Re: [Soekris] net5501: FreeBSD ipfw and the elusive 75Mbps throughput
Andrew Atrens [and...@atrens.ca] wrote: > BUGS > The vr driver always copies transmit mbuf chains into longword-aligned > buffers prior to transmission in order to pacify the Rhine chips. If > buffers are not aligned correctly, the chip will round the supplied > buffer address and begin DMAing from the wrong location. This buffer > copying impairs transmit performance on slower systems but cannot be > avoided. On faster machines (e.g. a Pentium II), the performance > impact > is much less noticeable. This is not true for the version of the chip included with Soekris and PC Engines boxes, nor is it true for the OpenBSD or FreeBSD drivers. OpenBSD's vr driver has interrupt mitigation and hw vlan tagging that FreeBSD does (did?) not. Both enable hw checksum capability for IPv4. ___ Soekris-tech mailing list Soekris-tech@lists.soekris.com http://lists.soekris.com/mailman/listinfo/soekris-tech
Re: [Soekris] net5501: FreeBSD ipfw and the elusive 75Mbps throughput
On 2016-06-10, Jed Clearwrote: >> The vr driver always copies transmit mbuf chains into longword-aligned >> buffers prior to transmission in order to pacify the Rhine chips. If >> buffers are not aligned correctly, the chip will round the supplied >> buffer address and begin DMAing from the wrong location. This buffer >> copying impairs transmit performance on slower systems but cannot be >> avoided. On faster machines (e.g. a Pentium II), the performance >> impact >> is much less noticeable. > > Interesting. You'd think the drivers would align the packets on Rx and I > vaguely recall that FreeBSD managed to go zero copy a few major versions ago. > Although that could be for straight forwarding. IPFW and NAT might behave > differently. Not sure I'd have any control over it. The manpage is outdated, this is only done for the old chips. It's not needed and isn't done for VT610x. ___ Soekris-tech mailing list Soekris-tech@lists.soekris.com http://lists.soekris.com/mailman/listinfo/soekris-tech