From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mail.mailmij.org (mail.mailmij.org [82.93.140.149]) by ozlabs.org (Postfix) with ESMTP id 2198D67B6A for ; Wed, 9 Aug 2006 02:50:25 +1000 (EST) Date: Tue, 8 Aug 2006 18:50:23 +0200 From: danny@mailmij.org To: linuxppc-dev@ozlabs.org Subject: Re: [PATCH] fix for firewire patch added in 2.6.17.2 that breaks things on ppc Message-ID: <20060808185023.B21209@luna.ellen.dexterslabs.com> References: <20060805151050.B24484@luna.ellen.dexterslabs.com> <20060807063854.B641@luna.ellen.dexterslabs.com> <17624.7051.905145.470927@cargo.ozlabs.ibm.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii In-Reply-To: <17624.7051.905145.470927@cargo.ozlabs.ibm.com>; from paulus@samba.org on Tue, Aug 08, 2006 at 03:05:15PM +1000 List-Id: Linux on PowerPC Developers Mail List List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , On Tue, Aug 08, 2006 at 03:05:15PM +1000, Paul Mackerras wrote: > danny@mailmij.org writes: > > > -#endif > > - > > +#else > > pci_save_state(pdev); > > - > > +#endif /* CONFIG_PPC_PMAC */ > > Shouldn't we just move the pci_save_state call up before the > pmac_call_feature() call? > > Paul. > I tested this. It works, but it has no effect (pci_restore_state does not report that anything changed, thus this means that the config space already contained the correct values). I discovered one additional problem with leaving a device running through a sleep/resume cycle: suspend: ohci1394: fw-host0: SelfID received outside of bus reset sequence resume: ohci1394: fw-host0: SelfID is inconsistent [0x7e80773f/0x7e80773f] The last leads to the node not being able to resume when the device is actually plugged. Anyway, it's a separate issue, I will investigate. Danny