From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: From: Arnd Bergmann To: Meador Inge Subject: Re: [PATCH v2 1/3] powerpc: Removing support for 'protected-sources' Date: Fri, 4 Feb 2011 13:17:32 +0100 References: <1296697900-14004-1-git-send-email-meador_inge@mentor.com> <201102031656.38222.arnd@arndb.de> <4D4B3A4E.2070106@mentor.com> In-Reply-To: <4D4B3A4E.2070106@mentor.com> MIME-Version: 1.0 Content-Type: Text/Plain; charset="iso-8859-1" Message-Id: <201102041317.32739.arnd@arndb.de> Cc: Hollis Blanchard , devicetree-discuss@lists.ozlabs.org, linuxppc-dev@lists.ozlabs.org List-Id: Linux on PowerPC Developers Mail List List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , On Friday 04 February 2011, Meador Inge wrote: > On 02/03/2011 09:56 AM, Arnd Bergmann wrote: > So in 'mpic_init' we don't initialize anything and then in > 'mpic_host_map' we lazily do the VECPRI and CPU binding initialization with: > > if (!(mpic->flags & MPIC_WANTS_RESET)) > if (!(mpic_is_ipi(mpic, hw) > || mpic_is_timer_interrupt(mpic, hw))) > mpic_init_vector(mpic, hw); > > Thus when 'no-reset' is thrown it ensures that only the sources which > are mentioned in the device tree are actually initialized. The net > effect should be the same as what 'protected-sources' was accomplishing, > but without having to maintain the list of sources in the property cell. That sounds like a good idea, but unfortunately, it's not what SLOF implements on QS21/QS22. It's a legacy product and there won't be any firmware updates. Moreover, it relies on the open firmware implementation and cannot boot with a flattened device tree image, so I don't see how your patch can work on the old systems. Maybe you can treat the presence of a 'protected-sources' property the same way that you treat the no-reset property? Arnd From mboxrd@z Thu Jan 1 00:00:00 1970 From: Arnd Bergmann Subject: Re: [PATCH v2 1/3] powerpc: Removing support for 'protected-sources' Date: Fri, 4 Feb 2011 13:17:32 +0100 Message-ID: <201102041317.32739.arnd@arndb.de> References: <1296697900-14004-1-git-send-email-meador_inge@mentor.com> <201102031656.38222.arnd@arndb.de> <4D4B3A4E.2070106@mentor.com> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Return-path: In-Reply-To: <4D4B3A4E.2070106-nmGgyN9QBj3QT0dZR+AlfA@public.gmane.org> List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Sender: devicetree-discuss-bounces+gldd-devicetree-discuss=m.gmane.org-uLR06cmDAlY/bJ5BZ2RsiQ@public.gmane.org Errors-To: devicetree-discuss-bounces+gldd-devicetree-discuss=m.gmane.org-uLR06cmDAlY/bJ5BZ2RsiQ@public.gmane.org To: Meador Inge Cc: Hollis Blanchard , devicetree-discuss-uLR06cmDAlY/bJ5BZ2RsiQ@public.gmane.org, linuxppc-dev-uLR06cmDAlY/bJ5BZ2RsiQ@public.gmane.org List-Id: devicetree@vger.kernel.org On Friday 04 February 2011, Meador Inge wrote: > On 02/03/2011 09:56 AM, Arnd Bergmann wrote: > So in 'mpic_init' we don't initialize anything and then in > 'mpic_host_map' we lazily do the VECPRI and CPU binding initialization with: > > if (!(mpic->flags & MPIC_WANTS_RESET)) > if (!(mpic_is_ipi(mpic, hw) > || mpic_is_timer_interrupt(mpic, hw))) > mpic_init_vector(mpic, hw); > > Thus when 'no-reset' is thrown it ensures that only the sources which > are mentioned in the device tree are actually initialized. The net > effect should be the same as what 'protected-sources' was accomplishing, > but without having to maintain the list of sources in the property cell. That sounds like a good idea, but unfortunately, it's not what SLOF implements on QS21/QS22. It's a legacy product and there won't be any firmware updates. Moreover, it relies on the open firmware implementation and cannot boot with a flattened device tree image, so I don't see how your patch can work on the old systems. Maybe you can treat the presence of a 'protected-sources' property the same way that you treat the no-reset property? Arnd