Usage with cudaHostAlloc

[htngr]

I'm trying to use this library on a Nvidia jetson nano. My goal is to go directly from v4l2 to the jetson's GPU without going through the CPU (zero-copy). There are official samples from nvidia which accomplishes this with a Userptr to GPU / shared memory (allocated via cudaHostAlloc). How can I use this crate for this purpose? I couldn't find a way to pass a pointer to the UserptrStream.
An alternative could be DMA but it seems to me that this isn't implemented yet?

[molysgaard]

I have done exactly the same. See this for reference: #103
This is another sample point that allowing the user of the library to specify how memory is allocated should be supported :)

[htngr]

I have done exactly the same. See this for reference: #103 This is another sample point that allowing the user of the library to specify how memory is allocated should be supported :)

Thanks, this seems to work (after writing a custom allocator)

[raymanfx]

dup: #103