libc/x86: ensure the stack 16-byte aligned when tasks created

Currently Renderscript sample code RsBalls crashed on x86 when SSE2
enabled. The root cause is that the stack was not 16-byte aligned
from the beginning when the processes/threads were created, so the
RsBalls crashed when SSE2 instructions tried to access the variables
on the stack.

- For the thread created by fork():
Its stack alignment is determined by crtbegin_{dynamic, static}.S

- For the thread created by pthread_create():
Its stack alignment is determined by clone.S. __thread_entry( ) is
a standard C function. In order to have its stack be aligned with
16 byte properly, __thread_entry() needs the stack with following
layout when it is called:
layout #1 (correct)
--------------
|            |
-------------- <--ESP (ECX - 20)
| ret EIP    |
-------------- <--ECX - 16
| arg0       |
-------------- <--ECX - 12
| arg1       |
-------------- <--ECX - 8
| arg2       |
-------------- <--ECX - 4
| unused     |
-------------- <--ECX (16-byte boundary)

But it has following layout for now:
layout #2: (incorrect)
--------------
|            |
-------------- <--ESP (ECX - 16)
| unused     |
-------------- <--ECX - 12
| arg0       |
-------------- <--ECX - 8
| arg1       |
-------------- <--ECX - 4
| arg2       |
-------------- <--ECX (16-byte boundary)

Fixed in this patch.

Change-Id: Ibe01f64db14be14033c505d854c73033556ddaa8
Signed-off-by: Michael Liao <michael.liao@intel.com>
Signed-off-by: H.J. Lu <hongjiu.lu@intel.com>
Signed-off-by: Jack Ren <jack.ren@intel.com>
Signed-off-by: Bruce Beare <bruce.j.beare@intel.com>

libc/arch-x86/bionic/clone.S

@@ -13,16 +13,21 @@ __pthread_clone:
         pushl   %ebx
         pushl   %ecx
         movl    16(%esp), %ecx
-        movl    20(%esp), %ebx
+
+        # save tls
+        movl    %ecx, %ebx
+        # 16-byte alignment on child stack
+        andl    $~15, %ecx
 
         # insert arguments onto the child stack
         movl    12(%esp), %eax
-        movl    %eax, -12(%ecx)
+        movl    %eax, -16(%ecx)
         movl    24(%esp), %eax
-        movl    %eax, -8(%ecx)
+        movl    %eax, -12(%ecx)
-        movl    %ecx, -4(%ecx)
+        movl    %ebx, -8(%ecx)
 
         subl    $16, %ecx
+        movl    20(%esp), %ebx
         movl    $__NR_clone, %eax
         int     $0x80
         test    %eax, %eax
@@ -40,7 +45,7 @@ __pthread_clone:
         # we're in the child thread now, call __thread_entry
         # with the appropriate arguments on the child stack
         # we already placed most of them
-        jmp     __thread_entry
+        call    __thread_entry
         hlt
 
 2:

libc/arch-x86/bionic/crtbegin_dynamic.S

@@ -51,6 +51,8 @@
 #
 _start:	
         mov     %esp, %eax
+        # before push arguments, align the stack to a 16 byte boundary
+        andl    $~15, %esp
         mov     $1f, %edx
         pushl   %edx
         mov     $0f, %edx

libc/arch-x86/bionic/crtbegin_static.S

@@ -51,6 +51,8 @@
 #
 _start:	
         mov     %esp, %eax
+        # before push arguments, align the stack to a 16 byte boundary
+        andl    $~15, %esp
         mov     $1f, %edx
         pushl   %edx
         mov     $0f, %edx