On Monday August 28, akpm@osdl.org wrote:
On Tue, 29 Aug 2006 15:39:24 +1000
NeilBrown [off-list ref] wrote:
quoted
Each backing_dev needs to be able to report whether it is congested,
either by modulating BDI_*_congested in ->state, or by
defining a ->congested_fn.
md/raid did neither of these. This patch add a congested_fn
which simply checks all component devices to see if they are
congested.
Signed-off-by: Neil Brown <redacted>
+static int linear_congested(void *data, int bits)
+{
+ mddev_t *mddev = data;
+ linear_conf_t *conf = mddev_to_conf(mddev);
+ int i, ret = 0;
+
+ for (i = 0; i < mddev->raid_disks && !ret ; i++) {
+ request_queue_t *q = bdev_get_queue(conf->disks[i].rdev->bdev);
+ ret |= bdi_congested(&q->backing_dev_info, bits);
nit: `ret = ' would suffice here.
Yeh.... I was looking at linear_issue_flush, which aborts when
ret == 0, and dm_table_any_congested, which 'or's together the results
and combining to two. And I kept changing my mind as to which way I
wanted to go for linear_congested et.al, and ended up going both ways
:-(
I hate indecision. I think. Well, probably I do. Maybe.
NeilBrown
--
VGER BF report: U 0.5