cmd/go, cmd/ld, runtime, os/user: TLS emulation for android

Based on cl/69170045 by Elias Naur.

There are currently several schemes for acquiring a TLS
slot to save the g register. None of them appear to work
for android. The closest are linux and darwin.

Linux uses a linker TLS relocation. This is not supported
by the android linker.

Darwin uses a fixed offset, and calls pthread_key_create
until it gets the slot it wants. As the runtime loads
late in the android process lifecycle, after an
arbitrary number of other libraries, we cannot rely on
any particular slot being available.

So we call pthread_key_create, take the first slot we are
given, and put it in runtime.tlsg, which we turn into a
regular variable in cmd/ld.

Makes android/arm cgo binaries work.

LGTM=minux
R=elias.naur, minux, dave, josharian
CC=golang-codereviews
https://golang.org/cl/106380043

src/cmd/go/build.go

@@ -2262,13 +2262,14 @@ func (b *builder) cgo(p *Package, cgoExe, obj string, gccfiles, gxxfiles, mfiles
 
 	linkobj = append(linkobj, p.SysoFiles...)
 	dynobj := obj + "_cgo_.o"
-	if goarch == "arm" && goos == "linux" { // we need to use -pie for Linux/ARM to get accurate imported sym
+	pie := goarch == "arm" && (goos == "linux" || goos == "android")
+	if pie { // we need to use -pie for Linux/ARM to get accurate imported sym
 		cgoLDFLAGS = append(cgoLDFLAGS, "-pie")
 	}
 	if err := b.gccld(p, dynobj, cgoLDFLAGS, linkobj); err != nil {
 		return nil, nil, err
 	}
-	if goarch == "arm" && goos == "linux" { // but we don't need -pie for normal cgo programs
+	if pie { // but we don't need -pie for normal cgo programs
 		cgoLDFLAGS = cgoLDFLAGS[0 : len(cgoLDFLAGS)-1]
 	}
 

src/cmd/ld/data.c

@@ -160,6 +160,10 @@ relocsym(LSym *s)
 		if(r->sym != S && r->sym->type != STLSBSS && !r->sym->reachable)
 			diag("unreachable sym in relocation: %s %s", s->name, r->sym->name);
 
+		// Android emulates runtime.tlsg as a regular variable.
+		if (r->type == R_TLS && strcmp(goos, "android") == 0)
+			r->type = R_ADDR;
+
 		switch(r->type) {
 		default:
 			o = 0;

src/cmd/ld/elf.c

@@ -776,7 +776,8 @@ elfshbits(Section *sect)
 	if(sect->rwx & 2)
 		sh->flags |= SHF_WRITE;
 	if(strcmp(sect->name, ".tbss") == 0) {
-		sh->flags |= SHF_TLS;
+		if(strcmp(goos, "android") != 0)
+			sh->flags |= SHF_TLS; // no TLS on android
 		sh->type = SHT_NOBITS;
 	}
 	if(linkmode != LinkExternal)

src/cmd/ld/lib.c

@@ -226,6 +226,10 @@ loadlib(void)
 			linkmode = LinkExternal;
 		else
 			linkmode = LinkInternal;
+
+		// Force external linking for android.
+		if(strcmp(goos, "android") == 0)
+			linkmode = LinkExternal;
 	}
 
 	if(linkmode == LinkInternal) {

src/cmd/ld/pobj.c

@@ -139,7 +139,7 @@ main(int argc, char *argv[])
 	if(HEADTYPE == -1)
 		HEADTYPE = headtype(goos);
 	ctxt->headtype = HEADTYPE;
-	if (headstring == nil)
+	if(headstring == nil)
 		headstring = headstr(HEADTYPE);
 
 	archinit();

src/cmd/ld/symtab.c

@@ -204,7 +204,12 @@ asmelfsym(void)
 			diag("missing section for %s", s->name);
 			errorexit();
 		}
-		putelfsyment(putelfstr(s->name), 0, s->size, (STB_LOCAL<<4)|STT_TLS, s->sect->elfsect->shnum, 0);
+		if (strcmp(goos, "android") == 0) {
+			// Android emulates runtime.tlsg as a regular variable.
+			putelfsyment(putelfstr(s->name), 0, s->size, (STB_LOCAL<<4)|STT_OBJECT, s->sect->elfsect->shnum, 0);
+		} else {
+			putelfsyment(putelfstr(s->name), 0, s->size, (STB_LOCAL<<4)|STT_TLS, s->sect->elfsect->shnum, 0);
+		}
 		s->elfsym = numelfsym++;
 	}
 

src/pkg/os/user/lookup_stubs.go

@@ -2,7 +2,7 @@
 // Use of this source code is governed by a BSD-style
 // license that can be found in the LICENSE file.
 
-// +build !cgo,!windows,!plan9
+// +build !cgo,!windows,!plan9 android
 
 package user
 

src/pkg/os/user/lookup_unix.go

@@ -2,7 +2,7 @@
 // Use of this source code is governed by a BSD-style
 // license that can be found in the LICENSE file.
 
-// +build darwin dragonfly freebsd linux netbsd openbsd solaris
+// +build darwin dragonfly freebsd !android,linux netbsd openbsd solaris
 // +build cgo
 
 package user

src/pkg/runtime/asm_arm.s

@@ -40,10 +40,12 @@ TEXT _rt0_go(SB),NOSPLIT,$-4
 	MOVW	_cgo_init(SB), R4
 	CMP	$0, R4
 	B.EQ	nocgo
-	BL		runtime·save_g(SB);
-	MOVW	g, R0 // first argument of _cgo_init is g
-	MOVW	$setg_gcc<>(SB), R1 // second argument is address of save_g
-	BL		(R4) // will clobber R0-R3
+	MRC     15, 0, R0, C13, C0, 3 	// load TLS base pointer
+	MOVW 	R0, R3 			// arg 3: TLS base pointer
+	MOVW 	$runtime·tlsg(SB), R2 	// arg 2: tlsg
+	MOVW	$setg_gcc<>(SB), R1 	// arg 1: setg
+	MOVW	g, R0 			// arg 0: G
+	BL	(R4) // will clobber R0-R3
 
 nocgo:
 	// update stackguard after _cgo_init
@@ -688,42 +690,6 @@ _eqnext:
 	MOVB	R7, v+16(FP)
 	RET
 
-// We have to resort to TLS variable to save g(R10).
-// One reason is that external code might trigger
-// SIGSEGV, and our runtime.sigtramp don't even know we
-// are in external code, and will continue to use R10,
-// this might as well result in another SIGSEGV.
-// Note: all three functions will clobber R0, and the last
-// two can be called from 5c ABI code.
-
-// save_g saves the g register into pthread-provided
-// thread-local memory, so that we can call externally compiled
-// ARM code that will overwrite those registers.
-// NOTE: runtime.gogo assumes that R1 is preserved by this function.
-//       runtime.mcall assumes this function only clobbers R0 and R11.
-TEXT runtime·save_g(SB),NOSPLIT,$0
-	MRC		15, 0, R0, C13, C0, 3 // fetch TLS base pointer
-	// $runtime.tlsg(SB) is a special linker symbol.
-	// It is the offset from the TLS base pointer to our
-	// thread-local storage for g.
-	MOVW	$runtime·tlsg(SB), R11
-	ADD	R11, R0
-	MOVW	g, 0(R0)
-	RET
-
-// load_g loads the g register from pthread-provided
-// thread-local memory, for use after calling externally compiled
-// ARM code that overwrote those registers.
-TEXT runtime·load_g(SB),NOSPLIT,$0
-	MRC		15, 0, R0, C13, C0, 3 // fetch TLS base pointer
-	// $runtime.tlsg(SB) is a special linker symbol.
-	// It is the offset from the TLS base pointer to our
-	// thread-local storage for g.
-	MOVW	$runtime·tlsg(SB), R11
-	ADD	R11, R0
-	MOVW	0(R0), g
-	RET
-
 // void setg_gcc(G*); set g called from gcc.
 TEXT setg_gcc<>(SB),NOSPLIT,$0
 	MOVW	R0, g

src/pkg/runtime/cgo/cgo.go

@@ -14,7 +14,8 @@ package cgo
 #cgo darwin LDFLAGS: -lpthread
 #cgo dragonfly LDFLAGS: -lpthread
 #cgo freebsd LDFLAGS: -lpthread
-#cgo linux LDFLAGS: -lpthread
+#cgo android LDFLAGS: -llog
+#cgo !android,linux LDFLAGS: -lpthread
 #cgo netbsd LDFLAGS: -lpthread
 #cgo openbsd LDFLAGS: -lpthread
 #cgo windows LDFLAGS: -lm -mthreads

src/pkg/runtime/cgo/gcc_android_arm.c

+// Copyright 2014 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.
+
+#include <android/log.h>
+#include <pthread.h>
+#include <signal.h>
+#include <stdio.h>
+#include <sys/limits.h>
+#include "libcgo.h"
+
+#define magic1 (0x23581321U)
+
+// PTHREAD_KEYS_MAX has been added to sys/limits.h at head in bionic:
+// https://android.googlesource.com/platform/bionic/+/master/libc/include/sys/limits.h
+// TODO(crawshaw): remove this definition when a new NDK is released.
+#define PTHREAD_KEYS_MAX 128
+
+// inittls allocates a thread-local storage slot for g.
+//
+// It finds the first available slot using pthread_key_create and uses
+// it as the offset value for runtime.tlsg.
+static void
+inittls(void **tlsg, void **tlsbase)
+{
+	pthread_key_t k;
+	int i, err;
+
+	err = pthread_key_create(&k, nil);
+	if(err != 0) {
+		fprintf(stderr, "runtime/cgo: pthread_key_create failed: %d\n", err);
+		__android_log_print(ANDROID_LOG_FATAL, "runtime/cgo", "pthread_key_create failed: %d", err);
+		abort();
+	}
+	pthread_setspecific(k, (void*)magic1);
+	for (i=0; i<PTHREAD_KEYS_MAX; i++) {
+		if (*(tlsbase+i) == (void*)magic1) {
+			*tlsg = (void*)(i*sizeof(void *));
+			pthread_setspecific(k, 0);
+			return;
+		}
+	}
+	fprintf(stderr, "runtime/cgo: could not find pthread key\n");
+	__android_log_print(ANDROID_LOG_FATAL, "runtime/cgo", "could not find pthread key");
+	abort();
+}
+
+void (*x_cgo_inittls)(void **tlsg, void **tlsbase) = inittls;

src/pkg/runtime/cgo/gcc_linux_arm.c

@@ -9,21 +9,7 @@
 
 static void *threadentry(void*);
 
-static void (*setg_gcc)(void*);
-
-void
-x_cgo_init(G *g, void (*setg)(void*))
-{
-	pthread_attr_t attr;
-	size_t size;
-
-	setg_gcc = setg;
-	pthread_attr_init(&attr);
-	pthread_attr_getstacksize(&attr, &size);
-	g->stackguard = (uintptr)&attr - size + 4096;
-	pthread_attr_destroy(&attr);
-}
-
+void (*setg_gcc)(void*);
 
 void
 _cgo_sys_thread_start(ThreadStart *ts)
@@ -50,8 +36,7 @@ _cgo_sys_thread_start(ThreadStart *ts)
 	pthread_sigmask(SIG_SETMASK, &oset, nil);
 
 	if (err != 0) {
-		fprintf(stderr, "runtime/cgo: pthread_create failed: %s\n", strerror(err));
-		abort();
+		fatalf("pthread_create failed: %s", strerror(err));
 	}
 }
 
@@ -75,3 +60,22 @@ threadentry(void *v)
 	crosscall_arm1(ts.fn, setg_gcc, (void*)ts.g);
 	return nil;
 }
+
+void (*x_cgo_inittls)(void **tlsg, void **tlsbase);
+
+void
+x_cgo_init(G *g, void (*setg)(void*), void **tlsg, void **tlsbase)
+{
+	pthread_attr_t attr;
+	size_t size;
+
+	setg_gcc = setg;
+	pthread_attr_init(&attr);
+	pthread_attr_getstacksize(&attr, &size);
+	g->stackguard = (uintptr)&attr - size + 4096;
+	pthread_attr_destroy(&attr);
+
+	if (x_cgo_inittls) {
+		x_cgo_inittls(tlsg, tlsbase);
+	}
+}

src/pkg/runtime/tls_arm.s

+// Copyright 2014 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.
+
+#include "zasm_GOOS_GOARCH.h"
+#include "funcdata.h"
+#include "../../cmd/ld/textflag.h"
+
+// We have to resort to TLS variable to save g(R10).
+// One reason is that external code might trigger
+// SIGSEGV, and our runtime.sigtramp don't even know we
+// are in external code, and will continue to use R10,
+// this might as well result in another SIGSEGV.
+// Note: both functions will clobber R0 and R11 and
+// can be called from 5c ABI code.
+
+// On android, runtime.tlsg is a normal variable.
+// TLS offset is computed in x_cgo_inittls.
+
+// save_g saves the g register into pthread-provided
+// thread-local memory, so that we can call externally compiled
+// ARM code that will overwrite those registers.
+// NOTE: runtime.gogo assumes that R1 is preserved by this function.
+//       runtime.mcall assumes this function only clobbers R0 and R11.
+TEXT runtime·save_g(SB),NOSPLIT,$0
+	MRC		15, 0, R0, C13, C0, 3 // fetch TLS base pointer
+	// $runtime.tlsg(SB) is a special linker symbol.
+	// It is the offset from the TLS base pointer to our
+	// thread-local storage for g.
+#ifdef GOOS_android
+	MOVW	runtime·tlsg(SB), R11
+#else
+	MOVW	$runtime·tlsg(SB), R11
+#endif
+	ADD	R11, R0
+	MOVW	g, 0(R0)
+	RET
+
+// load_g loads the g register from pthread-provided
+// thread-local memory, for use after calling externally compiled
+// ARM code that overwrote those registers.
+TEXT runtime·load_g(SB),NOSPLIT,$0
+	MRC		15, 0, R0, C13, C0, 3 // fetch TLS base pointer
+	// $runtime.tlsg(SB) is a special linker symbol.
+	// It is the offset from the TLS base pointer to our
+	// thread-local storage for g.
+#ifdef GOOS_android
+	MOVW	runtime·tlsg(SB), R11
+#else
+	MOVW	$runtime·tlsg(SB), R11
+#endif
+	ADD	R11, R0
+	MOVW	0(R0), g
+	RET