Created attachment 107991 [details] gdbtrace.log with segfault STR (1) In a bibisect repository, issue the command: opt/program/soffice --backtrace --norestore --nologo --base (2) In step 1 of Database Wizard accept all defaults, i.e., create embedded HSQLDB database; click <Next>. (3) In step 2 of Database Wizard select "No, do not register database"; click <Finish>. (4) In Save dialog, type the pathname of a non-existent file and click <Save>. Program creates file gdbtrace.log showing a SIGSEGV. The attached gdbtrace.log is from daily dbgutil bibisect repository, version (line wrapped for readability): HEAD is now at 72af794... 2014-10-17: source-hash-51b215902d8e1f415b5fef49fc50c7167f0c7787 but I see similar things in the 43all bibisect repository, which has product builds.
Terrence: I reproduced this. However with make debugrun, I always had this kind of segfault with Base: Program received signal SIGSEGV, Segmentation fault. [Switching to Thread 0x2aaad2d00700 (LWP 2203)] 0x00002aaad4437905 in ?? () I type "c" to continue and that's ok. Another one too when I launch wizard for table creation: Program received signal SIGSEGV, Segmentation fault. [Switching to Thread 0x2aaadcc1e700 (LWP 2215)] 0x00002aaad37035da in write_memory_serialize_page (thread=0x2b1b800) at /scratch/packages/openjdk/7u15/openjdk-7-7u65-2.5.2/build/openjdk/hotspot/src/share/vm/runtime/os.hpp:382 382 /scratch/packages/openjdk/7u15/openjdk-7-7u65-2.5.2/build/openjdk/hotspot/src/share/vm/runtime/os.hpp: Aucun fichier ou dossier de ce type. Since I reproduce this, I put it at New but I must recognize I don't know if it could be considered as a bug or not.
Thank you, Julien. I never thought of trying to continue past the segfault, but now that you mention it I think I remember that Java routinely creates and handles segfaults. I limited my search for potential duplicates before creating this report to just *open* bugs. Darn. Anyway, setting this report RESOLVED NOTABUG.