[Bug 258742] New: Kernel update during online update does not finish installation
https://bugzilla.novell.com/show_bug.cgi?id=258742 Summary: Kernel update during online update does not finish installation Product: openSUSE 10.3 Version: Alpha 2plus Platform: Other OS/Version: Other Status: NEW Severity: Major Priority: P5 - None Component: Installation AssignedTo: bnc-team-screening@forge.provo.novell.com ReportedBy: aj@novell.com QAContact: jsrain@novell.com CC: hmuelle@novell.com, visnov@novell.com If I ran the online update during installation, the machine reboots due to the kernel update test. But afterwards installation does not continue and directly boots into the system. I saw briefly a segmentation fault of yast after the boot :-( -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug, or are watching someone who is.
https://bugzilla.novell.com/show_bug.cgi?id=258742 ------- Comment #1 from aj@novell.com 2007-03-29 02:02 MST ------- Created an attachment (id=127263) --> (https://bugzilla.novell.com/attachment.cgi?id=127263&action=view) Screenshot of segmentation fault message -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug, or are watching someone who is.
https://bugzilla.novell.com/show_bug.cgi?id=258742 ------- Comment #2 from aj@novell.com 2007-03-29 02:26 MST ------- Created an attachment (id=127270) --> (https://bugzilla.novell.com/attachment.cgi?id=127270&action=view) yast2 log files -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug, or are watching someone who is.
https://bugzilla.novell.com/show_bug.cgi?id=258742 mhorvath@novell.com changed: What |Removed |Added ---------------------------------------------------------------------------- AssignedTo|bnc-team- |lslezak@novell.com |screening@forge.provo.novell| |.com | -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug, or are watching someone who is.
https://bugzilla.novell.com/show_bug.cgi?id=258742 lslezak@novell.com changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |lslezak@novell.com AssignedTo|lslezak@novell.com |sh@novell.com ------- Comment #3 from lslezak@novell.com 2007-04-02 03:11 MST -------
From the log:
clients/inst_auth.ycp is called, it registers Pkg callbacks and calls TargetInit, then it crashes in StartScanDb() callback: 2007-03-29 12:17:07 <1> e37(3911) [YCP] PackageCallbacks.ycp:1752 Scanning RPM DB... 2007-03-29 12:17:07 <3> e37(3911) [liby2] genericfrontend.cc(signal_handler):59 got signal 11 at YCP file PackageCallbacks.ycp: 1760 PackageCallbacks.ycp line 1760 contains UI::OpenDialog() call, it seems to be UI related. Andreas, could you reproduce it in ncurses UI or is it Qt specific? -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug, or are watching someone who is.
https://bugzilla.novell.com/show_bug.cgi?id=258742 ------- Comment #4 from aj@novell.com 2007-04-02 03:41 MST ------- I did not try ncurses - feel free to test it yourself ;-) -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug, or are watching someone who is.
https://bugzilla.novell.com/show_bug.cgi?id=258742 sh@novell.com changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |NEEDINFO Info Provider| |lslezak@novell.com ------- Comment #5 from sh@novell.com 2007-04-02 04:18 MST ------- The log line does not really give very much information. I don't see how an UI::OpenDialog() call could possibly cause a segfault there. It is no different than all other UI::OpenDialog() calls (which do not cause any such problem), of which we have dozens near that place. With the same reasoning we might file a bug report against the Qt libs, Xlib, glibc or gcc, every single maintainer of which will want an isolated test case. The library versions (of the Qt UI) seem to be the correct ones, otherwise you wouldn't even have come this far. But given the complexity of that callback mechanism and everything, there might be any amount of memory corruption happening before that which in turn causes one of the next functions to crash, which might be an UI call like OpenDialog() or anything else. On average, about 99% of all bug reports that claim the UI causes a crash or some strange bug turn out to be the fault of something completely different, yet they get assigned to me as the UI maintainer so I constantly have to prove that it's not the UI's fault. I believe it's a UI problem if and only if somebody brings positive proof. A minimal test case would be such evidence. But there is no way I can drop everything else and begin debugging something as convoluted as this peculiar update scenario 3 times a week every week. I am spending way too much time on things like that. I never get to do anything else like the development work I am supposed to do. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug, or are watching someone who is.
https://bugzilla.novell.com/show_bug.cgi?id=258742 sh@novell.com changed: What |Removed |Added ---------------------------------------------------------------------------- Info Provider|lslezak@novell.com |aj@novell.com ------- Comment #6 from sh@novell.com 2007-04-02 04:24 MST ------- The logs don't even contain debug log lines from the UI, the Qt UI or the YCP interpreter. It might be anything. Somebody please try to reproduce this with full debug logging and core dumps enabled. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug, or are watching someone who is.
https://bugzilla.novell.com/show_bug.cgi?id=258742 ------- Comment #7 from sh@novell.com 2007-04-03 05:18 MST ------- According to the y2logs, the Qt UI starts up correctly in the same process (PID 3911) that later segfaults: 2007-03-29 12:16:50 <2> linux-18uu(3911) [qt-ui] YQUI_x11.cc(loadPredefinedQtTranslations):434 Can't load translations for predefined Qt dialogs from /usr/lib/qt3/translations/qt_en.qm 2007-03-29 12:16:50 <1> linux-18uu(3911) [qt-ui] YQUI_x11.cc(setLangFonts):471 /usr/share/YaST2/data/lang_fonts read OK 2007-03-29 12:16:50 <1> linux-18uu(3911) [qt-ui] YQUI_x11.cc(setLangFonts):494 Using fallback for en: font = "Sans Serif" 2007-03-29 12:17:07 <1> e37(3911) [YCP] PackageCallbacks.ycp:1752 Scanning RPM DB... 2007-03-29 12:17:07 <3> e37(3911) [liby2] genericfrontend.cc(signal_handler):59 got signal 11 at YCP file PackageCallbacks.ycp:1760 -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug, or are watching someone who is.
https://bugzilla.novell.com/show_bug.cgi?id=258742 ------- Comment #8 from mmichna@novell.com 2007-04-03 05:31 MST ------- I would like to reproduce it but the information on how to reproduce this is missing. Is this a update from 10.2 or a installation from scratch. Which hardware was it? Gazillions of open questions to be answered before we can proceed! -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug, or are watching someone who is.
https://bugzilla.novell.com/show_bug.cgi?id=258742 ------- Comment #9 from sh@novell.com 2007-04-03 08:04 MST ------- Given the very limited amount of information in the bug description, trying to puzzle together what happened from the y2logs looks like this: - Initial installation is started, new installation, KDE desktop, time zone is set, most values left as defaults (-> macro_inst_initial.ycp) - 848 Packages are installed from ftp://dist/%2finstall/SLP/openSUSE-10.3-Alpha2plus-DVD/x86_64/DVD1 without problem, in particular withour segfault from PackageCallbacks.ycp (which is heavily used during package installation) - The second part of the installation ("continue mode") is started (PID 3959): 2007-03-28 22:37:35 <1> linux(3959) [liby2] genericfrontend.cc(main):256 Launched YaST2 component 'y2base' 'installation' '("continue")' 'qt' No more packages are left to be installed. - Root password, network etc. is configured. - Initial online update during installation is invoked. These packages/patches are marked for installation from http://gd.tuwien.ac.at/linux/suse.com/suse/update/10.3 : update-test-affects-package-manager-99-99.noarch update-test-interactive-99.99-99.99.noarch update-test-reboot-needed-99.99-99.99.noarch update-test-trival-99.99-99.99.noarch update-test-security-99.99-99.99.noarch One patch from these is installed (without problem - no segfault): update-test-affects-package-manager-99-99.noarch - Process 3959 terminates regularly because that patch has the "package management restart required after this patch" flag set. - The second part of the installation is started again (PID 6402): 2007-03-29 14:47:39 <1> e37(6402) [liby2] genericfrontend.cc(main):256 Launched YaST2 component 'y2base' 'installation' '("continue")' 'qt' - The leftover patches are installed (whithout problem - no segfault): update-test-interactive-99.99-99.99.noarch update-test-reboot-needed-99.99-99.99.noarch update-test-trival-99.99-99.99.noarch update-test-security-99.99-99.99.noarch - Process 6402 terminates regularly. - Since one of the patches (presumably update-test-reboot-needed) has the "need to reboot" flag set, the machine reboots or is manually rebooted (we can tell from the PIDs that again have lower numbers: 6402 -> 3911) - After reboot, the second part of the installation is started again, now for the third time: 2007-03-29 12:16:45 <1> linux-18uu(3911) [liby2] genericfrontend.cc(main):256 Launched YaST2 component 'y2base' 'installation' '("continue")' 'qt' - The Qt UI starts up regularly: 2007-03-29 12:16:47 <1> linux-18uu(3911) [qt-ui] YQUI_core.cc(processCommandLineArgs):240 Qt argument: qt ... 2007-03-29 12:16:48 <1> linux-18uu(3911) [qt-ui] YQUI_x11.cc(headingFont):190 Loaded 12 pixel bold font: Sans Serif,-1,12,5,75,0,0,0,0,0 - The last life signs of that process are: 2007-03-29 12:17:07 <1> e37(3911) [YCP] PackageCallbacks.ycp:1752 Scanning RPM DB... - Segfault: 2007-03-29 12:17:07 <3> e37(3911 [liby2]genericfrontend.cc(signal_handler):59 got signal 11 at YCP file PackageCallbacks.ycp:1760 -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug, or are watching someone who is.
https://bugzilla.novell.com/show_bug.cgi?id=258742 ------- Comment #10 from sh@novell.com 2007-04-03 08:14 MST ------- So that piece of code gets invoked no less than five (5) times in a row: - Initial package installation - Install leftover packages (even though there are none left, the dialog is still opened - a very old quirk that a number of users wanted to see changed several times) - Installing patch update-test-affects-package-manager - Installing the leftover patches - Installing - uh - whatever (why is it called again anyway? There should be nothing left to install) The first four invocations work without a problem. The 5th segfaults. There are no packages installed between the 4th and the 5th invocation, yet the 5th for mysterious reasons segfaults. Now somebody please explain why a generic call like UI::OpenDialog() would segfault under those circumstances. For all we know it might be the kernel; after all, the machine was rebooted. There is no substantial evidence whatsoever that points to a general UI problem. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug, or are watching someone who is.
https://bugzilla.novell.com/show_bug.cgi?id=258742 aj@novell.com changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEEDINFO |NEW Info Provider|aj@novell.com | ------- Comment #11 from aj@novell.com 2007-04-05 02:37 MST ------- What exactly do you need to know from me? Please set back to NEEDINFO and ask me a concrete question if still needed. Stefan your analysis in comment 9 on what's going on is correct. This is a clean NEW installation of openSUSE 10.3 Alpha2plus. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug, or are watching someone who is.
https://bugzilla.novell.com/show_bug.cgi?id=258742 ------- Comment #12 from aj@novell.com 2007-04-05 02:40 MST ------- The alpha2plus kernel had one strange behaviour in exit handling, I'm testing stable now to see whether I can reproduce this still. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug, or are watching someone who is.
https://bugzilla.novell.com/show_bug.cgi?id=258742 ------- Comment #15 from aj@novell.com 2007-04-05 03:52 MST ------- Btw. just tried current factory: It does not run the online update at all, so I cannot say whether it's fixed or not. If you think it's a kernel bug, feel free to set to NEEDINFO and I'll test next week again with Alpha3. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug, or are watching someone who is.
https://bugzilla.novell.com/show_bug.cgi?id=258742 ------- Comment #16 from sh@novell.com 2007-04-05 07:59 MST ------- Created an attachment (id=129350) --> (https://bugzilla.novell.com/attachment.cgi?id=129350&action=view) Reproduced problem on new i386 machine: y2logs.tar.bz2 I just reproduced the problem on new i386 machine. I meant to attach gdb to that y2base process, but there is no chance to start a debugger (or even just to log in) before the crash occurs. It happens immediately after reboot, even before there is a login prompt. I managed to supply Y2DEBUG=1 on the kernel command line of the boot just before the crash, however, so the last part of the y2logs are a little bit more verbose. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug, or are watching someone who is.
https://bugzilla.novell.com/show_bug.cgi?id=258742 ------- Comment #17 from sh@novell.com 2007-04-05 08:06 MST ------- The last lines of that y2log minus all [zypp] log lines: 2007-04-05 15:49:25 <0> e156(4477) [libycp] ExecutionEnvironment.cc(pushframe):78 Push frame CallbackProgressScanDb ("PackageCallbacks::ProgressScanDb") 2007-04-05 15:49:25 <1> e156(4477) [wfm] PackageCallbacksInit.ycp:40 Pkg Builtin called: CallbackProgressScanDb 2007-04-05 15:49:25 <0> e156(4477) [wfm] Callbacks.YCP.h(setCallback):227 Registering callback PackageCallbacks::ProgressScanDb 2007-04-05 15:49:25 <0> e156(4477) [liby2] Y2ComponentBroker.cc(getNamespaceComponent):108 getNamespace (PackageCallbacks) 2007-04-05 15:49:25 <0> e156(4477) [Y2Perl] Y2CCPerl.cc(provideNamespace):34 Y2CCPerl::provideNamespace PackageCallbacks 2007-04-05 15:49:25 <0> e156(4477) [libycp] pathsearch.cc(find):319 trying /y2update/modules/PackageCallbacks.pm 2007-04-05 15:49:25 <0> e156(4477) [libycp] pathsearch.cc(find):319 trying /root/.yast2/modules/PackageCallbacks.pm 2007-04-05 15:49:25 <0> e156(4477) [libycp] pathsearch.cc(find):319 trying /usr/share/YaST2/modules/PackageCallbacks.pm 2007-04-05 15:49:25 <0> e156(4477) [liby2] Y2CCPlugin.cc(provideNamespace):127 Y2PluginComponent tries to locate namespace 'PackageCallbacks' 2007-04-05 15:49:25 <0> e156(4477) [libycp] pathsearch.cc(findy2plugin):217 Testing existence of plugin /y2update/plugin/libpy2PackageCallbacks.so.2 2007-04-05 15:49:25 <0> e156(4477) [libycp] pathsearch.cc(findy2plugin):217 Testing existence of plugin Y2DIR/plugin/libpy2PackageCallbacks.so.2 2007-04-05 15:49:25 <0> e156(4477) [libycp] pathsearch.cc(findy2plugin):217 Testing existence of plugin /root/.yast2/plugin/libpy2PackageCallbacks.so.2 2007-04-05 15:49:25 <0> e156(4477) [libycp] pathsearch.cc(findy2plugin):217 Testing existence of plugin /usr/lib/YaST2/plugin/libpy2PackageCallbacks.so.2 2007-04-05 15:49:25 <0> e156(4477) [liby2] Y2CCPlugin.cc(provideNamespace):127 Y2PluginComponent tries to locate namespace 'PackageCallbacks' 2007-04-05 15:49:25 <0> e156(4477) [libycp] pathsearch.cc(findy2plugin):217 Testing existence of plugin /y2update/plugin/libpy2PackageCallbacks.so.2 2007-04-05 15:49:25 <0> e156(4477) [libycp] pathsearch.cc(findy2plugin):217 Testing existence of plugin Y2DIR/plugin/libpy2PackageCallbacks.so.2 2007-04-05 15:49:25 <0> e156(4477) [libycp] pathsearch.cc(findy2plugin):217 Testing existence of plugin /root/.yast2/plugin/libpy2PackageCallbacks.so.2 2007-04-05 15:49:25 <0> e156(4477) [libycp] pathsearch.cc(findy2plugin):217 Testing existence of plugin /usr/lib/YaST2/plugin/libpy2PackageCallbacks.so.2 2007-04-05 15:49:25 <0> e156(4477) [libycp] pathsearch.cc(find):319 trying /y2update/modules/PackageCallbacks.ybc 2007-04-05 15:49:25 <0> e156(4477) [libycp] pathsearch.cc(find):319 trying /root/.yast2/modules/PackageCallbacks.ybc 2007-04-05 15:49:25 <0> e156(4477) [libycp] pathsearch.cc(find):319 trying /usr/share/YaST2/modules/PackageCallbacks.ybc 2007-04-05 15:49:25 <0> e156(4477) [libycp] pathsearch.cc(find):322 ... success 2007-04-05 15:49:25 <0> e156(4477) [wfm] Y2CCWFM.cc(provideNamespace):203 Component to provide the namespace: 0x80931c8 2007-04-05 15:49:25 <0> e156(4477) [liby2] Y2ComponentBroker.cc(getNamespaceComponent):160 Component 0x80931c8 used for namespace PackageCallbacks 2007-04-05 15:49:25 <1> e156(4477) [wfm] Y2WFMComponent.cc(import):696 Y2WFMComponent::import (PackageCallbacks) 2007-04-05 15:49:25 <0> e156(4477) [libycp] pathsearch.cc(find):319 trying /y2update/modules/PackageCallbacks.ybc 2007-04-05 15:49:25 <0> e156(4477) [libycp] pathsearch.cc(find):319 trying /root/.yast2/modules/PackageCallbacks.ybc 2007-04-05 15:49:25 <0> e156(4477) [libycp] pathsearch.cc(find):319 trying /usr/share/YaST2/modules/PackageCallbacks.ybc 2007-04-05 15:49:25 <0> e156(4477) [libycp] pathsearch.cc(find):322 ... success 2007-04-05 15:49:25 <0> e156(4477) [libycp] ExecutionEnvironment.cc(popframe):93 Pop frame 0x8ed8a30 2007-04-05 15:49:25 <0> e156(4477) [libycp] ExecutionEnvironment.cc(pushframe):78 Push frame CallbackErrorScanDb ("PackageCallbacks::ErrorScanDb") 2007-04-05 15:49:25 <1> e156(4477) [wfm] PackageCallbacksInit.ycp:41 Pkg Builtin called: CallbackErrorScanDb 2007-04-05 15:49:25 <0> e156(4477) [wfm] Callbacks.YCP.h(setCallback):227 Registering callback PackageCallbacks::ErrorScanDb 2007-04-05 15:49:25 <0> e156(4477) [liby2] Y2ComponentBroker.cc(getNamespaceComponent):108 getNamespace (PackageCallbacks) 2007-04-05 15:49:25 <0> e156(4477) [Y2Perl] Y2CCPerl.cc(provideNamespace):34 Y2CCPerl::provideNamespace PackageCallbacks 2007-04-05 15:49:25 <0> e156(4477) [libycp] pathsearch.cc(find):319 trying /y2update/modules/PackageCallbacks.pm 2007-04-05 15:49:25 <0> e156(4477) [libycp] pathsearch.cc(find):319 trying /root/.yast2/modules/PackageCallbacks.pm 2007-04-05 15:49:25 <0> e156(4477) [libycp] pathsearch.cc(find):319 trying /usr/share/YaST2/modules/PackageCallbacks.pm 2007-04-05 15:49:25 <0> e156(4477) [liby2] Y2CCPlugin.cc(provideNamespace):127 Y2PluginComponent tries to locate namespace 'PackageCallbacks' 2007-04-05 15:49:25 <0> e156(4477) [libycp] pathsearch.cc(findy2plugin):217 Testing existence of plugin /y2update/plugin/libpy2PackageCallbacks.so.2 2007-04-05 15:49:25 <0> e156(4477) [libycp] pathsearch.cc(findy2plugin):217 Testing existence of plugin Y2DIR/plugin/libpy2PackageCallbacks.so.2 2007-04-05 15:49:25 <0> e156(4477) [libycp] pathsearch.cc(findy2plugin):217 Testing existence of plugin /root/.yast2/plugin/libpy2PackageCallbacks.so.2 2007-04-05 15:49:25 <0> e156(4477) [libycp] pathsearch.cc(findy2plugin):217 Testing existence of plugin /usr/lib/YaST2/plugin/libpy2PackageCallbacks.so.2 2007-04-05 15:49:25 <0> e156(4477) [liby2] Y2CCPlugin.cc(provideNamespace):127 Y2PluginComponent tries to locate namespace 'PackageCallbacks' 2007-04-05 15:49:25 <0> e156(4477) [libycp] pathsearch.cc(findy2plugin):217 Testing existence of plugin /y2update/plugin/libpy2PackageCallbacks.so.2 2007-04-05 15:49:25 <0> e156(4477) [libycp] pathsearch.cc(findy2plugin):217 Testing existence of plugin Y2DIR/plugin/libpy2PackageCallbacks.so.2 2007-04-05 15:49:25 <0> e156(4477) [libycp] pathsearch.cc(findy2plugin):217 Testing existence of plugin /root/.yast2/plugin/libpy2PackageCallbacks.so.2 2007-04-05 15:49:25 <0> e156(4477) [libycp] pathsearch.cc(findy2plugin):217 Testing existence of plugin /usr/lib/YaST2/plugin/libpy2PackageCallbacks.so.2 2007-04-05 15:49:25 <0> e156(4477) [libycp] pathsearch.cc(find):319 trying /y2update/modules/PackageCallbacks.ybc 2007-04-05 15:49:25 <0> e156(4477) [libycp] pathsearch.cc(find):319 trying /root/.yast2/modules/PackageCallbacks.ybc 2007-04-05 15:49:25 <0> e156(4477) [libycp] pathsearch.cc(find):319 trying /usr/share/YaST2/modules/PackageCallbacks.ybc 2007-04-05 15:49:25 <0> e156(4477) [libycp] pathsearch.cc(find):322 ... success 2007-04-05 15:49:25 <0> e156(4477) [wfm] Y2CCWFM.cc(provideNamespace):203 Component to provide the namespace: 0x80931c8 2007-04-05 15:49:25 <0> e156(4477) [liby2] Y2ComponentBroker.cc(getNamespaceComponent):160 Component 0x80931c8 used for namespace PackageCallbacks 2007-04-05 15:49:25 <1> e156(4477) [wfm] Y2WFMComponent.cc(import):696 Y2WFMComponent::import (PackageCallbacks) 2007-04-05 15:49:25 <0> e156(4477) [libycp] pathsearch.cc(find):319 trying /y2update/modules/PackageCallbacks.ybc 2007-04-05 15:49:25 <0> e156(4477) [libycp] pathsearch.cc(find):319 trying /root/.yast2/modules/PackageCallbacks.ybc 2007-04-05 15:49:25 <0> e156(4477) [libycp] pathsearch.cc(find):319 trying /usr/share/YaST2/modules/PackageCallbacks.ybc 2007-04-05 15:49:25 <0> e156(4477) [libycp] pathsearch.cc(find):322 ... success 2007-04-05 15:49:25 <0> e156(4477) [libycp] ExecutionEnvironment.cc(popframe):93 Pop frame 0x8ed8a30 2007-04-05 15:49:25 <0> e156(4477) [libycp] ExecutionEnvironment.cc(pushframe):78 Push frame CallbackDoneScanDb ("PackageCallbacks::DoneScanDb") 2007-04-05 15:49:25 <1> e156(4477) [wfm] PackageCallbacksInit.ycp:42 Pkg Builtin called: CallbackDoneScanDb 2007-04-05 15:49:25 <0> e156(4477) [wfm] Callbacks.YCP.h(setCallback):227 Registering callback PackageCallbacks::DoneScanDb 2007-04-05 15:49:25 <0> e156(4477) [liby2] Y2ComponentBroker.cc(getNamespaceComponent):108 getNamespace (PackageCallbacks) 2007-04-05 15:49:25 <0> e156(4477) [Y2Perl] Y2CCPerl.cc(provideNamespace):34 Y2CCPerl::provideNamespace PackageCallbacks 2007-04-05 15:49:25 <0> e156(4477) [libycp] pathsearch.cc(find):319 trying /y2update/modules/PackageCallbacks.pm 2007-04-05 15:49:25 <0> e156(4477) [libycp] pathsearch.cc(find):319 trying /root/.yast2/modules/PackageCallbacks.pm 2007-04-05 15:49:25 <0> e156(4477) [libycp] pathsearch.cc(find):319 trying /usr/share/YaST2/modules/PackageCallbacks.pm 2007-04-05 15:49:25 <0> e156(4477) [liby2] Y2CCPlugin.cc(provideNamespace):127 Y2PluginComponent tries to locate namespace 'PackageCallbacks' 2007-04-05 15:49:25 <0> e156(4477) [libycp] pathsearch.cc(findy2plugin):217 Testing existence of plugin /y2update/plugin/libpy2PackageCallbacks.so.2 2007-04-05 15:49:25 <0> e156(4477) [libycp] pathsearch.cc(findy2plugin):217 Testing existence of plugin Y2DIR/plugin/libpy2PackageCallbacks.so.2 2007-04-05 15:49:25 <0> e156(4477) [libycp] pathsearch.cc(findy2plugin):217 Testing existence of plugin /root/.yast2/plugin/libpy2PackageCallbacks.so.2 2007-04-05 15:49:25 <0> e156(4477) [libycp] pathsearch.cc(findy2plugin):217 Testing existence of plugin /usr/lib/YaST2/plugin/libpy2PackageCallbacks.so.2 2007-04-05 15:49:25 <0> e156(4477) [liby2] Y2CCPlugin.cc(provideNamespace):127 Y2PluginComponent tries to locate namespace 'PackageCallbacks' 2007-04-05 15:49:25 <0> e156(4477) [libycp] pathsearch.cc(findy2plugin):217 Testing existence of plugin /y2update/plugin/libpy2PackageCallbacks.so.2 2007-04-05 15:49:25 <0> e156(4477) [libycp] pathsearch.cc(findy2plugin):217 Testing existence of plugin Y2DIR/plugin/libpy2PackageCallbacks.so.2 2007-04-05 15:49:25 <0> e156(4477) [libycp] pathsearch.cc(findy2plugin):217 Testing existence of plugin /root/.yast2/plugin/libpy2PackageCallbacks.so.2 2007-04-05 15:49:25 <0> e156(4477) [libycp] pathsearch.cc(findy2plugin):217 Testing existence of plugin /usr/lib/YaST2/plugin/libpy2PackageCallbacks.so.2 2007-04-05 15:49:25 <0> e156(4477) [libycp] pathsearch.cc(find):319 trying /y2update/modules/PackageCallbacks.ybc 2007-04-05 15:49:25 <0> e156(4477) [libycp] pathsearch.cc(find):319 trying /root/.yast2/modules/PackageCallbacks.ybc 2007-04-05 15:49:25 <0> e156(4477) [libycp] pathsearch.cc(find):319 trying /usr/share/YaST2/modules/PackageCallbacks.ybc 2007-04-05 15:49:25 <0> e156(4477) [libycp] pathsearch.cc(find):322 ... success 2007-04-05 15:49:25 <0> e156(4477) [wfm] Y2CCWFM.cc(provideNamespace):203 Component to provide the namespace: 0x80931c8 2007-04-05 15:49:25 <0> e156(4477) [liby2] Y2ComponentBroker.cc(getNamespaceComponent):160 Component 0x80931c8 used for namespace PackageCallbacks 2007-04-05 15:49:25 <1> e156(4477) [wfm] Y2WFMComponent.cc(import):696 Y2WFMComponent::import (PackageCallbacks) 2007-04-05 15:49:25 <0> e156(4477) [libycp] pathsearch.cc(find):319 trying /y2update/modules/PackageCallbacks.ybc 2007-04-05 15:49:25 <0> e156(4477) [libycp] pathsearch.cc(find):319 trying /root/.yast2/modules/PackageCallbacks.ybc 2007-04-05 15:49:25 <0> e156(4477) [libycp] pathsearch.cc(find):319 trying /usr/share/YaST2/modules/PackageCallbacks.ybc 2007-04-05 15:49:25 <0> e156(4477) [libycp] pathsearch.cc(find):322 ... success 2007-04-05 15:49:25 <0> e156(4477) [libycp] ExecutionEnvironment.cc(popframe):93 Pop frame 0x8ed8a30 2007-04-05 15:49:25 <0> e156(4477) [libycp] ExecutionEnvironment.cc(popframe):93 Pop frame 0x8f00550 2007-04-05 15:49:25 <0> e156(4477) [libycp] ExecutionEnvironment.cc(popframe):93 Pop frame 0x8eda8d8 2007-04-05 15:49:25 <0> e156(4477) [libycp] ExecutionEnvironment.cc(pushframe):78 Push frame TargetInit ("/", false) 2007-04-05 15:49:25 <1> e156(4477) [wfm] PackageSystem.ycp:61 Pkg Builtin called: TargetInit 2007-04-05 15:49:26 <0> e156(4477) [wfm] Callbacks.YCP.h(evaluate):383 Evaluating callback (registered funciton: StartScanDb) 2007-04-05 15:49:26 <1> e156(4477) [YCP] PackageCallbacks.ycp:1752 Scanning RPM DB... 2007-04-05 15:49:26 <0> e156(4477) [libycp] ExecutionEnvironment.cc(pushframe):78 Push frame Mode::commandline () 2007-04-05 15:49:26 <0> e156(4477) [libycp] ExecutionEnvironment.cc(pushframe):78 Push frame Mode::ui () 2007-04-05 15:49:26 <0> e156(4477) [libycp] ExecutionEnvironment.cc(popframe):93 Pop frame 0x8f241c8 2007-04-05 15:49:26 <0> e156(4477) [libycp] ExecutionEnvironment.cc(popframe):93 Pop frame 0x8f7b1d0 2007-04-05 15:49:26 <3> e156(4477) [liby2] genericfrontend.cc(signal_handler):59 got signal 11 at YCP file PackageCallbacks.ycp:1760 -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug, or are watching someone who is.
https://bugzilla.novell.com/show_bug.cgi?id=258742 ------- Comment #18 from lslezak@novell.com 2007-04-05 08:26 MST ------- Um, that probably doesn't help much - the callback detects whether it is running in command line mode, if not it calls UI::OpenDialog() (and crashes there). Stefan, you can try to modify the start script to start yast in gdb, I think that could help to locate the problem... -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug, or are watching someone who is.
https://bugzilla.novell.com/show_bug.cgi?id=258742 ------- Comment #19 from sh@novell.com 2007-04-05 08:47 MST ------- The problem reoccurs after each reboot, and subsequent attempts even open a popup before the segfault to ask the user whether or not the remaining part of the installation should be started once more. This is the chance to enable debug logging (Shift-F7) and attach gdb from an xterm (for easier copy&paste) with Ctrl-Shift-Alt-X. Here is the output of that gdb: gdb) c Continuing. (no debugging symbols found) (no debugging symbols found) (no debugging symbols found) Program received signal SIGSEGV, Segmentation fault. [Switching to Thread -1258611824 (LWP 4446)] 0xb7f7ad4e in typeinfo for PtrBase<SymbolEntry> () from /usr/lib/YaST2/plugin/libpy2wfm.so.2 (gdb) bt #0 0xb7f7ad4e in typeinfo for PtrBase<SymbolEntry> () from /usr/lib/YaST2/plugin/libpy2wfm.so.2 #1 0xb5604d80 in QObject::QObject () from /usr/lib/qt3/lib/libqt-mt.so.3 #2 0xb563ad5d in QWidget::QWidget () from /usr/lib/qt3/lib/libqt-mt.so.3 #3 0xb5b28cdd in YQDialog::YQDialog () from /usr/lib/YaST2/plugin/libpy2qt.so.2 #4 0xb5b0d92d in YQUI::createDialog () from /usr/lib/YaST2/plugin/libpy2qt.so.2 #5 0xb7eab885 in YUI::evaluateOpenDialog () from /usr/lib/libyui.so.2 #6 0xb7ed0c1a in main () from /usr/lib/libyui.so.2 (gdb) -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug, or are watching someone who is.
https://bugzilla.novell.com/show_bug.cgi?id=258742 ------- Comment #20 from sh@novell.com 2007-04-05 08:49 MST ------- I wouldn't know what could case a crash in the QObject constructor other than a massive memory / stack corruption just prior to that. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug, or are watching someone who is.
https://bugzilla.novell.com/show_bug.cgi?id=258742 ------- Comment #21 from sh@novell.com 2007-04-05 09:17 MST ------- Duh. I overlooked the first stack frame #0: #0 0xb7f7ad4e in typeinfo for PtrBase<SymbolEntry> () from /usr/lib/YaST2/plugin/libpy2wfm.so.2 Whatever that does. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug, or are watching someone who is.
https://bugzilla.novell.com/show_bug.cgi?id=258742 ------- Comment #22 from sh@novell.com 2007-04-05 10:31 MST ------- I tried several scenarios now. I exchanged that offending UI::OpenDialog() call with UI::BusyCursor(). In that case, UI::BusyCursor() (which really doesn't do very much) is executed without problems (proving that UI calls in general still work), but the next UI::OpenDialog() call crashes. I started a plain X server manually with fvwm and xterm and started some YaST2 modules from there. They all worked flawlessly, even "sw_single" and "online_update". They never crashed at any of their many UI::OpenDialog() calls. PackageCallbacks reported the progress while scanning the RPM DB without problems in the proper progress popups. Rebooting the machine again once more started the second part of the installation, and the crash is there allright in the same location as before. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug, or are watching someone who is.
https://bugzilla.novell.com/show_bug.cgi?id=258742 ------- Comment #23 from sh@novell.com 2007-04-05 11:01 MST ------- I hacked up the start scripts and tried those scenarios: - qt with --nothreads. Crashes just the same. - ncurses. Does not crash. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug, or are watching someone who is.
https://bugzilla.novell.com/show_bug.cgi?id=258742 ------- Comment #24 from sh@novell.com 2007-04-05 11:27 MST ------- When I start the second stage of the installation manually (with Qt) calling /usr/lib/YaST2/startup/YaST2.Second-Stage in runlevel 3 , it works flawlessly. No crash. When I do "init S" and do the same, it crashes. After starting services present in runlevel 3, but not S, I found that "rcnetwork start" fixed the problem. Double-checking this, I simply invoked this on console 2: ping 10.10.0.1 -> Network is unreachable rcnetwork start when the popup was still there asking me if I wanted to retry the remaining part of the installation. And voila, it works. I can only assume exactly what happened: - We did a network installation. All installation sources were on the network. - Network was not started at that time. Thus all installation sources were unreachable. - Some libzypp call (in yast2-pkg-bindings?) returned an error or an exception that was ignored. Subsequent calls corrupted stack and/or memory. - The Qt UI was just unlucky enough to do something with the corrupted memory area(s). -> Segfault -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug, or are watching someone who is.
https://bugzilla.novell.com/show_bug.cgi?id=258742 sh@novell.com changed: What |Removed |Added ---------------------------------------------------------------------------- AssignedTo|sh@novell.com |locilka@novell.com ------- Comment #25 from sh@novell.com 2007-04-05 11:32 MST ------- The part about memory corruption due to ignored error returns is speculative. But I know for sure that network should be enabled when a network installation is to be completed. That part seems to be missing in our start scripts; we make sure that network is enabled for the second stage of a network installation, but obviously not if the online update that might also occur there requires another reboot. That part is missing. -> locilka as yast2-installation maintainer It might also be a very good idea to have a close look at possible error results and uncaught or poorly handled C++ exceptions in pkg-bindings. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug, or are watching someone who is.
https://bugzilla.novell.com/show_bug.cgi?id=258742 lslezak@novell.com changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |mvidner@novell.com ------- Comment #26 from lslezak@novell.com 2007-04-06 00:49 MST ------- Stefan, thank you very much for the debugging! Pkg-bindings catch all unhandled libzypp exceptions on the global level (in Y2PkgFunction::evaluateCall()). The log should contain error message "Caught an unhandled exception" if that happens. Y2log (from the second stage) doesn't contain any THROW text, so no libzypp exception occurred. IIRC an unhandled C++ exception causes SIGABORT not SEGFAULT. It's strange that it works in ncurses, another strange thing is that Pkg::TargetInit() should not use network at all... Martin, any idea what could happen in #0 0xb7f7ad4e in typeinfo for PtrBase<SymbolEntry> () from /usr/lib/YaST2/plugin/libpy2wfm.so.2 ? -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug, or are watching someone who is.
https://bugzilla.novell.com/show_bug.cgi?id=258742 locilka@novell.com changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |ASSIGNED ------- Comment #27 from locilka@novell.com 2007-04-06 02:32 MST ------- Installation ============ Problem description: -------------------- We have some YaST startup scripts. These scripts start the network but ... only for VNC or SSH installation. For the rest, inst_netsetup (part of the network) starts the network interfaces just before inst_rpmcopy is called (YaST starts *WITHOUT* network). According the the report and additional comments, it seems, that the network is not started after the online_update and reboot because it doesn't use SSH or VNC and inst_netsetup is not called because network is already configured. Problem Solution: ----------------- * After the online update, when YaST is forced to reboot the machine, test, whether network is up or down, remember the value. * Reboot * Check whether the network was up or down and adjust it (start it if it was up), probably, `rcnetwork start` * Start YaST Command that might say whether some network is up: /sbin/ip addr show | /bin/grep "^[ \t]\+inet.*scope global " Martin, is that command correct? Please note, that this is only a solution for YaST startup, it doesn't fix that SEGFAULT. `rcnetwork start` might fail (for whatever reason) and YaST might drop SEGFAULT again in that case. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug, or are watching someone who is.
https://bugzilla.novell.com/show_bug.cgi?id=258742 ------- Comment #28 from locilka@novell.com 2007-04-06 05:36 MST ------- Well, the proposed solution is not enough. Client inst_netsetup (for instance) calls also `rcportmap start` in case of NFS installation source. Firewall might but needn't be running. We need to call such function (start network, start services) every time computer is rebooted and YaST is started from another than the first step. Simple calling `inst_netsetup` migh be sufficient for that. yast2-installation requires yast2-network so it would be probably safe. However inst_netsetup starts network only in case of network installation source, which is insufficient for DVD installation + network Add-On and/or network online-update source. So the solution still seems to be: When stopping YaST to reboot the computer: - Remember network status (portmap as well, and maybe firewall, ...) When starting YaST after reboot: - Adjust network (and portmap, etc...) - Do not stop any service, just start if needed Computer can be rebooted at any installation step because, e.g, Add-On products can add such restart... -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug, or are watching someone who is.
https://bugzilla.novell.com/show_bug.cgi?id=258742 locilka@novell.com changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |locilka@novell.com AssignedTo|locilka@novell.com |mvidner@novell.com Status|ASSIGNED |NEW ------- Comment #29 from locilka@novell.com 2007-04-06 08:58 MST ------- Changes in yast2-installation.rpm: - In case of reboot during installation, network services status is stored to a reboot_network_settings file and their status is restored again when starting the installation after reboot. Restoring the status uses Progress library for user feedback (#258742). - Adjusted RPM dependencies. - 2.15.18 Changes in yast2.rpm: - Adding new Installation::reboot_net_settings that will point to a file storing current network services settings when rebooting a computer during installation (#258742). - 2.15.25 Reassigning to Mvidner to debug the SEGFAULT issue. Have a nice day. That's my easter egg for you :)))))) ;) -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug, or are watching someone who is.
https://bugzilla.novell.com/show_bug.cgi?id=258742 ------- Comment #30 from mvidner@novell.com 2007-04-10 01:28 MST ------- Created an attachment (id=129953) --> (https://bugzilla.novell.com/attachment.cgi?id=129953&action=view) valgrind.tar.gz drwxr-xr-x 0 2007-04-10 09:14:15 logs/ -rw------- 37494 2007-04-06 13:08:46 logs/valgrind.log.nonet -rw------- 27769 2007-04-06 12:56:16 logs/valgrind.log.net drwxr-xr-x 0 2007-04-10 09:14:26 diffs/ -rw-r--r-- 474 2007-04-10 07:13:54 diffs/installation.ycp.diff -rw-r--r-- 717 2007-04-10 07:16:22 diffs/YaST2.call.diff We have two bugs mixed here: (A) network not being started in certain cases (B) segfault (conditioned by A). If (A) needs to be reopened, please use a separate bugzilla entry. Concerning the segfault it seems that the problem really is somewhere under yast2-qt (qt, xorg, ...?). I patched yast to run under valgrind, optionally with network. I also added an early call to the package manager in hope to trigger the bug but that did not work. When yast runs with network, there is a valgrind error early on but then it runs fine. If network is not enabled, another error occurs and presumably is the immediate cause of the segfault. I think that the first error, which is a misplaced write, corrupts some memory area that is used only in the case without network. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug, or are watching someone who is.
https://bugzilla.novell.com/show_bug.cgi?id=258742 mvidner@novell.com changed: What |Removed |Added ---------------------------------------------------------------------------- AssignedTo|mvidner@novell.com |sh@novell.com ------- Comment #31 from mvidner@novell.com 2007-04-10 01:32 MST ------- Qt or X... Stefan please debug or pass it on. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug, or are watching someone who is.
https://bugzilla.novell.com/show_bug.cgi?id=258742 ------- Comment #32 from sh@novell.com 2007-04-10 07:14 MST ------- AFAICS both valgrind logs are pretty much the same at the beginng. I don't see any difference in the first error you mentioned. Both logs show the same valgrind warnings at the beginning -- which may or may not be relevant, but considering that they occur in standard Qt-Lib start-up code that is executed in both cases (probably upon every UI initialization everywhere), I can't quite believe that they would be a problem. What unsettles me a lot more, though, is that one (only in valgrind.log.nonet): ==4356== Address 0x1CA7B818 is 0 bytes inside a block of size 45 free'd ==4356== at 0x4C2257D: operator delete(void*) (in /usr/lib64/valgrind/amd64-linux/vgpreload_memcheck.so) ==4356== by 0x7A57E1D: std::string::reserve(unsigned long) (in /usr/lib64/libstdc++.so.6.0.8) ==4356== by 0x7A57FDE: std::string::append(char const*, unsigned long) (in /usr/lib64/libstdc++.so.6.0.8) ==4356== by 0x5772341: YECall::toString() const (basic_string.h:822) ==4356== by 0x577DCAA: YEFunction::evaluate(bool) (YExpression.cc:2845) ==4356== by 0x578F471: YSAssign::evaluate(bool) (YStatement.cc:579) ==4356== by 0x57A73F5: YBlock::evaluate(bool) (YBlock.cc:558) ==4356== by 0x576EEC2: YCPCodeRep::evaluate(bool) const (YCPCode.cc:81) ==4356== by 0x4E33268: Y2WFMComponent::doActualWork(YCPList const&, Y2Component*) (Y2WFMComponent.cc:177) ==4356== by 0x4E30530: Y2WFMComponent::CallFunction(YCPString const&, YCPList const&) (Y2WFMComponent.cc:651) ==4356== by 0x4E38781: WFMCallFunction1(YCPString const&, YCPList const&) (WFM.cc:80) ==4356== by 0x577EB1E: YEBuiltin::evaluate(bool) (YExpression.cc:2308) ==4356== by 0x5782F6A: YEPropagate::evaluate(bool) (YExpression.cc:1050) ==4356== by 0x578F471: YSAssign::evaluate(bool) (YStatement.cc:579) ==4356== by 0x57A73F5: YBlock::evaluate(bool) (YBlock.cc:558) ==4356== by 0x578EEA0: YSBlock::evaluate(bool) (YStatement.cc:291) ==4356== by 0x579DD17: YSWhile::evaluate(bool) (YStatement.cc:1065) ==4356== by 0x57A73F5: YBlock::evaluate(bool) (YBlock.cc:558) ==4356== by 0x5784136: Y2YCPFunction::evaluateCall() (YExpression.cc:3084) ==4356== by 0x577DCD5: YEFunction::evaluate(bool) (YExpression.cc:2847) ==4356== by 0x578F471: YSAssign::evaluate(bool) (YStatement.cc:579) ==4356== by 0x57A73F5: YBlock::evaluate(bool) (YBlock.cc:558) ==4356== by 0x578EEA0: YSBlock::evaluate(bool) (YStatement.cc:291) ==4356== by 0x578FA5B: YSIf::evaluate(bool) (YStatement.cc:948) ==4356== by 0x57A73F5: YBlock::evaluate(bool) (YBlock.cc:558) ==4356== by 0x578EEA0: YSBlock::evaluate(bool) (YStatement.cc:291) ==4356== by 0x578FAD1: YSIf::evaluate(bool) (YStatement.cc:943) ==4356== by 0x57A73F5: YBlock::evaluate(bool) (YBlock.cc:558) ==4356== by 0x576EEC2: YCPCodeRep::evaluate(bool) const (YCPCode.cc:81) ==4356== by 0x4E33268: Y2WFMComponent::doActualWork(YCPList const&, Y2Component*) (Y2WFMComponent.cc:177) ==4356== by 0x4E30530: Y2WFMComponent::CallFunction(YCPString const&, YCPList const&) (Y2WFMComponent.cc:651) ==4356== by 0x4E38781: WFMCallFunction1(YCPString const&, YCPList const&) (WFM.cc:80) ==4356== by 0x577EB1E: YEBuiltin::evaluate(bool) (YExpression.cc:2308) ==4356== by 0x578F471: YSAssign::evaluate(bool) (YStatement.cc:579) ==4356== by 0x57A73F5: YBlock::evaluate(bool) (YBlock.cc:558) ==4356== by 0x578EEA0: YSBlock::evaluate(bool) (YStatement.cc:291) ==4356== by 0x578FAD1: YSIf::evaluate(bool) (YStatement.cc:943) ==4356== by 0x578FA5B: YSIf::evaluate(bool) (YStatement.cc:948) ==4356== by 0x57A73F5: YBlock::evaluate(bool) (YBlock.cc:558) ==4356== by 0x576EEC2: YCPCodeRep::evaluate(bool) const (YCPCode.cc:81) ==4356== by 0x4E33268: Y2WFMComponent::doActualWork(YCPList const&, Y2Component*) (Y2WFMComponent.cc:177) ==4356== by 0x5A4AFB5: main (genericfrontend.cc:544) std::string::append() in connection with a block that was already deleted sounds like trouble. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug, or are watching someone who is.
https://bugzilla.novell.com/show_bug.cgi?id=258742 sh@novell.com changed: What |Removed |Added ---------------------------------------------------------------------------- AssignedTo|sh@novell.com |yast2-maintainers@suse.de ------- Comment #33 from sh@novell.com 2007-04-10 07:18 MST ------- Anyway, I already spent 3 days of last week with this. I am not prepared to spend any more time proving that it's not the UI. If nobody takes over, this will become a WONTFIX. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug, or are watching someone who is.
https://bugzilla.novell.com/show_bug.cgi?id=258742 mvidner@novell.com changed: What |Removed |Added ---------------------------------------------------------------------------- AssignedTo|yast2-maintainers@suse.de |mvidner@novell.com Status|NEW |ASSIGNED ------- Comment #34 from mvidner@novell.com 2007-04-10 10:20 MST ------- <mvidner> the invalid write at the beginning is worth debugging <HuHa> IIRC that is one of the cases where Qt / Xlib simply copies 8 bytes where only 4 are initialized <HuHa> not really intuitive, but it happens <HuHa> and since it happens in either case, I can't see a connection with that segfault <mvidner> are you sure? ok if that is so then it is probably unrelated <mvidner> i thought that it was overwriting a random memory location <HuHa> IIRC it's the read address where some bytes were never written to <HuHa> not an invalid write address <HuHa> I could be wrong there, though <HuHa> but that's one of the cases we had investigated further a long time ago <mvidner> hm, seems to make sense, "Syscall param writev(vector[...]) points to uninitialised byte(s)" <HuHa> right <mvidner> ok, i will apply more gdb tortu^Winvestigation meanwhile, valgrind refuses to start gdb, grrr. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug, or are watching someone who is.
https://bugzilla.novell.com/show_bug.cgi?id=258742 ------- Comment #35 from aj@novell.com 2007-04-10 11:22 MST ------- Just tried Build300 (candidate for Alpha3): This bug did not occur in my i386 installation. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug, or are watching someone who is.
https://bugzilla.novell.com/show_bug.cgi?id=258742 lslezak@novell.com changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |jkupec@novell.com ------- Comment #38 from lslezak@novell.com 2007-04-13 01:59 MST ------- *** Bug 263802 has been marked as a duplicate of this bug. *** -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug, or are watching someone who is.
https://bugzilla.novell.com/show_bug.cgi?id=258742 ------- Comment #39 from lslezak@novell.com 2007-04-13 02:03 MST ------- The duplicate is little different, it happened even if online update was skipped. But in my test installation it was OK. So it seems to be more serious... -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug, or are watching someone who is.
https://bugzilla.novell.com/show_bug.cgi?id=258742#c40
Christoph Thiel
https://bugzilla.novell.com/show_bug.cgi?id=258742#c41
Ladislav Slezak
https://bugzilla.novell.com/show_bug.cgi?id=258742#c42
Christoph Thiel
https://bugzilla.novell.com/show_bug.cgi?id=258742#c43
Holger Sickenberg
https://bugzilla.novell.com/show_bug.cgi?id=258742#c44
Martin Vidner
participants (1)
-
bugzilla_noreply@novell.com