> > Well, I don't see how this would help; after the serial is
> > disabled via MSR 0x119 magic, it can't be re-enabled until you reboot..
>
> I say remove the code which is disabling it.
> The kernel should not set the policy.
> If a user wants to disable it, they should use a userspace tool
> to interface the MSR. (But an MSR driver is needed first)
>
That would probably be the best thing.. Everyone with a pIII would have
to be notified of the change though, and some people might not want to run
a program 'just to disable the serial'.. and btw, you can simply write a
short program with iopl(3) (or you should be able to) that can write to
the MSRs..
> > Ploping the data to a file in /proc owned by root with 0400
> > perms would probably be the best thing if it needs to be
> > accessed (someone _has_ to have a use for it), but it has
> > to be read and saved before it's disabled..
>
> Yes, of course writing the MSRs should be a root-only thing.
>
> > As a side note, most bioses allow it to be disabled at boot
> > time (mine included), and I would rather linux left the bit
> > twiddling to me.. but I guess I can just go comment out the
> > relevant lines :)
>
> My point exactly.
>
> > Also, reading MSRs works fine (as a normal user) with the rdmsr
> > instruction; I haven't tried writing them though (I would
> > hope that isn't allowed in userland)
>
> You shouldn't be able to read them either, as AFAIR, it's a ring 0
> instruction. You should segfault immediately.
I might have been mistaken there (looks like I was getting if the serial
was enabled or not via eax == 1; CPUID then testing for bit 18 in edx,
I keep forgetting to read my late night messages twice ;)
>
> regards,
>
> Dave.
-- Zinx Verituse (finger @bliss.penguinpowered.com for pgp/gpg keys)(new jul10/99) pgp9FE5C9747EB8FF329BB13199C4008E67/gpg574673A12184A27A9EC0EDCCE132BCEF921B1558 0"2-1=0>0:1(2<192:0?0;0A0@2=0<0=1.0A2=0<2A0-">:#v_52*,@ 55*-3*\68*-+, v >
- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu Please read the FAQ at http://www.tux.org/lkml/