From mboxrd@z Thu Jan 1 00:00:00 1970 From: "Justin T. Gibbs" Subject: Re: Aic7xxx v6.2.23 and Aic79xx v1.3.0Alpha3 in bitkeeper Date: Mon, 16 Dec 2002 12:16:28 -0700 Sender: linux-scsi-owner@vger.kernel.org Message-ID: <46560000.1040066188@aslan.btc.adaptec.com> References: <200212161746.gBGHkjE03623@localhost.localdomain> <20021216185659.A10346@infradead.org> Reply-To: "Justin T. Gibbs" Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Return-path: In-Reply-To: <20021216185659.A10346@infradead.org> Content-Disposition: inline List-Id: linux-scsi@vger.kernel.org To: Christoph Hellwig , James Bottomley Cc: linux-scsi@vger.kernel.org > On Mon, Dec 16, 2002 at 11:46:45AM -0600, James Bottomley wrote: >> I've moved the previous aic7xxx repository to scsi-aic7xxx-old-2.5 and >> set up a brand new one in its place. This one should have the import >> of Justin's actual patch sets, so should be easier for us all to manage. >> >> The new repository is at >> >> http://linux-scsi.bkbits.net/scsi-aic7xxx-2.5 > > From a first look it seems almost fine. There's a small problem in > the ->biosparam implementations, and at least a part of that is > probably my fault. > > In aic7xxx there is a ahd_linux_biosparam prototype instead of > ahc_linux_biosparam and the actual implementation doesn't use the > sector_div wrapper I sent justing yet, so it probably fails on system > that are configured with 64bit block numbers. Also my cleanups to > get the ifdef mess down in ->biosparam don't seem to be there, maybe > my style doesn't match Justins (?). I just forgot to finish the integration into aic7xxx. Ooops. I'll fix that today. > I've seen Justin moved aic_sector_div into the common aiclib.c file, > which seems to be a good idea to me, and I think there's more in the > *_osm.c headers that should go this way. I think it should become > an module of it's own at least for the kernel tree, even if Adaptec > wants to keep it included in their tarballs. The aiclib stuff should become a module, but until we can validate that having it as a separate module works correctly for driver update diskette, and initrd scenarios, we've basically punted on that. There is a comment to that effect in aiclib.h under the heading "Module Library Hack". -- Justin