netfilter-devel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Marcelo Ricardo Leitner <mleitner@redhat.com>
To: netfilter-devel@vger.kernel.org
Subject: [PATCH v3 2/3] Do error handling if __build_packet_message fails
Date: Wed, 29 Oct 2014 10:04:53 -0200	[thread overview]
Message-ID: <c764c773bc964e4630761f5af7a0587c19d8d8b2.1414584011.git.mleitner@redhat.com> (raw)
In-Reply-To: <7be00bf663b212d4984414346da216cf17d3443d.1414584011.git.mleitner@redhat.com>

Currently, we don't check if __build_packet_message fails or not and
that leaves it prone to sending incomplete messages to userspace.

This commit fixes it by canceling the new message if there was any issue
while building it and makes sure the skb is freed if it was the only
messsage in it.

Signed-off-by: Marcelo Ricardo Leitner <mleitner@redhat.com>
---
 net/netfilter/nfnetlink_log.c | 21 +++++++++++++++------
 1 file changed, 15 insertions(+), 6 deletions(-)

diff --git a/net/netfilter/nfnetlink_log.c b/net/netfilter/nfnetlink_log.c
index b1e3a05794169283ed50d1c0fb4f44d9e7753eeb..05184a3d860885c10c3e977fdc47fb86e3528afa 100644
--- a/net/netfilter/nfnetlink_log.c
+++ b/net/netfilter/nfnetlink_log.c
@@ -568,10 +568,8 @@ __build_packet_message(struct nfnl_log_net *log,
 		struct nlattr *nla;
 		int size = nla_attr_size(data_len);
 
-		if (skb_tailroom(inst->skb) < nla_total_size(data_len)) {
-			printk(KERN_WARNING "nfnetlink_log: no tailroom!\n");
-			return -1;
-		}
+		if (skb_tailroom(inst->skb) < nla_total_size(data_len))
+			goto nla_put_failure;
 
 		nla = (struct nlattr *)skb_put(inst->skb, nla_total_size(data_len));
 		nla->nla_type = NFULA_PAYLOAD;
@@ -586,6 +584,7 @@ __build_packet_message(struct nfnl_log_net *log,
 
 nla_put_failure:
 	PRINTR(KERN_ERR "nfnetlink_log: error creating log nlmsg\n");
+	nlmsg_cancel(skb, nlh);
 	return -1;
 }
 
@@ -708,8 +707,9 @@ nfulnl_log_packet(struct net *net,
 
 	inst->qlen++;
 
-	__build_packet_message(log, inst, skb, data_len, pf,
-				hooknum, in, out, prefix, plen);
+	if (__build_packet_message(log, inst, skb, data_len, pf,
+				hooknum, in, out, prefix, plen))
+		goto build_failure;
 
 	if (inst->qlen >= qthreshold)
 		__nfulnl_flush(inst);
@@ -726,6 +726,15 @@ unlock_and_release:
 	instance_put(inst);
 	return;
 
+build_failure:
+	/* If no other messages in it, we're good to free it. */
+	if (!inst->skb->len) {
+		kfree_skb(inst->skb);
+		inst->skb = NULL;
+	}
+
+	inst->qlen--;
+
 alloc_failure:
 	/* FIXME: statistics */
 	goto unlock_and_release;
-- 
1.9.3


  reply	other threads:[~2014-10-29 12:05 UTC|newest]

Thread overview: 4+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2014-10-29 12:04 [PATCH v3 1/3] netfilter: log: protect nf_log_register against double registering Marcelo Ricardo Leitner
2014-10-29 12:04 ` Marcelo Ricardo Leitner [this message]
2014-10-29 12:46   ` [PATCH v3 2/3] Do error handling if __build_packet_message fails Marcelo Ricardo Leitner
2014-10-29 12:04 ` [PATCH v3 3/3] Make use of pr_fmt where applicable Marcelo Ricardo Leitner

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=c764c773bc964e4630761f5af7a0587c19d8d8b2.1414584011.git.mleitner@redhat.com \
    --to=mleitner@redhat.com \
    --cc=netfilter-devel@vger.kernel.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).