Re: [PATCH net-next] eth: bnxt: report TX HW GSO queue stats
From: Jakub Kicinski
Date: Fri Nov 20 2026 - 18:49:02 EST
On Thu, 19 Nov 2026 20:48:03 -0800 Narcisa Vasile wrote:
> > [Severity: Medium]
> > Can this over-count wire packets for GSO skbs that were sent through the
> > TX push path?
> >
> > This block runs before the is_push early-out. It treats every completed
> > skb with skb_is_gso() && !is_sw_gso as segmented by the hardware.
> > However, the push path gate in bnxt_start_xmit() does not exclude GSO
> > skbs:
> >
> > if (free_size == bp->tx_ring_size && length <= bp->tx_push_thresh &&
> > skb_frags_readable(skb) && !lflags) {
> >
> > The push BDs never set TX_BD_FLAGS_LSO or tx_bd_mss, so the NIC sends
> > the skb as a single frame.
> >
> > Take an idle ring with tx_push_thresh at 164 and a GSO skb of 164 bytes
> > or less. Two examples are TCP with a peer-advertised MSS of 48 and 96
> > bytes of payload, or UDP_SEGMENT with gso_size=8 and 80 bytes of
> > payload. Either skb would be counted as 1 hw_gso_packet plus gso_segs
> > hw_gso_wire_packets, even though only one frame went on the wire.
> >
> > tso.py reads these counters, so could this give misleading results
> > there? Fixing the push path issue described below would also fix this.
>
> This couldn't give misleading results there, since no GSO skb in tso.py
> can be 164 bytes or less, so none can take the push path.
Fair, but UDP segmentation may sneak in there. I will check with Joe.
Either way, I agree that this is orthogonal.