| Summary: | Valgrind chrashes with "Trace/BPT trap" | ||
|---|---|---|---|
| Product: | [Developer tools] valgrind | Reporter: | caulfield.holden |
| Component: | general | Assignee: | Julian Seward <jseward> |
| Status: | RESOLVED DUPLICATE | ||
| Severity: | crash | CC: | njn |
| Priority: | NOR | ||
| Version: | 3.5 SVN | ||
| Target Milestone: | --- | ||
| Platform: | Unlisted Binaries | ||
| OS: | macOS | ||
| Latest Commit: | Version Fixed In: | ||
| Sentry Crash Report: | |||
| Attachments: | Contents of the file config.log | ||
It may be that this is a dup of another recently discussed V-MacOSX bug, in which the problem is due to a conflict with some other package which also plays games with LD_PRELOAD (or is it dyld). But I don't remember much more than that (apart from, if you disable the other package, it's easily worked-around). It sounds like bug 193917 -- are you running a program called Instant Hijack? Problem solved. Just checked, and indeed I was running Instant Hijack. I did install it some time ago but it had since fallen off my radar. So, I disabled that one following the instructions posted in the comments attached to bug 193917. After that, valgrind seems to be working just fine. Thanks for the quick response. I took the liberty to mark this one as solved. *** This bug has been marked as a duplicate of bug 193917 *** |
Created attachment 35120 [details] Contents of the file config.log uname -a: Darwin <hostname has been removed> 9.7.1 Darwin Kernel Version 9.7.1: Thu Apr 23 13:52:18 PDT 2009; root:xnu-1228.14.1~1/RELEASE_I386 i386 valgrind -v: Trace/BPT trap G'day, I'm using revision 1906 from the SVN repository. I compiled and installed valgrind, with and without the configure flags "-enable-only64bit" or "-enable-only32bit", respectively, using the Apple supplied gcc-4.2 from XCode 3.1.3. Upon invocation of valgrind (e.g. just "valgrind") I get the message "Trace/BPT trap" in my terminal window. The system logs at the same time record the following messages: 07.07.09 11:51:31 kernel shared_region: enter(0xc73ef04,0x7c48058,0,7,0): vm_map_enter(0x90000000,0x20000000,0x6ebac94) error 0x3 07.07.09 11:51:31 ReportCrash[89417] Failed to create VMUSymbolicator for memcheck-x86-darwin[89416] 07.07.09 11:51:31 ReportCrash[89417] Formulating crash report for process memcheck-x86-darwin[89416] The crash reporter, as seen above, saves a report that contains the following information: Process: memcheck-x86-darwin [89416] Path: /usr/local/lib/valgrind/memcheck-x86-darwin Identifier: memcheck-x86-darwin Version: ??? (???) Code Type: X86 (Native) Parent Process: bash [28327] Date/Time: 2009-07-07 11:51:31.318 +0200 OS Version: Mac OS X 10.5.7 (9J3050) Report Version: 6 Anonymous UUID: 0B0F598F-4D68-4975-9CA5-DA6F34007C4C Exception Type: EXC_BREAKPOINT (SIGTRAP) Exception Codes: 0x0000000000000002, 0x0000000000000000 Crashed Thread: Unknown Backtrace not available Unknown thread crashed with X86 Thread State (32-bit): eax: 0x00000000 ebx: 0xf0adc286 ecx: 0x00000000 edx: 0xf0a985a8 edi: 0xf0ae59b0 esi: 0xf0adc278 ebp: 0xf007f4f8 esp: 0xf007f4e0 ss: 0x0000001f efl: 0x00000246 eip: 0xf0adc2a0 cs: 0x00000017 ds: 0x0000001f es: 0x0000001f fs: 0x00000000 gs: 0x00000037 cr2: 0xf0a985a8 Binary images description not available The described behavior is also encountered when invoking valgrind to use other tools like cachegrind. The system logs then show: 07.07.09 11:54:52 kernel shared_region: enter(0x123b0e88,0x7c48058,0,7,0): vm_map_enter(0x90000000,0x20000000,0x6ebac94) error 0x3 07.07.09 11:54:52 ReportCrash[89429] Failed to create VMUSymbolicator for cachegrind-x86-darwin[89428] 07.07.09 11:54:52 ReportCrash[89429] Formulating crash report for process cachegrind-x86-darwin[89428] So the problem seems to be unrelated to a specific valgrind tool. I'm using a MacBook Pro (mid 2009) with Core 2 Duo 2.66GHz processor and 4GB of RAM running Mac OS X 10.5.7. I have attached the contents of the file config.log after running "./configure". I have skimmed through the open bugs section but did not find anything similar, so I hope this is not a duplicate. Cheers, Jens