Re: [BLOCK] Unify the seperate read/write io stat fields into arrays

[Jens Axboe <axboe@suse.de>]

On Thu, Nov 03 2005, Jeff Garzik wrote:
> Linux Kernel Mailing List wrote:
> >tree fe4ce823e638ded151edcb142f28a240860f0d33
> >parent d72d904a5367ad4ca3f2c9a2ce8c3a68f0b28bf0
> >author Jens Axboe <axboe@suse.de> Tue, 01 Nov 2005 09:26:16 +0100
> >committer Jens Axboe <axboe@suse.de> Tue, 01 Nov 2005 09:26:16 +0100
> >
> >[BLOCK] Unify the seperate read/write io stat fields into arrays
> >
> >Instead of having ->read_sectors and ->write_sectors, combine the two
> >into ->sectors[2] and similar for the other fields. This saves a branch
> >several places in the io path, since we don't have to care for what the
> >actual io direction is. On my x86-64 box, that's 200 bytes less text in
> >just the core (not counting the various drivers).
> >
> >Signed-off-by: Jens Axboe <axboe@suse.de>
> >
> > drivers/block/genhd.c     |   29 ++++++++++++++---------------
> > drivers/block/ll_rw_blk.c |   40 ++++++++++++----------------------------
> > drivers/md/linear.c       |   10 +++-------
> > drivers/md/md.c           |    4 ++--
> > drivers/md/multipath.c    |   10 +++-------
> > drivers/md/raid0.c        |   10 +++-------
> > drivers/md/raid1.c        |   12 ++++--------
> > drivers/md/raid10.c       |   12 ++++--------
> > drivers/md/raid5.c        |   10 +++-------
> > drivers/md/raid6main.c    |   12 ++++--------
> > fs/partitions/check.c     |    7 ++++---
> > include/linux/genhd.h     |   10 +++++-----
> > 12 files changed, 61 insertions(+), 105 deletions(-)
> >
> >diff --git a/drivers/block/genhd.c b/drivers/block/genhd.c
> >index 486ce1f..54aec4a 100644
> >--- a/drivers/block/genhd.c
> >+++ b/drivers/block/genhd.c
> >@@ -391,13 +391,12 @@ static ssize_t disk_stats_read(struct ge
> > 		"%8u %8u %8llu %8u "
> > 		"%8u %8u %8u"
> > 		"\n",
> >-		disk_stat_read(disk, reads), disk_stat_read(disk, 
> >read_merges),
> >-		(unsigned long long)disk_stat_read(disk, read_sectors),
> >-		jiffies_to_msecs(disk_stat_read(disk, read_ticks)),
> >-		disk_stat_read(disk, writes), 
> >-		disk_stat_read(disk, write_merges),
> >-		(unsigned long long)disk_stat_read(disk, write_sectors),
> >-		jiffies_to_msecs(disk_stat_read(disk, write_ticks)),
> >+		disk_stat_read(disk, ios[0]), disk_stat_read(disk, 
> >merges[0]),
> >+		(unsigned long long)disk_stat_read(disk, sectors[0]),
> >+		jiffies_to_msecs(disk_stat_read(disk, ticks[0])),
> >+		disk_stat_read(disk, ios[1]), disk_stat_read(disk, 
> >merges[1]),
> >+		(unsigned long long)disk_stat_read(disk, sectors[1]),
> >+		jiffies_to_msecs(disk_stat_read(disk, ticks[1])),
> > 		disk->in_flight,
> > 		jiffies_to_msecs(disk_stat_read(disk, io_ticks)),
> > 		jiffies_to_msecs(disk_stat_read(disk, time_in_queue)));
> >@@ -583,12 +582,12 @@ static int diskstats_show(struct seq_fil
> > 	preempt_enable();
> > 	seq_printf(s, "%4d %4d %s %u %u %llu %u %u %u %llu %u %u %u %u\n",
> > 		gp->major, n + gp->first_minor, disk_name(gp, n, buf),
> >-		disk_stat_read(gp, reads), disk_stat_read(gp, read_merges),
> >-		(unsigned long long)disk_stat_read(gp, read_sectors),
> >-		jiffies_to_msecs(disk_stat_read(gp, read_ticks)),
> >-		disk_stat_read(gp, writes), disk_stat_read(gp, write_merges),
> >-		(unsigned long long)disk_stat_read(gp, write_sectors),
> >-		jiffies_to_msecs(disk_stat_read(gp, write_ticks)),
> >+		disk_stat_read(gp, ios[0]), disk_stat_read(gp, merges[0]),
> >+		(unsigned long long)disk_stat_read(gp, sectors[0]),
> >+		jiffies_to_msecs(disk_stat_read(gp, ticks[0])),
> >+		disk_stat_read(gp, ios[1]), disk_stat_read(gp, merges[1]),
> >+		(unsigned long long)disk_stat_read(gp, sectors[1]),
> >+		jiffies_to_msecs(disk_stat_read(gp, ticks[1])),
> > 		gp->in_flight,
> > 		jiffies_to_msecs(disk_stat_read(gp, io_ticks)),
> > 		jiffies_to_msecs(disk_stat_read(gp, time_in_queue)));
> 
> While the overall patch is OK, the use of magic numbers makes the code 
> less readable.  fsck if I know whether "[0]" represents read or write.

Does it matter, you know that they are reads and writes obviously. But
yes, using READ/WRITE would be nicer. I'll queue it up.

-- 
Jens Axboe