Re: [PATCH] v4l-compliance: re-introduce NON_COHERENT and cache hints tests

[Sergey Senozhatsky <senozhatsky@chromium.org>]

Hello,

On (21/09/10 14:20), Hans Verkuil wrote:
> Hi Sergey,
> 
> I've applied the "[PATCHv6 0/8] videobuf2: support new noncontiguous DMA API" series
> to my kernel, and applied this patch to v4l-utils.
> 
> Running 'test-media vim2m' in contrib/test results in the following compliance failures:
> 
> Streaming ioctls:
>         test read/write: OK (Not Supported)
>         test blocking wait: OK
>         Video Capture: Captured 8 buffers
>         test MMAP (no poll): OK
>         Video Capture: Captured 8 buffers
>         test MMAP (select): OK
>         Video Capture: Captured 8 buffers
>         test MMAP (epoll): OK
>         Video Capture: Captured 8 buffers
>         test USERPTR (no poll): OK
>         Video Capture: Captured 8 buffers
>         test USERPTR (select): OK
>                 fail: v4l2-test-buffers.cpp(1869): !(flags & V4L2_BUF_FLAG_NO_CACHE_INVALIDATE)
>                 fail: v4l2-test-buffers.cpp(1932): setupDmaBuf(expbuf_node, node, q, exp_q)
>         test DMABUF (no poll): FAIL
>                 fail: v4l2-test-buffers.cpp(1869): !(flags & V4L2_BUF_FLAG_NO_CACHE_INVALIDATE)
>                 fail: v4l2-test-buffers.cpp(1932): setupDmaBuf(expbuf_node, node, q, exp_q)
>         test DMABUF (select): FAIL
> 
> The same happens with e.g. vivid, but vim2m is quicker to test.
> 
> I'm not sure whether this is a bug in this v4l2-compliance patch or whether it is
> a bug in the v6 series, but it should be checked.

Looking into it now. I ran v4l2-compliance, but not "contrib/test/test-media"