[PATCH] arm: dma: Drop GFP_COMP for DMA memory allocations

[PATCH] arm: dma: Drop GFP_COMP for DMA memory allocations

[Stephen Warren]

From: Sumit Bhattacharya <sumitb@nvidia.com>

dma_alloc_coherent wants to split pages after allocation in order to
reduce the memory footprint. This does not work well with GFP_COMP
pages, so drop this flag before allocation.

This patch is ported from arch/avr32
(commit 3611553ef985ef7c5863c8a94641738addd04cff).

Signed-off-by: Sumit Bhattacharya <sumitb@nvidia.com>
Tested-by: Varun Colbert <vcolbert@nvidia.com>
[swarren: s/HUGETLB_PAGE/HUGETLBFS/ in comment, minor comment cleanup]
Signed-off-by: Stephen Warren <swarren@nvidia.com>
---
 arch/arm/mm/dma-mapping.c |    9 +++++++++
 1 files changed, 9 insertions(+), 0 deletions(-)

diff --git a/arch/arm/mm/dma-mapping.c b/arch/arm/mm/dma-mapping.c
index e4e7f6c..68d64b4 100644
--- a/arch/arm/mm/dma-mapping.c
+++ b/arch/arm/mm/dma-mapping.c
@@ -332,6 +332,15 @@ __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_HUGETLBFS.
+	 */
+	gfp &= ~(__GFP_COMP);
+
 	*handle = ~0;
 	size = PAGE_ALIGN(size);
 
-- 
1.7.0.4

RE: [PATCH] arm: dma: Drop GFP_COMP for DMA memory allocations

[Stephen Warren]

Stephen Warren wrote at Friday, November 04, 2011 3:36 PM:
> dma_alloc_coherent wants to split pages after allocation in order to
> reduce the memory footprint. This does not work well with GFP_COMP
> pages, so drop this flag before allocation.
> 
> This patch is ported from arch/avr32
> (commit 3611553ef985ef7c5863c8a94641738addd04cff).
> 
> Signed-off-by: Sumit Bhattacharya <sumitb@nvidia.com>
> Tested-by: Varun Colbert <vcolbert@nvidia.com>
> [swarren: s/HUGETLB_PAGE/HUGETLBFS/ in comment, minor comment cleanup]
> Signed-off-by: Stephen Warren <swarren@nvidia.com>

Russell, does this look good to go in the patch system now?

> ---
>  arch/arm/mm/dma-mapping.c |    9 +++++++++
>  1 files changed, 9 insertions(+), 0 deletions(-)
> 
> diff --git a/arch/arm/mm/dma-mapping.c b/arch/arm/mm/dma-mapping.c
> index e4e7f6c..68d64b4 100644
> --- a/arch/arm/mm/dma-mapping.c
> +++ b/arch/arm/mm/dma-mapping.c
> @@ -332,6 +332,15 @@ __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_HUGETLBFS.
> +	 */
> +	gfp &= ~(__GFP_COMP);
> +
>  	*handle = ~0;
>  	size = PAGE_ALIGN(size);
> 
> --
> 1.7.0.4

-- 
nvpublic

RE: [PATCH] arm: dma: Drop GFP_COMP for DMA memory allocations

[Stephen Warren]

Stephen Warren wrote at Wednesday, November 09, 2011 11:04 AM:
> Stephen Warren wrote at Friday, November 04, 2011 3:36 PM:
> > dma_alloc_coherent wants to split pages after allocation in order to
> > reduce the memory footprint. This does not work well with GFP_COMP
> > pages, so drop this flag before allocation.
> >
> > This patch is ported from arch/avr32
> > (commit 3611553ef985ef7c5863c8a94641738addd04cff).
> >
> > Signed-off-by: Sumit Bhattacharya <sumitb@nvidia.com>
> > Tested-by: Varun Colbert <vcolbert@nvidia.com>
> > [swarren: s/HUGETLB_PAGE/HUGETLBFS/ in comment, minor comment cleanup]
> > Signed-off-by: Stephen Warren <swarren@nvidia.com>
> 
> Russell, does this look good to go in the patch system now?

ping?

> > ---
> >  arch/arm/mm/dma-mapping.c |    9 +++++++++
> >  1 files changed, 9 insertions(+), 0 deletions(-)
> >
> > diff --git a/arch/arm/mm/dma-mapping.c b/arch/arm/mm/dma-mapping.c
> > index e4e7f6c..68d64b4 100644
> > --- a/arch/arm/mm/dma-mapping.c
> > +++ b/arch/arm/mm/dma-mapping.c
> > @@ -332,6 +332,15 @@ __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_HUGETLBFS.
> > +	 */
> > +	gfp &= ~(__GFP_COMP);
> > +
> >  	*handle = ~0;
> >  	size = PAGE_ALIGN(size);
> >
> > --
> > 1.7.0.4

-- 
nvpublic

Re: [PATCH] arm: dma: Drop GFP_COMP for DMA memory allocations

[Russell King - ARM Linux]

On Wed, Nov 23, 2011 at 11:10:03AM -0800, Stephen Warren wrote:
> Stephen Warren wrote at Wednesday, November 09, 2011 11:04 AM:
> > Stephen Warren wrote at Friday, November 04, 2011 3:36 PM:
> > > dma_alloc_coherent wants to split pages after allocation in order to
> > > reduce the memory footprint. This does not work well with GFP_COMP
> > > pages, so drop this flag before allocation.
> > >
> > > This patch is ported from arch/avr32
> > > (commit 3611553ef985ef7c5863c8a94641738addd04cff).
> > >
> > > Signed-off-by: Sumit Bhattacharya <sumitb@nvidia.com>
> > > Tested-by: Varun Colbert <vcolbert@nvidia.com>
> > > [swarren: s/HUGETLB_PAGE/HUGETLBFS/ in comment, minor comment cleanup]
> > > Signed-off-by: Stephen Warren <swarren@nvidia.com>
> > 
> > Russell, does this look good to go in the patch system now?
> 
> ping?

I think so.