| Message ID | 4f6a2b28-a137-2e19-bf62-5a8767d0d0ac@virtuozzo.com (mailing list archive) |
|---|---|
| State | Superseded |
| Delegated to: | Netdev Maintainers |
| Headers | show |
| Series | ipv6: allocate enough headroom in ip6_finish_output2() | expand |
| Context | Check | Description |
|---|---|---|
| netdev/cover_letter | success | Link |
| netdev/fixes_present | success | Link |
| netdev/patch_count | success | Link |
| netdev/tree_selection | success | Guessed tree name to be net-next |
| netdev/subject_prefix | warning | Target tree name not specified in the subject |
| netdev/cc_maintainers | success | CCed 5 of 5 maintainers |
| netdev/source_inline | fail | Was 0 now: 1 |
| netdev/verify_signedoff | success | Link |
| netdev/module_param | success | Was 0 now: 0 |
| netdev/build_32bit | success | Errors and warnings before: 0 this patch: 0 |
| 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, 48 lines checked |
| netdev/build_allmodconfig_warn | success | Errors and warnings before: 0 this patch: 0 |
| netdev/header_inline | success | Link |
Please do not use inline in foo.c files, let the compiler decde. Thank you.
Dear David,
I'm happy to hear you again.
On 7/9/21 8:58 PM, David Miller wrote:
> Please do not use inline in foo.c files, let the compiler decde.
Thank you for the hint, I did not know it, and will follow him next time.
This time I'm going to move this helper somewhere anyway:
either to net/core/skbuff.c as exported function where it will lost inline anyway,
or to include/linux/skbuff.h where inline is (it seems?) acceptable.
Could you please help me to find better name for this helper?
I would like to change its current name: 'skb_expand_head' looks very similar
to widely used 'pskb_expand_head' but have different semantic.
I afraid they can be accidentally misused in future.
Thank you,
Vasily Averin.
diff --git a/net/ipv6/ip6_output.c b/net/ipv6/ip6_output.c index ff4f9eb..6c5f85f 100644 --- a/net/ipv6/ip6_output.c +++ b/net/ipv6/ip6_output.c @@ -56,14 +56,48 @@ #include <net/lwtunnel.h> #include <net/ip_tunnels.h> +static inline struct sk_buff *skb_expand_head(struct sk_buff *skb, int delta) +{ + /* pskb_expand_head() might crash, if skb is shared */ + if (skb_shared(skb)) { + struct sk_buff *nskb = skb_clone(skb, GFP_ATOMIC); + + if (likely(nskb)) { + if (skb->sk) + skb_set_owner_w(skb, skb->sk); + consume_skb(skb); + } else { + kfree_skb(skb); + } + skb = nskb; + } + if (skb && + pskb_expand_head(skb, SKB_DATA_ALIGN(delta), 0, GFP_ATOMIC)) { + kfree_skb(skb); + skb = NULL; + } + return skb; +} + static int ip6_finish_output2(struct net *net, struct sock *sk, struct sk_buff *skb) { struct dst_entry *dst = skb_dst(skb); struct net_device *dev = dst->dev; + unsigned int hh_len = LL_RESERVED_SPACE(dev); + int delta = hh_len - skb_headroom(skb); const struct in6_addr *nexthop; struct neighbour *neigh; int ret; + /* Be paranoid, rather than too clever. */ + if (unlikely(delta > 0) && dev->header_ops) + skb = skb_expand_head(skb, delta); + + if (!skb) { + IP6_INC_STATS(net, ip6_dst_idev(dst), IPSTATS_MIB_OUTDISCARDS); + return -ENOMEM; + } + if (ipv6_addr_is_multicast(&ipv6_hdr(skb)->daddr)) { struct inet6_dev *idev = ip6_dst_idev(skb_dst(skb));
When TEE target mirrors traffic to another interface, sk_buff may not have enough headroom to be processed correctly. ip_finish_output2() detect this situation for ipv4 and allocates new skb with enogh headroom. However ipv6 lacks this logic in ip_finish_output2 and it leads to skb_under_panic: skbuff: skb_under_panic: text:ffffffffc0866ad4 len:96 put:24 head:ffff97be85e31800 data:ffff97be85e317f8 tail:0x58 end:0xc0 dev:gre0 ------------[ cut here ]------------ kernel BUG at net/core/skbuff.c:110! invalid opcode: 0000 [#1] SMP PTI CPU: 2 PID: 393 Comm: kworker/2:2 Tainted: G OE 5.13.0 #13 Hardware name: Virtuozzo KVM, BIOS 1.11.0-2.vz7.4 04/01/2014 Workqueue: ipv6_addrconf addrconf_dad_work RIP: 0010:skb_panic+0x48/0x4a Call Trace: skb_push.cold.111+0x10/0x10 ipgre_header+0x24/0xf0 [ip_gre] neigh_connected_output+0xae/0xf0 ip6_finish_output2+0x1a8/0x5a0 ip6_output+0x5c/0x110 nf_dup_ipv6+0x158/0x1000 [nf_dup_ipv6] tee_tg6+0x2e/0x40 [xt_TEE] ip6t_do_table+0x294/0x470 [ip6_tables] nf_hook_slow+0x44/0xc0 nf_hook.constprop.34+0x72/0xe0 ndisc_send_skb+0x20d/0x2e0 ndisc_send_ns+0xd1/0x210 addrconf_dad_work+0x3c8/0x540 process_one_work+0x1d1/0x370 worker_thread+0x30/0x390 kthread+0x116/0x130 ret_from_fork+0x22/0x30 This patch implement new helper that tries to expand headroom on current skb, if it is not possible (shared_skb) -- creates new one. v2 open questions: - currently helper name skb_expand_head is bad and should be changed to better one. Any suggestions? - proper location for new helper: in net/core/skbuff.c right below skb_realloc_headroom() ? - is it acceptable to free original skb inside helper ? Is it probably required to keep it in caller instead? Signed-off-by: Vasily Averin <vvs@virtuozzo.com> --- net/ipv6/ip6_output.c | 34 ++++++++++++++++++++++++++++++++++ 1 file changed, 34 insertions(+)