On Mon, 09 May 2011 12:00:28 -0400 Josef Bacik <jo...@redhat.com> wrote:
> On 05/07/2011 05:29 PM, sly...@gmail.com wrote:
> > Observed as a large delay when --mixed filesystem is filled up.
> > Test example:
> > 1. create tiny --mixed FS:
> > $ dd if=/dev/zero of=2G.img seek=$((2048 * 1024 * 1024 - 1)) count=1
> > bs=1
> > $ mkfs.btrfs --mixed 2G.img
> > $ mount -oloop 2G.img /mnt/ut/
> > 2. Try to fill it up:
> > $ dd if=/dev/urandom of=10M.file bs=10240 count=1024
> > $ seq 1 256 | while read file_no; do echo $file_no; time cp 10M.file
> > ${file_no}.copy; done
> >
> > Up to '200.copy' it goes fast, but when disk fills-up each -ENOSPC
> > message takes 3 seconds to pop-up _every_ ENOSPC (and in usermode linux
> > it's even more: 30-60 seconds!). (Maybe, time depends on kernel's timer
> > resolution).
> >
> > No IO, no CPU load, just rescheduling. Some debugging revealed busy spinning
> > in shrink_delalloc.
> >
> > Signed-off-by: Sergei Trofimovich<sly...@gentoo.org>
> > ---
> > fs/btrfs/extent-tree.c | 4 ++++
> > 1 files changed, 4 insertions(+), 0 deletions(-)
> >
> > diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c
> > index 9ee6bd5..9f5fdd3 100644
> > --- a/fs/btrfs/extent-tree.c
> > +++ b/fs/btrfs/extent-tree.c
> > @@ -3425,6 +3425,10 @@ static int shrink_delalloc(struct btrfs_trans_handle
> > *trans,
> > if (reserved == 0)
> > return 0;
> >
> > + /* nothing to shrink - nothing to reclaim */
> > + if (root->fs_info->delalloc_bytes == 0)
> > + return 0;
> > +
> > max_reclaim = min(reserved, to_reclaim);
> >
> > while (loops< 1024) {
>
> Nice catch, you can add
>
> Reviewed-by: Josef Bacik <jo...@redhat.com>
Will add it (and fix patch enumeration) and resend.
Thanks for the review!
--
Sergei
signature.asc
Description: PGP signature
