Messages in this thread | | | Date | Wed, 26 Aug 2015 09:10:09 -0700 | From | Dave Hansen <> | Subject | Re: [PATCH 02/11] x86, fpu: rename xfeature_bit |
| |
On 08/26/2015 09:06 AM, Borislav Petkov wrote: >> > -enum xfeature_bit { >> > - XSTATE_BIT_FP, >> > - XSTATE_BIT_SSE, >> > - XSTATE_BIT_YMM, >> > - XSTATE_BIT_BNDREGS, >> > - XSTATE_BIT_BNDCSR, >> > - XSTATE_BIT_OPMASK, >> > - XSTATE_BIT_ZMM_Hi256, >> > - XSTATE_BIT_Hi16_ZMM, >> > +enum xfeature_nr { >> > + XFEATURE_NR_FP, >> > + XFEATURE_NR_SSE, >> > + /* >> > + * Values above here are "legacy states". >> > + * Those below are "extended states". >> > + */ >> > + XFEATURE_NR_YMM, >> > + XFEATURE_NR_BNDREGS, >> > + XFEATURE_NR_BNDCSR, >> > + XFEATURE_NR_OPMASK, >> > + XFEATURE_NR_ZMM_Hi256, >> > + XFEATURE_NR_Hi16_ZMM, > Why not simply > > s/NR_// > > ? > > I.e., XFEATURE_FP, XFEATURE_SSE and so on...
I wanted to differentiate them from the
#define XSTATE_FP (1 << XSTATE_BIT_FP) #define XSTATE_SSE (1 << XSTATE_BIT_SSE)
defines below. Those are really XFEATURE_MASK_... I think this gets a lots more clear if you have a pair of #defines like:
XFEATURE_NR_BNDREGS XFEATURE_MASK_BNDREGS
It's pretty obvious what's going on there. What we have now is
XSTATE_BIT_BNDREGS XSTATE_BNDREGS
which I find much less clear.
| |