All of lore.kernel.org
 help / color / mirror / Atom feed
From: Chandan Rajendra <chandan@linux.vnet.ibm.com>
To: clm@fb.com, jbacik@fb.com, bo.li.liu@oracle.com, dsterba@suse.cz
Cc: Chandan Rajendra <chandan@linux.vnet.ibm.com>,
	linux-btrfs@vger.kernel.org, chandan@mykolab.com
Subject: [PATCH V15 10/15] Btrfs: subpage-blocksize: btrfs_punch_hole: Fix uptodate blocks check
Date: Thu, 11 Feb 2016 23:17:48 +0530	[thread overview]
Message-ID: <1455212873-14829-11-git-send-email-chandan@linux.vnet.ibm.com> (raw)
In-Reply-To: <1455212873-14829-1-git-send-email-chandan@linux.vnet.ibm.com>

In case of subpagesize-blocksize, the file blocks to be punched may map
only part of a page. For file blocks inside such pages, we need to check
for the presence of BLK_STATE_UPTODATE flag.

Signed-off-by: Chandan Rajendra <chandan@linux.vnet.ibm.com>
---
 fs/btrfs/file.c | 66 ++++++++++++++++++++++++++++++++++++++++++++++++++++++++-
 1 file changed, 65 insertions(+), 1 deletion(-)

diff --git a/fs/btrfs/file.c b/fs/btrfs/file.c
index b8b0eda..7f5e64c 100644
--- a/fs/btrfs/file.c
+++ b/fs/btrfs/file.c
@@ -2323,6 +2323,8 @@ static int btrfs_punch_hole(struct inode *inode, loff_t offset, loff_t len)
 	struct btrfs_path *path;
 	struct btrfs_block_rsv *rsv;
 	struct btrfs_trans_handle *trans;
+	struct address_space *mapping = inode->i_mapping;
+	pgoff_t start_index, end_index;
 	u64 lockstart;
 	u64 lockend;
 	u64 tail_start;
@@ -2335,6 +2337,7 @@ static int btrfs_punch_hole(struct inode *inode, loff_t offset, loff_t len)
 	int err = 0;
 	unsigned int rsv_count;
 	bool same_block;
+	bool same_page;
 	bool no_holes = btrfs_fs_incompat(root->fs_info, NO_HOLES);
 	u64 ino_size;
 	bool truncated_block = false;
@@ -2431,11 +2434,43 @@ static int btrfs_punch_hole(struct inode *inode, loff_t offset, loff_t len)
 		goto out_only_mutex;
 	}
 
+	start_index = lockstart >> PAGE_CACHE_SHIFT;
+	end_index = lockend >> PAGE_CACHE_SHIFT;
+
+	same_page = lockstart >> PAGE_CACHE_SHIFT
+		== lockend >> PAGE_CACHE_SHIFT;
+
 	while (1) {
 		struct btrfs_ordered_extent *ordered;
+		struct page *start_page = NULL;
+		struct page *end_page = NULL;
+		u64 nr_pages;
 
 		truncate_pagecache_range(inode, lockstart, lockend);
 
+		if (lockstart & (PAGE_CACHE_SIZE - 1)) {
+			start_page = find_or_create_page(mapping, start_index,
+							GFP_NOFS);
+			if (!start_page) {
+				mutex_unlock(&inode->i_mutex);
+				return -ENOMEM;
+			}
+		}
+
+		if (!same_page && ((lockend + 1) & (PAGE_CACHE_SIZE - 1))) {
+			end_page = find_or_create_page(mapping, end_index,
+						GFP_NOFS);
+			if (!end_page) {
+				if (start_page) {
+					unlock_page(start_page);
+					page_cache_release(start_page);
+				}
+				mutex_unlock(&inode->i_mutex);
+				return -ENOMEM;
+			}
+		}
+
+
 		lock_extent_bits(&BTRFS_I(inode)->io_tree, lockstart, lockend,
 				 &cached_state);
 		ordered = btrfs_lookup_first_ordered_extent(inode, lockend);
@@ -2445,18 +2480,47 @@ static int btrfs_punch_hole(struct inode *inode, loff_t offset, loff_t len)
 		 * and nobody raced in and read a page in this range, if we did
 		 * we need to try again.
 		 */
+		nr_pages = round_up(lockend, PAGE_CACHE_SIZE)
+			- round_down(lockstart, PAGE_CACHE_SIZE);
+		nr_pages >>= PAGE_CACHE_SHIFT;
+
 		if ((!ordered ||
 		    (ordered->file_offset + ordered->len <= lockstart ||
 		     ordered->file_offset > lockend)) &&
-		     !btrfs_page_exists_in_range(inode, lockstart, lockend)) {
+		     (!(start_page && PagePrivate(start_page) &&
+			test_page_blks_state(start_page, 1 << BLK_STATE_UPTODATE,
+			 lockstart,
+			 min(lockstart + PAGE_CACHE_SIZE - 1, lockend), 0)) &&
+		      !(end_page && PagePrivate(end_page) &&
+			test_page_blks_state(end_page, 1 << BLK_STATE_UPTODATE,
+			 page_offset(end_page), lockend, 0)) &&
+		      !(nr_pages > 2 && btrfs_page_exists_in_range(inode,
+					 round_up(lockstart, PAGE_CACHE_SIZE),
+					 round_down(lockend, PAGE_CACHE_SIZE) - 1)))) {
 			if (ordered)
 				btrfs_put_ordered_extent(ordered);
+			if (end_page) {
+				unlock_page(end_page);
+				page_cache_release(end_page);
+			}
+			if (start_page) {
+				unlock_page(start_page);
+				page_cache_release(start_page);
+			}
 			break;
 		}
 		if (ordered)
 			btrfs_put_ordered_extent(ordered);
 		unlock_extent_cached(&BTRFS_I(inode)->io_tree, lockstart,
 				     lockend, &cached_state, GFP_NOFS);
+		if (end_page) {
+			unlock_page(end_page);
+			page_cache_release(end_page);
+		}
+		if (start_page) {
+			unlock_page(start_page);
+			page_cache_release(start_page);
+		}
 		ret = btrfs_wait_ordered_range(inode, lockstart,
 					       lockend - lockstart + 1);
 		if (ret) {
-- 
2.1.0


  parent reply	other threads:[~2016-02-11 17:49 UTC|newest]

Thread overview: 23+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2016-02-11 17:47 [PATCH V15 00/15] Btrfs: Subpagesize-blocksize: Allow I/O on blocks whose size is less than page size Chandan Rajendra
2016-02-11 17:47 ` [PATCH V15 01/15] Btrfs: subpage-blocksize: Fix whole page read Chandan Rajendra
2016-02-11 17:47 ` [PATCH V15 02/15] Btrfs: subpage-blocksize: Fix whole page write Chandan Rajendra
2016-02-11 17:47 ` [PATCH V15 03/15] Btrfs: subpage-blocksize: Make sure delalloc range intersects with the locked page's range Chandan Rajendra
2016-02-11 17:47 ` [PATCH V15 04/15] Btrfs: subpage-blocksize: Define extent_buffer_head Chandan Rajendra
2016-02-11 17:47 ` [PATCH V15 05/15] Btrfs: subpage-blocksize: Read tree blocks whose size is < PAGE_SIZE Chandan Rajendra
2016-02-11 17:47 ` [PATCH V15 06/15] Btrfs: subpage-blocksize: Write only dirty extent buffers belonging to a page Chandan Rajendra
2016-02-11 17:47 ` [PATCH V15 07/15] Btrfs: subpage-blocksize: Allow mounting filesystems where sectorsize != PAGE_SIZE Chandan Rajendra
2016-02-11 17:47 ` [PATCH V15 08/15] Btrfs: subpage-blocksize: Deal with partial ordered extent allocations Chandan Rajendra
2016-02-11 17:47 ` [PATCH V15 09/15] Btrfs: subpage-blocksize: Explicitly track I/O status of blocks of an ordered extent Chandan Rajendra
2016-02-11 17:47 ` Chandan Rajendra [this message]
2016-02-11 17:47 ` [PATCH V15 11/15] Btrfs: subpage-blocksize: Prevent writes to an extent buffer when PG_writeback flag is set Chandan Rajendra
2016-02-11 17:47 ` [PATCH V15 12/15] Revert "btrfs: fix lockups from btrfs_clear_path_blocking" Chandan Rajendra
2016-02-11 17:47 ` [PATCH V15 13/15] Btrfs: subpage-blocksize: Fix file defragmentation code Chandan Rajendra
2016-02-11 17:47 ` [PATCH V15 14/15] Btrfs: subpage-blocksize: extent_clear_unlock_delalloc: Prevent page from being unlocked more than once Chandan Rajendra
2016-02-11 17:47 ` [PATCH V15 15/15] Btrfs: subpage-blocksize: Enable dedup ioctl Chandan Rajendra
2016-02-29  5:52 ` [PATCH V15 00/15] Btrfs: Subpagesize-blocksize: Allow I/O on blocks whose size is less than page size Chandan Rajendra
2016-02-29 12:11   ` David Sterba
2016-03-22 11:04 ` David Sterba
2016-03-22 13:20   ` Chandan Rajendra
2016-03-31  9:31     ` David Sterba
2016-03-31 13:59       ` David Sterba
2016-04-01  3:21         ` Chandan Rajendra

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=1455212873-14829-11-git-send-email-chandan@linux.vnet.ibm.com \
    --to=chandan@linux.vnet.ibm.com \
    --cc=bo.li.liu@oracle.com \
    --cc=chandan@mykolab.com \
    --cc=clm@fb.com \
    --cc=dsterba@suse.cz \
    --cc=jbacik@fb.com \
    --cc=linux-btrfs@vger.kernel.org \
    /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.