On Sat, Feb 06, 2016 at 11:44:22AM -0700, Sagar Dharia wrote:
> + spin_lock_irqsave(&ctrl->txn_lock, flags);
> + msg = ctrl->tid_tbl[tid];
> + if (msg == NULL || msg->rbuf == NULL) {
> + spin_unlock_irqrestore(&ctrl->txn_lock, flags);
> + dev_err(&ctrl->dev, "Got response to invalid TID:%d, len:%d\n",
> + tid, len);
> + return;
> + }
> + memcpy(msg->rbuf, reply, len);
> + ctrl->tid_tbl[tid] = NULL;
> + if (msg->comp_cb)
> + msg->comp_cb(msg->ctx, 0);
> + spin_unlock_irqrestore(&ctrl->txn_lock, flags);
Do we need to hold the lock for so long (especially with things like the
memcpy())? As far as I can tell we only need the lock for this:
> + msg = ctrl->tid_tbl[tid];
> + ctrl->tid_tbl[tid] = NULL;
> + if (mc == SLIM_MSG_MC_REQUEST_CHANGE_VALUE ||
> + mc == SLIM_MSG_MC_CHANGE_VALUE ||
> + mc == SLIM_MSG_MC_REQUEST_CLEAR_INFORMATION ||
> + mc == SLIM_MSG_MC_CLEAR_INFORMATION)
> + txn->rl += msg->num_bytes;
A switch statement might be nicer here.
Attachment:
signature.asc
Description: PGP signature
