cmd/asm, cmd/internal/obj/ppc64: Add ISA 2.05, 2.06 and 2.07 instructions.

This change adds instructions from ISA 2.05, 2.06 and 2.07 that are frequently
used in assembly optimizations for ppc64.

It also fixes two problems:

  * the implementation of RLDICR[CC]/RLDICL[CC] did not consider all possible
  cases for the bit mask.
  * removed two non-existing instructions that were added by mistake in the VMX
  implementation (VORL/VANDL).

Change-Id: Iaef4e5c6a5240c2156c6c0f28ad3bcd8780e9830
Reviewed-on: https://go-review.googlesource.com/36230
Run-TryBot: Lynn Boger <laboger@linux.vnet.ibm.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Lynn Boger <laboger@linux.vnet.ibm.com>

src/cmd/asm/internal/asm/testdata/ppc64.s

@@ -540,6 +540,14 @@ label1:
 //	}
 //	FCMPU	F1, F2, CR0
 
+//	FTDIV FRA, FRB, BF produces
+//	ftdiv BF, FRA, FRB
+	FTDIV F1,F2,$7
+
+//	FTSQRT	FRB, BF produces
+//	ftsqrt	BF, FRB
+	FTSQRT	F2,$7
+
 //
 // CMP
 //
@@ -567,6 +575,10 @@ label1:
 //	}
 	CMP	R1, $4, CR0 // CMP R1, CR0, $4
 
+//	CMPB  RS,RB,RA produces
+//	cmpb  RA,RS,RB
+	CMPB  R2,R2,R1
+
 //
 // rotate and mask
 //
@@ -673,6 +685,13 @@ label1:
 	DCBF	(R1)
 	DCBF	(R1+R2) // DCBF	(R1)(R2*1)
 
+//	Population count, X-form
+//	<MNEMONIC> RS,RA produces
+//	<mnemonic> RA,RS
+	POPCNTD	R1,R2
+	POPCNTW	R1,R2
+	POPCNTB R1,R2
+
 //	VMX instructions
 
 //	Described as:
@@ -703,14 +722,14 @@ label1:
 //	Vector AND, VX-form
 //	<MNEMONIC> VRA,VRB,VRT produces
 //	<mnemonic> VRT,VRA,VRB
-	VANDL	V10, V9, V8
+	VAND	V10, V9, V8
 	VANDC	V15, V14, V13
 	VNAND	V19, V18, V17
 
 //	Vector OR, VX-form
 //	<MNEMONIC> VRA,VRB,VRT produces
 //	<mnemonic> VRT,VRA,VRB
-	VORL	V26, V25, V24
+	VOR	V26, V25, V24
 	VORC	V23, V22, V21
 	VNOR	V20, V19, V18
 	VXOR	V17, V16, V15
@@ -739,6 +758,14 @@ label1:
 	VADDEUQM V4, V3, V2, V1
 	VADDECUQ V4, V3, V2, V1
 
+//	Vector polynomial multiply-sum, VX-form
+//	<MNEMONIC>  VRA,VRB,VRT produces
+//	<mnemonic>  VRT,VRA,VRB
+	VPMSUMB	V2, V3, V1
+	VPMSUMH	V2, V3, V1
+	VPMSUMW	V2, V3, V1
+	VPMSUMD	V2, V3, V1
+
 //	Vector SUB, VX-form
 //	<MNEMONIC> VRA,VRB,VRT produces
 //	<mnemonic> VRT,VRA,VRB

src/cmd/internal/obj/ppc64/a.out.go

@@ -626,6 +626,9 @@ const (
 	ACNTLZDCC
 	ACMPW /* CMP with L=0 */
 	ACMPWU
+	ACMPB
+	AFTDIV
+	AFTSQRT
 	ADIVD
 	ADIVDCC
 	ADIVDE
@@ -704,6 +707,9 @@ const (
 
 	/* more 64-bit operations */
 	AHRFID
+	APOPCNTD
+	APOPCNTW
+	APOPCNTB
 
 	/* Vector */
 	ALV
@@ -721,11 +727,9 @@ const (
 	ASTVX
 	ASTVXL
 	AVAND
-	AVANDL
 	AVANDC
 	AVNAND
 	AVOR
-	AVORL
 	AVORC
 	AVNOR
 	AVXOR
@@ -770,6 +774,11 @@ const (
 	AVSUBE
 	AVSUBEUQM
 	AVSUBECUQ
+	AVPMSUM
+	AVPMSUMB
+	AVPMSUMH
+	AVPMSUMW
+	AVPMSUMD
 	AVR
 	AVRLB
 	AVRLH
@@ -871,9 +880,13 @@ const (
 	ASTXSIWX
 	AMFVSR
 	AMFVSRD
+	AMFFPRD
+	AMFVRD
 	AMFVSRWZ
 	AMTVSR
 	AMTVSRD
+	AMTFPRD
+	AMTVRD
 	AMTVSRWA
 	AMTVSRWZ
 	AXXLAND

src/cmd/internal/obj/ppc64/anames.go

@@ -248,6 +248,9 @@ var Anames = []string{
 	"CNTLZDCC",
 	"CMPW",
 	"CMPWU",
+	"CMPB",
+	"FTDIV",
+	"FTSQRT",
 	"DIVD",
 	"DIVDCC",
 	"DIVDE",
@@ -321,6 +324,9 @@ var Anames = []string{
 	"REMDUV",
 	"REMDUVCC",
 	"HRFID",
+	"POPCNTD",
+	"POPCNTW",
+	"POPCNTB",
 	"LV",
 	"LVEBX",
 	"LVEHX",
@@ -336,11 +342,9 @@ var Anames = []string{
 	"STVX",
 	"STVXL",
 	"VAND",
-	"VANDL",
 	"VANDC",
 	"VNAND",
 	"VOR",
-	"VORL",
 	"VORC",
 	"VNOR",
 	"VXOR",
@@ -385,6 +389,11 @@ var Anames = []string{
 	"VSUBE",
 	"VSUBEUQM",
 	"VSUBECUQ",
+	"VPMSUM",
+	"VPMSUMB",
+	"VPMSUMH",
+	"VPMSUMW",
+	"VPMSUMD",
 	"VR",
 	"VRLB",
 	"VRLH",
@@ -484,9 +493,13 @@ var Anames = []string{
 	"STXSIWX",
 	"MFVSR",
 	"MFVSRD",
+	"MFFPRD",
+	"MFVRD",
 	"MFVSRWZ",
 	"MTVSR",
 	"MTVSRD",
+	"MTFPRD",
+	"MTVRD",
 	"MTVSRWA",
 	"MTVSRWZ",
 	"XXLAND",