Date: Sun, 11 Aug 2013 17:33:17 +1000 From: Darren Reed <darrenr@netbsd.org> To: Matt Thomas <matt@3am-software.com> Cc: tech-net@NetBSD.org, guy@alum.mit.edu, freebsd-net@freebsd.org Subject: Re: BPF_MISC+BPF_COP and BPF_COPX Message-ID: <52073E3D.8060008@netbsd.org> In-Reply-To: <D30E8B6A-A13F-4889-BA14-25D1955D0E97@3am-software.com> References: <20130804191310.2FFBB14A152@mail.netbsd.org> <5202693C.50608@netbsd.org> <20130807175548.1528014A21F@mail.netbsd.org> <5203535D.2040508@netbsd.org> <38CDC9BB-09C7-4241-8746-163BD15B80EC@cs.columbia.edu> <20130809203446.428A714A308@mail.netbsd.org> <20130809204436.GA3261@panix.com> <D30E8B6A-A13F-4889-BA14-25D1955D0E97@3am-software.com>
next in thread | previous in thread | raw e-mail | index | archive | help
On 10/08/2013 7:23 AM, Matt Thomas wrote: > ... > The possibility of the COP/COPX functions doing bad things is over wrought. It makes the assumption of avoiding BPF and then coding everything is safer than using BPF and COP/COPX functions. Depends on what you mean by "bad things." Thus far I haven't seen a proper problem statement, only this: > The problem is simple: I want a generic mechanism to offload more complex > packet inspection operations, e.g. lookup IP address in some container or > walk IPv6 headers and return some offsets. IMHO, the IPv6 problem is and will be common enough to deserve its own instructionas happened with IPv4 and determining the offset of the first byte after the IPv4 header. But the generic offload problem hasn't been explained nearly enough. Is it required for NPF? Is it required for tcpdump? Is it required for dhcpd? ... I don't think we know nearly enough about what the problem is in order to be able to judge whether or not the solution is acceptable. Darren
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?52073E3D.8060008>