From: Omar Sandoval <osandov@osandov.com>
To: Huijin Park <huijin.park@samsung.com>
Cc: Andreas Dilger <adilger.kernel@dilger.ca>,
Michael Callahan <michaelcallahan@fb.com>,
Omar Sandoval <osandov@fb.com>,
js07.lee@samsung.com, Huijin Park <bbanghj.park@gmail.com>,
linux-block@vger.kernel.org, linux-ext4@vger.kernel.org,
linux-kernel@vger.kernel.org
Subject: Re: [PATCH,1/2] genhd: avoid overflow of sectors in disk_stats
Date: Fri, 30 Nov 2018 11:56:07 -0800 [thread overview]
Message-ID: <20181130195607.GA27411@vader> (raw)
In-Reply-To: <1543570361-3168-1-git-send-email-huijin.park@samsung.com>
On Fri, Nov 30, 2018 at 04:32:40AM -0500, Huijin Park wrote:
> From: "huijin.park" <huijin.park@samsung.com>
>
> This patch changes the 'sectors' type to an u64.
> In 32 bit system, the 'sectors' can accumulate up to about 2TiB.
> If a 32 bit system makes i/o over 2TiB while running,
> the 'sectors' will overflow.
> As a result, the part_stat_read(sectors), the diskstats in proc and
> the (lifetime|session)_write_kbytes in sysfs return invalid statistic.
What about parsers which expect it to be an unsigned long? E.g., iostat:
https://github.com/sysstat/sysstat/blob/v12.1.1/rd_stats.c#L736
At least with glibc, scanf seems to truncate sanely, but this appears to
be undefined.
> Signed-off-by: huijin.park <huijin.park@samsung.com>
> ---
> block/genhd.c | 6 +++---
> include/linux/genhd.h | 2 +-
> 2 files changed, 4 insertions(+), 4 deletions(-)
>
> diff --git a/block/genhd.c b/block/genhd.c
> index 0145bcb..7518dcd 100644
> --- a/block/genhd.c
> +++ b/block/genhd.c
> @@ -1343,10 +1343,10 @@ static int diskstats_show(struct seq_file *seqf, void *v)
> part_stat_unlock();
> part_in_flight(gp->queue, hd, inflight);
> seq_printf(seqf, "%4d %7d %s "
> - "%lu %lu %lu %u "
> - "%lu %lu %lu %u "
> + "%lu %lu %llu %u "
> + "%lu %lu %llu %u "
> "%u %u %u "
> - "%lu %lu %lu %u\n",
> + "%lu %lu %llu %u\n",
> MAJOR(part_devt(hd)), MINOR(part_devt(hd)),
> disk_name(gp, hd->partno, buf),
> part_stat_read(hd, ios[STAT_READ]),
> diff --git a/include/linux/genhd.h b/include/linux/genhd.h
> index 0c5ee17..5bf86f9 100644
> --- a/include/linux/genhd.h
> +++ b/include/linux/genhd.h
> @@ -84,7 +84,7 @@ struct partition {
>
> struct disk_stats {
> u64 nsecs[NR_STAT_GROUPS];
> - unsigned long sectors[NR_STAT_GROUPS];
> + u64 sectors[NR_STAT_GROUPS];
> unsigned long ios[NR_STAT_GROUPS];
> unsigned long merges[NR_STAT_GROUPS];
> unsigned long io_ticks;
> --
> 1.7.9.5
>
next prev parent reply other threads:[~2018-11-30 19:56 UTC|newest]
Thread overview: 4+ messages / expand[flat|nested] mbox.gz Atom feed top
[not found] <CGME20181130093318eucas1p24feb8e649c8f244ce149a7f87e43c828@eucas1p2.samsung.com>
2018-11-30 9:32 ` [PATCH,1/2] genhd: avoid overflow of sectors in disk_stats Huijin Park
2018-11-30 9:32 ` [PATCH,2/2] ext4: change type to same as " Huijin Park
2018-11-30 19:56 ` Omar Sandoval [this message]
2018-12-09 14:44 ` [PATCH,1/2] genhd: avoid overflow of " Huijin Park
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20181130195607.GA27411@vader \
--to=osandov@osandov.com \
--cc=adilger.kernel@dilger.ca \
--cc=bbanghj.park@gmail.com \
--cc=huijin.park@samsung.com \
--cc=js07.lee@samsung.com \
--cc=linux-block@vger.kernel.org \
--cc=linux-ext4@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=michaelcallahan@fb.com \
--cc=osandov@fb.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.