src/cpu/ppc/vm/assembler_ppc.inline.hpp

rev 13437 : Implement SHA256 and SHA512 intrinsics for PPC64

Sample run reduced 6.5s run to 2.8s for SHA256.

*** 923,933 ****
  inline void Assembler::vncipher(    VectorRegister d, VectorRegister a, VectorRegister b) { emit_int32( VNCIPHER_OPCODE     | vrt(d) | vra(a) | vrb(b)); }
  inline void Assembler::vncipherlast(VectorRegister d, VectorRegister a, VectorRegister b) { emit_int32( VNCIPHERLAST_OPCODE | vrt(d) | vra(a) | vrb(b)); }
  inline void Assembler::vsbox(       VectorRegister d, VectorRegister a)                   { emit_int32( VSBOX_OPCODE        | vrt(d) | vra(a)         ); }
  
  // SHA (introduced with Power 8)
! // Not yet implemented.
  
  // Vector Binary Polynomial Multiplication (introduced with Power 8)
  inline void Assembler::vpmsumb(  VectorRegister d, VectorRegister a, VectorRegister b) { emit_int32( VPMSUMB_OPCODE | vrt(d) | vra(a) | vrb(b)); }
  inline void Assembler::vpmsumd(  VectorRegister d, VectorRegister a, VectorRegister b) { emit_int32( VPMSUMD_OPCODE | vrt(d) | vra(a) | vrb(b)); }
  inline void Assembler::vpmsumh(  VectorRegister d, VectorRegister a, VectorRegister b) { emit_int32( VPMSUMH_OPCODE | vrt(d) | vra(a) | vrb(b)); }
--- 923,934 ----
  inline void Assembler::vncipher(    VectorRegister d, VectorRegister a, VectorRegister b) { emit_int32( VNCIPHER_OPCODE     | vrt(d) | vra(a) | vrb(b)); }
  inline void Assembler::vncipherlast(VectorRegister d, VectorRegister a, VectorRegister b) { emit_int32( VNCIPHERLAST_OPCODE | vrt(d) | vra(a) | vrb(b)); }
  inline void Assembler::vsbox(       VectorRegister d, VectorRegister a)                   { emit_int32( VSBOX_OPCODE        | vrt(d) | vra(a)         ); }
  
  // SHA (introduced with Power 8)
! inline void Assembler::vshasigmad(VectorRegister d, VectorRegister a, bool st, int six) { emit_int32( VSHASIGMAD_OPCODE | vrt(d) | vra(a) | vst(st) | vsix(six)); }
! inline void Assembler::vshasigmaw(VectorRegister d, VectorRegister a, bool st, int six) { emit_int32( VSHASIGMAW_OPCODE | vrt(d) | vra(a) | vst(st) | vsix(six)); }
  
  // Vector Binary Polynomial Multiplication (introduced with Power 8)
  inline void Assembler::vpmsumb(  VectorRegister d, VectorRegister a, VectorRegister b) { emit_int32( VPMSUMB_OPCODE | vrt(d) | vra(a) | vrb(b)); }
  inline void Assembler::vpmsumd(  VectorRegister d, VectorRegister a, VectorRegister b) { emit_int32( VPMSUMD_OPCODE | vrt(d) | vra(a) | vrb(b)); }
  inline void Assembler::vpmsumh(  VectorRegister d, VectorRegister a, VectorRegister b) { emit_int32( VPMSUMH_OPCODE | vrt(d) | vra(a) | vrb(b)); }