Re: [PATCH V2] UefiCpuPkg: Keep library class header file definition independent

["Song, BinX" <binx.song@intel.com>]

Hi Mike,

Thanks for your suggestion.
After discussion with Eric, I know there is a function named IsCpuFeatureSupported() to do the feature valid/invalid check.
User should do IsCpuFeatureSupported() check first and then RegisterCpuFeature().
So there is no need to add any code to do the same work, I will roll back the patch which has been checked in before.

Best Regards,
Bell Song


> -----Original Message-----
> From: Kinney, Michael D
> Sent: Tuesday, December 19, 2017 7:21 AM
> To: Song, BinX <binx.song@intel.com>; edk2-devel@lists.01.org; Kinney,
> Michael D <michael.d.kinney@intel.com>
> Cc: Dong, Eric <eric.dong@intel.com>; lersek@redhat.com; Ni, Ruiyu
> <ruiyu.ni@intel.com>
> Subject: RE: [PATCH V2] UefiCpuPkg: Keep library class header file definition
> independent
> 
> This still does not work because a library instance that
> registers a new CPU feature may want to use a new feature
> bit larger than MAX.  This change moves the define
> from the RegisterCpuFeaturesLib library class to the
> implementation of the RegisterCpuFeaturesLibs.
> 
> The components that know the maximum bit number are the
> NULL lib instances that register CPU features (e.g.
> CpuCommonFeaturesLib) and the platform developer that
> that provides the set of NULL lib instances in a platform
> DSC file and the associated PCDs.  For example, the
> following DSC fragment only specifies a single NULL lib
> instance that registers CPU features.  However, it is
> legal to provide additional NULL lib instances from
> Si packages.
> 
>   UefiCpuPkg/CpuFeatures/CpuFeaturesDxe.inf {
>     <LibraryClasses>
> 
> NULL|UefiCpuPkg/Library/CpuCommonFeaturesLib/CpuCommonFeaturesLib
> .inf
>   }
> 
> The maximum size of the bitmask for CPU features is known
> by the size of the PCD called PcdCpuFeaturesSupport that
> has a DEC default value based on CpuCommonFeaturesLib but
> can be set to a larger size based on the maximum bit used
> by all the NULL lib instances.
> 
> So maybe the valid check should just verify that the bit
> number passed in is < PcdGetSize (PcdCpuFeaturesSupport) * 8.
> This eliminates the use of the #define value and uses an
> existing PCD.
> 
> Thanks,
> 
> Mike
> 
> > -----Original Message-----
> > From: Song, BinX
> > Sent: Sunday, December 17, 2017 9:37 PM
> > To: edk2-devel@lists.01.org
> > Cc: Dong, Eric <eric.dong@intel.com>;
> > lersek@redhat.com; Ni, Ruiyu <ruiyu.ni@intel.com>;
> > Kinney, Michael D <michael.d.kinney@intel.com>
> > Subject: [PATCH V2] UefiCpuPkg: Keep library class
> > header file definition independent
> >
> > V2:
> > Move CPU_FEATURE_MAX definition from header file to C
> > file.
> > V1:
> > Keep library class header file definition independent
> >
> > Cc: Eric Dong <eric.dong@intel.com>
> > Cc: Laszlo Ersek <lersek@redhat.com>
> > Contributed-under: TianoCore Contribution Agreement 1.1
> > Signed-off-by: Bell Song <binx.song@intel.com>
> > ---
> >  UefiCpuPkg/Include/Library/RegisterCpuFeaturesLib.h
> > |  5 -----
> >
> > .../Library/RegisterCpuFeaturesLib/RegisterCpuFeaturesL
> > ib.c   | 11 ++++++-----
> >  2 files changed, 6 insertions(+), 10 deletions(-)
> >
> > diff --git
> > a/UefiCpuPkg/Include/Library/RegisterCpuFeaturesLib.h
> > b/UefiCpuPkg/Include/Library/RegisterCpuFeaturesLib.h
> > index fc3ccda..9331e49 100644
> > ---
> > a/UefiCpuPkg/Include/Library/RegisterCpuFeaturesLib.h
> > +++
> > b/UefiCpuPkg/Include/Library/RegisterCpuFeaturesLib.h
> > @@ -71,11 +71,6 @@
> >  #define CPU_FEATURE_APIC_TPR_UPDATE_MESSAGE
> > (32+9)
> >  #define CPU_FEATURE_ENERGY_PERFORMANCE_BIAS
> > (32+10)
> >  #define CPU_FEATURE_PPIN
> > (32+11)
> > -//
> > -// Currently, CPU_FEATURE_PROC_TRACE is the MAX
> > feature we support.
> > -// If you define a feature bigger than it, please also
> > replace it
> > -// in RegisterCpuFeatureLibIsFeatureValid function.
> > -//
> >  #define CPU_FEATURE_PROC_TRACE
> > (32+12)
> >
> >  #define CPU_FEATURE_BEFORE_ALL
> > BIT27
> > diff --git
> > a/UefiCpuPkg/Library/RegisterCpuFeaturesLib/RegisterCpu
> > FeaturesLib.c
> > b/UefiCpuPkg/Library/RegisterCpuFeaturesLib/RegisterCpu
> > FeaturesLib.c
> > index 6ec26e1..afc424c 100644
> > ---
> > a/UefiCpuPkg/Library/RegisterCpuFeaturesLib/RegisterCpu
> > FeaturesLib.c
> > +++
> > b/UefiCpuPkg/Library/RegisterCpuFeaturesLib/RegisterCpu
> > FeaturesLib.c
> > @@ -13,6 +13,10 @@
> >  **/
> >
> >  #include "RegisterCpuFeatures.h"
> > +//
> > +// Please keep CPU_FEATURE_MAX as the max CPU feature
> > +//
> > +#define CPU_FEATURE_MAX              (32+12)
> >
> >  /**
> >    Checks if two CPU feature bit masks are equal.
> > @@ -97,11 +101,8 @@ RegisterCpuFeatureLibIsFeatureValid
> > (
> >
> >    Data = Feature;
> >    Data &= ~(CPU_FEATURE_BEFORE | CPU_FEATURE_AFTER |
> > CPU_FEATURE_BEFORE_ALL | CPU_FEATURE_AFTER_ALL);
> > -  //
> > -  // Currently, CPU_FEATURE_PROC_TRACE is the MAX
> > feature we support.
> > -  // If you define a feature bigger than it, please
> > replace it at below.
> > -  //
> > -  if (Data > CPU_FEATURE_PROC_TRACE) {
> > +
> > +  if (Data > CPU_FEATURE_MAX) {
> >      DEBUG ((DEBUG_ERROR, "Invalid CPU feature: 0x%x ",
> > Feature));
> >      return FALSE;
> >    }
> > --
> > 2.10.2.windows.1