Re: [Xen-devel] [PATCH v2 13/16] x86: rearrange x86_64/entry.S

[Wei Liu <wei.liu2@xxxxxxxxxx>]

On Mon, Oct 29, 2018 at 08:57:19AM -0600, Jan Beulich wrote:
> >>> On 19.10.18 at 16:28, <wei.liu2@xxxxxxxxxx> wrote:
> > @@ -845,19 +865,7 @@ handle_ist_exception:
> >          /* We want to get straight to the IRET on the NMI exit path. */
> >          testb $3,UREGS_cs(%rsp)
> >          jz    restore_all_xen
> > -        GET_CURRENT(bx)
> > -        /* Send an IPI to ourselves to cover for the lack of event 
> > checking. */
> > -        movl  VCPU_processor(%rbx),%eax
> > -        shll  $IRQSTAT_shift,%eax
> > -        leaq  irq_stat+IRQSTAT_softirq_pending(%rip),%rcx
> > -        cmpl  $0,(%rcx,%rax,1)
> > -        je    1f
> > -        movl  $EVENT_CHECK_VECTOR,%edi
> > -        call  send_IPI_self
> > -1:      movq  VCPU_domain(%rbx),%rax
> > -        cmpb  $0,DOMAIN_is_32bit_pv(%rax)
> > -        je    restore_all_guest
> > -        jmp   compat_restore_all_guest
> > +        jmp   self_ipi_restore_all_guest
> 
> I'm having some trouble understanding why you move this code:
> Without CONFIG_PV it is unreachable.

I thought this snippet was self-contained enough to move to the first
half of the file. But ...

> I'd prefer if it stayed in
> place and simply got an #ifdef around it.

... I'm fine with this too.

Wei.

> The one alternative
> option I could see is to move the restore_all_xen code block
> right here, so the JMP you put in could become JNZ.
> 
> Jan
> 
> 

_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxxxxxxxxx
https://lists.xenproject.org/mailman/listinfo/xen-devel