Re: OOPS in nf_ct_unlink_expect_report using Polycom RealPresence Mobile

Re: OOPS in nf_ct_unlink_expect_report using Polycom RealPresence Mobile

[Pablo Neira Ayuso]

[-- Attachment #1: Type: text/plain, Size: 234 bytes --]

On Fri, Jan 31, 2014 at 05:04:02PM +0100, astx wrote:
> Dear Alexey,
> 
> seems to help. Thank you for your quick response. Kernel 3.10.28 is
> now stable using h323 / Polycom.

Thanks, if no objection, will pass this patch to David.

[-- Attachment #2: 0001-netfilter-nf_nat_h323-fix-crash-in-nf_ct_unlink_expe.patch --]
[-- Type: text/x-diff, Size: 2377 bytes --]

>From d98506139d6e192705422ffba13bc2ff476ac513 Mon Sep 17 00:00:00 2001
From: Alexey Dobriyan <adobriyan@gmail.com>
Date: Mon, 3 Feb 2014 13:07:24 +0100
Subject: [PATCH] netfilter: nf_nat_h323: fix crash in
 nf_ct_unlink_expect_report()

Similar bug fixed in SIP module in 3f509c6 ("netfilter: nf_nat_sip: fix
incorrect handling of EBUSY for RTCP expectation").

BUG: unable to handle kernel paging request at 00100104
IP: [<f8214f07>] nf_ct_unlink_expect_report+0x57/0xf0 [nf_conntrack]
...
Call Trace:
  [<c0244bd8>] ? del_timer+0x48/0x70
  [<f8215687>] nf_ct_remove_expectations+0x47/0x60 [nf_conntrack]
  [<f8211c99>] nf_ct_delete_from_lists+0x59/0x90 [nf_conntrack]
  [<f8212e5e>] death_by_timeout+0x14e/0x1c0 [nf_conntrack]
  [<f8212d10>] ? nf_conntrack_set_hashsize+0x190/0x190 [nf_conntrack]
  [<c024442d>] call_timer_fn+0x1d/0x80
  [<c024461e>] run_timer_softirq+0x18e/0x1a0
  [<f8212d10>] ? nf_conntrack_set_hashsize+0x190/0x190 [nf_conntrack]
  [<c023e6f3>] __do_softirq+0xa3/0x170
  [<c023e650>] ? __local_bh_enable+0x70/0x70
  <IRQ>
  [<c023e587>] ? irq_exit+0x67/0xa0
  [<c0202af6>] ? do_IRQ+0x46/0xb0
  [<c027ad05>] ? clockevents_notify+0x35/0x110
  [<c066ac6c>] ? common_interrupt+0x2c/0x40
  [<c056e3c1>] ? cpuidle_enter_state+0x41/0xf0
  [<c056e6fb>] ? cpuidle_idle_call+0x8b/0x100
  [<c02085f8>] ? arch_cpu_idle+0x8/0x30
  [<c027314b>] ? cpu_idle_loop+0x4b/0x140
  [<c0273258>] ? cpu_startup_entry+0x18/0x20
  [<c066056d>] ? rest_init+0x5d/0x70
  [<c0813ac8>] ? start_kernel+0x2ec/0x2f2
  [<c081364f>] ? repair_env_string+0x5b/0x5b
  [<c0813269>] ? i386_start_kernel+0x33/0x35

Signed-off-by: Alexey Dobriyan <adobriyan@gmail.com>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
---
 net/ipv4/netfilter/nf_nat_h323.c |    5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)

diff --git a/net/ipv4/netfilter/nf_nat_h323.c b/net/ipv4/netfilter/nf_nat_h323.c
index 9eea059d..574f7eb 100644
--- a/net/ipv4/netfilter/nf_nat_h323.c
+++ b/net/ipv4/netfilter/nf_nat_h323.c
@@ -229,7 +229,10 @@ static int nat_rtp_rtcp(struct sk_buff *skb, struct nf_conn *ct,
 			ret = nf_ct_expect_related(rtcp_exp);
 			if (ret == 0)
 				break;
-			else if (ret != -EBUSY) {
+			else if (ret == -EBUSY) {
+				nf_ct_unexpect_related(rtp_exp);
+				continue;
+			} else if (ret < 0) {
 				nf_ct_unexpect_related(rtp_exp);
 				nated_port = 0;
 				break;
-- 
1.7.10.4

Re: OOPS in nf_ct_unlink_expect_report using Polycom RealPresence Mobile

[astx]

Test results / tested kernel versions:

3.2.54
3.8.13
3.10.28

Above kernel versions without patch are dying with same error on  
trying to start h323 connections using "Polycom RealPresence Mobile".

I can confirm that with this patch all three kernel versions are  
pretty stable now again.

Thank you all for your fast and competent help.

Best Regards,

Toni


Zitat von Pablo Neira Ayuso <pablo@netfilter.org>:

> On Fri, Jan 31, 2014 at 05:04:02PM +0100, astx wrote:
>> Dear Alexey,
>>
>> seems to help. Thank you for your quick response. Kernel 3.10.28 is
>> now stable using h323 / Polycom.
>
> Thanks, if no objection, will pass this patch to David.

Re: OOPS in nf_ct_unlink_expect_report using Polycom RealPresence Mobile

[Alexey Dobriyan]

[-- Attachment #1: Type: text/plain, Size: 853 bytes --]

> Disabling nf_nat_h323 and nf_conntrack_h323 avoids crash -
> but video conferencing software is no more usable.
>
> BUG: unable to handle kernel paging request at 00100104
> IP: [<f8214f07>] nf_ct_unlink_expect_report

This must be the same bug fixed in SIP module:
commit 3f509c689a07a4aa989b426893d8491a7ffcc410
netfilter: nf_nat_sip: fix incorrect handling of EBUSY for RTCP expectation

Try attached patch (if this is mangled):

--- a/net/ipv4/netfilter/nf_nat_h323.c
+++ b/net/ipv4/netfilter/nf_nat_h323.c
@@ -229,7 +229,10 @@ static int nat_rtp_rtcp(struct sk_buff *skb,
struct nf_conn *ct,
  ret = nf_ct_expect_related(rtcp_exp);
  if (ret == 0)
  break;
- else if (ret != -EBUSY) {
+ else if (ret == -EBUSY) {
+ nf_ct_unexpect_related(rtp_exp);
+ continue;
+ } else if (ret < 0) {
  nf_ct_unexpect_related(rtp_exp);
  nated_port = 0;
  break;

[-- Attachment #2: h323-expect.diff --]
[-- Type: text/plain, Size: 446 bytes --]

--- a/net/ipv4/netfilter/nf_nat_h323.c
+++ b/net/ipv4/netfilter/nf_nat_h323.c
@@ -229,7 +229,10 @@ static int nat_rtp_rtcp(struct sk_buff *skb, struct nf_conn *ct,
 			ret = nf_ct_expect_related(rtcp_exp);
 			if (ret == 0)
 				break;
-			else if (ret != -EBUSY) {
+			else if (ret == -EBUSY) {
+				nf_ct_unexpect_related(rtp_exp);
+				continue;
+			} else if (ret < 0) {
 				nf_ct_unexpect_related(rtp_exp);
 				nated_port = 0;
 				break;