Message ID | 20170112183848.23159-1-toshi.kani@hpe.com (mailing list archive) |
---|---|
State | New, archived |
Headers | show |
On Thu, 2017-01-12 at 11:38 -0700, Toshi Kani wrote: > DAX IO path does not support iostat, but its metadata IO path does. > Therefore, iostat shows metadata IO statistics only, which has been > confusing to users. [] > diff --git a/fs/dax.c b/fs/dax.c [] > @@ -1058,12 +1058,24 @@ dax_iomap_rw(struct kiocb *iocb, struct iov_iter *iter, > { > struct address_space *mapping = iocb->ki_filp->f_mapping; > struct inode *inode = mapping->host; > + struct gendisk *disk = inode->i_sb->s_bdev->bd_disk; > loff_t pos = iocb->ki_pos, ret = 0, done = 0; > unsigned flags = 0; > + unsigned long start = 0; > + int do_acct = blk_queue_io_stat(disk->queue); > > if (iov_iter_rw(iter) == WRITE) > flags |= IOMAP_WRITE; > > + if (do_acct) { > + sector_t sec = iov_iter_count(iter) >> 9; > + > + start = jiffies; > + generic_start_io_acct(iov_iter_rw(iter), > + min_t(unsigned long, 1, sec), I believe I mislead you with a thinko. Your original code was (!sec) ? 1 : sec and I suggested incorrectly using min_t It should of course be max_t. Sorry. Also, as sec is now sector_t (u64), perhaps this unsigned long cast is incorrect.
On Thu, 2017-01-12 at 10:02 -0800, Joe Perches wrote: > On Thu, 2017-01-12 at 11:38 -0700, Toshi Kani wrote: > > DAX IO path does not support iostat, but its metadata IO path does. > > Therefore, iostat shows metadata IO statistics only, which has been > > confusing to users. > > [] > > diff --git a/fs/dax.c b/fs/dax.c > > [] > > @@ -1058,12 +1058,24 @@ dax_iomap_rw(struct kiocb *iocb, struct > > iov_iter *iter, > > { > > struct address_space *mapping = iocb->ki_filp->f_mapping; > > struct inode *inode = mapping->host; > > + struct gendisk *disk = inode->i_sb->s_bdev->bd_disk; > > loff_t pos = iocb->ki_pos, ret = 0, done = 0; > > unsigned flags = 0; > > + unsigned long start = 0; > > + int do_acct = blk_queue_io_stat(disk->queue); > > > > if (iov_iter_rw(iter) == WRITE) > > flags |= IOMAP_WRITE; > > > > + if (do_acct) { > > + sector_t sec = iov_iter_count(iter) >> 9; > > + > > + start = jiffies; > > + generic_start_io_acct(iov_iter_rw(iter), > > + min_t(unsigned long, 1, > > sec), > > I believe I mislead you with a thinko. > > Your original code was > (!sec) ? 1 : sec > and I suggested incorrectly using min_t > > It should of course be max_t. Sorry. My bad. I should have caught it. > Also, as sec is now sector_t (u64), perhaps this > unsigned long cast is incorrect. I see. Since iov_iter_count() returns a size_t value, I will use 'size_t' for 'sec' as you originally suggested. Thanks, -Toshi
On Thu, Jan 12, 2017 at 11:38:48AM -0700, Toshi Kani wrote: > DAX IO path does not support iostat, but its metadata IO path does. > Therefore, iostat shows metadata IO statistics only, which has been > confusing to users. > > Add iostat support to the DAX read/write path. > > Note, iostat still does not support the DAX mmap path as it allows > user applications to access directly. NAK. DAX I/O should not be accounted for block device statistics.
Christoph Hellwig <hch@infradead.org> writes: > On Thu, Jan 12, 2017 at 11:38:48AM -0700, Toshi Kani wrote: >> DAX IO path does not support iostat, but its metadata IO path does. >> Therefore, iostat shows metadata IO statistics only, which has been >> confusing to users. >> >> Add iostat support to the DAX read/write path. >> >> Note, iostat still does not support the DAX mmap path as it allows >> user applications to access directly. > > NAK. DAX I/O should not be accounted for block device statistics. Agreed, this is a layering violation. Cheers, Jeff
On Wed, 2017-01-25 at 10:15 -0500, Jeff Moyer wrote: > Christoph Hellwig <hch@infradead.org> writes: > > > On Thu, Jan 12, 2017 at 11:38:48AM -0700, Toshi Kani wrote: > > > DAX IO path does not support iostat, but its metadata IO path > > > does. Therefore, iostat shows metadata IO statistics only, which > > > has been confusing to users. > > > > > > Add iostat support to the DAX read/write path. > > > > > > Note, iostat still does not support the DAX mmap path as it > > > allows user applications to access directly. > > > > NAK. DAX I/O should not be accounted for block device statistics. > > Agreed, this is a layering violation. I will check to see if it can fit on top of Dan's patch-set. Thanks, -Toshi
diff --git a/fs/dax.c b/fs/dax.c index 5c74f60..a3e406a 100644 --- a/fs/dax.c +++ b/fs/dax.c @@ -1058,12 +1058,24 @@ dax_iomap_rw(struct kiocb *iocb, struct iov_iter *iter, { struct address_space *mapping = iocb->ki_filp->f_mapping; struct inode *inode = mapping->host; + struct gendisk *disk = inode->i_sb->s_bdev->bd_disk; loff_t pos = iocb->ki_pos, ret = 0, done = 0; unsigned flags = 0; + unsigned long start = 0; + int do_acct = blk_queue_io_stat(disk->queue); if (iov_iter_rw(iter) == WRITE) flags |= IOMAP_WRITE; + if (do_acct) { + sector_t sec = iov_iter_count(iter) >> 9; + + start = jiffies; + generic_start_io_acct(iov_iter_rw(iter), + min_t(unsigned long, 1, sec), + &disk->part0); + } + while (iov_iter_count(iter)) { ret = iomap_apply(inode, pos, iov_iter_count(iter), flags, ops, iter, dax_iomap_actor); @@ -1073,6 +1085,9 @@ dax_iomap_rw(struct kiocb *iocb, struct iov_iter *iter, done += ret; } + if (do_acct) + generic_end_io_acct(iov_iter_rw(iter), &disk->part0, start); + iocb->ki_pos += done; return done ? done : ret; }
DAX IO path does not support iostat, but its metadata IO path does. Therefore, iostat shows metadata IO statistics only, which has been confusing to users. Add iostat support to the DAX read/write path. Note, iostat still does not support the DAX mmap path as it allows user applications to access directly. Signed-off-by: Toshi Kani <toshi.kani@hpe.com> Cc: Andrew Morton <akpm@linux-foundation.org> Cc: Dan Williams <dan.j.williams@intel.com> Cc: Alexander Viro <viro@zeniv.linux.org.uk> Cc: Dave Chinner <david@fromorbit.com> Cc: Ross Zwisler <ross.zwisler@linux.intel.com> Cc: Joe Perches <joe@perches.com> --- v5: - Add a flag in case 'start' is 0 after 'jiffies' rolls over. (Dan Williams) - Fix a signed/unsigned conversion. (Joe Perches) --- fs/dax.c | 15 +++++++++++++++ 1 file changed, 15 insertions(+)