Q: print_cfs_rq() in kernel/sched_debug.c (2.6.32.49-0.3, SLES10 SP1
From: Ulrich Windl
Date: Tue Dec 20 2011 - 08:18:14 EST
Hi!
I'm debuggin some problem with disk I/O hanging. So I dumped the blocked =
tasks (SysRq : Show Blocked State). I don't understand all the internals =
of the output, but I wondered why the vruntime max is smaller than the =
vruntime min in cfs_rq:
Dec 19 14:24:57 h01 kernel: [2241368.267723] cfs_rq[0]:/
Dec 19 14:24:57 h01 kernel: [2241368.267727] .exec_clock =
: 6494556.016152
Dec 19 14:24:57 h01 kernel: [2241368.267732] .MIN_vruntime =
: 0.000001
Dec 19 14:24:57 h01 kernel: [2241368.267736] .min_vruntime =
: 36574276.029743
Dec 19 14:24:57 h01 kernel: [2241368.267740] .max_vruntime =
: 0.000001
Dec 19 14:24:57 h01 kernel: [2241368.267743] .spread =
: 0.000000
Dec 19 14:24:57 h01 kernel: [2241368.267747] .spread0 =
: 0.000000
Dec 19 14:24:57 h01 kernel: [2241368.267751] .nr_running =
: 0
Dec 19 14:24:57 h01 kernel: [2241368.267754] .load =
: 0
Dec 19 14:24:57 h01 kernel: [2241368.267758] .nr_spread_over =
: 939
Dec 19 14:24:57 h01 kernel: [2241368.267762] .shares =
: 0
The code for the max looks like this:
if (last)
max_vruntime =3D last->vruntime;
First I suspected the error in the printf section.
Ist there anything wrong, or is it my misunderstanding?
More samples:
Dec 19 14:24:57 h01 kernel: [2241368.267937] .MIN_vruntime =
: 0.000001
Dec 19 14:24:57 h01 kernel: [2241368.267941] .min_vruntime =
: 48353350.463240
Dec 19 14:24:57 h01 kernel: [2241368.267945] .max_vruntime =
: 0.000001
Dec 19 14:24:57 h01 kernel: [2241368.268138] .MIN_vruntime =
: 0.000001
Dec 19 14:24:57 h01 kernel: [2241368.268142] .min_vruntime =
: 34269914.013038
Dec 19 14:24:57 h01 kernel: [2241368.268146] .max_vruntime =
: 0.000001
Regards,
Ulrich
P.S.: Please CC: any replies to me as I'm not subscribed to the kernel =
list
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/