y@vger.kernel.org
2011-May-10 19:57 UTC
[PATCH v2 0/3] btrfs: don''t spin in shrink_delalloc if there is nothing to free
From: Sergei Trofimovich <slyfox@gentoo.org> The really interesting commit is [PATCH 1/3] btrfs: don''t spin in shrink_delalloc if there is nothing to free which fixes hogs on ENOSPC for me. The rest of patches are cleanup. Change since v1: - Added Josef''s Reviewed-by - Fixed my MTA and patch numbering Thanks! -- To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
Sergei Trofimovich
2011-May-10 20:01 UTC
[PATCH v2 0/3] btrfs: don''t spin in shrink_delalloc if there is nothing to free
The really interesting commit is [PATCH 1/3] btrfs: don''t spin in shrink_delalloc if there is nothing to free which fixes hogs on ENOSPC for me. The rest of patches are cleanup. Change since v1: - Added Josef''s Reviewed-by - Fixed my MTA and patch numbering Thanks! -- To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
Sergei Trofimovich
2011-May-10 20:01 UTC
[PATCH v2 1/3] btrfs: don''t spin in shrink_delalloc if there is nothing to free
Observed as a large delay when --mixed filesystem is filled up. Test example: 1. create tiny --mixed FS: $ dd if=/dev/zero of=2G.img seek=$((2048 * 1024 * 1024 - 1)) count=1 bs=1 $ mkfs.btrfs --mixed 2G.img $ mount -oloop 2G.img /mnt/ut/ 2. Try to fill it up: $ dd if=/dev/urandom of=10M.file bs=10240 count=1024 $ seq 1 256 | while read file_no; do echo $file_no; time cp 10M.file ${file_no}.copy; done Up to ''200.copy'' it goes fast, but when disk fills-up each -ENOSPC message takes 3 seconds to pop-up _every_ ENOSPC (and in usermode linux it''s even more: 30-60 seconds!). (Maybe, time depends on kernel''s timer resolution). No IO, no CPU load, just rescheduling. Some debugging revealed busy spinning in shrink_delalloc. Signed-off-by: Sergei Trofimovich <slyfox@gentoo.org> Reviewed-by: Josef Bacik <josef@redhat.com> --- fs/btrfs/extent-tree.c | 4 ++++ 1 files changed, 4 insertions(+), 0 deletions(-) diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c index 9ee6bd5..9f5fdd3 100644 --- a/fs/btrfs/extent-tree.c +++ b/fs/btrfs/extent-tree.c @@ -3425,6 +3425,10 @@ static int shrink_delalloc(struct btrfs_trans_handle *trans, if (reserved == 0) return 0; + /* nothing to shrink - nothing to reclaim */ + if (root->fs_info->delalloc_bytes == 0) + return 0; + max_reclaim = min(reserved, to_reclaim); while (loops < 1024) { -- 1.7.3.4 -- To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
Sergei Trofimovich
2011-May-10 20:01 UTC
[PATCH v2 2/3] btrfs: typo: ''btrfS'' -> ''btrfs''
Signed-off-by: Sergei Trofimovich <slyfox@gentoo.org> --- fs/btrfs/dir-item.c | 2 +- 1 files changed, 1 insertions(+), 1 deletions(-) diff --git a/fs/btrfs/dir-item.c b/fs/btrfs/dir-item.c index c62f02f..dec9348 100644 --- a/fs/btrfs/dir-item.c +++ b/fs/btrfs/dir-item.c @@ -452,7 +452,7 @@ int verify_dir_item(struct btrfs_root *root, namelen = XATTR_NAME_MAX; if (btrfs_dir_name_len(leaf, dir_item) > namelen) { - printk(KERN_CRIT "btrfS: invalid dir item name len: %u\n", + printk(KERN_CRIT "btrfs: invalid dir item name len: %u\n", (unsigned)btrfs_dir_data_len(leaf, dir_item)); return 1; } -- 1.7.3.4 -- To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
Sergei Trofimovich
2011-May-10 20:01 UTC
[PATCH v2 3/3] btrfs: fix typo ''testeing'' -> ''testing''
Signed-off-by: Sergei Trofimovich <slyfox@gentoo.org> --- fs/btrfs/inode.c | 4 ++-- 1 files changed, 2 insertions(+), 2 deletions(-) diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index 7cd8ab0..72650ce 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -1310,7 +1310,7 @@ static int btrfs_set_bit_hook(struct inode *inode, /* * set_bit and clear bit hooks normally require _irqsave/restore - * but in this case, we are only testeing for the DELALLOC + * but in this case, we are only testing for the DELALLOC * bit, which is only set or cleared with irqs on */ if (!(state->state & EXTENT_DELALLOC) && (*bits & EXTENT_DELALLOC)) { @@ -1344,7 +1344,7 @@ static int btrfs_clear_bit_hook(struct inode *inode, { /* * set_bit and clear bit hooks normally require _irqsave/restore - * but in this case, we are only testeing for the DELALLOC + * but in this case, we are only testing for the DELALLOC * bit, which is only set or cleared with irqs on */ if ((state->state & EXTENT_DELALLOC) && (*bits & EXTENT_DELALLOC)) { -- 1.7.3.4 -- To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
Sergei Trofimovich
2011-May-20 20:20 UTC
[PATCH v2 0/3] btrfs: don''t spin in shrink_delalloc if there is nothing to free
The really interesting commit is btrfs: don''t spin in shrink_delalloc if there is nothing to free which fixes hogs on ENOSPC for me. The rest of patches are cleanup. Change since v2: - Rebased against 2.6.39 Change since v1: - Added Josef''s Reviewed-by - Fixed my MTA and patch numbering Thanks! P.S.: It''s also interesting to note, that ''shrink_delalloc'' tries to mess with ''bytes_reserved'' (which are metadata reservation right?) The more I look at ''bytes_reserved'' management paths the more I think ''shrink_delalloc'' should not touch them. -- To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
Sergei Trofimovich
2011-May-20 20:20 UTC
[[PATCH v3] 1/3] btrfs: don''t spin in shrink_delalloc if there is nothing to free
Observed as a large delay when --mixed filesystem is filled up. Test example: 1. create tiny --mixed FS: $ dd if=/dev/zero of=2G.img seek=$((2048 * 1024 * 1024 - 1)) count=1 bs=1 $ mkfs.btrfs --mixed 2G.img $ mount -oloop 2G.img /mnt/ut/ 2. Try to fill it up: $ dd if=/dev/urandom of=10M.file bs=10240 count=1024 $ seq 1 256 | while read file_no; do echo $file_no; time cp 10M.file ${file_no}.copy; done Up to ''200.copy'' it goes fast, but when disk fills-up each -ENOSPC message takes 3 seconds to pop-up _every_ ENOSPC (and in usermode linux it''s even more: 30-60 seconds!). (Maybe, time depends on kernel''s timer resolution). No IO, no CPU load, just rescheduling. Some debugging revealed busy spinning in shrink_delalloc. Signed-off-by: Sergei Trofimovich <slyfox@gentoo.org> Reviewed-by: Josef Bacik <josef@redhat.com> --- fs/btrfs/extent-tree.c | 4 ++++ 1 files changed, 4 insertions(+), 0 deletions(-) diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c index 9ee6bd5..9f5fdd3 100644 --- a/fs/btrfs/extent-tree.c +++ b/fs/btrfs/extent-tree.c @@ -3425,6 +3425,10 @@ static int shrink_delalloc(struct btrfs_trans_handle *trans, if (reserved == 0) return 0; + /* nothing to shrink - nothing to reclaim */ + if (root->fs_info->delalloc_bytes == 0) + return 0; + max_reclaim = min(reserved, to_reclaim); while (loops < 1024) { -- 1.7.3.4 -- To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
Sergei Trofimovich
2011-May-20 20:20 UTC
[[PATCH v3] 2/3] btrfs: typo: ''btrfS'' -> ''btrfs''
Signed-off-by: Sergei Trofimovich <slyfox@gentoo.org> --- fs/btrfs/dir-item.c | 2 +- 1 files changed, 1 insertions(+), 1 deletions(-) diff --git a/fs/btrfs/dir-item.c b/fs/btrfs/dir-item.c index c62f02f..dec9348 100644 --- a/fs/btrfs/dir-item.c +++ b/fs/btrfs/dir-item.c @@ -452,7 +452,7 @@ int verify_dir_item(struct btrfs_root *root, namelen = XATTR_NAME_MAX; if (btrfs_dir_name_len(leaf, dir_item) > namelen) { - printk(KERN_CRIT "btrfS: invalid dir item name len: %u\n", + printk(KERN_CRIT "btrfs: invalid dir item name len: %u\n", (unsigned)btrfs_dir_data_len(leaf, dir_item)); return 1; } -- 1.7.3.4 -- To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
Sergei Trofimovich
2011-May-20 20:20 UTC
[[PATCH v3] 3/3] btrfs: fix typo ''testeing'' -> ''testing''
Signed-off-by: Sergei Trofimovich <slyfox@gentoo.org> --- fs/btrfs/inode.c | 4 ++-- 1 files changed, 2 insertions(+), 2 deletions(-) diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index 7cd8ab0..72650ce 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -1310,7 +1310,7 @@ static int btrfs_set_bit_hook(struct inode *inode, /* * set_bit and clear bit hooks normally require _irqsave/restore - * but in this case, we are only testeing for the DELALLOC + * but in this case, we are only testing for the DELALLOC * bit, which is only set or cleared with irqs on */ if (!(state->state & EXTENT_DELALLOC) && (*bits & EXTENT_DELALLOC)) { @@ -1344,7 +1344,7 @@ static int btrfs_clear_bit_hook(struct inode *inode, { /* * set_bit and clear bit hooks normally require _irqsave/restore - * but in this case, we are only testeing for the DELALLOC + * but in this case, we are only testing for the DELALLOC * bit, which is only set or cleared with irqs on */ if ((state->state & EXTENT_DELALLOC) && (*bits & EXTENT_DELALLOC)) { -- 1.7.3.4 -- To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html