Messages in this thread | | | Date | Thu, 8 Jan 2009 10:55:04 +0900 | Subject | Re: [PATCH 0/13] IA64: unifying ways to handle multiple sets of dma mapping ops | From | FUJITA Tomonori <> |
| |
On Wed, 7 Jan 2009 20:56:46 +0100 Sam Ravnborg <sam@ravnborg.org> wrote:
> On Mon, Jan 05, 2009 at 11:36:05PM +0900, FUJITA Tomonori wrote: > > This patchset is the first part of the unification of ways to handle > > multiple sets of dma mapping API. The whole work consists of three > > patchset. This is for IA64 and can be applied independently. > > > > dma_mapping_ops (or dma_ops) struct is used to handle multiple sets of > > dma mapping API by X86, SPARC, and POWER. IA64 also handle multiple > > sets of dma mapping API but in a very different way (some define > > magic). > > > > X86 and IA64 share VT-d and SWIOTLB code. We need several workarounds > > for it because of the deference of ways to handle multiple sets of dma > > mapping API (e.g., X86 people can't freely change struct > > dma_mapping_ops in x86's dma-mapping.h now because it could break > > IA64). Seems POWER will use SWIOTLB code soon. I think that it's time > > to unify ways to handle multiple sets of dma mapping API. After > > applying the whole work, we have struct dma_map_ops > > include/linux/dma-mapping.h (I also dream of changing all the archs to > > use SWIOTLB in order to remove the bounce code in the block and > > network stacks...). > > > > This patchset changes IA64 to handle multiple sets of dma mapping API > > in the common way (as X86, SPARC, and POWER do): > > Do you have any plans to update sparc too? > Maybe it is not relevant.
I'll do after finishing IA64 and X86. It's not a huge gain since SPARC doesn't share IOMMU code with other architectures. But I think that it would be nice to remove arch/sparc/include/asm/dma-mapping_64.h
The long-term goal is removing the bounce code in the network and block stacks, and some drivers (using sorta SWIOTLB on non IOMMU systems instead). So I try to convert all the architectures to use this generic mechanism (though I'm not sure all the arch maintainers agree with it).
Thanks,
| |