Re: [PATCH rtems-source-builder] rtems-gcc-12-newlib-head.cfg: Update to include true long double math

Patch posted. Thanks. --joel On Fri, May 19, 2023 at 1:51 AM Sebastian Huber < sebastian.hu...@embedded-brains.de> wrote: > On 17.05.23 23:13, Joel Sherrill wrote: > > Previously newlib only had long double math functions when the > > sizeof(double) was equal to sizeof(long double). Newlib now

Re: [PATCH rtems-source-builder] rtems-gcc-12-newlib-head.cfg: Update to include true long double math

On 17.05.23 23:13, Joel Sherrill wrote: Previously newlib only had long double math functions when the sizeof(double) was equal to sizeof(long double). Newlib now includes support for true long double on x86, x86_64, and aarch64. --- rtems/config/tools/rtems-gcc-12-newlib-head.cfg | 4 ++-- 1

Re: [PATCH rtems-source-builder] rtems-gcc-12-newlib-head.cfg: Update to include true long double math

Thanks for the ACK. FreeBSD has support for two other architectures -- RISC-V and PowerPC. But they were going to require more work. It was a lot just to get the base infrastructure in. RISC-V only had 64-bit support in FreeBSD as I recall. I'm hoping we can politely beg Hesham to tweak that and

Re: [PATCH rtems-source-builder] rtems-gcc-12-newlib-head.cfg: Update to include true long double math

OK. Well done on the long doubles work. Chris On 18/5/2023 7:13 am, Joel Sherrill wrote: > Previously newlib only had long double math functions when the > sizeof(double) was equal to sizeof(long double). Newlib now includes > support for true long double on x86, x86_64, and aarch64. > --- >