powerpc/swiotlb: Fix off by one in determining boundary of which ops to use

Submitted by Kumar Gala on May 3, 2010, 12:43 p.m.


Message ID 1272890584-28631-1-git-send-email-galak@kernel.crashing.org
State Accepted, archived
Headers show

Commit Message

Kumar Gala May 3, 2010, 12:43 p.m.
When we compare the devices DMA mask to the amount of memory we need to
make sure we treat the DMA mask as an address boundary.  For example if
the DMA_MASK(32) and we have 4G of memory we'd incorrectly set the dma
ops to swiotlb.  We need to add one to the dma mask when we convert it.

Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
 arch/powerpc/kernel/dma-swiotlb.c |    2 +-
 1 files changed, 1 insertions(+), 1 deletions(-)

Patch hide | download patch | download mbox

diff --git a/arch/powerpc/kernel/dma-swiotlb.c b/arch/powerpc/kernel/dma-swiotlb.c
index 59c9285..f9ee4b0 100644
--- a/arch/powerpc/kernel/dma-swiotlb.c
+++ b/arch/powerpc/kernel/dma-swiotlb.c
@@ -70,7 +70,7 @@  static int ppc_swiotlb_bus_notify(struct notifier_block *nb,
 	sd->max_direct_dma_addr = 0;
 	/* May need to bounce if the device can't address all of DRAM */
-	if (dma_get_mask(dev) < lmb_end_of_DRAM())
+	if ((dma_get_mask(dev) + 1) < lmb_end_of_DRAM())
 		set_dma_ops(dev, &swiotlb_dma_ops);
 	return NOTIFY_DONE;