Use proper unaligned access attributes

Instead of using packed attribute hack, just use aligned attribute. It
improves code generation on armv6 and armv7, and slightly improves code
generation on aarch64. GCC generates identical code to regular aligned
access on ARMv6 for all versions between 4.5 and trunk, except GCC 5
which is buggy and generates the same (bad) code as packed access:
https://gcc.godbolt.org/z/hq37rz7sb

tests/fuzz/fuzz.h

@@ -26,8 +26,7 @@
  * @param MEM_FORCE_MEMORY_ACCESS:
  *        This flag controls how the zstd library accesses unaligned memory.
  *        It can be undefined, or 0 through 2. If it is undefined, it selects
- *        the method to use based on the compiler. If testing with UBSAN set
- *        MEM_FORCE_MEMORY_ACCESS=0 to use the standard compliant method.
+ *        the method to use based on the compiler.
  * @param FUZZING_BUILD_MODE_UNSAFE_FOR_PRODUCTION
  *        This is the canonical flag to enable deterministic builds for fuzzing.
  *        Changes to zstd for fuzzing are gated behind this define.