Align all arrays in mjData's buffer to 64 byte boundaries.

The introduction of additional mjData arrays for the implicit integrator caused in a shift in data alignment in memory. In certain combinations of machines and models, this change resulted in ~5% performance regression that appears to be attributable to L2 cache misses.

This change aligns all arrays in mjData to the boundary of a typical L2 cache line size. Benchmarking shows that this resolves the performance regression on a Broadwell machine.

PiperOrigin-RevId: 451652783
Change-Id: I4ed2c2d03453a83a0b4a9b8e224e2fd7c0dadea9

introspect/functions.py

@@ -4659,7 +4659,7 @@ FUNCTIONS: Mapping[str, FunctionDecl] = dict([
                  type=ValueType(name='size_t'),
              ),
          ),
-         doc='Allocate memory; byte-align on 8; pad size to multiple of 8.',
+         doc='Allocate memory; byte-align on 64; pad size to multiple of 64.',
      )),
     ('mju_free',
      FunctionDecl(