-----Original Message-----
> From: Paolo Abeni <pab...@redhat.com>
> Sent: Tuesday, May 20, 2025 12:04 PM
> To: Chia-Yu Chang (Nokia) <chia-yu.ch...@nokia-bell-labs.com>;
> linux-doc@vger.kernel.org; cor...@lwn.net; ho...@kernel.org;
> dsah...@kernel.org; kun...@amazon.com; b...@vger.kernel.org;
> net...@vger.kernel.org; dave.t...@gmail.com; j...@mojatatu.com;
> k...@kernel.org; step...@networkplumber.org; xiyou.wangc...@gmail.com;
> j...@resnulli.us; da...@davemloft.net; eduma...@google.com;
> andrew+net...@lunn.ch; donald.hun...@gmail.com; a...@fiberby.net;
> liuhang...@gmail.com; sh...@kernel.org; linux-kselft...@vger.kernel.org;
> i...@kernel.org; ncardw...@google.com; Koen De Schepper (Nokia)
> <koen.de_schep...@nokia-bell-labs.com>; g.white <g.wh...@cablelabs.com>;
> ingemar.s.johans...@ericsson.com; mirja.kuehlew...@ericsson.com;
> chesh...@apple.com; rs.i...@gmx.at; jason_living...@comcast.com; vidhi_goel
> <vidhi_g...@apple.com>
> Subject: Re: [PATCH v7 net-next 14/15] tcp: accecn: try to fit AccECN option
> with SACK
>
>
> CAUTION: This is an external email. Please be very careful when clicking
> links or opening attachments. See the URL nok.it/ext for additional
> information.
>
>
>
> On 5/14/25 3:56 PM, chia-yu.ch...@nokia-bell-labs.com wrote:
> > From: Ilpo Järvinen <i...@kernel.org>
> >
> > As SACK blocks tend to eat all option space when there are many holes,
> > it is useful to compromise on sending many SACK blocks in every ACK
> > and try to fit AccECN option there by reduction the number of SACK
> > blocks. But never go below two SACK blocks because of AccECN option.
> >
> > As AccECN option is often not put to every ACK, the space hijack is
> > usually only temporary.
> >
> > Signed-off-by: Ilpo Järvinen <i...@kernel.org>
> > Signed-off-by: Chia-Yu Chang <chia-yu.ch...@nokia-bell-labs.com>
> > ---
> > net/ipv4/tcp_output.c | 15 ++++++++++++++-
> > 1 file changed, 14 insertions(+), 1 deletion(-)
> >
> > diff --git a/net/ipv4/tcp_output.c b/net/ipv4/tcp_output.c index
> > b630923c4cef..d9d3cc8dbb5b 100644
> > --- a/net/ipv4/tcp_output.c
> > +++ b/net/ipv4/tcp_output.c
> > @@ -982,8 +982,21 @@ static int tcp_options_fit_accecn(struct
> > tcp_out_options *opts, int required,
> > opts->num_accecn_fields--;
> > size -= TCPOLEN_ACCECN_PERFIELD;
> > }
> > - if (opts->num_accecn_fields < required)
> > + if (opts->num_accecn_fields < required) {
> > + if (opts->num_sack_blocks > 2) {
> > + /* Try to fit the option by removing one SACK block */
> > + opts->num_sack_blocks--;
> > + size = tcp_options_fit_accecn(opts, required,
> > + remaining +
> > + TCPOLEN_SACK_PERBLOCK,
> > +
> > + max_combine_saving);
>
> How deep is the recursion level, worst case? In any case please try to avoid
> recursion entirely. Possibly a 'goto' statement would help.
>
> /P
Hi Paolo,
Thanks for the feedback, in this case up to 2 extra levels will be needed.
And the recursion will be avoided in the next version.
BRs,
Chia-Yu
