cmd/compile: intrinsify runtime/internal/atomic.{And,Or}{8,} on RISCV64

The 32 bit versions are easily implement with a single instruction, while the
8 bit versions require a bit more effort but use the same atomic instructions
via rewrite rules.

Change-Id: I42e8d457b239c8f75e39a8e282fc88c1bb292a99
Reviewed-on: https://go-review.googlesource.com/c/go/+/268098
Trust: Joel Sing <joel@sing.id.au>
Run-TryBot: Joel Sing <joel@sing.id.au>
TryBot-Result: Go Bot <gobot@golang.org>
Reviewed-by: Cherry Zhang <cherryyz@google.com>

src/cmd/compile/internal/riscv64/ssa.go

@@ -510,6 +510,14 @@ func ssaGenValue(s *ssagen.State, v *ssa.Value) {
 		p6 := s.Prog(obj.ANOP)
 		p2.To.SetTarget(p6)
 
+	case ssa.OpRISCV64LoweredAtomicAnd32, ssa.OpRISCV64LoweredAtomicOr32:
+		p := s.Prog(v.Op.Asm())
+		p.From.Type = obj.TYPE_REG
+		p.From.Reg = v.Args[1].Reg()
+		p.To.Type = obj.TYPE_MEM
+		p.To.Reg = v.Args[0].Reg()
+		p.RegTo2 = riscv.REG_ZERO
+
 	case ssa.OpRISCV64LoweredZero:
 		mov, sz := largestMove(v.AuxInt)
 

src/cmd/compile/internal/ssa/gen/RISCV64.rules

@@ -564,12 +564,28 @@
 (AtomicAdd32 ...) => (LoweredAtomicAdd32 ...)
 (AtomicAdd64 ...) => (LoweredAtomicAdd64 ...)
 
+// AtomicAnd8(ptr,val) => LoweredAtomicAnd32(ptr&^3, ^((uint8(val) ^ 0xff) << ((ptr & 3) * 8)))
+(AtomicAnd8 ptr val mem) =>
+	(LoweredAtomicAnd32 (ANDI <typ.Uintptr> [^3] ptr)
+		(NOT <typ.UInt32> (SLL <typ.UInt32> (XORI <typ.UInt32> [0xff] (ZeroExt8to32 val))
+			(SLLI <typ.UInt64> [3] (ANDI <typ.UInt64> [3] ptr)))) mem)
+
+(AtomicAnd32 ...) => (LoweredAtomicAnd32 ...)
+
 (AtomicCompareAndSwap32 ...) => (LoweredAtomicCas32 ...)
 (AtomicCompareAndSwap64 ...) => (LoweredAtomicCas64 ...)
 
 (AtomicExchange32 ...) => (LoweredAtomicExchange32 ...)
 (AtomicExchange64 ...) => (LoweredAtomicExchange64 ...)
 
+// AtomicOr8(ptr,val)  => LoweredAtomicOr32(ptr&^3, uint32(val)<<((ptr&3)*8))
+(AtomicOr8 ptr val mem) =>
+	(LoweredAtomicOr32 (ANDI <typ.Uintptr> [^3] ptr)
+		(SLL <typ.UInt32> (ZeroExt8to32 val)
+			(SLLI <typ.UInt64> [3] (ANDI <typ.UInt64> [3] ptr))) mem)
+
+(AtomicOr32  ...) => (LoweredAtomicOr32  ...)
+
 // Conditional branches
 (If cond yes no) => (BNEZ cond yes no)
 

src/cmd/compile/internal/ssa/gen/RISCV64Ops.go

@@ -126,6 +126,7 @@ func init() {
 		gp11sb   = regInfo{inputs: []regMask{gpspsbMask}, outputs: []regMask{gpMask}}
 		gpxchg   = regInfo{inputs: []regMask{gpspsbgMask, gpgMask}, outputs: []regMask{gpMask}}
 		gpcas    = regInfo{inputs: []regMask{gpspsbgMask, gpgMask, gpgMask}, outputs: []regMask{gpMask}}
+		gpatomic = regInfo{inputs: []regMask{gpspsbgMask, gpgMask}}
 
 		fp11    = regInfo{inputs: []regMask{fpMask}, outputs: []regMask{fpMask}}
 		fp21    = regInfo{inputs: []regMask{fpMask, fpMask}, outputs: []regMask{fpMask}}
@@ -335,7 +336,7 @@ func init() {
 		{name: "LoweredAtomicLoad64", argLength: 2, reg: gpload, faultOnNilArg0: true},
 
 		// Atomic stores.
-		// store arg1 to arg0. arg2=mem. returns memory.
+		// store arg1 to *arg0. arg2=mem. returns memory.
 		{name: "LoweredAtomicStore8", argLength: 3, reg: gpstore, faultOnNilArg0: true, hasSideEffects: true},
 		{name: "LoweredAtomicStore32", argLength: 3, reg: gpstore, faultOnNilArg0: true, hasSideEffects: true},
 		{name: "LoweredAtomicStore64", argLength: 3, reg: gpstore, faultOnNilArg0: true, hasSideEffects: true},
@@ -367,6 +368,11 @@ func init() {
 		{name: "LoweredAtomicCas32", argLength: 4, reg: gpcas, resultNotInArgs: true, faultOnNilArg0: true, hasSideEffects: true, unsafePoint: true},
 		{name: "LoweredAtomicCas64", argLength: 4, reg: gpcas, resultNotInArgs: true, faultOnNilArg0: true, hasSideEffects: true, unsafePoint: true},
 
+		// Atomic 32 bit AND/OR.
+		// *arg0 &= (|=) arg1. arg2=mem. returns nil.
+		{name: "LoweredAtomicAnd32", argLength: 3, reg: gpatomic, asm: "AMOANDW", faultOnNilArg0: true, hasSideEffects: true},
+		{name: "LoweredAtomicOr32", argLength: 3, reg: gpatomic, asm: "AMOORW", faultOnNilArg0: true, hasSideEffects: true},
+
 		// Lowering pass-throughs
 		{name: "LoweredNilCheck", argLength: 2, faultOnNilArg0: true, nilCheck: true, reg: regInfo{inputs: []regMask{gpspMask}}}, // arg0=ptr,arg1=mem, returns void.  Faults if ptr is nil.
 		{name: "LoweredGetClosurePtr", reg: regInfo{outputs: []regMask{regCtxt}}},                                                // scheduler ensures only at beginning of entry block