Bug 272053

Summary: x11/mate-screensver: dies with SIGBUS on initiation
Product: Ports & Packages Reporter: rkoberman
Component: Individual Port(s)Assignee: freebsd-ports-bugs (Nobody) <ports-bugs>
Status: Closed Overcome By Events    
Severity: Affects Only Me CC: fernape, gnome
Priority: --- Flags: fernape: maintainer-feedback? (gnome)
Version: Latest   
Hardware: amd64   
OS: Any   

Description rkoberman 2023-06-17 23:05:48 UTC
mate-screensaver dies in libmate-desktop-2.so.17 when the screensaver is activated.

I built mate-desktop and mate-screensaver WITH_DEBUG and got the following:
(lldb) target create "/usr/local/bin/mate-screensaver"
Current executable set to '/usr/local/bin/mate-screensaver' (x86_64).
(lldb) r
Process 70428 launched: '/usr/local/bin/mate-screensaver' (x86_64)
Process 70428 stopped
* thread #1, name = 'mate-screensaver', stop reason = signal SIGBUS: hardware error
    frame #0: 0x000000004f819272 libmate-desktop-2.so.17`blow_expensive_caches(data=0x0000000080350000) at mate-bg.c:2025:32

System is running HEAD:n262908-c16e08e5f324 on an Alder Lake (i5-1235U (2496.00-MHz K8-class CPU)

Since this is, in theory,  hardware error, is it an issue with the processor (2P cores and 8 E cores)? Maybe support for the Alder Lake GPU? I know that DRI is not working and the system freezes when it loads the firmware. Graphics works, badly, if I load drm-515-kmod without firmware.

If I am not running under lldb, invoking the screen saver causes in instant crash but running it under lldb the screen blanks until I kiil lldb. (I use a blank screen as my screensaver.)
Comment 1 rkoberman 2024-05-19 16:07:46 UTC
Last week I decided to check on whether the recent update to Mate might have corrected this issue and I re-enabled the screensaver. It is now back to working normally.