From: Daniel Ritz <daniel.ritz@gmx.ch>
It turns out that just blindly enabling read prefetch is wrong for the
older (and buggy) oz6833 bridge. fixed in the attached patch.
- don't touch the oz6832/6833
- more verbose
- only set the read prefetch and the write burst bits
it broke on some old laptop with a cardbus tulip.
Signed-off-by: Andrew Morton <akpm@osdl.org>
---
25-akpm/drivers/pcmcia/o2micro.h | 19 +++++++++++++++++--
1 files changed, 17 insertions(+), 2 deletions(-)
diff -puN drivers/pcmcia/o2micro.h~yenta-dont-enable-read-prefetch-on-older-o2-bridges drivers/pcmcia/o2micro.h
--- 25/drivers/pcmcia/o2micro.h~yenta-dont-enable-read-prefetch-on-older-o2-bridges 2004-10-26 19:02:15.983407360 -0700
+++ 25-akpm/drivers/pcmcia/o2micro.h 2004-10-26 19:02:15.987406752 -0700
@@ -129,9 +129,24 @@ static int o2micro_override(struct yenta
* ok to write to both registers on all O2 bridges.
* from Eric Still, 02Micro.
*/
+ u8 a, b;
+
if (PCI_FUNC(socket->dev->devfn) == 0) {
- config_writeb(socket, 0x94, 0xCA);
- config_writeb(socket, 0xD4, 0xCA);
+ a = config_readb(socket, 0x94);
+ b = config_readb(socket, 0xD4);
+
+ printk(KERN_INFO "Yenta O2: res at 0x94/0xD4: %02x/%02x\n", a, b);
+
+ switch (socket->dev->device) {
+ case PCI_DEVICE_ID_O2_6832:
+ printk(KERN_INFO "Yenta O2: old bridge, not enabling read prefetch / write burst\n");
+ break;
+
+ default:
+ printk(KERN_INFO "Yenta O2: enabling read prefetch/write burst\n");
+ config_writeb(socket, 0x94, a | 0x0a);
+ config_writeb(socket, 0xD4, b | 0x0a);
+ }
}
return 0;
_