Though that does beg the question of whether that is compatible with code that uses a different format for long-double.
On Fri, 19 Sept 2025 at 16:30, connor horman <[email protected]> wrote: > NVM, it fixed it, the file was just not being included b/c I didn't read > the config.host file correctly. > > On Thu, 18 Sept 2025 at 21:25, connor horman <[email protected]> wrote: > >> Setting a t- file with that flag doesn't fix it, same error. >> Also, as far as I can tell, that flag is a no-op other than undoing the >> other CLI flags. >> >> On Thu, Sep 18, 2025 at 15:00 H.J. Lu <[email protected]> wrote: >> >>> On Thu, Sep 18, 2025 at 11:51 AM connor horman <[email protected]> >>> wrote: >>> > >>> > FTR, using __float80 directly works. >>> >>> See -mlong-double-80 in libgcc/config/i386/t-linux >>> >>> -- >>> H.J. >>> >>
