From mboxrd@z Thu Jan 1 00:00:00 1970 Date: Fri, 11 May 2001 17:36:20 -0600 From: Val Henson To: Michael Schmitz Cc: Ken Offer , linuxppc-dev Subject: Re: Problem with linuxppc smp.c Message-ID: <20010511173620.K18959@boardwalk> References: <3AF8554D.BF250E0A@arlut.utexas.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii In-Reply-To: ; from schmitz@opal.biophys.uni-duesseldorf.de on Wed, May 09, 2001 at 12:54:36PM +0200 Sender: owner-linuxppc-dev@lists.linuxppc.org List-Id: On Wed, May 09, 2001 at 12:54:36PM +0200, Michael Schmitz wrote: > > > smp.c: In function `smp_gemini_kick_cpu': > > smp.c:670: `i' undeclared (first use in this function) > > smp.c:670: (Each undeclared identifier is reported only once > > smp.c:670: for each function it appears in.) > > smp.c: In function `smp_gemini_setup_cpu': > > smp.c:679: `cpu_nr' undeclared (first use in this function) > > > > Any thoughts? I guess I could try to grok the author's meaning and fix > > this... ;-) > > I guess it's obvious that both of these should be declared int. Actually, this dates back from the last big SMP reorganization (in September 2000 or so). The variable names weren't changed to match the variable name passed to the (newly created) function. Gemini hasn't compiled SMP since then. I've written one patch to make SMP work on the Gemini (similar to an earlier patch by Dan Cox) but Cort asked for some rewrites. I'm working on it as we speak. For the confused, I'm the official Gemini board maintainer now (taking over from Dan Cox). -VAL ** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/