Date: Fri, 7 May 2021 19:15:11 +0200 From: Michael Tuexen <michael.tuexen@macmic.franken.de> To: Gordon Bergling <gbe@FreeBSD.org> Cc: Brandon Bergren <bdragon@freebsd.org>, Warner Losh <imp@bsdimp.com>, Li-Wen Hsu <lwhsu@freebsd.org>, Randall Stewart <rrs@freebsd.org>, src-committers <src-committers@freebsd.org>, "<dev-commits-src-all@freebsd.org>" <dev-commits-src-all@freebsd.org>, dev-commits-src-main@freebsd.org Subject: Re: git: 5d8fd932e418 - main - This brings into sync FreeBSD with the netflix versions of rack and bbr. This fixes several breakages (panics) since the tcp_lro code was committed that have been reported. Quite a few new features are now in rack (prefecting of DGP -- Dynamic Goodput Pacing among the largest). There is also support for ack-war prevention. Documents comming soon on rack.. Message-ID: <72392D50-5388-488A-AB1A-0A242D54E8A8@macmic.franken.de> In-Reply-To: <YJVx34tBMk6n9CjU@lion.0xfce3.net> References: <202105061526.146FQS0E008755@gitrepo.freebsd.org> <CAKBkRUy-5MCGEOYRHGk%2BtC3FKjgXmhKg5ppvn1x9eFtbcK1=VQ@mail.gmail.com> <D6646C1F-0DD2-44E6-AF77-BD04D027C556@macmic.franken.de> <CANCZdfqh5fUwVu%2BLT7=Q-ip4_VuTQW%2BZiZbtXFQpMyan4ZW8Rw@mail.gmail.com> <454D2A05-9CE3-4E60-80FD-0C5D97093A3B@macmic.franken.de> <66dab209-3419-4c82-b5de-398b463610d0@www.fastmail.com> <YJVx34tBMk6n9CjU@lion.0xfce3.net>
next in thread | previous in thread | raw e-mail | index | archive | help
> On 7. May 2021, at 18:59, Gordon Bergling <gbe@FreeBSD.org> wrote: >=20 > On Fri, May 07, 2021 at 11:51:06AM -0500, Brandon Bergren wrote: >> On Fri, May 7, 2021, at 11:30 AM, Michael Tuexen wrote: >>>> On 7. May 2021, at 18:11, Warner Losh <imp@bsdimp.com> wrote: >>>>=20 >>>> Yea. If we can't enable it on powerpc, I have some ideas on how to = keep these modules >>>> from building when RATELIMIT isn't enabled. >>> What is the problem on 32-bit PPC? Unfortunately, my G4 MacMini died=20= >>> last year, but it >>> is fine on 64-bit PPC. Can test it on a POWER9 system. >>=20 >> The powerpc LINT config explicitly disables RATELIMIT. I'm not sure = why though, it looks like it was part of the conversion from = automatically generated LINT to the manually curated one. >=20 > Besides LINT build issues, it would maybe a good idea to enable the = build > of the extra TCP stacks by default to get further adaption of them. They are not build by default on purpose... That might change possibly = for RACK in the future... >=20 > That could also minimize build problems. Sure. But they have been sort of experimental up to now. Best regards Michael >=20 > --Gordon
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?72392D50-5388-488A-AB1A-0A242D54E8A8>