1 Release 3.23.0 (?? Apr 2024)
2 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~
4 This release supports X86/Linux, AMD64/Linux, ARM32/Linux, ARM64/Linux,
5 PPC32/Linux, PPC64BE/Linux, PPC64LE/Linux, S390X/Linux, MIPS32/Linux,
6 MIPS64/Linux, ARM/Android, ARM64/Android, MIPS32/Android, X86/Android,
7 X86/Solaris, AMD64/Solaris, AMD64/MacOSX 10.12, X86/FreeBSD and
8 AMD64/FreeBSD. There is also preliminary support for X86/macOS 10.13,
9 AMD64/macOS 10.13 and nanoMIPS/Linux.
11 * ==================== CORE CHANGES ===================
13 * ================== PLATFORM CHANGES =================
15 * ==================== TOOL CHANGES ===================
17 * ==================== FIXED BUGS ====================
19 The following bugs have been fixed or resolved. Note that "n-i-bz"
20 stands for "not in bugzilla" -- that is, a bug that was reported to us
21 but never got a bugzilla entry. We encourage you to file bugs in
22 bugzilla (https://bugs.kde.org/enter_bug.cgi?product=valgrind) rather
23 than mailing the developers (or mailing lists) directly -- bugs that
24 are not entered into bugzilla tend to get forgotten about or ignored.
27 401284 False positive "Source and destination overlap in strncat"
28 437790 valgrind reports "Conditional jump or move depends on uninitialised
29 value" in memchr of macOS 10.12-10.15
30 460616 disInstr(arm64): unhandled instruction 0x4E819402 (dotprod/ASIMDDP)
31 475498 Add reallocarray wrapper
32 476320 Build failure with GCC
33 476535 Difference in allocation size for massif/tests/overloaded-new between
34 clang++/libc++ and g++/libstdc++
35 476780 Extend strlcat and strlcpy wrappers to GNU libc
36 476787 Build of Valgrind 3.21.0 fails when SOLARIS_PT_SUNDWTRACE_THRP is
39 To see details of a given bug, visit
40 https://bugs.kde.org/show_bug.cgi?id=XXXXXX
41 where XXXXXX is the bug number as listed above.
43 (3.23.0.RC1: ?? Apr 2024)
45 Release 3.22.0 (31 Oct 2023)
46 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~
48 This release supports X86/Linux, AMD64/Linux, ARM32/Linux, ARM64/Linux,
49 PPC32/Linux, PPC64BE/Linux, PPC64LE/Linux, S390X/Linux, MIPS32/Linux,
50 MIPS64/Linux, ARM/Android, ARM64/Android, MIPS32/Android, X86/Android,
51 X86/Solaris, AMD64/Solaris, AMD64/MacOSX 10.12, X86/FreeBSD and
52 AMD64/FreeBSD. There is also preliminary support for X86/macOS 10.13,
53 AMD64/macOS 10.13 and nanoMIPS/Linux.
55 * ==================== CORE CHANGES ===================
57 * A new configure option --with-gdbscripts-dir lets you install
58 the gdb valgrind python monitor scripts in a specific location.
59 For example a distro could use it to install the scripts in a
60 safe load location --with-gdbscripts-dir=%{_datadir}/gdb/auto-load
61 It is also possible to configure --without-gdb-scripts-dir so no
62 .debug_gdb_scripts section is added to the vgpreload library and
63 no valgrind-monitor python scripts are installed at all.
65 * ================== PLATFORM CHANGES =================
67 * Support has been added for FreeBSD 14 and FreeBSD 15.
68 * Add support for the folllowing FreeBSD system calls:
69 close_range, kqueuex, membarrier, timerfd_create,
70 timerfd_settime and timerfd_gettime (all added in FreeBSD 15).
72 * ==================== TOOL CHANGES ===================
74 * Memcheck now tests and warns about the values used for
75 alignment and size. These apply to various functions: memalign,
76 posix_memalign and aligned_alloc in C and various overloads
77 of operators new and delete in C++. The kinds of error that can
79 - invalid alignment, for instance the alignment is usually required
81 - mismatched alignment between aligned allocation and aligned
83 - mismatched size when sized delete is used
84 - bad size for functions that have implementation defined behaviour
85 when the requested size is zero
88 - You can now profile part of a program's execution using the new
89 `CACHEGRIND_START_INSTRUMENTATION` and `CACHEGRIND_STOP_INSTRUMENTATION`
90 client requests, along with the new `--instr-at-start` option. The
91 behaviour is the same as Callgrind's equivalent functionality.
93 * ==================== FIXED BUGS ====================
95 The following bugs have been fixed or resolved. Note that "n-i-bz"
96 stands for "not in bugzilla" -- that is, a bug that was reported to us
97 but never got a bugzilla entry. We encourage you to file bugs in
98 bugzilla (https://bugs.kde.org/enter_bug.cgi?product=valgrind) rather
99 than mailing the developers (or mailing lists) directly -- bugs that
100 are not entered into bugzilla tend to get forgotten about or ignored.
102 390871 ELF debug info reader confused with multiple .rodata* sections
103 417993 vbit-test fail on s390x with Iop_Add32: spurious dependency on uninit
104 426751 Valgrind reports "still reachable" memory using musl
105 (alpine running inside docker)
106 432801 Valgrind 3.16.1 reports a jump based on uninitialized memory somehow
107 related to clang and signals
108 433857 Add validation to C++17 aligned new/delete alignment size
109 433859 Add mismatched detection to C++ 17 aligned new/delete
110 460192 Add epoll_pwait2
111 461074 DWARF2 CFI reader: unhandled DW_OP_ 0x11 (consts) DW_OP_ 0x92 (bregx)
112 465782 s390x: Valgrind doesn't compile with Clang on s390x
113 466105 aligned_alloc problems, part 2
114 467441 Add mismatched detection to C++ 14 sized delete
115 469049 link failure on ppc64 (big endian) valgrind 3.20
116 469146 massif --ignore-fn does not ignore inlined functions
117 469768 Make it possible to install gdb scripts in a different location
118 470121 Can't run callgrind_control with valgrind 3.21.0 because of perl errors
119 470132 s390x: Assertion failure on VGM instruction
120 470520 Multiple realloc zero errors crash in MC_(eq_Error)
121 470713 Failure on the Yosys project: valgrind: m_libcfile.c:1802
122 (Bool vgPlain_realpath(const HChar *, HChar *)):
123 Assertion 'resolved' failed
124 470830 Don't print actions vgdb me ... continue for vgdb --multi mode
125 470978 s390x: Valgrind cannot start qemu-kvm when "sysctl vm.allocate_pgste=0"
126 471311 gdb --multi mode stdout redirecting to stderr
127 471807 Add support for lazy reading and downloading of DWARF debuginfo
128 472219 Syscall param ppoll(ufds.events) points to uninitialised byte(s)
129 472875 none/tests/s390x/dfp-1 failure
130 472963 Broken regular expression in configure.ac
131 473604 Fix bug472219.c compile failure with Clang 16
132 473677 make check compile failure with Clang 16 based on GCC 13.x
133 473745 must-be-redirected function - strlen
134 473870 FreeBSD 14 applications fail early at startup
135 473944 Handle mold linker split RW PT_LOAD segments correctly
136 474332 aligned_alloc under Valgrind returns nullptr when alignment is not a multiple of sizeof(void *)
137 475650 DRD does not work with C11 threads
138 475652 Missing suppression for __wcsncpy_avx2 (strncpy-avx2.S:308)?
139 476108 vg_replace_malloc DELETE checks size
140 n-i-bz Allow arguments with spaces in .valgrindrc files
141 n-i-bz FreeBSD fixed reading of Valgrind tools own debuginfo
143 To see details of a given bug, visit
144 https://bugs.kde.org/show_bug.cgi?id=XXXXXX
145 where XXXXXX is the bug number as listed above.
147 (3.22.0.RC1: 17 Oct 2023)
148 (3.22.0.RC2: 26 Oct 2023)
150 Release 3.21.0 (28 Apr 2023)
151 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~
153 This release supports X86/Linux, AMD64/Linux, ARM32/Linux, ARM64/Linux,
154 PPC32/Linux, PPC64BE/Linux, PPC64LE/Linux, S390X/Linux, MIPS32/Linux,
155 MIPS64/Linux, ARM/Android, ARM64/Android, MIPS32/Android, X86/Android,
156 X86/Solaris, AMD64/Solaris, AMD64/MacOSX 10.12, X86/FreeBSD and
157 AMD64/FreeBSD. There is also preliminary support for X86/macOS 10.13,
158 AMD64/macOS 10.13 and nanoMIPS/Linux.
160 * ==================== CORE CHANGES ===================
162 * When GDB is used to debug a program running under valgrind using
163 the valgrind gdbserver, GDB will automatically load some
164 python code provided in valgrind defining GDB front end commands
165 corresponding to the valgrind monitor commands.
166 These GDB front end commands accept the same format as
167 the monitor commands directly sent to the Valgrind gdbserver.
168 These GDB front end commands provide a better integration
169 in the GDB command line interface, so as to use for example
170 GDB auto-completion, command specific help, searching for
171 a command or command help matching a regexp, ...
172 For relevant monitor commands, GDB will evaluate arguments
173 to make the use of monitor commands easier.
174 For example, instead of having to print the address of a variable
175 to pass it to a subsequent monitor command, the GDB front end
176 command will evaluate the address argument. It is for example
178 (gdb) memcheck who_points_at &some_struct sizeof(some_struct)
181 $2 = (some_struct_type *) 0x1130a0 <some_struct>
182 (gdb) p sizeof(some_struct)
184 (gdb) monitor who_point_at 0x1130a0 40
186 * The vgdb utility now supports extended-remote protocol when
187 invoked with --multi. In this mode the GDB run command is
188 supported. Which means you don't need to run gdb and valgrind
189 from different terminals. So for example to start your program
190 in gdb and run it under valgrind you can do:
192 (gdb) set remote exec-file prog
194 (gdb) target extended-remote | vgdb --multi
197 * The behaviour of realloc with a size of zero can now
198 be changed for tools that intercept malloc. Those
199 tools are memcheck, helgrind, drd, massif and dhat.
200 Realloc implementations generally do one of two things
201 - free the memory like free() and return NULL
202 (GNU libc and ptmalloc).
203 - either free the memory and then allocate a
204 minimum sized block or just return the
205 original pointer. Return NULL if the
206 allocation of the minimum sized block fails
207 (jemalloc, musl, snmalloc, Solaris, macOS).
208 When Valgrind is configured and built it will
209 try to match the OS and libc behaviour. However
210 if you are using a non-default library to replace
211 malloc and family (e.g., musl on a glibc Linux or
212 tcmalloc on FreeBSD) then you can use a command line
213 option to change the behaviour of Valgrind:
214 --realloc-zero-bytes-frees=yes|no [yes on Linux glibc, no otherwise]
216 * ================== PLATFORM CHANGES =================
218 * Make the address space limit on FreeBSD amd64 128Gbytes
219 (the same as Linux and Solaris, it was 32Gbytes)
221 * ==================== TOOL CHANGES ===================
224 - When doing a delta leak_search, it is now possible to only
225 output the new loss records compared to the previous leak search.
226 This is available in the memcheck monitor command 'leak_search'
227 by specifying the "new" keyword or in your program by using
228 the client request VALGRIND_DO_NEW_LEAK_CHECK.
229 Whenever a "delta" leak search is done (i.e. when specifying
230 "new" or "increased" or "changed" in the monitor command),
231 the new loss records have a "new" marker.
232 - Valgrind now contains python code that defines GDB memcheck
233 front end monitor commands. See CORE CHANGES.
234 - Performs checks for the use of realloc with a size of zero.
235 This is non-portable and a source of errors. If memcheck
236 detects such a usage it will generate an error
237 realloc() with size 0
238 followed by the usual callstacks.
239 A switch has been added to allow this to be turned off:
240 --show-realloc-size-zero=yes|no [yes]
243 - The option ---history-backtrace-size=<number> allows to configure
244 the number of entries to record in the stack traces of "old"
245 accesses. Previously, this number was hardcoded to 8.
246 - Valgrind now contains python code that defines GDB helgrind
247 front end monitor commands. See CORE CHANGES.
250 - `--cache-sim=no` is now the default. The cache simulation is old and
251 unlikely to match any real modern machine. This means only the `Ir`
252 event are gathered by default, but that is by far the most useful
254 - `cg_annotate`, `cg_diff`, and `cg_merge` have been rewritten in
255 Python. As a result, they all have more flexible command line
256 argument handling, e.g. supporting `--show-percs` and
257 `--no-show-percs` forms as well as the existing `--show-percs=yes`
258 and `--show-percs=no`.
259 - `cg_annotate` has some functional changes.
260 - It's much faster, e.g. 3-4x on common cases.
261 - It now supports diffing (with `--diff`, `--mod-filename`, and
262 `--mod-funcname`) and merging (by passing multiple data files).
263 - It now provides more information at the file and function level.
264 There are now "File:function" and "Function:file" sections. These
265 are very useful for programs that use inlining a lot.
266 - Support for user-annotated files and the `-I`/`--include` option
267 has been removed, because it was of little use and blocked other
269 - The `--auto` option is renamed `--annotate`, though the old
270 `--auto=yes`/`--auto=no` forms are still supported.
271 - `cg_diff` and `cg_merge` are now deprecated, because `cg_annotate`
272 now does a better job of diffing and merging.
273 - The Cachegrind output file format has changed very slightly, but in
274 ways nobody is likely to notice.
277 - Valgrind now contains python code that defines GDB callgrind
278 front end monitor commands. See CORE CHANGES.
281 - Valgrind now contains python code that defines GDB massif
282 front end monitor commands. See CORE CHANGES.
285 - A new kind of user request has been added which allows you to
286 override the 1024 byte limit on access count histograms for blocks
287 of memory. The client request is DHAT_HISTOGRAM_MEMORY.
289 * ==================== FIXED BUGS ====================
291 The following bugs have been fixed or resolved. Note that "n-i-bz"
292 stands for "not in bugzilla" -- that is, a bug that was reported to us
293 but never got a bugzilla entry. We encourage you to file bugs in
294 bugzilla (https://bugs.kde.org/enter_bug.cgi?product=valgrind) rather
295 than mailing the developers (or mailing lists) directly -- bugs that
296 are not entered into bugzilla tend to get forgotten about or ignored.
298 170510 Don't warn about ioctl of size 0 without direction hint
299 241072 List tools in --help output
300 327548 false positive while destroying mutex
301 382034 Testcases build fixes for musl
302 351857 confusing error message about valid command line option
303 374596 inconsistent RDTSCP support on x86_64
304 392331 Spurious lock not held error from inside pthread_cond_timedwait
305 397083 Likely false positive "uninitialised value(s)" for __wmemchr_avx2 and __wmemcmp_avx2_movbe
306 400793 pthread_rwlock_timedwrlock false positive
307 419054 Unhandled syscall getcpu on arm32
308 433873 openat2 syscall unimplemented on Linux
309 434057 Add stdio mode to valgrind's gdbserver
310 435441 valgrind fails to interpose malloc on musl 1.2.2 due to weak symbol name and no libc soname
311 436413 Warn about realloc of size zero
312 439685 compiler warning in callgrind/main.c
313 444110 priv/guest_ppc_toIR.c:36198:31: warning: duplicated 'if' condition.
314 444487 hginfo test detects an extra lock inside data symbol "_rtld_local"
315 444488 Use glibc.pthread.stack_cache_size tunable
316 444568 drd/tests/pth_barrier_thr_cr fails on Fedora 38
317 445743 "The impossible happened: mutex is locked simultaneously by two threads"
318 while using mutexes with priority inheritance and signals
319 449309 Missing loopback device ioctl(s)
320 459476 vgdb: allow address reuse to avoid "address already in use" errorsuse" errors
321 460356 s390: Sqrt32Fx4 -- cannot reduce tree
322 462830 WARNING: unhandled amd64-freebsd syscall: 474
323 463027 broken check for MPX instruction support in assembler
324 464103 Enhancement: add a client request to DHAT to mark memory to be histogrammed
325 464476 Firefox fails to start under Valgrind
326 464609 Valgrind memcheck should support Linux pidfd_open
327 464680 Show issues caused by memory policies like selinux deny_execmem
328 464859 Build failures with GCC-13 (drd tsan_unittest)
329 464969 D language demangling
330 465435 m_libcfile.c:66 (vgPlain_safe_fd): Assertion 'newfd >= VG_(fd_hard_limit)' failed.
331 466104 aligned_alloc problems, part 1
332 467036 Add time cost statistics for Regtest
333 467482 Build failure on aarch64 Alpine
334 467714 fdleak_* and rlimit tests fail when parent process has more than
335 64 descriptors opened
336 467839 Gdbserver: Improve compatibility of library directory name
337 468401 [PATCH] Add a style file for clang-format
338 468556 Build failure for vgdb
339 468606 build: remove "Valgrind relies on GCC" check/output
340 469097 ppc64(be) doesn't support SCV syscall instruction
341 n-i-bz FreeBSD rfork syscall fail with EINVAL or ENOSYS rather than VG_(unimplemented)
343 To see details of a given bug, visit
344 https://bugs.kde.org/show_bug.cgi?id=XXXXXX
345 where XXXXXX is the bug number as listed above.
347 * ==================== KNOWN ISSUES ===================
349 * configure --enable-lto=yes is know to not work in all setups.
350 See bug 469049. Workaround: Build without LTO.
352 (3.21.0.RC1: 14 Apr 2023)
353 (3.21.0.RC2: 21 Apr 2023)
355 Release 3.20.0 (24 Oct 2022)
356 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~
358 This release supports X86/Linux, AMD64/Linux, ARM32/Linux, ARM64/Linux,
359 PPC32/Linux, PPC64BE/Linux, PPC64LE/Linux, S390X/Linux, MIPS32/Linux,
360 MIPS64/Linux, ARM/Android, ARM64/Android, MIPS32/Android, X86/Android,
361 X86/Solaris, AMD64/Solaris, AMD64/MacOSX 10.12, X86/FreeBSD and
362 AMD64/FreeBSD. There is also preliminary support for X86/macOS 10.13,
363 AMD64/macOS 10.13 and nanoMIPS/Linux.
365 * ==================== CORE CHANGES ===================
367 * The option "--vgdb-stop-at=event1,event2,..." accepts the new value abexit.
368 This indicates to invoke gdbserver when your program exits abnormally
369 (i.e. with a non zero exit code).
370 * Fix Rust v0 name demangling.
371 * The Linux rseq syscall is now implemented as (silently) returning ENOSYS.
372 * Add FreeBSD syscall wrappers for __specialfd and __realpathat.
373 * Remove FreeBSD dependencies on COMPAT10, which fixes compatibility with
375 * The option --enable-debuginfod=<no|yes> [default: yes] has been added on
377 * More DWARF5 support as generated by clang14.
379 * ==================== FIXED BUGS ====================
381 The following bugs have been fixed or resolved. Note that "n-i-bz"
382 stands for "not in bugzilla" -- that is, a bug that was reported to us
383 but never got a bugzilla entry. We encourage you to file bugs in
384 bugzilla (https://bugs.kde.org/enter_bug.cgi?product=valgrind) rather
385 than mailing the developers (or mailing lists) directly -- bugs that
386 are not entered into bugzilla tend to get forgotten about or ignored.
388 131186 writev reports error in (vector[...])
389 434764 iconv_open causes ld.so v2.28+ to use optimised strncmp
390 446754 Improve error codes from alloc functions under memcheck
391 452274 memcheck crashes with Assertion 'sci->status.what == SsIdle' failed
392 452779 Valgrind fails to build on FreeBSD 13.0 with llvm-devel (15.0.0)
393 453055 shared_timed_mutex drd test fails with "Lock shared failed" message
394 453602 Missing command line option to enable/disable debuginfod
395 452802 Handle lld 9+ split RW PT_LOAD segments correctly
396 454040 s390x: False-positive memcheck:cond in memmem on arch13 systems
397 456171 [PATCH] FreeBSD: Don't record address errors when accessing the 'kern.ps_strings' sysctl struct
398 n-i-bz Implement vgdb invoker on FreeBSD
399 458845 PowerPC: The L field for the dcbf and sync instruction should be
401 458915 Remove register cache to fix 458915 gdbserver causes wrong syscall return
402 459031 Documentation on --error-exitcode incomplete
403 459477 XERROR messages lacks ending '\n' in vgdb
404 462007 Implicit int in none/tests/faultstatus.c
406 To see details of a given bug, visit
407 https://bugs.kde.org/show_bug.cgi?id=XXXXXX
408 where XXXXXX is the bug number as listed above.
410 (3.20.0.RC1: 20 Oct 2022)
413 Release 3.19.0 (11 Apr 2022)
414 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~
416 This release supports X86/Linux, AMD64/Linux, ARM32/Linux, ARM64/Linux,
417 PPC32/Linux, PPC64BE/Linux, PPC64LE/Linux, S390X/Linux, MIPS32/Linux,
418 MIPS64/Linux, ARM/Android, ARM64/Android, MIPS32/Android, X86/Android,
419 X86/Solaris, AMD64/Solaris, AMD64/MacOSX 10.12, X86/FreeBSD and
420 AMD64/FreeBSD. There is also preliminary support for X86/macOS 10.13,
421 AMD64/macOS 10.13 and nanoMIPS/Linux.
423 * ==================== CORE CHANGES ===================
425 * Fix Rust v0 name demangling.
426 * The Linux rseq syscall is now implemented as (silently) returning ENOSYS.
427 * Add FreeBSD syscall wrappers for __specialfd and __realpathat.
428 * Remove FreeBSD dependencies on COMPAT10, which fixes compatibility with HardenedBSD
430 * ================== PLATFORM CHANGES =================
433 - ignore the "v8.x" architecture levels, only look at actual CPU features
434 present. Fixes mismatch detected between RDMA and atomics features
435 preventing startup on some QEMU configurations.
436 - Implement LD{,A}XP and ST{,L}XP
437 - Fix incorrect code emitted for doubleword CAS.
440 - Fix sys_ipc semtimedop syscall
441 - Fix VFLRX and WFLRX instructions
442 - Fix EXRL instruction with negative offset
445 - Reimplement the vbpermq instruction support to generate less Iops and
446 avoid overflowing internal buffers.
447 - Fix checking for scv support to avoid "Facility 'SCV' unavailable (12),
448 exception" messages in dmsg.
449 - Fix setting condition code for Vector Compare quad word instructions.
450 - Fix fix lxsibzx, lxsihzx and lxsihzx instructions so they only load
451 their respective sized data.
452 - Fix the prefixed stq instruction in PC relative mode.
454 * ==================== TOOL CHANGES ===================
457 - Speed up --track-origins=yes for large (in the range of hundreds to
458 thousands of megabytes) mmap/munmaps.
460 - Several fixes for new versions of libstd++ using new posix try_lock
463 * ==================== FIXED BUGS ====================
465 The following bugs have been fixed or resolved. Note that "n-i-bz"
466 stands for "not in bugzilla" -- that is, a bug that was reported to us
467 but never got a bugzilla entry. We encourage you to file bugs in
468 bugzilla (https://bugs.kde.org/enter_bug.cgi?product=valgrind) rather
469 than mailing the developers (or mailing lists) directly -- bugs that
470 are not entered into bugzilla tend to get forgotten about or ignored.
472 403802 leak_cpp_interior fails with some reachable blocks different than expected
473 435732 memcheck/tests/leak_cpp_interior fails with gcc11
474 444242 s390x: Valgrind crashes on EXRL with negative offset
475 444399 arm64: unhandled instruction 0xC87F2D89 (LD{,A}XP and ST{,L}XP).
477 444481 gdb_server test failures on s390x
478 444495 dhat/tests/copy fails on s390x
479 444552 memcheck/tests/sem fails on s390x with glibc 2.34
480 444571 PPC, fix the lxsibzx and lxsihzx so they only load their respective
482 444836 PPC, pstq instruction for R=1 is not storing to the correct address.
483 444925 fexecve syscall wrapper not properly implemented
484 445032 valgrind/memcheck crash with SIGSEGV when SIGVTALRM timer used and
486 445211 Fix out of tree builds
487 445300 [PATCH] Fix building tests with Musl
488 445011 SIGCHLD is sent when valgrind uses debuginfod-find
489 445354 arm64 backend: incorrect code emitted for doubleword CAS
490 445415 arm64 front end: alignment checks missing for atomic instructions
491 445504 Using C++ condition_variable results in bogus "mutex is locked simultaneously by two threads" warning
492 445607 Unhandled amd64-freebsd syscall: 247
493 445668 Inline stack frame generation is broken for Rust binaries
494 445916 Demangle Rust v0 symbols with .llvm suffix
495 446139 DRD/Helgrind with std::shared_timed_mutex::try_lock_until and try_lock_shared_until false positives
496 446138 DRD/Helgrind with std::timed_mutex::try_lock_until false positives
497 446281 Add a DRD suppression for fwrite
498 446103 Memcheck: `--track-origins=yes` causes extreme slowdowns for large mmap/munmap
499 446139 DRD/Helgrind with std::shared_timed_mutex::try_lock_until and try_lock_shared_until false
500 446251 TARGET_SIGNAL_THR added to enum target_signal
501 446823 FreeBSD - missing syscalls when using libzm4
502 447991 s390x: Valgrind indicates illegal instruction on wflrx
503 447995 Valgrind segfault on power10 due to hwcap checking code
504 449483 Powerpc: vcmpgtsq., vcmpgtuq,, vcmpequq. instructions not setting the
505 condition code correctly.
506 449672 ppc64 --track-origins=yes failures because of bad cmov addHRegUse
507 449838 sigsegv liburing the 'impossible' happened for io_uring_setup
508 450025 Powerc: ACC file not implemented as a logical overlay of the VSR
510 450437 Warn for execve syscall with argv or argv[0] being NULL
511 450536 Powerpc: valgrind throws 'facility scv unavailable exception'
512 451626 Syscall param bpf(attr->raw_tracepoint.name) points to unaddressable byte(s)
513 451827 [ppc64le] VEX temporary storage exhausted with several vbpermq instructions
514 451843 valgrind fails to start on a FreeBSD system which enforces W^X
516 To see details of a given bug, visit
517 https://bugs.kde.org/show_bug.cgi?id=XXXXXX
518 where XXXXXX is the bug number as listed above.
520 (3.19.0.RC1: 02 Apr 2022)
521 (3.19.0.RC2: 08 Apr 2022)