dm: revert 8f50e358153d ("dm: limit the max bio size as BIO_MAX_PAGES * PAGE_SIZE")
authorMikulas Patocka <mpatocka@redhat.com>
Thu, 21 Mar 2019 20:46:12 +0000 (16:46 -0400)
committerMike Snitzer <snitzer@redhat.com>
Mon, 1 Apr 2019 20:20:36 +0000 (16:20 -0400)
commit75ae193626de3238ca5fb895868ec91c94e63b1b
tree03bbd7c98d4a514e69b71d17eeb2916d172db0e5
parent93fc91675a6c84d6ab355188aea398bda2cc51f8
dm: revert 8f50e358153d ("dm: limit the max bio size as BIO_MAX_PAGES * PAGE_SIZE")

The limit was already incorporated to dm-crypt with commit 4e870e948fba
("dm crypt: fix error with too large bios"), so we don't need to apply
it globally to all targets. The quantity BIO_MAX_PAGES * PAGE_SIZE is
wrong anyway because the variable ti->max_io_len it is supposed to be in
the units of 512-byte sectors not in bytes.

Reduction of the limit to 1048576 sectors could even cause data
corruption in rare cases - suppose that we have a dm-striped device with
stripe size 768MiB. The target will call dm_set_target_max_io_len with
the value 1572864. The buggy code would reduce it to 1048576. Now, the
dm-core will errorneously split the bios on 1048576-sector boundary
insetad of 1572864-sector boundary and pass these stripe-crossing bios
to the striped target.

Cc: stable@vger.kernel.org # v4.16+
Fixes: 8f50e358153d ("dm: limit the max bio size as BIO_MAX_PAGES * PAGE_SIZE")
Signed-off-by: Mikulas Patocka <mpatocka@redhat.com>
Acked-by: Ming Lei <ming.lei@redhat.com>
Signed-off-by: Mike Snitzer <snitzer@redhat.com>
drivers/md/dm.c