On Tue, 2010-08-10 at 00:04 +0200, Cesare Leonardi wrote: > Hi all. > > I've noted some changes in the scheduler configuration of > 2.6.35-trunk-686 and i wonder if they are erroneous or not, since their > effect are rather evident and i've seen no discussion on d-k. > I was filing a bug but i preferred to ask on the mailing list first. > > --- /boot/config-2.6.35-rc6-686 2010-07-26 13:09:56.000000000 +0200 > +++ /boot/config-2.6.35-trunk-686 2010-08-06 18:13:52.000000000 +0200 > @@ -221,11 +222,12 @@ > # > CONFIG_IOSCHED_NOOP=y > CONFIG_IOSCHED_DEADLINE=y > -CONFIG_IOSCHED_CFQ=y > -# CONFIG_DEFAULT_DEADLINE is not set > -CONFIG_DEFAULT_CFQ=y > +CONFIG_IOSCHED_CFQ=m > +CONFIG_CFQ_GROUP_IOSCHED=y > +CONFIG_DEFAULT_DEADLINE=y > +# CONFIG_DEFAULT_CFQ is not set > # CONFIG_DEFAULT_NOOP is not set > -CONFIG_DEFAULT_IOSCHED="cfq" > +CONFIG_DEFAULT_IOSCHED="deadline" > > Why setting deadline as default scheduler? And why configuring cfq as > module? This is due to enabling BLK_CGROUP: config IOSCHED_CFQ tristate "CFQ I/O scheduler" # If BLK_CGROUP is a module, CFQ has to be built as module. depends on (BLK_CGROUP=m && m) || !BLK_CGROUP || BLK_CGROUP=y Max, we have to make BLK_CGROUP=y or disable it completely. Ben. -- Ben Hutchings Once a job is fouled up, anything done to improve it makes it worse.
Attachment:
signature.asc
Description: This is a digitally signed message part