From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1756354Ab0EGJwW (ORCPT ); Fri, 7 May 2010 05:52:22 -0400 Received: from mail-pv0-f174.google.com ([74.125.83.174]:57979 "EHLO mail-pv0-f174.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1755441Ab0EGJwV (ORCPT ); Fri, 7 May 2010 05:52:21 -0400 Message-ID: <4BE3E208.1020002@vflare.org> Date: Fri, 07 May 2010 15:18:56 +0530 From: Nitin Gupta Reply-To: ngupta@vflare.org User-Agent: Mozilla/5.0 (X11; U; Linux x86_64; en-US; rv:1.9.1.9) Gecko/20100430 Fedora/3.0.4-2.fc11 Thunderbird/3.0.4 MIME-Version: 1.0 To: nigel@tuxonice.net CC: Greg KH , Linus Torvalds , Pekka Enberg , Hugh Dickins , Cyp , Minchan Kim , Al Viro , Christoph Hellwig , Jens Axboe , Andi Kleen , Andrew Morton , driverdev , linux-kernel Subject: Re: [PATCH 2/3] Add swap slot free callback to block_device_operations References: <1273217107-2023-1-git-send-email-ngupta@vflare.org> <1273217107-2023-3-git-send-email-ngupta@vflare.org> <4BE3DBD6.7070707@tuxonice.net> In-Reply-To: <4BE3DBD6.7070707@tuxonice.net> Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On 05/07/2010 02:52 PM, Nigel Cunningham wrote: > Hi. > > On 07/05/10 17:25, Nitin Gupta wrote: >> This callback is required when RAM based devices are used as swap disks. >> One such device is ramzswap which is used as compressed in-memory swap >> disk. For such devices, we need a callback as soon as a swap slot is no >> longer used to allow freeing memory allocated for this slot. Without >> this >> callback, stale data can quickly accumulate in memory defeating the whole >> purpose of such devices. >> >> Signed-off-by: Nitin Gupta >> --- >> include/linux/blkdev.h | 2 ++ >> mm/swapfile.c | 4 ++++ >> 2 files changed, 6 insertions(+), 0 deletions(-) >> >> diff --git a/include/linux/blkdev.h b/include/linux/blkdev.h >> index 6690e8b..413284a 100644 >> --- a/include/linux/blkdev.h >> +++ b/include/linux/blkdev.h >> @@ -1287,6 +1287,8 @@ struct block_device_operations { >> unsigned long long); >> int (*revalidate_disk) (struct gendisk *); >> int (*getgeo)(struct block_device *, struct hd_geometry *); >> + /* this callback is with swap_lock and sometimes page table lock >> held */ >> + void (*swap_slot_free_notify) (struct block_device *, unsigned >> long); >> struct module *owner; >> }; >> >> diff --git a/mm/swapfile.c b/mm/swapfile.c >> index ecb069e..f5ccc47 100644 >> --- a/mm/swapfile.c >> +++ b/mm/swapfile.c >> @@ -574,6 +574,7 @@ static unsigned char swap_entry_free(struct >> swap_info_struct *p, >> >> /* free if no reference */ >> if (!usage) { >> + struct gendisk *disk = p->bdev->bd_disk; >> if (offset< p->lowest_bit) >> p->lowest_bit = offset; >> if (offset> p->highest_bit) >> @@ -583,6 +584,9 @@ static unsigned char swap_entry_free(struct >> swap_info_struct *p, >> swap_list.next = p->type; >> nr_swap_pages++; >> p->inuse_pages--; >> + if ((p->flags& SWP_BLKDEV)&& >> + disk->fops->swap_slot_free_notify) >> + disk->fops->swap_slot_free_notify(p->bdev, offset); > > Is this p->flags & SWP_BLKDEV logic reversed? (Don't you want the > notifier called for devices that aren't backed by a block device?) > No, the logic here is correct: ramzswap is a block device for which we want this callback. Though its a RAM backed, it is still a block device. (I hope it answers your question in the other mail also). > I also wonder whether leaving the p->flags & SWP_BLKDEV part out might > be a good idea. Other potential notifier users? > For regular files, 'offset' used here makes little sense. For block devices, its simply offset in real device. Also, I doubt if *files* would ever like to have such a callback. Thanks, Nitin