cmd/compile: add intrinsic for reading caller's pc
First step towards removing the mandatory argument for
getcallerpc, which solves certain problems for the runtime.
This might also slightly improve performance.
Intrinsic enabled on 386, amd64, amd64p32,
runtime asm implementation removed on those architectures.
Now-superfluous argument remains in getcallerpc signature
(for a future CL; non-386/amd64 asm funcs ignore it).
Added getcallerpc to the "not a real function" test
in dcl.go, that story is a little odd with respect to
unexported functions but that is not this CL.
Fixes #17327.
Change-Id: I5df1ad91f27ee9ac1f0dd88fa48f1329d6306c3e
Reviewed-on: https://go-review.googlesource.com/31851
Run-TryBot: David Chase <drchase@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Austin Clements <austin@google.com>
diff --git a/src/cmd/compile/internal/amd64/ssa.go b/src/cmd/compile/internal/amd64/ssa.go
index 8c92f07..22e69aa 100644
--- a/src/cmd/compile/internal/amd64/ssa.go
+++ b/src/cmd/compile/internal/amd64/ssa.go
@@ -795,6 +795,15 @@
}
case ssa.OpAMD64CALLstatic, ssa.OpAMD64CALLclosure, ssa.OpAMD64CALLinter:
s.Call(v)
+
+ case ssa.OpAMD64LoweredGetCallerPC:
+ p := s.Prog(x86.AMOVQ)
+ p.From.Type = obj.TYPE_MEM
+ p.From.Offset = -8 // PC is stored 8 bytes below first parameter.
+ p.From.Name = obj.NAME_PARAM
+ p.To.Type = obj.TYPE_REG
+ p.To.Reg = v.Reg()
+
case ssa.OpAMD64NEGQ, ssa.OpAMD64NEGL,
ssa.OpAMD64BSWAPQ, ssa.OpAMD64BSWAPL,
ssa.OpAMD64NOTQ, ssa.OpAMD64NOTL:
diff --git a/src/cmd/compile/internal/gc/dcl.go b/src/cmd/compile/internal/gc/dcl.go
index 66caf2d..cf91f53 100644
--- a/src/cmd/compile/internal/gc/dcl.go
+++ b/src/cmd/compile/internal/gc/dcl.go
@@ -1077,9 +1077,9 @@
if s.IsBlank() {
return
}
- if compiling_runtime && (s.Name == "getg" || s.Name == "getclosureptr") {
- // runtime.getg() and getclosureptr are not real functions and so do not
- // get funcsyms.
+ if compiling_runtime && (s.Name == "getg" || s.Name == "getclosureptr" || s.Name == "getcallerpc") {
+ // runtime.getg(), getclosureptr(), and getcallerpc() are
+ // not real functions and so do not get funcsyms.
return
}
if _, existed := s.Pkg.LookupOK(funcsymname(s)); !existed {
diff --git a/src/cmd/compile/internal/gc/ssa.go b/src/cmd/compile/internal/gc/ssa.go
index 94446d8..1a96049 100644
--- a/src/cmd/compile/internal/gc/ssa.go
+++ b/src/cmd/compile/internal/gc/ssa.go
@@ -2584,6 +2584,11 @@
},
all...)
+ addF("runtime", "getcallerpc",
+ func(s *state, n *Node, args []*ssa.Value) *ssa.Value {
+ return s.newValue0(ssa.OpGetCallerPC, s.f.Config.Types.Uintptr)
+ }, sys.AMD64, sys.I386)
+
/******** runtime/internal/sys ********/
addF("runtime/internal/sys", "Ctz32",
func(s *state, n *Node, args []*ssa.Value) *ssa.Value {
@@ -2614,7 +2619,6 @@
return s.newValue1(ssa.OpSelect0, types.Types[TUINT32], v)
},
sys.AMD64, sys.ARM64, sys.S390X, sys.MIPS, sys.PPC64)
-
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())
diff --git a/src/cmd/compile/internal/ssa/gen/386.rules b/src/cmd/compile/internal/ssa/gen/386.rules
index 49fcd36..c961c0f 100644
--- a/src/cmd/compile/internal/ssa/gen/386.rules
+++ b/src/cmd/compile/internal/ssa/gen/386.rules
@@ -383,6 +383,7 @@
(NilCheck ptr mem) -> (LoweredNilCheck ptr mem)
(GetG mem) -> (LoweredGetG mem)
(GetClosurePtr) -> (LoweredGetClosurePtr)
+(GetCallerPC) -> (LoweredGetCallerPC)
(Addr {sym} base) -> (LEAL {sym} base)
// block rewrites
diff --git a/src/cmd/compile/internal/ssa/gen/386Ops.go b/src/cmd/compile/internal/ssa/gen/386Ops.go
index 25f3931..bea80fd 100644
--- a/src/cmd/compile/internal/ssa/gen/386Ops.go
+++ b/src/cmd/compile/internal/ssa/gen/386Ops.go
@@ -440,6 +440,11 @@
// and sorts it to the very beginning of the block to prevent other
// use of DX (the closure pointer)
{name: "LoweredGetClosurePtr", reg: regInfo{outputs: []regMask{buildReg("DX")}}},
+ // LoweredGetCallerPC evaluates to the PC to which its "caller" will return.
+ // I.e., if f calls g "calls" getcallerpc,
+ // the result should be the PC within f that g will return to.
+ // See runtime/stubs.go for a more detailed discussion.
+ {name: "LoweredGetCallerPC", reg: gp01},
//arg0=ptr,arg1=mem, returns void. Faults if ptr is nil.
{name: "LoweredNilCheck", argLength: 2, reg: regInfo{inputs: []regMask{gpsp}}, clobberFlags: true, nilCheck: true, faultOnNilArg0: true},
diff --git a/src/cmd/compile/internal/ssa/gen/AMD64.rules b/src/cmd/compile/internal/ssa/gen/AMD64.rules
index e4321cb..724b921 100644
--- a/src/cmd/compile/internal/ssa/gen/AMD64.rules
+++ b/src/cmd/compile/internal/ssa/gen/AMD64.rules
@@ -482,6 +482,7 @@
(NilCheck ptr mem) -> (LoweredNilCheck ptr mem)
(GetG mem) -> (LoweredGetG mem)
(GetClosurePtr) -> (LoweredGetClosurePtr)
+(GetCallerPC) -> (LoweredGetCallerPC)
(Addr {sym} base) && config.PtrSize == 8 -> (LEAQ {sym} base)
(Addr {sym} base) && config.PtrSize == 4 -> (LEAL {sym} base)
diff --git a/src/cmd/compile/internal/ssa/gen/AMD64Ops.go b/src/cmd/compile/internal/ssa/gen/AMD64Ops.go
index 8390fd0..51dc3d3 100644
--- a/src/cmd/compile/internal/ssa/gen/AMD64Ops.go
+++ b/src/cmd/compile/internal/ssa/gen/AMD64Ops.go
@@ -540,6 +540,11 @@
// and sorts it to the very beginning of the block to prevent other
// use of DX (the closure pointer)
{name: "LoweredGetClosurePtr", reg: regInfo{outputs: []regMask{buildReg("DX")}}},
+ // LoweredGetCallerPC evaluates to the PC to which its "caller" will return.
+ // I.e., if f calls g "calls" getcallerpc,
+ // the result should be the PC within f that g will return to.
+ // See runtime/stubs.go for a more detailed discussion.
+ {name: "LoweredGetCallerPC", reg: gp01},
//arg0=ptr,arg1=mem, returns void. Faults if ptr is nil.
{name: "LoweredNilCheck", argLength: 2, reg: regInfo{inputs: []regMask{gpsp}}, clobberFlags: true, nilCheck: true, faultOnNilArg0: true},
diff --git a/src/cmd/compile/internal/ssa/gen/genericOps.go b/src/cmd/compile/internal/ssa/gen/genericOps.go
index 2967d29..bec7ef9 100644
--- a/src/cmd/compile/internal/ssa/gen/genericOps.go
+++ b/src/cmd/compile/internal/ssa/gen/genericOps.go
@@ -378,6 +378,7 @@
// Pseudo-ops
{name: "GetG", argLength: 1}, // runtime.getg() (read g pointer). arg0=mem
{name: "GetClosurePtr"}, // get closure pointer from dedicated register
+ {name: "GetCallerPC"}, // for getcallerpc intrinsic
// Indexing operations
{name: "PtrIndex", argLength: 2}, // arg0=ptr, arg1=index. Computes ptr+sizeof(*v.type)*index, where index is extended to ptrwidth type
diff --git a/src/cmd/compile/internal/ssa/opGen.go b/src/cmd/compile/internal/ssa/opGen.go
index 1f0138b..39b9be3 100644
--- a/src/cmd/compile/internal/ssa/opGen.go
+++ b/src/cmd/compile/internal/ssa/opGen.go
@@ -396,6 +396,7 @@
Op386InvertFlags
Op386LoweredGetG
Op386LoweredGetClosurePtr
+ Op386LoweredGetCallerPC
Op386LoweredNilCheck
Op386MOVLconvert
Op386FlagEQ
@@ -653,6 +654,7 @@
OpAMD64InvertFlags
OpAMD64LoweredGetG
OpAMD64LoweredGetClosurePtr
+ OpAMD64LoweredGetCallerPC
OpAMD64LoweredNilCheck
OpAMD64MOVQconvert
OpAMD64MOVLconvert
@@ -1905,6 +1907,7 @@
OpNilCheck
OpGetG
OpGetClosurePtr
+ OpGetCallerPC
OpPtrIndex
OpOffPtr
OpSliceMake
@@ -4287,6 +4290,15 @@
},
},
{
+ name: "LoweredGetCallerPC",
+ argLen: 0,
+ reg: regInfo{
+ outputs: []outputInfo{
+ {0, 239}, // AX CX DX BX BP SI DI
+ },
+ },
+ },
+ {
name: "LoweredNilCheck",
argLen: 2,
clobberFlags: true,
@@ -7915,6 +7927,15 @@
},
},
{
+ name: "LoweredGetCallerPC",
+ argLen: 0,
+ reg: regInfo{
+ outputs: []outputInfo{
+ {0, 65519}, // AX CX DX BX BP SI DI R8 R9 R10 R11 R12 R13 R14 R15
+ },
+ },
+ },
+ {
name: "LoweredNilCheck",
argLen: 2,
clobberFlags: true,
@@ -22835,6 +22856,11 @@
generic: true,
},
{
+ name: "GetCallerPC",
+ argLen: 0,
+ generic: true,
+ },
+ {
name: "PtrIndex",
argLen: 2,
generic: true,
diff --git a/src/cmd/compile/internal/ssa/rewrite386.go b/src/cmd/compile/internal/ssa/rewrite386.go
index 6726dbd..b363dd3 100644
--- a/src/cmd/compile/internal/ssa/rewrite386.go
+++ b/src/cmd/compile/internal/ssa/rewrite386.go
@@ -329,6 +329,8 @@
return rewriteValue386_OpGeq8_0(v)
case OpGeq8U:
return rewriteValue386_OpGeq8U_0(v)
+ case OpGetCallerPC:
+ return rewriteValue386_OpGetCallerPC_0(v)
case OpGetClosurePtr:
return rewriteValue386_OpGetClosurePtr_0(v)
case OpGetG:
@@ -15052,6 +15054,15 @@
return true
}
}
+func rewriteValue386_OpGetCallerPC_0(v *Value) bool {
+ // match: (GetCallerPC)
+ // cond:
+ // result: (LoweredGetCallerPC)
+ for {
+ v.reset(Op386LoweredGetCallerPC)
+ return true
+ }
+}
func rewriteValue386_OpGetClosurePtr_0(v *Value) bool {
// match: (GetClosurePtr)
// cond:
diff --git a/src/cmd/compile/internal/ssa/rewriteAMD64.go b/src/cmd/compile/internal/ssa/rewriteAMD64.go
index 01d2f88..60d68db 100644
--- a/src/cmd/compile/internal/ssa/rewriteAMD64.go
+++ b/src/cmd/compile/internal/ssa/rewriteAMD64.go
@@ -563,6 +563,8 @@
return rewriteValueAMD64_OpGeq8_0(v)
case OpGeq8U:
return rewriteValueAMD64_OpGeq8U_0(v)
+ case OpGetCallerPC:
+ return rewriteValueAMD64_OpGetCallerPC_0(v)
case OpGetClosurePtr:
return rewriteValueAMD64_OpGetClosurePtr_0(v)
case OpGetG:
@@ -40851,6 +40853,15 @@
return true
}
}
+func rewriteValueAMD64_OpGetCallerPC_0(v *Value) bool {
+ // match: (GetCallerPC)
+ // cond:
+ // result: (LoweredGetCallerPC)
+ for {
+ v.reset(OpAMD64LoweredGetCallerPC)
+ return true
+ }
+}
func rewriteValueAMD64_OpGetClosurePtr_0(v *Value) bool {
// match: (GetClosurePtr)
// cond:
diff --git a/src/cmd/compile/internal/x86/ssa.go b/src/cmd/compile/internal/x86/ssa.go
index ca430d6..12cbac6 100644
--- a/src/cmd/compile/internal/x86/ssa.go
+++ b/src/cmd/compile/internal/x86/ssa.go
@@ -666,6 +666,15 @@
q.To.Type = obj.TYPE_REG
q.To.Reg = r
}
+
+ case ssa.Op386LoweredGetCallerPC:
+ p := s.Prog(x86.AMOVL)
+ p.From.Type = obj.TYPE_MEM
+ p.From.Offset = -4 // PC is stored 4 bytes below first parameter.
+ p.From.Name = obj.NAME_PARAM
+ p.To.Type = obj.TYPE_REG
+ p.To.Reg = v.Reg()
+
case ssa.Op386CALLstatic, ssa.Op386CALLclosure, ssa.Op386CALLinter:
s.Call(v)
case ssa.Op386NEGL,
