Hello, is it mandatory to do the following? I'm doing my lfs on a PC-32 bits.

5.4. Binutils-2.22 - Pass 1

case $(uname -m) in
x86_64) mkdir -v /tools/lib && ln -sv lib /tools/lib64 ;;
esac
------------------------------------------------------------------
5.8. Adjusting the Toolchain

SPECS=`dirname $($LFS_TGT-gcc -print-libgcc-file-name)`/specs
$LFS_TGT-gcc -dumpspecs | sed \
-e 's@/lib\(64\)\?/ld@/tools&@g' \<-- Is to locate the folder lib64 bits?
-e "/^\*cpp:$/{n;s,$, -isystem /tools/include,}" > $SPECS
echo "New specs file is: $SPECS"
unset SPECS
-------------------------------------------------------------------
6.5. Creating Directories
...
...
case $(uname -m) in
x86_64) ln -sv lib /lib64 && ln -sv lib /usr/lib64 ;;
esac
...
...
------------------
run normally the commands?

I build on a 32-bit machine. I omit the first one. That case statement is all by itself, and the book even says to do it "If building on x86_64...". I'm not, so I don't. But even if I did, it would not occur since the result of uname -m is i686 for a 32-bit machine.

The second one I do run because I copy and paste that entire block of commands all at once (again like the book says). That part you're interested in is a sed find and replace that searches for occurrences of /lib/ld or /lib64/ld. Then it adds /tools before occurrences of either of those. So everyone should run that whole block of commands.

I run that last one because it also is in a large block that I copy and paste all at once. On a 32-bit machine, that part you excerpted won't occur since the result of uname -m is i686. So everyone should run that whole block of commands.

Other people may do something different.

1 members found this post helpful.