| Message ID | 1620804453-57566-1-git-send-email-dseok.yi@samsung.com (mailing list archive) |
|---|---|
| State | Accepted |
| Commit | fa7b83bf3b156c767f3e4a25bbf3817b08f3ff8e |
| Delegated to: | BPF |
| Headers | show |
| Series | [bpf-next,v3] bpf: check for BPF_F_ADJ_ROOM_FIXED_GSO when bpf_skb_change_proto | expand |
| 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 bpf-next |
| netdev/subject_prefix | success | Link |
| netdev/cc_maintainers | success | CCed 12 of 12 maintainers |
| 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: 37 this patch: 37 |
| 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, 72 lines checked |
| netdev/build_allmodconfig_warn | success | Errors and warnings before: 37 this patch: 37 |
| netdev/header_inline | success | Link |
On Wed, May 12, 2021 at 3:41 AM Dongseok Yi <dseok.yi@samsung.com> wrote: > > In the forwarding path GRO -> BPF 6 to 4 -> GSO for TCP traffic, the > coalesced packet payload can be > MSS, but < MSS + 20. > bpf_skb_proto_6_to_4 will upgrade the MSS and it can be > the payload > length. After then tcp_gso_segment checks for the payload length if it > is <= MSS. The condition is causing the packet to be dropped. > > tcp_gso_segment(): > [...] > mss = skb_shinfo(skb)->gso_size; > if (unlikely(skb->len <= mss)) > goto out; > [...] > > Allow to upgrade/downgrade MSS only when BPF_F_ADJ_ROOM_FIXED_GSO is > not set. > > Signed-off-by: Dongseok Yi <dseok.yi@samsung.com> Acked-by: Willem de Bruijn <willemb@google.com>
Hello: This patch was applied to bpf/bpf-next.git (refs/heads/master): On Wed, 12 May 2021 16:27:33 +0900 you wrote: > In the forwarding path GRO -> BPF 6 to 4 -> GSO for TCP traffic, the > coalesced packet payload can be > MSS, but < MSS + 20. > bpf_skb_proto_6_to_4 will upgrade the MSS and it can be > the payload > length. After then tcp_gso_segment checks for the payload length if it > is <= MSS. The condition is causing the packet to be dropped. > > tcp_gso_segment(): > [...] > mss = skb_shinfo(skb)->gso_size; > if (unlikely(skb->len <= mss)) > goto out; > [...] > > [...] Here is the summary with links: - [bpf-next,v3] bpf: check for BPF_F_ADJ_ROOM_FIXED_GSO when bpf_skb_change_proto https://git.kernel.org/bpf/bpf-next/c/fa7b83bf3b15 You are awesome, thank you! -- Deet-doot-dot, I am a bot. https://korg.docs.kernel.org/patchwork/pwbot.html
diff --git a/net/core/filter.c b/net/core/filter.c index cae56d0..582ac19 100644 --- a/net/core/filter.c +++ b/net/core/filter.c @@ -3235,7 +3235,7 @@ static int bpf_skb_net_hdr_pop(struct sk_buff *skb, u32 off, u32 len) return ret; } -static int bpf_skb_proto_4_to_6(struct sk_buff *skb) +static int bpf_skb_proto_4_to_6(struct sk_buff *skb, u64 flags) { const u32 len_diff = sizeof(struct ipv6hdr) - sizeof(struct iphdr); u32 off = skb_mac_header_len(skb); @@ -3264,7 +3264,9 @@ static int bpf_skb_proto_4_to_6(struct sk_buff *skb) } /* Due to IPv6 header, MSS needs to be downgraded. */ - skb_decrease_gso_size(shinfo, len_diff); + if (!(flags & BPF_F_ADJ_ROOM_FIXED_GSO)) + skb_decrease_gso_size(shinfo, len_diff); + /* Header must be checked, and gso_segs recomputed. */ shinfo->gso_type |= SKB_GSO_DODGY; shinfo->gso_segs = 0; @@ -3276,7 +3278,7 @@ static int bpf_skb_proto_4_to_6(struct sk_buff *skb) return 0; } -static int bpf_skb_proto_6_to_4(struct sk_buff *skb) +static int bpf_skb_proto_6_to_4(struct sk_buff *skb, u64 flags) { const u32 len_diff = sizeof(struct ipv6hdr) - sizeof(struct iphdr); u32 off = skb_mac_header_len(skb); @@ -3305,7 +3307,9 @@ static int bpf_skb_proto_6_to_4(struct sk_buff *skb) } /* Due to IPv4 header, MSS can be upgraded. */ - skb_increase_gso_size(shinfo, len_diff); + if (!(flags & BPF_F_ADJ_ROOM_FIXED_GSO)) + skb_increase_gso_size(shinfo, len_diff); + /* Header must be checked, and gso_segs recomputed. */ shinfo->gso_type |= SKB_GSO_DODGY; shinfo->gso_segs = 0; @@ -3317,17 +3321,17 @@ static int bpf_skb_proto_6_to_4(struct sk_buff *skb) return 0; } -static int bpf_skb_proto_xlat(struct sk_buff *skb, __be16 to_proto) +static int bpf_skb_proto_xlat(struct sk_buff *skb, __be16 to_proto, u64 flags) { __be16 from_proto = skb->protocol; if (from_proto == htons(ETH_P_IP) && to_proto == htons(ETH_P_IPV6)) - return bpf_skb_proto_4_to_6(skb); + return bpf_skb_proto_4_to_6(skb, flags); if (from_proto == htons(ETH_P_IPV6) && to_proto == htons(ETH_P_IP)) - return bpf_skb_proto_6_to_4(skb); + return bpf_skb_proto_6_to_4(skb, flags); return -ENOTSUPP; } @@ -3337,7 +3341,7 @@ BPF_CALL_3(bpf_skb_change_proto, struct sk_buff *, skb, __be16, proto, { int ret; - if (unlikely(flags)) + if (unlikely(flags & ~(BPF_F_ADJ_ROOM_FIXED_GSO))) return -EINVAL; /* General idea is that this helper does the basic groundwork @@ -3357,7 +3361,7 @@ BPF_CALL_3(bpf_skb_change_proto, struct sk_buff *, skb, __be16, proto, * that. For offloads, we mark packet as dodgy, so that headers * need to be verified first. */ - ret = bpf_skb_proto_xlat(skb, proto); + ret = bpf_skb_proto_xlat(skb, proto, flags); bpf_compute_data_pointers(skb); return ret; }
In the forwarding path GRO -> BPF 6 to 4 -> GSO for TCP traffic, the coalesced packet payload can be > MSS, but < MSS + 20. bpf_skb_proto_6_to_4 will upgrade the MSS and it can be > the payload length. After then tcp_gso_segment checks for the payload length if it is <= MSS. The condition is causing the packet to be dropped. tcp_gso_segment(): [...] mss = skb_shinfo(skb)->gso_size; if (unlikely(skb->len <= mss)) goto out; [...] Allow to upgrade/downgrade MSS only when BPF_F_ADJ_ROOM_FIXED_GSO is not set. Signed-off-by: Dongseok Yi <dseok.yi@samsung.com> --- net/core/filter.c | 22 +++++++++++++--------- 1 file changed, 13 insertions(+), 9 deletions(-) v2: per Willem de Bruijn request, checked the flag instead of a generic approach. v3: per Willem de Bruijn request, moved to bpf-next, supported for both 6_to_4 and 4_to_6.