Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S934479AbXJQLKL (ORCPT ); Wed, 17 Oct 2007 07:10:11 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1759784AbXJQLJw (ORCPT ); Wed, 17 Oct 2007 07:09:52 -0400 Received: from 74-93-104-97-Washington.hfc.comcastbusiness.net ([74.93.104.97]:47332 "EHLO sunset.davemloft.net" rhost-flags-OK-FAIL-OK-OK) by vger.kernel.org with ESMTP id S1761587AbXJQLJu (ORCPT ); Wed, 17 Oct 2007 07:09:50 -0400 Date: Wed, 17 Oct 2007 04:10:04 -0700 (PDT) Message-Id: <20071017.041004.21594558.davem@davemloft.net> To: jens.axboe@oracle.com Cc: fujita.tomonori@lab.ntt.co.jp, linux-kernel@vger.kernel.org, linux-scsi@vger.kernel.org Subject: Re: [PATCH] SPARC64: fix iommu sg chaining From: David Miller In-Reply-To: <20071017110142.GU5043@kernel.dk> References: <20071017.035441.74746708.davem@davemloft.net> <20071017105840.GT5043@kernel.dk> <20071017110142.GU5043@kernel.dk> X-Mailer: Mew version 5.1.52 on Emacs 21.4 / Mule 5.0 (SAKAKI) Mime-Version: 1.0 Content-Type: Text/Plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 4469 Lines: 132 From: Jens Axboe Date: Wed, 17 Oct 2007 13:01:42 +0200 > Actually, just clearing AFTER sg_next() would be fine, since we know > that is not a link entry. Duh... Yes and I'm running a kernel successfully with this fix. Jens, please also add the following on top of Fujita-san's most recent sparc64 patch and we should be good to go. >From 0f6e2c3085ec57df78b249a8722323692f33a1b2 Mon Sep 17 00:00:00 2001 From: David S. Miller Date: Wed, 17 Oct 2007 04:08:48 -0700 Subject: [PATCH] [SPARC64]: Fix loop terminating conditions in fill_sg(). Signed-off-by: David S. Miller --- arch/sparc64/kernel/iommu.c | 12 +++++++----- arch/sparc64/kernel/pci_sun4v.c | 15 ++++++++------- 2 files changed, 15 insertions(+), 12 deletions(-) diff --git a/arch/sparc64/kernel/iommu.c b/arch/sparc64/kernel/iommu.c index 5d4e96d..29af777 100644 --- a/arch/sparc64/kernel/iommu.c +++ b/arch/sparc64/kernel/iommu.c @@ -475,12 +475,11 @@ static void dma_4u_unmap_single(struct device *dev, dma_addr_t bus_addr, #define SG_ENT_PHYS_ADDRESS(SG) \ (__pa(page_address((SG)->page)) + (SG)->offset) -static inline void fill_sg(iopte_t *iopte, struct scatterlist *sg, - int nused, int nelems, - unsigned long iopte_protection) +static void fill_sg(iopte_t *iopte, struct scatterlist *sg, + int nused, int nelems, + unsigned long iopte_protection) { struct scatterlist *dma_sg = sg; - struct scatterlist *sg_end = sg_last(sg, nelems); int i; for (i = 0; i < nused; i++) { @@ -516,6 +515,7 @@ static inline void fill_sg(iopte_t *iopte, struct scatterlist *sg, break; } sg = sg_next(sg); + nelems--; } pteval = iopte_protection | (pteval & IOPTE_PAGE); @@ -529,18 +529,20 @@ static inline void fill_sg(iopte_t *iopte, struct scatterlist *sg, pteval = (pteval & IOPTE_PAGE) + len; sg = sg_next(sg); + nelems--; /* Skip over any tail mappings we've fully mapped, * adjusting pteval along the way. Stop when we * detect a page crossing event. */ - while (sg != sg_end && + while (nelems && (pteval << (64 - IO_PAGE_SHIFT)) != 0UL && (pteval == SG_ENT_PHYS_ADDRESS(sg)) && ((pteval ^ (SG_ENT_PHYS_ADDRESS(sg) + sg->length - 1UL)) >> IO_PAGE_SHIFT) == 0UL) { pteval += sg->length; sg = sg_next(sg); + nelems--; } if ((pteval << (64 - IO_PAGE_SHIFT)) == 0UL) pteval = ~0UL; diff --git a/arch/sparc64/kernel/pci_sun4v.c b/arch/sparc64/kernel/pci_sun4v.c index 119f8ef..fe46ace 100644 --- a/arch/sparc64/kernel/pci_sun4v.c +++ b/arch/sparc64/kernel/pci_sun4v.c @@ -368,12 +368,11 @@ static void dma_4v_unmap_single(struct device *dev, dma_addr_t bus_addr, #define SG_ENT_PHYS_ADDRESS(SG) \ (__pa(page_address((SG)->page)) + (SG)->offset) -static inline long fill_sg(long entry, struct device *dev, - struct scatterlist *sg, - int nused, int nelems, unsigned long prot) +static long fill_sg(long entry, struct device *dev, + struct scatterlist *sg, + int nused, int nelems, unsigned long prot) { struct scatterlist *dma_sg = sg; - struct scatterlist *sg_end = sg_last(sg, nelems); unsigned long flags; int i; @@ -414,6 +413,7 @@ static inline long fill_sg(long entry, struct device *dev, break; } sg = sg_next(sg); + nelems--; } pteval = (pteval & IOPTE_PAGE); @@ -432,19 +432,20 @@ static inline long fill_sg(long entry, struct device *dev, pteval = (pteval & IOPTE_PAGE) + len; sg = sg_next(sg); + nelems--; /* Skip over any tail mappings we've fully mapped, * adjusting pteval along the way. Stop when we * detect a page crossing event. */ - while ((pteval << (64 - IO_PAGE_SHIFT)) != 0UL && + while (nelems && + (pteval << (64 - IO_PAGE_SHIFT)) != 0UL && (pteval == SG_ENT_PHYS_ADDRESS(sg)) && ((pteval ^ (SG_ENT_PHYS_ADDRESS(sg) + sg->length - 1UL)) >> IO_PAGE_SHIFT) == 0UL) { pteval += sg->length; - if (sg == sg_end) - break; sg = sg_next(sg); + nelems--; } if ((pteval << (64 - IO_PAGE_SHIFT)) == 0UL) pteval = ~0UL; -- 1.5.3.4 - 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/