Replaced more possible bit shifting into signed bit of stride values.

Leftover changes from a6766c94.

Use multiplication instead, allowing the compiler to optimize to
bitshifts if it believe it to be safe.

Partly solves github issue #18.

lib/c64x/c64xfrag.c

@@ -436,7 +436,7 @@ void oc_state_loop_filter_frag_rows_c64x(const oc_theora_state *_state,
           loop_filter_h(ref+8,ystride,ll);
         }
         if(fragi+nhfrags<fragi_bot&&!frags[fragi+nhfrags].coded){
-          loop_filter_v(ref+(ystride<<3),ystride,ll);
+          loop_filter_v(ref+(ystride*8),ystride,ll);
         }
       }
       fragi++;

lib/x86/mmxstate.c

@@ -146,7 +146,7 @@ void oc_state_loop_filter_frag_rows_mmx(const oc_theora_state *_state,
           OC_LOOP_FILTER_H(OC_LOOP_FILTER8_MMX,ref+8,ystride,ll);
         }
         if(fragi+nhfrags<fragi_bot&&!frags[fragi+nhfrags].coded){
-          OC_LOOP_FILTER_V(OC_LOOP_FILTER8_MMX,ref+(ystride<<3),ystride,ll);
+          OC_LOOP_FILTER_V(OC_LOOP_FILTER8_MMX,ref+(ystride*8),ystride,ll);
         }
       }
       fragi++;
@@ -213,7 +213,7 @@ void oc_state_loop_filter_frag_rows_mmxext(const oc_theora_state *_state,
           OC_LOOP_FILTER_H(OC_LOOP_FILTER8_MMXEXT,ref+8,ystride,_bv);
         }
         if(fragi+nhfrags<fragi_bot&&!frags[fragi+nhfrags].coded){
-          OC_LOOP_FILTER_V(OC_LOOP_FILTER8_MMXEXT,ref+(ystride<<3),ystride,_bv);
+          OC_LOOP_FILTER_V(OC_LOOP_FILTER8_MMXEXT,ref+(ystride*8),ystride,_bv);
         }
       }
       fragi++;

lib/x86_vc/mmxstate.c

@@ -157,7 +157,7 @@ void oc_state_loop_filter_frag_rows_mmx(const oc_theora_state *_state,
           OC_LOOP_FILTER_H_MMX(ref+8,ystride,_bv);
         }
         if(fragi+nhfrags<fragi_bot&&!frags[fragi+nhfrags].coded){
-          OC_LOOP_FILTER_V_MMX(ref+(ystride<<3),ystride,_bv);
+          OC_LOOP_FILTER_V_MMX(ref+(ystride*8),ystride,_bv);
         }
 #undef PIX
 #undef YSTRIDE3