Created attachment 676390 [details] emerge segfault Hi, Some days ago, while updating world on a system, I got a segfault from portage (output attached). I solved it by doing this : > # mv /var/cache/edb /var/cache/edb.old Then I ran emerge again to end the world update without any other problem. I still have : - /var/cache/edb.old folder - two dump core files (each 20 MB size) Do you want them ?
We need emerge --info too, but yes, we need the core files too (and a backtrace ideally). Could you tar all of that up and compress it?
Unfortunately I do not have any backtrace, unless it is automatically logged somewhere. Once the world updated and system rebooted, I can not reproduce it anymore. In system logs I can find this : > Nov 29 12:18:27 coreI3 kernel: emerge[19608]: segfault at 7fe52cdfaee0 ip 00007fc52db6aca8 sp 00007ffcb045a820 error 4 in libpython3.7m.so.1.0[7fc52da4c000+170000] > Nov 29 12:18:27 coreI3 kernel: Code: 00 48 89 04 24 48 8b 44 24 08 48 8d 04 40 48 c1 e0 04 4c 8b a4 05 60 01 00 00 4d 39 f4 0f 84 af 0a 00 00 4c 89 e0 0f 1f 40 00 <48> 8b 50 10 48 8b 70 20 83 e2 01 48 8d 0c 36 48 09 ca 48 89 50 10 > Nov 29 12:20:00 coreI3 kernel: emerge[30967]: segfault at 7f63160a8240 ip 00007f431b363ca8 sp 00007fffdd5d68e0 error 4 in libpython3.7m.so.1.0[7f431b245000+170000] > Nov 29 12:20:00 coreI3 kernel: Code: 00 48 89 04 24 48 8b 44 24 08 48 8d 04 40 48 c1 e0 04 4c 8b a4 05 60 01 00 00 4d 39 f4 0f 84 af 0a 00 00 4c 89 e0 0f 1f 40 00 <48> 8b 50 10 48 8b 70 20 83 e2 01 48 8d 0c 36 48 09 ca 48 89 50 10 All the stuff compressed weighs 7.6 MB, but it it still usable without backtrace ?
(In reply to Fab from comment #2) > Unfortunately I do not have any backtrace, unless it is automatically logged > somewhere. Once the world updated and system rebooted, I can not reproduce it > anymore. In system logs I can find this : > > > Nov 29 12:18:27 coreI3 kernel: emerge[19608]: segfault at 7fe52cdfaee0 ip 00007fc52db6aca8 sp 00007ffcb045a820 error 4 in libpython3.7m.so.1.0[7fc52da4c000+170000] > > Nov 29 12:18:27 coreI3 kernel: Code: 00 48 89 04 24 48 8b 44 24 08 48 8d 04 40 48 c1 e0 04 4c 8b a4 05 60 01 00 00 4d 39 f4 0f 84 af 0a 00 00 4c 89 e0 0f 1f 40 00 <48> 8b 50 10 48 8b 70 20 83 e2 01 48 8d 0c 36 48 09 ca 48 89 50 10 > > Nov 29 12:20:00 coreI3 kernel: emerge[30967]: segfault at 7f63160a8240 ip 00007f431b363ca8 sp 00007fffdd5d68e0 error 4 in libpython3.7m.so.1.0[7f431b245000+170000] > > Nov 29 12:20:00 coreI3 kernel: Code: 00 48 89 04 24 48 8b 44 24 08 48 8d 04 40 48 c1 e0 04 4c 8b a4 05 60 01 00 00 4d 39 f4 0f 84 af 0a 00 00 4c 89 e0 0f 1f 40 00 <48> 8b 50 10 48 8b 70 20 83 e2 01 48 8d 0c 36 48 09 ca 48 89 50 10 > > All the stuff compressed weighs 7.6 MB, but it it still usable without > backtrace ? The bt should be in the core dumps, it just would’ve made it easier to figure out where the problem lies without downloading them. Could you upload them with emerge —-info pasted here? Does it happen if you move the old cache back?
Created attachment 676396 [details] emerge --info Uploaded core dumps and cache folder : https://filebin.net/xkakenjw092xc591/segfault.tar.gz?t=pkoy2tym (In reply to Sam James from comment #3) > Does it happen if you move the old cache back? I tried, but no, it does not crash anymore (maybe because system was in the meantime rebooted).
A segmentation fault is more of a python problem than a portage problem. Since you can't reproduce it after the reboot, we should probably just close this bug.