Uwe Hermann wrote:
On Tue, Sep 29, 2009 at 01:30:04AM +0200, Stefan Reinauer wrote:
Patrick Georgi wrote:
Index: src/southbridge/via/Makefile.inc
--- src/southbridge/via/Makefile.inc (Revision 4672) +++ src/southbridge/via/Makefile.inc (Arbeitskopie) @@ -1,5 +1,4 @@ -#subdirs-y += k8t890 -#subdirs-y += vt8231 -#subdirs-y += vt8235 +subdirs-$(CONFIG_SOUTHBRIDGE_VIA_K8T890) += k8t890 +subdirs-$(CONFIG_SOUTHBRIDGE_VIA_VT8231) += vt8231 +subdirs-$(CONFIG_SOUTHBRIDGE_VIA_VT8235) += vt8235 subdirs-$(CONFIG_SOUTHBRIDGE_VIA_VT8237R) += vt8237r
-#subdirs-y += vt82c686
This will still be required later I guess (?).
vt82c686 is a weird stub of southbridge support: I found no use of it, and it seems to handle serial enable, only. My proposal? Unless it will actually be used, get rid of it.
Acked-by: Stefan Reinauer stepan@coresystems.de
Please don't drop this. I have some boards with this southbridge, and the code that's there looks OK (though unfinished, of course). I'll put this on my TODO list and improve the SB code one of these days.
I think we should not keep around broken components like that. An early serial init like that is trivial and not more than half an hour of effort. On the other side, the existence of the component in the tree will create a wrong impression about what coreboot can do or can not do.
This code is not unfinished, as it's not being worked on.. It's in svn history in case we'll some day have the time to support this.
Same reasoning as goes for the ppc port...
Stefan