Re: [PATCH v13 0/5] vfio/dma-buf: add TPH support for peer-to-peer access - ping

[Leon Romanovsky <leon@kernel.org>]

On Tue, Sep 15, 2026 at 02:59:43PM +0200, Christian König wrote:
> On 9/15/26 13:37, Leon Romanovsky wrote:
> > On Fri, Sep 11, 2026 at 09:58:28AM +0200, Christian König wrote:
> >> Hi Zhiping,
> >>
> >> sorry I'm completely underwater at the moment and don't have time to take another look at the full set.
> >>
> >> But IIRC you already fixed my documentation requirements and skimming over the patch once more I can't see anything wrong of hand.
> >>
> >> So feel free to add Acked-by: Christian König <christian.koenig@amd.com> to patch "dma-buf: add optional get_pci_tph() callback" and push it upstream through the VFIO channels.
> > 
> > Christian,
> > 
> > I'm slightly confused here. In the other thread, you expressed concerns
> > about having PCI details common to every device in the P2P world in the
> > dma-buf code, but here you seem fine with adding a niche P2P feature.
> 
> That one is only using u16 and u8 types and has no dependency on the PCI subsystem whatsoever.
> 
> Giving those adjacent information sounds perfectly fine to me, we just can't have any hard dependency on PCI subsystem types or functions.
> 
> DMA-buf itself is massively used with vendor specific P2P interconnects, PCI P2P is just a special case of that.

You are looking at it from the DRM perspective. Outside the DRM world, PCI is
THE interconnect, everything else is a special case :).

Thanks

> 
> Regards,
> Christian.
> 
> > 
> > Thanks,