux port to Valgrind<br> 270796 s390x: Removed broken support for the TS insn<br> 271438 Fix configure for proper SSE4.2 detection<br> 273114 s390x: Support TR, TRE, TROO, TROT, TRTO, and TRTT instructions<br> 273475 Add support for AVX instructions<br> 274078 improved configure logic for mpicc<br> 276993 fix mremap 'no thrash checks' <br> 278313 Fedora 15/x64: err read debug info with --read-var-info=yes flag<br> 281482 memcheck incorrect byte allocation count in realloc() for silly argument<br> 282230 group allocator for small fixed size, use it for MC_Chunk/SEc vbit<br> 283413 Fix wrong sanity check<br> 283671 Robustize alignment computation in LibVEX_Alloc<br> 283961 Adding support for some HCI IOCTLs<br> 284124 parse_type_DIE: confused by: DWARF 4<br> 284864 == 273475 (Add support for AVX instructions)<br> 285219 Too-restrictive constraints for Thumb2 "SP plus/minus register"<br> 285662 (MacOSX): Memcheck needs to replace memcpy/memmove<br> 285725 == 273475 (Add support for AVX instructions)<br> 286261 add wrapper for linux I2C_RDWR ioctl<br> 286270 vgpreload is not friendly to 64->32 bit execs, gives ld.so warnings<br> 286374 Running cachegrind with --branch-sim=yes on 64-bit PowerPC program fails<br> 286384 configure fails "checking for a supported version of gcc"<br> 286497 == 273475 (Add support for AVX instructions)<br> 286596 == 273475 (Add support for AVX instructions)<br> 286917 disInstr(arm): unhandled instruction: QADD (also QSUB)<br> 287175 ARM: scalar VFP fixed-point VCVT instructions not handled<br> 287260 Incorrect conditional jump or move depends on uninitialised value(s)<br> 287301 vex amd64->IR: 0x66 0xF 0x38 0x41 0xC0 0xB8 0x0 0x0 (PHMINPOSUW)<br> 287307 == 273475 (Add support for AVX instructions)<br> 287858 VG_(strerror): unknown error <br> 288298 (MacOSX) unhandled syscall shm_unlink<br> 288995 == 273475 (Add support for AVX instructions)<br> 289470 Loading of large Mach-O thin binaries fails.<br> 289656 == 273475 (Add support for AVX instructions)<br> 289699 vgdb connection in relay mode erroneously closed due to buffer overrun <br> 289823 == 293754 (PCMPxSTRx not implemented for 16-bit characters)<br> 289839 s390x: Provide support for unicode conversion instructions<br> 289939 monitor cmd 'leak_check' with details about leaked or reachable blocks<br> 290006 memcheck doesn't mark %xmm as initialized after "pcmpeqw %xmm %xmm"<br> 290655 Add support for AESKEYGENASSIST instruction <br> 290719 valgrind-3.7.0 fails with automake-1.11.2 due to"pkglibdir" usage<br> 290974 vgdb must align pages to VKI_SHMLBA (16KB) on ARM <br> 291253 ES register not initialised in valgrind simulation<br> 291568 Fix 3DNOW-related crashes with baseline x86_64 CPU (w patch)<br> 291865 s390x: Support the "Compare Double and Swap" family of instructions<br> 292300 == 273475 (Add support for AVX instructions)<br> 292430 unrecognized instruction in __intel_get_new_mem_ops_cpuid<br> 292493 == 273475 (Add support for AVX instructions)<br> 292626 Missing fcntl F_SETOWN_EX and F_GETOWN_EX support<br> 292627 Missing support for some SCSI ioctls<br> 292628 none/tests/x86/bug125959-x86.c triggers undefined behavior<br> 292841 == 273475 (Add support for AVX instructions)<br> 292993 implement the getcpu syscall on amd64-linux<br> 292995 Implement the “cross memory attach” syscalls introduced in Linux 3.2<br> 293088 Add some VEX sanity checks for ppc64 unhandled instructions<br> 293751 == 290655 (Add support for AESKEYGENASSIST instruction)<br> 293754 PCMPxSTRx not implemented for 16-bit characters<br> 293755 == 293754 (No tests for PCMPxSTRx on 16-bit characters)<br> 293808 CLFLUSH not supported by latest VEX for amd64<br> 294047 valgrind does not correctly emulate prlimit64(..., RLIMIT_NOFILE, ...)<br> 294048 MPSADBW instruction not implemented<br> 294055 regtest none/tests/shell fails when locale is not set to C<br> 294185 INT 0x44 (and others) not supported on x86 guest, but used by Jikes RVM<br> 294190 --vgdb-error=xxx can be out of sync with errors shown to the user<br> 294191 amd64: fnsave/frstor and 0x66 size prefixes on FP instructions<br> 294260 disInstr_AMD64: disInstr miscalculated next %rip<br> 294523 --partial-loads-ok=yes causes false negatives<br> 294617 vex amd64->IR: 0x66 0xF 0x3A 0xDF 0xD1 0x1 0xE8 0x6A<br> 294736 vex amd64->IR: 0x48 0xF 0xD7 0xD6 0x48 0x83<br> 294812 patch allowing to run (on x86 at least) helgrind/drd on tool.<br> 295089 can not annotate source for both helgrind and drd<br> 295221 POWER Processor decimal floating point instruction support missing<br> 295427 building for i386 with clang on darwin11 requires "-new_linker linker"<br> 295428 coregrind/m_main.c has incorrect x86 assembly for darwin<br> 295590 Helgrind: Assertion 'cvi->nWaiters > 0' failed<br> 295617 ARM - Add some missing syscalls<br> 295799 Missing \n with get_vbits in gdbserver when line is % 80 [...]<br> 296229 Linux user input device ioctls missing wrappers<br> 296318 ELF Debug info improvements (more than one rx/rw mapping)<br> 296422 Add translation chaining support<br> 296457 vex amd64->IR: 0x66 0xF 0x3A 0xDF 0xD1 0x1 0xE8 0x6A (dup of AES)<br> 296792 valgrind 3.7.0: add SIOCSHWTSTAMP (0x89B0) ioctl wrapper<br> 296983 Fix build issues on x86_64/ppc64 without 32-bit toolchains<br> 297078 gdbserver signal handling problems [..]<br> 297147 drd false positives on newly allocated memory<br> 297329 disallow decoding of IBM Power DFP insns on some machines<br> 297497 POWER Processor decimal floating point instruction support missing<br> 297701 Another alias for strncasecmp_l in libc-2.13.so<br> 297911 'invalid write' not reported when using APIs for custom mem allocators.<br> 297976 s390x: revisit EX implementation<br> 297991 Valgrind interferes with mmap()+ftell() <br> 297992 Support systems missing WIFCONTINUED (e.g. pre-2.6.10 Linux) <br> 297993 Fix compilation of valgrind with gcc -g3.<br> 298080 POWER Processor DFP support missing, part 3<br> 298227 == 273475 (Add support for AVX instructions)<br> 298335 == 273475 (Add support for AVX instructions)<br> 298354 Unhandled ARM Thumb instruction 0xEB0D 0x0585 (streq)<br> 298394 s390x: Don't bail out on an unknown machine model. [..]<br> 298421 accept4() syscall (366) support is missing for ARM<br> 298718 vex amd64->IR: 0xF 0xB1 0xCB 0x9C 0x8F 0x45<br> 298732 valgrind installation problem in ubuntu with kernel version 3.x<br> 298862 POWER Processor DFP instruction support missing, part 4<br> 298864 DWARF reader mis-parses DW_FORM_ref_addr<br> 298943 massif asserts with --pages-as-heap=yes when brk is changing [..]<br> 299053 Support DWARF4 DW_AT_high_pc constant form<br> 299104 == 273475 (Add support for AVX instructions)<br> 299316 Helgrind: hg_main.c:628 (map_threads_lookup): Assertion 'thr' failed.<br> 299629 dup3() syscall (358) support is missing for ARM<br> 299694 POWER Processor DFP instruction support missing, part 5<br> 299756 Ignore --free-fill for MEMPOOL_FREE and FREELIKE client requests<br> 299803 == 273475 (Add support for AVX instructions)<br> 299804 == 273475 (Add support for AVX instructions)<br> 299805 == 273475 (Add support for AVX instructions)<br> 300140 ARM - Missing (T1) SMMUL<br> 300195 == 296318 (ELF Debug info improvements (more than one rx/rw mapping))<br> 300389 Assertion `are_valid_hwcaps(VexArchAMD64, [..])' failed.<br> 300414 FCOM and FCOMP unimplemented for amd64 guest<br> 301204 infinite loop in canonicaliseSymtab with ifunc symbol<br> 301229 == 203877 (increase to 16Mb maximum allowed alignment for memalign etc)<br> 301265 add x86 support to Android build <br> 301984 configure script doesn't detect certain versions of clang<br> 302205 Fix compiler warnings for POWER VEX code and POWER test cases<br> 302287 Unhandled movbe instruction on Atom processors<br> 302370 PPC: fnmadd, fnmsub, fnmadds, fnmsubs insns always negate the result<br> 302536 Fix for the POWER Valgrind regression test: memcheck-ISA2.0.<br> 302578 Unrecognized isntruction 0xc5 0x32 0xc2 0xca 0x09 vcmpngess<br> 302656 == 273475 (Add support for AVX instructions)<br> 302709 valgrind for ARM needs extra tls support for android emulator [..]<br> 302827 add wrapper for CDROM_GET_CAPABILITY<br> 302901 Valgrind crashes with dwz optimized debuginfo<br> 302918 Enable testing of the vmaddfp and vnsubfp instructions in the testsuite<br> 303116 Add support for the POWER instruction popcntb<br> 303127 Power test suite fixes for frsqrte, vrefp, and vrsqrtefp instructions.<br> 303250 Assertion `instrs_in->arr_used <= 10000' failed w/ OpenSSL code<br> 303466 == 273475 (Add support for AVX instructions)<br> 303624 segmentation fault on Android 4.1 (e.g. on Galaxy Nexus OMAP) <br> 303963 strstr() function produces wrong results under valgrind callgrind<br> 304054 CALL_FN_xx macros need to enforce stack alignment<br> 304561 tee system call not supported<br> 715750 (MacOSX): Incorrect invalid-address errors near 0xFFFFxxxx (mozbug#)<br> n-i-bz Add missing gdbserver xml files for shadow registers for ppc32<br> n-i-bz Bypass gcc4.4/4.5 code gen bugs causing out of memory or asserts<br> n-i-bz Fix assert in gdbserver for watchpoints watching the same address<br> n-i-bz Fix false positive in sys_clone on amd64 when optional args [..]<br> n-i-bz s390x: Shadow registers can now be examined using vgdb<br> <br> (3.8.0-TEST3: 9 August 2012, vex r2465, valgrind r12865)<br> (3.8.0: 10 August 2012, vex r2465, valgrind r12866)<br> <br> <br> <br> Release 3.7.0 (5 November 2011)<br> ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~<br> 3.7.0 is a feature release with many significant improvements and the<br> usual collection of bug fixes.<br> <br> This release supports X86/Linux, AMD64/Linux, ARM/Linux, PPC32/Linux,<br> PPC64/Linux, S390X/Linux, ARM/Android, X86/Darwin and AMD64/Darwin.<br> Support for recent distros and toolchain components (glibc 2.14, gcc<br> 4.6, MacOSX 10.7) has been added.<br> <br> * ================== PLATFORM CHANGES =================<br> <br> * Support for IBM z/Architecture (s390x) running Linux. Valgrind can<br> analyse 64-bit programs running on z/Architecture. Most user space<br> instructions up to and including z10 are supported. Valgrind has<br> been tested extensively on z9, z10, and z196 machines running SLES<br> 10/11, RedHat 5/6m, and Fedora. The Memcheck and Massif tools are<br> known to work well. Callgrind, Helgrind, and DRD work reasonably<br> well on z9 and later models. See README.s390 for more details.<br> <br> * Preliminary support for MacOSX 10.7 and XCode 4. Both 32- and<br> 64-bit processes are supported. Some complex threaded applications<br> (Firefox) are observed to hang when run as 32 bit applications,<br> whereas 64-bit versions run OK. The cause is unknown. Memcheck<br> will likely report some false errors. In general, expect some rough<br> spots. This release also supports MacOSX 10.6, but drops support<br> for 10.5.<br> <br> * Preliminary support for Android (on ARM). Valgrind can now run<br> large applications (eg, Firefox) on (eg) a Samsung Nexus S. See<br> README.android for more details, plus instructions on how to get<br> started.<br> <br> * Support for the IBM Power ISA 2.06 (Power7 instructions)<br> <br> * General correctness and performance improvements for ARM/Linux, and,<br> by extension, ARM/Android.<br> <br> * Further solidification of support for SSE 4.2 in 64-bit mode. AVX<br> instruction set support is under development but is not available in<br> this release.<br> <br> * Support for AIX5 has been removed.<br> <br> * ==================== TOOL CHANGES ====================<br> <br> * Memcheck: some incremental changes:<br> <br> - reduction of memory use in some circumstances<br> <br> - improved handling of freed memory, which in some circumstances <br> can cause detection of use-after-free that would previously have<br> been missed<br> <br> - fix of a longstanding bug that could cause false negatives (missed<br> errors) in programs doing vector saturated narrowing instructions.<br> <br> * Helgrind: performance improvements and major memory use reductions,<br> particularly for large, long running applications which perform many<br> synchronisation (lock, unlock, etc) events. Plus many smaller<br> changes:<br> <br> - display of locksets for both threads involved in a race<br> <br> - general improvements in formatting/clarity of error messages<br> <br> - addition of facilities and documentation regarding annotation<br> of thread safe reference counted C++ classes<br> <br> - new flag --check-stack-refs=no|yes [yes], to disable race checking<br> on thread stacks (a performance hack)<br> <br> - new flag --free-is-write=no|yes [no], to enable detection of races<br> where one thread accesses heap memory but another one frees it,<br> without any coordinating synchronisation event<br> <br> * DRD: enabled XML output; added support for delayed thread deletion<br> in order to detect races that occur close to the end of a thread<br> (--join-list-vol); fixed a memory leak triggered by repeated client<br> memory allocatation and deallocation; improved Darwin support.<br> <br> * exp-ptrcheck: this tool has been renamed to exp-sgcheck<br> <br> * exp-sgcheck: this tool has been reduced in scope so as to improve<br> performance and remove checking that Memcheck does better.<br> Specifically, the ability to check for overruns for stack and global<br> arrays is unchanged, but the ability to check for overruns of heap<br> blocks has been removed. The tool has accordingly been renamed to<br> exp-sgcheck ("Stack and Global Array Checking").<br> <br> * ==================== OTHER CHANGES ====================<br> <br> * GDB server: Valgrind now has an embedded GDB server. That means it<br> is possible to control a Valgrind run from GDB, doing all the usual<br> things that GDB can do (single stepping, breakpoints, examining<br> data, etc). Tool-specific functionality is also available. For<br> example, it is possible to query the definedness state of variables<br> or memory from within GDB when running Memcheck; arbitrarily large<br> memory watchpoints are supported, etc. To use the GDB server, start<br> Valgrind with the flag --vgdb-error=0 and follow the on-screen<br> instructions.<br> <br> * Improved support for unfriendly self-modifying code: a new option<br> --smc-check=all-non-file is available. This adds the relevant<br> consistency checks only to code that originates in non-file-backed<br> mappings. In effect this confines the consistency checking only to<br> code that is or might be JIT generated, and avoids checks on code<br> that must have been compiled ahead of time. This significantly<br> improves performance on applications that generate code at run time.<br> <br> * It is now possible to build a working Valgrind using Clang-2.9 on<br> Linux.<br> <br> * new client requests VALGRIND_{DISABLE,ENABLE}_ERROR_REPORTING.<br> These enable and disable error reporting on a per-thread, and<br> nestable, basis. This is useful for hiding errors in particularly<br> troublesome pieces of code. The MPI wrapper library (libmpiwrap.c)<br> now uses this facility.<br> <br> * Added the --mod-funcname option to cg_diff.<br> <br> * ==================== FIXED BUGS ====================<br> <br> The following bugs have been fixed or resolved. Note that "n-i-bz"<br> stands for "not in bugzilla" -- that is, a bug that was reported to us<br> but never got a bugzilla entry. We encourage you to file bugs in<br> bugzilla (http://bugs.kde.org/enter_valgrind_bug.cgi) rather than<br> mailing the developers (or mailing lists) directly -- bugs that are<br> not entered into bugzilla tend to get forgotten about or ignored.<br> <br> To see details of a given bug, visit<br> https://bugs.kde.org/show_bug.cgi?id=XXXXXX<br> where XXXXXX is the bug number as listed below.<br> <br> 79311 malloc silly arg warning does not give stack trace<br> 210935 port valgrind.h (not valgrind) to win32 to support client requests<br> 214223 valgrind SIGSEGV on startup gcc 4.4.1 ppc32 (G4) Ubuntu 9.10<br> 243404 Port to zSeries<br> 243935 Helgrind: incorrect handling of ANNOTATE_HAPPENS_BEFORE()/AFTER()<br> 247223 non-x86: Suppress warning: 'regparm' attribute directive ignored<br> 250101 huge "free" memory usage due to m_mallocfree.c fragmentation<br> 253206 Some fixes for the faultstatus testcase<br> 255223 capget testcase fails when running as root<br> 256703 xlc_dbl_u32.c testcase broken<br> 256726 Helgrind tests have broken inline asm <br> 259977 == 214223 (Valgrind segfaults doing __builtin_longjmp)<br> 264800 testcase compile failure on zseries<br> 265762 make public VEX headers compilable by G++ 3.x<br> 265771 assertion in jumps.c (r11523) fails with glibc-2.3<br> 266753 configure script does not give the user the option to not use QtCore<br> 266931 gen_insn_test.pl is broken<br> 266961 ld-linux.so.2 i?86-linux strlen issues<br> 266990 setns instruction causes false positive<br> 267020 Make directory for temporary files configurable at run-time.<br> 267342 == 267997 (segmentation fault on Mac OS 10.6)<br> 267383 Assertion 'vgPlain_strlen(dir) + vgPlain_strlen(file) + 1 < 256' failed<br> 267413 Assertion 'DRD_(g_threadinfo)[tid].synchr_nesting >= 1' failed.<br> 267488 regtest: darwin support for 64-bit build<br> 267552 SIGSEGV (misaligned_stack_error) with DRD, but not with other tools<br> 267630 Add support for IBM Power ISA 2.06 -- stage 1<br> 267769 == 267997 (Darwin: memcheck triggers segmentation fault)<br> 267819 Add client request for informing the core about reallocation<br> 267925 laog data structure quadratic for a single sequence of lock<br> 267968 drd: (vgDrd_thread_set_joinable): Assertion '0 <= (int)tid ..' failed<br> 267997 MacOSX: 64-bit V segfaults on launch when built with Xcode 4.0.1<br> 268513 missed optimizations in fold_Expr<br> 268619 s390x: fpr - gpr transfer facility <br> 268620 s390x: reconsider "long displacement" requirement <br> 268621 s390x: improve IR generation for XC<br> 268715 s390x: FLOGR is not universally available<br> 268792 == 267997 (valgrind seg faults on startup when compiled with Xcode 4)<br> 268930 s390x: MHY is not universally available<br> 269078 arm->IR: unhandled instruction SUB (SP minus immediate/register) <br> 269079 Support ptrace system call on ARM<br> 269144 missing "Bad option" error message<br> 269209 conditional load and store facility (z196)<br> 269354 Shift by zero on x86 can incorrectly clobber CC_NDEP<br> 269641 == 267997 (valgrind segfaults immediately (segmentation fault))<br> 269736 s390x: minor code generation tweaks<br> 269778 == 272986 (valgrind.h: swap roles of VALGRIND_DO_CLIENT_REQUEST() ..)<br> 269863 s390x: remove unused function parameters<br> 269864 s390x: tweak s390_emit_load_cc <br> 269884 == 250101 (overhead for huge blocks exhausts space too soon)<br> 270082 s390x: Make sure to point the PSW address to the next address on SIGILL<br> 270115 s390x: rewrite some testcases<br> 270309 == 267997 (valgrind crash on startup)<br> 270320 add support for Linux FIOQSIZE ioctl() call<br> 270326 segfault while trying to sanitize the environment passed to execle<br> 270794 IBM POWER7 support patch causes regression in none/tests<br> 270851 IBM POWER7 fcfidus instruction causes memcheck to fail<br> 270856 IBM POWER7 xsnmaddadp instruction causes memcheck to fail on 32bit app <br> 270925 hyper-optimized strspn() in /lib64/libc-2.13.so needs fix<br> 270959 s390x: invalid use of R0 as base register<br> 271042 VSX configure check fails when it should not <br> 271043 Valgrind build fails with assembler error on ppc64 with binutils 2.21 <br> 271259 s390x: fix code confusion <br> 271337 == 267997 (Valgrind segfaults on MacOS X)<br> 271385 s390x: Implement Ist_MBE <br> 271501 s390x: misc cleanups <br> 271504 s390x: promote likely and unlikely <br> 271579 ppc: using wrong enum type <br> 271615 unhandled instruction "popcnt" (arch=amd10h) <br> 271730 Fix bug when checking ioctls: duplicate check <br> 271776 s390x: provide STFLE instruction support <br> 271779 s390x: provide clock instructions like STCK <br> 271799 Darwin: ioctls without an arg report a memory error <br> 271820 arm: fix type confusion <br> 271917 pthread_cond_timedwait failure leads to not-locked false positive <br> 272067 s390x: fix DISP20 macro <br> 272615 A typo in debug output in mc_leakcheck.c<br> 272661 callgrind_annotate chokes when run from paths containing regex chars<br> 272893 amd64->IR: 0x66 0xF 0x38 0x2B 0xC1 0x66 0xF 0x7F == (closed as dup)<br> 272955 Unhandled syscall error for pwrite64 on ppc64 arch <br> 272967 make documentation build-system more robust <br> 272986 Fix gcc-4.6 warnings with valgrind.h<br> 273318 amd64->IR: 0x66 0xF 0x3A 0x61 0xC1 0x38 (missing PCMPxSTRx case)<br> 273318 unhandled PCMPxSTRx case: vex amd64->IR: 0x66 0xF 0x3A 0x61 0xC1 0x38 <br> 273431 valgrind segfaults in evalCfiExpr (debuginfo.c:2039)<br> 273465 Callgrind: jumps.c:164 (new_jcc): Assertion '(0 <= jmp) && ...'<br> 273536 Build error: multiple definition of `vgDrd_pthread_cond_initializer'<br> 273640 ppc64-linux: unhandled syscalls setresuid(164) and setresgid(169)<br> 273729 == 283000 (Illegal opcode for SSE2 "roundsd" instruction)<br> 273778 exp-ptrcheck: unhandled sysno == 259<br> 274089 exp-ptrcheck: unhandled sysno == 208<br> 274378 s390x: Various dispatcher tweaks<br> 274447 WARNING: unhandled syscall: 340<br> 274776 amd64->IR: 0x66 0xF 0x38 0x2B 0xC5 0x66<br> 274784 == 267997 (valgrind ls -l results in Segmentation Fault)<br> 274926 valgrind does not build against linux-3<br> 275148 configure FAIL with glibc-2.14<br> 275151 Fedora 15 / glibc-2.14 'make regtest' FAIL<br> 275168 Make Valgrind work for MacOSX 10.7 Lion<br> 275212 == 275284 (lots of false positives from __memcpy_ssse3_back et al)<br> 275278 valgrind does not build on Linux kernel 3.0.* due to silly<br> 275284 Valgrind memcpy/memmove redirection stopped working in glibc 2.14/x86_64<br> 275308 Fix implementation for ppc64 fres instruc<br> 275339 s390x: fix testcase compile warnings<br> 275517 s390x: Provide support for CKSM instruction<br> 275710 s390x: get rid of redundant address mode calculation<br> 275815 == 247894 (Valgrind doesn't know about Linux readahead(2) syscall)<br> 275852 == 250101 (valgrind uses all swap space and is killed)<br> 276784 Add support for IBM Power ISA 2.06 -- stage 3<br> 276987 gdbsrv: fix tests following recent commits<br> 277045 Valgrind crashes with unhandled DW_OP_ opcode 0x2a<br> 277199 The test_isa_2_06_part1.c in none/tests/ppc64 should be a symlink<br> 277471 Unhandled syscall: 340<br> 277610 valgrind crashes in VG_(lseek)(core_fd, phdrs[idx].p_offset, ...)<br> 277653 ARM: support Thumb2 PLD instruction<br> 277663 ARM: NEON float VMUL by scalar incorrect<br> 277689 ARM: tests for VSTn with register post-index are broken<br> 277694 ARM: BLX LR instruction broken in ARM mode<br> 277780 ARM: VMOV.F32 (immediate) instruction is broken<br> 278057 fuse filesystem syscall deadlocks<br> 278078 Unimplemented syscall 280 on ppc32<br> 278349 F_GETPIPE_SZ and F_SETPIPE_SZ Linux fcntl commands<br> 278454 VALGRIND_STACK_DEREGISTER has wrong output type<br> 278502 == 275284 (Valgrind confuses memcpy() and memmove())<br> 278892 gdbsrv: factorize gdb version handling, fix doc and typos<br> 279027 Support for MVCL and CLCL instruction<br> 279027 s390x: Provide support for CLCL and MVCL instructions<br> 279062 Remove a redundant check in the insn selector for ppc.<br> 279071 JDK creates PTEST with redundant REX.W prefix<br> 279212 gdbsrv: add monitor cmd v.info scheduler.<br> 279378 exp-ptrcheck: the 'impossible' happened on mkfifo call<br> 279698 memcheck discards valid-bits for packuswb<br> 279795 memcheck reports uninitialised values for mincore on amd64<br> 279994 Add support for IBM Power ISA 2.06 -- stage 3<br> 280083 mempolicy syscall check errors<br> 280290 vex amd64->IR: 0x66 0xF 0x38 0x28 0xC1 0x66 0xF 0x6F<br> 280710 s390x: config files for nightly builds<br> 280757 /tmp dir still used by valgrind even if TMPDIR is specified<br> 280965 Valgrind breaks fcntl locks when program does mmap<br> 281138 WARNING: unhandled syscall: 340<br> 281241 == 275168 (valgrind useless on Macos 10.7.1 Lion)<br> 281304 == 275168 (Darwin: dyld "cannot load inserted library")<br> 281305 == 275168 (unhandled syscall: unix:357 on Darwin 11.1)<br> 281468 s390x: handle do_clone and gcc clones in call traces<br> 281488 ARM: VFP register corruption<br> 281828 == 275284 (false memmove warning: "Source and destination overlap")<br> 281883 s390x: Fix system call wrapper for "clone".<br> 282105 generalise 'reclaimSuperBlock' to also reclaim splittable superblock<br> 282112 Unhandled instruction bytes: 0xDE 0xD9 0x9B 0xDF (fcompp)<br> 282238 SLES10: make check fails<br> 282979 strcasestr needs replacement with recent(>=2.12) glibc<br> 283000 vex amd64->IR: 0x66 0xF 0x3A 0xA 0xC0 0x9 0xF3 0xF<br> 283243 Regression in ppc64 memcheck tests<br> 283325 == 267997 (Darwin: V segfaults on startup when built with Xcode 4.0)<br> 283427 re-connect epoll_pwait syscall on ARM linux<br> 283600 gdbsrv: android: port vgdb.c<br> 283709 none/tests/faultstatus needs to account for page size<br> 284305 filter_gdb needs enhancement to work on ppc64<br> 284384 clang 3.1 -Wunused-value warnings in valgrind.h, memcheck.h<br> 284472 Thumb2 ROR.W encoding T2 not implemented<br> 284621 XML-escape process command line in XML output<br> n-i-bz cachegrind/callgrind: handle CPUID information for Core iX Intel CPUs<br> that have non-power-of-2 sizes (also AMDs)<br> n-i-bz don't be spooked by libraries mashed by elfhack<br> n-i-bz don't be spooked by libxul.so linked with gold<br> n-i-bz improved checking for VALGRIND_CHECK_MEM_IS_DEFINED<br> <br> (3.7.0-TEST1: 27 October 2011, vex r2228, valgrind r12245)<br> (3.7.0.RC1: 1 November 2011, vex r2231, valgrind r12257)<br> (3.7.0: 5 November 2011, vex r2231, valgrind r12258)<br> <br> <br> <br> Release 3.6.1 (16 February 2011)<br> ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~<br> 3.6.1 is a bug fix release. It adds support for some SSE4<br> instructions that were omitted in 3.6.0 due to lack of time. Initial<br> support for glibc-2.13 has been added. A number of bugs causing<br> crashing or assertion failures have been fixed.<br> <br> The following bugs have been fixed or resolved. Note that "n-i-bz"<br> stands for "not in bugzilla" -- that is, a bug that was reported to us<br> but never got a bugzilla entry. We encourage you to file bugs in<br> bugzilla (http://bugs.kde.org/enter_valgrind_bug.cgi) rather than<br> mailing the developers (or mailing lists) directly -- bugs that are<br> not entered into bugzilla tend to get forgotten about or ignored.<br> <br> To see details of a given bug, visit<br> https://bugs.kde.org/show_bug.cgi?id=XXXXXX<br> where XXXXXX is the bug number as listed below.<br> <br> 188572 Valgrind on Mac should suppress setenv() mem leak<br> 194402 vex amd64->IR: 0x48 0xF 0xAE 0x4 (proper FX{SAVE,RSTOR} support)<br> 210481 vex amd64->IR: Assertion `sz == 2 || sz == 4' failed (REX.W POPQ)<br> 246152 callgrind internal error after pthread_cancel on 32 Bit Linux<br> 250038 ppc64: Altivec LVSR and LVSL instructions fail their regtest<br> 254420 memory pool tracking broken <br> 254957 Test code failing to compile due to changes in memcheck.h<br> 255009 helgrind/drd: crash on chmod with invalid parameter<br> 255130 readdwarf3.c parse_type_DIE confused by GNAT Ada types<br> 255355 helgrind/drd: crash on threaded programs doing fork<br> 255358 == 255355<br> 255418 (SSE4.x) rint call compiled with ICC<br> 255822 --gen-suppressions can create invalid files: "too many callers [...]"<br> 255888 closing valgrindoutput tag outputted to log-stream on error<br> 255963 (SSE4.x) vex amd64->IR: 0x66 0xF 0x3A 0x9 0xDB 0x0 (ROUNDPD)<br> 255966 Slowness when using mempool annotations<br> 256387 vex x86->IR: 0xD4 0xA 0x2 0x7 (AAD and AAM)<br> 256600 super-optimized strcasecmp() false positive<br> 256669 vex amd64->IR: Unhandled LOOPNEL insn on amd64<br> 256968 (SSE4.x) vex amd64->IR: 0x66 0xF 0x38 0x10 0xD3 0x66 (BLENDVPx)<br> 257011 (SSE4.x) vex amd64->IR: 0x66 0xF 0x3A 0xE 0xFD 0xA0 (PBLENDW)<br> 257063 (SSE4.x) vex amd64->IR: 0x66 0xF 0x3A 0x8 0xC0 0x0 (ROUNDPS)<br> 257276 Missing case in memcheck --track-origins=yes<br> 258870 (SSE4.x) Add support for EXTRACTPS SSE 4.1 instruction<br> 261966 (SSE4.x) support for CRC32B and CRC32Q is lacking (also CRC32{W,L})<br> 262985 VEX regression in valgrind 3.6.0 in handling PowerPC VMX<br> 262995 (SSE4.x) crash when trying to valgrind gcc-snapshot (PCMPxSTRx $0)<br> 263099 callgrind_annotate counts Ir improperly [...]<br> 263877 undefined coprocessor instruction on ARMv7<br> 265964 configure FAIL with glibc-2.13<br> n-i-bz Fix compile error w/ icc-12.x in guest_arm_toIR.c<br> n-i-bz Docs: fix bogus descriptions for VALGRIND_CREATE_BLOCK et al<br> n-i-bz Massif: don't assert on shmat() with --pages-as-heap=yes<br> n-i-bz Bug fixes and major speedups for the exp-DHAT space profiler<br> n-i-bz DRD: disable --free-is-write due to implementation difficulties<br> <br> (3.6.1: 16 February 2011, vex r2103, valgrind r11561).<br> <br> <br> <br> Release 3.6.0 (21 October 2010)<br> ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~<br> 3.6.0 is a feature release with many significant improvements and the<br> usual collection of bug fixes.<br> <br> This release supports X86/Linux, AMD64/Linux, ARM/Linux, PPC32/Linux,<br> PPC64/Linux, X86/Darwin and AMD64/Darwin. Support for recent distros<br> and toolchain components (glibc 2.12, gcc 4.5, OSX 10.6) has been added.<br> <br> -------------------------<br> <br> Here are some highlights. Details are shown further down:<br> <br> * Support for ARM/Linux.<br> <br> * Support for recent Linux distros: Ubuntu 10.10 and Fedora 14.<br> <br> * Support for Mac OS X 10.6, both 32- and 64-bit executables.<br> <br> * Support for the SSE4.2 instruction set.<br> <br> * Enhancements to the Callgrind profiler, including the ability to<br> handle CPUs with three levels of cache.<br> <br> * A new experimental heap profiler, DHAT.<br> <br> * A huge number of bug fixes and small enhancements.<br> <br> -------------------------<br> <br> Here are details of the above changes, together with descriptions of<br> many other changes, and a list of fixed bugs.<br> <br> * ================== PLATFORM CHANGES =================<br> <br> * Support for ARM/Linux. Valgrind now runs on ARMv7 capable CPUs<br> running Linux. It is known to work on Ubuntu 10.04, Ubuntu 10.10,<br> and Maemo 5, so you can run Valgrind on your Nokia N900 if you want.<br> <br> This requires a CPU capable of running the ARMv7-A instruction set<br> (Cortex A5, A8 and A9). Valgrind provides fairly complete coverage<br> of the user space instruction set, including ARM and Thumb integer<br> code, VFPv3, NEON and V6 media instructions. The Memcheck,<br> Cachegrind and Massif tools work properly; other tools work to<br> varying degrees.<br> <br> * Support for recent Linux distros (Ubuntu 10.10 and Fedora 14), along<br> with support for recent releases of the underlying toolchain<br> components, notably gcc-4.5 and glibc-2.12.<br> <br> * Support for Mac OS X 10.6, both 32- and 64-bit executables. 64-bit<br> support also works much better on OS X 10.5, and is as solid as<br> 32-bit support now.<br> <br> * Support for the SSE4.2 instruction set. SSE4.2 is supported in<br> 64-bit mode. In 32-bit mode, support is only available up to and<br> including SSSE3. Some exceptions: SSE4.2 AES instructions are not<br> supported in 64-bit mode, and 32-bit mode does in fact support the<br> bare minimum SSE4 instructions to needed to run programs on Mac OS X<br> 10.6 on 32-bit targets.<br> <br> * Support for IBM POWER6 cpus has been improved. The Power ISA up to<br> and including version 2.05 is supported.<br> <br> * ==================== TOOL CHANGES ====================<br> <br> * Cachegrind has a new processing script, cg_diff, which finds the<br> difference between two profiles. It's very useful for evaluating<br> the performance effects of a change in a program.<br> <br> Related to this change, the meaning of cg_annotate's (rarely-used)<br> --threshold option has changed; this is unlikely to affect many<br> people, if you do use it please see the user manual for details.<br> <br> * Callgrind now can do branch prediction simulation, similar to<br> Cachegrind. In addition, it optionally can count the number of<br> executed global bus events. Both can be used for a better<br> approximation of a "Cycle Estimation" as derived event (you need to<br> update the event formula in KCachegrind yourself).<br> <br> * Cachegrind and Callgrind now refer to the LL (last-level) cache<br> rather than the L2 cache. This is to accommodate machines with<br> three levels of caches -- if Cachegrind/Callgrind auto-detects the<br> cache configuration of such a machine it will run the simulation as<br> if the L2 cache isn't present. This means the results are less<br> likely to match the true result for the machine, but<br> Cachegrind/Callgrind's results are already only approximate, and<br> should not be considered authoritative. The results are still<br> useful for giving a general idea about a program's locality.<br> <br> * Massif has a new option, --pages-as-heap, which is disabled by<br> default. When enabled, instead of tracking allocations at the level<br> of heap blocks (as allocated with malloc/new/new[]), it instead<br> tracks memory allocations at the level of memory pages (as mapped by<br> mmap, brk, etc). Each mapped page is treated as its own block.<br> Interpreting the page-level output is harder than the heap-level<br> output, but this option is useful if you want to account for every<br> byte of memory used by a program.<br> <br> * DRD has two new command-line options: --free-is-write and<br> --trace-alloc. The former allows to detect reading from already freed<br> memory, and the latter allows tracing of all memory allocations and<br> deallocations.<br> <br> * DRD has several new annotations. Custom barrier implementations can<br> now be annotated, as well as benign races on static variables.<br> <br> * DRD's happens before / happens after annotations have been made more<br> powerful, so that they can now also be used to annotate e.g. a smart<br> pointer implementation.<br> <br> * Helgrind's annotation set has also been drastically improved, so as<br> to provide to users a general set of annotations to describe locks,<br> semaphores, barriers and condition variables. Annotations to<br> describe thread-safe reference counted heap objects have also been<br> added.<br> <br> * Memcheck has a new command-line option, --show-possibly-lost, which<br> is enabled by default. When disabled, the leak detector will not<br> show possibly-lost blocks.<br> <br> * A new experimental heap profiler, DHAT (Dynamic Heap Analysis Tool),<br> has been added. DHAT keeps track of allocated heap blocks, and also<br> inspects every memory reference to see which block (if any) is being<br> accessed. This gives a lot of insight into block lifetimes,<br> utilisation, turnover, liveness, and the location of hot and cold<br> fields. You can use DHAT to do hot-field profiling.<br> <br> * ==================== OTHER CHANGES ====================<br> <br> * Improved support for unfriendly self-modifying code: the extra<br> overhead incurred by --smc-check=all has been reduced by<br> approximately a factor of 5 as compared with 3.5.0.<br> <br> * Ability to show directory names for source files in error messages.<br> This is combined with a flexible mechanism for specifying which<br> parts of the paths should be shown. This is enabled by the new flag<br> --fullpath-after.<br> <br> * A new flag, --require-text-symbol, which will stop the run if a<br> specified symbol is not found it a given shared object when it is<br> loaded into the process. This makes advanced working with function<br> intercepting and wrapping safer and more reliable.<br> <br> * Improved support for the Valkyrie GUI, version 2.0.0. GUI output<br> and control of Valgrind is now available for the tools Memcheck and<br> Helgrind. XML output from Valgrind is available for Memcheck,<br> Helgrind and exp-Ptrcheck.<br> <br> * More reliable stack unwinding on amd64-linux, particularly in the<br> presence of function wrappers, and with gcc-4.5 compiled code.<br> <br> * Modest scalability (performance improvements) for massive<br> long-running applications, particularly for those with huge amounts<br> of code.<br> <br> * Support for analyzing programs running under Wine with has been<br> improved. The header files <valgrind/valgrind.h>,<br> <valgrind/memcheck.h> and <valgrind/drd.h> can now be used in<br> Windows-programs compiled with MinGW or one of the Microsoft Visual<br> Studio compilers.<br> <br> * A rare but serious error in the 64-bit x86 CPU simulation was fixed.<br> The 32-bit simulator was not affected. This did not occur often,<br> but when it did would usually crash the program under test.<br> Bug 245925.<br> <br> * A large number of bugs were fixed. These are shown below.<br> <br> * A number of bugs were investigated, and were candidates for fixing,<br> but are not fixed in 3.6.0, due to lack of developer time. They may<br> get fixed in later releases. They are:<br> <br> 194402 vex amd64->IR: 0x48 0xF 0xAE 0x4 0x24 0x49 (FXSAVE64)<br> 212419 false positive "lock order violated" (A+B vs A) <br> 213685 Undefined value propagates past dependency breaking instruction<br> 216837 Incorrect instrumentation of NSOperationQueue on Darwin <br> 237920 valgrind segfault on fork failure <br> 242137 support for code compiled by LLVM-2.8<br> 242423 Another unknown Intel cache config value <br> 243232 Inconsistent Lock Orderings report with trylock <br> 243483 ppc: callgrind triggers VEX assertion failure <br> 243935 Helgrind: implementation of ANNOTATE_HAPPENS_BEFORE() is wrong<br> 244677 Helgrind crash hg_main.c:616 (map_threads_lookup): Assertion<br> 'thr' failed. <br> 246152 callgrind internal error after pthread_cancel on 32 Bit Linux <br> 249435 Analyzing wine programs with callgrind triggers a crash <br> 250038 ppc64: Altivec lvsr and lvsl instructions fail their regtest<br> 250065 Handling large allocations <br> 250101 huge "free" memory usage due to m_mallocfree.c<br> "superblocks fragmentation"<br> 251569 vex amd64->IR: 0xF 0x1 0xF9 0x8B 0x4C 0x24 (RDTSCP)<br> 252091 Callgrind on ARM does not detect function returns correctly<br> 252600 [PATCH] Allow lhs to be a pointer for shl/shr<br> 254420 memory pool tracking broken<br> n-i-bz support for adding symbols for JIT generated code<br> <br> <br> The following bugs have been fixed or resolved. Note that "n-i-bz"<br> stands for "not in bugzilla" -- that is, a bug that was reported to us<br> but never got a bugzilla entry. We encourage you to file bugs in<br> bugzilla (http://bugs.kde.org/enter_valgrind_bug.cgi) rather than<br> mailing the developers (or mailing lists) directly -- bugs that are<br> not entered into bugzilla tend to get forgotten about or ignored.<br> <br> To see details of a given bug, visit<br> https://bugs.kde.org/show_bug.cgi?id=XXXXXX<br> where XXXXXX is the bug number as listed below.<br> <br> 135264 dcbzl instruction missing<br> 142688 == 250799<br> 153699 Valgrind should report unaligned reads with movdqa<br> 180217 == 212335<br> 190429 Valgrind reports lost of errors in ld.so<br> with x86_64 2.9.90 glibc <br> 197266 valgrind appears to choke on the xmms instruction<br> "roundsd" on x86_64 <br> 197988 Crash when demangling very large symbol names<br> 202315 unhandled syscall: 332 (inotify_init1)<br> 203256 Add page-level profiling to Massif<br> 205093 dsymutil=yes needs quotes, locking (partial fix)<br> 205241 Snow Leopard 10.6 support (partial fix)<br> 206600 Leak checker fails to upgrade indirect blocks when their<br> parent becomes reachable <br> 210935 port valgrind.h (not valgrind) to win32 so apps run under<br> wine can make client requests<br> 211410 vex amd64->IR: 0x15 0xFF 0xFF 0x0 0x0 0x89<br> within Linux ip-stack checksum functions <br> 212335 unhandled instruction bytes: 0xF3 0xF 0xBD 0xC0<br> (lzcnt %eax,%eax) <br> 213685 Undefined value propagates past dependency breaking instruction<br> (partial fix)<br> 215914 Valgrind inserts bogus empty environment variable <br> 217863 == 197988<br> 219538 adjtimex syscall wrapper wrong in readonly adjtime mode <br> 222545 shmat fails under valgind on some arm targets <br> 222560 ARM NEON support <br> 230407 == 202315<br> 231076 == 202315<br> 232509 Docs build fails with formatting inside <title></title> elements <br> 232793 == 202315<br> 235642 [PATCH] syswrap-linux.c: support evdev EVIOCG* ioctls <br> 236546 vex x86->IR: 0x66 0xF 0x3A 0xA<br> 237202 vex amd64->IR: 0xF3 0xF 0xB8 0xC0 0x49 0x3B <br> 237371 better support for VALGRIND_MALLOCLIKE_BLOCK <br> 237485 symlink (syscall 57) is not supported on Mac OS <br> 237723 sysno == 101 exp-ptrcheck: the 'impossible' happened:<br> unhandled syscall <br> 238208 is_just_below_ESP doesn't take into account red-zone <br> 238345 valgrind passes wrong $0 when executing a shell script <br> 238679 mq_timedreceive syscall doesn't flag the reception buffer<br> as "defined"<br> 238696 fcntl command F_DUPFD_CLOEXEC not supported <br> 238713 unhandled instruction bytes: 0x66 0xF 0x29 0xC6 <br> 238713 unhandled instruction bytes: 0x66 0xF 0x29 0xC6 <br> 238745 3.5.0 Make fails on PPC Altivec opcodes, though configure<br> says "Altivec off"<br> 239992 vex amd64->IR: 0x48 0xF 0xC4 0xC1 0x0 0x48 <br> 240488 == 197988<br> 240639 == 212335<br> 241377 == 236546<br> 241903 == 202315<br> 241920 == 212335<br> 242606 unhandled syscall: setegid (in Ptrcheck)<br> 242814 Helgrind "Impossible has happened" during<br> QApplication::initInstance(); <br> 243064 Valgrind attempting to read debug information from iso <br> 243270 Make stack unwinding in Valgrind wrappers more reliable<br> 243884 exp-ptrcheck: the 'impossible happened: unhandled syscall <br> sysno = 277 (mq_open)<br> 244009 exp-ptrcheck unknown syscalls in analyzing lighttpd<br> 244493 ARM VFP d16-d31 registers support <br> 244670 add support for audit_session_self syscall on Mac OS 10.6<br> 244921 The xml report of helgrind tool is not well format<br> 244923 In the xml report file, the <preamble> not escape the <br> xml char, eg '<','&','>'<br> 245535 print full path names in plain text reports <br> 245925 x86-64 red zone handling problem <br> 246258 Valgrind not catching integer underruns + new [] s<br> 246311 reg/reg cmpxchg doesn't work on amd64<br> 246549 unhandled syscall unix:277 while testing 32-bit Darwin app <br> 246888 Improve Makefile.vex.am <br> 247510 [OS X 10.6] Memcheck reports unaddressable bytes passed <br> to [f]chmod_extended<br> 247526 IBM POWER6 (ISA 2.05) support is incomplete<br> 247561 Some leak testcases fails due to reachable addresses in<br> caller save regs<br> 247875 sizeofIRType to handle Ity_I128 <br> 247894 [PATCH] unhandled syscall sys_readahead <br> 247980 Doesn't honor CFLAGS passed to configure <br> 248373 darwin10.supp is empty in the trunk <br> 248822 Linux FIBMAP ioctl has int parameter instead of long<br> 248893 [PATCH] make readdwarf.c big endianess safe to enable<br> unwinding on big endian systems<br> 249224 Syscall 336 not supported (SYS_proc_info) <br> 249359 == 245535<br> 249775 Incorrect scheme for detecting NEON capabilities of host CPU<br> 249943 jni JVM init fails when using valgrind<br> 249991 Valgrind incorrectly declares AESKEYGENASSIST support<br> since VEX r2011<br> 249996 linux/arm: unhandled syscall: 181 (__NR_pwrite64)<br> 250799 frexp$fenv_access_off function generates SIGILL <br> 250998 vex x86->IR: unhandled instruction bytes: 0x66 0x66 0x66 0x2E <br> 251251 support pclmulqdq insn <br> 251362 valgrind: ARM: attach to debugger either fails or provokes<br> kernel oops <br> 251674 Unhandled syscall 294<br> 251818 == 254550<br> <br> 254257 Add support for debugfiles found by build-id<br> 254550 [PATCH] Implement DW_ATE_UTF (DWARF4)<br> 254646 Wrapped functions cause stack misalignment on OS X<br> (and possibly Linux)<br> 254556 ARM: valgrinding anything fails with SIGSEGV for 0xFFFF0FA0<br> <br> (3.6.0: 21 October 2010, vex r2068, valgrind r11471).<br> <br> <br> <br> Release 3.5.0 (19 August 2009)<br> ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~<br> 3.5.0 is a feature release with many significant improvements and the<br> usual collection of bug fixes. The main improvement is that Valgrind<br> now works on Mac OS X.<br> <br> This release supports X86/Linux, AMD64/Linux, PPC32/Linux, PPC64/Linux<br> and X86/Darwin. Support for recent distros and toolchain components<br> (glibc 2.10, gcc 4.5) has been added.<br> <br> -------------------------<br> <br> Here is a short summary of the changes. Details are shown further<br> down:<br> <br> * Support for Mac OS X (10.5.x).<br> <br> * Improvements and simplifications to Memcheck's leak checker.<br> <br> * Clarification and simplifications in various aspects of Valgrind's<br> text output.<br> <br> * XML output for Helgrind and Ptrcheck.<br> <br> * Performance and stability improvements for Helgrind and DRD.<br> <br> * Genuinely atomic support for x86/amd64/ppc atomic instructions.<br> <br> * A new experimental tool, BBV, useful for computer architecture<br> research.<br> <br> * Improved Wine support, including ability to read Windows PDB<br> debuginfo.<br> <br> -------------------------<br> <br> Here are details of the above changes, followed by descriptions of<br> many other minor changes, and a list of fixed bugs.<br> <br> <br> * Valgrind now runs on Mac OS X. (Note that Mac OS X is sometimes<br> called "Darwin" because that is the name of the OS core, which is the<br> level that Valgrind works at.)<br> <br> Supported systems:<br> <br> - It requires OS 10.5.x (Leopard). Porting to 10.4.x is not planned<br> because it would require work and 10.4 is only becoming less common.<br> <br> - 32-bit programs on x86 and AMD64 (a.k.a x86-64) machines are supported<br> fairly well. For 10.5.x, 32-bit programs are the default even on<br> 64-bit machines, so it handles most current programs.<br> <br> - 64-bit programs on x86 and AMD64 (a.k.a x86-64) machines are not<br> officially supported, but simple programs at least will probably work.<br> However, start-up is slow.<br> <br> - PowerPC machines are not supported.<br> <br> Things that don't work:<br> <br> - The Ptrcheck tool.<br> <br> - Objective-C garbage collection.<br> <br> - --db-attach=yes.<br> <br> - If you have Rogue Amoeba's "Instant Hijack" program installed,<br> Valgrind will fail with a SIGTRAP at start-up. See<br> https://bugs.kde.org/show_bug.cgi?id=193917 for details and a<br> simple work-around.<br> <br> Usage notes:<br> <br> - You will likely find --dsymutil=yes a useful option, as error<br> messages may be imprecise without it.<br> <br> - Mac OS X support is new and therefore will be less robust than the<br> Linux support. Please report any bugs you find.<br> <br> - Threaded programs may run more slowly than on Linux.<br> <br> Many thanks to Greg Parker for developing this port over several years.<br> <br> <br> * Memcheck's leak checker has been improved. <br> <br> - The results for --leak-check=summary now match the summary results<br> for --leak-check=full. Previously they could differ because<br> --leak-check=summary counted "indirectly lost" blocks and<br> "suppressed" blocks as "definitely lost".<br> <br> - Blocks that are only reachable via at least one interior-pointer,<br> but are directly pointed to by a start-pointer, were previously<br> marked as "still reachable". They are now correctly marked as<br> "possibly lost".<br> <br> - The default value for the --leak-resolution option has been<br> changed from "low" to "high". In general, this means that more<br> leak reports will be produced, but each leak report will describe<br> fewer leaked blocks.<br> <br> - With --leak-check=full, "definitely lost" and "possibly lost"<br> leaks are now considered as proper errors, ie. they are counted<br> for the "ERROR SUMMARY" and affect the behaviour of<br> --error-exitcode. These leaks are not counted as errors if<br> --leak-check=summary is specified, however.<br> <br> - Documentation for the leak checker has been improved.<br> <br> <br> * Various aspects of Valgrind's text output have changed.<br> <br> - Valgrind's start-up message has changed. It is shorter but also<br> includes the command being run, which makes it easier to use<br> --trace-children=yes. An example:<br> <br> - Valgrind's shut-down messages have also changed. This is most<br> noticeable with Memcheck, where the leak summary now occurs before<br> the error summary. This change was necessary to allow leaks to be<br> counted as proper errors (see the description of the leak checker<br> changes above for more details). This was also necessary to fix a<br> longstanding bug in which uses of suppressions against leaks were<br> not "counted", leading to difficulties in maintaining suppression<br> files (see https://bugs.kde.org/show_bug.cgi?id=186790).<br> <br> - Behavior of -v has changed. In previous versions, -v printed out<br> a mixture of marginally-user-useful information, and tool/core<br> statistics. The statistics printing has now been moved to its own<br> flag, --stats=yes. This means -v is less verbose and more likely<br> to convey useful end-user information.<br> <br> - The format of some (non-XML) stack trace entries has changed a<br> little. Previously there were six possible forms:<br> <br> 0x80483BF: really (a.c:20)<br> 0x80483BF: really (in /foo/a.out)<br> 0x80483BF: really<br> 0x80483BF: (within /foo/a.out)<br> 0x80483BF: ??? (a.c:20)<br> 0x80483BF: ???<br> <br> The third and fourth of these forms have been made more consistent<br> with the others. The six possible forms are now:<br> <br> 0x80483BF: really (a.c:20)<br> 0x80483BF: really (in /foo/a.out)<br> 0x80483BF: really (in ???)<br> 0x80483BF: ??? (in /foo/a.out)<br> 0x80483BF: ??? (a.c:20)<br> 0x80483BF: ???<br> <br> Stack traces produced when --xml=yes is specified are different<br> and unchanged.<br> <br> <br> * Helgrind and Ptrcheck now support XML output, so they can be used<br> from GUI tools. Also, the XML output mechanism has been<br> overhauled.<br> <br> - The XML format has been overhauled and generalised, so it is more<br> suitable for error reporting tools in general. The Memcheck<br> specific aspects of it have been removed. The new format, which<br> is an evolution of the old format, is described in<br> docs/internals/xml-output-protocol4.txt.<br> <br> - Memcheck has been updated to use the new format.<br> <br> - Helgrind and Ptrcheck are now able to emit output in this format.<br> <br> - The XML output mechanism has been overhauled. XML is now output<br> to its own file descriptor, which means that:<br> <br> * Valgrind can output text and XML independently.<br> <br> * The longstanding problem of XML output being corrupted by <br> unexpected un-tagged text messages is solved.<br> <br> As before, the destination for text output is specified using<br> --log-file=, --log-fd= or --log-socket=.<br> <br> As before, XML output for a tool is enabled using --xml=yes.<br> <br> Because there's a new XML output channel, the XML output<br> destination is now specified by --xml-file=, --xml-fd= or<br> --xml-socket=.<br> <br> Initial feedback has shown this causes some confusion. To<br> clarify, the two envisaged usage scenarios are:<br> <br> (1) Normal text output. In this case, do not specify --xml=yes<br> nor any of --xml-file=, --xml-fd= or --xml-socket=.<br> <br> (2) XML output. In this case, specify --xml=yes, and one of<br> --xml-file=, --xml-fd= or --xml-socket= to select the XML<br> destination, one of --log-file=, --log-fd= or --log-socket=<br> to select the destination for any remaining text messages,<br> and, importantly, -q.<br> <br> -q makes Valgrind completely silent on the text channel,<br> except in the case of critical failures, such as Valgrind<br> itself segfaulting, or failing to read debugging information.<br> Hence, in this scenario, it suffices to check whether or not<br> any output appeared on the text channel. If yes, then it is<br> likely to be a critical error which should be brought to the<br> attention of the user. If no (the text channel produced no<br> output) then it can be assumed that the run was successful.<br> <br> This allows GUIs to make the critical distinction they need to<br> make (did the run fail or not?) without having to search or<br> filter the text output channel in any way.<br> <br> It is also recommended to use --child-silent-after-fork=yes in<br> scenario (2).<br> <br> <br> * Improvements and changes in Helgrind:<br> <br> - XML output, as described above<br> <br> - Checks for consistent association between pthread condition<br> variables and their associated mutexes are now performed.<br> <br> - pthread_spinlock functions are supported.<br> <br> - Modest performance improvements.<br> <br> - Initial (skeletal) support for describing the behaviour of<br> non-POSIX synchronisation objects through ThreadSanitizer<br> compatible ANNOTATE_* macros.<br> <br> - More controllable tradeoffs between performance and the level of<br> detail of "previous" accesses in a race. There are now three<br> settings:<br> <br> * --history-level=full. This is the default, and was also the<br> default in 3.4.x. It shows both stacks involved in a race, but<br> requires a lot of memory and can be very slow in programs that<br> do many inter-thread synchronisation events.<br> <br> * --history-level=none. This only shows the later stack involved<br> in a race. This can be much faster than --history-level=full,<br> but makes it much more difficult to find the other access<br> involved in the race.<br> <br> The new intermediate setting is<br> <br> * --history-level=approx<br> <br> For the earlier (other) access, two stacks are presented. The<br> earlier access is guaranteed to be somewhere in between the two<br> program points denoted by those stacks. This is not as useful<br> as showing the exact stack for the previous access (as per<br> --history-level=full), but it is better than nothing, and it's<br> almost as fast as --history-level=none.<br> <br> <br> * New features and improvements in DRD:<br> <br> - The error messages printed by DRD are now easier to interpret.<br> Instead of using two different numbers to identify each thread<br> (Valgrind thread ID and DRD thread ID), DRD does now identify<br> threads via a single number (the DRD thread ID). Furthermore<br> "first observed at" information is now printed for all error<br> messages related to synchronization objects.<br> <br> - Added support for named semaphores (sem_open() and sem_close()).<br> <br> - Race conditions between pthread_barrier_wait() and<br> pthread_barrier_destroy() calls are now reported.<br> <br> - Added support for custom allocators through the macros<br> VALGRIND_MALLOCLIKE_BLOCK() VALGRIND_FREELIKE_BLOCK() (defined in<br> in <valgrind/valgrind.h>). An alternative for these two macros is<br> the new client request VG_USERREQ__DRD_CLEAN_MEMORY (defined in<br> <valgrind/drd.h>).<br> <br> - Added support for annotating non-POSIX synchronization objects<br> through several new ANNOTATE_*() macros.<br> <br> - OpenMP: added support for the OpenMP runtime (libgomp) included<br> with gcc versions 4.3.0 and 4.4.0.<br> <br> - Faster operation.<br> <br> - Added two new command-line options (--first-race-only and<br> --segment-merging-interval).<br> <br> <br> * Genuinely atomic support for x86/amd64/ppc atomic instructions<br> <br> Valgrind will now preserve (memory-access) atomicity of LOCK-<br> prefixed x86/amd64 instructions, and any others implying a global<br> bus lock. Ditto for PowerPC l{w,d}arx/st{w,d}cx. instructions.<br> <br> This means that Valgrinded processes will "play nicely" in<br> situations where communication with other processes, or the kernel,<br> is done through shared memory and coordinated with such atomic<br> instructions. Prior to this change, such arrangements usually<br> resulted in hangs, races or other synchronisation failures, because<br> Valgrind did not honour atomicity of such instructions.<br> <br> <br> * A new experimental tool, BBV, has been added. BBV generates basic<br> block vectors for use with the SimPoint analysis tool, which allows<br> a program's overall behaviour to be approximated by running only a<br> fraction of it. This is useful for computer architecture<br> researchers. You can run BBV by specifying --tool=exp-bbv (the<br> "exp-" prefix is short for "experimental"). BBV was written by<br> Vince Weaver.<br> <br> <br> * Modestly improved support for running Windows applications under<br> Wine. In particular, initial support for reading Windows .PDB debug<br> information has been added.<br> <br> <br> * A new Memcheck client request VALGRIND_COUNT_LEAK_BLOCKS has been<br> added. It is similar to VALGRIND_COUNT_LEAKS but counts blocks<br> instead of bytes.<br> <br> <br> * The Valgrind client requests VALGRIND_PRINTF and<br> VALGRIND_PRINTF_BACKTRACE have been changed slightly. Previously,<br> the string was always printed immediately on its own line. Now, the<br> string will be added to a buffer but not printed until a newline is<br> encountered, or other Valgrind output is printed (note that for<br> VALGRIND_PRINTF_BACKTRACE, the back-trace itself is considered<br> "other Valgrind output"). This allows you to use multiple<br> VALGRIND_PRINTF calls to build up a single output line, and also to<br> print multiple output lines with a single request (by embedding<br> multiple newlines in the string).<br> <br> <br> * The graphs drawn by Massif's ms_print program have changed slightly:<br> <br> - The half-height chars '.' and ',' are no longer drawn, because<br> they are confusing. The --y option can be used if the default<br> y-resolution is not high enough.<br> <br> - Horizontal lines are now drawn after the top of a snapshot if<br> there is a gap until the next snapshot. This makes it clear that<br> the memory usage has not dropped to zero between snapshots.<br> <br> <br> * Something that happened in 3.4.0, but wasn't clearly announced: the<br> option --read-var-info=yes can be used by some tools (Memcheck,<br> Helgrind and DRD). When enabled, it causes Valgrind to read DWARF3<br> variable type and location information. This makes those tools<br> start up more slowly and increases memory consumption, but<br> descriptions of data addresses in error messages become more<br> detailed.<br> <br> <br> * exp-Omega, an experimental instantaneous leak-detecting tool, was<br> disabled in 3.4.0 due to a lack of interest and maintenance,<br> although the source code was still in the distribution. The source<br> code has now been removed from the distribution. For anyone<br> interested, the removal occurred in SVN revision r10247.<br> <br> <br> * Some changes have been made to the build system.<br> <br> - VEX/ is now integrated properly into the build system. This means<br> that dependency tracking within VEX/ now works properly, "make<br> install" will work without requiring "make" before it, and<br> parallel builds (ie. 'make -j') now work (previously a<br> .NOTPARALLEL directive was used to serialize builds, ie. 'make -j'<br> was effectively ignored).<br> <br> - The --with-vex configure option has been removed. It was of<br> little use and removing it simplified the build system.<br> <br> - The location of some install files has changed. This should not<br> affect most users. Those who might be affected:<br> <br> * For people who use Valgrind with MPI programs, the installed<br> libmpiwrap.so library has moved from<br> $(INSTALL)/<platform>/libmpiwrap.so to<br> $(INSTALL)/libmpiwrap-<platform>.so.<br> <br> * For people who distribute standalone Valgrind tools, the<br> installed libraries such as $(INSTALL)/<platform>/libcoregrind.a<br> have moved to $(INSTALL)/libcoregrind-<platform>.a.<br> <br> These changes simplify the build system.<br> <br> - Previously, all the distributed suppression (*.supp) files were<br> installed. Now, only default.supp is installed. This should not<br> affect users as the other installed suppression files were not<br> read; the fact that they were installed was a mistake.<br> <br> <br> * KNOWN LIMITATIONS:<br> <br> - Memcheck is unusable with the Intel compiler suite version 11.1,<br> when it generates code for SSE2-and-above capable targets. This<br> is because of icc's use of highly optimised inlined strlen<br> implementations. It causes Memcheck to report huge numbers of<br> false errors even in simple programs. Helgrind and DRD may also<br> have problems.<br> <br> Versions 11.0 and earlier may be OK, but this has not been<br> properly tested.<br> <br> <br> The following bugs have been fixed or resolved. Note that "n-i-bz"<br> stands for "not in bugzilla" -- that is, a bug that was reported to us<br> but never got a bugzilla entry. We encourage you to file bugs in<br> bugzilla (http://bugs.kde.org/enter_valgrind_bug.cgi) rather than<br> mailing the developers (or mailing lists) directly -- bugs that are<br> not entered into bugzilla tend to get forgotten about or ignored.<br> <br> To see details of a given bug, visit<br> https://bugs.kde.org/show_bug.cgi?id=XXXXXX<br> where XXXXXX is the bug number as listed below.<br> <br> 84303 How about a LockCheck tool? <br> 91633 dereference of null ptr in vgPlain_st_basetype <br> 97452 Valgrind doesn't report any pthreads problems <br> 100628 leak-check gets assertion failure when using <br> VALGRIND_MALLOCLIKE_BLOCK on malloc()ed memory <br> 108528 NPTL pthread cleanup handlers not called <br> 110126 Valgrind 2.4.1 configure.in tramples CFLAGS <br> 110128 mallinfo is not implemented... <br> 110770 VEX: Generated files not always updated when making valgrind<br> 111102 Memcheck: problems with large (memory footprint) applications <br> 115673 Vex's decoder should never assert <br> 117564 False positive: Syscall param clone(child_tidptr) contains<br> uninitialised byte(s) <br> 119404 executing ssh from inside valgrind fails <br> 133679 Callgrind does not write path names to sources with dwarf debug<br> info<br> 135847 configure.in problem with non gnu compilers (and possible fix) <br> 136154 threads.c:273 (vgCallgrind_post_signal): Assertion<br> '*(vgCallgrind_current_fn_stack.top) == 0' failed. <br> 136230 memcheck reports "possibly lost", should be "still reachable" <br> 137073 NULL arg to MALLOCLIKE_BLOCK causes crash <br> 137904 Valgrind reports a memory leak when using POSIX threads,<br> while it shouldn't <br> 139076 valgrind VT_GETSTATE error <br> 142228 complaint of elf_dynamic_do_rela in trivial usage <br> 145347 spurious warning with USBDEVFS_REAPURB <br> 148441 (wine) can't find memory leak in Wine, win32 binary <br> executable file.<br> 148742 Leak-check fails assert on exit <br> 149878 add (proper) check for calloc integer overflow <br> 150606 Call graph is broken when using callgrind control <br> 152393 leak errors produce an exit code of 0. I need some way to <br> cause leak errors to result in a nonzero exit code. <br> 157154 documentation (leak-resolution doc speaks about num-callers<br> def=4) + what is a loss record<br> 159501 incorrect handling of ALSA ioctls <br> 162020 Valgrinding an empty/zero-byte file crashes valgrind <br> 162482 ppc: Valgrind crashes while reading stabs information <br> 162718 x86: avoid segment selector 0 in sys_set_thread_area() <br> 163253 (wine) canonicaliseSymtab forgot some fields in DiSym <br> 163560 VEX/test_main.c is missing from valgrind-3.3.1 <br> 164353 malloc_usable_size() doesn't return a usable size <br> 165468 Inconsistent formatting in memcheck manual -- please fix <br> 169505 main.c:286 (endOfInstr):<br> Assertion 'ii->cost_offset == *cost_offset' failed <br> 177206 Generate default.supp during compile instead of configure<br> 177209 Configure valt_load_address based on arch+os <br> 177305 eventfd / syscall 323 patch lost<br> 179731 Tests fail to build because of inlining of non-local asm labels<br> 181394 helgrind: libhb_core.c:3762 (msm_write): Assertion <br> 'ordxx == POrd_EQ || ordxx == POrd_LT' failed. <br> 181594 Bogus warning for empty text segment <br> 181707 dwarf doesn't require enumerations to have name <br> 185038 exp-ptrcheck: "unhandled syscall: 285" (fallocate) on x86_64 <br> 185050 exp-ptrcheck: sg_main.c:727 (add_block_to_GlobalTree):<br> Assertion '!already_present' failed.<br> 185359 exp-ptrcheck: unhandled syscall getresuid()<br> 185794 "WARNING: unhandled syscall: 285" (fallocate) on x86_64<br> 185816 Valgrind is unable to handle debug info for files with split<br> debug info that are prelinked afterwards <br> 185980 [darwin] unhandled syscall: sem_open <br> 186238 bbToIR_AMD64: disInstr miscalculated next %rip<br> 186507 exp-ptrcheck unhandled syscalls prctl, etc. <br> 186790 Suppression pattern used for leaks are not reported <br> 186796 Symbols with length>200 in suppression files are ignored <br> 187048 drd: mutex PTHREAD_PROCESS_SHARED attribute missinterpretation<br> 187416 exp-ptrcheck: support for __NR_{setregid,setreuid,setresuid}<br> 188038 helgrind: hg_main.c:926: mk_SHVAL_fail: the 'impossible' happened<br> 188046 bashisms in the configure script<br> 188127 amd64->IR: unhandled instruction bytes: 0xF0 0xF 0xB0 0xA<br> 188161 memcheck: --track-origins=yes asserts "mc_machine.c:672<br> (get_otrack_shadow_offset_wrk): the 'impossible' happened."<br> 188248 helgrind: pthread_cleanup_push, pthread_rwlock_unlock, <br> assertion fail "!lock->heldBy" <br> 188427 Add support for epoll_create1 (with patch) <br> 188530 Support for SIOCGSTAMPNS<br> 188560 Include valgrind.spec in the tarball<br> 188572 Valgrind on Mac should suppress setenv() mem leak <br> 189054 Valgrind fails to build because of duplicate non-local asm labels <br> 189737 vex amd64->IR: unhandled instruction bytes: 0xAC<br> 189762 epoll_create syscall not handled (--tool=exp-ptrcheck)<br> 189763 drd assertion failure: s_threadinfo[tid].is_recording <br> 190219 unhandled syscall: 328 (x86-linux)<br> 190391 dup of 181394; see above<br> 190429 Valgrind reports lots of errors in ld.so with x86_64 2.9.90 glibc <br> 190820 No debug information on powerpc-linux<br> 191095 PATCH: Improve usbdevfs ioctl handling <br> 191182 memcheck: VALGRIND_LEAK_CHECK quadratic when big nr of chunks<br> or big nr of errors<br> 191189 --xml=yes should obey --gen-suppressions=all <br> 191192 syslog() needs a suppression on macosx <br> 191271 DARWIN: WARNING: unhandled syscall: 33554697 a.k.a.: 265 <br> 191761 getrlimit on MacOSX <br> 191992 multiple --fn-skip only works sometimes; dependent on order <br> 192634 V. reports "aspacem sync_check_mapping_callback: <br> segment mismatch" on Darwin<br> 192954 __extension__ missing on 2 client requests <br> 194429 Crash at start-up with glibc-2.10.1 and linux-2.6.29 <br> 194474 "INSTALL" file has different build instructions than "README"<br> 194671 Unhandled syscall (sem_wait?) from mac valgrind <br> 195069 memcheck: reports leak (memory still reachable) for <br> printf("%d', x) <br> 195169 drd: (vgDrd_barrier_post_wait):<br> Assertion 'r->sg[p->post_iteration]' failed. <br> 195268 valgrind --log-file doesn't accept ~/...<br> 195838 VEX abort: LibVEX_N_SPILL_BYTES too small for CPUID boilerplate <br> 195860 WARNING: unhandled syscall: unix:223 <br> 196528 need a error suppression for pthread_rwlock_init under os x? <br> 197227 Support aio_* syscalls on Darwin<br> 197456 valgrind should reject --suppressions=(directory) <br> 197512 DWARF2 CFI reader: unhandled CFI instruction 0:10 <br> 197591 unhandled syscall 27 (mincore) <br> 197793 Merge DCAS branch to the trunk == 85756, 142103<br> 197794 Avoid duplicate filenames in Vex <br> 197898 make check fails on current SVN <br> 197901 make check fails also under exp-ptrcheck in current SVN <br> 197929 Make --leak-resolution=high the default <br> 197930 Reduce spacing between leak reports <br> 197933 Print command line of client at start-up, and shorten preamble <br> 197966 unhandled syscall 205 (x86-linux, --tool=exp-ptrcheck)<br> 198395 add BBV to the distribution as an experimental tool <br> 198624 Missing syscalls on Darwin: 82, 167, 281, 347 <br> 198649 callgrind_annotate doesn't cumulate counters <br> 199338 callgrind_annotate sorting/thresholds are broken for all but Ir <br> 199977 Valgrind complains about an unrecognized instruction in the<br> atomic_incs test program<br> 200029 valgrind isn't able to read Fedora 12 debuginfo <br> 200760 darwin unhandled syscall: unix:284 <br> 200827 DRD doesn't work on Mac OS X <br> 200990 VG_(read_millisecond_timer)() does not work correctly <br> 201016 Valgrind does not support pthread_kill() on Mac OS <br> 201169 Document --read-var-info<br> 201323 Pre-3.5.0 performance sanity checking <br> 201384 Review user manual for the 3.5.0 release <br> 201585 mfpvr not implemented on ppc <br> 201708 tests failing because x86 direction flag is left set <br> 201757 Valgrind doesn't handle any recent sys_futex additions <br> 204377 64-bit valgrind can not start a shell script<br> (with #!/path/to/shell) if the shell is a 32-bit executable<br> n-i-bz drd: fixed assertion failure triggered by mutex reinitialization.<br> n-i-bz drd: fixed a bug that caused incorrect messages to be printed<br> about memory allocation events with memory access tracing enabled<br> n-i-bz drd: fixed a memory leak triggered by vector clock deallocation<br> <br> (3.5.0: 19 Aug 2009, vex r1913, valgrind r10846).<br> <br> <br> <br> Release 3.4.1 (28 February 2009)<br> ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~<br> 3.4.1 is a bug-fix release that fixes some regressions and assertion<br> failures in debug info reading in 3.4.0, most notably incorrect stack<br> traces on amd64-linux on older (glibc-2.3 based) systems. Various<br> other debug info problems are also fixed. A number of bugs in the<br> exp-ptrcheck tool introduced in 3.4.0 have been fixed.<br> <br> In view of the fact that 3.4.0 contains user-visible regressions<br> relative to 3.3.x, upgrading to 3.4.1 is recommended. Packagers are<br> encouraged to ship 3.4.1 in preference to 3.4.0.<br> <br> The fixed bugs are as follows. Note that "n-i-bz" stands for "not in<br> bugzilla" -- that is, a bug that was reported to us but never got a<br> bugzilla entry. We encourage you to file bugs in bugzilla<br> (http://bugs.kde.org/enter_valgrind_bug.cgi) rather than mailing the<br> developers (or mailing lists) directly -- bugs that are not entered<br> into bugzilla tend to get forgotten about or ignored.<br> <br> n-i-bz Fix various bugs reading icc-11 generated debug info<br> n-i-bz Fix various bugs reading gcc-4.4 generated debug info<br> n-i-bz Preliminary support for glibc-2.10 / Fedora 11<br> n-i-bz Cachegrind and Callgrind: handle non-power-of-two cache sizes,<br> so as to support (eg) 24k Atom D1 and Core2 with 3/6/12MB L2.<br> 179618 exp-ptrcheck crashed / exit prematurely<br> 179624 helgrind: false positive races with pthread_create and<br> recv/open/close/read<br> 134207 pkg-config output contains @VG_PLATFORM@<br> 176926 floating point exception at valgrind startup with PPC 440EPX<br> 181594 Bogus warning for empty text segment<br> 173751 amd64->IR: 0x48 0xF 0x6F 0x45 (even more redundant rex prefixes)<br> 181707 Dwarf3 doesn't require enumerations to have name<br> 185038 exp-ptrcheck: "unhandled syscall: 285" (fallocate) on x86_64<br> 185050 exp-ptrcheck: sg_main.c:727 (add_block_to_GlobalTree):<br> Assertion '!already_present' failed.<br> 185359 exp-ptrcheck unhandled syscall getresuid()<br> <br> (3.4.1.RC1: 24 Feb 2008, vex r1884, valgrind r9253).<br> (3.4.1: 28 Feb 2008, vex r1884, valgrind r9293).<br> <br> <br> <br> Release 3.4.0 (2 January 2009)<br> ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~<br> 3.4.0 is a feature release with many significant improvements and the<br> usual collection of bug fixes. This release supports X86/Linux,<br> AMD64/Linux, PPC32/Linux and PPC64/Linux. Support for recent distros<br> (using gcc 4.4, glibc 2.8 and 2.9) has been added.<br> <br> 3.4.0 brings some significant tool improvements. Memcheck can now<br> report the origin of uninitialised values, the thread checkers<br> Helgrind and DRD are much improved, and we have a new experimental<br> tool, exp-Ptrcheck, which is able to detect overruns of stack and<br> global arrays. In detail:<br> <br> * Memcheck is now able to track the origin of uninitialised values.<br> When it reports an uninitialised value error, it will try to show<br> the origin of the value, as either a heap or stack allocation.<br> Origin tracking is expensive and so is not enabled by default. To<br> use it, specify --track-origins=yes. Memcheck's speed will be<br> essentially halved, and memory usage will be significantly<br> increased. Nevertheless it can drastically reduce the effort<br> required to identify the root cause of uninitialised value errors,<br> and so is often a programmer productivity win, despite running more<br> slowly.<br> <br> * A version (1.4.0) of the Valkyrie GUI, that works with Memcheck in<br> 3.4.0, will be released shortly.<br> <br> * Helgrind's race detection algorithm has been completely redesigned<br> and reimplemented, to address usability and scalability concerns:<br> <br> - The new algorithm has a lower false-error rate: it is much less<br> likely to report races that do not really exist.<br> <br> - Helgrind will display full call stacks for both accesses involved<br> in a race. This makes it easier to identify the root causes of<br> races.<br> <br> - Limitations on the size of program that can run have been removed.<br> <br> - Performance has been modestly improved, although that is very<br> workload-dependent.<br> <br> - Direct support for Qt4 threading has been added.<br> <br> - pthread_barriers are now directly supported.<br> <br> - Helgrind works well on all supported Linux targets.<br> <br> * The DRD thread debugging tool has seen major improvements:<br> <br> - Greatly improved performance and significantly reduced memory<br> usage.<br> <br> - Support for several major threading libraries (Boost.Thread, Qt4,<br> glib, OpenMP) has been added.<br> <br> - Support for atomic instructions, POSIX semaphores, barriers and<br> reader-writer locks has been added.<br> <br> - Works now on PowerPC CPUs too.<br> <br> - Added support for printing thread stack usage at thread exit time.<br> <br> - Added support for debugging lock contention.<br> <br> - Added a manual for Drd.<br> <br> * A new experimental tool, exp-Ptrcheck, has been added. Ptrcheck<br> checks for misuses of pointers. In that sense it is a bit like<br> Memcheck. However, Ptrcheck can do things Memcheck can't: it can<br> detect overruns of stack and global arrays, it can detect<br> arbitrarily far out-of-bounds accesses to heap blocks, and it can<br> detect accesses heap blocks that have been freed a very long time<br> ago (millions of blocks in the past).<br> <br> Ptrcheck currently works only on x86-linux and amd64-linux. To use<br> it, use --tool=exp-ptrcheck. A simple manual is provided, as part<br> of the main Valgrind documentation. As this is an experimental<br> tool, we would be particularly interested in hearing about your<br> experiences with it.<br> <br> * exp-Omega, an experimental instantaneous leak-detecting tool, is no<br> longer built by default, although the code remains in the repository<br> and the tarball. This is due to three factors: a perceived lack of<br> users, a lack of maintenance, and concerns that it may not be<br> possible to achieve reliable operation using the existing design.<br> <br> * As usual, support for the latest Linux distros and toolchain<br> components has been added. It should work well on Fedora Core 10,<br> OpenSUSE 11.1 and Ubuntu 8.10. gcc-4.4 (in its current pre-release<br> state) is supported, as is glibc-2.9. The C++ demangler has been<br> updated so as to work well with C++ compiled by even the most recent<br> g++'s.<br> <br> * You can now use frame-level wildcards in suppressions. This was a<br> frequently-requested enhancement. A line "..." in a suppression now<br> matches zero or more frames. This makes it easier to write<br> suppressions which are precise yet insensitive to changes in<br> inlining behaviour.<br> <br> * 3.4.0 adds support on x86/amd64 for the SSSE3 instruction set.<br> <br> * Very basic support for IBM Power6 has been added (64-bit processes only).<br> <br> * Valgrind is now cross-compilable. For example, it is possible to<br> cross compile Valgrind on an x86/amd64-linux host, so that it runs<br> on a ppc32/64-linux target.<br> <br> * You can set the main thread's stack size at startup using the<br> new --main-stacksize= flag (subject of course to ulimit settings).<br> This is useful for running apps that need a lot of stack space.<br> <br> * The limitation that you can't use --trace-children=yes together<br> with --db-attach=yes has been removed.<br> <br> * The following bugs have been fixed. Note that "n-i-bz" stands for<br> "not in bugzilla" -- that is, a bug that was reported to us but<br> never got a bugzilla entry. We encourage you to file bugs in<br> bugzilla (http://bugs.kde.org/enter_valgrind_bug.cgi) rather than<br> mailing the developers (or mailing lists) directly.<br> <br> n-i-bz Make return types for some client requests 64-bit clean<br> n-i-bz glibc 2.9 support<br> n-i-bz ignore unsafe .valgrindrc's (CVE-2008-4865)<br> n-i-bz MPI_Init(0,0) is valid but libmpiwrap.c segfaults<br> n-i-bz Building in an env without gdb gives bogus gdb attach<br> 92456 Tracing the origin of uninitialised memory<br> 106497 Valgrind does not demangle some C++ template symbols<br> 162222 ==106497<br> 151612 Suppression with "..." (frame-level wildcards in .supp files)<br> 156404 Unable to start oocalc under memcheck on openSUSE 10.3 (64-bit)<br> 159285 unhandled syscall:25 (stime, on x86-linux)<br> 159452 unhandled ioctl 0x8B01 on "valgrind iwconfig"<br> 160954 ppc build of valgrind crashes with illegal instruction (isel)<br> 160956 mallinfo implementation, w/ patch<br> 162092 Valgrind fails to start gnome-system-monitor<br> 162819 malloc_free_fill test doesn't pass on glibc2.8 x86<br> 163794 assertion failure with "--track-origins=yes"<br> 163933 sigcontext.err and .trapno must be set together<br> 163955 remove constraint !(--db-attach=yes && --trace-children=yes)<br> 164476 Missing kernel module loading system calls<br> 164669 SVN regression: mmap() drops posix file locks<br> 166581 Callgrind output corruption when program forks<br> 167288 Patch file for missing system calls on Cell BE<br> 168943 unsupported scas instruction pentium<br> 171645 Unrecognised instruction (MOVSD, non-binutils encoding)<br> 172417 x86->IR: 0x82 ...<br> 172563 amd64->IR: 0xD9 0xF5 - fprem1<br> 173099 .lds linker script generation error<br> 173177 [x86_64] syscalls: 125/126/179 (capget/capset/quotactl)<br> 173751 amd64->IR: 0x48 0xF 0x6F 0x45 (even more redundant prefixes)<br> 174532 == 173751<br> 174908 --log-file value not expanded correctly for core file<br> 175044 Add lookup_dcookie for amd64<br> 175150 x86->IR: 0xF2 0xF 0x11 0xC1 (movss non-binutils encoding)<br> <br> Developer-visible changes:<br> <br> * Valgrind's debug-info reading machinery has been majorly overhauled.<br> It can now correctly establish the addresses for ELF data symbols,<br> which is something that has never worked properly before now.<br> <br> Also, Valgrind can now read DWARF3 type and location information for<br> stack and global variables. This makes it possible to use the<br> framework to build tools that rely on knowing the type and locations<br> of stack and global variables, for example exp-Ptrcheck.<br> <br> Reading of such information is disabled by default, because most<br> tools don't need it, and because it is expensive in space and time.<br> However, you can force Valgrind to read it, using the<br> --read-var-info=yes flag. Memcheck, Helgrind and DRD are able to<br> make use of such information, if present, to provide source-level<br> descriptions of data addresses in the error messages they create.<br> <br> (3.4.0.RC1: 24 Dec 2008, vex r1878, valgrind r8882).<br> (3.4.0: 3 Jan 2009, vex r1878, valgrind r8899).<br> <br> </p></div> </div> <div> <br><table class="nav" width="100%" cellspacing="3" cellpadding="2" border="0" summary="Navigation footer"> <tr> <td rowspan="2" width="40%" align="left"> <a accesskey="p" href="dist.authors.html"><< 1. AUTHORS</a> </td> <td width="20%" align="center"><a accesskey="u" href="dist.html">Up</a></td> <td rowspan="2" width="40%" align="right"> <a accesskey="n" href="dist.news.old.html">3. OLDER NEWS >></a> </td> </tr> <tr><td width="20%" align="center"><a accesskey="h" href="index.html">Home</a></td></tr> </table> </div> </body> </html>