Re: [PATCH v11 3/5] x86emul: Add return code information to error messages

[Petre Ovidiu PIRCALABU <ppircalabu@bitdefender.com>]

On Ma, 2017-09-19 at 09:22 -0600, Jan Beulich wrote:
> > 
> > > 
> > > > 
> > > > On 12.09.17 at 16:32, <ppircalabu@bitdefender.com> wrote:
> > --- a/xen/arch/x86/hvm/emulate.c
> > +++ b/xen/arch/x86/hvm/emulate.c
> > @@ -2055,7 +2055,7 @@ int hvm_emulate_one_mmio(unsigned long mfn,
> > unsigned long gla)
> >      {
> >      case X86EMUL_UNHANDLEABLE:
> >      case X86EMUL_UNIMPLEMENTED:
> > -        hvm_dump_emulation_state(XENLOG_G_WARNING, "MMCFG",
> > &ctxt);
> > +        hvm_dump_emulation_state(XENLOG_G_WARNING, "MMCFG", &ctxt,
> > rc);
> >          break;
> At the example of this one I think it is pretty clear that the order
> of patches would be the other way around. But I won't insist.
> 
> > 
> > @@ -2242,16 +2242,17 @@ static const char
> > *guest_x86_mode_to_str(int mode)
> >  }
> >  
> >  void hvm_dump_emulation_state(const char *loglvl, const char
> > *prefix,
> > -                              struct hvm_emulate_ctxt
> > *hvmemul_ctxt)
> > +                              struct hvm_emulate_ctxt
> > *hvmemul_ctxt, int rc)
> >  {
> >      struct vcpu *curr = current;
> >      const char *mode_str =
> > guest_x86_mode_to_str(hvm_guest_x86_mode(curr));
> >      const struct segment_register *cs =
> >          hvmemul_get_seg_reg(x86_seg_cs, hvmemul_ctxt);
> >  
> > -    printk("%s%s emulation failed: %pv %s @ %04x:%08lx -> %*ph\n",
> > -           loglvl, prefix, curr, mode_str, cs->sel, hvmemul_ctxt-
> > >insn_buf_eip,
> > -           hvmemul_ctxt->insn_buf_bytes, hvmemul_ctxt->insn_buf);
> > +    printk("%s%s emulation failed (rc=%d): %pv %s @ %04x:%08lx ->
> > %*ph\n",
> Please try to keep log messages short (but without losing relevant
> information). In the case here the "rc=" is unnecessary. With it
> dropped
I added the "rc=" to mark the distinction between "unimplemented" and
"unhandleable", as requested by Andrew Cooper for v10 

"Please modify hvm_dump_emulation_state to pass rc in, and distinguish
UNHANDLEABLE vs UNIMPLEMENTED in the printed message."


> Reviewed-by: Jan Beulich <jbeulich@suse.com>
> 
> Jan
> 
> 
> ________________________
> This email was scanned by Bitdefender
_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
https://lists.xen.org/xen-devel