Any time there is a CUDA device class that has virtual methods,
nvlink will issue a warning if the resulting library is linked
into code that does not use that method, so we basically have
to suppress these warnings for any execution environment class
with virtual methods.
- Reducing the stack allocation for CUDA for the BIH unit test
- Adding changes from Ken's review
- Suppress ptxas stack size warning for BoundingIntervalHierarchy
The PTX pass has pretty bad warning controls and we can't suppress
specific warnings for a single function. So instead we suppress
on dashboard the ptx warning for functions we know work
Sandia National Laboratories recently changed management from the
Sandia Corporation to the National Technology & Engineering Solutions
of Sandia, LLC (NTESS). The copyright statements need to be updated
accordingly.