From mboxrd@z Thu Jan 1 00:00:00 1970 From: Jesse Barnes Date: Sat, 05 Mar 2005 02:04:21 +0000 Subject: Re: need ia64 testers: no per-arch mem_map init Message-Id: <200503041804.21403.jbarnes@sgi.com> List-Id: References: <1109981490.1976.17.camel@localhost> In-Reply-To: <1109981490.1976.17.camel@localhost> MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit To: linux-ia64@vger.kernel.org On Friday, March 4, 2005 4:11 pm, Dave Hansen wrote: > I've sent this to some of you ia64 folks before, but this has changed > slightly. I'd like to send this to Andrew Morton for inclusion in -mm > in the next couple of days if everything goes OK. Bob Picco noticed > that it didn't work because of an oversight of mine with how vmem_map > works, but I think that's fixed now. > > In any case, this simplifies how all architectures set up the global > mem_map variable, and will help simplify interactions with the memory > hotplug code as well. > > However, ia64 has perhaps the most complex mem_map[ initialization, and > a diverse set of configurations. I'd much prefer to fix any corner > cases now, than try to fix it up after it's in -mm. I'd appreciate all > of the ia64 testing I can get, especially when CONFIG_NUMA is on. Sure, I'll test this on some Altix platforms (I'll try to find a big one), but I might not get to it until next Monday. Thanks, Jesse