aboutsummaryrefslogtreecommitdiffstats
path: root/fs/ext4/inode.c
diff options
context:
space:
mode:
authorJan Kara <jack@suse.cz>2022-05-20 13:14:02 +0200
committerTheodore Ts'o <tytso@mit.edu>2022-06-16 12:17:56 -0400
commit8d5459c11f548131ce48b2fbf45cccc5c382558f (patch)
tree2f05d59d2c4233c8aa5a4380ab562a64ee5b57cd /fs/ext4/inode.c
parent15baa7dcadf1c4f0b4f752dc054191855ff2d78e (diff)
downloadlinux-8d5459c11f548131ce48b2fbf45cccc5c382558f.tar.gz
ext4: improve write performance with disabled delalloc
When delayed allocation is disabled (either through mount option or because we are running low on free space), ext4_write_begin() allocates blocks with EXT4_GET_BLOCKS_IO_CREATE_EXT flag. With this flag extent merging is disabled and since ext4_write_begin() is called for each page separately, we end up with a *lot* of 1 block extents in the extent tree and following writeback is writing 1 block at a time which results in very poor write throughput (4 MB/s instead of 200 MB/s). These days when ext4_get_block_unwritten() is used only by ext4_write_begin(), ext4_page_mkwrite() and inline data conversion, we can safely allow extent merging to happen from these paths since following writeback will happen on different boundaries anyway. So use EXT4_GET_BLOCKS_CREATE_UNRIT_EXT instead which restores the performance. Signed-off-by: Jan Kara <jack@suse.cz> Link: https://lore.kernel.org/r/20220520111402.4252-1-jack@suse.cz Signed-off-by: Theodore Ts'o <tytso@mit.edu>
Diffstat (limited to 'fs/ext4/inode.c')
-rw-r--r--fs/ext4/inode.c2
1 files changed, 1 insertions, 1 deletions
diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c
index 3dce7d058985b..84c0eb55071d6 100644
--- a/fs/ext4/inode.c
+++ b/fs/ext4/inode.c
@@ -829,7 +829,7 @@ int ext4_get_block_unwritten(struct inode *inode, sector_t iblock,
ext4_debug("ext4_get_block_unwritten: inode %lu, create flag %d\n",
inode->i_ino, create);
return _ext4_get_block(inode, iblock, bh_result,
- EXT4_GET_BLOCKS_IO_CREATE_EXT);
+ EXT4_GET_BLOCKS_CREATE_UNWRIT_EXT);
}
/* Maximum number of blocks we map for direct IO at once. */