runtime: don't try to free OS-created signal stacks

Android's libc creates a signal stack for every thread it creates. In
Go, minitSignalStack picks up this existing signal stack and puts it
in m.gsignal.stack. However, if we later try to exit a thread (because
a locked goroutine is exiting), we'll attempt to stackfree this
libc-allocated signal stack and panic.

Fix this by clearing gsignal.stack when we unminitSignals in such a
situation.

This should fix the Android build, which is currently broken.

Change-Id: Ieea8d72ef063d22741c54c9daddd8bb84926a488
Reviewed-on: https://go-review.googlesource.com/70130
Reviewed-by: David Crawshaw <[email protected]>
Reviewed-by: Ian Lance Taylor <[email protected]>
Run-TryBot: David Crawshaw <[email protected]>
TryBot-Result: Gobot Gobot <[email protected]>

Author: aclements

src/runtime/proc.go

@@ -1268,7 +1268,13 @@ func mexit(osStack bool) {
 	unminit()
 
 	// Free the gsignal stack.
-	if m.gsignal != nil {
+	//
+	// If the signal stack was created outside Go, then gsignal
+	// will be non-nil, but unminitSignals set stack.lo to 0
+	// (e.g., Android's libc creates all threads with a signal
+	// stack, so it's possible for Go to exit them but not control
+	// the signal stack).
+	if m.gsignal != nil && m.gsignal.stack.lo != 0 {
 		stackfree(m.gsignal.stack)
 	}
 

src/runtime/signal_unix.go

@@ -744,6 +744,10 @@ func unminitSignals() {
 	if getg().m.newSigstack {
 		st := stackt{ss_flags: _SS_DISABLE}
 		sigaltstack(&st, nil)
+	} else {
+		// We got the signal stack from someone else. Clear it
+		// so we don't get confused.
+		getg().m.gsignal.stack = stack{}
 	}
 }