Hi All,
I have been trying to figure out which magic incantation I need to use _Float16
In the testsuite. The majority of the targets turn on -pedantic and unless you override it
(e.g. use dg-options instead of dg-additional-options) then the compile will always fail with
float16.c:1:1: warning: ISO C does not support the '_Float16' type [-Wpedantic]
1 | _Float16 foo (_Float16 a, _Float16 b)
Simple example
_Float16 foo (_Float16 a, _Float16 b)
{
return a + b;
}
And a trunk build
aarch64-none-elf-gcc -S -o - float16.c -O2 -march=armv8.2-a+fp16
works, but
aarch64-none-elf-gcc -S -o - float16.c -O2 -march=armv8.2-a+fp16 -pedantic-errors
fails.
Does that mean I have to use -Wno-pedantic anytime I want to use _Float16? No variation of -std=... seems to fix this either.
Thanks,
Tamar
