From 657a4cffde065beca7d919e867f61e7d322708b6 Mon Sep 17 00:00:00 2001 From: Eric Sandeen Date: Fri, 30 Apr 2010 03:42:49 +0000 Subject: xfs: replace E2BIG with EFBIG where appropriate Many places in the xfs code return E2BIG when they really mean EFBIG; trying to grow past 16T on a 32 bit machine, for example, says "Argument list too long" rather than "File too large" which is not particularly helpful. Some of these don't make perfect sense as EFBIG either, but still better than E2BIG IMHO. Signed-off-by: Eric Sandeen Reviewed-by: Christoph Hellwig Signed-off-by: Alex Elder --- fs/xfs/xfs_mount.c | 14 +++++++------- fs/xfs/xfs_rtalloc.c | 4 ++-- 2 files changed, 9 insertions(+), 9 deletions(-) (limited to 'fs') diff --git a/fs/xfs/xfs_mount.c b/fs/xfs/xfs_mount.c index d7bf38c8cd1c..2d811e57b332 100644 --- a/fs/xfs/xfs_mount.c +++ b/fs/xfs/xfs_mount.c @@ -268,10 +268,10 @@ xfs_sb_validate_fsb_count( #if XFS_BIG_BLKNOS /* Limited by ULONG_MAX of page cache index */ if (nblocks >> (PAGE_CACHE_SHIFT - sbp->sb_blocklog) > ULONG_MAX) - return E2BIG; + return EFBIG; #else /* Limited by UINT_MAX of sectors */ if (nblocks << (sbp->sb_blocklog - BBSHIFT) > UINT_MAX) - return E2BIG; + return EFBIG; #endif return 0; } @@ -393,7 +393,7 @@ xfs_mount_validate_sb( xfs_sb_validate_fsb_count(sbp, sbp->sb_rblocks)) { xfs_fs_mount_cmn_err(flags, "file system too large to be mounted on this system."); - return XFS_ERROR(E2BIG); + return XFS_ERROR(EFBIG); } if (unlikely(sbp->sb_inprogress)) { @@ -1009,7 +1009,7 @@ xfs_check_sizes(xfs_mount_t *mp) d = (xfs_daddr_t)XFS_FSB_TO_BB(mp, mp->m_sb.sb_dblocks); if (XFS_BB_TO_FSB(mp, d) != mp->m_sb.sb_dblocks) { cmn_err(CE_WARN, "XFS: size check 1 failed"); - return XFS_ERROR(E2BIG); + return XFS_ERROR(EFBIG); } error = xfs_read_buf(mp, mp->m_ddev_targp, d - XFS_FSS_TO_BB(mp, 1), @@ -1019,7 +1019,7 @@ xfs_check_sizes(xfs_mount_t *mp) } else { cmn_err(CE_WARN, "XFS: size check 2 failed"); if (error == ENOSPC) - error = XFS_ERROR(E2BIG); + error = XFS_ERROR(EFBIG); return error; } @@ -1027,7 +1027,7 @@ xfs_check_sizes(xfs_mount_t *mp) d = (xfs_daddr_t)XFS_FSB_TO_BB(mp, mp->m_sb.sb_logblocks); if (XFS_BB_TO_FSB(mp, d) != mp->m_sb.sb_logblocks) { cmn_err(CE_WARN, "XFS: size check 3 failed"); - return XFS_ERROR(E2BIG); + return XFS_ERROR(EFBIG); } error = xfs_read_buf(mp, mp->m_logdev_targp, d - XFS_FSB_TO_BB(mp, 1), @@ -1037,7 +1037,7 @@ xfs_check_sizes(xfs_mount_t *mp) } else { cmn_err(CE_WARN, "XFS: size check 3 failed"); if (error == ENOSPC) - error = XFS_ERROR(E2BIG); + error = XFS_ERROR(EFBIG); return error; } } diff --git a/fs/xfs/xfs_rtalloc.c b/fs/xfs/xfs_rtalloc.c index 6be05f756d59..16445518506d 100644 --- a/fs/xfs/xfs_rtalloc.c +++ b/fs/xfs/xfs_rtalloc.c @@ -2247,7 +2247,7 @@ xfs_rtmount_init( cmn_err(CE_WARN, "XFS: realtime mount -- %llu != %llu", (unsigned long long) XFS_BB_TO_FSB(mp, d), (unsigned long long) mp->m_sb.sb_rblocks); - return XFS_ERROR(E2BIG); + return XFS_ERROR(EFBIG); } error = xfs_read_buf(mp, mp->m_rtdev_targp, d - XFS_FSB_TO_BB(mp, 1), @@ -2256,7 +2256,7 @@ xfs_rtmount_init( cmn_err(CE_WARN, "XFS: realtime mount -- xfs_read_buf failed, returned %d", error); if (error == ENOSPC) - return XFS_ERROR(E2BIG); + return XFS_ERROR(EFBIG); return error; } xfs_buf_relse(bp); -- cgit v1.2.2 From 32891b292d6262d1db8e553cf3f4b38a91247b5a Mon Sep 17 00:00:00 2001 From: Eric Sandeen Date: Fri, 30 Apr 2010 16:43:48 +0000 Subject: xfs: be more explicit if RT mount fails due to config Recent testers were slightly confused that a realtime mount failed due to missing CONFIG_XFS_RT; we can make that a little more obvious. V2: drop the else as suggested by Christoph Signed-off-by: Eric Sandeen Reviewed-by: Christoph Hellwig Signed-off-by: Alex Elder --- fs/xfs/xfs_rtalloc.h | 11 ++++++++++- 1 file changed, 10 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/xfs/xfs_rtalloc.h b/fs/xfs/xfs_rtalloc.h index b2d67adb6a08..ff614c29b441 100644 --- a/fs/xfs/xfs_rtalloc.h +++ b/fs/xfs/xfs_rtalloc.h @@ -147,7 +147,16 @@ xfs_growfs_rt( # define xfs_rtfree_extent(t,b,l) (ENOSYS) # define xfs_rtpick_extent(m,t,l,rb) (ENOSYS) # define xfs_growfs_rt(mp,in) (ENOSYS) -# define xfs_rtmount_init(m) (((mp)->m_sb.sb_rblocks == 0)? 0 : (ENOSYS)) +static inline int /* error */ +xfs_rtmount_init( + xfs_mount_t *mp) /* file system mount structure */ +{ + if (mp->m_sb.sb_rblocks == 0) + return 0; + + cmn_err(CE_WARN, "XFS: Not built with CONFIG_XFS_RT"); + return ENOSYS; +} # define xfs_rtmount_inodes(m) (((mp)->m_sb.sb_rblocks == 0)? 0 : (ENOSYS)) # define xfs_rtunmount_inodes(m) #endif /* CONFIG_XFS_RT */ -- cgit v1.2.2 From 025101dca4480eff9da948405e872d5115030850 Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Tue, 4 May 2010 13:53:48 +0000 Subject: xfs: cleanup log reservation calculactions Instead of having small helper functions calling big macros do the calculations for the log reservations directly in the functions. These are mostly 1:1 from the macros execept that the macros kept the quota calculations in their callers. Signed-off-by: Christoph Hellwig Signed-off-by: Alex Elder --- fs/xfs/xfs_trans.c | 446 +++++++++++++++++++++++++++++++++++++++++++++++------ fs/xfs/xfs_trans.h | 411 ------------------------------------------------ 2 files changed, 400 insertions(+), 457 deletions(-) (limited to 'fs') diff --git a/fs/xfs/xfs_trans.c b/fs/xfs/xfs_trans.c index ce558efa2ea0..28547dfce037 100644 --- a/fs/xfs/xfs_trans.c +++ b/fs/xfs/xfs_trans.c @@ -48,134 +48,489 @@ kmem_zone_t *xfs_trans_zone; + /* - * Reservation functions here avoid a huge stack in xfs_trans_init - * due to register overflow from temporaries in the calculations. + * Various log reservation values. + * + * These are based on the size of the file system block because that is what + * most transactions manipulate. Each adds in an additional 128 bytes per + * item logged to try to account for the overhead of the transaction mechanism. + * + * Note: Most of the reservations underestimate the number of allocation + * groups into which they could free extents in the xfs_bmap_finish() call. + * This is because the number in the worst case is quite high and quite + * unusual. In order to fix this we need to change xfs_bmap_finish() to free + * extents in only a single AG at a time. This will require changes to the + * EFI code as well, however, so that the EFI for the extents not freed is + * logged again in each transaction. See SGI PV #261917. + * + * Reservation functions here avoid a huge stack in xfs_trans_init due to + * register overflow from temporaries in the calculations. + */ + + +/* + * In a write transaction we can allocate a maximum of 2 + * extents. This gives: + * the inode getting the new extents: inode size + * the inode's bmap btree: max depth * block size + * the agfs of the ags from which the extents are allocated: 2 * sector + * the superblock free block counter: sector size + * the allocation btrees: 2 exts * 2 trees * (2 * max depth - 1) * block size + * And the bmap_finish transaction can free bmap blocks in a join: + * the agfs of the ags containing the blocks: 2 * sector size + * the agfls of the ags containing the blocks: 2 * sector size + * the super block free block counter: sector size + * the allocation btrees: 2 exts * 2 trees * (2 * max depth - 1) * block size */ STATIC uint -xfs_calc_write_reservation(xfs_mount_t *mp) +xfs_calc_write_reservation( + struct xfs_mount *mp) { - return XFS_CALC_WRITE_LOG_RES(mp) + XFS_DQUOT_LOGRES(mp); + return XFS_DQUOT_LOGRES(mp) + + MAX((mp->m_sb.sb_inodesize + + XFS_FSB_TO_B(mp, XFS_BM_MAXLEVELS(mp, XFS_DATA_FORK)) + + 2 * mp->m_sb.sb_sectsize + + mp->m_sb.sb_sectsize + + XFS_ALLOCFREE_LOG_RES(mp, 2) + + 128 * (4 + XFS_BM_MAXLEVELS(mp, XFS_DATA_FORK) + + XFS_ALLOCFREE_LOG_COUNT(mp, 2))), + (2 * mp->m_sb.sb_sectsize + + 2 * mp->m_sb.sb_sectsize + + mp->m_sb.sb_sectsize + + XFS_ALLOCFREE_LOG_RES(mp, 2) + + 128 * (5 + XFS_ALLOCFREE_LOG_COUNT(mp, 2)))); } +/* + * In truncating a file we free up to two extents at once. We can modify: + * the inode being truncated: inode size + * the inode's bmap btree: (max depth + 1) * block size + * And the bmap_finish transaction can free the blocks and bmap blocks: + * the agf for each of the ags: 4 * sector size + * the agfl for each of the ags: 4 * sector size + * the super block to reflect the freed blocks: sector size + * worst case split in allocation btrees per extent assuming 4 extents: + * 4 exts * 2 trees * (2 * max depth - 1) * block size + * the inode btree: max depth * blocksize + * the allocation btrees: 2 trees * (max depth - 1) * block size + */ STATIC uint -xfs_calc_itruncate_reservation(xfs_mount_t *mp) +xfs_calc_itruncate_reservation( + struct xfs_mount *mp) { - return XFS_CALC_ITRUNCATE_LOG_RES(mp) + XFS_DQUOT_LOGRES(mp); + return XFS_DQUOT_LOGRES(mp) + + MAX((mp->m_sb.sb_inodesize + + XFS_FSB_TO_B(mp, XFS_BM_MAXLEVELS(mp, XFS_DATA_FORK) + 1) + + 128 * (2 + XFS_BM_MAXLEVELS(mp, XFS_DATA_FORK))), + (4 * mp->m_sb.sb_sectsize + + 4 * mp->m_sb.sb_sectsize + + mp->m_sb.sb_sectsize + + XFS_ALLOCFREE_LOG_RES(mp, 4) + + 128 * (9 + XFS_ALLOCFREE_LOG_COUNT(mp, 4)) + + 128 * 5 + + XFS_ALLOCFREE_LOG_RES(mp, 1) + + 128 * (2 + XFS_IALLOC_BLOCKS(mp) + mp->m_in_maxlevels + + XFS_ALLOCFREE_LOG_COUNT(mp, 1)))); } +/* + * In renaming a files we can modify: + * the four inodes involved: 4 * inode size + * the two directory btrees: 2 * (max depth + v2) * dir block size + * the two directory bmap btrees: 2 * max depth * block size + * And the bmap_finish transaction can free dir and bmap blocks (two sets + * of bmap blocks) giving: + * the agf for the ags in which the blocks live: 3 * sector size + * the agfl for the ags in which the blocks live: 3 * sector size + * the superblock for the free block count: sector size + * the allocation btrees: 3 exts * 2 trees * (2 * max depth - 1) * block size + */ STATIC uint -xfs_calc_rename_reservation(xfs_mount_t *mp) +xfs_calc_rename_reservation( + struct xfs_mount *mp) { - return XFS_CALC_RENAME_LOG_RES(mp) + XFS_DQUOT_LOGRES(mp); + return XFS_DQUOT_LOGRES(mp) + + MAX((4 * mp->m_sb.sb_inodesize + + 2 * XFS_DIROP_LOG_RES(mp) + + 128 * (4 + 2 * XFS_DIROP_LOG_COUNT(mp))), + (3 * mp->m_sb.sb_sectsize + + 3 * mp->m_sb.sb_sectsize + + mp->m_sb.sb_sectsize + + XFS_ALLOCFREE_LOG_RES(mp, 3) + + 128 * (7 + XFS_ALLOCFREE_LOG_COUNT(mp, 3)))); } +/* + * For creating a link to an inode: + * the parent directory inode: inode size + * the linked inode: inode size + * the directory btree could split: (max depth + v2) * dir block size + * the directory bmap btree could join or split: (max depth + v2) * blocksize + * And the bmap_finish transaction can free some bmap blocks giving: + * the agf for the ag in which the blocks live: sector size + * the agfl for the ag in which the blocks live: sector size + * the superblock for the free block count: sector size + * the allocation btrees: 2 trees * (2 * max depth - 1) * block size + */ STATIC uint -xfs_calc_link_reservation(xfs_mount_t *mp) +xfs_calc_link_reservation( + struct xfs_mount *mp) { - return XFS_CALC_LINK_LOG_RES(mp) + XFS_DQUOT_LOGRES(mp); + return XFS_DQUOT_LOGRES(mp) + + MAX((mp->m_sb.sb_inodesize + + mp->m_sb.sb_inodesize + + XFS_DIROP_LOG_RES(mp) + + 128 * (2 + XFS_DIROP_LOG_COUNT(mp))), + (mp->m_sb.sb_sectsize + + mp->m_sb.sb_sectsize + + mp->m_sb.sb_sectsize + + XFS_ALLOCFREE_LOG_RES(mp, 1) + + 128 * (3 + XFS_ALLOCFREE_LOG_COUNT(mp, 1)))); } +/* + * For removing a directory entry we can modify: + * the parent directory inode: inode size + * the removed inode: inode size + * the directory btree could join: (max depth + v2) * dir block size + * the directory bmap btree could join or split: (max depth + v2) * blocksize + * And the bmap_finish transaction can free the dir and bmap blocks giving: + * the agf for the ag in which the blocks live: 2 * sector size + * the agfl for the ag in which the blocks live: 2 * sector size + * the superblock for the free block count: sector size + * the allocation btrees: 2 exts * 2 trees * (2 * max depth - 1) * block size + */ STATIC uint -xfs_calc_remove_reservation(xfs_mount_t *mp) +xfs_calc_remove_reservation( + struct xfs_mount *mp) { - return XFS_CALC_REMOVE_LOG_RES(mp) + XFS_DQUOT_LOGRES(mp); + return XFS_DQUOT_LOGRES(mp) + + MAX((mp->m_sb.sb_inodesize + + mp->m_sb.sb_inodesize + + XFS_DIROP_LOG_RES(mp) + + 128 * (2 + XFS_DIROP_LOG_COUNT(mp))), + (2 * mp->m_sb.sb_sectsize + + 2 * mp->m_sb.sb_sectsize + + mp->m_sb.sb_sectsize + + XFS_ALLOCFREE_LOG_RES(mp, 2) + + 128 * (5 + XFS_ALLOCFREE_LOG_COUNT(mp, 2)))); } +/* + * For symlink we can modify: + * the parent directory inode: inode size + * the new inode: inode size + * the inode btree entry: 1 block + * the directory btree: (max depth + v2) * dir block size + * the directory inode's bmap btree: (max depth + v2) * block size + * the blocks for the symlink: 1 kB + * Or in the first xact we allocate some inodes giving: + * the agi and agf of the ag getting the new inodes: 2 * sectorsize + * the inode blocks allocated: XFS_IALLOC_BLOCKS * blocksize + * the inode btree: max depth * blocksize + * the allocation btrees: 2 trees * (2 * max depth - 1) * block size + */ STATIC uint -xfs_calc_symlink_reservation(xfs_mount_t *mp) +xfs_calc_symlink_reservation( + struct xfs_mount *mp) { - return XFS_CALC_SYMLINK_LOG_RES(mp) + XFS_DQUOT_LOGRES(mp); + return XFS_DQUOT_LOGRES(mp) + + MAX((mp->m_sb.sb_inodesize + + mp->m_sb.sb_inodesize + + XFS_FSB_TO_B(mp, 1) + + XFS_DIROP_LOG_RES(mp) + + 1024 + + 128 * (4 + XFS_DIROP_LOG_COUNT(mp))), + (2 * mp->m_sb.sb_sectsize + + XFS_FSB_TO_B(mp, XFS_IALLOC_BLOCKS(mp)) + + XFS_FSB_TO_B(mp, mp->m_in_maxlevels) + + XFS_ALLOCFREE_LOG_RES(mp, 1) + + 128 * (2 + XFS_IALLOC_BLOCKS(mp) + mp->m_in_maxlevels + + XFS_ALLOCFREE_LOG_COUNT(mp, 1)))); } +/* + * For create we can modify: + * the parent directory inode: inode size + * the new inode: inode size + * the inode btree entry: block size + * the superblock for the nlink flag: sector size + * the directory btree: (max depth + v2) * dir block size + * the directory inode's bmap btree: (max depth + v2) * block size + * Or in the first xact we allocate some inodes giving: + * the agi and agf of the ag getting the new inodes: 2 * sectorsize + * the superblock for the nlink flag: sector size + * the inode blocks allocated: XFS_IALLOC_BLOCKS * blocksize + * the inode btree: max depth * blocksize + * the allocation btrees: 2 trees * (max depth - 1) * block size + */ STATIC uint -xfs_calc_create_reservation(xfs_mount_t *mp) +xfs_calc_create_reservation( + struct xfs_mount *mp) { - return XFS_CALC_CREATE_LOG_RES(mp) + XFS_DQUOT_LOGRES(mp); + return XFS_DQUOT_LOGRES(mp) + + MAX((mp->m_sb.sb_inodesize + + mp->m_sb.sb_inodesize + + mp->m_sb.sb_sectsize + + XFS_FSB_TO_B(mp, 1) + + XFS_DIROP_LOG_RES(mp) + + 128 * (3 + XFS_DIROP_LOG_COUNT(mp))), + (3 * mp->m_sb.sb_sectsize + + XFS_FSB_TO_B(mp, XFS_IALLOC_BLOCKS(mp)) + + XFS_FSB_TO_B(mp, mp->m_in_maxlevels) + + XFS_ALLOCFREE_LOG_RES(mp, 1) + + 128 * (2 + XFS_IALLOC_BLOCKS(mp) + mp->m_in_maxlevels + + XFS_ALLOCFREE_LOG_COUNT(mp, 1)))); } +/* + * Making a new directory is the same as creating a new file. + */ STATIC uint -xfs_calc_mkdir_reservation(xfs_mount_t *mp) +xfs_calc_mkdir_reservation( + struct xfs_mount *mp) { - return XFS_CALC_MKDIR_LOG_RES(mp) + XFS_DQUOT_LOGRES(mp); + return xfs_calc_create_reservation(mp); } +/* + * In freeing an inode we can modify: + * the inode being freed: inode size + * the super block free inode counter: sector size + * the agi hash list and counters: sector size + * the inode btree entry: block size + * the on disk inode before ours in the agi hash list: inode cluster size + * the inode btree: max depth * blocksize + * the allocation btrees: 2 trees * (max depth - 1) * block size + */ STATIC uint -xfs_calc_ifree_reservation(xfs_mount_t *mp) +xfs_calc_ifree_reservation( + struct xfs_mount *mp) { - return XFS_CALC_IFREE_LOG_RES(mp) + XFS_DQUOT_LOGRES(mp); + return XFS_DQUOT_LOGRES(mp) + + mp->m_sb.sb_inodesize + + mp->m_sb.sb_sectsize + + mp->m_sb.sb_sectsize + + XFS_FSB_TO_B(mp, 1) + + MAX((__uint16_t)XFS_FSB_TO_B(mp, 1), + XFS_INODE_CLUSTER_SIZE(mp)) + + 128 * 5 + + XFS_ALLOCFREE_LOG_RES(mp, 1) + + 128 * (2 + XFS_IALLOC_BLOCKS(mp) + mp->m_in_maxlevels + + XFS_ALLOCFREE_LOG_COUNT(mp, 1)); } +/* + * When only changing the inode we log the inode and possibly the superblock + * We also add a bit of slop for the transaction stuff. + */ STATIC uint -xfs_calc_ichange_reservation(xfs_mount_t *mp) +xfs_calc_ichange_reservation( + struct xfs_mount *mp) { - return XFS_CALC_ICHANGE_LOG_RES(mp) + XFS_DQUOT_LOGRES(mp); + return XFS_DQUOT_LOGRES(mp) + + mp->m_sb.sb_inodesize + + mp->m_sb.sb_sectsize + + 512; + } +/* + * Growing the data section of the filesystem. + * superblock + * agi and agf + * allocation btrees + */ STATIC uint -xfs_calc_growdata_reservation(xfs_mount_t *mp) +xfs_calc_growdata_reservation( + struct xfs_mount *mp) { - return XFS_CALC_GROWDATA_LOG_RES(mp); + return mp->m_sb.sb_sectsize * 3 + + XFS_ALLOCFREE_LOG_RES(mp, 1) + + 128 * (3 + XFS_ALLOCFREE_LOG_COUNT(mp, 1)); } +/* + * Growing the rt section of the filesystem. + * In the first set of transactions (ALLOC) we allocate space to the + * bitmap or summary files. + * superblock: sector size + * agf of the ag from which the extent is allocated: sector size + * bmap btree for bitmap/summary inode: max depth * blocksize + * bitmap/summary inode: inode size + * allocation btrees for 1 block alloc: 2 * (2 * maxdepth - 1) * blocksize + */ STATIC uint -xfs_calc_growrtalloc_reservation(xfs_mount_t *mp) +xfs_calc_growrtalloc_reservation( + struct xfs_mount *mp) { - return XFS_CALC_GROWRTALLOC_LOG_RES(mp); + return 2 * mp->m_sb.sb_sectsize + + XFS_FSB_TO_B(mp, XFS_BM_MAXLEVELS(mp, XFS_DATA_FORK)) + + mp->m_sb.sb_inodesize + + XFS_ALLOCFREE_LOG_RES(mp, 1) + + 128 * (3 + XFS_BM_MAXLEVELS(mp, XFS_DATA_FORK) + + XFS_ALLOCFREE_LOG_COUNT(mp, 1)); } +/* + * Growing the rt section of the filesystem. + * In the second set of transactions (ZERO) we zero the new metadata blocks. + * one bitmap/summary block: blocksize + */ STATIC uint -xfs_calc_growrtzero_reservation(xfs_mount_t *mp) +xfs_calc_growrtzero_reservation( + struct xfs_mount *mp) { - return XFS_CALC_GROWRTZERO_LOG_RES(mp); + return mp->m_sb.sb_blocksize + 128; } +/* + * Growing the rt section of the filesystem. + * In the third set of transactions (FREE) we update metadata without + * allocating any new blocks. + * superblock: sector size + * bitmap inode: inode size + * summary inode: inode size + * one bitmap block: blocksize + * summary blocks: new summary size + */ STATIC uint -xfs_calc_growrtfree_reservation(xfs_mount_t *mp) +xfs_calc_growrtfree_reservation( + struct xfs_mount *mp) { - return XFS_CALC_GROWRTFREE_LOG_RES(mp); + return mp->m_sb.sb_sectsize + + 2 * mp->m_sb.sb_inodesize + + mp->m_sb.sb_blocksize + + mp->m_rsumsize + + 128 * 5; } +/* + * Logging the inode modification timestamp on a synchronous write. + * inode + */ STATIC uint -xfs_calc_swrite_reservation(xfs_mount_t *mp) +xfs_calc_swrite_reservation( + struct xfs_mount *mp) { - return XFS_CALC_SWRITE_LOG_RES(mp); + return mp->m_sb.sb_inodesize + 128; } +/* + * Logging the inode mode bits when writing a setuid/setgid file + * inode + */ STATIC uint xfs_calc_writeid_reservation(xfs_mount_t *mp) { - return XFS_CALC_WRITEID_LOG_RES(mp); + return mp->m_sb.sb_inodesize + 128; } +/* + * Converting the inode from non-attributed to attributed. + * the inode being converted: inode size + * agf block and superblock (for block allocation) + * the new block (directory sized) + * bmap blocks for the new directory block + * allocation btrees + */ STATIC uint -xfs_calc_addafork_reservation(xfs_mount_t *mp) +xfs_calc_addafork_reservation( + struct xfs_mount *mp) { - return XFS_CALC_ADDAFORK_LOG_RES(mp) + XFS_DQUOT_LOGRES(mp); + return XFS_DQUOT_LOGRES(mp) + + mp->m_sb.sb_inodesize + + mp->m_sb.sb_sectsize * 2 + + mp->m_dirblksize + + XFS_FSB_TO_B(mp, XFS_DAENTER_BMAP1B(mp, XFS_DATA_FORK) + 1) + + XFS_ALLOCFREE_LOG_RES(mp, 1) + + 128 * (4 + XFS_DAENTER_BMAP1B(mp, XFS_DATA_FORK) + 1 + + XFS_ALLOCFREE_LOG_COUNT(mp, 1)); } +/* + * Removing the attribute fork of a file + * the inode being truncated: inode size + * the inode's bmap btree: max depth * block size + * And the bmap_finish transaction can free the blocks and bmap blocks: + * the agf for each of the ags: 4 * sector size + * the agfl for each of the ags: 4 * sector size + * the super block to reflect the freed blocks: sector size + * worst case split in allocation btrees per extent assuming 4 extents: + * 4 exts * 2 trees * (2 * max depth - 1) * block size + */ STATIC uint -xfs_calc_attrinval_reservation(xfs_mount_t *mp) +xfs_calc_attrinval_reservation( + struct xfs_mount *mp) { - return XFS_CALC_ATTRINVAL_LOG_RES(mp); + return MAX((mp->m_sb.sb_inodesize + + XFS_FSB_TO_B(mp, XFS_BM_MAXLEVELS(mp, XFS_ATTR_FORK)) + + 128 * (1 + XFS_BM_MAXLEVELS(mp, XFS_ATTR_FORK))), + (4 * mp->m_sb.sb_sectsize + + 4 * mp->m_sb.sb_sectsize + + mp->m_sb.sb_sectsize + + XFS_ALLOCFREE_LOG_RES(mp, 4) + + 128 * (9 + XFS_ALLOCFREE_LOG_COUNT(mp, 4)))); } +/* + * Setting an attribute. + * the inode getting the attribute + * the superblock for allocations + * the agfs extents are allocated from + * the attribute btree * max depth + * the inode allocation btree + * Since attribute transaction space is dependent on the size of the attribute, + * the calculation is done partially at mount time and partially at runtime. + */ STATIC uint -xfs_calc_attrset_reservation(xfs_mount_t *mp) +xfs_calc_attrset_reservation( + struct xfs_mount *mp) { - return XFS_CALC_ATTRSET_LOG_RES(mp) + XFS_DQUOT_LOGRES(mp); + return XFS_DQUOT_LOGRES(mp) + + mp->m_sb.sb_inodesize + + mp->m_sb.sb_sectsize + + XFS_FSB_TO_B(mp, XFS_DA_NODE_MAXDEPTH) + + 128 * (2 + XFS_DA_NODE_MAXDEPTH); } +/* + * Removing an attribute. + * the inode: inode size + * the attribute btree could join: max depth * block size + * the inode bmap btree could join or split: max depth * block size + * And the bmap_finish transaction can free the attr blocks freed giving: + * the agf for the ag in which the blocks live: 2 * sector size + * the agfl for the ag in which the blocks live: 2 * sector size + * the superblock for the free block count: sector size + * the allocation btrees: 2 exts * 2 trees * (2 * max depth - 1) * block size + */ STATIC uint -xfs_calc_attrrm_reservation(xfs_mount_t *mp) +xfs_calc_attrrm_reservation( + struct xfs_mount *mp) { - return XFS_CALC_ATTRRM_LOG_RES(mp) + XFS_DQUOT_LOGRES(mp); + return XFS_DQUOT_LOGRES(mp) + + MAX((mp->m_sb.sb_inodesize + + XFS_FSB_TO_B(mp, XFS_DA_NODE_MAXDEPTH) + + XFS_FSB_TO_B(mp, XFS_BM_MAXLEVELS(mp, XFS_ATTR_FORK)) + + 128 * (1 + XFS_DA_NODE_MAXDEPTH + + XFS_BM_MAXLEVELS(mp, XFS_DATA_FORK))), + (2 * mp->m_sb.sb_sectsize + + 2 * mp->m_sb.sb_sectsize + + mp->m_sb.sb_sectsize + + XFS_ALLOCFREE_LOG_RES(mp, 2) + + 128 * (5 + XFS_ALLOCFREE_LOG_COUNT(mp, 2)))); } +/* + * Clearing a bad agino number in an agi hash bucket. + */ STATIC uint -xfs_calc_clear_agi_bucket_reservation(xfs_mount_t *mp) +xfs_calc_clear_agi_bucket_reservation( + struct xfs_mount *mp) { - return XFS_CALC_CLEAR_AGI_BUCKET_LOG_RES(mp); + return mp->m_sb.sb_sectsize + 128; } /* @@ -184,11 +539,10 @@ xfs_calc_clear_agi_bucket_reservation(xfs_mount_t *mp) */ void xfs_trans_init( - xfs_mount_t *mp) + struct xfs_mount *mp) { - xfs_trans_reservations_t *resp; + struct xfs_trans_reservations *resp = &mp->m_reservations; - resp = &(mp->m_reservations); resp->tr_write = xfs_calc_write_reservation(mp); resp->tr_itruncate = xfs_calc_itruncate_reservation(mp); resp->tr_rename = xfs_calc_rename_reservation(mp); diff --git a/fs/xfs/xfs_trans.h b/fs/xfs/xfs_trans.h index 8c69e7824f68..e639e8e9a2a9 100644 --- a/fs/xfs/xfs_trans.h +++ b/fs/xfs/xfs_trans.h @@ -299,24 +299,6 @@ xfs_lic_desc_to_chunk(xfs_log_item_desc_t *dp) #define XFS_TRANS_SB_REXTSLOG 0x00002000 -/* - * Various log reservation values. - * These are based on the size of the file system block - * because that is what most transactions manipulate. - * Each adds in an additional 128 bytes per item logged to - * try to account for the overhead of the transaction mechanism. - * - * Note: - * Most of the reservations underestimate the number of allocation - * groups into which they could free extents in the xfs_bmap_finish() - * call. This is because the number in the worst case is quite high - * and quite unusual. In order to fix this we need to change - * xfs_bmap_finish() to free extents in only a single AG at a time. - * This will require changes to the EFI code as well, however, so that - * the EFI for the extents not freed is logged again in each transaction. - * See bug 261917. - */ - /* * Per-extent log reservation for the allocation btree changes * involved in freeing or allocating an extent. @@ -341,429 +323,36 @@ xfs_lic_desc_to_chunk(xfs_log_item_desc_t *dp) (XFS_DAENTER_BLOCKS(mp, XFS_DATA_FORK) + \ XFS_DAENTER_BMAPS(mp, XFS_DATA_FORK) + 1) -/* - * In a write transaction we can allocate a maximum of 2 - * extents. This gives: - * the inode getting the new extents: inode size - * the inode's bmap btree: max depth * block size - * the agfs of the ags from which the extents are allocated: 2 * sector - * the superblock free block counter: sector size - * the allocation btrees: 2 exts * 2 trees * (2 * max depth - 1) * block size - * And the bmap_finish transaction can free bmap blocks in a join: - * the agfs of the ags containing the blocks: 2 * sector size - * the agfls of the ags containing the blocks: 2 * sector size - * the super block free block counter: sector size - * the allocation btrees: 2 exts * 2 trees * (2 * max depth - 1) * block size - */ -#define XFS_CALC_WRITE_LOG_RES(mp) \ - (MAX( \ - ((mp)->m_sb.sb_inodesize + \ - XFS_FSB_TO_B((mp), XFS_BM_MAXLEVELS(mp, XFS_DATA_FORK)) + \ - (2 * (mp)->m_sb.sb_sectsize) + \ - (mp)->m_sb.sb_sectsize + \ - XFS_ALLOCFREE_LOG_RES(mp, 2) + \ - (128 * (4 + XFS_BM_MAXLEVELS(mp, XFS_DATA_FORK) + XFS_ALLOCFREE_LOG_COUNT(mp, 2)))),\ - ((2 * (mp)->m_sb.sb_sectsize) + \ - (2 * (mp)->m_sb.sb_sectsize) + \ - (mp)->m_sb.sb_sectsize + \ - XFS_ALLOCFREE_LOG_RES(mp, 2) + \ - (128 * (5 + XFS_ALLOCFREE_LOG_COUNT(mp, 2)))))) #define XFS_WRITE_LOG_RES(mp) ((mp)->m_reservations.tr_write) - -/* - * In truncating a file we free up to two extents at once. We can modify: - * the inode being truncated: inode size - * the inode's bmap btree: (max depth + 1) * block size - * And the bmap_finish transaction can free the blocks and bmap blocks: - * the agf for each of the ags: 4 * sector size - * the agfl for each of the ags: 4 * sector size - * the super block to reflect the freed blocks: sector size - * worst case split in allocation btrees per extent assuming 4 extents: - * 4 exts * 2 trees * (2 * max depth - 1) * block size - * the inode btree: max depth * blocksize - * the allocation btrees: 2 trees * (max depth - 1) * block size - */ -#define XFS_CALC_ITRUNCATE_LOG_RES(mp) \ - (MAX( \ - ((mp)->m_sb.sb_inodesize + \ - XFS_FSB_TO_B((mp), XFS_BM_MAXLEVELS(mp, XFS_DATA_FORK) + 1) + \ - (128 * (2 + XFS_BM_MAXLEVELS(mp, XFS_DATA_FORK)))), \ - ((4 * (mp)->m_sb.sb_sectsize) + \ - (4 * (mp)->m_sb.sb_sectsize) + \ - (mp)->m_sb.sb_sectsize + \ - XFS_ALLOCFREE_LOG_RES(mp, 4) + \ - (128 * (9 + XFS_ALLOCFREE_LOG_COUNT(mp, 4))) + \ - (128 * 5) + \ - XFS_ALLOCFREE_LOG_RES(mp, 1) + \ - (128 * (2 + XFS_IALLOC_BLOCKS(mp) + (mp)->m_in_maxlevels + \ - XFS_ALLOCFREE_LOG_COUNT(mp, 1)))))) - #define XFS_ITRUNCATE_LOG_RES(mp) ((mp)->m_reservations.tr_itruncate) - -/* - * In renaming a files we can modify: - * the four inodes involved: 4 * inode size - * the two directory btrees: 2 * (max depth + v2) * dir block size - * the two directory bmap btrees: 2 * max depth * block size - * And the bmap_finish transaction can free dir and bmap blocks (two sets - * of bmap blocks) giving: - * the agf for the ags in which the blocks live: 3 * sector size - * the agfl for the ags in which the blocks live: 3 * sector size - * the superblock for the free block count: sector size - * the allocation btrees: 3 exts * 2 trees * (2 * max depth - 1) * block size - */ -#define XFS_CALC_RENAME_LOG_RES(mp) \ - (MAX( \ - ((4 * (mp)->m_sb.sb_inodesize) + \ - (2 * XFS_DIROP_LOG_RES(mp)) + \ - (128 * (4 + 2 * XFS_DIROP_LOG_COUNT(mp)))), \ - ((3 * (mp)->m_sb.sb_sectsize) + \ - (3 * (mp)->m_sb.sb_sectsize) + \ - (mp)->m_sb.sb_sectsize + \ - XFS_ALLOCFREE_LOG_RES(mp, 3) + \ - (128 * (7 + XFS_ALLOCFREE_LOG_COUNT(mp, 3)))))) - #define XFS_RENAME_LOG_RES(mp) ((mp)->m_reservations.tr_rename) - -/* - * For creating a link to an inode: - * the parent directory inode: inode size - * the linked inode: inode size - * the directory btree could split: (max depth + v2) * dir block size - * the directory bmap btree could join or split: (max depth + v2) * blocksize - * And the bmap_finish transaction can free some bmap blocks giving: - * the agf for the ag in which the blocks live: sector size - * the agfl for the ag in which the blocks live: sector size - * the superblock for the free block count: sector size - * the allocation btrees: 2 trees * (2 * max depth - 1) * block size - */ -#define XFS_CALC_LINK_LOG_RES(mp) \ - (MAX( \ - ((mp)->m_sb.sb_inodesize + \ - (mp)->m_sb.sb_inodesize + \ - XFS_DIROP_LOG_RES(mp) + \ - (128 * (2 + XFS_DIROP_LOG_COUNT(mp)))), \ - ((mp)->m_sb.sb_sectsize + \ - (mp)->m_sb.sb_sectsize + \ - (mp)->m_sb.sb_sectsize + \ - XFS_ALLOCFREE_LOG_RES(mp, 1) + \ - (128 * (3 + XFS_ALLOCFREE_LOG_COUNT(mp, 1)))))) - #define XFS_LINK_LOG_RES(mp) ((mp)->m_reservations.tr_link) - -/* - * For removing a directory entry we can modify: - * the parent directory inode: inode size - * the removed inode: inode size - * the directory btree could join: (max depth + v2) * dir block size - * the directory bmap btree could join or split: (max depth + v2) * blocksize - * And the bmap_finish transaction can free the dir and bmap blocks giving: - * the agf for the ag in which the blocks live: 2 * sector size - * the agfl for the ag in which the blocks live: 2 * sector size - * the superblock for the free block count: sector size - * the allocation btrees: 2 exts * 2 trees * (2 * max depth - 1) * block size - */ -#define XFS_CALC_REMOVE_LOG_RES(mp) \ - (MAX( \ - ((mp)->m_sb.sb_inodesize + \ - (mp)->m_sb.sb_inodesize + \ - XFS_DIROP_LOG_RES(mp) + \ - (128 * (2 + XFS_DIROP_LOG_COUNT(mp)))), \ - ((2 * (mp)->m_sb.sb_sectsize) + \ - (2 * (mp)->m_sb.sb_sectsize) + \ - (mp)->m_sb.sb_sectsize + \ - XFS_ALLOCFREE_LOG_RES(mp, 2) + \ - (128 * (5 + XFS_ALLOCFREE_LOG_COUNT(mp, 2)))))) - #define XFS_REMOVE_LOG_RES(mp) ((mp)->m_reservations.tr_remove) - -/* - * For symlink we can modify: - * the parent directory inode: inode size - * the new inode: inode size - * the inode btree entry: 1 block - * the directory btree: (max depth + v2) * dir block size - * the directory inode's bmap btree: (max depth + v2) * block size - * the blocks for the symlink: 1 kB - * Or in the first xact we allocate some inodes giving: - * the agi and agf of the ag getting the new inodes: 2 * sectorsize - * the inode blocks allocated: XFS_IALLOC_BLOCKS * blocksize - * the inode btree: max depth * blocksize - * the allocation btrees: 2 trees * (2 * max depth - 1) * block size - */ -#define XFS_CALC_SYMLINK_LOG_RES(mp) \ - (MAX( \ - ((mp)->m_sb.sb_inodesize + \ - (mp)->m_sb.sb_inodesize + \ - XFS_FSB_TO_B(mp, 1) + \ - XFS_DIROP_LOG_RES(mp) + \ - 1024 + \ - (128 * (4 + XFS_DIROP_LOG_COUNT(mp)))), \ - (2 * (mp)->m_sb.sb_sectsize + \ - XFS_FSB_TO_B((mp), XFS_IALLOC_BLOCKS((mp))) + \ - XFS_FSB_TO_B((mp), (mp)->m_in_maxlevels) + \ - XFS_ALLOCFREE_LOG_RES(mp, 1) + \ - (128 * (2 + XFS_IALLOC_BLOCKS(mp) + (mp)->m_in_maxlevels + \ - XFS_ALLOCFREE_LOG_COUNT(mp, 1)))))) - #define XFS_SYMLINK_LOG_RES(mp) ((mp)->m_reservations.tr_symlink) - -/* - * For create we can modify: - * the parent directory inode: inode size - * the new inode: inode size - * the inode btree entry: block size - * the superblock for the nlink flag: sector size - * the directory btree: (max depth + v2) * dir block size - * the directory inode's bmap btree: (max depth + v2) * block size - * Or in the first xact we allocate some inodes giving: - * the agi and agf of the ag getting the new inodes: 2 * sectorsize - * the superblock for the nlink flag: sector size - * the inode blocks allocated: XFS_IALLOC_BLOCKS * blocksize - * the inode btree: max depth * blocksize - * the allocation btrees: 2 trees * (max depth - 1) * block size - */ -#define XFS_CALC_CREATE_LOG_RES(mp) \ - (MAX( \ - ((mp)->m_sb.sb_inodesize + \ - (mp)->m_sb.sb_inodesize + \ - (mp)->m_sb.sb_sectsize + \ - XFS_FSB_TO_B(mp, 1) + \ - XFS_DIROP_LOG_RES(mp) + \ - (128 * (3 + XFS_DIROP_LOG_COUNT(mp)))), \ - (3 * (mp)->m_sb.sb_sectsize + \ - XFS_FSB_TO_B((mp), XFS_IALLOC_BLOCKS((mp))) + \ - XFS_FSB_TO_B((mp), (mp)->m_in_maxlevels) + \ - XFS_ALLOCFREE_LOG_RES(mp, 1) + \ - (128 * (2 + XFS_IALLOC_BLOCKS(mp) + (mp)->m_in_maxlevels + \ - XFS_ALLOCFREE_LOG_COUNT(mp, 1)))))) - #define XFS_CREATE_LOG_RES(mp) ((mp)->m_reservations.tr_create) - -/* - * Making a new directory is the same as creating a new file. - */ -#define XFS_CALC_MKDIR_LOG_RES(mp) XFS_CALC_CREATE_LOG_RES(mp) - #define XFS_MKDIR_LOG_RES(mp) ((mp)->m_reservations.tr_mkdir) - -/* - * In freeing an inode we can modify: - * the inode being freed: inode size - * the super block free inode counter: sector size - * the agi hash list and counters: sector size - * the inode btree entry: block size - * the on disk inode before ours in the agi hash list: inode cluster size - * the inode btree: max depth * blocksize - * the allocation btrees: 2 trees * (max depth - 1) * block size - */ -#define XFS_CALC_IFREE_LOG_RES(mp) \ - ((mp)->m_sb.sb_inodesize + \ - (mp)->m_sb.sb_sectsize + \ - (mp)->m_sb.sb_sectsize + \ - XFS_FSB_TO_B((mp), 1) + \ - MAX((__uint16_t)XFS_FSB_TO_B((mp), 1), XFS_INODE_CLUSTER_SIZE(mp)) + \ - (128 * 5) + \ - XFS_ALLOCFREE_LOG_RES(mp, 1) + \ - (128 * (2 + XFS_IALLOC_BLOCKS(mp) + (mp)->m_in_maxlevels + \ - XFS_ALLOCFREE_LOG_COUNT(mp, 1)))) - - #define XFS_IFREE_LOG_RES(mp) ((mp)->m_reservations.tr_ifree) - -/* - * When only changing the inode we log the inode and possibly the superblock - * We also add a bit of slop for the transaction stuff. - */ -#define XFS_CALC_ICHANGE_LOG_RES(mp) ((mp)->m_sb.sb_inodesize + \ - (mp)->m_sb.sb_sectsize + 512) - #define XFS_ICHANGE_LOG_RES(mp) ((mp)->m_reservations.tr_ichange) - -/* - * Growing the data section of the filesystem. - * superblock - * agi and agf - * allocation btrees - */ -#define XFS_CALC_GROWDATA_LOG_RES(mp) \ - ((mp)->m_sb.sb_sectsize * 3 + \ - XFS_ALLOCFREE_LOG_RES(mp, 1) + \ - (128 * (3 + XFS_ALLOCFREE_LOG_COUNT(mp, 1)))) - #define XFS_GROWDATA_LOG_RES(mp) ((mp)->m_reservations.tr_growdata) - -/* - * Growing the rt section of the filesystem. - * In the first set of transactions (ALLOC) we allocate space to the - * bitmap or summary files. - * superblock: sector size - * agf of the ag from which the extent is allocated: sector size - * bmap btree for bitmap/summary inode: max depth * blocksize - * bitmap/summary inode: inode size - * allocation btrees for 1 block alloc: 2 * (2 * maxdepth - 1) * blocksize - */ -#define XFS_CALC_GROWRTALLOC_LOG_RES(mp) \ - (2 * (mp)->m_sb.sb_sectsize + \ - XFS_FSB_TO_B((mp), XFS_BM_MAXLEVELS(mp, XFS_DATA_FORK)) + \ - (mp)->m_sb.sb_inodesize + \ - XFS_ALLOCFREE_LOG_RES(mp, 1) + \ - (128 * \ - (3 + XFS_BM_MAXLEVELS(mp, XFS_DATA_FORK) + \ - XFS_ALLOCFREE_LOG_COUNT(mp, 1)))) - #define XFS_GROWRTALLOC_LOG_RES(mp) ((mp)->m_reservations.tr_growrtalloc) - -/* - * Growing the rt section of the filesystem. - * In the second set of transactions (ZERO) we zero the new metadata blocks. - * one bitmap/summary block: blocksize - */ -#define XFS_CALC_GROWRTZERO_LOG_RES(mp) \ - ((mp)->m_sb.sb_blocksize + 128) - #define XFS_GROWRTZERO_LOG_RES(mp) ((mp)->m_reservations.tr_growrtzero) - -/* - * Growing the rt section of the filesystem. - * In the third set of transactions (FREE) we update metadata without - * allocating any new blocks. - * superblock: sector size - * bitmap inode: inode size - * summary inode: inode size - * one bitmap block: blocksize - * summary blocks: new summary size - */ -#define XFS_CALC_GROWRTFREE_LOG_RES(mp) \ - ((mp)->m_sb.sb_sectsize + \ - 2 * (mp)->m_sb.sb_inodesize + \ - (mp)->m_sb.sb_blocksize + \ - (mp)->m_rsumsize + \ - (128 * 5)) - #define XFS_GROWRTFREE_LOG_RES(mp) ((mp)->m_reservations.tr_growrtfree) - -/* - * Logging the inode modification timestamp on a synchronous write. - * inode - */ -#define XFS_CALC_SWRITE_LOG_RES(mp) \ - ((mp)->m_sb.sb_inodesize + 128) - #define XFS_SWRITE_LOG_RES(mp) ((mp)->m_reservations.tr_swrite) - /* * Logging the inode timestamps on an fsync -- same as SWRITE * as long as SWRITE logs the entire inode core */ #define XFS_FSYNC_TS_LOG_RES(mp) ((mp)->m_reservations.tr_swrite) - -/* - * Logging the inode mode bits when writing a setuid/setgid file - * inode - */ -#define XFS_CALC_WRITEID_LOG_RES(mp) \ - ((mp)->m_sb.sb_inodesize + 128) - #define XFS_WRITEID_LOG_RES(mp) ((mp)->m_reservations.tr_swrite) - -/* - * Converting the inode from non-attributed to attributed. - * the inode being converted: inode size - * agf block and superblock (for block allocation) - * the new block (directory sized) - * bmap blocks for the new directory block - * allocation btrees - */ -#define XFS_CALC_ADDAFORK_LOG_RES(mp) \ - ((mp)->m_sb.sb_inodesize + \ - (mp)->m_sb.sb_sectsize * 2 + \ - (mp)->m_dirblksize + \ - XFS_FSB_TO_B(mp, (XFS_DAENTER_BMAP1B(mp, XFS_DATA_FORK) + 1)) + \ - XFS_ALLOCFREE_LOG_RES(mp, 1) + \ - (128 * (4 + (XFS_DAENTER_BMAP1B(mp, XFS_DATA_FORK) + 1) + \ - XFS_ALLOCFREE_LOG_COUNT(mp, 1)))) - #define XFS_ADDAFORK_LOG_RES(mp) ((mp)->m_reservations.tr_addafork) - -/* - * Removing the attribute fork of a file - * the inode being truncated: inode size - * the inode's bmap btree: max depth * block size - * And the bmap_finish transaction can free the blocks and bmap blocks: - * the agf for each of the ags: 4 * sector size - * the agfl for each of the ags: 4 * sector size - * the super block to reflect the freed blocks: sector size - * worst case split in allocation btrees per extent assuming 4 extents: - * 4 exts * 2 trees * (2 * max depth - 1) * block size - */ -#define XFS_CALC_ATTRINVAL_LOG_RES(mp) \ - (MAX( \ - ((mp)->m_sb.sb_inodesize + \ - XFS_FSB_TO_B((mp), XFS_BM_MAXLEVELS(mp, XFS_ATTR_FORK)) + \ - (128 * (1 + XFS_BM_MAXLEVELS(mp, XFS_ATTR_FORK)))), \ - ((4 * (mp)->m_sb.sb_sectsize) + \ - (4 * (mp)->m_sb.sb_sectsize) + \ - (mp)->m_sb.sb_sectsize + \ - XFS_ALLOCFREE_LOG_RES(mp, 4) + \ - (128 * (9 + XFS_ALLOCFREE_LOG_COUNT(mp, 4)))))) - #define XFS_ATTRINVAL_LOG_RES(mp) ((mp)->m_reservations.tr_attrinval) - -/* - * Setting an attribute. - * the inode getting the attribute - * the superblock for allocations - * the agfs extents are allocated from - * the attribute btree * max depth - * the inode allocation btree - * Since attribute transaction space is dependent on the size of the attribute, - * the calculation is done partially at mount time and partially at runtime. - */ -#define XFS_CALC_ATTRSET_LOG_RES(mp) \ - ((mp)->m_sb.sb_inodesize + \ - (mp)->m_sb.sb_sectsize + \ - XFS_FSB_TO_B((mp), XFS_DA_NODE_MAXDEPTH) + \ - (128 * (2 + XFS_DA_NODE_MAXDEPTH))) - #define XFS_ATTRSET_LOG_RES(mp, ext) \ ((mp)->m_reservations.tr_attrset + \ (ext * (mp)->m_sb.sb_sectsize) + \ (ext * XFS_FSB_TO_B((mp), XFS_BM_MAXLEVELS(mp, XFS_ATTR_FORK))) + \ (128 * (ext + (ext * XFS_BM_MAXLEVELS(mp, XFS_ATTR_FORK))))) - -/* - * Removing an attribute. - * the inode: inode size - * the attribute btree could join: max depth * block size - * the inode bmap btree could join or split: max depth * block size - * And the bmap_finish transaction can free the attr blocks freed giving: - * the agf for the ag in which the blocks live: 2 * sector size - * the agfl for the ag in which the blocks live: 2 * sector size - * the superblock for the free block count: sector size - * the allocation btrees: 2 exts * 2 trees * (2 * max depth - 1) * block size - */ -#define XFS_CALC_ATTRRM_LOG_RES(mp) \ - (MAX( \ - ((mp)->m_sb.sb_inodesize + \ - XFS_FSB_TO_B((mp), XFS_DA_NODE_MAXDEPTH) + \ - XFS_FSB_TO_B((mp), XFS_BM_MAXLEVELS(mp, XFS_ATTR_FORK)) + \ - (128 * (1 + XFS_DA_NODE_MAXDEPTH + XFS_BM_MAXLEVELS(mp, XFS_DATA_FORK)))), \ - ((2 * (mp)->m_sb.sb_sectsize) + \ - (2 * (mp)->m_sb.sb_sectsize) + \ - (mp)->m_sb.sb_sectsize + \ - XFS_ALLOCFREE_LOG_RES(mp, 2) + \ - (128 * (5 + XFS_ALLOCFREE_LOG_COUNT(mp, 2)))))) - #define XFS_ATTRRM_LOG_RES(mp) ((mp)->m_reservations.tr_attrrm) - -/* - * Clearing a bad agino number in an agi hash bucket. - */ -#define XFS_CALC_CLEAR_AGI_BUCKET_LOG_RES(mp) \ - ((mp)->m_sb.sb_sectsize + 128) - #define XFS_CLEAR_AGI_BUCKET_LOG_RES(mp) ((mp)->m_reservations.tr_clearagi) -- cgit v1.2.2 From fdc07f44c891d3fdee7722a03e3881614a293b3c Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Mon, 10 May 2010 17:28:14 +0000 Subject: xfs: clean up xlog_align Add suggested cleanups to commit 29db3370a1369541d58d692fbfb168b8a0bd7f41 from review that didn't end up being commited. Signed-off-by: Christoph Hellwig Signed-off-by: Alex Elder --- fs/xfs/xfs_log_recover.c | 11 +++-------- 1 file changed, 3 insertions(+), 8 deletions(-) (limited to 'fs') diff --git a/fs/xfs/xfs_log_recover.c b/fs/xfs/xfs_log_recover.c index 14a69aec2c0b..ed0684cc50ee 100644 --- a/fs/xfs/xfs_log_recover.c +++ b/fs/xfs/xfs_log_recover.c @@ -132,15 +132,10 @@ xlog_align( int nbblks, xfs_buf_t *bp) { - xfs_daddr_t offset; - xfs_caddr_t ptr; + xfs_daddr_t offset = blk_no & ((xfs_daddr_t)log->l_sectBBsize - 1); - offset = blk_no & ((xfs_daddr_t) log->l_sectBBsize - 1); - ptr = XFS_BUF_PTR(bp) + BBTOB(offset); - - ASSERT(ptr + BBTOB(nbblks) <= XFS_BUF_PTR(bp) + XFS_BUF_SIZE(bp)); - - return ptr; + ASSERT(BBTOB(offset + nbblks) <= XFS_BUF_SIZE(bp)); + return XFS_BUF_PTR(bp) + BBTOB(offset); } -- cgit v1.2.2 From 07f1a4f5e89cd4e6c79d67d41e8a18c451214ae2 Mon Sep 17 00:00:00 2001 From: Dave Chinner Date: Fri, 21 May 2010 05:47:59 +0000 Subject: xfs: Check new inode size is OK before preallocating The new xfsqa test 228 tries to preallocate more space than the filesystem contains. it should fail, but instead triggers an assert about lock flags. The failure is due to the size extension failing in vmtruncate() due to rlimit being set. Check this before we start the preallocation to avoid allocating space that will never be used. Also the path through xfs_vn_allocate already holds the IO lock, so it should not be present in the lock flags when the setattr fails. Hence the assert needs to take this into account. This will prevent other such callers from hitting this incorrect ASSERT. (Fixed a reference to "newsize" to read "new_size". -Alex) Signed-off-by: Dave Chinner Reviewed-by: Christoph Hellwig Signed-off-by: Alex Elder --- fs/xfs/linux-2.6/xfs_iops.c | 16 +++++++++++++--- fs/xfs/xfs_vnodeops.c | 2 +- 2 files changed, 14 insertions(+), 4 deletions(-) (limited to 'fs') diff --git a/fs/xfs/linux-2.6/xfs_iops.c b/fs/xfs/linux-2.6/xfs_iops.c index 9c8019c78c92..44f0b2de153e 100644 --- a/fs/xfs/linux-2.6/xfs_iops.c +++ b/fs/xfs/linux-2.6/xfs_iops.c @@ -585,11 +585,20 @@ xfs_vn_fallocate( bf.l_len = len; xfs_ilock(ip, XFS_IOLOCK_EXCL); + + /* check the new inode size is valid before allocating */ + if (!(mode & FALLOC_FL_KEEP_SIZE) && + offset + len > i_size_read(inode)) { + new_size = offset + len; + error = inode_newsize_ok(inode, new_size); + if (error) + goto out_unlock; + } + error = -xfs_change_file_space(ip, XFS_IOC_RESVSP, &bf, 0, XFS_ATTR_NOLOCK); - if (!error && !(mode & FALLOC_FL_KEEP_SIZE) && - offset + len > i_size_read(inode)) - new_size = offset + len; + if (error) + goto out_unlock; /* Change file size if needed */ if (new_size) { @@ -600,6 +609,7 @@ xfs_vn_fallocate( error = -xfs_setattr(ip, &iattr, XFS_ATTR_NOLOCK); } +out_unlock: xfs_iunlock(ip, XFS_IOLOCK_EXCL); out_error: return error; diff --git a/fs/xfs/xfs_vnodeops.c b/fs/xfs/xfs_vnodeops.c index 9d376be0ea38..a06bd62504fc 100644 --- a/fs/xfs/xfs_vnodeops.c +++ b/fs/xfs/xfs_vnodeops.c @@ -267,7 +267,7 @@ xfs_setattr( if (code) { ASSERT(tp == NULL); lock_flags &= ~XFS_ILOCK_EXCL; - ASSERT(lock_flags == XFS_IOLOCK_EXCL); + ASSERT(lock_flags == XFS_IOLOCK_EXCL || !need_iolock); goto error_return; } tp = xfs_trans_alloc(mp, XFS_TRANS_SETATTR_SIZE); -- cgit v1.2.2 From 292ec4cf3536a5ed8809e8823341b203e497bbaf Mon Sep 17 00:00:00 2001 From: Huang Weiyi Date: Sat, 22 May 2010 17:13:20 +0000 Subject: xfs: xfs_trace.c: remove duplicated #include Remove duplicated #include('s) in fs/xfs/linux-2.6/xfs_trace.c Signed-off-by: Huang Weiyi Reviewed-by: Dave Chinner Signed-off-by: Alex Elder --- fs/xfs/linux-2.6/xfs_trace.c | 1 - 1 file changed, 1 deletion(-) (limited to 'fs') diff --git a/fs/xfs/linux-2.6/xfs_trace.c b/fs/xfs/linux-2.6/xfs_trace.c index 207fa77f63ae..d12be8470cba 100644 --- a/fs/xfs/linux-2.6/xfs_trace.c +++ b/fs/xfs/linux-2.6/xfs_trace.c @@ -50,7 +50,6 @@ #include "quota/xfs_dquot_item.h" #include "quota/xfs_dquot.h" #include "xfs_log_recover.h" -#include "xfs_buf_item.h" #include "xfs_inode_item.h" /* -- cgit v1.2.2 From f8adb4d574cf8c67f8391367136edc5469258fdc Mon Sep 17 00:00:00 2001 From: Li Zefan Date: Mon, 24 May 2010 08:25:57 +0000 Subject: xfs: convert more trace events to DEFINE_EVENT Use DECLARE_EVENT_CLASS, and save ~15K: text data bss dec hex filename 171949 43028 48 215025 347f1 fs/xfs/linux-2.6/xfs_trace.o.orig 156521 43028 36 199585 30ba1 fs/xfs/linux-2.6/xfs_trace.o No change in functionality. Signed-off-by: Li Zefan Acked-by: Steven Rostedt Signed-off-by: Alex Elder --- fs/xfs/linux-2.6/xfs_trace.h | 356 +++++++++++++++++++++++-------------------- 1 file changed, 188 insertions(+), 168 deletions(-) (limited to 'fs') diff --git a/fs/xfs/linux-2.6/xfs_trace.h b/fs/xfs/linux-2.6/xfs_trace.h index ff6bc797baf2..73d5aa117384 100644 --- a/fs/xfs/linux-2.6/xfs_trace.h +++ b/fs/xfs/linux-2.6/xfs_trace.h @@ -82,33 +82,6 @@ DECLARE_EVENT_CLASS(xfs_attr_list_class, ) ) -#define DEFINE_PERAG_REF_EVENT(name) \ -TRACE_EVENT(name, \ - TP_PROTO(struct xfs_mount *mp, xfs_agnumber_t agno, int refcount, \ - unsigned long caller_ip), \ - TP_ARGS(mp, agno, refcount, caller_ip), \ - TP_STRUCT__entry( \ - __field(dev_t, dev) \ - __field(xfs_agnumber_t, agno) \ - __field(int, refcount) \ - __field(unsigned long, caller_ip) \ - ), \ - TP_fast_assign( \ - __entry->dev = mp->m_super->s_dev; \ - __entry->agno = agno; \ - __entry->refcount = refcount; \ - __entry->caller_ip = caller_ip; \ - ), \ - TP_printk("dev %d:%d agno %u refcount %d caller %pf", \ - MAJOR(__entry->dev), MINOR(__entry->dev), \ - __entry->agno, \ - __entry->refcount, \ - (char *)__entry->caller_ip) \ -); - -DEFINE_PERAG_REF_EVENT(xfs_perag_get) -DEFINE_PERAG_REF_EVENT(xfs_perag_put) - #define DEFINE_ATTR_LIST_EVENT(name) \ DEFINE_EVENT(xfs_attr_list_class, name, \ TP_PROTO(struct xfs_attr_list_context *ctx), \ @@ -122,6 +95,37 @@ DEFINE_ATTR_LIST_EVENT(xfs_attr_list_add); DEFINE_ATTR_LIST_EVENT(xfs_attr_list_wrong_blk); DEFINE_ATTR_LIST_EVENT(xfs_attr_list_notfound); +DECLARE_EVENT_CLASS(xfs_perag_class, + TP_PROTO(struct xfs_mount *mp, xfs_agnumber_t agno, int refcount, + unsigned long caller_ip), + TP_ARGS(mp, agno, refcount, caller_ip), + TP_STRUCT__entry( + __field(dev_t, dev) + __field(xfs_agnumber_t, agno) + __field(int, refcount) + __field(unsigned long, caller_ip) + ), + TP_fast_assign( + __entry->dev = mp->m_super->s_dev; + __entry->agno = agno; + __entry->refcount = refcount; + __entry->caller_ip = caller_ip; + ), + TP_printk("dev %d:%d agno %u refcount %d caller %pf", + MAJOR(__entry->dev), MINOR(__entry->dev), + __entry->agno, + __entry->refcount, + (char *)__entry->caller_ip) +); + +#define DEFINE_PERAG_REF_EVENT(name) \ +DEFINE_EVENT(xfs_perag_class, name, \ + TP_PROTO(struct xfs_mount *mp, xfs_agnumber_t agno, int refcount, \ + unsigned long caller_ip), \ + TP_ARGS(mp, agno, refcount, caller_ip)) +DEFINE_PERAG_REF_EVENT(xfs_perag_get); +DEFINE_PERAG_REF_EVENT(xfs_perag_put); + TRACE_EVENT(xfs_attr_list_node_descend, TP_PROTO(struct xfs_attr_list_context *ctx, struct xfs_da_node_entry *btree), @@ -775,165 +779,181 @@ DEFINE_LOGGRANT_EVENT(xfs_log_ungrant_enter); DEFINE_LOGGRANT_EVENT(xfs_log_ungrant_exit); DEFINE_LOGGRANT_EVENT(xfs_log_ungrant_sub); -#define DEFINE_RW_EVENT(name) \ -TRACE_EVENT(name, \ - TP_PROTO(struct xfs_inode *ip, size_t count, loff_t offset, int flags), \ - TP_ARGS(ip, count, offset, flags), \ - TP_STRUCT__entry( \ - __field(dev_t, dev) \ - __field(xfs_ino_t, ino) \ - __field(xfs_fsize_t, size) \ - __field(xfs_fsize_t, new_size) \ - __field(loff_t, offset) \ - __field(size_t, count) \ - __field(int, flags) \ - ), \ - TP_fast_assign( \ - __entry->dev = VFS_I(ip)->i_sb->s_dev; \ - __entry->ino = ip->i_ino; \ - __entry->size = ip->i_d.di_size; \ - __entry->new_size = ip->i_new_size; \ - __entry->offset = offset; \ - __entry->count = count; \ - __entry->flags = flags; \ - ), \ - TP_printk("dev %d:%d ino 0x%llx size 0x%llx new_size 0x%llx " \ - "offset 0x%llx count 0x%zx ioflags %s", \ - MAJOR(__entry->dev), MINOR(__entry->dev), \ - __entry->ino, \ - __entry->size, \ - __entry->new_size, \ - __entry->offset, \ - __entry->count, \ - __print_flags(__entry->flags, "|", XFS_IO_FLAGS)) \ +DECLARE_EVENT_CLASS(xfs_file_class, + TP_PROTO(struct xfs_inode *ip, size_t count, loff_t offset, int flags), + TP_ARGS(ip, count, offset, flags), + TP_STRUCT__entry( + __field(dev_t, dev) + __field(xfs_ino_t, ino) + __field(xfs_fsize_t, size) + __field(xfs_fsize_t, new_size) + __field(loff_t, offset) + __field(size_t, count) + __field(int, flags) + ), + TP_fast_assign( + __entry->dev = VFS_I(ip)->i_sb->s_dev; + __entry->ino = ip->i_ino; + __entry->size = ip->i_d.di_size; + __entry->new_size = ip->i_new_size; + __entry->offset = offset; + __entry->count = count; + __entry->flags = flags; + ), + TP_printk("dev %d:%d ino 0x%llx size 0x%llx new_size 0x%llx " + "offset 0x%llx count 0x%zx ioflags %s", + MAJOR(__entry->dev), MINOR(__entry->dev), + __entry->ino, + __entry->size, + __entry->new_size, + __entry->offset, + __entry->count, + __print_flags(__entry->flags, "|", XFS_IO_FLAGS)) ) + +#define DEFINE_RW_EVENT(name) \ +DEFINE_EVENT(xfs_file_class, name, \ + TP_PROTO(struct xfs_inode *ip, size_t count, loff_t offset, int flags), \ + TP_ARGS(ip, count, offset, flags)) DEFINE_RW_EVENT(xfs_file_read); DEFINE_RW_EVENT(xfs_file_buffered_write); DEFINE_RW_EVENT(xfs_file_direct_write); DEFINE_RW_EVENT(xfs_file_splice_read); DEFINE_RW_EVENT(xfs_file_splice_write); - -#define DEFINE_PAGE_EVENT(name) \ -TRACE_EVENT(name, \ - TP_PROTO(struct inode *inode, struct page *page, unsigned long off), \ - TP_ARGS(inode, page, off), \ - TP_STRUCT__entry( \ - __field(dev_t, dev) \ - __field(xfs_ino_t, ino) \ - __field(pgoff_t, pgoff) \ - __field(loff_t, size) \ - __field(unsigned long, offset) \ - __field(int, delalloc) \ - __field(int, unmapped) \ - __field(int, unwritten) \ - ), \ - TP_fast_assign( \ - int delalloc = -1, unmapped = -1, unwritten = -1; \ - \ - if (page_has_buffers(page)) \ - xfs_count_page_state(page, &delalloc, \ - &unmapped, &unwritten); \ - __entry->dev = inode->i_sb->s_dev; \ - __entry->ino = XFS_I(inode)->i_ino; \ - __entry->pgoff = page_offset(page); \ - __entry->size = i_size_read(inode); \ - __entry->offset = off; \ - __entry->delalloc = delalloc; \ - __entry->unmapped = unmapped; \ - __entry->unwritten = unwritten; \ - ), \ - TP_printk("dev %d:%d ino 0x%llx pgoff 0x%lx size 0x%llx offset %lx " \ - "delalloc %d unmapped %d unwritten %d", \ - MAJOR(__entry->dev), MINOR(__entry->dev), \ - __entry->ino, \ - __entry->pgoff, \ - __entry->size, \ - __entry->offset, \ - __entry->delalloc, \ - __entry->unmapped, \ - __entry->unwritten) \ +DECLARE_EVENT_CLASS(xfs_page_class, + TP_PROTO(struct inode *inode, struct page *page, unsigned long off), + TP_ARGS(inode, page, off), + TP_STRUCT__entry( + __field(dev_t, dev) + __field(xfs_ino_t, ino) + __field(pgoff_t, pgoff) + __field(loff_t, size) + __field(unsigned long, offset) + __field(int, delalloc) + __field(int, unmapped) + __field(int, unwritten) + ), + TP_fast_assign( + int delalloc = -1, unmapped = -1, unwritten = -1; + + if (page_has_buffers(page)) + xfs_count_page_state(page, &delalloc, + &unmapped, &unwritten); + __entry->dev = inode->i_sb->s_dev; + __entry->ino = XFS_I(inode)->i_ino; + __entry->pgoff = page_offset(page); + __entry->size = i_size_read(inode); + __entry->offset = off; + __entry->delalloc = delalloc; + __entry->unmapped = unmapped; + __entry->unwritten = unwritten; + ), + TP_printk("dev %d:%d ino 0x%llx pgoff 0x%lx size 0x%llx offset %lx " + "delalloc %d unmapped %d unwritten %d", + MAJOR(__entry->dev), MINOR(__entry->dev), + __entry->ino, + __entry->pgoff, + __entry->size, + __entry->offset, + __entry->delalloc, + __entry->unmapped, + __entry->unwritten) ) + +#define DEFINE_PAGE_EVENT(name) \ +DEFINE_EVENT(xfs_page_class, name, \ + TP_PROTO(struct inode *inode, struct page *page, unsigned long off), \ + TP_ARGS(inode, page, off)) DEFINE_PAGE_EVENT(xfs_writepage); DEFINE_PAGE_EVENT(xfs_releasepage); DEFINE_PAGE_EVENT(xfs_invalidatepage); -#define DEFINE_IOMAP_EVENT(name) \ -TRACE_EVENT(name, \ - TP_PROTO(struct xfs_inode *ip, xfs_off_t offset, ssize_t count, \ - int flags, struct xfs_bmbt_irec *irec), \ - TP_ARGS(ip, offset, count, flags, irec), \ - TP_STRUCT__entry( \ - __field(dev_t, dev) \ - __field(xfs_ino_t, ino) \ - __field(loff_t, size) \ - __field(loff_t, new_size) \ - __field(loff_t, offset) \ - __field(size_t, count) \ - __field(int, flags) \ - __field(xfs_fileoff_t, startoff) \ - __field(xfs_fsblock_t, startblock) \ - __field(xfs_filblks_t, blockcount) \ - ), \ - TP_fast_assign( \ - __entry->dev = VFS_I(ip)->i_sb->s_dev; \ - __entry->ino = ip->i_ino; \ - __entry->size = ip->i_d.di_size; \ - __entry->new_size = ip->i_new_size; \ - __entry->offset = offset; \ - __entry->count = count; \ - __entry->flags = flags; \ - __entry->startoff = irec ? irec->br_startoff : 0; \ - __entry->startblock = irec ? irec->br_startblock : 0; \ - __entry->blockcount = irec ? irec->br_blockcount : 0; \ - ), \ - TP_printk("dev %d:%d ino 0x%llx size 0x%llx new_size 0x%llx " \ - "offset 0x%llx count %zd flags %s " \ - "startoff 0x%llx startblock %lld blockcount 0x%llx", \ - MAJOR(__entry->dev), MINOR(__entry->dev), \ - __entry->ino, \ - __entry->size, \ - __entry->new_size, \ - __entry->offset, \ - __entry->count, \ - __print_flags(__entry->flags, "|", BMAPI_FLAGS), \ - __entry->startoff, \ - (__int64_t)__entry->startblock, \ - __entry->blockcount) \ +DECLARE_EVENT_CLASS(xfs_iomap_class, + TP_PROTO(struct xfs_inode *ip, xfs_off_t offset, ssize_t count, + int flags, struct xfs_bmbt_irec *irec), + TP_ARGS(ip, offset, count, flags, irec), + TP_STRUCT__entry( + __field(dev_t, dev) + __field(xfs_ino_t, ino) + __field(loff_t, size) + __field(loff_t, new_size) + __field(loff_t, offset) + __field(size_t, count) + __field(int, flags) + __field(xfs_fileoff_t, startoff) + __field(xfs_fsblock_t, startblock) + __field(xfs_filblks_t, blockcount) + ), + TP_fast_assign( + __entry->dev = VFS_I(ip)->i_sb->s_dev; + __entry->ino = ip->i_ino; + __entry->size = ip->i_d.di_size; + __entry->new_size = ip->i_new_size; + __entry->offset = offset; + __entry->count = count; + __entry->flags = flags; + __entry->startoff = irec ? irec->br_startoff : 0; + __entry->startblock = irec ? irec->br_startblock : 0; + __entry->blockcount = irec ? irec->br_blockcount : 0; + ), + TP_printk("dev %d:%d ino 0x%llx size 0x%llx new_size 0x%llx " + "offset 0x%llx count %zd flags %s " + "startoff 0x%llx startblock %lld blockcount 0x%llx", + MAJOR(__entry->dev), MINOR(__entry->dev), + __entry->ino, + __entry->size, + __entry->new_size, + __entry->offset, + __entry->count, + __print_flags(__entry->flags, "|", BMAPI_FLAGS), + __entry->startoff, + (__int64_t)__entry->startblock, + __entry->blockcount) ) + +#define DEFINE_IOMAP_EVENT(name) \ +DEFINE_EVENT(xfs_iomap_class, name, \ + TP_PROTO(struct xfs_inode *ip, xfs_off_t offset, ssize_t count, \ + int flags, struct xfs_bmbt_irec *irec), \ + TP_ARGS(ip, offset, count, flags, irec)) DEFINE_IOMAP_EVENT(xfs_iomap_enter); DEFINE_IOMAP_EVENT(xfs_iomap_found); DEFINE_IOMAP_EVENT(xfs_iomap_alloc); -#define DEFINE_SIMPLE_IO_EVENT(name) \ -TRACE_EVENT(name, \ - TP_PROTO(struct xfs_inode *ip, xfs_off_t offset, ssize_t count), \ - TP_ARGS(ip, offset, count), \ - TP_STRUCT__entry( \ - __field(dev_t, dev) \ - __field(xfs_ino_t, ino) \ - __field(loff_t, size) \ - __field(loff_t, new_size) \ - __field(loff_t, offset) \ - __field(size_t, count) \ - ), \ - TP_fast_assign( \ - __entry->dev = VFS_I(ip)->i_sb->s_dev; \ - __entry->ino = ip->i_ino; \ - __entry->size = ip->i_d.di_size; \ - __entry->new_size = ip->i_new_size; \ - __entry->offset = offset; \ - __entry->count = count; \ - ), \ - TP_printk("dev %d:%d ino 0x%llx size 0x%llx new_size 0x%llx " \ - "offset 0x%llx count %zd", \ - MAJOR(__entry->dev), MINOR(__entry->dev), \ - __entry->ino, \ - __entry->size, \ - __entry->new_size, \ - __entry->offset, \ - __entry->count) \ +DECLARE_EVENT_CLASS(xfs_simple_io_class, + TP_PROTO(struct xfs_inode *ip, xfs_off_t offset, ssize_t count), + TP_ARGS(ip, offset, count), + TP_STRUCT__entry( + __field(dev_t, dev) + __field(xfs_ino_t, ino) + __field(loff_t, size) + __field(loff_t, new_size) + __field(loff_t, offset) + __field(size_t, count) + ), + TP_fast_assign( + __entry->dev = VFS_I(ip)->i_sb->s_dev; + __entry->ino = ip->i_ino; + __entry->size = ip->i_d.di_size; + __entry->new_size = ip->i_new_size; + __entry->offset = offset; + __entry->count = count; + ), + TP_printk("dev %d:%d ino 0x%llx size 0x%llx new_size 0x%llx " + "offset 0x%llx count %zd", + MAJOR(__entry->dev), MINOR(__entry->dev), + __entry->ino, + __entry->size, + __entry->new_size, + __entry->offset, + __entry->count) ); + +#define DEFINE_SIMPLE_IO_EVENT(name) \ +DEFINE_EVENT(xfs_simple_io_class, name, \ + TP_PROTO(struct xfs_inode *ip, xfs_off_t offset, ssize_t count), \ + TP_ARGS(ip, offset, count)) DEFINE_SIMPLE_IO_EVENT(xfs_delalloc_enospc); DEFINE_SIMPLE_IO_EVENT(xfs_unwritten_convert); -- cgit v1.2.2 From 3bd0946eb157e26240ca858d1a42738b095dc6f3 Mon Sep 17 00:00:00 2001 From: Huang Weiyi Date: Tue, 25 May 2010 22:15:26 +0000 Subject: xfs: remove duplicated #include Remove duplicated #include('s) in fs/xfs/linux-2.6/xfs_quotaops.c Signed-off-by: Huang Weiyi Reviewed-by: Dave Chinner Signed-off-by: Alex Elder --- fs/xfs/linux-2.6/xfs_quotaops.c | 1 - 1 file changed, 1 deletion(-) (limited to 'fs') diff --git a/fs/xfs/linux-2.6/xfs_quotaops.c b/fs/xfs/linux-2.6/xfs_quotaops.c index 2e73688dae9c..a184793f2399 100644 --- a/fs/xfs/linux-2.6/xfs_quotaops.c +++ b/fs/xfs/linux-2.6/xfs_quotaops.c @@ -23,7 +23,6 @@ #include "xfs_ag.h" #include "xfs_mount.h" #include "xfs_quota.h" -#include "xfs_log.h" #include "xfs_trans.h" #include "xfs_bmap_btree.h" #include "xfs_inode.h" -- cgit v1.2.2 From 38e712ab3d28d79725eaade02fe8aba51abac196 Mon Sep 17 00:00:00 2001 From: Julia Lawall Date: Wed, 26 May 2010 15:57:23 +0000 Subject: fs/xfs/quota: Add missing mutex_unlock Add a mutex_unlock missing on the error path. The use of this lock is balanced elsewhere in the file. The semantic match that finds this problem is as follows: (http://coccinelle.lip6.fr/) // @@ expression E1; @@ * mutex_lock(E1,...); <+... when != E1 if (...) { ... when != E1 * return ...; } ...+> * mutex_unlock(E1,...); // Signed-off-by: Julia Lawall Signed-off-by: Alex Elder --- fs/xfs/quota/xfs_qm.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/xfs/quota/xfs_qm.c b/fs/xfs/quota/xfs_qm.c index 38e764146644..2d8b7bc792c9 100644 --- a/fs/xfs/quota/xfs_qm.c +++ b/fs/xfs/quota/xfs_qm.c @@ -249,8 +249,10 @@ xfs_qm_hold_quotafs_ref( if (!xfs_Gqm) { xfs_Gqm = xfs_Gqm_init(); - if (!xfs_Gqm) + if (!xfs_Gqm) { + mutex_unlock(&xfs_Gqm_lock); return ENOMEM; + } } /* -- cgit v1.2.2 From 9b98b6f3e1534bba2efcd5b16318945cf2218d99 Mon Sep 17 00:00:00 2001 From: Dave Chinner Date: Thu, 27 May 2010 01:58:13 +0000 Subject: xfs: fix might_sleep() warning when initialising per-ag tree The use of radix_tree_preload() only works if the radix tree was initialised without the __GFP_WAIT flag. The per-ag tree uses GFP_NOFS, so does not trigger allocation of new tree nodes from the preloaded array. Hence it enters the allocator with a spinlock held and triggers the might_sleep() warnings. Reported-by; Chris Mason Signed-off-by: Dave Chinner Reviewed-by: Christoph Hellwig Signed-off-by: Alex Elder --- fs/xfs/xfs_mount.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/xfs/xfs_mount.c b/fs/xfs/xfs_mount.c index 2d811e57b332..ef1233b09683 100644 --- a/fs/xfs/xfs_mount.c +++ b/fs/xfs/xfs_mount.c @@ -1254,7 +1254,7 @@ xfs_mountfs( * Allocate and initialize the per-ag data. */ spin_lock_init(&mp->m_perag_lock); - INIT_RADIX_TREE(&mp->m_perag_tree, GFP_NOFS); + INIT_RADIX_TREE(&mp->m_perag_tree, GFP_ATOMIC); error = xfs_initialize_perag(mp, sbp->sb_agcount, &mp->m_maxagi); if (error) { cmn_err(CE_WARN, "XFS: Failed per-ag init: %d", error); -- cgit v1.2.2 From fb3b504adeee942e55393396fea8fdf406acf037 Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Fri, 28 May 2010 19:03:10 +0000 Subject: xfs: fix access to upper inodes without inode64 If a filesystem is mounted without the inode64 mount option we should still be able to access inodes not fitting into 32 bits, just not created new ones. For this to work we need to make sure the inode cache radix tree is initialized for all allocation groups, not just those we plan to allocate inodes from. This patch makes sure we initialize the inode cache radix tree for all allocation groups, and also cleans xfs_initialize_perag up a bit to separate the inode32 logical from the general perag structure setup. Signed-off-by: Christoph Hellwig Signed-off-by: Alex Elder --- fs/xfs/linux-2.6/xfs_sync.c | 9 -------- fs/xfs/xfs_ag.h | 1 - fs/xfs/xfs_iget.c | 3 --- fs/xfs/xfs_inode.c | 2 -- fs/xfs/xfs_mount.c | 52 ++++++++++++++++++--------------------------- 5 files changed, 21 insertions(+), 46 deletions(-) (limited to 'fs') diff --git a/fs/xfs/linux-2.6/xfs_sync.c b/fs/xfs/linux-2.6/xfs_sync.c index 3884e20bc14e..ef7f0218bccb 100644 --- a/fs/xfs/linux-2.6/xfs_sync.c +++ b/fs/xfs/linux-2.6/xfs_sync.c @@ -164,10 +164,6 @@ xfs_inode_ag_iterator( struct xfs_perag *pag; pag = xfs_perag_get(mp, ag); - if (!pag->pag_ici_init) { - xfs_perag_put(pag); - continue; - } error = xfs_inode_ag_walk(mp, pag, execute, flags, tag, exclusive, &nr); xfs_perag_put(pag); @@ -867,12 +863,7 @@ xfs_reclaim_inode_shrink( down_read(&xfs_mount_list_lock); list_for_each_entry(mp, &xfs_mount_list, m_mplist) { for (ag = 0; ag < mp->m_sb.sb_agcount; ag++) { - pag = xfs_perag_get(mp, ag); - if (!pag->pag_ici_init) { - xfs_perag_put(pag); - continue; - } reclaimable += pag->pag_ici_reclaimable; xfs_perag_put(pag); } diff --git a/fs/xfs/xfs_ag.h b/fs/xfs/xfs_ag.h index 401f364ad36c..4917d4eed4ed 100644 --- a/fs/xfs/xfs_ag.h +++ b/fs/xfs/xfs_ag.h @@ -227,7 +227,6 @@ typedef struct xfs_perag { atomic_t pagf_fstrms; /* # of filestreams active in this AG */ - int pag_ici_init; /* incore inode cache initialised */ rwlock_t pag_ici_lock; /* incore inode lock */ struct radix_tree_root pag_ici_root; /* incore inode cache root */ int pag_ici_reclaimable; /* reclaimable inodes */ diff --git a/fs/xfs/xfs_iget.c b/fs/xfs/xfs_iget.c index 6845db90818f..f44a367907a8 100644 --- a/fs/xfs/xfs_iget.c +++ b/fs/xfs/xfs_iget.c @@ -382,9 +382,6 @@ xfs_iget( /* get the perag structure and ensure that it's inode capable */ pag = xfs_perag_get(mp, XFS_INO_TO_AGNO(mp, ino)); - if (!pag->pagi_inodeok) - return EINVAL; - ASSERT(pag->pag_ici_init); agino = XFS_INO_TO_AGINO(mp, ino); again: diff --git a/fs/xfs/xfs_inode.c b/fs/xfs/xfs_inode.c index 8cd6e8d8fe9c..6ba7765026f6 100644 --- a/fs/xfs/xfs_inode.c +++ b/fs/xfs/xfs_inode.c @@ -2649,8 +2649,6 @@ xfs_iflush_cluster( int i; pag = xfs_perag_get(mp, XFS_INO_TO_AGNO(mp, ip->i_ino)); - ASSERT(pag->pagi_inodeok); - ASSERT(pag->pag_ici_init); inodes_per_cluster = XFS_INODE_CLUSTER_SIZE(mp) >> mp->m_sb.sb_inodelog; ilist_size = inodes_per_cluster * sizeof(xfs_inode_t *); diff --git a/fs/xfs/xfs_mount.c b/fs/xfs/xfs_mount.c index ef1233b09683..d59f4e8bedcf 100644 --- a/fs/xfs/xfs_mount.c +++ b/fs/xfs/xfs_mount.c @@ -413,17 +413,6 @@ xfs_mount_validate_sb( return 0; } -STATIC void -xfs_initialize_perag_icache( - xfs_perag_t *pag) -{ - if (!pag->pag_ici_init) { - rwlock_init(&pag->pag_ici_lock); - INIT_RADIX_TREE(&pag->pag_ici_root, GFP_ATOMIC); - pag->pag_ici_init = 1; - } -} - int xfs_initialize_perag( xfs_mount_t *mp, @@ -436,13 +425,8 @@ xfs_initialize_perag( xfs_agino_t agino; xfs_ino_t ino; xfs_sb_t *sbp = &mp->m_sb; - xfs_ino_t max_inum = XFS_MAXINUMBER_32; int error = -ENOMEM; - /* Check to see if the filesystem can overflow 32 bit inodes */ - agino = XFS_OFFBNO_TO_AGINO(mp, sbp->sb_agblocks - 1, 0); - ino = XFS_AGINO_TO_INO(mp, agcount - 1, agino); - /* * Walk the current per-ag tree so we don't try to initialise AGs * that already exist (growfs case). Allocate and insert all the @@ -456,11 +440,18 @@ xfs_initialize_perag( } if (!first_initialised) first_initialised = index; + pag = kmem_zalloc(sizeof(*pag), KM_MAYFAIL); if (!pag) goto out_unwind; + pag->pag_agno = index; + pag->pag_mount = mp; + rwlock_init(&pag->pag_ici_lock); + INIT_RADIX_TREE(&pag->pag_ici_root, GFP_ATOMIC); + if (radix_tree_preload(GFP_NOFS)) goto out_unwind; + spin_lock(&mp->m_perag_lock); if (radix_tree_insert(&mp->m_perag_tree, index, pag)) { BUG(); @@ -469,25 +460,26 @@ xfs_initialize_perag( error = -EEXIST; goto out_unwind; } - pag->pag_agno = index; - pag->pag_mount = mp; spin_unlock(&mp->m_perag_lock); radix_tree_preload_end(); } - /* Clear the mount flag if no inode can overflow 32 bits - * on this filesystem, or if specifically requested.. + /* + * If we mount with the inode64 option, or no inode overflows + * the legacy 32-bit address space clear the inode32 option. */ - if ((mp->m_flags & XFS_MOUNT_SMALL_INUMS) && ino > max_inum) { + agino = XFS_OFFBNO_TO_AGINO(mp, sbp->sb_agblocks - 1, 0); + ino = XFS_AGINO_TO_INO(mp, agcount - 1, agino); + + if ((mp->m_flags & XFS_MOUNT_SMALL_INUMS) && ino > XFS_MAXINUMBER_32) mp->m_flags |= XFS_MOUNT_32BITINODES; - } else { + else mp->m_flags &= ~XFS_MOUNT_32BITINODES; - } - /* If we can overflow then setup the ag headers accordingly */ if (mp->m_flags & XFS_MOUNT_32BITINODES) { - /* Calculate how much should be reserved for inodes to - * meet the max inode percentage. + /* + * Calculate how much should be reserved for inodes to meet + * the max inode percentage. */ if (mp->m_maxicount) { __uint64_t icount; @@ -500,30 +492,28 @@ xfs_initialize_perag( } else { max_metadata = agcount; } + for (index = 0; index < agcount; index++) { ino = XFS_AGINO_TO_INO(mp, index, agino); - if (ino > max_inum) { + if (ino > XFS_MAXINUMBER_32) { index++; break; } - /* This ag is preferred for inodes */ pag = xfs_perag_get(mp, index); pag->pagi_inodeok = 1; if (index < max_metadata) pag->pagf_metadata = 1; - xfs_initialize_perag_icache(pag); xfs_perag_put(pag); } } else { - /* Setup default behavior for smaller filesystems */ for (index = 0; index < agcount; index++) { pag = xfs_perag_get(mp, index); pag->pagi_inodeok = 1; - xfs_initialize_perag_icache(pag); xfs_perag_put(pag); } } + if (maxagi) *maxagi = index; return 0; -- cgit v1.2.2 From 84cb0999851e25bc4bd4aaa717cc8f8acbf42b2a Mon Sep 17 00:00:00 2001 From: Ryusuke Konishi Date: Sat, 22 May 2010 12:49:32 +0900 Subject: nilfs2: fix style issue in nilfs_destroy_cachep This gets rid of unwanted space chars in front of conditional sentences of nilfs_destroy_cachep(). Signed-off-by: Ryusuke Konishi --- fs/nilfs2/super.c | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) (limited to 'fs') diff --git a/fs/nilfs2/super.c b/fs/nilfs2/super.c index 03b34b738993..414ef68931cf 100644 --- a/fs/nilfs2/super.c +++ b/fs/nilfs2/super.c @@ -1130,13 +1130,13 @@ static void nilfs_segbuf_init_once(void *obj) static void nilfs_destroy_cachep(void) { - if (nilfs_inode_cachep) + if (nilfs_inode_cachep) kmem_cache_destroy(nilfs_inode_cachep); - if (nilfs_transaction_cachep) + if (nilfs_transaction_cachep) kmem_cache_destroy(nilfs_transaction_cachep); - if (nilfs_segbuf_cachep) + if (nilfs_segbuf_cachep) kmem_cache_destroy(nilfs_segbuf_cachep); - if (nilfs_btree_path_cache) + if (nilfs_btree_path_cache) kmem_cache_destroy(nilfs_btree_path_cache); } -- cgit v1.2.2 From c29684d6834af7b3792f2feb6bdcf8c906ad8db6 Mon Sep 17 00:00:00 2001 From: Ryusuke Konishi Date: Sun, 23 May 2010 20:40:32 +0900 Subject: nilfs2: remove obsolete declarations of cache constructor and destructor The commit 41c88bd7 ("nilfs2: cleanup multi kmem_cache_{create,destroy} code") consolidated slab constructors and destructors used in nilfs, but it left some declarations in header files. This gets rid of the obsolete declarations. Signed-off-by: Ryusuke Konishi --- fs/nilfs2/btree.h | 2 -- fs/nilfs2/segbuf.h | 2 -- fs/nilfs2/segment.h | 2 -- 3 files changed, 6 deletions(-) (limited to 'fs') diff --git a/fs/nilfs2/btree.h b/fs/nilfs2/btree.h index af638d59e3bf..43c8c5b541fd 100644 --- a/fs/nilfs2/btree.h +++ b/fs/nilfs2/btree.h @@ -75,8 +75,6 @@ struct nilfs_btree_path { extern struct kmem_cache *nilfs_btree_path_cache; -int nilfs_btree_path_cache_init(void); -void nilfs_btree_path_cache_destroy(void); int nilfs_btree_init(struct nilfs_bmap *); int nilfs_btree_convert_and_insert(struct nilfs_bmap *, __u64, __u64, const __u64 *, const __u64 *, int); diff --git a/fs/nilfs2/segbuf.h b/fs/nilfs2/segbuf.h index fdf1c3b6d673..85fbb66455e2 100644 --- a/fs/nilfs2/segbuf.h +++ b/fs/nilfs2/segbuf.h @@ -127,8 +127,6 @@ struct nilfs_segment_buffer { extern struct kmem_cache *nilfs_segbuf_cachep; -int __init nilfs_init_segbuf_cache(void); -void nilfs_destroy_segbuf_cache(void); struct nilfs_segment_buffer *nilfs_segbuf_new(struct super_block *); void nilfs_segbuf_free(struct nilfs_segment_buffer *); void nilfs_segbuf_map(struct nilfs_segment_buffer *, __u64, unsigned long, diff --git a/fs/nilfs2/segment.h b/fs/nilfs2/segment.h index dca142361ccf..01e20dbb217d 100644 --- a/fs/nilfs2/segment.h +++ b/fs/nilfs2/segment.h @@ -221,8 +221,6 @@ enum { extern struct kmem_cache *nilfs_transaction_cachep; /* segment.c */ -extern int nilfs_init_transaction_cache(void); -extern void nilfs_destroy_transaction_cache(void); extern void nilfs_relax_pressure_in_lock(struct super_block *); extern int nilfs_construct_segment(struct super_block *); -- cgit v1.2.2 From b160fdabe93a8a53094f90f02bf4dcb500782aab Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Tue, 1 Jun 2010 21:59:18 +0200 Subject: nfsd: nfsd_setattr needs to call commit_metadata The conversion of write_inode_now calls to commit_metadata in commit f501912a35c02eadc55ca9396ece55fe36f785d0 missed out the call in nfsd_setattr. But without this conversion we can't guarantee that a SETATTR request has actually been commited to disk with XFS, which causes a regression from 2.6.32 (only for NFSv2, but anyway). Signed-off-by: Christoph Hellwig Cc: stable@kernel.org Signed-off-by: J. Bruce Fields --- fs/nfsd/vfs.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/vfs.c b/fs/nfsd/vfs.c index 6dd5f1970e01..3440dd8a4fb3 100644 --- a/fs/nfsd/vfs.c +++ b/fs/nfsd/vfs.c @@ -443,8 +443,7 @@ nfsd_setattr(struct svc_rqst *rqstp, struct svc_fh *fhp, struct iattr *iap, if (size_change) put_write_access(inode); if (!err) - if (EX_ISSYNC(fhp->fh_export)) - write_inode_now(inode, 1); + commit_metadata(fhp); out: return err; -- cgit v1.2.2 From 13a4214cd9ec14d7b77e98bd3ee51f60f868a6e5 Mon Sep 17 00:00:00 2001 From: Henry C Chang Date: Tue, 1 Jun 2010 11:31:08 -0700 Subject: ceph: fix d_subdirs ordering problem We misused list_move_tail() to order the dentry in d_subdirs. This will screw up the d_subdirs order. This bug can be reliably reproduced by: 1. mount ceph fs. 2. on ceph fs, git clone git://ceph.newdream.net/git/ceph.git 3. Run autogen.sh in ceph directory. (Note: Errors only occur at the first time you run autogen.sh.) Signed-off-by: Henry C Chang Signed-off-by: Sage Weil --- fs/ceph/inode.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/ceph/inode.c b/fs/ceph/inode.c index 226f5a50d362..ab47f46ca282 100644 --- a/fs/ceph/inode.c +++ b/fs/ceph/inode.c @@ -827,7 +827,7 @@ static void ceph_set_dentry_offset(struct dentry *dn) spin_lock(&dcache_lock); spin_lock(&dn->d_lock); - list_move_tail(&dir->d_subdirs, &dn->d_u.d_child); + list_move(&dn->d_u.d_child, &dir->d_subdirs); dout("set_dentry_offset %p %lld (%p %p)\n", dn, di->offset, dn->d_u.d_child.prev, dn->d_u.d_child.next); spin_unlock(&dn->d_lock); -- cgit v1.2.2 From 205475679a74fe40b63a1c7f41110fdb64daa8b9 Mon Sep 17 00:00:00 2001 From: Yehuda Sadeh Date: Tue, 1 Jun 2010 10:37:40 -0700 Subject: ceph: fix memory leak in statfs Freeing the statfs request structure when required. Signed-off-by: Yehuda Sadeh Signed-off-by: Sage Weil --- fs/ceph/mon_client.c | 2 ++ 1 file changed, 2 insertions(+) (limited to 'fs') diff --git a/fs/ceph/mon_client.c b/fs/ceph/mon_client.c index 21c62e9b7d1d..07a539906e67 100644 --- a/fs/ceph/mon_client.c +++ b/fs/ceph/mon_client.c @@ -400,6 +400,8 @@ static void release_generic_request(struct kref *kref) ceph_msg_put(req->reply); if (req->request) ceph_msg_put(req->request); + + kfree(req); } static void put_generic_request(struct ceph_mon_generic_request *req) -- cgit v1.2.2 From 558d3499bd059d4534b1f2b69dc1c562acc733fe Mon Sep 17 00:00:00 2001 From: Sage Weil Date: Tue, 1 Jun 2010 12:51:12 -0700 Subject: ceph: fix f_namelen reported by statfs We were setting f_namelen in kstatfs to PATH_MAX instead of NAME_MAX. That disagrees with ceph_lookup behavior (which checks against NAME_MAX), and also makes the pjd posix test suite spit out ugly errors because with can't clean up its temporary files. Signed-off-by: Sage Weil --- fs/ceph/super.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/ceph/super.c b/fs/ceph/super.c index 4e0bee240b9d..8db88792b5ad 100644 --- a/fs/ceph/super.c +++ b/fs/ceph/super.c @@ -89,7 +89,7 @@ static int ceph_statfs(struct dentry *dentry, struct kstatfs *buf) buf->f_files = le64_to_cpu(st.num_objects); buf->f_ffree = -1; - buf->f_namelen = PATH_MAX; + buf->f_namelen = NAME_MAX; buf->f_frsize = PAGE_CACHE_SIZE; /* leave fsid little-endian, regardless of host endianness */ -- cgit v1.2.2 From 1f5a81e41f8b1a782c68d3843e9ec1bfaadf7d72 Mon Sep 17 00:00:00 2001 From: Theodore Ts'o Date: Wed, 2 Jun 2010 22:04:39 -0400 Subject: ext4: Make sure the MOVE_EXT ioctl can't overwrite append-only files Dan Roseberg has reported a problem with the MOVE_EXT ioctl. If the donor file is an append-only file, we should not allow the operation to proceed, lest we end up overwriting the contents of an append-only file. Signed-off-by: "Theodore Ts'o" Cc: Dan Rosenberg --- fs/ext4/move_extent.c | 3 +++ 1 file changed, 3 insertions(+) (limited to 'fs') diff --git a/fs/ext4/move_extent.c b/fs/ext4/move_extent.c index 3a6c92ac131c..52abfa12762a 100644 --- a/fs/ext4/move_extent.c +++ b/fs/ext4/move_extent.c @@ -960,6 +960,9 @@ mext_check_arguments(struct inode *orig_inode, return -EINVAL; } + if (IS_IMMUTABLE(donor_inode) || IS_APPEND(donor_inode)) + return -EPERM; + /* Ext4 move extent does not support swapfile */ if (IS_SWAPFILE(orig_inode) || IS_SWAPFILE(donor_inode)) { ext4_debug("ext4 move extent: The argument files should " -- cgit v1.2.2 From 5b257b4a1f9239624c6b5e669763de04e482c2b3 Mon Sep 17 00:00:00 2001 From: Dave Chinner Date: Thu, 3 Jun 2010 16:22:29 +1000 Subject: xfs: fix race in inode cluster freeing failing to stale inodes When an inode cluster is freed, it needs to mark all inodes in memory as XFS_ISTALE before marking the buffer as stale. This is eeded because the inodes have a different life cycle to the buffer, and once the buffer is torn down during transaction completion, we must ensure none of the inodes get written back (which is what XFS_ISTALE does). Unfortunately, xfs_ifree_cluster() has some bugs that lead to inodes not being marked with XFS_ISTALE. This shows up when xfs_iflush() is called on these inodes either during inode reclaim or tail pushing on the AIL. The buffer is read back, but no longer contains inodes and so triggers assert failures and shutdowns. This was reproducable with at run.dbench10 invocation from xfstests. There are two main causes of xfs_ifree_cluster() failing. The first is simple - it checks in-memory inodes it finds in the per-ag icache to see if they are clean without holding the flush lock. if they are clean it skips them completely. However, If an inode is flushed delwri, it will appear clean, but is not guaranteed to be written back until the flush lock has been dropped. Hence we may have raced on the clean check and the inode may actually be dirty. Hence always mark inodes found in memory stale before we check properly if they are clean. The second is more complex, and makes the first problem easier to hit. Basically the in-memory inode scan is done with full knowledge it can be racing with inode flushing and AIl tail pushing, which means that inodes that it can't get the flush lock on might not be attached to the buffer after then in-memory inode scan due to IO completion occurring. This is actually documented in the code as "needs better interlocking". i.e. this is a zero-day bug. Effectively, the in-memory scan must be done while the inode buffer is locked and Io cannot be issued on it while we do the in-memory inode scan. This ensures that inodes we couldn't get the flush lock on are guaranteed to be attached to the cluster buffer, so we can then catch all in-memory inodes and mark them stale. Now that the inode cluster buffer is locked before the in-memory scan is done, there is no need for the two-phase update of the in-memory inodes, so simplify the code into two loops and remove the allocation of the temporary buffer used to hold locked inodes across the phases. Signed-off-by: Dave Chinner Reviewed-by: Christoph Hellwig --- fs/xfs/xfs_inode.c | 142 +++++++++++++++++++++++------------------------------ 1 file changed, 62 insertions(+), 80 deletions(-) (limited to 'fs') diff --git a/fs/xfs/xfs_inode.c b/fs/xfs/xfs_inode.c index 6ba7765026f6..d53c39de7d05 100644 --- a/fs/xfs/xfs_inode.c +++ b/fs/xfs/xfs_inode.c @@ -1940,10 +1940,10 @@ xfs_ifree_cluster( int blks_per_cluster; int nbufs; int ninodes; - int i, j, found, pre_flushed; + int i, j; xfs_daddr_t blkno; xfs_buf_t *bp; - xfs_inode_t *ip, **ip_found; + xfs_inode_t *ip; xfs_inode_log_item_t *iip; xfs_log_item_t *lip; struct xfs_perag *pag; @@ -1960,114 +1960,97 @@ xfs_ifree_cluster( nbufs = XFS_IALLOC_BLOCKS(mp) / blks_per_cluster; } - ip_found = kmem_alloc(ninodes * sizeof(xfs_inode_t *), KM_NOFS); - for (j = 0; j < nbufs; j++, inum += ninodes) { + int found = 0; + blkno = XFS_AGB_TO_DADDR(mp, XFS_INO_TO_AGNO(mp, inum), XFS_INO_TO_AGBNO(mp, inum)); + /* + * We obtain and lock the backing buffer first in the process + * here, as we have to ensure that any dirty inode that we + * can't get the flush lock on is attached to the buffer. + * If we scan the in-memory inodes first, then buffer IO can + * complete before we get a lock on it, and hence we may fail + * to mark all the active inodes on the buffer stale. + */ + bp = xfs_trans_get_buf(tp, mp->m_ddev_targp, blkno, + mp->m_bsize * blks_per_cluster, + XBF_LOCK); + + /* + * Walk the inodes already attached to the buffer and mark them + * stale. These will all have the flush locks held, so an + * in-memory inode walk can't lock them. + */ + lip = XFS_BUF_FSPRIVATE(bp, xfs_log_item_t *); + while (lip) { + if (lip->li_type == XFS_LI_INODE) { + iip = (xfs_inode_log_item_t *)lip; + ASSERT(iip->ili_logged == 1); + lip->li_cb = (void(*)(xfs_buf_t*,xfs_log_item_t*)) xfs_istale_done; + xfs_trans_ail_copy_lsn(mp->m_ail, + &iip->ili_flush_lsn, + &iip->ili_item.li_lsn); + xfs_iflags_set(iip->ili_inode, XFS_ISTALE); + found++; + } + lip = lip->li_bio_list; + } /* - * Look for each inode in memory and attempt to lock it, - * we can be racing with flush and tail pushing here. - * any inode we get the locks on, add to an array of - * inode items to process later. + * For each inode in memory attempt to add it to the inode + * buffer and set it up for being staled on buffer IO + * completion. This is safe as we've locked out tail pushing + * and flushing by locking the buffer. * - * The get the buffer lock, we could beat a flush - * or tail pushing thread to the lock here, in which - * case they will go looking for the inode buffer - * and fail, we need some other form of interlock - * here. + * We have already marked every inode that was part of a + * transaction stale above, which means there is no point in + * even trying to lock them. */ - found = 0; for (i = 0; i < ninodes; i++) { read_lock(&pag->pag_ici_lock); ip = radix_tree_lookup(&pag->pag_ici_root, XFS_INO_TO_AGINO(mp, (inum + i))); - /* Inode not in memory or we found it already, - * nothing to do - */ + /* Inode not in memory or stale, nothing to do */ if (!ip || xfs_iflags_test(ip, XFS_ISTALE)) { read_unlock(&pag->pag_ici_lock); continue; } - if (xfs_inode_clean(ip)) { - read_unlock(&pag->pag_ici_lock); - continue; - } - - /* If we can get the locks then add it to the - * list, otherwise by the time we get the bp lock - * below it will already be attached to the - * inode buffer. - */ - - /* This inode will already be locked - by us, lets - * keep it that way. - */ - - if (ip == free_ip) { - if (xfs_iflock_nowait(ip)) { - xfs_iflags_set(ip, XFS_ISTALE); - if (xfs_inode_clean(ip)) { - xfs_ifunlock(ip); - } else { - ip_found[found++] = ip; - } - } + /* don't try to lock/unlock the current inode */ + if (ip != free_ip && + !xfs_ilock_nowait(ip, XFS_ILOCK_EXCL)) { read_unlock(&pag->pag_ici_lock); continue; } + read_unlock(&pag->pag_ici_lock); - if (xfs_ilock_nowait(ip, XFS_ILOCK_EXCL)) { - if (xfs_iflock_nowait(ip)) { - xfs_iflags_set(ip, XFS_ISTALE); - - if (xfs_inode_clean(ip)) { - xfs_ifunlock(ip); - xfs_iunlock(ip, XFS_ILOCK_EXCL); - } else { - ip_found[found++] = ip; - } - } else { + if (!xfs_iflock_nowait(ip)) { + if (ip != free_ip) xfs_iunlock(ip, XFS_ILOCK_EXCL); - } + continue; } - read_unlock(&pag->pag_ici_lock); - } - bp = xfs_trans_get_buf(tp, mp->m_ddev_targp, blkno, - mp->m_bsize * blks_per_cluster, - XBF_LOCK); - - pre_flushed = 0; - lip = XFS_BUF_FSPRIVATE(bp, xfs_log_item_t *); - while (lip) { - if (lip->li_type == XFS_LI_INODE) { - iip = (xfs_inode_log_item_t *)lip; - ASSERT(iip->ili_logged == 1); - lip->li_cb = (void(*)(xfs_buf_t*,xfs_log_item_t*)) xfs_istale_done; - xfs_trans_ail_copy_lsn(mp->m_ail, - &iip->ili_flush_lsn, - &iip->ili_item.li_lsn); - xfs_iflags_set(iip->ili_inode, XFS_ISTALE); - pre_flushed++; + xfs_iflags_set(ip, XFS_ISTALE); + if (xfs_inode_clean(ip)) { + ASSERT(ip != free_ip); + xfs_ifunlock(ip); + xfs_iunlock(ip, XFS_ILOCK_EXCL); + continue; } - lip = lip->li_bio_list; - } - for (i = 0; i < found; i++) { - ip = ip_found[i]; iip = ip->i_itemp; - if (!iip) { + /* inode with unlogged changes only */ + ASSERT(ip != free_ip); ip->i_update_core = 0; xfs_ifunlock(ip); xfs_iunlock(ip, XFS_ILOCK_EXCL); continue; } + found++; iip->ili_last_fields = iip->ili_format.ilf_fields; iip->ili_format.ilf_fields = 0; @@ -2078,17 +2061,16 @@ xfs_ifree_cluster( xfs_buf_attach_iodone(bp, (void(*)(xfs_buf_t*,xfs_log_item_t*)) xfs_istale_done, (xfs_log_item_t *)iip); - if (ip != free_ip) { + + if (ip != free_ip) xfs_iunlock(ip, XFS_ILOCK_EXCL); - } } - if (found || pre_flushed) + if (found) xfs_trans_stale_inode_buf(tp, bp); xfs_trans_binval(tp, bp); } - kmem_free(ip_found); xfs_perag_put(pag); } -- cgit v1.2.2 From 070ecdca54dde9577d2697088e74e45568f48efb Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Thu, 3 Jun 2010 16:22:29 +1000 Subject: xfs: skip writeback from reclaim context Allowing writeback from reclaim context causes massive problems with stack overflows as we can call into the writeback code which tends to be a heavy stack user both in the generic code and XFS from random contexts that perform memory allocations. Follow the example of btrfs (and in slightly different form ext4) and refuse to write out data from reclaim context. This issue should really be handled by the VM so that we can tune better for this case, but until we get it sorted out there we have to hack around this in each filesystem with a complex writeback path. Signed-off-by: Christoph Hellwig Reviewed-by: Dave Chinner --- fs/xfs/linux-2.6/xfs_aops.c | 15 +++++++++++++++ 1 file changed, 15 insertions(+) (limited to 'fs') diff --git a/fs/xfs/linux-2.6/xfs_aops.c b/fs/xfs/linux-2.6/xfs_aops.c index 089eaca860b4..a0fa3bf0d1bb 100644 --- a/fs/xfs/linux-2.6/xfs_aops.c +++ b/fs/xfs/linux-2.6/xfs_aops.c @@ -1332,6 +1332,21 @@ xfs_vm_writepage( trace_xfs_writepage(inode, page, 0); + /* + * Refuse to write the page out if we are called from reclaim context. + * + * This is primarily to avoid stack overflows when called from deep + * used stacks in random callers for direct reclaim, but disabling + * reclaim for kswap is a nice side-effect as kswapd causes rather + * suboptimal I/O patters, too. + * + * This should really be done by the core VM, but until that happens + * filesystems like XFS, btrfs and ext4 have to take care of this + * by themselves. + */ + if (current->flags & PF_MEMALLOC) + goto out_fail; + /* * We need a transaction if: * 1. There are delalloc buffers on the page -- cgit v1.2.2 From f9369729496a0f4c607a4cc1ea4dfeddbbfc505a Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Thu, 3 Jun 2010 16:22:29 +1000 Subject: xfs: improve xfs_isilocked Use rwsem_is_locked to make the assertations for shared locks work. Signed-off-by: Christoph Hellwig Reviewed-by: Dave Chinner --- fs/xfs/xfs_iget.c | 26 ++++++++++---------------- 1 file changed, 10 insertions(+), 16 deletions(-) (limited to 'fs') diff --git a/fs/xfs/xfs_iget.c b/fs/xfs/xfs_iget.c index f44a367907a8..75df75f43d48 100644 --- a/fs/xfs/xfs_iget.c +++ b/fs/xfs/xfs_iget.c @@ -741,30 +741,24 @@ xfs_ilock_demote( } #ifdef DEBUG -/* - * Debug-only routine, without additional rw_semaphore APIs, we can - * now only answer requests regarding whether we hold the lock for write - * (reader state is outside our visibility, we only track writer state). - * - * Note: this means !xfs_isilocked would give false positives, so don't do that. - */ int xfs_isilocked( xfs_inode_t *ip, uint lock_flags) { - if ((lock_flags & (XFS_ILOCK_EXCL|XFS_ILOCK_SHARED)) == - XFS_ILOCK_EXCL) { - if (!ip->i_lock.mr_writer) - return 0; + if (lock_flags & (XFS_ILOCK_EXCL|XFS_ILOCK_SHARED)) { + if (!(lock_flags & XFS_ILOCK_SHARED)) + return !!ip->i_lock.mr_writer; + return rwsem_is_locked(&ip->i_lock.mr_lock); } - if ((lock_flags & (XFS_IOLOCK_EXCL|XFS_IOLOCK_SHARED)) == - XFS_IOLOCK_EXCL) { - if (!ip->i_iolock.mr_writer) - return 0; + if (lock_flags & (XFS_IOLOCK_EXCL|XFS_IOLOCK_SHARED)) { + if (!(lock_flags & XFS_IOLOCK_SHARED)) + return !!ip->i_iolock.mr_writer; + return rwsem_is_locked(&ip->i_iolock.mr_lock); } - return 1; + ASSERT(0); + return 0; } #endif -- cgit v1.2.2 From f324e4cb2cadd9a42932c8a158e761ae31b88e72 Mon Sep 17 00:00:00 2001 From: David Woodhouse Date: Thu, 3 Jun 2010 08:03:39 +0100 Subject: jffs2: Fix in-core inode leaks on error paths Pointed out by Al Viro. Signed-off-by: David Woodhouse --- fs/jffs2/dir.c | 115 ++++++++++++++++++++++++++++----------------------------- 1 file changed, 56 insertions(+), 59 deletions(-) (limited to 'fs') diff --git a/fs/jffs2/dir.c b/fs/jffs2/dir.c index 7aa4417e085f..cb7ef34d384c 100644 --- a/fs/jffs2/dir.c +++ b/fs/jffs2/dir.c @@ -360,8 +360,8 @@ static int jffs2_symlink (struct inode *dir_i, struct dentry *dentry, const char /* Eeek. Wave bye bye */ mutex_unlock(&f->sem); jffs2_complete_reservation(c); - jffs2_clear_inode(inode); - return PTR_ERR(fn); + ret = PTR_ERR(fn); + goto fail; } /* We use f->target field to store the target path. */ @@ -370,8 +370,8 @@ static int jffs2_symlink (struct inode *dir_i, struct dentry *dentry, const char printk(KERN_WARNING "Can't allocate %d bytes of memory\n", targetlen + 1); mutex_unlock(&f->sem); jffs2_complete_reservation(c); - jffs2_clear_inode(inode); - return -ENOMEM; + ret = -ENOMEM; + goto fail; } memcpy(f->target, target, targetlen + 1); @@ -386,30 +386,24 @@ static int jffs2_symlink (struct inode *dir_i, struct dentry *dentry, const char jffs2_complete_reservation(c); ret = jffs2_init_security(inode, dir_i); - if (ret) { - jffs2_clear_inode(inode); - return ret; - } + if (ret) + goto fail; + ret = jffs2_init_acl_post(inode); - if (ret) { - jffs2_clear_inode(inode); - return ret; - } + if (ret) + goto fail; ret = jffs2_reserve_space(c, sizeof(*rd)+namelen, &alloclen, ALLOC_NORMAL, JFFS2_SUMMARY_DIRENT_SIZE(namelen)); - if (ret) { - /* Eep. */ - jffs2_clear_inode(inode); - return ret; - } + if (ret) + goto fail; rd = jffs2_alloc_raw_dirent(); if (!rd) { /* Argh. Now we treat it like a normal delete */ jffs2_complete_reservation(c); - jffs2_clear_inode(inode); - return -ENOMEM; + ret = -ENOMEM; + goto fail; } dir_f = JFFS2_INODE_INFO(dir_i); @@ -437,8 +431,8 @@ static int jffs2_symlink (struct inode *dir_i, struct dentry *dentry, const char jffs2_complete_reservation(c); jffs2_free_raw_dirent(rd); mutex_unlock(&dir_f->sem); - jffs2_clear_inode(inode); - return PTR_ERR(fd); + ret = PTR_ERR(fd); + goto fail; } dir_i->i_mtime = dir_i->i_ctime = ITIME(je32_to_cpu(rd->mctime)); @@ -454,6 +448,11 @@ static int jffs2_symlink (struct inode *dir_i, struct dentry *dentry, const char d_instantiate(dentry, inode); return 0; + + fail: + make_bad_inode(inode); + iput(inode); + return ret; } @@ -519,8 +518,8 @@ static int jffs2_mkdir (struct inode *dir_i, struct dentry *dentry, int mode) /* Eeek. Wave bye bye */ mutex_unlock(&f->sem); jffs2_complete_reservation(c); - jffs2_clear_inode(inode); - return PTR_ERR(fn); + ret = PTR_ERR(fn); + goto fail; } /* No data here. Only a metadata node, which will be obsoleted by the first data write @@ -531,30 +530,24 @@ static int jffs2_mkdir (struct inode *dir_i, struct dentry *dentry, int mode) jffs2_complete_reservation(c); ret = jffs2_init_security(inode, dir_i); - if (ret) { - jffs2_clear_inode(inode); - return ret; - } + if (ret) + goto fail; + ret = jffs2_init_acl_post(inode); - if (ret) { - jffs2_clear_inode(inode); - return ret; - } + if (ret) + goto fail; ret = jffs2_reserve_space(c, sizeof(*rd)+namelen, &alloclen, ALLOC_NORMAL, JFFS2_SUMMARY_DIRENT_SIZE(namelen)); - if (ret) { - /* Eep. */ - jffs2_clear_inode(inode); - return ret; - } + if (ret) + goto fail; rd = jffs2_alloc_raw_dirent(); if (!rd) { /* Argh. Now we treat it like a normal delete */ jffs2_complete_reservation(c); - jffs2_clear_inode(inode); - return -ENOMEM; + ret = -ENOMEM; + goto fail; } dir_f = JFFS2_INODE_INFO(dir_i); @@ -582,8 +575,8 @@ static int jffs2_mkdir (struct inode *dir_i, struct dentry *dentry, int mode) jffs2_complete_reservation(c); jffs2_free_raw_dirent(rd); mutex_unlock(&dir_f->sem); - jffs2_clear_inode(inode); - return PTR_ERR(fd); + ret = PTR_ERR(fd); + goto fail; } dir_i->i_mtime = dir_i->i_ctime = ITIME(je32_to_cpu(rd->mctime)); @@ -600,6 +593,11 @@ static int jffs2_mkdir (struct inode *dir_i, struct dentry *dentry, int mode) d_instantiate(dentry, inode); return 0; + + fail: + make_bad_inode(inode); + iput(inode); + return ret; } static int jffs2_rmdir (struct inode *dir_i, struct dentry *dentry) @@ -693,8 +691,8 @@ static int jffs2_mknod (struct inode *dir_i, struct dentry *dentry, int mode, de /* Eeek. Wave bye bye */ mutex_unlock(&f->sem); jffs2_complete_reservation(c); - jffs2_clear_inode(inode); - return PTR_ERR(fn); + ret = PTR_ERR(fn); + goto fail; } /* No data here. Only a metadata node, which will be obsoleted by the first data write @@ -705,30 +703,24 @@ static int jffs2_mknod (struct inode *dir_i, struct dentry *dentry, int mode, de jffs2_complete_reservation(c); ret = jffs2_init_security(inode, dir_i); - if (ret) { - jffs2_clear_inode(inode); - return ret; - } + if (ret) + goto fail; + ret = jffs2_init_acl_post(inode); - if (ret) { - jffs2_clear_inode(inode); - return ret; - } + if (ret) + goto fail; ret = jffs2_reserve_space(c, sizeof(*rd)+namelen, &alloclen, ALLOC_NORMAL, JFFS2_SUMMARY_DIRENT_SIZE(namelen)); - if (ret) { - /* Eep. */ - jffs2_clear_inode(inode); - return ret; - } + if (ret) + goto fail; rd = jffs2_alloc_raw_dirent(); if (!rd) { /* Argh. Now we treat it like a normal delete */ jffs2_complete_reservation(c); - jffs2_clear_inode(inode); - return -ENOMEM; + ret = -ENOMEM; + goto fail; } dir_f = JFFS2_INODE_INFO(dir_i); @@ -759,8 +751,8 @@ static int jffs2_mknod (struct inode *dir_i, struct dentry *dentry, int mode, de jffs2_complete_reservation(c); jffs2_free_raw_dirent(rd); mutex_unlock(&dir_f->sem); - jffs2_clear_inode(inode); - return PTR_ERR(fd); + ret = PTR_ERR(fd); + goto fail; } dir_i->i_mtime = dir_i->i_ctime = ITIME(je32_to_cpu(rd->mctime)); @@ -777,6 +769,11 @@ static int jffs2_mknod (struct inode *dir_i, struct dentry *dentry, int mode, de d_instantiate(dentry, inode); return 0; + + fail: + make_bad_inode(inode); + iput(inode); + return ret; } static int jffs2_rename (struct inode *old_dir_i, struct dentry *old_dentry, -- cgit v1.2.2 From e72e6497e74811e01d72b4c1b7537b3aea3ee857 Mon Sep 17 00:00:00 2001 From: David Woodhouse Date: Thu, 3 Jun 2010 08:09:12 +0100 Subject: jffs2: Fix NFS race by using insert_inode_locked() New inodes need to be locked as we're creating them, so they don't get used by other things (like NFSd) before they're ready. Pointed out by Al Viro. Signed-off-by: David Woodhouse --- fs/jffs2/dir.c | 12 ++++++++++-- fs/jffs2/fs.c | 7 ++++++- 2 files changed, 16 insertions(+), 3 deletions(-) (limited to 'fs') diff --git a/fs/jffs2/dir.c b/fs/jffs2/dir.c index cb7ef34d384c..166062a68230 100644 --- a/fs/jffs2/dir.c +++ b/fs/jffs2/dir.c @@ -222,15 +222,18 @@ static int jffs2_create(struct inode *dir_i, struct dentry *dentry, int mode, dir_i->i_mtime = dir_i->i_ctime = ITIME(je32_to_cpu(ri->ctime)); jffs2_free_raw_inode(ri); - d_instantiate(dentry, inode); D1(printk(KERN_DEBUG "jffs2_create: Created ino #%lu with mode %o, nlink %d(%d). nrpages %ld\n", inode->i_ino, inode->i_mode, inode->i_nlink, f->inocache->pino_nlink, inode->i_mapping->nrpages)); + + d_instantiate(dentry, inode); + unlock_new_inode(inode); return 0; fail: make_bad_inode(inode); + unlock_new_inode(inode); iput(inode); jffs2_free_raw_inode(ri); return ret; @@ -447,10 +450,12 @@ static int jffs2_symlink (struct inode *dir_i, struct dentry *dentry, const char jffs2_complete_reservation(c); d_instantiate(dentry, inode); + unlock_new_inode(inode); return 0; fail: make_bad_inode(inode); + unlock_new_inode(inode); iput(inode); return ret; } @@ -592,10 +597,12 @@ static int jffs2_mkdir (struct inode *dir_i, struct dentry *dentry, int mode) jffs2_complete_reservation(c); d_instantiate(dentry, inode); + unlock_new_inode(inode); return 0; fail: make_bad_inode(inode); + unlock_new_inode(inode); iput(inode); return ret; } @@ -767,11 +774,12 @@ static int jffs2_mknod (struct inode *dir_i, struct dentry *dentry, int mode, de jffs2_complete_reservation(c); d_instantiate(dentry, inode); - + unlock_new_inode(inode); return 0; fail: make_bad_inode(inode); + unlock_new_inode(inode); iput(inode); return ret; } diff --git a/fs/jffs2/fs.c b/fs/jffs2/fs.c index 86e0821fc989..26037e2d6154 100644 --- a/fs/jffs2/fs.c +++ b/fs/jffs2/fs.c @@ -465,7 +465,12 @@ struct inode *jffs2_new_inode (struct inode *dir_i, int mode, struct jffs2_raw_i inode->i_blocks = 0; inode->i_size = 0; - insert_inode_hash(inode); + if (insert_inode_locked(inode) < 0) { + make_bad_inode(inode); + unlock_new_inode(inode); + iput(inode); + return ERR_PTR(-EINVAL); + } return inode; } -- cgit v1.2.2 From 1e5ea23df11c7c90c7e7268dd3a6603bfa5aadf7 Mon Sep 17 00:00:00 2001 From: Sage Weil Date: Fri, 4 Jun 2010 10:05:40 -0700 Subject: ceph: fix lease revocation when seq doesn't match If the client revokes a lease with a higher seq than what we have, keep the mds's seq, so that it honors our release. Otherwise, we can hang indefinitely. Signed-off-by: Sage Weil --- fs/ceph/mds_client.c | 12 ++++++++---- 1 file changed, 8 insertions(+), 4 deletions(-) (limited to 'fs') diff --git a/fs/ceph/mds_client.c b/fs/ceph/mds_client.c index b49f12822cbc..29b4485cf1ca 100644 --- a/fs/ceph/mds_client.c +++ b/fs/ceph/mds_client.c @@ -2433,6 +2433,7 @@ static void handle_lease(struct ceph_mds_client *mdsc, struct ceph_dentry_info *di; int mds = session->s_mds; struct ceph_mds_lease *h = msg->front.iov_base; + u32 seq; struct ceph_vino vino; int mask; struct qstr dname; @@ -2446,6 +2447,7 @@ static void handle_lease(struct ceph_mds_client *mdsc, vino.ino = le64_to_cpu(h->ino); vino.snap = CEPH_NOSNAP; mask = le16_to_cpu(h->mask); + seq = le32_to_cpu(h->seq); dname.name = (void *)h + sizeof(*h) + sizeof(u32); dname.len = msg->front.iov_len - sizeof(*h) - sizeof(u32); if (dname.len != get_unaligned_le32(h+1)) @@ -2456,8 +2458,9 @@ static void handle_lease(struct ceph_mds_client *mdsc, /* lookup inode */ inode = ceph_find_inode(sb, vino); - dout("handle_lease '%s', mask %d, ino %llx %p\n", - ceph_lease_op_name(h->action), mask, vino.ino, inode); + dout("handle_lease %s, mask %d, ino %llx %p %.*s\n", + ceph_lease_op_name(h->action), mask, vino.ino, inode, + dname.len, dname.name); if (inode == NULL) { dout("handle_lease no inode %llx\n", vino.ino); goto release; @@ -2482,7 +2485,8 @@ static void handle_lease(struct ceph_mds_client *mdsc, switch (h->action) { case CEPH_MDS_LEASE_REVOKE: if (di && di->lease_session == session) { - h->seq = cpu_to_le32(di->lease_seq); + if (ceph_seq_cmp(di->lease_seq, seq) > 0) + h->seq = cpu_to_le32(di->lease_seq); __ceph_mdsc_drop_dentry_lease(dentry); } release = 1; @@ -2496,7 +2500,7 @@ static void handle_lease(struct ceph_mds_client *mdsc, unsigned long duration = le32_to_cpu(h->duration_ms) * HZ / 1000; - di->lease_seq = le32_to_cpu(h->seq); + di->lease_seq = seq; dentry->d_time = di->lease_renew_from + duration; di->lease_renew_after = di->lease_renew_from + (duration >> 1); -- cgit v1.2.2 From 84a8dce2710cc425089a2b92acc354d4fbb5788d Mon Sep 17 00:00:00 2001 From: Dmitry Monakhov Date: Sat, 5 Jun 2010 11:51:27 -0400 Subject: ext4: Fix remaining racy updates of EXT4_I(inode)->i_flags A few functions were still modifying i_flags in a racy manner. Signed-off-by: Dmitry Monakhov Signed-off-by: "Theodore Ts'o" --- fs/ext4/inode.c | 40 +++++++++++++++++++++++----------------- 1 file changed, 23 insertions(+), 17 deletions(-) (limited to 'fs') diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c index 19df61c321fd..42272d67955a 100644 --- a/fs/ext4/inode.c +++ b/fs/ext4/inode.c @@ -4942,20 +4942,26 @@ void ext4_set_inode_flags(struct inode *inode) /* Propagate flags from i_flags to EXT4_I(inode)->i_flags */ void ext4_get_inode_flags(struct ext4_inode_info *ei) { - unsigned int flags = ei->vfs_inode.i_flags; - - ei->i_flags &= ~(EXT4_SYNC_FL|EXT4_APPEND_FL| - EXT4_IMMUTABLE_FL|EXT4_NOATIME_FL|EXT4_DIRSYNC_FL); - if (flags & S_SYNC) - ei->i_flags |= EXT4_SYNC_FL; - if (flags & S_APPEND) - ei->i_flags |= EXT4_APPEND_FL; - if (flags & S_IMMUTABLE) - ei->i_flags |= EXT4_IMMUTABLE_FL; - if (flags & S_NOATIME) - ei->i_flags |= EXT4_NOATIME_FL; - if (flags & S_DIRSYNC) - ei->i_flags |= EXT4_DIRSYNC_FL; + unsigned int vfs_fl; + unsigned long old_fl, new_fl; + + do { + vfs_fl = ei->vfs_inode.i_flags; + old_fl = ei->i_flags; + new_fl = old_fl & ~(EXT4_SYNC_FL|EXT4_APPEND_FL| + EXT4_IMMUTABLE_FL|EXT4_NOATIME_FL| + EXT4_DIRSYNC_FL); + if (vfs_fl & S_SYNC) + new_fl |= EXT4_SYNC_FL; + if (vfs_fl & S_APPEND) + new_fl |= EXT4_APPEND_FL; + if (vfs_fl & S_IMMUTABLE) + new_fl |= EXT4_IMMUTABLE_FL; + if (vfs_fl & S_NOATIME) + new_fl |= EXT4_NOATIME_FL; + if (vfs_fl & S_DIRSYNC) + new_fl |= EXT4_DIRSYNC_FL; + } while (cmpxchg(&ei->i_flags, old_fl, new_fl) != old_fl); } static blkcnt_t ext4_inode_blocks(struct ext4_inode *raw_inode, @@ -5191,7 +5197,7 @@ static int ext4_inode_blocks_set(handle_t *handle, */ raw_inode->i_blocks_lo = cpu_to_le32(i_blocks); raw_inode->i_blocks_high = 0; - ei->i_flags &= ~EXT4_HUGE_FILE_FL; + ext4_clear_inode_flag(inode, EXT4_INODE_HUGE_FILE); return 0; } if (!EXT4_HAS_RO_COMPAT_FEATURE(sb, EXT4_FEATURE_RO_COMPAT_HUGE_FILE)) @@ -5204,9 +5210,9 @@ static int ext4_inode_blocks_set(handle_t *handle, */ raw_inode->i_blocks_lo = cpu_to_le32(i_blocks); raw_inode->i_blocks_high = cpu_to_le16(i_blocks >> 32); - ei->i_flags &= ~EXT4_HUGE_FILE_FL; + ext4_clear_inode_flag(inode, EXT4_INODE_HUGE_FILE); } else { - ei->i_flags |= EXT4_HUGE_FILE_FL; + ext4_set_inode_flag(inode, EXT4_INODE_HUGE_FILE); /* i_block is stored in file system block size */ i_blocks = i_blocks >> (inode->i_blkbits - 9); raw_inode->i_blocks_lo = cpu_to_le32(i_blocks); -- cgit v1.2.2 From 1c24d06f8e065023ebb428db5af5514500839ee6 Mon Sep 17 00:00:00 2001 From: Jan Kara Date: Fri, 4 Jun 2010 17:07:55 +0200 Subject: jffs2: update ctime when changing the file's permission by setfacl jffs2 didn't update the ctime of the file when its permission was changed. Steps to reproduce: # touch aaa # stat -c %Z aaa 1275289822 # setfacl -m 'u::x,g::x,o::x' aaa # stat -c %Z aaa 1275289822 <- unchanged But, according to the spec of the ctime, jffs2 must update it. Port of ext3 patch by Miao Xie . Signed-off-by: Jan Kara Signed-off-by: David Woodhouse --- fs/jffs2/acl.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/jffs2/acl.c b/fs/jffs2/acl.c index 7cdc3196476a..64d3b4b48dbf 100644 --- a/fs/jffs2/acl.c +++ b/fs/jffs2/acl.c @@ -234,8 +234,9 @@ static int jffs2_set_acl(struct inode *inode, int type, struct posix_acl *acl) if (inode->i_mode != mode) { struct iattr attr; - attr.ia_valid = ATTR_MODE; + attr.ia_valid = ATTR_MODE | ATTR_CTIME; attr.ia_mode = mode; + attr.ia_ctime = CURRENT_TIME_SEC; rc = jffs2_do_setattr(inode, &attr); if (rc < 0) return rc; -- cgit v1.2.2 From c3935e30495869dd611e1cd62253c94ebc7c6c04 Mon Sep 17 00:00:00 2001 From: "J. Bruce Fields" Date: Fri, 4 Jun 2010 16:42:08 -0400 Subject: nfsd4: shut down callback queue outside state lock This reportedly causes a lockdep warning on nfsd shutdown. That looks like a false positive to me, but there's no reason why this needs the state lock anyway. Reported-by: Jeff Layton Signed-off-by: J. Bruce Fields --- fs/nfsd/nfs4state.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c index 12f7109720c2..4a2734758778 100644 --- a/fs/nfsd/nfs4state.c +++ b/fs/nfsd/nfs4state.c @@ -4122,8 +4122,8 @@ nfs4_state_shutdown(void) nfs4_lock_state(); nfs4_release_reclaim(); __nfs4_state_shutdown(); - nfsd4_destroy_callback_queue(); nfs4_unlock_state(); + nfsd4_destroy_callback_queue(); } /* -- cgit v1.2.2 From 254c8c2dbf0e06a560a5814eb90cb628adb2de66 Mon Sep 17 00:00:00 2001 From: Dave Chinner Date: Wed, 9 Jun 2010 10:37:19 +1000 Subject: xfs: remove nr_to_write writeback windup. Now that the background flush code has been fixed, we shouldn't need to silently multiply the wbc->nr_to_write to get good writeback. Remove that code. Signed-off-by: Dave Chinner Reviewed-by: Christoph Hellwig Signed-off-by: Linus Torvalds --- fs/xfs/linux-2.6/xfs_aops.c | 8 -------- 1 file changed, 8 deletions(-) (limited to 'fs') diff --git a/fs/xfs/linux-2.6/xfs_aops.c b/fs/xfs/linux-2.6/xfs_aops.c index a0fa3bf0d1bb..34640d6dbdcb 100644 --- a/fs/xfs/linux-2.6/xfs_aops.c +++ b/fs/xfs/linux-2.6/xfs_aops.c @@ -1381,14 +1381,6 @@ xfs_vm_writepage( if (!page_has_buffers(page)) create_empty_buffers(page, 1 << inode->i_blkbits, 0); - - /* - * VM calculation for nr_to_write seems off. Bump it way - * up, this gets simple streaming writes zippy again. - * To be reviewed again after Jens' writeback changes. - */ - wbc->nr_to_write *= 4; - /* * Convert delayed allocate, unwritten or unmapped space * to real space and flush out to disk. -- cgit v1.2.2 From cf3425707ed9ce0d5ebaba20bc3d22dd39e52f2f Mon Sep 17 00:00:00 2001 From: Nick Piggin Date: Wed, 26 May 2010 01:50:21 +1000 Subject: block: bd_start_claiming fix module refcount bd_start_claiming has an unbalanced module_put introduced in 6b4517a79. Signed-off-by: Nick Piggin Acked-by: Tejun Heo Signed-off-by: Jens Axboe --- fs/block_dev.c | 1 + 1 file changed, 1 insertion(+) (limited to 'fs') diff --git a/fs/block_dev.c b/fs/block_dev.c index 7346c96308a5..204a7632c511 100644 --- a/fs/block_dev.c +++ b/fs/block_dev.c @@ -734,6 +734,7 @@ static struct block_device *bd_start_claiming(struct block_device *bdev, return ERR_PTR(-ENXIO); whole = bdget_disk(disk, 0); + module_put(disk->fops->owner); put_disk(disk); if (!whole) return ERR_PTR(-ENOMEM); -- cgit v1.2.2 From b0018361c3f934858592cbbb5e1a4f318c2a70ed Mon Sep 17 00:00:00 2001 From: Nick Piggin Date: Wed, 26 May 2010 01:51:19 +1000 Subject: block: bd_start_claiming cleanup I don't like the subtle multi-context code in bd_claim (ie. detects where it has been called based on bd_claiming). It seems clearer to just require a new function to finish a 2-part claim. Also improve commentary in bd_start_claiming as to how it should be used. Signed-off-by: Nick Piggin Acked-by: Tejun Heo Signed-off-by: Jens Axboe --- fs/block_dev.c | 72 ++++++++++++++++++++++++++++++++++++++-------------------- 1 file changed, 48 insertions(+), 24 deletions(-) (limited to 'fs') diff --git a/fs/block_dev.c b/fs/block_dev.c index 204a7632c511..a1642ef60197 100644 --- a/fs/block_dev.c +++ b/fs/block_dev.c @@ -706,8 +706,13 @@ retry: * @bdev is about to be opened exclusively. Check @bdev can be opened * exclusively and mark that an exclusive open is in progress. Each * successful call to this function must be matched with a call to - * either bd_claim() or bd_abort_claiming(). If this function - * succeeds, the matching bd_claim() is guaranteed to succeed. + * either bd_finish_claiming() or bd_abort_claiming() (which do not + * fail). + * + * This function is used to gain exclusive access to the block device + * without actually causing other exclusive open attempts to fail. It + * should be used when the open sequence itself requires exclusive + * access but may subsequently fail. * * CONTEXT: * Might sleep. @@ -783,15 +788,47 @@ static void bd_abort_claiming(struct block_device *whole, void *holder) __bd_abort_claiming(whole, holder); /* releases bdev_lock */ } +/* increment holders when we have a legitimate claim. requires bdev_lock */ +static void __bd_claim(struct block_device *bdev, struct block_device *whole, + void *holder) +{ + /* note that for a whole device bd_holders + * will be incremented twice, and bd_holder will + * be set to bd_claim before being set to holder + */ + whole->bd_holders++; + whole->bd_holder = bd_claim; + bdev->bd_holders++; + bdev->bd_holder = holder; +} + +/** + * bd_finish_claiming - finish claiming a block device + * @bdev: block device of interest (passed to bd_start_claiming()) + * @whole: whole block device returned by bd_start_claiming() + * @holder: holder trying to claim @bdev + * + * Finish a claiming block started by bd_start_claiming(). + * + * CONTEXT: + * Grabs and releases bdev_lock. + */ +static void bd_finish_claiming(struct block_device *bdev, + struct block_device *whole, void *holder) +{ + spin_lock(&bdev_lock); + BUG_ON(whole->bd_claiming != holder); + BUG_ON(!bd_may_claim(bdev, whole, holder)); + __bd_claim(bdev, whole, holder); + __bd_abort_claiming(whole, holder); /* not actually an abort */ +} + /** * bd_claim - claim a block device * @bdev: block device to claim * @holder: holder trying to claim @bdev * - * Try to claim @bdev which must have been opened successfully. This - * function may be called with or without preceding - * blk_start_claiming(). In the former case, this function is always - * successful and terminates the claiming block. + * Try to claim @bdev which must have been opened successfully. * * CONTEXT: * Might sleep. @@ -807,23 +844,10 @@ int bd_claim(struct block_device *bdev, void *holder) might_sleep(); spin_lock(&bdev_lock); - res = bd_prepare_to_claim(bdev, whole, holder); - if (res == 0) { - /* note that for a whole device bd_holders - * will be incremented twice, and bd_holder will - * be set to bd_claim before being set to holder - */ - whole->bd_holders++; - whole->bd_holder = bd_claim; - bdev->bd_holders++; - bdev->bd_holder = holder; - } - - if (whole->bd_claiming) - __bd_abort_claiming(whole, holder); /* releases bdev_lock */ - else - spin_unlock(&bdev_lock); + if (res == 0) + __bd_claim(bdev, whole, holder); + spin_unlock(&bdev_lock); return res; } @@ -1477,7 +1501,7 @@ static int blkdev_open(struct inode * inode, struct file * filp) if (whole) { if (res == 0) - BUG_ON(bd_claim(bdev, filp) != 0); + bd_finish_claiming(bdev, whole, filp); else bd_abort_claiming(whole, filp); } @@ -1713,7 +1737,7 @@ struct block_device *open_bdev_exclusive(const char *path, fmode_t mode, void *h if ((mode & FMODE_WRITE) && bdev_read_only(bdev)) goto out_blkdev_put; - BUG_ON(bd_claim(bdev, holder) != 0); + bd_finish_claiming(bdev, whole, holder); return bdev; out_blkdev_put: -- cgit v1.2.2 From 3e6c05052c262ebe7fdd85e75e9d4f956cdd8d82 Mon Sep 17 00:00:00 2001 From: Jens Axboe Date: Mon, 7 Jun 2010 20:17:38 +0200 Subject: block: remove duplicate BUG_ON() in bd_finish_claiming() We do the same BUG_ON() just a line later when calling into __bd_abort_claiming(). Reported-by: Tejun Heo Signed-off-by: Jens Axboe --- fs/block_dev.c | 1 - 1 file changed, 1 deletion(-) (limited to 'fs') diff --git a/fs/block_dev.c b/fs/block_dev.c index a1642ef60197..99d6af811747 100644 --- a/fs/block_dev.c +++ b/fs/block_dev.c @@ -817,7 +817,6 @@ static void bd_finish_claiming(struct block_device *bdev, struct block_device *whole, void *holder) { spin_lock(&bdev_lock); - BUG_ON(whole->bd_claiming != holder); BUG_ON(!bd_may_claim(bdev, whole, holder)); __bd_claim(bdev, whole, holder); __bd_abort_claiming(whole, holder); /* not actually an abort */ -- cgit v1.2.2 From 1d862f41222b7f385bada9f85a67ca5592ffd33e Mon Sep 17 00:00:00 2001 From: Miklos Szeredi Date: Tue, 8 Jun 2010 16:28:45 +0200 Subject: pipe: fix pipe buffer resizing pipe_set_size() needs to copy pipe bufs from the old circular buffer to the new. The current code gets this wrong in multiple ways, resulting in oops. Test program is available here: http://www.kernel.org/pub/linux/kernel/people/mszeredi/piperesize/ Signed-off-by: Miklos Szeredi Signed-off-by: Jens Axboe --- fs/pipe.c | 13 ++++++++++--- 1 file changed, 10 insertions(+), 3 deletions(-) (limited to 'fs') diff --git a/fs/pipe.c b/fs/pipe.c index 69c4c7c13ea9..f31e2d472984 100644 --- a/fs/pipe.c +++ b/fs/pipe.c @@ -1145,13 +1145,20 @@ static long pipe_set_size(struct pipe_inode_info *pipe, unsigned long nr_pages) * and adjust the indexes. */ if (pipe->nrbufs) { - const unsigned int tail = pipe->nrbufs & (pipe->buffers - 1); - const unsigned int head = pipe->nrbufs - tail; + unsigned int tail; + unsigned int head; + tail = pipe->curbuf + pipe->nrbufs; + if (tail < pipe->buffers) + tail = 0; + else + tail &= (pipe->buffers - 1); + + head = pipe->nrbufs - tail; if (head) memcpy(bufs, pipe->bufs + pipe->curbuf, head * sizeof(struct pipe_buffer)); if (tail) - memcpy(bufs + head, pipe->bufs + pipe->curbuf, tail * sizeof(struct pipe_buffer)); + memcpy(bufs + head, pipe->bufs, tail * sizeof(struct pipe_buffer)); } pipe->curbuf = 0; -- cgit v1.2.2 From 6db40cf047a8723095caf79f5569d21b388d7b31 Mon Sep 17 00:00:00 2001 From: Miklos Szeredi Date: Wed, 9 Jun 2010 09:27:57 +0200 Subject: pipe: fix check in "set size" fcntl As it stands this check compares the number of pages to the page size. This makes no sense and makes the fcntl fail in almost any sane case. Fix it by checking if nr_pages is not zero (it can become zero only if arg is too big and round_pipe_size() overflows). Signed-off-by: Miklos Szeredi Signed-off-by: Jens Axboe --- fs/pipe.c | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-) (limited to 'fs') diff --git a/fs/pipe.c b/fs/pipe.c index f31e2d472984..279eef96c51c 100644 --- a/fs/pipe.c +++ b/fs/pipe.c @@ -1215,12 +1215,13 @@ long pipe_fcntl(struct file *file, unsigned int cmd, unsigned long arg) size = round_pipe_size(arg); nr_pages = size >> PAGE_SHIFT; + ret = -EINVAL; + if (!nr_pages) + goto out; + if (!capable(CAP_SYS_RESOURCE) && size > pipe_max_size) { ret = -EPERM; goto out; - } else if (nr_pages < PAGE_SIZE) { - ret = -EINVAL; - goto out; } ret = pipe_set_size(pipe, nr_pages); break; -- cgit v1.2.2 From 00d5643e7c5ed4ae1bb0b385fe2f41bb951cc3cd Mon Sep 17 00:00:00 2001 From: Jeff Mahoney Date: Thu, 10 Jun 2010 11:13:58 -0400 Subject: ceph: fix atomic64_t initialization on ia64 bdi_seq is an atomic_long_t but we're using ATOMIC_INIT, which causes build failures on ia64. This patch fixes it to use ATOMIC_LONG_INIT. Signed-off-by: Jeff Mahoney Signed-off-by: Sage Weil --- fs/ceph/super.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/ceph/super.c b/fs/ceph/super.c index 8db88792b5ad..fa87f51e38e1 100644 --- a/fs/ceph/super.c +++ b/fs/ceph/super.c @@ -926,7 +926,7 @@ static int ceph_compare_super(struct super_block *sb, void *data) /* * construct our own bdi so we can control readahead, etc. */ -static atomic_long_t bdi_seq = ATOMIC_INIT(0); +static atomic_long_t bdi_seq = ATOMIC_LONG_INIT(0); static int ceph_register_bdi(struct super_block *sb, struct ceph_client *client) { -- cgit v1.2.2 From 9dbd412f56c453f15014396c6024b895c1485ccb Mon Sep 17 00:00:00 2001 From: Sage Weil Date: Thu, 10 Jun 2010 13:21:20 -0700 Subject: ceph: fix misleading/incorrect debug message Nothing is released here: the caps message is simply ignored in this case. Signed-off-by: Sage Weil --- fs/ceph/caps.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/ceph/caps.c b/fs/ceph/caps.c index ae3e3a306445..da2a0e3cb200 100644 --- a/fs/ceph/caps.c +++ b/fs/ceph/caps.c @@ -2714,7 +2714,7 @@ void ceph_handle_caps(struct ceph_mds_session *session, spin_lock(&inode->i_lock); cap = __get_cap_for_mds(ceph_inode(inode), mds); if (!cap) { - dout("no cap on %p ino %llx.%llx from mds%d, releasing\n", + dout(" no cap on %p ino %llx.%llx from mds%d\n", inode, ceph_ino(inode), ceph_snap(inode), mds); spin_unlock(&inode->i_lock); goto done; -- cgit v1.2.2 From 3d7ded4d81d807c2f75f310a8d74a5d72be13a1b Mon Sep 17 00:00:00 2001 From: Sage Weil Date: Wed, 9 Jun 2010 16:47:10 -0700 Subject: ceph: release cap on import if we don't have the inode If we get an IMPORT that give us a cap, but we don't have the inode, queue a release (and try to send it immediately) so that the MDS doesn't get stuck waiting for us. Signed-off-by: Sage Weil --- fs/ceph/caps.c | 90 ++++++++++++++++++++++++++++++++-------------------- fs/ceph/mds_client.c | 6 ++-- fs/ceph/mds_client.h | 3 ++ 3 files changed, 61 insertions(+), 38 deletions(-) (limited to 'fs') diff --git a/fs/ceph/caps.c b/fs/ceph/caps.c index da2a0e3cb200..7c692e746237 100644 --- a/fs/ceph/caps.c +++ b/fs/ceph/caps.c @@ -981,6 +981,46 @@ static int send_cap_msg(struct ceph_mds_session *session, return 0; } +static void __queue_cap_release(struct ceph_mds_session *session, + u64 ino, u64 cap_id, u32 migrate_seq, + u32 issue_seq) +{ + struct ceph_msg *msg; + struct ceph_mds_cap_release *head; + struct ceph_mds_cap_item *item; + + spin_lock(&session->s_cap_lock); + BUG_ON(!session->s_num_cap_releases); + msg = list_first_entry(&session->s_cap_releases, + struct ceph_msg, list_head); + + dout(" adding %llx release to mds%d msg %p (%d left)\n", + ino, session->s_mds, msg, session->s_num_cap_releases); + + BUG_ON(msg->front.iov_len + sizeof(*item) > PAGE_CACHE_SIZE); + head = msg->front.iov_base; + head->num = cpu_to_le32(le32_to_cpu(head->num) + 1); + item = msg->front.iov_base + msg->front.iov_len; + item->ino = cpu_to_le64(ino); + item->cap_id = cpu_to_le64(cap_id); + item->migrate_seq = cpu_to_le32(migrate_seq); + item->seq = cpu_to_le32(issue_seq); + + session->s_num_cap_releases--; + + msg->front.iov_len += sizeof(*item); + if (le32_to_cpu(head->num) == CEPH_CAPS_PER_RELEASE) { + dout(" release msg %p full\n", msg); + list_move_tail(&msg->list_head, &session->s_cap_releases_done); + } else { + dout(" release msg %p at %d/%d (%d)\n", msg, + (int)le32_to_cpu(head->num), + (int)CEPH_CAPS_PER_RELEASE, + (int)msg->front.iov_len); + } + spin_unlock(&session->s_cap_lock); +} + /* * Queue cap releases when an inode is dropped from our cache. Since * inode is about to be destroyed, there is no need for i_lock. @@ -994,41 +1034,9 @@ void ceph_queue_caps_release(struct inode *inode) while (p) { struct ceph_cap *cap = rb_entry(p, struct ceph_cap, ci_node); struct ceph_mds_session *session = cap->session; - struct ceph_msg *msg; - struct ceph_mds_cap_release *head; - struct ceph_mds_cap_item *item; - spin_lock(&session->s_cap_lock); - BUG_ON(!session->s_num_cap_releases); - msg = list_first_entry(&session->s_cap_releases, - struct ceph_msg, list_head); - - dout(" adding %p release to mds%d msg %p (%d left)\n", - inode, session->s_mds, msg, session->s_num_cap_releases); - - BUG_ON(msg->front.iov_len + sizeof(*item) > PAGE_CACHE_SIZE); - head = msg->front.iov_base; - head->num = cpu_to_le32(le32_to_cpu(head->num) + 1); - item = msg->front.iov_base + msg->front.iov_len; - item->ino = cpu_to_le64(ceph_ino(inode)); - item->cap_id = cpu_to_le64(cap->cap_id); - item->migrate_seq = cpu_to_le32(cap->mseq); - item->seq = cpu_to_le32(cap->issue_seq); - - session->s_num_cap_releases--; - - msg->front.iov_len += sizeof(*item); - if (le32_to_cpu(head->num) == CEPH_CAPS_PER_RELEASE) { - dout(" release msg %p full\n", msg); - list_move_tail(&msg->list_head, - &session->s_cap_releases_done); - } else { - dout(" release msg %p at %d/%d (%d)\n", msg, - (int)le32_to_cpu(head->num), - (int)CEPH_CAPS_PER_RELEASE, - (int)msg->front.iov_len); - } - spin_unlock(&session->s_cap_lock); + __queue_cap_release(session, ceph_ino(inode), cap->cap_id, + cap->mseq, cap->issue_seq); p = rb_next(p); __ceph_remove_cap(cap); } @@ -2655,7 +2663,7 @@ void ceph_handle_caps(struct ceph_mds_session *session, struct ceph_mds_caps *h; int mds = session->s_mds; int op; - u32 seq; + u32 seq, mseq; struct ceph_vino vino; u64 cap_id; u64 size, max_size; @@ -2675,6 +2683,7 @@ void ceph_handle_caps(struct ceph_mds_session *session, vino.snap = CEPH_NOSNAP; cap_id = le64_to_cpu(h->cap_id); seq = le32_to_cpu(h->seq); + mseq = le32_to_cpu(h->migrate_seq); size = le64_to_cpu(h->size); max_size = le64_to_cpu(h->max_size); @@ -2689,6 +2698,17 @@ void ceph_handle_caps(struct ceph_mds_session *session, vino.snap, inode); if (!inode) { dout(" i don't have ino %llx\n", vino.ino); + + if (op == CEPH_CAP_OP_IMPORT) + __queue_cap_release(session, vino.ino, cap_id, + mseq, seq); + + /* + * send any full release message to try to move things + * along for the mds (who clearly thinks we still have this + * cap). + */ + ceph_send_cap_releases(mdsc, session); goto done; } diff --git a/fs/ceph/mds_client.c b/fs/ceph/mds_client.c index 29b4485cf1ca..d28b6a9c0f96 100644 --- a/fs/ceph/mds_client.c +++ b/fs/ceph/mds_client.c @@ -1176,8 +1176,8 @@ static int check_cap_flush(struct ceph_mds_client *mdsc, u64 want_flush_seq) /* * called under s_mutex */ -static void send_cap_releases(struct ceph_mds_client *mdsc, - struct ceph_mds_session *session) +void ceph_send_cap_releases(struct ceph_mds_client *mdsc, + struct ceph_mds_session *session) { struct ceph_msg *msg; @@ -2693,7 +2693,7 @@ static void delayed_work(struct work_struct *work) add_cap_releases(mdsc, s, -1); if (s->s_state == CEPH_MDS_SESSION_OPEN || s->s_state == CEPH_MDS_SESSION_HUNG) - send_cap_releases(mdsc, s); + ceph_send_cap_releases(mdsc, s); mutex_unlock(&s->s_mutex); ceph_put_mds_session(s); diff --git a/fs/ceph/mds_client.h b/fs/ceph/mds_client.h index d9936c4f1212..e43752b52635 100644 --- a/fs/ceph/mds_client.h +++ b/fs/ceph/mds_client.h @@ -322,6 +322,9 @@ static inline void ceph_mdsc_put_request(struct ceph_mds_request *req) kref_put(&req->r_kref, ceph_mdsc_release_request); } +extern void ceph_send_cap_releases(struct ceph_mds_client *mdsc, + struct ceph_mds_session *session); + extern void ceph_mdsc_pre_umount(struct ceph_mds_client *mdsc); extern char *ceph_mdsc_build_path(struct dentry *dentry, int *plen, u64 *base, -- cgit v1.2.2 From 2b2300d62ea413bec631d5b880effa2cc5363acb Mon Sep 17 00:00:00 2001 From: Sage Weil Date: Wed, 9 Jun 2010 16:52:04 -0700 Subject: ceph: try to send partial cap release on cap message on missing inode If we have enough memory to allocate a new cap release message, do so, so that we can send a partial release message immediately. This keeps us from making the MDS wait when the cap release it needs is in a partially full release message. If we fail because of ENOMEM, oh well, they'll just have to wait a bit longer. Signed-off-by: Sage Weil --- fs/ceph/caps.c | 1 + fs/ceph/mds_client.c | 10 +++++----- fs/ceph/mds_client.h | 3 +++ 3 files changed, 9 insertions(+), 5 deletions(-) (limited to 'fs') diff --git a/fs/ceph/caps.c b/fs/ceph/caps.c index 7c692e746237..619b61655ee5 100644 --- a/fs/ceph/caps.c +++ b/fs/ceph/caps.c @@ -2708,6 +2708,7 @@ void ceph_handle_caps(struct ceph_mds_session *session, * along for the mds (who clearly thinks we still have this * cap). */ + ceph_add_cap_releases(mdsc, session, -1); ceph_send_cap_releases(mdsc, session); goto done; } diff --git a/fs/ceph/mds_client.c b/fs/ceph/mds_client.c index d28b6a9c0f96..1766947fc07a 100644 --- a/fs/ceph/mds_client.c +++ b/fs/ceph/mds_client.c @@ -1066,9 +1066,9 @@ static int trim_caps(struct ceph_mds_client *mdsc, * * Called under s_mutex. */ -static int add_cap_releases(struct ceph_mds_client *mdsc, - struct ceph_mds_session *session, - int extra) +int ceph_add_cap_releases(struct ceph_mds_client *mdsc, + struct ceph_mds_session *session, + int extra) { struct ceph_msg *msg; struct ceph_mds_cap_release *head; @@ -1980,7 +1980,7 @@ out_err: } mutex_unlock(&mdsc->mutex); - add_cap_releases(mdsc, req->r_session, -1); + ceph_add_cap_releases(mdsc, req->r_session, -1); mutex_unlock(&session->s_mutex); /* kick calling process */ @@ -2690,7 +2690,7 @@ static void delayed_work(struct work_struct *work) send_renew_caps(mdsc, s); else ceph_con_keepalive(&s->s_con); - add_cap_releases(mdsc, s, -1); + ceph_add_cap_releases(mdsc, s, -1); if (s->s_state == CEPH_MDS_SESSION_OPEN || s->s_state == CEPH_MDS_SESSION_HUNG) ceph_send_cap_releases(mdsc, s); diff --git a/fs/ceph/mds_client.h b/fs/ceph/mds_client.h index e43752b52635..b292fa42a66d 100644 --- a/fs/ceph/mds_client.h +++ b/fs/ceph/mds_client.h @@ -322,6 +322,9 @@ static inline void ceph_mdsc_put_request(struct ceph_mds_request *req) kref_put(&req->r_kref, ceph_mdsc_release_request); } +extern int ceph_add_cap_releases(struct ceph_mds_client *mdsc, + struct ceph_mds_session *session, + int extra); extern void ceph_send_cap_releases(struct ceph_mds_client *mdsc, struct ceph_mds_session *session); -- cgit v1.2.2 From 4a001071d3549f596c7c3736c5dda8a3a4aba9ed Mon Sep 17 00:00:00 2001 From: Miao Xie Date: Mon, 7 Jun 2010 03:38:51 +0000 Subject: Btrfs: fix loop device on top of btrfs We cannot use the loop device which has been connected to a file in the btrf The reproduce steps is following: # dd if=/dev/zero of=vdev0 bs=1M count=1024 # losetup /dev/loop0 vdev0 # mkfs.btrfs /dev/loop0 ... failed to zero device start -5 The reason is that the btrfs don't implement either ->write_begin or ->write the VFS API, so we fix it by setting ->write to do_sync_write(). Signed-off-by: Miao Xie Signed-off-by: Chris Mason --- fs/btrfs/file.c | 1 + 1 file changed, 1 insertion(+) (limited to 'fs') diff --git a/fs/btrfs/file.c b/fs/btrfs/file.c index 79437c5eeb1e..abcb91867b56 100644 --- a/fs/btrfs/file.c +++ b/fs/btrfs/file.c @@ -1197,6 +1197,7 @@ static int btrfs_file_mmap(struct file *filp, struct vm_area_struct *vma) const struct file_operations btrfs_file_operations = { .llseek = generic_file_llseek, .read = do_sync_read, + .write = do_sync_write, .aio_read = generic_file_aio_read, .splice_read = generic_file_splice_read, .aio_write = btrfs_file_aio_write, -- cgit v1.2.2 From 836097797236fd727f82ec2f3f376ac41a430876 Mon Sep 17 00:00:00 2001 From: Josef Bacik Date: Mon, 7 Jun 2010 18:26:37 +0000 Subject: Btrfs: fix fallocate regression Seems that when btrfs_fallocate was converted to use the new ENOSPC stuff we dropped passing the mode to the function that actually does the preallocation. This breaks anybody who wants to use FALLOC_FL_KEEP_SIZE. Thanks, Signed-off-by: Josef Bacik Signed-off-by: Chris Mason --- fs/btrfs/inode.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index 2551b8018399..d999c538cdc2 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -6893,7 +6893,7 @@ static long btrfs_fallocate(struct inode *inode, int mode, if (em->block_start == EXTENT_MAP_HOLE || (cur_offset >= inode->i_size && !test_bit(EXTENT_FLAG_PREALLOC, &em->flags))) { - ret = btrfs_prealloc_file_range(inode, 0, cur_offset, + ret = btrfs_prealloc_file_range(inode, mode, cur_offset, last_byte - cur_offset, 1 << inode->i_blkbits, offset + len, -- cgit v1.2.2 From 0e4dcbef1c0c3e29f9c7f824359445d385b2649a Mon Sep 17 00:00:00 2001 From: Dan Carpenter Date: Tue, 1 Jun 2010 08:23:11 +0000 Subject: Btrfs: uninitialized data is check_path_shared() refs can be used with uninitialized data if btrfs_lookup_extent_info() fails on the first pass through the loop. In the original code if that happens then check_path_shared() probably returns 1, this patch changes it to return 1 for safety. Signed-off-by: Dan Carpenter Signed-off-by: Chris Mason --- fs/btrfs/inode.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index d999c538cdc2..f08427c70a78 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -2673,7 +2673,7 @@ static int check_path_shared(struct btrfs_root *root, struct extent_buffer *eb; int level; int ret; - u64 refs; + u64 refs = 1; for (level = 0; level < BTRFS_MAX_LEVEL; level++) { if (!path->nodes[level]) -- cgit v1.2.2 From 058a457ef0ce28d595af53d6103db73332383cbc Mon Sep 17 00:00:00 2001 From: Miao Xie Date: Thu, 20 May 2010 07:21:50 +0000 Subject: Btrfs: fix remap_file_pages error when we use remap_file_pages() to remap a file, remap_file_pages always return error. It is because btrfs didn't set VM_CAN_NONLINEAR for vma. Signed-off-by: Miao Xie Signed-off-by: Chris Mason --- fs/btrfs/file.c | 9 ++++++++- 1 file changed, 8 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/btrfs/file.c b/fs/btrfs/file.c index abcb91867b56..ce0cd29efa9e 100644 --- a/fs/btrfs/file.c +++ b/fs/btrfs/file.c @@ -1189,8 +1189,15 @@ static const struct vm_operations_struct btrfs_file_vm_ops = { static int btrfs_file_mmap(struct file *filp, struct vm_area_struct *vma) { - vma->vm_ops = &btrfs_file_vm_ops; + struct address_space *mapping = filp->f_mapping; + + if (!mapping->a_ops->readpage) + return -ENOEXEC; + file_accessed(filp); + vma->vm_ops = &btrfs_file_vm_ops; + vma->vm_flags |= VM_CAN_NONLINEAR; + return 0; } -- cgit v1.2.2 From 046f264f6b3b2cf7e5a1769fc92335d8a9316282 Mon Sep 17 00:00:00 2001 From: "Yan, Zheng" Date: Mon, 31 May 2010 08:58:47 +0000 Subject: Btrfs: Fix null dereference in relocation.c Fix a potential null dereference in relocation.c Signed-off-by: Yan Zheng Acked-by: Dan Carpenter Signed-off-by: Chris Mason --- fs/btrfs/relocation.c | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/relocation.c b/fs/btrfs/relocation.c index 05d41e569236..b37d723b9d4a 100644 --- a/fs/btrfs/relocation.c +++ b/fs/btrfs/relocation.c @@ -784,16 +784,17 @@ again: struct btrfs_extent_ref_v0 *ref0; ref0 = btrfs_item_ptr(eb, path1->slots[0], struct btrfs_extent_ref_v0); - root = find_tree_root(rc, eb, ref0); - if (!root->ref_cows) - cur->cowonly = 1; if (key.objectid == key.offset) { + root = find_tree_root(rc, eb, ref0); if (root && !should_ignore_root(root)) cur->root = root; else list_add(&cur->list, &useless); break; } + if (is_cowonly_root(btrfs_ref_root_v0(eb, + ref0))) + cur->cowonly = 1; } #else BUG_ON(key.type == BTRFS_EXTENT_REF_V0_KEY); -- cgit v1.2.2 From 3bf84a5a834d13e7c5c3e8e5b5c6b26012118dd8 Mon Sep 17 00:00:00 2001 From: "Yan, Zheng" Date: Mon, 31 May 2010 09:04:46 +0000 Subject: Btrfs: Fix BUG_ON for fs converted from extN Tree blocks can live in data block groups in FS converted from extN. So it's easy to trigger the BUG_ON. Signed-off-by: Yan Zheng Signed-off-by: Chris Mason --- fs/btrfs/extent-tree.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c index 6c14101506e1..a46b64de8f02 100644 --- a/fs/btrfs/extent-tree.c +++ b/fs/btrfs/extent-tree.c @@ -4360,7 +4360,8 @@ void btrfs_free_tree_block(struct btrfs_trans_handle *trans, block_rsv = get_block_rsv(trans, root); cache = btrfs_lookup_block_group(root->fs_info, buf->start); - BUG_ON(block_rsv->space_info != cache->space_info); + if (block_rsv->space_info != cache->space_info) + goto out; if (btrfs_header_generation(buf) == trans->transid) { if (root->root_key.objectid != BTRFS_TREE_LOG_OBJECTID) { -- cgit v1.2.2 From fb4f6f910ca6f58564c31a680ef88940d8192713 Mon Sep 17 00:00:00 2001 From: Dan Carpenter Date: Sat, 29 May 2010 09:40:57 +0000 Subject: Btrfs: handle error returns from btrfs_lookup_dir_item() If btrfs_lookup_dir_item() fails, we should can just let the mount fail with an error. Signed-off-by: Dan Carpenter Signed-off-by: Chris Mason --- fs/btrfs/super.c | 2 ++ 1 file changed, 2 insertions(+) (limited to 'fs') diff --git a/fs/btrfs/super.c b/fs/btrfs/super.c index 574285c8cbd4..9ea711430466 100644 --- a/fs/btrfs/super.c +++ b/fs/btrfs/super.c @@ -360,6 +360,8 @@ static struct dentry *get_default_root(struct super_block *sb, */ dir_id = btrfs_super_root_dir(&root->fs_info->super_copy); di = btrfs_lookup_dir_item(NULL, root, path, dir_id, "default", 7, 0); + if (IS_ERR(di)) + return ERR_CAST(di); if (!di) { /* * Ok the default dir item isn't there. This is weird since -- cgit v1.2.2 From 676e4c86391936795c82ccd11ca9671ee6307936 Mon Sep 17 00:00:00 2001 From: Dan Carpenter Date: Sat, 29 May 2010 09:43:07 +0000 Subject: Btrfs: handle kzalloc() failure in open_ctree() Unwind and return -ENOMEM if the allocation fails here. Signed-off-by: Dan Carpenter Signed-off-by: Chris Mason --- fs/btrfs/disk-io.c | 7 +++++-- 1 file changed, 5 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c index f3b287c22caf..73895baf6e07 100644 --- a/fs/btrfs/disk-io.c +++ b/fs/btrfs/disk-io.c @@ -1941,8 +1941,11 @@ struct btrfs_root *open_ctree(struct super_block *sb, btrfs_level_size(tree_root, btrfs_super_log_root_level(disk_super)); - log_tree_root = kzalloc(sizeof(struct btrfs_root), - GFP_NOFS); + log_tree_root = kzalloc(sizeof(struct btrfs_root), GFP_NOFS); + if (!log_tree_root) { + err = -ENOMEM; + goto fail_trans_kthread; + } __setup_root(nodesize, leafsize, sectorsize, stripesize, log_tree_root, fs_info, BTRFS_TREE_LOG_OBJECTID); -- cgit v1.2.2 From 4cbd1149fbcc351bdf08ab749867d157905d0d35 Mon Sep 17 00:00:00 2001 From: Dan Carpenter Date: Sat, 29 May 2010 09:42:19 +0000 Subject: Btrfs: btrfs_iget() returns ERR_PTR btrfs_iget() returns an ERR_PTR() on failure and not null. Signed-off-by: Dan Carpenter Signed-off-by: Chris Mason --- fs/btrfs/super.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/super.c b/fs/btrfs/super.c index 9ea711430466..859ddaab5e02 100644 --- a/fs/btrfs/super.c +++ b/fs/btrfs/super.c @@ -392,8 +392,8 @@ setup_root: location.offset = 0; inode = btrfs_iget(sb, &location, new_root, &new); - if (!inode) - return ERR_PTR(-ENOMEM); + if (IS_ERR(inode)) + return ERR_CAST(inode); /* * If we're just mounting the root most subvol put the inode and return -- cgit v1.2.2 From d327099a23e3d0c8ec09137e9b4b115449d4eb29 Mon Sep 17 00:00:00 2001 From: Dan Carpenter Date: Sat, 29 May 2010 09:46:47 +0000 Subject: Btrfs: unwind after btrfs_start_transaction() errors This was added by a22285a6a3: "Btrfs: Integrate metadata reservation with start_transaction". If we goto out here then we skip all the unwinding and there are locks still held etc. Signed-off-by: Dan Carpenter Signed-off-by: Chris Mason --- fs/btrfs/ioctl.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/btrfs/ioctl.c b/fs/btrfs/ioctl.c index 4cdb98cf26de..9f9a1d9607a7 100644 --- a/fs/btrfs/ioctl.c +++ b/fs/btrfs/ioctl.c @@ -1280,7 +1280,7 @@ static noinline int btrfs_ioctl_snap_destroy(struct file *file, trans = btrfs_start_transaction(root, 0); if (IS_ERR(trans)) { err = PTR_ERR(trans); - goto out; + goto out_up_write; } trans->block_rsv = &root->fs_info->global_block_rsv; -- cgit v1.2.2 From 3140c9a34b44cd4013baae8704fdb34a93a44475 Mon Sep 17 00:00:00 2001 From: Dan Carpenter Date: Sat, 29 May 2010 09:44:10 +0000 Subject: Btrfs: btrfs_read_fs_root_no_name() returns ERR_PTRs btrfs_read_fs_root_no_name() returns ERR_PTRs on error so I added a check for that. It's not clear to me if it can also return NULL pointers or not so I left the original NULL pointer check as is. Signed-off-by: Dan Carpenter Signed-off-by: Chris Mason --- fs/btrfs/disk-io.c | 4 ++++ 1 file changed, 4 insertions(+) (limited to 'fs') diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c index 73895baf6e07..34f7c375567e 100644 --- a/fs/btrfs/disk-io.c +++ b/fs/btrfs/disk-io.c @@ -1985,6 +1985,10 @@ struct btrfs_root *open_ctree(struct super_block *sb, fs_info->fs_root = btrfs_read_fs_root_no_name(fs_info, &location); if (!fs_info->fs_root) goto fail_trans_kthread; + if (IS_ERR(fs_info->fs_root)) { + err = PTR_ERR(fs_info->fs_root); + goto fail_trans_kthread; + } if (!(sb->s_flags & MS_RDONLY)) { down_read(&fs_info->cleanup_work_sem); -- cgit v1.2.2 From cf1e99a4e0daa4a5623cd71108509823b9ff2d30 Mon Sep 17 00:00:00 2001 From: Dan Carpenter Date: Sat, 29 May 2010 09:47:24 +0000 Subject: Btrfs: btrfs_lookup_dir_item() can return ERR_PTR btrfs_lookup_dir_item() can return either ERR_PTRs or null. Signed-off-by: Dan Carpenter Signed-off-by: Chris Mason --- fs/btrfs/ioctl.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/btrfs/ioctl.c b/fs/btrfs/ioctl.c index 9f9a1d9607a7..4dbaf89b1337 100644 --- a/fs/btrfs/ioctl.c +++ b/fs/btrfs/ioctl.c @@ -1845,7 +1845,7 @@ static long btrfs_ioctl_default_subvol(struct file *file, void __user *argp) dir_id = btrfs_super_root_dir(&root->fs_info->super_copy); di = btrfs_lookup_dir_item(trans, root->fs_info->tree_root, path, dir_id, "default", 7, 1); - if (!di) { + if (IS_ERR_OR_NULL(di)) { btrfs_free_path(path); btrfs_end_transaction(trans, root); printk(KERN_ERR "Umm, you don't have the default dir item, " -- cgit v1.2.2 From 2f26afba46f0ebf155cf9be746496a0304a5b7cf Mon Sep 17 00:00:00 2001 From: Shi Weihua Date: Tue, 18 May 2010 00:50:32 +0000 Subject: Btrfs: should add a permission check for setfacl On btrfs, do the following ------------------ # su user1 # cd btrfs-part/ # touch aaa # getfacl aaa # file: aaa # owner: user1 # group: user1 user::rw- group::rw- other::r-- # su user2 # cd btrfs-part/ # setfacl -m u::rwx aaa # getfacl aaa # file: aaa # owner: user1 # group: user1 user::rwx <- successed to setfacl group::rw- other::r-- ------------------ but we should prohibit it that user2 changing user1's acl. In fact, on ext3 and other fs, a message occurs: setfacl: aaa: Operation not permitted This patch fixed it. Signed-off-by: Shi Weihua Signed-off-by: Chris Mason --- fs/btrfs/acl.c | 3 +++ 1 file changed, 3 insertions(+) (limited to 'fs') diff --git a/fs/btrfs/acl.c b/fs/btrfs/acl.c index 6ef7b26724ec..6b4d0cca5c7f 100644 --- a/fs/btrfs/acl.c +++ b/fs/btrfs/acl.c @@ -160,6 +160,9 @@ static int btrfs_xattr_acl_set(struct dentry *dentry, const char *name, int ret; struct posix_acl *acl = NULL; + if (!is_owner_or_cap(dentry->d_inode)) + return -EPERM; + if (value) { acl = posix_acl_from_xattr(value, size); if (acl == NULL) { -- cgit v1.2.2 From 731e3d1b4348a96d53de6c084774424dedc64a3b Mon Sep 17 00:00:00 2001 From: Shi Weihua Date: Tue, 18 May 2010 00:51:54 +0000 Subject: Btrfs: prohibit a operation of changing acl's mask when noacl mount option used when used Posix File System Test Suite(pjd-fstest) to test btrfs, some cases about setfacl failed when noacl mount option used. I simplified used commands in pjd-fstest, and the following steps can reproduce it. ------------------------ # cd btrfs-part/ # mkdir aaa # setfacl -m m::rw aaa <- successed, but not expected by pjd-fstest. ------------------------ I checked ext3, a warning message occured, like as: setfacl: aaa/: Operation not supported Certainly, it's expected by pjd-fstest. So, i compared acl.c of btrfs and ext3. Based on that, a patch created. Fortunately, it works. Signed-off-by: Shi Weihua Signed-off-by: Chris Mason --- fs/btrfs/acl.c | 3 +++ 1 file changed, 3 insertions(+) (limited to 'fs') diff --git a/fs/btrfs/acl.c b/fs/btrfs/acl.c index 6b4d0cca5c7f..a372985b3a9a 100644 --- a/fs/btrfs/acl.c +++ b/fs/btrfs/acl.c @@ -163,6 +163,9 @@ static int btrfs_xattr_acl_set(struct dentry *dentry, const char *name, if (!is_owner_or_cap(dentry->d_inode)) return -EPERM; + if (!IS_POSIXACL(dentry->d_inode)) + return -EOPNOTSUPP; + if (value) { acl = posix_acl_from_xattr(value, size); if (acl == NULL) { -- cgit v1.2.2 From 15e7000095e6fc9ad07e476a100c900c72c14225 Mon Sep 17 00:00:00 2001 From: Sage Weil Date: Mon, 17 May 2010 17:15:27 +0000 Subject: Btrfs: avoid BUG when dropping root and reference in same transaction If btrfs_ioctl_snap_destroy() deletes a snapshot but finishes with end_transaction(), the cleaner kthread may come in and drop the root in the same transaction. If that's the case, the root's refs still == 1 in the tree when btrfs_del_root() deletes the item, because commit_fs_roots() hasn't updated it yet (that happens during the commit). This wasn't a problem before only because btrfs_ioctl_snap_destroy() would commit the transaction before dropping the dentry reference, so the dead root wouldn't get queued up until after the fs root item was updated in the btree. Since it is not an error to drop the root reference and the root in the same transaction, just drop the BUG_ON() in btrfs_del_root(). Signed-off-by: Sage Weil Signed-off-by: Chris Mason --- fs/btrfs/root-tree.c | 3 --- 1 file changed, 3 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/root-tree.c b/fs/btrfs/root-tree.c index b91ccd972644..2d958be761c8 100644 --- a/fs/btrfs/root-tree.c +++ b/fs/btrfs/root-tree.c @@ -330,7 +330,6 @@ int btrfs_del_root(struct btrfs_trans_handle *trans, struct btrfs_root *root, { struct btrfs_path *path; int ret; - u32 refs; struct btrfs_root_item *ri; struct extent_buffer *leaf; @@ -344,8 +343,6 @@ int btrfs_del_root(struct btrfs_trans_handle *trans, struct btrfs_root *root, leaf = path->nodes[0]; ri = btrfs_item_ptr(leaf, path->slots[0], struct btrfs_root_item); - refs = btrfs_disk_root_refs(leaf, ri); - BUG_ON(refs != 0); ret = btrfs_del_item(trans, root, path); out: btrfs_free_path(path); -- cgit v1.2.2 From 834e74759a473f8101a273e843d1edec2778801d Mon Sep 17 00:00:00 2001 From: Dan Carpenter Date: Sat, 29 May 2010 09:48:35 +0000 Subject: Btrfs: handle ERR_PTR from posix_acl_from_xattr() posix_acl_from_xattr() returns both ERR_PTRs and null, but it's OK to pass null values to set_cached_acl() Signed-off-by: Dan Carpenter Signed-off-by: Chris Mason --- fs/btrfs/acl.c | 2 ++ 1 file changed, 2 insertions(+) (limited to 'fs') diff --git a/fs/btrfs/acl.c b/fs/btrfs/acl.c index a372985b3a9a..1606dc1e8d4a 100644 --- a/fs/btrfs/acl.c +++ b/fs/btrfs/acl.c @@ -60,6 +60,8 @@ static struct posix_acl *btrfs_get_acl(struct inode *inode, int type) size = __btrfs_getxattr(inode, name, value, size); if (size > 0) { acl = posix_acl_from_xattr(value, size); + if (IS_ERR(acl)) + return acl; set_cached_acl(inode, type, acl); } kfree(value); -- cgit v1.2.2 From 6f902af400b2499c80865c62a06fbbd15cf804fd Mon Sep 17 00:00:00 2001 From: Dan Carpenter Date: Sat, 29 May 2010 09:49:07 +0000 Subject: Btrfs: The file argument for fsync() is never null The "file" argument for fsync is never null so we can remove this check. What drew my attention here is that 7ea8085910e: "drop unused dentry argument to ->fsync" introduced an unconditional dereference at the start of the function and that generated a smatch warning. Signed-off-by: Dan Carpenter Signed-off-by: Chris Mason --- fs/btrfs/file.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/btrfs/file.c b/fs/btrfs/file.c index ce0cd29efa9e..7f29464c0ebf 100644 --- a/fs/btrfs/file.c +++ b/fs/btrfs/file.c @@ -1139,7 +1139,7 @@ int btrfs_sync_file(struct file *file, struct dentry *dentry, int datasync) /* * ok we haven't committed the transaction yet, lets do a commit */ - if (file && file->private_data) + if (file->private_data) btrfs_ioctl_trans_end(file); trans = btrfs_start_transaction(root, 0); -- cgit v1.2.2