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 281059 Cannot connect to Oracle using valgrind
28 390269 unhandled amd64-darwin syscall: unix:464 (openat_nocancel)
29 401284 False positive "Source and destination overlap in strncat"
30 428364 Signals inside io_uring_enter not handled
31 437790 valgrind reports "Conditional jump or move depends on uninitialised
32 value" in memchr of macOS 10.12-10.15
33 460616 disInstr(arm64): unhandled instruction 0x4E819402 (dotprod/ASIMDDP)
34 466884 Missing writev uninit padding suppression for _XSend
35 475498 Add reallocarray wrapper
36 476320 Build failure with GCC
37 476535 Difference in allocation size for massif/tests/overloaded-new between
38 clang++/libc++ and g++/libstdc++
39 476548 valgrind 3.22.0 fails on assertion when loading debuginfo file produced by mold
40 476708 valgrind-monitor.py regular expressions should use raw strings
41 476780 Extend strlcat and strlcpy wrappers to GNU libc
42 476787 Build of Valgrind 3.21.0 fails when SOLARIS_PT_SUNDWTRACE_THRP is
44 476887 WARNING: unhandled amd64-freebsd syscall: 578
45 477198 Add fchmodat2 syscall on linux
46 477628 Add mremap support for Solaris
47 477630 Include ucontext.h rather than sys/ucontext.h in Solaris sources
48 477719 vgdb incorrectly replies to qRcmd packet
49 478211 Redundant code for vgdb.c and Valgrind core tools
50 n-i-bz Add redirect for memccpy
52 To see details of a given bug, visit
53 https://bugs.kde.org/show_bug.cgi?id=XXXXXX
54 where XXXXXX is the bug number as listed above.
56 (3.23.0.RC1: ?? Apr 2024)
58 Release 3.22.0 (31 Oct 2023)
59 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~
61 This release supports X86/Linux, AMD64/Linux, ARM32/Linux, ARM64/Linux,
62 PPC32/Linux, PPC64BE/Linux, PPC64LE/Linux, S390X/Linux, MIPS32/Linux,
63 MIPS64/Linux, ARM/Android, ARM64/Android, MIPS32/Android, X86/Android,
64 X86/Solaris, AMD64/Solaris, AMD64/MacOSX 10.12, X86/FreeBSD and
65 AMD64/FreeBSD. There is also preliminary support for X86/macOS 10.13,
66 AMD64/macOS 10.13 and nanoMIPS/Linux.
68 * ==================== CORE CHANGES ===================
70 * A new configure option --with-gdbscripts-dir lets you install
71 the gdb valgrind python monitor scripts in a specific location.
72 For example a distro could use it to install the scripts in a
73 safe load location --with-gdbscripts-dir=%{_datadir}/gdb/auto-load
74 It is also possible to configure --without-gdb-scripts-dir so no
75 .debug_gdb_scripts section is added to the vgpreload library and
76 no valgrind-monitor python scripts are installed at all.
78 * ================== PLATFORM CHANGES =================
80 * Support has been added for FreeBSD 14 and FreeBSD 15.
81 * Add support for the folllowing FreeBSD system calls:
82 close_range, kqueuex, membarrier, timerfd_create,
83 timerfd_settime and timerfd_gettime (all added in FreeBSD 15).
85 * ==================== TOOL CHANGES ===================
87 * Memcheck now tests and warns about the values used for
88 alignment and size. These apply to various functions: memalign,
89 posix_memalign and aligned_alloc in C and various overloads
90 of operators new and delete in C++. The kinds of error that can
92 - invalid alignment, for instance the alignment is usually required
94 - mismatched alignment between aligned allocation and aligned
96 - mismatched size when sized delete is used
97 - bad size for functions that have implementation defined behaviour
98 when the requested size is zero
101 - You can now profile part of a program's execution using the new
102 `CACHEGRIND_START_INSTRUMENTATION` and `CACHEGRIND_STOP_INSTRUMENTATION`
103 client requests, along with the new `--instr-at-start` option. The
104 behaviour is the same as Callgrind's equivalent functionality.
106 * ==================== FIXED BUGS ====================
108 The following bugs have been fixed or resolved. Note that "n-i-bz"
109 stands for "not in bugzilla" -- that is, a bug that was reported to us
110 but never got a bugzilla entry. We encourage you to file bugs in
111 bugzilla (https://bugs.kde.org/enter_bug.cgi?product=valgrind) rather
112 than mailing the developers (or mailing lists) directly -- bugs that
113 are not entered into bugzilla tend to get forgotten about or ignored.
115 390871 ELF debug info reader confused with multiple .rodata* sections
116 417993 vbit-test fail on s390x with Iop_Add32: spurious dependency on uninit
117 426751 Valgrind reports "still reachable" memory using musl
118 (alpine running inside docker)
119 432801 Valgrind 3.16.1 reports a jump based on uninitialized memory somehow
120 related to clang and signals
121 433857 Add validation to C++17 aligned new/delete alignment size
122 433859 Add mismatched detection to C++ 17 aligned new/delete
123 460192 Add epoll_pwait2
124 461074 DWARF2 CFI reader: unhandled DW_OP_ 0x11 (consts) DW_OP_ 0x92 (bregx)
125 465782 s390x: Valgrind doesn't compile with Clang on s390x
126 466105 aligned_alloc problems, part 2
127 467441 Add mismatched detection to C++ 14 sized delete
128 469049 link failure on ppc64 (big endian) valgrind 3.20
129 469146 massif --ignore-fn does not ignore inlined functions
130 469768 Make it possible to install gdb scripts in a different location
131 470121 Can't run callgrind_control with valgrind 3.21.0 because of perl errors
132 470132 s390x: Assertion failure on VGM instruction
133 470520 Multiple realloc zero errors crash in MC_(eq_Error)
134 470713 Failure on the Yosys project: valgrind: m_libcfile.c:1802
135 (Bool vgPlain_realpath(const HChar *, HChar *)):
136 Assertion 'resolved' failed
137 470830 Don't print actions vgdb me ... continue for vgdb --multi mode
138 470978 s390x: Valgrind cannot start qemu-kvm when "sysctl vm.allocate_pgste=0"
139 471311 gdb --multi mode stdout redirecting to stderr
140 471807 Add support for lazy reading and downloading of DWARF debuginfo
141 472219 Syscall param ppoll(ufds.events) points to uninitialised byte(s)
142 472875 none/tests/s390x/dfp-1 failure
143 472963 Broken regular expression in configure.ac
144 473604 Fix bug472219.c compile failure with Clang 16
145 473677 make check compile failure with Clang 16 based on GCC 13.x
146 473745 must-be-redirected function - strlen
147 473870 FreeBSD 14 applications fail early at startup
148 473944 Handle mold linker split RW PT_LOAD segments correctly
149 474332 aligned_alloc under Valgrind returns nullptr when alignment is not a multiple of sizeof(void *)
150 475650 DRD does not work with C11 threads
151 475652 Missing suppression for __wcsncpy_avx2 (strncpy-avx2.S:308)?
152 476108 vg_replace_malloc DELETE checks size
153 n-i-bz Allow arguments with spaces in .valgrindrc files
154 n-i-bz FreeBSD fixed reading of Valgrind tools own debuginfo
156 To see details of a given bug, visit
157 https://bugs.kde.org/show_bug.cgi?id=XXXXXX
158 where XXXXXX is the bug number as listed above.
160 (3.22.0.RC1: 17 Oct 2023)
161 (3.22.0.RC2: 26 Oct 2023)
163 Release 3.21.0 (28 Apr 2023)
164 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~
166 This release supports X86/Linux, AMD64/Linux, ARM32/Linux, ARM64/Linux,
167 PPC32/Linux, PPC64BE/Linux, PPC64LE/Linux, S390X/Linux, MIPS32/Linux,
168 MIPS64/Linux, ARM/Android, ARM64/Android, MIPS32/Android, X86/Android,
169 X86/Solaris, AMD64/Solaris, AMD64/MacOSX 10.12, X86/FreeBSD and
170 AMD64/FreeBSD. There is also preliminary support for X86/macOS 10.13,
171 AMD64/macOS 10.13 and nanoMIPS/Linux.
173 * ==================== CORE CHANGES ===================
175 * When GDB is used to debug a program running under valgrind using
176 the valgrind gdbserver, GDB will automatically load some
177 python code provided in valgrind defining GDB front end commands
178 corresponding to the valgrind monitor commands.
179 These GDB front end commands accept the same format as
180 the monitor commands directly sent to the Valgrind gdbserver.
181 These GDB front end commands provide a better integration
182 in the GDB command line interface, so as to use for example
183 GDB auto-completion, command specific help, searching for
184 a command or command help matching a regexp, ...
185 For relevant monitor commands, GDB will evaluate arguments
186 to make the use of monitor commands easier.
187 For example, instead of having to print the address of a variable
188 to pass it to a subsequent monitor command, the GDB front end
189 command will evaluate the address argument. It is for example
191 (gdb) memcheck who_points_at &some_struct sizeof(some_struct)
194 $2 = (some_struct_type *) 0x1130a0 <some_struct>
195 (gdb) p sizeof(some_struct)
197 (gdb) monitor who_point_at 0x1130a0 40
199 * The vgdb utility now supports extended-remote protocol when
200 invoked with --multi. In this mode the GDB run command is
201 supported. Which means you don't need to run gdb and valgrind
202 from different terminals. So for example to start your program
203 in gdb and run it under valgrind you can do:
205 (gdb) set remote exec-file prog
207 (gdb) target extended-remote | vgdb --multi
210 * The behaviour of realloc with a size of zero can now
211 be changed for tools that intercept malloc. Those
212 tools are memcheck, helgrind, drd, massif and dhat.
213 Realloc implementations generally do one of two things
214 - free the memory like free() and return NULL
215 (GNU libc and ptmalloc).
216 - either free the memory and then allocate a
217 minimum sized block or just return the
218 original pointer. Return NULL if the
219 allocation of the minimum sized block fails
220 (jemalloc, musl, snmalloc, Solaris, macOS).
221 When Valgrind is configured and built it will
222 try to match the OS and libc behaviour. However
223 if you are using a non-default library to replace
224 malloc and family (e.g., musl on a glibc Linux or
225 tcmalloc on FreeBSD) then you can use a command line
226 option to change the behaviour of Valgrind:
227 --realloc-zero-bytes-frees=yes|no [yes on Linux glibc, no otherwise]
229 * ================== PLATFORM CHANGES =================
231 * Make the address space limit on FreeBSD amd64 128Gbytes
232 (the same as Linux and Solaris, it was 32Gbytes)
234 * ==================== TOOL CHANGES ===================
237 - When doing a delta leak_search, it is now possible to only
238 output the new loss records compared to the previous leak search.
239 This is available in the memcheck monitor command 'leak_search'
240 by specifying the "new" keyword or in your program by using
241 the client request VALGRIND_DO_NEW_LEAK_CHECK.
242 Whenever a "delta" leak search is done (i.e. when specifying
243 "new" or "increased" or "changed" in the monitor command),
244 the new loss records have a "new" marker.
245 - Valgrind now contains python code that defines GDB memcheck
246 front end monitor commands. See CORE CHANGES.
247 - Performs checks for the use of realloc with a size of zero.
248 This is non-portable and a source of errors. If memcheck
249 detects such a usage it will generate an error
250 realloc() with size 0
251 followed by the usual callstacks.
252 A switch has been added to allow this to be turned off:
253 --show-realloc-size-zero=yes|no [yes]
256 - The option ---history-backtrace-size=<number> allows to configure
257 the number of entries to record in the stack traces of "old"
258 accesses. Previously, this number was hardcoded to 8.
259 - Valgrind now contains python code that defines GDB helgrind
260 front end monitor commands. See CORE CHANGES.
263 - `--cache-sim=no` is now the default. The cache simulation is old and
264 unlikely to match any real modern machine. This means only the `Ir`
265 event are gathered by default, but that is by far the most useful
267 - `cg_annotate`, `cg_diff`, and `cg_merge` have been rewritten in
268 Python. As a result, they all have more flexible command line
269 argument handling, e.g. supporting `--show-percs` and
270 `--no-show-percs` forms as well as the existing `--show-percs=yes`
271 and `--show-percs=no`.
272 - `cg_annotate` has some functional changes.
273 - It's much faster, e.g. 3-4x on common cases.
274 - It now supports diffing (with `--diff`, `--mod-filename`, and
275 `--mod-funcname`) and merging (by passing multiple data files).
276 - It now provides more information at the file and function level.
277 There are now "File:function" and "Function:file" sections. These
278 are very useful for programs that use inlining a lot.
279 - Support for user-annotated files and the `-I`/`--include` option
280 has been removed, because it was of little use and blocked other
282 - The `--auto` option is renamed `--annotate`, though the old
283 `--auto=yes`/`--auto=no` forms are still supported.
284 - `cg_diff` and `cg_merge` are now deprecated, because `cg_annotate`
285 now does a better job of diffing and merging.
286 - The Cachegrind output file format has changed very slightly, but in
287 ways nobody is likely to notice.
290 - Valgrind now contains python code that defines GDB callgrind
291 front end monitor commands. See CORE CHANGES.
294 - Valgrind now contains python code that defines GDB massif
295 front end monitor commands. See CORE CHANGES.
298 - A new kind of user request has been added which allows you to
299 override the 1024 byte limit on access count histograms for blocks
300 of memory. The client request is DHAT_HISTOGRAM_MEMORY.
302 * ==================== FIXED BUGS ====================
304 The following bugs have been fixed or resolved. Note that "n-i-bz"
305 stands for "not in bugzilla" -- that is, a bug that was reported to us
306 but never got a bugzilla entry. We encourage you to file bugs in
307 bugzilla (https://bugs.kde.org/enter_bug.cgi?product=valgrind) rather
308 than mailing the developers (or mailing lists) directly -- bugs that
309 are not entered into bugzilla tend to get forgotten about or ignored.
311 170510 Don't warn about ioctl of size 0 without direction hint
312 241072 List tools in --help output
313 327548 false positive while destroying mutex
314 382034 Testcases build fixes for musl
315 351857 confusing error message about valid command line option
316 374596 inconsistent RDTSCP support on x86_64
317 392331 Spurious lock not held error from inside pthread_cond_timedwait
318 397083 Likely false positive "uninitialised value(s)" for __wmemchr_avx2 and __wmemcmp_avx2_movbe
319 400793 pthread_rwlock_timedwrlock false positive
320 419054 Unhandled syscall getcpu on arm32
321 433873 openat2 syscall unimplemented on Linux
322 434057 Add stdio mode to valgrind's gdbserver
323 435441 valgrind fails to interpose malloc on musl 1.2.2 due to weak symbol name and no libc soname
324 436413 Warn about realloc of size zero
325 439685 compiler warning in callgrind/main.c
326 444110 priv/guest_ppc_toIR.c:36198:31: warning: duplicated 'if' condition.
327 444487 hginfo test detects an extra lock inside data symbol "_rtld_local"
328 444488 Use glibc.pthread.stack_cache_size tunable
329 444568 drd/tests/pth_barrier_thr_cr fails on Fedora 38
330 445743 "The impossible happened: mutex is locked simultaneously by two threads"
331 while using mutexes with priority inheritance and signals
332 449309 Missing loopback device ioctl(s)
333 459476 vgdb: allow address reuse to avoid "address already in use" errorsuse" errors
334 460356 s390: Sqrt32Fx4 -- cannot reduce tree
335 462830 WARNING: unhandled amd64-freebsd syscall: 474
336 463027 broken check for MPX instruction support in assembler
337 464103 Enhancement: add a client request to DHAT to mark memory to be histogrammed
338 464476 Firefox fails to start under Valgrind
339 464609 Valgrind memcheck should support Linux pidfd_open
340 464680 Show issues caused by memory policies like selinux deny_execmem
341 464859 Build failures with GCC-13 (drd tsan_unittest)
342 464969 D language demangling
343 465435 m_libcfile.c:66 (vgPlain_safe_fd): Assertion 'newfd >= VG_(fd_hard_limit)' failed.
344 466104 aligned_alloc problems, part 1
345 467036 Add time cost statistics for Regtest
346 467482 Build failure on aarch64 Alpine
347 467714 fdleak_* and rlimit tests fail when parent process has more than
348 64 descriptors opened
349 467839 Gdbserver: Improve compatibility of library directory name
350 468401 [PATCH] Add a style file for clang-format
351 468556 Build failure for vgdb
352 468606 build: remove "Valgrind relies on GCC" check/output
353 469097 ppc64(be) doesn't support SCV syscall instruction
354 n-i-bz FreeBSD rfork syscall fail with EINVAL or ENOSYS rather than VG_(unimplemented)
356 To see details of a given bug, visit
357 https://bugs.kde.org/show_bug.cgi?id=XXXXXX
358 where XXXXXX is the bug number as listed above.
360 * ==================== KNOWN ISSUES ===================
362 * configure --enable-lto=yes is know to not work in all setups.
363 See bug 469049. Workaround: Build without LTO.
365 (3.21.0.RC1: 14 Apr 2023)
366 (3.21.0.RC2: 21 Apr 2023)
368 Release 3.20.0 (24 Oct 2022)
369 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~
371 This release supports X86/Linux, AMD64/Linux, ARM32/Linux, ARM64/Linux,
372 PPC32/Linux, PPC64BE/Linux, PPC64LE/Linux, S390X/Linux, MIPS32/Linux,
373 MIPS64/Linux, ARM/Android, ARM64/Android, MIPS32/Android, X86/Android,
374 X86/Solaris, AMD64/Solaris, AMD64/MacOSX 10.12, X86/FreeBSD and
375 AMD64/FreeBSD. There is also preliminary support for X86/macOS 10.13,
376 AMD64/macOS 10.13 and nanoMIPS/Linux.
378 * ==================== CORE CHANGES ===================
380 * The option "--vgdb-stop-at=event1,event2,..." accepts the new value abexit.
381 This indicates to invoke gdbserver when your program exits abnormally
382 (i.e. with a non zero exit code).
383 * Fix Rust v0 name demangling.
384 * The Linux rseq syscall is now implemented as (silently) returning ENOSYS.
385 * Add FreeBSD syscall wrappers for __specialfd and __realpathat.
386 * Remove FreeBSD dependencies on COMPAT10, which fixes compatibility with
388 * The option --enable-debuginfod=<no|yes> [default: yes] has been added on
390 * More DWARF5 support as generated by clang14.
392 * ==================== FIXED BUGS ====================
394 The following bugs have been fixed or resolved. Note that "n-i-bz"
395 stands for "not in bugzilla" -- that is, a bug that was reported to us
396 but never got a bugzilla entry. We encourage you to file bugs in
397 bugzilla (https://bugs.kde.org/enter_bug.cgi?product=valgrind) rather
398 than mailing the developers (or mailing lists) directly -- bugs that
399 are not entered into bugzilla tend to get forgotten about or ignored.
401 131186 writev reports error in (vector[...])
402 434764 iconv_open causes ld.so v2.28+ to use optimised strncmp
403 446754 Improve error codes from alloc functions under memcheck
404 452274 memcheck crashes with Assertion 'sci->status.what == SsIdle' failed
405 452779 Valgrind fails to build on FreeBSD 13.0 with llvm-devel (15.0.0)
406 453055 shared_timed_mutex drd test fails with "Lock shared failed" message
407 453602 Missing command line option to enable/disable debuginfod
408 452802 Handle lld 9+ split RW PT_LOAD segments correctly
409 454040 s390x: False-positive memcheck:cond in memmem on arch13 systems
410 456171 [PATCH] FreeBSD: Don't record address errors when accessing the 'kern.ps_strings' sysctl struct
411 n-i-bz Implement vgdb invoker on FreeBSD
412 458845 PowerPC: The L field for the dcbf and sync instruction should be
414 458915 Remove register cache to fix 458915 gdbserver causes wrong syscall return
415 459031 Documentation on --error-exitcode incomplete
416 459477 XERROR messages lacks ending '\n' in vgdb
417 462007 Implicit int in none/tests/faultstatus.c
419 To see details of a given bug, visit
420 https://bugs.kde.org/show_bug.cgi?id=XXXXXX
421 where XXXXXX is the bug number as listed above.
423 (3.20.0.RC1: 20 Oct 2022)
426 Release 3.19.0 (11 Apr 2022)
427 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~
429 This release supports X86/Linux, AMD64/Linux, ARM32/Linux, ARM64/Linux,
430 PPC32/Linux, PPC64BE/Linux, PPC64LE/Linux, S390X/Linux, MIPS32/Linux,
431 MIPS64/Linux, ARM/Android, ARM64/Android, MIPS32/Android, X86/Android,
432 X86/Solaris, AMD64/Solaris, AMD64/MacOSX 10.12, X86/FreeBSD and
433 AMD64/FreeBSD. There is also preliminary support for X86/macOS 10.13,
434 AMD64/macOS 10.13 and nanoMIPS/Linux.
436 * ==================== CORE CHANGES ===================
438 * Fix Rust v0 name demangling.
439 * The Linux rseq syscall is now implemented as (silently) returning ENOSYS.
440 * Add FreeBSD syscall wrappers for __specialfd and __realpathat.
441 * Remove FreeBSD dependencies on COMPAT10, which fixes compatibility with HardenedBSD
443 * ================== PLATFORM CHANGES =================
446 - ignore the "v8.x" architecture levels, only look at actual CPU features
447 present. Fixes mismatch detected between RDMA and atomics features
448 preventing startup on some QEMU configurations.
449 - Implement LD{,A}XP and ST{,L}XP
450 - Fix incorrect code emitted for doubleword CAS.
453 - Fix sys_ipc semtimedop syscall
454 - Fix VFLRX and WFLRX instructions
455 - Fix EXRL instruction with negative offset
458 - Reimplement the vbpermq instruction support to generate less Iops and
459 avoid overflowing internal buffers.
460 - Fix checking for scv support to avoid "Facility 'SCV' unavailable (12),
461 exception" messages in dmsg.
462 - Fix setting condition code for Vector Compare quad word instructions.
463 - Fix fix lxsibzx, lxsihzx and lxsihzx instructions so they only load
464 their respective sized data.
465 - Fix the prefixed stq instruction in PC relative mode.
467 * ==================== TOOL CHANGES ===================
470 - Speed up --track-origins=yes for large (in the range of hundreds to
471 thousands of megabytes) mmap/munmaps.
473 - Several fixes for new versions of libstd++ using new posix try_lock
476 * ==================== FIXED BUGS ====================
478 The following bugs have been fixed or resolved. Note that "n-i-bz"
479 stands for "not in bugzilla" -- that is, a bug that was reported to us
480 but never got a bugzilla entry. We encourage you to file bugs in
481 bugzilla (https://bugs.kde.org/enter_bug.cgi?product=valgrind) rather
482 than mailing the developers (or mailing lists) directly -- bugs that
483 are not entered into bugzilla tend to get forgotten about or ignored.
485 403802 leak_cpp_interior fails with some reachable blocks different than expected
486 435732 memcheck/tests/leak_cpp_interior fails with gcc11
487 444242 s390x: Valgrind crashes on EXRL with negative offset
488 444399 arm64: unhandled instruction 0xC87F2D89 (LD{,A}XP and ST{,L}XP).
490 444481 gdb_server test failures on s390x
491 444495 dhat/tests/copy fails on s390x
492 444552 memcheck/tests/sem fails on s390x with glibc 2.34
493 444571 PPC, fix the lxsibzx and lxsihzx so they only load their respective
495 444836 PPC, pstq instruction for R=1 is not storing to the correct address.
496 444925 fexecve syscall wrapper not properly implemented
497 445032 valgrind/memcheck crash with SIGSEGV when SIGVTALRM timer used and
499 445211 Fix out of tree builds
500 445300 [PATCH] Fix building tests with Musl
501 445011 SIGCHLD is sent when valgrind uses debuginfod-find
502 445354 arm64 backend: incorrect code emitted for doubleword CAS
503 445415 arm64 front end: alignment checks missing for atomic instructions
504 445504 Using C++ condition_variable results in bogus "mutex is locked simultaneously by two threads" warning
505 445607 Unhandled amd64-freebsd syscall: 247
506 445668 Inline stack frame generation is broken for Rust binaries
507 445916 Demangle Rust v0 symbols with .llvm suffix
508 446139 DRD/Helgrind with std::shared_timed_mutex::try_lock_until and try_lock_shared_until false positives
509 446138 DRD/Helgrind with std::timed_mutex::try_lock_until false positives
510 446281 Add a DRD suppression for fwrite
511 446103 Memcheck: `--track-origins=yes` causes extreme slowdowns for large mmap/munmap
512 446139 DRD/Helgrind with std::shared_timed_mutex::try_lock_until and try_lock_shared_until false
513 446251 TARGET_SIGNAL_THR added to enum target_signal
514 446823 FreeBSD - missing syscalls when using libzm4
515 447991 s390x: Valgrind indicates illegal instruction on wflrx
516 447995 Valgrind segfault on power10 due to hwcap checking code
517 449483 Powerpc: vcmpgtsq., vcmpgtuq,, vcmpequq. instructions not setting the
518 condition code correctly.
519 449672 ppc64 --track-origins=yes failures because of bad cmov addHRegUse
520 449838 sigsegv liburing the 'impossible' happened for io_uring_setup
521 450025 Powerc: ACC file not implemented as a logical overlay of the VSR
523 450437 Warn for execve syscall with argv or argv[0] being NULL
524 450536 Powerpc: valgrind throws 'facility scv unavailable exception'
525 451626 Syscall param bpf(attr->raw_tracepoint.name) points to unaddressable byte(s)
526 451827 [ppc64le] VEX temporary storage exhausted with several vbpermq instructions
527 451843 valgrind fails to start on a FreeBSD system which enforces W^X
529 To see details of a given bug, visit
530 https://bugs.kde.org/show_bug.cgi?id=XXXXXX
531 where XXXXXX is the bug number as listed above.
533 (3.19.0.RC1: 02 Apr 2022)
534 (3.19.0.RC2: 08 Apr 2022)