"tar up /tools" = bad advice

[alexander@…]

"6.63. Cleaning Up" currently says:

Since they are no longer needed you can delete the /tools directory if so desired or tar it up and keep it to build another final system.

But, since the contents of /tools are irreversibly modified in "6.12. Re-adjusting the Toolchain", such re-adjusted /tools are probably useless for building another final system.

[chris@…]

The statement about tarring up /tools could just be moved to Chapter 5, but it would also need to say that you need the binutils source and build dirs as well.

[alexander@…]

If we implement the "ld-new" hack from DIY, there is no longer any need to tar up binutils build dir. And indeed, the note could be moved to Chapter 5 then.

[Matthew Burgess]

You've touched on one of my pet peeves in the current build method, i.e. having to keep build directories around between chapters 5 and 6. Would you mind pointing me to the relevant bit of DIY-Linux that resolves this, so we can assess its impact on/suitability for LFS?

[n-roeser@…]

Tushar's post <URL:http://linuxfromscratch.org/pipermail/lfs-dev/2006-January/055024.html> looks good.

[tushar@…]

(In reply to comment #3)

You've touched on one of my pet peeves in the current build method, i.e. having to keep build directories around between chapters 5 and 6. Would you mind pointing me to the relevant bit of DIY-Linux that resolves this, so we can assess its impact on/suitability for LFS?

It has been my pet peeve since the day it was proposed :-) But no wanted to change it at that time.

I use an additional fix to avoid modifying /tools during Ch 6. I change the specs file at the end of Ch 5. In the section "Creating essential symlinks", I create an additional symlink "ln -sf /tools/lib/ld-linux.so.2 /lib/". Greg uses another technique of using a temporary specs file by passing the -specs= option to gcc.

[matt@…]

Additional output - a user requested help with this.

Looks like the perl component from his toolchain does not like being used on the new machine.

cd /sources/glibc-build/posix && gcc -O -include ../config.h /sources/glibc- 2.3.4-20040701/posix/annexc.c -o annexc /tmp/ccWfumu7.o(.text+0x1ae): In function `get_null_defines': : warning: the use of tmpnam' is dangerous, better use mkstemp' /sources/glibc-build/posix/annexc 'gcc' \

'-I../csu -I../iconv -I../iconvdata -I../locale -I../localedata -I../assert -

I../ctype -I../intl -I../catgets -I../math -I../setjmp -I../signal - I../stdlib -I../stdio-common -I../libio -I../dlfcn -I../malloc -I../string - I../wcsmbs -I../timezone -I../time -I../dirent -I../grp -I../pwd -I../posix - I../io -I../termios -I../resource -I../misc -I../socket -I../sysvipc - I../gmon -I../gnulib -I../wctype -I../manual -I../shadow -I../po -I../argp - I../crypt -I../nptl -I../resolv -I../nss -I../rt -I../conform -I../debug - I../nptl_db -I../inet -I../hesiod -I../sunrpc -I../nis -I../nscd -I../streams - I../login -I../elf -I../include -I.. -I/sources/glibc-build - I../sysdeps/i386/elf -I../nptl/sysdeps/unix/sysv/linux/i386/i686 - I../nptl/sysdeps/unix/sysv/linux/i386 -I../nptl/sysdeps/unix/sysv/linux - I../nptl/sysdeps/pthread -I../sysdeps/pthread -I../nptl/sysdeps/unix/sysv - I../nptl/sysdeps/unix -I../nptl/sysdeps/i386/i686 -I../nptl/sysdeps/i386 - I../sysdeps/unix/sysv/linux/i386 -I../sysdeps/unix/sysv/linux - I../sysdeps/gnu -I../sysdeps/unix/common -I../sysdeps/unix/mman - I../sysdeps/unix/inet -I../sysdeps/unix/sysv/i386 -I../sysdeps/unix/sysv - I../sysdeps/unix/i386 -I../sysdeps/unix -I../sysdeps/posix - I../sysdeps/i386/i686/fpu -I../sysdeps/i386/i686 -I../sysdeps/i386/i486 - I../nptl/sysdeps/i386/i486 -I../sysdeps/i386/fpu -I../sysdeps/i386 - I../sysdeps/wordsize-32 -I../sysdeps/ieee754/ldbl-96 -I../sysdeps/ieee754/dbl- 64 -I../sysdeps/ieee754/flt-32 -I../sysdeps/ieee754 -I../sysdeps/generic/elf - I../sysdeps/generic -nostdinc -isystem /tools/lib/gcc/i686-pc-linux- gnu/3.4.1/include -isystem /tools/glibc-kernheaders' > /sources/glibc- build/posix/annexc.out make[2]: sources/glibc-build/posix/annexc.out Error 1 (ignored) /sources/glibc-build/malloc/mtrace /sources/glibc-build/posix/bug- regex2.mtrace > /sources/glibc-build/posix/bug-regex2-mem /bin/sh: /sources/glibc-build/malloc/mtrace: No such file or directory make[2]: * sources/glibc-build/posix/bug-regex2-mem Error 127 make[2]: Leaving directory `/sources/glibc-2.3.4-20040701/posix' make[1]: * [posix/tests] Error 2 make[1]: Leaving directory `/sources/glibc-2.3.4-20040701' make: * [check] Error 2 root:/sources/glibc-build#

[alexander@…]

Matt, sorry for wrongly relating this error output with this bug. Looks like the user actually followed our bad advice. But the point made by me on IRC is still valid: we should either not suggest to tar up tools, or say that such tarred up /tools are useful only for building exactly the same LFS version, which means that it is more productive to tar up the final (full) LFS image.