blob: dadc12f4dbfba5a28c5fac8b731dc8d7f1cf0ac3 [file] [log] [blame]
// Copyright 2021 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
//go:build race
#include "go_asm.h"
#include "funcdata.h"
#include "textflag.h"
// The following thunks allow calling the gcc-compiled race runtime directly
// from Go code without going all the way through cgo.
// First, it's much faster (up to 50% speedup for real Go programs).
// Second, it eliminates race-related special cases from cgocall and scheduler.
// Third, in long-term it will allow to remove cyclic runtime/race dependency on cmd/go.
// A brief recap of the s390x C calling convention.
// Arguments are passed in R2...R6, the rest is on stack.
// Callee-saved registers are: R6...R13, R15.
// Temporary registers are: R0...R5, R14.
// When calling racecalladdr, R1 is the call target address.
// The race ctx, ThreadState *thr below, is passed in R2 and loaded in racecalladdr.
// func runtime·raceread(addr uintptr)
// Called from instrumented code.
TEXT runtime·raceread(SB), NOSPLIT, $0-8
// void __tsan_read(ThreadState *thr, void *addr, void *pc);
MOVD $__tsan_read(SB), R1
MOVD addr+0(FP), R3
MOVD R14, R4
JMP racecalladdr<>(SB)
// func runtime·RaceRead(addr uintptr)
TEXT runtime·RaceRead(SB), NOSPLIT, $0-8
// This needs to be a tail call, because raceread reads caller pc.
JMP runtime·raceread(SB)
// func runtime·racereadpc(void *addr, void *callpc, void *pc)
TEXT runtime·racereadpc(SB), NOSPLIT, $0-24
// void __tsan_read_pc(ThreadState *thr, void *addr, void *callpc, void *pc);
MOVD $__tsan_read_pc(SB), R1
LMG addr+0(FP), R3, R5
JMP racecalladdr<>(SB)
// func runtime·racewrite(addr uintptr)
// Called from instrumented code.
TEXT runtime·racewrite(SB), NOSPLIT, $0-8
// void __tsan_write(ThreadState *thr, void *addr, void *pc);
MOVD $__tsan_write(SB), R1
MOVD addr+0(FP), R3
MOVD R14, R4
JMP racecalladdr<>(SB)
// func runtime·RaceWrite(addr uintptr)
TEXT runtime·RaceWrite(SB), NOSPLIT, $0-8
// This needs to be a tail call, because racewrite reads caller pc.
JMP runtime·racewrite(SB)
// func runtime·racewritepc(void *addr, void *callpc, void *pc)
TEXT runtime·racewritepc(SB), NOSPLIT, $0-24
// void __tsan_write_pc(ThreadState *thr, void *addr, void *callpc, void *pc);
MOVD $__tsan_write_pc(SB), R1
LMG addr+0(FP), R3, R5
JMP racecalladdr<>(SB)
// func runtime·racereadrange(addr, size uintptr)
// Called from instrumented code.
TEXT runtime·racereadrange(SB), NOSPLIT, $0-16
// void __tsan_read_range(ThreadState *thr, void *addr, uintptr size, void *pc);
MOVD $__tsan_read_range(SB), R1
LMG addr+0(FP), R3, R4
MOVD R14, R5
JMP racecalladdr<>(SB)
// func runtime·RaceReadRange(addr, size uintptr)
TEXT runtime·RaceReadRange(SB), NOSPLIT, $0-16
// This needs to be a tail call, because racereadrange reads caller pc.
JMP runtime·racereadrange(SB)
// func runtime·racereadrangepc1(void *addr, uintptr sz, void *pc)
TEXT runtime·racereadrangepc1(SB), NOSPLIT, $0-24
// void __tsan_read_range(ThreadState *thr, void *addr, uintptr size, void *pc);
MOVD $__tsan_read_range(SB), R1
LMG addr+0(FP), R3, R5
// pc is an interceptor address, but TSan expects it to point to the
// middle of an interceptor (see LLVM's SCOPED_INTERCEPTOR_RAW).
ADD $2, R5
JMP racecalladdr<>(SB)
// func runtime·racewriterange(addr, size uintptr)
// Called from instrumented code.
TEXT runtime·racewriterange(SB), NOSPLIT, $0-16
// void __tsan_write_range(ThreadState *thr, void *addr, uintptr size, void *pc);
MOVD $__tsan_write_range(SB), R1
LMG addr+0(FP), R3, R4
MOVD R14, R5
JMP racecalladdr<>(SB)
// func runtime·RaceWriteRange(addr, size uintptr)
TEXT runtime·RaceWriteRange(SB), NOSPLIT, $0-16
// This needs to be a tail call, because racewriterange reads caller pc.
JMP runtime·racewriterange(SB)
// func runtime·racewriterangepc1(void *addr, uintptr sz, void *pc)
TEXT runtime·racewriterangepc1(SB), NOSPLIT, $0-24
// void __tsan_write_range(ThreadState *thr, void *addr, uintptr size, void *pc);
MOVD $__tsan_write_range(SB), R1
LMG addr+0(FP), R3, R5
// pc is an interceptor address, but TSan expects it to point to the
// middle of an interceptor (see LLVM's SCOPED_INTERCEPTOR_RAW).
ADD $2, R5
JMP racecalladdr<>(SB)
// If R3 is out of range, do nothing. Otherwise, setup goroutine context and
// invoke racecall. Other arguments are already set.
TEXT racecalladdr<>(SB), NOSPLIT, $0-0
MOVD runtime·racearenastart(SB), R0
CMPUBLT R3, R0, data // Before racearena start?
MOVD runtime·racearenaend(SB), R0
CMPUBLT R3, R0, call // Before racearena end?
data:
MOVD runtime·racedatastart(SB), R0
CMPUBLT R3, R0, ret // Before racedata start?
MOVD runtime·racedataend(SB), R0
CMPUBGE R3, R0, ret // At or after racedata end?
call:
MOVD g_racectx(g), R2
JMP racecall<>(SB)
ret:
RET
// func runtime·racefuncenter(pc uintptr)
// Called from instrumented code.
TEXT runtime·racefuncenter(SB), NOSPLIT, $0-8
MOVD callpc+0(FP), R3
JMP racefuncenter<>(SB)
// Common code for racefuncenter
// R3 = caller's return address
TEXT racefuncenter<>(SB), NOSPLIT, $0-0
// void __tsan_func_enter(ThreadState *thr, void *pc);
MOVD $__tsan_func_enter(SB), R1
MOVD g_racectx(g), R2
BL racecall<>(SB)
RET
// func runtime·racefuncexit()
// Called from instrumented code.
TEXT runtime·racefuncexit(SB), NOSPLIT, $0-0
// void __tsan_func_exit(ThreadState *thr);
MOVD $__tsan_func_exit(SB), R1
MOVD g_racectx(g), R2
JMP racecall<>(SB)
// Atomic operations for sync/atomic package.
// Load
TEXT sync∕atomic·LoadInt32(SB), NOSPLIT, $0-12
GO_ARGS
MOVD $__tsan_go_atomic32_load(SB), R1
BL racecallatomic<>(SB)
RET
TEXT sync∕atomic·LoadInt64(SB), NOSPLIT, $0-16
GO_ARGS
MOVD $__tsan_go_atomic64_load(SB), R1
BL racecallatomic<>(SB)
RET
TEXT sync∕atomic·LoadUint32(SB), NOSPLIT, $0-12
GO_ARGS
JMP sync∕atomic·LoadInt32(SB)
TEXT sync∕atomic·LoadUint64(SB), NOSPLIT, $0-16
GO_ARGS
JMP sync∕atomic·LoadInt64(SB)
TEXT sync∕atomic·LoadUintptr(SB), NOSPLIT, $0-16
GO_ARGS
JMP sync∕atomic·LoadInt64(SB)
TEXT sync∕atomic·LoadPointer(SB), NOSPLIT, $0-16
GO_ARGS
JMP sync∕atomic·LoadInt64(SB)
// Store
TEXT sync∕atomic·StoreInt32(SB), NOSPLIT, $0-12
GO_ARGS
MOVD $__tsan_go_atomic32_store(SB), R1
BL racecallatomic<>(SB)
RET
TEXT sync∕atomic·StoreInt64(SB), NOSPLIT, $0-16
GO_ARGS
MOVD $__tsan_go_atomic64_store(SB), R1
BL racecallatomic<>(SB)
RET
TEXT sync∕atomic·StoreUint32(SB), NOSPLIT, $0-12
GO_ARGS
JMP sync∕atomic·StoreInt32(SB)
TEXT sync∕atomic·StoreUint64(SB), NOSPLIT, $0-16
GO_ARGS
JMP sync∕atomic·StoreInt64(SB)
TEXT sync∕atomic·StoreUintptr(SB), NOSPLIT, $0-16
GO_ARGS
JMP sync∕atomic·StoreInt64(SB)
// Swap
TEXT sync∕atomic·SwapInt32(SB), NOSPLIT, $0-20
GO_ARGS
MOVD $__tsan_go_atomic32_exchange(SB), R1
BL racecallatomic<>(SB)
RET
TEXT sync∕atomic·SwapInt64(SB), NOSPLIT, $0-24
GO_ARGS
MOVD $__tsan_go_atomic64_exchange(SB), R1
BL racecallatomic<>(SB)
RET
TEXT sync∕atomic·SwapUint32(SB), NOSPLIT, $0-20
GO_ARGS
JMP sync∕atomic·SwapInt32(SB)
TEXT sync∕atomic·SwapUint64(SB), NOSPLIT, $0-24
GO_ARGS
JMP sync∕atomic·SwapInt64(SB)
TEXT sync∕atomic·SwapUintptr(SB), NOSPLIT, $0-24
GO_ARGS
JMP sync∕atomic·SwapInt64(SB)
// Add
TEXT sync∕atomic·AddInt32(SB), NOSPLIT, $0-20
GO_ARGS
MOVD $__tsan_go_atomic32_fetch_add(SB), R1
BL racecallatomic<>(SB)
// TSan performed fetch_add, but Go needs add_fetch.
MOVW add+8(FP), R0
MOVW ret+16(FP), R1
ADD R0, R1, R0
MOVW R0, ret+16(FP)
RET
TEXT sync∕atomic·AddInt64(SB), NOSPLIT, $0-24
GO_ARGS
MOVD $__tsan_go_atomic64_fetch_add(SB), R1
BL racecallatomic<>(SB)
// TSan performed fetch_add, but Go needs add_fetch.
MOVD add+8(FP), R0
MOVD ret+16(FP), R1
ADD R0, R1, R0
MOVD R0, ret+16(FP)
RET
TEXT sync∕atomic·AddUint32(SB), NOSPLIT, $0-20
GO_ARGS
JMP sync∕atomic·AddInt32(SB)
TEXT sync∕atomic·AddUint64(SB), NOSPLIT, $0-24
GO_ARGS
JMP sync∕atomic·AddInt64(SB)
TEXT sync∕atomic·AddUintptr(SB), NOSPLIT, $0-24
GO_ARGS
JMP sync∕atomic·AddInt64(SB)
// CompareAndSwap
TEXT sync∕atomic·CompareAndSwapInt32(SB), NOSPLIT, $0-17
GO_ARGS
MOVD $__tsan_go_atomic32_compare_exchange(SB), R1
BL racecallatomic<>(SB)
RET
TEXT sync∕atomic·CompareAndSwapInt64(SB), NOSPLIT, $0-25
GO_ARGS
MOVD $__tsan_go_atomic64_compare_exchange(SB), R1
BL racecallatomic<>(SB)
RET
TEXT sync∕atomic·CompareAndSwapUint32(SB), NOSPLIT, $0-17
GO_ARGS
JMP sync∕atomic·CompareAndSwapInt32(SB)
TEXT sync∕atomic·CompareAndSwapUint64(SB), NOSPLIT, $0-25
GO_ARGS
JMP sync∕atomic·CompareAndSwapInt64(SB)
TEXT sync∕atomic·CompareAndSwapUintptr(SB), NOSPLIT, $0-25
GO_ARGS
JMP sync∕atomic·CompareAndSwapInt64(SB)
// Common code for atomic operations. Calls R1.
TEXT racecallatomic<>(SB), NOSPLIT, $0
MOVD 24(R15), R5 // Address (arg1, after 2xBL).
// If we pass an invalid pointer to the TSan runtime, it will cause a
// "fatal error: unknown caller pc". So trigger a SEGV here instead.
MOVB (R5), R0
MOVD runtime·racearenastart(SB), R0
CMPUBLT R5, R0, racecallatomic_data // Before racearena start?
MOVD runtime·racearenaend(SB), R0
CMPUBLT R5, R0, racecallatomic_ok // Before racearena end?
racecallatomic_data:
MOVD runtime·racedatastart(SB), R0
CMPUBLT R5, R0, racecallatomic_ignore // Before racedata start?
MOVD runtime·racedataend(SB), R0
CMPUBGE R5, R0, racecallatomic_ignore // At or after racearena end?
racecallatomic_ok:
MOVD g_racectx(g), R2 // ThreadState *.
MOVD 8(R15), R3 // Caller PC.
MOVD R14, R4 // PC.
ADD $24, R15, R5 // Arguments.
// Tail call fails to restore R15, so use a normal one.
BL racecall<>(SB)
RET
racecallatomic_ignore:
// Call __tsan_go_ignore_sync_begin to ignore synchronization during
// the atomic op. An attempt to synchronize on the address would cause
// a crash.
MOVD R1, R6 // Save target function.
MOVD R14, R7 // Save PC.
MOVD $__tsan_go_ignore_sync_begin(SB), R1
MOVD g_racectx(g), R2 // ThreadState *.
BL racecall<>(SB)
MOVD R6, R1 // Restore target function.
MOVD g_racectx(g), R2 // ThreadState *.
MOVD 8(R15), R3 // Caller PC.
MOVD R7, R4 // PC.
ADD $24, R15, R5 // Arguments.
BL racecall<>(SB)
MOVD $__tsan_go_ignore_sync_end(SB), R1
MOVD g_racectx(g), R2 // ThreadState *.
BL racecall<>(SB)
RET
// func runtime·racecall(void(*f)(...), ...)
// Calls C function f from race runtime and passes up to 4 arguments to it.
// The arguments are never heap-object-preserving pointers, so we pretend there
// are no arguments.
TEXT runtime·racecall(SB), NOSPLIT, $0-0
MOVD fn+0(FP), R1
MOVD arg0+8(FP), R2
MOVD arg1+16(FP), R3
MOVD arg2+24(FP), R4
MOVD arg3+32(FP), R5
JMP racecall<>(SB)
// Switches SP to g0 stack and calls R1. Arguments are already set.
TEXT racecall<>(SB), NOSPLIT, $0-0
BL runtime·save_g(SB) // Save g for callbacks.
MOVD R15, R7 // Save SP.
MOVD g_m(g), R8 // R8 = thread.
MOVD m_g0(R8), R8 // R8 = g0.
CMPBEQ R8, g, call // Already on g0?
MOVD (g_sched+gobuf_sp)(R8), R15 // Switch SP to g0.
call: SUB $160, R15 // Allocate C frame.
BL R1 // Call C code.
MOVD R7, R15 // Restore SP.
RET // Return to Go.
// C->Go callback thunk that allows to call runtime·racesymbolize from C
// code. racecall has only switched SP, finish g->g0 switch by setting correct
// g. R2 contains command code, R3 contains command-specific context. See
// racecallback for command codes.
TEXT runtime·racecallbackthunk(SB), NOSPLIT|NOFRAME, $0
STMG R6, R15, 48(R15) // Save non-volatile regs.
BL runtime·load_g(SB) // Saved by racecall.
CMPBNE R2, $0, rest // raceGetProcCmd?
MOVD g_m(g), R2 // R2 = thread.
MOVD m_p(R2), R2 // R2 = processor.
MVC $8, p_raceprocctx(R2), (R3) // *R3 = ThreadState *.
LMG 48(R15), R6, R15 // Restore non-volatile regs.
BR R14 // Return to C.
rest: MOVD g_m(g), R4 // R4 = current thread.
MOVD m_g0(R4), g // Switch to g0.
SUB $24, R15 // Allocate Go argument slots.
STMG R2, R3, 8(R15) // Fill Go frame.
BL runtime·racecallback(SB) // Call Go code.
LMG 72(R15), R6, R15 // Restore non-volatile regs.
BR R14 // Return to C.