[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

What is wrong with kernels 5.X on sparc64 ?



Hi people!

Couple months ago, I've got two Ultra 45 workstations.

Cleaned them up, filled with memory and disks, and since Solaris 10 is
too old, 11 doesn't work on UltraSparc IIIi procs, Linux, namely Debian
was most logical choice.

I have bullseye/sid installed in md raid1 + LVM configuration and hoping
to convert this into LXC containered services machines - a system level
virtualization.

I've managed to get some things with mmap MAP_FIXED on sparc64 fixed by
lxc people ( https://github.com/lxc/lxc/issues/3262 ) and successfuly
created first test containers today!

But I have one nightmare with this machines: 4.19.0-5-sparc64-smp is the
last really stable kernel I can get to work there. With all other
updates to kernel image, system panics often, specially when running
systemctl command manually (enable/disable actions) or when called from
apt. I have even tried with manually compiled newest 5.4.1 and 5.5
kernels, but no luck.

The blues usually goes like this (OBP on serial):

Processing triggers for libc-bin (2.31-0+sparc64) ...
Processing triggers for systemd (245.2-1) ...

varnava login: [12438.966956] Unable to handle kernel paging request in
mna handler
[12438.966960]  at virtual address 2a70e013abac6164
[12439.095778] current->{active_,}mm->context = 00000000000000cb
[12439.164784] current->{active_,}mm->pgd = fff0001375fcc000
[12439.229899] Kernel panic - not syncing: Attempted to kill init!
exitcode=0x00000009
[12439.321907] Press Stop-A (L1-A) from sun keyboard or send break
[12439.321907] twice on console to return to the boot prom
[12439.321938] kernel BUG at kernel/cpu.c:941!

... or this:

(Reading database ... 56196 files and directories currently installed.)
Removing exim4-daemon-light (4.93-10) ...
Removing exim4-base (4.93-10) ...


varnava login: [ 4888.989100] Unable to handle kernel NULL pointer
dereference
[ 4889.057151] tsk->{mm,active_mm}->context = 00000000000000ce
[ 4889.124099] tsk->{mm,active_mm}->pgd = fff0001375888000
[ 4889.187349] Kernel panic - not syncing: Attempted to kill init!
exitcode=0x00000009
[ 4889.279315] Press Stop-A (L1-A) from sun keyboard or send break
[ 4889.279315] twice on console to return to the boot prom
[ 4889.279345] kernel BUG at kernel/cpu.c:941!


Looks like almost always this is triggered from systemd.


Does anybody have some information or experience with this phenomenon?


-- 
Not every internet quote is true.
	-- Oscar Wilde


Reply to: