On Wed, Jan 12, 2022 at 05:28:46PM +0800, Shiyang Ruan wrote:
> mread copies data one-byte-at-a-time, so it may races with
> reflink_range() who invalidates page cache of the dest file.
> Allow this race by adjusting the egrep regexp.
>
> Reported-by: Darrick J. Wong <djwong@xxxxxxxxxx>
> Suggested-by: Darrick J. Wong <djwong@xxxxxxxxxx>
> Signed-off-by: Shiyang Ruan <ruansy.fnst@xxxxxxxxxxx>
Looks reasonable; I'll put this in the test system for tonight and
report back if I see anything strange. I didn't see anything obvious
after a dozen or so rounds.
Reviewed-by: Darrick J. Wong <djwong@xxxxxxxxxx>
--D
> ---
> tests/generic/670 | 5 ++++-
> 1 file changed, 4 insertions(+), 1 deletion(-)
>
> diff --git a/tests/generic/670 b/tests/generic/670
> index 83fbd0b9..4a895d90 100755
> --- a/tests/generic/670
> +++ b/tests/generic/670
> @@ -40,7 +40,10 @@ _cp_reflink $testdir/file1 $testdir/file3
> _scratch_cycle_mount
>
> fbytes() {
> - egrep -v '(61 61 61 61 61 61 61 61 61 61 61 61 61 61 61 61|62 62 62 62 62 62 62 62 62 62 62 62 62 62 62 62)'
> + # Different with generic/164,165, mread copies data from mmapped area
> + # one-byte-at-a-time, which may cause races during reflink_range().
> + # So the result of _mread_range() may be a mix of 61 and 62.
> + egrep -v '((61|62) ){15}(61|62)'
> }
>
> reader() {
> --
> 2.34.1
>
>
>
