Re: [PATCH] cfq: improve fsync performance for small files

[Corrado Zoccolo]

[+tytso]
Hi Theodore,
in order to fix a performance problem with fsync in cfq, we need to
mark write requests sent by jbd/jbd2 as synchronous, so cfq knows that
some process may be waiting on them, and therefore will avoid idling.
Is this change reasonable for you? Should I send the change to
jbd*/commit.cc in a separated patch through your tree?

Thanks
Corrado

On Tue, Jul 20, 2010 at 10:09 PM, Corrado Zoccolo <czoccolo@xxxxxxxxx> wrote:
> Fsync performance for small files achieved by cfq on high-end disks is
> lower than what deadline can achieve, due to idling introduced between
> the sync write happening in process context and the journal commit.
>
> Moreover, when competing with a sequential reader, a process writing
> small files and fsync-ing them is starved.
>
> This patch fixes the two problems by:
> - marking journal commits as WRITE_SYNC, so that they get the REQ_NOIDLE
> Âflag set,
> - force all queues that have REQ_NOIDLE requests to be put in the noidle
> Âtree.
>
> Having the queue associated to the fsync-ing process and the one associated
> Âto journal commits in the noidle tree allows:
> - switching between them without idling,
> - fairness vs. competing idling queues, since they will be serviced only
> Âafter the noidle tree expires its slice.
>
> Acked-by: Vivek Goyal <vgoyal@xxxxxxxxxx>
> Reviewed-by: Jeff Moyer <jmoyer@xxxxxxxxxx>
> Tested-by: Jeff Moyer <jmoyer@xxxxxxxxxx>
> Signed-off-by: Corrado Zoccolo <czoccolo@xxxxxxxxx>
> ---
> Âblock/cfq-iosched.c | Â 18 ++++--------------
> Âfs/jbd/commit.c   |  Â2 +-
> Âfs/jbd2/commit.c  Â|  Â2 +-
> Â3 files changed, 6 insertions(+), 16 deletions(-)
>
> diff --git a/block/cfq-iosched.c b/block/cfq-iosched.c
> index eb4086f..49dada4 100644
> --- a/block/cfq-iosched.c
> +++ b/block/cfq-iosched.c
> @@ -216,7 +216,6 @@ struct cfq_data {
> Â Â Â Âenum wl_type_t serving_type;
> Â Â Â Âunsigned long workload_expires;
> Â Â Â Âstruct cfq_group *serving_group;
> - Â Â Â bool noidle_tree_requires_idle;
>
> Â Â Â Â/*
> Â Â Â Â * Each priority tree is sorted by next_request position. ÂThese
> @@ -2126,7 +2125,6 @@ static void choose_service_tree(struct cfq_data *cfqd, struct cfq_group *cfqg)
> Â Â Â Âslice = max_t(unsigned, slice, CFQ_MIN_TT);
> Â Â Â Âcfq_log(cfqd, "workload slice:%d", slice);
> Â Â Â Âcfqd->workload_expires = jiffies + slice;
> - Â Â Â cfqd->noidle_tree_requires_idle = false;
> Â}
>
> Âstatic struct cfq_group *cfq_get_next_cfqg(struct cfq_data *cfqd)
> @@ -3108,7 +3106,9 @@ cfq_update_idle_window(struct cfq_data *cfqd, struct cfq_queue *cfqq,
> Â Â Â Âif (cfqq->queued[0] + cfqq->queued[1] >= 4)
> Â Â Â Â Â Â Â Âcfq_mark_cfqq_deep(cfqq);
>
> - Â Â Â if (!atomic_read(&cic->ioc->nr_tasks) || !cfqd->cfq_slice_idle ||
> + Â Â Â if (cfqq->next_rq && (cfqq->next_rq->cmd_flags & REQ_NOIDLE))
> + Â Â Â Â Â Â Â enable_idle = 0;
> + Â Â Â else if (!atomic_read(&cic->ioc->nr_tasks) || !cfqd->cfq_slice_idle ||
> Â Â Â Â Â Â(!cfq_cfqq_deep(cfqq) && CFQQ_SEEKY(cfqq)))
> Â Â Â Â Â Â Â Âenable_idle = 0;
> Â Â Â Âelse if (sample_valid(cic->ttime_samples)) {
> @@ -3421,17 +3421,7 @@ static void cfq_completed_request(struct request_queue *q, struct request *rq)
> Â Â Â Â Â Â Â Â Â Â Â Âcfq_slice_expired(cfqd, 1);
> Â Â Â Â Â Â Â Âelse if (sync && cfqq_empty &&
> Â Â Â Â Â Â Â Â Â Â Â Â !cfq_close_cooperator(cfqd, cfqq)) {
> - Â Â Â Â Â Â Â Â Â Â Â cfqd->noidle_tree_requires_idle |=
> - Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â !(rq->cmd_flags & REQ_NOIDLE);
> - Â Â Â Â Â Â Â Â Â Â Â /*
> - Â Â Â Â Â Â Â Â Â Â Â Â* Idling is enabled for SYNC_WORKLOAD.
> - Â Â Â Â Â Â Â Â Â Â Â Â* SYNC_NOIDLE_WORKLOAD idles at the end of the tree
> - Â Â Â Â Â Â Â Â Â Â Â Â* only if we processed at least one !REQ_NOIDLE request
> - Â Â Â Â Â Â Â Â Â Â Â Â*/
> - Â Â Â Â Â Â Â Â Â Â Â if (cfqd->serving_type == SYNC_WORKLOAD
> - Â Â Â Â Â Â Â Â Â Â Â Â Â || cfqd->noidle_tree_requires_idle
> - Â Â Â Â Â Â Â Â Â Â Â Â Â || cfqq->cfqg->nr_cfqq == 1)
> - Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â cfq_arm_slice_timer(cfqd);
> + Â Â Â Â Â Â Â Â Â Â Â cfq_arm_slice_timer(cfqd);
> Â Â Â Â Â Â Â Â}
> Â Â Â Â}
>
> diff --git a/fs/jbd/commit.c b/fs/jbd/commit.c
> index 28a9dda..d97a0c6 100644
> --- a/fs/jbd/commit.c
> +++ b/fs/jbd/commit.c
> @@ -317,7 +317,7 @@ void journal_commit_transaction(journal_t *journal)
> Â Â Â Âint first_tag = 0;
> Â Â Â Âint tag_flag;
> Â Â Â Âint i;
> - Â Â Â int write_op = WRITE;
> + Â Â Â int write_op = WRITE_SYNC;
>
> Â Â Â Â/*
> Â Â Â Â * First job: lock down the current transaction and wait for
> diff --git a/fs/jbd2/commit.c b/fs/jbd2/commit.c
> index 75716d3..a078744 100644
> --- a/fs/jbd2/commit.c
> +++ b/fs/jbd2/commit.c
> @@ -369,7 +369,7 @@ void jbd2_journal_commit_transaction(journal_t *journal)
> Â Â Â Âint tag_bytes = journal_tag_bytes(journal);
> Â Â Â Âstruct buffer_head *cbh = NULL; /* For transactional checksums */
> Â Â Â Â__u32 crc32_sum = ~0;
> - Â Â Â int write_op = WRITE;
> + Â Â Â int write_op = WRITE_SYNC;
>
> Â Â Â Â/*
> Â Â Â Â * First job: lock down the current transaction and wait for
> --
> 1.6.2.5
>
>
>



-- 
__________________________________________________________________________

dott. Corrado Zoccolo             mailto:czoccolo@xxxxxxxxx
PhD - Department of Computer Science - University of Pisa, Italy
--------------------------------------------------------------------------
The self-confidence of a warrior is not the self-confidence of the average
man. The average man seeks certainty in the eyes of the onlooker and calls
that self-confidence. The warrior seeks impeccability in his own eyes and
calls that humbleness.
              Â Tales of Power - C. Castaneda
--
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/