C23 mode: <float.h> missing FLT_SNAN macro #162830
Description
This program:
#include <float.h>
int main(void){
if(1){
static float f = FLT_SNAN + FLT_SNAN;
}
return 0;
}gets this error:
. /usr/bin/../lib/clang/20/share/msan_ignorelist.txt
. /usr/bin/../lib/clang/20/include/float.h
test190.c:4:22: error: use of undeclared identifier 'FLT_SNAN'
4 | static float f = FLT_SNAN + FLT_SNAN;
| ^
test190.c:4:33: error: use of undeclared identifier 'FLT_SNAN'
4 | static float f = FLT_SNAN + FLT_SNAN;
| ^
2 errors generated.
The flags passed to clang are:
# flags common to both compiler and linker
export CLFLAGS="-g \
-fsanitize=memory \
-fsanitize=undefined \
-fno-sanitize=float-cast-overflow \
-fno-sanitize=float-divide-by-zero"
# -fsanitize=address added 2019/04/28 => kills compile; not w/ memory
# flags for compiler
export CFLAGS="-H -std=c23 -m64 -mfpmath=sse -msse2 -pedantic -O0 \
-fno-fast-math \
-fdenormal-fp-math=ieee \
-fstrict-float-cast-overflow \
-fmath-errno \
-ftrapping-math \
-ffp-contract=on \
-fhonor-infinities \
-fhonor-nans \
-fsigned-zeros \
-fno-associative-math \
-fno-reciprocal-math \
-fno-unsafe-math-optimizations \
-fno-finite-math-only \
-frounding-math \
-ffp-model=strict \
-ffp-exception-behavior=strict \
-Xclang -disable-llvm-optzns ${INCS}"
This is on Intel x86_64 Core i5 running Fedora Linux 42.
clang -v gets as it first line:
clang version 20.1.8 (Fedora 20.1.8-4.fc42)
Aside: If I use the older signaling NAN macro in <math.h> SNANF, the above code cannot be compiled for a different reason.