DPDK patches and discussions
 help / color / mirror / Atom feed
From: Viacheslav Ovsiienko <viacheslavo@nvidia.com>
To: <dev@dpdk.org>
Cc: <rasland@nvidia.com>, <matan@nvidia.com>, <stable@dpdk.org>
Subject: [dpdk-dev] [PATCH] common/mlx5: fix Netlink port name attribute padding
Date: Sat, 19 Jun 2021 16:56:28 +0300	[thread overview]
Message-ID: <20210619135628.1626-1-viacheslavo@nvidia.com> (raw)

On some kernels the string attributes within Netlink
reply messages might be not padded with zeroes (in cases
when string length is aligned with 4-byte boundary).
This caused wrong physical port names recognition and
mlx5 PMD load failures.

Fixes: 30a86157f6d5 ("net/mlx5: support PF representor")
Cc: stable@dpdk.org

Signed-off-by: Viacheslav Ovsiienko <viacheslavo@nvidia.com>
---
 drivers/common/mlx5/linux/mlx5_nl.c | 22 +++++++++++++++++++++-
 1 file changed, 21 insertions(+), 1 deletion(-)

diff --git a/drivers/common/mlx5/linux/mlx5_nl.c b/drivers/common/mlx5/linux/mlx5_nl.c
index f0d04f9473..3f1912d078 100644
--- a/drivers/common/mlx5/linux/mlx5_nl.c
+++ b/drivers/common/mlx5/linux/mlx5_nl.c
@@ -33,6 +33,8 @@
 #define MLX5_SEND_BUF_SIZE 32768
 /* Receive buffer size for the Netlink socket */
 #define MLX5_RECV_BUF_SIZE 32768
+/* Maximal physical port name length. */
+#define MLX5_PHYS_PORT_NAME_MAX 128
 
 /** Parameters of VLAN devices created by driver. */
 #define MLX5_VMWA_VLAN_DEVICE_PFX "evmlx"
@@ -1191,6 +1193,7 @@ mlx5_nl_switch_info_cb(struct nlmsghdr *nh, void *arg)
 	size_t off = NLMSG_LENGTH(sizeof(struct ifinfomsg));
 	bool switch_id_set = false;
 	bool num_vf_set = false;
+	int len;
 
 	if (nh->nlmsg_type != RTM_NEWLINK)
 		goto error;
@@ -1206,7 +1209,24 @@ mlx5_nl_switch_info_cb(struct nlmsghdr *nh, void *arg)
 			num_vf_set = true;
 			break;
 		case IFLA_PHYS_PORT_NAME:
-			mlx5_translate_port_name((char *)payload, &info);
+			len = RTA_PAYLOAD(ra);
+			/* Some kernels do not pad attributes with zero. */
+			if (len > 0 && len < MLX5_PHYS_PORT_NAME_MAX) {
+				char name[MLX5_PHYS_PORT_NAME_MAX];
+
+				/*
+				 * We can't just patch the message with padding
+				 * zero - it might corrupt the following items
+				 * in the message, we have to copy the string
+				 * by attribute length and pad the copied one.
+				 */
+				memcpy(name, payload, len);
+				name[len] = 0;
+				mlx5_translate_port_name(name, &info);
+			} else {
+				info.name_type =
+					MLX5_PHYS_PORT_NAME_TYPE_UNKNOWN;
+			}
 			break;
 		case IFLA_PHYS_SWITCH_ID:
 			info.switch_id = 0;
-- 
2.18.1


             reply	other threads:[~2021-06-19 13:56 UTC|newest]

Thread overview: 3+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-06-19 13:56 Viacheslav Ovsiienko [this message]
2021-06-20  6:07 ` Matan Azrad
2021-06-20 12:02 ` Raslan Darawsheh

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20210619135628.1626-1-viacheslavo@nvidia.com \
    --to=viacheslavo@nvidia.com \
    --cc=dev@dpdk.org \
    --cc=matan@nvidia.com \
    --cc=rasland@nvidia.com \
    --cc=stable@dpdk.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).