From mboxrd@z Thu Jan 1 00:00:00 1970 From: Jeff Moyer Subject: Re: [patch|rfc] block: don't mark buffers beyond end of disk as mapped Date: Tue, 01 May 2012 17:40:35 -0400 Message-ID: References: Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Cc: linux-kernel@vger.kernel.org, linux-fsdevel@vger.kernel.org, jaxboe@fusionio.com, Kyle McMartin To: Nick Piggin Return-path: Received: from mx1.redhat.com ([209.132.183.28]:3356 "EHLO mx1.redhat.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753089Ab2EAVki (ORCPT ); Tue, 1 May 2012 17:40:38 -0400 In-Reply-To: (Jeff Moyer's message of "Tue, 01 May 2012 16:37:10 -0400") Sender: linux-fsdevel-owner@vger.kernel.org List-ID: Jeff Moyer writes: > Nick Piggin writes: > >> On 2 May 2012 00:08, Jeff Moyer wrote: >>> Nick Piggin writes: >>> >>>> Not a bad fix. But it's kind of sad to have i_size checking logic also in >>>> block_read_full_page, that does not cope with this. >>>> >>>> I have found there are parts of the kernel (readahead) that try to read >>>> beyond EOF and seem to get angry if we return an error (by not >>>> marking uptodate in readpage) in that case though :( >>>> >>>> But, either way, I think it's very reasonable to not mark buffers beyond >>>> end of device as mapped. So I think your patch is fine. >>>> >>>> I guess for ext[234], it does not read metadata close to the end of the >>>> device or you were using 4K sized blocks? >>> >>> Well, the test case just reads directly from the loop device, bypassing >>> the file system, and I did use 1KB blocks when making the file system, so >>> it is quite puzzling. >> >> It's because buffer_head creation does not go through the same paths >> for bdev file access versus getblk APIs. >> >> blkdev_get_block does the right thing there >> >> In fact, it's probably good to unify the checks here, i.e., use max_blocks() > > You really think it's worth it? I mean, it's just an i_size_read and a > shift, and there is precedent for it inside fs/buffer.c. I'd prefer to > keep the patch as-is, but will change it if you feel that strongly about > it. Anyway, here is the other version of the patch, using max_block as you suggested. Cheers, Jeff Signed-off-by: Jeff Moyer diff --git a/fs/block_dev.c b/fs/block_dev.c index e08f6a2..ba11c30 100644 --- a/fs/block_dev.c +++ b/fs/block_dev.c @@ -70,7 +70,7 @@ static void bdev_inode_switch_bdi(struct inode *inode, spin_unlock(&dst->wb.list_lock); } -static sector_t max_block(struct block_device *bdev) +sector_t blkdev_max_block(struct block_device *bdev) { sector_t retval = ~((sector_t)0); loff_t sz = i_size_read(bdev->bd_inode); @@ -163,7 +163,7 @@ static int blkdev_get_block(struct inode *inode, sector_t iblock, struct buffer_head *bh, int create) { - if (iblock >= max_block(I_BDEV(inode))) { + if (iblock >= blkdev_max_block(I_BDEV(inode))) { if (create) return -EIO; @@ -185,7 +185,7 @@ static int blkdev_get_blocks(struct inode *inode, sector_t iblock, struct buffer_head *bh, int create) { - sector_t end_block = max_block(I_BDEV(inode)); + sector_t end_block = blkdev_max_block(I_BDEV(inode)); unsigned long max_blocks = bh->b_size >> inode->i_blkbits; if ((iblock + max_blocks) > end_block) { diff --git a/fs/buffer.c b/fs/buffer.c index 351e18e..ad5938c 100644 --- a/fs/buffer.c +++ b/fs/buffer.c @@ -921,6 +921,7 @@ init_page_buffers(struct page *page, struct block_device *bdev, struct buffer_head *head = page_buffers(page); struct buffer_head *bh = head; int uptodate = PageUptodate(page); + sector_t end_block = blkdev_max_block(I_BDEV(bdev->bd_inode)); do { if (!buffer_mapped(bh)) { @@ -929,7 +930,8 @@ init_page_buffers(struct page *page, struct block_device *bdev, bh->b_blocknr = block; if (uptodate) set_buffer_uptodate(bh); - set_buffer_mapped(bh); + if (block < end_block) + set_buffer_mapped(bh); } block++; bh = bh->b_this_page; diff --git a/include/linux/fs.h b/include/linux/fs.h index 8de6755..25c40b9 100644 --- a/include/linux/fs.h +++ b/include/linux/fs.h @@ -2051,6 +2051,7 @@ extern void unregister_blkdev(unsigned int, const char *); extern struct block_device *bdget(dev_t); extern struct block_device *bdgrab(struct block_device *bdev); extern void bd_set_size(struct block_device *, loff_t size); +extern sector_t blkdev_max_block(struct block_device *bdev); extern void bd_forget(struct inode *inode); extern void bdput(struct block_device *); extern void invalidate_bdev(struct block_device *);