Date: Fri, 24 Apr 1998 16:32:50 -0600 From: "Justin T. Gibbs" <gibbs@plutotech.com> To: Julian Elischer <julian@whistle.com> Cc: "Justin T. Gibbs" <gibbs@plutotech.com>, Garrett Wollman <wollman@khavrinen.lcs.mit.edu>, current@FreeBSD.ORG Subject: Re: Bandwidth throttling etc. Message-ID: <199804242236.QAA15340@pluto.plutotech.com> In-Reply-To: Your message of "Fri, 24 Apr 1998 15:21:21 PDT." <35411060.102F11D5@whistle.com>
next in thread | previous in thread | raw e-mail | index | archive | help
>> How do deal with a route change between the time you determine the >> target interface, ask it to allocate space for you, and construct >> the packet? The race window becomes much larger if the application >> asks to pre-allocate space that it will reuse repeatedly to achieve >> zero copy. Whatever the interface is, it must also handle external >> memory allocations such as dmaing the payload out of the memory on >> another device. > >of course.. each driver has to be able to handle the case where >the mbuf is not one of it's own. It goes beyond this. What if the original interface had the ability to compute TCP checksums in hardware, but the second interface could not. Not only do you have to be able to send the data, but you may have to kick it back to a routine to perform the checksums. >As I mentioned in another mail, we've been doing this for 10 years >at TFS. it works fine. Does TFS do hardware based checksumming too? -- Justin To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-current" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199804242236.QAA15340>