* [PATCH net] net: mask NETIF_F_IPV6_CSUM flag on irregular packet header size
@ 2025-08-05 12:16 Jakub Ramaseuski
2025-08-06 13:32 ` Willem de Bruijn
0 siblings, 1 reply; 5+ messages in thread
From: Jakub Ramaseuski @ 2025-08-05 12:16 UTC (permalink / raw)
To: netdev; +Cc: kuba, horms, pabeni, Jakub Ramaseuski, Tianhao Zhao,
Michal Schmidt
Throughput with GRE on IPv6 drops to 0 on NICs that use ice/bnxt_en
or any driver with NETIF_F_IP_CSUM and NETIF_F_IPV6_CSUM set and with
NETIF_F_HW_CSUM unset, see following dmesg output for more info.
bnxt_en: caps=(0x009201c01dd14833, 0x0000000e401d4869)
WARNING: CPU: 1 PID: 5273 at net/core/dev.c:3535 skb_warn_bad_offload+0x81/0x140
Modules linked in: ip6_gre gre ip6_tunnel tunnel6 mlx5_ib mlx5_fwctl macsec fwctl rfkill irdma ib_uverbs ib_core amd_atl intel_rapl_msr intel_rapl_common amd64_edac edac_mce_amd kvm_amd ipmi_ssif mlx5_core kvm mlxfw psample acpi_ipmi ice tls ast irqbypass i40e tg3 bnxt_en rapl wmi_bmof i2c_algo_bit pcspkr acpi_cpufreq pci_hyperv_intf ipmi_si i2c_piix4 gnss k10temp i2c_smbus libie ptdma ipmi_devintf ipmi_msghandler joydev sg loop fuse nfnetlink xfs sd_mod ahci libahci libata ghash_clmulni_intel ccp sp5100_tco wmi dm_mirror dm_region_hash dm_log dm_mod
CPU: 1 UID: 0 PID: 5273 Comm: iperf3 Kdump: loaded Not tainted 6.16.0-0.rc7.60.eln150.x86_64 #1 PREEMPT(lazy)
Hardware name: Abacus electric, s.r.o. - servis@abacus.cz Super Server/H12SSW-iN, BIOS 2.7 10/25/2023
RIP: 0010:skb_warn_bad_offload+0x81/0x140
Code: 8d 88 18 02 00 00 48 85 c0 48 c7 c0 28 43 41 9e 48 0f 44 c8 48 8d 93 b8 00 00 00 4c 89 c6 48 c7 c7 b3 92 a3 9e e8 cf 0e 4e ff <0f> 0b 48 83 c4 08 5b 5d e9 6d b5 2a ff 80 bb 20 01 00 00 00 74 1d
RSP: 0018:ffffcfed417aef70 EFLAGS: 00010246
RAX: 0000000000000000 RBX: ffff8ec94e59c000 RCX: 0000000000000000
RDX: ffff8ed84e66a500 RSI: 0000000000000001 RDI: ffff8ed84e65c200
RBP: ffff8ecb1716e0e8 R08: 0000000000000000 R09: 00000000ffff7fff
R10: ffffffff9f265700 R11: ffffcfed417aee08 R12: ffff8ec94e59c000
R13: ffffcfed417af023 R14: ffff8ec94e59c000 R15: ffffcfed417af023
FS: 00007f7ee69516c0(0000) GS:ffff8ed8ae504000(0000) knlGS:0000000000000000
CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
CR2: 00007f7ee6950f78 CR3: 000000011436c006 CR4: 0000000000f70ef0
PKRU: 55555554
Call Trace:
<TASK>
skb_checksum_help+0x12a/0x1f0
? netif_skb_features+0xc1/0x2e0
validate_xmit_skb+0x1a3/0x2d0
validate_xmit_skb_list+0x4f/0x80
sch_direct_xmit+0x1a2/0x380
__dev_xmit_skb+0x242/0x670
__dev_queue_xmit+0x3fc/0x7f0
? srso_alias_return_thunk+0x5/0xfbef5
? ip6_rt_copy_init+0xf0/0x290
? srso_alias_return_thunk+0x5/0xfbef5
? selinux_ip_postroute+0x1c5/0x420
? srso_alias_return_thunk+0x5/0xfbef5
ip6_finish_output2+0x25e/0x5d0
? srso_alias_return_thunk+0x5/0xfbef5
? nf_hook_slow+0x47/0xf0
? srso_alias_return_thunk+0x5/0xfbef5
ip6_finish_output+0x1fc/0x3f0
ip6_tnl_xmit+0x608/0xc00 [ip6_tunnel]
? srso_alias_return_thunk+0x5/0xfbef5
ip6gre_tunnel_xmit+0x1c0/0x390 [ip6_gre]
dev_hard_start_xmit+0x63/0x1c0
__dev_queue_xmit+0x6d0/0x7f0
? srso_alias_return_thunk+0x5/0xfbef5
? chacha_block_generic+0x72/0xd0
? srso_alias_return_thunk+0x5/0xfbef5
? selinux_ip_postroute+0x1c5/0x420
ip6_finish_output2+0x214/0x5d0
? srso_alias_return_thunk+0x5/0xfbef5
? nf_hook_slow+0x47/0xf0
ip6_finish_output+0x1fc/0x3f0
ip6_xmit+0x2ca/0x6f0
? __pfx_dst_output+0x10/0x10
? srso_alias_return_thunk+0x5/0xfbef5
? __sk_dst_check+0x41/0xc0
? srso_alias_return_thunk+0x5/0xfbef5
? inet6_csk_route_socket+0x12e/0x200
inet6_csk_xmit+0xeb/0x150
__tcp_transmit_skb+0x555/0xa80
tcp_write_xmit+0x32a/0xe90
? srso_alias_return_thunk+0x5/0xfbef5
? skb_do_copy_data_nocache+0xc9/0x150
tcp_sendmsg_locked+0x437/0x1110
? srso_alias_return_thunk+0x5/0xfbef5
tcp_sendmsg+0x2f/0x50
sock_write_iter+0x126/0x1a0
vfs_write+0x3c8/0x480
ksys_write+0xbf/0xf0
do_syscall_64+0x7c/0x970
? srso_alias_return_thunk+0x5/0xfbef5
? syscall_exit_work+0x143/0x1b0
? srso_alias_return_thunk+0x5/0xfbef5
? do_syscall_64+0xaf/0x970
? srso_alias_return_thunk+0x5/0xfbef5
? __rseq_handle_notify_resume+0x39/0x60
? srso_alias_return_thunk+0x5/0xfbef5
? exit_to_user_mode_loop+0xbf/0x120
? srso_alias_return_thunk+0x5/0xfbef5
? do_syscall_64+0xaf/0x970
? srso_alias_return_thunk+0x5/0xfbef5
? drain_stock+0x79/0xa0
? srso_alias_return_thunk+0x5/0xfbef5
? __mem_cgroup_threshold+0x18/0xf0
? srso_alias_return_thunk+0x5/0xfbef5
? srso_alias_return_thunk+0x5/0xfbef5
? memcg1_check_events+0x60/0x1b0
? srso_alias_return_thunk+0x5/0xfbef5
? memcg1_commit_charge+0x6f/0x90
? srso_alias_return_thunk+0x5/0xfbef5
? srso_alias_return_thunk+0x5/0xfbef5
? mod_memcg_lruvec_state+0x1a4/0x200
? srso_alias_return_thunk+0x5/0xfbef5
? __lruvec_stat_mod_folio+0x85/0xd0
? srso_alias_return_thunk+0x5/0xfbef5
? srso_alias_return_thunk+0x5/0xfbef5
? __folio_mod_stat+0x2d/0x90
? srso_alias_return_thunk+0x5/0xfbef5
? folio_add_new_anon_rmap+0x72/0x1b0
? srso_alias_return_thunk+0x5/0xfbef5
? do_anonymous_page+0x49c/0x710
? srso_alias_return_thunk+0x5/0xfbef5
? syscall_exit_work+0x143/0x1b0
? srso_alias_return_thunk+0x5/0xfbef5
? do_syscall_64+0xaf/0x970
? srso_alias_return_thunk+0x5/0xfbef5
? count_memcg_events+0x14d/0x1a0
? srso_alias_return_thunk+0x5/0xfbef5
? handle_mm_fault+0x247/0x360
? srso_alias_return_thunk+0x5/0xfbef5
? do_user_addr_fault+0x20f/0x6a0
? srso_alias_return_thunk+0x5/0xfbef5
? srso_alias_return_thunk+0x5/0xfbef5
? irqentry_exit_to_user_mode+0x2c/0x180
entry_SYSCALL_64_after_hwframe+0x76/0x7e
RIP: 0033:0x7f7ee713098f
Code: 89 54 24 18 48 89 74 24 10 89 7c 24 08 e8 39 7a f9 ff 48 8b 54 24 18 48 8b 74 24 10 41 89 c0 8b 7c 24 08 b8 01 00 00 00 0f 05 <48> 3d 00 f0 ff ff 77 31 44 89 c7 48 89 44 24 08 e8 8c 7a f9 ff 48
RSP: 002b:00007f7ee6950e00 EFLAGS: 00000246 ORIG_RAX: 0000000000000001
RAX: ffffffffffffffda RBX: 0000000000020000 RCX: 00007f7ee713098f
RDX: 0000000000020000 RSI: 00007f7ee6f11000 RDI: 0000000000000005
RBP: 00007f7ee6f11000 R08: 0000000000000002 R09: 00007f7ee69516c0
R10: 0000000000000008 R11: 0000000000000246 R12: 0000000000000005
R13: 0000000000020000 R14: 00007ffea6ee3400 R15: 00007ffea6ee3507
</TASK>
---[ end trace 0000000000000000 ]---
skb len=7018 headroom=182 headlen=138 tailroom=0
mac=(182,14) mac_len=0 net=(196,48) trans=244
shinfo(txflags=0 nr_frags=1 gso(size=1376 type=80 segs=5))
csum(0x100120 start=288 offset=16 ip_summed=3 complete_sw=0 valid=0 level=0)
hash(0x7ec90a00 sw=0 l4=1) proto=0x86dd pkttype=0 iif=0
priority=0x0 mark=0x0 alloc_cpu=1 vlan_all=0x0
encapsulation=1 inner(proto=0xdd86, mac=248, net=248, trans=288)
dev name=enp1s0f0np0 feat=0x009201c01dd14833
sk family=10 type=1 proto=6
skb linear: 00000000: e4 3d 1a 7d ec 30 e4 3d 1a 7e 5d 90 86 dd 60 0e
skb linear: 00000010: 00 0a 1b 34 3c 40 20 11 00 00 00 00 00 00 00 00
skb linear: 00000020: 00 00 00 00 00 12 20 11 00 00 00 00 00 00 00 00
skb linear: 00000030: 00 00 00 00 00 11 2f 00 04 01 04 01 01 00 00 00
skb linear: 00000040: 86 dd 60 0e 00 0a 1b 00 06 40 20 23 00 00 00 00
skb linear: 00000050: 00 00 00 00 00 00 00 00 00 12 20 23 00 00 00 00
skb linear: 00000060: 00 00 00 00 00 00 00 00 00 11 bf 96 14 51 13 f9
skb linear: 00000070: ae 27 a0 a8 2b e3 80 18 00 40 5b 6f 00 00 01 01
skb linear: 00000080: 08 0a 42 d4 50 d5 4b 70 f8 1a
skb frag: 00000000: 80 de bb bd 51 f8 32 e7 5d f5 65 65 a1 22 76 05
skb frag: 00000010: 02 f8 60 25 e6 37 0b b3 90 05 8e 7f f4 c2 5d 9c
skb frag: 00000020: e3 24 84 ac 0e 03 9d 14 ac 1e e2 18 4c 45 ef 5f
skb frag: 00000030: db 95 db ab 1f c9 7f 6d 19 70 1f 0c e7 6e fd 6e
skb frag: 00000040: f4 ff 73 1e 06 8d a8 06 53 ba bf 58 12 cb b9 59
skb frag: 00000050: f0 71 7e c3 69 0a f5 19 8b b3 eb b1 fa e5 9c 59
skb frag: 00000060: 40 bb 1d 11 88 f4 c1 cc 77 91 41 2c bb 7e 9d b7
skb frag: 00000070: ac 50 1a e3 d1 ce f7 f9 58 e4 d5 5c 62 f4 eb 39
skb frag: 00000080: 0d 13 2a 31 2c ec
------------[ cut here ]------------
Mask NETIF_F_IPV6_CSUM in gso_features_check if the IPv6 header contains
extension headers. This flag indicates that the network interface
is capable of computing the checksum only for plain IPv6 headers
without any extension headers.
The exception is a BIG TCP extension, which, as stated in 68e068cabd2c6c53:
"The feature is only enabled on devices that support BIG TCP TSO.
The header is only present for PF_PACKET taps like tcpdump,
and not transmitted by physical devices."
Fixes: 04c20a9356f283da ("net: skip offload for NETIF_F_IPV6_CSUM if ipv6 header contains extension")
Reported-by: Tianhao Zhao <tizhao@redhat.com>
Suggested-by: Michal Schmidt <mschmidt@redhat.com>
Signed-off-by: Jakub Ramaseuski <jramaseu@redhat.com>
---
---
net/core/dev.c | 4 ++++
1 file changed, 4 insertions(+)
diff --git a/net/core/dev.c b/net/core/dev.c
index b28ce68830b2b..118c433c2cb9b 100644
--- a/net/core/dev.c
+++ b/net/core/dev.c
@@ -3778,6 +3778,10 @@ static netdev_features_t gso_features_check(const struct sk_buff *skb,
if (!(iph->frag_off & htons(IP_DF)))
features &= ~NETIF_F_TSO_MANGLEID;
}
+ if (vlan_get_protocol(skb) == htons(ETH_P_IPV6) &&
+ skb_network_header_len(skb) != sizeof(struct ipv6hdr) &&
+ !ipv6_has_hopopt_jumbo(skb))
+ features &= ~NETIF_F_IPV6_CSUM;
return features;
}
--
2.50.1
^ permalink raw reply related [flat|nested] 5+ messages in thread
* Re: [PATCH net] net: mask NETIF_F_IPV6_CSUM flag on irregular packet header size
2025-08-05 12:16 [PATCH net] net: mask NETIF_F_IPV6_CSUM flag on irregular packet header size Jakub Ramaseuski
@ 2025-08-06 13:32 ` Willem de Bruijn
2025-08-06 13:46 ` Willem de Bruijn
0 siblings, 1 reply; 5+ messages in thread
From: Willem de Bruijn @ 2025-08-06 13:32 UTC (permalink / raw)
To: Jakub Ramaseuski, netdev
Cc: kuba, horms, pabeni, Jakub Ramaseuski, Tianhao Zhao,
Michal Schmidt
Jakub Ramaseuski wrote:
> Throughput with GRE on IPv6 drops to 0 on NICs that use ice/bnxt_en
> or any driver with NETIF_F_IP_CSUM and NETIF_F_IPV6_CSUM set and with
> NETIF_F_HW_CSUM unset, see following dmesg output for more info.
HW_CSUM is generally not advertised if IP_CSUM and IPV6_CSUM are.
HW_CSUM is a superset, and generally advisable. But some devices
only support specific protocols..
The pertinent comment here is that SKB_GSO_TCPV6 cannot be segmented
if NETIF_F_IPV6_CSUM is set. And the same for SKB_GSO_UDP_L4 if
ETH_P_IPV6.
> bnxt_en: caps=(0x009201c01dd14833, 0x0000000e401d4869)
> WARNING: CPU: 1 PID: 5273 at net/core/dev.c:3535 skb_warn_bad_offload+0x81/0x140
> Modules linked in: ip6_gre gre ip6_tunnel tunnel6 mlx5_ib mlx5_fwctl macsec fwctl rfkill irdma ib_uverbs ib_core amd_atl intel_rapl_msr intel_rapl_common amd64_edac edac_mce_amd kvm_amd ipmi_ssif mlx5_core kvm mlxfw psample acpi_ipmi ice tls ast irqbypass i40e tg3 bnxt_en rapl wmi_bmof i2c_algo_bit pcspkr acpi_cpufreq pci_hyperv_intf ipmi_si i2c_piix4 gnss k10temp i2c_smbus libie ptdma ipmi_devintf ipmi_msghandler joydev sg loop fuse nfnetlink xfs sd_mod ahci libahci libata ghash_clmulni_intel ccp sp5100_tco wmi dm_mirror dm_region_hash dm_log dm_mod
> CPU: 1 UID: 0 PID: 5273 Comm: iperf3 Kdump: loaded Not tainted 6.16.0-0.rc7.60.eln150.x86_64 #1 PREEMPT(lazy)
> Hardware name: Abacus electric, s.r.o. - servis@abacus.cz Super Server/H12SSW-iN, BIOS 2.7 10/25/2023
> RIP: 0010:skb_warn_bad_offload+0x81/0x140
> Code: 8d 88 18 02 00 00 48 85 c0 48 c7 c0 28 43 41 9e 48 0f 44 c8 48 8d 93 b8 00 00 00 4c 89 c6 48 c7 c7 b3 92 a3 9e e8 cf 0e 4e ff <0f> 0b 48 83 c4 08 5b 5d e9 6d b5 2a ff 80 bb 20 01 00 00 00 74 1d
> RSP: 0018:ffffcfed417aef70 EFLAGS: 00010246
> RAX: 0000000000000000 RBX: ffff8ec94e59c000 RCX: 0000000000000000
> RDX: ffff8ed84e66a500 RSI: 0000000000000001 RDI: ffff8ed84e65c200
> RBP: ffff8ecb1716e0e8 R08: 0000000000000000 R09: 00000000ffff7fff
> R10: ffffffff9f265700 R11: ffffcfed417aee08 R12: ffff8ec94e59c000
> R13: ffffcfed417af023 R14: ffff8ec94e59c000 R15: ffffcfed417af023
> FS: 00007f7ee69516c0(0000) GS:ffff8ed8ae504000(0000) knlGS:0000000000000000
> CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
> CR2: 00007f7ee6950f78 CR3: 000000011436c006 CR4: 0000000000f70ef0
> PKRU: 55555554
> Call Trace:
> <TASK>
> skb_checksum_help+0x12a/0x1f0
> ? netif_skb_features+0xc1/0x2e0
> validate_xmit_skb+0x1a3/0x2d0
> validate_xmit_skb_list+0x4f/0x80
> sch_direct_xmit+0x1a2/0x380
> __dev_xmit_skb+0x242/0x670
> __dev_queue_xmit+0x3fc/0x7f0
> ? srso_alias_return_thunk+0x5/0xfbef5
> ? ip6_rt_copy_init+0xf0/0x290
> ? srso_alias_return_thunk+0x5/0xfbef5
> ? selinux_ip_postroute+0x1c5/0x420
> ? srso_alias_return_thunk+0x5/0xfbef5
> ip6_finish_output2+0x25e/0x5d0
> ? srso_alias_return_thunk+0x5/0xfbef5
> ? nf_hook_slow+0x47/0xf0
> ? srso_alias_return_thunk+0x5/0xfbef5
> ip6_finish_output+0x1fc/0x3f0
> ip6_tnl_xmit+0x608/0xc00 [ip6_tunnel]
> ? srso_alias_return_thunk+0x5/0xfbef5
> ip6gre_tunnel_xmit+0x1c0/0x390 [ip6_gre]
> dev_hard_start_xmit+0x63/0x1c0
> __dev_queue_xmit+0x6d0/0x7f0
> ? srso_alias_return_thunk+0x5/0xfbef5
> ? chacha_block_generic+0x72/0xd0
> ? srso_alias_return_thunk+0x5/0xfbef5
> ? selinux_ip_postroute+0x1c5/0x420
> ip6_finish_output2+0x214/0x5d0
> ? srso_alias_return_thunk+0x5/0xfbef5
> ? nf_hook_slow+0x47/0xf0
> ip6_finish_output+0x1fc/0x3f0
> ip6_xmit+0x2ca/0x6f0
> ? __pfx_dst_output+0x10/0x10
> ? srso_alias_return_thunk+0x5/0xfbef5
> ? __sk_dst_check+0x41/0xc0
> ? srso_alias_return_thunk+0x5/0xfbef5
> ? inet6_csk_route_socket+0x12e/0x200
> inet6_csk_xmit+0xeb/0x150
> __tcp_transmit_skb+0x555/0xa80
> tcp_write_xmit+0x32a/0xe90
> ? srso_alias_return_thunk+0x5/0xfbef5
> ? skb_do_copy_data_nocache+0xc9/0x150
> tcp_sendmsg_locked+0x437/0x1110
> ? srso_alias_return_thunk+0x5/0xfbef5
> tcp_sendmsg+0x2f/0x50
Report can be truncated here.
> sock_write_iter+0x126/0x1a0
> vfs_write+0x3c8/0x480
> ksys_write+0xbf/0xf0
> do_syscall_64+0x7c/0x970
> ? srso_alias_return_thunk+0x5/0xfbef5
> ? syscall_exit_work+0x143/0x1b0
> ? srso_alias_return_thunk+0x5/0xfbef5
> ? do_syscall_64+0xaf/0x970
> ? srso_alias_return_thunk+0x5/0xfbef5
> ? __rseq_handle_notify_resume+0x39/0x60
> ? srso_alias_return_thunk+0x5/0xfbef5
> ? exit_to_user_mode_loop+0xbf/0x120
> ? srso_alias_return_thunk+0x5/0xfbef5
> ? do_syscall_64+0xaf/0x970
> ? srso_alias_return_thunk+0x5/0xfbef5
> ? drain_stock+0x79/0xa0
> ? srso_alias_return_thunk+0x5/0xfbef5
> ? __mem_cgroup_threshold+0x18/0xf0
> ? srso_alias_return_thunk+0x5/0xfbef5
> ? srso_alias_return_thunk+0x5/0xfbef5
> ? memcg1_check_events+0x60/0x1b0
> ? srso_alias_return_thunk+0x5/0xfbef5
> ? memcg1_commit_charge+0x6f/0x90
> ? srso_alias_return_thunk+0x5/0xfbef5
> ? srso_alias_return_thunk+0x5/0xfbef5
> ? mod_memcg_lruvec_state+0x1a4/0x200
> ? srso_alias_return_thunk+0x5/0xfbef5
> ? __lruvec_stat_mod_folio+0x85/0xd0
> ? srso_alias_return_thunk+0x5/0xfbef5
> ? srso_alias_return_thunk+0x5/0xfbef5
> ? __folio_mod_stat+0x2d/0x90
> ? srso_alias_return_thunk+0x5/0xfbef5
> ? folio_add_new_anon_rmap+0x72/0x1b0
> ? srso_alias_return_thunk+0x5/0xfbef5
> ? do_anonymous_page+0x49c/0x710
> ? srso_alias_return_thunk+0x5/0xfbef5
> ? syscall_exit_work+0x143/0x1b0
> ? srso_alias_return_thunk+0x5/0xfbef5
> ? do_syscall_64+0xaf/0x970
> ? srso_alias_return_thunk+0x5/0xfbef5
> ? count_memcg_events+0x14d/0x1a0
> ? srso_alias_return_thunk+0x5/0xfbef5
> ? handle_mm_fault+0x247/0x360
> ? srso_alias_return_thunk+0x5/0xfbef5
> ? do_user_addr_fault+0x20f/0x6a0
> ? srso_alias_return_thunk+0x5/0xfbef5
> ? srso_alias_return_thunk+0x5/0xfbef5
> ? irqentry_exit_to_user_mode+0x2c/0x180
> entry_SYSCALL_64_after_hwframe+0x76/0x7e
> RIP: 0033:0x7f7ee713098f
> Code: 89 54 24 18 48 89 74 24 10 89 7c 24 08 e8 39 7a f9 ff 48 8b 54 24 18 48 8b 74 24 10 41 89 c0 8b 7c 24 08 b8 01 00 00 00 0f 05 <48> 3d 00 f0 ff ff 77 31 44 89 c7 48 89 44 24 08 e8 8c 7a f9 ff 48
> RSP: 002b:00007f7ee6950e00 EFLAGS: 00000246 ORIG_RAX: 0000000000000001
> RAX: ffffffffffffffda RBX: 0000000000020000 RCX: 00007f7ee713098f
> RDX: 0000000000020000 RSI: 00007f7ee6f11000 RDI: 0000000000000005
> RBP: 00007f7ee6f11000 R08: 0000000000000002 R09: 00007f7ee69516c0
> R10: 0000000000000008 R11: 0000000000000246 R12: 0000000000000005
> R13: 0000000000020000 R14: 00007ffea6ee3400 R15: 00007ffea6ee3507
> </TASK>
> ---[ end trace 0000000000000000 ]---
> skb len=7018 headroom=182 headlen=138 tailroom=0
> mac=(182,14) mac_len=0 net=(196,48) trans=244
> shinfo(txflags=0 nr_frags=1 gso(size=1376 type=80 segs=5))
> csum(0x100120 start=288 offset=16 ip_summed=3 complete_sw=0 valid=0 level=0)
> hash(0x7ec90a00 sw=0 l4=1) proto=0x86dd pkttype=0 iif=0
> priority=0x0 mark=0x0 alloc_cpu=1 vlan_all=0x0
> encapsulation=1 inner(proto=0xdd86, mac=248, net=248, trans=288)
> dev name=enp1s0f0np0 feat=0x009201c01dd14833
> sk family=10 type=1 proto=6
In this case we can trim the payload. It can often help debug these
issues, but is not needed to understand the root cause here.
> skb linear: 00000000: e4 3d 1a 7d ec 30 e4 3d 1a 7e 5d 90 86 dd 60 0e
> skb linear: 00000010: 00 0a 1b 34 3c 40 20 11 00 00 00 00 00 00 00 00
> skb linear: 00000020: 00 00 00 00 00 12 20 11 00 00 00 00 00 00 00 00
> skb linear: 00000030: 00 00 00 00 00 11 2f 00 04 01 04 01 01 00 00 00
> skb linear: 00000040: 86 dd 60 0e 00 0a 1b 00 06 40 20 23 00 00 00 00
> skb linear: 00000050: 00 00 00 00 00 00 00 00 00 12 20 23 00 00 00 00
> skb linear: 00000060: 00 00 00 00 00 00 00 00 00 11 bf 96 14 51 13 f9
> skb linear: 00000070: ae 27 a0 a8 2b e3 80 18 00 40 5b 6f 00 00 01 01
> skb linear: 00000080: 08 0a 42 d4 50 d5 4b 70 f8 1a
> skb frag: 00000000: 80 de bb bd 51 f8 32 e7 5d f5 65 65 a1 22 76 05
> skb frag: 00000010: 02 f8 60 25 e6 37 0b b3 90 05 8e 7f f4 c2 5d 9c
> skb frag: 00000020: e3 24 84 ac 0e 03 9d 14 ac 1e e2 18 4c 45 ef 5f
> skb frag: 00000030: db 95 db ab 1f c9 7f 6d 19 70 1f 0c e7 6e fd 6e
> skb frag: 00000040: f4 ff 73 1e 06 8d a8 06 53 ba bf 58 12 cb b9 59
> skb frag: 00000050: f0 71 7e c3 69 0a f5 19 8b b3 eb b1 fa e5 9c 59
> skb frag: 00000060: 40 bb 1d 11 88 f4 c1 cc 77 91 41 2c bb 7e 9d b7
> skb frag: 00000070: ac 50 1a e3 d1 ce f7 f9 58 e4 d5 5c 62 f4 eb 39
> skb frag: 00000080: 0d 13 2a 31 2c ec
> ------------[ cut here ]------------
>
> Mask NETIF_F_IPV6_CSUM in gso_features_check if the IPv6 header contains
> extension headers. This flag indicates that the network interface
> is capable of computing the checksum only for plain IPv6 headers
> without any extension headers.
>
> The exception is a BIG TCP extension, which, as stated in 68e068cabd2c6c53:
> "The feature is only enabled on devices that support BIG TCP TSO.
> The header is only present for PF_PACKET taps like tcpdump,
> and not transmitted by physical devices."
>
> Fixes: 04c20a9356f283da ("net: skip offload for NETIF_F_IPV6_CSUM if ipv6 header contains extension")
That fix is still a valid fix to fall back onto software checksum
offload for non GSO packets.
This fix addresses the same for GSO packets.
That fix is not wrong, and still needed (this patch also does not modify or remove its code).
> Reported-by: Tianhao Zhao <tizhao@redhat.com>
> Suggested-by: Michal Schmidt <mschmidt@redhat.com>
> Signed-off-by: Jakub Ramaseuski <jramaseu@redhat.com>
> ---
> ---
> net/core/dev.c | 4 ++++
> 1 file changed, 4 insertions(+)
>
> diff --git a/net/core/dev.c b/net/core/dev.c
> index b28ce68830b2b..118c433c2cb9b 100644
> --- a/net/core/dev.c
> +++ b/net/core/dev.c
> @@ -3778,6 +3778,10 @@ static netdev_features_t gso_features_check(const struct sk_buff *skb,
> if (!(iph->frag_off & htons(IP_DF)))
> features &= ~NETIF_F_TSO_MANGLEID;
> }
> + if (vlan_get_protocol(skb) == htons(ETH_P_IPV6) &&
> + skb_network_header_len(skb) != sizeof(struct ipv6hdr) &&
> + !ipv6_has_hopopt_jumbo(skb))
> + features &= ~NETIF_F_IPV6_CSUM;
Exit the branch as soon as possible. Avoid the Ethernet header lookup
if possible.
And unfortunately transport header cannot be guaranted with
virtio_net_hdr_to_skb.
/* NETIF_F_IPV6_CSUM does not support IPv6 extension headers,
* so neither does TSO that depends on it.
*/
if (features & NETIF_F_IPV6_CSUM &&
(skb_shinfo(skb)->gso_type & SKB_GSO_TCPV6 ||
(skb_shinfo(skb)->gso_type & SKB_GSO_UDP_L4 &&
vlan_get_protocol(skb) == htons(ETH_P_IPV6))) &&
skb_transport_header_was_set(skb) &&
skb_network_header_len(skb) != sizeof(struct ipv6hdr) &&
!ipv6_has_hopopt_jumbo(skb))
features &= ~(NETIF_F_IPV6_CSUM | NETIF_F_TSO6 | NETIF_F_GSO_UDP_L4);
> return features;
> }
> --
> 2.50.1
>
^ permalink raw reply [flat|nested] 5+ messages in thread
* Re: [PATCH net] net: mask NETIF_F_IPV6_CSUM flag on irregular packet header size
2025-08-06 13:32 ` Willem de Bruijn
@ 2025-08-06 13:46 ` Willem de Bruijn
2025-08-29 13:18 ` Jakub Ramaseuski
0 siblings, 1 reply; 5+ messages in thread
From: Willem de Bruijn @ 2025-08-06 13:46 UTC (permalink / raw)
To: Willem de Bruijn, Jakub Ramaseuski, netdev
Cc: kuba, horms, pabeni, Jakub Ramaseuski, Tianhao Zhao,
Michal Schmidt
Willem de Bruijn wrote:
> Jakub Ramaseuski wrote:
> > Throughput with GRE on IPv6 drops to 0 on NICs that use ice/bnxt_en
> > or any driver with NETIF_F_IP_CSUM and NETIF_F_IPV6_CSUM set and with
> > NETIF_F_HW_CSUM unset, see following dmesg output for more info.
>
> HW_CSUM is generally not advertised if IP_CSUM and IPV6_CSUM are.
> HW_CSUM is a superset, and generally advisable. But some devices
> only support specific protocols..
>
> The pertinent comment here is that SKB_GSO_TCPV6 cannot be segmented
> if NETIF_F_IPV6_CSUM is set
and the IPv6 packet has extension headers, I mean to add.
V> And the same for SKB_GSO_UDP_L4 if ETH_P_IPV6.
>
> > Mask NETIF_F_IPV6_CSUM in gso_features_check if the IPv6 header contains
> > extension headers. This flag indicates that the network interface
> > is capable of computing the checksum only for plain IPv6 headers
> > without any extension headers.
> >
> > The exception is a BIG TCP extension, which, as stated in 68e068cabd2c6c53:
> > "The feature is only enabled on devices that support BIG TCP TSO.
> > The header is only present for PF_PACKET taps like tcpdump,
> > and not transmitted by physical devices."
> >
> > Fixes: 04c20a9356f283da ("net: skip offload for NETIF_F_IPV6_CSUM if ipv6 header contains extension")
>
> That fix is still a valid fix to fall back onto software checksum
> offload for non GSO packets.
>
> This fix addresses the same for GSO packets.
>
> That fix is not wrong, and still needed (this patch also does not modify or remove its code).
The question is what the right tag.
Your code also shows a tunnel, but not the one implicated on the above patch:
Fixes: aa3463d65e7b ("fou: Add encap ops for IPv6 tunnels")
But conceivably this is not tunnel related at all, if any TSO6 traffic can
cause IPv6 extension headers to be present.
Minor: when respinning please put this brief explanation first, and
the log output (stack trace, skb_dump), second.
> > Reported-by: Tianhao Zhao <tizhao@redhat.com>
> > Suggested-by: Michal Schmidt <mschmidt@redhat.com>
> > Signed-off-by: Jakub Ramaseuski <jramaseu@redhat.com>
> > ---
> > ---
> > net/core/dev.c | 4 ++++
> > 1 file changed, 4 insertions(+)
> >
> > diff --git a/net/core/dev.c b/net/core/dev.c
> > index b28ce68830b2b..118c433c2cb9b 100644
> > --- a/net/core/dev.c
> > +++ b/net/core/dev.c
> > @@ -3778,6 +3778,10 @@ static netdev_features_t gso_features_check(const struct sk_buff *skb,
> > if (!(iph->frag_off & htons(IP_DF)))
> > features &= ~NETIF_F_TSO_MANGLEID;
> > }
> > + if (vlan_get_protocol(skb) == htons(ETH_P_IPV6) &&
> > + skb_network_header_len(skb) != sizeof(struct ipv6hdr) &&
> > + !ipv6_has_hopopt_jumbo(skb))
> > + features &= ~NETIF_F_IPV6_CSUM;
>
> Exit the branch as soon as possible. Avoid the Ethernet header lookup
> if possible.
>
> And unfortunately transport header cannot be guaranted with
> virtio_net_hdr_to_skb.
>
> /* NETIF_F_IPV6_CSUM does not support IPv6 extension headers,
> * so neither does TSO that depends on it.
> */
> if (features & NETIF_F_IPV6_CSUM &&
> (skb_shinfo(skb)->gso_type & SKB_GSO_TCPV6 ||
> (skb_shinfo(skb)->gso_type & SKB_GSO_UDP_L4 &&
> vlan_get_protocol(skb) == htons(ETH_P_IPV6))) &&
> skb_transport_header_was_set(skb) &&
> skb_network_header_len(skb) != sizeof(struct ipv6hdr) &&
> !ipv6_has_hopopt_jumbo(skb))
> features &= ~(NETIF_F_IPV6_CSUM | NETIF_F_TSO6 | NETIF_F_GSO_UDP_L4);
>
>
> > return features;
> > }
> > --
> > 2.50.1
> >
>
>
^ permalink raw reply [flat|nested] 5+ messages in thread
* Re: [PATCH net] net: mask NETIF_F_IPV6_CSUM flag on irregular packet header size
2025-08-06 13:46 ` Willem de Bruijn
@ 2025-08-29 13:18 ` Jakub Ramaseuski
2025-09-01 15:15 ` Willem de Bruijn
0 siblings, 1 reply; 5+ messages in thread
From: Jakub Ramaseuski @ 2025-08-29 13:18 UTC (permalink / raw)
To: willemdebruijn.kernel; +Cc: horms, jramaseu, netdev, tizhao
For whom it may concern,
same problem currently appears when GREoIPv6 encapsulates IPv4 traffic,
i.e. iperf3 -c IPV4_GRE_TUNNEL_ADDRESS, even with current 6.17.0-rc3.
Willem de Bruijn wrote:
> if (features & NETIF_F_IPV6_CSUM &&
> (skb_shinfo(skb)->gso_type & SKB_GSO_TCPV6 ||
> (skb_shinfo(skb)->gso_type & SKB_GSO_UDP_L4 &&
> vlan_get_protocol(skb) == htons(ETH_P_IPV6))) &&
> skb_transport_header_was_set(skb) &&
> skb_network_header_len(skb) != sizeof(struct ipv6hdr) &&
> !ipv6_has_hopopt_jumbo(skb))
> features &= ~(NETIF_F_IPV6_CSUM | NETIF_F_TSO6 | NETIF_F_GSO_UDP_L4);
This part has been accepted as
864e3396976e ("net: gso: Forbid IPv6 TSO with extensions on devices with only IPV6_CSUM")
I am currently looking for suggestions on how to loosen the suggested
solution, I added check for SKB_GSO_TCPV4, which worked fine (though with
SW checksum calculation), but I believe there could be more elegant solution.
I should also ask on next steps which should support IPv6 checksum offload
with devices capable of it, but not capable of NETIF_F_HW_CSUM,
and throughput seen prior to 04c20a9356f283da, whether it is possible.
Thank you in advance!
Here is the kernel log, note the gso_type in shinfo has flags set
for SKB_GSO_TCPV4 and SKB_GSO_GRE, but also the skb content:
gre: GRE over IPv4 demultiplexer driver
ip6_gre: GRE over IPv6 tunneling driver
ip6_tunnel: gre1 xmit: Local address not yet configured!
skb len=7098 headroom=202 headlen=118 tailroom=0
mac=(202,14) mac_len=0 net=(216,48) trans=264
shinfo(txflags=0 nr_frags=1 gso(size=1396 type=65 segs=5))
csum(0x100120 start=288 offset=16 ip_summed=3 complete_sw=0 valid=0 level=0)
hash(0x4dab788f sw=0 l4=1) proto=0x86dd pkttype=0 iif=0
priority=0x0 mark=0x0 alloc_cpu=20 vlan_all=0x0
encapsulation=1 inner(proto=0x0008, mac=268, net=268, trans=288)
dev name=enp65s0f0np0 feat=0x001081e21fd54bb3
sk family=2 type=1 proto=6
skb linear: 00000000: b4 96 91 db db e8 b4 96 91 db dc b0 86 dd 60 0d
skb linear: 00000010: 8f 78 1b 84 3c 40 20 11 00 00 00 00 00 00 00 00
skb linear: 00000020: 00 00 00 00 00 11 20 11 00 00 00 00 00 00 00 00
skb linear: 00000030: 00 00 00 00 00 12 2f 00 04 01 04 01 01 00 00 00
skb linear: 00000040: 08 00 45 00 1b 78 34 84 40 00 40 06 15 94 c0 a8
skb linear: 00000050: 2a 0b c0 a8 2a 0c 88 e2 14 51 0d 2e f4 85 98 ca
skb linear: 00000060: 81 5e 80 18 00 40 f0 d2 00 00 01 01 08 0a f8 ea
skb linear: 00000070: b5 98 43 9a 51 5d
...
ice: caps=(0x001081e21fd54bb3, 0x0000000e401d4869)
WARNING: CPU: 20 PID: 5341 at net/core/dev.c:3532 skb_warn_bad_offload+0x81/0x140
...
Call Trace:
<TASK>
skb_checksum_help+0x12a/0x1f0
? mod_memcg_lruvec_state+0xc5/0x1f0
validate_xmit_skb+0x1a3/0x2d0
validate_xmit_skb_list+0x4f/0x80
sch_direct_xmit+0x1a2/0x380
__dev_xmit_skb+0x242/0x670
__dev_queue_xmit+0x3fc/0x7f0
? srso_alias_return_thunk+0x5/0xfbef5
? ip6_rt_copy_init+0xed/0x2a0
? srso_alias_return_thunk+0x5/0xfbef5
? selinux_ip_postroute+0x1c5/0x420
? srso_alias_return_thunk+0x5/0xfbef5
ip6_finish_output2+0x25e/0x5d0
? srso_alias_return_thunk+0x5/0xfbef5
? nf_hook_slow+0x47/0xf0
ip6_finish_output+0x1fc/0x410
ip6_tnl_xmit+0x60a/0xc10 [ip6_tunnel]
? srso_alias_return_thunk+0x5/0xfbef5
ip6gre_xmit_ipv4+0xa8/0x120 [ip6_gre]
ip6gre_tunnel_xmit+0x2d5/0x390 [ip6_gre]
? srso_alias_return_thunk+0x5/0xfbef5
? chacha_block_generic+0x72/0xd0
? srso_alias_return_thunk+0x5/0xfbef5
dev_hard_start_xmit+0x63/0x1c0
__dev_queue_xmit+0x6d0/0x7f0
? srso_alias_return_thunk+0x5/0xfbef5
? selinux_ip_postroute+0x1c5/0x420
ip_finish_output2+0x1a9/0x520
? __ip_finish_output+0xba/0x180
ip_output+0x63/0x110
? __pfx_ip_finish_output+0x10/0x10
__ip_queue_xmit+0x369/0x4f0
__tcp_transmit_skb+0x8cc/0xa80
tcp_write_xmit+0x32a/0xea0
? srso_alias_return_thunk+0x5/0xfbef5
? skb_do_copy_data_nocache+0xc9/0x150
tcp_sendmsg_locked+0x434/0x10e0
? srso_alias_return_thunk+0x5/0xfbef5
tcp_sendmsg+0x2f/0x50
...
And aformentioned iperf3:
ip a output, iperf3 server side (truncated), with enp65s0f0np0 using ice driver:
8: enp65s0f0np0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc mq state UP group default qlen 1000
link/ether b4:96:91:db:db:e8 brd ff:ff:ff:ff:ff:ff
altname enxb49691dbdbe8
inet 192.168.44.12/24 scope global enp65s0f0np0
valid_lft forever preferred_lft forever
inet6 2011::12/64 scope global
valid_lft forever preferred_lft forever
19: gre1@enp65s0f0np0: <POINTOPOINT,NOARP,UP,LOWER_UP> mtu 1448 qdisc noqueue state UNKNOWN group default qlen 1000
link/gre6 2011::12 peer 2011::11 permaddr 429b:2cf8:de35::
inet 192.168.42.12/24 scope global gre1
valid_lft forever preferred_lft forever
inet6 2023::12/64 scope global
valid_lft forever preferred_lft forever
inet6 fe80::12/64 scope link
valid_lft forever preferred_lft forever
iperf3 output (server side)
-----------------------------------------------------------
Server listening on 5201 (test #3)
-----------------------------------------------------------
Accepted connection from 2023::11, port 44632
[ 5] local 2023::12 port 5201 connected to 2023::11 port 44642
[ ID] Interval Transfer Bitrate
[ 5] 0.00-1.00 sec 822 MBytes 6.88 Gbits/sec
[ 5] 1.00-2.00 sec 864 MBytes 7.25 Gbits/sec
[ 5] 2.00-3.00 sec 850 MBytes 7.13 Gbits/sec
[ 5] 3.00-4.00 sec 860 MBytes 7.22 Gbits/sec
[ 5] 4.00-5.00 sec 868 MBytes 7.28 Gbits/sec
[ 5] 5.00-6.00 sec 866 MBytes 7.26 Gbits/sec
[ 5] 6.00-7.00 sec 868 MBytes 7.28 Gbits/sec
[ 5] 7.00-8.00 sec 854 MBytes 7.17 Gbits/sec
[ 5] 8.00-9.00 sec 858 MBytes 7.20 Gbits/sec
[ 5] 9.00-10.00 sec 871 MBytes 7.30 Gbits/sec
[ 5] 10.00-10.00 sec 384 KBytes 9.17 Gbits/sec
- - - - - - - - - - - - - - - - - - - - - - - - -
[ ID] Interval Transfer Bitrate
[ 5] 0.00-10.00 sec 8.38 GBytes 7.20 Gbits/sec receiver
-----------------------------------------------------------
Server listening on 5201 (test #4)
-----------------------------------------------------------
Accepted connection from 192.168.42.11, port 44614
[ 5] local 192.168.42.12 port 5201 connected to 192.168.42.11 port 44616
[ ID] Interval Transfer Bitrate
[ 5] 0.00-1.00 sec 0.00 Bytes 0.00 bits/sec
[ 5] 1.00-2.00 sec 0.00 Bytes 0.00 bits/sec
[ 5] 2.00-3.00 sec 0.00 Bytes 0.00 bits/sec
[ 5] 3.00-4.00 sec 0.00 Bytes 0.00 bits/sec
[ 5] 4.00-5.00 sec 0.00 Bytes 0.00 bits/sec
[ 5] 5.00-6.00 sec 0.00 Bytes 0.00 bits/sec
[ 5] 6.00-7.00 sec 0.00 Bytes 0.00 bits/sec
[ 5] 7.00-8.00 sec 0.00 Bytes 0.00 bits/sec
[ 5] 8.00-9.00 sec 0.00 Bytes 0.00 bits/sec
[ 5] 9.00-10.00 sec 91.3 KBytes 748 Kbits/sec
[ 5] 10.00-10.53 sec 0.00 Bytes 0.00 bits/sec
- - - - - - - - - - - - - - - - - - - - - - - - -
[ ID] Interval Transfer Bitrate
[ 5] 0.00-10.53 sec 91.3 KBytes 71.1 Kbits/sec receiver
^ permalink raw reply [flat|nested] 5+ messages in thread
* Re: [PATCH net] net: mask NETIF_F_IPV6_CSUM flag on irregular packet header size
2025-08-29 13:18 ` Jakub Ramaseuski
@ 2025-09-01 15:15 ` Willem de Bruijn
0 siblings, 0 replies; 5+ messages in thread
From: Willem de Bruijn @ 2025-09-01 15:15 UTC (permalink / raw)
To: Jakub Ramaseuski, willemdebruijn.kernel; +Cc: horms, jramaseu, netdev, tizhao
Jakub Ramaseuski wrote:
> For whom it may concern,
> same problem currently appears when GREoIPv6 encapsulates IPv4 traffic,
> i.e. iperf3 -c IPV4_GRE_TUNNEL_ADDRESS, even with current 6.17.0-rc3.
>
> Willem de Bruijn wrote:
> > if (features & NETIF_F_IPV6_CSUM &&
> > (skb_shinfo(skb)->gso_type & SKB_GSO_TCPV6 ||
> > (skb_shinfo(skb)->gso_type & SKB_GSO_UDP_L4 &&
> > vlan_get_protocol(skb) == htons(ETH_P_IPV6))) &&
> > skb_transport_header_was_set(skb) &&
> > skb_network_header_len(skb) != sizeof(struct ipv6hdr) &&
> > !ipv6_has_hopopt_jumbo(skb))
> > features &= ~(NETIF_F_IPV6_CSUM | NETIF_F_TSO6 | NETIF_F_GSO_UDP_L4);
> This part has been accepted as
> 864e3396976e ("net: gso: Forbid IPv6 TSO with extensions on devices with only IPV6_CSUM")
>
> I am currently looking for suggestions on how to loosen the suggested
> solution, I added check for SKB_GSO_TCPV4, which worked fine (though with
> SW checksum calculation), but I believe there could be more elegant solution.
Per the skb_dump (thanks), gso_type is SKB_GSO_GRE | SKB_GSO_TCPV4.
The gso_type does not capture explicitly that the outer network type
is IPv6.
But nor should it care? NETIF_F_IPV6_CSUM affects the (inner)
transport layer checksum, which only depends on the inner network
layer header for input to the pseudo-header.
> I should also ask on next steps which should support IPv6 checksum offload
> with devices capable of it, but not capable of NETIF_F_HW_CSUM,
> and throughput seen prior to 04c20a9356f283da, whether it is possible.
> Thank you in advance!
Do you mean that we lack a way for devices to communicate that they
only support IPv4/IPv6 checksums, but do support this IPv6 checksums
with extension headers? Something like NETIF_F_IPV6_EXTHDR_CSUM?
Or perhaps these can just advertise NETIF_F_HW_CSUM and in their
ndo_features_check drop the capability for all packets not IPv4/IPv6.
> Here is the kernel log, note the gso_type in shinfo has flags set
> for SKB_GSO_TCPV4 and SKB_GSO_GRE, but also the skb content:
>
> gre: GRE over IPv4 demultiplexer driver
> ip6_gre: GRE over IPv6 tunneling driver
> ip6_tunnel: gre1 xmit: Local address not yet configured!
> skb len=7098 headroom=202 headlen=118 tailroom=0
> mac=(202,14) mac_len=0 net=(216,48) trans=264
> shinfo(txflags=0 nr_frags=1 gso(size=1396 type=65 segs=5))
> csum(0x100120 start=288 offset=16 ip_summed=3 complete_sw=0 valid=0 level=0)
> hash(0x4dab788f sw=0 l4=1) proto=0x86dd pkttype=0 iif=0
> priority=0x0 mark=0x0 alloc_cpu=20 vlan_all=0x0
> encapsulation=1 inner(proto=0x0008, mac=268, net=268, trans=288)
> dev name=enp65s0f0np0 feat=0x001081e21fd54bb3
> sk family=2 type=1 proto=6
> skb linear: 00000000: b4 96 91 db db e8 b4 96 91 db dc b0 86 dd 60 0d
> skb linear: 00000010: 8f 78 1b 84 3c 40 20 11 00 00 00 00 00 00 00 00
> skb linear: 00000020: 00 00 00 00 00 11 20 11 00 00 00 00 00 00 00 00
> skb linear: 00000030: 00 00 00 00 00 12 2f 00 04 01 04 01 01 00 00 00
> skb linear: 00000040: 08 00 45 00 1b 78 34 84 40 00 40 06 15 94 c0 a8
> skb linear: 00000050: 2a 0b c0 a8 2a 0c 88 e2 14 51 0d 2e f4 85 98 ca
> skb linear: 00000060: 81 5e 80 18 00 40 f0 d2 00 00 01 01 08 0a f8 ea
> skb linear: 00000070: b5 98 43 9a 51 5d
> ...
> ice: caps=(0x001081e21fd54bb3, 0x0000000e401d4869)
> WARNING: CPU: 20 PID: 5341 at net/core/dev.c:3532 skb_warn_bad_offload+0x81/0x140
> ...
> Call Trace:
> <TASK>
> skb_checksum_help+0x12a/0x1f0
> ? mod_memcg_lruvec_state+0xc5/0x1f0
> validate_xmit_skb+0x1a3/0x2d0
> validate_xmit_skb_list+0x4f/0x80
> sch_direct_xmit+0x1a2/0x380
> __dev_xmit_skb+0x242/0x670
> __dev_queue_xmit+0x3fc/0x7f0
> ? srso_alias_return_thunk+0x5/0xfbef5
> ? ip6_rt_copy_init+0xed/0x2a0
> ? srso_alias_return_thunk+0x5/0xfbef5
> ? selinux_ip_postroute+0x1c5/0x420
> ? srso_alias_return_thunk+0x5/0xfbef5
> ip6_finish_output2+0x25e/0x5d0
> ? srso_alias_return_thunk+0x5/0xfbef5
> ? nf_hook_slow+0x47/0xf0
> ip6_finish_output+0x1fc/0x410
> ip6_tnl_xmit+0x60a/0xc10 [ip6_tunnel]
> ? srso_alias_return_thunk+0x5/0xfbef5
> ip6gre_xmit_ipv4+0xa8/0x120 [ip6_gre]
> ip6gre_tunnel_xmit+0x2d5/0x390 [ip6_gre]
> ? srso_alias_return_thunk+0x5/0xfbef5
> ? chacha_block_generic+0x72/0xd0
> ? srso_alias_return_thunk+0x5/0xfbef5
> dev_hard_start_xmit+0x63/0x1c0
> __dev_queue_xmit+0x6d0/0x7f0
> ? srso_alias_return_thunk+0x5/0xfbef5
> ? selinux_ip_postroute+0x1c5/0x420
> ip_finish_output2+0x1a9/0x520
> ? __ip_finish_output+0xba/0x180
> ip_output+0x63/0x110
> ? __pfx_ip_finish_output+0x10/0x10
> __ip_queue_xmit+0x369/0x4f0
> __tcp_transmit_skb+0x8cc/0xa80
> tcp_write_xmit+0x32a/0xea0
> ? srso_alias_return_thunk+0x5/0xfbef5
> ? skb_do_copy_data_nocache+0xc9/0x150
> tcp_sendmsg_locked+0x434/0x10e0
> ? srso_alias_return_thunk+0x5/0xfbef5
> tcp_sendmsg+0x2f/0x50
> ...
>
> And aformentioned iperf3:
> ip a output, iperf3 server side (truncated), with enp65s0f0np0 using ice driver:
> 8: enp65s0f0np0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc mq state UP group default qlen 1000
> link/ether b4:96:91:db:db:e8 brd ff:ff:ff:ff:ff:ff
> altname enxb49691dbdbe8
> inet 192.168.44.12/24 scope global enp65s0f0np0
> valid_lft forever preferred_lft forever
> inet6 2011::12/64 scope global
> valid_lft forever preferred_lft forever
> 19: gre1@enp65s0f0np0: <POINTOPOINT,NOARP,UP,LOWER_UP> mtu 1448 qdisc noqueue state UNKNOWN group default qlen 1000
> link/gre6 2011::12 peer 2011::11 permaddr 429b:2cf8:de35::
> inet 192.168.42.12/24 scope global gre1
> valid_lft forever preferred_lft forever
> inet6 2023::12/64 scope global
> valid_lft forever preferred_lft forever
> inet6 fe80::12/64 scope link
> valid_lft forever preferred_lft forever
>
> iperf3 output (server side)
> -----------------------------------------------------------
> Server listening on 5201 (test #3)
> -----------------------------------------------------------
> Accepted connection from 2023::11, port 44632
> [ 5] local 2023::12 port 5201 connected to 2023::11 port 44642
> [ ID] Interval Transfer Bitrate
> [ 5] 0.00-1.00 sec 822 MBytes 6.88 Gbits/sec
> [ 5] 1.00-2.00 sec 864 MBytes 7.25 Gbits/sec
> [ 5] 2.00-3.00 sec 850 MBytes 7.13 Gbits/sec
> [ 5] 3.00-4.00 sec 860 MBytes 7.22 Gbits/sec
> [ 5] 4.00-5.00 sec 868 MBytes 7.28 Gbits/sec
> [ 5] 5.00-6.00 sec 866 MBytes 7.26 Gbits/sec
> [ 5] 6.00-7.00 sec 868 MBytes 7.28 Gbits/sec
> [ 5] 7.00-8.00 sec 854 MBytes 7.17 Gbits/sec
> [ 5] 8.00-9.00 sec 858 MBytes 7.20 Gbits/sec
> [ 5] 9.00-10.00 sec 871 MBytes 7.30 Gbits/sec
> [ 5] 10.00-10.00 sec 384 KBytes 9.17 Gbits/sec
> - - - - - - - - - - - - - - - - - - - - - - - - -
> [ ID] Interval Transfer Bitrate
> [ 5] 0.00-10.00 sec 8.38 GBytes 7.20 Gbits/sec receiver
> -----------------------------------------------------------
> Server listening on 5201 (test #4)
> -----------------------------------------------------------
> Accepted connection from 192.168.42.11, port 44614
> [ 5] local 192.168.42.12 port 5201 connected to 192.168.42.11 port 44616
> [ ID] Interval Transfer Bitrate
> [ 5] 0.00-1.00 sec 0.00 Bytes 0.00 bits/sec
> [ 5] 1.00-2.00 sec 0.00 Bytes 0.00 bits/sec
> [ 5] 2.00-3.00 sec 0.00 Bytes 0.00 bits/sec
> [ 5] 3.00-4.00 sec 0.00 Bytes 0.00 bits/sec
> [ 5] 4.00-5.00 sec 0.00 Bytes 0.00 bits/sec
> [ 5] 5.00-6.00 sec 0.00 Bytes 0.00 bits/sec
> [ 5] 6.00-7.00 sec 0.00 Bytes 0.00 bits/sec
> [ 5] 7.00-8.00 sec 0.00 Bytes 0.00 bits/sec
> [ 5] 8.00-9.00 sec 0.00 Bytes 0.00 bits/sec
> [ 5] 9.00-10.00 sec 91.3 KBytes 748 Kbits/sec
> [ 5] 10.00-10.53 sec 0.00 Bytes 0.00 bits/sec
> - - - - - - - - - - - - - - - - - - - - - - - - -
> [ ID] Interval Transfer Bitrate
> [ 5] 0.00-10.53 sec 91.3 KBytes 71.1 Kbits/sec receiver
>
^ permalink raw reply [flat|nested] 5+ messages in thread
end of thread, other threads:[~2025-09-01 15:15 UTC | newest]
Thread overview: 5+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2025-08-05 12:16 [PATCH net] net: mask NETIF_F_IPV6_CSUM flag on irregular packet header size Jakub Ramaseuski
2025-08-06 13:32 ` Willem de Bruijn
2025-08-06 13:46 ` Willem de Bruijn
2025-08-29 13:18 ` Jakub Ramaseuski
2025-09-01 15:15 ` Willem de Bruijn
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).