diff options
| author | Ilya Dryomov <idryomov@gmail.com> | 2018-02-06 19:26:34 +0100 |
|---|---|---|
| committer | Ilya Dryomov <idryomov@gmail.com> | 2018-04-02 10:12:43 +0200 |
| commit | 2bb1e56ec6450ce533c644c5bfa548dc34c551a0 (patch) | |
| tree | 10b591e350fc80d19b5a46c3798f3be5fa3acbc8 /drivers/block | |
| parent | rbd: move to obj_req->img_extents (diff) | |
| download | linux-2bb1e56ec6450ce533c644c5bfa548dc34c551a0.tar.xz linux-2bb1e56ec6450ce533c644c5bfa548dc34c551a0.zip | |
rbd: create+truncate for whole-object layered discards
A whole-object layered discard is implemented as a truncate rather
than a delete: a dummy object is needed to prevent the CoW machinery
from kicking in. However, a truncate on a non-existent object is
a no-op. If the object doesn't exist in HEAD, a discard request is
effectively ignored, which violates our "discard zeroes data" promise
and breaks REQ_OP_WRITE_ZEROES implementation.
A non-exclusive create on an existing object is also a no-op, so the
fix is to do a compound create+truncate instead.
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
Diffstat (limited to 'drivers/block')
| -rw-r--r-- | drivers/block/rbd.c | 7 |
1 files changed, 6 insertions, 1 deletions
diff --git a/drivers/block/rbd.c b/drivers/block/rbd.c index 0aa95e08664d..fc94e2c45e28 100644 --- a/drivers/block/rbd.c +++ b/drivers/block/rbd.c @@ -1884,6 +1884,8 @@ static void __rbd_obj_setup_discard(struct rbd_obj_request *obj_req, if (rbd_obj_is_entire(obj_req)) { if (obj_req->num_img_extents) { + osd_req_op_init(obj_req->osd_req, which++, + CEPH_OSD_OP_CREATE, 0); opcode = CEPH_OSD_OP_TRUNCATE; } else { osd_req_op_init(obj_req->osd_req, which++, @@ -1917,7 +1919,10 @@ static int rbd_obj_setup_discard(struct rbd_obj_request *obj_req) if (rbd_obj_is_entire(obj_req)) { obj_req->write_state = RBD_OBJ_WRITE_FLAT; - num_osd_ops = 1; /* truncate/delete */ + if (obj_req->num_img_extents) + num_osd_ops = 2; /* create + truncate */ + else + num_osd_ops = 1; /* delete */ } else { if (obj_req->num_img_extents) { obj_req->write_state = RBD_OBJ_WRITE_GUARD; |