Messages in this thread | ![/](/images/icornerl.gif) | | From | Logan Gunthorpe <> | Date | Wed, 9 Oct 2019 10:17:36 -0600 | Subject | Re: [PATCH 1/3] iommu/amd: Implement dma_[un]map_resource() |
| |
On 2019-10-09 12:57 a.m., Christoph Hellwig wrote: > On Tue, Oct 08, 2019 at 04:18:35PM -0600, Logan Gunthorpe wrote: >> From: Kit Chow <kchow@gigaio.com> >> >> Currently the Intel IOMMU uses the default dma_[un]map_resource() > > s/Intel/AMD/ ?
Oops, yes, my mistake.
>> +static dma_addr_t map_resource(struct device *dev, phys_addr_t paddr, >> + size_t size, enum dma_data_direction dir, unsigned long attrs) >> +{ >> + struct protection_domain *domain; >> + struct dma_ops_domain *dma_dom; >> + >> + domain = get_domain(dev); >> + if (PTR_ERR(domain) == -EINVAL) >> + return (dma_addr_t)paddr; > > I thought that case can't happen anymore? > > Also note that Joerg just applied the patch to convert the AMD iommu > driver to use the dma-iommu ops. Can you test that series and check > it does the right thing for your use case? From looking at the code > I think it should.
Yes, looking at the new code, it looks like this patch will not be needed. So we can drop it. We'll test it to make sure.
I believe the other two patches in this series are still needed though.
Thanks,
Logan
| ![\](/images/icornerr.gif) |