[lfs-support] gcc 4.7.1 second pass
failfailfailfailfailfailfailfailfailfailfailfailfailfailfailfailfailfailfailfailfailfailfailfailfailfailfailfail make[2]: Entering directory `/mnt/erraticOS/usr/src/gcc-build/x86_64-linux-gnu/libgcc' # If this is the top-level multilib, build all the other # multilibs. DEFINES='' HEADERS='../../../gcc/libgcc/config/i386/value-unwind.h' \ ../../../gcc/libgcc/mkheader.sh tmp-libgcc_tm.h /bin/bash ../../../gcc/libgcc/../move-if-change tmp-libgcc_tm.h libgcc_tm.h echo timestamp libgcc_tm.stamp cc -g -O2 -O2 -g -O2 -DIN_GCC -fomit-frame-pointer -W -Wall -Wwrite-strings -Wcast-qual -Wstrict-prototypes -Wmissing-prototypes -Wold-style-definition -isystem ./include -fpic -g -DIN_LIBGCC2 -fbuilding-libgcc -fno-stack-protector -fpic -I. -I. -I../.././gcc -I../../../gcc/libgcc -I../../../gcc/libgcc/. -I../../../gcc/libgcc/../gcc -I../../../gcc/libgcc/../include -I../../../gcc/libgcc/config/libbid -DENABLE_DECIMAL_BID_FORMAT -DHAVE_CC_TLS -DUSE_TLS -o _muldi3.o -MT _muldi3.o -MD -MP -MF _muldi3.dep -DL_muldi3 -c ../../../gcc/libgcc/libgcc2.c -fvisibility=hidden -DHIDE_EXPORTS cc1: error: unrecognized command line option ‘-fbuilding-libgcc’ make[2]: *** [_muldi3.o] Error 1 make[2]: Leaving directory `/mnt/erraticOS/usr/src/gcc-build/x86_64-linux-gnu/libgcc' make[1]: *** [all-target-libgcc] Error 2 make[1]: Leaving directory `/mnt/erraticOS/usr/src/gcc-build' make: *** [all] Error 2 compile failed in /mnt/erraticOS/usr/src/gcc-build failfailfailfailfailfailfailfailfailfailfailfailfailfailfailfailfailfailfailfailfailfailfailfailfailfailfailfail when I built the second pass of binutils, this broke too: #make -C ld clean #make -C ld LIB_PATH=/usr/lib:/lib kept saying as segmentation fault something ernother heres that there script i been usin: https://github.com/paigeadele/erraticOS/blob/master/usr/src/bootstrap_userland.sh ps I want to gently shove my mac pro out of the window behind my desk so it can fall several stories and explode and make some kid cry because of this, I hope it makes you want to cry too knowing that my preference to windows is much stronger now that I know that regardless of the fact that I have the source, I still can't build it... what makes me even more sad is that LFS is the only resource I can find on the internet for how to do this. How do the ubuttnut and debian people do it.. how does the slackware guy do it... hell the how the hell does gentoo do it I guess I'll take a look... any advice is welcome I would really appreciate the help because i know getting past this toolchain is the hard part and it'll be leaps and bounds better in the chroot. thanks. -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: [lfs-support] gcc 4.7.1 second pass
aha, once again, dev@dev-VirtualBox /mnt/erraticOS/usr/src/gcc-build $ gcc --version gcc (Ubuntu/Linaro 4.6.3-1ubuntu5) 4.6.3 Copyright (C) 2011 Free Software Foundation, Inc. This is free software; see the source for copying conditions. There is NO warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. dev@dev-VirtualBox /mnt/erraticOS/usr/src/gcc-build $ apt-get I knew I figured this out once before, and the reason I knew is because I removed the -fbuilding-libgcc (just now and before) from the makefile and came to this problem) cc1plus: error: unrecognized command line option ‘-std=gnu++11’ which has to do with the fact that cc1plus 4.6 doesn't have support for that ;) odd that they didn't include the build process for compiling 4.7 from 4.6 huh? simple symlink (after installing the gcc.47 ppa fixes itbecause in the beginning god crapped out a gcc 4.7 for everyone to compile the new gcc 4.7 that they wrote with.) It'd totally be legit if it was because you need gcc 4.7 (not g++ 4.7) to compile g++ 4.7 but its totally not that way apparently. I was really disappointed to find this... pisses me off I kinda have lost a little faith in this but then again, I don't really have much anyway considering the circumstances. NOW to figure out why previously the #make -C ld clean #make -C ld LIB_PATH=/usr/lib:/lib was causing segfault, and also why I have to specify --host=x86_64-linux-gnu because for whatever reason config.guess doesn't work on the second pass of binutils. then I guess once I have everything else built I should be able to chroot into it in theory but we'll see, I have a feeling its going to be a huge waste of time. On Nov 14, 2012, at 6:49 AM, Paige Thompson erra...@devel.ws wrote: failfailfailfailfailfailfailfailfailfailfailfailfailfailfailfailfailfailfailfailfailfailfailfailfailfailfailfail make[2]: Entering directory `/mnt/erraticOS/usr/src/gcc-build/x86_64-linux-gnu/libgcc' # If this is the top-level multilib, build all the other # multilibs. DEFINES='' HEADERS='../../../gcc/libgcc/config/i386/value-unwind.h' \ ../../../gcc/libgcc/mkheader.sh tmp-libgcc_tm.h /bin/bash ../../../gcc/libgcc/../move-if-change tmp-libgcc_tm.h libgcc_tm.h echo timestamp libgcc_tm.stamp cc -g -O2 -O2 -g -O2 -DIN_GCC -fomit-frame-pointer -W -Wall -Wwrite-strings -Wcast-qual -Wstrict-prototypes -Wmissing-prototypes -Wold-style-definition -isystem ./include -fpic -g -DIN_LIBGCC2 -fbuilding-libgcc -fno-stack-protector -fpic -I. -I. -I../.././gcc -I../../../gcc/libgcc -I../../../gcc/libgcc/. -I../../../gcc/libgcc/../gcc -I../../../gcc/libgcc/../include -I../../../gcc/libgcc/config/libbid -DENABLE_DECIMAL_BID_FORMAT -DHAVE_CC_TLS -DUSE_TLS -o _muldi3.o -MT _muldi3.o -MD -MP -MF _muldi3.dep -DL_muldi3 -c ../../../gcc/libgcc/libgcc2.c -fvisibility=hidden -DHIDE_EXPORTS cc1: error: unrecognized command line option ‘-fbuilding-libgcc’ make[2]: *** [_muldi3.o] Error 1 make[2]: Leaving directory `/mnt/erraticOS/usr/src/gcc-build/x86_64-linux-gnu/libgcc' make[1]: *** [all-target-libgcc] Error 2 make[1]: Leaving directory `/mnt/erraticOS/usr/src/gcc-build' make: *** [all] Error 2 compile failed in /mnt/erraticOS/usr/src/gcc-build failfailfailfailfailfailfailfailfailfailfailfailfailfailfailfailfailfailfailfailfailfailfailfailfailfailfailfail when I built the second pass of binutils, this broke too: #make -C ld clean #make -C ld LIB_PATH=/usr/lib:/lib kept saying as segmentation fault something ernother heres that there script i been usin: https://github.com/paigeadele/erraticOS/blob/master/usr/src/bootstrap_userland.sh ps I want to gently shove my mac pro out of the window behind my desk so it can fall several stories and explode and make some kid cry because of this, I hope it makes you want to cry too knowing that my preference to windows is much stronger now that I know that regardless of the fact that I have the source, I still can't build it... what makes me even more sad is that LFS is the only resource I can find on the internet for how to do this. How do the ubuttnut and debian people do it.. how does the slackware guy do it... hell the how the hell does gentoo do it I guess I'll take a look... any advice is welcome I would really appreciate the help because i know getting past this toolchain is the hard part and it'll be leaps and bounds better in the chroot. thanks. -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: [lfs-support] gcc 4.7.1 second pass
On Wed, Nov 14, 2012 at 06:49:34AM -0800, Paige Thompson wrote: heres that there script i been usin: https://github.com/paigeadele/erraticOS/blob/master/usr/src/bootstrap_userland.sh The *first* time you build LFS, run all the commands through by hand - pasting from the browser is perfectly ok. But before you restart, check the host-prerequisites section and fix up any missing packages. Yes, doing it unscripted takes longer. While packages are building, take the time to read the page about what you are doing, and make sure you don't omit any steps. When you have built LFS successfully, scripts are good. By that stage you should have a deeper understanding of what *can* go wrong, and also of what logging (if any) you wish to achieve. Equally, you should then be in a position to diagnose where the error is likely to be. I don't claim to understand your script (the use of compile.list looks cool, I just don't see where it ever gets updated), but I note two things : 1. You are using apt-get before you begin. Is /bin/sh a symlink to dash ? If so, that is not adequate. 2. You seem to be using git checkouts instead of tarballs - fair enough, if you have the bandwidth and if the versions agree. This does mean that your scripts become excessively long. Earlier, you had the 'No such file' error - that is because something has linked to a host library that no longer exists in chroot : it's in the FAQ. You mentioned using the host 'ld', I think. When you do things differently, you *will* get unusual errors. If you can't fix them after googling, it will often be some hours before you get any response : we aren't all keeping the same hours ĸen -- das eine Mal als Tragödie, das andere Mal als Farce -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: [lfs-support] gcc 4.7.1 second pass
thank you for your response! sorry if I come across as angry at the list I'm not, I just wish this process was something somebody had figured out by now given all the distress that exist out there.. it seems far fetched that I have to use pretty much the extent of my skills to do this sort of thing. I've definitely had to stop and do things by hands every so often... having trouble on this though, dev@dev-VirtualBox /mnt/erraticOS/usr/src/tcl-mirror/unix $ make install Installing libtcl8.5.so to /tools/lib/ Installing tclsh as /tools/bin/tclsh8.5 Installing tclConfig.sh to /tools/lib/ Installing libtclstub8.5.a to /tools/lib/ Installing message catalogs ./tclsh: error while loading shared libraries: __vdso_time: invalid mode for dlopen(): Invalid argument make: *** [install-msgs] Error 127 dev@dev-VirtualBox /mnt/erraticOS/usr/src/tcl-mirror/unix $ I noticed if I try to skip this one and substitute tclsh that I can get by but I keep running into problems similar to this one like with check and nurses: make[1]: Entering directory `/mnt/erraticOS/usr/src/ncurses/progs' gcc ../obj_s/tic.o ../obj_s/dump_entry.o ../obj_s/transform.o -I../progs -I. -DHAVE_CONFIG_H -D_GNU_SOURCE -DNDEBUG -I. -I../include -I/tools/include -O2 --param max-inline-insns-single=1200 -fPIC -L../lib -lncurses -L../lib -lncurses-o tic ../lib/libncurses.so: undefined reference to `__poll_chk' collect2: error: ld returned 1 exit status make[1]: *** [tic] Error 1 make[1]: Leaving directory `/mnt/erraticOS/usr/src/ncurses/progs' make: *** [all] Error 2 not real sure what to make of it but I'm sure I'll figure it out eventually if nobody here knows.. it's been an interesting learning experience if nothing else :) On Nov 14, 2012, at 7:11 AM, Paige Thompson erra...@devel.ws wrote: aha, once again, dev@dev-VirtualBox /mnt/erraticOS/usr/src/gcc-build $ gcc --version gcc (Ubuntu/Linaro 4.6.3-1ubuntu5) 4.6.3 Copyright (C) 2011 Free Software Foundation, Inc. This is free software; see the source for copying conditions. There is NO warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. dev@dev-VirtualBox /mnt/erraticOS/usr/src/gcc-build $ apt-get I knew I figured this out once before, and the reason I knew is because I removed the -fbuilding-libgcc (just now and before) from the makefile and came to this problem) cc1plus: error: unrecognized command line option ‘-std=gnu++11’ which has to do with the fact that cc1plus 4.6 doesn't have support for that ;) odd that they didn't include the build process for compiling 4.7 from 4.6 huh? simple symlink (after installing the gcc.47 ppa fixes itbecause in the beginning god crapped out a gcc 4.7 for everyone to compile the new gcc 4.7 that they wrote with.) It'd totally be legit if it was because you need gcc 4.7 (not g++ 4.7) to compile g++ 4.7 but its totally not that way apparently. I was really disappointed to find this... pisses me off I kinda have lost a little faith in this but then again, I don't really have much anyway considering the circumstances. NOW to figure out why previously the #make -C ld clean #make -C ld LIB_PATH=/usr/lib:/lib was causing segfault, and also why I have to specify --host=x86_64-linux-gnu because for whatever reason config.guess doesn't work on the second pass of binutils. then I guess once I have everything else built I should be able to chroot into it in theory but we'll see, I have a feeling its going to be a huge waste of time. On Nov 14, 2012, at 6:49 AM, Paige Thompson erra...@devel.ws wrote: failfailfailfailfailfailfailfailfailfailfailfailfailfailfailfailfailfailfailfailfailfailfailfailfailfailfailfail make[2]: Entering directory `/mnt/erraticOS/usr/src/gcc-build/x86_64-linux-gnu/libgcc' # If this is the top-level multilib, build all the other # multilibs. DEFINES='' HEADERS='../../../gcc/libgcc/config/i386/value-unwind.h' \ ../../../gcc/libgcc/mkheader.sh tmp-libgcc_tm.h /bin/bash ../../../gcc/libgcc/../move-if-change tmp-libgcc_tm.h libgcc_tm.h echo timestamp libgcc_tm.stamp cc -g -O2 -O2 -g -O2 -DIN_GCC -fomit-frame-pointer -W -Wall -Wwrite-strings -Wcast-qual -Wstrict-prototypes -Wmissing-prototypes -Wold-style-definition -isystem ./include -fpic -g -DIN_LIBGCC2 -fbuilding-libgcc -fno-stack-protector -fpic -I. -I. -I../.././gcc -I../../../gcc/libgcc -I../../../gcc/libgcc/. -I../../../gcc/libgcc/../gcc -I../../../gcc/libgcc/../include -I../../../gcc/libgcc/config/libbid -DENABLE_DECIMAL_BID_FORMAT -DHAVE_CC_TLS -DUSE_TLS -o _muldi3.o -MT _muldi3.o -MD -MP -MF _muldi3.dep -DL_muldi3 -c ../../../gcc/libgcc/libgcc2.c -fvisibility=hidden -DHIDE_EXPORTS cc1: error: unrecognized command line option ‘-fbuilding-libgcc’ make[2]: *** [_muldi3.o] Error 1 make[2]: Leaving directory `/mnt/erraticOS/usr/src/gcc-build/x86_64-linux-gnu/libgcc' make[1]: *** [all-target-libgcc]
Re: [lfs-support] gcc 4.7.1 second pass
ah interesting: gcc ../obj_s/tic.o ../obj_s/dump_entry.o ../obj_s/transform.o -I../progs -I. -DHAVE_CONFIG_H -D_GNU_SOURCE -DNDEBUG -I. -I../include -I/tools/include -O2 --param max-inline-insns-single=1200 -fPIC -L../lib -lncurses -L../lib -L /tools/lib -lncurses-o tic I added -L /tools/lib and it compiled... interesting #ncurses grep -x ncurses compiled.list || ( cd $CWD cd ncurses ./configure --prefix=/tools --with-shared \ --without-debug --without-ada --enable-overwrite ((make pwd | awk -F'/' '{for(i=NF;i=NF;i++) print $i}' $CWD/compiled.list true) || (echo compile failed in `pwd` ; false)) || exit make install ) || exit seems like --prefix would have implied that On Nov 14, 2012, at 7:31 AM, Paige Thompson erra...@devel.ws wrote: thank you for your response! sorry if I come across as angry at the list I'm not, I just wish this process was something somebody had figured out by now given all the distress that exist out there.. it seems far fetched that I have to use pretty much the extent of my skills to do this sort of thing. I've definitely had to stop and do things by hands every so often... having trouble on this though, dev@dev-VirtualBox /mnt/erraticOS/usr/src/tcl-mirror/unix $ make install Installing libtcl8.5.so to /tools/lib/ Installing tclsh as /tools/bin/tclsh8.5 Installing tclConfig.sh to /tools/lib/ Installing libtclstub8.5.a to /tools/lib/ Installing message catalogs ./tclsh: error while loading shared libraries: __vdso_time: invalid mode for dlopen(): Invalid argument make: *** [install-msgs] Error 127 dev@dev-VirtualBox /mnt/erraticOS/usr/src/tcl-mirror/unix $ I noticed if I try to skip this one and substitute tclsh that I can get by but I keep running into problems similar to this one like with check and nurses: make[1]: Entering directory `/mnt/erraticOS/usr/src/ncurses/progs' gcc ../obj_s/tic.o ../obj_s/dump_entry.o ../obj_s/transform.o -I../progs -I. -DHAVE_CONFIG_H -D_GNU_SOURCE -DNDEBUG -I. -I../include -I/tools/include -O2 --param max-inline-insns-single=1200 -fPIC -L../lib -lncurses -L../lib -lncurses-o tic ../lib/libncurses.so: undefined reference to `__poll_chk' collect2: error: ld returned 1 exit status make[1]: *** [tic] Error 1 make[1]: Leaving directory `/mnt/erraticOS/usr/src/ncurses/progs' make: *** [all] Error 2 not real sure what to make of it but I'm sure I'll figure it out eventually if nobody here knows.. it's been an interesting learning experience if nothing else :) On Nov 14, 2012, at 7:11 AM, Paige Thompson erra...@devel.ws wrote: aha, once again, dev@dev-VirtualBox /mnt/erraticOS/usr/src/gcc-build $ gcc --version gcc (Ubuntu/Linaro 4.6.3-1ubuntu5) 4.6.3 Copyright (C) 2011 Free Software Foundation, Inc. This is free software; see the source for copying conditions. There is NO warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. dev@dev-VirtualBox /mnt/erraticOS/usr/src/gcc-build $ apt-get I knew I figured this out once before, and the reason I knew is because I removed the -fbuilding-libgcc (just now and before) from the makefile and came to this problem) cc1plus: error: unrecognized command line option ‘-std=gnu++11’ which has to do with the fact that cc1plus 4.6 doesn't have support for that ;) odd that they didn't include the build process for compiling 4.7 from 4.6 huh? simple symlink (after installing the gcc.47 ppa fixes itbecause in the beginning god crapped out a gcc 4.7 for everyone to compile the new gcc 4.7 that they wrote with.) It'd totally be legit if it was because you need gcc 4.7 (not g++ 4.7) to compile g++ 4.7 but its totally not that way apparently. I was really disappointed to find this... pisses me off I kinda have lost a little faith in this but then again, I don't really have much anyway considering the circumstances. NOW to figure out why previously the #make -C ld clean #make -C ld LIB_PATH=/usr/lib:/lib was causing segfault, and also why I have to specify --host=x86_64-linux-gnu because for whatever reason config.guess doesn't work on the second pass of binutils. then I guess once I have everything else built I should be able to chroot into it in theory but we'll see, I have a feeling its going to be a huge waste of time. On Nov 14, 2012, at 6:49 AM, Paige Thompson erra...@devel.ws wrote: failfailfailfailfailfailfailfailfailfailfailfailfailfailfailfailfailfailfailfailfailfailfailfailfailfailfailfail make[2]: Entering directory `/mnt/erraticOS/usr/src/gcc-build/x86_64-linux-gnu/libgcc' # If this is the top-level multilib, build all the other # multilibs. DEFINES='' HEADERS='../../../gcc/libgcc/config/i386/value-unwind.h' \ ../../../gcc/libgcc/mkheader.sh tmp-libgcc_tm.h /bin/bash ../../../gcc/libgcc/../move-if-change tmp-libgcc_tm.h libgcc_tm.h echo timestamp libgcc_tm.stamp cc -g -O2 -O2
Re: [lfs-support] gcc 4.7.1 second pass
On Wed, Nov 14, 2012 at 07:11:23AM -0800, Paige Thompson wrote: aha, once again, dev@dev-VirtualBox /mnt/erraticOS/usr/src/gcc-build $ gcc --version gcc (Ubuntu/Linaro 4.6.3-1ubuntu5) 4.6.3 dev@dev-VirtualBox /mnt/erraticOS/usr/src/gcc-build $ apt-get I knew I figured this out once before, and the reason I knew is because I removed the -fbuilding-libgcc (just now and before) from the makefile and came to this problem) cc1plus: error: unrecognized command line option ‘-std=gnu++11’ which has to do with the fact that cc1plus 4.6 doesn't have support for that ;) odd that they didn't include the build process for compiling 4.7 from 4.6 huh? Ah ha! You are using the host's compiler in pass 2. Wrong path, failed to correctly specify a variable, omitting +h, or something else. The first pass of gcc builds a pseudo cross-compiler, hence the LFS setting : just enough of a name change to fool the build system into using the cross-compile routines (for separation from the host). This compiler only supports 'C'. The second pass uses the 'cross' compiler to build a native C and C++ compiler. ĸen -- das eine Mal als Tragödie, das andere Mal als Farce -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: [lfs-support] gcc 4.7.1 second pass
thats what I thought you were supposed to do but for some reason I thought that was a bad idea On Nov 14, 2012, at 7:28 AM, Ken Moffat zarniwh...@ntlworld.com wrote: On Wed, Nov 14, 2012 at 07:11:23AM -0800, Paige Thompson wrote: aha, once again, dev@dev-VirtualBox /mnt/erraticOS/usr/src/gcc-build $ gcc --version gcc (Ubuntu/Linaro 4.6.3-1ubuntu5) 4.6.3 dev@dev-VirtualBox /mnt/erraticOS/usr/src/gcc-build $ apt-get I knew I figured this out once before, and the reason I knew is because I removed the -fbuilding-libgcc (just now and before) from the makefile and came to this problem) cc1plus: error: unrecognized command line option ‘-std=gnu++11’ which has to do with the fact that cc1plus 4.6 doesn't have support for that ;) odd that they didn't include the build process for compiling 4.7 from 4.6 huh? Ah ha! You are using the host's compiler in pass 2. Wrong path, failed to correctly specify a variable, omitting +h, or something else. The first pass of gcc builds a pseudo cross-compiler, hence the LFS setting : just enough of a name change to fool the build system into using the cross-compile routines (for separation from the host). This compiler only supports 'C'. The second pass uses the 'cross' compiler to build a native C and C++ compiler. ĸen -- das eine Mal als Tragödie, das andere Mal als Farce -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: [lfs-support] gcc 4.7.1 second pass
so what about compiling the rest of the stuff am I supposed to be using the host compiler or the 2nd pass built gcc for that? On Nov 14, 2012, at 7:43 AM, Paige Thompson erra...@devel.ws wrote: thats what I thought you were supposed to do but for some reason I thought that was a bad idea On Nov 14, 2012, at 7:28 AM, Ken Moffat zarniwh...@ntlworld.com wrote: On Wed, Nov 14, 2012 at 07:11:23AM -0800, Paige Thompson wrote: aha, once again, dev@dev-VirtualBox /mnt/erraticOS/usr/src/gcc-build $ gcc --version gcc (Ubuntu/Linaro 4.6.3-1ubuntu5) 4.6.3 dev@dev-VirtualBox /mnt/erraticOS/usr/src/gcc-build $ apt-get I knew I figured this out once before, and the reason I knew is because I removed the -fbuilding-libgcc (just now and before) from the makefile and came to this problem) cc1plus: error: unrecognized command line option ‘-std=gnu++11’ which has to do with the fact that cc1plus 4.6 doesn't have support for that ;) odd that they didn't include the build process for compiling 4.7 from 4.6 huh? Ah ha! You are using the host's compiler in pass 2. Wrong path, failed to correctly specify a variable, omitting +h, or something else. The first pass of gcc builds a pseudo cross-compiler, hence the LFS setting : just enough of a name change to fool the build system into using the cross-compile routines (for separation from the host). This compiler only supports 'C'. The second pass uses the 'cross' compiler to build a native C and C++ compiler. ĸen -- das eine Mal als Tragödie, das andere Mal als Farce -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page