diff options
| author | Lars Ellenberg <lars.ellenberg@linbit.com> | 2015-05-18 15:08:46 +0300 |
|---|---|---|
| committer | Jens Axboe <axboe@fb.com> | 2015-11-25 19:22:03 +0300 |
| commit | f85d9f2d02cdcd1b79e00fccd667b37b251ba3ac (patch) | |
| tree | 44a62bbab00dc8eecfbdd9ab312eb4686994704c | |
| parent | 706447861bb210cf2fb6a58bc1d29a6636175987 (diff) | |
| download | linux-f85d9f2d02cdcd1b79e00fccd667b37b251ba3ac.tar.xz | |
drbd: fix "endless" transfer log walk in protocol A
Don't remember a DRBD request as ack_pending, if it is not.
In protocol A, we usually clear RQ_NET_PENDING at the same time we set
RQ_NET_SENT, so when deciding to remember it as ack_pending,
mod_rq_state needs to look at the current request state,
not at the previous state before the current modification was applied.
This should prevent advance_conn_req_ack_pending() from walking the full
transfer log just to find NULL in protocol A, which would cause serious
performance degradation with many "in-flight" requests, e.g. when
working via DRBD-proxy, or with a huge bandwidth-delay product.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
| -rw-r--r-- | drivers/block/drbd/drbd_req.c | 2 |
1 files changed, 1 insertions, 1 deletions
diff --git a/drivers/block/drbd/drbd_req.c b/drivers/block/drbd/drbd_req.c index 7907fb562388..2255dcfebd2b 100644 --- a/drivers/block/drbd/drbd_req.c +++ b/drivers/block/drbd/drbd_req.c @@ -458,7 +458,7 @@ static void mod_rq_state(struct drbd_request *req, struct bio_and_error *m, atomic_add(req->i.size >> 9, &device->ap_in_flight); set_if_null_req_not_net_done(peer_device, req); } - if (s & RQ_NET_PENDING) + if (req->rq_state & RQ_NET_PENDING) set_if_null_req_ack_pending(peer_device, req); } |