Hi Daniel..
On Wed, Dec 22, 2010 at 08:40, Daniel (Youngwhan) Song
wrote:
> As we've talked, it looks the code would be ok, but I guess architecture
> doesn't handle the 16 bit address correctly. Our team will find a way to get
> through this by modifying the PCI module.
OK, got it...just keep u
> As we've talked, it looks the code would be ok, but I guess architecture
> doesn't handle the 16 bit address correctly. Our team will find a way to get
> through this by modifying the PCI module.
Which is the platform/arch you are seeing this problem as opposed to
intel(Where the code works?).
Re
Hi Mulyadi,
On Tue, Dec 21, 2010 at 9:59 AM, Mulyadi Santosa
wrote:
> Hi Daniel.
>
> On Tue, Dec 21, 2010 at 16:04, YOUNGWHAN SONG wrote:
> > test_foo = (struct foo*)regsva;
>
> I was thinking differently, could it be that casting...as shown
> above...introduce this behaviour? what if yo
Hi Daniel.
On Tue, Dec 21, 2010 at 16:04, YOUNGWHAN SONG wrote:
> test_foo = (struct foo*)regsva;
I was thinking differently, could it be that casting...as shown
above...introduce this behaviour? what if you just point to the
address...directly without any "container" such as struct foo?