Attached patch moves support for Deschutes Slot 1 CPUs (model_65x) into its own home.
This is similar to the Katmai patch I submitted before, but without microcode this time. With 17 of them they will make the patch too large. I think I'll defer them for later or someone else*, hehe. :D
abuild-tested. I have no Deschutes CPUs to boot test this with.
Signed-off-by: Keith Hui buurin@gmail.com
* also because I like to keep all microcodes for the same family within one file instead of one microcode per file.
Keith Hui wrote:
Attached patch moves support for Deschutes Slot 1 CPUs (model_65x) into its own home.
Then I guess the patch should also delete some lines in existing files?
This is similar to the Katmai patch I submitted before, but without microcode this time. With 17 of them they will make the patch too large. I think I'll defer them for later or someone else*, hehe. :D
abuild-tested.
It doesn't look like the new code is used anywhere? Should it be?
- also because I like to keep all microcodes for the same family
within one file instead of one microcode per file.
We want microcode updates to be as close as possible to upstream format. Eventually I think we want to put them verbatim into cbfs.
//Peter
On 10/15/10 9:04 PM, Peter Stuge wrote:
- also because I like to keep all microcodes for the same family
within one file instead of one microcode per file.
We want microcode updates to be as close as possible to upstream format. Eventually I think we want to put them verbatim into cbfs.
Yes. Please use the microcode extraction script in our tree to put them into the right places.
On Fri, Oct 15, 2010 at 11:25:02PM -0400, Keith Hui wrote:
Attached patch moves support for Deschutes Slot 1 CPUs (model_65x) into its own home.
Thanks, r5954.
This is similar to the Katmai patch I submitted before, but without microcode this time. With 17 of them they will make the patch too large. I think I'll defer them for later or someone else*, hehe. :D
[...]
- also because I like to keep all microcodes for the same family
within one file instead of one microcode per file.
As Stefan mentioned, we should always use the script from src/cpu/intel/microcode/update-microcodes.sh to create/update the microcode files.
We're leaving model_6xx untouched for now, as it is still being used at the moment and duplicate CPUs in model_6xx and model_65x for example should not matter I think. When the splitting is done, we'll probably get rid of model_6xx entirely and have slot1/slot2/socket370 etc. include only the proper families they need.
Uwe.