Re: [PATCH v2 RESEND 4/5] fstests: btrfs: Add per inode dedup flag test

[Dave Chinner <david@fromorbit.com>]

On Wed, Feb 24, 2016 at 04:06:35PM +0800, Qu Wenruo wrote:
> +# File size is twice the maximum file extent of btrfs
> +# So even fallbacked to non-dedup, it will have at least 2 extents
> +file_size=$(( 256 * 1024 * 1024 ))
> +dedup_bs=$(( 64 * 1024 ))

256m, 64k.

> +_scratch_mkfs "-O dedup" >> $seqres.full 2>&1
> +_scratch_mount
> +
> +# Return 0 for not deduped at all , return 1 for part or full deduped
> +test_file_deduped () {
> +	file=$1
> +

echo -n "$file: "

> +	nr_uniq_extents=$(_uniq_extent_count $file)
> +	nr_total_extents=$(_extent_count $file)
> +
> +	if [ $nr_uniq_extents -eq $nr_total_extents ]; then
> +		echo "not de-duplicated"
> +	else
> +		echo "de-duplicated"
> +	fi

> +}
> +
> +dedup_write_file () {
> +	file=$1
> +	size=$2
> +
> +	$XFS_IO_PROG -f -c "pwrite -b $dedup_bs 0 $size" $file | _filter_xfs_io
> +}
> +
> +print_result () {
> +	file=$1
> +
> +	echo "$(basename $file): $(test_file_deduped $file)"
> +}
> +_run_btrfs_util_prog dedup enable -b $dedup_bs $SCRATCH_MNT
> +touch $SCRATCH_MNT/dedup_file
> +touch $SCRATCH_MNT/no_dedup_file
> +mkdir $SCRATCH_MNT/dedup_dir
> +mkdir $SCRATCH_MNT/no_dedup_dir
> +
> +_run_btrfs_util_prog property set $SCRATCH_MNT/no_dedup_file dedup disable
> +_run_btrfs_util_prog property set $SCRATCH_MNT/no_dedup_dir dedup disable
> +
> +dedup_write_file $SCRATCH_MNT/tmp $dedup_bs

Just call xfs_io directly - wrapper is not necessary for a single
line command.

> +# sync to ensure hash is added to dedup tree
> +sync
> +
> +dedup_write_file $SCRATCH_MNT/dedup_file $file_size
> +dedup_write_file $SCRATCH_MNT/no_dedup_file $file_size
> +dedup_write_file $SCRATCH_MNT/dedup_dir/dedup_dir_default_file $file_size
> +dedup_write_file $SCRATCH_MNT/no_dedup_dir/no_dedup_dir_default_file $file_size
> +
> +print_result $SCRATCH_MNT/dedup_file

test_file_deduped $SCRATCH_MNT/dedup_file | _filter_scratch

Cheers,

Dave.
-- 
Dave Chinner
david@fromorbit.com