On Wed, Jul 18, 2007 at 01:26:48PM +0530, Dhaval Giani wrote:
[refer http://marc.info/?l=linux-kernel&m=118474574807055 for complete
report of this bug]
Ingo,
Dhaval tracked the root cause of this problem to be in cfs (btw
cfs patches weren't git-bisect safe).
Basically, "make -s -j" workload hanged the machine, leading to lot of
OOM killings. This was on a 8-cpu machine with no swap space configured and
4GB RAM. The same workload works "fine" (runs to completion) on 2.6.22.
I played with the scheduler tunables a bit and found that the problem
goes away if I set sched_granularity_ns to 100ms (default value 32ms).
So my theory is this: 32ms preemption granularity is too low value for
any compile thread to make "usefull" progress. As a result of this rapid
context switch, job retiral rate slows down compared to job arrival
rate. This builds up job pressure on the system very quickly (than may
have happened with 100ms default granularity_ns or 2.6.22 kernel),
leading to OOM killings (and hang).
value, this may be seen as a regression.
Perhaps, these new tunables in cfs are something for users to become used to
and tune it to approp setting for their system. It would have been
nice for kernel to auto-tune the settings based on workload, but I guess
that's harder.
--
Regards,
vatsa
-
