From mboxrd@z Thu Jan 1 00:00:00 1970 From: linux@arm.linux.org.uk (Russell King - ARM Linux) Date: Mon, 31 Oct 2011 10:36:10 +0000 Subject: [RFC PATCH] arm: dma: Drop GFP_COMP for DMA memory allocations In-Reply-To: <1319836732-31687-1-git-send-email-swarren@nvidia.com> References: <1319836732-31687-1-git-send-email-swarren@nvidia.com> Message-ID: <20111031103610.GG19187@n2100.arm.linux.org.uk> To: linux-arm-kernel@lists.infradead.org List-Id: linux-arm-kernel.lists.infradead.org On Fri, Oct 28, 2011 at 03:18:52PM -0600, Stephen Warren wrote: > I believe the patch below was once posted to solve this for ARM, but never > got picked up. Evidently, AVR32 did pick it up. I'm wondering if we should > pick this up for ARM now, or if there's some other solution? Well, Nick did talk about a solution using GFP_USERMAP in one of those threads, which doesn't seem to have gone anywhere. So I guess we need this patch. > diff --git a/arch/arm/mm/dma-mapping.c b/arch/arm/mm/dma-mapping.c > index 82a093c..93b86e6 100644 > --- a/arch/arm/mm/dma-mapping.c > +++ b/arch/arm/mm/dma-mapping.c > @@ -308,6 +308,13 @@ __dma_alloc(struct device *dev, size_t size, dma_addr_t *handle, gfp_t gfp, > struct page *page; > void *addr; > > + /* Following is a work-around (a.k.a. hack) to prevent pages > + * with __GFP_COMP being passed to split_page() which cannot > + * handle them. The real problem is that this flag probably > + * should be 0 on ARM as it is not supported on this > + * platform--see CONFIG_HUGETLB_PAGE. */ I don't think this comment makes sense anymore - HUGETLB_PAGE doesn't have any help text. > + gfp &= ~(__GFP_COMP); > + > *handle = ~0; > size = PAGE_ALIGN(size); > > -- > 1.7.0.4 > > From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S933743Ab1JaKjg (ORCPT ); Mon, 31 Oct 2011 06:39:36 -0400 Received: from caramon.arm.linux.org.uk ([78.32.30.218]:37086 "EHLO caramon.arm.linux.org.uk" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S932676Ab1JaKjf (ORCPT ); Mon, 31 Oct 2011 06:39:35 -0400 Date: Mon, 31 Oct 2011 10:36:10 +0000 From: Russell King - ARM Linux To: Stephen Warren Cc: Marc Singer , LKML , linux-arm-kernel@lists.infradead.org, Sumit Bhattacharya Subject: Re: [RFC PATCH] arm: dma: Drop GFP_COMP for DMA memory allocations Message-ID: <20111031103610.GG19187@n2100.arm.linux.org.uk> References: <1319836732-31687-1-git-send-email-swarren@nvidia.com> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <1319836732-31687-1-git-send-email-swarren@nvidia.com> User-Agent: Mutt/1.5.19 (2009-01-05) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Fri, Oct 28, 2011 at 03:18:52PM -0600, Stephen Warren wrote: > I believe the patch below was once posted to solve this for ARM, but never > got picked up. Evidently, AVR32 did pick it up. I'm wondering if we should > pick this up for ARM now, or if there's some other solution? Well, Nick did talk about a solution using GFP_USERMAP in one of those threads, which doesn't seem to have gone anywhere. So I guess we need this patch. > diff --git a/arch/arm/mm/dma-mapping.c b/arch/arm/mm/dma-mapping.c > index 82a093c..93b86e6 100644 > --- a/arch/arm/mm/dma-mapping.c > +++ b/arch/arm/mm/dma-mapping.c > @@ -308,6 +308,13 @@ __dma_alloc(struct device *dev, size_t size, dma_addr_t *handle, gfp_t gfp, > struct page *page; > void *addr; > > + /* Following is a work-around (a.k.a. hack) to prevent pages > + * with __GFP_COMP being passed to split_page() which cannot > + * handle them. The real problem is that this flag probably > + * should be 0 on ARM as it is not supported on this > + * platform--see CONFIG_HUGETLB_PAGE. */ I don't think this comment makes sense anymore - HUGETLB_PAGE doesn't have any help text. > + gfp &= ~(__GFP_COMP); > + > *handle = ~0; > size = PAGE_ALIGN(size); > > -- > 1.7.0.4 > >