From mboxrd@z Thu Jan 1 00:00:00 1970 Date: Thu, 20 Nov 2003 13:17:17 +0100 From: Christoph Hellwig To: Wolfgang Denk Cc: Christoph Hellwig , Tom Rini , Dale Farnsworth , linuxppc-dev@lists.linuxppc.org Subject: Re: MPC5200 Patches Message-ID: <20031120121717.GA7764@lst.de> References: <20031120115315.GA7411@lst.de> <20031120121037.90F93C5F5F@atlas.denx.de> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii In-Reply-To: <20031120121037.90F93C5F5F@atlas.denx.de> Sender: owner-linuxppc-dev@lists.linuxppc.org List-Id: On Thu, Nov 20, 2003 at 01:10:32PM +0100, Wolfgang Denk wrote: > In message <20031120115315.GA7411@lst.de> you wrote: > > > > Umm, the code needs a rewrite before it's suitable for mainline. In > > Are you volunteering? ;-) Sure, feel free to contract me for that. > > it's current state is just bogus utter crap. The bestcomm API needs > > to go away and the dma engine used directly. > > It would be nice if you explained your reasons for such a statement. Because Linux is about doing things right, not fast? Look at the bogus taks abstractions - you just call into some deep multiplexer calls to load pregenerated and (accoring to benh who worked with the even more buggy predessecor) often bogus. It looks like MOT once again produced a really crappy chip and now hides the interface behind a huge bloated API. That might be find for them to sell the POS to someone wearing Suits, but it's not really enough to get it included into the kernel tree. ** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/