: Netfilter fixes
Subsystems (3)
cover : Netfilter fixes
Hi Dave,
following are a few netfilter fixes for 2.6.25, fixing a BUG() when mangling non-linear skbs in {ip,ip6,nfnetlink}_queue, a compilation error in busybox, IPv6 address range matching in xt_iprange, missing unidef'ing of xt_policy.h. Additionally there is a patch to remove an unnecessary member from the xt_hashlimit matchinfo, which we need to do know before it hits a release.
Please apply, thanks.
include/linux/netfilter.h | 2 +- include/linux/netfilter/Kbuild | 2 +- include/linux/netfilter/xt_hashlimit.h | 1 - net/bridge/netfilter/ebt_dnat.c | 2 +- net/bridge/netfilter/ebt_redirect.c | 2 +- net/bridge/netfilter/ebt_snat.c | 2 +- net/ipv4/netfilter/arpt_mangle.c | 2 +- net/ipv4/netfilter/ip_queue.c | 12 +++++++----- net/ipv6/netfilter/ip6_queue.c | 10 ++++++---- net/netfilter/nfnetlink_queue.c | 10 ++++++---- net/netfilter/xt_hashlimit.c | 3 --- net/netfilter/xt_iprange.c | 2 +- net/netfilter/xt_u32.c | 11 +++++------ 13 files changed, 31 insertions(+), 30 deletions(-)
Jan Engelhardt (2):
[NETFILTER]: xt_hashlimit: remove unneeded struct member [NETFILTER]: xt_iprange: fix subtraction-based comparison
Joonwoo Park (1):
[NETFILTER]: Fix incorrect use of skb_make_writable
Patrick McHardy (2):
[NETFILTER]: {ip,ip6,nfnetlink}_queue: fix SKB_LINEAR_ASSERT when mangling packet data [NETFILTER]: Use __u32 in struct nf_inet_addr
Pavel Emelyanov (1):
[NETFILTER]: xt_u32: drop the actually unused variable from u32_match_it
Robert P. J. Day (1):
[NETFILTER]: Make sure xt_policy.h is unifdef'ed.
PATCH 1/7 : nfnetlink_queue: fix SKB_LINEAR_ASSERT when mangling packet data
[NETFILTER]: {ip,ip6,nfnetlink}_queue: fix SKB_LINEAR_ASSERT when mangling packet data
As reported by Tomas Simonaitis <[email protected]>, inserting new data in skbs queued over {ip,ip6,nfnetlink}_queue triggers a SKB_LINEAR_ASSERT in skb_put().
Going back through the git history, it seems this bug is present since at least 2.6.12-rc2, probably even since the removal of skb_linearize() for netfilter.
Linearize non-linear skbs through skb_copy_expand() when enlarging them. Tested by Thomas, fixes bugzilla #9933.
---
commit 2292fd9623ce0391217dfe8cec57f70fd90da538
tree ba852b5da64fac4aaec85ecf9622cca31e3ffb0d
parent 73b9987a76c8e0b5326afcc3632f79be425676bc
author Patrick McHardy <[email protected]> Tue, 19 Feb 2008 17:15:26 +0100 committer Patrick McHardy <[email protected]> Tue, 19 Feb 2008 17:15:26 +0100
net/ipv4/netfilter/ip_queue.c | 12 +++++++----- net/ipv6/netfilter/ip6_queue.c | 10 ++++++---- net/netfilter/nfnetlink_queue.c | 10 ++++++---- 3 files changed, 19 insertions(+), 13 deletions(-)
net/ipv4/netfilter/ip_queue.c +7 -5
@@ -283,8 +283,8 @@ static int ipq_mangle_ipv4(ipq_verdict_msg_t *v, struct nf_queue_entry *e) {int diff; -int err;struct iphdr *user_iph = (struct iphdr *)v->payload; +struct sk_buff *nskb;if (v->data_len <sizeof (*user_iph))return 0 ;
@@ -296,14 +296,16 @@ ipq_mangle_ipv4(ipq_verdict_msg_t *v, struct nf_queue_entry *e)if (v->data_len >0xFFFF )return -EINVAL ;if (diff >skb_tailroom (e->skb)) { - err =pskb_expand_head (e->skb,0 , + nskb =skb_copy_expand (e->skb,0 , diff -skb_tailroom (e->skb),GFP_ATOMIC ); -if (err) { +if (!nskb) {printk (KERN_WARNING "ip_queue: error " -"in mangle, dropping packet: %d\n" , -err); -return err; +"in mangle, dropping packet\n" ); +return -ENOMEM ; } +kfree_skb (e->skb); + e->skb = nskb; }skb_put (e->skb, diff); }
net/ipv6/netfilter/ip6_queue.c +6 -4
@@ -285,8 +285,8 @@ static int ipq_mangle_ipv6(ipq_verdict_msg_t *v, struct nf_queue_entry *e) {int diff; -int err;struct ipv6hdr *user_iph = (struct ipv6hdr *)v->payload; +struct sk_buff *nskb;if (v->data_len <sizeof (*user_iph))return 0 ;
@@ -298,14 +298,16 @@ ipq_mangle_ipv6(ipq_verdict_msg_t *v, struct nf_queue_entry *e)if (v->data_len >0xFFFF )return -EINVAL ;if (diff >skb_tailroom (e->skb)) { - err =pskb_expand_head (e->skb,0 , + nskb =skb_copy_expand (e->skb,0 , diff -skb_tailroom (e->skb),GFP_ATOMIC ); -if (err) { +if (!nskb) {printk (KERN_WARNING "ip6_queue: OOM " "in mangle, dropping packet\n" ); -return err; +return -ENOMEM ; } +kfree_skb (e->skb); + e->skb = nskb; }skb_put (e->skb, diff); }
net/netfilter/nfnetlink_queue.c +6 -4
@@ -443,8 +443,8 @@ err_out:static int nfqnl_mangle (void *data,int data_len,struct nf_queue_entry *e) { +struct sk_buff *nskb;int diff; -int err; diff = data_len - e->skb->len;if (diff <0 ) {
@@ -454,14 +454,16 @@ nfqnl_mangle(void *data, int data_len, struct nf_queue_entry *e)if (data_len >0xFFFF )return -EINVAL ;if (diff >skb_tailroom (e->skb)) { - err =pskb_expand_head (e->skb,0 , + nskb =skb_copy_expand (e->skb,0 , diff -skb_tailroom (e->skb),GFP_ATOMIC ); -if (err) { +if (!nskb) {printk (KERN_WARNING "nf_queue: OOM " "in mangle, dropping packet\n" ); -return err; +return -ENOMEM ; } +kfree_skb (e->skb); + e->skb = nskb; }skb_put (e->skb, diff); }
PATCH 2/7 : xt_u32: drop the actually unused variable from u32_match_it
In reply to Patrick McHardy[NETFILTER]: xt_u32: drop the actually unused variable from u32_match_it
The int ret variable is used only to trigger the BUG_ON() after the skb_copy_bits() call, so check the call failure directly and drop the variable.
---
commit 30c709c51fa0f400859890d534cfa02eb8d06cc7
tree 6df1f3fd873cae02ae066e2a76bafd1ef32e055a
parent 2292fd9623ce0391217dfe8cec57f70fd90da538
author Pavel Emelyanov <[email protected]> Tue, 19 Feb 2008 17:15:26 +0100 committer Patrick McHardy <[email protected]> Tue, 19 Feb 2008 17:15:26 +0100
net/netfilter/xt_u32.c | 11 +++++------ 1 files changed, 5 insertions(+), 6 deletions(-)
net/netfilter/xt_u32.c +5 -6
@@ -26,7 +26,6 @@ static bool u32_match_it(const struct xt_u32 *data,u_int32_t pos;u_int32_t val;u_int32_t at; -int ret;/* * Small example: "0 >> 28 == 4 && 8 & 0xFF0000 >> 16 = 6, 17"
@@ -40,8 +39,8 @@ static bool u32_match_it(const struct xt_u32 *data,if (skb->len <4 || pos > skb->len -4 )return false; - ret =skb_copy_bits (skb, pos, &n,sizeof (n)); -BUG_ON (ret <0 ); +if (skb_copy_bits (skb, pos, &n,sizeof (n)) <0 ) +BUG (); val =ntohl (n); nnums = ct->nnums;
@@ -67,9 +66,9 @@ static bool u32_match_it(const struct xt_u32 *data, pos > skb->len - at - 4)return false; - ret =skb_copy_bits (skb, at + pos, &n, -sizeof (n)); -BUG_ON (ret <0 ); +if (skb_copy_bits (skb, at + pos, &n, +sizeof (n)) <0 ) +BUG (); val =ntohl (n);break ; }
PATCH 3/7 : Fix incorrect use of skb_make_writable
In reply to Patrick McHardy[NETFILTER]: Fix incorrect use of skb_make_writable
http://bugzilla.kernel.org/show_bug.cgi?id=9920
The function skb_make_writable returns true or false.
---
commit 02413e566cd9a49bcf1f15f69cdc749780181412
tree 9ec6b55a428fbfdd24da2bc619ebf4c5dabb0354
parent 30c709c51fa0f400859890d534cfa02eb8d06cc7
author Joonwoo Park <[email protected]> Tue, 19 Feb 2008 17:15:27 +0100 committer Patrick McHardy <[email protected]> Tue, 19 Feb 2008 17:15:27 +0100
net/bridge/netfilter/ebt_dnat.c | 2 +- net/bridge/netfilter/ebt_redirect.c | 2 +- net/bridge/netfilter/ebt_snat.c | 2 +- net/ipv4/netfilter/arpt_mangle.c | 2 +- 4 files changed, 4 insertions(+), 4 deletions(-)
net/bridge/netfilter/ebt_dnat.c +1 -1
@@ -20,7 +20,7 @@ static int ebt_target_dnat(struct sk_buff *skb, unsigned int hooknr, {const struct ebt_nat_info *info = data; -if (skb_make_writable (skb,0 )) +if (!skb_make_writable (skb,0 ))return NF_DROP ;memcpy (eth_hdr (skb)->h_dest, info->mac,ETH_ALEN );
net/bridge/netfilter/ebt_redirect.c +1 -1
@@ -21,7 +21,7 @@ static int ebt_target_redirect(struct sk_buff *skb, unsigned int hooknr, {const struct ebt_redirect_info *info = data; -if (skb_make_writable (skb,0 )) +if (!skb_make_writable (skb,0 ))return NF_DROP ;if (hooknr !=NF_BR_BROUTING )
net/bridge/netfilter/ebt_snat.c +1 -1
@@ -22,7 +22,7 @@ static int ebt_target_snat(struct sk_buff *skb, unsigned int hooknr, {const struct ebt_nat_info *info = data; -if (skb_make_writable (skb,0 )) +if (!skb_make_writable (skb,0 ))return NF_DROP ;memcpy (eth_hdr (skb)->h_source, info->mac,ETH_ALEN );
net/ipv4/netfilter/arpt_mangle.c +1 -1
@@ -19,7 +19,7 @@ target(struct sk_buff *skb,unsigned char *arpptr;int pln, hln; -if (skb_make_writable (skb, skb->len)) +if (!skb_make_writable (skb, skb->len))return NF_DROP ; arp =arp_hdr (skb);
PATCH 4/7 : Make sure xt_policy.h is unifdef'ed.
In reply to Patrick McHardy[NETFILTER]: Make sure xt_policy.h is unifdef'ed.
Since the header file xt_policy.h tests __KERNEL__, it should be unifdef'ed before exporting to userspace.
---
commit 0776e52cdb04aad5b59c4e12dc16e7b58737c2d2
tree d8f3cb734bad2c4df073262ab9ac14989f28eb28
parent 02413e566cd9a49bcf1f15f69cdc749780181412
author Robert P. J. Day <[email protected]> Tue, 19 Feb 2008 17:15:28 +0100 committer Patrick McHardy <[email protected]> Tue, 19 Feb 2008 17:15:28 +0100
include/linux/netfilter/Kbuild | 2 +- 1 files changed, 1 insertions(+), 1 deletions(-)
include/linux/netfilter/Kbuild +1 -1
@@ -30,7 +30,6 @@ header-y += xt_mark.h header-y += xt_multiport.h header-y += xt_owner.h header-y += xt_pkttype.h -header-y += xt_policy.h header-y += xt_rateest.h header-y += xt_realm.h header-y += xt_sctp.h
@@ -47,3 +46,4 @@ unifdef-y += nfnetlink.h unifdef-y += nfnetlink_compat.h unifdef-y += x_tables.h unifdef-y += xt_physdev.h +unifdef-y += xt_policy.h
PATCH 5/7 : xt_hashlimit: remove unneeded struct member
In reply to Patrick McHardy[NETFILTER]: xt_hashlimit: remove unneeded struct member
By allocating ->hinfo, we already have the needed indirection to cope with the per-cpu xtables struct match_entry.
[Patrick: do this now before the revision 1 struct is used by userspace]
---
commit 4e2e7aba6471fcd945a3f8fdc40d5cf11e7039b3
tree fee4fdd8dfbf77ac747e89a316b591fab09bf884
parent 0776e52cdb04aad5b59c4e12dc16e7b58737c2d2
author Jan Engelhardt <[email protected]> Tue, 19 Feb 2008 17:15:28 +0100 committer Patrick McHardy <[email protected]> Tue, 19 Feb 2008 17:15:28 +0100
include/linux/netfilter/xt_hashlimit.h | 1 - net/netfilter/xt_hashlimit.c | 3 --- 2 files changed, 0 insertions(+), 4 deletions(-)
include/linux/netfilter/xt_hashlimit.h +0 -1
@@ -61,7 +61,6 @@ struct xt_hashlimit_mtinfo1 {/* Used internally by the kernel */ struct xt_hashlimit_htable *hinfo __attribute__((aligned(8))); - struct xt_hashlimit_mtinfo1 *master __attribute__((aligned(8))); };#endif /*_XT_HASHLIMIT_H*/
net/netfilter/xt_hashlimit.c +0 -3
@@ -774,9 +774,6 @@ hashlimit_mt_check(const char *tablename, const void *inf,return false; }mutex_unlock (&hlimit_mutex); - -/* Ugly hack: For SMP, we only want to use one set */ - info->master = info;return true; }
PATCH 6/7 : xt_iprange: fix subtraction-based comparison
In reply to Patrick McHardy[NETFILTER]: xt_iprange: fix subtraction-based comparison
The host address parts need to be converted to host-endian first before arithmetic makes any sense on them.
---
commit 5a090319a35aac1052102d97c431230d3140daf1
tree 72fb9bcb9ebf94fad277a3461288d23841f50662
parent 4e2e7aba6471fcd945a3f8fdc40d5cf11e7039b3
author Jan Engelhardt <[email protected]> Tue, 19 Feb 2008 17:15:29 +0100 committer Patrick McHardy <[email protected]> Tue, 19 Feb 2008 17:15:29 +0100
net/netfilter/xt_iprange.c | 2 +- 1 files changed, 1 insertions(+), 1 deletions(-)
net/netfilter/xt_iprange.c +1 -1
@@ -102,7 +102,7 @@ iprange_ipv6_sub(const struct in6_addr *a, const struct in6_addr *b)int r;for (i =0 ; i <4 ; ++i) { - r = (__force u32 )a->s6_addr32[i] - (__force u32 )b->s6_addr32[i]; + r =ntohl (a->s6_addr32[i]) -ntohl (b->s6_addr32[i]);if (r !=0 )return r; }
PATCH 7/7 : Use __u32 in struct nf_inet_addr
In reply to Patrick McHardy[NETFILTER]: Use __u32 in struct nf_inet_addr
As reported by David Woodhouse <[email protected]>, using u_int32_t in struct nf_inet_addr breaks the busybox build. Fix by using __u32.
---
commit 4f4e9be33da8cc1a297c4d7ef51c1fa1effcbace
tree a856969ae8edeebca7c8018e18e4304cd198e529
parent 5a090319a35aac1052102d97c431230d3140daf1
author Patrick McHardy <[email protected]> Tue, 19 Feb 2008 17:15:29 +0100 committer Patrick McHardy <[email protected]> Tue, 19 Feb 2008 17:15:29 +0100
include/linux/netfilter.h | 2 +- 1 files changed, 1 insertions(+), 1 deletions(-)
include/linux/netfilter.h +1 -1
@@ -51,7 +51,7 @@ enum nf_inet_hooks { };union nf_inet_addr { -u_int32_t all[4 ]; +__u32 all[4 ];__be32 ip;__be32 ip6[4 ];struct in_addr in;
Re: [NETFILTER 07/07]: Use __u32 in struct nf_inet_addr
Quoted 4 lines · Patrick McHardy
[NETFILTER]: Use
__u32in structnf_inet_addrAs reported by David Woodhouse <[email protected]>, using
u_int32_tin structnf_inet_addrbreaks the busybox build. Fix by using__u32.
... that's why I say we should use uint32_t (with just one underscore), because that is available in userspace thorugh <stdint.h>, while u_int32_t is not (at first glance²).
² db1/db.h and ALSA do seem to use it, though.
-
To unsubscribe from this list: send the line "unsubscribe netfilter-devel" in the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html
Re: [NETFILTER 07/07]: Use __u32 in struct nf_inet_addr
Quoted 9 lines · Jan Engelhardt
Quoted 4 lines · Patrick McHardy
[NETFILTER]: Use
__u32in structnf_inet_addrAs reported by David Woodhouse <[email protected]>, using
u_int32_tin structnf_inet_addrbreaks the busybox build. Fix by using__u32.... that's why I say we should use
uint32_t(with just one underscore), because that is available in userspace thorugh <stdint.h>, whileu_int32_tis not (at first glance²).
As David explained, kernel headers shouldn't assume that userspace has pulled in any of these typedefs and therefore should use the kernel-internally defined types.
-
To unsubscribe from this list: send the line "unsubscribe netfilter-devel" in the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html
David Miller
maintainerFrom: Patrick McHardy <[email protected]>
Date: Tue, 19 Feb 2008 17:25:44 +0100 (MET)
Quoted 8 lines
following are a few netfilter fixes for 2.6.25, fixing a
BUG()when mangling non-linear skbs in {ip,ip6,nfnetlink}_queue, a compilation error in busybox, IPv6 address range matching inxt_iprange, missing unidef'ing ofxt_policy.h. Additionally there is a patch to remove an unnecessary member from thext_hashlimitmatchinfo, which we need to do know before it hits a release.Please apply, thanks.
All applied, thanks Patrick.