Date: Tue, 21 Sep 2004 18:26:10 -0500 (EST) From: Sam <sah@softcardsystems.com> To: Julian Elischer <julian@elischer.org> Cc: freebsd-arch@freebsd.org Subject: Re: AoE for 4.x Message-ID: <Pine.LNX.4.60.0409211808020.2887@athena> In-Reply-To: <41508FEB.6030203@elischer.org> References: <Pine.LNX.4.60.0409211531450.32120@athena> <415083CC.7C041DFD@freebsd.org><41508827.8010007@elischer.org> <41508FEB.6030203@elischer.org>
next in thread | previous in thread | raw e-mail | index | archive | help
> Sam wrote: > >>> I'm sure you mentionned this before, but why ar eyou not doing it against >>> -current first? >>> >>> We generally don't allow changes to 4.x untill they have settled a bit in >>> -current. (or 5.x) >> >> >> I decided to write a 4.x driver thinking it would a) be easier to port >> forward than backward, and b) would give me experience in the old kernel >> to understand why certain changes were made for the new. >> >> [Thompson's Rule for First-Time Telescope Makers] >> It is faster to make a four-inch mirror and then a six-inch mirror than >> to make a six-inch mirror. > > > hmmmm it might be the other way around here... > >> >> >> As for the single patch, I'll cook one up shortly. I just figured it >> would be more of a pain to look through. > > > I guess we are used to it. > > I can't recommend enough the utility of having a cvs mirror and using cvs to > keep your > work trees sync'd against FreeBSD. It is also a goodd way of making sure that > you don't > forget parts of the diff (e.g. Makefiles etc.) as CVS will pick up all the > changed files. > And I can't thank you enough. I forgot about a few other (rather) important file changes this picked up. A patchfile for aoe against today's 4.10 source tree is at: http://www.coraid.com/support/freebsd/aoe.patch A few notes: The file sys/dev/aoe/aoe4bsd.dd is basic documentation for the driver. All references to chr major 200 will have to be changed to whatever gets assigned, obviously. The netisr selected in sys/net/netisr.h was picked at random. Is it OK? The label typeunknown in sys/net/if_ethersubr.c is added to give netgraph a chance to get the packets if AoE is not loaded. Appropriate? I dinna know how to fill out the copyright/license comment in the headers, but i gave it a shot. Thoughts here would be appreciated. TIA to all the curious eyes that skim the driver for glaring flaws. Cheers, Sam
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?Pine.LNX.4.60.0409211808020.2887>