On 3/20/2018 4:07 PM, Jason Gunthorpe wrote:
On Tue, Mar 20, 2018 at 02:38:17PM +0200, Yishai Hadas wrote:
On 3/14/2018 12:07 AM, Jason Gunthorpe wrote:
From: Jason Gunthorpe <jgg@xxxxxxxxxxxx>
Instead of creating a dummy structure in kern-abi.h, just use the
normal command structure provided by the kernel with some simple
container_of help.
We didn't come up to finalize the review on this patch, however it was
already merged..
It was on the list for a week and is minor, if anything is found we
can fix in a followup. It is holding up sending the patches for the
32/64 compat fixes.
diff --git a/libibverbs/cmd.c b/libibverbs/cmd.c
index e30603dd5d4194..36d62eb09ad107 100644
@@ -896,7 +896,8 @@ int ibv_cmd_create_qp_ex2(struct ibv_context *context,
IBV_INIT_CMD_RESP_EX_V(cmd, cmd_core_size, cmd_size, CREATE_QP, resp,
resp_core_size, resp_size);
- err = create_qp_ex_common(qp, qp_attr, vxrcd, &cmd->base);
+ err = create_qp_ex_common(qp, qp_attr, vxrcd,
+ ibv_create_cq_ex_to_reg(cmd));
Why ibv_create_cq_ex_to_reg ? we are in create_qp_ex. (same note for its
usage around).
I don't really understand this question.. None of the function
argument types were really changed, the use of ibv_create_cq_ex_to_reg
is just replacing cmd->base, which matches the way the existing code
already flowed.
Yes, it is confusing and weird that something called
create_qp_ex_common accepts a mixture of _ex and not _ex structures,
but that is how it is..
I'm referring to the 'cq' vs 'qp' mismatch, typo ?
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
