From mboxrd@z Thu Jan 1 00:00:00 1970 From: William Lee Irwin III Subject: Re: net patches Date: Tue, 4 Jan 2005 22:48:57 -0800 Message-ID: <20050105064857.GD7961@holomorphy.com> References: <20050104215913.274e6567.akpm@osdl.org> <20050104223104.2ec87f5b.davem@davemloft.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Cc: Andrew Morton , jgarzik@pobox.com, netdev@oss.sgi.com Return-path: To: "David S. Miller" Content-Disposition: inline In-Reply-To: <20050104223104.2ec87f5b.davem@davemloft.net> Sender: netdev-bounce@oss.sgi.com Errors-to: netdev-bounce@oss.sgi.com List-Id: netdev.vger.kernel.org On Tue, 4 Jan 2005 21:59:13 -0800 Andrew Morton wrote: >> I'll unload some accumulated net patches. I can't say that I've looked at >> them very closely. On Tue, Jan 04, 2005 at 10:31:04PM -0800, David S. Miller wrote: > I'll be quiet for the next few days as I try to discover why > sparc64 now explodes after the past week of changes. I'm > heavily suspecting the 4-level page table stuff. wli kept > saying that he was having trouble getting his sparc boxes > working with the -mm tree over the past month and that's about > how long the 4-level page table stuff has been in there. > What's odd is that Andi's original 4-level page table patch > he gave to me worked just fine on sparc64, but it's been > changed a lot. I eventually narrowed this down to something associated with some unusual core fault handling changes with an unclear relationship to Andi's 4-level code. Something odd was going on with spec-guided definitions of pte_read(), pte_write(), et al. sun4u seemed to do okay with Andi's earlier code in my testing; later problems seemed to stem from elsewhere (e.g. the ioctl bogon). $ uname -a Linux analyticity 2.6.10-rc2-mm2 #8 SMP Sat Dec 18 07:57:42 PST 2004 sparc64 GNU/Linux -- wli