Re: [PATCH v4 19/27] s390: Introduce Start Arm Execution instruction

[Steffen Eiden]

On Mon, Jul 06, 2026 at 01:30:20PM +0200, Christian Borntraeger wrote:
> 
> 
> Am 06.07.26 um 13:17 schrieb Janosch Frank:
> 
> > > +SYM_FUNC_START(__sae64a)
> > > +    stmg    %r6,%r14,__SF_GPRS(%r15)    # store kernel registers
> > > +    STBEAR    __SF_SAE_BEAR(%r15)        # save breaking event address register
> > > +    .insn    rre,0xb9a50000,%r2,0        # Start Arm Execution
> > > +# Let the next instruction be NOP to avoid triggering a machine check
> > > +# and handling it in a guest as result of the instruction execution.
> > 
> > @Christian:
> > I have the feeling things got lost in translation when this comment was written.
> > How would we trigger a MCHECK by having an instruction after SIE/SAE?
> > AFAIU this allows us to have a label which we can check in the mcheck handler to decide who triggered the mcheck.
> 
> the original rationale was that some machine checks are nullifying some are terminating. So the PSW might point
> to SAE or the instruction afterwards for a guest machine check.
> In SIE, we use that nop to account any machine check on that nop as a guest machine check.
> 
> > 
> > Without the label the comment and NOP are useless as far as I can see.
> > So we need to decide if this still applies and then either remove it or add sae handling to mcck_int_handler and maybe other mcheck functions.
> 
> We plan to get more information on the machine checks for sae so we might no longer need that.
> But this series does not yet contain the machine check handling. Maybe its better to remove comment and
> nop and introduce that later when we add machine check handling and if still needed.

Ok I'll remove the comment and the nop for the next round.