Hi Ken,
The libgcc_s files should have been installed in $LFS/usr in gcc
> pass 2 at the end of chapter 6. $LFS should be /mnt/lfs. The only
> packages installed in /mnt/lfs/tools are binutils pass 1 and gcc
> pass 1.
>
That's interesting. When I check the LFS users history, I find the
following commands:
```
lfs:~$ cat ~/.bash_history | grep -- --enable-languages=c,c++
../configure --target=$LFS_TGT
--prefix=$LFS/tools
--with-glibc-version=2.11
--with-sysroot=$LFS --with-newlib
--without-headers
--enable-initfini-array --disable-nls
--disable-shared
--disable-multilib
--disable-decimal-float --disable-threads
--disable-libatomic
--disable-libgomp --disable-libquadmath
--disable-libssp
--disable-libvtv --disable-libstdcxx
--enable-languages=c,c++
../configure
--build=$(../config.guess) --host=$LFS_TGT
--prefix=/usr
CC_FOR_TARGET=$LFS_TGT-gcc
--with-build-sysroot=$LFS --enable-initfini-array
--disable-nls
--disable-multilib
--disable-decimal-float --disable-libatomic
--disable-libgomp
--disable-libquadmath --disable-libssp
--disable-libvtv
--disable-libstdcxx
--enable-languages=c,c++
```
No idea why yours is different, but I wonder if you have passed
> 'tools' to a different package (before 10.0 we put a lot of things
> in tools, now we DESTDIR to $LFS/usr).
>
For what it's worth here, I am using 10.0
I looked through all of the places I used destdir during temp toolchain
onwards. I see the following:
```
lfs:~$ cat ~/.bash_history | grep -- DESTDIR
make DESTDIR=$LFS install -j8
make DESTDIR=$LFS install -j8
make DESTDIR=$LFS install
make DESTDIR=$LFS TIC_PATH=$(pwd)/build/progs/tic install
make DESTDIR=$LFS TIC_PATH=$(pwd)/build/progs/tic install -j8
make DESTDIR=$LFS install -j8
make DESTDIR=$LFS install -j8
make DESTDIR=$LFS install -j8
make DESTDIR=$LFS install -j8
make DESTDIR=$LFS install
make DESTDIR=$LFS install -j8
make DESTDIR=$LFS install -j8
make DESTDIR=$LFS install -j8
./configure --prefix=/usr --host=$LFS_TGT && make -j8 && make DESTDIR=$LFS
install -j8
make DESTDIR=$LFS install -j8
make DESTDIR=$LFS install -j8
make DESTDIR=$LFS install -j8
make DESTDIR=$LFS install -j8
make -j8 && make DESTDIR=$LFS install -j8
make DESTDIR=$LFS install
make DESTDIR=$LFS install -j8
```
Don't seem to be any locations where `tool` is passed, and I have $LFS in
my /etc/environment (checked to make sure it is correctly set to `/mnt/lfs`)
When looking at chapter 6.18 (second pass of gcc) is the `--prefix=usr`
directive in configure be responsible for ensuring that the gcc shared libs
are installed correctly?
Additionally, is there any extra, helpful info that I could provide to
you? I've proactively provided the contents of /usr, /lib, and /usr/lib.
```
lfs:~$ ls $LFS/usr
bin include lib lib64 libexec local sbin share src var
x86_64-lfs-linux-gnu
```
```
lfs:~$ ls $LFS/lib
firmware libbz2.so.1.0 libgcc_s.so libmount.so.1.1.0
libnss_db.so.2 libreadline.so.8 libutil-2.32.so
ld-2.32.so libbz2.so.1.0.8 libgcc_s.so.1 libmvec-2.32.so
libnss_dns-2.32.so libreadline.so.8.0 libutil.so.1
ld-linux-x86-64.so.2 libc-2.32.so libhistory.so.8 libmvec.so.1
libnss_dns.so.2 libresolv-2.32.so libuuid.so.1
libBrokenLocale-2.32.so libc.so.6 libhistory.so.8.0 libncursesw.so.6
libnss_files-2.32.so libresolv.so.2 libuuid.so.1.3.0
libBrokenLocale.so.1 libcrypt-2.32.so liblzma.so.5
libncursesw.so.6.2 libnss_files.so.2 librt-2.32.so libz.so.1
libSegFault.so libcrypt.so.1 liblzma.so.5.2.5 libnsl-2.32.so
libnss_hesiod-2.32.so librt.so.1 libz.so.1.2.11
libanl-2.32.so libdl-2.32.so libm-2.32.so libnsl.so.1
libnss_hesiod.so.2 libsmartcols.so.1 libzstd.so.1
libanl.so.1 libdl.so.2 libm.so.6 libnss_compat-2.32.so
libpcprofile.so libsmartcols.so.1.1.0 libzstd.so.1.4.5
libblkid.so.1 libfdisk.so.1 libmemusage.so libnss_compat.so.2
libpthread-2.32.so libthread_db-1.0.so
libblkid.so.1.1.0 libfdisk.so.1.1.0 libmount.so.1 libnss_db-2.32.so
libpthread.so.0 libthread_db.so.1
```
```
lfs:~$ ls $LFS/usr/lib
Mcrt1.o libbz2.a libfl.so.2 libmenuw.so.6.2 libpython3.so
libz.a
Scrt1.o libbz2.so libfl.so.2.0.0 libmount.so
libreadline.so libz.so
audit libc.a libformw.so 'libmount.so*' libresolv.a
libzstd.so
bash libc.so libformw.so.6 libmvec.a libresolv.so
locale
crt1.o libc_nonshared.a libformw.so.6.2 libmvec.so
librt.a perl5
crti.o libcrypt.a libg.a libncurses.so librt.so
pkgconfig
crtn.o libcrypt.so libgcc_s.so libncursesw.so
libsmartcols.so python3.8
expect5.45.4 libctf-nobfd.a libgcc_s.so.1 libnss_compat.so
libstdc++.a sqlite3.30.1.2
gawk libctf-nobfd.so libhistory.so libnss_db.so
libstdc++.so tcl8
gcc libctf-nobfd.so.0 liblzma.la libnss_dns.so
libstdc++.so.6 tcl8.6
gconv libctf-nobfd.so.0.0.0 liblzma.so libnss_files.so
libstdc++.so.6.0.28 tclConfig.sh
gcrt1.o libctf.a 'liblzma.so*' libnss_hesiod.so libstdc++.
so.6.0.28-gdb.py tclooConfig.sh
itcl4.2.0 libctf.so libm-2.32.a libopcodes-2.35.so
libstdc++fs.a tdbc1.1.1
libBrokenLocale.a libctf.so.0 libm.a libopcodes.a
libsupc++.a tdbcmysql1.1.1
libBrokenLocale.so libctf.so.0.0.0 libm.so libopcodes.so
libtcl8.6.so tdbcodbc1.1.1
libanl.a libdl.a libmagic.la libpanelw.so
libtclstub8.6.a tdbcpostgres1.1.1
libanl.so libdl.so libmagic.so libpanelw.so.6
libthread_db.so terminfo
libbfd-2.35.so libexpect5.45.4.so libmagic.so.1
libpanelw.so.6.2 libutil.a texinfo
libbfd.a libfdisk.so libmagic.so.1.0.0 libpthread.a
libutil.so thread2.8.5
libbfd.so libfl.a libmcheck.a libpthread.so libuuid.so
libblkid.so libfl.la libmenuw.so libpython3.8.so
'libuuid.so*'
'libblkid.so*' libfl.so libmenuw.so.6 libpython3.8.so.1.0
liby.a
```
On Mon, Oct 26, 2020 at 1:31 PM Ken Moffat <zarniwh...@ntlworld.com> wrote:
> On Mon, Oct 26, 2020 at 11:46:23AM -0600, Alexander Brassel wrote:
> > When running the check, I see errors reported about missing libgcc_s.so.1
> >
> > ```
> > make[5]: *** [Makefile:9399: gdb_index_test_1] Error 1
> > make[5]: Target 'gdb_index_test_1.stdout' not remade because of errors.
> > `echo g++ -W -Wall -Wstack-usage=262144 -D_LARGEFILE_SOURCE
> > -D_FILE_OFFSET_BITS=64 -fmerge-constants -g -O2 -fno-use-linker-plugin
> -o
> > gdb_index_test_2 | sed -e 's/\([^ ]*\)\(.*\)/\1 -Bgcctestdir\/\2/' -e
> > 's/-Wp,-D_FORTIFY_SOURCE=[0-9][0-9]*//'` -Wl,--gdb-index
> > gdb_index_test_cdebug.o
> > gcctestdir/collect-ld: error: cannot open
> > /usr/lib/gcc/x86_64-lfs-linux-gnu/10.2.0/../../../../lib/libgcc_s.so: No
> > such file or directory
> > gcctestdir/collect-ld: error: cannot open
> > /usr/lib/gcc/x86_64-lfs-linux-gnu/10.2.0/../../../../lib/libgcc_s.so: No
> > such file or directory
> > collect2: error: ld returned 1 exit status
> > make[5]: *** [Makefile:9405: gdb_index_test_2] Error 1
> > ```
> >
> > So I checked the directory layout and usr/lib/libgcc_s.so does exist.
> > However, it's a pointer to:
> >
> > `readlink libgcc_s.so`
> > produces: `/tools/lib/libgcc_s.so`
> >
> > Navigating to /tools/lib:
> > ```
> > (lfs chroot) root:/tools/lib# ls
> > gcc libcc1.la libcc1.so libcc1.so.0 libcc1.so.0.0.0
> > ```
> >
> > Why might this setup be occuring?
>
> The libgcc_s files should have been installed in $LFS/usr in gcc
> pass 2 at the end of chapter 6. $LFS should be /mnt/lfs. The only
> packages installed in /mnt/lfs/tools are binutils pass 1 and gcc
> pass 1.
>
> No idea why yours is different, but I wonder if you have passed
> 'tools' to a different package (before 10.0 we put a lot of things
> in tools, now we DESTDIR to $LFS/usr).
>
> ĸen
> --
> The people next door oppress me all night long. I tell them: I work
> all day, a man's got to have some time to learn to play the tuba.
> That's oppression, that is. [ Guards! Guards! ]
> --
> http://lists.linuxfromscratch.org/listinfo/lfs-support
> FAQ: http://www.linuxfromscratch.org/blfs/faq.html
> Unsubscribe: See the above information page
>
> Do not top post on this list.
>
> A: Because it messes up the order in which people normally read text.
> Q: Why is top-posting such a bad thing?
> A: Top-posting.
> Q: What is the most annoying thing in e-mail?
>
> http://en.wikipedia.org/wiki/Posting_style
>
--
http://lists.linuxfromscratch.org/listinfo/lfs-support
FAQ: http://www.linuxfromscratch.org/blfs/faq.html
Unsubscribe: See the above information page
Do not top post on this list.
A: Because it messes up the order in which people normally read text.
Q: Why is top-posting such a bad thing?
A: Top-posting.
Q: What is the most annoying thing in e-mail?
http://en.wikipedia.org/wiki/Posting_style
