Re:C++'s `Noexcept` Can (Sometimes) Help (Or Hurt) Performance
by terrymah on Monday August 05, 2024 @10:25PM
(#41167342)
Nah that was mostly about extern "C" functions which technically can't throw (so the noexcept runtime stuff would be optimized out) but in practice there is a ton of code marked extern "C" which throws
[ Parent
| on HN ]
Re:C++'s `Noexcept` Can (Sometimes) Help (Or Hurt) Performance
by dataflow on Monday August 05, 2024 @10:28PM
(#41167353)
extern "C" seems related to the other flags, not 'r'?
Re:C++'s `Noexcept` Can (Sometimes) Help (Or Hurt) Performance
by terrymah on Monday August 05, 2024 @10:57PM
(#41167446)
Well, yeah, things can be related to many things, but throwing extern "C"s was one of the motivations as I recall for 'r'. r is about a compiler optimization where we elide the runtime terminate check if we can statically "prove" a function can never throw. To prove it statically we depend on things like extern "C" functions not throwing, even though users can (and do) totally write that code.
Re:C++'s `Noexcept` Can (Sometimes) Help (Or Hurt) Performance
by Arech on Tuesday August 06, 2024 @04:14AM
(#41168779)
> in practice there is a ton of code marked extern "C" which throws
Obviously, a random programmer could do any evil things, but does that apply to standard code, such as C standard library used from C++?
Re:C++'s `Noexcept` Can (Sometimes) Help (Or Hurt) Performance
by leni536 on Tuesday August 06, 2024 @06:03AM
(#41169298)
Yes, qsort and bsearch can throw.
Re:C++'s `Noexcept` Can (Sometimes) Help (Or Hurt) Performance
by Arech on Tuesday August 06, 2024 @06:16AM
(#41169384)
jfc... Can you give more info how did you learn this and to which lib implementation this applies?
Re:C++'s `Noexcept` Can (Sometimes) Help (Or Hurt) Performance
by tsimionescu on Tuesday August 06, 2024 @06:24AM
(#41169449)
Well, given that qsort and bsearch take a function pointer and call it, that function pointer can easily point to a function that throws. So I think this applies to all implementations of qsort and bsearch. Especially since there is no way to mark a function pointer as noexcept.
Re:C++'s `Noexcept` Can (Sometimes) Help (Or Hurt) Performance
by leni536 on Tuesday August 06, 2024 @12:32PM
(#41172437)
> Especially since there is no way to mark a function pointer as noexcept.
There is, noexcept is part of the type since C++17. In fact, I prefer noexcept function pointer parameters for C library wrappers, as I don't expect most libraries written in C to deal with stack unwinding at all.
Re:C++'s `Noexcept` Can (Sometimes) Help (Or Hurt) Performance
by tsimionescu on Wednesday August 07, 2024 @06:00AM
(#41179793)
Oops, you're right, I hadn't checked that this had changed in C++ 17.
Re:C++'s `Noexcept` Can (Sometimes) Help (Or Hurt) Performance
by leni536 on Tuesday August 06, 2024 @12:29PM
(#41172414)
https://eel.is/c++draft/alg.c.library#4
Any library implementation that is C++ compliant must implement this. I'm pretty sure that libstdc++ + glibc is compliant, assuming sane glibc compiler options.
Re:C++'s `Noexcept` Can (Sometimes) Help (Or Hurt) Performance
by Arech on Tuesday August 06, 2024 @01:11PM
(#41172873)
but to me these are - again - a user induced problems. I'm interested if a user doesn't do stupid things, should they still afraid that a standard extern C code could throw? Say, std::sprintf() which if I'm not mistaken boils down to C directly? Are there cases where C std lib could throw without a "help from a user"?
Re:C++'s `Noexcept` Can (Sometimes) Help (Or Hurt) Performance
by leni536 on Tuesday August 06, 2024 @04:06PM
(#41174761)
https://eel.is/c++draft/res.on.exception.handling#2
I general functions from the C standard library must not throw, except for the ones taking function pointers that they execute.
Re:C++'s `Noexcept` Can (Sometimes) Help (Or Hurt) Performance
by Arech on Wednesday August 07, 2024 @06:32AM
(#41179925)
perfect again, thank you!
Re:C++'s `Noexcept` Can (Sometimes) Help (Or Hurt) Performance
by tsimionescu on Wednesday August 07, 2024 @06:27AM
(#41179896)
I don't think anything in the C parts of the C++ standard library throws its own exceptions. However, it's not completely unreasonable for a third party C library to use a C++ library underneath, and that might propagate any exceptions that the C++ side throws. This would be especially true if the C library were designed with some kind of plugin support, and someone supplied a C++ plugin.