btrfs: zoned: optimize hint byte for zoned allocator
authorNaohiro Aota <naohiro.aota@wdc.com>
Mon, 18 Dec 2023 16:02:29 +0000 (01:02 +0900)
committerGreg Kroah-Hartman <gregkh@linuxfoundation.org>
Thu, 1 Feb 2024 00:19:10 +0000 (16:19 -0800)
commit46ac4e1f9a9ddfbc1dee75d3880966084b7723bd
tree5fd656daee9e4565f67683e1aa68755e1a2f7c7d
parentb1e30e2ff6eb1b200220036b22066f762ceceea2
btrfs: zoned: optimize hint byte for zoned allocator

[ Upstream commit 02444f2ac26eae6385a65fcd66915084d15dffba ]

Writing sequentially to a huge file on btrfs on a SMR HDD revealed a
decline of the performance (220 MiB/s to 30 MiB/s after 500 minutes).

The performance goes down because of increased latency of the extent
allocation, which is induced by a traversing of a lot of full block groups.

So, this patch optimizes the ffe_ctl->hint_byte by choosing a block group
with sufficient size from the active block group list, which does not
contain full block groups.

After applying the patch, the performance is maintained well.

Fixes: 2eda57089ea3 ("btrfs: zoned: implement sequential extent allocation")
CC: stable@vger.kernel.org # 5.15+
Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Signed-off-by: Naohiro Aota <naohiro.aota@wdc.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
fs/btrfs/extent-tree.c