timing.hpp: Casting to atomic uint64_t before Running/last_ assignment

[spchamp]

This patch adds a cast for an assignment to an atomic integer value in timing.hpp

Updated: This patch serves as a duplicate or alternate approach to how the issue is addressed in Pull #3836

Previous to the patch, compiling Rubinius v5.0 with clang 11 on FreeBSD 11.1:

4: CXX machine/accessor_primitives.cpp
In file included from machine/accessor_primitives.cpp:9:
In file included from /usr/home/u1000/wk/dist_wk/ports_devo/lang/rubinius/work/rubinius-5.0/machine/includes.hpp:15:
In file included from ./machine/class/block_environment.hpp:12:
In file included from /usr/home/u1000/wk/dist_wk/ports_devo/lang/rubinius/work/rubinius-5.0/machine/class/variable_scope.hpp:9:
In file included from /usr/home/u1000/wk/dist_wk/ports_devo/lang/rubinius/work/rubinius-5.0/machine/class/fiber.hpp:13:
/usr/home/u1000/wk/dist_wk/ports_devo/lang/rubinius/work/rubinius-5.0/machine/diagnostics/timing.hpp:116:25: error: incompatible integer to pointer conversion assigning to 'std::atomic<uint>
      if(last_) last_ = run;
                        ^~~
In file included from machine/bytecode_verifier.cpp:17:
/usr/home/u1000/wk/dist_wk/ports_devo/lang/rubinius/work/rubinius-5.0/machine/diagnostics/timing.hpp:116:25: error: incompatible integer to pointer conversion assigning to 'std::atomic<uint>
      if(last_) last_ = run;
                        ^~~

---

1. Is this pull-request complete?

• Yes, this pull-request is ready to be reviewed and merged.
• No, this pull-request is a work-in-progress.

2. Does this pull request fix an issue with an existing feature or introduce a new feature?

• It fixes an issue with an existing features.
• It introduces a new features.

3. Does this pull-request include tests?

• Yes, it includes tests.
• No, it does not include tests because the tests already exist.
• No, it does not include tests because tests are too difficult to write.
• No, it does not include tests because ...