From: Hugh Dickins <hugh.dickins@tiscali.co.uk>
To: Nitin Gupta <ngupta@vflare.org>
Cc: Pekka Enberg <penberg@cs.helsinki.fi>,
Al Viro <viro@ZenIV.linux.org.uk>,
Christoph Hellwig <hch@infradead.org>,
Jens Axboe <jens.axboe@oracle.com>,
Andi Kleen <andi@firstfloor.org>,
Andrew Morton <akpm@linux-foundation.org>,
linux-kernel <linux-kernel@vger.kernel.org>
Subject: Re: [PATCH 2/2] [mmotm] ramzswap: use slot free callback to eliminate stale data
Date: Tue, 5 Jan 2010 10:58:58 +0000 (GMT) [thread overview]
Message-ID: <alpine.LSU.2.00.1001051058070.26135@sister.anvils> (raw)
In-Reply-To: <1262661514-4148-2-git-send-email-ngupta@vflare.org>
On Tue, 5 Jan 2010, Nitin Gupta wrote:
> ramzswap driver creates RAM backed block devices which are
> used as swap disks. Pages swapped to these disks are compressed
> and stored in memory itself. However, when a swap page becomes
> stale i.e. it is no longer referenced by any process (say, when
> owning process exits), the driver does not get any notification
> about this. So, it has to keep such pages in memory until kernel
> swaps to the same swap slot again thereby overwriting previous
> (stale) page.
>
> Often, a large number of such stale pages accumulate which defeats
> the whole purpose of in-memory compressed swapping and it begins
> to have a negative impact on system performance.
>
> To overcome this problem, we now register a callback function
> which is called as soon as a swap slot is freed which allows
> us to free corresponding memory, eliminating any stale data
> in (compressed) memory.
>
> Signed-off-by: Nitin Gupta <ngupta@vflare.org>
Acked-by: Hugh Dickins <hugh.dickins@tiscali.co.uk>
> ---
> drivers/staging/ramzswap/ramzswap_drv.c | 15 ++++++++++++++-
> drivers/staging/ramzswap/ramzswap_drv.h | 1 +
> drivers/staging/ramzswap/ramzswap_ioctl.h | 1 +
> 3 files changed, 16 insertions(+), 1 deletions(-)
>
> diff --git a/drivers/staging/ramzswap/ramzswap_drv.c b/drivers/staging/ramzswap/ramzswap_drv.c
> index b839f05..bada4ae 100644
> --- a/drivers/staging/ramzswap/ramzswap_drv.c
> +++ b/drivers/staging/ramzswap/ramzswap_drv.c
> @@ -276,6 +276,7 @@ void ramzswap_ioctl_get_stats(struct ramzswap *rzs,
> s->failed_reads = rs->failed_reads;
> s->failed_writes = rs->failed_writes;
> s->invalid_io = rs->invalid_io;
> + s->notify_free = rs->notify_free;
> s->pages_zero = rs->pages_zero;
>
> s->good_compress_pct = good_compress_perc;
> @@ -1310,9 +1311,21 @@ out:
> return ret;
> }
>
> +void ramzswap_slot_free_notify(struct block_device *bdev, unsigned long index)
> +{
> + struct ramzswap *rzs;
> +
> + rzs = bdev->bd_disk->private_data;
> + ramzswap_free_page(rzs, index);
> + stat_inc(rzs->stats.notify_free);
> +
> + return;
> +}
> +
> static struct block_device_operations ramzswap_devops = {
> .ioctl = ramzswap_ioctl,
> - .owner = THIS_MODULE,
> + .swap_slot_free_notify = ramzswap_slot_free_notify,
> + .owner = THIS_MODULE
> };
>
> static void create_device(struct ramzswap *rzs, int device_id)
> diff --git a/drivers/staging/ramzswap/ramzswap_drv.h b/drivers/staging/ramzswap/ramzswap_drv.h
> index a6ea240..adc841a 100644
> --- a/drivers/staging/ramzswap/ramzswap_drv.h
> +++ b/drivers/staging/ramzswap/ramzswap_drv.h
> @@ -124,6 +124,7 @@ struct ramzswap_stats {
> u64 failed_reads; /* can happen when memory is too low */
> u64 failed_writes; /* should NEVER! happen */
> u64 invalid_io; /* non-swap I/O requests */
> + u64 notify_free; /* no. of swap slot free notifications */
> u32 pages_zero; /* no. of zero filled pages */
> u32 pages_stored; /* no. of pages currently stored */
> u32 good_compress; /* % of pages with compression ratio<=50% */
> diff --git a/drivers/staging/ramzswap/ramzswap_ioctl.h b/drivers/staging/ramzswap/ramzswap_ioctl.h
> index c713a09..ec50416 100644
> --- a/drivers/staging/ramzswap/ramzswap_ioctl.h
> +++ b/drivers/staging/ramzswap/ramzswap_ioctl.h
> @@ -27,6 +27,7 @@ struct ramzswap_ioctl_stats {
> u64 failed_reads; /* can happen when memory is too low */
> u64 failed_writes; /* should NEVER! happen */
> u64 invalid_io; /* non-swap I/O requests */
> + u64 notify_free; /* no. of swap slot free notifications */
> u32 pages_zero; /* no. of zero filled pages */
> u32 good_compress_pct; /* no. of pages with compression ratio<=50% */
> u32 pages_expand_pct; /* no. of incompressible pages */
> --
> 1.6.2.5
next prev parent reply other threads:[~2010-01-05 10:59 UTC|newest]
Thread overview: 5+ messages / expand[flat|nested] mbox.gz Atom feed top
2010-01-05 3:18 [PATCH 1/2] [mmotm] Add swap slot free callback to block_device_operations Nitin Gupta
2010-01-05 3:18 ` [PATCH 2/2] [mmotm] ramzswap: use slot free callback to eliminate stale data Nitin Gupta
2010-01-05 10:58 ` Hugh Dickins [this message]
2010-01-05 10:58 ` [PATCH 1/2] [mmotm] Add swap slot free callback to block_device_operations Hugh Dickins
2010-01-06 13:40 ` Nitin Gupta
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=alpine.LSU.2.00.1001051058070.26135@sister.anvils \
--to=hugh.dickins@tiscali.co.uk \
--cc=akpm@linux-foundation.org \
--cc=andi@firstfloor.org \
--cc=hch@infradead.org \
--cc=jens.axboe@oracle.com \
--cc=linux-kernel@vger.kernel.org \
--cc=ngupta@vflare.org \
--cc=penberg@cs.helsinki.fi \
--cc=viro@ZenIV.linux.org.uk \
/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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox
all inboxes | Powered by JetHome®