JackDoan 921ed4360a overlay/tio: validate WriteGSO geometry instead of silently dropping ...

The length checks were fishy on four counts: an empty hdr/transportHdr
with real payload returned nil (silent drop with a success signal); the
HdrLen/GSOSize/CsumStart uint16 conversions could wrap unchecked;
nothing verified transportHdr covers csum_start+csum_offset, so the
kernel's NEEDS_CSUM write could land in payload bytes; and there was no
total-size bound even though every length field involved is 16-bit.

Malformed geometry is now a real error, and a single 65535 total-length
guard makes all the u16 conversions exact.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

2026-07-27 16:21:10 -05:00

..

virtio

another VIRTIO_NET_HDR_GSO_ECN mistake

2026-07-27 16:20:10 -05:00

blockon_linux.go

unslop some comments

2026-07-24 16:39:03 -05:00

queueset_gso_linux.go

datapath: fix 12 correctness findings from tun/UDP offload review

2026-07-24 16:39:03 -05:00

queueset_poll_linux.go

datapath: fix 12 correctness findings from tun/UDP offload review

2026-07-24 16:39:03 -05:00

segment_bench_test.go

GSO/GRO offloads, with TCP+ECN and UDP support

2026-07-24 16:37:30 -05:00

segment_other.go

scoot stuff around for e2e

2026-07-24 16:37:30 -05:00

single.go

checkpt

2026-07-24 16:39:03 -05:00

tio_gso_linux.go

overlay/tio: validate WriteGSO geometry instead of silently dropping

2026-07-27 16:21:10 -05:00

tio_poll_linux.go

more fixes!

2026-07-24 16:39:03 -05:00

tio.go

more fixes!

2026-07-24 16:39:03 -05:00

tun_file_linux_test.go

more fixes!

2026-07-24 16:39:03 -05:00

tun_linux_offload_test.go

overlay/tio: validate WriteGSO geometry instead of silently dropping

2026-07-27 16:21:10 -05:00

tun_linux_offload.go

tio: accept VIRTIO_NET_HDR_GSO_ECN-qualified superpackets

2026-07-24 16:39:03 -05:00