Add mempcpy implementation

A GNU extension that returns dst + size instead of dst. It's a nice
shortcut when copying strings with a known size or back-to-back blocks
and you have to do it often.

May of course be called directly or alternately through
__builtin_mempcpy in some compiler versions.

For ASM on native targets, it is implemented as an alternate entrypoint
to memcpy which adds minimal code and overhead.

Change-Id: I4cbb3483f6df3c1007247fe0a95fd7078737462b

1 files changed, 10 insertions, 1 deletions

diff --git a/firmware/asm/mips/memcpy.S b/firmware/asm/mips/memcpy.S
index edbf5ac5eb..ec1625bb4f 100644
--- a/firmware/asm/mips/memcpy.S
+++ b/firmware/asm/mips/memcpy.S
@@ -43,8 +43,16 @@
 
     .global    memcpy
     .type      memcpy, %function
+    .global    mempcpy
+    .type      mempcpy, %function
     
     .set       noreorder
+mempcpy:
+    slti       t0, a2, 8                # Less than 8?
+    bne        t0, zero, last8
+     addu      v0, a0, a2               # exit value = s1 + n
+    b          1f
+     xor       t0, a1, a0               # Find a0/a1 displacement (fill delay)
 
 memcpy:
     slti       t0, a2, 8                # Less than 8?
@@ -52,7 +60,8 @@ memcpy:
     move       v0, a0                   # Setup exit value before too late
 
     xor        t0, a1, a0               # Find a0/a1 displacement
-    andi       t0, 0x3
+
+1:  andi       t0, 0x3
     bne        t0, zero, shift          # Go handle the unaligned case
     subu       t1, zero, a1
     andi       t1, 0x3                  # a0/a1 are aligned, but are we