| Elliott Hughes | a0664b9 | 2017-04-18 17:46:52 -0700 | [diff] [blame] | 1 | <html> |
| 2 | <head> |
| 3 | <meta http-equiv="Content-Type" content="text/html; charset=ISO-8859-1"> |
| 4 | <title>2. NEWS</title> |
| 5 | <link rel="stylesheet" type="text/css" href="vg_basic.css"> |
| Elliott Hughes | ed39800 | 2017-06-21 14:41:24 -0700 | [diff] [blame^] | 6 | <meta name="generator" content="DocBook XSL Stylesheets V1.79.1"> |
| Elliott Hughes | a0664b9 | 2017-04-18 17:46:52 -0700 | [diff] [blame] | 7 | <link rel="home" href="index.html" title="Valgrind Documentation"> |
| 8 | <link rel="up" href="dist.html" title="Valgrind Distribution Documents"> |
| 9 | <link rel="prev" href="dist.authors.html" title="1. AUTHORS"> |
| 10 | <link rel="next" href="dist.news.old.html" title="3. OLDER NEWS"> |
| 11 | </head> |
| 12 | <body bgcolor="white" text="black" link="#0000FF" vlink="#840084" alink="#0000FF"> |
| 13 | <div><table class="nav" width="100%" cellspacing="3" cellpadding="3" border="0" summary="Navigation header"><tr> |
| 14 | <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> |
| 15 | <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> |
| 16 | <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> |
| 17 | <th align="center" valign="middle">Valgrind Distribution Documents</th> |
| 18 | <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> |
| 19 | </tr></table></div> |
| 20 | <div class="chapter"> |
| 21 | <div class="titlepage"><div><div><h1 class="title"> |
| 22 | <a name="dist.news"></a>2. NEWS</h1></div></div></div> |
| 23 | <div class="literallayout"><p><br> |
| Elliott Hughes | ed39800 | 2017-06-21 14:41:24 -0700 | [diff] [blame^] | 24 | Release 3.13.0 (15 June 2017)<br> |
| 25 | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~<br> |
| 26 | <br> |
| 27 | 3.13.0 is a feature release with many improvements and the usual collection of<br> |
| 28 | bug fixes.<br> |
| 29 | <br> |
| 30 | This release supports X86/Linux, AMD64/Linux, ARM32/Linux, ARM64/Linux,<br> |
| 31 | PPC32/Linux, PPC64BE/Linux, PPC64LE/Linux, S390X/Linux, MIPS32/Linux,<br> |
| 32 | MIPS64/Linux, ARM/Android, ARM64/Android, MIPS32/Android, X86/Android,<br> |
| 33 | X86/Solaris, AMD64/Solaris and AMD64/MacOSX 10.12.<br> |
| 34 | <br> |
| 35 | * ==================== CORE CHANGES ===================<br> |
| 36 | <br> |
| 37 | * The translation cache size has been increased to keep up with the demands of<br> |
| 38 | large applications. The maximum number of sectors has increased from 24 to<br> |
| 39 | 48. The default number of sectors has increased from 16 to 32 on all<br> |
| 40 | targets except Android, where the increase is from 6 to 12.<br> |
| 41 | <br> |
| 42 | * The amount of memory that Valgrind can use has been increased from 64GB to<br> |
| 43 | 128GB. In particular this means your application can allocate up to about<br> |
| 44 | 60GB when running on Memcheck.<br> |
| 45 | <br> |
| 46 | * Valgrind's default load address has been changed from 0x3800'0000 to<br> |
| 47 | 0x5800'0000, so as to make it possible to load larger executables. This<br> |
| 48 | should make it possible to load executables of size at least 1200MB.<br> |
| 49 | <br> |
| 50 | * A massive spaceleak caused by reading compressed debuginfo files has been<br> |
| 51 | fixed. Valgrind should now be entirely usable with gcc-7.0 "-gz" created<br> |
| 52 | debuginfo.<br> |
| 53 | <br> |
| 54 | * The C++ demangler has been updated.<br> |
| 55 | <br> |
| 56 | * Support for demangling Rust symbols has been added.<br> |
| 57 | <br> |
| 58 | * A new representation of stack traces, the "XTree", has been added. An XTree<br> |
| 59 | is a tree of stacktraces with data associated with the stacktraces. This is<br> |
| 60 | used by various tools (Memcheck, Helgrind, Massif) to report on the heap<br> |
| 61 | consumption of your program. Reporting is controlled by the new options<br> |
| 62 | --xtree-memory=none|allocs|full and --xtree-memory-file=<file>.<br> |
| 63 | <br> |
| 64 | A report can also be produced on demand using the gdbserver monitor command<br> |
| 65 | 'xtmemory [<filename>]>'. The XTree can be output in 2 formats: 'callgrind<br> |
| 66 | format' and 'massif format. The existing visualisers for these formats (e.g.<br> |
| 67 | callgrind_annotate, KCachegrind, ms_print) can be used to visualise and<br> |
| 68 | analyse these reports.<br> |
| 69 | <br> |
| 70 | Memcheck can also produce XTree leak reports using the Callgrind file<br> |
| 71 | format. For more details, see the user manual.<br> |
| 72 | <br> |
| 73 | * ================== PLATFORM CHANGES =================<br> |
| 74 | <br> |
| 75 | * ppc64: support for ISA 3.0B and various fixes for existing 3.0 support<br> |
| 76 | <br> |
| 77 | * amd64: fixes for JIT failure problems on long AVX2 code blocks<br> |
| 78 | <br> |
| 79 | * amd64 and x86: support for CET prefixes has been added<br> |
| 80 | <br> |
| 81 | * arm32: a few missing ARMv8 instructions have been implemented<br> |
| 82 | <br> |
| 83 | * arm64, mips64, mips32: an alternative implementation of Load-Linked and<br> |
| 84 | Store-Conditional instructions has been added. This is to deal with<br> |
| 85 | processor implementations that implement the LL/SC specifications strictly<br> |
| 86 | and as a result cause Valgrind to hang in certain situations. The<br> |
| 87 | alternative implementation is automatically enabled at startup, as required.<br> |
| 88 | You can use the option --sim-hints=fallback-llsc to force-enable it if you<br> |
| 89 | want.<br> |
| 90 | <br> |
| 91 | * Support for OSX 10.12 has been improved.<br> |
| 92 | <br> |
| 93 | * On Linux, clone handling has been improved to honour CLONE_VFORK that<br> |
| 94 | involves a child stack. Note however that CLONE_VFORK | CLONE_VM is handled<br> |
| 95 | like CLONE_VFORK (by removing CLONE_VM), so applications that depend on<br> |
| 96 | CLONE_VM exact semantics will (still) not work.<br> |
| 97 | <br> |
| 98 | * The TileGX/Linux port has been removed because it appears to be both unused<br> |
| 99 | and unsupported.<br> |
| 100 | <br> |
| 101 | * ==================== TOOL CHANGES ====================<br> |
| 102 | <br> |
| 103 | * Memcheck:<br> |
| 104 | <br> |
| 105 | - Memcheck should give fewer false positives when running optimised<br> |
| 106 | Clang/LLVM generated code.<br> |
| 107 | <br> |
| 108 | - Support for --xtree-memory and 'xtmemory [<filename>]>'.<br> |
| 109 | <br> |
| 110 | - New command line options --xtree-leak=no|yes and --xtree-leak-file=<file><br> |
| 111 | to produce the end of execution leak report in a xtree callgrind format<br> |
| 112 | file.<br> |
| 113 | <br> |
| 114 | - New option 'xtleak' in the memcheck leak_check monitor command, to produce<br> |
| 115 | the leak report in an xtree file.<br> |
| 116 | <br> |
| 117 | * Massif:<br> |
| 118 | <br> |
| 119 | - Support for --xtree-memory and 'xtmemory [<filename>]>'.<br> |
| 120 | <br> |
| 121 | - For some workloads (typically, for big applications), Massif memory<br> |
| 122 | consumption and CPU consumption has decreased significantly.<br> |
| 123 | <br> |
| 124 | * Helgrind:<br> |
| 125 | <br> |
| 126 | - Support for --xtree-memory and 'xtmemory [<filename>]>'.<br> |
| 127 | <br> |
| 128 | - addition of client request VALGRIND_HG_GNAT_DEPENDENT_MASTER_JOIN, useful<br> |
| 129 | for Ada gnat compiled applications.<br> |
| 130 | <br> |
| 131 | * ==================== OTHER CHANGES ====================<br> |
| 132 | <br> |
| 133 | * For Valgrind developers: in an outer/inner setup, the outer Valgrind will<br> |
| 134 | append the inner guest stacktrace to the inner host stacktrace. This helps<br> |
| 135 | to investigate the errors reported by the outer, when they are caused by the<br> |
| 136 | inner guest program (such as an inner regtest). See README_DEVELOPERS for<br> |
| 137 | more info.<br> |
| 138 | <br> |
| 139 | * To allow fast detection of callgrind files by desktop environments and file<br> |
| 140 | managers, the format was extended to have an optional first line that<br> |
| 141 | uniquely identifies the format ("# callgrind format"). Callgrind creates<br> |
| 142 | this line now, as does the new xtree functionality.<br> |
| 143 | <br> |
| 144 | * File name template arguments (such as --log-file, --xtree-memory-file, ...)<br> |
| 145 | have a new %n format letter that is replaced by a sequence number.<br> |
| 146 | <br> |
| 147 | * "--version -v" now shows the SVN revision numbers from which Valgrind was<br> |
| 148 | built.<br> |
| 149 | <br> |
| 150 | * ==================== FIXED BUGS ====================<br> |
| 151 | <br> |
| 152 | The following bugs have been fixed or resolved. Note that "n-i-bz"<br> |
| 153 | stands for "not in bugzilla" -- that is, a bug that was reported to us<br> |
| 154 | but never got a bugzilla entry. We encourage you to file bugs in<br> |
| 155 | bugzilla (https://bugs.kde.org/enter_bug.cgi?product=valgrind) rather<br> |
| 156 | than mailing the developers (or mailing lists) directly -- bugs that<br> |
| 157 | are not entered into bugzilla tend to get forgotten about or ignored.<br> |
| 158 | <br> |
| 159 | To see details of a given bug, visit<br> |
| 160 | https://bugs.kde.org/show_bug.cgi?id=XXXXXX<br> |
| 161 | where XXXXXX is the bug number as listed below.<br> |
| 162 | <br> |
| 163 | 162848 --log-file output isn't split when a program forks<br> |
| 164 | 340777 Illegal instruction on mips (ar71xx)<br> |
| 165 | 341481 MIPS64: Iop_CmpNE32 triggers false warning on MIPS64 platforms<br> |
| 166 | 342040 Valgrind mishandles clone with CLONE_VFORK | CLONE_VM that clones<br> |
| 167 | to a different stack.<br> |
| 168 | 344139 x86 stack-seg overrides, needed by the Wine people<br> |
| 169 | 344524 store conditional of guest applications always fail - observed on<br> |
| 170 | Octeon3(MIPS)<br> |
| 171 | 348616 Wine/valgrind: noted but unhandled ioctl 0x5390 [..] (DVD_READ_STRUCT)<br> |
| 172 | 352395 Please provide SVN revision info in --version -v<br> |
| 173 | 352767 Wine/valgrind: noted but unhandled ioctl 0x5307 [..] (CDROMSTOP)<br> |
| 174 | 356374 Assertion 'DRD_(g_threadinfo)[tid].pt_threadid !=<br> |
| 175 | INVALID_POSIX_THREADID' failed<br> |
| 176 | 358213 helgrind/drd bar_bad testcase hangs or crashes with new glibc pthread<br> |
| 177 | barrier implementation<br> |
| 178 | 358697 valgrind.h: Some code remains even when defining NVALGRIND<br> |
| 179 | 359202 Add musl libc configure/compile<br> |
| 180 | 360415 amd64 instructions ADCX and ADOX are not implemented in VEX<br> |
| 181 | == 372828 (vex amd64->IR: 0x66 0xF 0x3A 0x62 0x4A 0x10)<br> |
| 182 | 360429 unhandled ioctl 0x530d with no size/direction hints (CDROMREADMODE1)<br> |
| 183 | 362223 assertion failed when .valgrindrc is a directory instead of a file<br> |
| 184 | 367543 bt/btc/btr/bts x86/x86_64 instructions are poorly-handled wrt flags<br> |
| 185 | 367942 Segfault vgPlain_do_sys_sigaction (m_signals.c:1138)<br> |
| 186 | 368507 can't malloc chunks larger than about 34GB<br> |
| 187 | 368529 Android arm target link error, missing atexit and pthread_atfork<br> |
| 188 | 368863 WARNING: unhandled arm64-linux syscall: 100 (get_robust_list)<br> |
| 189 | 368865 WARNING: unhandled arm64-linux syscall: 272 (kcmp)<br> |
| 190 | 368868 disInstr(arm64): unhandled instruction 0xD53BE000 = cntfrq_el0 (ARMv8)<br> |
| 191 | 368917 WARNING: unhandled arm64-linux syscall: 218 (request_key)<br> |
| 192 | 368918 WARNING: unhandled arm64-linux syscall: 127 (sched_rr_get_interval)<br> |
| 193 | 368922 WARNING: unhandled arm64-linux syscall: 161 (sethostname)<br> |
| 194 | 368924 WARNING: unhandled arm64-linux syscall: 84 (sync_file_range)<br> |
| 195 | 368925 WARNING: unhandled arm64-linux syscall: 130 (tkill)<br> |
| 196 | 368926 WARNING: unhandled arm64-linux syscall: 97 (unshare)<br> |
| 197 | 369459 valgrind on arm64 violates the ARMv8 spec (ldxr/stxr)<br> |
| 198 | 370028 Reduce the number of compiler warnings on MIPS platforms<br> |
| 199 | 370635 arm64 missing syscall getcpu<br> |
| 200 | 371225 Fix order of timer_{gettime,getoverrun,settime} syscalls on arm64<br> |
| 201 | 371227 Clean AArch64 syscall table<br> |
| 202 | 371412 Rename wrap_sys_shmat to sys_shmat like other wrappers<br> |
| 203 | 371471 Valgrind complains about non legit memory leaks on placement new (C++)<br> |
| 204 | 371491 handleAddrOverrides() is [incorrect] when ASO prefix is used<br> |
| 205 | 371503 disInstr(arm64): unhandled instruction 0xF89F0000<br> |
| 206 | 371869 support '%' in symbol Z-encoding<br> |
| 207 | 371916 execution tree xtree concept<br> |
| 208 | 372120 c++ demangler demangles symbols which are not c++<br> |
| 209 | 372185 Support of valgrind on ARMv8 with 32 bit executable<br> |
| 210 | 372188 vex amd64->IR: 0x66 0xF 0x3A 0x62 0x4A 0x10 0x10 0x48 (PCMPxSTRx $0x10)<br> |
| 211 | 372195 Power PC, xxsel instruction is not always recognized.<br> |
| 212 | 372504 Hanging on exit_group<br> |
| 213 | 372600 process loops forever when fatal signals are arriving quickly<br> |
| 214 | 372794 LibVEX (arm32 front end): 'Assertion szBlg2 <= 3' failed<br> |
| 215 | 373046 Stacks registered by core are never deregistered<br> |
| 216 | 373069 memcheck/tests/leak_cpp_interior fails with GCC 5.1+<br> |
| 217 | 373086 Implement additional Xen hypercalls<br> |
| 218 | 373192 Calling posix_spawn in glibc 2.24 completely broken<br> |
| 219 | 373488 Support for fanotify API on ARM64 architecture<br> |
| 220 | == 368864 WARNING: unhandled arm64-linux syscall: 262 (fanotify_init)<br> |
| 221 | 373555 Rename BBPTR to GSPTR as it denotes guest state pointer only<br> |
| 222 | 373938 const IRExpr arguments for matchIRExpr()<br> |
| 223 | 374719 some spelling fixes<br> |
| 224 | 374963 increase valgrind's load address to prevent mmap failure<br> |
| 225 | 375514 valgrind_get_tls_addr() does not work in case of static TLS<br> |
| 226 | 375772 +1 error in get_elf_symbol_info() when computing value of 'hi' address<br> |
| 227 | for ML_(find_rx_mapping)()<br> |
| 228 | 375806 Test helgrind/tests/tc22_exit_w_lock fails with glibc 2.24<br> |
| 229 | 375839 Temporary storage exhausted, with long sequence of vfmadd231ps insns<br> |
| 230 | == 377159 "vex: the `impossible' happened" still present<br> |
| 231 | == 375150 Assertion 'tres.status == VexTransOK' failed<br> |
| 232 | == 378068 valgrind crashes on AVX2 function in FFmpeg<br> |
| 233 | 376142 Segfaults on MIPS Cavium Octeon boards<br> |
| 234 | 376279 disInstr(arm64): unhandled instruction 0xD50320FF<br> |
| 235 | 376455 Solaris: unhandled syscall lgrpsys(180)<br> |
| 236 | 376518 Solaris: unhandled fast trap getlgrp(6)<br> |
| 237 | 376611 ppc64 and arm64 don't know about prlimit64 syscall<br> |
| 238 | 376729 PPC64, remove R2 from the clobber list<br> |
| 239 | == 371668<br> |
| 240 | 376956 syswrap of SNDDRV and DRM_IOCTL_VERSION causing some addresses<br> |
| 241 | to be wrongly marked as addressable<br> |
| 242 | 377066 Some Valgrind unit tests fail to compile on Ubuntu 16.10 with<br> |
| 243 | PIE enabled by default<br> |
| 244 | 377376 memcheck/tests/linux/getregset fails with glibc2.24<br> |
| 245 | 377427 PPC64, lxv instruction failing on odd destination register <br> |
| 246 | 377478 PPC64: ISA 3.0 setup fixes<br> |
| 247 | 377698 Missing memory check for futex() uaddr arg for FUTEX_WAKE<br> |
| 248 | and FUTEX_WAKE_BITSET, check only 4 args for FUTEX_WAKE_BITSET,<br> |
| 249 | and 2 args for FUTEX_TRYLOCK_PI<br> |
| 250 | 377717 Fix massive space leak when reading compressed debuginfo sections<br> |
| 251 | 377891 Update Xen 4.6 domctl wrappers<br> |
| 252 | 377930 fcntl syscall wrapper is missing flock structure check<br> |
| 253 | 378524 libvexmultiarch_test regression on s390x and ppc64<br> |
| 254 | 378535 Valgrind reports INTERNAL ERROR in execve syscall wrapper<br> |
| 255 | 378673 Update libiberty demangler<br> |
| 256 | 378931 Add ISA 3.0B additional isnstructions, add OV32, CA32 setting support<br> |
| 257 | 379039 syscall wrapper for prctl(PR_SET_NAME) must not check more than 16 bytes<br> |
| 258 | 379094 Valgrind reports INTERNAL ERROR in rt_sigsuspend syscall wrapper<br> |
| 259 | 379371 UNKNOWN task message [id 3444, to mach_task_self(), reply 0x603]<br> |
| 260 | (task_register_dyld_image_infos)<br> |
| 261 | 379372 UNKNOWN task message [id 3447, to mach_task_self(), reply 0x603]<br> |
| 262 | (task_register_dyld_shared_cache_image_info)<br> |
| 263 | 379390 unhandled syscall: mach:70 (host_create_mach_voucher_trap)<br> |
| 264 | 379473 MIPS: add support for rdhwr cycle counter register<br> |
| 265 | 379504 remove TileGX/Linux port<br> |
| 266 | 379525 Support more x86 nop opcodes<br> |
| 267 | 379838 disAMode(x86): not an addr!<br> |
| 268 | 379703 PC ISA 3.0 fixes: stxvx, stxv, xscmpexpdp instructions<br> |
| 269 | 379890 arm: unhandled instruction: 0xEBAD 0x1B05 (sub.w fp, sp, r5, lsl #4)<br> |
| 270 | 379895 clock_gettime does not execute POST syscall wrapper<br> |
| 271 | 379925 PPC64, mtffs does not set the FPCC and C bits in the FPSCR correctly<br> |
| 272 | 379966 WARNING: unhandled amd64-linux syscall: 313 (finit_module)<br> |
| 273 | 380200 xtree generated callgrind files refer to files without directory name<br> |
| 274 | 380202 Assertion failure for cache line size (cls == 64) on aarch64.<br> |
| 275 | 380397 s390x: __GI_strcspn() replacement needed<br> |
| 276 | n-i-bz Fix pub_tool_basics.h build issue with g++ 4.4.7.<br> |
| 277 | <br> |
| 278 | (3.13.0.RC1: 2 June 2017, vex r3386, valgrind r16434)<br> |
| 279 | (3.13.0.RC2: 9 June 2017, vex r3389, valgrind r16443)<br> |
| 280 | (3.13.0: 14 June 2017, vex r3396, valgrind r16446)<br> |
| 281 | <br> |
| 282 | <br> |
| 283 | <br> |
| Elliott Hughes | a0664b9 | 2017-04-18 17:46:52 -0700 | [diff] [blame] | 284 | Release 3.12.0 (20 October 2016)<br> |
| 285 | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~<br> |
| 286 | <br> |
| 287 | 3.12.0 is a feature release with many improvements and the usual<br> |
| 288 | collection of bug fixes.<br> |
| 289 | <br> |
| 290 | This release supports X86/Linux, AMD64/Linux, ARM32/Linux,<br> |
| 291 | ARM64/Linux, PPC32/Linux, PPC64BE/Linux, PPC64LE/Linux, S390X/Linux,<br> |
| 292 | MIPS32/Linux, MIPS64/Linux, ARM/Android, ARM64/Android,<br> |
| 293 | MIPS32/Android, X86/Android, X86/Solaris, AMD64/Solaris, X86/MacOSX<br> |
| 294 | 10.10 and AMD64/MacOSX 10.10. There is also preliminary support for<br> |
| 295 | X86/MacOSX 10.11/12, AMD64/MacOSX 10.11/12 and TILEGX/Linux.<br> |
| 296 | <br> |
| 297 | * ================== PLATFORM CHANGES =================<br> |
| 298 | <br> |
| 299 | * POWER: Support for ISA 3.0 has been added<br> |
| 300 | <br> |
| 301 | * mips: support for O32 FPXX ABI has been added.<br> |
| 302 | * mips: improved recognition of different processors<br> |
| 303 | * mips: determination of page size now done at run time<br> |
| 304 | <br> |
| 305 | * amd64: Partial support for AMD FMA4 instructions.<br> |
| 306 | <br> |
| 307 | * arm, arm64: Support for v8 crypto and CRC instructions.<br> |
| 308 | <br> |
| 309 | * Improvements and robustification of the Solaris port.<br> |
| 310 | <br> |
| 311 | * Preliminary support for MacOS 10.12 (Sierra) has been added.<br> |
| 312 | <br> |
| 313 | Whilst 3.12.0 continues to support the 32-bit x86 instruction set, we<br> |
| 314 | would prefer users to migrate to 64-bit x86 (a.k.a amd64 or x86_64)<br> |
| 315 | where possible. Valgrind's support for 32-bit x86 has stagnated in<br> |
| 316 | recent years and has fallen far behind that for 64-bit x86<br> |
| 317 | instructions. By contrast 64-bit x86 is well supported, up to and<br> |
| 318 | including AVX2.<br> |
| 319 | <br> |
| 320 | * ==================== TOOL CHANGES ====================<br> |
| 321 | <br> |
| 322 | * Memcheck:<br> |
| 323 | <br> |
| 324 | - Added meta mempool support for describing a custom allocator which:<br> |
| 325 | - Auto-frees all chunks assuming that destroying a pool destroys all<br> |
| 326 | objects in the pool<br> |
| 327 | - Uses itself to allocate other memory blocks<br> |
| 328 | <br> |
| 329 | - New flag --ignore-range-below-sp to ignore memory accesses below<br> |
| 330 | the stack pointer, if you really have to. The related flag<br> |
| 331 | --workaround-gcc296-bugs=yes is now deprecated. Use<br> |
| 332 | --ignore-range-below-sp=1024-1 as a replacement.<br> |
| 333 | <br> |
| 334 | * DRD:<br> |
| 335 | <br> |
| 336 | - Improved thread startup time significantly on non-Linux platforms.<br> |
| 337 | <br> |
| 338 | * DHAT<br> |
| 339 | <br> |
| 340 | - Added collection of the metric "tot-blocks-allocd"<br> |
| 341 | <br> |
| 342 | * ==================== OTHER CHANGES ====================<br> |
| 343 | <br> |
| 344 | * Replacement/wrapping of malloc/new related functions is now done not just<br> |
| 345 | for system libraries by default, but for any globally defined malloc/new<br> |
| 346 | related function (both in shared libraries and statically linked alternative<br> |
| 347 | malloc implementations). The dynamic (runtime) linker is excluded, though.<br> |
| 348 | To only intercept malloc/new related functions in<br> |
| 349 | system libraries use --soname-synonyms=somalloc=nouserintercepts (where<br> |
| 350 | "nouserintercepts" can be any non-existing library name).<br> |
| 351 | This new functionality is not implemented for MacOS X.<br> |
| 352 | <br> |
| 353 | * The maximum number of callers in a suppression entry is now equal to<br> |
| 354 | the maximum size for --num-callers (500).<br> |
| 355 | Note that --gen-suppressions=yes|all similarly generates suppressions<br> |
| 356 | containing up to --num-callers frames.<br> |
| 357 | <br> |
| 358 | * New and modified GDB server monitor features:<br> |
| 359 | <br> |
| 360 | - Valgrind's gdbserver now accepts the command 'catch syscall'.<br> |
| 361 | Note that you must have GDB >= 7.11 to use 'catch syscall' with<br> |
| 362 | gdbserver.<br> |
| 363 | <br> |
| 364 | * New option --run-cxx-freeres=<yes|no> can be used to change whether<br> |
| 365 | __gnu_cxx::__freeres() cleanup function is called or not. Default is<br> |
| 366 | 'yes'.<br> |
| 367 | <br> |
| 368 | * Valgrind is able to read compressed debuginfo sections in two formats:<br> |
| 369 | - zlib ELF gABI format with SHF_COMPRESSED flag (gcc option -gz=zlib)<br> |
| 370 | - zlib GNU format with .zdebug sections (gcc option -gz=zlib-gnu)<br> |
| 371 | <br> |
| 372 | * Modest JIT-cost improvements: the cost of instrumenting code blocks<br> |
| 373 | for the most common use case (x86_64-linux, Memcheck) has been<br> |
| 374 | reduced by 10%-15%.<br> |
| 375 | <br> |
| 376 | * Improved performance for programs that do a lot of discarding of<br> |
| 377 | instruction address ranges of 8KB or less.<br> |
| 378 | <br> |
| 379 | * The C++ symbol demangler has been updated.<br> |
| 380 | <br> |
| 381 | * More robustness against invalid syscall parameters on Linux.<br> |
| 382 | <br> |
| 383 | * ==================== FIXED BUGS ====================<br> |
| 384 | <br> |
| 385 | The following bugs have been fixed or resolved. Note that "n-i-bz"<br> |
| 386 | stands for "not in bugzilla" -- that is, a bug that was reported to us<br> |
| 387 | but never got a bugzilla entry. We encourage you to file bugs in<br> |
| 388 | bugzilla (https://bugs.kde.org/enter_bug.cgi?product=valgrind) rather<br> |
| 389 | than mailing the developers (or mailing lists) directly -- bugs that<br> |
| 390 | are not entered into bugzilla tend to get forgotten about or ignored.<br> |
| 391 | <br> |
| 392 | To see details of a given bug, visit<br> |
| 393 | https://bugs.kde.org/show_bug.cgi?id=XXXXXX<br> |
| 394 | where XXXXXX is the bug number as listed below.<br> |
| 395 | <br> |
| 396 | 191069 Exiting due to signal not reported in XML output<br> |
| 397 | 199468 Suppressions: stack size limited to 25<br> |
| 398 | while --num-callers allows more frames<br> |
| 399 | 212352 vex amd64 unhandled opc_aux = 0x 2, first_opcode == 0xDC (FCOM)<br> |
| 400 | 278744 cvtps2pd with redundant RexW<br> |
| 401 | 303877 valgrind doesn't support compressed debuginfo sections.<br> |
| 402 | 345307 Warning about "still reachable" memory when using libstdc++ from gcc 5<br> |
| 403 | 348345 Assertion fails for negative lineno<br> |
| Elliott Hughes | ed39800 | 2017-06-21 14:41:24 -0700 | [diff] [blame^] | 404 | 348924 MIPS: Load doubles through memory so the code compiles with the FPXX ABI<br> |
| Elliott Hughes | a0664b9 | 2017-04-18 17:46:52 -0700 | [diff] [blame] | 405 | 351282 V 3.10.1 MIPS softfloat build broken with GCC 4.9.3 / binutils 2.25.1<br> |
| 406 | 351692 Dumps created by valgrind are not readable by gdb (mips32 specific)<br> |
| 407 | 351804 Crash on generating suppressions for "printf" call on OS X 10.10<br> |
| 408 | 352197 mips: mmap2() not wrapped correctly for page size > 4096<br> |
| 409 | 353083 arm64 doesn't implement various xattr system calls<br> |
| 410 | 353084 arm64 doesn't support sigpending system call<br> |
| 411 | 353137 www: update info for Supported Platforms<br> |
| 412 | 353138 www: update "The Valgrind Developers" page<br> |
| 413 | 353370 don't advertise RDRAND in cpuid for Core-i7-4910-like avx2 machine<br> |
| 414 | == 365325<br> |
| 415 | == 357873<br> |
| 416 | 353384 amd64->IR: 0x66 0xF 0x3A 0x62 0xD1 0x62 (pcmpXstrX $0x62)<br> |
| 417 | 353398 WARNING: unhandled amd64-solaris syscall: 207<br> |
| 418 | 353660 XML in auxwhat tag not escaping reserved symbols properly<br> |
| 419 | 353680 s390x: Crash with certain glibc versions due to non-implemented TBEGIN<br> |
| 420 | 353727 amd64->IR: 0x66 0xF 0x3A 0x62 0xD1 0x72 (pcmpXstrX $0x72)<br> |
| 421 | 353802 ELF debug info reader confused with multiple .rodata sections<br> |
| 422 | 353891 Assert 'bad_scanned_addr < VG_ROUNDDN(start+len, sizeof(Addr))' failed<br> |
| 423 | 353917 unhandled amd64-solaris syscall fchdir(120)<br> |
| 424 | 353920 unhandled amd64-solaris syscall: 170<br> |
| 425 | 354274 arm: unhandled instruction: 0xEBAD 0x0AC1 (sub.w sl, sp, r1, lsl #3)<br> |
| 426 | 354392 unhandled amd64-solaris syscall: 171<br> |
| 427 | 354797 Vbit test does not include Iops for Power 8 instruction support<br> |
| 428 | 354883 tst->os_state.pthread - magic_delta assertion failure on OSX 10.11<br> |
| 429 | == 361351<br> |
| 430 | == 362920<br> |
| 431 | == 366222<br> |
| 432 | 354933 Fix documentation of --kernel-variant=android-no-hw-tls option<br> |
| 433 | 355188 valgrind should intercept all malloc related global functions<br> |
| 434 | 355454 do not intercept malloc related symbols from the runtime linker<br> |
| 435 | 355455 stderr.exp of test cases wrapmalloc and wrapmallocstatic overconstrained<br> |
| 436 | 356044 Dwarf line info reader misinterprets is_stmt register<br> |
| 437 | 356112 mips: replace addi with addiu<br> |
| 438 | 356393 valgrind (vex) crashes because isZeroU happened<br> |
| 439 | == 363497<br> |
| 440 | == 364497<br> |
| 441 | 356676 arm64-linux: unhandled syscalls 125, 126 (sched_get_priority_max/min)<br> |
| 442 | 356678 arm64-linux: unhandled syscall 232 (mincore)<br> |
| 443 | 356817 valgrind.h triggers compiler errors on MSVC when defining NVALGRIND<br> |
| 444 | 356823 Unsupported ARM instruction: stlex<br> |
| 445 | 357059 x86/amd64: SSE cvtpi2ps with memory source does transition to MMX state<br> |
| 446 | 357338 Unhandled instruction for SHA instructions libcrypto Boring SSL<br> |
| 447 | 357673 crash if I try to run valgrind with a binary link with libcurl<br> |
| 448 | 357833 Setting RLIMIT_DATA to zero breaks with linux 4.5+<br> |
| 449 | 357871 pthread_spin_destroy not properly wrapped<br> |
| 450 | 357887 Calls to VG_(fclose) do not close the file descriptor<br> |
| 451 | 357932 amd64->IR: accept redundant REX prefixes for {minsd,maxsd} m128, xmm.<br> |
| 452 | 358030 support direct socket calls on x86 32bit (new in linux 4.3)<br> |
| 453 | 358478 drd/tests/std_thread.cpp doesn't build with GCC6<br> |
| 454 | 359133 Assertion 'eltSzB <= ddpa->poolSzB' failed<br> |
| 455 | 359181 Buffer Overflow during Demangling<br> |
| 456 | 359201 futex syscall "skips" argument 5 if op is FUTEX_WAIT_BITSET<br> |
| 457 | 359289 s390x: popcnt (B9E1) not implemented<br> |
| 458 | 359472 The Power PC vsubuqm instruction doesn't always give the correct result<br> |
| 459 | 359503 Add missing syscalls for aarch64 (arm64)<br> |
| 460 | 359645 "You need libc6-dbg" help message could be more helpful<br> |
| 461 | 359703 s390: wire up separate socketcalls system calls<br> |
| 462 | 359724 getsockname might crash - deref_UInt should call safe_to_deref<br> |
| 463 | 359733 amd64 implement ld.so strchr/index override like x86<br> |
| 464 | 359767 Valgrind does not support the IBM POWER ISA 3.0 instructions, part 1/5<br> |
| 465 | 359829 Power PC test suite none/tests/ppc64/test_isa_2_07.c uses<br> |
| 466 | uninitialized data<br> |
| 467 | 359838 arm64: Unhandled instruction 0xD5033F5F (clrex)<br> |
| 468 | 359871 Incorrect mask handling in ppoll<br> |
| 469 | 359952 Unrecognised PCMPESTRM variants (0x70, 0x19)<br> |
| 470 | 360008 Contents of Power vr registers contents is not printed correctly when<br> |
| 471 | the --vgdb-shadow-registers=yes option is used<br> |
| 472 | 360035 POWER PC instruction bcdadd and bcdsubtract generate result with<br> |
| 473 | non-zero shadow bits<br> |
| 474 | 360378 arm64: Unhandled instruction 0x5E280844 (sha1h s4, s2)<br> |
| 475 | 360425 arm64 unsupported instruction ldpsw<br> |
| 476 | == 364435<br> |
| 477 | 360519 none/tests/arm64/memory.vgtest might fail with newer gcc<br> |
| 478 | 360571 Error about the Android Runtime reading below the stack pointer on ARM<br> |
| 479 | 360574 Wrong parameter type for an ashmem ioctl() call on Android and ARM64<br> |
| 480 | 360749 kludge for multiple .rodata sections on Solaris no longer needed<br> |
| 481 | 360752 raise the number of reserved fds in m_main.c from 10 to 12<br> |
| 482 | 361207 Valgrind does not support the IBM POWER ISA 3.0 instructions, part 2/5<br> |
| 483 | 361226 s390x: risbgn (EC59) not implemented<br> |
| 484 | 361253 [s390x] ex_clone.c:42: undefined reference to `pthread_create'<br> |
| 485 | 361354 ppc64[le]: wire up separate socketcalls system calls<br> |
| 486 | 361615 Inconsistent termination for multithreaded process terminated by signal<br> |
| 487 | 361926 Unhandled Solaris syscall: sysfs(84)<br> |
| 488 | 362009 V dumps core on unimplemented functionality before threads are created<br> |
| 489 | 362329 Valgrind does not support the IBM POWER ISA 3.0 instructions, part 3/5<br> |
| 490 | 362894 missing (broken) support for wbit field on mtfsfi instruction (ppc64)<br> |
| 491 | 362935 [AsusWRT] Assertion 'sizeof(TTEntryC) <= 88' failed<br> |
| 492 | 362953 Request for an update to the Valgrind Developers page<br> |
| 493 | 363680 add renameat2() support<br> |
| 494 | 363705 arm64 missing syscall name_to_handle_at and open_by_handle_at<br> |
| 495 | 363714 ppc64 missing syscalls sync, waitid and name_to/open_by_handle_at<br> |
| 496 | 363858 Valgrind does not support the IBM POWER ISA 3.0 instructions, part 4/5<br> |
| 497 | 364058 clarify in manual limitations of array overruns detections<br> |
| 498 | 364413 pselect sycallwrapper mishandles NULL sigmask<br> |
| 499 | 364728 Power PC, missing support for several HW registers in<br> |
| 500 | get_otrack_shadow_offset_wrk()<br> |
| 501 | 364948 Valgrind does not support the IBM POWER ISA 3.0 instructions, part 5/5<br> |
| 502 | 365273 Invalid write to stack location reported after signal handler runs<br> |
| 503 | 365912 ppc64BE segfault during jm-insns test (RELRO)<br> |
| 504 | 366079 FPXX Support for MIPS32 Valgrind<br> |
| 505 | 366138 Fix configure errors out when using Xcode 8 (clang 8.0.0)<br> |
| 506 | 366344 Multiple unhandled instruction for Aarch64<br> |
| 507 | (0x0EE0E020, 0x1AC15800, 0x4E284801, 0x5E040023, 0x5E056060)<br> |
| 508 | 367995 Integration of memcheck with custom memory allocator<br> |
| 509 | 368120 x86_linux asm _start functions do not keep 16-byte aligned stack pointer<br> |
| 510 | 368412 False positive result for altivec capability check<br> |
| 511 | 368416 Add tc06_two_races_xml.exp output for ppc64<br> |
| 512 | 368419 Perf Events ioctls not implemented<br> |
| 513 | 368461 mmapunmap test fails on ppc64<br> |
| 514 | 368823 run_a_thread_NORETURN assembly code typo for VGP_arm64_linux target<br> |
| 515 | 369000 AMD64 fma4 instructions unsupported.<br> |
| 516 | 369169 ppc64 fails jm_int_isa_2_07 test<br> |
| 517 | 369175 jm_vec_isa_2_07 test crashes on ppc64<br> |
| 518 | 369209 valgrind loops and eats up all memory if cwd doesn't exist.<br> |
| 519 | 369356 pre_mem_read_sockaddr syscall wrapper can crash with bad sockaddr<br> |
| 520 | 369359 msghdr_foreachfield can crash when handling bad iovec<br> |
| 521 | 369360 Bad sigprocmask old or new sets can crash valgrind<br> |
| 522 | 369361 vmsplice syscall wrapper crashes on bad iovec<br> |
| 523 | 369362 Bad sigaction arguments crash valgrind<br> |
| 524 | 369383 x86 sys_modify_ldt wrapper crashes on bad ptr<br> |
| 525 | 369402 Bad set/get_thread_area pointer crashes valgrind<br> |
| 526 | 369441 bad lvec argument crashes process_vm_readv/writev syscall wrappers<br> |
| 527 | 369446 valgrind crashes on unknown fcntl command<br> |
| 528 | 369439 S390x: Unhandled insns RISBLG/RISBHG and LDE/LDER <br> |
| 529 | 369468 Remove quadratic metapool algorithm using VG_(HT_remove_at_Iter)<br> |
| 530 | 370265 ISA 3.0 HW cap stuff needs updating<br> |
| 531 | 371128 BCD add and subtract instructions on Power BE in 32-bit mode do not work<br> |
| Elliott Hughes | ed39800 | 2017-06-21 14:41:24 -0700 | [diff] [blame^] | 532 | 372195 Power PC, xxsel instruction is not always recognized<br> |
| 533 | <br> |
| Elliott Hughes | a0664b9 | 2017-04-18 17:46:52 -0700 | [diff] [blame] | 534 | n-i-bz Fix incorrect (or infinite loop) unwind on RHEL7 x86 and amd64<br> |
| 535 | n-i-bz massif --pages-as-heap=yes does not report peak caused by mmap+munmap<br> |
| 536 | n-i-bz false positive leaks due to aspacemgr merging heap & non heap segments<br> |
| 537 | n-i-bz Fix ppoll_alarm exclusion on OS X<br> |
| 538 | n-i-bz Document brk segment limitation, reference manual in limit reached msg.<br> |
| 539 | n-i-bz Fix clobber list in none/tests/amd64/xacq_xrel.c [valgrind r15737]<br> |
| 540 | n-i-bz Bump allowed shift value for "add.w reg, sp, reg, lsl #N" [vex r3206]<br> |
| 541 | n-i-bz amd64: memcheck false positive with shr %edx<br> |
| 542 | n-i-bz arm3: Allow early writeback of SP base register in "strd rD, [sp, #-16]"<br> |
| 543 | n-i-bz ppc: Fix two cases of PPCAvFpOp vs PPCFpOp enum confusion<br> |
| 544 | n-i-bz arm: Fix incorrect register-number constraint check for LDAEX{,B,H,D}<br> |
| 545 | n-i-bz DHAT: added collection of the metric "tot-blocks-allocd" <br> |
| 546 | <br> |
| 547 | (3.12.0.RC1: 20 October 2016, vex r3282, valgrind r16094)<br> |
| 548 | (3.12.0.RC2: 20 October 2016, vex r3282, valgrind r16096)<br> |
| 549 | (3.12.0: 21 October 2016, vex r3282, valgrind r16098)<br> |
| 550 | <br> |
| 551 | <br> |
| 552 | <br> |
| 553 | Release 3.11.0 (22 September 2015)<br> |
| 554 | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~<br> |
| 555 | <br> |
| 556 | 3.11.0 is a feature release with many improvements and the usual<br> |
| 557 | collection of bug fixes.<br> |
| 558 | <br> |
| 559 | This release supports X86/Linux, AMD64/Linux, ARM32/Linux,<br> |
| 560 | ARM64/Linux, PPC32/Linux, PPC64BE/Linux, PPC64LE/Linux, S390X/Linux,<br> |
| 561 | MIPS32/Linux, MIPS64/Linux, ARM/Android, ARM64/Android,<br> |
| 562 | MIPS32/Android, X86/Android, X86/Solaris, AMD64/Solaris, X86/MacOSX<br> |
| 563 | 10.10 and AMD64/MacOSX 10.10. There is also preliminary support for<br> |
| 564 | X86/MacOSX 10.11, AMD64/MacOSX 10.11 and TILEGX/Linux.<br> |
| 565 | <br> |
| 566 | * ================== PLATFORM CHANGES =================<br> |
| 567 | <br> |
| 568 | * Support for Solaris/x86 and Solaris/amd64 has been added.<br> |
| 569 | <br> |
| 570 | * Preliminary support for Mac OS X 10.11 (El Capitan) has been added.<br> |
| 571 | <br> |
| 572 | * Preliminary support for the Tilera TileGX architecture has been added.<br> |
| 573 | <br> |
| 574 | * s390x: It is now required for the host to have the "long displacement"<br> |
| 575 | facility. The oldest supported machine model is z990.<br> |
| 576 | <br> |
| 577 | * x86: on an SSE2 only host, Valgrind in 32 bit mode now claims to be a<br> |
| 578 | Pentium 4. 3.10.1 wrongly claimed to be a Core 2, which is SSSE3.<br> |
| 579 | <br> |
| 580 | * The JIT's register allocator is significantly faster, making the JIT<br> |
| 581 | as a whole somewhat faster, so JIT-intensive activities, for example<br> |
| 582 | program startup, are modestly faster, around 5%.<br> |
| 583 | <br> |
| 584 | * There have been changes to the default settings of several command<br> |
| 585 | line flags, as detailed below.<br> |
| 586 | <br> |
| 587 | * Intel AVX2 support is more complete (64 bit targets only). On AVX2<br> |
| 588 | capable hosts, the simulated CPUID will now indicate AVX2 support.<br> |
| 589 | <br> |
| 590 | * ==================== TOOL CHANGES ====================<br> |
| 591 | <br> |
| 592 | * Memcheck:<br> |
| 593 | <br> |
| 594 | - The default value for --leak-check-heuristics has been changed from<br> |
| 595 | "none" to "all". This helps to reduce the number of possibly<br> |
| 596 | lost blocks, in particular for C++ applications.<br> |
| 597 | <br> |
| 598 | - The default value for --keep-stacktraces has been changed from<br> |
| 599 | "malloc-then-free" to "malloc-and-free". This has a small cost in<br> |
| 600 | memory (one word per malloc-ed block) but allows Memcheck to show the<br> |
| 601 | 3 stacktraces of a dangling reference: where the block was allocated,<br> |
| 602 | where it was freed, and where it is acccessed after being freed.<br> |
| 603 | <br> |
| 604 | - The default value for --partial-loads-ok has been changed from "no" to <br> |
| 605 | "yes", so as to avoid false positive errors resulting from some kinds<br> |
| 606 | of vectorised loops.<br> |
| 607 | <br> |
| 608 | - A new monitor command 'xb <addr> <len>' shows the validity bits of<br> |
| 609 | <len> bytes at <addr>. The monitor command 'xb' is easier to use<br> |
| 610 | than get_vbits when you need to associate byte data value with<br> |
| 611 | their corresponding validity bits.<br> |
| 612 | <br> |
| 613 | - The 'block_list' monitor command has been enhanced:<br> |
| 614 | o it can print a range of loss records<br> |
| 615 | o it now accepts an optional argument 'limited <max_blocks>'<br> |
| 616 | to control the number of blocks printed.<br> |
| 617 | o if a block has been found using a heuristic, then<br> |
| 618 | 'block_list' now shows the heuristic after the block size.<br> |
| 619 | o the loss records/blocks to print can be limited to the blocks<br> |
| 620 | found via specified heuristics.<br> |
| 621 | <br> |
| 622 | - The C helper functions used to instrument loads on<br> |
| 623 | x86-{linux,solaris} and arm-linux (both 32-bit only) have been<br> |
| 624 | replaced by handwritten assembly sequences. This gives speedups<br> |
| 625 | in the region of 0% to 7% for those targets only.<br> |
| 626 | <br> |
| 627 | - A new command line option, --expensive-definedness-checks=yes|no,<br> |
| 628 | has been added. This is useful for avoiding occasional invalid<br> |
| 629 | uninitialised-value errors in optimised code. Watch out for<br> |
| 630 | runtime degradation, as this can be up to 25%. As always, though,<br> |
| 631 | the slowdown is highly application specific. The default setting<br> |
| 632 | is "no".<br> |
| 633 | <br> |
| 634 | * Massif:<br> |
| 635 | <br> |
| 636 | - A new monitor command 'all_snapshots <filename>' dumps all<br> |
| 637 | snapshots taken so far.<br> |
| 638 | <br> |
| 639 | * Helgrind:<br> |
| 640 | <br> |
| 641 | - Significant memory reduction and moderate speedups for<br> |
| 642 | --history-level=full for applications accessing a lot of memory<br> |
| 643 | with many different stacktraces.<br> |
| 644 | <br> |
| 645 | - The default value for --conflict-cache-size=N has been doubled to<br> |
| 646 | 2000000. Users that were not using the default value should<br> |
| 647 | preferably also double the value they give.<br> |
| 648 | <br> |
| 649 | The default was changed due to the changes in the "full history"<br> |
| 650 | implementation. Doubling the value gives on average a slightly more<br> |
| 651 | complete history and uses similar memory (or significantly less memory<br> |
| 652 | in the worst case) than the previous implementation.<br> |
| 653 | <br> |
| 654 | - The Helgrind monitor command 'info locks' now accepts an optional<br> |
| 655 | argument 'lock_addr', which shows information about the lock at the<br> |
| 656 | given address only.<br> |
| 657 | <br> |
| 658 | - When using --history-level=full, the new Helgrind monitor command<br> |
| 659 | 'accesshistory <addr> [<len>]' will show the recorded accesses for<br> |
| 660 | <len> (or 1) bytes at <addr>.<br> |
| 661 | <br> |
| 662 | * ==================== OTHER CHANGES ====================<br> |
| 663 | <br> |
| 664 | * The default value for the --smc-check option has been changed from<br> |
| 665 | "stack" to "all-non-file" on targets that provide automatic D-I<br> |
| 666 | cache coherence (x86, amd64 and s390x). The result is to provide,<br> |
| 667 | by default, transparent support for JIT generated and self-modifying<br> |
| 668 | code on all targets.<br> |
| 669 | <br> |
| 670 | * Mac OS X only: the default value for the --dsymutil option has been<br> |
| 671 | changed from "no" to "yes", since any serious usage on Mac OS X<br> |
| 672 | always required it to be "yes".<br> |
| 673 | <br> |
| 674 | * The command line options --db-attach and --db-command have been removed.<br> |
| 675 | They were deprecated in 3.10.0.<br> |
| 676 | <br> |
| 677 | * When a process dies due to a signal, Valgrind now shows the signal<br> |
| 678 | and the stacktrace at default verbosity (i.e. verbosity 1).<br> |
| 679 | <br> |
| 680 | * The address description logic used by Memcheck and Helgrind now<br> |
| 681 | describes addresses in anonymous segments, file mmap-ed segments,<br> |
| 682 | shared memory segments and the brk data segment.<br> |
| 683 | <br> |
| 684 | * The new option --error-markers=<begin>,<end> can be used to mark the<br> |
| 685 | begin/end of errors in textual output mode, to facilitate<br> |
| 686 | searching/extracting errors in output files that mix valgrind errors<br> |
| 687 | with program output.<br> |
| 688 | <br> |
| 689 | * The new option --max-threads=<number> can be used to change the number<br> |
| 690 | of threads valgrind can handle. The default is 500 threads which<br> |
| 691 | should be more than enough for most applications.<br> |
| 692 | <br> |
| 693 | * The new option --valgrind-stacksize=<number> can be used to change the<br> |
| 694 | size of the private thread stacks used by Valgrind. This is useful<br> |
| 695 | for reducing memory use or increasing the stack size if Valgrind<br> |
| 696 | segfaults due to stack overflow.<br> |
| 697 | <br> |
| 698 | * The new option --avg-transtab-entry-size=<number> can be used to specify<br> |
| 699 | the expected instrumented block size, either to reduce memory use or<br> |
| 700 | to avoid excessive retranslation.<br> |
| 701 | <br> |
| 702 | * Valgrind can be built with Intel's ICC compiler, version 14.0 or later.<br> |
| 703 | <br> |
| 704 | * New and modified GDB server monitor features:<br> |
| 705 | <br> |
| 706 | - When a signal is reported in GDB, you can now use the GDB convenience<br> |
| 707 | variable $_siginfo to examine detailed signal information.<br> |
| 708 | <br> |
| 709 | - Valgrind's gdbserver now allows the user to change the signal<br> |
| 710 | to deliver to the process. So, use 'signal SIGNAL' to continue execution<br> |
| 711 | with SIGNAL instead of the signal reported to GDB. Use 'signal 0' to<br> |
| 712 | continue without passing the signal to the process.<br> |
| 713 | <br> |
| 714 | - With GDB >= 7.10, the command 'target remote'<br> |
| 715 | will automatically load the executable file of the process running<br> |
| 716 | under Valgrind. This means you do not need to specify the executable<br> |
| 717 | file yourself, GDB will discover it itself. See GDB documentation about<br> |
| 718 | 'qXfer:exec-file:read' packet for more info.<br> |
| 719 | <br> |
| 720 | * ==================== FIXED BUGS ====================<br> |
| 721 | <br> |
| 722 | The following bugs have been fixed or resolved. Note that "n-i-bz"<br> |
| 723 | stands for "not in bugzilla" -- that is, a bug that was reported to us<br> |
| 724 | but never got a bugzilla entry. We encourage you to file bugs in<br> |
| 725 | bugzilla (https://bugs.kde.org/enter_bug.cgi?product=valgrind) rather<br> |
| 726 | than mailing the developers (or mailing lists) directly -- bugs that<br> |
| 727 | are not entered into bugzilla tend to get forgotten about or ignored.<br> |
| 728 | <br> |
| 729 | To see details of a given bug, visit<br> |
| 730 | https://bugs.kde.org/show_bug.cgi?id=XXXXXX<br> |
| 731 | where XXXXXX is the bug number as listed below.<br> |
| 732 | <br> |
| 733 | 116002 VG_(printf): Problems with justification of strings and integers<br> |
| 734 | 155125 avoid cutting away file:lineno after long function name<br> |
| 735 | 197259 Unsupported arch_prtctl PR_SET_GS option<br> |
| 736 | 201152 ppc64: Assertion in ppc32g_dirtyhelper_MFSPR_268_269<br> |
| 737 | 201216 Fix Valgrind does not support pthread_sigmask() on OS X<br> |
| 738 | 201435 Fix Darwin: -v does not show kernel version<br> |
| 739 | 208217 "Warning: noted but unhandled ioctl 0x2000747b" on Mac OS X<br> |
| 740 | 211256 Fixed an outdated comment regarding the default platform.<br> |
| 741 | 211529 Incomplete call stacks for code compiled by newer versions of MSVC<br> |
| 742 | 211926 Avoid compilation warnings in valgrind.h with -pedantic<br> |
| 743 | 212291 Fix unhandled syscall: unix:132 (mkfifo) on OS X<br> |
| 744 | == 263119<br> |
| 745 | 226609 Crediting upstream authors in man page<br> |
| 746 | 231257 Valgrind omits path when executing script from shebang line<br> |
| 747 | 254164 OS X task_info: UNKNOWN task message [id 3405, to mach_task_self() [..]<br> |
| 748 | 294065 Improve the pdb file reader by avoiding hardwired absolute pathnames<br> |
| 749 | 269360 s390x: Fix addressing mode selection for compare-and-swap<br> |
| 750 | 302630 Memcheck: Assertion failed: 'sizeof(UWord) == sizeof(UInt)'<br> |
| 751 | == 326797<br> |
| 752 | 312989 ioctl handling needs to do POST handling on generic ioctls and [..]<br> |
| 753 | 319274 Fix unhandled syscall: unix:410 (sigsuspend_nocancel) on OS X<br> |
| 754 | 324181 mmap does not handle MAP_32BIT (handle it now, rather than fail it)<br> |
| 755 | 327745 Fix valgrind 3.9.0 build fails on Mac OS X 10.6.8<br> |
| 756 | 330147 libmpiwrap PMPI_Get_count returns undefined value<br> |
| 757 | 333051 mmap of huge pages fails due to incorrect alignment<br> |
| 758 | == 339163<br> |
| 759 | 334802 valgrind does not always explain why a given option is bad<br> |
| 760 | 335618 mov.w rN, pc/sp (ARM32)<br> |
| 761 | 335785 amd64->IR 0xC4 0xE2 0x75 0x2F (vmaskmovpd)<br> |
| 762 | == 307399<br> |
| 763 | == 343175<br> |
| 764 | == 342740<br> |
| 765 | == 346912<br> |
| 766 | 335907 segfault when running wine's ddrawex/tests/surface.c under valgrind<br> |
| 767 | 338602 AVX2 bit in CPUID missing<br> |
| 768 | 338606 Strange message for scripts with invalid interpreter<br> |
| 769 | 338731 ppc: Fix testuite build for toolchains not supporting -maltivec<br> |
| 770 | 338995 shmat with hugepages (SHM_HUGETLB) fails with EINVAL<br> |
| 771 | 339045 Getting valgrind to compile and run on OS X Yosemite (10.10)<br> |
| 772 | == 340252<br> |
| 773 | 339156 gdbsrv not called for fatal signal<br> |
| 774 | 339215 Valgrind 3.10.0 contain 2013 in copyrights notice<br> |
| 775 | 339288 support Cavium Octeon MIPS specific BBIT*32 instructions<br> |
| 776 | 339636 Use fxsave64 and fxrstor64 mnemonics instead of old-school rex64 prefix<br> |
| 777 | 339442 Fix testsuite build failure on OS X 10.9<br> |
| 778 | 339542 Enable compilation with Intel's ICC compiler<br> |
| 779 | 339563 The DVB demux DMX_STOP ioctl doesn't have a wrapper<br> |
| 780 | 339688 Mac-specific ASM does not support .version directive (cpuid,<br> |
| 781 | tronical and pushfpopf tests)<br> |
| 782 | 339745 Valgrind crash when check Marmalade app (partial fix)<br> |
| 783 | 339755 Fix known deliberate memory leak in setenv() on Mac OS X 10.9<br> |
| 784 | 339778 Linux/TileGx platform support for Valgrind<br> |
| 785 | 339780 Fix known uninitialised read in pthread_rwlock_init() on Mac OS X 10.9 <br> |
| 786 | 339789 Fix none/tests/execve test on Mac OS X 10.9<br> |
| 787 | 339808 Fix none/tests/rlimit64_nofile test on Mac OS X 10.9<br> |
| 788 | 339820 vex amd64->IR: 0x66 0xF 0x3A 0x63 0xA 0x42 0x74 0x9 (pcmpistri $0x42)<br> |
| 789 | 340115 Fix none/tests/cmdline[1|2] tests on systems which define TMPDIR<br> |
| 790 | 340392 Allow user to select more accurate definedness checking in memcheck<br> |
| 791 | to avoid invalid complaints on optimised code<br> |
| 792 | 340430 Fix some grammatical weirdness in the manual.<br> |
| 793 | 341238 Recognize GCC5/DWARFv5 DW_LANG constants (Go, C11, C++11, C++14)<br> |
| 794 | 341419 Signal handler ucontext_t not filled out correctly on OS X<br> |
| 795 | 341539 VG_(describe_addr) should not describe address as belonging to client<br> |
| 796 | segment if it is past the heap end<br> |
| 797 | 341613 Enable building of manythreads and thread-exits tests on Mac OS X<br> |
| 798 | 341615 Fix none/tests/darwin/access_extended test on Mac OS X<br> |
| 799 | 341698 Valgrind's AESKEYGENASSIST gives wrong result in words 0 and 2 [..]<br> |
| 800 | 341789 aarch64: shmat fails with valgrind on ARMv8<br> |
| 801 | 341997 MIPS64: Cavium OCTEON insns - immediate operand handled incorrectly<br> |
| 802 | 342008 valgrind.h needs type cast [..] for clang/llvm in 64-bit mode<br> |
| 803 | 342038 Unhandled syscalls on aarch64 (mbind/get/set_mempolicy)<br> |
| 804 | 342063 wrong format specifier for test mcblocklistsearch in gdbserver_tests<br> |
| 805 | 342117 Hang when loading PDB file for MSVC compiled Firefox under Wine<br> |
| 806 | 342221 socket connect false positive uninit memory for unknown af family<br> |
| 807 | 342353 Allow dumping full massif output while valgrind is still running<br> |
| 808 | 342571 Valgrind chokes on AVX compare intrinsic with _CMP_GE_QS<br> |
| 809 | == 346476<br> |
| 810 | == 348387<br> |
| 811 | == 350593<br> |
| 812 | 342603 Add I2C_SMBUS ioctl support<br> |
| 813 | 342635 OS X 10.10 (Yosemite) - missing system calls and fcntl code<br> |
| 814 | 342683 Mark memory past the initial brk limit as unaddressable<br> |
| 815 | 342783 arm: unhandled instruction 0xEEFE1ACA = "vcvt.s32.f32 s3, s3, #12"<br> |
| 816 | 342795 Internal glibc __GI_mempcpy call should be intercepted<br> |
| 817 | 342841 s390x: Support instructions fiebr(a) and fidbr(a)<br> |
| 818 | 343012 Unhandled syscall 319 (memfd_create)<br> |
| 819 | 343069 Patch updating v4l2 API support<br> |
| 820 | 343173 helgrind crash during stack unwind<br> |
| 821 | 343219 fix GET_STARTREGS for arm<br> |
| 822 | 343303 Fix known deliberate memory leak in setenv() on Mac OS X 10.10<br> |
| 823 | 343306 OS X 10.10: UNKNOWN mach_msg unhandled MACH_SEND_TRAILER option<br> |
| 824 | 343332 Unhandled instruction 0x9E310021 (fcvtmu) on aarch64<br> |
| 825 | 343335 unhandled instruction 0x1E638400 (fccmp) aarch64<br> |
| 826 | 343523 OS X mach_ports_register: UNKNOWN task message [id 3403, to [..]<br> |
| 827 | 343525 OS X host_get_special_port: UNKNOWN host message [id 412, to [..]<br> |
| 828 | 343597 ppc64le: incorrect use of offseof macro<br> |
| 829 | 343649 OS X host_create_mach_voucher: UNKNOWN host message [id 222, to [..]<br> |
| 830 | 343663 OS X 10.10 Memchecj always reports a leak regardless of [..]<br> |
| 831 | 343732 Unhandled syscall 144 (setgid) on aarch64<br> |
| 832 | 343733 Unhandled syscall 187 (msgctl and related) on aarch64<br> |
| 833 | 343802 s390x: False positive "conditional jump or move depends on [..]<br> |
| 834 | 343902 --vgdb=yes doesn't break when --xml=yes is used<br> |
| 835 | 343967 Don't warn about setuid/setgid/setcap executable for directories<br> |
| 836 | 343978 Recognize DWARF5/GCC5 DW_LANG_Fortran 2003 and 2008 constants<br> |
| 837 | 344007 accept4 syscall unhandled on arm64 (242) and ppc64 (344)<br> |
| 838 | 344033 Helgrind on ARM32 loses track of mutex state in pthread_cond_wait<br> |
| 839 | 344054 www - update info for Solaris/illumos<br> |
| 840 | 344416 'make regtest' does not work cleanly on OS X<br> |
| 841 | 344235 Remove duplicate include of pub_core_aspacemgr.h<br> |
| 842 | 344279 syscall sendmmsg on arm64 (269) and ppc32/64 (349) unhandled<br> |
| 843 | 344295 syscall recvmmsg on arm64 (243) and ppc32/64 (343) unhandled<br> |
| 844 | 344307 2 unhandled syscalls on aarch64/arm64: umount2(39), mount (40)<br> |
| 845 | 344314 callgrind_annotate ... warnings about commands containing newlines<br> |
| 846 | 344318 socketcall should wrap recvmmsg and sendmmsg<br> |
| 847 | 344337 Fix unhandled syscall: mach:41 (_kernelrpc_mach_port_guard_trap)<br> |
| 848 | 344416 Fix 'make regtest' does not work cleanly on OS X<br> |
| 849 | 344499 Fix compilation for Linux kernel >= 4.0.0<br> |
| 850 | 344512 OS X: unhandled syscall: unix:348 (__pthread_chdir), <br> |
| 851 | unix:349 (__pthread_fchdir)<br> |
| 852 | 344559 Garbage collection of unused segment names in address space manager<br> |
| 853 | 344560 Fix stack traces missing penultimate frame on OS X<br> |
| 854 | 344621 Fix memcheck/tests/err_disable4 test on OS X<br> |
| 855 | 344686 Fix suppression for pthread_rwlock_init on OS X 10.10<br> |
| 856 | 344702 Fix missing libobjc suppressions on OS X 10.10<br> |
| 857 | == 344543<br> |
| 858 | 344936 Fix unhandled syscall: unix:473 (readlinkat) on OS X 10.10<br> |
| 859 | 344939 Fix memcheck/tests/xml1 on OS X 10.10<br> |
| 860 | 345016 helgrind/tests/locked_vs_unlocked2 is failing sometimes<br> |
| 861 | 345079 Fix build problems in VEX/useful/test_main.c<br> |
| 862 | 345126 Incorrect handling of VIDIOC_G_AUDIO and G_AUDOUT<br> |
| 863 | 345177 arm64: prfm (reg) not implemented<br> |
| 864 | 345215 Performance improvements for the register allocator<br> |
| 865 | 345248 add support for Solaris OS in valgrind<br> |
| 866 | 345338 TIOCGSERIAL and TIOCSSERIAL ioctl support on Linux<br> |
| 867 | 345394 Fix memcheck/tests/strchr on OS X<br> |
| 868 | 345637 Fix memcheck/tests/sendmsg on OS X<br> |
| 869 | 345695 Add POWERPC support for AT_DCACHESIZE and HWCAP2<br> |
| 870 | 345824 Fix aspacem segment mismatch: seen with none/tests/bigcode<br> |
| 871 | 345887 Fix an assertion in the address space manager<br> |
| 872 | 345928 amd64: callstack only contains current function for small stacks<br> |
| 873 | 345984 disInstr(arm): unhandled instruction: 0xEE193F1E<br> |
| 874 | 345987 MIPS64: Implement cavium LHX instruction<br> |
| 875 | 346031 MIPS: Implement support for the CvmCount register (rhwr %0, 31)<br> |
| 876 | 346185 Fix typo saving altivec register v24<br> |
| 877 | 346267 Compiler warnings for PPC64 code on call to LibVEX_GuestPPC64_get_XER()<br> |
| 878 | and LibVEX_GuestPPC64_get_CR()<br> |
| 879 | 346270 Regression tests none/tests/jm_vec/isa_2_07 and<br> |
| 880 | none/tests/test_isa_2_07_part2 have failures on PPC64 little endian<br> |
| 881 | 346307 fuse filesystem syscall deadlocks<br> |
| 882 | 346324 PPC64 missing support for lbarx, lharx, stbcx and sthcx instructions<br> |
| 883 | 346411 MIPS: SysRes::_valEx handling is incorrect<br> |
| 884 | 346416 Add support for LL_IOC_PATH2FID and LL_IOC_GETPARENT Lustre ioctls<br> |
| 885 | 346474 PPC64 Power 8, spr TEXASRU register not supported<br> |
| 886 | 346487 Compiler generates "note" about a future ABI change for PPC64<br> |
| 887 | 346562 MIPS64: lwl/lwr instructions are performing 64bit loads<br> |
| 888 | and causing spurious "invalid read of size 8" warnings<br> |
| 889 | 346801 Fix link error on OS X: _vgModuleLocal_sf_maybe_extend_stack<br> |
| 890 | 347151 Fix suppression for pthread_rwlock_init on OS X 10.8<br> |
| 891 | 347233 Fix memcheck/tests/strchr on OS X 10.10 (Haswell) <br> |
| 892 | 347322 Power PC regression test cleanup<br> |
| 893 | 347379 valgrind --leak-check=full leak errors from system libs on OS X 10.8<br> |
| 894 | == 217236<br> |
| 895 | 347389 unhandled syscall: 373 (Linux ARM syncfs)<br> |
| 896 | 347686 Patch set to cleanup PPC64 regtests<br> |
| 897 | 347978 Remove bash dependencies where not needed<br> |
| 898 | 347982 OS X: undefined symbols for architecture x86_64: "_global" [..]<br> |
| 899 | 347988 Memcheck: the 'impossible' happened: unexpected size for Addr (OSX/wine)<br> |
| 900 | == 345929<br> |
| 901 | 348102 Patch updating v4l2 API support<br> |
| 902 | 348247 amd64 front end: jno jumps wrongly when overflow is not set<br> |
| 903 | 348269 Improve mmap MAP_HUGETLB support.<br> |
| 904 | 348334 (ppc) valgrind does not simulate dcbfl - then my program terminates<br> |
| 905 | 348345 Assertion fails for negative lineno<br> |
| 906 | 348377 Unsupported ARM instruction: yield<br> |
| 907 | 348565 Fix detection of command line option availability for clang<br> |
| 908 | 348574 vex amd64->IR pcmpistri SSE4.2 unsupported (pcmpistri $0x18)<br> |
| 909 | 348728 Fix broken check for VIDIOC_G_ENC_INDEX<br> |
| 910 | 348748 Fix redundant condition<br> |
| 911 | 348890 Fix clang warning about unsupported --param inline-unit-growth=900<br> |
| 912 | 348949 Bogus "ERROR: --ignore-ranges: suspiciously large range"<br> |
| 913 | 349034 Add Lustre ioctls LL_IOC_GROUP_LOCK and LL_IOC_GROUP_UNLOCK<br> |
| 914 | 349086 Fix UNKNOWN task message [id 3406, to mach_task_self(), [..]<br> |
| 915 | 349087 Fix UNKNOWN task message [id 3410, to mach_task_self(), [..]<br> |
| 916 | 349626 Implemented additional Xen hypercalls<br> |
| 917 | 349769 Clang/osx: ld: warning: -read_only_relocs cannot be used with x86_64<br> |
| 918 | 349790 Clean up of the hardware capability checking utilities.<br> |
| 919 | 349828 memcpy intercepts memmove causing src/dst overlap error (ppc64 ld.so)<br> |
| 920 | 349874 Fix typos in source code<br> |
| 921 | 349879 memcheck: add handwritten assembly for helperc_LOADV*<br> |
| 922 | 349941 di_notify_mmap might create wrong start/size DebugInfoMapping<br> |
| 923 | 350062 vex x86->IR: 0x66 0xF 0x3A 0xB (ROUNDSD) on OS X<br> |
| 924 | 350202 Add limited param to 'monitor block_list'<br> |
| 925 | 350290 s390x: Support instructions fixbr(a)<br> |
| 926 | 350359 memcheck/tests/x86/fxsave hangs indefinetely on OS X<br> |
| 927 | 350809 Fix none/tests/async-sigs for Solaris<br> |
| 928 | 350811 Remove reference to --db-attach which has been removed.<br> |
| 929 | 350813 Memcheck/x86: enable handwritten assembly helpers for x86/Solaris too<br> |
| 930 | 350854 hard-to-understand code in VG_(load_ELF)()<br> |
| 931 | 351140 arm64 syscalls setuid (146) and setresgid (149) not implemented<br> |
| 932 | 351386 Solaris: Cannot run ld.so.1 under Valgrind<br> |
| 933 | 351474 Fix VG_(iseqsigset) as obvious<br> |
| 934 | 351531 Typo in /include/vki/vki-xen-physdev.h header guard<br> |
| 935 | 351756 Intercept platform_memchr$VARIANT$Haswell on OS X<br> |
| 936 | 351858 ldsoexec support on Solaris<br> |
| 937 | 351873 Newer gcc doesn't allow __builtin_tabortdc[i] in ppc32 mode<br> |
| 938 | 352130 helgrind reports false races for printfs using mempcpy on FILE* state<br> |
| 939 | 352284 s390: Conditional jump depends on uninitialised value(s) in vfprintf <br> |
| 940 | 352320 arm64 crash on none/tests/nestedfs<br> |
| 941 | 352765 Vbit test fails on Power 6<br> |
| 942 | 352768 The mbar instruction is missing from the Power PC support<br> |
| 943 | 352769 Power PC program priority register (PPR) is not supported<br> |
| 944 | n-i-bz Provide implementations of certain compiler builtins to support<br> |
| 945 | compilers that may not provide those<br> |
| 946 | n-i-bz Old STABS code is still being compiled, but never used. Remove it.<br> |
| 947 | n-i-bz Fix compilation on distros with glibc < 2.5<br> |
| 948 | n-i-bz (vex 3098) Avoid generation of Neon insns on non-Neon hosts<br> |
| 949 | n-i-bz Enable rt_sigpending syscall on ppc64 linux.<br> |
| 950 | n-i-bz mremap did not work properly on shared memory<br> |
| 951 | n-i-bz Fix incorrect sizeof expression in syswrap-xen.c reported by Coverity<br> |
| 952 | n-i-bz In VALGRIND_PRINTF write out thread name, if any, to xml<br> |
| 953 | <br> |
| 954 | (3.11.0.TEST1: 8 September 2015, vex r3187, valgrind r15646)<br> |
| 955 | (3.11.0.TEST2: 21 September 2015, vex r3193, valgrind r15667)<br> |
| 956 | (3.11.0: 22 September 2015, vex r3195, valgrind r15674)<br> |
| 957 | <br> |
| 958 | <br> |
| 959 | <br> |
| 960 | Release 3.10.1 (25 November 2014)<br> |
| 961 | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~<br> |
| 962 | 3.10.1 is a bug fix release. It fixes various bugs reported in 3.10.0<br> |
| 963 | and backports fixes for all reported missing AArch64 ARMv8 instructions<br> |
| 964 | and syscalls from the trunk. If you package or deliver 3.10.0 for others<br> |
| 965 | to use, you might want to consider upgrading to 3.10.1 instead.<br> |
| 966 | <br> |
| 967 | The following bugs have been fixed or resolved. Note that "n-i-bz"<br> |
| 968 | stands for "not in bugzilla" -- that is, a bug that was reported to us<br> |
| 969 | but never got a bugzilla entry. We encourage you to file bugs in<br> |
| 970 | bugzilla (https://bugs.kde.org/enter_bug.cgi?product=valgrind) rather<br> |
| 971 | than mailing the developers (or mailing lists) directly -- bugs that<br> |
| 972 | are not entered into bugzilla tend to get forgotten about or ignored.<br> |
| 973 | <br> |
| 974 | To see details of a given bug, visit<br> |
| 975 | https://bugs.kde.org/show_bug.cgi?id=XXXXXX<br> |
| 976 | where XXXXXX is the bug number as listed below.<br> |
| 977 | <br> |
| 978 | 335440 arm64: ld1 (single structure) is not implemented<br> |
| 979 | 335713 arm64: unhanded instruction: prfm (immediate)<br> |
| 980 | 339020 ppc64: memcheck/tests/ppc64/power_ISA2_05 failing in nightly build<br> |
| 981 | 339182 ppc64: AvSplat ought to load destination vector register with [..]<br> |
| 982 | 339336 PPC64 store quad instruction (stq) is not supposed to change [..]<br> |
| 983 | 339433 ppc64 lxvw4x instruction uses four 32-byte loads<br> |
| 984 | 339645 Use correct tag names in sys_getdents/64 wrappers<br> |
| 985 | 339706 Fix false positive for ioctl(TIOCSIG) on linux<br> |
| 986 | 339721 assertion 'check_sibling == sibling' failed in readdwarf3.c ...<br> |
| 987 | 339853 arm64 times syscall unknown<br> |
| 988 | 339855 arm64 unhandled getsid/setsid syscalls<br> |
| 989 | 339858 arm64 dmb sy not implemented<br> |
| 990 | 339926 Unhandled instruction 0x1E674001 (frintx) on aarm64<br> |
| 991 | 339927 Unhandled instruction 0x9E7100C6 (fcvtmu) on aarch64<br> |
| 992 | 339938 disInstr(arm64): unhandled instruction 0x4F8010A4 (fmla)<br> |
| 993 | == 339950<br> |
| 994 | 339940 arm64: unhandled syscall: 83 (sys_fdatasync) + patch<br> |
| 995 | 340033 arm64: unhandled insn dmb ishld and some other isb-dmb-dsb variants<br> |
| 996 | 340028 unhandled syscalls for arm64 (msync, pread64, setreuid and setregid)<br> |
| 997 | 340036 arm64: Unhandled instruction ld4 (multiple structures, no offset)<br> |
| 998 | 340236 arm64: unhandled syscalls: mknodat, fchdir, chroot, fchownat<br> |
| 999 | 340509 arm64: unhandled instruction fcvtas<br> |
| 1000 | 340630 arm64: fchmod (52) and fchown (55) syscalls not recognized<br> |
| 1001 | 340632 arm64: unhandled instruction fcvtas<br> |
| 1002 | 340722 Resolve "UNKNOWN attrlist flags 0:0x10000000"<br> |
| 1003 | 340725 AVX2: Incorrect decoding of vpbroadcast{b,w} reg,reg forms<br> |
| 1004 | 340788 warning: unhandled syscall: 318 (getrandom)<br> |
| 1005 | 340807 disInstr(arm): unhandled instruction: 0xEE989B20<br> |
| 1006 | 340856 disInstr(arm64): unhandled instruction 0x1E634C45 (fcsel)<br> |
| 1007 | 340922 arm64: unhandled getgroups/setgroups syscalls<br> |
| 1008 | 350251 Fix typo in VEX utility program (test_main.c).<br> |
| 1009 | 350407 arm64: unhandled instruction ucvtf (vector, integer)<br> |
| 1010 | 350809 none/tests/async-sigs breaks when run under cron on Solaris<br> |
| 1011 | 350811 update README.solaris after r15445<br> |
| 1012 | 350813 Use handwritten memcheck assembly helpers on x86/Solaris [..]<br> |
| 1013 | 350854 strange code in VG_(load_ELF)()<br> |
| 1014 | 351140 arm64 syscalls setuid (146) and setresgid (149) not implemented<br> |
| 1015 | n-i-bz DRD and Helgrind: Handle Imbe_CancelReservation (clrex on ARM)<br> |
| 1016 | n-i-bz Add missing ]] to terminate CDATA.<br> |
| 1017 | n-i-bz Glibc versions prior to 2.5 do not define PTRACE_GETSIGINFO<br> |
| 1018 | n-i-bz Enable sys_fadvise64_64 on arm32.<br> |
| 1019 | n-i-bz Add test cases for all remaining AArch64 SIMD, FP and memory insns.<br> |
| 1020 | n-i-bz Add test cases for all known arm64 load/store instructions.<br> |
| 1021 | n-i-bz PRE(sys_openat): when checking whether ARG1 == VKI_AT_FDCWD [..]<br> |
| 1022 | n-i-bz Add detection of old ppc32 magic instructions from bug 278808.<br> |
| 1023 | n-i-bz exp-dhat: Implement missing function "dh_malloc_usable_size".<br> |
| 1024 | n-i-bz arm64: Implement "fcvtpu w, s".<br> |
| 1025 | n-i-bz arm64: implement ADDP and various others<br> |
| 1026 | n-i-bz arm64: Implement {S,U}CVTF (scalar, fixedpt).<br> |
| 1027 | n-i-bz arm64: enable FCVT{A,N}S X,S.<br> |
| 1028 | <br> |
| 1029 | (3.10.1: 25 November 2014, vex r3026, valgrind r14785)<br> |
| 1030 | <br> |
| 1031 | <br> |
| 1032 | <br> |
| 1033 | Release 3.10.0 (10 September 2014)<br> |
| 1034 | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~<br> |
| 1035 | <br> |
| 1036 | 3.10.0 is a feature release with many improvements and the usual<br> |
| 1037 | collection of bug fixes.<br> |
| 1038 | <br> |
| 1039 | This release supports X86/Linux, AMD64/Linux, ARM32/Linux, ARM64/Linux,<br> |
| 1040 | PPC32/Linux, PPC64BE/Linux, PPC64LE/Linux, S390X/Linux, MIPS32/Linux,<br> |
| 1041 | MIPS64/Linux, ARM/Android, MIPS32/Android, X86/Android, X86/MacOSX 10.9<br> |
| 1042 | and AMD64/MacOSX 10.9. Support for MacOSX 10.8 and 10.9 is<br> |
| 1043 | significantly improved relative to the 3.9.0 release.<br> |
| 1044 | <br> |
| 1045 | * ================== PLATFORM CHANGES =================<br> |
| 1046 | <br> |
| 1047 | * Support for the 64-bit ARM Architecture (AArch64 ARMv8). This port<br> |
| 1048 | is mostly complete, and is usable, but some SIMD instructions are as<br> |
| 1049 | yet unsupported.<br> |
| 1050 | <br> |
| 1051 | * Support for little-endian variant of the 64-bit POWER architecture.<br> |
| 1052 | <br> |
| 1053 | * Support for Android on MIPS32.<br> |
| 1054 | <br> |
| 1055 | * Support for 64bit FPU on MIPS32 platforms.<br> |
| 1056 | <br> |
| 1057 | * Both 32- and 64-bit executables are supported on MacOSX 10.8 and 10.9.<br> |
| 1058 | <br> |
| 1059 | * Configuration for and running on Android targets has changed.<br> |
| 1060 | See README.android in the source tree for details.<br> |
| 1061 | <br> |
| 1062 | * ================== DEPRECATED FEATURES =================<br> |
| 1063 | <br> |
| 1064 | * --db-attach is now deprecated and will be removed in the next<br> |
| 1065 | valgrind feature release. The built-in GDB server capabilities are<br> |
| 1066 | superior and should be used instead. Learn more here:<br> |
| 1067 | http://valgrind.org/docs/manual/manual-core-adv.html#manual-core-adv.gdbserver<br> |
| 1068 | <br> |
| 1069 | * ==================== TOOL CHANGES ====================<br> |
| 1070 | <br> |
| 1071 | * Memcheck:<br> |
| 1072 | <br> |
| 1073 | - Client code can now selectively disable and re-enable reporting of<br> |
| 1074 | invalid address errors in specific ranges using the new client<br> |
| 1075 | requests VALGRIND_DISABLE_ADDR_ERROR_REPORTING_IN_RANGE and<br> |
| 1076 | VALGRIND_ENABLE_ADDR_ERROR_REPORTING_IN_RANGE.<br> |
| 1077 | <br> |
| 1078 | - Leak checker: there is a new leak check heuristic called<br> |
| 1079 | "length64". This is used to detect interior pointers pointing 8<br> |
| 1080 | bytes inside a block, on the assumption that the first 8 bytes<br> |
| 1081 | holds the value "block size - 8". This is used by<br> |
| 1082 | sqlite3MemMalloc, for example.<br> |
| 1083 | <br> |
| 1084 | - Checking of system call parameters: if a syscall parameter<br> |
| 1085 | (e.g. bind struct sockaddr, sendmsg struct msghdr, ...) has<br> |
| 1086 | several fields not initialised, an error is now reported for each<br> |
| 1087 | field. Previously, an error was reported only for the first<br> |
| 1088 | uninitialised field.<br> |
| 1089 | <br> |
| 1090 | - Mismatched alloc/free checking: a new flag<br> |
| 1091 | --show-mismatched-frees=no|yes [yes] makes it possible to turn off<br> |
| 1092 | such checks if necessary.<br> |
| 1093 | <br> |
| 1094 | * Helgrind:<br> |
| 1095 | <br> |
| 1096 | - Improvements to error messages:<br> |
| 1097 | <br> |
| 1098 | o Race condition error message involving heap allocated blocks also<br> |
| 1099 | show the thread number that allocated the raced-on block.<br> |
| 1100 | <br> |
| 1101 | o All locks referenced by an error message are now announced.<br> |
| 1102 | Previously, some error messages only showed the lock addresses.<br> |
| 1103 | <br> |
| 1104 | o The message indicating where a lock was first observed now also<br> |
| 1105 | describes the address/location of the lock.<br> |
| 1106 | <br> |
| 1107 | - Helgrind now understands the Ada task termination rules and<br> |
| 1108 | creates a happens-before relationship between a terminated task<br> |
| 1109 | and its master. This avoids some false positives and avoids a big<br> |
| 1110 | memory leak when a lot of Ada tasks are created and terminated.<br> |
| 1111 | The interceptions are only activated with forthcoming releases of<br> |
| 1112 | gnatpro >= 7.3.0w-20140611 and gcc >= 5.0.<br> |
| 1113 | <br> |
| 1114 | - A new GDB server monitor command "info locks" giving the list of<br> |
| 1115 | locks, their location, and their status.<br> |
| 1116 | <br> |
| 1117 | * Callgrind:<br> |
| 1118 | <br> |
| 1119 | - callgrind_control now supports the --vgdb-prefix argument,<br> |
| 1120 | which is needed if valgrind was started with this same argument.<br> |
| 1121 | <br> |
| 1122 | * ==================== OTHER CHANGES ====================<br> |
| 1123 | <br> |
| 1124 | * Unwinding through inlined function calls. Stack unwinding can now<br> |
| 1125 | make use of Dwarf3 inlined-unwind information if it is available.<br> |
| 1126 | The practical effect is that inlined calls become visible in stack<br> |
| 1127 | traces. The suppression matching machinery has been adjusted<br> |
| 1128 | accordingly. This is controlled by the new option<br> |
| 1129 | --read-inline-info=yes|no. Currently this is enabled by default<br> |
| 1130 | only on Linux and Android targets and only for the tools Memcheck,<br> |
| 1131 | Helgrind and DRD.<br> |
| 1132 | <br> |
| 1133 | * Valgrind can now read EXIDX unwind information on 32-bit ARM<br> |
| 1134 | targets. If an object contains both CFI and EXIDX unwind<br> |
| 1135 | information, Valgrind will prefer the CFI over the EXIDX. This<br> |
| 1136 | facilitates unwinding through system libraries on arm-android<br> |
| 1137 | targets.<br> |
| 1138 | <br> |
| 1139 | * Address description logic has been improved and is now common<br> |
| 1140 | between Memcheck and Helgrind, resulting in better address<br> |
| 1141 | descriptions for some kinds of error messages.<br> |
| 1142 | <br> |
| 1143 | * Error messages about dubious arguments (eg, to malloc or calloc) are<br> |
| 1144 | output like other errors. This means that they can be suppressed<br> |
| 1145 | and they have a stack trace.<br> |
| 1146 | <br> |
| 1147 | * The C++ demangler has been updated for better C++11 support.<br> |
| 1148 | <br> |
| 1149 | * New and modified GDB server monitor features:<br> |
| 1150 | <br> |
| 1151 | - Thread local variables/storage (__thread) can now be displayed.<br> |
| 1152 | <br> |
| 1153 | - The GDB server monitor command "v.info location <address>"<br> |
| 1154 | displays information about an address. The information produced<br> |
| 1155 | depends on the tool and on the options given to valgrind.<br> |
| 1156 | Possibly, the following are described: global variables, local<br> |
| 1157 | (stack) variables, allocated or freed blocks, ...<br> |
| 1158 | <br> |
| 1159 | - The option "--vgdb-stop-at=event1,event2,..." allows the user to<br> |
| 1160 | ask the GDB server to stop at the start of program execution, at<br> |
| 1161 | the end of the program execution and on Valgrind internal errors.<br> |
| 1162 | <br> |
| 1163 | - A new monitor command "v.info stats" shows various Valgrind core<br> |
| 1164 | and tool statistics.<br> |
| 1165 | <br> |
| 1166 | - A new monitor command "v.set hostvisibility" allows the GDB server<br> |
| 1167 | to provide access to Valgrind internal host status/memory.<br> |
| 1168 | <br> |
| 1169 | * A new option "--aspace-minaddr=<address>" can in some situations<br> |
| 1170 | allow the use of more memory by decreasing the address above which<br> |
| 1171 | Valgrind maps memory. It can also be used to solve address<br> |
| 1172 | conflicts with system libraries by increasing the default value.<br> |
| 1173 | See user manual for details.<br> |
| 1174 | <br> |
| 1175 | * The amount of memory used by Valgrind to store debug info (unwind<br> |
| 1176 | info, line number information and symbol data) has been<br> |
| 1177 | significantly reduced, even though Valgrind now reads more<br> |
| 1178 | information in order to support unwinding of inlined function calls.<br> |
| 1179 | <br> |
| 1180 | * Dwarf3 handling with --read-var-info=yes has been improved:<br> |
| 1181 | <br> |
| 1182 | - Ada and C struct containing VLAs no longer cause a "bad DIE" error<br> |
| 1183 | <br> |
| 1184 | - Code compiled with<br> |
| 1185 | -ffunction-sections -fdata-sections -Wl,--gc-sections<br> |
| 1186 | no longer causes assertion failures.<br> |
| 1187 | <br> |
| 1188 | * Improved checking for the --sim-hints= and --kernel-variant=<br> |
| 1189 | options. Unknown strings are now detected and reported to the user<br> |
| 1190 | as a usage error.<br> |
| 1191 | <br> |
| 1192 | * The semantics of stack start/end boundaries in the valgrind.h<br> |
| 1193 | VALGRIND_STACK_REGISTER client request has been clarified and<br> |
| 1194 | documented. The convention is that start and end are respectively<br> |
| 1195 | the lowest and highest addressable bytes of the stack.<br> |
| 1196 | <br> |
| 1197 | * ==================== FIXED BUGS ====================<br> |
| 1198 | <br> |
| 1199 | The following bugs have been fixed or resolved. Note that "n-i-bz"<br> |
| 1200 | stands for "not in bugzilla" -- that is, a bug that was reported to us<br> |
| 1201 | but never got a bugzilla entry. We encourage you to file bugs in<br> |
| 1202 | bugzilla (https://bugs.kde.org/enter_bug.cgi?product=valgrind) rather<br> |
| 1203 | than mailing the developers (or mailing lists) directly -- bugs that<br> |
| 1204 | are not entered into bugzilla tend to get forgotten about or ignored.<br> |
| 1205 | <br> |
| 1206 | To see details of a given bug, visit<br> |
| 1207 | https://bugs.kde.org/show_bug.cgi?id=XXXXXX<br> |
| 1208 | where XXXXXX is the bug number as listed below.<br> |
| 1209 | <br> |
| 1210 | 175819 Support for ipv6 socket reporting with --track-fds<br> |
| 1211 | 232510 make distcheck fails<br> |
| 1212 | 249435 Analyzing wine programs with callgrind triggers a crash<br> |
| 1213 | 278972 support for inlined function calls in stacktraces and suppression<br> |
| 1214 | == 199144<br> |
| 1215 | 291310 FXSAVE instruction marks memory as undefined on amd64<br> |
| 1216 | 303536 ioctl for SIOCETHTOOL (ethtool(8)) isn't wrapped<br> |
| 1217 | 308729 vex x86->IR: unhandled instruction bytes 0xf 0x5 (syscall) <br> |
| 1218 | 315199 vgcore file for threaded app does not show which thread crashed<br> |
| 1219 | 315952 tun/tap ioctls are not supported<br> |
| 1220 | 323178 Unhandled instruction: PLDW register (ARM) <br> |
| 1221 | 323179 Unhandled instruction: PLDW immediate (ARM)<br> |
| 1222 | 324050 Helgrind: SEGV because of unaligned stack when using movdqa<br> |
| 1223 | 325110 Add test-cases for Power ISA 2.06 insns: divdo/divdo. and divduo/divduo.<br> |
| 1224 | 325124 [MIPSEL] Compilation error<br> |
| 1225 | 325477 Phase 4 support for IBM Power ISA 2.07<br> |
| 1226 | 325538 cavium octeon mips64, valgrind reported "dumping core" [...]<br> |
| 1227 | 325628 Phase 5 support for IBM Power ISA 2.07<br> |
| 1228 | 325714 Empty vgcore but RLIMIT_CORE is big enough (too big) <br> |
| 1229 | 325751 Missing the two privileged Power PC Transactional Memory Instructions<br> |
| 1230 | 325816 Phase 6 support for IBM Power ISA 2.07<br> |
| 1231 | 325856 Make SGCheck fail gracefully on unsupported platforms<br> |
| 1232 | 326026 Iop names for count leading zeros/sign bits incorrectly imply [..]<br> |
| 1233 | 326436 DRD: False positive in libstdc++ std::list::push_back<br> |
| 1234 | 326444 Cavium MIPS Octeon Specific Load Indexed Instructions<br> |
| 1235 | 326462 Refactor vgdb to isolate invoker stuff into separate module<br> |
| 1236 | 326469 amd64->IR: 0x66 0xF 0x3A 0x63 0xC1 0xE (pcmpistri 0x0E)<br> |
| 1237 | 326623 DRD: false positive conflict report in a field assignment<br> |
| 1238 | 326724 Valgrind does not compile on OSX 1.9 Mavericks<br> |
| 1239 | 326816 Intercept for __strncpy_sse2_unaligned missing?<br> |
| 1240 | 326921 coregrind fails to compile m_trampoline.S with MIPS/Linux port of V<br> |
| 1241 | 326983 Clear direction flag after tests on amd64.<br> |
| 1242 | 327212 Do not prepend the current directory to absolute path names.<br> |
| 1243 | 327223 Support for Cavium MIPS Octeon Atomic and Count Instructions<br> |
| 1244 | 327238 Callgrind Assertion 'passed <= last_bb->cjmp_count' failed<br> |
| 1245 | 327284 s390x: Fix translation of the risbg instruction<br> |
| 1246 | 327639 vex amd64->IR pcmpestri SSE4.2 instruction is unsupported 0x34<br> |
| 1247 | 327837 dwz compressed alternate .debug_info and .debug_str not read correctly<br> |
| 1248 | 327916 DW_TAG_typedef may have no name<br> |
| 1249 | 327943 s390x: add a redirection for the 'index' function<br> |
| 1250 | 328100 XABORT not implemented<br> |
| 1251 | 328205 Implement additional Xen hypercalls<br> |
| 1252 | 328454 add support Backtraces with ARM unwind tables (EXIDX)<br> |
| 1253 | 328455 s390x: SIGILL after emitting wrong register pair for ldxbr<br> |
| 1254 | 328711 valgrind.1 manpage "memcheck options" section is badly generated<br> |
| 1255 | 328878 vex amd64->IR pcmpestri SSE4.2 instruction is unsupported 0x14<br> |
| 1256 | 329612 Incorrect handling of AT_BASE for image execution <br> |
| 1257 | 329694 clang warns about using uninitialized variable <br> |
| 1258 | 329956 valgrind crashes when lmw/stmw instructions are used on ppc64<br> |
| 1259 | 330228 mmap must align to VKI_SHMLBA on mips32<br> |
| 1260 | 330257 LLVM does not support `-mno-dynamic-no-pic` option<br> |
| 1261 | 330319 amd64->IR: unhandled instruction bytes: 0xF 0x1 0xD5 (xend)<br> |
| 1262 | 330459 --track-fds=yes doesn't track eventfds<br> |
| 1263 | 330469 Add clock_adjtime syscall support<br> |
| 1264 | 330594 Missing sysalls on PowerPC / uClibc<br> |
| 1265 | 330622 Add test to regression suite for POWER instruction: dcbzl<br> |
| 1266 | 330939 Support for AMD's syscall instruction on x86<br> |
| 1267 | == 308729<br> |
| 1268 | 330941 Typo in PRE(poll) syscall wrapper<br> |
| 1269 | 331057 unhandled instruction: 0xEEE01B20 (vfma.f64) (has patch)<br> |
| 1270 | 331254 Fix expected output for memcheck/tests/dw4<br> |
| 1271 | 331255 Fix race condition in test none/tests/coolo_sigaction<br> |
| 1272 | 331257 Fix type of jump buffer in test none/tests/faultstatus<br> |
| 1273 | 331305 configure uses bash specific syntax<br> |
| 1274 | 331337 s390x WARNING: unhandled syscall: 326 (dup3)<br> |
| 1275 | 331380 Syscall param timer_create(evp) points to uninitialised byte(s)<br> |
| 1276 | 331476 Patch to handle ioctl 0x5422 on Linux (x86 and amd64)<br> |
| 1277 | 331829 Unexpected ioctl opcode sign extension<br> |
| 1278 | 331830 ppc64: WARNING: unhandled syscall: 96/97<br> |
| 1279 | 331839 drd/tests/sem_open specifies invalid semaphore name <br> |
| 1280 | 331847 outcome of drd/tests/thread_name is nondeterministic<br> |
| 1281 | 332037 Valgrind cannot handle Thumb "add pc, reg"<br> |
| 1282 | 332055 drd asserts on platforms with VG_STACK_REDZONE_SZB == 0 and<br> |
| 1283 | consistency checks enabled<br> |
| 1284 | 332263 intercepts for pthread_rwlock_timedrdlock and<br> |
| 1285 | pthread_rwlock_timedwrlock are incorrect<br> |
| 1286 | 332265 drd could do with post-rwlock_init and pre-rwlock_destroy<br> |
| 1287 | client requests<br> |
| 1288 | 332276 Implement additional Xen hypercalls<br> |
| 1289 | 332658 ldrd.w r1, r2, [PC, #imm] does not adjust for 32bit alignment<br> |
| 1290 | 332765 Fix ms_print to create temporary files in a proper directory<br> |
| 1291 | 333072 drd: Add semaphore annotations<br> |
| 1292 | 333145 Tests for missaligned PC+#imm access for arm<br> |
| 1293 | 333228 AAarch64 Missing instruction encoding: mrs %[reg], ctr_el0<br> |
| 1294 | 333230 AAarch64 missing instruction encodings: dc, ic, dsb.<br> |
| 1295 | 333248 WARNING: unhandled syscall: unix:443<br> |
| 1296 | 333428 ldr.w pc [rD, #imm] instruction leads to assertion<br> |
| 1297 | 333501 cachegrind: assertion: Cache set count is not a power of two.<br> |
| 1298 | == 336577<br> |
| 1299 | == 292281<br> |
| 1300 | 333666 Recognize MPX instructions and bnd prefix.<br> |
| 1301 | 333788 Valgrind does not support the CDROM_DISC_STATUS ioctl (has patch)<br> |
| 1302 | 333817 Valgrind reports the memory areas written to by the SG_IO<br> |
| 1303 | ioctl as untouched<br> |
| 1304 | 334049 lzcnt fails silently (x86_32)<br> |
| 1305 | 334384 Valgrind does not have support Little Endian support for<br> |
| 1306 | IBM POWER PPC 64<br> |
| 1307 | 334585 recvmmsg unhandled (+patch) (arm)<br> |
| 1308 | 334705 sendmsg and recvmsg should guard against bogus msghdr fields.<br> |
| 1309 | 334727 Build fails with -Werror=format-security<br> |
| 1310 | 334788 clarify doc about --log-file initial program directory<br> |
| 1311 | 334834 PPC64 Little Endian support, patch 2<br> |
| 1312 | 334836 PPC64 Little Endian support, patch 3 testcase fixes<br> |
| 1313 | 334936 patch to fix false positives on alsa SNDRV_CTL_* ioctls<br> |
| 1314 | 335034 Unhandled ioctl: HCIGETDEVLIST<br> |
| 1315 | 335155 vgdb, fix error print statement.<br> |
| 1316 | 335262 arm64: movi 8bit version is not supported<br> |
| 1317 | 335263 arm64: dmb instruction is not implemented<br> |
| 1318 | 335441 unhandled ioctl 0x8905 (SIOCATMARK) when running wine under valgrind<br> |
| 1319 | 335496 arm64: sbc/abc instructions are not implemented<br> |
| 1320 | 335554 arm64: unhandled instruction: abs<br> |
| 1321 | 335564 arm64: unhandled instruction: fcvtpu Xn, Sn<br> |
| 1322 | 335735 arm64: unhandled instruction: cnt<br> |
| 1323 | 335736 arm64: unhandled instruction: uaddlv<br> |
| 1324 | 335848 arm64: unhandled instruction: {s,u}cvtf<br> |
| 1325 | 335902 arm64: unhandled instruction: sli<br> |
| 1326 | 335903 arm64: unhandled instruction: umull (vector)<br> |
| 1327 | 336055 arm64: unhandled instruction: mov (element)<br> |
| 1328 | 336062 arm64: unhandled instruction: shrn{,2}<br> |
| 1329 | 336139 mip64: [...] valgrind hangs and spins on a single core [...]<br> |
| 1330 | 336189 arm64: unhandled Instruction: mvn<br> |
| 1331 | 336435 Valgrind hangs in pthread_spin_lock consuming 100% CPU<br> |
| 1332 | 336619 valgrind --read-var-info=yes doesn't handle DW_TAG_restrict_type<br> |
| 1333 | 336772 Make moans about unknown ioctls more informative<br> |
| 1334 | 336957 Add a section about the Solaris/illumos port on the webpage<br> |
| 1335 | 337094 ifunc wrapper is broken on ppc64<br> |
| 1336 | 337285 fcntl commands F_OFD_SETLK, F_OFD_SETLKW, and F_OFD_GETLK not supported<br> |
| 1337 | 337528 leak check heuristic for block prefixed by length as 64bit number<br> |
| 1338 | 337740 Implement additional Xen hypercalls<br> |
| 1339 | 337762 guest_arm64_toIR.c:4166 (dis_ARM64_load_store): Assertion `0' failed.<br> |
| 1340 | 337766 arm64-linux: unhandled syscalls mlock (228) and mlockall (230)<br> |
| 1341 | 337871 deprecate --db-attach<br> |
| 1342 | 338023 Add support for all V4L2/media ioctls<br> |
| 1343 | 338024 inlined functions are not shown if DW_AT_ranges is used<br> |
| 1344 | 338106 Add support for 'kcmp' syscall<br> |
| 1345 | 338115 DRD: computed conflict set differs from actual after fork<br> |
| 1346 | 338160 implement display of thread local storage in gdbsrv<br> |
| 1347 | 338205 configure.ac and check for -Wno-tautological-compare<br> |
| 1348 | 338300 coredumps are missing one byte of every segment<br> |
| 1349 | 338445 amd64 vbit-test fails with unknown opcodes used by arm64 VEX<br> |
| 1350 | 338499 --sim-hints parsing broken due to wrong order in tokens<br> |
| 1351 | 338615 suppress glibc 2.20 optimized strcmp implementation for ARMv7<br> |
| 1352 | 338681 Unable to unwind through clone thread created on i386-linux<br> |
| 1353 | 338698 race condition between gdbsrv and vgdb on startup<br> |
| 1354 | 338703 helgrind on arm-linux gets false positives in dynamic loader<br> |
| 1355 | 338791 alt dwz files can be relative of debug/main file<br> |
| 1356 | 338878 on MacOS: assertion 'VG_IS_PAGE_ALIGNED(clstack_end+1)' failed<br> |
| 1357 | 338932 build V-trunk with gcc-trunk<br> |
| 1358 | 338974 glibc 2.20 changed size of struct sigaction sa_flags field on s390<br> |
| 1359 | 345079 Fix build problems in VEX/useful/test_main.c<br> |
| 1360 | n-i-bz Fix KVM_CREATE_IRQCHIP ioctl handling<br> |
| 1361 | n-i-bz s390x: Fix memory corruption for multithreaded applications<br> |
| 1362 | n-i-bz vex arm->IR: allow PC as basereg in some LDRD cases<br> |
| 1363 | n-i-bz internal error in Valgrind if vgdb transmit signals when ptrace invoked<br> |
| 1364 | n-i-bz Fix mingw64 support in valgrind.h (dev@, 9 May 2014)<br> |
| 1365 | n-i-bz drd manual: Document how to C++11 programs that use class "std::thread"<br> |
| 1366 | n-i-bz Add command-line option --default-suppressions<br> |
| 1367 | n-i-bz Add support for BLKDISCARDZEROES ioctl<br> |
| 1368 | n-i-bz ppc32/64: fix a regression with the mtfsb0/mtfsb1 instructions<br> |
| 1369 | n-i-bz Add support for sys_pivot_root and sys_unshare<br> |
| 1370 | <br> |
| 1371 | (3.10.0.BETA1: 2 September 2014, vex r2940, valgrind r14428)<br> |
| 1372 | (3.10.0.BETA2: 8 September 2014, vex r2950, valgrind r14503)<br> |
| 1373 | (3.10.0: 10 September 2014, vex r2950, valgrind r14514)<br> |
| 1374 | <br> |
| 1375 | <br> |
| 1376 | <br> |
| 1377 | Release 3.9.0 (31 October 2013)<br> |
| 1378 | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~<br> |
| 1379 | 3.9.0 is a feature release with many improvements and the usual<br> |
| 1380 | collection of bug fixes.<br> |
| 1381 | <br> |
| 1382 | This release supports X86/Linux, AMD64/Linux, ARM/Linux, PPC32/Linux,<br> |
| 1383 | PPC64/Linux, S390X/Linux, MIPS32/Linux, MIPS64/Linux, ARM/Android,<br> |
| 1384 | X86/Android, X86/MacOSX 10.7 and AMD64/MacOSX 10.7. Support for<br> |
| 1385 | MacOSX 10.8 is significantly improved relative to the 3.8.0 release.<br> |
| 1386 | <br> |
| 1387 | * ================== PLATFORM CHANGES =================<br> |
| 1388 | <br> |
| 1389 | * Support for MIPS64 LE and BE running Linux. Valgrind has been<br> |
| 1390 | tested on MIPS64 Debian Squeeze and Debian Wheezy distributions.<br> |
| 1391 | <br> |
| 1392 | * Support for MIPS DSP ASE on MIPS32 platforms.<br> |
| 1393 | <br> |
| 1394 | * Support for s390x Decimal Floating Point instructions on hosts that<br> |
| 1395 | have the DFP facility installed.<br> |
| 1396 | <br> |
| 1397 | * Support for POWER8 (Power ISA 2.07) instructions<br> |
| 1398 | <br> |
| 1399 | * Support for Intel AVX2 instructions. This is available only on 64<br> |
| 1400 | bit code.<br> |
| 1401 | <br> |
| 1402 | * Initial support for Intel Transactional Synchronization Extensions,<br> |
| 1403 | both RTM and HLE.<br> |
| 1404 | <br> |
| 1405 | * Initial support for Hardware Transactional Memory on POWER.<br> |
| 1406 | <br> |
| 1407 | * Improved support for MacOSX 10.8 (64-bit only). Memcheck can now<br> |
| 1408 | run large GUI apps tolerably well.<br> |
| 1409 | <br> |
| 1410 | * ==================== TOOL CHANGES ====================<br> |
| 1411 | <br> |
| 1412 | * Memcheck:<br> |
| 1413 | <br> |
| 1414 | - Improvements in handling of vectorised code, leading to<br> |
| 1415 | significantly fewer false error reports. You need to use the flag<br> |
| 1416 | --partial-loads-ok=yes to get the benefits of these changes.<br> |
| 1417 | <br> |
| 1418 | - Better control over the leak checker. It is now possible to<br> |
| 1419 | specify which leak kinds (definite/indirect/possible/reachable)<br> |
| 1420 | should be displayed, which should be regarded as errors, and which<br> |
| 1421 | should be suppressed by a given leak suppression. This is done<br> |
| 1422 | using the options --show-leak-kinds=kind1,kind2,..,<br> |
| 1423 | --errors-for-leak-kinds=kind1,kind2,.. and an optional<br> |
| 1424 | "match-leak-kinds:" line in suppression entries, respectively.<br> |
| 1425 | <br> |
| 1426 | Note that generated leak suppressions contain this new line and<br> |
| 1427 | are therefore more specific than in previous releases. To get the<br> |
| 1428 | same behaviour as previous releases, remove the "match-leak-kinds:"<br> |
| 1429 | line from generated suppressions before using them.<br> |
| 1430 | <br> |
| 1431 | - Reduced "possible leak" reports from the leak checker by the use<br> |
| 1432 | of better heuristics. The available heuristics provide detection<br> |
| 1433 | of valid interior pointers to std::stdstring, to new[] allocated<br> |
| 1434 | arrays with elements having destructors and to interior pointers<br> |
| 1435 | pointing to an inner part of a C++ object using multiple<br> |
| 1436 | inheritance. They can be selected individually using the<br> |
| 1437 | option --leak-check-heuristics=heur1,heur2,...<br> |
| 1438 | <br> |
| 1439 | - Better control of stacktrace acquisition for heap-allocated<br> |
| 1440 | blocks. Using the --keep-stacktraces option, it is possible to<br> |
| 1441 | control independently whether a stack trace is acquired for each<br> |
| 1442 | allocation and deallocation. This can be used to create better<br> |
| 1443 | "use after free" errors or to decrease Valgrind's resource<br> |
| 1444 | consumption by recording less information.<br> |
| 1445 | <br> |
| 1446 | - Better reporting of leak suppression usage. The list of used<br> |
| 1447 | suppressions (shown when the -v option is given) now shows, for<br> |
| 1448 | each leak suppressions, how many blocks and bytes it suppressed<br> |
| 1449 | during the last leak search.<br> |
| 1450 | <br> |
| 1451 | * Helgrind:<br> |
| 1452 | <br> |
| 1453 | - False errors resulting from the use of statically initialised<br> |
| 1454 | mutexes and condition variables (PTHREAD_MUTEX_INITIALISER, etc)<br> |
| 1455 | have been removed.<br> |
| 1456 | <br> |
| 1457 | - False errors resulting from the use of pthread_cond_waits that<br> |
| 1458 | timeout, have been removed.<br> |
| 1459 | <br> |
| 1460 | * ==================== OTHER CHANGES ====================<br> |
| 1461 | <br> |
| 1462 | * Some attempt to tune Valgrind's space requirements to the expected<br> |
| 1463 | capabilities of the target:<br> |
| 1464 | <br> |
| 1465 | - The default size of the translation cache has been reduced from 8<br> |
| 1466 | sectors to 6 on Android platforms, since each sector occupies<br> |
| 1467 | about 40MB when using Memcheck.<br> |
| 1468 | <br> |
| 1469 | - The default size of the translation cache has been increased to 16<br> |
| 1470 | sectors on all other platforms, reflecting the fact that large<br> |
| 1471 | applications require instrumentation and storage of huge amounts<br> |
| 1472 | of code. For similar reasons, the number of memory mapped<br> |
| 1473 | segments that can be tracked has been increased by a factor of 6.<br> |
| 1474 | <br> |
| 1475 | - In all cases, the maximum number of sectors in the translation<br> |
| 1476 | cache can be controlled by the new flag --num-transtab-sectors.<br> |
| 1477 | <br> |
| 1478 | * Changes in how debug info (line numbers, etc) is read:<br> |
| 1479 | <br> |
| 1480 | - Valgrind no longer temporarily mmaps the entire object to read<br> |
| 1481 | from it. Instead, reading is done through a small fixed sized<br> |
| 1482 | buffer. This avoids virtual memory usage spikes when Valgrind<br> |
| 1483 | reads debuginfo from large shared objects.<br> |
| 1484 | <br> |
| 1485 | - A new experimental remote debug info server. Valgrind can read<br> |
| 1486 | debug info from a different machine (typically, a build host)<br> |
| 1487 | where debuginfo objects are stored. This can save a lot of time<br> |
| 1488 | and hassle when running Valgrind on resource-constrained targets<br> |
| 1489 | (phones, tablets) when the full debuginfo objects are stored<br> |
| 1490 | somewhere else. This is enabled by the --debuginfo-server=<br> |
| 1491 | option.<br> |
| 1492 | <br> |
| 1493 | - Consistency checking between main and debug objects can be<br> |
| 1494 | disabled using the --allow-mismatched-debuginfo option.<br> |
| 1495 | <br> |
| 1496 | * Stack unwinding by stack scanning, on ARM. Unwinding by stack<br> |
| 1497 | scanning can recover stack traces in some cases when the normal<br> |
| 1498 | unwind mechanisms fail. Stack scanning is best described as "a<br> |
| 1499 | nasty, dangerous and misleading hack" and so is disabled by default.<br> |
| 1500 | Use --unw-stack-scan-thresh and --unw-stack-scan-frames to enable<br> |
| 1501 | and control it.<br> |
| 1502 | <br> |
| 1503 | * Detection and merging of recursive stack frame cycles. When your<br> |
| 1504 | program has recursive algorithms, this limits the memory used by<br> |
| 1505 | Valgrind for recorded stack traces and avoids recording<br> |
| 1506 | uninteresting repeated calls. This is controlled by the command<br> |
| 1507 | line option --merge-recursive-frame and by the monitor command<br> |
| 1508 | "v.set merge-recursive-frames".<br> |
| 1509 | <br> |
| 1510 | * File name and line numbers for used suppressions. The list of used<br> |
| 1511 | suppressions (shown when the -v option is given) now shows, for each<br> |
| 1512 | used suppression, the file name and line number where the suppression<br> |
| 1513 | is defined.<br> |
| 1514 | <br> |
| 1515 | * New and modified GDB server monitor features:<br> |
| 1516 | <br> |
| 1517 | - valgrind.h has a new client request, VALGRIND_MONITOR_COMMAND,<br> |
| 1518 | that can be used to execute gdbserver monitor commands from the<br> |
| 1519 | client program.<br> |
| 1520 | <br> |
| 1521 | - A new monitor command, "v.info open_fds", that gives the list of<br> |
| 1522 | open file descriptors and additional details.<br> |
| 1523 | <br> |
| 1524 | - An optional message in the "v.info n_errs_found" monitor command,<br> |
| 1525 | for example "v.info n_errs_found test 1234 finished", allowing a<br> |
| 1526 | comment string to be added to the process output, perhaps for the<br> |
| 1527 | purpose of separating errors of different tests or test phases.<br> |
| 1528 | <br> |
| 1529 | - A new monitor command "v.info execontext" that shows information<br> |
| 1530 | about the stack traces recorded by Valgrind.<br> |
| 1531 | <br> |
| 1532 | - A new monitor command "v.do expensive_sanity_check_general" to run<br> |
| 1533 | some internal consistency checks.<br> |
| 1534 | <br> |
| 1535 | * New flag --sigill-diagnostics to control whether a diagnostic<br> |
| 1536 | message is printed when the JIT encounters an instruction it can't<br> |
| 1537 | translate. The actual behavior -- delivery of SIGILL to the<br> |
| 1538 | application -- is unchanged.<br> |
| 1539 | <br> |
| 1540 | * The maximum amount of memory that Valgrind can use on 64 bit targets<br> |
| 1541 | has been increased from 32GB to 64GB. This should make it possible<br> |
| 1542 | to run applications on Memcheck that natively require up to about 35GB.<br> |
| 1543 | <br> |
| 1544 | * ==================== FIXED BUGS ====================<br> |
| 1545 | <br> |
| 1546 | The following bugs have been fixed or resolved. Note that "n-i-bz"<br> |
| 1547 | stands for "not in bugzilla" -- that is, a bug that was reported to us<br> |
| 1548 | but never got a bugzilla entry. We encourage you to file bugs in<br> |
| 1549 | bugzilla (https://bugs.kde.org/enter_bug.cgi?product=valgrind) rather<br> |
| 1550 | than mailing the developers (or mailing lists) directly -- bugs that<br> |
| 1551 | are not entered into bugzilla tend to get forgotten about or ignored.<br> |
| 1552 | <br> |
| 1553 | To see details of a given bug, visit<br> |
| 1554 | https://bugs.kde.org/show_bug.cgi?id=XXXXXX<br> |
| 1555 | where XXXXXX is the bug number as listed below.<br> |
| 1556 | <br> |
| 1557 | 123837 system call: 4th argument is optional, depending on cmd<br> |
| 1558 | 135425 memcheck should tell you where Freed blocks were Mallocd<br> |
| 1559 | 164485 VG_N_SEGNAMES and VG_N_SEGMENTS are (still) too small<br> |
| 1560 | 207815 Adds some of the drm ioctls to syswrap-linux.c <br> |
| 1561 | 251569 vex amd64->IR: 0xF 0x1 0xF9 0xBF 0x90 0xD0 0x3 0x0 (RDTSCP)<br> |
| 1562 | 252955 Impossible to compile with ccache<br> |
| 1563 | 253519 Memcheck reports auxv pointer accesses as invalid reads.<br> |
| 1564 | 263034 Crash when loading some PPC64 binaries<br> |
| 1565 | 269599 Increase deepest backtrace<br> |
| 1566 | 274695 s390x: Support "compare to/from logical" instructions (z196)<br> |
| 1567 | 275800 s390x: Autodetect cache info (part 2)<br> |
| 1568 | 280271 Valgrind reports possible memory leaks on still-reachable std::string<br> |
| 1569 | 284540 Memcheck shouldn't count suppressions matching still-reachable [..]<br> |
| 1570 | 289578 Backtraces with ARM unwind tables (stack scan flags)<br> |
| 1571 | 296311 Wrong stack traces due to -fomit-frame-pointer (x86) <br> |
| 1572 | 304832 ppc32: build failure<br> |
| 1573 | 305431 Use find_buildid shdr fallback for separate .debug files<br> |
| 1574 | 305728 Add support for AVX2 instructions<br> |
| 1575 | 305948 ppc64: code generation for ShlD64 / ShrD64 asserts<br> |
| 1576 | 306035 s390x: Fix IR generation for LAAG and friends<br> |
| 1577 | 306054 s390x: Condition code computation for convert-to-int/logical<br> |
| 1578 | 306098 s390x: alternate opcode form for convert to/from fixed<br> |
| 1579 | 306587 Fix cache line detection from auxiliary vector for PPC.<br> |
| 1580 | 306783 Mips unhandled syscall : 4025 / 4079 / 4182<br> |
| 1581 | 307038 DWARF2 CFI reader: unhandled DW_OP_ opcode 0x8 (DW_OP_const1u et al)<br> |
| 1582 | 307082 HG false positive: pthread_cond_destroy: destruction of unknown CV<br> |
| 1583 | 307101 sys_capget second argument can be NULL<br> |
| 1584 | 307103 sys_openat: If pathname is absolute, then dirfd is ignored.<br> |
| 1585 | 307106 amd64->IR: f0 0f c0 02 (lock xadd byte)<br> |
| 1586 | 307113 s390x: DFP support<br> |
| 1587 | 307141 valgrind does't work in mips-linux system<br> |
| 1588 | 307155 filter_gdb should filter out syscall-template.S T_PSEUDO<br> |
| 1589 | 307285 x86_amd64 feature test for avx in test suite is wrong<br> |
| 1590 | 307290 memcheck overlap testcase needs memcpy version filter<br> |
| 1591 | 307463 Please add "&limit=0" to the "all open bugs" link<br> |
| 1592 | 307465 --show-possibly-lost=no should reduce the error count / exit code<br> |
| 1593 | 307557 Leaks on Mac OS X 10.7.5 libraries at ImageLoader::recursiveInit[..]<br> |
| 1594 | 307729 pkgconfig support broken valgrind.pc<br> |
| 1595 | 307828 Memcheck false errors SSE optimized wcscpy, wcscmp, wcsrchr, wcschr<br> |
| 1596 | 307955 Building valgrind 3.7.0-r4 fails in Gentoo AMD64 when using clang<br> |
| 1597 | 308089 Unhandled syscall on ppc64: prctl<br> |
| 1598 | 308135 PPC32 MPC8xx has 16 bytes cache size<br> |
| 1599 | 308321 testsuite memcheck filter interferes with gdb_filter <br> |
| 1600 | 308333 == 307106<br> |
| 1601 | 308341 vgdb should report process exit (or fatal signal)<br> |
| 1602 | 308427 s390 memcheck reports tsearch cjump/cmove depends on uninit<br> |
| 1603 | 308495 Remove build dependency on installed Xen headers<br> |
| 1604 | 308573 Internal error on 64-bit instruction executed in 32-bit mode<br> |
| 1605 | 308626 == 308627<br> |
| 1606 | 308627 pmovmskb validity bit propagation is imprecise<br> |
| 1607 | 308644 vgdb command for having the info for the track-fds option<br> |
| 1608 | 308711 give more info about aspacemgr and arenas in out_of_memory<br> |
| 1609 | 308717 ARM: implement fixed-point VCVT.F64.[SU]32<br> |
| 1610 | 308718 ARM implement SMLALBB family of instructions<br> |
| 1611 | 308886 Missing support for PTRACE_SET/GETREGSET <br> |
| 1612 | 308930 syscall name_to_handle_at (303 on amd64) not handled<br> |
| 1613 | 309229 V-bit tester does not report number of tests generated<br> |
| 1614 | 309323 print unrecognized instuction on MIPS<br> |
| 1615 | 309425 Provide a --sigill-diagnostics flag to suppress illegal [..]<br> |
| 1616 | 309427 SSE optimized stpncpy trigger uninitialised value [..] errors<br> |
| 1617 | 309430 Self hosting ppc64 encounters a vassert error on operand type<br> |
| 1618 | 309600 valgrind is a bit confused about 0-sized sections<br> |
| 1619 | 309823 Generate errors for still reachable blocks<br> |
| 1620 | 309921 PCMPISTRI validity bit propagation is imprecise<br> |
| 1621 | 309922 none/tests/ppc64/test_dfp5 sometimes fails<br> |
| 1622 | 310169 The Iop_CmpORD class of Iops is not supported by the vbit checker.<br> |
| 1623 | 310424 --read-var-info does not properly describe static variables <br> |
| 1624 | 310792 search additional path for debug symbols<br> |
| 1625 | 310931 s390x: Message-security assist (MSA) instruction extension [..]<br> |
| 1626 | 311100 PPC DFP implementation of the integer operands is inconsistent [..]<br> |
| 1627 | 311318 ARM: "128-bit constant is not implemented" error message<br> |
| 1628 | 311407 ssse3 bcopy (actually converted memcpy) causes invalid read [..]<br> |
| 1629 | 311690 V crashes because it redirects branches inside of a redirected function<br> |
| 1630 | 311880 x86_64: make regtest hangs at shell_valid1<br> |
| 1631 | 311922 WARNING: unhandled syscall: 170<br> |
| 1632 | 311933 == 251569<br> |
| 1633 | 312171 ppc: insn selection for DFP<br> |
| 1634 | 312571 Rounding mode call wrong for the DFP Iops [..]<br> |
| 1635 | 312620 Change to Iop_D32toD64 [..] for s390 DFP support broke ppc [..]<br> |
| 1636 | 312913 Dangling pointers error should also report the alloc stack trace<br> |
| 1637 | 312980 Building on Mountain Lion generates some compiler warnings<br> |
| 1638 | 313267 Adding MIPS64/Linux port to Valgrind<br> |
| 1639 | 313348 == 251569<br> |
| 1640 | 313354 == 251569<br> |
| 1641 | 313811 Buffer overflow in assert_fail<br> |
| 1642 | 314099 coverity pointed out error in VEX guest_ppc_toIR.c insn_suffix<br> |
| 1643 | 314269 ppc: dead code in insn selection<br> |
| 1644 | 314718 ARM: implement integer divide instruction (sdiv and udiv)<br> |
| 1645 | 315345 cl-format.xml and callgrind/dump.c don't agree on using cfl= or cfi=<br> |
| 1646 | 315441 sendmsg syscall should ignore unset msghdr msg_flags<br> |
| 1647 | 315534 msgrcv inside a thread causes valgrind to hang (block)<br> |
| 1648 | 315545 Assertion '(UChar*)sec->tt[tteNo].tcptr <= (UChar*)hcode' failed<br> |
| 1649 | 315689 disInstr(thumb): unhandled instruction: 0xF852 0x0E10 (LDRT)<br> |
| 1650 | 315738 disInstr(arm): unhandled instruction: 0xEEBE0BEE (vcvt.s32.f64)<br> |
| 1651 | 315959 valgrind man page has bogus SGCHECK (and no BBV) OPTIONS section<br> |
| 1652 | 316144 valgrind.1 manpage contains unknown ??? strings [..]<br> |
| 1653 | 316145 callgrind command line options in manpage reference (unknown) [..]<br> |
| 1654 | 316145 callgrind command line options in manpage reference [..]<br> |
| 1655 | 316181 drd: Fixed a 4x slowdown for certain applications<br> |
| 1656 | 316503 Valgrind does not support SSE4 "movntdqa" instruction<br> |
| 1657 | 316535 Use of |signed int| instead of |size_t| in valgrind messages<br> |
| 1658 | 316696 fluidanimate program of parsec 2.1 stuck <br> |
| 1659 | 316761 syscall open_by_handle_at (304 on amd64, 342 on x86) not handled<br> |
| 1660 | 317091 Use -Wl,-Ttext-segment when static linking if possible [..]<br> |
| 1661 | 317186 "Impossible happens" when occurs VCVT instruction on ARM<br> |
| 1662 | 317318 Support for Threading Building Blocks "scalable_malloc"<br> |
| 1663 | 317444 amd64->IR: 0xC4 0x41 0x2C 0xC2 0xD2 0x8 (vcmpeq_uqps)<br> |
| 1664 | 317461 Fix BMI assembler configure check and avx2/bmi/fma vgtest prereqs<br> |
| 1665 | 317463 bmi testcase IR SANITY CHECK FAILURE<br> |
| 1666 | 317506 memcheck/tests/vbit-test fails with unknown opcode after [..]<br> |
| 1667 | 318050 libmpiwrap fails to compile with out-of-source build<br> |
| 1668 | 318203 setsockopt handling needs to handle SOL_SOCKET/SO_ATTACH_FILTER<br> |
| 1669 | 318643 annotate_trace_memory tests infinite loop on arm and ppc [..]<br> |
| 1670 | 318773 amd64->IR: 0xF3 0x48 0x0F 0xBC 0xC2 0xC3 0x66 0x0F<br> |
| 1671 | 318929 Crash with: disInstr(thumb): 0xF321 0x0001 (ssat16)<br> |
| 1672 | 318932 Add missing PPC64 and PPC32 system call support<br> |
| 1673 | 319235 --db-attach=yes is broken with Yama (ptrace scoping) enabled<br> |
| 1674 | 319395 Crash with unhandled instruction on STRT (Thumb) instructions<br> |
| 1675 | 319494 VEX Makefile-gcc standalone build update after r2702<br> |
| 1676 | 319505 [MIPSEL] Crash: unhandled UNRAY operator.<br> |
| 1677 | 319858 disInstr(thumb): unhandled instruction on instruction STRBT<br> |
| 1678 | 319932 disInstr(thumb): unhandled instruction on instruction STRHT<br> |
| 1679 | 320057 Problems when we try to mmap more than 12 memory pages on MIPS32<br> |
| 1680 | 320063 Memory from PTRACE_GET_THREAD_AREA is reported uninitialised<br> |
| 1681 | 320083 disInstr(thumb): unhandled instruction on instruction LDRBT<br> |
| 1682 | 320116 bind on AF_BLUETOOTH produces warnings because of sockaddr_rc padding<br> |
| 1683 | 320131 WARNING: unhandled syscall: 369 on ARM (prlimit64)<br> |
| 1684 | 320211 Stack buffer overflow in ./coregrind/m_main.c with huge TMPDIR<br> |
| 1685 | 320661 vgModuleLocal_read_elf_debug_info(): "Assertion '!di->soname'<br> |
| 1686 | 320895 add fanotify support (patch included)<br> |
| 1687 | 320998 vex amd64->IR pcmpestri and pcmpestrm SSE4.2 instruction<br> |
| 1688 | 321065 Valgrind updates for Xen 4.3<br> |
| 1689 | 321148 Unhandled instruction: PLI (Thumb 1, 2, 3)<br> |
| 1690 | 321363 Unhandled instruction: SSAX (ARM + Thumb)<br> |
| 1691 | 321364 Unhandled instruction: SXTAB16 (ARM + Thumb)<br> |
| 1692 | 321466 Unhandled instruction: SHASX (ARM + Thumb)<br> |
| 1693 | 321467 Unhandled instruction: SHSAX (ARM + Thumb)<br> |
| 1694 | 321468 Unhandled instruction: SHSUB16 (ARM + Thumb)<br> |
| 1695 | 321619 Unhandled instruction: SHSUB8 (ARM + Thumb)<br> |
| 1696 | 321620 Unhandled instruction: UASX (ARM + Thumb)<br> |
| 1697 | 321621 Unhandled instruction: USAX (ARM + Thumb)<br> |
| 1698 | 321692 Unhandled instruction: UQADD16 (ARM + Thumb)<br> |
| 1699 | 321693 Unhandled instruction: LDRSBT (Thumb)<br> |
| 1700 | 321694 Unhandled instruction: UQASX (ARM + Thumb)<br> |
| 1701 | 321696 Unhandled instruction: UQSAX (Thumb + ARM)<br> |
| 1702 | 321697 Unhandled instruction: UHASX (ARM + Thumb)<br> |
| 1703 | 321703 Unhandled instruction: UHSAX (ARM + Thumb)<br> |
| 1704 | 321704 Unhandled instruction: REVSH (ARM + Thumb)<br> |
| 1705 | 321730 Add cg_diff and cg_merge man pages<br> |
| 1706 | 321738 Add vgdb and valgrind-listener man pages<br> |
| 1707 | 321814 == 315545<br> |
| 1708 | 321891 Unhandled instruction: LDRHT (Thumb)<br> |
| 1709 | 321960 pthread_create() then alloca() causing invalid stack write errors<br> |
| 1710 | 321969 ppc32 and ppc64 don't support [lf]setxattr<br> |
| 1711 | 322254 Show threadname together with tid if set by application<br> |
| 1712 | 322294 Add initial support for IBM Power ISA 2.07<br> |
| 1713 | 322368 Assertion failure in wqthread_hijack under OS X 10.8<br> |
| 1714 | 322563 vex mips->IR: 0x70 0x83 0xF0 0x3A<br> |
| 1715 | 322807 VALGRIND_PRINTF_BACKTRACE writes callstack to xml and text to stderr<br> |
| 1716 | 322851 0bXXX binary literal syntax is not standard <br> |
| 1717 | 323035 Unhandled instruction: LDRSHT(Thumb)<br> |
| 1718 | 323036 Unhandled instruction: SMMLS (ARM and Thumb)<br> |
| 1719 | 323116 The memcheck/tests/ppc64/power_ISA2_05.c fails to build [..]<br> |
| 1720 | 323175 Unhandled instruction: SMLALD (ARM + Thumb)<br> |
| 1721 | 323177 Unhandled instruction: SMLSLD (ARM + Thumb)<br> |
| 1722 | 323432 Calling pthread_cond_destroy() or pthread_mutex_destroy() [..]<br> |
| 1723 | 323437 Phase 2 support for IBM Power ISA 2.07<br> |
| 1724 | 323713 Support mmxext (integer sse) subset on i386 (athlon)<br> |
| 1725 | 323803 Transactional memory instructions are not supported for Power<br> |
| 1726 | 323893 SSE3 not available on amd cpus in valgrind<br> |
| 1727 | 323905 Probable false positive from Valgrind/drd on close()<br> |
| 1728 | 323912 valgrind.h header isn't compatible for mingw64<br> |
| 1729 | 324047 Valgrind doesn't support [LDR,ST]{S}[B,H]T ARM instructions<br> |
| 1730 | 324149 helgrind: When pthread_cond_timedwait returns ETIMEDOUT [..]<br> |
| 1731 | 324181 mmap does not handle MAP_32BIT<br> |
| 1732 | 324227 memcheck false positive leak when a thread calls exit+block [..]<br> |
| 1733 | 324421 Support for fanotify API on ARM architecture<br> |
| 1734 | 324514 gdbserver monitor cmd output behaviour consistency [..]<br> |
| 1735 | 324518 ppc64: Emulation of dcbt instructions does not handle [..]<br> |
| 1736 | 324546 none/tests/ppc32 test_isa_2_07_part2 requests -m64<br> |
| 1737 | 324582 When access is made to freed memory, report both allocation [..]<br> |
| 1738 | 324594 Fix overflow computation for Power ISA 2.06 insns: mulldo/mulldo.<br> |
| 1739 | 324765 ppc64: illegal instruction when executing none/tests/ppc64/jm-misc<br> |
| 1740 | 324816 Incorrect VEX implementation for xscvspdp/xvcvspdp for SNaN inputs<br> |
| 1741 | 324834 Unhandled instructions in Microsoft C run-time for x86_64<br> |
| 1742 | 324894 Phase 3 support for IBM Power ISA 2.07<br> |
| 1743 | 326091 drd: Avoid false race reports from optimized strlen() impls<br> |
| 1744 | 326113 valgrind libvex hwcaps error on AMD64 <br> |
| 1745 | n-i-bz Some wrong command line options could be ignored<br> |
| 1746 | n-i-bz patch to allow fair-sched on android<br> |
| 1747 | n-i-bz report error for vgdb snapshot requested before execution<br> |
| 1748 | n-i-bz same as 303624 (fixed in 3.8.0), but for x86 android<br> |
| 1749 | <br> |
| 1750 | (3.9.0: 31 October 2013, vex r2796, valgrind r13708)<br> |
| 1751 | <br> |
| 1752 | <br> |
| 1753 | <br> |
| 1754 | Release 3.8.1 (19 September 2012)<br> |
| 1755 | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~<br> |
| 1756 | 3.8.1 is a bug fix release. It fixes some assertion failures in 3.8.0<br> |
| 1757 | that occur moderately frequently in real use cases, adds support for<br> |
| 1758 | some missing instructions on ARM, and fixes a deadlock condition on<br> |
| 1759 | MacOSX. If you package or deliver 3.8.0 for others to use, you might<br> |
| 1760 | want to consider upgrading to 3.8.1 instead.<br> |
| 1761 | <br> |
| 1762 | The following bugs have been fixed or resolved. Note that "n-i-bz"<br> |
| 1763 | stands for "not in bugzilla" -- that is, a bug that was reported to us<br> |
| 1764 | but never got a bugzilla entry. We encourage you to file bugs in<br> |
| 1765 | bugzilla (https://bugs.kde.org/enter_bug.cgi?product=valgrind) rather<br> |
| 1766 | than mailing the developers (or mailing lists) directly -- bugs that<br> |
| 1767 | are not entered into bugzilla tend to get forgotten about or ignored.<br> |
| 1768 | <br> |
| 1769 | To see details of a given bug, visit<br> |
| 1770 | https://bugs.kde.org/show_bug.cgi?id=XXXXXX<br> |
| 1771 | where XXXXXX is the bug number as listed below.<br> |
| 1772 | <br> |
| 1773 | 284004 == 301281<br> |
| 1774 | 289584 Unhandled instruction: 0xF 0x29 0xE5 (MOVAPS)<br> |
| 1775 | 295808 amd64->IR: 0xF3 0xF 0xBC 0xC0 (TZCNT)<br> |
| 1776 | 298281 wcslen causes false(?) uninitialised value warnings<br> |
| 1777 | 301281 valgrind hangs on OS X when the process calls system()<br> |
| 1778 | 304035 disInstr(arm): unhandled instruction 0xE1023053<br> |
| 1779 | 304867 implement MOVBE instruction in x86 mode<br> |
| 1780 | 304980 Assertion 'lo <= hi' failed in vgModuleLocal_find_rx_mapping<br> |
| 1781 | 305042 amd64: implement 0F 7F encoding of movq between two registers<br> |
| 1782 | 305199 ARM: implement QDADD and QDSUB<br> |
| 1783 | 305321 amd64->IR: 0xF 0xD 0xC (prefetchw)<br> |
| 1784 | 305513 killed by fatal signal: SIGSEGV<br> |
| 1785 | 305690 DRD reporting invalid semaphore when sem_trywait fails<br> |
| 1786 | 305926 Invalid alignment checks for some AVX instructions<br> |
| 1787 | 306297 disInstr(thumb): unhandled instruction 0xE883 0x000C<br> |
| 1788 | 306310 3.8.0 release tarball missing some files<br> |
| 1789 | 306612 RHEL 6 glibc-2.X default suppressions need /lib*/libc-*patterns<br> |
| 1790 | 306664 vex amd64->IR: 0x66 0xF 0x3A 0x62 0xD1 0x46 0x66 0xF<br> |
| 1791 | n-i-bz shmat of a segment > 4Gb does not work <br> |
| 1792 | n-i-bz simulate_control_c script wrong USR1 signal number on mips<br> |
| 1793 | n-i-bz vgdb ptrace calls wrong on mips [...]<br> |
| 1794 | n-i-bz Fixes for more MPI false positives<br> |
| 1795 | n-i-bz exp-sgcheck's memcpy causes programs to segfault<br> |
| 1796 | n-i-bz OSX build w/ clang: asserts at startup<br> |
| 1797 | n-i-bz Incorrect undef'dness prop for Iop_DPBtoBCD and Iop_BCDtoDPB<br> |
| 1798 | n-i-bz fix a couple of union tag-vs-field mixups<br> |
| 1799 | n-i-bz OSX: use __NR_poll_nocancel rather than __NR_poll<br> |
| 1800 | <br> |
| 1801 | The following bugs were fixed in 3.8.0 but not listed in this NEWS<br> |
| 1802 | file at the time:<br> |
| 1803 | <br> |
| 1804 | 254088 Valgrind should know about UD2 instruction<br> |
| 1805 | 301280 == 254088<br> |
| 1806 | 301902 == 254088<br> |
| 1807 | 304754 NEWS blows TeX's little mind<br> |
| 1808 | <br> |
| 1809 | (3.8.1: 19 September 2012, vex r2537, valgrind r12996)<br> |
| 1810 | <br> |
| 1811 | <br> |
| 1812 | <br> |
| 1813 | Release 3.8.0 (10 August 2012)<br> |
| 1814 | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~<br> |
| 1815 | 3.8.0 is a feature release with many improvements and the usual<br> |
| 1816 | collection of bug fixes.<br> |
| 1817 | <br> |
| 1818 | This release supports X86/Linux, AMD64/Linux, ARM/Linux, PPC32/Linux,<br> |
| 1819 | PPC64/Linux, S390X/Linux, MIPS/Linux, ARM/Android, X86/Android,<br> |
| 1820 | X86/MacOSX 10.6/10.7 and AMD64/MacOSX 10.6/10.7. Support for recent<br> |
| 1821 | distros and toolchain components (glibc 2.16, gcc 4.7) has been added.<br> |
| 1822 | There is initial support for MacOSX 10.8, but it is not usable for<br> |
| 1823 | serious work at present.<br> |
| 1824 | <br> |
| 1825 | * ================== PLATFORM CHANGES =================<br> |
| 1826 | <br> |
| 1827 | * Support for MIPS32 platforms running Linux. Valgrind has been<br> |
| 1828 | tested on MIPS32 and MIPS32r2 platforms running different Debian<br> |
| 1829 | Squeeze and MeeGo distributions. Both little-endian and big-endian<br> |
| 1830 | cores are supported. The tools Memcheck, Massif and Lackey have<br> |
| 1831 | been tested and are known to work. See README.mips for more details.<br> |
| 1832 | <br> |
| 1833 | * Preliminary support for Android running on x86.<br> |
| 1834 | <br> |
| 1835 | * Preliminary (as-yet largely unusable) support for MacOSX 10.8.<br> |
| 1836 | <br> |
| 1837 | * Support for Intel AVX instructions and for AES instructions. This<br> |
| 1838 | support is available only for 64 bit code.<br> |
| 1839 | <br> |
| 1840 | * Support for POWER Decimal Floating Point instructions.<br> |
| 1841 | <br> |
| 1842 | * ==================== TOOL CHANGES ====================<br> |
| 1843 | <br> |
| 1844 | * Non-libc malloc implementations are now supported. This is useful<br> |
| 1845 | for tools that replace malloc (Memcheck, Massif, DRD, Helgrind).<br> |
| 1846 | Using the new option --soname-synonyms, such tools can be informed<br> |
| 1847 | that the malloc implementation is either linked statically into the<br> |
| 1848 | executable, or is present in some other shared library different<br> |
| 1849 | from libc.so. This makes it possible to process statically linked<br> |
| 1850 | programs, and programs using other malloc libraries, for example<br> |
| 1851 | TCMalloc or JEMalloc.<br> |
| 1852 | <br> |
| 1853 | * For tools that provide their own replacement for malloc et al, the<br> |
| 1854 | option --redzone-size=<number> allows users to specify the size of<br> |
| 1855 | the padding blocks (redzones) added before and after each client<br> |
| 1856 | allocated block. Smaller redzones decrease the memory needed by<br> |
| 1857 | Valgrind. Bigger redzones increase the chance to detect blocks<br> |
| 1858 | overrun or underrun. Prior to this change, the redzone size was<br> |
| 1859 | hardwired to 16 bytes in Memcheck.<br> |
| 1860 | <br> |
| 1861 | * Memcheck:<br> |
| 1862 | <br> |
| 1863 | - The leak_check GDB server monitor command now can<br> |
| 1864 | control the maximum nr of loss records to output.<br> |
| 1865 | <br> |
| 1866 | - Reduction of memory use for applications allocating<br> |
| 1867 | many blocks and/or having many partially defined bytes.<br> |
| 1868 | <br> |
| 1869 | - Addition of GDB server monitor command 'block_list' that lists<br> |
| 1870 | the addresses/sizes of the blocks of a leak search loss record.<br> |
| 1871 | <br> |
| 1872 | - Addition of GDB server monitor command 'who_points_at' that lists<br> |
| 1873 | the locations pointing at a block.<br> |
| 1874 | <br> |
| 1875 | - If a redzone size > 0 is given, VALGRIND_MALLOCLIKE_BLOCK now will<br> |
| 1876 | detect an invalid access of these redzones, by marking them<br> |
| 1877 | noaccess. Similarly, if a redzone size is given for a memory<br> |
| 1878 | pool, VALGRIND_MEMPOOL_ALLOC will mark the redzones no access.<br> |
| 1879 | This still allows to find some bugs if the user has forgotten to<br> |
| 1880 | mark the pool superblock noaccess.<br> |
| 1881 | <br> |
| 1882 | - Performance of memory leak check has been improved, especially in<br> |
| 1883 | cases where there are many leaked blocks and/or many suppression<br> |
| 1884 | rules used to suppress leak reports.<br> |
| 1885 | <br> |
| 1886 | - Reduced noise (false positive) level on MacOSX 10.6/10.7, due to<br> |
| 1887 | more precise analysis, which is important for LLVM/Clang<br> |
| 1888 | generated code. This is at the cost of somewhat reduced<br> |
| 1889 | performance. Note there is no change to analysis precision or<br> |
| 1890 | costs on Linux targets.<br> |
| 1891 | <br> |
| 1892 | * DRD:<br> |
| 1893 | <br> |
| 1894 | - Added even more facilities that can help finding the cause of a data<br> |
| 1895 | race, namely the command-line option --ptrace-addr and the macro<br> |
| 1896 | DRD_STOP_TRACING_VAR(x). More information can be found in the manual.<br> |
| 1897 | <br> |
| 1898 | - Fixed a subtle bug that could cause false positive data race reports.<br> |
| 1899 | <br> |
| 1900 | * ==================== OTHER CHANGES ====================<br> |
| 1901 | <br> |
| 1902 | * The C++ demangler has been updated so as to work well with C++ <br> |
| 1903 | compiled by up to at least g++ 4.6.<br> |
| 1904 | <br> |
| 1905 | * Tool developers can make replacement/wrapping more flexible thanks<br> |
| 1906 | to the new option --soname-synonyms. This was reported above, but<br> |
| 1907 | in fact is very general and applies to all function<br> |
| 1908 | replacement/wrapping, not just to malloc-family functions.<br> |
| 1909 | <br> |
| 1910 | * Round-robin scheduling of threads can be selected, using the new<br> |
| 1911 | option --fair-sched= yes. Prior to this change, the pipe-based<br> |
| 1912 | thread serialisation mechanism (which is still the default) could<br> |
| 1913 | give very unfair scheduling. --fair-sched=yes improves<br> |
| 1914 | responsiveness of interactive multithreaded applications, and<br> |
| 1915 | improves repeatability of results from the thread checkers Helgrind<br> |
| 1916 | and DRD.<br> |
| 1917 | <br> |
| 1918 | * For tool developers: support to run Valgrind on Valgrind has been<br> |
| 1919 | improved. We can now routinely Valgrind on Helgrind or Memcheck.<br> |
| 1920 | <br> |
| 1921 | * gdbserver now shows the float shadow registers as integer<br> |
| 1922 | rather than float values, as the shadow values are mostly<br> |
| 1923 | used as bit patterns.<br> |
| 1924 | <br> |
| 1925 | * Increased limit for the --num-callers command line flag to 500.<br> |
| 1926 | <br> |
| 1927 | * Performance improvements for error matching when there are many<br> |
| 1928 | suppression records in use.<br> |
| 1929 | <br> |
| 1930 | * Improved support for DWARF4 debugging information (bug 284184).<br> |
| 1931 | <br> |
| 1932 | * Initial support for DWZ compressed Dwarf debug info.<br> |
| 1933 | <br> |
| 1934 | * Improved control over the IR optimiser's handling of the tradeoff<br> |
| 1935 | between performance and precision of exceptions. Specifically,<br> |
| 1936 | --vex-iropt-precise-memory-exns has been removed and replaced by<br> |
| 1937 | --vex-iropt-register-updates, with extended functionality. This<br> |
| 1938 | allows the Valgrind gdbserver to always show up to date register<br> |
| 1939 | values to GDB.<br> |
| 1940 | <br> |
| 1941 | * Modest performance gains through the use of translation chaining for<br> |
| 1942 | JIT-generated code.<br> |
| 1943 | <br> |
| 1944 | * ==================== FIXED BUGS ====================<br> |
| 1945 | <br> |
| 1946 | The following bugs have been fixed or resolved. Note that "n-i-bz"<br> |
| 1947 | stands for "not in bugzilla" -- that is, a bug that was reported to us<br> |
| 1948 | but never got a bugzilla entry. We encourage you to file bugs in<br> |
| 1949 | bugzilla (https://bugs.kde.org/enter_bug.cgi?product=valgrind) rather<br> |
| 1950 | than mailing the developers (or mailing lists) directly -- bugs that<br> |
| 1951 | are not entered into bugzilla tend to get forgotten about or ignored.<br> |
| 1952 | <br> |
| 1953 | To see details of a given bug, visit<br> |
| 1954 | https://bugs.kde.org/show_bug.cgi?id=XXXXXX<br> |
| 1955 | where XXXXXX is the bug number as listed below.<br> |
| 1956 | <br> |
| 1957 | 197914 Building valgrind from svn now requires automake-1.10<br> |
| 1958 | 203877 increase to 16Mb maximum allowed alignment for memalign et al<br> |
| 1959 | 219156 Handle statically linked malloc or other malloc lib (e.g. tcmalloc) <br> |
| 1960 | 247386 make perf does not run all performance tests<br> |
| 1961 | 270006 Valgrind scheduler unfair <br> |
| 1962 | 270777 Adding MIPS/Linux port to Valgrind<br> |
| 1963 | 270796 s390x: Removed broken support for the TS insn<br> |
| 1964 | 271438 Fix configure for proper SSE4.2 detection<br> |
| 1965 | 273114 s390x: Support TR, TRE, TROO, TROT, TRTO, and TRTT instructions<br> |
| 1966 | 273475 Add support for AVX instructions<br> |
| 1967 | 274078 improved configure logic for mpicc<br> |
| 1968 | 276993 fix mremap 'no thrash checks' <br> |
| 1969 | 278313 Fedora 15/x64: err read debug info with --read-var-info=yes flag<br> |
| 1970 | 281482 memcheck incorrect byte allocation count in realloc() for silly argument<br> |
| 1971 | 282230 group allocator for small fixed size, use it for MC_Chunk/SEc vbit<br> |
| 1972 | 283413 Fix wrong sanity check<br> |
| 1973 | 283671 Robustize alignment computation in LibVEX_Alloc<br> |
| 1974 | 283961 Adding support for some HCI IOCTLs<br> |
| 1975 | 284124 parse_type_DIE: confused by: DWARF 4<br> |
| 1976 | 284864 == 273475 (Add support for AVX instructions)<br> |
| 1977 | 285219 Too-restrictive constraints for Thumb2 "SP plus/minus register"<br> |
| 1978 | 285662 (MacOSX): Memcheck needs to replace memcpy/memmove<br> |
| 1979 | 285725 == 273475 (Add support for AVX instructions)<br> |
| 1980 | 286261 add wrapper for linux I2C_RDWR ioctl<br> |
| 1981 | 286270 vgpreload is not friendly to 64->32 bit execs, gives ld.so warnings<br> |
| 1982 | 286374 Running cachegrind with --branch-sim=yes on 64-bit PowerPC program fails<br> |
| 1983 | 286384 configure fails "checking for a supported version of gcc"<br> |
| 1984 | 286497 == 273475 (Add support for AVX instructions)<br> |
| 1985 | 286596 == 273475 (Add support for AVX instructions)<br> |
| 1986 | 286917 disInstr(arm): unhandled instruction: QADD (also QSUB)<br> |
| 1987 | 287175 ARM: scalar VFP fixed-point VCVT instructions not handled<br> |
| 1988 | 287260 Incorrect conditional jump or move depends on uninitialised value(s)<br> |
| 1989 | 287301 vex amd64->IR: 0x66 0xF 0x38 0x41 0xC0 0xB8 0x0 0x0 (PHMINPOSUW)<br> |
| 1990 | 287307 == 273475 (Add support for AVX instructions)<br> |
| 1991 | 287858 VG_(strerror): unknown error <br> |
| 1992 | 288298 (MacOSX) unhandled syscall shm_unlink<br> |
| 1993 | 288995 == 273475 (Add support for AVX instructions)<br> |
| 1994 | 289470 Loading of large Mach-O thin binaries fails.<br> |
| 1995 | 289656 == 273475 (Add support for AVX instructions)<br> |
| 1996 | 289699 vgdb connection in relay mode erroneously closed due to buffer overrun <br> |
| 1997 | 289823 == 293754 (PCMPxSTRx not implemented for 16-bit characters)<br> |
| 1998 | 289839 s390x: Provide support for unicode conversion instructions<br> |
| 1999 | 289939 monitor cmd 'leak_check' with details about leaked or reachable blocks<br> |
| 2000 | 290006 memcheck doesn't mark %xmm as initialized after "pcmpeqw %xmm %xmm"<br> |
| 2001 | 290655 Add support for AESKEYGENASSIST instruction <br> |
| 2002 | 290719 valgrind-3.7.0 fails with automake-1.11.2 due to"pkglibdir" usage<br> |
| 2003 | 290974 vgdb must align pages to VKI_SHMLBA (16KB) on ARM <br> |
| 2004 | 291253 ES register not initialised in valgrind simulation<br> |
| 2005 | 291568 Fix 3DNOW-related crashes with baseline x86_64 CPU (w patch)<br> |
| 2006 | 291865 s390x: Support the "Compare Double and Swap" family of instructions<br> |
| 2007 | 292300 == 273475 (Add support for AVX instructions)<br> |
| 2008 | 292430 unrecognized instruction in __intel_get_new_mem_ops_cpuid<br> |
| 2009 | 292493 == 273475 (Add support for AVX instructions)<br> |
| 2010 | 292626 Missing fcntl F_SETOWN_EX and F_GETOWN_EX support<br> |
| 2011 | 292627 Missing support for some SCSI ioctls<br> |
| 2012 | 292628 none/tests/x86/bug125959-x86.c triggers undefined behavior<br> |
| 2013 | 292841 == 273475 (Add support for AVX instructions)<br> |
| 2014 | 292993 implement the getcpu syscall on amd64-linux<br> |
| 2015 | 292995 Implement the “cross memory attach” syscalls introduced in Linux 3.2<br> |
| 2016 | 293088 Add some VEX sanity checks for ppc64 unhandled instructions<br> |
| 2017 | 293751 == 290655 (Add support for AESKEYGENASSIST instruction)<br> |
| 2018 | 293754 PCMPxSTRx not implemented for 16-bit characters<br> |
| 2019 | 293755 == 293754 (No tests for PCMPxSTRx on 16-bit characters)<br> |
| 2020 | 293808 CLFLUSH not supported by latest VEX for amd64<br> |
| 2021 | 294047 valgrind does not correctly emulate prlimit64(..., RLIMIT_NOFILE, ...)<br> |
| 2022 | 294048 MPSADBW instruction not implemented<br> |
| 2023 | 294055 regtest none/tests/shell fails when locale is not set to C<br> |
| 2024 | 294185 INT 0x44 (and others) not supported on x86 guest, but used by Jikes RVM<br> |
| 2025 | 294190 --vgdb-error=xxx can be out of sync with errors shown to the user<br> |
| 2026 | 294191 amd64: fnsave/frstor and 0x66 size prefixes on FP instructions<br> |
| 2027 | 294260 disInstr_AMD64: disInstr miscalculated next %rip<br> |
| 2028 | 294523 --partial-loads-ok=yes causes false negatives<br> |
| 2029 | 294617 vex amd64->IR: 0x66 0xF 0x3A 0xDF 0xD1 0x1 0xE8 0x6A<br> |
| 2030 | 294736 vex amd64->IR: 0x48 0xF 0xD7 0xD6 0x48 0x83<br> |
| 2031 | 294812 patch allowing to run (on x86 at least) helgrind/drd on tool.<br> |
| 2032 | 295089 can not annotate source for both helgrind and drd<br> |
| 2033 | 295221 POWER Processor decimal floating point instruction support missing<br> |
| 2034 | 295427 building for i386 with clang on darwin11 requires "-new_linker linker"<br> |
| 2035 | 295428 coregrind/m_main.c has incorrect x86 assembly for darwin<br> |
| 2036 | 295590 Helgrind: Assertion 'cvi->nWaiters > 0' failed<br> |
| 2037 | 295617 ARM - Add some missing syscalls<br> |
| 2038 | 295799 Missing \n with get_vbits in gdbserver when line is % 80 [...]<br> |
| 2039 | 296229 Linux user input device ioctls missing wrappers<br> |
| 2040 | 296318 ELF Debug info improvements (more than one rx/rw mapping)<br> |
| 2041 | 296422 Add translation chaining support<br> |
| 2042 | 296457 vex amd64->IR: 0x66 0xF 0x3A 0xDF 0xD1 0x1 0xE8 0x6A (dup of AES)<br> |
| 2043 | 296792 valgrind 3.7.0: add SIOCSHWTSTAMP (0x89B0) ioctl wrapper<br> |
| 2044 | 296983 Fix build issues on x86_64/ppc64 without 32-bit toolchains<br> |
| 2045 | 297078 gdbserver signal handling problems [..]<br> |
| 2046 | 297147 drd false positives on newly allocated memory<br> |
| 2047 | 297329 disallow decoding of IBM Power DFP insns on some machines<br> |
| 2048 | 297497 POWER Processor decimal floating point instruction support missing<br> |
| 2049 | 297701 Another alias for strncasecmp_l in libc-2.13.so<br> |
| 2050 | 297911 'invalid write' not reported when using APIs for custom mem allocators.<br> |
| 2051 | 297976 s390x: revisit EX implementation<br> |
| 2052 | 297991 Valgrind interferes with mmap()+ftell() <br> |
| 2053 | 297992 Support systems missing WIFCONTINUED (e.g. pre-2.6.10 Linux) <br> |
| 2054 | 297993 Fix compilation of valgrind with gcc -g3.<br> |
| 2055 | 298080 POWER Processor DFP support missing, part 3<br> |
| 2056 | 298227 == 273475 (Add support for AVX instructions)<br> |
| 2057 | 298335 == 273475 (Add support for AVX instructions)<br> |
| 2058 | 298354 Unhandled ARM Thumb instruction 0xEB0D 0x0585 (streq)<br> |
| 2059 | 298394 s390x: Don't bail out on an unknown machine model. [..]<br> |
| 2060 | 298421 accept4() syscall (366) support is missing for ARM<br> |
| 2061 | 298718 vex amd64->IR: 0xF 0xB1 0xCB 0x9C 0x8F 0x45<br> |
| 2062 | 298732 valgrind installation problem in ubuntu with kernel version 3.x<br> |
| 2063 | 298862 POWER Processor DFP instruction support missing, part 4<br> |
| 2064 | 298864 DWARF reader mis-parses DW_FORM_ref_addr<br> |
| 2065 | 298943 massif asserts with --pages-as-heap=yes when brk is changing [..]<br> |
| 2066 | 299053 Support DWARF4 DW_AT_high_pc constant form<br> |
| 2067 | 299104 == 273475 (Add support for AVX instructions)<br> |
| 2068 | 299316 Helgrind: hg_main.c:628 (map_threads_lookup): Assertion 'thr' failed.<br> |
| 2069 | 299629 dup3() syscall (358) support is missing for ARM<br> |
| 2070 | 299694 POWER Processor DFP instruction support missing, part 5<br> |
| 2071 | 299756 Ignore --free-fill for MEMPOOL_FREE and FREELIKE client requests<br> |
| 2072 | 299803 == 273475 (Add support for AVX instructions)<br> |
| 2073 | 299804 == 273475 (Add support for AVX instructions)<br> |
| 2074 | 299805 == 273475 (Add support for AVX instructions)<br> |
| 2075 | 300140 ARM - Missing (T1) SMMUL<br> |
| 2076 | 300195 == 296318 (ELF Debug info improvements (more than one rx/rw mapping))<br> |
| 2077 | 300389 Assertion `are_valid_hwcaps(VexArchAMD64, [..])' failed.<br> |
| 2078 | 300414 FCOM and FCOMP unimplemented for amd64 guest<br> |
| 2079 | 301204 infinite loop in canonicaliseSymtab with ifunc symbol<br> |
| 2080 | 301229 == 203877 (increase to 16Mb maximum allowed alignment for memalign etc)<br> |
| 2081 | 301265 add x86 support to Android build <br> |
| 2082 | 301984 configure script doesn't detect certain versions of clang<br> |
| 2083 | 302205 Fix compiler warnings for POWER VEX code and POWER test cases<br> |
| 2084 | 302287 Unhandled movbe instruction on Atom processors<br> |
| 2085 | 302370 PPC: fnmadd, fnmsub, fnmadds, fnmsubs insns always negate the result<br> |
| 2086 | 302536 Fix for the POWER Valgrind regression test: memcheck-ISA2.0.<br> |
| 2087 | 302578 Unrecognized isntruction 0xc5 0x32 0xc2 0xca 0x09 vcmpngess<br> |
| 2088 | 302656 == 273475 (Add support for AVX instructions)<br> |
| 2089 | 302709 valgrind for ARM needs extra tls support for android emulator [..]<br> |
| 2090 | 302827 add wrapper for CDROM_GET_CAPABILITY<br> |
| 2091 | 302901 Valgrind crashes with dwz optimized debuginfo<br> |
| 2092 | 302918 Enable testing of the vmaddfp and vnsubfp instructions in the testsuite<br> |
| 2093 | 303116 Add support for the POWER instruction popcntb<br> |
| 2094 | 303127 Power test suite fixes for frsqrte, vrefp, and vrsqrtefp instructions.<br> |
| 2095 | 303250 Assertion `instrs_in->arr_used <= 10000' failed w/ OpenSSL code<br> |
| 2096 | 303466 == 273475 (Add support for AVX instructions)<br> |
| 2097 | 303624 segmentation fault on Android 4.1 (e.g. on Galaxy Nexus OMAP) <br> |
| 2098 | 303963 strstr() function produces wrong results under valgrind callgrind<br> |
| 2099 | 304054 CALL_FN_xx macros need to enforce stack alignment<br> |
| 2100 | 304561 tee system call not supported<br> |
| 2101 | 715750 (MacOSX): Incorrect invalid-address errors near 0xFFFFxxxx (mozbug#)<br> |
| 2102 | n-i-bz Add missing gdbserver xml files for shadow registers for ppc32<br> |
| 2103 | n-i-bz Bypass gcc4.4/4.5 code gen bugs causing out of memory or asserts<br> |
| 2104 | n-i-bz Fix assert in gdbserver for watchpoints watching the same address<br> |
| 2105 | n-i-bz Fix false positive in sys_clone on amd64 when optional args [..]<br> |
| 2106 | n-i-bz s390x: Shadow registers can now be examined using vgdb<br> |
| 2107 | <br> |
| 2108 | (3.8.0-TEST3: 9 August 2012, vex r2465, valgrind r12865)<br> |
| 2109 | (3.8.0: 10 August 2012, vex r2465, valgrind r12866)<br> |
| 2110 | <br> |
| 2111 | <br> |
| 2112 | <br> |
| 2113 | Release 3.7.0 (5 November 2011)<br> |
| 2114 | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~<br> |
| 2115 | 3.7.0 is a feature release with many significant improvements and the<br> |
| 2116 | usual collection of bug fixes.<br> |
| 2117 | <br> |
| 2118 | This release supports X86/Linux, AMD64/Linux, ARM/Linux, PPC32/Linux,<br> |
| 2119 | PPC64/Linux, S390X/Linux, ARM/Android, X86/Darwin and AMD64/Darwin.<br> |
| 2120 | Support for recent distros and toolchain components (glibc 2.14, gcc<br> |
| 2121 | 4.6, MacOSX 10.7) has been added.<br> |
| 2122 | <br> |
| 2123 | * ================== PLATFORM CHANGES =================<br> |
| 2124 | <br> |
| 2125 | * Support for IBM z/Architecture (s390x) running Linux. Valgrind can<br> |
| 2126 | analyse 64-bit programs running on z/Architecture. Most user space<br> |
| 2127 | instructions up to and including z10 are supported. Valgrind has<br> |
| 2128 | been tested extensively on z9, z10, and z196 machines running SLES<br> |
| 2129 | 10/11, RedHat 5/6m, and Fedora. The Memcheck and Massif tools are<br> |
| 2130 | known to work well. Callgrind, Helgrind, and DRD work reasonably<br> |
| 2131 | well on z9 and later models. See README.s390 for more details.<br> |
| 2132 | <br> |
| 2133 | * Preliminary support for MacOSX 10.7 and XCode 4. Both 32- and<br> |
| 2134 | 64-bit processes are supported. Some complex threaded applications<br> |
| 2135 | (Firefox) are observed to hang when run as 32 bit applications,<br> |
| 2136 | whereas 64-bit versions run OK. The cause is unknown. Memcheck<br> |
| 2137 | will likely report some false errors. In general, expect some rough<br> |
| 2138 | spots. This release also supports MacOSX 10.6, but drops support<br> |
| 2139 | for 10.5.<br> |
| 2140 | <br> |
| 2141 | * Preliminary support for Android (on ARM). Valgrind can now run<br> |
| 2142 | large applications (eg, Firefox) on (eg) a Samsung Nexus S. See<br> |
| 2143 | README.android for more details, plus instructions on how to get<br> |
| 2144 | started.<br> |
| 2145 | <br> |
| 2146 | * Support for the IBM Power ISA 2.06 (Power7 instructions)<br> |
| 2147 | <br> |
| 2148 | * General correctness and performance improvements for ARM/Linux, and,<br> |
| 2149 | by extension, ARM/Android.<br> |
| 2150 | <br> |
| 2151 | * Further solidification of support for SSE 4.2 in 64-bit mode. AVX<br> |
| 2152 | instruction set support is under development but is not available in<br> |
| 2153 | this release.<br> |
| 2154 | <br> |
| 2155 | * Support for AIX5 has been removed.<br> |
| 2156 | <br> |
| 2157 | * ==================== TOOL CHANGES ====================<br> |
| 2158 | <br> |
| 2159 | * Memcheck: some incremental changes:<br> |
| 2160 | <br> |
| 2161 | - reduction of memory use in some circumstances<br> |
| 2162 | <br> |
| 2163 | - improved handling of freed memory, which in some circumstances <br> |
| 2164 | can cause detection of use-after-free that would previously have<br> |
| 2165 | been missed<br> |
| 2166 | <br> |
| 2167 | - fix of a longstanding bug that could cause false negatives (missed<br> |
| 2168 | errors) in programs doing vector saturated narrowing instructions.<br> |
| 2169 | <br> |
| 2170 | * Helgrind: performance improvements and major memory use reductions,<br> |
| 2171 | particularly for large, long running applications which perform many<br> |
| 2172 | synchronisation (lock, unlock, etc) events. Plus many smaller<br> |
| 2173 | changes:<br> |
| 2174 | <br> |
| 2175 | - display of locksets for both threads involved in a race<br> |
| 2176 | <br> |
| 2177 | - general improvements in formatting/clarity of error messages<br> |
| 2178 | <br> |
| 2179 | - addition of facilities and documentation regarding annotation<br> |
| 2180 | of thread safe reference counted C++ classes<br> |
| 2181 | <br> |
| 2182 | - new flag --check-stack-refs=no|yes [yes], to disable race checking<br> |
| 2183 | on thread stacks (a performance hack)<br> |
| 2184 | <br> |
| 2185 | - new flag --free-is-write=no|yes [no], to enable detection of races<br> |
| 2186 | where one thread accesses heap memory but another one frees it,<br> |
| 2187 | without any coordinating synchronisation event<br> |
| 2188 | <br> |
| 2189 | * DRD: enabled XML output; added support for delayed thread deletion<br> |
| 2190 | in order to detect races that occur close to the end of a thread<br> |
| 2191 | (--join-list-vol); fixed a memory leak triggered by repeated client<br> |
| 2192 | memory allocatation and deallocation; improved Darwin support.<br> |
| 2193 | <br> |
| 2194 | * exp-ptrcheck: this tool has been renamed to exp-sgcheck<br> |
| 2195 | <br> |
| 2196 | * exp-sgcheck: this tool has been reduced in scope so as to improve<br> |
| 2197 | performance and remove checking that Memcheck does better.<br> |
| 2198 | Specifically, the ability to check for overruns for stack and global<br> |
| 2199 | arrays is unchanged, but the ability to check for overruns of heap<br> |
| 2200 | blocks has been removed. The tool has accordingly been renamed to<br> |
| 2201 | exp-sgcheck ("Stack and Global Array Checking").<br> |
| 2202 | <br> |
| 2203 | * ==================== OTHER CHANGES ====================<br> |
| 2204 | <br> |
| 2205 | * GDB server: Valgrind now has an embedded GDB server. That means it<br> |
| 2206 | is possible to control a Valgrind run from GDB, doing all the usual<br> |
| 2207 | things that GDB can do (single stepping, breakpoints, examining<br> |
| 2208 | data, etc). Tool-specific functionality is also available. For<br> |
| 2209 | example, it is possible to query the definedness state of variables<br> |
| 2210 | or memory from within GDB when running Memcheck; arbitrarily large<br> |
| 2211 | memory watchpoints are supported, etc. To use the GDB server, start<br> |
| 2212 | Valgrind with the flag --vgdb-error=0 and follow the on-screen<br> |
| 2213 | instructions.<br> |
| 2214 | <br> |
| 2215 | * Improved support for unfriendly self-modifying code: a new option<br> |
| 2216 | --smc-check=all-non-file is available. This adds the relevant<br> |
| 2217 | consistency checks only to code that originates in non-file-backed<br> |
| 2218 | mappings. In effect this confines the consistency checking only to<br> |
| 2219 | code that is or might be JIT generated, and avoids checks on code<br> |
| 2220 | that must have been compiled ahead of time. This significantly<br> |
| 2221 | improves performance on applications that generate code at run time.<br> |
| 2222 | <br> |
| 2223 | * It is now possible to build a working Valgrind using Clang-2.9 on<br> |
| 2224 | Linux.<br> |
| 2225 | <br> |
| 2226 | * new client requests VALGRIND_{DISABLE,ENABLE}_ERROR_REPORTING.<br> |
| 2227 | These enable and disable error reporting on a per-thread, and<br> |
| 2228 | nestable, basis. This is useful for hiding errors in particularly<br> |
| 2229 | troublesome pieces of code. The MPI wrapper library (libmpiwrap.c)<br> |
| 2230 | now uses this facility.<br> |
| 2231 | <br> |
| 2232 | * Added the --mod-funcname option to cg_diff.<br> |
| 2233 | <br> |
| 2234 | * ==================== FIXED BUGS ====================<br> |
| 2235 | <br> |
| 2236 | The following bugs have been fixed or resolved. Note that "n-i-bz"<br> |
| 2237 | stands for "not in bugzilla" -- that is, a bug that was reported to us<br> |
| 2238 | but never got a bugzilla entry. We encourage you to file bugs in<br> |
| 2239 | bugzilla (http://bugs.kde.org/enter_valgrind_bug.cgi) rather than<br> |
| 2240 | mailing the developers (or mailing lists) directly -- bugs that are<br> |
| 2241 | not entered into bugzilla tend to get forgotten about or ignored.<br> |
| 2242 | <br> |
| 2243 | To see details of a given bug, visit<br> |
| 2244 | https://bugs.kde.org/show_bug.cgi?id=XXXXXX<br> |
| 2245 | where XXXXXX is the bug number as listed below.<br> |
| 2246 | <br> |
| 2247 | 79311 malloc silly arg warning does not give stack trace<br> |
| 2248 | 210935 port valgrind.h (not valgrind) to win32 to support client requests<br> |
| 2249 | 214223 valgrind SIGSEGV on startup gcc 4.4.1 ppc32 (G4) Ubuntu 9.10<br> |
| 2250 | 243404 Port to zSeries<br> |
| 2251 | 243935 Helgrind: incorrect handling of ANNOTATE_HAPPENS_BEFORE()/AFTER()<br> |
| 2252 | 247223 non-x86: Suppress warning: 'regparm' attribute directive ignored<br> |
| 2253 | 250101 huge "free" memory usage due to m_mallocfree.c fragmentation<br> |
| 2254 | 253206 Some fixes for the faultstatus testcase<br> |
| 2255 | 255223 capget testcase fails when running as root<br> |
| 2256 | 256703 xlc_dbl_u32.c testcase broken<br> |
| 2257 | 256726 Helgrind tests have broken inline asm <br> |
| 2258 | 259977 == 214223 (Valgrind segfaults doing __builtin_longjmp)<br> |
| 2259 | 264800 testcase compile failure on zseries<br> |
| 2260 | 265762 make public VEX headers compilable by G++ 3.x<br> |
| 2261 | 265771 assertion in jumps.c (r11523) fails with glibc-2.3<br> |
| 2262 | 266753 configure script does not give the user the option to not use QtCore<br> |
| 2263 | 266931 gen_insn_test.pl is broken<br> |
| 2264 | 266961 ld-linux.so.2 i?86-linux strlen issues<br> |
| 2265 | 266990 setns instruction causes false positive<br> |
| 2266 | 267020 Make directory for temporary files configurable at run-time.<br> |
| 2267 | 267342 == 267997 (segmentation fault on Mac OS 10.6)<br> |
| 2268 | 267383 Assertion 'vgPlain_strlen(dir) + vgPlain_strlen(file) + 1 < 256' failed<br> |
| 2269 | 267413 Assertion 'DRD_(g_threadinfo)[tid].synchr_nesting >= 1' failed.<br> |
| 2270 | 267488 regtest: darwin support for 64-bit build<br> |
| 2271 | 267552 SIGSEGV (misaligned_stack_error) with DRD, but not with other tools<br> |
| 2272 | 267630 Add support for IBM Power ISA 2.06 -- stage 1<br> |
| 2273 | 267769 == 267997 (Darwin: memcheck triggers segmentation fault)<br> |
| 2274 | 267819 Add client request for informing the core about reallocation<br> |
| 2275 | 267925 laog data structure quadratic for a single sequence of lock<br> |
| 2276 | 267968 drd: (vgDrd_thread_set_joinable): Assertion '0 <= (int)tid ..' failed<br> |
| 2277 | 267997 MacOSX: 64-bit V segfaults on launch when built with Xcode 4.0.1<br> |
| 2278 | 268513 missed optimizations in fold_Expr<br> |
| 2279 | 268619 s390x: fpr - gpr transfer facility <br> |
| 2280 | 268620 s390x: reconsider "long displacement" requirement <br> |
| 2281 | 268621 s390x: improve IR generation for XC<br> |
| 2282 | 268715 s390x: FLOGR is not universally available<br> |
| 2283 | 268792 == 267997 (valgrind seg faults on startup when compiled with Xcode 4)<br> |
| 2284 | 268930 s390x: MHY is not universally available<br> |
| 2285 | 269078 arm->IR: unhandled instruction SUB (SP minus immediate/register) <br> |
| 2286 | 269079 Support ptrace system call on ARM<br> |
| 2287 | 269144 missing "Bad option" error message<br> |
| 2288 | 269209 conditional load and store facility (z196)<br> |
| 2289 | 269354 Shift by zero on x86 can incorrectly clobber CC_NDEP<br> |
| 2290 | 269641 == 267997 (valgrind segfaults immediately (segmentation fault))<br> |
| 2291 | 269736 s390x: minor code generation tweaks<br> |
| 2292 | 269778 == 272986 (valgrind.h: swap roles of VALGRIND_DO_CLIENT_REQUEST() ..)<br> |
| 2293 | 269863 s390x: remove unused function parameters<br> |
| 2294 | 269864 s390x: tweak s390_emit_load_cc <br> |
| 2295 | 269884 == 250101 (overhead for huge blocks exhausts space too soon)<br> |
| 2296 | 270082 s390x: Make sure to point the PSW address to the next address on SIGILL<br> |
| 2297 | 270115 s390x: rewrite some testcases<br> |
| 2298 | 270309 == 267997 (valgrind crash on startup)<br> |
| 2299 | 270320 add support for Linux FIOQSIZE ioctl() call<br> |
| 2300 | 270326 segfault while trying to sanitize the environment passed to execle<br> |
| 2301 | 270794 IBM POWER7 support patch causes regression in none/tests<br> |
| 2302 | 270851 IBM POWER7 fcfidus instruction causes memcheck to fail<br> |
| 2303 | 270856 IBM POWER7 xsnmaddadp instruction causes memcheck to fail on 32bit app <br> |
| 2304 | 270925 hyper-optimized strspn() in /lib64/libc-2.13.so needs fix<br> |
| 2305 | 270959 s390x: invalid use of R0 as base register<br> |
| 2306 | 271042 VSX configure check fails when it should not <br> |
| 2307 | 271043 Valgrind build fails with assembler error on ppc64 with binutils 2.21 <br> |
| 2308 | 271259 s390x: fix code confusion <br> |
| 2309 | 271337 == 267997 (Valgrind segfaults on MacOS X)<br> |
| 2310 | 271385 s390x: Implement Ist_MBE <br> |
| 2311 | 271501 s390x: misc cleanups <br> |
| 2312 | 271504 s390x: promote likely and unlikely <br> |
| 2313 | 271579 ppc: using wrong enum type <br> |
| 2314 | 271615 unhandled instruction "popcnt" (arch=amd10h) <br> |
| 2315 | 271730 Fix bug when checking ioctls: duplicate check <br> |
| 2316 | 271776 s390x: provide STFLE instruction support <br> |
| 2317 | 271779 s390x: provide clock instructions like STCK <br> |
| 2318 | 271799 Darwin: ioctls without an arg report a memory error <br> |
| 2319 | 271820 arm: fix type confusion <br> |
| 2320 | 271917 pthread_cond_timedwait failure leads to not-locked false positive <br> |
| 2321 | 272067 s390x: fix DISP20 macro <br> |
| 2322 | 272615 A typo in debug output in mc_leakcheck.c<br> |
| 2323 | 272661 callgrind_annotate chokes when run from paths containing regex chars<br> |
| 2324 | 272893 amd64->IR: 0x66 0xF 0x38 0x2B 0xC1 0x66 0xF 0x7F == (closed as dup)<br> |
| 2325 | 272955 Unhandled syscall error for pwrite64 on ppc64 arch <br> |
| 2326 | 272967 make documentation build-system more robust <br> |
| 2327 | 272986 Fix gcc-4.6 warnings with valgrind.h<br> |
| 2328 | 273318 amd64->IR: 0x66 0xF 0x3A 0x61 0xC1 0x38 (missing PCMPxSTRx case)<br> |
| 2329 | 273318 unhandled PCMPxSTRx case: vex amd64->IR: 0x66 0xF 0x3A 0x61 0xC1 0x38 <br> |
| 2330 | 273431 valgrind segfaults in evalCfiExpr (debuginfo.c:2039)<br> |
| 2331 | 273465 Callgrind: jumps.c:164 (new_jcc): Assertion '(0 <= jmp) && ...'<br> |
| 2332 | 273536 Build error: multiple definition of `vgDrd_pthread_cond_initializer'<br> |
| 2333 | 273640 ppc64-linux: unhandled syscalls setresuid(164) and setresgid(169)<br> |
| 2334 | 273729 == 283000 (Illegal opcode for SSE2 "roundsd" instruction)<br> |
| 2335 | 273778 exp-ptrcheck: unhandled sysno == 259<br> |
| 2336 | 274089 exp-ptrcheck: unhandled sysno == 208<br> |
| 2337 | 274378 s390x: Various dispatcher tweaks<br> |
| 2338 | 274447 WARNING: unhandled syscall: 340<br> |
| 2339 | 274776 amd64->IR: 0x66 0xF 0x38 0x2B 0xC5 0x66<br> |
| 2340 | 274784 == 267997 (valgrind ls -l results in Segmentation Fault)<br> |
| 2341 | 274926 valgrind does not build against linux-3<br> |
| 2342 | 275148 configure FAIL with glibc-2.14<br> |
| 2343 | 275151 Fedora 15 / glibc-2.14 'make regtest' FAIL<br> |
| 2344 | 275168 Make Valgrind work for MacOSX 10.7 Lion<br> |
| 2345 | 275212 == 275284 (lots of false positives from __memcpy_ssse3_back et al)<br> |
| 2346 | 275278 valgrind does not build on Linux kernel 3.0.* due to silly<br> |
| 2347 | 275284 Valgrind memcpy/memmove redirection stopped working in glibc 2.14/x86_64<br> |
| 2348 | 275308 Fix implementation for ppc64 fres instruc<br> |
| 2349 | 275339 s390x: fix testcase compile warnings<br> |
| 2350 | 275517 s390x: Provide support for CKSM instruction<br> |
| 2351 | 275710 s390x: get rid of redundant address mode calculation<br> |
| 2352 | 275815 == 247894 (Valgrind doesn't know about Linux readahead(2) syscall)<br> |
| 2353 | 275852 == 250101 (valgrind uses all swap space and is killed)<br> |
| 2354 | 276784 Add support for IBM Power ISA 2.06 -- stage 3<br> |
| 2355 | 276987 gdbsrv: fix tests following recent commits<br> |
| 2356 | 277045 Valgrind crashes with unhandled DW_OP_ opcode 0x2a<br> |
| 2357 | 277199 The test_isa_2_06_part1.c in none/tests/ppc64 should be a symlink<br> |
| 2358 | 277471 Unhandled syscall: 340<br> |
| 2359 | 277610 valgrind crashes in VG_(lseek)(core_fd, phdrs[idx].p_offset, ...)<br> |
| 2360 | 277653 ARM: support Thumb2 PLD instruction<br> |
| 2361 | 277663 ARM: NEON float VMUL by scalar incorrect<br> |
| 2362 | 277689 ARM: tests for VSTn with register post-index are broken<br> |
| 2363 | 277694 ARM: BLX LR instruction broken in ARM mode<br> |
| 2364 | 277780 ARM: VMOV.F32 (immediate) instruction is broken<br> |
| 2365 | 278057 fuse filesystem syscall deadlocks<br> |
| 2366 | 278078 Unimplemented syscall 280 on ppc32<br> |
| 2367 | 278349 F_GETPIPE_SZ and F_SETPIPE_SZ Linux fcntl commands<br> |
| 2368 | 278454 VALGRIND_STACK_DEREGISTER has wrong output type<br> |
| 2369 | 278502 == 275284 (Valgrind confuses memcpy() and memmove())<br> |
| 2370 | 278892 gdbsrv: factorize gdb version handling, fix doc and typos<br> |
| 2371 | 279027 Support for MVCL and CLCL instruction<br> |
| 2372 | 279027 s390x: Provide support for CLCL and MVCL instructions<br> |
| 2373 | 279062 Remove a redundant check in the insn selector for ppc.<br> |
| 2374 | 279071 JDK creates PTEST with redundant REX.W prefix<br> |
| 2375 | 279212 gdbsrv: add monitor cmd v.info scheduler.<br> |
| 2376 | 279378 exp-ptrcheck: the 'impossible' happened on mkfifo call<br> |
| 2377 | 279698 memcheck discards valid-bits for packuswb<br> |
| 2378 | 279795 memcheck reports uninitialised values for mincore on amd64<br> |
| 2379 | 279994 Add support for IBM Power ISA 2.06 -- stage 3<br> |
| 2380 | 280083 mempolicy syscall check errors<br> |
| 2381 | 280290 vex amd64->IR: 0x66 0xF 0x38 0x28 0xC1 0x66 0xF 0x6F<br> |
| 2382 | 280710 s390x: config files for nightly builds<br> |
| 2383 | 280757 /tmp dir still used by valgrind even if TMPDIR is specified<br> |
| 2384 | 280965 Valgrind breaks fcntl locks when program does mmap<br> |
| 2385 | 281138 WARNING: unhandled syscall: 340<br> |
| 2386 | 281241 == 275168 (valgrind useless on Macos 10.7.1 Lion)<br> |
| 2387 | 281304 == 275168 (Darwin: dyld "cannot load inserted library")<br> |
| 2388 | 281305 == 275168 (unhandled syscall: unix:357 on Darwin 11.1)<br> |
| 2389 | 281468 s390x: handle do_clone and gcc clones in call traces<br> |
| 2390 | 281488 ARM: VFP register corruption<br> |
| 2391 | 281828 == 275284 (false memmove warning: "Source and destination overlap")<br> |
| 2392 | 281883 s390x: Fix system call wrapper for "clone".<br> |
| 2393 | 282105 generalise 'reclaimSuperBlock' to also reclaim splittable superblock<br> |
| 2394 | 282112 Unhandled instruction bytes: 0xDE 0xD9 0x9B 0xDF (fcompp)<br> |
| 2395 | 282238 SLES10: make check fails<br> |
| 2396 | 282979 strcasestr needs replacement with recent(>=2.12) glibc<br> |
| 2397 | 283000 vex amd64->IR: 0x66 0xF 0x3A 0xA 0xC0 0x9 0xF3 0xF<br> |
| 2398 | 283243 Regression in ppc64 memcheck tests<br> |
| 2399 | 283325 == 267997 (Darwin: V segfaults on startup when built with Xcode 4.0)<br> |
| 2400 | 283427 re-connect epoll_pwait syscall on ARM linux<br> |
| 2401 | 283600 gdbsrv: android: port vgdb.c<br> |
| 2402 | 283709 none/tests/faultstatus needs to account for page size<br> |
| 2403 | 284305 filter_gdb needs enhancement to work on ppc64<br> |
| 2404 | 284384 clang 3.1 -Wunused-value warnings in valgrind.h, memcheck.h<br> |
| 2405 | 284472 Thumb2 ROR.W encoding T2 not implemented<br> |
| 2406 | 284621 XML-escape process command line in XML output<br> |
| 2407 | n-i-bz cachegrind/callgrind: handle CPUID information for Core iX Intel CPUs<br> |
| 2408 | that have non-power-of-2 sizes (also AMDs)<br> |
| 2409 | n-i-bz don't be spooked by libraries mashed by elfhack<br> |
| 2410 | n-i-bz don't be spooked by libxul.so linked with gold<br> |
| 2411 | n-i-bz improved checking for VALGRIND_CHECK_MEM_IS_DEFINED<br> |
| 2412 | <br> |
| 2413 | (3.7.0-TEST1: 27 October 2011, vex r2228, valgrind r12245)<br> |
| 2414 | (3.7.0.RC1: 1 November 2011, vex r2231, valgrind r12257)<br> |
| 2415 | (3.7.0: 5 November 2011, vex r2231, valgrind r12258)<br> |
| 2416 | <br> |
| 2417 | <br> |
| 2418 | <br> |
| 2419 | Release 3.6.1 (16 February 2011)<br> |
| 2420 | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~<br> |
| 2421 | 3.6.1 is a bug fix release. It adds support for some SSE4<br> |
| 2422 | instructions that were omitted in 3.6.0 due to lack of time. Initial<br> |
| 2423 | support for glibc-2.13 has been added. A number of bugs causing<br> |
| 2424 | crashing or assertion failures have been fixed.<br> |
| 2425 | <br> |
| 2426 | The following bugs have been fixed or resolved. Note that "n-i-bz"<br> |
| 2427 | stands for "not in bugzilla" -- that is, a bug that was reported to us<br> |
| 2428 | but never got a bugzilla entry. We encourage you to file bugs in<br> |
| 2429 | bugzilla (http://bugs.kde.org/enter_valgrind_bug.cgi) rather than<br> |
| 2430 | mailing the developers (or mailing lists) directly -- bugs that are<br> |
| 2431 | not entered into bugzilla tend to get forgotten about or ignored.<br> |
| 2432 | <br> |
| 2433 | To see details of a given bug, visit<br> |
| 2434 | https://bugs.kde.org/show_bug.cgi?id=XXXXXX<br> |
| 2435 | where XXXXXX is the bug number as listed below.<br> |
| 2436 | <br> |
| 2437 | 188572 Valgrind on Mac should suppress setenv() mem leak<br> |
| 2438 | 194402 vex amd64->IR: 0x48 0xF 0xAE 0x4 (proper FX{SAVE,RSTOR} support)<br> |
| 2439 | 210481 vex amd64->IR: Assertion `sz == 2 || sz == 4' failed (REX.W POPQ)<br> |
| 2440 | 246152 callgrind internal error after pthread_cancel on 32 Bit Linux<br> |
| 2441 | 250038 ppc64: Altivec LVSR and LVSL instructions fail their regtest<br> |
| 2442 | 254420 memory pool tracking broken <br> |
| 2443 | 254957 Test code failing to compile due to changes in memcheck.h<br> |
| 2444 | 255009 helgrind/drd: crash on chmod with invalid parameter<br> |
| 2445 | 255130 readdwarf3.c parse_type_DIE confused by GNAT Ada types<br> |
| 2446 | 255355 helgrind/drd: crash on threaded programs doing fork<br> |
| 2447 | 255358 == 255355<br> |
| 2448 | 255418 (SSE4.x) rint call compiled with ICC<br> |
| 2449 | 255822 --gen-suppressions can create invalid files: "too many callers [...]"<br> |
| 2450 | 255888 closing valgrindoutput tag outputted to log-stream on error<br> |
| 2451 | 255963 (SSE4.x) vex amd64->IR: 0x66 0xF 0x3A 0x9 0xDB 0x0 (ROUNDPD)<br> |
| 2452 | 255966 Slowness when using mempool annotations<br> |
| 2453 | 256387 vex x86->IR: 0xD4 0xA 0x2 0x7 (AAD and AAM)<br> |
| 2454 | 256600 super-optimized strcasecmp() false positive<br> |
| 2455 | 256669 vex amd64->IR: Unhandled LOOPNEL insn on amd64<br> |
| 2456 | 256968 (SSE4.x) vex amd64->IR: 0x66 0xF 0x38 0x10 0xD3 0x66 (BLENDVPx)<br> |
| 2457 | 257011 (SSE4.x) vex amd64->IR: 0x66 0xF 0x3A 0xE 0xFD 0xA0 (PBLENDW)<br> |
| 2458 | 257063 (SSE4.x) vex amd64->IR: 0x66 0xF 0x3A 0x8 0xC0 0x0 (ROUNDPS)<br> |
| 2459 | 257276 Missing case in memcheck --track-origins=yes<br> |
| 2460 | 258870 (SSE4.x) Add support for EXTRACTPS SSE 4.1 instruction<br> |
| 2461 | 261966 (SSE4.x) support for CRC32B and CRC32Q is lacking (also CRC32{W,L})<br> |
| 2462 | 262985 VEX regression in valgrind 3.6.0 in handling PowerPC VMX<br> |
| 2463 | 262995 (SSE4.x) crash when trying to valgrind gcc-snapshot (PCMPxSTRx $0)<br> |
| 2464 | 263099 callgrind_annotate counts Ir improperly [...]<br> |
| 2465 | 263877 undefined coprocessor instruction on ARMv7<br> |
| 2466 | 265964 configure FAIL with glibc-2.13<br> |
| 2467 | n-i-bz Fix compile error w/ icc-12.x in guest_arm_toIR.c<br> |
| 2468 | n-i-bz Docs: fix bogus descriptions for VALGRIND_CREATE_BLOCK et al<br> |
| 2469 | n-i-bz Massif: don't assert on shmat() with --pages-as-heap=yes<br> |
| 2470 | n-i-bz Bug fixes and major speedups for the exp-DHAT space profiler<br> |
| 2471 | n-i-bz DRD: disable --free-is-write due to implementation difficulties<br> |
| 2472 | <br> |
| 2473 | (3.6.1: 16 February 2011, vex r2103, valgrind r11561).<br> |
| 2474 | <br> |
| 2475 | <br> |
| 2476 | <br> |
| 2477 | Release 3.6.0 (21 October 2010)<br> |
| 2478 | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~<br> |
| 2479 | 3.6.0 is a feature release with many significant improvements and the<br> |
| 2480 | usual collection of bug fixes.<br> |
| 2481 | <br> |
| 2482 | This release supports X86/Linux, AMD64/Linux, ARM/Linux, PPC32/Linux,<br> |
| 2483 | PPC64/Linux, X86/Darwin and AMD64/Darwin. Support for recent distros<br> |
| 2484 | and toolchain components (glibc 2.12, gcc 4.5, OSX 10.6) has been added.<br> |
| 2485 | <br> |
| 2486 | -------------------------<br> |
| 2487 | <br> |
| 2488 | Here are some highlights. Details are shown further down:<br> |
| 2489 | <br> |
| 2490 | * Support for ARM/Linux.<br> |
| 2491 | <br> |
| 2492 | * Support for recent Linux distros: Ubuntu 10.10 and Fedora 14.<br> |
| 2493 | <br> |
| 2494 | * Support for Mac OS X 10.6, both 32- and 64-bit executables.<br> |
| 2495 | <br> |
| 2496 | * Support for the SSE4.2 instruction set.<br> |
| 2497 | <br> |
| 2498 | * Enhancements to the Callgrind profiler, including the ability to<br> |
| 2499 | handle CPUs with three levels of cache.<br> |
| 2500 | <br> |
| 2501 | * A new experimental heap profiler, DHAT.<br> |
| 2502 | <br> |
| 2503 | * A huge number of bug fixes and small enhancements.<br> |
| 2504 | <br> |
| 2505 | -------------------------<br> |
| 2506 | <br> |
| 2507 | Here are details of the above changes, together with descriptions of<br> |
| 2508 | many other changes, and a list of fixed bugs.<br> |
| 2509 | <br> |
| 2510 | * ================== PLATFORM CHANGES =================<br> |
| 2511 | <br> |
| 2512 | * Support for ARM/Linux. Valgrind now runs on ARMv7 capable CPUs<br> |
| 2513 | running Linux. It is known to work on Ubuntu 10.04, Ubuntu 10.10,<br> |
| 2514 | and Maemo 5, so you can run Valgrind on your Nokia N900 if you want.<br> |
| 2515 | <br> |
| 2516 | This requires a CPU capable of running the ARMv7-A instruction set<br> |
| 2517 | (Cortex A5, A8 and A9). Valgrind provides fairly complete coverage<br> |
| 2518 | of the user space instruction set, including ARM and Thumb integer<br> |
| 2519 | code, VFPv3, NEON and V6 media instructions. The Memcheck,<br> |
| 2520 | Cachegrind and Massif tools work properly; other tools work to<br> |
| 2521 | varying degrees.<br> |
| 2522 | <br> |
| 2523 | * Support for recent Linux distros (Ubuntu 10.10 and Fedora 14), along<br> |
| 2524 | with support for recent releases of the underlying toolchain<br> |
| 2525 | components, notably gcc-4.5 and glibc-2.12.<br> |
| 2526 | <br> |
| 2527 | * Support for Mac OS X 10.6, both 32- and 64-bit executables. 64-bit<br> |
| 2528 | support also works much better on OS X 10.5, and is as solid as<br> |
| 2529 | 32-bit support now.<br> |
| 2530 | <br> |
| 2531 | * Support for the SSE4.2 instruction set. SSE4.2 is supported in<br> |
| 2532 | 64-bit mode. In 32-bit mode, support is only available up to and<br> |
| 2533 | including SSSE3. Some exceptions: SSE4.2 AES instructions are not<br> |
| 2534 | supported in 64-bit mode, and 32-bit mode does in fact support the<br> |
| 2535 | bare minimum SSE4 instructions to needed to run programs on Mac OS X<br> |
| 2536 | 10.6 on 32-bit targets.<br> |
| 2537 | <br> |
| 2538 | * Support for IBM POWER6 cpus has been improved. The Power ISA up to<br> |
| 2539 | and including version 2.05 is supported.<br> |
| 2540 | <br> |
| 2541 | * ==================== TOOL CHANGES ====================<br> |
| 2542 | <br> |
| 2543 | * Cachegrind has a new processing script, cg_diff, which finds the<br> |
| 2544 | difference between two profiles. It's very useful for evaluating<br> |
| 2545 | the performance effects of a change in a program.<br> |
| 2546 | <br> |
| 2547 | Related to this change, the meaning of cg_annotate's (rarely-used)<br> |
| 2548 | --threshold option has changed; this is unlikely to affect many<br> |
| 2549 | people, if you do use it please see the user manual for details.<br> |
| 2550 | <br> |
| 2551 | * Callgrind now can do branch prediction simulation, similar to<br> |
| 2552 | Cachegrind. In addition, it optionally can count the number of<br> |
| 2553 | executed global bus events. Both can be used for a better<br> |
| 2554 | approximation of a "Cycle Estimation" as derived event (you need to<br> |
| 2555 | update the event formula in KCachegrind yourself).<br> |
| 2556 | <br> |
| 2557 | * Cachegrind and Callgrind now refer to the LL (last-level) cache<br> |
| 2558 | rather than the L2 cache. This is to accommodate machines with<br> |
| 2559 | three levels of caches -- if Cachegrind/Callgrind auto-detects the<br> |
| 2560 | cache configuration of such a machine it will run the simulation as<br> |
| 2561 | if the L2 cache isn't present. This means the results are less<br> |
| 2562 | likely to match the true result for the machine, but<br> |
| 2563 | Cachegrind/Callgrind's results are already only approximate, and<br> |
| 2564 | should not be considered authoritative. The results are still<br> |
| 2565 | useful for giving a general idea about a program's locality.<br> |
| 2566 | <br> |
| 2567 | * Massif has a new option, --pages-as-heap, which is disabled by<br> |
| 2568 | default. When enabled, instead of tracking allocations at the level<br> |
| 2569 | of heap blocks (as allocated with malloc/new/new[]), it instead<br> |
| 2570 | tracks memory allocations at the level of memory pages (as mapped by<br> |
| 2571 | mmap, brk, etc). Each mapped page is treated as its own block.<br> |
| 2572 | Interpreting the page-level output is harder than the heap-level<br> |
| 2573 | output, but this option is useful if you want to account for every<br> |
| 2574 | byte of memory used by a program.<br> |
| 2575 | <br> |
| 2576 | * DRD has two new command-line options: --free-is-write and<br> |
| 2577 | --trace-alloc. The former allows to detect reading from already freed<br> |
| 2578 | memory, and the latter allows tracing of all memory allocations and<br> |
| 2579 | deallocations.<br> |
| 2580 | <br> |
| 2581 | * DRD has several new annotations. Custom barrier implementations can<br> |
| 2582 | now be annotated, as well as benign races on static variables.<br> |
| 2583 | <br> |
| 2584 | * DRD's happens before / happens after annotations have been made more<br> |
| 2585 | powerful, so that they can now also be used to annotate e.g. a smart<br> |
| 2586 | pointer implementation.<br> |
| 2587 | <br> |
| 2588 | * Helgrind's annotation set has also been drastically improved, so as<br> |
| 2589 | to provide to users a general set of annotations to describe locks,<br> |
| 2590 | semaphores, barriers and condition variables. Annotations to<br> |
| 2591 | describe thread-safe reference counted heap objects have also been<br> |
| 2592 | added.<br> |
| 2593 | <br> |
| 2594 | * Memcheck has a new command-line option, --show-possibly-lost, which<br> |
| 2595 | is enabled by default. When disabled, the leak detector will not<br> |
| 2596 | show possibly-lost blocks.<br> |
| 2597 | <br> |
| 2598 | * A new experimental heap profiler, DHAT (Dynamic Heap Analysis Tool),<br> |
| 2599 | has been added. DHAT keeps track of allocated heap blocks, and also<br> |
| 2600 | inspects every memory reference to see which block (if any) is being<br> |
| 2601 | accessed. This gives a lot of insight into block lifetimes,<br> |
| 2602 | utilisation, turnover, liveness, and the location of hot and cold<br> |
| 2603 | fields. You can use DHAT to do hot-field profiling.<br> |
| 2604 | <br> |
| 2605 | * ==================== OTHER CHANGES ====================<br> |
| 2606 | <br> |
| 2607 | * Improved support for unfriendly self-modifying code: the extra<br> |
| 2608 | overhead incurred by --smc-check=all has been reduced by<br> |
| 2609 | approximately a factor of 5 as compared with 3.5.0.<br> |
| 2610 | <br> |
| 2611 | * Ability to show directory names for source files in error messages.<br> |
| 2612 | This is combined with a flexible mechanism for specifying which<br> |
| 2613 | parts of the paths should be shown. This is enabled by the new flag<br> |
| 2614 | --fullpath-after.<br> |
| 2615 | <br> |
| 2616 | * A new flag, --require-text-symbol, which will stop the run if a<br> |
| 2617 | specified symbol is not found it a given shared object when it is<br> |
| 2618 | loaded into the process. This makes advanced working with function<br> |
| 2619 | intercepting and wrapping safer and more reliable.<br> |
| 2620 | <br> |
| 2621 | * Improved support for the Valkyrie GUI, version 2.0.0. GUI output<br> |
| 2622 | and control of Valgrind is now available for the tools Memcheck and<br> |
| 2623 | Helgrind. XML output from Valgrind is available for Memcheck,<br> |
| 2624 | Helgrind and exp-Ptrcheck.<br> |
| 2625 | <br> |
| 2626 | * More reliable stack unwinding on amd64-linux, particularly in the<br> |
| 2627 | presence of function wrappers, and with gcc-4.5 compiled code.<br> |
| 2628 | <br> |
| 2629 | * Modest scalability (performance improvements) for massive<br> |
| 2630 | long-running applications, particularly for those with huge amounts<br> |
| 2631 | of code.<br> |
| 2632 | <br> |
| 2633 | * Support for analyzing programs running under Wine with has been<br> |
| 2634 | improved. The header files <valgrind/valgrind.h>,<br> |
| 2635 | <valgrind/memcheck.h> and <valgrind/drd.h> can now be used in<br> |
| 2636 | Windows-programs compiled with MinGW or one of the Microsoft Visual<br> |
| 2637 | Studio compilers.<br> |
| 2638 | <br> |
| 2639 | * A rare but serious error in the 64-bit x86 CPU simulation was fixed.<br> |
| 2640 | The 32-bit simulator was not affected. This did not occur often,<br> |
| 2641 | but when it did would usually crash the program under test.<br> |
| 2642 | Bug 245925.<br> |
| 2643 | <br> |
| 2644 | * A large number of bugs were fixed. These are shown below.<br> |
| 2645 | <br> |
| 2646 | * A number of bugs were investigated, and were candidates for fixing,<br> |
| 2647 | but are not fixed in 3.6.0, due to lack of developer time. They may<br> |
| 2648 | get fixed in later releases. They are:<br> |
| 2649 | <br> |
| 2650 | 194402 vex amd64->IR: 0x48 0xF 0xAE 0x4 0x24 0x49 (FXSAVE64)<br> |
| 2651 | 212419 false positive "lock order violated" (A+B vs A) <br> |
| 2652 | 213685 Undefined value propagates past dependency breaking instruction<br> |
| 2653 | 216837 Incorrect instrumentation of NSOperationQueue on Darwin <br> |
| 2654 | 237920 valgrind segfault on fork failure <br> |
| 2655 | 242137 support for code compiled by LLVM-2.8<br> |
| 2656 | 242423 Another unknown Intel cache config value <br> |
| 2657 | 243232 Inconsistent Lock Orderings report with trylock <br> |
| 2658 | 243483 ppc: callgrind triggers VEX assertion failure <br> |
| 2659 | 243935 Helgrind: implementation of ANNOTATE_HAPPENS_BEFORE() is wrong<br> |
| 2660 | 244677 Helgrind crash hg_main.c:616 (map_threads_lookup): Assertion<br> |
| 2661 | 'thr' failed. <br> |
| 2662 | 246152 callgrind internal error after pthread_cancel on 32 Bit Linux <br> |
| 2663 | 249435 Analyzing wine programs with callgrind triggers a crash <br> |
| 2664 | 250038 ppc64: Altivec lvsr and lvsl instructions fail their regtest<br> |
| 2665 | 250065 Handling large allocations <br> |
| 2666 | 250101 huge "free" memory usage due to m_mallocfree.c<br> |
| 2667 | "superblocks fragmentation"<br> |
| 2668 | 251569 vex amd64->IR: 0xF 0x1 0xF9 0x8B 0x4C 0x24 (RDTSCP)<br> |
| 2669 | 252091 Callgrind on ARM does not detect function returns correctly<br> |
| 2670 | 252600 [PATCH] Allow lhs to be a pointer for shl/shr<br> |
| 2671 | 254420 memory pool tracking broken<br> |
| 2672 | n-i-bz support for adding symbols for JIT generated code<br> |
| 2673 | <br> |
| 2674 | <br> |
| 2675 | The following bugs have been fixed or resolved. Note that "n-i-bz"<br> |
| 2676 | stands for "not in bugzilla" -- that is, a bug that was reported to us<br> |
| 2677 | but never got a bugzilla entry. We encourage you to file bugs in<br> |
| 2678 | bugzilla (http://bugs.kde.org/enter_valgrind_bug.cgi) rather than<br> |
| 2679 | mailing the developers (or mailing lists) directly -- bugs that are<br> |
| 2680 | not entered into bugzilla tend to get forgotten about or ignored.<br> |
| 2681 | <br> |
| 2682 | To see details of a given bug, visit<br> |
| 2683 | https://bugs.kde.org/show_bug.cgi?id=XXXXXX<br> |
| 2684 | where XXXXXX is the bug number as listed below.<br> |
| 2685 | <br> |
| 2686 | 135264 dcbzl instruction missing<br> |
| 2687 | 142688 == 250799<br> |
| 2688 | 153699 Valgrind should report unaligned reads with movdqa<br> |
| 2689 | 180217 == 212335<br> |
| 2690 | 190429 Valgrind reports lost of errors in ld.so<br> |
| 2691 | with x86_64 2.9.90 glibc <br> |
| 2692 | 197266 valgrind appears to choke on the xmms instruction<br> |
| 2693 | "roundsd" on x86_64 <br> |
| 2694 | 197988 Crash when demangling very large symbol names<br> |
| 2695 | 202315 unhandled syscall: 332 (inotify_init1)<br> |
| 2696 | 203256 Add page-level profiling to Massif<br> |
| 2697 | 205093 dsymutil=yes needs quotes, locking (partial fix)<br> |
| 2698 | 205241 Snow Leopard 10.6 support (partial fix)<br> |
| 2699 | 206600 Leak checker fails to upgrade indirect blocks when their<br> |
| 2700 | parent becomes reachable <br> |
| 2701 | 210935 port valgrind.h (not valgrind) to win32 so apps run under<br> |
| 2702 | wine can make client requests<br> |
| 2703 | 211410 vex amd64->IR: 0x15 0xFF 0xFF 0x0 0x0 0x89<br> |
| 2704 | within Linux ip-stack checksum functions <br> |
| 2705 | 212335 unhandled instruction bytes: 0xF3 0xF 0xBD 0xC0<br> |
| 2706 | (lzcnt %eax,%eax) <br> |
| 2707 | 213685 Undefined value propagates past dependency breaking instruction<br> |
| 2708 | (partial fix)<br> |
| 2709 | 215914 Valgrind inserts bogus empty environment variable <br> |
| 2710 | 217863 == 197988<br> |
| 2711 | 219538 adjtimex syscall wrapper wrong in readonly adjtime mode <br> |
| 2712 | 222545 shmat fails under valgind on some arm targets <br> |
| 2713 | 222560 ARM NEON support <br> |
| 2714 | 230407 == 202315<br> |
| 2715 | 231076 == 202315<br> |
| 2716 | 232509 Docs build fails with formatting inside <title></title> elements <br> |
| 2717 | 232793 == 202315<br> |
| 2718 | 235642 [PATCH] syswrap-linux.c: support evdev EVIOCG* ioctls <br> |
| 2719 | 236546 vex x86->IR: 0x66 0xF 0x3A 0xA<br> |
| 2720 | 237202 vex amd64->IR: 0xF3 0xF 0xB8 0xC0 0x49 0x3B <br> |
| 2721 | 237371 better support for VALGRIND_MALLOCLIKE_BLOCK <br> |
| 2722 | 237485 symlink (syscall 57) is not supported on Mac OS <br> |
| 2723 | 237723 sysno == 101 exp-ptrcheck: the 'impossible' happened:<br> |
| 2724 | unhandled syscall <br> |
| 2725 | 238208 is_just_below_ESP doesn't take into account red-zone <br> |
| 2726 | 238345 valgrind passes wrong $0 when executing a shell script <br> |
| 2727 | 238679 mq_timedreceive syscall doesn't flag the reception buffer<br> |
| 2728 | as "defined"<br> |
| 2729 | 238696 fcntl command F_DUPFD_CLOEXEC not supported <br> |
| 2730 | 238713 unhandled instruction bytes: 0x66 0xF 0x29 0xC6 <br> |
| 2731 | 238713 unhandled instruction bytes: 0x66 0xF 0x29 0xC6 <br> |
| 2732 | 238745 3.5.0 Make fails on PPC Altivec opcodes, though configure<br> |
| 2733 | says "Altivec off"<br> |
| 2734 | 239992 vex amd64->IR: 0x48 0xF 0xC4 0xC1 0x0 0x48 <br> |
| 2735 | 240488 == 197988<br> |
| 2736 | 240639 == 212335<br> |
| 2737 | 241377 == 236546<br> |
| 2738 | 241903 == 202315<br> |
| 2739 | 241920 == 212335<br> |
| 2740 | 242606 unhandled syscall: setegid (in Ptrcheck)<br> |
| 2741 | 242814 Helgrind "Impossible has happened" during<br> |
| 2742 | QApplication::initInstance(); <br> |
| 2743 | 243064 Valgrind attempting to read debug information from iso <br> |
| 2744 | 243270 Make stack unwinding in Valgrind wrappers more reliable<br> |
| 2745 | 243884 exp-ptrcheck: the 'impossible happened: unhandled syscall <br> |
| 2746 | sysno = 277 (mq_open)<br> |
| 2747 | 244009 exp-ptrcheck unknown syscalls in analyzing lighttpd<br> |
| 2748 | 244493 ARM VFP d16-d31 registers support <br> |
| 2749 | 244670 add support for audit_session_self syscall on Mac OS 10.6<br> |
| 2750 | 244921 The xml report of helgrind tool is not well format<br> |
| 2751 | 244923 In the xml report file, the <preamble> not escape the <br> |
| 2752 | xml char, eg '<','&','>'<br> |
| 2753 | 245535 print full path names in plain text reports <br> |
| 2754 | 245925 x86-64 red zone handling problem <br> |
| 2755 | 246258 Valgrind not catching integer underruns + new [] s<br> |
| 2756 | 246311 reg/reg cmpxchg doesn't work on amd64<br> |
| 2757 | 246549 unhandled syscall unix:277 while testing 32-bit Darwin app <br> |
| 2758 | 246888 Improve Makefile.vex.am <br> |
| 2759 | 247510 [OS X 10.6] Memcheck reports unaddressable bytes passed <br> |
| 2760 | to [f]chmod_extended<br> |
| 2761 | 247526 IBM POWER6 (ISA 2.05) support is incomplete<br> |
| 2762 | 247561 Some leak testcases fails due to reachable addresses in<br> |
| 2763 | caller save regs<br> |
| 2764 | 247875 sizeofIRType to handle Ity_I128 <br> |
| 2765 | 247894 [PATCH] unhandled syscall sys_readahead <br> |
| 2766 | 247980 Doesn't honor CFLAGS passed to configure <br> |
| 2767 | 248373 darwin10.supp is empty in the trunk <br> |
| 2768 | 248822 Linux FIBMAP ioctl has int parameter instead of long<br> |
| 2769 | 248893 [PATCH] make readdwarf.c big endianess safe to enable<br> |
| 2770 | unwinding on big endian systems<br> |
| 2771 | 249224 Syscall 336 not supported (SYS_proc_info) <br> |
| 2772 | 249359 == 245535<br> |
| 2773 | 249775 Incorrect scheme for detecting NEON capabilities of host CPU<br> |
| 2774 | 249943 jni JVM init fails when using valgrind<br> |
| 2775 | 249991 Valgrind incorrectly declares AESKEYGENASSIST support<br> |
| 2776 | since VEX r2011<br> |
| 2777 | 249996 linux/arm: unhandled syscall: 181 (__NR_pwrite64)<br> |
| 2778 | 250799 frexp$fenv_access_off function generates SIGILL <br> |
| 2779 | 250998 vex x86->IR: unhandled instruction bytes: 0x66 0x66 0x66 0x2E <br> |
| 2780 | 251251 support pclmulqdq insn <br> |
| 2781 | 251362 valgrind: ARM: attach to debugger either fails or provokes<br> |
| 2782 | kernel oops <br> |
| 2783 | 251674 Unhandled syscall 294<br> |
| 2784 | 251818 == 254550<br> |
| 2785 | <br> |
| 2786 | 254257 Add support for debugfiles found by build-id<br> |
| 2787 | 254550 [PATCH] Implement DW_ATE_UTF (DWARF4)<br> |
| 2788 | 254646 Wrapped functions cause stack misalignment on OS X<br> |
| 2789 | (and possibly Linux)<br> |
| 2790 | 254556 ARM: valgrinding anything fails with SIGSEGV for 0xFFFF0FA0<br> |
| 2791 | <br> |
| 2792 | (3.6.0: 21 October 2010, vex r2068, valgrind r11471).<br> |
| 2793 | <br> |
| 2794 | <br> |
| 2795 | <br> |
| 2796 | Release 3.5.0 (19 August 2009)<br> |
| 2797 | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~<br> |
| 2798 | 3.5.0 is a feature release with many significant improvements and the<br> |
| 2799 | usual collection of bug fixes. The main improvement is that Valgrind<br> |
| 2800 | now works on Mac OS X.<br> |
| 2801 | <br> |
| 2802 | This release supports X86/Linux, AMD64/Linux, PPC32/Linux, PPC64/Linux<br> |
| 2803 | and X86/Darwin. Support for recent distros and toolchain components<br> |
| 2804 | (glibc 2.10, gcc 4.5) has been added.<br> |
| 2805 | <br> |
| 2806 | -------------------------<br> |
| 2807 | <br> |
| 2808 | Here is a short summary of the changes. Details are shown further<br> |
| 2809 | down:<br> |
| 2810 | <br> |
| 2811 | * Support for Mac OS X (10.5.x).<br> |
| 2812 | <br> |
| 2813 | * Improvements and simplifications to Memcheck's leak checker.<br> |
| 2814 | <br> |
| 2815 | * Clarification and simplifications in various aspects of Valgrind's<br> |
| 2816 | text output.<br> |
| 2817 | <br> |
| 2818 | * XML output for Helgrind and Ptrcheck.<br> |
| 2819 | <br> |
| 2820 | * Performance and stability improvements for Helgrind and DRD.<br> |
| 2821 | <br> |
| 2822 | * Genuinely atomic support for x86/amd64/ppc atomic instructions.<br> |
| 2823 | <br> |
| 2824 | * A new experimental tool, BBV, useful for computer architecture<br> |
| 2825 | research.<br> |
| 2826 | <br> |
| 2827 | * Improved Wine support, including ability to read Windows PDB<br> |
| 2828 | debuginfo.<br> |
| 2829 | <br> |
| 2830 | -------------------------<br> |
| 2831 | <br> |
| 2832 | Here are details of the above changes, followed by descriptions of<br> |
| 2833 | many other minor changes, and a list of fixed bugs.<br> |
| 2834 | <br> |
| 2835 | <br> |
| 2836 | * Valgrind now runs on Mac OS X. (Note that Mac OS X is sometimes<br> |
| 2837 | called "Darwin" because that is the name of the OS core, which is the<br> |
| 2838 | level that Valgrind works at.)<br> |
| 2839 | <br> |
| 2840 | Supported systems:<br> |
| 2841 | <br> |
| 2842 | - It requires OS 10.5.x (Leopard). Porting to 10.4.x is not planned<br> |
| 2843 | because it would require work and 10.4 is only becoming less common.<br> |
| 2844 | <br> |
| 2845 | - 32-bit programs on x86 and AMD64 (a.k.a x86-64) machines are supported<br> |
| 2846 | fairly well. For 10.5.x, 32-bit programs are the default even on<br> |
| 2847 | 64-bit machines, so it handles most current programs.<br> |
| 2848 | <br> |
| 2849 | - 64-bit programs on x86 and AMD64 (a.k.a x86-64) machines are not<br> |
| 2850 | officially supported, but simple programs at least will probably work.<br> |
| 2851 | However, start-up is slow.<br> |
| 2852 | <br> |
| 2853 | - PowerPC machines are not supported.<br> |
| 2854 | <br> |
| 2855 | Things that don't work:<br> |
| 2856 | <br> |
| 2857 | - The Ptrcheck tool.<br> |
| 2858 | <br> |
| 2859 | - Objective-C garbage collection.<br> |
| 2860 | <br> |
| 2861 | - --db-attach=yes.<br> |
| 2862 | <br> |
| 2863 | - If you have Rogue Amoeba's "Instant Hijack" program installed,<br> |
| 2864 | Valgrind will fail with a SIGTRAP at start-up. See<br> |
| 2865 | https://bugs.kde.org/show_bug.cgi?id=193917 for details and a<br> |
| 2866 | simple work-around.<br> |
| 2867 | <br> |
| 2868 | Usage notes:<br> |
| 2869 | <br> |
| 2870 | - You will likely find --dsymutil=yes a useful option, as error<br> |
| 2871 | messages may be imprecise without it.<br> |
| 2872 | <br> |
| 2873 | - Mac OS X support is new and therefore will be less robust than the<br> |
| 2874 | Linux support. Please report any bugs you find.<br> |
| 2875 | <br> |
| 2876 | - Threaded programs may run more slowly than on Linux.<br> |
| 2877 | <br> |
| 2878 | Many thanks to Greg Parker for developing this port over several years.<br> |
| 2879 | <br> |
| 2880 | <br> |
| 2881 | * Memcheck's leak checker has been improved. <br> |
| 2882 | <br> |
| 2883 | - The results for --leak-check=summary now match the summary results<br> |
| 2884 | for --leak-check=full. Previously they could differ because<br> |
| 2885 | --leak-check=summary counted "indirectly lost" blocks and<br> |
| 2886 | "suppressed" blocks as "definitely lost".<br> |
| 2887 | <br> |
| 2888 | - Blocks that are only reachable via at least one interior-pointer,<br> |
| 2889 | but are directly pointed to by a start-pointer, were previously<br> |
| 2890 | marked as "still reachable". They are now correctly marked as<br> |
| 2891 | "possibly lost".<br> |
| 2892 | <br> |
| 2893 | - The default value for the --leak-resolution option has been<br> |
| 2894 | changed from "low" to "high". In general, this means that more<br> |
| 2895 | leak reports will be produced, but each leak report will describe<br> |
| 2896 | fewer leaked blocks.<br> |
| 2897 | <br> |
| 2898 | - With --leak-check=full, "definitely lost" and "possibly lost"<br> |
| 2899 | leaks are now considered as proper errors, ie. they are counted<br> |
| 2900 | for the "ERROR SUMMARY" and affect the behaviour of<br> |
| 2901 | --error-exitcode. These leaks are not counted as errors if<br> |
| 2902 | --leak-check=summary is specified, however.<br> |
| 2903 | <br> |
| 2904 | - Documentation for the leak checker has been improved.<br> |
| 2905 | <br> |
| 2906 | <br> |
| 2907 | * Various aspects of Valgrind's text output have changed.<br> |
| 2908 | <br> |
| 2909 | - Valgrind's start-up message has changed. It is shorter but also<br> |
| 2910 | includes the command being run, which makes it easier to use<br> |
| 2911 | --trace-children=yes. An example:<br> |
| 2912 | <br> |
| 2913 | - Valgrind's shut-down messages have also changed. This is most<br> |
| 2914 | noticeable with Memcheck, where the leak summary now occurs before<br> |
| 2915 | the error summary. This change was necessary to allow leaks to be<br> |
| 2916 | counted as proper errors (see the description of the leak checker<br> |
| 2917 | changes above for more details). This was also necessary to fix a<br> |
| 2918 | longstanding bug in which uses of suppressions against leaks were<br> |
| 2919 | not "counted", leading to difficulties in maintaining suppression<br> |
| 2920 | files (see https://bugs.kde.org/show_bug.cgi?id=186790).<br> |
| 2921 | <br> |
| 2922 | - Behavior of -v has changed. In previous versions, -v printed out<br> |
| 2923 | a mixture of marginally-user-useful information, and tool/core<br> |
| 2924 | statistics. The statistics printing has now been moved to its own<br> |
| 2925 | flag, --stats=yes. This means -v is less verbose and more likely<br> |
| 2926 | to convey useful end-user information.<br> |
| 2927 | <br> |
| 2928 | - The format of some (non-XML) stack trace entries has changed a<br> |
| 2929 | little. Previously there were six possible forms:<br> |
| 2930 | <br> |
| 2931 | 0x80483BF: really (a.c:20)<br> |
| 2932 | 0x80483BF: really (in /foo/a.out)<br> |
| 2933 | 0x80483BF: really<br> |
| 2934 | 0x80483BF: (within /foo/a.out)<br> |
| 2935 | 0x80483BF: ??? (a.c:20)<br> |
| 2936 | 0x80483BF: ???<br> |
| 2937 | <br> |
| 2938 | The third and fourth of these forms have been made more consistent<br> |
| 2939 | with the others. The six possible forms are now:<br> |
| 2940 | <br> |
| 2941 | 0x80483BF: really (a.c:20)<br> |
| 2942 | 0x80483BF: really (in /foo/a.out)<br> |
| 2943 | 0x80483BF: really (in ???)<br> |
| 2944 | 0x80483BF: ??? (in /foo/a.out)<br> |
| 2945 | 0x80483BF: ??? (a.c:20)<br> |
| 2946 | 0x80483BF: ???<br> |
| 2947 | <br> |
| 2948 | Stack traces produced when --xml=yes is specified are different<br> |
| 2949 | and unchanged.<br> |
| 2950 | <br> |
| 2951 | <br> |
| 2952 | * Helgrind and Ptrcheck now support XML output, so they can be used<br> |
| 2953 | from GUI tools. Also, the XML output mechanism has been<br> |
| 2954 | overhauled.<br> |
| 2955 | <br> |
| 2956 | - The XML format has been overhauled and generalised, so it is more<br> |
| 2957 | suitable for error reporting tools in general. The Memcheck<br> |
| 2958 | specific aspects of it have been removed. The new format, which<br> |
| 2959 | is an evolution of the old format, is described in<br> |
| 2960 | docs/internals/xml-output-protocol4.txt.<br> |
| 2961 | <br> |
| 2962 | - Memcheck has been updated to use the new format.<br> |
| 2963 | <br> |
| 2964 | - Helgrind and Ptrcheck are now able to emit output in this format.<br> |
| 2965 | <br> |
| 2966 | - The XML output mechanism has been overhauled. XML is now output<br> |
| 2967 | to its own file descriptor, which means that:<br> |
| 2968 | <br> |
| 2969 | * Valgrind can output text and XML independently.<br> |
| 2970 | <br> |
| 2971 | * The longstanding problem of XML output being corrupted by <br> |
| 2972 | unexpected un-tagged text messages is solved.<br> |
| 2973 | <br> |
| 2974 | As before, the destination for text output is specified using<br> |
| 2975 | --log-file=, --log-fd= or --log-socket=.<br> |
| 2976 | <br> |
| 2977 | As before, XML output for a tool is enabled using --xml=yes.<br> |
| 2978 | <br> |
| 2979 | Because there's a new XML output channel, the XML output<br> |
| 2980 | destination is now specified by --xml-file=, --xml-fd= or<br> |
| 2981 | --xml-socket=.<br> |
| 2982 | <br> |
| 2983 | Initial feedback has shown this causes some confusion. To<br> |
| 2984 | clarify, the two envisaged usage scenarios are:<br> |
| 2985 | <br> |
| 2986 | (1) Normal text output. In this case, do not specify --xml=yes<br> |
| 2987 | nor any of --xml-file=, --xml-fd= or --xml-socket=.<br> |
| 2988 | <br> |
| 2989 | (2) XML output. In this case, specify --xml=yes, and one of<br> |
| 2990 | --xml-file=, --xml-fd= or --xml-socket= to select the XML<br> |
| 2991 | destination, one of --log-file=, --log-fd= or --log-socket=<br> |
| 2992 | to select the destination for any remaining text messages,<br> |
| 2993 | and, importantly, -q.<br> |
| 2994 | <br> |
| 2995 | -q makes Valgrind completely silent on the text channel,<br> |
| 2996 | except in the case of critical failures, such as Valgrind<br> |
| 2997 | itself segfaulting, or failing to read debugging information.<br> |
| 2998 | Hence, in this scenario, it suffices to check whether or not<br> |
| 2999 | any output appeared on the text channel. If yes, then it is<br> |
| 3000 | likely to be a critical error which should be brought to the<br> |
| 3001 | attention of the user. If no (the text channel produced no<br> |
| 3002 | output) then it can be assumed that the run was successful.<br> |
| 3003 | <br> |
| 3004 | This allows GUIs to make the critical distinction they need to<br> |
| 3005 | make (did the run fail or not?) without having to search or<br> |
| 3006 | filter the text output channel in any way.<br> |
| 3007 | <br> |
| 3008 | It is also recommended to use --child-silent-after-fork=yes in<br> |
| 3009 | scenario (2).<br> |
| 3010 | <br> |
| 3011 | <br> |
| 3012 | * Improvements and changes in Helgrind:<br> |
| 3013 | <br> |
| 3014 | - XML output, as described above<br> |
| 3015 | <br> |
| 3016 | - Checks for consistent association between pthread condition<br> |
| 3017 | variables and their associated mutexes are now performed.<br> |
| 3018 | <br> |
| 3019 | - pthread_spinlock functions are supported.<br> |
| 3020 | <br> |
| 3021 | - Modest performance improvements.<br> |
| 3022 | <br> |
| 3023 | - Initial (skeletal) support for describing the behaviour of<br> |
| 3024 | non-POSIX synchronisation objects through ThreadSanitizer<br> |
| 3025 | compatible ANNOTATE_* macros.<br> |
| 3026 | <br> |
| 3027 | - More controllable tradeoffs between performance and the level of<br> |
| 3028 | detail of "previous" accesses in a race. There are now three<br> |
| 3029 | settings:<br> |
| 3030 | <br> |
| 3031 | * --history-level=full. This is the default, and was also the<br> |
| 3032 | default in 3.4.x. It shows both stacks involved in a race, but<br> |
| 3033 | requires a lot of memory and can be very slow in programs that<br> |
| 3034 | do many inter-thread synchronisation events.<br> |
| 3035 | <br> |
| 3036 | * --history-level=none. This only shows the later stack involved<br> |
| 3037 | in a race. This can be much faster than --history-level=full,<br> |
| 3038 | but makes it much more difficult to find the other access<br> |
| 3039 | involved in the race.<br> |
| 3040 | <br> |
| 3041 | The new intermediate setting is<br> |
| 3042 | <br> |
| 3043 | * --history-level=approx<br> |
| 3044 | <br> |
| 3045 | For the earlier (other) access, two stacks are presented. The<br> |
| 3046 | earlier access is guaranteed to be somewhere in between the two<br> |
| 3047 | program points denoted by those stacks. This is not as useful<br> |
| 3048 | as showing the exact stack for the previous access (as per<br> |
| 3049 | --history-level=full), but it is better than nothing, and it's<br> |
| 3050 | almost as fast as --history-level=none.<br> |
| 3051 | <br> |
| 3052 | <br> |
| 3053 | * New features and improvements in DRD:<br> |
| 3054 | <br> |
| 3055 | - The error messages printed by DRD are now easier to interpret.<br> |
| 3056 | Instead of using two different numbers to identify each thread<br> |
| 3057 | (Valgrind thread ID and DRD thread ID), DRD does now identify<br> |
| 3058 | threads via a single number (the DRD thread ID). Furthermore<br> |
| 3059 | "first observed at" information is now printed for all error<br> |
| 3060 | messages related to synchronization objects.<br> |
| 3061 | <br> |
| 3062 | - Added support for named semaphores (sem_open() and sem_close()).<br> |
| 3063 | <br> |
| 3064 | - Race conditions between pthread_barrier_wait() and<br> |
| 3065 | pthread_barrier_destroy() calls are now reported.<br> |
| 3066 | <br> |
| 3067 | - Added support for custom allocators through the macros<br> |
| 3068 | VALGRIND_MALLOCLIKE_BLOCK() VALGRIND_FREELIKE_BLOCK() (defined in<br> |
| 3069 | in <valgrind/valgrind.h>). An alternative for these two macros is<br> |
| 3070 | the new client request VG_USERREQ__DRD_CLEAN_MEMORY (defined in<br> |
| 3071 | <valgrind/drd.h>).<br> |
| 3072 | <br> |
| 3073 | - Added support for annotating non-POSIX synchronization objects<br> |
| 3074 | through several new ANNOTATE_*() macros.<br> |
| 3075 | <br> |
| 3076 | - OpenMP: added support for the OpenMP runtime (libgomp) included<br> |
| 3077 | with gcc versions 4.3.0 and 4.4.0.<br> |
| 3078 | <br> |
| 3079 | - Faster operation.<br> |
| 3080 | <br> |
| 3081 | - Added two new command-line options (--first-race-only and<br> |
| 3082 | --segment-merging-interval).<br> |
| 3083 | <br> |
| 3084 | <br> |
| 3085 | * Genuinely atomic support for x86/amd64/ppc atomic instructions<br> |
| 3086 | <br> |
| 3087 | Valgrind will now preserve (memory-access) atomicity of LOCK-<br> |
| 3088 | prefixed x86/amd64 instructions, and any others implying a global<br> |
| 3089 | bus lock. Ditto for PowerPC l{w,d}arx/st{w,d}cx. instructions.<br> |
| 3090 | <br> |
| 3091 | This means that Valgrinded processes will "play nicely" in<br> |
| 3092 | situations where communication with other processes, or the kernel,<br> |
| 3093 | is done through shared memory and coordinated with such atomic<br> |
| 3094 | instructions. Prior to this change, such arrangements usually<br> |
| 3095 | resulted in hangs, races or other synchronisation failures, because<br> |
| 3096 | Valgrind did not honour atomicity of such instructions.<br> |
| 3097 | <br> |
| 3098 | <br> |
| 3099 | * A new experimental tool, BBV, has been added. BBV generates basic<br> |
| 3100 | block vectors for use with the SimPoint analysis tool, which allows<br> |
| 3101 | a program's overall behaviour to be approximated by running only a<br> |
| 3102 | fraction of it. This is useful for computer architecture<br> |
| 3103 | researchers. You can run BBV by specifying --tool=exp-bbv (the<br> |
| 3104 | "exp-" prefix is short for "experimental"). BBV was written by<br> |
| 3105 | Vince Weaver.<br> |
| 3106 | <br> |
| 3107 | <br> |
| 3108 | * Modestly improved support for running Windows applications under<br> |
| 3109 | Wine. In particular, initial support for reading Windows .PDB debug<br> |
| 3110 | information has been added.<br> |
| 3111 | <br> |
| 3112 | <br> |
| 3113 | * A new Memcheck client request VALGRIND_COUNT_LEAK_BLOCKS has been<br> |
| 3114 | added. It is similar to VALGRIND_COUNT_LEAKS but counts blocks<br> |
| 3115 | instead of bytes.<br> |
| 3116 | <br> |
| 3117 | <br> |
| 3118 | * The Valgrind client requests VALGRIND_PRINTF and<br> |
| 3119 | VALGRIND_PRINTF_BACKTRACE have been changed slightly. Previously,<br> |
| 3120 | the string was always printed immediately on its own line. Now, the<br> |
| 3121 | string will be added to a buffer but not printed until a newline is<br> |
| 3122 | encountered, or other Valgrind output is printed (note that for<br> |
| 3123 | VALGRIND_PRINTF_BACKTRACE, the back-trace itself is considered<br> |
| 3124 | "other Valgrind output"). This allows you to use multiple<br> |
| 3125 | VALGRIND_PRINTF calls to build up a single output line, and also to<br> |
| 3126 | print multiple output lines with a single request (by embedding<br> |
| 3127 | multiple newlines in the string).<br> |
| 3128 | <br> |
| 3129 | <br> |
| 3130 | * The graphs drawn by Massif's ms_print program have changed slightly:<br> |
| 3131 | <br> |
| 3132 | - The half-height chars '.' and ',' are no longer drawn, because<br> |
| 3133 | they are confusing. The --y option can be used if the default<br> |
| 3134 | y-resolution is not high enough.<br> |
| 3135 | <br> |
| 3136 | - Horizontal lines are now drawn after the top of a snapshot if<br> |
| 3137 | there is a gap until the next snapshot. This makes it clear that<br> |
| 3138 | the memory usage has not dropped to zero between snapshots.<br> |
| 3139 | <br> |
| 3140 | <br> |
| 3141 | * Something that happened in 3.4.0, but wasn't clearly announced: the<br> |
| 3142 | option --read-var-info=yes can be used by some tools (Memcheck,<br> |
| 3143 | Helgrind and DRD). When enabled, it causes Valgrind to read DWARF3<br> |
| 3144 | variable type and location information. This makes those tools<br> |
| 3145 | start up more slowly and increases memory consumption, but<br> |
| 3146 | descriptions of data addresses in error messages become more<br> |
| 3147 | detailed.<br> |
| 3148 | <br> |
| 3149 | <br> |
| 3150 | * exp-Omega, an experimental instantaneous leak-detecting tool, was<br> |
| 3151 | disabled in 3.4.0 due to a lack of interest and maintenance,<br> |
| 3152 | although the source code was still in the distribution. The source<br> |
| 3153 | code has now been removed from the distribution. For anyone<br> |
| 3154 | interested, the removal occurred in SVN revision r10247.<br> |
| 3155 | <br> |
| 3156 | <br> |
| 3157 | * Some changes have been made to the build system.<br> |
| 3158 | <br> |
| 3159 | - VEX/ is now integrated properly into the build system. This means<br> |
| 3160 | that dependency tracking within VEX/ now works properly, "make<br> |
| 3161 | install" will work without requiring "make" before it, and<br> |
| 3162 | parallel builds (ie. 'make -j') now work (previously a<br> |
| 3163 | .NOTPARALLEL directive was used to serialize builds, ie. 'make -j'<br> |
| 3164 | was effectively ignored).<br> |
| 3165 | <br> |
| 3166 | - The --with-vex configure option has been removed. It was of<br> |
| 3167 | little use and removing it simplified the build system.<br> |
| 3168 | <br> |
| 3169 | - The location of some install files has changed. This should not<br> |
| 3170 | affect most users. Those who might be affected:<br> |
| 3171 | <br> |
| 3172 | * For people who use Valgrind with MPI programs, the installed<br> |
| 3173 | libmpiwrap.so library has moved from<br> |
| 3174 | $(INSTALL)/<platform>/libmpiwrap.so to<br> |
| 3175 | $(INSTALL)/libmpiwrap-<platform>.so.<br> |
| 3176 | <br> |
| 3177 | * For people who distribute standalone Valgrind tools, the<br> |
| 3178 | installed libraries such as $(INSTALL)/<platform>/libcoregrind.a<br> |
| 3179 | have moved to $(INSTALL)/libcoregrind-<platform>.a.<br> |
| 3180 | <br> |
| 3181 | These changes simplify the build system.<br> |
| 3182 | <br> |
| 3183 | - Previously, all the distributed suppression (*.supp) files were<br> |
| 3184 | installed. Now, only default.supp is installed. This should not<br> |
| 3185 | affect users as the other installed suppression files were not<br> |
| 3186 | read; the fact that they were installed was a mistake.<br> |
| 3187 | <br> |
| 3188 | <br> |
| 3189 | * KNOWN LIMITATIONS:<br> |
| 3190 | <br> |
| 3191 | - Memcheck is unusable with the Intel compiler suite version 11.1,<br> |
| 3192 | when it generates code for SSE2-and-above capable targets. This<br> |
| 3193 | is because of icc's use of highly optimised inlined strlen<br> |
| 3194 | implementations. It causes Memcheck to report huge numbers of<br> |
| 3195 | false errors even in simple programs. Helgrind and DRD may also<br> |
| 3196 | have problems.<br> |
| 3197 | <br> |
| 3198 | Versions 11.0 and earlier may be OK, but this has not been<br> |
| 3199 | properly tested.<br> |
| 3200 | <br> |
| 3201 | <br> |
| 3202 | The following bugs have been fixed or resolved. Note that "n-i-bz"<br> |
| 3203 | stands for "not in bugzilla" -- that is, a bug that was reported to us<br> |
| 3204 | but never got a bugzilla entry. We encourage you to file bugs in<br> |
| 3205 | bugzilla (http://bugs.kde.org/enter_valgrind_bug.cgi) rather than<br> |
| 3206 | mailing the developers (or mailing lists) directly -- bugs that are<br> |
| 3207 | not entered into bugzilla tend to get forgotten about or ignored.<br> |
| 3208 | <br> |
| 3209 | To see details of a given bug, visit<br> |
| 3210 | https://bugs.kde.org/show_bug.cgi?id=XXXXXX<br> |
| 3211 | where XXXXXX is the bug number as listed below.<br> |
| 3212 | <br> |
| 3213 | 84303 How about a LockCheck tool? <br> |
| 3214 | 91633 dereference of null ptr in vgPlain_st_basetype <br> |
| 3215 | 97452 Valgrind doesn't report any pthreads problems <br> |
| 3216 | 100628 leak-check gets assertion failure when using <br> |
| 3217 | VALGRIND_MALLOCLIKE_BLOCK on malloc()ed memory <br> |
| 3218 | 108528 NPTL pthread cleanup handlers not called <br> |
| 3219 | 110126 Valgrind 2.4.1 configure.in tramples CFLAGS <br> |
| 3220 | 110128 mallinfo is not implemented... <br> |
| 3221 | 110770 VEX: Generated files not always updated when making valgrind<br> |
| 3222 | 111102 Memcheck: problems with large (memory footprint) applications <br> |
| 3223 | 115673 Vex's decoder should never assert <br> |
| 3224 | 117564 False positive: Syscall param clone(child_tidptr) contains<br> |
| 3225 | uninitialised byte(s) <br> |
| 3226 | 119404 executing ssh from inside valgrind fails <br> |
| 3227 | 133679 Callgrind does not write path names to sources with dwarf debug<br> |
| 3228 | info<br> |
| 3229 | 135847 configure.in problem with non gnu compilers (and possible fix) <br> |
| 3230 | 136154 threads.c:273 (vgCallgrind_post_signal): Assertion<br> |
| 3231 | '*(vgCallgrind_current_fn_stack.top) == 0' failed. <br> |
| 3232 | 136230 memcheck reports "possibly lost", should be "still reachable" <br> |
| 3233 | 137073 NULL arg to MALLOCLIKE_BLOCK causes crash <br> |
| 3234 | 137904 Valgrind reports a memory leak when using POSIX threads,<br> |
| 3235 | while it shouldn't <br> |
| 3236 | 139076 valgrind VT_GETSTATE error <br> |
| 3237 | 142228 complaint of elf_dynamic_do_rela in trivial usage <br> |
| 3238 | 145347 spurious warning with USBDEVFS_REAPURB <br> |
| 3239 | 148441 (wine) can't find memory leak in Wine, win32 binary <br> |
| 3240 | executable file.<br> |
| 3241 | 148742 Leak-check fails assert on exit <br> |
| 3242 | 149878 add (proper) check for calloc integer overflow <br> |
| 3243 | 150606 Call graph is broken when using callgrind control <br> |
| 3244 | 152393 leak errors produce an exit code of 0. I need some way to <br> |
| 3245 | cause leak errors to result in a nonzero exit code. <br> |
| 3246 | 157154 documentation (leak-resolution doc speaks about num-callers<br> |
| 3247 | def=4) + what is a loss record<br> |
| 3248 | 159501 incorrect handling of ALSA ioctls <br> |
| 3249 | 162020 Valgrinding an empty/zero-byte file crashes valgrind <br> |
| 3250 | 162482 ppc: Valgrind crashes while reading stabs information <br> |
| 3251 | 162718 x86: avoid segment selector 0 in sys_set_thread_area() <br> |
| 3252 | 163253 (wine) canonicaliseSymtab forgot some fields in DiSym <br> |
| 3253 | 163560 VEX/test_main.c is missing from valgrind-3.3.1 <br> |
| 3254 | 164353 malloc_usable_size() doesn't return a usable size <br> |
| 3255 | 165468 Inconsistent formatting in memcheck manual -- please fix <br> |
| 3256 | 169505 main.c:286 (endOfInstr):<br> |
| 3257 | Assertion 'ii->cost_offset == *cost_offset' failed <br> |
| 3258 | 177206 Generate default.supp during compile instead of configure<br> |
| 3259 | 177209 Configure valt_load_address based on arch+os <br> |
| 3260 | 177305 eventfd / syscall 323 patch lost<br> |
| 3261 | 179731 Tests fail to build because of inlining of non-local asm labels<br> |
| 3262 | 181394 helgrind: libhb_core.c:3762 (msm_write): Assertion <br> |
| 3263 | 'ordxx == POrd_EQ || ordxx == POrd_LT' failed. <br> |
| 3264 | 181594 Bogus warning for empty text segment <br> |
| 3265 | 181707 dwarf doesn't require enumerations to have name <br> |
| 3266 | 185038 exp-ptrcheck: "unhandled syscall: 285" (fallocate) on x86_64 <br> |
| 3267 | 185050 exp-ptrcheck: sg_main.c:727 (add_block_to_GlobalTree):<br> |
| 3268 | Assertion '!already_present' failed.<br> |
| 3269 | 185359 exp-ptrcheck: unhandled syscall getresuid()<br> |
| 3270 | 185794 "WARNING: unhandled syscall: 285" (fallocate) on x86_64<br> |
| 3271 | 185816 Valgrind is unable to handle debug info for files with split<br> |
| 3272 | debug info that are prelinked afterwards <br> |
| 3273 | 185980 [darwin] unhandled syscall: sem_open <br> |
| 3274 | 186238 bbToIR_AMD64: disInstr miscalculated next %rip<br> |
| 3275 | 186507 exp-ptrcheck unhandled syscalls prctl, etc. <br> |
| 3276 | 186790 Suppression pattern used for leaks are not reported <br> |
| 3277 | 186796 Symbols with length>200 in suppression files are ignored <br> |
| 3278 | 187048 drd: mutex PTHREAD_PROCESS_SHARED attribute missinterpretation<br> |
| 3279 | 187416 exp-ptrcheck: support for __NR_{setregid,setreuid,setresuid}<br> |
| 3280 | 188038 helgrind: hg_main.c:926: mk_SHVAL_fail: the 'impossible' happened<br> |
| 3281 | 188046 bashisms in the configure script<br> |
| 3282 | 188127 amd64->IR: unhandled instruction bytes: 0xF0 0xF 0xB0 0xA<br> |
| 3283 | 188161 memcheck: --track-origins=yes asserts "mc_machine.c:672<br> |
| 3284 | (get_otrack_shadow_offset_wrk): the 'impossible' happened."<br> |
| 3285 | 188248 helgrind: pthread_cleanup_push, pthread_rwlock_unlock, <br> |
| 3286 | assertion fail "!lock->heldBy" <br> |
| 3287 | 188427 Add support for epoll_create1 (with patch) <br> |
| 3288 | 188530 Support for SIOCGSTAMPNS<br> |
| 3289 | 188560 Include valgrind.spec in the tarball<br> |
| 3290 | 188572 Valgrind on Mac should suppress setenv() mem leak <br> |
| 3291 | 189054 Valgrind fails to build because of duplicate non-local asm labels <br> |
| 3292 | 189737 vex amd64->IR: unhandled instruction bytes: 0xAC<br> |
| 3293 | 189762 epoll_create syscall not handled (--tool=exp-ptrcheck)<br> |
| 3294 | 189763 drd assertion failure: s_threadinfo[tid].is_recording <br> |
| 3295 | 190219 unhandled syscall: 328 (x86-linux)<br> |
| 3296 | 190391 dup of 181394; see above<br> |
| 3297 | 190429 Valgrind reports lots of errors in ld.so with x86_64 2.9.90 glibc <br> |
| 3298 | 190820 No debug information on powerpc-linux<br> |
| 3299 | 191095 PATCH: Improve usbdevfs ioctl handling <br> |
| 3300 | 191182 memcheck: VALGRIND_LEAK_CHECK quadratic when big nr of chunks<br> |
| 3301 | or big nr of errors<br> |
| 3302 | 191189 --xml=yes should obey --gen-suppressions=all <br> |
| 3303 | 191192 syslog() needs a suppression on macosx <br> |
| 3304 | 191271 DARWIN: WARNING: unhandled syscall: 33554697 a.k.a.: 265 <br> |
| 3305 | 191761 getrlimit on MacOSX <br> |
| 3306 | 191992 multiple --fn-skip only works sometimes; dependent on order <br> |
| 3307 | 192634 V. reports "aspacem sync_check_mapping_callback: <br> |
| 3308 | segment mismatch" on Darwin<br> |
| 3309 | 192954 __extension__ missing on 2 client requests <br> |
| 3310 | 194429 Crash at start-up with glibc-2.10.1 and linux-2.6.29 <br> |
| 3311 | 194474 "INSTALL" file has different build instructions than "README"<br> |
| 3312 | 194671 Unhandled syscall (sem_wait?) from mac valgrind <br> |
| 3313 | 195069 memcheck: reports leak (memory still reachable) for <br> |
| 3314 | printf("%d', x) <br> |
| 3315 | 195169 drd: (vgDrd_barrier_post_wait):<br> |
| 3316 | Assertion 'r->sg[p->post_iteration]' failed. <br> |
| 3317 | 195268 valgrind --log-file doesn't accept ~/...<br> |
| 3318 | 195838 VEX abort: LibVEX_N_SPILL_BYTES too small for CPUID boilerplate <br> |
| 3319 | 195860 WARNING: unhandled syscall: unix:223 <br> |
| 3320 | 196528 need a error suppression for pthread_rwlock_init under os x? <br> |
| 3321 | 197227 Support aio_* syscalls on Darwin<br> |
| 3322 | 197456 valgrind should reject --suppressions=(directory) <br> |
| 3323 | 197512 DWARF2 CFI reader: unhandled CFI instruction 0:10 <br> |
| 3324 | 197591 unhandled syscall 27 (mincore) <br> |
| 3325 | 197793 Merge DCAS branch to the trunk == 85756, 142103<br> |
| 3326 | 197794 Avoid duplicate filenames in Vex <br> |
| 3327 | 197898 make check fails on current SVN <br> |
| 3328 | 197901 make check fails also under exp-ptrcheck in current SVN <br> |
| 3329 | 197929 Make --leak-resolution=high the default <br> |
| 3330 | 197930 Reduce spacing between leak reports <br> |
| 3331 | 197933 Print command line of client at start-up, and shorten preamble <br> |
| 3332 | 197966 unhandled syscall 205 (x86-linux, --tool=exp-ptrcheck)<br> |
| 3333 | 198395 add BBV to the distribution as an experimental tool <br> |
| 3334 | 198624 Missing syscalls on Darwin: 82, 167, 281, 347 <br> |
| 3335 | 198649 callgrind_annotate doesn't cumulate counters <br> |
| 3336 | 199338 callgrind_annotate sorting/thresholds are broken for all but Ir <br> |
| 3337 | 199977 Valgrind complains about an unrecognized instruction in the<br> |
| 3338 | atomic_incs test program<br> |
| 3339 | 200029 valgrind isn't able to read Fedora 12 debuginfo <br> |
| 3340 | 200760 darwin unhandled syscall: unix:284 <br> |
| 3341 | 200827 DRD doesn't work on Mac OS X <br> |
| 3342 | 200990 VG_(read_millisecond_timer)() does not work correctly <br> |
| 3343 | 201016 Valgrind does not support pthread_kill() on Mac OS <br> |
| 3344 | 201169 Document --read-var-info<br> |
| 3345 | 201323 Pre-3.5.0 performance sanity checking <br> |
| 3346 | 201384 Review user manual for the 3.5.0 release <br> |
| 3347 | 201585 mfpvr not implemented on ppc <br> |
| 3348 | 201708 tests failing because x86 direction flag is left set <br> |
| 3349 | 201757 Valgrind doesn't handle any recent sys_futex additions <br> |
| 3350 | 204377 64-bit valgrind can not start a shell script<br> |
| 3351 | (with #!/path/to/shell) if the shell is a 32-bit executable<br> |
| 3352 | n-i-bz drd: fixed assertion failure triggered by mutex reinitialization.<br> |
| 3353 | n-i-bz drd: fixed a bug that caused incorrect messages to be printed<br> |
| 3354 | about memory allocation events with memory access tracing enabled<br> |
| 3355 | n-i-bz drd: fixed a memory leak triggered by vector clock deallocation<br> |
| 3356 | <br> |
| 3357 | (3.5.0: 19 Aug 2009, vex r1913, valgrind r10846).<br> |
| 3358 | <br> |
| 3359 | <br> |
| 3360 | <br> |
| 3361 | Release 3.4.1 (28 February 2009)<br> |
| 3362 | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~<br> |
| 3363 | 3.4.1 is a bug-fix release that fixes some regressions and assertion<br> |
| 3364 | failures in debug info reading in 3.4.0, most notably incorrect stack<br> |
| 3365 | traces on amd64-linux on older (glibc-2.3 based) systems. Various<br> |
| 3366 | other debug info problems are also fixed. A number of bugs in the<br> |
| 3367 | exp-ptrcheck tool introduced in 3.4.0 have been fixed.<br> |
| 3368 | <br> |
| 3369 | In view of the fact that 3.4.0 contains user-visible regressions<br> |
| 3370 | relative to 3.3.x, upgrading to 3.4.1 is recommended. Packagers are<br> |
| 3371 | encouraged to ship 3.4.1 in preference to 3.4.0.<br> |
| 3372 | <br> |
| 3373 | The fixed bugs are as follows. Note that "n-i-bz" stands for "not in<br> |
| 3374 | bugzilla" -- that is, a bug that was reported to us but never got a<br> |
| 3375 | bugzilla entry. We encourage you to file bugs in bugzilla<br> |
| 3376 | (http://bugs.kde.org/enter_valgrind_bug.cgi) rather than mailing the<br> |
| 3377 | developers (or mailing lists) directly -- bugs that are not entered<br> |
| 3378 | into bugzilla tend to get forgotten about or ignored.<br> |
| 3379 | <br> |
| 3380 | n-i-bz Fix various bugs reading icc-11 generated debug info<br> |
| 3381 | n-i-bz Fix various bugs reading gcc-4.4 generated debug info<br> |
| 3382 | n-i-bz Preliminary support for glibc-2.10 / Fedora 11<br> |
| 3383 | n-i-bz Cachegrind and Callgrind: handle non-power-of-two cache sizes,<br> |
| 3384 | so as to support (eg) 24k Atom D1 and Core2 with 3/6/12MB L2.<br> |
| 3385 | 179618 exp-ptrcheck crashed / exit prematurely<br> |
| 3386 | 179624 helgrind: false positive races with pthread_create and<br> |
| 3387 | recv/open/close/read<br> |
| 3388 | 134207 pkg-config output contains @VG_PLATFORM@<br> |
| 3389 | 176926 floating point exception at valgrind startup with PPC 440EPX<br> |
| 3390 | 181594 Bogus warning for empty text segment<br> |
| 3391 | 173751 amd64->IR: 0x48 0xF 0x6F 0x45 (even more redundant rex prefixes)<br> |
| 3392 | 181707 Dwarf3 doesn't require enumerations to have name<br> |
| 3393 | 185038 exp-ptrcheck: "unhandled syscall: 285" (fallocate) on x86_64<br> |
| 3394 | 185050 exp-ptrcheck: sg_main.c:727 (add_block_to_GlobalTree):<br> |
| 3395 | Assertion '!already_present' failed.<br> |
| 3396 | 185359 exp-ptrcheck unhandled syscall getresuid()<br> |
| 3397 | <br> |
| 3398 | (3.4.1.RC1: 24 Feb 2008, vex r1884, valgrind r9253).<br> |
| 3399 | (3.4.1: 28 Feb 2008, vex r1884, valgrind r9293).<br> |
| 3400 | <br> |
| 3401 | <br> |
| 3402 | <br> |
| 3403 | Release 3.4.0 (2 January 2009)<br> |
| 3404 | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~<br> |
| 3405 | 3.4.0 is a feature release with many significant improvements and the<br> |
| 3406 | usual collection of bug fixes. This release supports X86/Linux,<br> |
| 3407 | AMD64/Linux, PPC32/Linux and PPC64/Linux. Support for recent distros<br> |
| 3408 | (using gcc 4.4, glibc 2.8 and 2.9) has been added.<br> |
| 3409 | <br> |
| 3410 | 3.4.0 brings some significant tool improvements. Memcheck can now<br> |
| 3411 | report the origin of uninitialised values, the thread checkers<br> |
| 3412 | Helgrind and DRD are much improved, and we have a new experimental<br> |
| 3413 | tool, exp-Ptrcheck, which is able to detect overruns of stack and<br> |
| 3414 | global arrays. In detail:<br> |
| 3415 | <br> |
| 3416 | * Memcheck is now able to track the origin of uninitialised values.<br> |
| 3417 | When it reports an uninitialised value error, it will try to show<br> |
| 3418 | the origin of the value, as either a heap or stack allocation.<br> |
| 3419 | Origin tracking is expensive and so is not enabled by default. To<br> |
| 3420 | use it, specify --track-origins=yes. Memcheck's speed will be<br> |
| 3421 | essentially halved, and memory usage will be significantly<br> |
| 3422 | increased. Nevertheless it can drastically reduce the effort<br> |
| 3423 | required to identify the root cause of uninitialised value errors,<br> |
| 3424 | and so is often a programmer productivity win, despite running more<br> |
| 3425 | slowly.<br> |
| 3426 | <br> |
| 3427 | * A version (1.4.0) of the Valkyrie GUI, that works with Memcheck in<br> |
| 3428 | 3.4.0, will be released shortly.<br> |
| 3429 | <br> |
| 3430 | * Helgrind's race detection algorithm has been completely redesigned<br> |
| 3431 | and reimplemented, to address usability and scalability concerns:<br> |
| 3432 | <br> |
| 3433 | - The new algorithm has a lower false-error rate: it is much less<br> |
| 3434 | likely to report races that do not really exist.<br> |
| 3435 | <br> |
| 3436 | - Helgrind will display full call stacks for both accesses involved<br> |
| 3437 | in a race. This makes it easier to identify the root causes of<br> |
| 3438 | races.<br> |
| 3439 | <br> |
| 3440 | - Limitations on the size of program that can run have been removed.<br> |
| 3441 | <br> |
| 3442 | - Performance has been modestly improved, although that is very<br> |
| 3443 | workload-dependent.<br> |
| 3444 | <br> |
| 3445 | - Direct support for Qt4 threading has been added.<br> |
| 3446 | <br> |
| 3447 | - pthread_barriers are now directly supported.<br> |
| 3448 | <br> |
| 3449 | - Helgrind works well on all supported Linux targets.<br> |
| 3450 | <br> |
| 3451 | * The DRD thread debugging tool has seen major improvements:<br> |
| 3452 | <br> |
| 3453 | - Greatly improved performance and significantly reduced memory<br> |
| 3454 | usage.<br> |
| 3455 | <br> |
| 3456 | - Support for several major threading libraries (Boost.Thread, Qt4,<br> |
| 3457 | glib, OpenMP) has been added.<br> |
| 3458 | <br> |
| 3459 | - Support for atomic instructions, POSIX semaphores, barriers and<br> |
| 3460 | reader-writer locks has been added.<br> |
| 3461 | <br> |
| 3462 | - Works now on PowerPC CPUs too.<br> |
| 3463 | <br> |
| 3464 | - Added support for printing thread stack usage at thread exit time.<br> |
| 3465 | <br> |
| 3466 | - Added support for debugging lock contention.<br> |
| 3467 | <br> |
| 3468 | - Added a manual for Drd.<br> |
| 3469 | <br> |
| 3470 | * A new experimental tool, exp-Ptrcheck, has been added. Ptrcheck<br> |
| 3471 | checks for misuses of pointers. In that sense it is a bit like<br> |
| 3472 | Memcheck. However, Ptrcheck can do things Memcheck can't: it can<br> |
| 3473 | detect overruns of stack and global arrays, it can detect<br> |
| 3474 | arbitrarily far out-of-bounds accesses to heap blocks, and it can<br> |
| 3475 | detect accesses heap blocks that have been freed a very long time<br> |
| 3476 | ago (millions of blocks in the past).<br> |
| 3477 | <br> |
| 3478 | Ptrcheck currently works only on x86-linux and amd64-linux. To use<br> |
| 3479 | it, use --tool=exp-ptrcheck. A simple manual is provided, as part<br> |
| 3480 | of the main Valgrind documentation. As this is an experimental<br> |
| 3481 | tool, we would be particularly interested in hearing about your<br> |
| 3482 | experiences with it.<br> |
| 3483 | <br> |
| 3484 | * exp-Omega, an experimental instantaneous leak-detecting tool, is no<br> |
| 3485 | longer built by default, although the code remains in the repository<br> |
| 3486 | and the tarball. This is due to three factors: a perceived lack of<br> |
| 3487 | users, a lack of maintenance, and concerns that it may not be<br> |
| 3488 | possible to achieve reliable operation using the existing design.<br> |
| 3489 | <br> |
| 3490 | * As usual, support for the latest Linux distros and toolchain<br> |
| 3491 | components has been added. It should work well on Fedora Core 10,<br> |
| 3492 | OpenSUSE 11.1 and Ubuntu 8.10. gcc-4.4 (in its current pre-release<br> |
| 3493 | state) is supported, as is glibc-2.9. The C++ demangler has been<br> |
| 3494 | updated so as to work well with C++ compiled by even the most recent<br> |
| 3495 | g++'s.<br> |
| 3496 | <br> |
| 3497 | * You can now use frame-level wildcards in suppressions. This was a<br> |
| 3498 | frequently-requested enhancement. A line "..." in a suppression now<br> |
| 3499 | matches zero or more frames. This makes it easier to write<br> |
| 3500 | suppressions which are precise yet insensitive to changes in<br> |
| 3501 | inlining behaviour.<br> |
| 3502 | <br> |
| 3503 | * 3.4.0 adds support on x86/amd64 for the SSSE3 instruction set.<br> |
| 3504 | <br> |
| 3505 | * Very basic support for IBM Power6 has been added (64-bit processes only).<br> |
| 3506 | <br> |
| 3507 | * Valgrind is now cross-compilable. For example, it is possible to<br> |
| 3508 | cross compile Valgrind on an x86/amd64-linux host, so that it runs<br> |
| 3509 | on a ppc32/64-linux target.<br> |
| 3510 | <br> |
| 3511 | * You can set the main thread's stack size at startup using the<br> |
| 3512 | new --main-stacksize= flag (subject of course to ulimit settings).<br> |
| 3513 | This is useful for running apps that need a lot of stack space.<br> |
| 3514 | <br> |
| 3515 | * The limitation that you can't use --trace-children=yes together<br> |
| 3516 | with --db-attach=yes has been removed.<br> |
| 3517 | <br> |
| 3518 | * The following bugs have been fixed. Note that "n-i-bz" stands for<br> |
| 3519 | "not in bugzilla" -- that is, a bug that was reported to us but<br> |
| 3520 | never got a bugzilla entry. We encourage you to file bugs in<br> |
| 3521 | bugzilla (http://bugs.kde.org/enter_valgrind_bug.cgi) rather than<br> |
| 3522 | mailing the developers (or mailing lists) directly.<br> |
| 3523 | <br> |
| 3524 | n-i-bz Make return types for some client requests 64-bit clean<br> |
| 3525 | n-i-bz glibc 2.9 support<br> |
| 3526 | n-i-bz ignore unsafe .valgrindrc's (CVE-2008-4865)<br> |
| 3527 | n-i-bz MPI_Init(0,0) is valid but libmpiwrap.c segfaults<br> |
| 3528 | n-i-bz Building in an env without gdb gives bogus gdb attach<br> |
| 3529 | 92456 Tracing the origin of uninitialised memory<br> |
| 3530 | 106497 Valgrind does not demangle some C++ template symbols<br> |
| 3531 | 162222 ==106497<br> |
| 3532 | 151612 Suppression with "..." (frame-level wildcards in .supp files)<br> |
| 3533 | 156404 Unable to start oocalc under memcheck on openSUSE 10.3 (64-bit)<br> |
| 3534 | 159285 unhandled syscall:25 (stime, on x86-linux)<br> |
| 3535 | 159452 unhandled ioctl 0x8B01 on "valgrind iwconfig"<br> |
| 3536 | 160954 ppc build of valgrind crashes with illegal instruction (isel)<br> |
| 3537 | 160956 mallinfo implementation, w/ patch<br> |
| 3538 | 162092 Valgrind fails to start gnome-system-monitor<br> |
| 3539 | 162819 malloc_free_fill test doesn't pass on glibc2.8 x86<br> |
| 3540 | 163794 assertion failure with "--track-origins=yes"<br> |
| 3541 | 163933 sigcontext.err and .trapno must be set together<br> |
| 3542 | 163955 remove constraint !(--db-attach=yes && --trace-children=yes)<br> |
| 3543 | 164476 Missing kernel module loading system calls<br> |
| 3544 | 164669 SVN regression: mmap() drops posix file locks<br> |
| 3545 | 166581 Callgrind output corruption when program forks<br> |
| 3546 | 167288 Patch file for missing system calls on Cell BE<br> |
| 3547 | 168943 unsupported scas instruction pentium<br> |
| 3548 | 171645 Unrecognised instruction (MOVSD, non-binutils encoding)<br> |
| 3549 | 172417 x86->IR: 0x82 ...<br> |
| 3550 | 172563 amd64->IR: 0xD9 0xF5 - fprem1<br> |
| 3551 | 173099 .lds linker script generation error<br> |
| 3552 | 173177 [x86_64] syscalls: 125/126/179 (capget/capset/quotactl)<br> |
| 3553 | 173751 amd64->IR: 0x48 0xF 0x6F 0x45 (even more redundant prefixes)<br> |
| 3554 | 174532 == 173751<br> |
| 3555 | 174908 --log-file value not expanded correctly for core file<br> |
| 3556 | 175044 Add lookup_dcookie for amd64<br> |
| 3557 | 175150 x86->IR: 0xF2 0xF 0x11 0xC1 (movss non-binutils encoding)<br> |
| 3558 | <br> |
| 3559 | Developer-visible changes:<br> |
| 3560 | <br> |
| 3561 | * Valgrind's debug-info reading machinery has been majorly overhauled.<br> |
| 3562 | It can now correctly establish the addresses for ELF data symbols,<br> |
| 3563 | which is something that has never worked properly before now.<br> |
| 3564 | <br> |
| 3565 | Also, Valgrind can now read DWARF3 type and location information for<br> |
| 3566 | stack and global variables. This makes it possible to use the<br> |
| 3567 | framework to build tools that rely on knowing the type and locations<br> |
| 3568 | of stack and global variables, for example exp-Ptrcheck.<br> |
| 3569 | <br> |
| 3570 | Reading of such information is disabled by default, because most<br> |
| 3571 | tools don't need it, and because it is expensive in space and time.<br> |
| 3572 | However, you can force Valgrind to read it, using the<br> |
| 3573 | --read-var-info=yes flag. Memcheck, Helgrind and DRD are able to<br> |
| 3574 | make use of such information, if present, to provide source-level<br> |
| 3575 | descriptions of data addresses in the error messages they create.<br> |
| 3576 | <br> |
| 3577 | (3.4.0.RC1: 24 Dec 2008, vex r1878, valgrind r8882).<br> |
| 3578 | (3.4.0: 3 Jan 2009, vex r1878, valgrind r8899).<br> |
| 3579 | <br> |
| 3580 | </p></div> |
| 3581 | </div> |
| 3582 | <div> |
| 3583 | <br><table class="nav" width="100%" cellspacing="3" cellpadding="2" border="0" summary="Navigation footer"> |
| 3584 | <tr> |
| 3585 | <td rowspan="2" width="40%" align="left"> |
| 3586 | <a accesskey="p" href="dist.authors.html"><< 1. AUTHORS</a> </td> |
| 3587 | <td width="20%" align="center"><a accesskey="u" href="dist.html">Up</a></td> |
| 3588 | <td rowspan="2" width="40%" align="right"> <a accesskey="n" href="dist.news.old.html">3. OLDER NEWS >></a> |
| 3589 | </td> |
| 3590 | </tr> |
| 3591 | <tr><td width="20%" align="center"><a accesskey="h" href="index.html">Home</a></td></tr> |
| 3592 | </table> |
| 3593 | </div> |
| 3594 | </body> |
| 3595 | </html> |