Date: Sat, 24 May 2025 16:38:21 +0100 From: void <void@f-m.fm> To: freebsd-net@freebsd.org Cc: Peter Blok <pblok@bsd4all.org>, Ronald Klop <ronald-lists@klop.ws>, net@freebsd.org, stable@freebsd.org Subject: Re: net.link.bridge.member_ifaddrs has landed in stable/14 Message-ID: <aDHn7fivh7nn6MLG@int21h> In-Reply-To: <aDCy5LDMq4gr3yoF@ragweed.eden.le-fay.org> References: <aC1Q4u97cnqCZed3@ragweed.eden.le-fay.org> <aDB2MNbtGiBLW4kK@int21h> <1456605763.8629.1748007827309@localhost> <187BDEF1-CA44-4BF4-B6F3-3CEEEDA8F802@bsd4all.org> <aDCy5LDMq4gr3yoF@ragweed.eden.le-fay.org>
next in thread | previous in thread | raw e-mail | index | archive | help
On Fri, May 23, 2025 at 06:39:48PM +0100, ivy@FreeBSD.org wrote: >this should be fixed by https://reviews.freebsd.org/D50477, as long as >you have allow_llz_overlap set to 0 (which is the default, and there's >almost never any reason to change it). Just built against your patch and it's all working, TYVM. FreeBSD 14.3-STABLE #2 stable/14-n271551-4e0a5fd7b608-dirty: Sat May 24 15:16:54 BST 2025 also: net.link.bridge.allow_llz_overlap: 0 net.link.bridge.member_ifaddrs="0" my ifconfig_em0 looks like this: ifconfig_em0="media 1000baseT mediaopt full-duplex up" bridge0: ifconfig_bridge0="inet 192.168.1.2/24 addm em0 addm tap0 addm tap1" speed tests against a freebsd vm installed on tap0 showed no significant speed reduction compared to baremetal, which is fantastic. This sort of thing I was looking at back in October last year with iperf3, when freebsd guests had sometimes less than half the expected throughput. Next up to test is a -current system with many VMs. --
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?aDHn7fivh7nn6MLG>