Am 23.06.2016 um 10:14 schrieb akhiezer:
From: Edgar Alwers <[email protected]>
Date: Wed, 22 Jun 2016 22:18:09 +0200
Subject: Re: [lfs-support] 6.7 Linux-4.6 API headers make mrproper fails
Am 22.06.2016 um 18:35 schrieb Pierre Labastie:
Did you type the CC=... CXX=... etc lines? They are needed (with
backslash as continuation character) to get the behavior I
described... Maybe LFS_TGT is not set? Or $LFS_TGT-gcc is not found in
the PATH? To be clear, you can unpack gcc outside chroot and run
configure as you did (hopefully with the CC= etc lines), but as user
"lfs", and with the specific environment for that user.
Yes. As user lfs i did as the book says.
Your following wording does not make that clear.
CC=$LFS_TGT-gcc \
CXX=$LFS_TGT-g++
...
Then compilation as indicated in the book:
Your word 'Then' there, suggests that the env-vars command was issued
as a separate command from the '../configure ...' that you then quote
below: whereas of course they are all the one cmdline:
====
Ref:
http://www.linuxfromscratch.org/lfs/view/development/chapter05/gcc-pass2.html
==
CC=$LFS_TGT-gcc \
CXX=$LFS_TGT-g++ \
AR=$LFS_TGT-ar \
RANLIB=$LFS_TGT-ranlib \
../configure \
--prefix=/tools \
--with-local-prefix=/tools \
--with-native-system-header-dir=/tools/include \
--enable-languages=c,c++ \
--disable-libstdcxx-pch \
--disable-multilib \
--disable-bootstrap \
--disable-libgomp
====
../configure \ --prefix=/tools \ --with-local-prefix=/tools \
--with-native-system-header-dir=/tools/include \
--enable-languages=c,c++ \ --disable-libstdcxx-pch \ --disable-multilib
\ --disable-bootstrap \ --disable-libgomp
goes OK.
The mentioned line is "checking for gcc... gcc"
LFS_TGT was not set. I got on the corresponding command:
bash: `dirname $($LFS_TGT-gcc
-print-libgcc-file-name)`/include-fixed/limits.h: Permission denied
I anyway think that structures arent more clean now, after going from
root to lfs, no chroot, again lfs. I do not trust anymore.
I hope, the book can be changed to include 32 bit ;-)
You've said a few times that you definitely didn't make any mistakes.
I do not speak "ex catedra" As far as the building up to ยง 6 concerns,
everything went OK, no errorrs were reported, exact the same behaviour
in both trials, starting once from a systemd host, a second time from a
sysvinit one. configure, make, make install, all OK. To this refers my
comment
If so, then next thing would be to pinpoint why "LFS_TGT was not set".
Dont mix the things. LFS_TG was set in both trials. As I said, no
errors. I was just telling you, that in my effort to check the line you
wanted to see, I repeated, not chrooted, a ./compile, as root or as lfs
or I do not know more what. Forgett it.
I deleted the archiv gcc-6.1 after pass-2, as instructed by the book. So
clearly: I cannot tell you the content of the line "checking for gcc" in
the compilation process.
But another comment: it should be the same if I enter the four line
commands
CC=$LFS_TGT-gcc
CXX=$LFS_TGT-g++
AR=$LFS_TGT-ar
RANLIB=$LFS_TGT-ranlib
as single commands and then enter the .compile one or if I give the complete
command as indicated by the book ?
I did the first one. If this was wrong, voila, we have one error !
regards,
Edgar
--
Dr.-Ing. Edgar Alwers <[email protected]>
GPG Key ID:AD5C6F70
--
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
