[00/12] x86: various improvements

Message ID 5B0EB75802000078001C70B7@prv1-mh.provo.novell.com
Headers show
Series
  • x86: various improvements
Related show

Message

Jan Beulich May 30, 2018, 2:38 p.m.
In particular, a lot of template folding. But some of the then available infrastructure
also allows to fix a couple of bugs. Not all of patches are tightly coupled together,
but I'm afraid they need to be applied strictly in the order given.

01: fold MOV to/from segment register templates
02: fold various AVX512VL templates into their AVX512F counterparts
03: fold various AVX512CD templates
04: fold various AVX512BW templates
05: fold various AVX512DQ templates
06: fold various AVX512* templates
07: fold VFPCLASSP{D,S} templates
08: fold narrowing VCVT* templates
09: correct AVX512F vcvtsi2s{d,s} handling
10: drop CpuVREX
11: don't abort() upon DATA16 prefix on (E)VEX encoded insn
12: correct permitted operand sizes for AVX512 scatter/gather

Jan