Hi, On 05/05/2017 10:39 AM, Mathias Nyman wrote: > Hi > > On 04.05.2017 22:13, Matthias Lange wrote: >> Hi, >> >> I am wondering what the reason for the GFP_DMA flag in >> xhci_segment_alloc() (drivers/usb/host/xhci-mem.c:59) is. >> >> It is causing me problems in a virtualization setup with enabled IOMMU >> (x86_64). The result is that USB is not working in the VM. >> >> Could someone please shed some light on this? >> > > You refer to this line, right: > seg->bounce_buf = kzalloc(max_packet, flags | GFP_DMA); Yes. > GFP_DMA was set as to make sure seg->bounce_buf is within > the dma_mask as it is used as the *cpu_addr in dma_map_single() later. > > But looks like this is complete overkill and it will make sure it is > within the first 16MB ISA DMA addresses instead. Exactly that came to my mind. Thanks for pointing it out. > DMA-API.txt states for dma_map_single that: > > "To ensure that the memory allocated by kmalloc is within the dma_mask, > the driver may specify various platform-dependent flags to restrict > the DMA address range of the allocation (e.g., on x86, GFP_DMA > guarantees to be within the first 16MB of available DMA addresses, > as required by ISA devices)." > > So I don't think there is any reason it couldn't be just GFP_KERNEL instead I will send a patch. Best, Matthias. -- To unsubscribe from this list: send the line "unsubscribe linux-usb" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
