Re: [RFC PATCH 01/15] PCI/P2PDMA: Don't sleep in upstream_bridge_distance_warn()

[Christoph Hellwig]

On Fri, Nov 06, 2020 at 10:00:22AM -0700, Logan Gunthorpe wrote:
> In order to call this function from a dma_map function, it must not sleep.
> The only reason it does sleep so to allocate the seqbuf to print
> which devices are within the ACS path.
> 
> Switch the kmalloc call to use GFP_NOWAIT and simply not print that
> message if the buffer fails to be allocated.

Please pass in the actual gfp_t.  Especially from an I/O path
GFP_NOWAIT is not the right gfp_t anyway, you probably want GFP_ATOMIC
there.  But also for the path where we can sleep we should allow that.