Re: 3.17-rc6: bcache_gc: BUG: soft lockup - CPU#2 stuck for 23s!

[Stefan Seyfried <stefan.seyfried@googlemail.com>]

Hi all,

Am 24.11.2014 um 19:52 schrieb Eric Wheeler:
>>>> Before the patch goes upstream I need it tested so I know if it fixes
>>>> the actual issue or not.
>>>
>>> We have been using the rcu_sched patch and the cond_resched patch
>>> together
>>> (both attached) since November 3rd on 3.17.2 without any bcache
>>> backtraces. bcache is running in writeback mode. The server is
>>> predominantly write-only with relatively few reads.
>>
>> Is the rcu_sched patch supposed to help the same or a totally different
>> problem? Means: should I also apply it, rebuild the module and reboot
>> (resetting the test time to zero :-)

Ok, I am also running both patches, my uptime is 10 days and I'm right
now doing a bit of the usual stress-testing on the bcached partitions
(find /m1 /m2 /m3 > /dev/null which does read-mostly, and a rebuild of
all my yocto-project embedded board targets, which does read-write)

I have not seen any problems yet, so the patches don't seem to hurt.

>> I had the "deadly" soft lockup once per week since updating to
>> 3.16/3.17, so I can only tell if it might help in two weeks earliest.

It's not two weeks yet, but getting close.

Load average never goes below 1 (with 3 bcache-backed mounts).

Best regards,

	Stefan
-- 
Stefan Seyfried
Linux Consultant & Developer -- GPG Key: 0x731B665B

B1 Systems GmbH
Osterfeldstraße 7 / 85088 Vohburg / http://www.b1-systems.de
GF: Ralph Dehner / Unternehmenssitz: Vohburg / AG: Ingolstadt,HRB 3537