SUMMARY When I edit a formula in a cell in a certain way, Calc crashes. STEPS TO REPRODUCE 1. Create a formula (eg. =A1+A2+A3+A4+A5 is enough) 2. Click on the cell and press F2 to edit the cell 3. Click in the edit field above the sheet to edit the formula 4. Press F2 again (cursor will jump to the beginning of the edit field) 5. Select a part that you want to change with your mouse cursor (eg. "3" in A3 that you want to change into A7) 6. Press a key to change that part The bug is always reproducible, but may have to be done multiple times to get the crash. If you don't get a crash on pressing (from the example:) 7, you can abort the test by pressing ESC and retry with the exact same way. Note that it seems to make a difference where you click and what you edit. When you have "=A1+A2+A3+A4+A5" it seemed to me that the crash occurs more consistent if you press in between the "3" of A3 and the "+" symbol. OBSERVED RESULT Calc crashes on button press (in the provided example this would be on pressing "7") EXPECTED RESULT Calc does not crash. SOFTWARE/OS VERSIONS LibreOffice: - Version: 25.8.3.2 (X86_64) / LibreOffice Community - Build ID: 8ca8d55c161d602844f5428fa4b58097424e324e - CPU threads: 12; OS: Linux 6.14; UI render: default; VCL: gtk3 - Locale: de-AT (de_AT.UTF-8); UI: de-DE - Flatpak - Calc: threaded Operating system: - Operating System: KDE neon User Edition - KDE Plasma Version: 6.5.4 - KDE Frameworks Version: 6.21.0 - Qt Version: 6.10.1 - Kernel Version: 6.14.0-37-generic (64-bit) - Graphics Platform: Wayland
Created attachment 204664 [details] CALC - bug 169995 - crash on formula edit - test file test file for the bug
Created attachment 204665 [details] CALC - bug 169995 - crash on formula edit - demo video Demo video of the bug Note that on the first try I didn't click between "3" and "+" at 0:49, but between "A" and "3". As described: The bug looks like the bug is more reliable to reproduce when you click between "3" and "+". On the second try at 1:21 I clicked between "3" and "+" and was able to cause the crash. At 0:59 Calc didn't crash so you press ESC and restart the same process.
ADDITIONAL NOTE After more testing it looks like as if the second run will always be successful no matter where you click or what you select. So the causing cycle seems to be: F2 -> click in line -> F2 -> mark & edit -> fail -> ESC -> F2 -> click in line -> F2 -> mark & edit -> crash
Please test Menu>Help>Restart in Safe Mode. No crash with: Version: 25.8.4.2 (X86_64) Build ID: 290daaa01b999472f0c7a3890eb6a550fd74c6df CPU threads: 4; OS: Windows 10 X86_64 (build 19045); UI render: Skia/Raster; VCL: win Locale: es-ES (es_ES); UI: en-US Calc: CL threaded
== C5, 1 == (In reply to m_a_riosv from comment #4) > Please test Menu>Help>Restart in Safe Mode. Crash still appears in safe mode. Even I wasn't able to get this working the last time (mainly because of flatpack, but also because I wasn't able to reproduce the crash) I maybe am able to retrieve a backtrace this time as described on https://wiki.documentfoundation.org/QA/BugReport/Debug_Information#GNU/Linux:_How_to_get_a_backtrace == C5, 2 == I also tested this bug on my PC at work and I couldn't reproduce it. => HOWEVER, as m_a_riosv test my work PC also runs under Windows 10. LibreOffice: - Version: 24.2.6.3_AA (X86_64) / LibreOffice Community - Build ID: 23735e228767740599e3f00b9371b05bcbf243d2 - CPU threads: 4; OS: Windows 10.0 Build 19045; UI render: default; VCL: win - Locale: de-AT (de_AT); UI: de-DE - Calc: threaded Operating System: - OS: Microsoft Windows 10 Enterprise - Version: 10.0.19045 Build 19045
[Automated Action] NeedInfo-To-Unconfirmed
I tested some more. As I saw in the past that some bug were Linux specific (and sometimes tied to one specific desktop environment) I wanted to try different setups. All setups were run as live USBs. == C7, 1 == # System Details Report ## Software Information: - **OS Name:** Ubuntu 25.10 - **OS Type:** 64-bit - **GNOME Version:** 49 - **Windowing System:** Wayland - **Kernel Version:** Linux 6.17.0-5-generic # LibreOffice - Version: 25.8.1.1 (X86_64) / LibreOffice Community - Build ID: 580(Build:1) - CPU threads: 12; OS: Linux 6.17; UI render: default; VCL: gtk3 - Locale: en-US (C.UTF-8); UI: en-US - Ubuntu package version: 4:25.8.1~rc1-0ubuntu1 - Calc: threaded This test was done on a stock Ubuntu 25.10 with the preinstalled LibreOffice package (I guess it's a snap package). => I was able to reproduce the bug == C7, 2 == # System Details Report - Operating System: KDE neon User Edition - KDE Plasma Version: 6.5.3 - KDE Frameworks Version: 6.20.0 - Qt Version: 6.10.1 - Kernel Version: 6.14.0-37-generic (64-bit) - Graphics Platform: Wayland This test was done on a stock KDEneon (neon-user-20251211-1320.iso) - so Ubuntu base - with LibreOffice installed from the repo. # LibreOffice (1) - Version: 24.2.7.2 (X86_64) / LibreOffice Community - Build ID: 420(Build:2) - CPU threads: 12; OS: Linux 6.14; UI render: default; VCL: gtk3 - Locale: en-US (C.UTF-8); UI: en-US - Ubuntu package version: 4:24.2.7-0ubuntu0.24.04.4 - Calc: threaded I installed this version from the repo via the terminal (that's why it's a bit old) # LibreOffice (2) - Version: 25.8.4.2 (X86_64) - Build ID: 290daaa01b999472f0c7a3890eb6a550fd74c6df - CPU threads: 12; OS: Linux 6.14; UI render: default; VCL: kf5 (cairo+wayland) - Locale: en-US (C.UTF-8); UI: en-US - Calc: threaded I installed this version from the deb files you can find on the libreoffice website (eg. LibreOffice_25.8.4_Linux_aarch64_deb.tar.gz from https://de.libreoffice.org/download/download/?type=deb-aarch64&version=25.8.4) As my current version is 25.8.3.2 I also tested this download (LibreOffice_25.8.3_Linux_aarch64_deb.tar.gz from https://www.libreoffice.org/download/download-libreoffice/?type=deb-aarch64&version=25.8.3) => In both cases (1 and 2) I was NOT able to reproduce the bug (I got the backtrace working, but could not cause the same crash). == C7, 3 == I thought if it was may just a strange configuration or download that was may already resolved. I restarted my PC and tested the bug again in my usual productive system. => I was able to reproduce the bug I thought that it's maybe because I am using libreoffice as flatpack on my productive system. I did not use flatpacks while bugtesting because I couldn't get the backtrace to work the last time (see https://bugs.documentfoundation.org/show_bug.cgi?id=164482#c14). I would be very thankful if anybody had more information or a website that explains or helps with backtrace with flatpacks. I got into a fresh live USB session again and installed LibreOffice as flatpack. => I was able to reproduce the bug On flathub the LibreOffice package (https://flathub.org/en/apps/org.libreoffice.LibreOffice) seems to be an official package: * blue checkmark indicates an 'official' package by original developers * Test on hover: "The ownership of the org.libreoffice.LibreOffice app ID has been verified using libreoffice.org" TLDR: The bug seems to be only present under Linux and here the flatpack version seems to be effected.
More testing. The following files are from an attempt to do a backtrace under Linux with the latest flatpack version. I created this backtrace with the help of ChatGPT so I have absolutely no idea if this is any helpful or not (please let me know if it is or isn't! - it would help me next time). Attached files: * gdb output of command 'bt full' * LibreOffice version * operating system version * gdb version
Created attachment 204706 [details] CALC - bug 169995 - crash on formula edit - flatpak backtrace-gdb output file The output of gdb with 'bt full' after the I recreated the crash.
Created attachment 204707 [details] CALC - bug 169995 - crash on formula edit - flatpak backtrace-LibreOffice version
Created attachment 204708 [details] CALC - bug 169995 - crash on formula edit - flatpak backtrace-operating system version The operating system version
Created attachment 204709 [details] CALC - bug 169995 - crash on formula edit - flatpak backtrace-gdb version The gdb version
Created attachment 204710 [details] CALC - bug 169995 - crash on formula edit - flatpak backtrace-gdb thread apply all output file The gdb output with command 'thread apply all bt full'
(In reply to BDF from comment #3) > > So the causing cycle seems to be: F2 -> click in line -> F2 -> mark & edit > -> fail -> ESC -> F2 -> click in line -> F2 -> mark & edit -> crash No crash with Version: 26.2.0.1.0+ (X86_64) Build ID: bdc4b4cafd092ead7eb5a0fc67d02b9c289f8df7 CPU threads: 4; OS: Linux 6.8; UI render: default; VCL: gtk3 Locale: cs-CZ (cs_CZ.UTF-8); UI: en-US Calc: threaded
(In reply to raal from comment #14) > (In reply to BDF from comment #3) > > > > So the causing cycle seems to be: F2 -> click in line -> F2 -> mark & edit > > -> fail -> ESC -> F2 -> click in line -> F2 -> mark & edit -> crash > > No crash with Version: 26.2.0.1.0+ (X86_64) > Build ID: bdc4b4cafd092ead7eb5a0fc67d02b9c289f8df7 > CPU threads: 4; OS: Linux 6.8; UI render: default; VCL: gtk3 > Locale: cs-CZ (cs_CZ.UTF-8); UI: en-US > Calc: threaded From where did you install LibreOffice? Note that it seems to be only reproducible if you install the application from flathub.