https://gcc.gnu.org/bugzilla/show_bug.cgi?id=128472
Bug ID: 128472
Summary: [16/17 Regression] libgcc build fails on s390x with
--disable-decimal-float
Product: gcc
Version: 16.2.0
Status: UNCONFIRMED
Severity: normal
Priority: P3
Component: target
Assignee: unassigned at gcc dot gnu.org
Reporter: fuchsfl at gmail dot com
Target Milestone: ---
When configured with --disable-decimal-float, the libgcc build fails because
config/s390/t-float16 unconditionally adds the conversion routines between
_Float16 and decimal float (the _dpd_ files), which require decimal float
support. This is a regression introduced by r16-6682-g5d6d56d837c3db ("s390:
Add HF mode support").
configure:
../configure --target=s390x-unknown-linux-musl
--prefix=/media/flo/nvme0-ssd/gcc/s390x-unknown-linux-musl
--with-sysroot=/media/flo/nvme0-ssd/gcc/s390x-unknown-linux-musl/sysroot
--build=x86_64-unknown-linux-gnu --enable-languages=c --with-arch=arch13
--disable-shared --disable-threads --disable-libssp --disable-libquadmath
--disable-libgomp --disable-libatomic --disable-libsanitizer --disable-multilib
--disable-nls --disable-decimal-float
Build error:
make all-target-libgcc
../../../libgcc/config/s390/_dpd_hf_to_sd.c:27:1: error: decimal floating-point
not supported for this target
27 | _Decimal32 __dpd_extendhfsd (HFtype);
| ^~~~~~~~~~
../../../libgcc/config/s390/_dpd_hf_to_sd.c:34:1: error: decimal floating-point
not supported for this target
34 | _Decimal32
| ^~~~~~~~~~
make[1]: *** [../../../libgcc/static-object.mk:17: _dpd_hf_to_sd.o] Error 1
make[1]: Leaving directory
'/media/flo/nvme0-ssd/gcc/gcc-git-s390musl/build-gcc1/s390x-unknown-linux-musl/libgcc'
make: *** [Makefile:13238: all-target-libgcc] Error 2
