cmd/compile: provide Load8/Store8 atomic intrinsics on riscv64

Updates #36765

Change-Id: Ieeb6bbc54e4841a1348ad50e80342ec4bc675e07
Reviewed-on: https://go-review.googlesource.com/c/go/+/223557
Reviewed-by: Cherry Zhang <[email protected]>
Run-TryBot: Cherry Zhang <[email protected]>
TryBot-Result: Gobot Gobot <[email protected]>

Author: 4a6f656c

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

@@ -3341,7 +3341,7 @@ func init() {
 			s.vars[&memVar] = s.newValue1(ssa.OpSelect1, types.TypeMem, v)
 			return s.newValue1(ssa.OpSelect0, types.Types[TUINT8], v)
 		},
-		sys.AMD64, sys.ARM64, sys.S390X, sys.MIPS, sys.MIPS64, sys.PPC64)
+		sys.AMD64, sys.ARM64, sys.MIPS, sys.MIPS64, sys.PPC64, sys.RISCV64, sys.S390X)
 	addF("runtime/internal/atomic", "Load64",
 		func(s *state, n *Node, args []*ssa.Value) *ssa.Value {
 			v := s.newValue2(ssa.OpAtomicLoad64, types.NewTuple(types.Types[TUINT64], types.TypeMem), args[0], s.mem())
@@ -3375,7 +3375,7 @@ func init() {
 			s.vars[&memVar] = s.newValue3(ssa.OpAtomicStore8, types.TypeMem, args[0], args[1], s.mem())
 			return nil
 		},
-		sys.AMD64, sys.ARM64, sys.S390X, sys.MIPS, sys.MIPS64, sys.PPC64)
+		sys.AMD64, sys.ARM64, sys.MIPS, sys.MIPS64, sys.PPC64, sys.RISCV64, sys.S390X)
 	addF("runtime/internal/atomic", "Store64",
 		func(s *state, n *Node, args []*ssa.Value) *ssa.Value {
 			s.vars[&memVar] = s.newValue3(ssa.OpAtomicStore64, types.TypeMem, args[0], args[1], s.mem())

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

@@ -341,6 +341,25 @@ func ssaGenValue(s *gc.SSAGenState, v *ssa.Value) {
 		p.To.Name = obj.NAME_EXTERN
 		p.To.Sym = gc.BoundsCheckFunc[v.AuxInt]
 		s.UseArgs(16) // space used in callee args area by assembly stubs
+
+	case ssa.OpRISCV64LoweredAtomicLoad8:
+		s.Prog(riscv.AFENCE)
+		p := s.Prog(riscv.AMOVBU)
+		p.From.Type = obj.TYPE_MEM
+		p.From.Reg = v.Args[0].Reg()
+		p.To.Type = obj.TYPE_REG
+		p.To.Reg = v.Reg0()
+		s.Prog(riscv.AFENCE)
+
+	case ssa.OpRISCV64LoweredAtomicStore8:
+		s.Prog(riscv.AFENCE)
+		p := s.Prog(riscv.AMOVB)
+		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()
+		s.Prog(riscv.AFENCE)
+
 	case ssa.OpRISCV64LoweredZero:
 		mov, sz := largestMove(v.AuxInt)
 

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

@@ -468,6 +468,11 @@
 (ClosureCall ...) -> (CALLclosure ...)
 (InterCall   ...) -> (CALLinter   ...)
 
+// Atomic Intrinsics
+(AtomicLoad8 ...) -> (LoweredAtomicLoad8 ...)
+
+(AtomicStore8 ...) -> (LoweredAtomicStore8 ...)
+
 // Optimizations
 
 // Absorb SNEZ into branch.

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

@@ -265,6 +265,15 @@ func init() {
 			faultOnNilArg1: true,
 		},
 
+		// Atomic loads.
+		// load from arg0. arg1=mem.
+		// returns <value,memory> so they can be properly ordered with other loads.
+		{name: "LoweredAtomicLoad8", argLength: 2, reg: gpload, faultOnNilArg0: true},
+
+		// Atomic stores.
+		// store arg1 to arg0. arg2=mem. returns memory.
+		{name: "LoweredAtomicStore8", argLength: 3, reg: gpstore, 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