Hi Rob,

Rob McMahon píše v po 12. 07. 2010 v 10:05 +0100:
> On 10/07/2010 15:35, Milan Jurik wrote:
> > > I managed to install OSOL build 134 on a Toshiba Tecra M11:
> > >  - using Vesa mode
> > > 
> > > Im seeing a lot of messages like:
> > >  Unknown domain coordination type: 4 ... !? 
> > > 
> > This message?
> > 
> > http://src.opensolaris.org/source/xref/onnv/onnv-gate/usr/src/uts/i86pc/os/cpupm/cpupm_mach.c#661
> Hmm, ...
> 616   cpupm_state_domains_t   *state_domain;
>     617       cpuset_t                set;
>     618 ...
>     624 
>     625       switch (state) {
>     626       case CPUPM_P_STATES:
>     627               state_ops = mach_state->ms_pstate.cma_ops;
>     628               state_domain = mach_state->ms_pstate.cma_domain;
>     629               break;
>     630       case CPUPM_T_STATES:
>     631               state_ops = mach_state->ms_tstate.cma_ops;
>     632               state_domain = mach_state->ms_tstate.cma_domain;
>     633               break;
>     634       default:
>     635               break;
>     636       }
>     637 
>     638       switch (state_domain->pm_type) {
> That can't be good.  If state is not CPUPM_P_STATES or CPUPM_T_STATES,
> don't bother setting state_domain and then immediately dereference
> some garbage from the stack.  Surely there should be at least an
> cmn_err and return there, if not an assert ... I'm sure the function
> is always called with state being one of the two, but still.
> 

Good point. Yes, it is called only with one of these STATES:

http://src.opensolaris.org/source/search?q=cpupm_state_change&defs=&refs=&path=&hist=&project=%2Fonnv

But at least ASSERT() would be nice there. Feel free to submit bug to
bugs.opensolaris.org

Best regards,

Milan

> Sorry, reading it just raised the hairs on the back of my neck.
> 
> Rob


_______________________________________________
pm-discuss mailing list
[email protected]
http://mail.opensolaris.org/mailman/listinfo/pm-discuss

Reply via email to