[Bug libstdc++/113283] missing C++26 freestanding headers.
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=113283 --- Comment #11 from Arsen Arsenović --- could be implemented in libstdc++ when no libc impl is present
[Bug libstdc++/113283] missing C++26 freestanding headers.
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=113283 --- Comment #10 from cqwrteur --- also. how do deal with other headers. like cstdlib which C++26 requires qsort to be freestanding.
[Bug libstdc++/113283] missing C++26 freestanding headers.
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=113283 --- Comment #9 from Arsen Arsenović --- (In reply to cqwrteur from comment #6) > if someone writes an operating system or libc,he can ensure the abi being > the same. that's a job of theirs. they can and should add OS defines for their OS.
[Bug libstdc++/113283] missing C++26 freestanding headers.
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=113283 Christopher Di Bella changed: What|Removed |Added CC||cjdb.ns at gmail dot com --- Comment #8 from Christopher Di Bella --- I briefly suggested moving `` into Clang like `` is, but nothing concrete was ever put into place. As it stands, I do not expect this to be actioned. For folks who are interested in what LLVM folks have discussed, please see the relevant [GitHub issue](https://github.com/llvm/llvm-project/issues/77373).
[Bug libstdc++/113283] missing C++26 freestanding headers.
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=113283 --- Comment #7 from cqwrteur --- (In reply to Arsen Arsenović from comment #5) > C does not have a freestanding error.h, indeed. > > We were considering making up some numbers in a high-up range so that we can > provide some non-OS-provided errors that the standard requires, too > > why do you think having the same numbers for both libraries in freestanding > would be beneficial? this seems odd to me Whatever, i will make clang/libc++ be compatible to gcc/libstdc++ here. The LLVM folks wants to put errno.h into the compiler side as an extension for a freestanding C header.
[Bug libstdc++/113283] missing C++26 freestanding headers.
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=113283 --- Comment #6 from cqwrteur --- if someone writes an operating system or libc,he can ensure the abi being the same. Get Outlook for Android<https://aka.ms/AAb9ysg> From: arsen at gcc dot gnu.org Sent: Tuesday, January 9, 2024 7:18:46 AM To: unlv...@live.com Subject: [Bug libstdc++/113283] missing C++26 freestanding headers. https://gcc.gnu.org/bugzilla/show_bug.cgi?id=113283 Arsen Arsenović changed: What|Removed |Added CC||arsen at gcc dot gnu.org --- Comment #5 from Arsen Arsenović --- C does not have a freestanding error.h, indeed. We were considering making up some numbers in a high-up range so that we can provide some non-OS-provided errors that the standard requires, too why do you think having the same numbers for both libraries in freestanding would be beneficial? this seems odd to me -- You are receiving this mail because: You reported the bug. You are on the CC list for the bug.
[Bug libstdc++/113283] missing C++26 freestanding headers.
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=113283 Arsen Arsenović changed: What|Removed |Added CC||arsen at gcc dot gnu.org --- Comment #5 from Arsen Arsenović --- C does not have a freestanding error.h, indeed. We were considering making up some numbers in a high-up range so that we can provide some non-OS-provided errors that the standard requires, too why do you think having the same numbers for both libraries in freestanding would be beneficial? this seems odd to me
[Bug libstdc++/113283] missing C++26 freestanding headers.
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=113283 Andrew Pinski changed: What|Removed |Added Severity|normal |enhancement
[Bug libstdc++/113283] missing C++26 freestanding headers.
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=113283 --- Comment #4 from cqwrteur --- Created attachment 57012 --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=57012&action=edit Proposed errno numbers These numbers come from Linux Kernel Headers and redhat Newlib-Cygwin. I guess these numbers would be ok and commonly agreed. If you think this is ok, i will start working on clang and libcxx to use these numbers.
[Bug libstdc++/113283] missing C++26 freestanding headers.
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=113283 --- Comment #3 from cqwrteur --- I do not see errno.h in freestanding C. Am i correct? Should compiler provide an errno.h? https://en.cppreference.com/w/c/language/conformance
[Bug libstdc++/113283] missing C++26 freestanding headers.
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=113283 --- Comment #2 from cqwrteur --- (In reply to Jonathan Wakely from comment #1) > We can just make up our own numbers, since we won't be getting errno values > from the OS. i would like to make libcxx to have the same numbers as libstdc++. My proposed patches before to libcxx was using linux's errno numbers as the fallback if __has_include_next() does not exist. What do you think?
[Bug libstdc++/113283] missing C++26 freestanding headers.
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=113283 Jonathan Wakely changed: What|Removed |Added Blocks||110339 --- Comment #1 from Jonathan Wakely --- We can just make up our own numbers, since we won't be getting errno values from the OS. Referenced Bugs: https://gcc.gnu.org/bugzilla/show_bug.cgi?id=110339 [Bug 110339] Implement C++26 library features