[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <ZGIv4WAAV088VdSx@boxer>
Date: Mon, 15 May 2023 15:13:05 +0200
From: Maciej Fijalkowski <maciej.fijalkowski@...el.com>
To: Piotr Raczynski <piotr.raczynski@...el.com>
CC: <intel-wired-lan@...ts.osuosl.org>, <netdev@...r.kernel.org>,
<bpf@...r.kernel.org>, <anthony.l.nguyen@...el.com>,
<magnus.karlsson@...el.com>
Subject: Re: [PATCH iwl-net] ice: recycle/free all of the fragments from
multi-buffer packet
On Fri, May 12, 2023 at 05:40:11PM +0200, Piotr Raczynski wrote:
> > @@ -1162,6 +1162,9 @@ int ice_clean_rx_irq(struct ice_rx_ring *rx_ring, int budget)
> > bool failure;
> > u32 first;
> >
> > + if (ntc != rx_ring->first_desc)
> > + cached_ntc = rx_ring->first_desc;
> > +
> Would it make sense to apply likely/unlikely here since we check it per
> packet?
This was done before running Rx processing loop so this was not per
packet, only a single check, but in the end we don't need this, see
Simon's reply.
>
> Piotr
> > /* Frame size depend on rx_ring setup when PAGE_SIZE=4K */
> > #if (PAGE_SIZE < 8192)
>
> >
Powered by blists - more mailing lists