Date: Mon, 14 Dec 2009 19:14:16 -0800 From: Pyun YongHyeon <pyunyh@gmail.com> To: Carey Jones <mcj@bluetonic.org> Cc: freebsd-net@freebsd.org, yongari@freebsd.org Subject: Re: kern/141276: [vge] Strange behavior with vge gigabit ethernet adapter Message-ID: <20091215031416.GE1122@michelle.cdnetworks.com> In-Reply-To: <77e6ee410912141858w2395cefcrcd689ba0f028aec6@mail.gmail.com> References: <200912142238.nBEMcbT4080874@freefall.freebsd.org> <77e6ee410912141536i7731618xe8ec0553ac5c4956@mail.gmail.com> <20091215010641.GD1122@michelle.cdnetworks.com> <77e6ee410912141858w2395cefcrcd689ba0f028aec6@mail.gmail.com>
next in thread | previous in thread | raw e-mail | index | archive | help
On Mon, Dec 14, 2009 at 08:58:19PM -0600, Carey Jones wrote: > Hello, > > The version from HEAD seems to work better than any of the previous patches > I've tried, though it still doesn't seem to get quite the throughput that > the adapter managed under RELENG_7. Unfortunately I don't have any > quantitative data for you. However, the symptoms from my pr are definitely > gone. > Glad to hear that. I have strong evidence that the vge(4) can't saturate the link for any frame size. I still have no idea why it shows poor TX performance(about 650Mbps) compared to RX(about 930Mbps). The best performance number I can get from my local patch was about 800Mbps which is still very lower than any other gigabit controllers. I'm working on it with various scenarios but I couldn't find clue. > Thanks,
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20091215031416.GE1122>