Re: [Qemu-ppc] [RFC for-2.12 8/8] spapr: Handle Decimal Floating Point (DFP) as an optional capability

[David Gibson]

On Tue, Dec 12, 2017 at 01:58:48PM +0100, Greg Kurz wrote:
> On Mon, 11 Dec 2017 18:08:08 +1100
> David Gibson <address@hidden> wrote:
> 
> > Decimal Floating Point has been available on POWER7 and later (server)
> > cpus.  However, it can be disabled on the hypervisor, meaning that it's
> > not available to guests.
> > 
> > We currently handle this by conditionally advertising DFP support in the
> > device tree depending on whether the guest CPU model supports it - which
> > can also depend on what's allowed in the host for -cpu host.  That can lead
> > to confusion on migration, since host properties are silently affecting
> > guest visible properties.
> > 
> > This patch handles it by treating it as an optional capability for the
> > pseries machine type.
> > 
> > Signed-off-by: David Gibson <address@hidden>
> > ---
> >  hw/ppc/spapr.c         |  5 +++--
> >  hw/ppc/spapr_caps.c    | 17 +++++++++++++++++
> >  include/hw/ppc/spapr.h |  3 +++
> >  3 files changed, 23 insertions(+), 2 deletions(-)
> > 
> > diff --git a/hw/ppc/spapr.c b/hw/ppc/spapr.c
> > index 6b37511e8f..cca3789b35 100644
> > --- a/hw/ppc/spapr.c
> > +++ b/hw/ppc/spapr.c
> > @@ -571,7 +571,7 @@ static void spapr_populate_cpu_dt(CPUState *cs, void 
> > *fdt, int offset,
> >      /* Advertise DFP (Decimal Floating Point) if available
> >       *   0 / no property == no DFP
> >       *   1               == DFP available */
> > -    if (env->insns_flags2 & PPC2_DFP) {
> > +    if (spapr_has_cap(spapr, SPAPR_CAP_DFP)) {
> >          _FDT((fdt_setprop_cell(fdt, offset, "ibm,dfp", 1)));
> >      }
> >  
> > @@ -3831,7 +3831,8 @@ static void spapr_machine_class_init(ObjectClass *oc, 
> > void *data)
> >       */
> >      mc->numa_mem_align_shift = 28;
> >  
> > -    smc->default_caps = spapr_caps(SPAPR_CAP_HTM | SPAPR_CAP_VSX);
> > +    smc->default_caps = spapr_caps(SPAPR_CAP_HTM | SPAPR_CAP_VSX
> > +                                   | SPAPR_CAP_DFP);
> >      spapr_capability_properties(smc, &error_abort);
> >  }
> >  
> > diff --git a/hw/ppc/spapr_caps.c b/hw/ppc/spapr_caps.c
> > index b19a9f4417..49c2e10d27 100644
> > --- a/hw/ppc/spapr_caps.c
> > +++ b/hw/ppc/spapr_caps.c
> > @@ -49,6 +49,11 @@ static sPAPRCapabilityInfo capability_table[] = {
> >          .description = "Allow Vector Scalar Extensions (VSX)",
> >          .bit = SPAPR_CAP_VSX,
> >      },
> > +    {
> > +        .name = "dfp",
> > +        .description = "Allow Decimal Floating Point (DFP)",
> > +        .bit = SPAPR_CAP_DFP,
> > +    },
> >  };
> >  
> >  static sPAPRCapabilities default_caps_with_cpu(sPAPRMachineState *spapr, 
> > CPUState *cs)
> > @@ -67,6 +72,7 @@ static sPAPRCapabilities 
> > default_caps_with_cpu(sPAPRMachineState *spapr, CPUStat
> >      if (!ppc_check_compat(cpu, CPU_POWERPC_LOGICAL_2_06,
> >                            0, spapr->max_compat_pvr)) {
> >          caps.mask &= ~SPAPR_CAP_VSX;
> > +        caps.mask &= ~SPAPR_CAP_DFP;
> >      }
> >  
> >      return caps;
> > @@ -184,6 +190,13 @@ static void spapr_allow_caps(sPAPRMachineState *spapr, 
> > Error **errp)
> >          }
> >      }
> >  
> > +    if (spapr_has_cap(spapr, SPAPR_CAP_DFP)) {
> > +        if (!(env->insns_flags2 & PPC2_DFP)) {
> > +            error_setg(&err, "DFP support not available, try cap-dfp=off");
> > +            goto out;
> > +        }
> > +    }
> > +
> >  out:
> >      if (err) {
> >          error_propagate(errp, err);
> > @@ -199,6 +212,10 @@ static void spapr_enforce_caps(sPAPRMachineState 
> > *spapr, Error **errp)
> >      if (!spapr_has_cap(spapr, SPAPR_CAP_VSX)) {
> >          /* TODO: Any way to disable VSX? */
> >      }
> > +
> > +    if (!spapr_has_cap(spapr, SPAPR_CAP_VSX)) {
> 
> s/SPAPR_CAP_VSX/SPAPR_CAP_DFP/

Oops, fixed.

> 
> With that fixed:
> 
> Reviewed-by: Greg Kurz <address@hidden>
> 
> > +        /* TODO: Any way to disable DFP? */
> > +    }
> >  }
> >  
> >  void spapr_caps_reset(sPAPRMachineState *spapr)
> > diff --git a/include/hw/ppc/spapr.h b/include/hw/ppc/spapr.h
> > index 4c173cb40c..ef4f7427de 100644
> > --- a/include/hw/ppc/spapr.h
> > +++ b/include/hw/ppc/spapr.h
> > @@ -62,6 +62,9 @@ typedef enum {
> >  /* Vector Scalar Extensions */
> >  #define SPAPR_CAP_VSX               0x0000000000000002ULL
> >  
> > +/* Decimal Floating Point */
> > +#define SPAPR_CAP_DFP               0x0000000000000004ULL
> > +
> >  typedef struct sPAPRCapabilities sPAPRCapabilities;
> >  struct sPAPRCapabilities {
> >      uint64_t mask;
> 

-- 
David Gibson                    | I'll have my music baroque, and my code
david AT gibson.dropbear.id.au  | minimalist, thank you.  NOT _the_ _other_
                                | _way_ _around_!
http://www.ozlabs.org/~dgibson

signature.asc
Description: PGP signature