summaryrefslogtreecommitdiff
diff options
context:
space:
mode:
authorChris Snook <csnook@redhat.com>2007-09-20 15:57:15 -0400
committerDavid S. Miller <davem@sunset.davemloft.net>2007-10-10 16:53:49 -0700
commitcdcc520d7b73445c3552a70786afed9a2b22c010 (patch)
tree948eb73c7ca53c27736151739d99655e909e5b63
parent01faccbf866195831af202de59f37e29467a3d74 (diff)
downloadlinux-3.10-cdcc520d7b73445c3552a70786afed9a2b22c010.tar.gz
linux-3.10-cdcc520d7b73445c3552a70786afed9a2b22c010.tar.bz2
linux-3.10-cdcc520d7b73445c3552a70786afed9a2b22c010.zip
atl1: explain 32-bit DMA restriction
Document the fact that atl1 uses a single shared register for the high 32 bits of 64-bit DMA addresses, making 64-bit DMA more trouble than it's worth. Signed-off-by: Chris Snook <csnook@redhat.com> Signed-off-by: Jeff Garzik <jeff@garzik.org>
-rw-r--r--drivers/net/atl1/atl1_main.c10
1 files changed, 8 insertions, 2 deletions
diff --git a/drivers/net/atl1/atl1_main.c b/drivers/net/atl1/atl1_main.c
index e1a9223d0c1..4c728f1169c 100644
--- a/drivers/net/atl1/atl1_main.c
+++ b/drivers/net/atl1/atl1_main.c
@@ -2209,8 +2209,14 @@ static int __devinit atl1_probe(struct pci_dev *pdev,
return err;
/*
- * 64-bit DMA currently has data corruption problems, so let's just
- * use 32-bit DMA for now. This is a big hack that is probably wrong.
+ * The atl1 chip can DMA to 64-bit addresses, but it uses a single
+ * shared register for the high 32 bits, so only a single, aligned,
+ * 4 GB physical address range can be used at a time.
+ *
+ * Supporting 64-bit DMA on this hardware is more trouble than it's
+ * worth. It is far easier to limit to 32-bit DMA than update
+ * various kernel subsystems to support the mechanics required by a
+ * fixed-high-32-bit system.
*/
err = pci_set_dma_mask(pdev, DMA_32BIT_MASK);
if (err) {