<p dir="ltr"><br>
Il 16 apr 2017 5:36 PM, "Michele Bucca" <<a href="mailto:michele.bucca@gmail.com">michele.bucca@gmail.com</a>> ha scritto:<br>
><br>
> 2017-04-16 9:15 GMT+02:00 Michele Bucca <<a href="mailto:michele.bucca@gmail.com">michele.bucca@gmail.com</a>>:<br>
> > 2017-04-15 21:58 GMT+02:00 Michele Bucca <<a href="mailto:michele.bucca@gmail.com">michele.bucca@gmail.com</a>>:<br>
> >><br>
> >> Il 15 apr 2017 9:56 PM, "akhiezer" <<a href="mailto:lfs65@cruziero.com">lfs65@cruziero.com</a>> ha scritto:<br>
> >>><br>
> >>> > From: Michele Bucca <<a href="mailto:michele.bucca@gmail.com">michele.bucca@gmail.com</a>><br>
> >>> > Date: Sat, 15 Apr 2017 13:15:01 +0200<br>
> >>> > To: CLFS Support <<a href="mailto:clfs-support@lists.clfs.org">clfs-support@lists.clfs.org</a>><br>
> >>> > Subject: Re: [Clfs-support] [Native Compiler] Problem Cross-compiling<br>
> >>> > GMP<br>
> >>> > MPFR and MPC<br>
> >>> ><br>
> >>> > 2017-04-15 11:22 GMT+02:00 Michele Bucca <<a href="mailto:michele.bucca@gmail.com">michele.bucca@gmail.com</a>>:<br>
> >>> .<br>
> >>> .<br>
> >>> ><br>
> >>> > I've solved the problem with the libraries but the problem while<br>
> >>> > compiling the host-x-host compiler still persists<br>
> >>><br>
> >>><br>
> >>> OK: what you're aiming for, depends of course on a long chain of details<br>
> >>> being all exactly correct. So let's assume for now that all has been<br>
> >>> done correctly thus far: and deal with the gcc error 'at face value'.<br>
> >>><br>
> >>><br>
> >>> ><br>
> >>> > i486-linux-musl-gcc -g -O2 -O2 -g -O2 -DIN_GCC -W -Wall<br>
> >>> > -Wno-narrowing -Wwrite-strings -Wcast-qual -Wstrict-prototypes<br>
> >>> > -Wmissing-prototypes -Wold-style-definition -isystem ./include<br>
> >>> > -fpic -mlong-double-80 -DUSE_ELF_SYMVER -g -DIN_LIBGCC2<br>
> >>> > -fbuilding-libgcc -fno-stack-protector -Dinhibit_libc -fpic<br>
> >>> > -mlong-double-80 -DUSE_ELF_SYMVER -I. -I. -I../.././gcc<br>
> >>> > -I../../../gcc-src/libgcc -I../../../gcc-src/libgcc/.<br>
> >>> > -I../../../gcc-src/libgcc/../gcc -I../../../gcc-src/libgcc/../include<br>
> >>> > -I../../../gcc-src/libgcc/../libdecnumber/no<br>
> >>> > -I../../../gcc-src/libgcc/../libdecnumber -DHAVE_CC_TLS -DUSE_TLS -o<br>
> >>> > decLibrary.o -MT decLibrary.o -MD -MP -MF decLibrary.dep -c<br>
> >>> > ../../../gcc-src/libgcc/../libdecnumber/decLibrary.c<br>
> >>> > ../../../gcc-src/libgcc/../libdecnumber/decLibrary.c:27:24: fatal<br>
> >>> > error: decimal128.h: No such file or directory<br>
> >>> > #include "decimal128.h"<br>
> >>> > ^<br>
> >>><br>
> >>><br>
> >>> Is this an issue where 'decimal128.h' does exist, but is just not<br>
> >>> being found. Do you have a 'decimal128.h' file anywhere under your<br>
> >>> compile/target areas - just do 'find -P / -iname decimal128.h -ls'<br>
> >>> to get them all.<br>
> >><br>
> >> Compile/target areas?<br>
> ><br>
> > I did this:<br>
> ><br>
> > michele@glados:~/tmp/musl-native-toolchain$ find -P $PWD -iname decimal128.h -ls<br>
> > 22569295 4 -rw-r--r-- 1 michele michele 4054 gen 4 2016<br>
> > /home/michele/tmp/musl-native-toolchain/gcc-src/libdecnumber/dpd/decimal128.h<br>
> > 22569320 4 -rw-r--r-- 1 michele michele 28 mar 24 2007<br>
> > /home/michele/tmp/musl-native-toolchain/gcc-src/libdecnumber/bid/decimal128.h<br>
> > 10093214 4 -rw-r--r-- 1 michele michele 4054 gen 4 2016<br>
> > /home/michele/tmp/musl-native-toolchain/gcc-6.2.0/libdecnumber/dpd/decimal128.h<br>
> > 10093239 4 -rw-r--r-- 1 michele michele 28 mar 24 2007<br>
> > /home/michele/tmp/musl-native-toolchain/gcc-6.2.0/libdecnumber/bid/decimal128.h<br>
> ><br>
> > the GCC source was unpacked twice because of me trying to fix this<br>
> > problem. I thought that there was a problem with the source code of<br>
> > GCC<br>
> ><br>
> >>><br>
> >>><br>
> >>> If it does exist, then it may be worth trying - at least, as a debugging<br>
> >>> step - a symlink to it from where gcc is looking for it.<br>
> >>><br>
><br>
><br>
> I made sure that GCC could find the decimal128.h and the other files<br>
> that it was looking for but I got this error.<br>
><br>
> i486-linux-musl-gcc -g -O2 -O2 -g -O2 -DIN_GCC -W -Wall<br>
> -Wno-narrowing -Wwrite-strings -Wcast-qual -Wstrict-prototypes<br>
> -Wmissing-prototypes -Wold-style-definition -isystem ./include<br>
> -fpic -mlong-double-80 -DUSE_ELF_SYMVER -g -DIN_LIBGCC2<br>
> -fbuilding-libgcc -fno-stack-protector -Dinhibit_libc -fpic<br>
> -mlong-double-80 -DUSE_ELF_SYMVER -I. -I. -I../.././gcc<br>
> -I../../../gcc-src/libgcc -I../../../gcc-src/libgcc/.<br>
> -I../../../gcc-src/libgcc/../gcc -I../../../gcc-src/libgcc/../include<br>
> -I../../../gcc-src/libgcc/../libdecnumber/no<br>
> -I../../../gcc-src/libgcc/../libdecnumber -DHAVE_CC_TLS -DUSE_TLS -o<br>
> decSingle.o -MT decSingle.o -MD -MP -MF decSingle.dep -c<br>
> ../../../gcc-src/libgcc/../libdecnumber/decSingle.c<br>
> make[1]: *** Nessuna regola per generare l'obiettivo<br>
> "../../../gcc-src/libgcc/../libdecnumber/no/decimal32.c", necessario<br>
> per "decimal32.o". Arresto.<br>
> make[1]: uscita dalla directory<br>
> "/home/michele/tmp/musl-native-toolchain/2-obj-gcc/i486-linux-musl/libgcc"<br>
> Makefile:11384: set di istruzioni per l'obiettivo "all-target-libgcc"<br>
> non riuscito<br>
> make: *** [all-target-libgcc] Errore 2<br>
><br>
> it says that there's no rule to generate the object decimal32.o</p>
<p dir="ltr">problem solved removing the --disable-decimal-float switch from configure.</p>
<p dir="ltr">to build C++ support in the host-x-host relocatable compiler use </p>
<p dir="ltr">--build=${CLFS_HOST}<br>
--host=${CLFS_TARGET}<br>
--target=${CLFS_TARGET}<br>
--enable-languages=c,c++ \<br>
--disable-libmudflap \<br>
--disable-libsatinizer \<br>
--disable-libmpx \<br>
--enable-c99 \<br>
--enable-long-long \<br>
--disable-nls \<br>
--disable-multilib<br>
</p>