Awesome, thanks! On Wed, 12 Apr 2023 at 21:44, Jonathan Wakely <jwakely.gcc@xxxxxxxxx> wrote: > On Tue, 11 Apr 2023 at 10:49, Jonathan Wakely <jwakely.gcc@xxxxxxxxx> > wrote: > > > > On Mon, 10 Apr 2023 at 13:34, Kalamatee <kalamatee@xxxxxxxxx> wrote: > > > > > > Thanks, I'll check that out. > > > > Actually, it looks like the problem might be much simpler: > > fast_float.h fails to include <stdint.h> > > > > Libstdc++ provides its own <stdint.h> header which should solve the > > __STDC_LIMITS_MACROS issue with non-conforming C libraries. > > > > I'll look into it ... > > > Now fixed for gcc-12 and trunk: > https://gcc.gnu.org/pipermail/gcc-patches/2023-April/615601.html > > > > > > > > On Mon, 10 Apr 2023, 13:15 Jonathan Wakely, <jwakely.gcc@xxxxxxxxx> > wrote: > > >> > > >> > > >> > > >> On Mon, 10 Apr 2023, 12:34 Kalamatee via Gcc-help, < > gcc-help@xxxxxxxxxxx> wrote: > > >>> > > >>> Hi, > > >>> > > >>> When im trying to build gcc12, I get the following error -: > > >>> > > >>> /bin/bash ../../libtool --tag CXX --tag disable-shared > --mode=compile > > >>> /home/test/gcc12/./gcc/xgcc -shared-libgcc -B/home/test/gcc12/./gcc > > >>> -nostdinc++ -L/home/test/gcc12/x86_64-aros/libstdc++-v3/src > > >>> -L/home/test/gcc12/x86_64-aros/libstdc++-v3/src/.libs > > >>> -L/home/test/gcc12/x86_64-aros/libstdc++-v3/libsupc++/.libs > > >>> > -B/home/kalam/builds/pc-x86_64-gcc12-clean/bin/linux-x86_64/tools/crosstools/x86_64-aros/bin/ > > >>> > -B/home/kalam/builds/pc-x86_64-gcc12-clean/bin/linux-x86_64/tools/crosstools/x86_64-aros/lib/ > > >>> -isystem > > >>> > /home/kalam/builds/pc-x86_64-gcc12-clean/bin/linux-x86_64/tools/crosstools/x86_64-aros/include > > >>> -isystem > > >>> > /home/kalam/builds/pc-x86_64-gcc12-clean/bin/linux-x86_64/tools/crosstools/x86_64-aros/sys-include > > >>> -I/home/test/gcc12/src/gcc-12.2.0/libstdc++-v3/../libgcc > > >>> -I/home/test/gcc12/x86_64-aros/libstdc++-v3/include/x86_64-aros > > >>> -I/home/test/gcc12/x86_64-aros/libstdc++-v3/include > > >>> -I/home/test/gcc12/src/gcc-12.2.0/libstdc++-v3/libsupc++ > -std=gnu++17 > > >>> -nostdinc++ -fno-implicit-templates -Wall -Wextra -Wwrite-strings > > >>> -Wcast-qual -Wabi=2 -fdiagnostics-show-location=once > -ffunction-sections > > >>> -fdata-sections -frandom-seed=floating_from_chars.lo > -fimplicit-templates > > >>> -g -O2 -c -o floating_from_chars.lo > > >>> > /home/test/gcc12/src/gcc-12.2.0/libstdc++-v3/src/c++17/floating_from_chars.cc > > >>> libtool: compile: /home/test/gcc12/./gcc/xgcc -shared-libgcc > > >>> -B/home/test/gcc12/./gcc -nostdinc++ > > >>> -L/home/test/gcc12/x86_64-aros/libstdc++-v3/src > > >>> -L/home/test/gcc12/x86_64-aros/libstdc++-v3/src/.libs > > >>> -L/home/test/gcc12/x86_64-aros/libstdc++-v3/libsupc++/.libs > > >>> > -B/home/kalam/builds/pc-x86_64-gcc12-clean/bin/linux-x86_64/tools/crosstools/x86_64-aros/bin/ > > >>> > -B/home/kalam/builds/pc-x86_64-gcc12-clean/bin/linux-x86_64/tools/crosstools/x86_64-aros/lib/ > > >>> -isystem > > >>> > /home/kalam/builds/pc-x86_64-gcc12-clean/bin/linux-x86_64/tools/crosstools/x86_64-aros/include > > >>> -isystem > > >>> > /home/kalam/builds/pc-x86_64-gcc12-clean/bin/linux-x86_64/tools/crosstools/x86_64-aros/sys-include > > >>> -I/home/test/gcc12/src/gcc-12.2.0/libstdc++-v3/../libgcc > > >>> -I/home/test/gcc12/x86_64-aros/libstdc++-v3/include/x86_64-aros > > >>> -I/home/test/gcc12/x86_64-aros/libstdc++-v3/include > > >>> -I/home/test/gcc12/src/gcc-12.2.0/libstdc++-v3/libsupc++ -std=gnu++17 > > >>> -nostdinc++ -fno-implicit-templates -Wall -Wextra -Wwrite-strings > > >>> -Wcast-qual -Wabi=2 -fdiagnostics-show-location=once > -ffunction-sections > > >>> -fdata-sections -frandom-seed=floating_from_chars.lo > -fimplicit-templates > > >>> -g -O2 -c > > >>> > /home/test/gcc12/src/gcc-12.2.0/libstdc++-v3/src/c++17/floating_from_chars.cc > > >>> -o floating_from_chars.o > > >>> In file included from > > >>> > /home/test/gcc12/src/gcc-12.2.0/libstdc++-v3/src/c++17/floating_from_chars.cc:77: > > >>> > /home/test/gcc12/src/gcc-12.2.0/libstdc++-v3/src/c++17/fast_float/fast_float.h: > > >>> In function 'void > > >>> {anonymous}::fast_float::round_nearest_tie_even(adjusted_mantissa&, > > >>> int32_t, callback)': > > >>> > /home/test/gcc12/src/gcc-12.2.0/libstdc++-v3/src/c++17/fast_float/fast_float.h:2500:12: > > >>> error: 'UINT64_MAX' was not declared in this scope > > >>> 2500 | mask = UINT64_MAX; > > >>> | ^~~~~~~~~~ > > >>> > /home/test/gcc12/src/gcc-12.2.0/libstdc++-v3/src/c++17/fast_float/fast_float.h:1:1: > > >>> note: 'UINT64_MAX' is defined in header '<cstdint>'; did you forget > to > > >>> '#include <cstdint>'? > > >>> +++ |+#include <cstdint> > > >>> 1 | // fast_float by Daniel Lemire > > >>> make[7]: *** [Makefile:585: floating_from_chars.lo] Error 1 > > >>> > > >>> I can see from the preprocessor output, the necessary header is > included - > > >>> but is protected with > > >>> #if !defined(__cplusplus) || defined(__STDC_LIMIT_MACROS) > > >>> #endif > > >>> > > >>> And since the compiler doesn't define __STDC_LIMIT_MACROS on the > command > > >>> line, or in the source file - it is resolved. > > >>> > > >>> I'm not sure though, where is the correct place to fix this? > > >> > > >> > > >> > > >> In the C library. The C++ standard is very clear that > __STDC_LIMIT_MACROS has no meaning whatsoever in C++. The C library's > <stdint.h> must define those macros unconditionally for C++. The C99 > standard said the macros are required for C++, so some C libraries follow > that rule, but they should follow the C++ standard for C++ code, since C99 > only defines C and had no business trying to specify how C++ works. > > >> > > >> If you cannot fix the libc headers then you can #define > __STDC_LIMIT_MACROS in your target's os_defines.h header in libstdc++ (I > think that's the right workaround, but I'm not at my desk to check it ... > Search for that macro in the libstdc++ source to see how other targets > handle non-conforming libc headers. >
