On Sat, Jul 05, 2025 at 10:18:06PM +0200, Scott Kostyshak wrote:
> On Sat, Jul 05, 2025 at 09:35:58PM +0200, Kornel Benko wrote:
> > Am Sat, 5 Jul 2025 21:18:52 +0200
> > schrieb Kornel Benko <kor...@lyx.org>:
> >
> > > Am Sat, 5 Jul 2025 20:59:42 +0200
> > > schrieb Scott Kostyshak <skost...@lyx.org>:
> > >
> > > > On Sat, Jul 05, 2025 at 08:14:49PM +0200, Kornel Benko wrote:
> > > > > Am Sat, 05 Jul 2025 18:05:27 +0000
> > > > > schrieb Scott Kostyshak <skost...@lyx.org>:
> > > > >
> > > > > > commit cd5fb84098f46738a45699aeb1491b52282e3c2d
> > > > > > Author: Scott Kostyshak <skost...@lyx.org>
> > > > > > Date: Sat Jul 5 20:04:28 2025 +0200
> > > > > >
> > > > > > Fix the last remaining -Wmissing-noreturn
> > > > > >
> > > > > > lyxbreaker() calls LBUFERR() which calls doBufErr() which was
> > > > > > previously
> > > > > > marked as [[noreturn] at 37b9655f.
> > > > > > ---
> > > > > > src/CoordCache.h | 2 +-
> > > > > > 1 file changed, 1 insertion(+), 1 deletion(-)
> > > > > >
> > > > > > diff --git a/src/CoordCache.h b/src/CoordCache.h
> > > > > > index 8893aebe48..c732e6f042 100644
> > > > > > --- a/src/CoordCache.h
> > > > > > +++ b/src/CoordCache.h
> > > > > > @@ -38,7 +38,7 @@ class MathData;
> > > > > > #define ASSERT_HAS_POS(thing, hint)
> > > > > > #endif
> > > > > >
> > > > > > -void lyxbreaker(void const * data, const char * hint, int size);
> > > > > > +[[noreturn]] void lyxbreaker(void const * data, const char * hint,
> > > > > > int size);
> > > > > >
> > > > > > struct Geometry {
> > > > > > Point pos = {-10000, -10000 };
> > > > >
> > > > > /usr2/src/lyx/lyx-git/src/LyX.cpp:270:1: error: function declared
> > > > > 'noreturn' should
> > > > > not return [-Werror,-Winvalid-noreturn] 270 | }
> > > >
> > > > Dang. What is your compiler and compiler version number?
> > > >
> > > > I tested on GCC 13.3.0 and Clang 18.1.3.
> > > >
> > > > I have access to other versions so I'll test those when I can.
> > > >
> > > > Scott
> > >
> > > clang18.1.3
> > > -fno-strict-aliasing -O3 -DNDEBUG -std=c++20 -Wall -Wextra
> > > -Wno-deprecated-copy -Werror
> > > -pedantic -Wundef -Wno-nested-anon-types -Wno-deprecated-declarations
> > >
> > > Kornel
> >
> > This 'cures' it
> > diff --git a/src/LyX.cpp b/src/LyX.cpp
> > index 8c467f5e6a..c847de0d3e 100644
> > --- a/src/LyX.cpp
> > +++ b/src/LyX.cpp
> > @@ -263,12 +263,11 @@ void lyx_exit(int exit_code)
> > #endif
> >
> > // In release mode, try to exit gracefully.
> > if (theApp())
> > theApp()->exit(exit_code);
> > - else
> > - exit(exit_code);
> > + exit(exit_code);
>
> Interesting! Thanks for figuring that out. It makes sense from a cursor
> point of view, but I also am hesitant to touch this code.
>
> I suggest we wait to see if anyone else gives a +1 to your patch.
> Otherwise I will revert my change to that function and put this warning
> category on hold.I can reproduce the warning you get if I set -DNDEBUG. I wonder if what's going on is that -DNDEBUG turns off some underlying assertions so that the function does actually return? Knowing that, does that tell us whether your patch makes sense to commit? I still don't know. Either way, it seems like we should either both set -DNDEBUG or both not set it. From what I understand, -DNDEBUG is more of a release flag. So I wonder if it makes sense for you to remove it from your set of development flags looking at warnings. That said, I don't know what I'm talking about so I could have misunderstood. Scott
signature.asc
Description: PGP signature
-- lyx-devel mailing list lyx-devel@lists.lyx.org https://lists.lyx.org/mailman/listinfo/lyx-devel
