RDMA/bnxt_re: Fix the sideband buffer size handling for FW commands

bnxt_qplib_rcfw_alloc_sbuf allocates 24 bytes and it is better to fit
on stack variables. This way we can avoid unwanted kmalloc call.
Call dma_alloc_coherent directly instead of wrapper
bnxt_qplib_rcfw_alloc_sbuf.

Also, FW expects the side buffer  needs to be aligned to
BNXT_QPLIB_CMDQE_UNITS(16B). So align the size to have the
extra padding bytes.

Signed-off-by: Kashyap Desai <kashyap.desai@broadcom.com>
Signed-off-by: Hongguang Gao <hongguang.gao@broadcom.com>
Signed-off-by: Selvin Xavier <selvin.xavier@broadcom.com>
Link: https://lore.kernel.org/r/1691052326-32143-4-git-send-email-selvin.xavier@broadcom.com
Signed-off-by: Leon Romanovsky <leon@kernel.org>

3 files changed, 55 insertions, 80 deletions

diff --git a/drivers/infiniband/hw/bnxt_re/qplib_fp.c b/drivers/infiniband/hw/bnxt_re/qplib_fp.c
index f9dee0d2da9c..282e34e8739e 100644
--- a/drivers/infiniband/hw/bnxt_re/qplib_fp.c
+++ b/drivers/infiniband/hw/bnxt_re/qplib_fp.c
@@ -709,7 +709,7 @@ int bnxt_qplib_query_srq(struct bnxt_qplib_res *res,
 	struct bnxt_qplib_rcfw *rcfw = res->rcfw;
 	struct creq_query_srq_resp resp = {};
 	struct bnxt_qplib_cmdqmsg msg = {};
-	struct bnxt_qplib_rcfw_sbuf *sbuf;
+	struct bnxt_qplib_rcfw_sbuf sbuf;
 	struct creq_query_srq_resp_sb *sb;
 	struct cmdq_query_srq req = {};
 	int rc = 0;
@@ -719,17 +719,20 @@ int bnxt_qplib_query_srq(struct bnxt_qplib_res *res,
 				 sizeof(req));
 
 	/* Configure the request */
-	sbuf = bnxt_qplib_rcfw_alloc_sbuf(rcfw, sizeof(*sb));
-	if (!sbuf)
+	sbuf.size = ALIGN(sizeof(*sb), BNXT_QPLIB_CMDQE_UNITS);
+	sbuf.sb = dma_alloc_coherent(&rcfw->pdev->dev, sbuf.size,
+				     &sbuf.dma_addr, GFP_KERNEL);
+	if (!sbuf.sb)
 		return -ENOMEM;
-	req.resp_size = sizeof(*sb) / BNXT_QPLIB_CMDQE_UNITS;
+	req.resp_size = sbuf.size / BNXT_QPLIB_CMDQE_UNITS;
 	req.srq_cid = cpu_to_le32(srq->id);
-	sb = sbuf->sb;
-	bnxt_qplib_fill_cmdqmsg(&msg, &req, &resp, sbuf, sizeof(req),
+	sb = sbuf.sb;
+	bnxt_qplib_fill_cmdqmsg(&msg, &req, &resp, &sbuf, sizeof(req),
 				sizeof(resp), 0);
 	rc = bnxt_qplib_rcfw_send_message(rcfw, &msg);
 	srq->threshold = le16_to_cpu(sb->srq_limit);
-	bnxt_qplib_rcfw_free_sbuf(rcfw, sbuf);
+	dma_free_coherent(&rcfw->pdev->dev, sbuf.size,
+			  sbuf.sb, sbuf.dma_addr);
 
 	return rc;
 }
@@ -1347,24 +1350,26 @@ int bnxt_qplib_query_qp(struct bnxt_qplib_res *res, struct bnxt_qplib_qp *qp)
 	struct bnxt_qplib_rcfw *rcfw = res->rcfw;
 	struct creq_query_qp_resp resp = {};
 	struct bnxt_qplib_cmdqmsg msg = {};
-	struct bnxt_qplib_rcfw_sbuf *sbuf;
+	struct bnxt_qplib_rcfw_sbuf sbuf;
 	struct creq_query_qp_resp_sb *sb;
 	struct cmdq_query_qp req = {};
 	u32 temp32[4];
 	int i, rc = 0;
 
+	sbuf.size = ALIGN(sizeof(*sb), BNXT_QPLIB_CMDQE_UNITS);
+	sbuf.sb = dma_alloc_coherent(&rcfw->pdev->dev, sbuf.size,
+				     &sbuf.dma_addr, GFP_KERNEL);
+	if (!sbuf.sb)
+		return -ENOMEM;
+	sb = sbuf.sb;
+
 	bnxt_qplib_rcfw_cmd_prep((struct cmdq_base *)&req,
 				 CMDQ_BASE_OPCODE_QUERY_QP,
 				 sizeof(req));
 
-	sbuf = bnxt_qplib_rcfw_alloc_sbuf(rcfw, sizeof(*sb));
-	if (!sbuf)
-		return -ENOMEM;
-	sb = sbuf->sb;
-
 	req.qp_cid = cpu_to_le32(qp->id);
-	req.resp_size = sizeof(*sb) / BNXT_QPLIB_CMDQE_UNITS;
-	bnxt_qplib_fill_cmdqmsg(&msg, &req, &resp, sbuf, sizeof(req),
+	req.resp_size = sbuf.size / BNXT_QPLIB_CMDQE_UNITS;
+	bnxt_qplib_fill_cmdqmsg(&msg, &req, &resp, &sbuf, sizeof(req),
 				sizeof(resp), 0);
 	rc = bnxt_qplib_rcfw_send_message(rcfw, &msg);
 	if (rc)
@@ -1423,7 +1428,8 @@ int bnxt_qplib_query_qp(struct bnxt_qplib_res *res, struct bnxt_qplib_qp *qp)
 	memcpy(qp->smac, sb->src_mac, 6);
 	qp->vlan_id = le16_to_cpu(sb->vlan_pcp_vlan_dei_vlan_id);
 bail:
-	bnxt_qplib_rcfw_free_sbuf(rcfw, sbuf);
+	dma_free_coherent(&rcfw->pdev->dev, sbuf.size,
+			  sbuf.sb, sbuf.dma_addr);
 	return rc;
 }
 
diff --git a/drivers/infiniband/hw/bnxt_re/qplib_rcfw.c b/drivers/infiniband/hw/bnxt_re/qplib_rcfw.c
index b30e66b64827..9d26871af296 100644
--- a/drivers/infiniband/hw/bnxt_re/qplib_rcfw.c
+++ b/drivers/infiniband/hw/bnxt_re/qplib_rcfw.c
@@ -335,7 +335,8 @@ static int __send_message(struct bnxt_qplib_rcfw *rcfw,
 					  cpu_to_le64(sbuf->dma_addr));
 		__set_cmdq_base_resp_size(msg->req, msg->req_sz,
 					  ALIGN(sbuf->size,
-						BNXT_QPLIB_CMDQE_UNITS));
+						BNXT_QPLIB_CMDQE_UNITS) /
+						BNXT_QPLIB_CMDQE_UNITS);
 	}
 
 	preq = (u8 *)msg->req;
@@ -1196,34 +1197,3 @@ int bnxt_qplib_enable_rcfw_channel(struct bnxt_qplib_rcfw *rcfw,
 
 	return 0;
 }
-
-struct bnxt_qplib_rcfw_sbuf *bnxt_qplib_rcfw_alloc_sbuf(
-		struct bnxt_qplib_rcfw *rcfw,
-		u32 size)
-{
-	struct bnxt_qplib_rcfw_sbuf *sbuf;
-
-	sbuf = kzalloc(sizeof(*sbuf), GFP_KERNEL);
-	if (!sbuf)
-		return NULL;
-
-	sbuf->size = size;
-	sbuf->sb = dma_alloc_coherent(&rcfw->pdev->dev, sbuf->size,
-				      &sbuf->dma_addr, GFP_KERNEL);
-	if (!sbuf->sb)
-		goto bail;
-
-	return sbuf;
-bail:
-	kfree(sbuf);
-	return NULL;
-}
-
-void bnxt_qplib_rcfw_free_sbuf(struct bnxt_qplib_rcfw *rcfw,
-			       struct bnxt_qplib_rcfw_sbuf *sbuf)
-{
-	if (sbuf->sb)
-		dma_free_coherent(&rcfw->pdev->dev, sbuf->size,
-				  sbuf->sb, sbuf->dma_addr);
-	kfree(sbuf);
-}
diff --git a/drivers/infiniband/hw/bnxt_re/qplib_sp.c b/drivers/infiniband/hw/bnxt_re/qplib_sp.c
index b77928ac0217..05ee8fdb44ad 100644
--- a/drivers/infiniband/hw/bnxt_re/qplib_sp.c
+++ b/drivers/infiniband/hw/bnxt_re/qplib_sp.c
@@ -94,7 +94,7 @@ int bnxt_qplib_get_dev_attr(struct bnxt_qplib_rcfw *rcfw,
 	struct creq_query_func_resp resp = {};
 	struct bnxt_qplib_cmdqmsg msg = {};
 	struct creq_query_func_resp_sb *sb;
-	struct bnxt_qplib_rcfw_sbuf *sbuf;
+	struct bnxt_qplib_rcfw_sbuf sbuf;
 	struct cmdq_query_func req = {};
 	u8 *tqm_alloc;
 	int i, rc = 0;
@@ -104,16 +104,14 @@ int bnxt_qplib_get_dev_attr(struct bnxt_qplib_rcfw *rcfw,
 				 CMDQ_BASE_OPCODE_QUERY_FUNC,
 				 sizeof(req));
 
-	sbuf = bnxt_qplib_rcfw_alloc_sbuf(rcfw, sizeof(*sb));
-	if (!sbuf) {
-		dev_err(&rcfw->pdev->dev,
-			"SP: QUERY_FUNC alloc side buffer failed\n");
+	sbuf.size = ALIGN(sizeof(*sb), BNXT_QPLIB_CMDQE_UNITS);
+	sbuf.sb = dma_alloc_coherent(&rcfw->pdev->dev, sbuf.size,
+				     &sbuf.dma_addr, GFP_KERNEL);
+	if (!sbuf.sb)
 		return -ENOMEM;
-	}
-
-	sb = sbuf->sb;
-	req.resp_size = sizeof(*sb) / BNXT_QPLIB_CMDQE_UNITS;
-	bnxt_qplib_fill_cmdqmsg(&msg, &req, &resp, sbuf, sizeof(req),
+	sb = sbuf.sb;
+	req.resp_size = sbuf.size / BNXT_QPLIB_CMDQE_UNITS;
+	bnxt_qplib_fill_cmdqmsg(&msg, &req, &resp, &sbuf, sizeof(req),
 				sizeof(resp), 0);
 	rc = bnxt_qplib_rcfw_send_message(rcfw, &msg);
 	if (rc)
@@ -174,7 +172,8 @@ int bnxt_qplib_get_dev_attr(struct bnxt_qplib_rcfw *rcfw,
 
 	attr->is_atomic = bnxt_qplib_is_atomic_cap(rcfw);
 bail:
-	bnxt_qplib_rcfw_free_sbuf(rcfw, sbuf);
+	dma_free_coherent(&rcfw->pdev->dev, sbuf.size,
+			  sbuf.sb, sbuf.dma_addr);
 	return rc;
 }
 
@@ -717,23 +716,22 @@ int bnxt_qplib_get_roce_stats(struct bnxt_qplib_rcfw *rcfw,
 	struct creq_query_roce_stats_resp_sb *sb;
 	struct cmdq_query_roce_stats req = {};
 	struct bnxt_qplib_cmdqmsg msg = {};
-	struct bnxt_qplib_rcfw_sbuf *sbuf;
+	struct bnxt_qplib_rcfw_sbuf sbuf;
 	int rc = 0;
 
 	bnxt_qplib_rcfw_cmd_prep((struct cmdq_base *)&req,
 				 CMDQ_BASE_OPCODE_QUERY_ROCE_STATS,
 				 sizeof(req));
 
-	sbuf = bnxt_qplib_rcfw_alloc_sbuf(rcfw, sizeof(*sb));
-	if (!sbuf) {
-		dev_err(&rcfw->pdev->dev,
-			"SP: QUERY_ROCE_STATS alloc side buffer failed\n");
+	sbuf.size = ALIGN(sizeof(*sb), BNXT_QPLIB_CMDQE_UNITS);
+	sbuf.sb = dma_alloc_coherent(&rcfw->pdev->dev, sbuf.size,
+				     &sbuf.dma_addr, GFP_KERNEL);
+	if (!sbuf.sb)
 		return -ENOMEM;
-	}
+	sb = sbuf.sb;
 
-	sb = sbuf->sb;
-	req.resp_size = sizeof(*sb) / BNXT_QPLIB_CMDQE_UNITS;
-	bnxt_qplib_fill_cmdqmsg(&msg, &req, &resp, sbuf, sizeof(req),
+	req.resp_size = sbuf.size / BNXT_QPLIB_CMDQE_UNITS;
+	bnxt_qplib_fill_cmdqmsg(&msg, &req, &resp, &sbuf, sizeof(req),
 				sizeof(resp), 0);
 	rc = bnxt_qplib_rcfw_send_message(rcfw, &msg);
 	if (rc)
@@ -789,7 +787,8 @@ int bnxt_qplib_get_roce_stats(struct bnxt_qplib_rcfw *rcfw,
 	}
 
 bail:
-	bnxt_qplib_rcfw_free_sbuf(rcfw, sbuf);
+	dma_free_coherent(&rcfw->pdev->dev, sbuf.size,
+			  sbuf.sb, sbuf.dma_addr);
 	return rc;
 }
 
@@ -800,32 +799,31 @@ int bnxt_qplib_qext_stat(struct bnxt_qplib_rcfw *rcfw, u32 fid,
 	struct creq_query_roce_stats_ext_resp_sb *sb;
 	struct cmdq_query_roce_stats_ext req = {};
 	struct bnxt_qplib_cmdqmsg msg = {};
-	struct bnxt_qplib_rcfw_sbuf *sbuf;
+	struct bnxt_qplib_rcfw_sbuf sbuf;
 	int rc;
 
-	sbuf = bnxt_qplib_rcfw_alloc_sbuf(rcfw, sizeof(*sb));
-	if (!sbuf) {
-		dev_err(&rcfw->pdev->dev,
-			"SP: QUERY_ROCE_STATS_EXT alloc sb failed");
+	sbuf.size = ALIGN(sizeof(*sb), BNXT_QPLIB_CMDQE_UNITS);
+	sbuf.sb = dma_alloc_coherent(&rcfw->pdev->dev, sbuf.size,
+				     &sbuf.dma_addr, GFP_KERNEL);
+	if (!sbuf.sb)
 		return -ENOMEM;
-	}
 
+	sb = sbuf.sb;
 	bnxt_qplib_rcfw_cmd_prep((struct cmdq_base *)&req,
 				 CMDQ_QUERY_ROCE_STATS_EXT_OPCODE_QUERY_ROCE_STATS,
 				 sizeof(req));
 
-	req.resp_size = ALIGN(sizeof(*sb), BNXT_QPLIB_CMDQE_UNITS);
-	req.resp_addr = cpu_to_le64(sbuf->dma_addr);
+	req.resp_size = sbuf.size / BNXT_QPLIB_CMDQE_UNITS;
+	req.resp_addr = cpu_to_le64(sbuf.dma_addr);
 	req.function_id = cpu_to_le32(fid);
 	req.flags = cpu_to_le16(CMDQ_QUERY_ROCE_STATS_EXT_FLAGS_FUNCTION_ID);
 
-	bnxt_qplib_fill_cmdqmsg(&msg, &req, &resp, sbuf, sizeof(req),
+	bnxt_qplib_fill_cmdqmsg(&msg, &req, &resp, &sbuf, sizeof(req),
 				sizeof(resp), 0);
 	rc = bnxt_qplib_rcfw_send_message(rcfw, &msg);
 	if (rc)
 		goto bail;
 
-	sb = sbuf->sb;
 	estat->tx_atomic_req = le64_to_cpu(sb->tx_atomic_req_pkts);
 	estat->tx_read_req = le64_to_cpu(sb->tx_read_req_pkts);
 	estat->tx_read_res = le64_to_cpu(sb->tx_read_res_pkts);
@@ -849,7 +847,8 @@ int bnxt_qplib_qext_stat(struct bnxt_qplib_rcfw *rcfw, u32 fid,
 	estat->rx_ecn_marked = le64_to_cpu(sb->rx_ecn_marked_pkts);
 
 bail:
-	bnxt_qplib_rcfw_free_sbuf(rcfw, sbuf);
+	dma_free_coherent(&rcfw->pdev->dev, sbuf.size,
+			  sbuf.sb, sbuf.dma_addr);
 	return rc;
 }