<html> <head> <meta http-equiv="Content-Type" content="text/html; charset=utf-8"> </head> <body smarttemplateinserted="true" text="#000000" bgcolor="#FFFFFF"> <div id="smartTemplate4-template"> Many thanks! mmap succeeds - I removed the checks only to post the code on this list. I use this now: if ( !ioctl(container, VFIO_IOMMU_MAP_DMA, &dma_map) ) { if ( !ioctl(container, VFIO_IOMMU_MAP_DMA, &dma_map) ) { perror("mmap ioctl"); return 1; } } It mostly succeeds only on second try but I never needed more than two tries. So it's ok. </div> Re. IOMMU type 1 recognition: You are right: It is independent of kernel version. It seems to be related to auto-loading vfio_pci via /etc/modules.conf. Installing the module later gets it working. <div id="smartTemplate4-quoteHeader"> <hr style="color:#62B3FF"> <div style="background-color: #DDDDDD; font-size:10pt">From: <a title="alex.williamson@redhat.com" href="mailto:alex.williamson@redhat.com">Alex Williamson</a> Sent: Monday, March 12, 2018 6:57AM To: <a title="Oliver.Heid@h-next.de" href="mailto:Oliver.Heid@h-next.de">Heid, Oliver</a> Cc: <a title="vfio-users@redhat.com" href="mailto:vfio-users@redhat.com">Vfio-users</a> Subject: Re: [vfio-users] VFIO_IOMMU_MAP_DMA succeeds only on second try?</div> </div> <pre wrap="">On Mon, 5 Mar 2018 12:37:24 +0100 Oliver Heid <a class="moz-txt-link-rfc2396E" href="mailto:Oliver.Heid@h-next.de"><Oliver.Heid@h-next.de></a> wrote: </pre> <blockquote type="cite"> <pre wrap="">Do I miss something here? I want to allow RW access of a peripheral device to a memory region via struct vfio_iommu_type1_dma_map dma_map = { .argsz = sizeof(dma_map) }; __u32* mem = (__u32*)mmap(NULL,size,PROT_READ|PROT_WRITE, MAP_SHARED|MAP_ANONYMOUS|MAP_LOCKED, -1, 0); dma_map.argsz = sizeof(dma_map); dma_map.flags = VFIO_DMA_MAP_FLAG_READ | VFIO_DMA_MAP_FLAG_WRITE; dma_map.vaddr = (__u64)mem; dma_map.iova = 0; dma_map.size = size; ioctl(container, VFIO_IOMMU_MAP_DMA, &dma_map); Any idea why theVFIO_IOMMU_MAP_DMA ioctl fails with EINVAL, but then a second, identical call succeeds? Does it actually succeed then? QEMU re-tries if the first attempt fails with EBUSY, but not with EINVAL. </pre> </blockquote> <pre wrap=""> I don't think the QEMU -EBUSY behavior is related, that handles previous mappings and may no longer even be necessary. One of my unit tests[1] does a similar map with no issues. Your code excerpt doesn't check whether mmap succeeds, so for all we know mmap is failing too. </pre> <blockquote type="cite"> <pre wrap="">The latest 4.15.7 kernel (4.15.6 too IIRC) does not recognize the AMD Ryzen 3 1200 as VFIO_TYPE1_IOMMU so I am using version 4.15.4. </pre> </blockquote> <pre wrap=""> This doesn't make sense, anything implementing the kernel's IOMMU API in the kernel is compatible with type1, that includes anything supporting AMD-Vi. There is no whitelist or specific processor support selection. There are no iommu or vfio changes between the kernels you're quoting, perhaps you could do a bisect. Thanks, Alex [1] <a class="moz-txt-link-freetext" href="https://github.com/awilliam/tests/blob/master/vfio-iommu-map-unmap.c">https://github.com/awilliam/tests/blob/master/vfio-iommu-map-unmap.c</a> </pre> </body> </html>