Implement blm and bls using an inline loop.

This replaces a call to memmove() in libc.  That was working for me,
but it can fail because EM programs don't always link to libc.

blm and bls only need to copy aligned words.  They don't need to copy
bytes, and they don't need to copy between overlapping buffers, as
memmove() does.  So the new loop is simpler than memmove().

diff --git a/mach/powerpc/ncg/table b/mach/powerpc/ncg/table
index 1ba20a8..569f3dd 100644 (file)
--- a/mach/powerpc/ncg/table
+++ b/mach/powerpc/ncg/table
@@ -289,6 +289,7 @@ INSTRUCTIONS
   andisX "andis." GPR:wo:cc, GPR:ro, CONST:ro.
   b               LABEL:ro.
   bc              CONST:ro, CONST:ro, LABEL:ro.
+    bdnz          LABEL:ro.
     beq           LABEL:ro.
     bne           LABEL:ro.
     bgt           LABEL:ro.
@@ -362,6 +363,7 @@ INSTRUCTIONS
   rlwinm          GPR:wo, GPR:ro, CONST:ro, CONST:ro, CONST:ro.
     extlwi        GPR:wo, GPR:ro, CONST:ro, CONST:ro.
     extrwi        GPR:wo, GPR:ro, CONST:ro, CONST:ro.
+    srwi          GPR:wo, GPR:ro, CONST:ro.
   slw             GPR:wo, GPR:ro, GPR:ro.
   subf            GPR:wo, GPR:ro, GPR:ro.
   sraw            GPR:wo, GPR:ro, GPR:ro cost(4, 2).
@@ -1930,24 +1932,22 @@ PATTERNS
                        b {LABEL, ".ret"}
 
        pat blm                            /* Block move constant length */
-               with REG REG STACK
-                       uses REG
-                       gen
-                               move {CONST, $1}, %a
-                               stwu %a, {IND_RC_W, SP, 0-4}
-                               stwu %2, {IND_RC_W, SP, 0-4}
-                               stwu %1, {IND_RC_W, SP, 0-4}
-                               bl {LABEL, "_memmove"}
-                               addi SP, SP, {CONST, 12}
+               leaving
+                       loc $1
+                       bls
 
        pat bls                            /* Block move variable length */
-               with REG REG REG STACK
+               with REG REG REG
+                       uses reusing %1, REG, REG={CONST_0000_7FFF, 0}
                        gen
-                               stwu %1, {IND_RC_W, SP, 0-4}
-                               stwu %3, {IND_RC_W, SP, 0-4}
-                               stwu %2, {IND_RC_W, SP, 0-4}
-                               bl {LABEL, "_memmove"}
-                               addi SP, SP, {CONST, 12}
+                               /* Wrong if size is zero */
+                               srwi %1, %1, {CONST, 2}
+                               mtspr CTR, %1
+                               1:
+                               lwzx %a, %3, %b
+                               stwx %a, %2, %b
+                               addi %b, %b, {CONST, 4}
+                               bdnz {LABEL, "1b"}
 
        pat csa                            /* Array-lookup switch */
                with STACK