RDMA/siw: Always report immediate post SQ errors

In siw_post_send(), any immediate error encountered during processing of
the work request list must be reported to the caller, even if previous
work requests in that list were just accepted and added to the send queue.

Not reporting those errors confuses the caller, which would wait
indefinitely for the failing and potentially subsequently aborted work
requests completion.

This fixes a case where immediate errors were overwritten by subsequent
code in siw_post_send().

Fixes: 303ae1cdfdf7 ("rdma/siw: application interface")
Link: https://patch.msgid.link/r/[e-mail]
Suggested-by: Stefan Metzmacher <[e-mail]>
Signed-off-by: Bernard Metzler <[e-mail]>
Signed-off-by: Jason Gunthorpe <[e-mail]>

drivers/infiniband/sw/siw/siw_verbs.c

···
 	struct siw_wqe *wqe = tx_wqe(qp);
 
 	unsigned long flags;
-	int rv = 0;
+	int rv = 0, imm_err = 0;
 
 	if (wr && !rdma_is_kernel_res(&qp->base_qp.res)) {
 		siw_dbg_qp(qp, "wr must be empty for user mapped sq\n");
···
 	 * Send directly if SQ processing is not in progress.
 	 * Eventual immediate errors (rv < 0) do not affect the involved
 	 * RI resources (Verbs, 8.3.1) and thus do not prevent from SQ
-	 * processing, if new work is already pending. But rv must be passed
-	 * to caller.
+	 * processing, if new work is already pending. But rv and pointer
+	 * to failed work request must be passed to caller.
 	 */
+	if (unlikely(rv < 0)) {
+		/*
+		 * Immediate error
+		 */
+		siw_dbg_qp(qp, "Immediate error %d\n", rv);
+		imm_err = rv;
+		*bad_wr = wr;
+	}
 	if (wqe->wr_status != SIW_WR_IDLE) {
 		spin_unlock_irqrestore(&qp->sq_lock, flags);
 		goto skip_direct_sending;
···
 
 	up_read(&qp->state_lock);
 
-	if (rv >= 0)
-		return 0;
-	/*
-	 * Immediate error
-	 */
-	siw_dbg_qp(qp, "error %d\n", rv);
+	if (unlikely(imm_err))
+		return imm_err;
 
-	*bad_wr = wr;
-	return rv;
+	return (rv >= 0) ? 0 : rv;
 }
 
 /*