On 03/08/2016 10:16 AM, Christoph Hellwig wrote:
+ reg->mr = ib_mr_pool_get(qp, &qp->rdma_mrs);
+ if (!reg->mr) {
+ pr_info("failed to allocate MR from pool\n");
+ ret = -EAGAIN;
+ goto out_free;
+ }
The above pr_info() message does not provide enough context information
to allow a user to figure out why that message was reported. Please
leave that message out and make the callers of this function report MR
allocation failure wherever useful.
+ ret = ib_map_mr_sg(reg->mr, sg, nents, offset,
+ PAGE_SIZE);
[ ... ]
+ sg = sg_next(sg);
ib_map_mr_sg() processed 'ret' elements of scatterlist 'sg'. So why is
'sg' only advanced by one element at the end of the loop?
+ for (j = 0; j < nr_sge; j++, sg = sg_next(sg)) {
+ rdma_wr->wr.num_sge++;
+
+ sge->addr = ib_sg_dma_address(dev, sg) + offset;
+ sge->length = ib_sg_dma_len(dev, sg) - offset;
+ sge->lkey = qp->pd->local_dma_lkey;
+
+ total_len += sge->length;
+ sge++;
+ sge_left--;
+ offset = 0;
+ }
Shouldn't there be a break statement in the above loop that stops this
loop if the end of the sg list has been reached? Less than nr_sge
iterations will be needed to reach the end of the sg list if the length
of the sg list is not a multiple of nr_sge.
Bart.
--
To unsubscribe from this list: send the line "unsubscribe target-devel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
