Re: [PATCH 3/3] x86: make GART to respect device's dma_mask about virtual mappings

[FUJITA Tomonori <fujita.tomonori@lab.ntt.co.jp>]

On Fri, 12 Sep 2008 16:52:27 +0200
Joerg Roedel <joerg.roedel@amd.com> wrote:

> On Fri, Sep 12, 2008 at 07:42:35PM +0900, FUJITA Tomonori wrote:
> > Currently, GART IOMMU ingores device's dma_mask when it does virtual
> > mappings. So it could give a device a virtual address that the device
> > can't access to.
> > 
> > This patch fixes the above problem.
> > 
> > Signed-off-by: FUJITA Tomonori <fujita.tomonori@lab.ntt.co.jp>
> > ---
> >  arch/x86/kernel/pci-gart_64.c |   39 ++++++++++++++++++++++++++++-----------
> >  1 files changed, 28 insertions(+), 11 deletions(-)
> > 
> > diff --git a/arch/x86/kernel/pci-gart_64.c b/arch/x86/kernel/pci-gart_64.c
> > index 47abe43..9739d56 100644
> > --- a/arch/x86/kernel/pci-gart_64.c
> > +++ b/arch/x86/kernel/pci-gart_64.c
> > @@ -83,23 +83,34 @@ static unsigned long next_bit;  /* protected by iommu_bitmap_lock */
> >  static int need_flush;		/* global flush state. set for each gart wrap */
> >  
> >  static unsigned long alloc_iommu(struct device *dev, int size,
> > -				 unsigned long align_mask)
> > +				 unsigned long align_mask, u64 dma_mask)
> 
> You can calculate the dma_mask in this function from the dev parameter.
> There is no need to pass it two levels down to this function extending
> various parameter lists.

No, we can't because we need to use dev->coherent_dma_mask for
alloc_coherent and dev->dma_mask for the rest.