[dpdk-stable] patch 'common/mlx5: fix Netlink receive message buffer size' has been queued to stable release 20.11.3

[luca.boccassi at gmail.com]

Hi,

FYI, your patch has been queued to stable release 20.11.3

Note it hasn't been pushed to http://dpdk.org/browse/dpdk-stable yet.
It will be pushed if I get no objections before 07/14/21. So please
shout if anyone has objections.

Also note that after the patch there's a diff of the upstream commit vs the
patch applied to the branch. This will indicate if there was any rebasing
needed to apply to the stable branch. If there were code changes for rebasing
(ie: not only metadata diffs), please double check that the rebase was
correctly done.

Queued patches are on a temporary branch at:
https://github.com/bluca/dpdk-stable

This queued commit can be viewed at:
https://github.com/bluca/dpdk-stable/commit/3c47e4e2e5fcb69131d2a7bdfa2a7659e187a8f2

Thanks.

Luca Boccassi

---
>From 3c47e4e2e5fcb69131d2a7bdfa2a7659e187a8f2 Mon Sep 17 00:00:00 2001
From: Viacheslav Ovsiienko <viacheslavo at nvidia.com>
Date: Thu, 1 Jul 2021 10:31:33 +0300
Subject: [PATCH] common/mlx5: fix Netlink receive message buffer size

[ upstream commit 32d1e4dbadba4bc3523d8b354458a4b979e5c2e6 ]

If there are many VFs the Netlink message length sent by kernel
in reply to RTM_GETLINK request can be large. We should query
the size of message being received in advance and allocate
the large enough buffer to handle these large messages.

Fixes: ccdcba53a3f4 ("net/mlx5: use Netlink to add/remove MAC addresses")

Signed-off-by: Viacheslav Ovsiienko <viacheslavo at nvidia.com>
---
 drivers/common/mlx5/linux/mlx5_nl.c | 69 +++++++++++++++++++++++------
 1 file changed, 55 insertions(+), 14 deletions(-)

diff --git a/drivers/common/mlx5/linux/mlx5_nl.c b/drivers/common/mlx5/linux/mlx5_nl.c
index 94afa04b85..1f765dca07 100644
--- a/drivers/common/mlx5/linux/mlx5_nl.c
+++ b/drivers/common/mlx5/linux/mlx5_nl.c
@@ -189,8 +189,8 @@ int
 mlx5_nl_init(int protocol)
 {
 	int fd;
-	int sndbuf_size = MLX5_SEND_BUF_SIZE;
-	int rcvbuf_size = MLX5_RECV_BUF_SIZE;
+	int buf_size;
+	socklen_t opt_size;
 	struct sockaddr_nl local = {
 		.nl_family = AF_NETLINK,
 	};
@@ -201,16 +201,36 @@ mlx5_nl_init(int protocol)
 		rte_errno = errno;
 		return -rte_errno;
 	}
-	ret = setsockopt(fd, SOL_SOCKET, SO_SNDBUF, &sndbuf_size, sizeof(int));
+	opt_size = sizeof(buf_size);
+	ret = getsockopt(fd, SOL_SOCKET, SO_SNDBUF, &buf_size, &opt_size);
 	if (ret == -1) {
 		rte_errno = errno;
 		goto error;
 	}
-	ret = setsockopt(fd, SOL_SOCKET, SO_RCVBUF, &rcvbuf_size, sizeof(int));
+	DRV_LOG(DEBUG, "Netlink socket send buffer: %d", buf_size);
+	if (buf_size < MLX5_SEND_BUF_SIZE) {
+		ret = setsockopt(fd, SOL_SOCKET, SO_SNDBUF,
+				 &buf_size, sizeof(buf_size));
+		if (ret == -1) {
+			rte_errno = errno;
+			goto error;
+		}
+	}
+	opt_size = sizeof(buf_size);
+	ret = getsockopt(fd, SOL_SOCKET, SO_RCVBUF, &buf_size, &opt_size);
 	if (ret == -1) {
 		rte_errno = errno;
 		goto error;
 	}
+	DRV_LOG(DEBUG, "Netlink socket recv buffer: %d", buf_size);
+	if (buf_size < MLX5_RECV_BUF_SIZE) {
+		ret = setsockopt(fd, SOL_SOCKET, SO_RCVBUF,
+				 &buf_size, sizeof(buf_size));
+		if (ret == -1) {
+			rte_errno = errno;
+			goto error;
+		}
+	}
 	ret = bind(fd, (struct sockaddr *)&local, sizeof(local));
 	if (ret == -1) {
 		rte_errno = errno;
@@ -332,11 +352,7 @@ mlx5_nl_recv(int nlsk_fd, uint32_t sn, int (*cb)(struct nlmsghdr *, void *arg),
 	     void *arg)
 {
 	struct sockaddr_nl sa;
-	void *buf = mlx5_malloc(0, MLX5_RECV_BUF_SIZE, 0, SOCKET_ID_ANY);
-	struct iovec iov = {
-		.iov_base = buf,
-		.iov_len = MLX5_RECV_BUF_SIZE,
-	};
+	struct iovec iov;
 	struct msghdr msg = {
 		.msg_name = &sa,
 		.msg_namelen = sizeof(sa),
@@ -344,18 +360,43 @@ mlx5_nl_recv(int nlsk_fd, uint32_t sn, int (*cb)(struct nlmsghdr *, void *arg),
 		/* One message at a time */
 		.msg_iovlen = 1,
 	};
+	void *buf = NULL;
 	int multipart = 0;
 	int ret = 0;
 
-	if (!buf) {
-		rte_errno = ENOMEM;
-		return -rte_errno;
-	}
 	do {
 		struct nlmsghdr *nh;
-		int recv_bytes = 0;
+		int recv_bytes;
 
 		do {
+			/* Query length of incoming message. */
+			iov.iov_base = NULL;
+			iov.iov_len = 0;
+			recv_bytes = recvmsg(nlsk_fd, &msg,
+					     MSG_PEEK | MSG_TRUNC);
+			if (recv_bytes < 0) {
+				rte_errno = errno;
+				ret = -rte_errno;
+				goto exit;
+			}
+			if (recv_bytes == 0) {
+				rte_errno = ENODATA;
+				ret = -rte_errno;
+				goto exit;
+			}
+			/* Allocate buffer to fetch the message. */
+			if (recv_bytes < MLX5_RECV_BUF_SIZE)
+				recv_bytes = MLX5_RECV_BUF_SIZE;
+			mlx5_free(buf);
+			buf = mlx5_malloc(0, recv_bytes, 0, SOCKET_ID_ANY);
+			if (!buf) {
+				rte_errno = ENOMEM;
+				ret = -rte_errno;
+				goto exit;
+			}
+			/* Fetch the message. */
+			iov.iov_base = buf;
+			iov.iov_len = recv_bytes;
 			recv_bytes = recvmsg(nlsk_fd, &msg, 0);
 			if (recv_bytes == -1) {
 				rte_errno = errno;
-- 
2.30.2

---
  Diff of the applied patch vs upstream commit (please double-check if non-empty:
---
--- -	2021-07-12 13:41:41.699249959 +0100
+++ 0099-common-mlx5-fix-Netlink-receive-message-buffer-size.patch	2021-07-12 13:41:36.826129610 +0100
@@ -1 +1 @@
-From 32d1e4dbadba4bc3523d8b354458a4b979e5c2e6 Mon Sep 17 00:00:00 2001
+From 3c47e4e2e5fcb69131d2a7bdfa2a7659e187a8f2 Mon Sep 17 00:00:00 2001
@@ -5,0 +6,2 @@
+[ upstream commit 32d1e4dbadba4bc3523d8b354458a4b979e5c2e6 ]
+
@@ -12 +13,0 @@
-Cc: stable at dpdk.org
@@ -20 +21 @@
-index 3f1912d078..dc8dafd0a8 100644
+index 94afa04b85..1f765dca07 100644