bionic: Add ARM optimized strcpy()

Reference results of the experiments on Qualcomm MSM7x25 (524MHz):

[original C code]
             prc thr   usecs/call      samples   errors cnt/samp
size
strcpy_1k      1   1     14.56159           99        0     1000
1024

[ARM optimized code]
             prc thr   usecs/call      samples   errors cnt/samp
size
strcpy_1k      1   1      3.46653           99        0     1000
1024

The work was derived from ARM Ltd.

Change-Id: I906ac53bb7a7285e14693c77d3ce8d4ed6f98bfd
diff --git a/libc/Android.mk b/libc/Android.mk
index 39c63a2..71a8941 100644
--- a/libc/Android.mk
+++ b/libc/Android.mk
@@ -186,7 +186,6 @@
 	string/strcat.c \
 	string/strchr.c \
 	string/strcoll.c \
-	string/strcpy.c \
 	string/strcspn.c \
 	string/strdup.c \
 	string/strerror.c \
@@ -309,6 +308,7 @@
 	arch-arm/bionic/setjmp.S \
 	arch-arm/bionic/sigsetjmp.S \
 	arch-arm/bionic/strlen.c.arm \
+	arch-arm/bionic/strcpy.S \
 	arch-arm/bionic/syscall.S \
 	arch-arm/bionic/sigaction.c \
 	arch-arm/bionic/__sig_restorer.S \
@@ -357,6 +357,7 @@
 	arch-x86/string/strcmp_wrapper.S \
 	arch-x86/string/strncmp_wrapper.S \
 	arch-x86/string/strlen_wrapper.S \
+	string/strcpy.c \
 	bionic/pthread.c \
 	bionic/pthread-atfork.c \
 	bionic/pthread-timers.c \
@@ -394,6 +395,7 @@
 	string/strncmp.c \
 	string/memcmp.c \
 	string/strlen.c \
+	string/strcpy.c \
 	bionic/pthread.c \
 	bionic/pthread-atfork.c \
 	bionic/pthread-timers.c \