From 4e857c58efeb99393cba5a5d0d8ec7117183137c Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Mon, 17 Mar 2014 18:06:10 +0100 Subject: arch: Mass conversion of smp_mb__*() Mostly scripted conversion of the smp_mb__* barriers. Signed-off-by: Peter Zijlstra Acked-by: Paul E. McKenney Link: http://lkml.kernel.org/n/tip-55dhyhocezdw1dg7u19hmh1u@git.kernel.org Cc: Linus Torvalds Cc: linux-arch@vger.kernel.org Signed-off-by: Ingo Molnar --- fs/btrfs/inode.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) (limited to 'fs/btrfs/inode.c') diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index 5f805bc944f..5a3b8371772 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -7126,7 +7126,7 @@ static void btrfs_end_dio_bio(struct bio *bio, int err) * before atomic variable goto zero, we must make sure * dip->errors is perceived to be set. */ - smp_mb__before_atomic_dec(); + smp_mb__before_atomic(); } /* if there are more bios still pending for this dio, just exit */ @@ -7306,7 +7306,7 @@ out_err: * before atomic variable goto zero, we must * make sure dip->errors is perceived to be set. */ - smp_mb__before_atomic_dec(); + smp_mb__before_atomic(); if (atomic_dec_and_test(&dip->pending_bios)) bio_io_error(dip->orig_bio); @@ -7449,7 +7449,7 @@ static ssize_t btrfs_direct_IO(int rw, struct kiocb *iocb, return 0; atomic_inc(&inode->i_dio_count); - smp_mb__after_atomic_inc(); + smp_mb__after_atomic(); /* * The generic stuff only does filemap_write_and_wait_range, which -- cgit v1.2.3-70-g09d2 From d8d3d94b80aa1a1c0ca75c58b8abdc7356f38418 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Tue, 4 Mar 2014 21:27:34 -0500 Subject: pass iov_iter to ->direct_IO() unmodified, for now Signed-off-by: Al Viro --- Documentation/filesystems/Locking | 3 +-- Documentation/filesystems/vfs.txt | 3 +-- drivers/staging/lustre/lustre/llite/rw26.c | 17 ++++++++--------- fs/9p/vfs_addr.c | 5 ++--- fs/block_dev.c | 9 +++++---- fs/btrfs/inode.c | 12 ++++++------ fs/ceph/addr.c | 4 ++-- fs/cifs/file.c | 4 ++-- fs/exofs/inode.c | 2 +- fs/ext2/inode.c | 11 ++++++----- fs/ext3/inode.c | 16 +++++++--------- fs/ext4/inode.c | 11 +++++------ fs/f2fs/data.c | 8 ++++---- fs/fat/inode.c | 13 +++++++------ fs/fuse/file.c | 10 +++++----- fs/gfs2/aops.c | 11 +++++------ fs/hfs/inode.c | 8 ++++---- fs/hfsplus/inode.c | 6 +++--- fs/jfs/inode.c | 8 ++++---- fs/nfs/direct.c | 8 ++++---- fs/nilfs2/inode.c | 10 +++++----- fs/ocfs2/aops.c | 7 +++---- fs/reiserfs/inode.c | 9 ++++----- fs/udf/file.c | 4 ++-- fs/udf/inode.c | 8 ++++---- fs/xfs/xfs_aops.c | 15 +++++++-------- include/linux/fs.h | 3 +-- include/linux/nfs_fs.h | 3 +-- mm/filemap.c | 9 ++++----- mm/page_io.c | 6 ++++-- 30 files changed, 117 insertions(+), 126 deletions(-) (limited to 'fs/btrfs/inode.c') diff --git a/Documentation/filesystems/Locking b/Documentation/filesystems/Locking index eba79013425..9b0d5a33c8b 100644 --- a/Documentation/filesystems/Locking +++ b/Documentation/filesystems/Locking @@ -196,8 +196,7 @@ prototypes: void (*invalidatepage) (struct page *, unsigned int, unsigned int); int (*releasepage) (struct page *, int); void (*freepage)(struct page *); - int (*direct_IO)(int, struct kiocb *, const struct iovec *iov, - loff_t offset, unsigned long nr_segs); + int (*direct_IO)(int, struct kiocb *, struct iov_iter *iter, loff_t offset); int (*get_xip_mem)(struct address_space *, pgoff_t, int, void **, unsigned long *); int (*migratepage)(struct address_space *, struct page *, struct page *); diff --git a/Documentation/filesystems/vfs.txt b/Documentation/filesystems/vfs.txt index 617f6d70c07..1846374a5ad 100644 --- a/Documentation/filesystems/vfs.txt +++ b/Documentation/filesystems/vfs.txt @@ -589,8 +589,7 @@ struct address_space_operations { void (*invalidatepage) (struct page *, unsigned int, unsigned int); int (*releasepage) (struct page *, int); void (*freepage)(struct page *); - ssize_t (*direct_IO)(int, struct kiocb *, const struct iovec *iov, - loff_t offset, unsigned long nr_segs); + ssize_t (*direct_IO)(int, struct kiocb *, struct iov_iter *iter, loff_t offset); struct page* (*get_xip_page)(struct address_space *, sector_t, int); /* migrate the contents of a page to the specified target */ diff --git a/drivers/staging/lustre/lustre/llite/rw26.c b/drivers/staging/lustre/lustre/llite/rw26.c index 7e3e0967993..66e05c6f4d2 100644 --- a/drivers/staging/lustre/lustre/llite/rw26.c +++ b/drivers/staging/lustre/lustre/llite/rw26.c @@ -363,15 +363,14 @@ static ssize_t ll_direct_IO_26_seg(const struct lu_env *env, struct cl_io *io, #define MAX_DIO_SIZE ((MAX_MALLOC / sizeof(struct brw_page) * PAGE_CACHE_SIZE) & \ ~(DT_MAX_BRW_SIZE - 1)) static ssize_t ll_direct_IO_26(int rw, struct kiocb *iocb, - const struct iovec *iov, loff_t file_offset, - unsigned long nr_segs) + struct iov_iter *iter, loff_t file_offset) { struct lu_env *env; struct cl_io *io; struct file *file = iocb->ki_filp; struct inode *inode = file->f_mapping->host; struct ccc_object *obj = cl_inode2ccc(inode); - long count = iov_length(iov, nr_segs); + long count = iov_length(iter->iov, iter->nr_segs); long tot_bytes = 0, result = 0; struct ll_inode_info *lli = ll_i2info(inode); unsigned long seg = 0; @@ -392,9 +391,9 @@ static ssize_t ll_direct_IO_26(int rw, struct kiocb *iocb, MAX_DIO_SIZE >> PAGE_CACHE_SHIFT); /* Check that all user buffers are aligned as well */ - for (seg = 0; seg < nr_segs; seg++) { - if (((unsigned long)iov[seg].iov_base & ~CFS_PAGE_MASK) || - (iov[seg].iov_len & ~CFS_PAGE_MASK)) + for (seg = 0; seg < iter->nr_segs; seg++) { + if (((unsigned long)iter->iov[seg].iov_base & ~CFS_PAGE_MASK) || + (iter->iov[seg].iov_len & ~CFS_PAGE_MASK)) return -EINVAL; } @@ -411,9 +410,9 @@ static ssize_t ll_direct_IO_26(int rw, struct kiocb *iocb, mutex_lock(&inode->i_mutex); LASSERT(obj->cob_transient_pages == 0); - for (seg = 0; seg < nr_segs; seg++) { - long iov_left = iov[seg].iov_len; - unsigned long user_addr = (unsigned long)iov[seg].iov_base; + for (seg = 0; seg < iter->nr_segs; seg++) { + long iov_left = iter->iov[seg].iov_len; + unsigned long user_addr = (unsigned long)iter->iov[seg].iov_base; if (rw == READ) { if (file_offset >= i_size_read(inode)) diff --git a/fs/9p/vfs_addr.c b/fs/9p/vfs_addr.c index c71e88602ff..cc1cfae726b 100644 --- a/fs/9p/vfs_addr.c +++ b/fs/9p/vfs_addr.c @@ -259,8 +259,7 @@ static int v9fs_launder_page(struct page *page) * */ static ssize_t -v9fs_direct_IO(int rw, struct kiocb *iocb, const struct iovec *iov, - loff_t pos, unsigned long nr_segs) +v9fs_direct_IO(int rw, struct kiocb *iocb, struct iov_iter *iter, loff_t pos) { /* * FIXME @@ -269,7 +268,7 @@ v9fs_direct_IO(int rw, struct kiocb *iocb, const struct iovec *iov, */ p9_debug(P9_DEBUG_VFS, "v9fs_direct_IO: v9fs_direct_IO (%s) off/no(%lld/%lu) EINVAL\n", iocb->ki_filp->f_path.dentry->d_name.name, - (long long)pos, nr_segs); + (long long)pos, iter->nr_segs); return -EINVAL; } diff --git a/fs/block_dev.c b/fs/block_dev.c index 552a8d13bc3..938fc707d76 100644 --- a/fs/block_dev.c +++ b/fs/block_dev.c @@ -165,14 +165,15 @@ blkdev_get_block(struct inode *inode, sector_t iblock, } static ssize_t -blkdev_direct_IO(int rw, struct kiocb *iocb, const struct iovec *iov, - loff_t offset, unsigned long nr_segs) +blkdev_direct_IO(int rw, struct kiocb *iocb, struct iov_iter *iter, + loff_t offset) { struct file *file = iocb->ki_filp; struct inode *inode = file->f_mapping->host; - return __blockdev_direct_IO(rw, iocb, inode, I_BDEV(inode), iov, offset, - nr_segs, blkdev_get_block, NULL, NULL, 0); + return __blockdev_direct_IO(rw, iocb, inode, I_BDEV(inode), iter->iov, + offset, iter->nr_segs, blkdev_get_block, + NULL, NULL, 0); } int __sync_blockdev(struct block_device *bdev, int wait) diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index 5f805bc944f..30a6cc51f32 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -7433,8 +7433,7 @@ out: } static ssize_t btrfs_direct_IO(int rw, struct kiocb *iocb, - const struct iovec *iov, loff_t offset, - unsigned long nr_segs) + struct iov_iter *iter, loff_t offset) { struct file *file = iocb->ki_filp; struct inode *inode = file->f_mapping->host; @@ -7444,8 +7443,8 @@ static ssize_t btrfs_direct_IO(int rw, struct kiocb *iocb, bool relock = false; ssize_t ret; - if (check_direct_IO(BTRFS_I(inode)->root, rw, iocb, iov, - offset, nr_segs)) + if (check_direct_IO(BTRFS_I(inode)->root, rw, iocb, iter->iov, + offset, iter->nr_segs)) return 0; atomic_inc(&inode->i_dio_count); @@ -7457,7 +7456,7 @@ static ssize_t btrfs_direct_IO(int rw, struct kiocb *iocb, * we need to flush the dirty pages again to make absolutely sure * that any outstanding dirty pages are on disk. */ - count = iov_length(iov, nr_segs); + count = iov_length(iter->iov, iter->nr_segs); if (test_bit(BTRFS_INODE_HAS_ASYNC_EXTENT, &BTRFS_I(inode)->runtime_flags)) filemap_fdatawrite_range(inode->i_mapping, offset, count); @@ -7484,7 +7483,8 @@ static ssize_t btrfs_direct_IO(int rw, struct kiocb *iocb, ret = __blockdev_direct_IO(rw, iocb, inode, BTRFS_I(inode)->root->fs_info->fs_devices->latest_bdev, - iov, offset, nr_segs, btrfs_get_blocks_direct, NULL, + iter->iov, offset, iter->nr_segs, + btrfs_get_blocks_direct, NULL, btrfs_submit_direct, flags); if (rw & WRITE) { if (ret < 0 && ret != -EIOCBQUEUED) diff --git a/fs/ceph/addr.c b/fs/ceph/addr.c index b53278c9fd9..342ca5e423f 100644 --- a/fs/ceph/addr.c +++ b/fs/ceph/addr.c @@ -1187,8 +1187,8 @@ static int ceph_write_end(struct file *file, struct address_space *mapping, * never get called. */ static ssize_t ceph_direct_io(int rw, struct kiocb *iocb, - const struct iovec *iov, - loff_t pos, unsigned long nr_segs) + struct iov_iter *iter, + loff_t pos) { WARN_ON(1); return -EINVAL; diff --git a/fs/cifs/file.c b/fs/cifs/file.c index 2900d150654..a4ccc39e6c1 100644 --- a/fs/cifs/file.c +++ b/fs/cifs/file.c @@ -3702,8 +3702,8 @@ void cifs_oplock_break(struct work_struct *work) * Direct IO is not yet supported in the cached mode. */ static ssize_t -cifs_direct_io(int rw, struct kiocb *iocb, const struct iovec *iov, - loff_t pos, unsigned long nr_segs) +cifs_direct_io(int rw, struct kiocb *iocb, struct iov_iter *iter, + loff_t pos) { /* * FIXME diff --git a/fs/exofs/inode.c b/fs/exofs/inode.c index d1c244d6766..3f9cafd7393 100644 --- a/fs/exofs/inode.c +++ b/fs/exofs/inode.c @@ -964,7 +964,7 @@ static void exofs_invalidatepage(struct page *page, unsigned int offset, /* TODO: Should be easy enough to do proprly */ static ssize_t exofs_direct_IO(int rw, struct kiocb *iocb, - const struct iovec *iov, loff_t offset, unsigned long nr_segs) + struct iov_iter *iter, loff_t offset) { return 0; } diff --git a/fs/ext2/inode.c b/fs/ext2/inode.c index b1d2a4675d4..47fbe760a7f 100644 --- a/fs/ext2/inode.c +++ b/fs/ext2/inode.c @@ -850,18 +850,19 @@ static sector_t ext2_bmap(struct address_space *mapping, sector_t block) } static ssize_t -ext2_direct_IO(int rw, struct kiocb *iocb, const struct iovec *iov, - loff_t offset, unsigned long nr_segs) +ext2_direct_IO(int rw, struct kiocb *iocb, struct iov_iter *iter, + loff_t offset) { struct file *file = iocb->ki_filp; struct address_space *mapping = file->f_mapping; struct inode *inode = mapping->host; ssize_t ret; - ret = blockdev_direct_IO(rw, iocb, inode, iov, offset, nr_segs, - ext2_get_block); + ret = blockdev_direct_IO(rw, iocb, inode, iter->iov, offset, + iter->nr_segs, ext2_get_block); if (ret < 0 && (rw & WRITE)) - ext2_write_failed(mapping, offset + iov_length(iov, nr_segs)); + ext2_write_failed(mapping, offset + + iov_length(iter->iov, iter->nr_segs)); return ret; } diff --git a/fs/ext3/inode.c b/fs/ext3/inode.c index f5157d0d1b4..7a5c501dc31 100644 --- a/fs/ext3/inode.c +++ b/fs/ext3/inode.c @@ -1820,8 +1820,7 @@ static int ext3_releasepage(struct page *page, gfp_t wait) * VFS code falls back into buffered path in that case so we are safe. */ static ssize_t ext3_direct_IO(int rw, struct kiocb *iocb, - const struct iovec *iov, loff_t offset, - unsigned long nr_segs) + struct iov_iter *iter, loff_t offset) { struct file *file = iocb->ki_filp; struct inode *inode = file->f_mapping->host; @@ -1829,10 +1828,10 @@ static ssize_t ext3_direct_IO(int rw, struct kiocb *iocb, handle_t *handle; ssize_t ret; int orphan = 0; - size_t count = iov_length(iov, nr_segs); + size_t count = iov_length(iter->iov, iter->nr_segs); int retries = 0; - trace_ext3_direct_IO_enter(inode, offset, iov_length(iov, nr_segs), rw); + trace_ext3_direct_IO_enter(inode, offset, count, rw); if (rw == WRITE) { loff_t final_size = offset + count; @@ -1856,15 +1855,15 @@ static ssize_t ext3_direct_IO(int rw, struct kiocb *iocb, } retry: - ret = blockdev_direct_IO(rw, iocb, inode, iov, offset, nr_segs, - ext3_get_block); + ret = blockdev_direct_IO(rw, iocb, inode, iter->iov, offset, + iter->nr_segs, ext3_get_block); /* * In case of error extending write may have instantiated a few * blocks outside i_size. Trim these off again. */ if (unlikely((rw & WRITE) && ret < 0)) { loff_t isize = i_size_read(inode); - loff_t end = offset + iov_length(iov, nr_segs); + loff_t end = offset + count; if (end > isize) ext3_truncate_failed_direct_write(inode); @@ -1909,8 +1908,7 @@ retry: ret = err; } out: - trace_ext3_direct_IO_exit(inode, offset, - iov_length(iov, nr_segs), rw, ret); + trace_ext3_direct_IO_exit(inode, offset, count, rw, ret); return ret; } diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c index d7b7462a0e1..f51db730da3 100644 --- a/fs/ext4/inode.c +++ b/fs/ext4/inode.c @@ -3222,8 +3222,7 @@ retake_lock: } static ssize_t ext4_direct_IO(int rw, struct kiocb *iocb, - const struct iovec *iov, loff_t offset, - unsigned long nr_segs) + struct iov_iter *iter, loff_t offset) { struct file *file = iocb->ki_filp; struct inode *inode = file->f_mapping->host; @@ -3239,13 +3238,13 @@ static ssize_t ext4_direct_IO(int rw, struct kiocb *iocb, if (ext4_has_inline_data(inode)) return 0; - trace_ext4_direct_IO_enter(inode, offset, iov_length(iov, nr_segs), rw); + trace_ext4_direct_IO_enter(inode, offset, iov_length(iter->iov, iter->nr_segs), rw); if (ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS)) - ret = ext4_ext_direct_IO(rw, iocb, iov, offset, nr_segs); + ret = ext4_ext_direct_IO(rw, iocb, iter->iov, offset, iter->nr_segs); else - ret = ext4_ind_direct_IO(rw, iocb, iov, offset, nr_segs); + ret = ext4_ind_direct_IO(rw, iocb, iter->iov, offset, iter->nr_segs); trace_ext4_direct_IO_exit(inode, offset, - iov_length(iov, nr_segs), rw, ret); + iov_length(iter->iov, iter->nr_segs), rw, ret); return ret; } diff --git a/fs/f2fs/data.c b/fs/f2fs/data.c index 45abd60e2bf..3a6ef121c09 100644 --- a/fs/f2fs/data.c +++ b/fs/f2fs/data.c @@ -1010,7 +1010,7 @@ static int check_direct_IO(struct inode *inode, int rw, } static ssize_t f2fs_direct_IO(int rw, struct kiocb *iocb, - const struct iovec *iov, loff_t offset, unsigned long nr_segs) + struct iov_iter *iter, loff_t offset) { struct file *file = iocb->ki_filp; struct inode *inode = file->f_mapping->host; @@ -1019,11 +1019,11 @@ static ssize_t f2fs_direct_IO(int rw, struct kiocb *iocb, if (f2fs_has_inline_data(inode)) return 0; - if (check_direct_IO(inode, rw, iov, offset, nr_segs)) + if (check_direct_IO(inode, rw, iter->iov, offset, iter->nr_segs)) return 0; - return blockdev_direct_IO(rw, iocb, inode, iov, offset, nr_segs, - get_data_block); + return blockdev_direct_IO(rw, iocb, inode, iter->iov, offset, + iter->nr_segs, get_data_block); } static void f2fs_invalidate_data_page(struct page *page, unsigned int offset, diff --git a/fs/fat/inode.c b/fs/fat/inode.c index b3361fe2bcb..d5237a19905 100644 --- a/fs/fat/inode.c +++ b/fs/fat/inode.c @@ -185,8 +185,8 @@ static int fat_write_end(struct file *file, struct address_space *mapping, } static ssize_t fat_direct_IO(int rw, struct kiocb *iocb, - const struct iovec *iov, - loff_t offset, unsigned long nr_segs) + struct iov_iter *iter, + loff_t offset) { struct file *file = iocb->ki_filp; struct address_space *mapping = file->f_mapping; @@ -203,7 +203,7 @@ static ssize_t fat_direct_IO(int rw, struct kiocb *iocb, * * Return 0, and fallback to normal buffered write. */ - loff_t size = offset + iov_length(iov, nr_segs); + loff_t size = offset + iov_length(iter->iov, iter->nr_segs); if (MSDOS_I(inode)->mmu_private < size) return 0; } @@ -212,10 +212,11 @@ static ssize_t fat_direct_IO(int rw, struct kiocb *iocb, * FAT need to use the DIO_LOCKING for avoiding the race * condition of fat_get_block() and ->truncate(). */ - ret = blockdev_direct_IO(rw, iocb, inode, iov, offset, nr_segs, - fat_get_block); + ret = blockdev_direct_IO(rw, iocb, inode, iter->iov, offset, + iter->nr_segs, fat_get_block); if (ret < 0 && (rw & WRITE)) - fat_write_failed(mapping, offset + iov_length(iov, nr_segs)); + fat_write_failed(mapping, offset + + iov_length(iter->iov, iter->nr_segs)); return ret; } diff --git a/fs/fuse/file.c b/fs/fuse/file.c index 9c7f346879e..17d96f36df1 100644 --- a/fs/fuse/file.c +++ b/fs/fuse/file.c @@ -2890,8 +2890,8 @@ static inline loff_t fuse_round_up(loff_t off) } static ssize_t -fuse_direct_IO(int rw, struct kiocb *iocb, const struct iovec *iov, - loff_t offset, unsigned long nr_segs) +fuse_direct_IO(int rw, struct kiocb *iocb, struct iov_iter *iter, + loff_t offset) { ssize_t ret = 0; struct file *file = iocb->ki_filp; @@ -2900,7 +2900,7 @@ fuse_direct_IO(int rw, struct kiocb *iocb, const struct iovec *iov, loff_t pos = 0; struct inode *inode; loff_t i_size; - size_t count = iov_length(iov, nr_segs); + size_t count = iov_length(iter->iov, iter->nr_segs); struct fuse_io_priv *io; pos = offset; @@ -2944,9 +2944,9 @@ fuse_direct_IO(int rw, struct kiocb *iocb, const struct iovec *iov, io->async = false; if (rw == WRITE) - ret = __fuse_direct_write(io, iov, nr_segs, &pos); + ret = __fuse_direct_write(io, iter->iov, iter->nr_segs, &pos); else - ret = __fuse_direct_read(io, iov, nr_segs, &pos, count); + ret = __fuse_direct_read(io, iter->iov, iter->nr_segs, &pos, count); if (io->async) { fuse_aio_complete(io, ret < 0 ? ret : 0, -1); diff --git a/fs/gfs2/aops.c b/fs/gfs2/aops.c index ce62dcac90b..e84ddaa4210 100644 --- a/fs/gfs2/aops.c +++ b/fs/gfs2/aops.c @@ -1041,8 +1041,7 @@ static int gfs2_ok_for_dio(struct gfs2_inode *ip, int rw, loff_t offset) static ssize_t gfs2_direct_IO(int rw, struct kiocb *iocb, - const struct iovec *iov, loff_t offset, - unsigned long nr_segs) + struct iov_iter *iter, loff_t offset) { struct file *file = iocb->ki_filp; struct inode *inode = file->f_mapping->host; @@ -1082,7 +1081,7 @@ static ssize_t gfs2_direct_IO(int rw, struct kiocb *iocb, */ if (mapping->nrpages) { loff_t lstart = offset & (PAGE_CACHE_SIZE - 1); - loff_t len = iov_length(iov, nr_segs); + loff_t len = iov_length(iter->iov, iter->nr_segs); loff_t end = PAGE_ALIGN(offset + len) - 1; rv = 0; @@ -1097,9 +1096,9 @@ static ssize_t gfs2_direct_IO(int rw, struct kiocb *iocb, truncate_inode_pages_range(mapping, lstart, end); } - rv = __blockdev_direct_IO(rw, iocb, inode, inode->i_sb->s_bdev, iov, - offset, nr_segs, gfs2_get_block_direct, - NULL, NULL, 0); + rv = __blockdev_direct_IO(rw, iocb, inode, inode->i_sb->s_bdev, + iter->iov, offset, iter->nr_segs, + gfs2_get_block_direct, NULL, NULL, 0); out: gfs2_glock_dq(&gh); gfs2_holder_uninit(&gh); diff --git a/fs/hfs/inode.c b/fs/hfs/inode.c index 9e2fecd62f6..09cff13528c 100644 --- a/fs/hfs/inode.c +++ b/fs/hfs/inode.c @@ -125,15 +125,15 @@ static int hfs_releasepage(struct page *page, gfp_t mask) } static ssize_t hfs_direct_IO(int rw, struct kiocb *iocb, - const struct iovec *iov, loff_t offset, unsigned long nr_segs) + struct iov_iter *iter, loff_t offset) { struct file *file = iocb->ki_filp; struct address_space *mapping = file->f_mapping; struct inode *inode = file_inode(file)->i_mapping->host; ssize_t ret; - ret = blockdev_direct_IO(rw, iocb, inode, iov, offset, nr_segs, - hfs_get_block); + ret = blockdev_direct_IO(rw, iocb, inode, iter->iov, offset, + iter->nr_segs, hfs_get_block); /* * In case of error extending write may have instantiated a few @@ -141,7 +141,7 @@ static ssize_t hfs_direct_IO(int rw, struct kiocb *iocb, */ if (unlikely((rw & WRITE) && ret < 0)) { loff_t isize = i_size_read(inode); - loff_t end = offset + iov_length(iov, nr_segs); + loff_t end = offset + iov_length(iter->iov, iter->nr_segs); if (end > isize) hfs_write_failed(mapping, end); diff --git a/fs/hfsplus/inode.c b/fs/hfsplus/inode.c index a4f45bd88a6..7f894a5b5ea 100644 --- a/fs/hfsplus/inode.c +++ b/fs/hfsplus/inode.c @@ -123,14 +123,14 @@ static int hfsplus_releasepage(struct page *page, gfp_t mask) } static ssize_t hfsplus_direct_IO(int rw, struct kiocb *iocb, - const struct iovec *iov, loff_t offset, unsigned long nr_segs) + struct iov_iter *iter, loff_t offset) { struct file *file = iocb->ki_filp; struct address_space *mapping = file->f_mapping; struct inode *inode = file_inode(file)->i_mapping->host; ssize_t ret; - ret = blockdev_direct_IO(rw, iocb, inode, iov, offset, nr_segs, + ret = blockdev_direct_IO(rw, iocb, inode, iter->iov, offset, iter->nr_segs, hfsplus_get_block); /* @@ -139,7 +139,7 @@ static ssize_t hfsplus_direct_IO(int rw, struct kiocb *iocb, */ if (unlikely((rw & WRITE) && ret < 0)) { loff_t isize = i_size_read(inode); - loff_t end = offset + iov_length(iov, nr_segs); + loff_t end = offset + iov_length(iter->iov, iter->nr_segs); if (end > isize) hfsplus_write_failed(mapping, end); diff --git a/fs/jfs/inode.c b/fs/jfs/inode.c index 6f8fe72c2a7..7052744d510 100644 --- a/fs/jfs/inode.c +++ b/fs/jfs/inode.c @@ -331,15 +331,15 @@ static sector_t jfs_bmap(struct address_space *mapping, sector_t block) } static ssize_t jfs_direct_IO(int rw, struct kiocb *iocb, - const struct iovec *iov, loff_t offset, unsigned long nr_segs) + struct iov_iter *iter, loff_t offset) { struct file *file = iocb->ki_filp; struct address_space *mapping = file->f_mapping; struct inode *inode = file->f_mapping->host; ssize_t ret; - ret = blockdev_direct_IO(rw, iocb, inode, iov, offset, nr_segs, - jfs_get_block); + ret = blockdev_direct_IO(rw, iocb, inode, iter->iov, offset, + iter->nr_segs, jfs_get_block); /* * In case of error extending write may have instantiated a few @@ -347,7 +347,7 @@ static ssize_t jfs_direct_IO(int rw, struct kiocb *iocb, */ if (unlikely((rw & WRITE) && ret < 0)) { loff_t isize = i_size_read(inode); - loff_t end = offset + iov_length(iov, nr_segs); + loff_t end = offset + iov_length(iter->iov, iter->nr_segs); if (end > isize) jfs_write_failed(mapping, end); diff --git a/fs/nfs/direct.c b/fs/nfs/direct.c index b8797ae6831..e9cde393500 100644 --- a/fs/nfs/direct.c +++ b/fs/nfs/direct.c @@ -121,20 +121,20 @@ static inline int put_dreq(struct nfs_direct_req *dreq) * shunt off direct read and write requests before the VFS gets them, * so this method is only ever called for swap. */ -ssize_t nfs_direct_IO(int rw, struct kiocb *iocb, const struct iovec *iov, loff_t pos, unsigned long nr_segs) +ssize_t nfs_direct_IO(int rw, struct kiocb *iocb, struct iov_iter *iter, loff_t pos) { #ifndef CONFIG_NFS_SWAP dprintk("NFS: nfs_direct_IO (%pD) off/no(%Ld/%lu) EINVAL\n", - iocb->ki_filp, (long long) pos, nr_segs); + iocb->ki_filp, (long long) pos, iter->nr_segs); return -EINVAL; #else VM_BUG_ON(iocb->ki_nbytes != PAGE_SIZE); if (rw == READ || rw == KERNEL_READ) - return nfs_file_direct_read(iocb, iov, nr_segs, pos, + return nfs_file_direct_read(iocb, iter->iov, iter->nr_segs, pos, rw == READ ? true : false); - return nfs_file_direct_write(iocb, iov, nr_segs, pos, + return nfs_file_direct_write(iocb, iter->iov, iter->nr_segs, pos, rw == WRITE ? true : false); #endif /* CONFIG_NFS_SWAP */ } diff --git a/fs/nilfs2/inode.c b/fs/nilfs2/inode.c index b9c5726120e..1c0e8fedc09 100644 --- a/fs/nilfs2/inode.c +++ b/fs/nilfs2/inode.c @@ -298,8 +298,8 @@ static int nilfs_write_end(struct file *file, struct address_space *mapping, } static ssize_t -nilfs_direct_IO(int rw, struct kiocb *iocb, const struct iovec *iov, - loff_t offset, unsigned long nr_segs) +nilfs_direct_IO(int rw, struct kiocb *iocb, struct iov_iter *iter, + loff_t offset) { struct file *file = iocb->ki_filp; struct address_space *mapping = file->f_mapping; @@ -310,8 +310,8 @@ nilfs_direct_IO(int rw, struct kiocb *iocb, const struct iovec *iov, return 0; /* Needs synchronization with the cleaner */ - size = blockdev_direct_IO(rw, iocb, inode, iov, offset, nr_segs, - nilfs_get_block); + size = blockdev_direct_IO(rw, iocb, inode, iter->iov, offset, + iter->nr_segs, nilfs_get_block); /* * In case of error extending write may have instantiated a few @@ -319,7 +319,7 @@ nilfs_direct_IO(int rw, struct kiocb *iocb, const struct iovec *iov, */ if (unlikely((rw & WRITE) && size < 0)) { loff_t isize = i_size_read(inode); - loff_t end = offset + iov_length(iov, nr_segs); + loff_t end = offset + iov_length(iter->iov, iter->nr_segs); if (end > isize) nilfs_write_failed(mapping, end); diff --git a/fs/ocfs2/aops.c b/fs/ocfs2/aops.c index d310d12a9ad..799fd0afcb3 100644 --- a/fs/ocfs2/aops.c +++ b/fs/ocfs2/aops.c @@ -599,9 +599,8 @@ static int ocfs2_releasepage(struct page *page, gfp_t wait) static ssize_t ocfs2_direct_IO(int rw, struct kiocb *iocb, - const struct iovec *iov, - loff_t offset, - unsigned long nr_segs) + struct iov_iter *iter, + loff_t offset) { struct file *file = iocb->ki_filp; struct inode *inode = file_inode(file)->i_mapping->host; @@ -618,7 +617,7 @@ static ssize_t ocfs2_direct_IO(int rw, return 0; return __blockdev_direct_IO(rw, iocb, inode, inode->i_sb->s_bdev, - iov, offset, nr_segs, + iter->iov, offset, iter->nr_segs, ocfs2_direct_IO_get_blocks, ocfs2_dio_end_io, NULL, 0); } diff --git a/fs/reiserfs/inode.c b/fs/reiserfs/inode.c index bc8b8009897..17bf4c41a50 100644 --- a/fs/reiserfs/inode.c +++ b/fs/reiserfs/inode.c @@ -3083,15 +3083,14 @@ static int reiserfs_releasepage(struct page *page, gfp_t unused_gfp_flags) /* We thank Mingming Cao for helping us understand in great detail what to do in this section of the code. */ static ssize_t reiserfs_direct_IO(int rw, struct kiocb *iocb, - const struct iovec *iov, loff_t offset, - unsigned long nr_segs) + struct iov_iter *iter, loff_t offset) { struct file *file = iocb->ki_filp; struct inode *inode = file->f_mapping->host; ssize_t ret; - ret = blockdev_direct_IO(rw, iocb, inode, iov, offset, nr_segs, - reiserfs_get_blocks_direct_io); + ret = blockdev_direct_IO(rw, iocb, inode, iter->iov, offset, + iter->nr_segs, reiserfs_get_blocks_direct_io); /* * In case of error extending write may have instantiated a few @@ -3099,7 +3098,7 @@ static ssize_t reiserfs_direct_IO(int rw, struct kiocb *iocb, */ if (unlikely((rw & WRITE) && ret < 0)) { loff_t isize = i_size_read(inode); - loff_t end = offset + iov_length(iov, nr_segs); + loff_t end = offset + iov_length(iter->iov, iter->nr_segs); if ((end > isize) && inode_newsize_ok(inode, isize) == 0) { truncate_setsize(inode, isize); diff --git a/fs/udf/file.c b/fs/udf/file.c index d2c170f8b03..ade88640165 100644 --- a/fs/udf/file.c +++ b/fs/udf/file.c @@ -119,8 +119,8 @@ static int udf_adinicb_write_end(struct file *file, } static ssize_t udf_adinicb_direct_IO(int rw, struct kiocb *iocb, - const struct iovec *iov, - loff_t offset, unsigned long nr_segs) + struct iov_iter *iter, + loff_t offset) { /* Fallback to buffered I/O. */ return 0; diff --git a/fs/udf/inode.c b/fs/udf/inode.c index 5d643706212..5b184c7f7dc 100644 --- a/fs/udf/inode.c +++ b/fs/udf/inode.c @@ -217,18 +217,18 @@ static int udf_write_begin(struct file *file, struct address_space *mapping, } static ssize_t udf_direct_IO(int rw, struct kiocb *iocb, - const struct iovec *iov, - loff_t offset, unsigned long nr_segs) + struct iov_iter *iter, + loff_t offset) { struct file *file = iocb->ki_filp; struct address_space *mapping = file->f_mapping; struct inode *inode = mapping->host; ssize_t ret; - ret = blockdev_direct_IO(rw, iocb, inode, iov, offset, nr_segs, + ret = blockdev_direct_IO(rw, iocb, inode, iter->iov, offset, iter->nr_segs, udf_get_block); if (unlikely(ret < 0 && (rw & WRITE))) - udf_write_failed(mapping, offset + iov_length(iov, nr_segs)); + udf_write_failed(mapping, offset + iov_length(iter->iov, iter->nr_segs)); return ret; } diff --git a/fs/xfs/xfs_aops.c b/fs/xfs/xfs_aops.c index 0479c32c5eb..330d7b1c4be 100644 --- a/fs/xfs/xfs_aops.c +++ b/fs/xfs/xfs_aops.c @@ -1449,9 +1449,8 @@ STATIC ssize_t xfs_vm_direct_IO( int rw, struct kiocb *iocb, - const struct iovec *iov, - loff_t offset, - unsigned long nr_segs) + struct iov_iter *iter, + loff_t offset) { struct inode *inode = iocb->ki_filp->f_mapping->host; struct block_device *bdev = xfs_find_bdev_for_inode(inode); @@ -1459,7 +1458,7 @@ xfs_vm_direct_IO( ssize_t ret; if (rw & WRITE) { - size_t size = iov_length(iov, nr_segs); + size_t size = iov_length(iter->iov, iter->nr_segs); /* * We cannot preallocate a size update transaction here as we @@ -1471,16 +1470,16 @@ xfs_vm_direct_IO( if (offset + size > XFS_I(inode)->i_d.di_size) ioend->io_isdirect = 1; - ret = __blockdev_direct_IO(rw, iocb, inode, bdev, iov, - offset, nr_segs, + ret = __blockdev_direct_IO(rw, iocb, inode, bdev, iter->iov, + offset, iter->nr_segs, xfs_get_blocks_direct, xfs_end_io_direct_write, NULL, DIO_ASYNC_EXTEND); if (ret != -EIOCBQUEUED && iocb->private) goto out_destroy_ioend; } else { - ret = __blockdev_direct_IO(rw, iocb, inode, bdev, iov, - offset, nr_segs, + ret = __blockdev_direct_IO(rw, iocb, inode, bdev, iter->iov, + offset, iter->nr_segs, xfs_get_blocks_direct, NULL, NULL, 0); } diff --git a/include/linux/fs.h b/include/linux/fs.h index 796de742fe4..399a338c92b 100644 --- a/include/linux/fs.h +++ b/include/linux/fs.h @@ -343,8 +343,7 @@ struct address_space_operations { void (*invalidatepage) (struct page *, unsigned int, unsigned int); int (*releasepage) (struct page *, gfp_t); void (*freepage)(struct page *); - ssize_t (*direct_IO)(int, struct kiocb *, const struct iovec *iov, - loff_t offset, unsigned long nr_segs); + ssize_t (*direct_IO)(int, struct kiocb *, struct iov_iter *iter, loff_t offset); int (*get_xip_mem)(struct address_space *, pgoff_t, int, void **, unsigned long *); /* diff --git a/include/linux/nfs_fs.h b/include/linux/nfs_fs.h index fa6918b0f82..5a0d78ec739 100644 --- a/include/linux/nfs_fs.h +++ b/include/linux/nfs_fs.h @@ -459,8 +459,7 @@ extern int nfs3_removexattr (struct dentry *, const char *name); /* * linux/fs/nfs/direct.c */ -extern ssize_t nfs_direct_IO(int, struct kiocb *, const struct iovec *, loff_t, - unsigned long); +extern ssize_t nfs_direct_IO(int, struct kiocb *, struct iov_iter *, loff_t); extern ssize_t nfs_file_direct_read(struct kiocb *iocb, const struct iovec *iov, unsigned long nr_segs, loff_t pos, bool uio); diff --git a/mm/filemap.c b/mm/filemap.c index 7c1417b0bd7..139641274f1 100644 --- a/mm/filemap.c +++ b/mm/filemap.c @@ -1699,10 +1699,9 @@ generic_file_aio_read(struct kiocb *iocb, const struct iovec *iov, size = i_size_read(inode); retval = filemap_write_and_wait_range(mapping, pos, pos + iov_length(iov, nr_segs) - 1); - if (!retval) { - retval = mapping->a_ops->direct_IO(READ, iocb, - iov, pos, nr_segs); - } + if (!retval) + retval = mapping->a_ops->direct_IO(READ, iocb, &i, pos); + if (retval > 0) { *ppos = pos + retval; count -= retval; @@ -2383,7 +2382,7 @@ generic_file_direct_write(struct kiocb *iocb, struct iov_iter *from, } } - written = mapping->a_ops->direct_IO(WRITE, iocb, from->iov, pos, from->nr_segs); + written = mapping->a_ops->direct_IO(WRITE, iocb, from, pos); /* * Finally, try again to invalidate clean pages which might have been diff --git a/mm/page_io.c b/mm/page_io.c index 7c59ef68138..0ed0644c73d 100644 --- a/mm/page_io.c +++ b/mm/page_io.c @@ -263,16 +263,18 @@ int __swap_writepage(struct page *page, struct writeback_control *wbc, .iov_base = kmap(page), .iov_len = PAGE_SIZE, }; + struct iov_iter from; init_sync_kiocb(&kiocb, swap_file); kiocb.ki_pos = page_file_offset(page); kiocb.ki_nbytes = PAGE_SIZE; + iov_iter_init(&from, &iov, 1, PAGE_SIZE, 0); set_page_writeback(page); unlock_page(page); ret = mapping->a_ops->direct_IO(KERNEL_WRITE, - &kiocb, &iov, - kiocb.ki_pos, 1); + &kiocb, &from, + kiocb.ki_pos); kunmap(page); if (ret == PAGE_SIZE) { count_vm_event(PSWPOUT); -- cgit v1.2.3-70-g09d2 From a6cbcd4a4a85e2fdb0b3344b88df2e8b3d526b9e Mon Sep 17 00:00:00 2001 From: Al Viro Date: Tue, 4 Mar 2014 22:38:00 -0500 Subject: get rid of pointless iov_length() in ->direct_IO() all callers have iov_length(iter->iov, iter->nr_segs) == iov_iter_count(iter) Signed-off-by: Al Viro --- drivers/staging/lustre/lustre/llite/rw26.c | 2 +- fs/btrfs/inode.c | 2 +- fs/ext2/inode.c | 4 ++-- fs/ext3/inode.c | 2 +- fs/ext4/indirect.c | 2 +- fs/ext4/inode.c | 8 ++++---- fs/fat/inode.c | 6 +++--- fs/fuse/file.c | 2 +- fs/gfs2/aops.c | 2 +- fs/hfs/inode.c | 3 ++- fs/hfsplus/inode.c | 3 ++- fs/jfs/inode.c | 3 ++- fs/nfs/direct.c | 10 +++------- fs/nilfs2/inode.c | 3 ++- fs/reiserfs/inode.c | 3 ++- fs/udf/inode.c | 3 ++- fs/xfs/xfs_aops.c | 2 +- mm/filemap.c | 2 +- 18 files changed, 32 insertions(+), 30 deletions(-) (limited to 'fs/btrfs/inode.c') diff --git a/drivers/staging/lustre/lustre/llite/rw26.c b/drivers/staging/lustre/lustre/llite/rw26.c index 66e05c6f4d2..38a5b580e7f 100644 --- a/drivers/staging/lustre/lustre/llite/rw26.c +++ b/drivers/staging/lustre/lustre/llite/rw26.c @@ -370,7 +370,7 @@ static ssize_t ll_direct_IO_26(int rw, struct kiocb *iocb, struct file *file = iocb->ki_filp; struct inode *inode = file->f_mapping->host; struct ccc_object *obj = cl_inode2ccc(inode); - long count = iov_length(iter->iov, iter->nr_segs); + long count = iov_iter_count(iter); long tot_bytes = 0, result = 0; struct ll_inode_info *lli = ll_i2info(inode); unsigned long seg = 0; diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index 30a6cc51f32..c46a025d0c4 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -7456,7 +7456,7 @@ static ssize_t btrfs_direct_IO(int rw, struct kiocb *iocb, * we need to flush the dirty pages again to make absolutely sure * that any outstanding dirty pages are on disk. */ - count = iov_length(iter->iov, iter->nr_segs); + count = iov_iter_count(iter); if (test_bit(BTRFS_INODE_HAS_ASYNC_EXTENT, &BTRFS_I(inode)->runtime_flags)) filemap_fdatawrite_range(inode->i_mapping, offset, count); diff --git a/fs/ext2/inode.c b/fs/ext2/inode.c index 47fbe760a7f..116e809aa7c 100644 --- a/fs/ext2/inode.c +++ b/fs/ext2/inode.c @@ -856,13 +856,13 @@ ext2_direct_IO(int rw, struct kiocb *iocb, struct iov_iter *iter, struct file *file = iocb->ki_filp; struct address_space *mapping = file->f_mapping; struct inode *inode = mapping->host; + size_t count = iov_iter_count(iter); ssize_t ret; ret = blockdev_direct_IO(rw, iocb, inode, iter->iov, offset, iter->nr_segs, ext2_get_block); if (ret < 0 && (rw & WRITE)) - ext2_write_failed(mapping, offset + - iov_length(iter->iov, iter->nr_segs)); + ext2_write_failed(mapping, offset + count); return ret; } diff --git a/fs/ext3/inode.c b/fs/ext3/inode.c index 7a5c501dc31..8582ae2c80b 100644 --- a/fs/ext3/inode.c +++ b/fs/ext3/inode.c @@ -1828,7 +1828,7 @@ static ssize_t ext3_direct_IO(int rw, struct kiocb *iocb, handle_t *handle; ssize_t ret; int orphan = 0; - size_t count = iov_length(iter->iov, iter->nr_segs); + size_t count = iov_iter_count(iter); int retries = 0; trace_ext3_direct_IO_enter(inode, offset, count, rw); diff --git a/fs/ext4/indirect.c b/fs/ext4/indirect.c index eb5ae16902d..123898a6af0 100644 --- a/fs/ext4/indirect.c +++ b/fs/ext4/indirect.c @@ -647,7 +647,7 @@ ssize_t ext4_ind_direct_IO(int rw, struct kiocb *iocb, handle_t *handle; ssize_t ret; int orphan = 0; - size_t count = iov_length(iter->iov, iter->nr_segs); + size_t count = iov_iter_count(iter); int retries = 0; if (rw == WRITE) { diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c index 29996c1b673..2b993579a96 100644 --- a/fs/ext4/inode.c +++ b/fs/ext4/inode.c @@ -3090,7 +3090,7 @@ static ssize_t ext4_ext_direct_IO(int rw, struct kiocb *iocb, struct file *file = iocb->ki_filp; struct inode *inode = file->f_mapping->host; ssize_t ret; - size_t count = iov_length(iter->iov, iter->nr_segs); + size_t count = iov_iter_count(iter); int overwrite = 0; get_block_t *get_block_func = NULL; int dio_flags = 0; @@ -3225,6 +3225,7 @@ static ssize_t ext4_direct_IO(int rw, struct kiocb *iocb, { struct file *file = iocb->ki_filp; struct inode *inode = file->f_mapping->host; + size_t count = iov_iter_count(iter); ssize_t ret; /* @@ -3237,13 +3238,12 @@ static ssize_t ext4_direct_IO(int rw, struct kiocb *iocb, if (ext4_has_inline_data(inode)) return 0; - trace_ext4_direct_IO_enter(inode, offset, iov_length(iter->iov, iter->nr_segs), rw); + trace_ext4_direct_IO_enter(inode, offset, count, rw); if (ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS)) ret = ext4_ext_direct_IO(rw, iocb, iter, offset); else ret = ext4_ind_direct_IO(rw, iocb, iter, offset); - trace_ext4_direct_IO_exit(inode, offset, - iov_length(iter->iov, iter->nr_segs), rw, ret); + trace_ext4_direct_IO_exit(inode, offset, count, rw, ret); return ret; } diff --git a/fs/fat/inode.c b/fs/fat/inode.c index d5237a19905..154a6f9d318 100644 --- a/fs/fat/inode.c +++ b/fs/fat/inode.c @@ -191,6 +191,7 @@ static ssize_t fat_direct_IO(int rw, struct kiocb *iocb, struct file *file = iocb->ki_filp; struct address_space *mapping = file->f_mapping; struct inode *inode = mapping->host; + size_t count = iov_iter_count(iter); ssize_t ret; if (rw == WRITE) { @@ -203,7 +204,7 @@ static ssize_t fat_direct_IO(int rw, struct kiocb *iocb, * * Return 0, and fallback to normal buffered write. */ - loff_t size = offset + iov_length(iter->iov, iter->nr_segs); + loff_t size = offset + count; if (MSDOS_I(inode)->mmu_private < size) return 0; } @@ -215,8 +216,7 @@ static ssize_t fat_direct_IO(int rw, struct kiocb *iocb, ret = blockdev_direct_IO(rw, iocb, inode, iter->iov, offset, iter->nr_segs, fat_get_block); if (ret < 0 && (rw & WRITE)) - fat_write_failed(mapping, offset + - iov_length(iter->iov, iter->nr_segs)); + fat_write_failed(mapping, offset + count); return ret; } diff --git a/fs/fuse/file.c b/fs/fuse/file.c index 17d96f36df1..36f15f9c1de 100644 --- a/fs/fuse/file.c +++ b/fs/fuse/file.c @@ -2900,7 +2900,7 @@ fuse_direct_IO(int rw, struct kiocb *iocb, struct iov_iter *iter, loff_t pos = 0; struct inode *inode; loff_t i_size; - size_t count = iov_length(iter->iov, iter->nr_segs); + size_t count = iov_iter_count(iter); struct fuse_io_priv *io; pos = offset; diff --git a/fs/gfs2/aops.c b/fs/gfs2/aops.c index e84ddaa4210..228a12d2afa 100644 --- a/fs/gfs2/aops.c +++ b/fs/gfs2/aops.c @@ -1081,7 +1081,7 @@ static ssize_t gfs2_direct_IO(int rw, struct kiocb *iocb, */ if (mapping->nrpages) { loff_t lstart = offset & (PAGE_CACHE_SIZE - 1); - loff_t len = iov_length(iter->iov, iter->nr_segs); + loff_t len = iov_iter_count(iter); loff_t end = PAGE_ALIGN(offset + len) - 1; rv = 0; diff --git a/fs/hfs/inode.c b/fs/hfs/inode.c index 09cff13528c..dc69e8f3158 100644 --- a/fs/hfs/inode.c +++ b/fs/hfs/inode.c @@ -130,6 +130,7 @@ static ssize_t hfs_direct_IO(int rw, struct kiocb *iocb, struct file *file = iocb->ki_filp; struct address_space *mapping = file->f_mapping; struct inode *inode = file_inode(file)->i_mapping->host; + size_t count = iov_iter_count(iter); ssize_t ret; ret = blockdev_direct_IO(rw, iocb, inode, iter->iov, offset, @@ -141,7 +142,7 @@ static ssize_t hfs_direct_IO(int rw, struct kiocb *iocb, */ if (unlikely((rw & WRITE) && ret < 0)) { loff_t isize = i_size_read(inode); - loff_t end = offset + iov_length(iter->iov, iter->nr_segs); + loff_t end = offset + count; if (end > isize) hfs_write_failed(mapping, end); diff --git a/fs/hfsplus/inode.c b/fs/hfsplus/inode.c index 7f894a5b5ea..e6b1251af47 100644 --- a/fs/hfsplus/inode.c +++ b/fs/hfsplus/inode.c @@ -128,6 +128,7 @@ static ssize_t hfsplus_direct_IO(int rw, struct kiocb *iocb, struct file *file = iocb->ki_filp; struct address_space *mapping = file->f_mapping; struct inode *inode = file_inode(file)->i_mapping->host; + size_t count = iov_iter_count(iter); ssize_t ret; ret = blockdev_direct_IO(rw, iocb, inode, iter->iov, offset, iter->nr_segs, @@ -139,7 +140,7 @@ static ssize_t hfsplus_direct_IO(int rw, struct kiocb *iocb, */ if (unlikely((rw & WRITE) && ret < 0)) { loff_t isize = i_size_read(inode); - loff_t end = offset + iov_length(iter->iov, iter->nr_segs); + loff_t end = offset + count; if (end > isize) hfsplus_write_failed(mapping, end); diff --git a/fs/jfs/inode.c b/fs/jfs/inode.c index 7052744d510..6cde5928693 100644 --- a/fs/jfs/inode.c +++ b/fs/jfs/inode.c @@ -336,6 +336,7 @@ static ssize_t jfs_direct_IO(int rw, struct kiocb *iocb, struct file *file = iocb->ki_filp; struct address_space *mapping = file->f_mapping; struct inode *inode = file->f_mapping->host; + size_t count = iov_iter_count(iter); ssize_t ret; ret = blockdev_direct_IO(rw, iocb, inode, iter->iov, offset, @@ -347,7 +348,7 @@ static ssize_t jfs_direct_IO(int rw, struct kiocb *iocb, */ if (unlikely((rw & WRITE) && ret < 0)) { loff_t isize = i_size_read(inode); - loff_t end = offset + iov_length(iter->iov, iter->nr_segs); + loff_t end = offset + count; if (end > isize) jfs_write_failed(mapping, end); diff --git a/fs/nfs/direct.c b/fs/nfs/direct.c index 21723149668..1d34f454989 100644 --- a/fs/nfs/direct.c +++ b/fs/nfs/direct.c @@ -486,9 +486,7 @@ ssize_t nfs_file_direct_read(struct kiocb *iocb, struct iov_iter *iter, struct nfs_direct_req *dreq; struct nfs_lock_context *l_ctx; ssize_t result = -EINVAL; - size_t count; - - count = iov_length(iter->iov, iter->nr_segs); + size_t count = iov_iter_count(iter); nfs_add_stats(mapping->host, NFSIOS_DIRECTREADBYTES, count); dfprintk(FILE, "NFS: direct read(%pD2, %zd@%Ld)\n", @@ -877,7 +875,7 @@ static ssize_t nfs_direct_write_schedule_iovec(struct nfs_direct_req *dreq, get_dreq(dreq); atomic_inc(&inode->i_dio_count); - NFS_I(dreq->inode)->write_io += iov_length(iter->iov, iter->nr_segs); + NFS_I(dreq->inode)->write_io += iov_iter_count(iter); for (seg = 0; seg < iter->nr_segs; seg++) { const struct iovec *vec = &iter->iov[seg]; result = nfs_direct_write_schedule_segment(&desc, vec, pos, uio); @@ -936,9 +934,7 @@ ssize_t nfs_file_direct_write(struct kiocb *iocb, struct iov_iter *iter, struct nfs_direct_req *dreq; struct nfs_lock_context *l_ctx; loff_t end; - size_t count; - - count = iov_length(iter->iov, iter->nr_segs); + size_t count = iov_iter_count(iter); end = (pos + count - 1) >> PAGE_CACHE_SHIFT; nfs_add_stats(mapping->host, NFSIOS_DIRECTWRITTENBYTES, count); diff --git a/fs/nilfs2/inode.c b/fs/nilfs2/inode.c index 1c0e8fedc09..7aaf913e870 100644 --- a/fs/nilfs2/inode.c +++ b/fs/nilfs2/inode.c @@ -304,6 +304,7 @@ nilfs_direct_IO(int rw, struct kiocb *iocb, struct iov_iter *iter, struct file *file = iocb->ki_filp; struct address_space *mapping = file->f_mapping; struct inode *inode = file->f_mapping->host; + size_t count = iov_iter_count(iter); ssize_t size; if (rw == WRITE) @@ -319,7 +320,7 @@ nilfs_direct_IO(int rw, struct kiocb *iocb, struct iov_iter *iter, */ if (unlikely((rw & WRITE) && size < 0)) { loff_t isize = i_size_read(inode); - loff_t end = offset + iov_length(iter->iov, iter->nr_segs); + loff_t end = offset + count; if (end > isize) nilfs_write_failed(mapping, end); diff --git a/fs/reiserfs/inode.c b/fs/reiserfs/inode.c index 17bf4c41a50..723affe921f 100644 --- a/fs/reiserfs/inode.c +++ b/fs/reiserfs/inode.c @@ -3087,6 +3087,7 @@ static ssize_t reiserfs_direct_IO(int rw, struct kiocb *iocb, { struct file *file = iocb->ki_filp; struct inode *inode = file->f_mapping->host; + size_t count = iov_iter_count(iter); ssize_t ret; ret = blockdev_direct_IO(rw, iocb, inode, iter->iov, offset, @@ -3098,7 +3099,7 @@ static ssize_t reiserfs_direct_IO(int rw, struct kiocb *iocb, */ if (unlikely((rw & WRITE) && ret < 0)) { loff_t isize = i_size_read(inode); - loff_t end = offset + iov_length(iter->iov, iter->nr_segs); + loff_t end = offset + count; if ((end > isize) && inode_newsize_ok(inode, isize) == 0) { truncate_setsize(inode, isize); diff --git a/fs/udf/inode.c b/fs/udf/inode.c index 5b184c7f7dc..28984baf619 100644 --- a/fs/udf/inode.c +++ b/fs/udf/inode.c @@ -223,12 +223,13 @@ static ssize_t udf_direct_IO(int rw, struct kiocb *iocb, struct file *file = iocb->ki_filp; struct address_space *mapping = file->f_mapping; struct inode *inode = mapping->host; + size_t count = iov_iter_count(iter); ssize_t ret; ret = blockdev_direct_IO(rw, iocb, inode, iter->iov, offset, iter->nr_segs, udf_get_block); if (unlikely(ret < 0 && (rw & WRITE))) - udf_write_failed(mapping, offset + iov_length(iter->iov, iter->nr_segs)); + udf_write_failed(mapping, offset + count); return ret; } diff --git a/fs/xfs/xfs_aops.c b/fs/xfs/xfs_aops.c index 330d7b1c4be..6462b318678 100644 --- a/fs/xfs/xfs_aops.c +++ b/fs/xfs/xfs_aops.c @@ -1458,7 +1458,7 @@ xfs_vm_direct_IO( ssize_t ret; if (rw & WRITE) { - size_t size = iov_length(iter->iov, iter->nr_segs); + size_t size = iov_iter_count(iter); /* * We cannot preallocate a size update transaction here as we diff --git a/mm/filemap.c b/mm/filemap.c index 139641274f1..70c048ea36e 100644 --- a/mm/filemap.c +++ b/mm/filemap.c @@ -1698,7 +1698,7 @@ generic_file_aio_read(struct kiocb *iocb, const struct iovec *iov, goto out; /* skip atime */ size = i_size_read(inode); retval = filemap_write_and_wait_range(mapping, pos, - pos + iov_length(iov, nr_segs) - 1); + pos + count - 1); if (!retval) retval = mapping->a_ops->direct_IO(READ, iocb, &i, pos); -- cgit v1.2.3-70-g09d2 From 31b140398ce56ab41646eda7f02bcb78d6a4c916 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Wed, 5 Mar 2014 01:33:16 -0500 Subject: switch {__,}blockdev_direct_IO() to iov_iter Signed-off-by: Al Viro --- fs/block_dev.c | 4 ++-- fs/btrfs/inode.c | 3 +-- fs/direct-io.c | 33 ++++++++++++++++----------------- fs/ext2/inode.c | 3 +-- fs/ext3/inode.c | 3 +-- fs/ext4/indirect.c | 7 +++---- fs/ext4/inode.c | 4 ++-- fs/f2fs/data.c | 4 ++-- fs/fat/inode.c | 3 +-- fs/gfs2/aops.c | 2 +- fs/hfs/inode.c | 3 +-- fs/hfsplus/inode.c | 2 +- fs/jfs/inode.c | 3 +-- fs/nilfs2/inode.c | 4 ++-- fs/ocfs2/aops.c | 2 +- fs/reiserfs/inode.c | 4 ++-- fs/udf/inode.c | 3 +-- fs/xfs/xfs_aops.c | 10 ++++------ include/linux/fs.h | 12 ++++++------ 19 files changed, 49 insertions(+), 60 deletions(-) (limited to 'fs/btrfs/inode.c') diff --git a/fs/block_dev.c b/fs/block_dev.c index 938fc707d76..937e3011ed5 100644 --- a/fs/block_dev.c +++ b/fs/block_dev.c @@ -171,8 +171,8 @@ blkdev_direct_IO(int rw, struct kiocb *iocb, struct iov_iter *iter, struct file *file = iocb->ki_filp; struct inode *inode = file->f_mapping->host; - return __blockdev_direct_IO(rw, iocb, inode, I_BDEV(inode), iter->iov, - offset, iter->nr_segs, blkdev_get_block, + return __blockdev_direct_IO(rw, iocb, inode, I_BDEV(inode), iter, + offset, blkdev_get_block, NULL, NULL, 0); } diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index c46a025d0c4..b0b8fa0efba 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -7483,8 +7483,7 @@ static ssize_t btrfs_direct_IO(int rw, struct kiocb *iocb, ret = __blockdev_direct_IO(rw, iocb, inode, BTRFS_I(inode)->root->fs_info->fs_devices->latest_bdev, - iter->iov, offset, iter->nr_segs, - btrfs_get_blocks_direct, NULL, + iter, offset, btrfs_get_blocks_direct, NULL, btrfs_submit_direct, flags); if (rw & WRITE) { if (ret < 0 && ret != -EIOCBQUEUED) diff --git a/fs/direct-io.c b/fs/direct-io.c index 31ba0935e32..1c677899b98 100644 --- a/fs/direct-io.c +++ b/fs/direct-io.c @@ -1107,8 +1107,8 @@ static inline int drop_refcount(struct dio *dio) */ static inline ssize_t do_blockdev_direct_IO(int rw, struct kiocb *iocb, struct inode *inode, - struct block_device *bdev, const struct iovec *iov, loff_t offset, - unsigned long nr_segs, get_block_t get_block, dio_iodone_t end_io, + struct block_device *bdev, struct iov_iter *iter, loff_t offset, + get_block_t get_block, dio_iodone_t end_io, dio_submit_t submit_io, int flags) { int seg; @@ -1143,9 +1143,9 @@ do_blockdev_direct_IO(int rw, struct kiocb *iocb, struct inode *inode, } /* Check the memory alignment. Blocks cannot straddle pages */ - for (seg = 0; seg < nr_segs; seg++) { - addr = (unsigned long)iov[seg].iov_base; - size = iov[seg].iov_len; + for (seg = 0; seg < iter->nr_segs; seg++) { + addr = (unsigned long)iter->iov[seg].iov_base; + size = iter->iov[seg].iov_len; end += size; if (unlikely((addr & blocksize_mask) || (size & blocksize_mask))) { @@ -1256,18 +1256,18 @@ do_blockdev_direct_IO(int rw, struct kiocb *iocb, struct inode *inode, if (unlikely(sdio.blkfactor)) sdio.pages_in_io = 2; - for (seg = 0; seg < nr_segs; seg++) { - user_addr = (unsigned long)iov[seg].iov_base; + for (seg = 0; seg < iter->nr_segs; seg++) { + user_addr = (unsigned long)iter->iov[seg].iov_base; sdio.pages_in_io += - ((user_addr + iov[seg].iov_len + PAGE_SIZE-1) / + ((user_addr + iter->iov[seg].iov_len + PAGE_SIZE-1) / PAGE_SIZE - user_addr / PAGE_SIZE); } blk_start_plug(&plug); - for (seg = 0; seg < nr_segs; seg++) { - user_addr = (unsigned long)iov[seg].iov_base; - sdio.size += bytes = iov[seg].iov_len; + for (seg = 0; seg < iter->nr_segs; seg++) { + user_addr = (unsigned long)iter->iov[seg].iov_base; + sdio.size += bytes = iter->iov[seg].iov_len; /* Index into the first page of the first block */ sdio.first_block_in_page = (user_addr & ~PAGE_MASK) >> blkbits; @@ -1288,7 +1288,7 @@ do_blockdev_direct_IO(int rw, struct kiocb *iocb, struct inode *inode, retval = do_direct_IO(dio, &sdio, &map_bh); - dio->result += iov[seg].iov_len - + dio->result += iter->iov[seg].iov_len - ((sdio.final_block_in_request - sdio.block_in_file) << blkbits); @@ -1365,8 +1365,8 @@ out: ssize_t __blockdev_direct_IO(int rw, struct kiocb *iocb, struct inode *inode, - struct block_device *bdev, const struct iovec *iov, loff_t offset, - unsigned long nr_segs, get_block_t get_block, dio_iodone_t end_io, + struct block_device *bdev, struct iov_iter *iter, loff_t offset, + get_block_t get_block, dio_iodone_t end_io, dio_submit_t submit_io, int flags) { /* @@ -1381,9 +1381,8 @@ __blockdev_direct_IO(int rw, struct kiocb *iocb, struct inode *inode, prefetch(bdev->bd_queue); prefetch((char *)bdev->bd_queue + SMP_CACHE_BYTES); - return do_blockdev_direct_IO(rw, iocb, inode, bdev, iov, offset, - nr_segs, get_block, end_io, - submit_io, flags); + return do_blockdev_direct_IO(rw, iocb, inode, bdev, iter, offset, + get_block, end_io, submit_io, flags); } EXPORT_SYMBOL(__blockdev_direct_IO); diff --git a/fs/ext2/inode.c b/fs/ext2/inode.c index 116e809aa7c..36d35c36311 100644 --- a/fs/ext2/inode.c +++ b/fs/ext2/inode.c @@ -859,8 +859,7 @@ ext2_direct_IO(int rw, struct kiocb *iocb, struct iov_iter *iter, size_t count = iov_iter_count(iter); ssize_t ret; - ret = blockdev_direct_IO(rw, iocb, inode, iter->iov, offset, - iter->nr_segs, ext2_get_block); + ret = blockdev_direct_IO(rw, iocb, inode, iter, offset, ext2_get_block); if (ret < 0 && (rw & WRITE)) ext2_write_failed(mapping, offset + count); return ret; diff --git a/fs/ext3/inode.c b/fs/ext3/inode.c index 8582ae2c80b..4d32133a76c 100644 --- a/fs/ext3/inode.c +++ b/fs/ext3/inode.c @@ -1855,8 +1855,7 @@ static ssize_t ext3_direct_IO(int rw, struct kiocb *iocb, } retry: - ret = blockdev_direct_IO(rw, iocb, inode, iter->iov, offset, - iter->nr_segs, ext3_get_block); + ret = blockdev_direct_IO(rw, iocb, inode, iter, offset, ext3_get_block); /* * In case of error extending write may have instantiated a few * blocks outside i_size. Trim these off again. diff --git a/fs/ext4/indirect.c b/fs/ext4/indirect.c index 123898a6af0..8a57e9fcd1b 100644 --- a/fs/ext4/indirect.c +++ b/fs/ext4/indirect.c @@ -686,14 +686,13 @@ retry: goto locked; } ret = __blockdev_direct_IO(rw, iocb, inode, - inode->i_sb->s_bdev, iter->iov, - offset, iter->nr_segs, + inode->i_sb->s_bdev, iter, offset, ext4_get_block, NULL, NULL, 0); inode_dio_done(inode); } else { locked: - ret = blockdev_direct_IO(rw, iocb, inode, iter->iov, - offset, iter->nr_segs, ext4_get_block); + ret = blockdev_direct_IO(rw, iocb, inode, iter, + offset, ext4_get_block); if (unlikely((rw & WRITE) && ret < 0)) { loff_t isize = i_size_read(inode); diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c index 2b993579a96..e5718385a03 100644 --- a/fs/ext4/inode.c +++ b/fs/ext4/inode.c @@ -3166,8 +3166,8 @@ static ssize_t ext4_ext_direct_IO(int rw, struct kiocb *iocb, dio_flags = DIO_LOCKING; } ret = __blockdev_direct_IO(rw, iocb, inode, - inode->i_sb->s_bdev, iter->iov, - offset, iter->nr_segs, + inode->i_sb->s_bdev, iter, + offset, get_block_func, ext4_end_io_dio, NULL, diff --git a/fs/f2fs/data.c b/fs/f2fs/data.c index 3a6ef121c09..151488f2775 100644 --- a/fs/f2fs/data.c +++ b/fs/f2fs/data.c @@ -1022,8 +1022,8 @@ static ssize_t f2fs_direct_IO(int rw, struct kiocb *iocb, if (check_direct_IO(inode, rw, iter->iov, offset, iter->nr_segs)) return 0; - return blockdev_direct_IO(rw, iocb, inode, iter->iov, offset, - iter->nr_segs, get_data_block); + return blockdev_direct_IO(rw, iocb, inode, iter, offset, + get_data_block); } static void f2fs_invalidate_data_page(struct page *page, unsigned int offset, diff --git a/fs/fat/inode.c b/fs/fat/inode.c index 154a6f9d318..385cce464e8 100644 --- a/fs/fat/inode.c +++ b/fs/fat/inode.c @@ -213,8 +213,7 @@ static ssize_t fat_direct_IO(int rw, struct kiocb *iocb, * FAT need to use the DIO_LOCKING for avoiding the race * condition of fat_get_block() and ->truncate(). */ - ret = blockdev_direct_IO(rw, iocb, inode, iter->iov, offset, - iter->nr_segs, fat_get_block); + ret = blockdev_direct_IO(rw, iocb, inode, iter, offset, fat_get_block); if (ret < 0 && (rw & WRITE)) fat_write_failed(mapping, offset + count); diff --git a/fs/gfs2/aops.c b/fs/gfs2/aops.c index 228a12d2afa..910838951d6 100644 --- a/fs/gfs2/aops.c +++ b/fs/gfs2/aops.c @@ -1097,7 +1097,7 @@ static ssize_t gfs2_direct_IO(int rw, struct kiocb *iocb, } rv = __blockdev_direct_IO(rw, iocb, inode, inode->i_sb->s_bdev, - iter->iov, offset, iter->nr_segs, + iter, offset, gfs2_get_block_direct, NULL, NULL, 0); out: gfs2_glock_dq(&gh); diff --git a/fs/hfs/inode.c b/fs/hfs/inode.c index dc69e8f3158..f5fb09ebc85 100644 --- a/fs/hfs/inode.c +++ b/fs/hfs/inode.c @@ -133,8 +133,7 @@ static ssize_t hfs_direct_IO(int rw, struct kiocb *iocb, size_t count = iov_iter_count(iter); ssize_t ret; - ret = blockdev_direct_IO(rw, iocb, inode, iter->iov, offset, - iter->nr_segs, hfs_get_block); + ret = blockdev_direct_IO(rw, iocb, inode, iter, offset, hfs_get_block); /* * In case of error extending write may have instantiated a few diff --git a/fs/hfsplus/inode.c b/fs/hfsplus/inode.c index e6b1251af47..76b930ff58a 100644 --- a/fs/hfsplus/inode.c +++ b/fs/hfsplus/inode.c @@ -131,7 +131,7 @@ static ssize_t hfsplus_direct_IO(int rw, struct kiocb *iocb, size_t count = iov_iter_count(iter); ssize_t ret; - ret = blockdev_direct_IO(rw, iocb, inode, iter->iov, offset, iter->nr_segs, + ret = blockdev_direct_IO(rw, iocb, inode, iter, offset, hfsplus_get_block); /* diff --git a/fs/jfs/inode.c b/fs/jfs/inode.c index 6cde5928693..bd3df1ca3c9 100644 --- a/fs/jfs/inode.c +++ b/fs/jfs/inode.c @@ -339,8 +339,7 @@ static ssize_t jfs_direct_IO(int rw, struct kiocb *iocb, size_t count = iov_iter_count(iter); ssize_t ret; - ret = blockdev_direct_IO(rw, iocb, inode, iter->iov, offset, - iter->nr_segs, jfs_get_block); + ret = blockdev_direct_IO(rw, iocb, inode, iter, offset, jfs_get_block); /* * In case of error extending write may have instantiated a few diff --git a/fs/nilfs2/inode.c b/fs/nilfs2/inode.c index 7aaf913e870..6252b173a46 100644 --- a/fs/nilfs2/inode.c +++ b/fs/nilfs2/inode.c @@ -311,8 +311,8 @@ nilfs_direct_IO(int rw, struct kiocb *iocb, struct iov_iter *iter, return 0; /* Needs synchronization with the cleaner */ - size = blockdev_direct_IO(rw, iocb, inode, iter->iov, offset, - iter->nr_segs, nilfs_get_block); + size = blockdev_direct_IO(rw, iocb, inode, iter, offset, + nilfs_get_block); /* * In case of error extending write may have instantiated a few diff --git a/fs/ocfs2/aops.c b/fs/ocfs2/aops.c index 799fd0afcb3..4a231a166cf 100644 --- a/fs/ocfs2/aops.c +++ b/fs/ocfs2/aops.c @@ -617,7 +617,7 @@ static ssize_t ocfs2_direct_IO(int rw, return 0; return __blockdev_direct_IO(rw, iocb, inode, inode->i_sb->s_bdev, - iter->iov, offset, iter->nr_segs, + iter, offset, ocfs2_direct_IO_get_blocks, ocfs2_dio_end_io, NULL, 0); } diff --git a/fs/reiserfs/inode.c b/fs/reiserfs/inode.c index 723affe921f..b8003e8dd1f 100644 --- a/fs/reiserfs/inode.c +++ b/fs/reiserfs/inode.c @@ -3090,8 +3090,8 @@ static ssize_t reiserfs_direct_IO(int rw, struct kiocb *iocb, size_t count = iov_iter_count(iter); ssize_t ret; - ret = blockdev_direct_IO(rw, iocb, inode, iter->iov, offset, - iter->nr_segs, reiserfs_get_blocks_direct_io); + ret = blockdev_direct_IO(rw, iocb, inode, iter, offset, + reiserfs_get_blocks_direct_io); /* * In case of error extending write may have instantiated a few diff --git a/fs/udf/inode.c b/fs/udf/inode.c index 28984baf619..236cd48184c 100644 --- a/fs/udf/inode.c +++ b/fs/udf/inode.c @@ -226,8 +226,7 @@ static ssize_t udf_direct_IO(int rw, struct kiocb *iocb, size_t count = iov_iter_count(iter); ssize_t ret; - ret = blockdev_direct_IO(rw, iocb, inode, iter->iov, offset, iter->nr_segs, - udf_get_block); + ret = blockdev_direct_IO(rw, iocb, inode, iter, offset, udf_get_block); if (unlikely(ret < 0 && (rw & WRITE))) udf_write_failed(mapping, offset + count); return ret; diff --git a/fs/xfs/xfs_aops.c b/fs/xfs/xfs_aops.c index 6462b318678..08d13e39525 100644 --- a/fs/xfs/xfs_aops.c +++ b/fs/xfs/xfs_aops.c @@ -1470,17 +1470,15 @@ xfs_vm_direct_IO( if (offset + size > XFS_I(inode)->i_d.di_size) ioend->io_isdirect = 1; - ret = __blockdev_direct_IO(rw, iocb, inode, bdev, iter->iov, - offset, iter->nr_segs, - xfs_get_blocks_direct, + ret = __blockdev_direct_IO(rw, iocb, inode, bdev, iter, + offset, xfs_get_blocks_direct, xfs_end_io_direct_write, NULL, DIO_ASYNC_EXTEND); if (ret != -EIOCBQUEUED && iocb->private) goto out_destroy_ioend; } else { - ret = __blockdev_direct_IO(rw, iocb, inode, bdev, iter->iov, - offset, iter->nr_segs, - xfs_get_blocks_direct, + ret = __blockdev_direct_IO(rw, iocb, inode, bdev, iter, + offset, xfs_get_blocks_direct, NULL, NULL, 0); } diff --git a/include/linux/fs.h b/include/linux/fs.h index 399a338c92b..946a9484844 100644 --- a/include/linux/fs.h +++ b/include/linux/fs.h @@ -2474,16 +2474,16 @@ enum { void dio_end_io(struct bio *bio, int error); ssize_t __blockdev_direct_IO(int rw, struct kiocb *iocb, struct inode *inode, - struct block_device *bdev, const struct iovec *iov, loff_t offset, - unsigned long nr_segs, get_block_t get_block, dio_iodone_t end_io, + struct block_device *bdev, struct iov_iter *iter, loff_t offset, + get_block_t get_block, dio_iodone_t end_io, dio_submit_t submit_io, int flags); static inline ssize_t blockdev_direct_IO(int rw, struct kiocb *iocb, - struct inode *inode, const struct iovec *iov, loff_t offset, - unsigned long nr_segs, get_block_t get_block) + struct inode *inode, struct iov_iter *iter, loff_t offset, + get_block_t get_block) { - return __blockdev_direct_IO(rw, iocb, inode, inode->i_sb->s_bdev, iov, - offset, nr_segs, get_block, NULL, NULL, + return __blockdev_direct_IO(rw, iocb, inode, inode->i_sb->s_bdev, iter, + offset, get_block, NULL, NULL, DIO_LOCKING | DIO_SKIP_HOLES); } #endif -- cgit v1.2.3-70-g09d2 From 28060d5d9b261da110afe48aae7a2aa6555f798f Mon Sep 17 00:00:00 2001 From: Al Viro Date: Sat, 22 Mar 2014 05:15:17 -0400 Subject: btrfs: switch check_direct_IO() to iov_iter ... and don't open-code iov_iter_alignment() there Signed-off-by: Al Viro --- fs/btrfs/inode.c | 40 +++++++++++++++------------------------- 1 file changed, 15 insertions(+), 25 deletions(-) (limited to 'fs/btrfs/inode.c') diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index b0b8fa0efba..c8386f1961f 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -7391,39 +7391,30 @@ free_ordered: } static ssize_t check_direct_IO(struct btrfs_root *root, int rw, struct kiocb *iocb, - const struct iovec *iov, loff_t offset, - unsigned long nr_segs) + const struct iov_iter *iter, loff_t offset) { int seg; int i; - size_t size; - unsigned long addr; unsigned blocksize_mask = root->sectorsize - 1; ssize_t retval = -EINVAL; - loff_t end = offset; if (offset & blocksize_mask) goto out; - /* Check the memory alignment. Blocks cannot straddle pages */ - for (seg = 0; seg < nr_segs; seg++) { - addr = (unsigned long)iov[seg].iov_base; - size = iov[seg].iov_len; - end += size; - if ((addr & blocksize_mask) || (size & blocksize_mask)) - goto out; - - /* If this is a write we don't need to check anymore */ - if (rw & WRITE) - continue; + if (iov_iter_alignment(iter) & blocksize_mask) + goto out; - /* - * Check to make sure we don't have duplicate iov_base's in this - * iovec, if so return EINVAL, otherwise we'll get csum errors - * when reading back. - */ - for (i = seg + 1; i < nr_segs; i++) { - if (iov[seg].iov_base == iov[i].iov_base) + /* If this is a write we don't need to check anymore */ + if (rw & WRITE) + return 0; + /* + * Check to make sure we don't have duplicate iov_base's in this + * iovec, if so return EINVAL, otherwise we'll get csum errors + * when reading back. + */ + for (seg = 0; seg < iter->nr_segs; seg++) { + for (i = seg + 1; i < iter->nr_segs; i++) { + if (iter->iov[seg].iov_base == iter->iov[i].iov_base) goto out; } } @@ -7443,8 +7434,7 @@ static ssize_t btrfs_direct_IO(int rw, struct kiocb *iocb, bool relock = false; ssize_t ret; - if (check_direct_IO(BTRFS_I(inode)->root, rw, iocb, iter->iov, - offset, iter->nr_segs)) + if (check_direct_IO(BTRFS_I(inode)->root, rw, iocb, iter, offset)) return 0; atomic_inc(&inode->i_dio_count); -- cgit v1.2.3-70-g09d2 From 166ae5a418756dacd2e8883c6f562c2d0ce2db2d Mon Sep 17 00:00:00 2001 From: Zach Brown Date: Fri, 9 May 2014 17:15:10 -0400 Subject: btrfs: fix inline compressed read err corruption uncompress_inline() is dropping the error from btrfs_decompress() after testing it and zeroing the page that was supposed to hold decompressed data. This can silently turn compressed inline data in to zeros if decompression fails due to corrupt compressed data or memory allocation failure. I verified this by manually forcing the error from btrfs_decompress() for a silly named copy of od: if (!strcmp(current->comm, "failod")) ret = -ENOMEM; # od -x /mnt/btrfs/dir/80 | head -1 0000000 3031 3038 310a 2d30 6f70 6e69 0a74 3031 # echo 3 > /proc/sys/vm/drop_caches # cp $(which od) /tmp/failod # /tmp/failod -x /mnt/btrfs/dir/80 | head -1 0000000 0000 0000 0000 0000 0000 0000 0000 0000 The fix is to pass the error to its caller. Which still has a BUG_ON(). So we fix that too. There seems to be no reason for the zeroing of the page on the error from btrfs_decompress() but not from the allocation error a few lines above. So the page zeroing is removed. Signed-off-by: Zach Brown Reviewed-by: David Sterba Signed-off-by: Chris Mason --- fs/btrfs/inode.c | 15 +++++---------- 1 file changed, 5 insertions(+), 10 deletions(-) (limited to 'fs/btrfs/inode.c') diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index 5f805bc944f..24dfa27a4f7 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -6086,16 +6086,8 @@ static noinline int uncompress_inline(struct btrfs_path *path, max_size = min_t(unsigned long, PAGE_CACHE_SIZE, max_size); ret = btrfs_decompress(compress_type, tmp, page, extent_offset, inline_size, max_size); - if (ret) { - char *kaddr = kmap_atomic(page); - unsigned long copy_size = min_t(u64, - PAGE_CACHE_SIZE - pg_offset, - max_size - extent_offset); - memset(kaddr + pg_offset, 0, copy_size); - kunmap_atomic(kaddr); - } kfree(tmp); - return 0; + return ret; } /* @@ -6296,7 +6288,10 @@ next: ret = uncompress_inline(path, inode, page, pg_offset, extent_offset, item); - BUG_ON(ret); /* -ENOMEM */ + if (ret) { + err = ret; + goto out; + } } else { map = kmap(page); read_extent_buffer(leaf, map + pg_offset, ptr, -- cgit v1.2.3-70-g09d2 From ef3b9af50bfa6a1f02cd7b3f5124b712b1ba3e3c Mon Sep 17 00:00:00 2001 From: Filipe Manana Date: Sun, 27 Apr 2014 20:40:45 +0100 Subject: Btrfs: implement inode_operations callback tmpfile This implements the tmpfile callback of struct inode_operations, introduced in the linux kernel 3.11, and implemented already by some filesystems. This callback is invoked by the VFS when the flag O_TMPFILE is passed to the open system call. Signed-off-by: Filipe David Borba Manana Signed-off-by: Chris Mason Reviewed-by: David Sterba --- fs/btrfs/inode.c | 118 +++++++++++++++++++++++++++++++++++++++++++++---------- 1 file changed, 98 insertions(+), 20 deletions(-) (limited to 'fs/btrfs/inode.c') diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index 24dfa27a4f7..5c0def8dd76 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -5553,6 +5553,7 @@ static struct inode *btrfs_new_inode(struct btrfs_trans_handle *trans, struct btrfs_inode_ref *ref; struct btrfs_key key[2]; u32 sizes[2]; + int nitems = name ? 2 : 1; unsigned long ptr; int ret; @@ -5572,7 +5573,7 @@ static struct inode *btrfs_new_inode(struct btrfs_trans_handle *trans, */ inode->i_ino = objectid; - if (dir) { + if (dir && name) { trace_btrfs_inode_request(dir); ret = btrfs_set_inode_index(dir, index); @@ -5581,6 +5582,8 @@ static struct inode *btrfs_new_inode(struct btrfs_trans_handle *trans, iput(inode); return ERR_PTR(ret); } + } else if (dir) { + *index = 0; } /* * index_cnt is ignored for everything but a dir, @@ -5605,21 +5608,24 @@ static struct inode *btrfs_new_inode(struct btrfs_trans_handle *trans, btrfs_set_key_type(&key[0], BTRFS_INODE_ITEM_KEY); key[0].offset = 0; - /* - * Start new inodes with an inode_ref. This is slightly more - * efficient for small numbers of hard links since they will - * be packed into one item. Extended refs will kick in if we - * add more hard links than can fit in the ref item. - */ - key[1].objectid = objectid; - btrfs_set_key_type(&key[1], BTRFS_INODE_REF_KEY); - key[1].offset = ref_objectid; - sizes[0] = sizeof(struct btrfs_inode_item); - sizes[1] = name_len + sizeof(*ref); + + if (name) { + /* + * Start new inodes with an inode_ref. This is slightly more + * efficient for small numbers of hard links since they will + * be packed into one item. Extended refs will kick in if we + * add more hard links than can fit in the ref item. + */ + key[1].objectid = objectid; + btrfs_set_key_type(&key[1], BTRFS_INODE_REF_KEY); + key[1].offset = ref_objectid; + + sizes[1] = name_len + sizeof(*ref); + } path->leave_spinning = 1; - ret = btrfs_insert_empty_items(trans, root, path, key, sizes, 2); + ret = btrfs_insert_empty_items(trans, root, path, key, sizes, nitems); if (ret != 0) goto fail; @@ -5632,12 +5638,14 @@ static struct inode *btrfs_new_inode(struct btrfs_trans_handle *trans, sizeof(*inode_item)); fill_inode_item(trans, path->nodes[0], inode_item, inode); - ref = btrfs_item_ptr(path->nodes[0], path->slots[0] + 1, - struct btrfs_inode_ref); - btrfs_set_inode_ref_name_len(path->nodes[0], ref, name_len); - btrfs_set_inode_ref_index(path->nodes[0], ref, *index); - ptr = (unsigned long)(ref + 1); - write_extent_buffer(path->nodes[0], name, ptr, name_len); + if (name) { + ref = btrfs_item_ptr(path->nodes[0], path->slots[0] + 1, + struct btrfs_inode_ref); + btrfs_set_inode_ref_name_len(path->nodes[0], ref, name_len); + btrfs_set_inode_ref_index(path->nodes[0], ref, *index); + ptr = (unsigned long)(ref + 1); + write_extent_buffer(path->nodes[0], name, ptr, name_len); + } btrfs_mark_buffer_dirty(path->nodes[0]); btrfs_free_path(path); @@ -5673,7 +5681,7 @@ static struct inode *btrfs_new_inode(struct btrfs_trans_handle *trans, return inode; fail: - if (dir) + if (dir && name) BTRFS_I(dir)->index_cnt--; btrfs_free_path(path); iput(inode); @@ -5958,6 +5966,15 @@ static int btrfs_link(struct dentry *old_dentry, struct inode *dir, err = btrfs_update_inode(trans, root, inode); if (err) goto fail; + if (inode->i_nlink == 1) { + /* + * If new hard link count is 1, it's a file created + * with open(2) O_TMPFILE flag. + */ + err = btrfs_orphan_del(trans, inode); + if (err) + goto fail; + } d_instantiate(dentry, inode); btrfs_log_new_name(trans, inode, NULL, parent); } @@ -8884,6 +8901,66 @@ static int btrfs_permission(struct inode *inode, int mask) return generic_permission(inode, mask); } +static int btrfs_tmpfile(struct inode *dir, struct dentry *dentry, umode_t mode) +{ + struct btrfs_trans_handle *trans; + struct btrfs_root *root = BTRFS_I(dir)->root; + struct inode *inode = NULL; + u64 objectid; + u64 index; + int ret = 0; + + /* + * 5 units required for adding orphan entry + */ + trans = btrfs_start_transaction(root, 5); + if (IS_ERR(trans)) + return PTR_ERR(trans); + + ret = btrfs_find_free_ino(root, &objectid); + if (ret) + goto out; + + inode = btrfs_new_inode(trans, root, dir, NULL, 0, + btrfs_ino(dir), objectid, mode, &index); + if (IS_ERR(inode)) { + ret = PTR_ERR(inode); + inode = NULL; + goto out; + } + + ret = btrfs_init_inode_security(trans, inode, dir, NULL); + if (ret) + goto out; + + ret = btrfs_update_inode(trans, root, inode); + if (ret) + goto out; + + inode->i_fop = &btrfs_file_operations; + inode->i_op = &btrfs_file_inode_operations; + + inode->i_mapping->a_ops = &btrfs_aops; + inode->i_mapping->backing_dev_info = &root->fs_info->bdi; + BTRFS_I(inode)->io_tree.ops = &btrfs_extent_io_ops; + + ret = btrfs_orphan_add(trans, inode); + if (ret) + goto out; + + d_tmpfile(dentry, inode); + mark_inode_dirty(inode); + +out: + btrfs_end_transaction(trans, root); + if (ret) + iput(inode); + btrfs_balance_delayed_items(root); + btrfs_btree_balance_dirty(root); + + return ret; +} + static const struct inode_operations btrfs_dir_inode_operations = { .getattr = btrfs_getattr, .lookup = btrfs_lookup, @@ -8904,6 +8981,7 @@ static const struct inode_operations btrfs_dir_inode_operations = { .get_acl = btrfs_get_acl, .set_acl = btrfs_set_acl, .update_time = btrfs_update_time, + .tmpfile = btrfs_tmpfile, }; static const struct inode_operations btrfs_dir_ro_inode_operations = { .lookup = btrfs_lookup, -- cgit v1.2.3-70-g09d2 From 944a4515b27302ac42d5964b7400c72f4d692782 Mon Sep 17 00:00:00 2001 From: Daeseok Youn Date: Mon, 14 Apr 2014 15:37:02 +0900 Subject: btrfs: remove redundant null check in btrfs_dentry_release() It doesn't need to check NULL for kfree() Signed-off-by: Daeseok Youn Signed-off-by: Chris Mason --- fs/btrfs/inode.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) (limited to 'fs/btrfs/inode.c') diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index 5c0def8dd76..36c13e391ae 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -5168,8 +5168,7 @@ static int btrfs_dentry_delete(const struct dentry *dentry) static void btrfs_dentry_release(struct dentry *dentry) { - if (dentry->d_fsdata) - kfree(dentry->d_fsdata); + kfree(dentry->d_fsdata); } static struct dentry *btrfs_lookup(struct inode *dir, struct dentry *dentry, -- cgit v1.2.3-70-g09d2 From 27cdeb7096b86f05ad018a24cdb63acdf0850a5d Mon Sep 17 00:00:00 2001 From: Miao Xie Date: Wed, 2 Apr 2014 19:51:05 +0800 Subject: Btrfs: use bitfield instead of integer data type for the some variants in btrfs_root Signed-off-by: Miao Xie Signed-off-by: Wang Shilong Signed-off-by: Chris Mason --- fs/btrfs/ctree.c | 25 ++++++++++++++----------- fs/btrfs/ctree.h | 39 +++++++++++++++++++++------------------ fs/btrfs/disk-io.c | 33 +++++++++++++++------------------ fs/btrfs/extent-tree.c | 6 +++--- fs/btrfs/file.c | 4 +++- fs/btrfs/inode.c | 29 ++++++++++++++++++----------- fs/btrfs/ioctl.c | 4 ++-- fs/btrfs/relocation.c | 17 +++++++++-------- fs/btrfs/root-tree.c | 2 +- fs/btrfs/transaction.c | 33 +++++++++++++++++---------------- fs/btrfs/tree-defrag.c | 2 +- fs/btrfs/tree-log.c | 9 +++++---- 12 files changed, 109 insertions(+), 94 deletions(-) (limited to 'fs/btrfs/inode.c') diff --git a/fs/btrfs/ctree.c b/fs/btrfs/ctree.c index 1bcfcdb23cf..2f10e12ae94 100644 --- a/fs/btrfs/ctree.c +++ b/fs/btrfs/ctree.c @@ -224,7 +224,8 @@ static struct extent_buffer *btrfs_read_lock_root_node(struct btrfs_root *root) static void add_root_to_dirty_list(struct btrfs_root *root) { spin_lock(&root->fs_info->trans_lock); - if (root->track_dirty && list_empty(&root->dirty_list)) { + if (test_bit(BTRFS_ROOT_TRACK_DIRTY, &root->state) && + list_empty(&root->dirty_list)) { list_add(&root->dirty_list, &root->fs_info->dirty_cowonly_roots); } @@ -246,9 +247,10 @@ int btrfs_copy_root(struct btrfs_trans_handle *trans, int level; struct btrfs_disk_key disk_key; - WARN_ON(root->ref_cows && trans->transid != - root->fs_info->running_transaction->transid); - WARN_ON(root->ref_cows && trans->transid != root->last_trans); + WARN_ON(test_bit(BTRFS_ROOT_REF_COWS, &root->state) && + trans->transid != root->fs_info->running_transaction->transid); + WARN_ON(test_bit(BTRFS_ROOT_REF_COWS, &root->state) && + trans->transid != root->last_trans); level = btrfs_header_level(buf); if (level == 0) @@ -997,14 +999,14 @@ int btrfs_block_can_be_shared(struct btrfs_root *root, * snapshot and the block was not allocated by tree relocation, * we know the block is not shared. */ - if (root->ref_cows && + if (test_bit(BTRFS_ROOT_REF_COWS, &root->state) && buf != root->node && buf != root->commit_root && (btrfs_header_generation(buf) <= btrfs_root_last_snapshot(&root->root_item) || btrfs_header_flag(buf, BTRFS_HEADER_FLAG_RELOC))) return 1; #ifdef BTRFS_COMPAT_EXTENT_TREE_V0 - if (root->ref_cows && + if (test_bit(BTRFS_ROOT_REF_COWS, &root->state) && btrfs_header_backref_rev(buf) < BTRFS_MIXED_BACKREF_REV) return 1; #endif @@ -1146,9 +1148,10 @@ static noinline int __btrfs_cow_block(struct btrfs_trans_handle *trans, btrfs_assert_tree_locked(buf); - WARN_ON(root->ref_cows && trans->transid != - root->fs_info->running_transaction->transid); - WARN_ON(root->ref_cows && trans->transid != root->last_trans); + WARN_ON(test_bit(BTRFS_ROOT_REF_COWS, &root->state) && + trans->transid != root->fs_info->running_transaction->transid); + WARN_ON(test_bit(BTRFS_ROOT_REF_COWS, &root->state) && + trans->transid != root->last_trans); level = btrfs_header_level(buf); @@ -1193,7 +1196,7 @@ static noinline int __btrfs_cow_block(struct btrfs_trans_handle *trans, return ret; } - if (root->ref_cows) { + if (test_bit(BTRFS_ROOT_REF_COWS, &root->state)) { ret = btrfs_reloc_cow_block(trans, root, buf, cow); if (ret) return ret; @@ -1556,7 +1559,7 @@ static inline int should_cow_block(struct btrfs_trans_handle *trans, !btrfs_header_flag(buf, BTRFS_HEADER_FLAG_WRITTEN) && !(root->root_key.objectid != BTRFS_TREE_RELOC_OBJECTID && btrfs_header_flag(buf, BTRFS_HEADER_FLAG_RELOC)) && - !root->force_cow) + !test_bit(BTRFS_ROOT_FORCE_COW, &root->state)) return 0; return 1; } diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h index 0a805b8d61c..efd3bf61696 100644 --- a/fs/btrfs/ctree.h +++ b/fs/btrfs/ctree.h @@ -1710,6 +1710,26 @@ struct btrfs_subvolume_writers { wait_queue_head_t wait; }; +/* + * The state of btrfs root + */ +/* + * btrfs_record_root_in_trans is a multi-step process, + * and it can race with the balancing code. But the + * race is very small, and only the first time the root + * is added to each transaction. So IN_TRANS_SETUP + * is used to tell us when more checks are required + */ +#define BTRFS_ROOT_IN_TRANS_SETUP 0 +#define BTRFS_ROOT_REF_COWS 1 +#define BTRFS_ROOT_TRACK_DIRTY 2 +#define BTRFS_ROOT_IN_RADIX 3 +#define BTRFS_ROOT_DUMMY_ROOT 4 +#define BTRFS_ROOT_ORPHAN_ITEM_INSERTED 5 +#define BTRFS_ROOT_DEFRAG_RUNNING 6 +#define BTRFS_ROOT_FORCE_COW 7 +#define BTRFS_ROOT_MULTI_LOG_TASKS 8 + /* * in ram representation of the tree. extent_root is used for all allocations * and for the extent tree extent_root root. @@ -1721,6 +1741,7 @@ struct btrfs_root { struct btrfs_root *log_root; struct btrfs_root *reloc_root; + unsigned long state; struct btrfs_root_item root_item; struct btrfs_key root_key; struct btrfs_fs_info *fs_info; @@ -1755,7 +1776,6 @@ struct btrfs_root { /* Just be updated when the commit succeeds. */ int last_log_commit; pid_t log_start_pid; - bool log_multiple_pids; u64 objectid; u64 last_trans; @@ -1775,23 +1795,9 @@ struct btrfs_root { u64 highest_objectid; - /* btrfs_record_root_in_trans is a multi-step process, - * and it can race with the balancing code. But the - * race is very small, and only the first time the root - * is added to each transaction. So in_trans_setup - * is used to tell us when more checks are required - */ - unsigned long in_trans_setup; - int ref_cows; - int track_dirty; - int in_radix; -#ifdef CONFIG_BTRFS_FS_RUN_SANITY_TESTS - int dummy_root; -#endif u64 defrag_trans_start; struct btrfs_key defrag_progress; struct btrfs_key defrag_max; - int defrag_running; char *name; /* the dirty list is only used by non-reference counted roots */ @@ -1805,7 +1811,6 @@ struct btrfs_root { spinlock_t orphan_lock; atomic_t orphan_inodes; struct btrfs_block_rsv *orphan_block_rsv; - int orphan_item_inserted; int orphan_cleanup_state; spinlock_t inode_lock; @@ -1823,8 +1828,6 @@ struct btrfs_root { */ dev_t anon_dev; - int force_cow; - spinlock_t root_item_lock; atomic_t refs; diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c index 85f514483bd..cf8427a7a61 100644 --- a/fs/btrfs/disk-io.c +++ b/fs/btrfs/disk-io.c @@ -1201,10 +1201,7 @@ static void __setup_root(u32 nodesize, u32 leafsize, u32 sectorsize, root->nodesize = nodesize; root->leafsize = leafsize; root->stripesize = stripesize; - root->ref_cows = 0; - root->track_dirty = 0; - root->in_radix = 0; - root->orphan_item_inserted = 0; + root->state = 0; root->orphan_cleanup_state = 0; root->objectid = objectid; @@ -1265,7 +1262,6 @@ static void __setup_root(u32 nodesize, u32 leafsize, u32 sectorsize, else root->defrag_trans_start = 0; init_completion(&root->kobj_unregister); - root->defrag_running = 0; root->root_key.objectid = objectid; root->anon_dev = 0; @@ -1290,7 +1286,7 @@ struct btrfs_root *btrfs_alloc_dummy_root(void) if (!root) return ERR_PTR(-ENOMEM); __setup_root(4096, 4096, 4096, 4096, root, NULL, 1); - root->dummy_root = 1; + set_bit(BTRFS_ROOT_DUMMY_ROOT, &root->state); return root; } @@ -1341,8 +1337,7 @@ struct btrfs_root *btrfs_create_tree(struct btrfs_trans_handle *trans, btrfs_mark_buffer_dirty(leaf); root->commit_root = btrfs_root_node(root); - root->track_dirty = 1; - + set_bit(BTRFS_ROOT_TRACK_DIRTY, &root->state); root->root_item.flags = 0; root->root_item.byte_limit = 0; @@ -1396,13 +1391,15 @@ static struct btrfs_root *alloc_log_tree(struct btrfs_trans_handle *trans, root->root_key.objectid = BTRFS_TREE_LOG_OBJECTID; root->root_key.type = BTRFS_ROOT_ITEM_KEY; root->root_key.offset = BTRFS_TREE_LOG_OBJECTID; + /* + * DON'T set REF_COWS for log trees + * * log trees do not get reference counted because they go away * before a real commit is actually done. They do store pointers * to file data extents, and those reference counts still get * updated (along with back refs to the log tree). */ - root->ref_cows = 0; leaf = btrfs_alloc_free_block(trans, root, root->leafsize, 0, BTRFS_TREE_LOG_OBJECTID, NULL, @@ -1536,7 +1533,7 @@ struct btrfs_root *btrfs_read_fs_root(struct btrfs_root *tree_root, return root; if (root->root_key.objectid != BTRFS_TREE_LOG_OBJECTID) { - root->ref_cows = 1; + set_bit(BTRFS_ROOT_REF_COWS, &root->state); btrfs_check_and_init_root_item(&root->root_item); } @@ -1606,7 +1603,7 @@ int btrfs_insert_fs_root(struct btrfs_fs_info *fs_info, (unsigned long)root->root_key.objectid, root); if (ret == 0) - root->in_radix = 1; + set_bit(BTRFS_ROOT_IN_RADIX, &root->state); spin_unlock(&fs_info->fs_roots_radix_lock); radix_tree_preload_end(); @@ -1662,7 +1659,7 @@ again: if (ret < 0) goto fail; if (ret == 0) - root->orphan_item_inserted = 1; + set_bit(BTRFS_ROOT_ORPHAN_ITEM_INSERTED, &root->state); ret = btrfs_insert_fs_root(fs_info, root); if (ret) { @@ -2101,7 +2098,7 @@ static void del_fs_roots(struct btrfs_fs_info *fs_info) struct btrfs_root, root_list); list_del(&gang[0]->root_list); - if (gang[0]->in_radix) { + if (test_bit(BTRFS_ROOT_IN_RADIX, &gang[0]->state)) { btrfs_drop_and_free_fs_root(fs_info, gang[0]); } else { free_extent_buffer(gang[0]->node); @@ -2694,7 +2691,7 @@ retry_root_backup: ret = PTR_ERR(extent_root); goto recovery_tree_root; } - extent_root->track_dirty = 1; + set_bit(BTRFS_ROOT_TRACK_DIRTY, &extent_root->state); fs_info->extent_root = extent_root; location.objectid = BTRFS_DEV_TREE_OBJECTID; @@ -2703,7 +2700,7 @@ retry_root_backup: ret = PTR_ERR(dev_root); goto recovery_tree_root; } - dev_root->track_dirty = 1; + set_bit(BTRFS_ROOT_TRACK_DIRTY, &dev_root->state); fs_info->dev_root = dev_root; btrfs_init_devices_late(fs_info); @@ -2713,13 +2710,13 @@ retry_root_backup: ret = PTR_ERR(csum_root); goto recovery_tree_root; } - csum_root->track_dirty = 1; + set_bit(BTRFS_ROOT_TRACK_DIRTY, &csum_root->state); fs_info->csum_root = csum_root; location.objectid = BTRFS_QUOTA_TREE_OBJECTID; quota_root = btrfs_read_tree_root(tree_root, &location); if (!IS_ERR(quota_root)) { - quota_root->track_dirty = 1; + set_bit(BTRFS_ROOT_TRACK_DIRTY, "a_root->state); fs_info->quota_enabled = 1; fs_info->pending_quota_state = 1; fs_info->quota_root = quota_root; @@ -2734,7 +2731,7 @@ retry_root_backup: create_uuid_tree = true; check_uuid_tree = false; } else { - uuid_root->track_dirty = 1; + set_bit(BTRFS_ROOT_TRACK_DIRTY, &uuid_root->state); fs_info->uuid_root = uuid_root; create_uuid_tree = false; check_uuid_tree = diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c index 15467e3f587..7c18566fd2b 100644 --- a/fs/btrfs/extent-tree.c +++ b/fs/btrfs/extent-tree.c @@ -2983,7 +2983,7 @@ static int __btrfs_mod_ref(struct btrfs_trans_handle *trans, nritems = btrfs_header_nritems(buf); level = btrfs_header_level(buf); - if (!root->ref_cows && level == 0) + if (!test_bit(BTRFS_ROOT_REF_COWS, &root->state) && level == 0) return 0; if (inc) @@ -4472,7 +4472,7 @@ static struct btrfs_block_rsv *get_block_rsv( { struct btrfs_block_rsv *block_rsv = NULL; - if (root->ref_cows) + if (test_bit(BTRFS_ROOT_REF_COWS, &root->state)) block_rsv = trans->block_rsv; if (root == root->fs_info->csum_root && trans->adding_csums) @@ -7838,7 +7838,7 @@ int btrfs_drop_snapshot(struct btrfs_root *root, } } - if (root->in_radix) { + if (test_bit(BTRFS_ROOT_IN_RADIX, &root->state)) { btrfs_drop_and_free_fs_root(tree_root->fs_info, root); } else { free_extent_buffer(root->node); diff --git a/fs/btrfs/file.c b/fs/btrfs/file.c index 3029925e96d..5c6947dbc94 100644 --- a/fs/btrfs/file.c +++ b/fs/btrfs/file.c @@ -714,7 +714,7 @@ int __btrfs_drop_extents(struct btrfs_trans_handle *trans, int recow; int ret; int modify_tree = -1; - int update_refs = (root->ref_cows || root == root->fs_info->tree_root); + int update_refs; int found = 0; int leafs_visited = 0; @@ -724,6 +724,8 @@ int __btrfs_drop_extents(struct btrfs_trans_handle *trans, if (start >= BTRFS_I(inode)->disk_i_size && !replace_extent) modify_tree = 0; + update_refs = (test_bit(BTRFS_ROOT_REF_COWS, &root->state) || + root == root->fs_info->tree_root); while (1) { recow = 0; ret = btrfs_lookup_file_extent(trans, root, path, ino, diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index 36c13e391ae..ef3bd3deca1 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -2947,14 +2947,15 @@ void btrfs_orphan_commit_root(struct btrfs_trans_handle *trans, root->orphan_block_rsv = NULL; spin_unlock(&root->orphan_lock); - if (root->orphan_item_inserted && + if (test_bit(BTRFS_ROOT_ORPHAN_ITEM_INSERTED, &root->state) && btrfs_root_refs(&root->root_item) > 0) { ret = btrfs_del_orphan_item(trans, root->fs_info->tree_root, root->root_key.objectid); if (ret) btrfs_abort_transaction(trans, root, ret); else - root->orphan_item_inserted = 0; + clear_bit(BTRFS_ROOT_ORPHAN_ITEM_INSERTED, + &root->state); } if (block_rsv) { @@ -3271,7 +3272,8 @@ int btrfs_orphan_cleanup(struct btrfs_root *root) btrfs_block_rsv_release(root, root->orphan_block_rsv, (u64)-1); - if (root->orphan_block_rsv || root->orphan_item_inserted) { + if (root->orphan_block_rsv || + test_bit(BTRFS_ROOT_ORPHAN_ITEM_INSERTED, &root->state)) { trans = btrfs_join_transaction(root); if (!IS_ERR(trans)) btrfs_end_transaction(trans, root); @@ -3998,7 +4000,8 @@ int btrfs_truncate_inode_items(struct btrfs_trans_handle *trans, * not block aligned since we will be keeping the last block of the * extent just the way it is. */ - if (root->ref_cows || root == root->fs_info->tree_root) + if (test_bit(BTRFS_ROOT_REF_COWS, &root->state) || + root == root->fs_info->tree_root) btrfs_drop_extent_cache(inode, ALIGN(new_size, root->sectorsize), (u64)-1, 0); @@ -4091,7 +4094,9 @@ search_again: extent_num_bytes); num_dec = (orig_num_bytes - extent_num_bytes); - if (root->ref_cows && extent_start != 0) + if (test_bit(BTRFS_ROOT_REF_COWS, + &root->state) && + extent_start != 0) inode_sub_bytes(inode, num_dec); btrfs_mark_buffer_dirty(leaf); } else { @@ -4105,7 +4110,8 @@ search_again: num_dec = btrfs_file_extent_num_bytes(leaf, fi); if (extent_start != 0) { found_extent = 1; - if (root->ref_cows) + if (test_bit(BTRFS_ROOT_REF_COWS, + &root->state)) inode_sub_bytes(inode, num_dec); } } @@ -4120,10 +4126,9 @@ search_again: btrfs_file_extent_other_encoding(leaf, fi) == 0) { u32 size = new_size - found_key.offset; - if (root->ref_cows) { + if (test_bit(BTRFS_ROOT_REF_COWS, &root->state)) inode_sub_bytes(inode, item_end + 1 - new_size); - } /* * update the ram bytes to properly reflect @@ -4133,7 +4138,8 @@ search_again: size = btrfs_file_extent_calc_inline_size(size); btrfs_truncate_item(root, path, size, 1); - } else if (root->ref_cows) { + } else if (test_bit(BTRFS_ROOT_REF_COWS, + &root->state)) { inode_sub_bytes(inode, item_end + 1 - found_key.offset); } @@ -4155,8 +4161,9 @@ delete: } else { break; } - if (found_extent && (root->ref_cows || - root == root->fs_info->tree_root)) { + if (found_extent && + (test_bit(BTRFS_ROOT_REF_COWS, &root->state) || + root == root->fs_info->tree_root)) { btrfs_set_path_blocking(path); ret = btrfs_free_extent(trans, root, extent_start, extent_num_bytes, 0, diff --git a/fs/btrfs/ioctl.c b/fs/btrfs/ioctl.c index ebac486f59a..242a37cd26b 100644 --- a/fs/btrfs/ioctl.c +++ b/fs/btrfs/ioctl.c @@ -638,7 +638,7 @@ static int create_snapshot(struct btrfs_root *root, struct inode *dir, struct btrfs_trans_handle *trans; int ret; - if (!root->ref_cows) + if (!test_bit(BTRFS_ROOT_REF_COWS, &root->state)) return -EINVAL; atomic_inc(&root->will_be_snapshoted); @@ -2369,7 +2369,7 @@ static noinline int btrfs_ioctl_snap_destroy(struct file *file, dest->root_item.drop_level = 0; btrfs_set_root_refs(&dest->root_item, 0); - if (!xchg(&dest->orphan_item_inserted, 1)) { + if (!test_and_set_bit(BTRFS_ROOT_ORPHAN_ITEM_INSERTED, &dest->state)) { ret = btrfs_insert_orphan_item(trans, root->fs_info->tree_root, dest->root_key.objectid); diff --git a/fs/btrfs/relocation.c b/fs/btrfs/relocation.c index 7f92ab1daa8..b9cf0f52251 100644 --- a/fs/btrfs/relocation.c +++ b/fs/btrfs/relocation.c @@ -528,7 +528,7 @@ static int should_ignore_root(struct btrfs_root *root) { struct btrfs_root *reloc_root; - if (!root->ref_cows) + if (!test_bit(BTRFS_ROOT_REF_COWS, &root->state)) return 0; reloc_root = root->reloc_root; @@ -610,7 +610,7 @@ struct btrfs_root *find_tree_root(struct reloc_control *rc, root = read_fs_root(rc->extent_root->fs_info, root_objectid); BUG_ON(IS_ERR(root)); - if (root->ref_cows && + if (test_bit(BTRFS_ROOT_REF_COWS, &root->state) && generation != btrfs_root_generation(&root->root_item)) return NULL; @@ -887,7 +887,7 @@ again: goto out; } - if (!root->ref_cows) + if (!test_bit(BTRFS_ROOT_REF_COWS, &root->state)) cur->cowonly = 1; if (btrfs_root_level(&root->root_item) == cur->level) { @@ -954,7 +954,8 @@ again: upper->bytenr = eb->start; upper->owner = btrfs_header_owner(eb); upper->level = lower->level + 1; - if (!root->ref_cows) + if (!test_bit(BTRFS_ROOT_REF_COWS, + &root->state)) upper->cowonly = 1; /* @@ -2441,7 +2442,7 @@ struct btrfs_root *select_reloc_root(struct btrfs_trans_handle *trans, next = walk_up_backref(next, edges, &index); root = next->root; BUG_ON(!root); - BUG_ON(!root->ref_cows); + BUG_ON(!test_bit(BTRFS_ROOT_REF_COWS, &root->state)); if (root->root_key.objectid == BTRFS_TREE_RELOC_OBJECTID) { record_reloc_root_in_trans(trans, root); @@ -2506,7 +2507,7 @@ struct btrfs_root *select_one_root(struct btrfs_trans_handle *trans, BUG_ON(!root); /* no other choice for non-references counted tree */ - if (!root->ref_cows) + if (!test_bit(BTRFS_ROOT_REF_COWS, &root->state)) return root; if (root->root_key.objectid != BTRFS_TREE_RELOC_OBJECTID) @@ -2893,14 +2894,14 @@ static int relocate_tree_block(struct btrfs_trans_handle *trans, goto out; } - if (!root || root->ref_cows) { + if (!root || test_bit(BTRFS_ROOT_REF_COWS, &root->state)) { ret = reserve_metadata_space(trans, rc, node); if (ret) goto out; } if (root) { - if (root->ref_cows) { + if (test_bit(BTRFS_ROOT_REF_COWS, &root->state)) { BUG_ON(node->new_bytenr); BUG_ON(!list_empty(&node->list)); btrfs_record_root_in_trans(trans, root); diff --git a/fs/btrfs/root-tree.c b/fs/btrfs/root-tree.c index 38bb47e7d6b..360a728a639 100644 --- a/fs/btrfs/root-tree.c +++ b/fs/btrfs/root-tree.c @@ -306,7 +306,7 @@ int btrfs_find_orphan_roots(struct btrfs_root *tree_root) break; } - root->orphan_item_inserted = 1; + set_bit(BTRFS_ROOT_ORPHAN_ITEM_INSERTED, &root->state); err = btrfs_insert_fs_root(root->fs_info, root); if (err) { diff --git a/fs/btrfs/transaction.c b/fs/btrfs/transaction.c index c30815e7923..7c4c049da87 100644 --- a/fs/btrfs/transaction.c +++ b/fs/btrfs/transaction.c @@ -241,18 +241,19 @@ loop: static int record_root_in_trans(struct btrfs_trans_handle *trans, struct btrfs_root *root) { - if (root->ref_cows && root->last_trans < trans->transid) { + if (test_bit(BTRFS_ROOT_REF_COWS, &root->state) && + root->last_trans < trans->transid) { WARN_ON(root == root->fs_info->extent_root); WARN_ON(root->commit_root != root->node); /* - * see below for in_trans_setup usage rules + * see below for IN_TRANS_SETUP usage rules * we have the reloc mutex held now, so there * is only one writer in this function */ - root->in_trans_setup = 1; + set_bit(BTRFS_ROOT_IN_TRANS_SETUP, &root->state); - /* make sure readers find in_trans_setup before + /* make sure readers find IN_TRANS_SETUP before * they find our root->last_trans update */ smp_wmb(); @@ -279,7 +280,7 @@ static int record_root_in_trans(struct btrfs_trans_handle *trans, * But, we have to set root->last_trans before we * init the relocation root, otherwise, we trip over warnings * in ctree.c. The solution used here is to flag ourselves - * with root->in_trans_setup. When this is 1, we're still + * with root IN_TRANS_SETUP. When this is 1, we're still * fixing up the reloc trees and everyone must wait. * * When this is zero, they can trust root->last_trans and fly @@ -288,8 +289,8 @@ static int record_root_in_trans(struct btrfs_trans_handle *trans, * done before we pop in the zero below */ btrfs_init_reloc_root(trans, root); - smp_wmb(); - root->in_trans_setup = 0; + smp_mb__before_clear_bit(); + clear_bit(BTRFS_ROOT_IN_TRANS_SETUP, &root->state); } return 0; } @@ -298,16 +299,16 @@ static int record_root_in_trans(struct btrfs_trans_handle *trans, int btrfs_record_root_in_trans(struct btrfs_trans_handle *trans, struct btrfs_root *root) { - if (!root->ref_cows) + if (!test_bit(BTRFS_ROOT_REF_COWS, &root->state)) return 0; /* - * see record_root_in_trans for comments about in_trans_setup usage + * see record_root_in_trans for comments about IN_TRANS_SETUP usage * and barriers */ smp_rmb(); if (root->last_trans == trans->transid && - !root->in_trans_setup) + !test_bit(BTRFS_ROOT_IN_TRANS_SETUP, &root->state)) return 0; mutex_lock(&root->fs_info->reloc_mutex); @@ -365,7 +366,7 @@ static int may_wait_transaction(struct btrfs_root *root, int type) static inline bool need_reserve_reloc_root(struct btrfs_root *root) { if (!root->fs_info->reloc_ctl || - !root->ref_cows || + !test_bit(BTRFS_ROOT_REF_COWS, &root->state) || root->root_key.objectid == BTRFS_TREE_RELOC_OBJECTID || root->reloc_root) return false; @@ -1049,8 +1050,8 @@ static noinline int commit_fs_roots(struct btrfs_trans_handle *trans, btrfs_save_ino_cache(root, trans); /* see comments in should_cow_block() */ - root->force_cow = 0; - smp_wmb(); + clear_bit(BTRFS_ROOT_FORCE_COW, &root->state); + smp_mb__after_clear_bit(); if (root->commit_root != root->node) { list_add_tail(&root->dirty_list, @@ -1081,7 +1082,7 @@ int btrfs_defrag_root(struct btrfs_root *root) struct btrfs_trans_handle *trans; int ret; - if (xchg(&root->defrag_running, 1)) + if (test_and_set_bit(BTRFS_ROOT_DEFRAG_RUNNING, &root->state)) return 0; while (1) { @@ -1104,7 +1105,7 @@ int btrfs_defrag_root(struct btrfs_root *root) break; } } - root->defrag_running = 0; + clear_bit(BTRFS_ROOT_DEFRAG_RUNNING, &root->state); return ret; } @@ -1271,7 +1272,7 @@ static noinline int create_pending_snapshot(struct btrfs_trans_handle *trans, } /* see comments in should_cow_block() */ - root->force_cow = 1; + set_bit(BTRFS_ROOT_FORCE_COW, &root->state); smp_wmb(); btrfs_set_root_node(new_root_item, tmp); diff --git a/fs/btrfs/tree-defrag.c b/fs/btrfs/tree-defrag.c index 76928ca9774..a63719cc957 100644 --- a/fs/btrfs/tree-defrag.c +++ b/fs/btrfs/tree-defrag.c @@ -49,7 +49,7 @@ int btrfs_defrag_leaves(struct btrfs_trans_handle *trans, goto out; } - if (root->ref_cows == 0) + if (!test_bit(BTRFS_ROOT_REF_COWS, &root->state)) goto out; if (btrfs_test_opt(root, SSD)) diff --git a/fs/btrfs/tree-log.c b/fs/btrfs/tree-log.c index e2f45fc0261..ffee1585613 100644 --- a/fs/btrfs/tree-log.c +++ b/fs/btrfs/tree-log.c @@ -152,9 +152,9 @@ static int start_log_trans(struct btrfs_trans_handle *trans, if (!root->log_start_pid) { root->log_start_pid = current->pid; - root->log_multiple_pids = false; + clear_bit(BTRFS_ROOT_MULTI_LOG_TASKS, &root->state); } else if (root->log_start_pid != current->pid) { - root->log_multiple_pids = true; + set_bit(BTRFS_ROOT_MULTI_LOG_TASKS, &root->state); } atomic_inc(&root->log_batch); @@ -181,7 +181,7 @@ static int start_log_trans(struct btrfs_trans_handle *trans, if (ret) goto out; } - root->log_multiple_pids = false; + clear_bit(BTRFS_ROOT_MULTI_LOG_TASKS, &root->state); root->log_start_pid = current->pid; atomic_inc(&root->log_batch); atomic_inc(&root->log_writers); @@ -2500,7 +2500,8 @@ int btrfs_sync_log(struct btrfs_trans_handle *trans, while (1) { int batch = atomic_read(&root->log_batch); /* when we're on an ssd, just kick the log commit out */ - if (!btrfs_test_opt(root, SSD) && root->log_multiple_pids) { + if (!btrfs_test_opt(root, SSD) && + test_bit(BTRFS_ROOT_MULTI_LOG_TASKS, &root->state)) { mutex_unlock(&root->log_mutex); schedule_timeout_uninterruptible(1); mutex_lock(&root->log_mutex); -- cgit v1.2.3-70-g09d2 From 995946dd2961776bb45e653810ae96643fbf4ce8 Mon Sep 17 00:00:00 2001 From: Miao Xie Date: Wed, 2 Apr 2014 19:51:06 +0800 Subject: Btrfs: use helpers for last_trans_log_full_commit instead of opencode Signed-off-by: Miao Xie Signed-off-by: Wang Shilong Signed-off-by: Chris Mason --- fs/btrfs/extent-tree.c | 5 ++--- fs/btrfs/inode.c | 2 +- fs/btrfs/tree-log.c | 40 +++++++++++++++++----------------------- fs/btrfs/tree-log.h | 16 ++++++++++++++++ 4 files changed, 36 insertions(+), 27 deletions(-) (limited to 'fs/btrfs/inode.c') diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c index 7c18566fd2b..203d9a5a403 100644 --- a/fs/btrfs/extent-tree.c +++ b/fs/btrfs/extent-tree.c @@ -26,10 +26,9 @@ #include #include #include "hash.h" -#include "ctree.h" +#include "tree-log.h" #include "disk-io.h" #include "print-tree.h" -#include "transaction.h" #include "volumes.h" #include "raid56.h" #include "locking.h" @@ -8714,7 +8713,7 @@ int btrfs_make_block_group(struct btrfs_trans_handle *trans, extent_root = root->fs_info->extent_root; - root->fs_info->last_trans_log_full_commit = trans->transid; + btrfs_set_log_full_commit(root->fs_info, trans); cache = btrfs_create_block_group_cache(root, chunk_offset, size); if (!cache) diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index ef3bd3deca1..95ac6b4f239 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -8329,7 +8329,7 @@ static int btrfs_rename(struct inode *old_dir, struct dentry *old_dentry, BTRFS_I(old_inode)->dir_index = 0ULL; if (unlikely(old_ino == BTRFS_FIRST_FREE_OBJECTID)) { /* force full log commit if subvolume involved. */ - root->fs_info->last_trans_log_full_commit = trans->transid; + btrfs_set_log_full_commit(root->fs_info, trans); } else { ret = btrfs_insert_inode_ref(trans, dest, new_dentry->d_name.name, diff --git a/fs/btrfs/tree-log.c b/fs/btrfs/tree-log.c index ffee1585613..9e1f2cd5e67 100644 --- a/fs/btrfs/tree-log.c +++ b/fs/btrfs/tree-log.c @@ -20,13 +20,11 @@ #include #include #include -#include "ctree.h" -#include "transaction.h" +#include "tree-log.h" #include "disk-io.h" #include "locking.h" #include "print-tree.h" #include "backref.h" -#include "tree-log.h" #include "hash.h" /* magic values for the inode_only field in btrfs_log_inode: @@ -144,12 +142,10 @@ static int start_log_trans(struct btrfs_trans_handle *trans, mutex_lock(&root->log_mutex); if (root->log_root) { - if (ACCESS_ONCE(root->fs_info->last_trans_log_full_commit) == - trans->transid) { + if (btrfs_need_log_full_commit(root->fs_info, trans)) { ret = -EAGAIN; goto out; } - if (!root->log_start_pid) { root->log_start_pid = current->pid; clear_bit(BTRFS_ROOT_MULTI_LOG_TASKS, &root->state); @@ -2512,8 +2508,7 @@ int btrfs_sync_log(struct btrfs_trans_handle *trans, } /* bail out if we need to do a full commit */ - if (ACCESS_ONCE(root->fs_info->last_trans_log_full_commit) == - trans->transid) { + if (btrfs_need_log_full_commit(root->fs_info, trans)) { ret = -EAGAIN; btrfs_free_logged_extents(log, log_transid); mutex_unlock(&root->log_mutex); @@ -2534,8 +2529,7 @@ int btrfs_sync_log(struct btrfs_trans_handle *trans, blk_finish_plug(&plug); btrfs_abort_transaction(trans, root, ret); btrfs_free_logged_extents(log, log_transid); - ACCESS_ONCE(root->fs_info->last_trans_log_full_commit) = - trans->transid; + btrfs_set_log_full_commit(root->fs_info, trans); mutex_unlock(&root->log_mutex); goto out; } @@ -2578,8 +2572,8 @@ int btrfs_sync_log(struct btrfs_trans_handle *trans, list_del_init(&root_log_ctx.list); blk_finish_plug(&plug); - ACCESS_ONCE(root->fs_info->last_trans_log_full_commit) = - trans->transid; + btrfs_set_log_full_commit(root->fs_info, trans); + if (ret != -ENOSPC) { btrfs_abort_transaction(trans, root, ret); mutex_unlock(&log_root_tree->log_mutex); @@ -2623,8 +2617,7 @@ int btrfs_sync_log(struct btrfs_trans_handle *trans, * now that we've moved on to the tree of log tree roots, * check the full commit flag again */ - if (ACCESS_ONCE(root->fs_info->last_trans_log_full_commit) == - trans->transid) { + if (btrfs_need_log_full_commit(root->fs_info, trans)) { blk_finish_plug(&plug); btrfs_wait_marked_extents(log, &log->dirty_log_pages, mark); btrfs_free_logged_extents(log, log_transid); @@ -2638,8 +2631,7 @@ int btrfs_sync_log(struct btrfs_trans_handle *trans, EXTENT_DIRTY | EXTENT_NEW); blk_finish_plug(&plug); if (ret) { - ACCESS_ONCE(root->fs_info->last_trans_log_full_commit) = - trans->transid; + btrfs_set_log_full_commit(root->fs_info, trans); btrfs_abort_transaction(trans, root, ret); btrfs_free_logged_extents(log, log_transid); mutex_unlock(&log_root_tree->log_mutex); @@ -2668,8 +2660,7 @@ int btrfs_sync_log(struct btrfs_trans_handle *trans, */ ret = write_ctree_super(trans, root->fs_info->tree_root, 1); if (ret) { - ACCESS_ONCE(root->fs_info->last_trans_log_full_commit) = - trans->transid; + btrfs_set_log_full_commit(root->fs_info, trans); btrfs_abort_transaction(trans, root, ret); goto out_wake_log_root; } @@ -2887,7 +2878,7 @@ fail: out_unlock: mutex_unlock(&BTRFS_I(dir)->log_mutex); if (ret == -ENOSPC) { - root->fs_info->last_trans_log_full_commit = trans->transid; + btrfs_set_log_full_commit(root->fs_info, trans); ret = 0; } else if (ret < 0) btrfs_abort_transaction(trans, root, ret); @@ -2920,7 +2911,7 @@ int btrfs_del_inode_ref_in_log(struct btrfs_trans_handle *trans, dirid, &index); mutex_unlock(&BTRFS_I(inode)->log_mutex); if (ret == -ENOSPC) { - root->fs_info->last_trans_log_full_commit = trans->transid; + btrfs_set_log_full_commit(root->fs_info, trans); ret = 0; } else if (ret < 0 && ret != -ENOENT) btrfs_abort_transaction(trans, root, ret); @@ -4131,8 +4122,7 @@ static noinline int check_parent_dirs_for_sync(struct btrfs_trans_handle *trans, * make sure any commits to the log are forced * to be full commits */ - root->fs_info->last_trans_log_full_commit = - trans->transid; + btrfs_set_log_full_commit(root->fs_info, trans); ret = 1; break; } @@ -4178,6 +4168,10 @@ static int btrfs_log_inode_parent(struct btrfs_trans_handle *trans, goto end_no_trans; } + /* + * The prev transaction commit doesn't complete, we need do + * full commit by ourselves. + */ if (root->fs_info->last_trans_log_full_commit > root->fs_info->last_trans_committed) { ret = 1; @@ -4247,7 +4241,7 @@ static int btrfs_log_inode_parent(struct btrfs_trans_handle *trans, end_trans: dput(old_parent); if (ret < 0) { - root->fs_info->last_trans_log_full_commit = trans->transid; + btrfs_set_log_full_commit(root->fs_info, trans); ret = 1; } diff --git a/fs/btrfs/tree-log.h b/fs/btrfs/tree-log.h index 91b145fce33..7f5b41bd537 100644 --- a/fs/btrfs/tree-log.h +++ b/fs/btrfs/tree-log.h @@ -19,6 +19,9 @@ #ifndef __TREE_LOG_ #define __TREE_LOG_ +#include "ctree.h" +#include "transaction.h" + /* return value for btrfs_log_dentry_safe that means we don't need to log it at all */ #define BTRFS_NO_LOG_SYNC 256 @@ -35,6 +38,19 @@ static inline void btrfs_init_log_ctx(struct btrfs_log_ctx *ctx) INIT_LIST_HEAD(&ctx->list); } +static inline void btrfs_set_log_full_commit(struct btrfs_fs_info *fs_info, + struct btrfs_trans_handle *trans) +{ + ACCESS_ONCE(fs_info->last_trans_log_full_commit) = trans->transid; +} + +static inline int btrfs_need_log_full_commit(struct btrfs_fs_info *fs_info, + struct btrfs_trans_handle *trans) +{ + return ACCESS_ONCE(fs_info->last_trans_log_full_commit) == + trans->transid; +} + int btrfs_sync_log(struct btrfs_trans_handle *trans, struct btrfs_root *root, struct btrfs_log_ctx *ctx); int btrfs_free_log(struct btrfs_trans_handle *trans, struct btrfs_root *root); -- cgit v1.2.3-70-g09d2 From 351fd3532159441e810d458a5b681090ff8449fd Mon Sep 17 00:00:00 2001 From: David Sterba Date: Thu, 15 May 2014 16:48:20 +0200 Subject: btrfs: remove stale newlines from log messages I've noticed an extra line after "use no compression", but search revealed much more in messages of more critical levels and rare errors. Signed-off-by: David Sterba Signed-off-by: Chris Mason --- fs/btrfs/extent-tree.c | 2 +- fs/btrfs/inode.c | 4 ++-- fs/btrfs/ordered-data.c | 2 +- fs/btrfs/relocation.c | 4 ++-- fs/btrfs/send.c | 4 ++-- fs/btrfs/super.c | 2 +- fs/btrfs/volumes.c | 10 +++++----- 7 files changed, 14 insertions(+), 14 deletions(-) (limited to 'fs/btrfs/inode.c') diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c index ddf16bfb927..bb5b3067ddc 100644 --- a/fs/btrfs/extent-tree.c +++ b/fs/btrfs/extent-tree.c @@ -5932,7 +5932,7 @@ static int __btrfs_free_extent(struct btrfs_trans_handle *trans, refs = btrfs_extent_refs(leaf, ei); if (refs < refs_to_drop) { btrfs_err(info, "trying to drop %d refs but we only have %Lu " - "for bytenr %Lu\n", refs_to_drop, refs, bytenr); + "for bytenr %Lu", refs_to_drop, refs, bytenr); ret = -EINVAL; btrfs_abort_transaction(trans, extent_root, ret); goto out; diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index 95ac6b4f239..0b8ce3002cf 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -3475,7 +3475,7 @@ cache_acl: ret = btrfs_load_inode_props(inode, path); if (ret) btrfs_err(root->fs_info, - "error loading props for ino %llu (root %llu): %d\n", + "error loading props for ino %llu (root %llu): %d", btrfs_ino(inode), root->root_key.objectid, ret); } @@ -8010,7 +8010,7 @@ int btrfs_create_subvol_root(struct btrfs_trans_handle *trans, err = btrfs_subvol_inherit_props(trans, new_root, parent_root); if (err) btrfs_err(new_root->fs_info, - "error inheriting subvolume %llu properties: %d\n", + "error inheriting subvolume %llu properties: %d", new_root->root_key.objectid, err); err = btrfs_update_inode(trans, new_root, inode); diff --git a/fs/btrfs/ordered-data.c b/fs/btrfs/ordered-data.c index a94b05f7286..e12441c7cf1 100644 --- a/fs/btrfs/ordered-data.c +++ b/fs/btrfs/ordered-data.c @@ -67,7 +67,7 @@ static void ordered_data_tree_panic(struct inode *inode, int errno, { struct btrfs_fs_info *fs_info = btrfs_sb(inode->i_sb); btrfs_panic(fs_info, errno, "Inconsistency in ordered tree at offset " - "%llu\n", offset); + "%llu", offset); } /* diff --git a/fs/btrfs/relocation.c b/fs/btrfs/relocation.c index b9cf0f52251..65245a07275 100644 --- a/fs/btrfs/relocation.c +++ b/fs/btrfs/relocation.c @@ -337,7 +337,7 @@ static void backref_tree_panic(struct rb_node *rb_node, int errno, u64 bytenr) if (bnode->root) fs_info = bnode->root->fs_info; btrfs_panic(fs_info, errno, "Inconsistency in backref cache " - "found at offset %llu\n", bytenr); + "found at offset %llu", bytenr); } /* @@ -1259,7 +1259,7 @@ static int __must_check __add_reloc_root(struct btrfs_root *root) if (rb_node) { btrfs_panic(root->fs_info, -EEXIST, "Duplicate root found " "for start=%llu while inserting into relocation " - "tree\n", node->bytenr); + "tree", node->bytenr); kfree(node); return -EEXIST; } diff --git a/fs/btrfs/send.c b/fs/btrfs/send.c index 3f14b31c0c9..a86c049932f 100644 --- a/fs/btrfs/send.c +++ b/fs/btrfs/send.c @@ -1349,7 +1349,7 @@ static int find_extent_clone(struct send_ctx *sctx, ret = -EIO; btrfs_err(sctx->send_root->fs_info, "did not find backref in " "send_root. inode=%llu, offset=%llu, " - "disk_byte=%llu found extent=%llu\n", + "disk_byte=%llu found extent=%llu", ino, data_offset, disk_byte, found_key.objectid); goto out; } @@ -5472,7 +5472,7 @@ static void btrfs_root_dec_send_in_progress(struct btrfs_root* root) */ if (root->send_in_progress < 0) btrfs_err(root->fs_info, - "send_in_progres unbalanced %d root %llu\n", + "send_in_progres unbalanced %d root %llu", root->send_in_progress, root->root_key.objectid); spin_unlock(&root->root_item_lock); } diff --git a/fs/btrfs/super.c b/fs/btrfs/super.c index 77fcb570670..4662d92a4b7 100644 --- a/fs/btrfs/super.c +++ b/fs/btrfs/super.c @@ -511,7 +511,7 @@ int btrfs_parse_options(struct btrfs_root *root, char *options) } else if (compress) { if (!btrfs_test_opt(root, COMPRESS)) btrfs_info(root->fs_info, - "btrfs: use %s compression\n", + "btrfs: use %s compression", compress_type); } break; diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c index be8d6714494..7d725a9ce67 100644 --- a/fs/btrfs/volumes.c +++ b/fs/btrfs/volumes.c @@ -4133,7 +4133,7 @@ static int __btrfs_alloc_chunk(struct btrfs_trans_handle *trans, if (!devs_max) devs_max = BTRFS_MAX_DEVS_SYS_CHUNK; } else { - btrfs_err(info, "invalid chunk type 0x%llx requested\n", + btrfs_err(info, "invalid chunk type 0x%llx requested", type); BUG_ON(1); } @@ -4381,7 +4381,7 @@ int btrfs_finish_chunk_alloc(struct btrfs_trans_handle *trans, if (em->start != chunk_offset || em->len != chunk_size) { btrfs_crit(extent_root->fs_info, "found a bad mapping, wanted" - " %Lu-%Lu, found %Lu-%Lu\n", chunk_offset, + " %Lu-%Lu, found %Lu-%Lu", chunk_offset, chunk_size, em->start, em->len); free_extent_map(em); return -EINVAL; @@ -4583,14 +4583,14 @@ int btrfs_num_copies(struct btrfs_fs_info *fs_info, u64 logical, u64 len) * and exit, so return 1 so the callers don't try to use other copies. */ if (!em) { - btrfs_crit(fs_info, "No mapping for %Lu-%Lu\n", logical, + btrfs_crit(fs_info, "No mapping for %Lu-%Lu", logical, logical+len); return 1; } if (em->start > logical || em->start + em->len < logical) { btrfs_crit(fs_info, "Invalid mapping for %Lu-%Lu, got " - "%Lu-%Lu\n", logical, logical+len, em->start, + "%Lu-%Lu", logical, logical+len, em->start, em->start + em->len); free_extent_map(em); return 1; @@ -4771,7 +4771,7 @@ static int __btrfs_map_block(struct btrfs_fs_info *fs_info, int rw, if (em->start > logical || em->start + em->len < logical) { btrfs_crit(fs_info, "found a bad mapping, wanted %Lu, " - "found %Lu-%Lu\n", logical, em->start, + "found %Lu-%Lu", logical, em->start, em->start + em->len); free_extent_map(em); return -EINVAL; -- cgit v1.2.3-70-g09d2 From fc4adbff823f76577ece26dcb88bf6f8392dbd43 Mon Sep 17 00:00:00 2001 From: Alex Gartrell Date: Tue, 20 May 2014 13:07:56 -0700 Subject: btrfs: Drop EXTENT_UPTODATE check in hole punching and direct locking In these instances, we are trying to determine if a page has been accessed since we began the operation for the sake of retry. This is easily accomplished by doing a gang lookup in the page mapping radix tree, and it saves us the dependency on the flag (so that we might eventually delete it). btrfs_page_exists_in_range borrows heavily from find_get_page, replacing the radix tree look up with a gang lookup of 1, so that we can find the next highest page >= index and see if it falls into our lock range. Signed-off-by: Chris Mason Signed-off-by: Alex Gartrell --- fs/btrfs/btrfs_inode.h | 2 ++ fs/btrfs/file.c | 4 +-- fs/btrfs/inode.c | 72 +++++++++++++++++++++++++++++++++++++++++++++++--- 3 files changed, 71 insertions(+), 7 deletions(-) (limited to 'fs/btrfs/inode.c') diff --git a/fs/btrfs/btrfs_inode.h b/fs/btrfs/btrfs_inode.h index c9a24444ec9..a0cf3e56fe2 100644 --- a/fs/btrfs/btrfs_inode.h +++ b/fs/btrfs/btrfs_inode.h @@ -284,4 +284,6 @@ static inline void btrfs_inode_resume_unlocked_dio(struct inode *inode) &BTRFS_I(inode)->runtime_flags); } +bool btrfs_page_exists_in_range(struct inode *inode, loff_t start, loff_t end); + #endif diff --git a/fs/btrfs/file.c b/fs/btrfs/file.c index 8accf94ef22..e46bfaf6cde 100644 --- a/fs/btrfs/file.c +++ b/fs/btrfs/file.c @@ -2266,9 +2266,7 @@ static int btrfs_punch_hole(struct inode *inode, loff_t offset, loff_t len) if ((!ordered || (ordered->file_offset + ordered->len <= lockstart || ordered->file_offset > lockend)) && - !test_range_bit(&BTRFS_I(inode)->io_tree, lockstart, - lockend, EXTENT_UPTODATE, 0, - cached_state)) { + !btrfs_page_exists_in_range(inode, lockstart, lockend)) { if (ordered) btrfs_put_ordered_extent(ordered); break; diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index 0b8ce3002cf..a462da1a3e6 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -6735,6 +6735,71 @@ out: return ret; } +bool btrfs_page_exists_in_range(struct inode *inode, loff_t start, loff_t end) +{ + struct radix_tree_root *root = &inode->i_mapping->page_tree; + int found = false; + void **pagep = NULL; + struct page *page = NULL; + int start_idx; + int end_idx; + + start_idx = start >> PAGE_CACHE_SHIFT; + + /* + * end is the last byte in the last page. end == start is legal + */ + end_idx = end >> PAGE_CACHE_SHIFT; + + rcu_read_lock(); + + /* Most of the code in this while loop is lifted from + * find_get_page. It's been modified to begin searching from a + * page and return just the first page found in that range. If the + * found idx is less than or equal to the end idx then we know that + * a page exists. If no pages are found or if those pages are + * outside of the range then we're fine (yay!) */ + while (page == NULL && + radix_tree_gang_lookup_slot(root, &pagep, NULL, start_idx, 1)) { + page = radix_tree_deref_slot(pagep); + if (unlikely(!page)) + break; + + if (radix_tree_exception(page)) { + if (radix_tree_deref_retry(page)) + continue; + /* + * Otherwise, shmem/tmpfs must be storing a swap entry + * here as an exceptional entry: so return it without + * attempting to raise page count. + */ + break; /* TODO: Is this relevant for this use case? */ + } + + if (!page_cache_get_speculative(page)) + continue; + + /* + * Has the page moved? + * This is part of the lockless pagecache protocol. See + * include/linux/pagemap.h for details. + */ + if (unlikely(page != *pagep)) { + page_cache_release(page); + page = NULL; + } + } + + if (page) { + if (page->index <= end_idx) + found = true; + page_cache_release(page); + } + + rcu_read_unlock(); + return found; +} + static int lock_extent_direct(struct inode *inode, u64 lockstart, u64 lockend, struct extent_state **cached_state, int writing) { @@ -6759,10 +6824,9 @@ static int lock_extent_direct(struct inode *inode, u64 lockstart, u64 lockend, * invalidate needs to happen so that reads after a write do not * get stale data. */ - if (!ordered && (!writing || - !test_range_bit(&BTRFS_I(inode)->io_tree, - lockstart, lockend, EXTENT_UPTODATE, 0, - *cached_state))) + if (!ordered && + (!writing || + !btrfs_page_exists_in_range(inode, lockstart, lockend))) break; unlock_extent_cached(&BTRFS_I(inode)->io_tree, lockstart, lockend, -- cgit v1.2.3-70-g09d2 From 40f765805f082ed679c55bf6ab60212e55fb6fc1 Mon Sep 17 00:00:00 2001 From: Chris Mason Date: Wed, 21 May 2014 13:35:51 -0700 Subject: Btrfs: split up __extent_writepage to lower stack usage __extent_writepage has two unrelated parts. First it does the delayed allocation dance and second it does the mapping and IO for the page we're actually writing. This splits it up into those two parts so the stack from one doesn't impact the stack from the other. Signed-off-by: Chris Mason --- fs/btrfs/extent_io.c | 330 ++++++++++++++++++++++++++++++--------------------- fs/btrfs/inode.c | 2 +- 2 files changed, 194 insertions(+), 138 deletions(-) (limited to 'fs/btrfs/inode.c') diff --git a/fs/btrfs/extent_io.c b/fs/btrfs/extent_io.c index 51299c261d5..0b5fa91d9a8 100644 --- a/fs/btrfs/extent_io.c +++ b/fs/btrfs/extent_io.c @@ -3101,143 +3101,130 @@ static noinline void update_nr_written(struct page *page, } /* - * the writepage semantics are similar to regular writepage. extent - * records are inserted to lock ranges in the tree, and as dirty areas - * are found, they are marked writeback. Then the lock bits are removed - * and the end_io handler clears the writeback ranges + * helper for __extent_writepage, doing all of the delayed allocation setup. + * + * This returns 1 if our fill_delalloc function did all the work required + * to write the page (copy into inline extent). In this case the IO has + * been started and the page is already unlocked. + * + * This returns 0 if all went well (page still locked) + * This returns < 0 if there were errors (page still locked) */ -static int __extent_writepage(struct page *page, struct writeback_control *wbc, - void *data) +static noinline_for_stack int writepage_delalloc(struct inode *inode, + struct page *page, struct writeback_control *wbc, + struct extent_page_data *epd, + u64 delalloc_start, + unsigned long *nr_written) +{ + struct extent_io_tree *tree = epd->tree; + u64 page_end = delalloc_start + PAGE_CACHE_SIZE - 1; + u64 nr_delalloc; + u64 delalloc_to_write = 0; + u64 delalloc_end = 0; + int ret; + int page_started = 0; + + if (epd->extent_locked || !tree->ops || !tree->ops->fill_delalloc) + return 0; + + while (delalloc_end < page_end) { + nr_delalloc = find_lock_delalloc_range(inode, tree, + page, + &delalloc_start, + &delalloc_end, + 128 * 1024 * 1024); + if (nr_delalloc == 0) { + delalloc_start = delalloc_end + 1; + continue; + } + ret = tree->ops->fill_delalloc(inode, page, + delalloc_start, + delalloc_end, + &page_started, + nr_written); + /* File system has been set read-only */ + if (ret) { + SetPageError(page); + /* fill_delalloc should be return < 0 for error + * but just in case, we use > 0 here meaning the + * IO is started, so we don't want to return > 0 + * unless things are going well. + */ + ret = ret < 0 ? ret : -EIO; + goto done; + } + /* + * delalloc_end is already one less than the total + * length, so we don't subtract one from + * PAGE_CACHE_SIZE + */ + delalloc_to_write += (delalloc_end - delalloc_start + + PAGE_CACHE_SIZE) >> + PAGE_CACHE_SHIFT; + delalloc_start = delalloc_end + 1; + } + if (wbc->nr_to_write < delalloc_to_write) { + int thresh = 8192; + + if (delalloc_to_write < thresh * 2) + thresh = delalloc_to_write; + wbc->nr_to_write = min_t(u64, delalloc_to_write, + thresh); + } + + /* did the fill delalloc function already unlock and start + * the IO? + */ + if (page_started) { + /* + * we've unlocked the page, so we can't update + * the mapping's writeback index, just update + * nr_to_write. + */ + wbc->nr_to_write -= *nr_written; + return 1; + } + + ret = 0; + +done: + return ret; +} + +/* + * helper for __extent_writepage. This calls the writepage start hooks, + * and does the loop to map the page into extents and bios. + * + * We return 1 if the IO is started and the page is unlocked, + * 0 if all went well (page still locked) + * < 0 if there were errors (page still locked) + */ +static noinline_for_stack int __extent_writepage_io(struct inode *inode, + struct page *page, + struct writeback_control *wbc, + struct extent_page_data *epd, + loff_t i_size, + unsigned long nr_written, + int write_flags, int *nr_ret) { - struct inode *inode = page->mapping->host; - struct extent_page_data *epd = data; struct extent_io_tree *tree = epd->tree; u64 start = page_offset(page); - u64 delalloc_start; u64 page_end = start + PAGE_CACHE_SIZE - 1; u64 end; u64 cur = start; u64 extent_offset; - u64 last_byte = i_size_read(inode); u64 block_start; u64 iosize; sector_t sector; struct extent_state *cached_state = NULL; struct extent_map *em; struct block_device *bdev; - int ret; - int nr = 0; size_t pg_offset = 0; size_t blocksize; - loff_t i_size = i_size_read(inode); - unsigned long end_index = i_size >> PAGE_CACHE_SHIFT; - u64 nr_delalloc; - u64 delalloc_end; - int page_started; - int compressed; - int write_flags; - unsigned long nr_written = 0; - bool fill_delalloc = true; - - if (wbc->sync_mode == WB_SYNC_ALL) - write_flags = WRITE_SYNC; - else - write_flags = WRITE; - - trace___extent_writepage(page, inode, wbc); - - WARN_ON(!PageLocked(page)); - - ClearPageError(page); - - pg_offset = i_size & (PAGE_CACHE_SIZE - 1); - if (page->index > end_index || - (page->index == end_index && !pg_offset)) { - page->mapping->a_ops->invalidatepage(page, 0, PAGE_CACHE_SIZE); - unlock_page(page); - return 0; - } - - if (page->index == end_index) { - char *userpage; - - userpage = kmap_atomic(page); - memset(userpage + pg_offset, 0, - PAGE_CACHE_SIZE - pg_offset); - kunmap_atomic(userpage); - flush_dcache_page(page); - } - pg_offset = 0; - - set_page_extent_mapped(page); - - if (!tree->ops || !tree->ops->fill_delalloc) - fill_delalloc = false; - - delalloc_start = start; - delalloc_end = 0; - page_started = 0; - if (!epd->extent_locked && fill_delalloc) { - u64 delalloc_to_write = 0; - /* - * make sure the wbc mapping index is at least updated - * to this page. - */ - update_nr_written(page, wbc, 0); - - while (delalloc_end < page_end) { - nr_delalloc = find_lock_delalloc_range(inode, tree, - page, - &delalloc_start, - &delalloc_end, - 128 * 1024 * 1024); - if (nr_delalloc == 0) { - delalloc_start = delalloc_end + 1; - continue; - } - ret = tree->ops->fill_delalloc(inode, page, - delalloc_start, - delalloc_end, - &page_started, - &nr_written); - /* File system has been set read-only */ - if (ret) { - SetPageError(page); - goto done; - } - /* - * delalloc_end is already one less than the total - * length, so we don't subtract one from - * PAGE_CACHE_SIZE - */ - delalloc_to_write += (delalloc_end - delalloc_start + - PAGE_CACHE_SIZE) >> - PAGE_CACHE_SHIFT; - delalloc_start = delalloc_end + 1; - } - if (wbc->nr_to_write < delalloc_to_write) { - int thresh = 8192; - - if (delalloc_to_write < thresh * 2) - thresh = delalloc_to_write; - wbc->nr_to_write = min_t(u64, delalloc_to_write, - thresh); - } + int ret = 0; + int nr = 0; + bool compressed; - /* did the fill delalloc function already unlock and start - * the IO? - */ - if (page_started) { - ret = 0; - /* - * we've unlocked the page, so we can't update - * the mapping's writeback index, just update - * nr_to_write. - */ - wbc->nr_to_write -= nr_written; - goto done_unlocked; - } - } if (tree->ops && tree->ops->writepage_start_hook) { ret = tree->ops->writepage_start_hook(page, start, page_end); @@ -3247,9 +3234,10 @@ static int __extent_writepage(struct page *page, struct writeback_control *wbc, wbc->pages_skipped++; else redirty_page_for_writepage(wbc, page); + update_nr_written(page, wbc, nr_written); unlock_page(page); - ret = 0; + ret = 1; goto done_unlocked; } } @@ -3261,7 +3249,7 @@ static int __extent_writepage(struct page *page, struct writeback_control *wbc, update_nr_written(page, wbc, nr_written + 1); end = page_end; - if (last_byte <= start) { + if (i_size <= start) { if (tree->ops && tree->ops->writepage_end_io_hook) tree->ops->writepage_end_io_hook(page, start, page_end, NULL, 1); @@ -3271,7 +3259,8 @@ static int __extent_writepage(struct page *page, struct writeback_control *wbc, blocksize = inode->i_sb->s_blocksize; while (cur <= end) { - if (cur >= last_byte) { + u64 em_end; + if (cur >= i_size) { if (tree->ops && tree->ops->writepage_end_io_hook) tree->ops->writepage_end_io_hook(page, cur, page_end, NULL, 1); @@ -3286,9 +3275,10 @@ static int __extent_writepage(struct page *page, struct writeback_control *wbc, } extent_offset = cur - em->start; - BUG_ON(extent_map_end(em) <= cur); + em_end = extent_map_end(em); + BUG_ON(em_end <= cur); BUG_ON(end < cur); - iosize = min(extent_map_end(em) - cur, end - cur + 1); + iosize = min(em_end - cur, end - cur + 1); iosize = ALIGN(iosize, blocksize); sector = (em->block_start + extent_offset) >> 9; bdev = em->bdev; @@ -3324,13 +3314,6 @@ static int __extent_writepage(struct page *page, struct writeback_control *wbc, pg_offset += iosize; continue; } - /* leave this out until we have a page_mkwrite call */ - if (0 && !test_range_bit(tree, cur, cur + iosize - 1, - EXTENT_DIRTY, 0, NULL)) { - cur = cur + iosize; - pg_offset += iosize; - continue; - } if (tree->ops && tree->ops->writepage_io_hook) { ret = tree->ops->writepage_io_hook(page, cur, @@ -3341,7 +3324,7 @@ static int __extent_writepage(struct page *page, struct writeback_control *wbc, if (ret) { SetPageError(page); } else { - unsigned long max_nr = end_index + 1; + unsigned long max_nr = (i_size >> PAGE_CACHE_SHIFT) + 1; set_range_writeback(tree, cur, cur + iosize - 1); if (!PageWriteback(page)) { @@ -3362,6 +3345,81 @@ static int __extent_writepage(struct page *page, struct writeback_control *wbc, pg_offset += iosize; nr++; } +done: + *nr_ret = nr; + +done_unlocked: + + /* drop our reference on any cached states */ + free_extent_state(cached_state); + return ret; +} + +/* + * the writepage semantics are similar to regular writepage. extent + * records are inserted to lock ranges in the tree, and as dirty areas + * are found, they are marked writeback. Then the lock bits are removed + * and the end_io handler clears the writeback ranges + */ +static int __extent_writepage(struct page *page, struct writeback_control *wbc, + void *data) +{ + struct inode *inode = page->mapping->host; + struct extent_page_data *epd = data; + u64 start = page_offset(page); + u64 page_end = start + PAGE_CACHE_SIZE - 1; + int ret; + int nr = 0; + size_t pg_offset = 0; + loff_t i_size = i_size_read(inode); + unsigned long end_index = i_size >> PAGE_CACHE_SHIFT; + int write_flags; + unsigned long nr_written = 0; + + if (wbc->sync_mode == WB_SYNC_ALL) + write_flags = WRITE_SYNC; + else + write_flags = WRITE; + + trace___extent_writepage(page, inode, wbc); + + WARN_ON(!PageLocked(page)); + + ClearPageError(page); + + pg_offset = i_size & (PAGE_CACHE_SIZE - 1); + if (page->index > end_index || + (page->index == end_index && !pg_offset)) { + page->mapping->a_ops->invalidatepage(page, 0, PAGE_CACHE_SIZE); + unlock_page(page); + return 0; + } + + if (page->index == end_index) { + char *userpage; + + userpage = kmap_atomic(page); + memset(userpage + pg_offset, 0, + PAGE_CACHE_SIZE - pg_offset); + kunmap_atomic(userpage); + flush_dcache_page(page); + } + + pg_offset = 0; + + set_page_extent_mapped(page); + + ret = writepage_delalloc(inode, page, wbc, epd, start, &nr_written); + if (ret == 1) + goto done_unlocked; + if (ret) + goto done; + + ret = __extent_writepage_io(inode, page, wbc, epd, + i_size, nr_written, write_flags, &nr); + if (ret == 1) + goto done_unlocked; + done: if (nr == 0) { /* make sure the mapping tag for page dirty gets cleared */ @@ -3373,12 +3431,10 @@ done: end_extent_writepage(page, ret, start, page_end); } unlock_page(page); + return ret; done_unlocked: - - /* drop our reference on any cached states */ - free_extent_state(cached_state); - return ret; + return 0; } static int eb_wait(void *word) diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index a462da1a3e6..992aae6c00b 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -125,7 +125,7 @@ static int btrfs_init_inode_security(struct btrfs_trans_handle *trans, * the btree. The caller should have done a btrfs_drop_extents so that * no overlapping inline items exist in the btree */ -static noinline int insert_inline_extent(struct btrfs_trans_handle *trans, +static int insert_inline_extent(struct btrfs_trans_handle *trans, struct btrfs_path *path, int extent_inserted, struct btrfs_root *root, struct inode *inode, u64 start, size_t size, size_t compressed_size, -- cgit v1.2.3-70-g09d2 From a79b7d4b3e8118f265dcb4bdf9a572c392f02708 Mon Sep 17 00:00:00 2001 From: Chris Mason Date: Thu, 22 May 2014 16:18:52 -0700 Subject: Btrfs: async delayed refs Delayed extent operations are triggered during transaction commits. The goal is to queue up a healthly batch of changes to the extent allocation tree and run through them in bulk. This farms them off to async helper threads. The goal is to have the bulk of the delayed operations being done in the background, but this is also important to limit our stack footprint. Signed-off-by: Chris Mason --- fs/btrfs/ctree.h | 5 ++++ fs/btrfs/disk-io.c | 6 ++++ fs/btrfs/extent-tree.c | 79 ++++++++++++++++++++++++++++++++++++++++++++++++++ fs/btrfs/inode.c | 1 + fs/btrfs/transaction.c | 19 ++++++++++-- 5 files changed, 107 insertions(+), 3 deletions(-) (limited to 'fs/btrfs/inode.c') diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h index 4896d7a947e..02895a126ab 100644 --- a/fs/btrfs/ctree.h +++ b/fs/btrfs/ctree.h @@ -1546,6 +1546,9 @@ struct btrfs_fs_info { */ struct btrfs_workqueue *fixup_workers; struct btrfs_workqueue *delayed_workers; + + /* the extent workers do delayed refs on the extent allocation tree */ + struct btrfs_workqueue *extent_workers; struct task_struct *transaction_kthread; struct task_struct *cleaner_kthread; int thread_pool_size; @@ -3268,6 +3271,8 @@ int btrfs_check_space_for_delayed_refs(struct btrfs_trans_handle *trans, void btrfs_put_block_group(struct btrfs_block_group_cache *cache); int btrfs_run_delayed_refs(struct btrfs_trans_handle *trans, struct btrfs_root *root, unsigned long count); +int btrfs_async_run_delayed_refs(struct btrfs_root *root, + unsigned long count, int wait); int btrfs_lookup_extent(struct btrfs_root *root, u64 start, u64 len); int btrfs_lookup_extent_info(struct btrfs_trans_handle *trans, struct btrfs_root *root, u64 bytenr, diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c index 0c0fa78ef45..8bb4aa19898 100644 --- a/fs/btrfs/disk-io.c +++ b/fs/btrfs/disk-io.c @@ -2069,6 +2069,7 @@ static void btrfs_stop_all_workers(struct btrfs_fs_info *fs_info) btrfs_destroy_workqueue(fs_info->readahead_workers); btrfs_destroy_workqueue(fs_info->flush_workers); btrfs_destroy_workqueue(fs_info->qgroup_rescan_workers); + btrfs_destroy_workqueue(fs_info->extent_workers); } static void free_root_extent_buffers(struct btrfs_root *root) @@ -2586,6 +2587,10 @@ int open_ctree(struct super_block *sb, btrfs_alloc_workqueue("readahead", flags, max_active, 2); fs_info->qgroup_rescan_workers = btrfs_alloc_workqueue("qgroup-rescan", flags, 1, 0); + fs_info->extent_workers = + btrfs_alloc_workqueue("extent-refs", flags, + min_t(u64, fs_devices->num_devices, + max_active), 8); if (!(fs_info->workers && fs_info->delalloc_workers && fs_info->submit_workers && fs_info->flush_workers && @@ -2595,6 +2600,7 @@ int open_ctree(struct super_block *sb, fs_info->endio_freespace_worker && fs_info->rmw_workers && fs_info->caching_workers && fs_info->readahead_workers && fs_info->fixup_workers && fs_info->delayed_workers && + fs_info->fixup_workers && fs_info->extent_workers && fs_info->qgroup_rescan_workers)) { err = -ENOMEM; goto fail_sb_buffer; diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c index bb5b3067ddc..6caddd5970e 100644 --- a/fs/btrfs/extent-tree.c +++ b/fs/btrfs/extent-tree.c @@ -2674,15 +2674,94 @@ int btrfs_should_throttle_delayed_refs(struct btrfs_trans_handle *trans, u64 num_entries = atomic_read(&trans->transaction->delayed_refs.num_entries); u64 avg_runtime; + u64 val; smp_mb(); avg_runtime = fs_info->avg_delayed_ref_runtime; + val = num_entries * avg_runtime; if (num_entries * avg_runtime >= NSEC_PER_SEC) return 1; + if (val >= NSEC_PER_SEC / 2) + return 2; return btrfs_check_space_for_delayed_refs(trans, root); } +struct async_delayed_refs { + struct btrfs_root *root; + int count; + int error; + int sync; + struct completion wait; + struct btrfs_work work; +}; + +static void delayed_ref_async_start(struct btrfs_work *work) +{ + struct async_delayed_refs *async; + struct btrfs_trans_handle *trans; + int ret; + + async = container_of(work, struct async_delayed_refs, work); + + trans = btrfs_join_transaction(async->root); + if (IS_ERR(trans)) { + async->error = PTR_ERR(trans); + goto done; + } + + /* + * trans->sync means that when we call end_transaciton, we won't + * wait on delayed refs + */ + trans->sync = true; + ret = btrfs_run_delayed_refs(trans, async->root, async->count); + if (ret) + async->error = ret; + + ret = btrfs_end_transaction(trans, async->root); + if (ret && !async->error) + async->error = ret; +done: + if (async->sync) + complete(&async->wait); + else + kfree(async); +} + +int btrfs_async_run_delayed_refs(struct btrfs_root *root, + unsigned long count, int wait) +{ + struct async_delayed_refs *async; + int ret; + + async = kmalloc(sizeof(*async), GFP_NOFS); + if (!async) + return -ENOMEM; + + async->root = root->fs_info->tree_root; + async->count = count; + async->error = 0; + if (wait) + async->sync = 1; + else + async->sync = 0; + init_completion(&async->wait); + + btrfs_init_work(&async->work, delayed_ref_async_start, + NULL, NULL); + + btrfs_queue_work(root->fs_info->extent_workers, &async->work); + + if (wait) { + wait_for_completion(&async->wait); + ret = async->error; + kfree(async); + return ret; + } + return 0; +} + /* * this starts processing the delayed reference count updates and * extent insertions we have queued up so far. count can be diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index 992aae6c00b..38d1e7b976d 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -2678,6 +2678,7 @@ static int btrfs_finish_ordered_io(struct btrfs_ordered_extent *ordered_extent) trans = NULL; goto out_unlock; } + trans->block_rsv = &root->fs_info->delalloc_block_rsv; if (test_bit(BTRFS_ORDERED_COMPRESSED, &ordered_extent->flags)) diff --git a/fs/btrfs/transaction.c b/fs/btrfs/transaction.c index 3aafbde8b63..1c54e2eb74a 100644 --- a/fs/btrfs/transaction.c +++ b/fs/btrfs/transaction.c @@ -697,6 +697,7 @@ static int __btrfs_end_transaction(struct btrfs_trans_handle *trans, unsigned long cur = trans->delayed_ref_updates; int lock = (trans->type != TRANS_JOIN_NOLOCK); int err = 0; + int must_run_delayed_refs = 0; if (trans->use_count > 1) { trans->use_count--; @@ -711,10 +712,18 @@ static int __btrfs_end_transaction(struct btrfs_trans_handle *trans, btrfs_create_pending_block_groups(trans, root); trans->delayed_ref_updates = 0; - if (!trans->sync && btrfs_should_throttle_delayed_refs(trans, root)) { + if (!trans->sync) { + must_run_delayed_refs = + btrfs_should_throttle_delayed_refs(trans, root); cur = max_t(unsigned long, cur, 32); - trans->delayed_ref_updates = 0; - btrfs_run_delayed_refs(trans, root, cur); + + /* + * don't make the caller wait if they are from a NOLOCK + * or ATTACH transaction, it will deadlock with commit + */ + if (must_run_delayed_refs == 1 && + (trans->type & (__TRANS_JOIN_NOLOCK | __TRANS_ATTACH))) + must_run_delayed_refs = 2; } if (trans->qgroup_reserved) { @@ -775,6 +784,10 @@ static int __btrfs_end_transaction(struct btrfs_trans_handle *trans, assert_qgroups_uptodate(trans); kmem_cache_free(btrfs_trans_handle_cachep, trans); + if (must_run_delayed_refs) { + btrfs_async_run_delayed_refs(root, cur, + must_run_delayed_refs == 1); + } return err; } -- cgit v1.2.3-70-g09d2 From 91405151eb7005949072bee751343390757f8417 Mon Sep 17 00:00:00 2001 From: Filipe Manana Date: Thu, 5 Jun 2014 13:22:24 +0100 Subject: Btrfs: make sure we retry if we couldn't get the page In inode.c:btrfs_page_exists_in_range(), if we can't get the page we need to retry. However we weren't retrying because we weren't setting page to NULL, which makes the while loop exit immediately and will make us call page_cache_release after exiting the loop which is incorrect because our page get didn't succeed. This could also make us return true when we shouldn't. Signed-off-by: Filipe David Borba Manana Signed-off-by: Chris Mason --- fs/btrfs/inode.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) (limited to 'fs/btrfs/inode.c') diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index 38d1e7b976d..cdbd20e62bb 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -6777,8 +6777,10 @@ bool btrfs_page_exists_in_range(struct inode *inode, loff_t start, loff_t end) break; /* TODO: Is this relevant for this use case? */ } - if (!page_cache_get_speculative(page)) + if (!page_cache_get_speculative(page)) { + page = NULL; continue; + } /* * Has the page moved? -- cgit v1.2.3-70-g09d2 From 809f9016253dbfec00f71e98e11c4bcbf01b6c26 Mon Sep 17 00:00:00 2001 From: Filipe Manana Date: Thu, 5 Jun 2014 13:22:25 +0100 Subject: Btrfs: make sure we retry if page is a retriable exception In inode.c:btrfs_page_exists_in_range(), if the page we get from the radix tree is an exception which should make us retry, set page to NULL in order to really retry, because otherwise we don't get another loop iteration executed (page != NULL makes the while loop exit). This also was making us call page_cache_release after exiting the loop, which isn't correct because page doesn't point to a valid page, and possibly return true from the function when we shouldn't. Signed-off-by: Filipe David Borba Manana Signed-off-by: Chris Mason --- fs/btrfs/inode.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) (limited to 'fs/btrfs/inode.c') diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index cdbd20e62bb..f265f413acb 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -6767,8 +6767,10 @@ bool btrfs_page_exists_in_range(struct inode *inode, loff_t start, loff_t end) break; if (radix_tree_exception(page)) { - if (radix_tree_deref_retry(page)) + if (radix_tree_deref_retry(page)) { + page = NULL; continue; + } /* * Otherwise, shmem/tmpfs must be storing a swap entry * here as an exceptional entry: so return it without -- cgit v1.2.3-70-g09d2 From 6fdef6d43cb897bcbbed13786e8cc8cccabc442a Mon Sep 17 00:00:00 2001 From: Filipe Manana Date: Thu, 5 Jun 2014 13:22:26 +0100 Subject: Btrfs: don't release invalid page in btrfs_page_exists_in_range() In inode.c:btrfs_page_exists_in_range(), if the page we got from the radix tree is an exception entry, which can't be retried, we exit the loop with a non-NULL page and then call page_cache_release against it, which is not ok since it's not a valid page. This could also make us return true when we shouldn't. Signed-off-by: Filipe David Borba Manana Signed-off-by: Chris Mason --- fs/btrfs/inode.c | 1 + 1 file changed, 1 insertion(+) (limited to 'fs/btrfs/inode.c') diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index f265f413acb..477e64a1146 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -6776,6 +6776,7 @@ bool btrfs_page_exists_in_range(struct inode *inode, loff_t start, loff_t end) * here as an exceptional entry: so return it without * attempting to raise page count. */ + page = NULL; break; /* TODO: Is this relevant for this use case? */ } -- cgit v1.2.3-70-g09d2 From 7ffbb598a059b73487909619d73150f99b50337a Mon Sep 17 00:00:00 2001 From: Filipe Manana Date: Mon, 9 Jun 2014 03:48:05 +0100 Subject: Btrfs: make fsync work after cloning into a file When cloning into a file, we were correctly replacing the extent items in the target range and removing the extent maps. However we weren't replacing the extent maps with new ones that point to the new extents - as a consequence, an incremental fsync (when the inode doesn't have the full sync flag) was a NOOP, since it relies on the existence of extent maps in the modified list of the inode's extent map tree, which was empty. Therefore add new extent maps to reflect the target clone range. A test case for xfstests follows. Signed-off-by: Filipe David Borba Manana Signed-off-by: Chris Mason --- fs/btrfs/ctree.h | 6 +++++ fs/btrfs/file-item.c | 76 ++++++++++++++++++++++++++++++++++++++++++++++++++++ fs/btrfs/inode.c | 42 +++-------------------------- fs/btrfs/ioctl.c | 69 +++++++++++++++++++++++++++++++++++++++++++++++ 4 files changed, 155 insertions(+), 38 deletions(-) (limited to 'fs/btrfs/inode.c') diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h index af523d69543..b7e2c1c1ef3 100644 --- a/fs/btrfs/ctree.h +++ b/fs/btrfs/ctree.h @@ -3749,6 +3749,12 @@ int btrfs_csum_one_bio(struct btrfs_root *root, struct inode *inode, struct bio *bio, u64 file_start, int contig); int btrfs_lookup_csums_range(struct btrfs_root *root, u64 start, u64 end, struct list_head *list, int search_commit); +void btrfs_extent_item_to_extent_map(struct inode *inode, + const struct btrfs_path *path, + struct btrfs_file_extent_item *fi, + const bool new_inline, + struct extent_map *em); + /* inode.c */ struct btrfs_delalloc_work { struct inode *inode; diff --git a/fs/btrfs/file-item.c b/fs/btrfs/file-item.c index 609d56b9fd8..f46cfe45d68 100644 --- a/fs/btrfs/file-item.c +++ b/fs/btrfs/file-item.c @@ -885,3 +885,79 @@ out: fail_unlock: goto out; } + +void btrfs_extent_item_to_extent_map(struct inode *inode, + const struct btrfs_path *path, + struct btrfs_file_extent_item *fi, + const bool new_inline, + struct extent_map *em) +{ + struct btrfs_root *root = BTRFS_I(inode)->root; + struct extent_buffer *leaf = path->nodes[0]; + const int slot = path->slots[0]; + struct btrfs_key key; + u64 extent_start, extent_end; + u64 bytenr; + u8 type = btrfs_file_extent_type(leaf, fi); + int compress_type = btrfs_file_extent_compression(leaf, fi); + + em->bdev = root->fs_info->fs_devices->latest_bdev; + btrfs_item_key_to_cpu(leaf, &key, slot); + extent_start = key.offset; + + if (type == BTRFS_FILE_EXTENT_REG || + type == BTRFS_FILE_EXTENT_PREALLOC) { + extent_end = extent_start + + btrfs_file_extent_num_bytes(leaf, fi); + } else if (type == BTRFS_FILE_EXTENT_INLINE) { + size_t size; + size = btrfs_file_extent_inline_len(leaf, slot, fi); + extent_end = ALIGN(extent_start + size, root->sectorsize); + } + + em->ram_bytes = btrfs_file_extent_ram_bytes(leaf, fi); + if (type == BTRFS_FILE_EXTENT_REG || + type == BTRFS_FILE_EXTENT_PREALLOC) { + em->start = extent_start; + em->len = extent_end - extent_start; + em->orig_start = extent_start - + btrfs_file_extent_offset(leaf, fi); + em->orig_block_len = btrfs_file_extent_disk_num_bytes(leaf, fi); + bytenr = btrfs_file_extent_disk_bytenr(leaf, fi); + if (bytenr == 0) { + em->block_start = EXTENT_MAP_HOLE; + return; + } + if (compress_type != BTRFS_COMPRESS_NONE) { + set_bit(EXTENT_FLAG_COMPRESSED, &em->flags); + em->compress_type = compress_type; + em->block_start = bytenr; + em->block_len = em->orig_block_len; + } else { + bytenr += btrfs_file_extent_offset(leaf, fi); + em->block_start = bytenr; + em->block_len = em->len; + if (type == BTRFS_FILE_EXTENT_PREALLOC) + set_bit(EXTENT_FLAG_PREALLOC, &em->flags); + } + } else if (type == BTRFS_FILE_EXTENT_INLINE) { + em->block_start = EXTENT_MAP_INLINE; + em->start = extent_start; + em->len = extent_end - extent_start; + /* + * Initialize orig_start and block_len with the same values + * as in inode.c:btrfs_get_extent(). + */ + em->orig_start = EXTENT_MAP_HOLE; + em->block_len = (u64)-1; + if (!new_inline && compress_type != BTRFS_COMPRESS_NONE) { + set_bit(EXTENT_FLAG_COMPRESSED, &em->flags); + em->compress_type = compress_type; + } + } else { + btrfs_err(root->fs_info, + "unknown file extent item type %d, inode %llu, offset %llu, root %llu", + type, btrfs_ino(inode), extent_start, + root->root_key.objectid); + } +} diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index 477e64a1146..372b2cb2b29 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -6129,7 +6129,6 @@ struct extent_map *btrfs_get_extent(struct inode *inode, struct page *page, { int ret; int err = 0; - u64 bytenr; u64 extent_start = 0; u64 extent_end = 0; u64 objectid = btrfs_ino(inode); @@ -6143,7 +6142,7 @@ struct extent_map *btrfs_get_extent(struct inode *inode, struct page *page, struct extent_map_tree *em_tree = &BTRFS_I(inode)->extent_tree; struct extent_io_tree *io_tree = &BTRFS_I(inode)->io_tree; struct btrfs_trans_handle *trans = NULL; - int compress_type; + const bool new_inline = !page || create; again: read_lock(&em_tree->lock); @@ -6217,7 +6216,6 @@ again: found_type = btrfs_file_extent_type(leaf, item); extent_start = found_key.offset; - compress_type = btrfs_file_extent_compression(leaf, item); if (found_type == BTRFS_FILE_EXTENT_REG || found_type == BTRFS_FILE_EXTENT_PREALLOC) { extent_end = extent_start + @@ -6252,32 +6250,10 @@ next: goto not_found_em; } - em->ram_bytes = btrfs_file_extent_ram_bytes(leaf, item); + btrfs_extent_item_to_extent_map(inode, path, item, new_inline, em); + if (found_type == BTRFS_FILE_EXTENT_REG || found_type == BTRFS_FILE_EXTENT_PREALLOC) { - em->start = extent_start; - em->len = extent_end - extent_start; - em->orig_start = extent_start - - btrfs_file_extent_offset(leaf, item); - em->orig_block_len = btrfs_file_extent_disk_num_bytes(leaf, - item); - bytenr = btrfs_file_extent_disk_bytenr(leaf, item); - if (bytenr == 0) { - em->block_start = EXTENT_MAP_HOLE; - goto insert; - } - if (compress_type != BTRFS_COMPRESS_NONE) { - set_bit(EXTENT_FLAG_COMPRESSED, &em->flags); - em->compress_type = compress_type; - em->block_start = bytenr; - em->block_len = em->orig_block_len; - } else { - bytenr += btrfs_file_extent_offset(leaf, item); - em->block_start = bytenr; - em->block_len = em->len; - if (found_type == BTRFS_FILE_EXTENT_PREALLOC) - set_bit(EXTENT_FLAG_PREALLOC, &em->flags); - } goto insert; } else if (found_type == BTRFS_FILE_EXTENT_INLINE) { unsigned long ptr; @@ -6286,12 +6262,8 @@ next: size_t extent_offset; size_t copy_size; - em->block_start = EXTENT_MAP_INLINE; - if (!page || create) { - em->start = extent_start; - em->len = extent_end - extent_start; + if (new_inline) goto out; - } size = btrfs_file_extent_inline_len(leaf, path->slots[0], item); extent_offset = page_offset(page) + pg_offset - extent_start; @@ -6301,10 +6273,6 @@ next: em->len = ALIGN(copy_size, root->sectorsize); em->orig_block_len = em->len; em->orig_start = em->start; - if (compress_type) { - set_bit(EXTENT_FLAG_COMPRESSED, &em->flags); - em->compress_type = compress_type; - } ptr = btrfs_file_extent_inline_start(item) + extent_offset; if (create == 0 && !PageUptodate(page)) { if (btrfs_file_extent_compression(leaf, item) != @@ -6351,8 +6319,6 @@ next: set_extent_uptodate(io_tree, em->start, extent_map_end(em) - 1, NULL, GFP_NOFS); goto insert; - } else { - WARN(1, KERN_ERR "btrfs unknown found_type %d\n", found_type); } not_found: em->start = start; diff --git a/fs/btrfs/ioctl.c b/fs/btrfs/ioctl.c index dd9a02a53c7..c2e796b664c 100644 --- a/fs/btrfs/ioctl.c +++ b/fs/btrfs/ioctl.c @@ -3043,6 +3043,60 @@ out: return ret; } +static void clone_update_extent_map(struct inode *inode, + const struct btrfs_trans_handle *trans, + const struct btrfs_path *path, + struct btrfs_file_extent_item *fi, + const u64 hole_offset, + const u64 hole_len) +{ + struct extent_map_tree *em_tree = &BTRFS_I(inode)->extent_tree; + struct extent_map *em; + int ret; + + em = alloc_extent_map(); + if (!em) { + set_bit(BTRFS_INODE_NEEDS_FULL_SYNC, + &BTRFS_I(inode)->runtime_flags); + return; + } + + if (fi) { + btrfs_extent_item_to_extent_map(inode, path, fi, false, em); + em->generation = -1; + if (btrfs_file_extent_type(path->nodes[0], fi) == + BTRFS_FILE_EXTENT_INLINE) + set_bit(BTRFS_INODE_NEEDS_FULL_SYNC, + &BTRFS_I(inode)->runtime_flags); + } else { + em->start = hole_offset; + em->len = hole_len; + em->ram_bytes = em->len; + em->orig_start = hole_offset; + em->block_start = EXTENT_MAP_HOLE; + em->block_len = 0; + em->orig_block_len = 0; + em->compress_type = BTRFS_COMPRESS_NONE; + em->generation = trans->transid; + } + + while (1) { + write_lock(&em_tree->lock); + ret = add_extent_mapping(em_tree, em, 1); + write_unlock(&em_tree->lock); + if (ret != -EEXIST) { + free_extent_map(em); + break; + } + btrfs_drop_extent_cache(inode, em->start, + em->start + em->len - 1, 0); + } + + if (unlikely(ret)) + set_bit(BTRFS_INODE_NEEDS_FULL_SYNC, + &BTRFS_I(inode)->runtime_flags); +} + /** * btrfs_clone() - clone a range from inode file to another * @@ -3361,8 +3415,19 @@ process_slot: btrfs_item_ptr_offset(leaf, slot), size); inode_add_bytes(inode, datal); + extent = btrfs_item_ptr(leaf, slot, + struct btrfs_file_extent_item); } + /* If we have an implicit hole (NO_HOLES feature). */ + if (drop_start < new_key.offset) + clone_update_extent_map(inode, trans, + path, NULL, drop_start, + new_key.offset - drop_start); + + clone_update_extent_map(inode, trans, path, + extent, 0, 0); + btrfs_mark_buffer_dirty(leaf); btrfs_release_path(path); @@ -3406,6 +3471,10 @@ process_slot: } ret = clone_finish_inode_update(trans, inode, destoff + len, destoff, olen); + if (ret) + goto out; + clone_update_extent_map(inode, trans, path, NULL, last_dest_end, + destoff + len - last_dest_end); } out: -- cgit v1.2.3-70-g09d2 From e570fd27f2c5d7eac3876bccf99e9838d7f911a3 Mon Sep 17 00:00:00 2001 From: Miao Xie Date: Thu, 19 Jun 2014 10:42:50 +0800 Subject: Btrfs: fix broken free space cache after the system crashed When we mounted the filesystem after the crash, we got the following message: BTRFS error (device xxx): block group xxxx has wrong amount of free space BTRFS error (device xxx): failed to load free space cache for block group xxx It is because we didn't update the metadata of the allocated space (in extent tree) until the file data was written into the disk. During this time, there was no information about the allocated spaces in either the extent tree nor the free space cache. when we wrote out the free space cache at this time (commit transaction), those spaces were lost. In fact, only the free space that is used to store the file data had this problem, the others didn't because the metadata of them is updated in the same transaction context. There are many methods which can fix the above problem - track the allocated space, and write it out when we write out the free space cache - account the size of the allocated space that is used to store the file data, if the size is not zero, don't write out the free space cache. The first one is complex and may make the performance drop down. This patch chose the second method, we use a per-block-group variant to account the size of that allocated space. Besides that, we also introduce a per-block-group read-write semaphore to avoid the race between the allocation and the free space cache write out. Signed-off-by: Miao Xie Signed-off-by: Chris Mason --- fs/btrfs/ctree.h | 13 +++- fs/btrfs/extent-tree.c | 143 ++++++++++++++++++++++++++++++++++---------- fs/btrfs/free-space-cache.c | 33 ++++++++++ fs/btrfs/inode.c | 41 +++++++++---- 4 files changed, 186 insertions(+), 44 deletions(-) (limited to 'fs/btrfs/inode.c') diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h index b7e2c1c1ef3..be91397f4e9 100644 --- a/fs/btrfs/ctree.h +++ b/fs/btrfs/ctree.h @@ -1259,11 +1259,19 @@ struct btrfs_block_group_cache { spinlock_t lock; u64 pinned; u64 reserved; + u64 delalloc_bytes; u64 bytes_super; u64 flags; u64 sectorsize; u64 cache_generation; + /* + * It is just used for the delayed data space allocation because + * only the data space allocation and the relative metadata update + * can be done cross the transaction. + */ + struct rw_semaphore data_rwsem; + /* for raid56, this is a full stripe, without parity */ unsigned long full_stripe_len; @@ -3316,7 +3324,7 @@ int btrfs_alloc_logged_file_extent(struct btrfs_trans_handle *trans, struct btrfs_key *ins); int btrfs_reserve_extent(struct btrfs_root *root, u64 num_bytes, u64 min_alloc_size, u64 empty_size, u64 hint_byte, - struct btrfs_key *ins, int is_data); + struct btrfs_key *ins, int is_data, int delalloc); int btrfs_inc_ref(struct btrfs_trans_handle *trans, struct btrfs_root *root, struct extent_buffer *buf, int full_backref, int no_quota); int btrfs_dec_ref(struct btrfs_trans_handle *trans, struct btrfs_root *root, @@ -3330,7 +3338,8 @@ int btrfs_free_extent(struct btrfs_trans_handle *trans, u64 bytenr, u64 num_bytes, u64 parent, u64 root_objectid, u64 owner, u64 offset, int no_quota); -int btrfs_free_reserved_extent(struct btrfs_root *root, u64 start, u64 len); +int btrfs_free_reserved_extent(struct btrfs_root *root, u64 start, u64 len, + int delalloc); int btrfs_free_and_pin_reserved_extent(struct btrfs_root *root, u64 start, u64 len); void btrfs_prepare_extent_commit(struct btrfs_trans_handle *trans, diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c index fafb3e53ecd..99c25391820 100644 --- a/fs/btrfs/extent-tree.c +++ b/fs/btrfs/extent-tree.c @@ -105,7 +105,8 @@ static int find_next_key(struct btrfs_path *path, int level, static void dump_space_info(struct btrfs_space_info *info, u64 bytes, int dump_block_groups); static int btrfs_update_reserved_bytes(struct btrfs_block_group_cache *cache, - u64 num_bytes, int reserve); + u64 num_bytes, int reserve, + int delalloc); static int block_rsv_use_bytes(struct btrfs_block_rsv *block_rsv, u64 num_bytes); int btrfs_pin_extent(struct btrfs_root *root, @@ -3260,7 +3261,8 @@ again: spin_lock(&block_group->lock); if (block_group->cached != BTRFS_CACHE_FINISHED || - !btrfs_test_opt(root, SPACE_CACHE)) { + !btrfs_test_opt(root, SPACE_CACHE) || + block_group->delalloc_bytes) { /* * don't bother trying to write stuff out _if_ * a) we're not cached, @@ -5613,6 +5615,7 @@ int btrfs_exclude_logged_extents(struct btrfs_root *log, * @cache: The cache we are manipulating * @num_bytes: The number of bytes in question * @reserve: One of the reservation enums + * @delalloc: The blocks are allocated for the delalloc write * * This is called by the allocator when it reserves space, or by somebody who is * freeing space that was never actually used on disk. For example if you @@ -5631,7 +5634,7 @@ int btrfs_exclude_logged_extents(struct btrfs_root *log, * succeeds. */ static int btrfs_update_reserved_bytes(struct btrfs_block_group_cache *cache, - u64 num_bytes, int reserve) + u64 num_bytes, int reserve, int delalloc) { struct btrfs_space_info *space_info = cache->space_info; int ret = 0; @@ -5650,12 +5653,18 @@ static int btrfs_update_reserved_bytes(struct btrfs_block_group_cache *cache, num_bytes, 0); space_info->bytes_may_use -= num_bytes; } + + if (delalloc) + cache->delalloc_bytes += num_bytes; } } else { if (cache->ro) space_info->bytes_readonly += num_bytes; cache->reserved -= num_bytes; space_info->bytes_reserved -= num_bytes; + + if (delalloc) + cache->delalloc_bytes -= num_bytes; } spin_unlock(&cache->lock); spin_unlock(&space_info->lock); @@ -6206,7 +6215,7 @@ void btrfs_free_tree_block(struct btrfs_trans_handle *trans, WARN_ON(test_bit(EXTENT_BUFFER_DIRTY, &buf->bflags)); btrfs_add_free_space(cache, buf->start, buf->len); - btrfs_update_reserved_bytes(cache, buf->len, RESERVE_FREE); + btrfs_update_reserved_bytes(cache, buf->len, RESERVE_FREE, 0); trace_btrfs_reserved_extent_free(root, buf->start, buf->len); pin = 0; } @@ -6365,6 +6374,70 @@ enum btrfs_loop_type { LOOP_NO_EMPTY_SIZE = 3, }; +static inline void +btrfs_lock_block_group(struct btrfs_block_group_cache *cache, + int delalloc) +{ + if (delalloc) + down_read(&cache->data_rwsem); +} + +static inline void +btrfs_grab_block_group(struct btrfs_block_group_cache *cache, + int delalloc) +{ + btrfs_get_block_group(cache); + if (delalloc) + down_read(&cache->data_rwsem); +} + +static struct btrfs_block_group_cache * +btrfs_lock_cluster(struct btrfs_block_group_cache *block_group, + struct btrfs_free_cluster *cluster, + int delalloc) +{ + struct btrfs_block_group_cache *used_bg; + bool locked = false; +again: + spin_lock(&cluster->refill_lock); + if (locked) { + if (used_bg == cluster->block_group) + return used_bg; + + up_read(&used_bg->data_rwsem); + btrfs_put_block_group(used_bg); + } + + used_bg = cluster->block_group; + if (!used_bg) + return NULL; + + if (used_bg == block_group) + return used_bg; + + btrfs_get_block_group(used_bg); + + if (!delalloc) + return used_bg; + + if (down_read_trylock(&used_bg->data_rwsem)) + return used_bg; + + spin_unlock(&cluster->refill_lock); + down_read(&used_bg->data_rwsem); + locked = true; + goto again; +} + +static inline void +btrfs_release_block_group(struct btrfs_block_group_cache *cache, + int delalloc) +{ + if (delalloc) + up_read(&cache->data_rwsem); + btrfs_put_block_group(cache); +} + /* * walks the btree of allocated extents and find a hole of a given size. * The key ins is changed to record the hole: @@ -6379,7 +6452,7 @@ enum btrfs_loop_type { static noinline int find_free_extent(struct btrfs_root *orig_root, u64 num_bytes, u64 empty_size, u64 hint_byte, struct btrfs_key *ins, - u64 flags) + u64 flags, int delalloc) { int ret = 0; struct btrfs_root *root = orig_root->fs_info->extent_root; @@ -6467,6 +6540,7 @@ static noinline int find_free_extent(struct btrfs_root *orig_root, up_read(&space_info->groups_sem); } else { index = get_block_group_index(block_group); + btrfs_lock_block_group(block_group, delalloc); goto have_block_group; } } else if (block_group) { @@ -6481,7 +6555,7 @@ search: u64 offset; int cached; - btrfs_get_block_group(block_group); + btrfs_grab_block_group(block_group, delalloc); search_start = block_group->key.objectid; /* @@ -6529,16 +6603,16 @@ have_block_group: * the refill lock keeps out other * people trying to start a new cluster */ - spin_lock(&last_ptr->refill_lock); - used_block_group = last_ptr->block_group; - if (used_block_group != block_group && - (!used_block_group || - used_block_group->ro || - !block_group_bits(used_block_group, flags))) + used_block_group = btrfs_lock_cluster(block_group, + last_ptr, + delalloc); + if (!used_block_group) goto refill_cluster; - if (used_block_group != block_group) - btrfs_get_block_group(used_block_group); + if (used_block_group != block_group && + (used_block_group->ro || + !block_group_bits(used_block_group, flags))) + goto release_cluster; offset = btrfs_alloc_from_cluster(used_block_group, last_ptr, @@ -6552,16 +6626,15 @@ have_block_group: used_block_group, search_start, num_bytes); if (used_block_group != block_group) { - btrfs_put_block_group(block_group); + btrfs_release_block_group(block_group, + delalloc); block_group = used_block_group; } goto checks; } WARN_ON(last_ptr->block_group != used_block_group); - if (used_block_group != block_group) - btrfs_put_block_group(used_block_group); -refill_cluster: +release_cluster: /* If we are on LOOP_NO_EMPTY_SIZE, we can't * set up a new clusters, so lets just skip it * and let the allocator find whatever block @@ -6578,8 +6651,10 @@ refill_cluster: * succeeding in the unclustered * allocation. */ if (loop >= LOOP_NO_EMPTY_SIZE && - last_ptr->block_group != block_group) { + used_block_group != block_group) { spin_unlock(&last_ptr->refill_lock); + btrfs_release_block_group(used_block_group, + delalloc); goto unclustered_alloc; } @@ -6589,6 +6664,10 @@ refill_cluster: */ btrfs_return_cluster_to_free_space(NULL, last_ptr); + if (used_block_group != block_group) + btrfs_release_block_group(used_block_group, + delalloc); +refill_cluster: if (loop >= LOOP_NO_EMPTY_SIZE) { spin_unlock(&last_ptr->refill_lock); goto unclustered_alloc; @@ -6696,7 +6775,7 @@ checks: BUG_ON(offset > search_start); ret = btrfs_update_reserved_bytes(block_group, num_bytes, - alloc_type); + alloc_type, delalloc); if (ret == -EAGAIN) { btrfs_add_free_space(block_group, offset, num_bytes); goto loop; @@ -6708,13 +6787,13 @@ checks: trace_btrfs_reserve_extent(orig_root, block_group, search_start, num_bytes); - btrfs_put_block_group(block_group); + btrfs_release_block_group(block_group, delalloc); break; loop: failed_cluster_refill = false; failed_alloc = false; BUG_ON(index != get_block_group_index(block_group)); - btrfs_put_block_group(block_group); + btrfs_release_block_group(block_group, delalloc); } up_read(&space_info->groups_sem); @@ -6827,7 +6906,7 @@ again: int btrfs_reserve_extent(struct btrfs_root *root, u64 num_bytes, u64 min_alloc_size, u64 empty_size, u64 hint_byte, - struct btrfs_key *ins, int is_data) + struct btrfs_key *ins, int is_data, int delalloc) { bool final_tried = false; u64 flags; @@ -6837,7 +6916,7 @@ int btrfs_reserve_extent(struct btrfs_root *root, again: WARN_ON(num_bytes < root->sectorsize); ret = find_free_extent(root, num_bytes, empty_size, hint_byte, ins, - flags); + flags, delalloc); if (ret == -ENOSPC) { if (!final_tried && ins->offset) { @@ -6862,7 +6941,8 @@ again: } static int __btrfs_free_reserved_extent(struct btrfs_root *root, - u64 start, u64 len, int pin) + u64 start, u64 len, + int pin, int delalloc) { struct btrfs_block_group_cache *cache; int ret = 0; @@ -6881,7 +6961,7 @@ static int __btrfs_free_reserved_extent(struct btrfs_root *root, pin_down_extent(root, cache, start, len, 1); else { btrfs_add_free_space(cache, start, len); - btrfs_update_reserved_bytes(cache, len, RESERVE_FREE); + btrfs_update_reserved_bytes(cache, len, RESERVE_FREE, delalloc); } btrfs_put_block_group(cache); @@ -6891,15 +6971,15 @@ static int __btrfs_free_reserved_extent(struct btrfs_root *root, } int btrfs_free_reserved_extent(struct btrfs_root *root, - u64 start, u64 len) + u64 start, u64 len, int delalloc) { - return __btrfs_free_reserved_extent(root, start, len, 0); + return __btrfs_free_reserved_extent(root, start, len, 0, delalloc); } int btrfs_free_and_pin_reserved_extent(struct btrfs_root *root, u64 start, u64 len) { - return __btrfs_free_reserved_extent(root, start, len, 1); + return __btrfs_free_reserved_extent(root, start, len, 1, 0); } static int alloc_reserved_file_extent(struct btrfs_trans_handle *trans, @@ -7114,7 +7194,7 @@ int btrfs_alloc_logged_file_extent(struct btrfs_trans_handle *trans, return -EINVAL; ret = btrfs_update_reserved_bytes(block_group, ins->offset, - RESERVE_ALLOC_NO_ACCOUNT); + RESERVE_ALLOC_NO_ACCOUNT, 0); BUG_ON(ret); /* logic error */ ret = alloc_reserved_file_extent(trans, root, 0, root_objectid, 0, owner, offset, ins, 1); @@ -7256,7 +7336,7 @@ struct extent_buffer *btrfs_alloc_free_block(struct btrfs_trans_handle *trans, return ERR_CAST(block_rsv); ret = btrfs_reserve_extent(root, blocksize, blocksize, - empty_size, hint, &ins, 0); + empty_size, hint, &ins, 0, 0); if (ret) { unuse_block_rsv(root->fs_info, block_rsv, blocksize); return ERR_PTR(ret); @@ -8659,6 +8739,7 @@ btrfs_create_block_group_cache(struct btrfs_root *root, u64 start, u64 size) start); atomic_set(&cache->count, 1); spin_lock_init(&cache->lock); + init_rwsem(&cache->data_rwsem); INIT_LIST_HEAD(&cache->list); INIT_LIST_HEAD(&cache->cluster_list); INIT_LIST_HEAD(&cache->new_bg_list); diff --git a/fs/btrfs/free-space-cache.c b/fs/btrfs/free-space-cache.c index a852e15173e..2b0a627cb5f 100644 --- a/fs/btrfs/free-space-cache.c +++ b/fs/btrfs/free-space-cache.c @@ -680,6 +680,13 @@ static int __load_free_space_cache(struct btrfs_root *root, struct inode *inode, generation = btrfs_free_space_generation(leaf, header); btrfs_release_path(path); + if (!BTRFS_I(inode)->generation) { + btrfs_info(root->fs_info, + "The free space cache file (%llu) is invalid. skip it\n", + offset); + return 0; + } + if (BTRFS_I(inode)->generation != generation) { btrfs_err(root->fs_info, "free space inode generation (%llu) " @@ -1107,6 +1114,20 @@ static int __btrfs_write_out_cache(struct btrfs_root *root, struct inode *inode, if (ret) return -1; + if (block_group && (block_group->flags & BTRFS_BLOCK_GROUP_DATA)) { + down_write(&block_group->data_rwsem); + spin_lock(&block_group->lock); + if (block_group->delalloc_bytes) { + block_group->disk_cache_state = BTRFS_DC_WRITTEN; + spin_unlock(&block_group->lock); + up_write(&block_group->data_rwsem); + BTRFS_I(inode)->generation = 0; + ret = 0; + goto out; + } + spin_unlock(&block_group->lock); + } + /* Lock all pages first so we can lock the extent safely. */ io_ctl_prepare_pages(&io_ctl, inode, 0); @@ -1145,6 +1166,8 @@ static int __btrfs_write_out_cache(struct btrfs_root *root, struct inode *inode, if (ret) goto out_nospc; + if (block_group && (block_group->flags & BTRFS_BLOCK_GROUP_DATA)) + up_write(&block_group->data_rwsem); /* * Release the pages and unlock the extent, we will flush * them out later @@ -1173,6 +1196,10 @@ out: out_nospc: cleanup_write_cache_enospc(inode, &io_ctl, &cached_state, &bitmap_list); + + if (block_group && (block_group->flags & BTRFS_BLOCK_GROUP_DATA)) + up_write(&block_group->data_rwsem); + goto out; } @@ -1192,6 +1219,12 @@ int btrfs_write_out_cache(struct btrfs_root *root, spin_unlock(&block_group->lock); return 0; } + + if (block_group->delalloc_bytes) { + block_group->disk_cache_state = BTRFS_DC_WRITTEN; + spin_unlock(&block_group->lock); + return 0; + } spin_unlock(&block_group->lock); inode = lookup_free_space_inode(root, block_group, path); diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index 372b2cb2b29..6b65fab27a1 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -693,7 +693,7 @@ retry: ret = btrfs_reserve_extent(root, async_extent->compressed_size, async_extent->compressed_size, - 0, alloc_hint, &ins, 1); + 0, alloc_hint, &ins, 1, 1); if (ret) { int i; @@ -794,7 +794,7 @@ retry: out: return ret; out_free_reserve: - btrfs_free_reserved_extent(root, ins.objectid, ins.offset); + btrfs_free_reserved_extent(root, ins.objectid, ins.offset, 1); out_free: extent_clear_unlock_delalloc(inode, async_extent->start, async_extent->start + @@ -917,7 +917,7 @@ static noinline int cow_file_range(struct inode *inode, cur_alloc_size = disk_num_bytes; ret = btrfs_reserve_extent(root, cur_alloc_size, root->sectorsize, 0, alloc_hint, - &ins, 1); + &ins, 1, 1); if (ret < 0) goto out_unlock; @@ -995,7 +995,7 @@ out: return ret; out_reserve: - btrfs_free_reserved_extent(root, ins.objectid, ins.offset); + btrfs_free_reserved_extent(root, ins.objectid, ins.offset, 1); out_unlock: extent_clear_unlock_delalloc(inode, start, end, locked_page, EXTENT_LOCKED | EXTENT_DO_ACCOUNTING | @@ -2599,6 +2599,21 @@ out_kfree: return NULL; } +static void btrfs_release_delalloc_bytes(struct btrfs_root *root, + u64 start, u64 len) +{ + struct btrfs_block_group_cache *cache; + + cache = btrfs_lookup_block_group(root->fs_info, start); + ASSERT(cache); + + spin_lock(&cache->lock); + cache->delalloc_bytes -= len; + spin_unlock(&cache->lock); + + btrfs_put_block_group(cache); +} + /* as ordered data IO finishes, this gets called so we can finish * an ordered extent if the range of bytes in the file it covers are * fully written. @@ -2698,6 +2713,10 @@ static int btrfs_finish_ordered_io(struct btrfs_ordered_extent *ordered_extent) logical_len, logical_len, compress_type, 0, 0, BTRFS_FILE_EXTENT_REG); + if (!ret) + btrfs_release_delalloc_bytes(root, + ordered_extent->start, + ordered_extent->disk_len); } unpin_extent_cache(&BTRFS_I(inode)->extent_tree, ordered_extent->file_offset, ordered_extent->len, @@ -2750,7 +2769,7 @@ out: !test_bit(BTRFS_ORDERED_NOCOW, &ordered_extent->flags) && !test_bit(BTRFS_ORDERED_PREALLOC, &ordered_extent->flags)) btrfs_free_reserved_extent(root, ordered_extent->start, - ordered_extent->disk_len); + ordered_extent->disk_len, 1); } @@ -6535,21 +6554,21 @@ static struct extent_map *btrfs_new_extent_direct(struct inode *inode, alloc_hint = get_extent_allocation_hint(inode, start, len); ret = btrfs_reserve_extent(root, len, root->sectorsize, 0, - alloc_hint, &ins, 1); + alloc_hint, &ins, 1, 1); if (ret) return ERR_PTR(ret); em = create_pinned_em(inode, start, ins.offset, start, ins.objectid, ins.offset, ins.offset, ins.offset, 0); if (IS_ERR(em)) { - btrfs_free_reserved_extent(root, ins.objectid, ins.offset); + btrfs_free_reserved_extent(root, ins.objectid, ins.offset, 1); return em; } ret = btrfs_add_ordered_extent_dio(inode, start, ins.objectid, ins.offset, ins.offset, 0); if (ret) { - btrfs_free_reserved_extent(root, ins.objectid, ins.offset); + btrfs_free_reserved_extent(root, ins.objectid, ins.offset, 1); free_extent_map(em); return ERR_PTR(ret); } @@ -7437,7 +7456,7 @@ free_ordered: if (!test_bit(BTRFS_ORDERED_PREALLOC, &ordered->flags) && !test_bit(BTRFS_ORDERED_NOCOW, &ordered->flags)) btrfs_free_reserved_extent(root, ordered->start, - ordered->disk_len); + ordered->disk_len, 1); btrfs_put_ordered_extent(ordered); btrfs_put_ordered_extent(ordered); } @@ -8819,7 +8838,7 @@ static int __btrfs_prealloc_file_range(struct inode *inode, int mode, cur_bytes = min(num_bytes, 256ULL * 1024 * 1024); cur_bytes = max(cur_bytes, min_size); ret = btrfs_reserve_extent(root, cur_bytes, min_size, 0, - *alloc_hint, &ins, 1); + *alloc_hint, &ins, 1, 0); if (ret) { if (own_trans) btrfs_end_transaction(trans, root); @@ -8833,7 +8852,7 @@ static int __btrfs_prealloc_file_range(struct inode *inode, int mode, BTRFS_FILE_EXTENT_PREALLOC); if (ret) { btrfs_free_reserved_extent(root, ins.objectid, - ins.offset); + ins.offset, 0); btrfs_abort_transaction(trans, root, ret); if (own_trans) btrfs_end_transaction(trans, root); -- cgit v1.2.3-70-g09d2