ttm: Utilize the DMA API for pages that have TTM_PAGE_FLAG_DMA32 set.
authorKonrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Mon, 29 Nov 2010 19:03:30 +0000 (14:03 -0500)
committerKonrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Thu, 27 Jan 2011 21:03:09 +0000 (16:03 -0500)
commit69a07f0b117a40fcc1a479358d8e1f41793617f2
treefc3b827b9a9c7898b35d5cffd5995e56805cc4d8
parentf9820a46dd7888b05a36e81166fb1abcc47dcc3f
ttm: Utilize the DMA API for pages that have TTM_PAGE_FLAG_DMA32 set.

For pages that have the TTM_PAGE_FLAG_DMA32 flag set we
use the DMA API. We save the bus address in our array which we
use to program the GART (see "radeon/ttm/PCIe: Use dma_addr if TTM
has set it." and "nouveau/ttm/PCIe: Use dma_addr if TTM has set it.").

The reason behind using the DMA API is that under Xen we would
end up programming the GART with the bounce buffer (SWIOTLB)
DMA address instead of the physical DMA address of the TTM page.
The reason being that alloc_page with GFP_DMA32 does not allocate
pages under the the 4GB mark when running under Xen hypervisor.

Under baremetal this means we do the DMA API call earlier instead
of when we program the GART.

For details please refer to:
https://lkml.org/lkml/2011/1/7/251

[v2: Fixed indentation, revised desc, added Reviewed-by]
Reviewed-by: Thomas Hellstrom <thomas@shipmail.org>
Signed-off-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Tested-by: Ian Campbell <ian.campbell@citrix.com>
drivers/gpu/drm/ttm/ttm_page_alloc.c