From mboxrd@z Thu Jan 1 00:00:00 1970 From: Tony Lindgren Subject: Re: [PATCH 1/4] OMAP: DMA: Cleanup of DMA Request Lines Date: Wed, 31 Oct 2007 06:26:24 -0700 Message-ID: <20071031132624.GK32547@atomide.com> References: <010C7BAE6783F34D9AC336EE5A01A08804AF985E@dbde01.ent.ti.com> <20071031131059.GG32547@atomide.com> <3B6D69C3A9EBCA4BA5DA60D913027429025C0830@dlee13.ent.ti.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Return-path: Content-Disposition: inline In-Reply-To: <3B6D69C3A9EBCA4BA5DA60D913027429025C0830@dlee13.ent.ti.com> List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Sender: linux-omap-open-source-bounces@linux.omap.com Errors-To: linux-omap-open-source-bounces@linux.omap.com To: "Woodruff, Richard" Cc: linux-omap-open-source@linux.omap.com List-Id: linux-omap@vger.kernel.org * Woodruff, Richard [071031 06:19]: > > Let's remove the #ifdefs here, we want to be able to compile support > for > > 24xx and 34xx into the same kernel eventually. It's OK to have the > same > > DMA request channel defined in different way for 24xx and 34xx. > > One question here is will that happen practically? Maybe it will, not > sure. The same sentiments were made about OMAP1 vs. OMAP2. The > structure which results from the sharing intent perhaps is net positive > but I don't see them converging. > > People would likely compiled 3430 for ARMv7 and ARMv6 for 24xx. Some > new instructions and the like are around. I don't that its fully > pluggable at the architecture level even. Recent gcc allows mixing in assembly for later cores, so it's not a problem. Sure you can optimize it for one core by not selecting anything else. > EABI if really used might allow some fancy solutions in theory but I > wonder if those will develop. Not to say this the only way. It may take a while, but let's not add any new barriers to compile in multiple cores. Regards, Tony