Den 2023-10-05 kl. 04:50, skrev Yu Kuai:
Hi,
在 2023/09/30 3:44, Joel Parthemore 写道:
Den 2023-09-26 kl. 03:10, skrev Yu Kuai:
It'll be much helper for developers to collect kernel stack for
all stuck thread(and it'll be much better to use add2line).
Presuming I can re-create the problem, let me know what I should do
to collect that information for you. I'm very much a newbie in that
area.
You can use following cmd:
for pid in `ps -elf | grep " D " | awk '{print $4}'`; do ps $pid;
cat /proc/$pid/stack; done
Thanks,
Kuai
for pid in `ps -elf | grep " D " | awk '{print $4}'`; do ps $pid; cat
/proc/$pid/stack; done
PID TTY STAT TIME COMMAND
4017 tty1 D+ 0:00 e2fsck /dev/md126
Do you check that this thread is hanged and the e2fsck doesn't make
progress?
I wasn't trying to run e2fsck at the time. Instead I had run mdadm
--stop /dev/md126, which was hanging. (Sorry; looks like I forgot to
include that!) mdadm -D /md126 showed no re-syncing happening at the time.
Unfortunately, I had not heard back from you, and I felt like I couldn't
keep the problematic RAID array around any longer, so I wiped it and
replaced it with a fully native-Linux RAID5 array instead of one using
IMSM metadata.
Joel
