[llvm] [NVPTX] Improve device function byval parameter lowering (PR #129188)

[Artem Belevich via llvm-commits]

Artem-B wrote:

> Unfortunately the criteria for determining which case is possible are not correct, leading to miscompilations (https://godbolt.org/z/Gq1fP7a3G). 

Can you elaborate on what exactly is incorrect about the example?
AFAICT, the code is still valid, even if ptxas itself has to make a copy: https://godbolt.org/z/ffd3d3G6z

My understanding is that ld.param from the parameter address is still legal and that we assumed that ptxas is smart enough to avoid local copies. Even if it does not, the code is still valid, even if it may be suboptimal.

https://github.com/llvm/llvm-project/pull/129188