On Tue, Jan 20, 2009 at 5:32 AM, Carl-Daniel Hailfinger c-d.hailfinger.devel.2006@gmx.net wrote:
Marc,
I believe you are the person who is most proficient in K8 RAM init (no offense to anyone else intended). Could you take a look at the patch and give me a few comments?
I think that the code flow is equivalent. Maybe add a comment about the return values for the functions. The case statement could have a default to catch an unexpected return. Maybe make the hw_error the default?
if (latencies == 0) {
That should probably cause a disable dimm.
I think that spd_set_memclk() has the same CAS checking so you could do the fix their as well. Search for SPD_CAS_LAT.
Marc