Summary: | kded crashes when pressing eject on CD/DVD player while CD/DVD is mounted | ||
---|---|---|---|
Product: | [Frameworks and Libraries] kdelibs | Reporter: | Naga <nagatoro> |
Component: | kded | Assignee: | David Faure <faure> |
Status: | RESOLVED FIXED | ||
Severity: | crash | CC: | development--2015--bugs.kde.org, inspirra, isj-bugzilla, maxim |
Priority: | NOR | ||
Version: | unspecified | ||
Target Milestone: | --- | ||
Platform: | Gentoo Packages | ||
OS: | Linux | ||
Latest Commit: | Version Fixed In: |
Description
Naga
2007-01-26 15:55:03 UTC
Same behaviour with Ubuntu Edgy and KDE 3.5.6. Problem appears here too: Gentoo KDE-3.5.6 hal-0.5.9 dbus-1.0.2 Kernel: 2.6.19-suspend2 The backtrace seems to be unusable here although it's a debug build, but valgrind looks better: elias@nx9420 ~ $ valgrind --tool=memcheck kded ==28051== Memcheck, a memory error detector. ==28051== Copyright (C) 2002-2006, and GNU GPL'd, by Julian Seward et al. ==28051== Using LibVEX rev 1658, a library for dynamic binary translation. ==28051== Copyright (C) 2004-2006, and GNU GPL'd, by OpenWorks LLP. ==28051== Using valgrind-3.2.1, a dynamic binary instrumentation framework. ==28051== Copyright (C) 2000-2006, and GNU GPL'd, by Julian Seward et al. ==28051== For more details, rerun with: -v ==28051== ==28052== Syscall param write(buf) points to uninitialised byte(s) ==28052== at 0x54C4C83: __write_nocancel (in /lib/libpthread-2.5.so) ==28052== by 0x53EE7C7: (within /usr/lib/libX11.so.6.2.0) ==28052== Address 0x582BF58 is 184 bytes inside a block of size 16,384 alloc'd ==28052== at 0x4020753: calloc (in /usr/lib/valgrind/x86-linux/vgpreload_memcheck.so) ==28052== by 0x53D4FA5: XOpenDisplay (in /usr/lib/libX11.so.6.2.0) ==28051== ==28051== ERROR SUMMARY: 0 errors from 0 contexts (suppressed: 7 from 1) ==28051== malloc/free: in use at exit: 49,843 bytes in 2,467 blocks. ==28051== malloc/free: 16,471 allocs, 14,004 frees, 316,438 bytes allocated. ==28051== For counts of detected errors, rerun with: -v ==28051== searching for pointers to 2,467 not-freed blocks. ==28051== checked 774,256 bytes. ==28051== ==28051== LEAK SUMMARY: ==28051== definitely lost: 0 bytes in 0 blocks. ==28051== possibly lost: 0 bytes in 0 blocks. ==28051== still reachable: 49,843 bytes in 2,467 blocks. ==28051== suppressed: 0 bytes in 0 blocks. ==28051== Reachable blocks (those to which a pointer was found) are not shown. ==28051== To see them, rerun with: --show-reachable=yes elias@nx9420 ~ $ ==28055== ==28055== ERROR SUMMARY: 6 errors from 1 contexts (suppressed: 7 from 1) ==28055== malloc/free: in use at exit: 900,104 bytes in 14,642 blocks. ==28055== malloc/free: 154,525 allocs, 139,883 frees, 7,441,019 bytes allocated. ==28055== For counts of detected errors, rerun with: -v ==28055== searching for pointers to 14,642 not-freed blocks. ==28055== checked 1,604,344 bytes. ==28055== ==28055== LEAK SUMMARY: ==28055== definitely lost: 6,372 bytes in 164 blocks. ==28055== possibly lost: 68 bytes in 1 blocks. ==28055== still reachable: 893,664 bytes in 14,477 blocks. ==28055== suppressed: 0 bytes in 0 blocks. ==28055== Use --leak-check=full to see details of leaked memory. ==28052== ==28052== Syscall param writev(vector[...]) points to uninitialised byte(s) ==28052== at 0x569442E: (within /lib/libc-2.5.so) ==28052== by 0x53EE761: (within /usr/lib/libX11.so.6.2.0) ==28052== Address 0x582BF9C is 252 bytes inside a block of size 16,384 alloc'd ==28052== at 0x4020753: calloc (in /usr/lib/valgrind/x86-linux/vgpreload_memcheck.so) ==28052== by 0x53D4FA5: XOpenDisplay (in /usr/lib/libX11.so.6.2.0) ==28052== Warning: noted but unhandled ioctl 0x5327 with no size/direction hints ==28052== This could cause spurious value errors to appear. ==28052== See README_MISSING_SYSCALL_OR_IOCTL for guidance on writing a proper wrapper. ==28052== Warning: noted but unhandled ioctl 0x5327 with no size/direction hints ==28052== This could cause spurious value errors to appear. ==28052== See README_MISSING_SYSCALL_OR_IOCTL for guidance on writing a proper wrapper. ==28052== Warning: noted but unhandled ioctl 0x5327 with no size/direction hints ==28052== This could cause spurious value errors to appear. ==28052== See README_MISSING_SYSCALL_OR_IOCTL for guidance on writing a proper wrapper. process 28052: arguments to dbus_message_new_method_call() were incorrect, assertion "_dbus_check_is_valid_path (path)" failed in file dbus-message.c line 1074. This is normally a bug in some application using the D-Bus library. /usr/lib/libdbus-1.so.3 [0x672d8ee] Last DCOP call before KDED crash was from application 'kded' to object 'KDirNotify-4', function 'FilesChanged(KURL::List)'. KCrash: Application 'kded' crashing... ==28052== ==28052== ERROR SUMMARY: 23 errors from 2 contexts (suppressed: 7 from 1) ==28052== malloc/free: in use at exit: 1,818,914 bytes in 39,035 blocks. ==28052== malloc/free: 660,318 allocs, 621,283 frees, 19,018,297 bytes allocated. ==28052== For counts of detected errors, rerun with: -v ==28052== searching for pointers to 39,035 not-freed blocks. ==28052== checked 2,468,500 bytes. ==28052== ==28052== LEAK SUMMARY: ==28052== definitely lost: 47,936 bytes in 361 blocks. ==28052== possibly lost: 68 bytes in 1 blocks. ==28052== still reachable: 1,770,910 bytes in 38,673 blocks. ==28052== suppressed: 0 bytes in 0 blocks. ==28052== Use --leak-check=full to see details of leaked memory. Tried to reproduce it several times. Everytime kded crashed, it was one of this both reasons: process 16426: arguments to dbus_message_new_method_call() were incorrect, assertion "_dbus_check_is_valid_path (path)" failed in file dbus-message.c line 1074. This is normally a bug in some application using the D-Bus library. /usr/lib/libdbus-1.so.3 [0x672d8ee] Last DCOP call before KDED crash was from application 'DCOPServer' to object '', function 'applicationRemoved(QCString)'. KCrash: Application 'kded' crashing... process 17530: arguments to dbus_message_new_method_call() were incorrect, assertion "_dbus_check_is_valid_path (path)" failed in file dbus-message.c line 1074. This is normally a bug in some application using the D-Bus library. /usr/lib/libdbus-1.so.3 [0x672d8ee] Last DCOP call before KDED crash was from application 'kded' to object 'KDirNotify-4', function 'FilesChanged(KURL::List)'. KCrash: Application 'kded' crashing... Sadly I don't know enough about getting more details using dbg. Maybe somebody can reproduce this or tell me, how to provide more details. Thank you! Elias P. Earlier, I had a similar mistake: http://bugs.kde.org/show_bug.cgi?id=143070 But in now mistake has disappeared. And it seemed to me, that this mistake was from for lines in a file /etc/fstab: =============================== //user@host/C$ /mount/point smbfs rw,noauto,-I192.168.2.5,-Ekoi8-r:cp866,-f666,-N 0 0 =============================== But I can and be mistaken, as it does not repeat any more. Probably, the mistake has ceased appears from for that that in FreeBSD, changes in a file "/etc/fstab" more are not processed. $ pkg_info -E hal\* dbus\* kde-3\* ; uname -rms dbus-1.0.2_1 dbus-glib-0.73 dbus-qt3-0.70 hal-0.5.8.20070403 hal-device-manager-0.5.8.20070403 kde-3.5.6 FreeBSD 6.2-STABLE i386 See bug 151038 ( https://bugs.kde.org/show_bug.cgi?id=151038 ). Naga, your HAL path is "dx:mntbackupmntbackup". It looks like you have a colon ':' somewhere in your fstab. Remove that colon, and your problem might go away. For the other affected users: Watch out for characters in the first and second column in your "/etc/fstab" file which do not match "[A-Za-z0-9_]". Try again with these characters removed. Well I do have a ":" in my fstab since I mount some dirs using nfs. ie: host:/path/to/dir syntax. I can consistently recreate the problem with KDE 3.5.7 (openSuse 10.3) I can confirm that removing any entries in /etc/fstab which do not match "[A-Za-z0-9_]" eleminates the crash. But that workaround is not usable because my /etc/fstab contains both suse-generated lines like: /dev/disk/by-id/scsi-SATA_SanDisk_SDCFX3-_103401D2697X4431-part6 swap ... and also NFS mounts like: isjfs-i1:/home /mnt/isjfs_home ... Recreated a problem on openSUSE 10.3 Absolutely right, the necessary conditions (as for me) are: 1. you need to burn cd/dvd with k3b once 2. you need to have network paths in your /etc/fstab Removing network paths is a TEMPORARY solution. Let's vote for this bug) *** Bug 150804 has been marked as a duplicate of this bug. *** Already fixed. *** Bug 151038 has been marked as a duplicate of this bug. *** *** Bug 143070 has been marked as a duplicate of this bug. *** This happens for me EVERY time I log out. I use startkde to start my kde session: startkde: Starting up... kbuildsycoca running... Desktop Layout OK No running windows found QApplication::postEvent: Unexpected null receiver startkde: Shutting down... klauncher: Exiting on signal 1 kded: ERROR: : couldn't create slave : No se puede dialogar con klauncher process 19965: arguments to dbus_message_new_method_call() were incorrect, assertion "_dbus_check_is_valid_path (path)" failed in file dbus-message.c line 1074. This is normally a bug in some application using the D-Bus library. D-Bus not built with -rdynamic so unable to print a backtrace Last DCOP call before KDED crash was from application 'DCOPServer' to object '', function 'applicationRemoved(QCString)'. KCrash: Application 'kded' crashing... Warning: connect() failed: : No existe el fichero o el directorio KCrash cannot reach kdeinit, launching directly. startkde: Running shutdown scripts... startkde: Done. It takes like 10 seconds or so to exit kde, and then it exits with this error. I saw Lubos's comment http://bugs.kde.org/show_bug.cgi?id=140668#c10 about the fact that this bug is now fixed, but I am wondering which version of KDE fixes the problem. In fact I am using KDE 3.5.8 and I run into the problem reported here. Maybe it was only fixed in KDE 3.5.9 ? More precisely, I am looking for the revision of the fix (somewhere in http://websvn.kde.org/). Is there a way to find it or could someone provide the information ? |