[coreboot] cleanup

ron minnich rminnich at gmail.com
Wed Oct 29 05:20:35 CET 2008


On Tue, Oct 28, 2008 at 5:55 PM, Uwe Hermann <uwe at hermann-uwe.de> wrote:

> Can you elaborate? What was 'initialized' used for specifically and why is
> it not needed anymore?

in v2 the init function was called several times. That is not true in
v3. It only gets called once. Hence we do not need to know that we
called it once.



>
>
>> Signed-off-by: Ronald G. Minnich <rminnich at gmail.com>
>>
>>
>> Index: southbridge/intel/i82371eb/i82371eb.c
>> ===================================================================
>> --- southbridge/intel/i82371eb/i82371eb.c     (revision 954)
>> +++ southbridge/intel/i82371eb/i82371eb.c     (working copy)
>> @@ -43,6 +43,7 @@
>>  {
>>       unsigned short c;
>>
>> +     /* These should be controlled in the dts. */
>
> Btw, don't waste too much time on 82371EB in v3, I'll move the v2 version
> (which is a lot more complete) to v3 soonish. That'll work for the
> 440BX (which will be moved to v3 sooner or later too), but also for QEMU
> of course.

great to hear.




>> --- device/device.c   (revision 954)
>> +++ device/device.c   (working copy)
>> @@ -805,6 +805,7 @@
>>  #warning do we call phase3_enable here.
>>               new_max = busdevice->ops->phase3_scan(busdevice, max);
>>               do_phase3 = 0;
>> +             /* do we *ever* use this path */
>
> Is this a question or observation?

a question. We believe this path is never used.
               dev->ops && dev->ops->phase6_init) {

>> +                     memmove(resource, resource + 1, (dev->resources-i)* sizeof(*resource));
>
> Specific bug or is this "just in case"?

peter pointed out that is is most likely a bug.

Thanks

ron




More information about the coreboot mailing list