From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from the.earth.li ([193.201.200.66]) by canuck.infradead.org with esmtps (Exim 4.61 #1 (Red Hat Linux)) id 1FYODA-0003CY-VO for linux-mtd@lists.infradead.org; Tue, 25 Apr 2006 10:08:25 -0400 Date: Tue, 25 Apr 2006 15:03:11 +0100 From: Jonathan McDowell To: David Woodhouse Message-ID: <20060425140311.GP7570@earth.li> References: <20060424161529.GR7570@earth.li> <1145972592.11909.392.camel@pmac.infradead.org> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <1145972592.11909.392.camel@pmac.infradead.org> Cc: linux-mtd@lists.infradead.org Subject: Re: [PATCH/RFC] Add Amstrad Delta NAND support. List-Id: Linux MTD discussion mailing list List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , On Tue, Apr 25, 2006 at 02:43:12PM +0100, David Woodhouse wrote: > On Mon, 2006-04-24 at 17:15 +0100, Jonathan McDowell wrote: > > In particular I'm using "udelay(1);" in the read/write routines to wait > > for the bus to settle after setting the read/write enable lines. Really > > I want "ndelay(40);", but ndelay doesn't seem to be implemented on ARM, > > except in terms of udelay. Can I get away with udelay(0.04)? Should I be > > doing something else entirely? > > That's a question best directed at linux-arm-kernel list. Ok, ta. > > +/* > > + * These 3 functions are basically identical to the ones in nand_base.c, but > > + * we have to call our read_byte/write_byte functions instead of readb/writeb > > + * directly. I submitted a patch to MTD to move this to nand_base.c, but it > > + * was deemed too intrusive. > > + */ > > Thomas' decision, I assume? I think he's right -- this way it can be > inlined in the loop rather than being a real function call for every > byte. Ben Dooks. I'd actually forgotten I'd left that comment in. I'll remove it before I submit the patch proper; it was mainly a note to people who might have picked up my earlier version that patches nand-core as well. J. -- /-\ | What's the worse that could |@/ Debian GNU/Linux Developer | happen? Smoke. - Anonymous \- | HWHacker