Or maybe I am overselling, perhaps the hours of ABD fixes in raidz was
*because* of the crypto changes. But it did lead us to doing something like
https://github.com/openzfsonosx/zfs/commit/5995f89c1eeae7f029b3537564780458c2920f44
so we could be sure the wrong buffer was not passed into ABD.
I believe we corrected those over in the crypto PR, and I think there were a
few in vdev_raidz's parity work as well, if you want to fix those too.
--
You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub:
Closed #490 via f6a0bdbc540ca599334534616dcfd0a425166e4b.
--
You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub:
https://github.com/openzfs/openzfs/pull/490#event-1354787518
--
Another case when this patch helps is short sequential sub-block reads, which
previously could be handled as misses for the same reason. Prefetch in that
case may be not so important for performance, but it filled stream cache with
useless garbage, that could affect prefetch of other valid
https://www.illumos.org/issues/8856
--
You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub:
https://github.com/openzfs/openzfs/pull/498#issuecomment-346394055
--
openzfs-developer
Archives:
Ramzec commented on this pull request.
> @@ -1567,7 +1569,7 @@ arc_cksum_is_equal(arc_buf_hdr_t *hdr, zio_t *zio)
bzero((char *)cbuf + csize, HDR_GET_PSIZE(hdr) - csize);
ok. will do.
--
You are receiving this because you are subscribed to this thread.
Reply to this
grwilson commented on this pull request.
> @@ -1567,7 +1569,7 @@ arc_cksum_is_equal(arc_buf_hdr_t *hdr, zio_t *zio)
bzero((char *)cbuf + csize, HDR_GET_PSIZE(hdr) - csize);
BTW, this wasn't part of your change but this also looks broken. I believe that
we should be
grwilson approved this pull request.
Changes look good. Make sure you open up an illumos bug for this this.
--
You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub:
We found this issue during merging of ABD into our gate
You can view, comment on, or merge this pull request online at:
https://github.com/openzfs/openzfs/pull/498
-- Commit Summary --
* fixed arc_cksum_is_equal() that doesn't take into account ABD-logic
-- File Changes --
M
