Patchwork Kconfig patches

login
register
about
Submitter Myles Watson
Date 2009-09-22 16:51:30
Message ID <2831fecf0909220951i40c96683na8e2f699bd82d47b@mail.gmail.com>
Download mbox | patch
Permalink /patch/275/
State Accepted
Headers show

Comments

Myles Watson - 2009-09-22 16:51:30
I'm still trying to update my Tyan boards so they don't get pruned.  They're
pretty close, there is still an issue with ACPI that I'm working on, but I'd
like to start shrinking my patch stack.

tables.diff: Add Kconfig dialogues for ACPI, MP_TABLE, ...
Kconfig_bools.diff: Change some more ints to bools, change some default
values.
xip_size.diff: Make XIP_SIZE + XIP_BASE add up to 4GB so that the end of the
ROM gets cached (speeds up booting immensely.)
failoverR.diff: Revert my failover change from before since Kconfig only
supports fallback.
kconfig_s2892.dif: Add support for Tyan s2891, s2892, and s2895 to Kconfig.
smp.diff: set CONFIG_SMP based on MAX_CPUS.  I think maybe CONFIG_SMP should
just die and test for CONFIG_MAX_CPUS.

Signed-off-by: Myles Watson <mylesgw@gmail.com>

Thanks,
Myles
ron minnich - 2009-09-22 17:36:29
All are
Acked-by: Ronald G. Minnich <rminnich@gmail.com>

I really like the tables cleanup too.

We do need to get some of those k8 options out of mainboards, but
that's a future fix. These patches are a big improvement.

ron
Cristi Magherusan - 2009-09-22 18:40:29
On Tue, 2009-09-22 at 10:36 -0700, ron minnich wrote:
> All are
> Acked-by: Ronald G. Minnich <rminnich@gmail.com>
> 
> I really like the tables cleanup too.
> 
> We do need to get some of those k8 options out of mainboards, but
> that's a future fix. These patches are a big improvement.
> 
> ron
> 
Great work! 

Kconfig_bools.diff defines LOGICAL_CPUS as bool, shouldn't it be left as
int instead?

Regards,
Cristi
Myles Watson - 2009-09-22 18:45:07
>
> Great work!
>
> Kconfig_bools.diff defines LOGICAL_CPUS as bool, shouldn't it be left as
> int instead?


It is used as either true or false.  The way to tell how many logical CPUs
there might be is dividing CONFIG_MAX_CPUS and CONFIG_MAX_PHYSICAL_CPUS.

from  src/northbridge/amd/amdk8/.svn/text-base/northbridge.c.svn-base:
    disable_siblings = !CONFIG_LOGICAL_CPUS;

Thanks for the review,
Myles
Myles Watson - 2009-09-22 18:56:27
On Tue, Sep 22, 2009 at 11:36 AM, ron minnich <rminnich@gmail.com> wrote:

> All are
> Acked-by: Ronald G. Minnich <rminnich@gmail.com>
>
> I really like the tables cleanup too.
>
Thanks.


> We do need to get some of those k8 options out of mainboards, but
> that's a future fix.
>
I'm not sure where they should go.  I guess we'd have to go through them one
by one and suggest places.

General Kconfig patches committed in Rev 4655.  Tyan-specific patches in Rev
4656.

Thanks,
Myles

Patch

Index: svn/src/cpu/Kconfig
===================================================================
--- svn.orig/src/cpu/Kconfig
+++ svn/src/cpu/Kconfig
@@ -19,7 +19,7 @@  config DCACHE_RAM_SIZE
 
 config SMP
 	bool
-	default n
+	default y if MAX_CPUS != 1
 
 config CPU_SOCKET_TYPE
 	hex