|
From: Michele G. <mi...@gm...> - 2008-12-23 12:44:17
|
Hi, I'm trying to use valgrind with maya (a 3d animation software): I developed a plugin that I want valgrind to check for memory lacks. That's the error i get: ==17112== Memcheck, a memory error detector. ==17112== Copyright (C) 2002-2008, and GNU GPL'd, by Julian Seward et al. ==17112== Using LibVEX rev 1878, a library for dynamic binary translation. ==17112== Copyright (C) 2004-2008, and GNU GPL'd, by OpenWorks LLP. ==17112== Using valgrind-3.4.0.SVN, a dynamic binary instrumentation framework. ==17112== Copyright (C) 2000-2008, and GNU GPL'd, by Julian Seward et al. ==17112== For more details, rerun with: -v ==17112== valgrind: m_debuginfo/debuginfo.c:527 (check_CFSI_related_invariants): Assertion 'di->cfsi_minavma >= di->rx_map_avma' failed. ==17112== at 0x3802810C: report_and_quit (m_libcassert.c:140) ==17112== by 0x3802841A: vgPlain_assert_fail (m_libcassert.c:205) ==17112== by 0x38046A5A: vgPlain_di_notify_mmap (debuginfo.c:527) ==17112== by 0x38085B5E: vgModuleLocal_generic_PRE_sys_mmap (syswrap-generic.c:1974) ==17112== by 0x3808E042: vgSysWrap_amd64_linux_sys_mmap_before (syswrap-amd64-linux.c:958) ==17112== by 0x3804C4E6: vgPlain_client_syscall (syswrap-main.c:942) ==17112== by 0x3804A39D: vgPlain_scheduler (scheduler.c:824) ==17112== by 0x3805C32E: run_a_thread_NORETURN (syswrap-linux.c:89) ==17112== by 0xFFFFFFFFFFFFFFFF: ??? ==17112== by 0xDEADBEEFDEADBEEE: ??? ==17112== by 0x20: ??? ==17112== by 0x406A96338: ??? ==17112== by 0x406A962D8: ??? ==17112== by 0x38027DAB: bm_qsort (m_libcbase.c:612) ==17112== by 0xFFFFFFFFFFFFFFE0: ??? ==17112== by 0xFFFFFFFFFFFFFFE0: ??? ==17112== by 0x40: ??? ==17112== by 0x38097900: compare_DiCfSI (storage.c:1268) ==17112== by 0x4: ??? ==17112== by 0x406A94E58: ??? ==17112== by 0x406A96258: ??? ==17112== by 0x380396E5: vgPlain_do_syscall (m_syscall.c:531) sched status: running_tid=1 Thread 1: status = VgTs_Runnable ==17112== at 0x401072C: (within /lib/ld-2.3.6.so) ==17112== by 0x4005122: (within /lib/ld-2.3.6.so) ==17112== by 0x4006D6C: (within /lib/ld-2.3.6.so) ==17112== by 0x4009EEC: (within /lib/ld-2.3.6.so) ==17112== by 0x400B7D0: (within /lib/ld-2.3.6.so) ==17112== by 0x400A5C3: (within /lib/ld-2.3.6.so) ==17112== by 0x40028FB: (within /lib/ld-2.3.6.so) ==17112== by 0x400F7DF: (within /lib/ld-2.3.6.so) ==17112== by 0x400127C: (within /lib/ld-2.3.6.so) ==17112== by 0x4000A87: (within /lib/ld-2.3.6.so) ==17112== by 0x1: ??? ==17112== by 0x7FF000892: ??? ==17112== by 0x7FF0008B6: ??? Any help would be appreciated |
|
From: Julian S. <js...@ac...> - 2008-12-23 12:59:15
|
This is a bug in Valgrind. > valgrind: m_debuginfo/debuginfo.c:527 (check_CFSI_related_invariants): > Assertion 'di->cfsi_minavma >= di->rx_map_avma' failed. This happened when Valgrind was reading debugging information from a shared object (.so), yes? You can check by rerunning with -v. If so, can you compress the .so with bzip2 and send it to me in private email? J |
|
From: Michele G. <mi...@gm...> - 2008-12-23 15:43:48
|
I've sent the file... Hope this help On Tue, Dec 23, 2008 at 1:45 PM, Julian Seward <js...@ac...> wrote: > > This is a bug in Valgrind. > > > valgrind: m_debuginfo/debuginfo.c:527 (check_CFSI_related_invariants): > > Assertion 'di->cfsi_minavma >= di->rx_map_avma' failed. > > This happened when Valgrind was reading debugging information from a > shared object (.so), yes? You can check by rerunning with -v. > > If so, can you compress the .so with bzip2 and send it to me in private > email? > > J > > ------------------------------------------------------------------------------ > _______________________________________________ > Valgrind-users mailing list > Val...@li... > https://lists.sourceforge.net/lists/listinfo/valgrind-users |
|
From: Michele G. <mi...@gm...> - 2008-12-24 08:36:06
|
Wow thats great! Now it works perfectly Thanks very much! Really appreciated On Tue, Dec 23, 2008 at 7:39 PM, Julian Seward <js...@ac...> wrote: > > Looks like you got lucky in the "priority support" department. Try the > attached patch and let me know ASAP if it does/does not fix the problem. > Certainly it fixes the problem with the .so you sent. > > J > > > On Tuesday 23 December 2008, Wayne Arnold wrote: >> If you need more info on Maya itself and Valgrind, I may be able to help >> >> I know the latest Nvidia drivers 180.xx don't mix with Maya and valgrind, >> reported to Nvidia, I haven't had a resolution yet. As well the newer Intel >> compilers 10.1.x/11.0 can have problems with debug symbol generation, with >> -O3 -g compile options >> >> >> >> -----Original Message----- >> From: Michele Gregori [mailto:mi...@gm...] >> Sent: Tuesday, December 23, 2008 10:44 AM >> To: val...@li... >> Subject: Re: [Valgrind-users] valgrind and maya >> >> I've sent the file... Hope this help >> >> On Tue, Dec 23, 2008 at 1:45 PM, Julian Seward <js...@ac...> wrote: >> > This is a bug in Valgrind. >> > >> > > valgrind: m_debuginfo/debuginfo.c:527 (check_CFSI_related_invariants): >> > > Assertion 'di->cfsi_minavma >= di->rx_map_avma' failed. >> > >> > This happened when Valgrind was reading debugging information from a >> > shared object (.so), yes? You can check by rerunning with -v. >> > >> > If so, can you compress the .so with bzip2 and send it to me in private >> > email? >> > >> > J >> > >> > ------------------------------------------------------------------------- >> >----- _______________________________________________ >> > Valgrind-users mailing list >> > Val...@li... >> > https://lists.sourceforge.net/lists/listinfo/valgrind-users >> >> --------------------------------------------------------------------------- >>--- _______________________________________________ >> Valgrind-users mailing list >> Val...@li... >> https://lists.sourceforge.net/lists/listinfo/valgrind-users > > > |
|
From: Michele G. <mi...@gm...> - 2008-12-24 08:37:52
|
The problem was the .so library... Thanks anyway :) On Tue, Dec 23, 2008 at 5:11 PM, Wayne Arnold <Way...@au...> wrote: > If you need more info on Maya itself and Valgrind, I may be able to help > > I know the latest Nvidia drivers 180.xx don't mix with Maya and valgrind, reported to Nvidia, I haven't had a resolution yet. > As well the newer Intel compilers 10.1.x/11.0 can have problems with debug symbol generation, with -O3 -g compile options > > > > -----Original Message----- > From: Michele Gregori [mailto:mi...@gm...] > Sent: Tuesday, December 23, 2008 10:44 AM > To: val...@li... > Subject: Re: [Valgrind-users] valgrind and maya > > I've sent the file... Hope this help > > > On Tue, Dec 23, 2008 at 1:45 PM, Julian Seward <js...@ac...> wrote: >> >> This is a bug in Valgrind. >> >> > valgrind: m_debuginfo/debuginfo.c:527 (check_CFSI_related_invariants): >> > Assertion 'di->cfsi_minavma >= di->rx_map_avma' failed. >> >> This happened when Valgrind was reading debugging information from a >> shared object (.so), yes? You can check by rerunning with -v. >> >> If so, can you compress the .so with bzip2 and send it to me in private >> email? >> >> J >> >> ------------------------------------------------------------------------------ >> _______________________________________________ >> Valgrind-users mailing list >> Val...@li... >> https://lists.sourceforge.net/lists/listinfo/valgrind-users > > ------------------------------------------------------------------------------ > _______________________________________________ > Valgrind-users mailing list > Val...@li... > https://lists.sourceforge.net/lists/listinfo/valgrind-users > |