* Re: [RESEND PATCH 4/4] aio: use iovec array rather than the single one [not found] ` <1406109834-4414-4-git-send-email-guz.fnst@cn.fujitsu.com> @ 2014-07-23 13:25 ` Jeff Moyer 2014-07-24 1:15 ` Gu Zheng 2014-07-24 15:56 ` Benjamin LaHaise 1 sibling, 1 reply; 6+ messages in thread From: Jeff Moyer @ 2014-07-23 13:25 UTC (permalink / raw) To: Gu Zheng; +Cc: bcrl, axboe, akpm, linux-aio, linux-kernel Gu Zheng <guz.fnst@cn.fujitsu.com> writes: > Previously, we only offer a single iovec to handle all the read/write cases, so > the PREADV/PWRITEV request always need to alloc more iovec buffer when copying > user vectors. > If we use a tmp iovec array rather than the single one, some small PREADV/PWRITEV > workloads(vector size small than the tmp buffer) will not need to alloc more > iovec buffer when copying user vectors. Hi, Gu, This still doesn't explain why you decided to look into this. Did you notice a performance issue in this path? Do you have benchmarks that show some speedup due to this change? Thanks, Jeff > > Reviewed-by: Jeff Moyer <jmoyer@redhat.com> > Signed-off-by: Gu Zheng <guz.fnst@cn.fujitsu.com> > --- > fs/aio.c | 10 +++++----- > 1 files changed, 5 insertions(+), 5 deletions(-) > > diff --git a/fs/aio.c b/fs/aio.c > index f1fede2..df3491a 100644 > --- a/fs/aio.c > +++ b/fs/aio.c > @@ -1267,12 +1267,12 @@ static ssize_t aio_setup_vectored_rw(struct kiocb *kiocb, > if (compat) > ret = compat_rw_copy_check_uvector(rw, > (struct compat_iovec __user *)buf, > - *nr_segs, 1, *iovec, iovec); > + *nr_segs, UIO_FASTIOV, *iovec, iovec); > else > #endif > ret = rw_copy_check_uvector(rw, > (struct iovec __user *)buf, > - *nr_segs, 1, *iovec, iovec); > + *nr_segs, UIO_FASTIOV, *iovec, iovec); > if (ret < 0) > return ret; > > @@ -1309,7 +1309,7 @@ static ssize_t aio_run_iocb(struct kiocb *req, unsigned opcode, > fmode_t mode; > aio_rw_op *rw_op; > rw_iter_op *iter_op; > - struct iovec inline_vec, *iovec = &inline_vec; > + struct iovec inline_vecs[UIO_FASTIOV], *iovec = inline_vecs; > struct iov_iter iter; > > switch (opcode) { > @@ -1344,7 +1344,7 @@ rw_common: > if (!ret) > ret = rw_verify_area(rw, file, &req->ki_pos, req->ki_nbytes); > if (ret < 0) { > - if (iovec != &inline_vec) > + if (iovec != inline_vecs) > kfree(iovec); > return ret; > } > @@ -1391,7 +1391,7 @@ rw_common: > return -EINVAL; > } > > - if (iovec != &inline_vec) > + if (iovec != inline_vecs) > kfree(iovec); > > if (ret != -EIOCBQUEUED) { ^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: [RESEND PATCH 4/4] aio: use iovec array rather than the single one 2014-07-23 13:25 ` [RESEND PATCH 4/4] aio: use iovec array rather than the single one Jeff Moyer @ 2014-07-24 1:15 ` Gu Zheng 0 siblings, 0 replies; 6+ messages in thread From: Gu Zheng @ 2014-07-24 1:15 UTC (permalink / raw) To: Jeff Moyer; +Cc: bcrl, axboe, akpm, linux-aio, linux-kernel Hi Jeff, On 07/23/2014 09:25 PM, Jeff Moyer wrote: > Gu Zheng <guz.fnst@cn.fujitsu.com> writes: > >> Previously, we only offer a single iovec to handle all the read/write cases, so >> the PREADV/PWRITEV request always need to alloc more iovec buffer when copying >> user vectors. >> If we use a tmp iovec array rather than the single one, some small PREADV/PWRITEV >> workloads(vector size small than the tmp buffer) will not need to alloc more >> iovec buffer when copying user vectors. > > Hi, Gu, > > This still doesn't explain why you decided to look into this. The comment is clear, just want to avoid some needless memory allocation in the io submit path. > Did you > notice a performance issue in this path? Do you have benchmarks that > show some speedup due to this change? Just some common tests based on fio, it gains a slight improvement(~3%) when the iodepth in [5,6,7] than before. I did not paste these info here, because I think other guys(especially the guys have production environment) who are interested in this can give us more meaningful feedback. Thanks, Gu > > Thanks, > Jeff > >> >> Reviewed-by: Jeff Moyer <jmoyer@redhat.com> >> Signed-off-by: Gu Zheng <guz.fnst@cn.fujitsu.com> >> --- >> fs/aio.c | 10 +++++----- >> 1 files changed, 5 insertions(+), 5 deletions(-) >> >> diff --git a/fs/aio.c b/fs/aio.c >> index f1fede2..df3491a 100644 >> --- a/fs/aio.c >> +++ b/fs/aio.c >> @@ -1267,12 +1267,12 @@ static ssize_t aio_setup_vectored_rw(struct kiocb *kiocb, >> if (compat) >> ret = compat_rw_copy_check_uvector(rw, >> (struct compat_iovec __user *)buf, >> - *nr_segs, 1, *iovec, iovec); >> + *nr_segs, UIO_FASTIOV, *iovec, iovec); >> else >> #endif >> ret = rw_copy_check_uvector(rw, >> (struct iovec __user *)buf, >> - *nr_segs, 1, *iovec, iovec); >> + *nr_segs, UIO_FASTIOV, *iovec, iovec); >> if (ret < 0) >> return ret; >> >> @@ -1309,7 +1309,7 @@ static ssize_t aio_run_iocb(struct kiocb *req, unsigned opcode, >> fmode_t mode; >> aio_rw_op *rw_op; >> rw_iter_op *iter_op; >> - struct iovec inline_vec, *iovec = &inline_vec; >> + struct iovec inline_vecs[UIO_FASTIOV], *iovec = inline_vecs; >> struct iov_iter iter; >> >> switch (opcode) { >> @@ -1344,7 +1344,7 @@ rw_common: >> if (!ret) >> ret = rw_verify_area(rw, file, &req->ki_pos, req->ki_nbytes); >> if (ret < 0) { >> - if (iovec != &inline_vec) >> + if (iovec != inline_vecs) >> kfree(iovec); >> return ret; >> } >> @@ -1391,7 +1391,7 @@ rw_common: >> return -EINVAL; >> } >> >> - if (iovec != &inline_vec) >> + if (iovec != inline_vecs) >> kfree(iovec); >> >> if (ret != -EIOCBQUEUED) { > . > ^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: [RESEND PATCH 4/4] aio: use iovec array rather than the single one [not found] ` <1406109834-4414-4-git-send-email-guz.fnst@cn.fujitsu.com> 2014-07-23 13:25 ` [RESEND PATCH 4/4] aio: use iovec array rather than the single one Jeff Moyer @ 2014-07-24 15:56 ` Benjamin LaHaise 1 sibling, 0 replies; 6+ messages in thread From: Benjamin LaHaise @ 2014-07-24 15:56 UTC (permalink / raw) To: Gu Zheng; +Cc: axboe, akpm, jmoyer, linux-aio, linux-kernel On Wed, Jul 23, 2014 at 06:03:54PM +0800, Gu Zheng wrote: > Previously, we only offer a single iovec to handle all the read/write cases, so > the PREADV/PWRITEV request always need to alloc more iovec buffer when copying > user vectors. > If we use a tmp iovec array rather than the single one, some small PREADV/PWRITEV > workloads(vector size small than the tmp buffer) will not need to alloc more > iovec buffer when copying user vectors. Thanks Gu -- applied to my aio-next tree at git://git.kvack.org/~bcrl/aio-next.git . -ben > Reviewed-by: Jeff Moyer <jmoyer@redhat.com> > Signed-off-by: Gu Zheng <guz.fnst@cn.fujitsu.com> > --- > fs/aio.c | 10 +++++----- > 1 files changed, 5 insertions(+), 5 deletions(-) > > diff --git a/fs/aio.c b/fs/aio.c > index f1fede2..df3491a 100644 > --- a/fs/aio.c > +++ b/fs/aio.c > @@ -1267,12 +1267,12 @@ static ssize_t aio_setup_vectored_rw(struct kiocb *kiocb, > if (compat) > ret = compat_rw_copy_check_uvector(rw, > (struct compat_iovec __user *)buf, > - *nr_segs, 1, *iovec, iovec); > + *nr_segs, UIO_FASTIOV, *iovec, iovec); > else > #endif > ret = rw_copy_check_uvector(rw, > (struct iovec __user *)buf, > - *nr_segs, 1, *iovec, iovec); > + *nr_segs, UIO_FASTIOV, *iovec, iovec); > if (ret < 0) > return ret; > > @@ -1309,7 +1309,7 @@ static ssize_t aio_run_iocb(struct kiocb *req, unsigned opcode, > fmode_t mode; > aio_rw_op *rw_op; > rw_iter_op *iter_op; > - struct iovec inline_vec, *iovec = &inline_vec; > + struct iovec inline_vecs[UIO_FASTIOV], *iovec = inline_vecs; > struct iov_iter iter; > > switch (opcode) { > @@ -1344,7 +1344,7 @@ rw_common: > if (!ret) > ret = rw_verify_area(rw, file, &req->ki_pos, req->ki_nbytes); > if (ret < 0) { > - if (iovec != &inline_vec) > + if (iovec != inline_vecs) > kfree(iovec); > return ret; > } > @@ -1391,7 +1391,7 @@ rw_common: > return -EINVAL; > } > > - if (iovec != &inline_vec) > + if (iovec != inline_vecs) > kfree(iovec); > > if (ret != -EIOCBQUEUED) { > -- > 1.7.7 -- "Thought is the essence of where you are now." ^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: [RESEND PATCH 1/4] aio: remove the needless registration of ring file's private_data [not found] <1406109834-4414-1-git-send-email-guz.fnst@cn.fujitsu.com> [not found] ` <1406109834-4414-4-git-send-email-guz.fnst@cn.fujitsu.com> @ 2014-07-24 15:55 ` Benjamin LaHaise [not found] ` <1406109834-4414-2-git-send-email-guz.fnst@cn.fujitsu.com> [not found] ` <1406109834-4414-3-git-send-email-guz.fnst@cn.fujitsu.com> 3 siblings, 0 replies; 6+ messages in thread From: Benjamin LaHaise @ 2014-07-24 15:55 UTC (permalink / raw) To: Gu Zheng; +Cc: axboe, akpm, jmoyer, linux-aio, linux-kernel On Wed, Jul 23, 2014 at 06:03:51PM +0800, Gu Zheng wrote: > Remove the registration of ring file's private_data, we do not use > it. Thanks Gu -- applied to my aio-next tree at git://git.kvack.org/~bcrl/aio-next.git . -ben > Reviewed-by: Jeff Moyer <jmoyer@redhat.com> > Signed-off-by: Gu Zheng <guz.fnst@cn.fujitsu.com> > --- > fs/aio.c | 1 - > 1 files changed, 0 insertions(+), 1 deletions(-) > > diff --git a/fs/aio.c b/fs/aio.c > index 1c9c5f0..94546ab 100644 > --- a/fs/aio.c > +++ b/fs/aio.c > @@ -192,7 +192,6 @@ static struct file *aio_private_file(struct kioctx *ctx, loff_t nr_pages) > } > > file->f_flags = O_RDWR; > - file->private_data = ctx; > return file; > } > > -- > 1.7.7 -- "Thought is the essence of where you are now." ^ permalink raw reply [flat|nested] 6+ messages in thread
[parent not found: <1406109834-4414-2-git-send-email-guz.fnst@cn.fujitsu.com>]
* Re: [RESEND PATCH 2/4] aio: use the macro rather than the inline magic number [not found] ` <1406109834-4414-2-git-send-email-guz.fnst@cn.fujitsu.com> @ 2014-07-24 15:55 ` Benjamin LaHaise 0 siblings, 0 replies; 6+ messages in thread From: Benjamin LaHaise @ 2014-07-24 15:55 UTC (permalink / raw) To: Gu Zheng; +Cc: axboe, akpm, jmoyer, linux-aio, linux-kernel On Wed, Jul 23, 2014 at 06:03:52PM +0800, Gu Zheng wrote: > Replace the inline magic number with the ready-made macro(AIO_RING_MAGIC), > just clean up. Thanks Gu -- applied to my aio-next tree at git://git.kvack.org/~bcrl/aio-next.git . -ben > Reviewed-by: Jeff Moyer <jmoyer@redhat.com> > Signed-off-by: Gu Zheng <guz.fnst@cn.fujitsu.com> > --- > fs/aio.c | 2 +- > 1 files changed, 1 insertions(+), 1 deletions(-) > > diff --git a/fs/aio.c b/fs/aio.c > index 94546ab..e0f838d 100644 > --- a/fs/aio.c > +++ b/fs/aio.c > @@ -201,7 +201,7 @@ static struct dentry *aio_mount(struct file_system_type *fs_type, > static const struct dentry_operations ops = { > .d_dname = simple_dname, > }; > - return mount_pseudo(fs_type, "aio:", NULL, &ops, 0xa10a10a1); > + return mount_pseudo(fs_type, "aio:", NULL, &ops, AIO_RING_MAGIC); > } > > /* aio_setup > -- > 1.7.7 -- "Thought is the essence of where you are now." ^ permalink raw reply [flat|nested] 6+ messages in thread
[parent not found: <1406109834-4414-3-git-send-email-guz.fnst@cn.fujitsu.com>]
* Re: [RESEND PATCH 3/4] aio: fix some comments [not found] ` <1406109834-4414-3-git-send-email-guz.fnst@cn.fujitsu.com> @ 2014-07-24 15:56 ` Benjamin LaHaise 0 siblings, 0 replies; 6+ messages in thread From: Benjamin LaHaise @ 2014-07-24 15:56 UTC (permalink / raw) To: Gu Zheng; +Cc: axboe, akpm, jmoyer, linux-aio, linux-kernel On Wed, Jul 23, 2014 at 06:03:53PM +0800, Gu Zheng wrote: > The function comments of aio_run_iocb and aio_read_events are out of date, so > fix them here. Thanks Gu -- applied to my aio-next tree at git://git.kvack.org/~bcrl/aio-next.git . -ben > Reviewed-by: Jeff Moyer <jmoyer@redhat.com> > Signed-off-by: Gu Zheng <guz.fnst@cn.fujitsu.com> > --- > fs/aio.c | 7 +++---- > 1 files changed, 3 insertions(+), 4 deletions(-) > > diff --git a/fs/aio.c b/fs/aio.c > index e0f838d..f1fede2 100644 > --- a/fs/aio.c > +++ b/fs/aio.c > @@ -1044,7 +1044,7 @@ void aio_complete(struct kiocb *iocb, long res, long res2) > } > EXPORT_SYMBOL(aio_complete); > > -/* aio_read_events > +/* aio_read_events_ring > * Pull an event off of the ioctx's event ring. Returns the number of > * events fetched > */ > @@ -1296,9 +1296,8 @@ static ssize_t aio_setup_single_vector(struct kiocb *kiocb, > } > > /* > - * aio_setup_iocb: > - * Performs the initial checks and aio retry method > - * setup for the kiocb at the time of io submission. > + * aio_run_iocb: > + * Performs the initial checks and io submission. > */ > static ssize_t aio_run_iocb(struct kiocb *req, unsigned opcode, > char __user *buf, bool compat) > -- > 1.7.7 -- "Thought is the essence of where you are now." ^ permalink raw reply [flat|nested] 6+ messages in thread
end of thread, other threads:[~2014-07-24 15:56 UTC | newest] Thread overview: 6+ messages (download: mbox.gz / follow: Atom feed) -- links below jump to the message on this page -- [not found] <1406109834-4414-1-git-send-email-guz.fnst@cn.fujitsu.com> [not found] ` <1406109834-4414-4-git-send-email-guz.fnst@cn.fujitsu.com> 2014-07-23 13:25 ` [RESEND PATCH 4/4] aio: use iovec array rather than the single one Jeff Moyer 2014-07-24 1:15 ` Gu Zheng 2014-07-24 15:56 ` Benjamin LaHaise 2014-07-24 15:55 ` [RESEND PATCH 1/4] aio: remove the needless registration of ring file's private_data Benjamin LaHaise [not found] ` <1406109834-4414-2-git-send-email-guz.fnst@cn.fujitsu.com> 2014-07-24 15:55 ` [RESEND PATCH 2/4] aio: use the macro rather than the inline magic number Benjamin LaHaise [not found] ` <1406109834-4414-3-git-send-email-guz.fnst@cn.fujitsu.com> 2014-07-24 15:56 ` [RESEND PATCH 3/4] aio: fix some comments Benjamin LaHaise
This is a public inbox, see mirroring instructions for how to clone and mirror all data and code used for this inbox; as well as URLs for NNTP newsgroup(s).