amdgpu/TTM oopses since merging swiotlb_dma_ops into the dma_direct code

[Sibren Vasse]

On Mon, 14 Jan 2019 at 19:13, Christoph Hellwig <hch at lst.de> wrote:
>
> Hmm, I wonder if we are not actually using swiotlb in the end,
> can you check if your dmesg contains this line or not?
>
> PCI-DMA: Using software bounce buffering for IO (SWIOTLB)
This line does not appear in my dmesg.

>
> If not I guess we found a bug in swiotlb exit vs is_swiotlb_buffer,
> and you can try this patch:
>
> diff --git a/kernel/dma/swiotlb.c b/kernel/dma/swiotlb.c
> index d6361776dc5c..1fb6fd68b9c7 100644
> --- a/kernel/dma/swiotlb.c
> +++ b/kernel/dma/swiotlb.c
> @@ -378,6 +378,8 @@ void __init swiotlb_exit(void)
>                 memblock_free_late(io_tlb_start,
>                                    PAGE_ALIGN(io_tlb_nslabs << IO_TLB_SHIFT));
>         }
> +       io_tlb_start = 0;
> +       io_tlb_end = 0;
>         io_tlb_nslabs = 0;
>         max_segment = 0;
>  }
With the patch applied to v5.0-rc2 I can no longer reproduce the issue.