Add --opt=force-aligned-memory option.

This forces all vector loads/stores to be done assuming that the given
pointer is aligned to the vector size, thus allowing the use of sometimes
more-efficient instructions.  (If it isn't the case that the memory is
aligned, the program will fail!).

ispc.cpp

@@ -640,6 +640,7 @@ Opt::Opt() {
     unrollLoops = true;
     disableAsserts = false;
     disableFMA = false;
+    forceAlignedMemory = false;
     disableMaskAllOnOptimizations = false;
     disableHandlePseudoMemoryOps = false;
     disableBlendedMaskedStores = false;