[PATCH] D66862: Make lround builtin constexpr (and others)

[Zoe Carver via Phabricator via cfe-commits]

zoecarver marked 3 inline comments as done.
zoecarver added inline comments.


================
Comment at: clang/lib/AST/ExprConstant.cpp:9612
 
+  case Builtin::BIlround:
+  case Builtin::BI__builtin_lround: {
----------------
rsmith wrote:
> It's non-conforming to accept calls to these non-`__builtin` functions in constant expression evaluation, but it's fine to constant-fold them. Please issue a `CCEDiag` diagnostic for them (take a look at what we do for `Builtin::BIstrlen` or `Builtin::BIstrcmp` for an example; it would probably be reasonable to factor out a local lambda to issue suitable diagnostics for this case).
> 
> (For example, we are required to produce a diagnostic for `constexpr long n = lround(1.2);`, but we are permitted to constant-fold `long n = lround(1.2);` and emit it with static initialization.)
Thank you for pointing this out. I will do that in a separate patch.


================
Comment at: clang/lib/AST/ExprConstant.cpp:9616
+    return EvaluateFloat(E->getArg(0), Val, Info) &&
+           Success(lround(Val.convertToDouble()), E);
+  }
----------------
rsmith wrote:
> This assumes that the host `double` has the same behavior as the target `double`, which is not true in general. Generally the right thing to do is to implement the relevant functionality on `APFloat`, taking care to produce the right answer for the floating-point model being used by that `APFloat`. This also assumes that the host `long` is the same size as the target `long`, which again is not true in general. Finally, you need to catch the case where the `APFloat` does not fit into the target type; in that case, the best response is probably to treat the evaluation as non-constant and leave it to the runtime library to implement whatever error response it chooses.
I've updated it to do just that: if there is an error, it will treat this evaluation as non-constant and leave it for the runtime library to handle. 

How should I determine the width of the APInt? Because it is run at compile-time, can I use the max size (long long)? Either way, how should I get the correct size, maybe from a builtin long type? 


================
Comment at: clang/test/SemaCXX/math-builtins.cpp:20
+
+  static_assert(__builtin_llround(float()) == 0, ""); // expected-error {{static_assert expression is not an integral constant expression}}
+  static_assert(__builtin_llroundf(float()) == 0, ""); // expected-error {{static_assert expression is not an integral constant expression}}
----------------
For some reasons, these `expected-error`s aren't being caught by lit. Any ideas? Do I need to put it in a `.fail.cpp` file or add something to the `RUN:` comment? 


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D66862/new/

https://reviews.llvm.org/D66862