Date: Sat, 10 Jul 2021 14:33:51 +0200 From: Lutz Donnerhacke <lutz@donnerhacke.de> To: Stefan Esser <se@freebsd.org> Cc: Karl Denninger <karl@denninger.net>, stable@freebsd.org Subject: Re: [PATCH] Re: 12.2 Splay Tree ipfw potential panic source Message-ID: <20210710123351.GA30826@belenus.iks-jena.de> In-Reply-To: <a06435bb-65c4-c645-031a-dc1bbf121b20@freebsd.org> References: <2e3dcd4d-c8e6-8381-0010-d0844c99901e@denninger.net> <20210708221134.GA32658@belenus.iks-jena.de> <a6a9c220-fee6-a0ea-7721-f88ff865a6a8@denninger.net> <CAFMmRNy9K-1mTDoqQhgdChWV5f_n4QhNesz%2B6xWywn_TQ43xng@mail.gmail.com> <ca5beb7c-db38-1d3c-0f3c-b1b6a12c311e@denninger.net> <7bfda38b-cf81-d8be-7691-e18946e6b56e@denninger.net> <dde6a01e-c41f-19be-593c-246eef11ea3b@freebsd.org> <a06435bb-65c4-c645-031a-dc1bbf121b20@freebsd.org>
next in thread | previous in thread | raw e-mail | index | archive | help
On Sat, Jul 10, 2021 at 10:52:48AM +0200, Stefan Esser wrote:
> > /sys/netinet/libalias/alias_db.c:1753:
> >
> > if (packets % packet_limit == 0) {
> >
> > Seems that packet_limit can become zero, there ...
> >
> > At line 1780 within that function:
> >
> > if (now != LibAliasTime) {
> > /* retry three times a second */
> > packet_limit = packets / 3;
> > packets = 0;
> > LibAliasTime = now;
> > }
Thank you for deducing the reason.
Seem to be my fault.
I only tested it with heavy traffic.
The +1 solution seems to be the best one, because the real number is not
important (+100 works, too) and this way the per packet code path is kept to
a minimum of instructions.
Does a bug ticket exists?
I'll issue a review tomorrow.
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20210710123351.GA30826>