[LinuxBIOS] r2670 - in trunk/LinuxBIOSv2/src/mainboard: asus/p2b bitworks/ims densitron/dpx114 via/epia via/epia-m

Corey Osgood corey_osgood at verizon.net
Wed May 16 05:01:48 CEST 2007


Uwe Hermann wrote:
> On Tue, May 15, 2007 at 03:58:40PM -0400, Corey Osgood wrote:
>> Hmm, why don't we just fix raminit.c and debug.c (and northbridge.c, if
>> applicable) for those northbridges (vt8601, vt8623, 440bx) to just use
>> smbus_read_byte instead of spd_read_byte? Eliminate any chance of a
>> problem altogether?
> 
> spd_read_byte() is the correct name, so it should stay IMO. Is there a
> reason why all/most boards define this as a wrapper for smbus_read_byte()?
> 
> Could this be something else than smbus_read_byte() for some board?
> If no, maybe a #define is easier? Or move the wrapper to a central
> location?

okay, I guess this is the real question: is there ever any case where or
reason why spd_read_byte doesn't just return smbus_read_byte? And if
not, then why should we use a wrapper if it isn't necessary? I mean, it
would be even easier just to change the south bridges from
smbus_read_byte to spd_read_byte. This wrapper makes the code less
readable, and evidently causes problems, so I can't understand why
having it is a good thing, unless there's some other driving need for it.

-Corey




More information about the coreboot mailing list