Align the child stack in clone(2).

Also let clone(2) set the TLS for x86.

Also ensure we initialize the TLS before we clone(2) for all architectures.

Change-Id: Ie5fa4466e1c9ee116a281dfedef574c5ba60c0b5

libc/arch-x86_64/bionic/__bionic_clone.S

@@ -30,9 +30,6 @@
 
 // pid_t __bionic_clone(int flags, void* child_stack, pid_t* parent_tid, void* tls, pid_t* child_tid, int (*fn)(void*), void* arg);
 ENTRY(__bionic_clone)
-        # Align 'child_stack' to 16 bytes.
-        andq    $~15, %rsi
-
         # Copy 'fn' and 'arg' onto the child stack.
         movq    %r9, -16(%rsi)  # fn
         movq    8(%rsp), %rax   # Read 'arg'.

libc/arch-x86_64/bionic/__set_tls.c

@@ -32,7 +32,5 @@
 extern int __arch_prctl(int, unsigned long);
 
 int __set_tls(void* ptr) {
-  // We also need to write the location of the tls to ptr[0].
-  *(void**) ptr = ptr;
   return __arch_prctl(ARCH_SET_FS, (uintptr_t) ptr);
 }