[release-branch.go1.13] runtime: fix darwin syscall performance regression

While understanding why syscall.Read is 2x slower on darwin/amd64, I found
out that, contrary to popular belief, the slowdown is not due to the migration
to use libSystem.dylib instead of direct SYSCALLs, i.e., CL 141639 (and #17490),
but due to a subtle change introduced in CL 141639.

Previously, syscall.Read used syscall.Syscall(SYS_READ), whose preamble called
runtime.entersyscall, but after CL 141639, syscall.Read changes to call
runtime.syscall_syscall instead, which in turn calls runtime.entersyscallblock
instead of runtime.entersyscall. And the entire 2x slow down can be attributed
to this change.

I think this is unnecessary as even though syscalls like Read might block, it
does not always block, so there is no need to handoff P proactively for each
Read. Additionally, we have been fine with not handing off P for each Read
prior to Go 1.12, so we probably don't need to change it. This changes restores
the pre-Go 1.12 behavior, where syscall preamble uses runtime.entersyscall,
and we rely on sysmon to take P back from g blocked in syscalls.

Updates #34712

Change-Id: If76e97b5a7040cf1c10380a567c4f5baec3121ba
Reviewed-on: https://go-review.googlesource.com/c/go/+/197938
Run-TryBot: Minux Ma <[email protected]>
TryBot-Result: Gobot Gobot <[email protected]>
Reviewed-by: Keith Randall <[email protected]>
Reviewed-by: Ian Lance Taylor <[email protected]>
(cherry picked from commit c1635ad)
Reviewed-on: https://go-review.googlesource.com/c/go/+/200103
Run-TryBot: Alexander Rakoczy <[email protected]>
Reviewed-by: Cherry Zhang <[email protected]>

Author: minux

src/runtime/sys_darwin.go

@@ -60,7 +60,7 @@ func libcCall(fn, arg unsafe.Pointer) int32 {
 //go:nosplit
 //go:cgo_unsafe_args
 func syscall_syscall(fn, a1, a2, a3 uintptr) (r1, r2, err uintptr) {
-	entersyscallblock()
+	entersyscall()
 	libcCall(unsafe.Pointer(funcPC(syscall)), unsafe.Pointer(&fn))
 	exitsyscall()
 	return
@@ -71,7 +71,7 @@ func syscall()
 //go:nosplit
 //go:cgo_unsafe_args
 func syscall_syscall6(fn, a1, a2, a3, a4, a5, a6 uintptr) (r1, r2, err uintptr) {
-	entersyscallblock()
+	entersyscall()
 	libcCall(unsafe.Pointer(funcPC(syscall6)), unsafe.Pointer(&fn))
 	exitsyscall()
 	return
@@ -82,7 +82,7 @@ func syscall6()
 //go:nosplit
 //go:cgo_unsafe_args
 func syscall_syscall6X(fn, a1, a2, a3, a4, a5, a6 uintptr) (r1, r2, err uintptr) {
-	entersyscallblock()
+	entersyscall()
 	libcCall(unsafe.Pointer(funcPC(syscall6X)), unsafe.Pointer(&fn))
 	exitsyscall()
 	return
@@ -93,7 +93,7 @@ func syscall6X()
 //go:nosplit
 //go:cgo_unsafe_args
 func syscall_syscallPtr(fn, a1, a2, a3 uintptr) (r1, r2, err uintptr) {
-	entersyscallblock()
+	entersyscall()
 	libcCall(unsafe.Pointer(funcPC(syscallPtr)), unsafe.Pointer(&fn))
 	exitsyscall()
 	return