[mythtv-users] [SPAM?]: Re: questions about suspend with ACPI

[Leighton Brough]

> Wiki-ing the above certanly would be nice. I also had some problems with
> suspend resume functionality which took me ages. I still have problems with
> nVidia and X. Right now I'm quiting X before suspend. I wasn't aware of the
> nVidia driver hack and will check this out later this week.
> Thanks.
>
>   
The hack is my own, inspired by a patch I found via some Googling late 
one night. You need to make a few trivial changes to the source file 
nv.c. Then reinstall.
Here's a diff of the source for the 7676 drivers. The changes are 
similar for different versions, although the line numbers will differ:

225c225
< int nv_kern_acpi_standby(struct pci_dev *, u32);
---
 > int nv_kern_acpi_standby(struct pci_dev *, pm_message_t);
3627c3627
<         case PM_SUSPEND_MEM:
---
 >         case PM_EVENT_SUSPEND:
3632c3632
<         case PM_SUSPEND_ON:
---
 >         case PM_EVENT_ON:
3652c3652
<     u32 state
---
 >     pm_message_t state
3655c3655
<     return nv_acpi_event(dev, state);
---
 >     return nv_acpi_event(dev, state.event);
3664c3664
<     return nv_acpi_event(dev, PM_SUSPEND_ON);
---
 >     return nv_acpi_event(dev, PM_EVENT_ON);

I hope it's clear from the above that you need to modify the 
PM_SUSPEND... identifiers to be PM_EVENT... and change the type of the 
nv_acpi_event function.

These changes mean that the nvidia drivers correctly handle the ACPI 
events for S3, instead of throwing an error and preventing suspend.