[PATCH] RDMA/mana_ib: Fix a bug when the PF indicates more entries for registering memory on first packet

[PATCH] RDMA/mana_ib: Fix a bug when the PF indicates more entries for registering memory on first packet

[longli]

From: Long Li <longli@microsoft.com>

When registering memory in a large chunk that doesn't fit into a single PF
message, the PF may return GDMA_STATUS_MORE_ENTRIES on the first message if
there are more messages needed for registering more chunks.

Fix the VF to make it process the correct return code.

Fixes: 0266a177631d ("RDMA/mana_ib: Add a driver for Microsoft Azure Network Adapter")
Signed-off-by: Long Li <longli@microsoft.com>
---
 drivers/infiniband/hw/mana/main.c | 22 ++++++++++++----------
 1 file changed, 12 insertions(+), 10 deletions(-)

diff --git a/drivers/infiniband/hw/mana/main.c b/drivers/infiniband/hw/mana/main.c
index 8b3bc302d6f3..7be4c3adb4e2 100644
--- a/drivers/infiniband/hw/mana/main.c
+++ b/drivers/infiniband/hw/mana/main.c
@@ -249,7 +249,8 @@ static int
 mana_ib_gd_first_dma_region(struct mana_ib_dev *dev,
 			    struct gdma_context *gc,
 			    struct gdma_create_dma_region_req *create_req,
-			    size_t num_pages, mana_handle_t *gdma_region)
+			    size_t num_pages, mana_handle_t *gdma_region,
+			    u32 expected_status)
 {
 	struct gdma_create_dma_region_resp create_resp = {};
 	unsigned int create_req_msg_size;
@@ -261,7 +262,7 @@ mana_ib_gd_first_dma_region(struct mana_ib_dev *dev,
 
 	err = mana_gd_send_request(gc, create_req_msg_size, create_req,
 				   sizeof(create_resp), &create_resp);
-	if (err || create_resp.hdr.status) {
+	if (err || create_resp.hdr.status != expected_status) {
 		ibdev_dbg(&dev->ib_dev,
 			  "Failed to create DMA region: %d, 0x%x\n",
 			  err, create_resp.hdr.status);
@@ -372,14 +373,21 @@ int mana_ib_gd_create_dma_region(struct mana_ib_dev *dev, struct ib_umem *umem,
 
 	page_addr_list = create_req->page_addr_list;
 	rdma_umem_for_each_dma_block(umem, &biter, page_sz) {
+		u32 expected_status = 0;
+
 		page_addr_list[tail++] = rdma_block_iter_dma_address(&biter);
 		if (tail < num_pages_to_handle)
 			continue;
 
+		if (num_pages_processed + num_pages_to_handle <
+		    num_pages_total)
+			expected_status = GDMA_STATUS_MORE_ENTRIES;
+
 		if (!num_pages_processed) {
 			/* First create message */
 			err = mana_ib_gd_first_dma_region(dev, gc, create_req,
-							  tail, gdma_region);
+							  tail, gdma_region,
+							  expected_status);
 			if (err)
 				goto out;
 
@@ -392,14 +400,8 @@ int mana_ib_gd_create_dma_region(struct mana_ib_dev *dev, struct ib_umem *umem,
 			page_addr_list = add_req->page_addr_list;
 		} else {
 			/* Subsequent create messages */
-			u32 expected_s = 0;
-
-			if (num_pages_processed + num_pages_to_handle <
-			    num_pages_total)
-				expected_s = GDMA_STATUS_MORE_ENTRIES;
-
 			err = mana_ib_gd_add_dma_region(dev, gc, add_req, tail,
-							expected_s);
+							expected_status);
 			if (err)
 				break;
 		}
-- 
2.17.1

Re: [PATCH] RDMA/mana_ib: Fix a bug when the PF indicates more entries for registering memory on first packet

[Jason Gunthorpe]

On Wed, Feb 15, 2023 at 04:32:02PM -0800, longli@linuxonhyperv.com wrote:
> From: Long Li <longli@microsoft.com>
> 
> When registering memory in a large chunk that doesn't fit into a single PF
> message, the PF may return GDMA_STATUS_MORE_ENTRIES on the first message if
> there are more messages needed for registering more chunks.
> 
> Fix the VF to make it process the correct return code.
> 
> Fixes: 0266a177631d ("RDMA/mana_ib: Add a driver for Microsoft Azure Network Adapter")
> Signed-off-by: Long Li <longli@microsoft.com>
> ---
>  drivers/infiniband/hw/mana/main.c | 22 ++++++++++++----------
>  1 file changed, 12 insertions(+), 10 deletions(-)

Applied to for-next, thanks

Jason