Date: Mon, 19 Dec 2005 14:13:46 -0500 From: Jung-uk Kim <jkim@FreeBSD.org> To: Jeremie Le Hen <jeremie@le-hen.org> Cc: cvs-src@FreeBSD.org, src-committers@FreeBSD.org, cvs-all@FreeBSD.org Subject: Re: cvs commit: src/sys/amd64/amd64 bpf_jit_machdep.c bpf_jit_machdep.h src/sys/conf files files.amd64 files.i386 options.amd64 options.i386 src/sys/i386/i386 bpf_jit_machdep.c bpf_jit_machdep.h src/sys/net bpf.c bpf_jitter.c bpf_jitter.h bpfdesc.h Message-ID: <200512191413.52731.jkim@FreeBSD.org> In-Reply-To: <20051219183835.GC3512@obiwan.tataz.chchile.org> References: <200512060258.jB62wCnk084452@repoman.freebsd.org> <20051219183835.GC3512@obiwan.tataz.chchile.org>
next in thread | previous in thread | raw e-mail | index | archive | help
On Monday 19 December 2005 01:38 pm, Jeremie Le Hen wrote: > Hi, Jung-uk, > > On Tue, Dec 06, 2005 at 02:58:12AM +0000, Jung-uk Kim wrote: > > jkim 2005-12-06 02:58:12 UTC > > > > FreeBSD src repository > > > > Modified files: > > sys/conf files files.amd64 files.i386 > > options.amd64 options.i386 > > sys/net bpf.c bpfdesc.h > > Added files: > > sys/amd64/amd64 bpf_jit_machdep.c bpf_jit_machdep.h > > sys/i386/i386 bpf_jit_machdep.c bpf_jit_machdep.h > > sys/net bpf_jitter.c bpf_jitter.h > > Log: > > Add experimental BPF Just-In-Time compiler for amd64 and i386. > > > > Use the following kernel configuration option to enable: > > > > options BPF_JITTER > > > > If you want to use bpf_filter() instead (e. g., debugging), do: > > > > sysctl net.bpf.jitter.enable=0 > > > > to turn it off. > > > > Currently BIOCSETWF and bpf_mtap2() are unsupported, and > > bpf_mtap() is partially supported because 1) no need, 2) avoid > > expensive m_copydata(9). > > > > Obtained from: WinPcap 3.1 (for i386) > > Though the name looks quite exciting, I don't really know what it > is. I tried to look on WinPcap's website as well as searching on > Google, but found nothing relevant. Could you explain in a few > words what it is and the difference with the old bpf(4) behaviour, > please ? BPF JIT compiler converts BPF instructions into native machine code when BIOCSETF ioctl is issued, and then the native filter is used to filter packets instead of bpf_filter(), which is a simple virtual machine. > Does this change will lead to a new note in bpf(4) manual > page once it won't be experimental any more ? Yes. Maybe bpf(9), too. Jung-uk Kim
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?200512191413.52731.jkim>