[PATCH] libstdc++: testsuite: don't test for __unused

Jonathan Wakely jwakely.gcc@gmail.com
Sun Jan 25 19:28:31 GMT 2026


On Sun, 25 Jan 2026, 16:06 Xi Ruoyao, <xry111@xry111.site> wrote:

> x86_64 glibc has started to use it since the 2.43 release.
>

The correct fix is to #undef it for glibc then. That way it's still defined
(and tested) for other targets.




> Link:
> https://sourceware.org/git/?p=glibc.git;a=commitdiff;h=b9579342c68b#patch67
> ---
>  libstdc++-v3/testsuite/17_intro/badnames.cc | 2 +-
>  libstdc++-v3/testsuite/17_intro/names.cc    | 2 +-
>  2 files changed, 2 insertions(+), 2 deletions(-)
>
> diff --git a/libstdc++-v3/testsuite/17_intro/badnames.cc
> b/libstdc++-v3/testsuite/17_intro/badnames.cc
> index ff590fcdf4d..de4d94faf14 100644
> --- a/libstdc++-v3/testsuite/17_intro/badnames.cc
> +++ b/libstdc++-v3/testsuite/17_intro/badnames.cc
> @@ -67,7 +67,7 @@
>
>  // BSD adds:
>  #define __used         __used is a BADNAME
> -#define __unused       __unused is a BADNAME
> +        // __unused    (glibc uses this so can't test here)
>          // __inline    (glibc uses this so can't test here)
>          // _Complex    (glibc uses this so can't test here)
>  #define __istype       __istype is a BADNAME
> diff --git a/libstdc++-v3/testsuite/17_intro/names.cc
> b/libstdc++-v3/testsuite/17_intro/names.cc
> index 106acfa641b..2249735baae 100644
> --- a/libstdc++-v3/testsuite/17_intro/names.cc
> +++ b/libstdc++-v3/testsuite/17_intro/names.cc
> @@ -150,7 +150,7 @@
>  # define __lockable            cannot be used as an identifier
>  # define __null_sentinel       cannot be used as an identifier
>  # define __packed              cannot be used as an identifier
> -# define __unused              cannot be used as an identifier
> +      // __unused              (glibc uses this so can't test here)
>  # define __used                        cannot be used as an identifier
>
>  #ifndef __APPLE__
> --
> 2.52.0
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://gcc.gnu.org/pipermail/libstdc++/attachments/20260125/81395daf/attachment-0001.htm>


More information about the Libstdc++ mailing list