From 5d0af85cd0964bb845b63d5059bb20e8f7731e65 Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Thu, 28 Oct 2010 21:37:10 +0000 Subject: xfs: remove experimental tag from the delaylog option We promised to do this for 2.6.37, and the code looks stable enough to keep that promise. Signed-off-by: Christoph Hellwig Reviewed-by: Dave Chinner Signed-off-by: Alex Elder --- Documentation/filesystems/xfs-delayed-logging-design.txt | 11 ----------- fs/xfs/linux-2.6/xfs_super.c | 3 --- 2 files changed, 14 deletions(-) diff --git a/Documentation/filesystems/xfs-delayed-logging-design.txt b/Documentation/filesystems/xfs-delayed-logging-design.txt index 96d0df28bed3..7445bf335dae 100644 --- a/Documentation/filesystems/xfs-delayed-logging-design.txt +++ b/Documentation/filesystems/xfs-delayed-logging-design.txt @@ -794,17 +794,6 @@ designed. Roadmap: -2.6.37 Remove experimental tag from mount option - => should be roughly 6 months after initial merge - => enough time to: - => gain confidence and fix problems reported by early - adopters (a.k.a. guinea pigs) - => address worst performance regressions and undesired - behaviours - => start tuning/optimising code for parallelism - => start tuning/optimising algorithms consuming - excessive CPU time - 2.6.39 Switch default mount option to use delayed logging => should be roughly 12 months after initial merge => enough time to shake out remaining problems before next round of diff --git a/fs/xfs/linux-2.6/xfs_super.c b/fs/xfs/linux-2.6/xfs_super.c index 9f3a78fe6ae4..064f964d4f3c 100644 --- a/fs/xfs/linux-2.6/xfs_super.c +++ b/fs/xfs/linux-2.6/xfs_super.c @@ -353,9 +353,6 @@ xfs_parseargs( mp->m_qflags &= ~XFS_OQUOTA_ENFD; } else if (!strcmp(this_char, MNTOPT_DELAYLOG)) { mp->m_flags |= XFS_MOUNT_DELAYLOG; - cmn_err(CE_WARN, - "Enabling EXPERIMENTAL delayed logging feature " - "- use at your own risk.\n"); } else if (!strcmp(this_char, MNTOPT_NODELAYLOG)) { mp->m_flags &= ~XFS_MOUNT_DELAYLOG; } else if (!strcmp(this_char, "ihashsize")) { -- cgit v1.2.3 From 6762b938eac878a30a90e770ac655874c36bc642 Mon Sep 17 00:00:00 2001 From: Kulikov Vasiliy Date: Sat, 30 Oct 2010 14:26:17 +0000 Subject: xfs: xfs_ioctl: fix information leak to userland al_hreq is copied from userland. If al_hreq.buflen is not properly aligned then xfs_attr_list will ignore the last bytes of kbuf. These bytes are unitialized. It leads to leaking of contents of kernel stack memory. Signed-off-by: Vasiliy Kulikov Signed-off-by: Alex Elder --- fs/xfs/linux-2.6/xfs_ioctl.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/fs/xfs/linux-2.6/xfs_ioctl.c b/fs/xfs/linux-2.6/xfs_ioctl.c index 2ea238f6d38e..ad442d9e392e 100644 --- a/fs/xfs/linux-2.6/xfs_ioctl.c +++ b/fs/xfs/linux-2.6/xfs_ioctl.c @@ -416,7 +416,7 @@ xfs_attrlist_by_handle( if (IS_ERR(dentry)) return PTR_ERR(dentry); - kbuf = kmalloc(al_hreq.buflen, GFP_KERNEL); + kbuf = kzalloc(al_hreq.buflen, GFP_KERNEL); if (!kbuf) goto out_dput; -- cgit v1.2.3 From f83282a8ef799c0bdcb0c32971487087da1bc216 Mon Sep 17 00:00:00 2001 From: Dave Chinner Date: Mon, 8 Nov 2010 08:55:04 +0000 Subject: xfs: fix per-ag reference counting in inode reclaim tree walking The walk fails to decrement the per-ag reference count when the non-blocking walk fails to obtain the per-ag reclaim lock, leading to an assert failure on debug kernels when unmounting a filesystem. Signed-off-by: Dave Chinner Reviewed-by: Christoph Hellwig Signed-off-by: Alex Elder --- fs/xfs/linux-2.6/xfs_sync.c | 1 + fs/xfs/xfs_mount.c | 1 + 2 files changed, 2 insertions(+) diff --git a/fs/xfs/linux-2.6/xfs_sync.c b/fs/xfs/linux-2.6/xfs_sync.c index 37d33254981d..afb0d7cfad1c 100644 --- a/fs/xfs/linux-2.6/xfs_sync.c +++ b/fs/xfs/linux-2.6/xfs_sync.c @@ -853,6 +853,7 @@ restart: if (trylock) { if (!mutex_trylock(&pag->pag_ici_reclaim_lock)) { skipped++; + xfs_perag_put(pag); continue; } first_index = pag->pag_ici_reclaim_cursor; diff --git a/fs/xfs/xfs_mount.c b/fs/xfs/xfs_mount.c index b1498ab5a399..19e9dfa1c254 100644 --- a/fs/xfs/xfs_mount.c +++ b/fs/xfs/xfs_mount.c @@ -275,6 +275,7 @@ xfs_free_perag( pag = radix_tree_delete(&mp->m_perag_tree, agno); spin_unlock(&mp->m_perag_lock); ASSERT(pag); + ASSERT(atomic_read(&pag->pag_ref) == 0); call_rcu(&pag->rcu_head, __xfs_free_perag); } } -- cgit v1.2.3 From bfe2741967eaa3434fa9b3d8f24b1422d4540e7d Mon Sep 17 00:00:00 2001 From: Dave Chinner Date: Mon, 8 Nov 2010 08:55:05 +0000 Subject: xfs: move delayed write buffer trace The delayed write buffer split trace currently issues a trace for every buffer it scans. These buffers are not necessarily queued for delayed write. Indeed, when buffers are pinned, there can be thousands of traces of buffers that aren't actually queued for delayed write and the ones that are are lost in the noise. Move the trace point to record only buffers that are split out for IO to be issued on. Signed-off-by: Dave Chinner Reviewed-by: Christoph Hellwig Signed-off-by: Alex Elder --- fs/xfs/linux-2.6/xfs_buf.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/fs/xfs/linux-2.6/xfs_buf.c b/fs/xfs/linux-2.6/xfs_buf.c index 63fd2c07cb57..aa1d353def29 100644 --- a/fs/xfs/linux-2.6/xfs_buf.c +++ b/fs/xfs/linux-2.6/xfs_buf.c @@ -1781,7 +1781,6 @@ xfs_buf_delwri_split( INIT_LIST_HEAD(list); spin_lock(dwlk); list_for_each_entry_safe(bp, n, dwq, b_list) { - trace_xfs_buf_delwri_split(bp, _RET_IP_); ASSERT(bp->b_flags & XBF_DELWRI); if (!XFS_BUF_ISPINNED(bp) && !xfs_buf_cond_lock(bp)) { @@ -1795,6 +1794,7 @@ xfs_buf_delwri_split( _XBF_RUN_QUEUES); bp->b_flags |= XBF_WRITE; list_move_tail(&bp->b_list, list); + trace_xfs_buf_delwri_split(bp, _RET_IP_); } else skipped++; } -- cgit v1.2.3 From 785ce41805ea7b6a9b2775ed9f4cf10cd7a90c03 Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Sat, 6 Nov 2010 11:42:44 +0000 Subject: xfs: tell lockdep about parent iolock usage in filestreams The filestreams code may take the iolock on the parent inode while holding it on a child. This is the only place in XFS where we take both the child and parent iolock, so just telling lockdep about it is enough. The lock flag required for that was already added as part of the ilock lockdep annotations and unused so far. Signed-off-by: Christoph Hellwig Signed-off-by: Alex Elder --- fs/xfs/xfs_filestream.c | 8 +++++++- 1 file changed, 7 insertions(+), 1 deletion(-) diff --git a/fs/xfs/xfs_filestream.c b/fs/xfs/xfs_filestream.c index 9b715dce5699..9124425b7f2f 100644 --- a/fs/xfs/xfs_filestream.c +++ b/fs/xfs/xfs_filestream.c @@ -744,9 +744,15 @@ xfs_filestream_new_ag( * If the file's parent directory is known, take its iolock in exclusive * mode to prevent two sibling files from racing each other to migrate * themselves and their parent to different AGs. + * + * Note that we lock the parent directory iolock inside the child + * iolock here. That's fine as we never hold both parent and child + * iolock in any other place. This is different from the ilock, + * which requires locking of the child after the parent for namespace + * operations. */ if (pip) - xfs_ilock(pip, XFS_IOLOCK_EXCL); + xfs_ilock(pip, XFS_IOLOCK_EXCL | XFS_IOLOCK_PARENT); /* * A new AG needs to be found for the file. If the file's parent -- cgit v1.2.3 From 5d2bf8a55e03b0e59ed5a4ac2ff7f9ee3ba7e40d Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Sat, 6 Nov 2010 11:42:56 +0000 Subject: xfs: fix a few compiler warnings with CONFIG_XFS_QUOTA=n Andi Kleen reported that gcc-4.5 gives lots of warnings for him inside the XFS code. It turned out most of them are due to the quota stubs beeing macros, and gcc now complaining about macros evaluating to 0 that are not assigned to variables. Signed-off-by: Christoph Hellwig Signed-off-by: Alex Elder --- fs/xfs/xfs_quota.h | 20 ++++++++++++++++---- 1 file changed, 16 insertions(+), 4 deletions(-) diff --git a/fs/xfs/xfs_quota.h b/fs/xfs/xfs_quota.h index e0e64b113bd6..9bb6eda4cd21 100644 --- a/fs/xfs/xfs_quota.h +++ b/fs/xfs/xfs_quota.h @@ -346,8 +346,17 @@ xfs_qm_vop_dqalloc(struct xfs_inode *ip, uid_t uid, gid_t gid, prid_t prid, #define xfs_trans_mod_dquot_byino(tp, ip, fields, delta) #define xfs_trans_apply_dquot_deltas(tp) #define xfs_trans_unreserve_and_mod_dquots(tp) -#define xfs_trans_reserve_quota_nblks(tp, ip, nblks, ninos, flags) (0) -#define xfs_trans_reserve_quota_bydquots(tp, mp, u, g, nb, ni, fl) (0) +static inline int xfs_trans_reserve_quota_nblks(struct xfs_trans *tp, + struct xfs_inode *ip, long nblks, long ninos, uint flags) +{ + return 0; +} +static inline int xfs_trans_reserve_quota_bydquots(struct xfs_trans *tp, + struct xfs_mount *mp, struct xfs_dquot *udqp, + struct xfs_dquot *gdqp, long nblks, long nions, uint flags) +{ + return 0; +} #define xfs_qm_vop_create_dqattach(tp, ip, u, g) #define xfs_qm_vop_rename_dqattach(it) (0) #define xfs_qm_vop_chown(tp, ip, old, new) (NULL) @@ -357,11 +366,14 @@ xfs_qm_vop_dqalloc(struct xfs_inode *ip, uid_t uid, gid_t gid, prid_t prid, #define xfs_qm_dqdetach(ip) #define xfs_qm_dqrele(d) #define xfs_qm_statvfs(ip, s) -#define xfs_qm_sync(mp, fl) (0) +static inline int xfs_qm_sync(struct xfs_mount *mp, int flags) +{ + return 0; +} #define xfs_qm_newmount(mp, a, b) (0) #define xfs_qm_mount_quotas(mp) #define xfs_qm_unmount(mp) -#define xfs_qm_unmount_quotas(mp) (0) +#define xfs_qm_unmount_quotas(mp) #endif /* CONFIG_XFS_QUOTA */ #define xfs_trans_unreserve_quota_nblks(tp, ip, nblks, ninos, flags) \ -- cgit v1.2.3 From c6f6cd0608b1826ee1797cf57a808416e4bdb806 Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Sat, 6 Nov 2010 11:43:08 +0000 Subject: xfs: use hlist_add_fake XFS does not need it's inodes to actuall be hashed in the VFS inode cache, but we require the inode to be marked hashed for the writeback code to work. Insted of using insert_inode_hash, which requires a second inode_lock roundtrip after the partial merge of the inode scalability patches in 2.6.37-rc simply use the new hlist_add_fake helper to mark it hashed without requiring a lock or touching a global cache line. Signed-off-by: Christoph Hellwig Signed-off-by: Alex Elder --- fs/xfs/linux-2.6/xfs_iops.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/fs/xfs/linux-2.6/xfs_iops.c b/fs/xfs/linux-2.6/xfs_iops.c index 96107efc0c61..94d5fd6a2973 100644 --- a/fs/xfs/linux-2.6/xfs_iops.c +++ b/fs/xfs/linux-2.6/xfs_iops.c @@ -762,7 +762,8 @@ xfs_setup_inode( inode->i_state = I_NEW; inode_sb_list_add(inode); - insert_inode_hash(inode); + /* make the inode look hashed for the writeback code */ + hlist_add_fake(&inode->i_hash); inode->i_mode = ip->i_d.di_mode; inode->i_nlink = ip->i_d.di_nlink; -- cgit v1.2.3 From ece413f59f257682de4a2e2e42af33b016af53f3 Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Wed, 10 Nov 2010 21:39:11 +0000 Subject: xfs: remove incorrect assert in xfs_vm_writepage In commit 20cb52ebd1b5ca6fa8a5d9b6b1392292f5ca8a45, titled "xfs: simplify xfs_vm_writepage" I added an assert that any !mapped and uptodate buffers are not dirty. That asserts turns out to trigger a lot when running fsx on filesystems with small block sizes. The reason for that is that the assert is simply incorrect. !mapped and uptodate just mean this buffer covers a hole, and whenever we do a set_page_dirty we mark all blocks in the page dirty, no matter if they have data or not. So remove the assert, and update the comment above the condition to match reality. Signed-off-by: Christoph Hellwig Signed-off-by: Alex Elder --- fs/xfs/linux-2.6/xfs_aops.c | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-) diff --git a/fs/xfs/linux-2.6/xfs_aops.c b/fs/xfs/linux-2.6/xfs_aops.c index c9af48fffcd7..7d287afccde5 100644 --- a/fs/xfs/linux-2.6/xfs_aops.c +++ b/fs/xfs/linux-2.6/xfs_aops.c @@ -1111,11 +1111,12 @@ xfs_vm_writepage( uptodate = 0; /* - * A hole may still be marked uptodate because discard_buffer - * leaves the flag set. + * set_page_dirty dirties all buffers in a page, independent + * of their state. The dirty state however is entirely + * meaningless for holes (!mapped && uptodate), so skip + * buffers covering holes here. */ if (!buffer_mapped(bh) && buffer_uptodate(bh)) { - ASSERT(!buffer_dirty(bh)); imap_valid = 0; continue; } -- cgit v1.2.3