a lot simpler VLA code

Author: Philip <pipcet@gmail.com>
    Our VLA code can be made a lot simpler (simple enough for
    even me to understand it) by giving up on the optimization idea, which
    is very tempting. There's a patch to do that attached, feel free to
    test and commit it if you like. (It passes all the tests, at least

x86_64-gen.c

@@ -2240,11 +2240,7 @@ ST_FUNC void gen_vla_alloc(CType *type, int align) {
     /* We align to 16 bytes rather than align */
     /* and ~15, %rsp */
     o(0xf0e48348);
-    /* mov %rsp, r */
-    o(0x8948);
-    o(0xe0 | REG_VALUE(r));
     vpop();
-    vset(type, r, 0);
 #endif
 }