On Sun, Feb 27, 2022 at 7:31 PM Christoph Hellwig wrote:
>
> The IOMMU table tries to separate the different IOMMUs into different
> backends, but actually requires various cross calls.
>
> Rewrite the code to do the generic swiotlb/swiotlb-xen setup directly
> in pci-dma.c and then just call into
On Wed, Mar 02, 2022 at 12:18:26PM +0300, Anatoly Pugachev wrote:
> Is it possible to keep documentation comments in source files? Or are
> they completely irrelevant now?
That ones you quoted are very much irrelevant now. And the behaviour
of the swiotlb disabling will have to change (this patch
On Tue, Mar 01, 2022 at 01:20:22PM -0500, Konrad Rzeszutek Wilk wrote:
> I think you also need to check for IBM Calgary?
The IBM Calgary IOMMU support is long gone.
___
iommu mailing list
iommu@lists.linux-foundation.org
https://lists.linuxfoundation.org
> -#include
> -
> -/*
> - * History lesson:
> - * The execution chain of IOMMUs in 2.6.36 looks as so:
> - *
> - *[xen-swiotlb]
> - * |
> - * +[swiotlb *]--+
> - */ | \
> - * / | \
> - *[GART] [Calgary]
The IOMMU table tries to separate the different IOMMUs into different
backends, but actually requires various cross calls.
Rewrite the code to do the generic swiotlb/swiotlb-xen setup directly
in pci-dma.c and then just call into the IOMMU drivers.
Signed-off-by: Christoph Hellwig
---
arch/ia64
On 2/22/22 9:05 PM, Christoph Hellwig wrote:
> The IOMMU table tries to separate the different IOMMUs into different
> backends, but actually requires various cross calls.
>
> Rewrite the code to do the generic swiotlb/swiotlb-xen setup directly
> in pci-dma.c and then just call into the IOMMU d
The IOMMU table tries to separate the different IOMMUs into different
backends, but actually requires various cross calls.
Rewrite the code to do the generic swiotlb/swiotlb-xen setup directly
in pci-dma.c and then just call into the IOMMU drivers.
Signed-off-by: Christoph Hellwig
---
arch/ia64
