From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from xyzzy.farnsworth.org (xyzzy.farnsworth.org [65.39.95.219]) (using TLSv1 with cipher AES256-SHA (256/256 bits)) (Client did not present a certificate) by ozlabs.org (Postfix) with ESMTPS id 7DF40DDDF7 for ; Sat, 5 Apr 2008 16:41:54 +1100 (EST) Date: Fri, 4 Apr 2008 22:41:46 -0700 From: Dale Farnsworth To: Benjamin Herrenschmidt Message-ID: <20080405054146.GA10233@farnsworth.org> References: <20080404213932.GA15847@farnsworth.org> <1207351869.6675.0.camel@johannes.berg> <1207372496.10388.441.camel@pasglop> MIME-Version: 1.0 In-Reply-To: <1207372496.10388.441.camel@pasglop> Subject: Re: [PATCH v2] powerpc: Add irqtrace support for 32-bit powerpc Content-Type: text/plain; charset=us-ascii Cc: linuxppc-dev@ozlabs.org, Johannes Berg List-Id: Linux on PowerPC Developers Mail List List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , On Sat, Apr 05, 2008 at 04:14:56PM +1100, Benjamin Herrenschmidt wrote: > > On Sat, 2008-04-05 at 01:31 +0200, Johannes Berg wrote: > > > This version takes care to save and restore *all* the volatile registers > > > around the call to trace_hardirqs_off in transfer_to_handler, and should > > > fix the anomalies seen by Johannes Berg. I've tested on 7447- and 8548- > > > based boards. > > > > Still has problems on my machine. Maybe to do with non-maskable PMU > > interrupts? (saw a patch about something like that float by, don't > > really have any idea whether it's related to this work or not) > > Or more likely he missed some clobber somewhere, it took me a while to > get that right on ppc64 (well... hopefully it -is- right by now :-) Yep, likely. I'd appreciate help finding it. I'm at a loss. :-) -Dale