Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1753572AbYKROIZ (ORCPT ); Tue, 18 Nov 2008 09:08:25 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1752523AbYKROIP (ORCPT ); Tue, 18 Nov 2008 09:08:15 -0500 Received: from netops-testserver-3-out.sgi.com ([192.48.171.28]:45824 "EHLO relay.sgi.com" rhost-flags-OK-OK-OK-FAIL) by vger.kernel.org with ESMTP id S1752267AbYKROIO (ORCPT ); Tue, 18 Nov 2008 09:08:14 -0500 From: John Keller Message-Id: <200811181408.mAIE8BHH057134@fcbayern.americas.sgi.com> Subject: Re: [PATCH] ia64: SN specific version of dma_get_required_mask() To: bwalle@suse.de (Bernhard Walle) Date: Tue, 18 Nov 2008 08:08:11 -0600 (CST) Cc: linux-ia64@vger.kernel.org, linux-kernel@vger.kernel.org In-Reply-To: <20081117173930.35342b31@hale.suse.de> from "Bernhard Walle" at Nov 17, 2008 05:39:30 PM X-Mailer: ELM [version 2.5 PL2] MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 1647 Lines: 45 > > * John Keller [2008-11-17 10:24]: > > > > Create a platform specific version of dma_get_required_mask() > > for ia64 SN Altix. All SN Altix platforms support 64 bit DMA > > addressing regardless of the size of system memory. > > Create a ia64 machvec for dma_get_required_mask, with the > > SN version unconditionally returning DMA_64BIT_MASK. > > Maybe we should then switch the ia64_platform_is("sn2") > in check_crashkernel_memory() to dma_get_required_mask() == > DMA_64BIT_MASK because that's basically what the check is about. > > BTW: What about UV? > > > Regards, > Bernhard > -- > Bernhard Walle, SUSE Linux Products GmbH, Architecture Development > This patch addresses a problem on SN Altix systems with < 4GB, where device drivers using the dma_get_required_mask() API would be told to use 32 bit DMA, when 64 bit is more efficient. How exactly the use of dma_get_required_mask() relates to the crash kernel code you refer to is unclear to me. If, for all platforms, the crash kernel code could use the mask returned from dma_get_required_mask() to do its check, then switching the code might be OK. But, if that's not possible for some platforms, then I'd wonder if dma_get_required_mask() is being used in the wrong context in this case. For UV systems, the default/generic dma_get_required_mask() will be used, returning a value based on the size of system memory. John -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/