login
Search
I'm guessing that the merge window opening might not be too far away now, and in any case, I won't have quite my normal internet access next week. So I'm pushing out the current GFS2 tree so that (I hope) there will be time to fix any issues. Again, there are fewer patches here. A lot of them are fairly small too. The most noteable item deals with the meta filesystem which was in response to Al Viro's suggestions concerning a better way to structure that code. It certainly results in a much cleaner implementation, so thanks go to Al for pointing that out. A couple of new features: I/O barrier support (needs no user configuration, see the patch for details) and UUID support (no code changes, its all userland but we reserve space in the super block, again details in the patch itself). I know that I hardly need say, but please let me know if you have any comments :-) Steve. --
Due to an incorrect iterator, some glocks were being missed from the
glock dumps obtained via debugfs. This patch fixes the problem and
ensures that we don't miss any glocks in future.
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
diff --git a/fs/gfs2/glock.c b/fs/gfs2/glock.c
index 13391e5..4cbb695 100644
--- a/fs/gfs2/glock.c
+++ b/fs/gfs2/glock.c
@@ -1816,15 +1816,17 @@ restart:
if (gl) {
gi->gl = hlist_entry(gl->gl_list.next,
struct gfs2_glock, gl_list);
- if (gi->gl)
- gfs2_glock_hold(gi->gl);
+ } else {
+ gi->gl = hlist_entry(gl_hash_table[gi->hash].hb_list.first,
+ struct gfs2_glock, gl_list);
}
+ if (gi->gl)
+ gfs2_glock_hold(gi->gl);
read_unlock(gl_lock_addr(gi->hash));
if (gl)
gfs2_glock_put(gl);
- if (gl && gi->gl == NULL)
- gi->hash++;
while (gi->gl == NULL) {
+ gi->hash++;
if (gi->hash >= GFS2_GL_HASH_SIZE)
return 1;
read_lock(gl_lock_addr(gi->hash));
@@ -1833,7 +1835,6 @@ restart:
if (gi->gl)
gfs2_glock_hold(gi->gl);
read_unlock(gl_lock_addr(gi->hash));
- gi->hash++;
}
if (gi->sdp != gi->gl->gl_sbd)
--
1.5.5.1
--
From: Bob Peterson <rpeterso@redhat.com>
This patch fixes a problem whereby simultaneous unlink, rmdir,
rename and link operations (e.g. rm -fR *) from multiple nodes
on the same GFS2 file system can cause kernel panics, hangs,
and/or memory corruption. It also gets rid of all the non-rgrp
calls to gfs2_glock_nq_m.
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
diff --git a/fs/gfs2/ops_inode.c b/fs/gfs2/ops_inode.c
index e2c62f7..35f6f03 100644
--- a/fs/gfs2/ops_inode.c
+++ b/fs/gfs2/ops_inode.c
@@ -159,9 +159,13 @@ static int gfs2_link(struct dentry *old_dentry, struct inode *dir,
gfs2_holder_init(dip->i_gl, LM_ST_EXCLUSIVE, 0, ghs);
gfs2_holder_init(ip->i_gl, LM_ST_EXCLUSIVE, 0, ghs + 1);
- error = gfs2_glock_nq_m(2, ghs);
+ error = gfs2_glock_nq(ghs); /* parent */
if (error)
- goto out;
+ goto out_parent;
+
+ error = gfs2_glock_nq(ghs + 1); /* child */
+ if (error)
+ goto out_child;
error = gfs2_permission(dir, MAY_WRITE | MAY_EXEC);
if (error)
@@ -245,8 +249,10 @@ out_alloc:
if (alloc_required)
gfs2_alloc_put(dip);
out_gunlock:
- gfs2_glock_dq_m(2, ghs);
-out:
+ gfs2_glock_dq(ghs + 1);
+out_child:
+ gfs2_glock_dq(ghs);
+out_parent:
gfs2_holder_uninit(ghs);
gfs2_holder_uninit(ghs + 1);
if (!error) {
@@ -302,7 +308,7 @@ static int gfs2_unlink(struct inode *dir, struct dentry *dentry)
error = gfs2_unlink_ok(dip, &dentry->d_name, ip);
if (error)
- goto out_rgrp;
+ goto out_gunlock;
error = gfs2_trans_begin(sdp, 2*RES_DINODE + RES_LEAF + RES_RG_BIT, 0);
if (error)
@@ -316,6 +322,7 @@ static int gfs2_unlink(struct inode *dir, struct dentry *dentry)
out_end_trans:
gfs2_trans_end(sdp);
+out_gunlock:
gfs2_glock_dq(ghs + 2);
out_rgrp:
gfs2_holder_uninit(ghs + 2);
@@ -485,7 +492,6 @@ static int gfs2_rmdir(struct inode *dir, struct dentry *dentry)
struct gfs2_holder ri_gh;
int error;
-
error = gfs2_rindex_hold(sdp, &ri_gh);
if ...This patch fixes a locking issue in the rename code by ensuring that we hold
the per sb rename lock over both directory and "other" renames which involve
different parent directories.
At the same time, this moved the (only called from one place) function
gfs2_ok_to_move into the file that its called from, so we can mark it
static. This should make a code a bit easier to follow.
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
Cc: Peter Staubach <staubach@redhat.com>
diff --git a/fs/gfs2/inode.c b/fs/gfs2/inode.c
index 8b0806a..8752552 100644
--- a/fs/gfs2/inode.c
+++ b/fs/gfs2/inode.c
@@ -1140,54 +1140,6 @@ int gfs2_unlink_ok(struct gfs2_inode *dip, const struct qstr *name,
return 0;
}
-/*
- * gfs2_ok_to_move - check if it's ok to move a directory to another directory
- * @this: move this
- * @to: to here
- *
- * Follow @to back to the root and make sure we don't encounter @this
- * Assumes we already hold the rename lock.
- *
- * Returns: errno
- */
-
-int gfs2_ok_to_move(struct gfs2_inode *this, struct gfs2_inode *to)
-{
- struct inode *dir = &to->i_inode;
- struct super_block *sb = dir->i_sb;
- struct inode *tmp;
- struct qstr dotdot;
- int error = 0;
-
- gfs2_str2qstr(&dotdot, "..");
-
- igrab(dir);
-
- for (;;) {
- if (dir == &this->i_inode) {
- error = -EINVAL;
- break;
- }
- if (dir == sb->s_root->d_inode) {
- error = 0;
- break;
- }
-
- tmp = gfs2_lookupi(dir, &dotdot, 1);
- if (IS_ERR(tmp)) {
- error = PTR_ERR(tmp);
- break;
- }
-
- iput(dir);
- dir = tmp;
- }
-
- iput(dir);
-
- return error;
-}
-
/**
* gfs2_readlinki - return the contents of a symlink
* @ip: the symlink's inode
diff --git a/fs/gfs2/inode.h b/fs/gfs2/inode.h
index 58f9607..bfd2afc 100644
--- a/fs/gfs2/inode.h
+++ b/fs/gfs2/inode.h
@@ -91,7 +91,6 @@ int gfs2_rmdiri(struct gfs2_inode *dip, const struct qstr *name,
int gfs2_unlink_ok(struct gfs2_inode *dip, const struct qstr *name,
const struct gfs2_inode *ip);
...From: Julien Brunel <brunel@diku.dk> In case of error, the function gfs2_inode_lookup returns an ERR pointer, but never returns a NULL pointer. So a NULL test that necessarily comes after an IS_ERR test should be deleted, and a NULL test that may come after a call to this function should be strengthened by an IS_ERR test. The semantic match that finds this problem is as follows: (http://www.emn.fr/x-info/coccinelle/) // <smpl> @match_bad_null_test@ expression x, E; statement S1,S2; @@ x = gfs2_inode_lookup(...) ... when != x = E * if (x != NULL) S1 else S2 // </smpl> Signed-off-by: Julien Brunel <brunel@diku.dk> Signed-off-by: Julia Lawall <julia@diku.dk> Signed-off-by: Steven Whitehouse <swhiteho@redhat.com> diff --git a/fs/gfs2/inode.c b/fs/gfs2/inode.c index 8752552..c8a959c 100644 --- a/fs/gfs2/inode.c +++ b/fs/gfs2/inode.c @@ -1033,13 +1033,11 @@ struct inode *gfs2_createi(struct gfs2_holder *ghs, const struct qstr *name, if (bh) brelse(bh); - if (!inode) - return ERR_PTR(-ENOMEM); return inode; fail_gunlock2: gfs2_glock_dq_uninit(ghs + 1); - if (inode) + if (inode && !IS_ERR(inode)) iput(inode); fail_gunlock: gfs2_glock_dq(ghs); -- 1.5.5.1 --
From: Bob Peterson <rpeterso@redhat.com> This patch fixes a problem whereby a direct_io write doesn't fall back to buffered write properly at end of file. Signed-off-by: Bob Peterson <rpeterso@redhat.com> Signed-off-by: Steven Whitehouse <swhiteho@redhat.com> diff --git a/fs/gfs2/ops_address.c b/fs/gfs2/ops_address.c index e64a1b0..ae7126a 100644 --- a/fs/gfs2/ops_address.c +++ b/fs/gfs2/ops_address.c @@ -975,7 +975,7 @@ static int gfs2_ok_for_dio(struct gfs2_inode *ip, int rw, loff_t offset) if (gfs2_is_stuffed(ip)) return 0; - if (offset > i_size_read(&ip->i_inode)) + if (offset >= i_size_read(&ip->i_inode)) return 0; return 1; } -- 1.5.5.1 --
From: Abhijith Das <adas@redhat.com>
The gfs2 superblock pointer is NULL after a failed mount. When control
eventually goes to gfs2_kill_sb, we dereference this NULL pointer. This
patch ensures that the gfs2 superblock pointer is not NULL before being
dereferenced in gfs2_kill_sb.
Signed-off-by: Abhijith Das <adas@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
diff --git a/fs/gfs2/ops_fstype.c b/fs/gfs2/ops_fstype.c
index a6225cc..ae35f09 100644
--- a/fs/gfs2/ops_fstype.c
+++ b/fs/gfs2/ops_fstype.c
@@ -1237,14 +1237,17 @@ static int gfs2_get_sb_meta(struct file_system_type *fs_type, int flags,
static void gfs2_kill_sb(struct super_block *sb)
{
struct gfs2_sbd *sdp = sb->s_fs_info;
- gfs2_meta_syncfs(sdp);
- dput(sdp->sd_root_dir);
- dput(sdp->sd_master_dir);
- sdp->sd_root_dir = NULL;
- sdp->sd_master_dir = NULL;
+ if (sdp) {
+ gfs2_meta_syncfs(sdp);
+ dput(sdp->sd_root_dir);
+ dput(sdp->sd_master_dir);
+ sdp->sd_root_dir = NULL;
+ sdp->sd_master_dir = NULL;
+ }
shrink_dcache_sb(sb);
kill_block_super(sb);
- gfs2_delete_debugfs_file(sdp);
+ if (sdp)
+ gfs2_delete_debugfs_file(sdp);
}
struct file_system_type gfs2_fs_type = {
--
1.5.5.1
--
The following patch shrinks the gfs2_args structure which is embedded in
every GFS2 superblock. It cuts down the size of the options to a single
unsigned int (the 13 bits of bitfields will be rounded up to that size
by the compiler) from the current 11 unsigned ints. So on x86 thats 44
bytes shrinking to 4 bytes, in each and every GFS2 superblock.
Signed-off-by: Steven Whitehouse <swhitho@redhat.com>
diff --git a/fs/gfs2/incore.h b/fs/gfs2/incore.h
index a1777a1..b2c5784 100644
--- a/fs/gfs2/incore.h
+++ b/fs/gfs2/incore.h
@@ -386,21 +386,21 @@ struct gfs2_statfs_change_host {
#define GFS2_DATA_ORDERED 2
struct gfs2_args {
- char ar_lockproto[GFS2_LOCKNAME_LEN]; /* Name of the Lock Protocol */
- char ar_locktable[GFS2_LOCKNAME_LEN]; /* Name of the Lock Table */
- char ar_hostdata[GFS2_LOCKNAME_LEN]; /* Host specific data */
- int ar_spectator; /* Don't get a journal because we're always RO */
- int ar_ignore_local_fs; /* Don't optimize even if local_fs is 1 */
- int ar_localflocks; /* Let the VFS do flock|fcntl locks for us */
- int ar_localcaching; /* Local-style caching (dangerous on multihost) */
- int ar_debug; /* Oops on errors instead of trying to be graceful */
- int ar_upgrade; /* Upgrade ondisk/multihost format */
- unsigned int ar_num_glockd; /* Number of glockd threads */
- int ar_posix_acl; /* Enable posix acls */
- int ar_quota; /* off/account/on */
- int ar_suiddir; /* suiddir support */
- int ar_data; /* ordered/writeback */
- int ar_meta; /* mount metafs */
+ char ar_lockproto[GFS2_LOCKNAME_LEN]; /* Name of the Lock Protocol */
+ char ar_locktable[GFS2_LOCKNAME_LEN]; /* Name of the Lock Table */
+ char ar_hostdata[GFS2_LOCKNAME_LEN]; /* Host specific data */
+ unsigned int ar_spectator:1; /* Don't get a journal */
+ unsigned int ar_ignore_local_fs:1; /* Ignore optimisations */
+ unsigned int ar_localflocks:1; /* Let the VFS do flock|fcntl */
+ unsigned int ar_localcaching:1; /* Local caching */
+ unsigned int ar_debug:1; /* Oops on errors ...Until now, we've used the same scheme as GFS1 for atime. This has failed since atime is a per vfsmnt flag, not a per fs flag and as such the "noatime" flag was not getting passed down to the filesystems. This patch removes all the "special casing" around atime updates and we simply use the VFS's atime code. The net result is that GFS2 will now support all the same atime related mount options of any other filesystem on a per-vfsmnt basis. We do lose the "lazy atime" updates, but we gain "relatime". We could add lazy atime to the VFS at a later date, if there is a requirement for that variant still - I suspect relatime will be enough. Also we lose about 100 lines of code after this patch has been applied, and I have a suspicion that it will speed things up a bit, even when atime is "on". So it seems like a nice clean up as well. From a user perspective, everything stays the same except the loss of the per-fs atime quantum tweekable (ought to be per-vfsmnt at the very least, and to be honest I don't think anybody ever used it) and that a number of options which were ignored before now work correctly. Please let me know if you've got any comments. I'm pushing this out early so that you can all see what my plans are. Signed-off-by: Steven Whitehouse <swhiteho@redhat.com> diff --git a/fs/gfs2/glock.c b/fs/gfs2/glock.c index 806e1eb..c962283 100644 --- a/fs/gfs2/glock.c +++ b/fs/gfs2/glock.c @@ -1580,8 +1580,6 @@ static const char *hflags2str(char *buf, unsigned flags, unsigned long iflags) *p++ = 'a'; if (flags & GL_EXACT) *p++ = 'E'; - if (flags & GL_ATIME) - *p++ = 'a'; if (flags & GL_NOCACHE) *p++ = 'c'; if (test_bit(HIF_HOLDER, &iflags)) diff --git a/fs/gfs2/glock.h b/fs/gfs2/glock.h index 971d92a..695c6b1 100644 --- a/fs/gfs2/glock.h +++ b/fs/gfs2/glock.h @@ -24,7 +24,6 @@ #define GL_ASYNC 0x00000040 #define GL_EXACT 0x00000080 #define GL_SKIP 0x00000100 -#define GL_ATIME 0x00000200 #define GL_NOCACHE 0x00000400 #define ...
This patch adds a UUID to the GFS2 sb structure. This field is not
actually referenced from kernel space at all, but is added for
completeness and due to the userland tools which get their on-disk
structure information from the gfs2_ondisk.h header file.
Since we have to be backwards compatible, we will assume that any GFS2
sb for which the UUID is all 0 does not have a UUID as such.
We should then be (after some userland changes) able to support the -U
mount option. This addresses Fedora bugzilla #242689
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
diff --git a/include/linux/gfs2_ondisk.h b/include/linux/gfs2_ondisk.h
index c3c19f9..14d0df0 100644
--- a/include/linux/gfs2_ondisk.h
+++ b/include/linux/gfs2_ondisk.h
@@ -118,7 +118,11 @@ struct gfs2_sb {
char sb_lockproto[GFS2_LOCKNAME_LEN];
char sb_locktable[GFS2_LOCKNAME_LEN];
- /* In gfs1, quota and license dinodes followed */
+
+ struct gfs2_inum __pad3; /* Was quota inode in gfs1 */
+ struct gfs2_inum __pad4; /* Was licence inode in gfs1 */
+#define GFS2_HAS_UUID 1
+ __u8 sb_uuid[16]; /* The UUID, maybe 0 for backwards compat */
};
/*
--
1.5.5.1
--
This patch adds barrier support to GFS2. There is not a lot of change
really... we just add the barrier flag when we write journal header
blocks. If the underlying device refuses to support them, we fall back
to the previous way of doing things (wait for the I/O and hope) since
there is nothing else we can do. There is no user configuration,
barriers will always be on unless the device refuses to support them.
This seems a reasonable solution to me since this is a correctness
issue.
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
diff --git a/fs/gfs2/incore.h b/fs/gfs2/incore.h
index f1ed3a1..f566ec1 100644
--- a/fs/gfs2/incore.h
+++ b/fs/gfs2/incore.h
@@ -432,6 +432,7 @@ enum {
SDF_JOURNAL_CHECKED = 0,
SDF_JOURNAL_LIVE = 1,
SDF_SHUTDOWN = 2,
+ SDF_NOBARRIERS = 3,
};
#define GFS2_FSNAME_LEN 256
diff --git a/fs/gfs2/log.c b/fs/gfs2/log.c
index 6c6af9f..ad30585 100644
--- a/fs/gfs2/log.c
+++ b/fs/gfs2/log.c
@@ -18,6 +18,7 @@
#include <linux/delay.h>
#include <linux/kthread.h>
#include <linux/freezer.h>
+#include <linux/bio.h>
#include "gfs2.h"
#include "incore.h"
@@ -584,7 +585,6 @@ static void log_write_header(struct gfs2_sbd *sdp, u32 flags, int pull)
memset(bh->b_data, 0, bh->b_size);
set_buffer_uptodate(bh);
clear_buffer_dirty(bh);
- unlock_buffer(bh);
gfs2_ail1_empty(sdp, 0);
tail = current_tail(sdp);
@@ -601,8 +601,23 @@ static void log_write_header(struct gfs2_sbd *sdp, u32 flags, int pull)
hash = gfs2_disk_hash(bh->b_data, sizeof(struct gfs2_log_header));
lh->lh_hash = cpu_to_be32(hash);
- set_buffer_dirty(bh);
- if (sync_dirty_buffer(bh))
+ bh->b_end_io = end_buffer_write_sync;
+ if (test_bit(SDF_NOBARRIERS, &sdp->sd_flags))
+ goto skip_barrier;
+ get_bh(bh);
+ submit_bh(WRITE_BARRIER | (1 << BIO_RW_META), bh);
+ wait_on_buffer(bh);
+ if (buffer_eopnotsupp(bh)) {
+ clear_buffer_eopnotsupp(bh);
+ set_buffer_uptodate(bh);
+ set_bit(SDF_NOBARRIERS, ...
