Re: mlx4 fix for 36-bit bus addresses on 32-bit arch (was: [PATCH 1/1 (resend)] mthca: Modify to support embedded PowerPC platforms)

[Eli Cohen <eli-LDSdmyG8hGV8YrgS2mwiifqBs+8SCbDb@public.gmane.org>]

On Wed, Jan 12, 2011 at 09:53:00AM -0800, Roland Dreier wrote:
>  
> -	ibdev->uar_map = ioremap(ibdev->priv_uar.pfn << PAGE_SHIFT, PAGE_SIZE);
> +	ibdev->uar_map = ioremap((phys_addr_t) ibdev->priv_uar.pfn << PAGE_SHIFT,
> +				 PAGE_SIZE);
>  	if (!ibdev->uar_map)
>  		goto err_uar;

CX devices are 64 bit devices with respect to the PCI bus. As far as I
know, a 64 bit PCI device memory space, could be placed anywhere in
that space regardless of the CPU architecture. So I think it would be
more appropriate to to use u64 casting. I know that ioremap accepts
phys_addr_t as its first argument but I wonder if ioremap shouldn't be
changed to accept u64 for that argument.

Regardless of whether what I say above is correct or not, wouldn't it
be nicer to to define pfn as either u64 or phys_addr_t and avoid the
casting?

--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html