From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([2001:4830:134:3::10]:47705) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1anL1e-0000Fd-Nv for qemu-devel@nongnu.org; Tue, 05 Apr 2016 03:03:47 -0400 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1anL1a-00040K-O1 for qemu-devel@nongnu.org; Tue, 05 Apr 2016 03:03:46 -0400 Received: from e06smtp06.uk.ibm.com ([195.75.94.102]:53727) by eggs.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1anL1a-0003zM-DZ for qemu-devel@nongnu.org; Tue, 05 Apr 2016 03:03:42 -0400 Received: from localhost by e06smtp06.uk.ibm.com with IBM ESMTP SMTP Gateway: Authorized Use Only! Violators will be prosecuted for from ; Tue, 5 Apr 2016 08:03:38 +0100 References: <1459822643-4770-1-git-send-email-david@gibson.dropbear.id.au> <1459822643-4770-2-git-send-email-david@gibson.dropbear.id.au> <1459822790.20087.65.camel@kernel.crashing.org> <20160405032533.GS16485@voom.fritz.box> From: =?UTF-8?Q?C=c3=a9dric_Le_Goater?= Message-ID: <5703633D.8080101@fr.ibm.com> Date: Tue, 5 Apr 2016 09:03:25 +0200 MIME-Version: 1.0 In-Reply-To: <20160405032533.GS16485@voom.fritz.box> Content-Type: text/plain; charset=windows-1252 Content-Transfer-Encoding: 8bit Subject: Re: [Qemu-devel] [PULL 1/3] ppc: Rework POWER7 & POWER8 exception model List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: David Gibson , Benjamin Herrenschmidt Cc: peter.maydell@linaro.org, qemu-devel@nongnu.org, aik@ozlabs.ru, mdroth@linux.vnet.ibm.com, agraf@suse.de, qemu-ppc@nongnu.org, pbonzini@redhat.com On 04/05/2016 05:25 AM, David Gibson wrote: > On Tue, Apr 05, 2016 at 12:19:50PM +1000, Benjamin Herrenschmidt wrote: >> On Tue, 2016-04-05 at 12:17 +1000, David Gibson wrote: >>> From: Cédric Le Goater >>> >>> From: Benjamin Herrenschmidt >>> >>> This patch fixes the current AIL implementation for POWER8. The >>> interrupt vector address can be calculated directly from LPCR when >>> the >>> exception is handled. The excp_prefix update becomes useless and we >>> can cleanup the H_SET_MODE hcall. >> >> Beware, iirc, this depends on the new cpu_set_papr() stuff I did so we >> get the right LPCR values in PAPR mode. > > Right, Cédric already submitted that before the 2.6 freeze, and it was > merged as 26a7f129, AFAICT. Well, yes, but cpu_ppc_set_papr() only handles the AMOR setting, the LPCR settings were kept for later as they were not bug fixes. As for now, powerpc_excp() checks the ILE bit and uses the AIL bits to calculate the vector address, which was done before in the H_SET_MODE hcall. This allows some simplification, getting rid of excp_prefix, and fixes a migration bug in TCG. C.