On Mon, 21 Dec 2020 19:46:23 +0100 Eric Dumazet wrote:

So maybe the issue is coming from traffic coming from a VM through a
tun device or something,
and our handling of GSO_ROBUST / DODGY never cared about setting
SKB_GSO_TCPV4 or SKB_GSO_TCPV6 if not already given by user space ?

Or a plain bug somewhere, possibly overwriting  gso_type with 0 or garbage...

Hm, the Arch Linux forum thread linked from the kernel bugzilla
has the user complaining about NFS performance.