Re: [PATCH v2 3/4] arm: add ISAR2, MMFR0 and MMFR1 fields in cpufeature

[Stefano Stabellini <sstabellini@xxxxxxxxxx>]

On Mon, 6 Jun 2022, Bertrand Marquis wrote:
> Hi Stefano,
> 
> > On 3 Jun 2022, at 02:45, Stefano Stabellini <sstabellini@xxxxxxxxxx> wrote:
> > 
> > On Tue, 31 May 2022, Bertrand Marquis wrote:
> >> Complete AA64ISAR2 and AA64MMFR[0-1] with more fields.
> >> While there add a comment for MMFR bitfields as for other registers in
> >> the cpuinfo structure definition.
> >> 
> >> Signed-off-by: Bertrand Marquis <bertrand.marquis@xxxxxxx>
> >> ---
> >> Changes in v2:
> >> - patch introduced to isolate changes in cpufeature.h
> >> - complete MMFR0 and ISAR2 to sync with sysregs.h status
> >> ---
> >> xen/arch/arm/include/asm/cpufeature.h | 28 ++++++++++++++++++++++-----
> >> 1 file changed, 23 insertions(+), 5 deletions(-)
> >> 
> >> diff --git a/xen/arch/arm/include/asm/cpufeature.h 
> >> b/xen/arch/arm/include/asm/cpufeature.h
> >> index 9649a7afee..57eb6773d3 100644
> >> --- a/xen/arch/arm/include/asm/cpufeature.h
> >> +++ b/xen/arch/arm/include/asm/cpufeature.h
> >> @@ -234,6 +234,7 @@ struct cpuinfo_arm {
> >> union {
> >> register_t bits[3];
> >> struct {
> >> + /* MMFR0 */
> >> unsigned long pa_range:4;
> >> unsigned long asid_bits:4;
> >> unsigned long bigend:4;
> >> @@ -242,18 +243,31 @@ struct cpuinfo_arm {
> >> unsigned long tgranule_16K:4;
> >> unsigned long tgranule_64K:4;
> >> unsigned long tgranule_4K:4;
> >> - unsigned long __res0:32;
> >> -
> >> + unsigned long tgranule_16k_2:4;
> >> + unsigned long tgranule_64k_2:4;
> >> + unsigned long tgranule_4k:4;
> > 
> > Should be tgranule_4k_2:4
> 
> Right I will fix that.
> 
> > 
> > 
> >> + unsigned long exs:4;
> >> + unsigned long __res0:8;
> >> + unsigned long fgt:4;
> >> + unsigned long ecv:4;
> >> +
> >> + /* MMFR1 */
> >> unsigned long hafdbs:4;
> >> unsigned long vmid_bits:4;
> >> unsigned long vh:4;
> >> unsigned long hpds:4;
> >> unsigned long lo:4;
> >> unsigned long pan:4;
> >> - unsigned long __res1:8;
> >> - unsigned long __res2:28;
> >> + unsigned long specsei:4;
> >> + unsigned long xnx:4;
> >> + unsigned long twed:4;
> >> + unsigned long ets:4;
> >> + unsigned long __res1:4;
> > 
> > hcx?
> > 
> > 
> >> + unsigned long afp:4;
> >> + unsigned long __res2:12;
> > 
> > ntlbpa
> > tidcp1
> > cmow
> > 
> >> unsigned long ecbhb:4;
> > 
> > Strangely enough I am looking at DDI0487H and ecbhb is not there
> > (D13.2.65). Am I looking at the wrong location?
> 
> Right now I declared here only the values which have a corresponding
> declaration in sysregs.h
> If I add more fields here we will not be in sync with it anymore.
> 
> And on ecbhb it will be in the next revision of the manual yes.
> 
> 
> > 
> > 
> >> + /* MMFR2 */
> >> unsigned long __res3:64;
> >> };
> >> } mm64;
> >> @@ -297,7 +311,11 @@ struct cpuinfo_arm {
> >> unsigned long __res2:8;
> >> 
> >> /* ISAR2 */
> >> - unsigned long __res3:28;
> >> + unsigned long wfxt:4;
> >> + unsigned long rpres:4;
> >> + unsigned long gpa3:4;
> >> + unsigned long apa3:4;
> >> + unsigned long __res3:12;
> > 
> > mops
> > bc
> > pac_frac
> > 
> > 
> >> unsigned long clearbhb:4;
> > 
> > And again this is not described at D13.2.63. Probably the bhb stuff
> > didn't make it into the ARM ARM yet.
> 
> As said before, are you ok with only adding stuff declared in sysregs
> to make it simpler to sync with Linux ?

Yes, that makes sense. In that case just fix tgranule_4k_2 and you can
add my

Reviewed-by: Stefano Stabellini <sstabellini@xxxxxxxxxx>