On: C++'s `noexcept` can sometimes help or hurt performance

[rockwotj]

Added to my list why I compile with -fno-exceptions

[jcelerier → rockwotj]

-fno-exceptions only prevents you from calling throw. If you don't want overhead likely you want -fno-asynchronous-unwind-tables + that clang flag that specifies that extern "C" functions don't throw

[flohofwoe → jcelerier]

I'm pretty sure I could see a roughly 10% binary size decrease in my C++ projcts just by setting -fno-exceptions, and that was for C++ code that didn't use exceptions in the first place, so there must be more to it then just forbidding throw. Last time I tinkered with this stuff was around 2017 though.

[gpderetta → flohofwoe]

You do not need unwind tables for noexcept functions, that can be a significant space saving.

[rockwotj → gpderetta]

sure but the unwind flags don’t prevent optimizations like exceptions

[rockwotj → jcelerier]

Looking how the docs for https://gcc.gnu.org/onlinedocs/gcc/Code-Gen-Options.html#ind...

How does that affect code generation? It reads as only effecting debug information and binary size to my untrained eyes.

[rockwotj → rockwotj]

And based on a few clang discourse threads, it only removes .eh_frame

I think this only effects binary size, which I understand smaller binaries can load faster but not being able to get stacktraces for debuggers and profilers seems like a loss