On Mon, Apr 26, 2004 at 11:48:21PM +0200, Guido Guenther wrote:
> When suspending pmdisk calls resume_device again. This resumes the usb
> devices too which is fine for the first two hubs ( 0001:01:1{8,9}.0 but
> I'm catching a machine check exception when it tries to resume
> 0001:01:1a.0. It happens when usb_hcd_pci_resume() calls
> hcd->driver-resume() for that device.
Replying to myself from a freshly suspended/resumed system: the attached
patch fixes a typo in pmac_features that didn't resume the 3rd USB
channel properly and so caused the above exception. It works now...
Cheers,
-- Guido
Index: arch/ppc/platforms/pmac_feature.c
===================================================================
--- arch/ppc/platforms/pmac_feature.c (revision 26)
+++ arch/ppc/platforms/pmac_feature.c (working copy)
@@ -1156,7 +1158,7 @@
(void)MACIO_IN32(KEYLARGO_FCR1);
mdelay(1);
LOCK(flags);
- MACIO_BIS(KEYLARGO_FCR0, KL1_USB2_CELL_ENABLE);
+ MACIO_BIS(KEYLARGO_FCR1, KL1_USB2_CELL_ENABLE);
}
if (number < 4) {
reg = MACIO_IN32(KEYLARGO_FCR4);
Attachment:
signature.asc
Description: Digital signature