Re: [PATCH v2 1/2] ext4: avoid RWM atomic in EXT4_MB_GRP_TEST_AND_SET_READ

[Jan Kara]

On Wed 27-05-26 11:03:26, Bohdan Trach wrote:
> EXT4_MB_GRP_TEST_AND_SET_READ uses test_and_set_bit function which
> issues an atomic write. This can cause high overhead due to cache
> contention when multiple threads iterate over groups in a tight loop,
> as is the case for ext4_mb_prefetch(). We have seen this to be a
> problem for Kunpeng 920b CPUs which uses a single ARM LSE instruction
> for this purpose.
> 
> Avoid this unconditional atomic write by testing the bit first without
> changing its value. This is OK for this use case as this bit is never
> unset.
> 
> This change significantly reduces costs of fallocate() operations which
> trigger linear group scans on large multicore machines where
> test_and_set_bit issues an atomic write operation unconditionally.
> 
> Signed-off-by: Bohdan Trach <bohdan.trach@xxxxxxxxxxxxxxx>

Looks good to me. Feel free to add:

Reviewed-by: Jan Kara <jack@xxxxxxx>

								Honza

> ---
>  fs/ext4/ext4.h | 8 +++++++-
>  1 file changed, 7 insertions(+), 1 deletion(-)
> 
> diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h
> index 56b82d4a15d7..f8eacf1375f8 100644
> --- a/fs/ext4/ext4.h
> +++ b/fs/ext4/ext4.h
> @@ -3551,7 +3551,13 @@ struct ext4_group_info {
>  #define EXT4_MB_GRP_CLEAR_TRIMMED(grp)	\
>  	(clear_bit(EXT4_GROUP_INFO_WAS_TRIMMED_BIT, &((grp)->bb_state)))
>  #define EXT4_MB_GRP_TEST_AND_SET_READ(grp)	\
> -	(test_and_set_bit(EXT4_GROUP_INFO_BBITMAP_READ_BIT, &((grp)->bb_state)))
> +	(ext4_mb_grp_test_and_set_read((grp)))
> +
> +static inline int ext4_mb_grp_test_and_set_read(struct ext4_group_info *grp)
> +{
> +	return (test_bit(EXT4_GROUP_INFO_BBITMAP_READ_BIT, &grp->bb_state) ||
> +		test_and_set_bit(EXT4_GROUP_INFO_BBITMAP_READ_BIT, &grp->bb_state));
> +}
>  
>  #define EXT4_MAX_CONTENTION		8
>  #define EXT4_CONTENTION_THRESHOLD	2
> -- 
> 2.43.0
> 
-- 
Jan Kara <jack@xxxxxxxx>
SUSE Labs, CR