linux-kernel - [PATCH AUTOSEL 4.19 099/237] ceph: only allow punch hole mode in fallocate

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:   Sat, 16 Nov 2019 10:38:54 -0500
From:   Sasha Levin <sashal@...nel.org>
To:     linux-kernel@...r.kernel.org, stable@...r.kernel.org
Cc:     Luis Henriques <lhenriques@...e.com>,
        "Yan, Zheng" <zyan@...hat.com>, Ilya Dryomov <idryomov@...il.com>,
        Sasha Levin <sashal@...nel.org>, ceph-devel@...r.kernel.org
Subject: [PATCH AUTOSEL 4.19 099/237] ceph: only allow punch hole mode in fallocate

From: Luis Henriques <lhenriques@...e.com>

[ Upstream commit bddff633ab7bc60a18a86ac8b322695b6f8594d0 ]

Current implementation of cephfs fallocate isn't correct as it doesn't
really reserve the space in the cluster, which means that a subsequent
call to a write may actually fail due to lack of space.  In fact, it is
currently possible to fallocate an amount space that is larger than the
free space in the cluster.  It has behaved this way since the initial
commit ad7a60de882a ("ceph: punch hole support").

Since there's no easy solution to fix this at the moment, this patch
simply removes support for all fallocate operations but
FALLOC_FL_PUNCH_HOLE (which implies FALLOC_FL_KEEP_SIZE).

Link: https://tracker.ceph.com/issues/36317
Signed-off-by: Luis Henriques <lhenriques@...e.com>
Reviewed-by: "Yan, Zheng" <zyan@...hat.com>
Signed-off-by: Ilya Dryomov <idryomov@...il.com>
Signed-off-by: Sasha Levin <sashal@...nel.org>
---
 fs/ceph/file.c | 45 +++++++++------------------------------------
 1 file changed, 9 insertions(+), 36 deletions(-)

diff --git a/fs/ceph/file.c b/fs/ceph/file.c
index 92ab204336829..91a7ad259bcf2 100644
--- a/fs/ceph/file.c
+++ b/fs/ceph/file.c
@@ -1735,7 +1735,6 @@ static long ceph_fallocate(struct file *file, int mode,
 	struct ceph_file_info *fi = file->private_data;
 	struct inode *inode = file_inode(file);
 	struct ceph_inode_info *ci = ceph_inode(inode);
-	struct ceph_fs_client *fsc = ceph_inode_to_client(inode);
 	struct ceph_cap_flush *prealloc_cf;
 	int want, got = 0;
 	int dirty;
@@ -1743,10 +1742,7 @@ static long ceph_fallocate(struct file *file, int mode,
 	loff_t endoff = 0;
 	loff_t size;
 
-	if ((offset + length) > max(i_size_read(inode), fsc->max_file_size))
-		return -EFBIG;
-
-	if (mode & ~(FALLOC_FL_KEEP_SIZE | FALLOC_FL_PUNCH_HOLE))
+	if (mode != (FALLOC_FL_KEEP_SIZE | FALLOC_FL_PUNCH_HOLE))
 		return -EOPNOTSUPP;
 
 	if (!S_ISREG(inode->i_mode))
@@ -1763,18 +1759,6 @@ static long ceph_fallocate(struct file *file, int mode,
 		goto unlock;
 	}
 
-	if (!(mode & (FALLOC_FL_PUNCH_HOLE | FALLOC_FL_KEEP_SIZE)) &&
-	    ceph_quota_is_max_bytes_exceeded(inode, offset + length)) {
-		ret = -EDQUOT;
-		goto unlock;
-	}
-
-	if (ceph_osdmap_flag(&fsc->client->osdc, CEPH_OSDMAP_FULL) &&
-	    !(mode & FALLOC_FL_PUNCH_HOLE)) {
-		ret = -ENOSPC;
-		goto unlock;
-	}
-
 	if (ci->i_inline_version != CEPH_INLINE_NONE) {
 		ret = ceph_uninline_data(file, NULL);
 		if (ret < 0)
@@ -1782,12 +1766,12 @@ static long ceph_fallocate(struct file *file, int mode,
 	}
 
 	size = i_size_read(inode);
-	if (!(mode & FALLOC_FL_KEEP_SIZE)) {
-		endoff = offset + length;
-		ret = inode_newsize_ok(inode, endoff);
-		if (ret)
-			goto unlock;
-	}
+
+	/* Are we punching a hole beyond EOF? */
+	if (offset >= size)
+		goto unlock;
+	if ((offset + length) > size)
+		length = size - offset;
 
 	if (fi->fmode & CEPH_FILE_MODE_LAZY)
 		want = CEPH_CAP_FILE_BUFFER | CEPH_CAP_FILE_LAZYIO;
@@ -1798,16 +1782,8 @@ static long ceph_fallocate(struct file *file, int mode,
 	if (ret < 0)
 		goto unlock;
 
-	if (mode & FALLOC_FL_PUNCH_HOLE) {
-		if (offset < size)
-			ceph_zero_pagecache_range(inode, offset, length);
-		ret = ceph_zero_objects(inode, offset, length);
-	} else if (endoff > size) {
-		truncate_pagecache_range(inode, size, -1);
-		if (ceph_inode_set_size(inode, endoff))
-			ceph_check_caps(ceph_inode(inode),
-				CHECK_CAPS_AUTHONLY, NULL);
-	}
+	ceph_zero_pagecache_range(inode, offset, length);
+	ret = ceph_zero_objects(inode, offset, length);
 
 	if (!ret) {
 		spin_lock(&ci->i_ceph_lock);
@@ -1817,9 +1793,6 @@ static long ceph_fallocate(struct file *file, int mode,
 		spin_unlock(&ci->i_ceph_lock);
 		if (dirty)
 			__mark_inode_dirty(inode, dirty);
-		if ((endoff > size) &&
-		    ceph_quota_is_max_bytes_approaching(inode, endoff))
-			ceph_check_caps(ci, CHECK_CAPS_NODELAY, NULL);
 	}
 
 	ceph_put_cap_refs(ci, got);
-- 
2.20.1