On Wed, 23 Jan 2019, hch@infradead.org wrote:
> On Wed, Jan 23, 2019 at 01:04:33PM -0800, Stefano Stabellini wrote:
> > If vring_use_dma_api is actually supposed to return true when
> > dma_dev->dma_mem is set, then both Peng's patch and the patch I wrote
> > are not fixing the real issue here.
> > 
> > I don't know enough about remoteproc to know where the problem actually
> > lies though.
> 
> The problem is the following:
> 
> Devices can declare a specific memory region that they want to use when
> the driver calls dma_alloc_coherent for the device, this is done using
> the shared-dma-pool DT attribute, which comes in two variants that
> would be a little to much to explain here.
> 
> remoteproc makes use of that because apparently the device can
> only communicate using that region.  But it then feeds back memory
> obtained with dma_alloc_coherent into the virtio code.  For that
> it calls vmalloc_to_page on the dma_alloc_coherent, which is a huge
> no-go for the ĐMA API and only worked accidentally on a few platform,
> and apparently arm64 just changed a few internals that made it stop
> working for remoteproc.
> 
> The right answer is to not use the DMA API to allocate memory from
> a device-speficic region, but to tie the driver directly into the
> DT reserved memory API in a way that allows it to easilt obtain
> a struct device for it.

If I understand correctly, Peng should be able to reproduce the problem
on native Linux without any Xen involvement simply by forcing
vring_use_dma_api to return true. Peng, can you confirm?

And the right fix is not to call vmalloc_to_page on a dma_alloc_coherent
buffer -- I don't know about the recent changes on arm64, but that's not
going to work with arm32 either AFAIK. Given that I don't have a repro,
I'll leave it to Peng and/or others to send the appropriate patch for
remoteproc.


> This is orthogonal to another issue, and that is that hardware
> virtio devices really always need to use the DMA API, otherwise
> we'll bypass such features as the device specific DMA pools,
> DMA offsets, cache flushing, etc, etc.

I understand, I'll drop my patch.