diff mbox series

[net-next,2/4] rtnetlink: extend RTEXT_FILTER_SKIP_STATS to IFLA_VF_INFO

Message ID 20210123045321.2797360-3-edwin.peer@broadcom.com (mailing list archive)
State Changes Requested
Delegated to: Netdev Maintainers
Headers show
Series support for 256 VFs in RTM_GETLINK | expand

Checks

Context Check Description
netdev/cover_letter success Link
netdev/fixes_present success Link
netdev/patch_count success Link
netdev/tree_selection success Clearly marked for net-next
netdev/subject_prefix success Link
netdev/cc_maintainers warning 13 maintainers not CCed: leon@kernel.org hadarh@mellanox.com ogerlitz@mellanox.com toke@redhat.com davem@davemloft.net zhudi21@huawei.com jwi@linux.ibm.com saeedm@mellanox.com edumazet@google.com roopa@cumulusnetworks.com ast@kernel.org eranbe@mellanox.com eugenia@mellanox.com
netdev/source_inline success Was 0 now: 0
netdev/verify_signedoff success Link
netdev/module_param success Was 0 now: 0
netdev/build_32bit success Errors and warnings before: 1 this patch: 1
netdev/kdoc success Errors and warnings before: 0 this patch: 0
netdev/verify_fixes success Link
netdev/checkpatch success total: 0 errors, 0 warnings, 0 checks, 122 lines checked
netdev/build_allmodconfig_warn success Errors and warnings before: 1 this patch: 1
netdev/header_inline success Link
netdev/stable success Stable not CCed

Commit Message

Edwin Peer Jan. 23, 2021, 4:53 a.m. UTC 
This filter already exists for excluding IPv6 SNMP stats. Extend its
definition to also exclude IFLA_VF_INFO stats in RTM_GETLINK.

This patch constitutes a partial fix for a netlink attribute nesting
overflow bug in IFLA_VFINFO_LIST. By excluding the stats when the
requester doesn't need them, the truncation of the VF list is avoided.

While it was technically only the stats added in commit c5a9f6f0ab40
("net/core: Add drop counters to VF statistics") breaking the camel's
back, the appreciable size of the stats data should never have been
included without due consideration for the maximum number of VFs
supported by PCI.

Fixes: 3b766cd83232 ("net/core: Add reading VF statistics through the PF netdevice")
Fixes: c5a9f6f0ab40 ("net/core: Add drop counters to VF statistics")
Signed-off-by: Edwin Peer <edwin.peer@broadcom.com>
---
 net/core/rtnetlink.c | 96 +++++++++++++++++++++++---------------------
 1 file changed, 51 insertions(+), 45 deletions(-)

Comments

Jakub Kicinski Jan. 26, 2021, 1:55 a.m. UTC | #1 
On Fri, 22 Jan 2021 20:53:19 -0800 Edwin Peer wrote:
> This filter already exists for excluding IPv6 SNMP stats. Extend its
> definition to also exclude IFLA_VF_INFO stats in RTM_GETLINK.
> 
> This patch constitutes a partial fix for a netlink attribute nesting
> overflow bug in IFLA_VFINFO_LIST. By excluding the stats when the
> requester doesn't need them, the truncation of the VF list is avoided.
> 
> While it was technically only the stats added in commit c5a9f6f0ab40
> ("net/core: Add drop counters to VF statistics") breaking the camel's
> back, the appreciable size of the stats data should never have been
> included without due consideration for the maximum number of VFs
> supported by PCI.
> 
> Fixes: 3b766cd83232 ("net/core: Add reading VF statistics through the PF netdevice")
> Fixes: c5a9f6f0ab40 ("net/core: Add drop counters to VF statistics")
> Signed-off-by: Edwin Peer <edwin.peer@broadcom.com>

Could you include in the commit message the size breakdown of a single
VF nest? With and without efficient unaligned 64b access?
Edwin Peer Jan. 26, 2021, 10:48 p.m. UTC | #2 
On Mon, Jan 25, 2021 at 5:56 PM Jakub Kicinski <kuba@kernel.org> wrote:

> > Fixes: 3b766cd83232 ("net/core: Add reading VF statistics through the PF netdevice")
> > Fixes: c5a9f6f0ab40 ("net/core: Add drop counters to VF statistics")
> > Signed-off-by: Edwin Peer <edwin.peer@broadcom.com>
>
> Could you include in the commit message the size breakdown of a single
> VF nest? With and without efficient unaligned 64b access?

Oh, apologies. I see missed this question. I can certainly do that.

Regards,
Edwin Peer
diff mbox series

Patch

diff --git a/net/core/rtnetlink.c b/net/core/rtnetlink.c
index 3d6ab194d0f5..466f920ac974 100644
--- a/net/core/rtnetlink.c
+++ b/net/core/rtnetlink.c
@@ -931,24 +931,27 @@  static inline int rtnl_vfinfo_size(const struct net_device *dev,
 			 nla_total_size(sizeof(struct ifla_vf_rate)) +
 			 nla_total_size(sizeof(struct ifla_vf_link_state)) +
 			 nla_total_size(sizeof(struct ifla_vf_rss_query_en)) +
-			 nla_total_size(0) + /* nest IFLA_VF_STATS */
-			 /* IFLA_VF_STATS_RX_PACKETS */
-			 nla_total_size_64bit(sizeof(__u64)) +
-			 /* IFLA_VF_STATS_TX_PACKETS */
-			 nla_total_size_64bit(sizeof(__u64)) +
-			 /* IFLA_VF_STATS_RX_BYTES */
-			 nla_total_size_64bit(sizeof(__u64)) +
-			 /* IFLA_VF_STATS_TX_BYTES */
-			 nla_total_size_64bit(sizeof(__u64)) +
-			 /* IFLA_VF_STATS_BROADCAST */
-			 nla_total_size_64bit(sizeof(__u64)) +
-			 /* IFLA_VF_STATS_MULTICAST */
-			 nla_total_size_64bit(sizeof(__u64)) +
-			 /* IFLA_VF_STATS_RX_DROPPED */
-			 nla_total_size_64bit(sizeof(__u64)) +
-			 /* IFLA_VF_STATS_TX_DROPPED */
-			 nla_total_size_64bit(sizeof(__u64)) +
 			 nla_total_size(sizeof(struct ifla_vf_trust)));
+		if (~ext_filter_mask & RTEXT_FILTER_SKIP_STATS) {
+			size += num_vfs *
+				(nla_total_size(0) + /* nest IFLA_VF_STATS */
+				 /* IFLA_VF_STATS_RX_PACKETS */
+				 nla_total_size_64bit(sizeof(__u64)) +
+				 /* IFLA_VF_STATS_TX_PACKETS */
+				 nla_total_size_64bit(sizeof(__u64)) +
+				 /* IFLA_VF_STATS_RX_BYTES */
+				 nla_total_size_64bit(sizeof(__u64)) +
+				 /* IFLA_VF_STATS_TX_BYTES */
+				 nla_total_size_64bit(sizeof(__u64)) +
+				 /* IFLA_VF_STATS_BROADCAST */
+				 nla_total_size_64bit(sizeof(__u64)) +
+				 /* IFLA_VF_STATS_MULTICAST */
+				 nla_total_size_64bit(sizeof(__u64)) +
+				 /* IFLA_VF_STATS_RX_DROPPED */
+				 nla_total_size_64bit(sizeof(__u64)) +
+				 /* IFLA_VF_STATS_TX_DROPPED */
+				 nla_total_size_64bit(sizeof(__u64)));
+		}
 		return size;
 	} else
 		return 0;
@@ -1223,7 +1226,8 @@  static noinline_for_stack int rtnl_fill_stats(struct sk_buff *skb,
 static noinline_for_stack int rtnl_fill_vfinfo(struct sk_buff *skb,
 					       struct net_device *dev,
 					       int vfs_num,
-					       struct nlattr *vfinfo)
+					       struct nlattr *vfinfo,
+					       u32 ext_filter_mask)
 {
 	struct ifla_vf_rss_query_en vf_rss_query_en;
 	struct nlattr *vf, *vfstats, *vfvlanlist;
@@ -1329,33 +1333,35 @@  static noinline_for_stack int rtnl_fill_vfinfo(struct sk_buff *skb,
 		goto nla_put_vf_failure;
 	}
 	nla_nest_end(skb, vfvlanlist);
-	memset(&vf_stats, 0, sizeof(vf_stats));
-	if (dev->netdev_ops->ndo_get_vf_stats)
-		dev->netdev_ops->ndo_get_vf_stats(dev, vfs_num,
-						&vf_stats);
-	vfstats = nla_nest_start_noflag(skb, IFLA_VF_STATS);
-	if (!vfstats)
-		goto nla_put_vf_failure;
-	if (nla_put_u64_64bit(skb, IFLA_VF_STATS_RX_PACKETS,
-			      vf_stats.rx_packets, IFLA_VF_STATS_PAD) ||
-	    nla_put_u64_64bit(skb, IFLA_VF_STATS_TX_PACKETS,
-			      vf_stats.tx_packets, IFLA_VF_STATS_PAD) ||
-	    nla_put_u64_64bit(skb, IFLA_VF_STATS_RX_BYTES,
-			      vf_stats.rx_bytes, IFLA_VF_STATS_PAD) ||
-	    nla_put_u64_64bit(skb, IFLA_VF_STATS_TX_BYTES,
-			      vf_stats.tx_bytes, IFLA_VF_STATS_PAD) ||
-	    nla_put_u64_64bit(skb, IFLA_VF_STATS_BROADCAST,
-			      vf_stats.broadcast, IFLA_VF_STATS_PAD) ||
-	    nla_put_u64_64bit(skb, IFLA_VF_STATS_MULTICAST,
-			      vf_stats.multicast, IFLA_VF_STATS_PAD) ||
-	    nla_put_u64_64bit(skb, IFLA_VF_STATS_RX_DROPPED,
-			      vf_stats.rx_dropped, IFLA_VF_STATS_PAD) ||
-	    nla_put_u64_64bit(skb, IFLA_VF_STATS_TX_DROPPED,
-			      vf_stats.tx_dropped, IFLA_VF_STATS_PAD)) {
-		nla_nest_cancel(skb, vfstats);
-		goto nla_put_vf_failure;
+	if (~ext_filter_mask & RTEXT_FILTER_SKIP_STATS) {
+		memset(&vf_stats, 0, sizeof(vf_stats));
+		if (dev->netdev_ops->ndo_get_vf_stats)
+			dev->netdev_ops->ndo_get_vf_stats(dev, vfs_num,
+							  &vf_stats);
+		vfstats = nla_nest_start_noflag(skb, IFLA_VF_STATS);
+		if (!vfstats)
+			goto nla_put_vf_failure;
+		if (nla_put_u64_64bit(skb, IFLA_VF_STATS_RX_PACKETS,
+				      vf_stats.rx_packets, IFLA_VF_STATS_PAD) ||
+		    nla_put_u64_64bit(skb, IFLA_VF_STATS_TX_PACKETS,
+				      vf_stats.tx_packets, IFLA_VF_STATS_PAD) ||
+		    nla_put_u64_64bit(skb, IFLA_VF_STATS_RX_BYTES,
+				      vf_stats.rx_bytes, IFLA_VF_STATS_PAD) ||
+		    nla_put_u64_64bit(skb, IFLA_VF_STATS_TX_BYTES,
+				      vf_stats.tx_bytes, IFLA_VF_STATS_PAD) ||
+		    nla_put_u64_64bit(skb, IFLA_VF_STATS_BROADCAST,
+				      vf_stats.broadcast, IFLA_VF_STATS_PAD) ||
+		    nla_put_u64_64bit(skb, IFLA_VF_STATS_MULTICAST,
+				      vf_stats.multicast, IFLA_VF_STATS_PAD) ||
+		    nla_put_u64_64bit(skb, IFLA_VF_STATS_RX_DROPPED,
+				      vf_stats.rx_dropped, IFLA_VF_STATS_PAD) ||
+		    nla_put_u64_64bit(skb, IFLA_VF_STATS_TX_DROPPED,
+				      vf_stats.tx_dropped, IFLA_VF_STATS_PAD)) {
+			nla_nest_cancel(skb, vfstats);
+			goto nla_put_vf_failure;
+		}
+		nla_nest_end(skb, vfstats);
 	}
-	nla_nest_end(skb, vfstats);
 	nla_nest_end(skb, vf);
 	return 0;
 
@@ -1388,7 +1394,7 @@  static noinline_for_stack int rtnl_fill_vf(struct sk_buff *skb,
 		return -EMSGSIZE;
 
 	for (i = 0; i < num_vfs; i++) {
-		if (rtnl_fill_vfinfo(skb, dev, i, vfinfo))
+		if (rtnl_fill_vfinfo(skb, dev, i, vfinfo, ext_filter_mask))
 			return -EMSGSIZE;
 	}