[PATCH v2 6/7] iommu/rockchip: use DMA API to map, to flush cache
shunqian.zheng at gmail.com
Mon Jun 13 03:31:11 PDT 2016
On 2016年06月13日 18:21, Tomasz Figa wrote:
> On Mon, Jun 13, 2016 at 6:56 PM, Shunqian Zheng
> <shunqian.zheng at gmail.com> wrote:
>> On 2016年06月10日 17:10, Tomasz Figa wrote:
>>> On Wed, Jun 8, 2016 at 10:26 PM, Shunqian Zheng <zhengsq at rock-chips.com>
>>>> Use DMA API instead of architecture internal functions like
>>>> __cpuc_flush_dcache_area() etc.
>>>> To support the virtual device like DRM the virtual slave iommu
>>>> added in the previous patch, attaching to which the DRM can use
>>>> it own domain->dev for dma_map_*(), dma_sync_*() even VOP is disabled.
>>>> With this patch, this driver is available for ARM64 like RK3399.
>>> Could we instead simply allocate coherent memory for page tables using
>>> dma_alloc_coherent() and skip any flushing on CPU side completely? If
>>> I'm looking correctly, the driver only reads back the page directory
>>> when checking if there is a need to allocate new page table, so there
>>> shouldn't be any significant penalty for disabling the cache.
>> I try to use dma_alloc_coherent() to replace the dma_map_single(),
>> but it doesn't work for me properly.
>> Because the DRM uses the iommu_dma_ops instead the swiotlb_dma_ops after
>> to iommu, so when the iommu domain need to alloc a new page in
>> it would call:
>> rk_iommu_map() --> dma_alloc_coherent() --> ops->alloc() --> iommu_map()
>> --> rk_iommu_map()
> It shouldn't call iommu_map(), because the IOMMU is not behind another
> IOMMU. Are you sure you called dma_alloc_coherent() on behalf of the
> IOMMU struct device and not the DRM device?
I called dma_alloc_coherent() with DRM device but not the IOMMU device,
because DRM didn't attach to any iommu. Even allocating an virtual one
when attaching, the iommu->dev
is DRM device though.
Am I right here?
Thank you very much,
> Best regards,
More information about the Linux-rockchip