GNOME Bugzilla – Bug 410242
crash on log file replay
Last modified: 2018-06-29 21:27:29 UTC
To: submit@bugs.gnome.org X-Mailer: bug-buddy 2.2.0 Subject: Random crash Package: GnuCash Severity: normal Version: 2.0.5 Synopsis: Random crash Bugzilla-Product: GnuCash Bugzilla-Component: General BugBuddy-GnomeVersion: 2.0 (2.4.0.1) Description: Description of Problem: GnuCash 2.0.5. After compiling and installing, I created and opened a new set of business accounts (no file yet). I created a handful of simple transactions. Then the Gnome Kiss of Death screen came up telling me my application has creashed. Here is the output from the console: lots of chatter, which may or may not be related to the crash. I am most worried about the failed assertion near the end. v$ gnucash --nofile ** (process:2210): WARNING **: Failed to dlopen() '/usr/lib/gnucash/libgncmod-backend-file.la': /usr/lib/gnucash/libgncmod-backend-file.so.0: undefined symbol: gnc_commodity_set_exchange_code ** (process:2210): WARNING **: Failed to dlopen() '/usr/lib/gnucash/libgncmod-app-file.la': /usr/lib/gnucash/libgncmod-app-file.so.0: undefined symbol: gnc_mdi_show_progress ** (process:2210): WARNING **: Failed to dlopen() '/usr/lib/gnucash/libgncmod-binary-import.la': /usr/lib/gnucash/libgncmod-app-file.so.0: undefined symbol: gnc_mdi_show_progress ** (process:2210): WARNING **: Failed to dlopen() '/usr/lib/gnucash/libgw-binary-import.la': /usr/lib/gnucash/libgncmod-app-file.so.0: undefined symbol: gnc_mdi_show_progress (gnucash:2210): GLib-GObject-WARNING **: g_object_set_valist: object class `GtkCellRendererToggle' has no property named `sensitive' (gnucash:2210): GLib-GObject-WARNING **: g_object_set_valist: object class `GtkCellRendererText' has no property named `sensitive' [ message repeated a few hundred times ] GnomeCanvas-ERROR **: file gnome-canvas.c: line 3509 (gnome_canvas_request_update_real): assertion failed: (canvas->idle_id != 0) aborting... "/usr/bin/gnucash": not in executable format: File format not recognized Steps to reproduce the problem: 1. 2. 3. Actual Results: Expected Results: How often does this happen? Additional Information: Debugging Information: Backtrace was generated from '/usr/bin/gnucash' [New Thread 16384 (LWP 2210)] 0x40d91766 in waitpid () from /lib/libpthread.so.0
+ Trace 112467
------- Bug created by bug-buddy at 2007-02-21 02:00 -------
Thanks for taking the time to report this bug. Unfortunately, that stack trace is missing some elements that will help a lot to solve the problem, so it will be hard for the developers to fix that crash. Can you get us a stack trace with debugging symbols? Please see http://live.gnome.org/GettingTraces for more information on how to do so. Thanks in advance!
Sorry, I can't upload another stack trace at the moment, because I have already uninstalled GnuCash 2.0.5. (I need a working version, and I was unable to install two versions simultaneously, as each overwrites the other. I was also unable to run GnuCash without installing it first). -- Peter
Ok, thanks for reporting this bug. Just in case you ever try gnucash again: It is important to know whether you clicked "Save" before this crash occurred or whether you didn't save your data at all. Whatever. Feel free to reopen this bug or open new bugs if you find any problems again. Thanks!
Thanks. As a matter of fact, I did reinstall 2.0.5 yesterday, with debugging symbols, but I did not get another crash. I will post a stack trace as soon as I do. I don't think I was saving at the time (I was working with an unsaved new file). As far as I remember, the crash occurred as I was double-clicking on an account in the accounts chart to change to another account. -- Peter
I am reopening this bug because I experienced another, very similar crash. Only this time, I have debugging symbols, so I hope the backtrace will be more useful. Description of Problem: (a) I opened a set of accounts (from the command line) that had been created with gnucash 1.8.*. (b) I went through the dialog to resolve "Ambiguous character encoding". (c) Then I replayed three log files, (d) selected "Quit", selected "Save" from the "Save changes before closing?" popup. The application saved successfully, then crashed instead of quitting. The terminal showed this: ** ERROR **: file Account.c: line 278 (xaccFreeAccount): assertion failed: (acc->splits == NULL) aborting... Steps to reproduce: I can reproduce the bug by following the above steps exactly, if I start with the same file. Also, (a) and (b) are not really necessary to reproduce the crash, i.e., I get the same crash by starting directly with the file that was saved as part of (b). Also, the crash also occurs if I only replay the second of the three log files, instead of all three (but it doesn't work if I just replay the first one). I can reproduce the crash any number of times, but I can't send my accounts file because it contains too much personal information. If there is an "account anonymizer" tool that changes all the amounts and descriptions, then I might be able to send a sample file. Debugging Information: Backtrace was generated from '/usr/local/bin/gnucash' [New Thread 16384 (LWP 17769)] 0x40d87766 in waitpid () from /lib/libpthread.so.0
+ Trace 118164
name=0x80e0f30 "H\017\016\b\200\017\016\b
I just verified that the crash also occurs if I first save (from the menu), then quit. It happens during quitting, not saving. However, I can reopen the saved file and then quit without crashing.
Created attachment 85284 [details] GnuCash accounts file to reproduce the crash To reproduce the crash, open crashfile.xac in GnuCash 2.0.5. File -> Import -> Replay GnuCash .log file: crashfile.log (see next attachment). File -> Save. File -> Quit. A fatal error occurs (see the stack trace posted above).
Created attachment 85285 [details] Log file needed to reproduce crash.
As you might have guessed, the actual problem is related to the log file replay. Quoting Chris: This crash happens upon shutdown whenever something in GnuCash opens a transaction for editing and doesn't commit it or roll it back. It indicates a bug somewhere else that could result in incorrect or inconsistent data being saved (if it didn't crash). The report suggests that this might be related to log replay. Thanks for the example file. We'll try to fix this soon.
*** Bug 410153 has been marked as a duplicate of this bug. ***
Obsolete - doesn't happen with 2.2.9
GnuCash bug tracking has moved to a new Bugzilla host. This bug has been copied to https://bugs.gnucash.org/show_bug.cgi?id=410242. Please update any external references or bookmarks.