Make proto2 benchmarks use ParsePartial, to make results vs upb more comparable.

ParsePartial does not check that required fields are set, which upb
also currently does not do.

This sped up proto2 reflection by 40%, but upb is still 7x faster in
a comparable test (instead of 11x, like before).

benchmarks/parsetostruct.proto2_compiled.cc

@@ -29,7 +29,7 @@ static void cleanup()
 
 static size_t run(int i)
 {
-  if(!msg[i%NUM_MESSAGES].ParseFromString(str)) {
+  if(!msg[i%NUM_MESSAGES].ParsePartialFromString(str)) {
     fprintf(stderr, "Error parsing with proto2.\n");
     return 0;
   }

benchmarks/parsetostruct.proto2_table.cc

@@ -38,7 +38,7 @@ static void cleanup()
 
 static size_t run(int i)
 {
-  if(!msg[i%NUM_MESSAGES]->ParseFromString(str)) {
+  if(!msg[i%NUM_MESSAGES]->ParsePartialFromString(str)) {
     fprintf(stderr, "Error parsing with proto2.\n");
     return 0;
   }