[PlanetCCRMA] Fw: Kernel 2.4.20-4.ll.acpi(smp) panic!

[Rui Nuno Capela]

Glad to hear from you, nando,

>> And the guilty one is: ... High Memory support (4GB).
>
> Wow, good detective work!

Tks. I think this problem relies soleley on this machine of mine, not on
planetccrma's, redhat's or even suse's (where this kernel problem was also
replicated). Don't even know what this 4GB stuff is all about, since my
machines (I have 5, of which 3 are dedicated to my home studio) only allow
a maximum of 3GB physical SDRAM each.

>> At last, nando, can you tell what precise sources and/or patches have
>> you based your kernel builds? I would appreciate very much if you can.
>
> Yes, of course, it is all available in the source rpm for the kernel
> (which is not the same as the binary rpm for the kernel sources - I
> know, very confusing). As with any other rpm, the kernel is rebuilt from
> a source rpm, in this case kernel-2.4.20-4.ll.acpi.src.rpm. It is
> available on the web site, go to the kernel page and follow the link
> that's there.
>
> 2.4.20-4.ll.acpi is (more or less)
> 2.4.20 + 2.4.21-pre6 + lowlat + preempt + capabilities + variable HZ

What is this capabilities thingie? Are we talking about the same when it
gets to change include/linux/capability.h header file? I mean forcing the
following two definitions?

#define CAP_INIT_EFF_SET    to_cap_t(~0)
#define CAP_INIT_INH_SET    to_cap_t(~0)

This stuff is indeed recomended when compiling MusE's, as suggested on
site (http://muse.seh.de) on Tommi Ilmonenīs behalf. BTW, MusE 0.6.0pre8
has been announced -- gotta get back to my SuSE to build it too, sometime
:)

Best regards,

-- 
rncbc aka Rui Nuno Capela
rncbc@rncbc.org

P.S. Notice I've changed email address from M$N hotmail. This is my
official one, for any further exchange....