Re: Running rc.d programs under a different user
On 02/06/10 22:11, Charles Solar wrote: LFS is working great, and I can safely say I know much more about linux then I did before going through the book, so thanks to all those who contributed. I am wrapping up my fileserver built from LFS SVN 2010-01-09 and I am trying to find a way to run a daemon process as a different user. In the book we only used the apps config file to specify a different user, ex, svn, proftpd. But I want to run Deluge under a different user and it does not have a config option for this. So I am wondering if there is some way I am 'supposed' to setup the init script to do this. I am trying to avoid installing any more unnecessary programs like daemon and start-stop-daemon so any advice is appreciated. Basically I want to mimic the provided example init scripts from here http://dev.deluge-torrent.org/wiki/UserGuide/InitScript This is what I am using at the moment #!/bin/sh # Deluge daemon startup . /etc/sysconfig/rc . $rc_functions case $1 in start) boot_mesg Starting deluged service... loadproc -p /var/run/deluged.pid /usr/sbin/deluged -c /etc/deluge/ -p 58840 -l /var/log/deluged.log RETVAL=$? if [ $RETVAL -eq 0 ] ; then boot_mesg Starting deluge web ui... loadproc -p /var/run/webluge.pid /usr/sbin/deluge-web -f -c /etc/deluge/ -l /var/log/webluged.log fi ;; stop) boot_mesg Stopping deluged service... killproc /usr/sbin/deluged boot_mesg Stopping deluge web ui... killproc /usr/sbin/deluge-web ;; restart) $0 stop sleep 1 $0 start ;; status) statusproc /usr/bin/deluged ;; *) echo Usage: $0 {start|stop|restart|status} exit 1 ;; esac I checked out the functions library and did not see anything jump out at me. Thanks not sure.. first is you probably should post this on blfs or something(so people don't bi*ch and complain), but as for different user tough to say I would test/try setting the user as root maybe or something(but if this daemon checks /etc/group*shadow for the user logging in then things might be trick). hope this helps. Justin P. Mattock -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: Has anyone been able to build LFS on a 64bit system?
On 02/03/10 23:33, xsystem wrote: I cant not Access http://www.torproject.org/ and I cant not Access facebook and even more . if i can Access http://www.torproject.org/ ,I wiil do not send this email . I rest my case.. bottoms up people lifes too short for this crap bye: Justin P. Mattock -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: Has anyone been able to build LFS on a 64bit system?
On 02/03/10 20:25, brown wrap wrote: Yesterday I thought I had made it through pass one. That was probably because any of the compiles I made didn't require the tests being run. So today I started off with pass 2, and the I think it was glibc failed right off the bat. So when I saw the new release of 6.6, I thought I'd start all over with it. Now I can't get gcc to compile: checking for x86_64-lfs-linux-gnu-gcc... /mnt/lfs/sources/6.6/gcc-build/./gcc/xgcc -B/mnt/lfs/sources/6.6/gcc-build/./gcc/ -B/tools/x86_64-lfs-linux-gnu/bin/ -B/tools/x86_64-lfs-linux-gnu/lib/ -isystem /tools/x86_64-lfs-linux-gnu/include -isystem /tools/x86_64-lfs-linux-gnu/sys-include checking for suffix of object files... configure: error: in `/mnt/lfs/sources/6.6/gcc-build/x86_64-lfs-linux-gnu/libgcc': configure: error: cannot compute suffix of object files: cannot compile See `config.log' for more details. make[1]: *** [configure-target-libgcc] Error 1 make[1]: Leaving directory `/mnt/lfs/sources/6.6/gcc-build' make: *** [all] Error 2 if your wanting to build an x86_64 look at clfs there you choose if you want pure64 or multilib Justin P. Mattock -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: Has anyone been able to build LFS on a 64bit system?
On 02/03/10 21:02, brown wrap wrote: OK, thank you. I just need a pure 64 bit machine. I tried to build LFS on 32 bit and it was way too slow. really depends on what your wanting to create. In my case wanted pure64(used ubuntu 64bit as the host), but if I wanted i3,4,586 etc.. could have done so by just making sure the CFLAGS are set correctly (as well as all other relevant cross compiling switches). Justin P. Mattock -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: Has anyone been able to build LFS on a 64bit system?
On 02/03/10 22:03, xsystem wrote: Who can give me a tor . Thanks !!! well tor would be later down the line i.g. probably in blfs. Justin P. Mattock -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: Has anyone been able to build LFS on a 64bit system?
On 02/03/10 22:19, Bruce Dubbs wrote: Justin P. Mattock wrote: On 02/03/10 20:25, brown wrap wrote: Yesterday I thought I had made it through pass one. That was probably because any of the compiles I made didn't require the tests being run. So today I started off with pass 2, and the I think it was glibc failed right off the bat. So when I saw the new release of 6.6, I thought I'd start all over with it. Now I can't get gcc to compile: checking for x86_64-lfs-linux-gnu-gcc... /mnt/lfs/sources/6.6/gcc-build/./gcc/xgcc -B/mnt/lfs/sources/6.6/gcc-build/./gcc/ -B/tools/x86_64-lfs-linux-gnu/bin/ -B/tools/x86_64-lfs-linux-gnu/lib/ -isystem /tools/x86_64-lfs-linux-gnu/include -isystem /tools/x86_64-lfs-linux-gnu/sys-include checking for suffix of object files... configure: error: in `/mnt/lfs/sources/6.6/gcc-build/x86_64-lfs-linux-gnu/libgcc': configure: error: cannot compute suffix of object files: cannot compile See `config.log' for more details. make[1]: *** [configure-target-libgcc] Error 1 make[1]: Leaving directory `/mnt/lfs/sources/6.6/gcc-build' make: *** [all] Error 2 if your wanting to build an x86_64 look at clfs there you choose if you want pure64 or multilib No, that's not necessary. You only need to follow those instructions if you are trying to do 32-64 or want to build a multi-lib system. I'm currently building 6.6-rc1 as a pure 64-bit system, its working on Chapter 6 GCC now. The GCC build was fine. It's doing the tests. There are no problems so far. The only requirement is to start from a 64-bit distro (and to follow directions). -- Bruce yeah!! your right.. I figured clfs would be more suitable (especially if they decide to build /lib64 and only lib64(like fedora)); main concern though is the dynamic linker Justin P. Mattock -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: Has anyone been able to build LFS on a 64bit system?
On 02/03/10 22:33, xsystem wrote: I can not download tor . You can to me via gmail. The best is the latest version of Official. Thank you very much !!! sure you can: http://www.torproject.org/ Justin P. Mattock -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: Trying to follow 6.5 instructions, but
On 01/31/10 21:46, brown wrap wrote: I am trying to follow the 6.5 instructions, but running iinto problems. First in trying to build gmp and mpfr, there are instructions to untar their files and move their directories to new names. Tha's it, no further instructions that I see. BTW, I am using the online instructions: http://www.linuxfromscratch.org/lfs/view/6.5/ So I used the instructions found in the 2nd build. Then when trying to build gcc I get the following error. Have these errors been fixed or am I doing something wrong. Thanks. checking for x86_64-lfs-linux-gnu-gcc... /mnt/lfs/sources/6.4/gcc-build/./gcc/xgcc -B/mnt/lfs/sources/6.4/gcc-build/./gcc/ -B/tools/x86_64-lfs-linux-gnu/bin/ -B/tools/x86_64-lfs-linux-gnu/lib/ -isystem /tools/x86_64-lfs-linux-gnu/include -isystem /tools/x86_64-lfs-linux-gnu/sys-include checking for suffix of object files... configure: error: in `/mnt/lfs/sources/6.4/gcc-build/x86_64-lfs-linux-gnu/libgcc': configure: error: cannot compute suffix of object files: cannot compile See `config.log' for more details. make[1]: *** [configure-target-libgcc] Error 1 make[1]: Leaving directory `/mnt/lfs/sources/6.4/gcc-build' make: *** [all] Error 2 your supposed to unpackage(rename etc) the packages inside the source of gcc, then gcc should automatically know what they are. or(the long way) compile individually and when getting to gcc use the --with-gmp/mpfr* switches. hope this helps Justin P. Mattock -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: LFS6.4 step 5.17 compiling coreutils-6.12 error
On 01/29/10 05:13, haiye1018 wrote: Thank everyone!!! I have fixed it. Coreutils whose version is greater than 6.0 needs selinux, which can be see from source file: #include selinux/selinux.h or / and #include selinux/context.h These are host system's header files,so we must replace these file with se-selinux.in.h se-context.in.h They are coreutils's built-in header file for host in which no selinux is not supported. My scripts: for file in $(find ./ -name *.c -o -name *.h) do cp -uv $file{,.orig} \ sed -e 's...@#include selinux/selinux.h@#include se-selinux.in.h@g' \ -e 's...@#include selinux/context.h@#include se-context.in.h@g' $file.orig $file \ rm $file.orig \ touch $file done I have constructed the temporary system and entered the chroot environment successfully! 在2010-01-27 10:47:02,Trent Shea trents...@gmail.com 写道: Ack, I keep hitting reply and pulling stuff off the list. Anyhow, thanks for poking me in the right direction. On Tuesday 26 January 2010 19:20:48 you wrote: No, you aren't isolated from the host until you enter chroot in Chapter Fair enough, I should have said fairly isolated at this point. 6.As you build Chapter 5, you get more of what you previously build, but there are libraries and headers from the host that are still potentially available.The linker still uses the libraries in /etc/ld.so.conf for your tools when building other programs. Ah, I was thinking about that, too.There's probably some rpath risks and pkgconfig magic, too. -- Regards, Trent. -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page 网易邮箱,没有垃圾邮件的免费电子邮箱! http://yeah.net/?from=b1 so if you built a system using a host cd without(libselinux.so) -lselinux you probably would of been fine. but if building with ubuntu/fedora /lib/libselinux.so is there causing coreutils to enable all support for SELinux (but could be wrong); Justin P. Mattock -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: Step 6.7. Linux-2.6.30.2 API Headers [Error]
On 01/27/10 07:04, Dmitry Sokolov wrote: I tried to unpack (unbzip) package linux-2.6.30.2.tar.bz2, result: tar: bzip2: Cannot exec: Permission denied. ...Error is not recoverable: exiting now. ... Because of what it on yours? LFS 6.5 Stable fromhttp://www.linuxfromscratch.org/lfs/downloads/stable/ Packages downloaded from wget list LiveCD lfslivecd-x86_64-6.3-r2160 -- Dmitry does your home directory have the proper permissions? i.g. chown name:name /home/name Justin P. Mattock -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: Step 6.7. Linux-2.6.30.2 API Headers [Error]
On 01/27/10 13:27, Andrew Benton wrote: On 27/01/10 19:13, Justin P. Mattock wrote: does your home directory have the proper permissions? i.g. chown name:name /home/name In LFS we compile in $LFS/sources which should have been chown lfs when you created the user lfs Andy Oops.. I didn't really follow the book step by step. ended up building in my home dir instead of /tools as for the kernel even a tar xvjf *.tar.bz2 (breaks?) Justin P. Mattock -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: Step 6.7. Linux-2.6.30.2 API Headers [Error]
On 01/27/10 14:23, linux fan wrote: On 1/27/10, Justin P. Mattockjustinmatt...@gmail.com wrote: Oops.. I didn't really follow the book step by step. The best results are by following the book instructions. Especially in chapter 6 those first steps must be done correctly: Set $LFS correctly Preparing Virtual Kernel File Systems Entering the Chroot Environment yeah your right.. In this case I just made sure the permissions are set correctly so I don't have to build anything as root.(only root command is make install); Justin P. Mattock -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: LFS6.4 step 5.17 compiling coreutils-6.12 error
On 01/23/10 08:38, Trent Shea wrote: On Saturday 23 January 2010 09:32:12 Trent Shea wrote: echo 'main(){}' dummy.c cc -v dummy.c 21 |grep include rm dummy.c or don't forget: rm a.out getfilecon sounds something from an SELinux header file. do you have an updated libselinux?(or maybe build disabling audit/SELinux) Justin P. Mattock -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: LFS6.4 step 5.17 compiling coreutils-6.12 error
On 01/23/10 09:21, Trent Shea wrote: On Saturday 23 January 2010 10:17:21 you wrote: I don't see any flags to disable the search for SELinux. I should mention it's my understanding that gcc shouldn't be searching for host include files at this point, maybe someone can clarify for me though. I haven't the source in front of me at the moment for coreutils. but from looking at the errors with chcon getfilecon etc.. those are all of the tools to manage SELinux(i.g. ls -Z etc..) your host system probably has libselinux installed and my guess is coureutils has some mechanisms to check for that, and if present enables SELinux support(without any switches in configure). my guess is either find a host system to build on that has no libselinux so coreutils is built without SELinux support, or with the host system your using now install libselinux,libsepol,libsemamane(dev's) this way the header files are there coreutils builds with SELinux support, but you'll just never really use it(but if you decide the option is there). hope this helps Justin P. Mattock -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: 70-persistent-net.rules not created
On 01/08/10 21:25, jmsc...@setex.ipcallback.com wrote: On Thu, Jan 07, 2010 at 11:15:23PM -0600, Bruce Dubbs wrote: jmsc...@setex.ipcallback.com wrote: Hello, I'm having a problem with configuring the network scripts in the LFS book 6.5. The file /etc/udev/rules.d/70-persistent-net.rules doesn't appear to be created. In particular, in chapter 7.13.1 is the blurb of code to pre-generate the udev rules file for NIC in /sys/class/net/* ; do INTERFACE=${NIC##*/} udevadm test --action=add $NIC done followed by the blurb to inspect the 70-persistent-net.rules file. cat /etc/udev/rules.d/70-persistent-net.rules The problem is that /etc/udev/rules.d/70-persistent-net.rules is not created. Did I miss running a command in the previous chapters ... perhaps /lib/udev/write_net_rules? Here is the link to the web page with the code in question http://www.linuxfromscratch.org/lfs/view/stable/chapter07/network.html Also, the host os is a freshly installed Centos 5.4 server in a virtualbox.com VM, all running under MAC Snow Leopard. Does the VM have a NIC? What are the results of ls /sys/class/net/ Yes, the NIC is configured and working in the host. I'm acccessing the host via ssh. The output of the ls and cat commands follow: root:~# ls /sys/class/net/ eth0 lo sit0 root:~# cat /etc/udev/rules.d/70-persistent-net.rules cat: /etc/udev/rules.d/70-persistent-net.rules: No such file or directory You should have a eth0. The udevadm test generates the 70-persistent-net.rules file: Nope, appears that /etc/udev/rules.d/70-persistent-net.rules is missing. I see many but no obvious errors in the output of of the 'for' loop in the blurb in paragraph 7.13.1: for NIC in /sys/class/net/* ; do INTERFACE=${NIC##*/} udevadm test --action=add $NIC done Perplexing. $ cat /etc/udev/rules.d/70-persistent-net.rules # This file was automatically generated by the /lib/udev/write_net_rules # program, run by the persistent-net-generator.rules rules file. # # You can modify it, as long as you keep each rule on a single # line, and change only the value of the NAME= key. # PCI device 0x8086:0x10de (e1000e) SUBSYSTEM==net, ACTION==add, DRIVERS==?*, ATTR{address}==00:25:64:38:ec:dd, ATTR{type}==1, KERNEL==eth*, NAME=eth0 -- Bruce -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page Since udev isn't critical to booting a kernel, I suppose I can keep soldiering forward, but I'd prefer to follow the letter of the Book ... -j you should be fine. over here I think I was hitting some odd quirk(but could be wrong), i.g. two iMac's both the same system yet one gave a long pause during boot(you gots me!!) Justin P. Mattock -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: 70-persistent-net.rules not created
On 01/07/10 16:07, jmsc...@setex.ipcallback.com wrote: Hello, I'm having a problem with configuring the network scripts in the LFS book 6.5. The file /etc/udev/rules.d/70-persistent-net.rules doesn't appear to be created. In particular, in chapter 7.13.1 is the blurb of code to pre-generate the udev rules file for NIC in /sys/class/net/* ; do INTERFACE=${NIC##*/} udevadm test --action=add $NIC done followed by the blurb to inspect the 70-persistent-net.rules file. cat /etc/udev/rules.d/70-persistent-net.rules The problem is that /etc/udev/rules.d/70-persistent-net.rules is not created. Did I miss running a command in the previous chapters ... perhaps /lib/udev/write_net_rules? Here is the link to the web page with the code in question http://www.linuxfromscratch.org/lfs/view/stable/chapter07/network.html Also, the host os is a freshly installed Centos 5.4 server in a virtualbox.com VM, all running under MAC Snow Leopard. Any help is appreciated. John Scott jmsc...@setspace.com that rule might already be there i.g. under /lib/udev/rules.d/* also check in the source code for this. Justin P. Mattock -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: Knoppix is missing the patch command
On 12/15/09 22:34, stosss wrote: uh,I have never use a live cd to build my LFS,it too troublesome! maybe you should install a linux distro . It is much better. Why do you say that? Just curious, because i used the LFS LiveCD 6.3 to build my first LFS 6.5 build. you can use a livecd, just remember that some packages will max out your memory because they expand. as for patch, sure. doo this all the time on livecd's Justin P. Mattock -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: Version 6.5 - 6.15.1 GCC compile failure
On 12/01/09 19:22, Sean Porterfield wrote: As background, I'm using VirtualBox with the LFS LiveCD. I haven't learned how to copy/paste from my real PC to the VirtualBox, so I'm typing everything by hand. After compiling and installing gcc, I tried to compile dummy.c but it failed. Any suggestions on how to figure out where I screwed up? dummy.log ends with: /usr/bin/ld: : No such file: No such file or directory collect2: ld returned 1 exit status Earlier in dummy.log I see an odd message that might be related: ignoring nonexistent directory /usr/lib/gcc/i686-pc-linux-gnu/4.4.1/../../../../i686-pc-linux-gnu/include -- Sean Porterfield you don't have ld installed(or have it in another location) I think it's binutils that provides that. (just look at the bottom of the tutorial for the list of apps they provide) then see if it goes through. Justin P. Mattock -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: Grub-1.97 problems
linux fan wrote: On 11/11/09, Bruce Dubbsbruce.du...@gmail.com wrote: # grub-install --grub-setup=/bin/true /dev/sda11 This should be /dev/sda, but I see you tried that too. Now I really _want_ to install grub on individual partitions so that I can chainloader to different installations like grub-legacy has let me do for years. grub-install /dev/sdb8 grub-setup: warn: Attempting to install GRUB to a partition instead of the MBR. This is a BAD idea. grub-setup: warn: Embedding is not possible. GRUB can only be installed in this setup by using blocklists. However, blocklists are UNRELIABLE and its use is discouraged. grub-setup: error: If you really want blocklists, use --force. USE THE FORCE, LUKE ... grub-setup --force /dev/sdb8 grub-setup: warn: Attempting to install GRUB to a partition instead of the MBR. This is a BAD idea. grub-setup: warn: Embedding is not possible. GRUB can only be installed in this setup by using blocklists. However, blocklists are UNRELIABLE and its use is discouraged. It didn't ... It it didn't put grub on /dev/sdb8 ... the grub-legacy is still there GNU Grub 0.97 Plus, grub-legacy used to let you setup grub from the grub command line is you had booted from a floppy, or a cd, or some other way. I think that is a major function loss. I usually do grub-setup on the main partition /dev/sda then on the os partition /dev/sda* (still uncertain which it should be though). Justin P. Mattock -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: LFS svn-20091012, 4.3 adding the lfs user
Philippe Delavalade wrote: Hi. Two weeks ago I tried to install LFS svn-20091012 on a computer with no internet connection... So my questions won't be very precise because I forgot one error message :-( The host was debian squeeze. Fisrt in 4.3 and 4.4, I didn't get the correct prompt although I used the -k /dev/null option in useradd ; .bash_profile was conform to 4.4. But my $PS1 was in fact '\...@\h:\w$ '. What to think about this ? Second, in 5.14, make install failed but I cann't remember the error message, maybe sothing like 'machine failure'... I stopped installing but I shall try another time in some months when I'm back to this holidays computer. And third, on that machine with no internet, is it possible to avoid installing packages which deal with networking as iana, inetutils or iproute ? Thanks for help and advices. -- Ph. Delavalade you can install those without internet. just download everything, or use the livCD that has all the packages on it. Justin P. Mattock -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: Make command is gone
pieter blomme wrote: Hi, I finally got to building the real LFS-environment. However, after I entered chroot, I can no longer issue the make command (bash: command not found). I retried everything from chapter 6 onwards, so the error isn't there. I suppose I made an error while compiling the make-package in chapter 5, but I ran all the test-suites in that chapter and they didn't produce any errors. Should I recompile that one anyhow? -- Groeten, Pieter Frontal Noize Mailorder http://frontalnoize.blogspot.com - http://www.myspace.com/frontalnoizedistribution I hit something like this similar. ended up (in my case) being the dynamic linker i.g. lib64 - lib Justin P. Mattock -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: 5.5.1 Installation of Cross GCC - 6.5 PDF Manual - Compilation Error
Owen Lopez wrote: Preparing GCC for Compilation was successful using the following params: ../gcc-4.4.1/configure \ ---target=$LFS_TGT --prefix=/tools \ --disable-nls --disable-shared --disable-multilib \ --disable-decimal-float --disable-threads \ --disable-libmudflap --disable-libssp \ --disable-libgomp --enable-languages=c I am also in a seperate gcc-build directory: /mnt/lfs/sources/gcc-build$ Error when trying to compile gcc using make: checking for i686-lfs-linux-gnu-gcc... /mnt/lfs/sources/gcc-build/./gcc/xgcc -B/mnt/lfs/sources/gcc-build/./gcc/ -B/tools/i686-lfs-linux-gnu/bin/ -B/tools/i686-lfs-linux-gnu/lib/ -isystem /tools/i686-lfs-linux-gnu/include -isystem /tools/i686-lfs-linux-gnu/sys-include checking for suffix of object files... configure: error: in `/mnt/lfs/sources/gcc-build/i686-lfs-linux-gnu/libgcc': configure: error: cannot compute suffix of object files: cannot compile See `config.log' for more details. make[1]: *** [configure-target-libgcc] Error 1 make[1]: Leaving directory `/mnt/lfs/sources/gcc-build' make: *** [all] Error 2 gcc -v returns l...@owen-laptop:/mnt/lfs/sources/gcc-build$ gcc -v Using built-in specs. Target: i486-linux-gnu Configured with: ../src/configure -v --with-pkgversion='Ubuntu 4.3.3-5ubuntu4' --with-bugurl=file:///usr/share/doc/gcc-4.3/README.Bugs --enable-languages=c,c++,fortran,objc,obj-c++ --prefix=/usr --enable-shared --with-system-zlib --libexecdir=/usr/lib --without-included-gettext --enable-threads=posix --enable-nls --with-gxx-include-dir=/usr/include/c++/4.3 --program-suffix=-4.3 --enable-clocale=gnu --enable-libstdcxx-debug --enable-objc-gc --enable-mpfr --enable-targets=all --with-tune=generic --enable-checking=release --build=i486-linux-gnu --host=i486-linux-gnu --target=i486-linux-gnu Thread model: posix gcc version 4.3.3 (Ubuntu 4.3.3-5ubuntu4) l...@owen-laptop:/mnt/lfs/sources/gcc-build$ Any help is appreciated, I did not skip any steps in the manual and read all IMPORTANT emphasized boxes. Currently logged in as the LFS user that I created with the custom bash profile for LFS and everything else the book told me to do, I am pretty sure I had no typos in the configuration. Attached is my config.log for GCC. Thanks in advanced to anyone who helps! not sure, but do you have the development packages install? (object files are located in /usr/lib/*.o) Justin P. Mattock -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: lfs 6.5 binutils
Geoff Howell wrote: Fairly certain I didn't, I followed the book religously (copy/paste instructions). I will of course check tonight. I guess a good way of checking is the install verbosity. Justin P. Mattock I've checked the tools directory that I backed up at the end of ch5, and the supposedly missing libs are there after all - but I can guarantee you that they were not there after the binutils install, so some other process has put them there. I didn't bother to check after I decided to push on... So, now for the next problem - I'll post it in a new mail. Thanks for your assistance. Geoff. Tough to say what happened.. in any case checking the verbosity is a good thing. Justin P. Mattock -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: lfs 6.5 binutils
Chris Staub wrote: On 11/01/2009 10:33 PM, Geoff Howell wrote: Hi everyone I'm new here, and first time trying to build a linux system. I'm trying to build lfs6.5 on a debian lenny system inside a ViritualBox vm. I'm following the book to the letter (copy/paste instructions rather than typing by hand), and have not got beyond glibc in chapter 6. My system meets or exceeds software requirements from version-check.sh. First problem: binutils in ch5 - it compiles successfully, but when I check what's installed I find no libopcodes or libbfd in /tools/lib (or anywhere else for that matter). So, is this an issue (ie are these lib files required or can I proceed without them?) Second problem: will bring it up after first problem is solved :) as it may be related. I don't think that my host system has these libs either (but can't confirm that right now as I'm at work). I've installed build-essential and linux headers (2.6.26-2). After much searching I can't find any answer on this. Does anyone have any idea why these libs are not installed? I can provide further information later this evening when I'm at home. Thanks Geoff I don't know for sure whether they are supposed to be in /tools, but...is there a specific reason you're asking? Are you getting errors? are you sure you didn't install libop* unto your host system i.g. DESTDIR=/some/location/otherthan/thehost Justin P. Mattock -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: lfs 6.5 binutils
Geoff Howell wrote: Justin, I may have misinterpreted your reply - are you asking whether I altered the lfs book's instructions to install libop* to a specific location other than it's default? If so, the answer is no, I haven't modified the book's instructions at all. I recall checking to see whether the result of my binutils installation looked like a) the debian 'lenny' installation I'm working from, and from my kubuntu host (on which my VirtualBox - hosted debian system is running) - I don't recall seeing the missing lib's anywhere on any of the systems - but will of course double check tonight. my bad(bad advice on my part). What I was asking is maybe you installed those libs in the same location of the host instead of /tool.(DESTDIR=* is how define the destination of where you want your new build installed, keep in mind libc use install_root=). I did follow the book, but then again was so anxious to get the thing booted I skipped parts here and there. Justin P. Mattock -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: lfs 6.5 binutils
Geoff Howell wrote: Justin, I may have misinterpreted your reply - are you asking whether I altered the lfs book's instructions to install libop* to a specific location other than it's default? If so, the answer is no, I haven't modified the book's instructions at all. I recall checking to see whether the result of my binutils installation looked like a) the debian 'lenny' installation I'm working from, and from my kubuntu host (on which my VirtualBox - hosted debian system is running) - I don't recall seeing the missing lib's anywhere on any of the systems - but will of course double check tonight. my bad(bad advice on my part). What I was asking is maybe you installed those libs in the same location of the host instead of /tool.(DESTDIR=* is how define the destination of where you want your new build installed, keep in mind libc use install_root=). I did follow the book, but then again was so anxious to get the thing booted I skipped parts here and there. Justin P. Mattock Fairly certain I didn't, I followed the book religously (copy/paste instructions). I will of course check tonight. -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page I guess a good way of checking is the install verbosity. Justin P. Mattock -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: Compiling GLIBC under chroot
Ken Turner wrote: When I try to compile the glibc-2.10.1 package I get the following error: error: bad value (switch) for -mtune= switch When I did my case statement before compiling, I added the parameter -mtune=native. This is what is in the document. Has anyone encourtered this and what did you do resolve it? Thanks David I ended up using -m64 -march=core2 -mtune=core2 (for an imac) not sure what you have, but here have a look to figure your processor: http://gcc.gnu.org/onlinedocs/gcc/i386-and-x86_002d64-Options.html Justin P. Mattock -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: Brick Wall on Network Device Congifuration
Dan McGhee wrote: Justin P. Mattock wrote: I'm wondering if it's the dhclient-script i.g. had something similar and it ended up being that.(threw in ubuntu's and bam internet). Just to make sure--you're talking about the one in /etc/sysconfing/network-devices/service right? Dan mine is located in /sbin/dhclient-script (same as debian/ubuntu) Justin P. Mattock -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: Stuck on Page 43 (chapter 5.8 Toolchain)
Ken Turner wrote: I am trying to the following and getting the subsequent error: lfs:/mnt/lfs/sources$ $LFS_TGT-gcc -B/tools/lib dummy.c/mnt/lfs/tools/bin/../lib/gcc/i686-lfs-linux-gnu/4.3.2/../../../../i686-lfs-linux-gnu/bin/ld: crt1.o: No such file: No such file or directory collect2: ld returned 1 exit status Can anyone point me the right direction. I tried google but the threads are pretty vague Thanks David your missing the object files in /usr/lib ls -l /usr/lib/*.o /usr/lib/Mcrt1.o /usr/lib/Scrt1.o /usr/lib/crt1.o /usr/lib/crti.o /usr/lib/crtn.o /usr/lib/gcrt1.o Justin P. Mattock -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: lfs 6.4 live on usb flash
William Immendorf wrote: On Wed, Sep 23, 2009 at 1:54 PM, William Immendorf will.immend...@gmail.com wrote: Good. Now, put LFS 6.5 on your system. Then, you are happy , ow. I meant if you put LFS 6.5 on your system, you will be happy. William haven't looked into livecd(although there was discussion) if you just want a quick transfer just: tar -pczf some_name.tar.gz bin boot etc home (and the rest of dirs) then unpack on the usb. as for the 500 mb that's depending on what you as the builder decides. keep in mind there probably is better ways of doing this. Justin P. Mattock -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: Host System
Marcus Wanner wrote: On 9/17/2009 11:12 PM, Shawn Eary wrote: Sir: I think that you will get more consistent and reliable results if you burn the LFS CD ISO to a disk and boot from that for the host http://www.linuxfromscratch.org/livecd/download.html In theory, I suppose building from an alternate host using an optimized gcc compiler could generate a slightly faster LFS target, but I don't think the minimal extra speed would be worth risking the chance of small corruptions along the way. Even if the original host system was optimized better than that of the tools on the LFS Live CD, I don't think that this would matter much because the LFS build seems be multi-pass. You construct temporary LFS tools that are used to compile the final LFS system. By the time the final LFS passes are done, I think that you are building with pretty good tools. This is just a wild oppinion, however, I may be missing something that others aren't. Best! Shawn Theoretically, the host system does not matter at all. That is why there are three passes: to get away from the host. If the host can make it through pass one without errors, then it shouldn't affect the resulting lfs system in the slightest. If you are really paranoid about it, I would think that clfs would get even farther from the host, since you are cross-compiling anyway. However, I'm relatively new to this, so I may not be correct. Marcus One thing that I experienced was the dynamic linker. (keep in mind I am a newbie) Somewhere along the way with one of my builds, is was I was trying to have the linker go in a certain direction. But ended up having the dynamic linker go in another direction; same as the host system.(Ill have to build another later on just to get things correct). In any case I agree, it doesn't matter what the host system is, as long as it supplies the right tools for you to get the job done, or built correctly.(the lfs livecd is ideal because it supplies all the tools, patches, packages etc...) Justin P. Mattock -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Check out my photos on Facebook
Hi lfs-support@linuxfromscratch.org, I set up a Facebook profile where I can post my pictures, videos and events and I want to add you as a friend so you can see it. First, you need to join Facebook! Once you join, you can also create your own profile. Thanks, Justin To sign up for Facebook, follow the link below: http://www.facebook.com/p.php?i=1358683573k=Z31T6VT5R4T16FBCRB2Tr lfs-support@linuxfromscratch.org was invited to join Facebook by Justin P. Mattock. If you do not wish to receive this type of email from Facebook in the future, please click on the link below to unsubscribe. http://www.facebook.com/o.php?k=438a11u=704633048mid=10cc6d2G29ffd8d8G0G8 Facebook's offices are located at 1601 S. California Ave., Palo Alto, CA 94304. -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: Problem compiling Glibc Lfs 6.5
Giorgio Gambino wrote: *i've resolved the problem modifyng this commands:* case `uname -m` in i?86) echo CFLAGS += -march=i686 -mtune=native configparms ;; esac cp -v nptl/sysdeps/unix/sysv/linux/i386/i686/lowlevellock.S{,.orig} sed -e 's/FUTEX_WAIT\( | FUTEX_CLOCK_REALTIME, reg\)/FUTEX_WAIT_BITSET\1/' \ nptl/sysdeps/unix/sysv/linux/i386/i686/lowlevellock.S.orig \ nptl/sysdeps/unix/sysv/linux/i386/i686/lowlevellock.S *now the compilation finishes with success but i don't know if my changes have jeopardized something* Cool. I'm not sure why some programs will fail like that.example mplayer would do that if I had the flags after ./configure but before configure mplayer would go through. Justin P. Mattock -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: dynamic linker problem with pure64
Bruce Dubbs wrote: Justin P. Mattock wrote: I think I screwed up here, when building a pure64 I used ubuntu x86_64(which is multilib) after compiling and building everything I chrooted into the directory to build the kernel. Im noticing I have some commands that work like mv, ls dmesg, and probably some others. but when I make menuconfig I receive a command not found. when I do ls -l /usr/bin/make I see: -rwxr-xr-x 1 root:root /usr/bin/make but ldd /usr/bin/make gives: not a dynamic executable (seems othere apps/libs are in the same state) My guess is during building some programs get hard wired(maybe by pkgconfig) to reading the link to lib which gets symlinked to lib64 (this is how ubuntu has it setup). is there a way to adjust this, or should I start over on a regular 686 and build a pure64 from there so the linkers get directed to /lib without having the multilib obstruction? I just went through this for the first time on a 64-bit system. I started with the Ubuntu 64-bit server and everything built fine. There were some problems in BLFS with xorg that we're trying to get straightened out but that was overcome. For LFS, I just followed the book (actually with jhalfs). bdu...@core2:~$ ldd /usr/bin/make linux-vdso.so.1 = (0x7fff95b3f000) librt.so.1 = /lib/librt.so.1 (0x7fe720da) libc.so.6 = /lib/libc.so.6 (0x7fe720a4d000) libpthread.so.0 = /lib/libpthread.so.0 (0x7fe720832000) /lib64/ld-linux-x86-64.so.2 (0x7fe720fa8000) I don't know what you did, but a pure 64-bit system can be done. -- Bruce Thanks for the response, at the moment I just did the multilib x86_64 without any of the 32bit libs just lib64 kinda like what fedora has done. seems to be o.k. Anyways: I think my mistake was not doing the adjusting the toolchain section.(causing everything to be linked wrong during compile time). Probably would have been o.k. if I was using a pure64 host system instead of the multilib/symlink system (but could be wrong). when doing a ldd I see the same as you see for the version of make on the new system: /lib64/ld-linux-x86-64.so.2 (0x7f07fae5) if this is the hard coded(irregardless of where make is at)in make, then this is why I'm not getting a proper reaction out of make, and most of these programs..(could probably be fixed by creating a symlink to a lib64, but for now I'm trying to avoid that. My best bet is to probably go back and readjust the toolchain then recompile and see if the system works as it should. Justin P. Mattock -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: dynamic linker problem with pure64
Bruce Dubbs wrote: Justin P. Mattock wrote: I think I screwed up here, when building a pure64 I used ubuntu x86_64(which is multilib) after compiling and building everything I chrooted into the directory to build the kernel. Im noticing I have some commands that work like mv, ls dmesg, and probably some others. but when I make menuconfig I receive a command not found. when I do ls -l /usr/bin/make I see: -rwxr-xr-x 1 root:root /usr/bin/make but ldd /usr/bin/make gives: not a dynamic executable (seems othere apps/libs are in the same state) My guess is during building some programs get hard wired(maybe by pkgconfig) to reading the link to lib which gets symlinked to lib64 (this is how ubuntu has it setup). is there a way to adjust this, or should I start over on a regular 686 and build a pure64 from there so the linkers get directed to /lib without having the multilib obstruction? I just went through this for the first time on a 64-bit system. I started with the Ubuntu 64-bit server and everything built fine. There were some problems in BLFS with xorg that we're trying to get straightened out but that was overcome. For LFS, I just followed the book (actually with jhalfs). bdu...@core2:~$ ldd /usr/bin/make linux-vdso.so.1 = (0x7fff95b3f000) librt.so.1 = /lib/librt.so.1 (0x7fe720da) libc.so.6 = /lib/libc.so.6 (0x7fe720a4d000) libpthread.so.0 = /lib/libpthread.so.0 (0x7fe720832000) /lib64/ld-linux-x86-64.so.2 (0x7fe720fa8000) I don't know what you did, but a pure 64-bit system can be done. -- Bruce Seems to get things working I have to create the symlink to lib64 (leave ld.so.conf empty) then issue ldconfig for everything to function properly. Is there a way to have just /lib /usr/lib or is this a no go on 64bit builds? Justin P. Mattock -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: Problem compiling Glibc Lfs 6.5
Giorgio Gambino wrote: *Hi when i try to compile glibc I receive this error:* mawk: scripts/gen-sorted.awk: line 19: regular expression compile failed (bad class -- [], [^] or [) /[^ mawk: scripts/gen-sorted.awk: line 19: syntax error at or near ] mawk: scripts/gen-sorted.awk: line 19: runaway regular expression /, , subd ... make[1]: Leaving directory `/mnt/lfs/sources/glibc-2.10.1' make[1]: Entering directory `/mnt/lfs/sources/glibc-2.10.1' mawk -f scripts/gen-sorted.awk \ -v subdirs='csu assert ctype locale intl catgets math setjmp signal stdlib stdio-common libio malloc string wcsmbs time dirent grp pwd posix io termios resource misc socket sysvipc gmon gnulib iconv iconvdata wctype manual shadow gshadow po argp crypt nss localedata timezone rt conform debug dlfcn elf' \ -v srcpfx='' \ nptl/sysdeps/pthread/Subdirs sysdeps/unix/inet/Subdirs sysdeps/unix/Subdirs assert/Depend intl/Depend catgets/Depend stdlib/Depend stdio-common/Depend libio/Depend malloc/Depend string/Depend wcsmbs/Depend time/Depend posix/Depend iconvdata/Depend nss/Depend localedata/Depend rt/Depend debug/Depend /mnt/lfs/sources/glibc-build/sysd-sorted-tmp mawk: scripts/gen-sorted.awk: line 19: regular expression compile failed (bad class -- [], [^] or [) /[^ mawk: scripts/gen-sorted.awk: line 19: syntax error at or near ] mawk: scripts/gen-sorted.awk: line 19: runaway regular expression /, , subd ... rm -f /mnt/lfs/sources/glibc-build/stamp.o; /mnt/lfs/sources/glibc-build/stamp.o rm -f /mnt/lfs/sources/glibc-build/stamp.os; /mnt/lfs/sources/glibc-build/stamp.os rm -f /mnt/lfs/sources/glibc-build/stamp.oS; /mnt/lfs/sources/glibc-build/stamp.oS cd /mnt/lfs/sources/glibc-build /mnt/lfs/tools/bin/../lib/gcc/i686-lfs-linux-gnu/4.4.1/../../../../i686-lfs-linux-gnu/bin/ar cruv libc.a `cat stamp.o` : /mnt/lfs/sources/glibc-build/libc.a cd /mnt/lfs/sources/glibc-build /mnt/lfs/tools/bin/../lib/gcc/i686-lfs-linux-gnu/4.4.1/../../../../i686-lfs-linux-gnu/bin/ar cruv libc_pic.a `cat stamp.os` : /mnt/lfs/sources/glibc-build/libc_pic.a cd /mnt/lfs/sources/glibc-build /mnt/lfs/tools/bin/../lib/gcc/i686-lfs-linux-gnu/4.4.1/../../../../i686-lfs-linux-gnu/bin/ar cruv libc_nonshared.a `cat stamp.oS` : /mnt/lfs/sources/glibc-build/libc_nonshared.a make[1]: *** No rule to make target `/mnt/lfs/sources/glibc-build/Versions.all', needed by `/mnt/lfs/sources/glibc-build/abi-versions.h'. Stop. make[1]: Leaving directory `/mnt/lfs/sources/glibc-2.10.1' make: *** [all] Error 2 *which problem can be? I have not changed anything i've followed the guide* *Thanks* do you have gawk installed, as well as bison, flex Justin P. Mattock -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: Problem compiling Glibc Lfs 6.5
Giorgio Gambino wrote: *Thanks i've resolved this problem installing gawk but there's another..* *sorry for the banality of these problems but I am a beginner :)* make[2]: Leaving directory `/mnt/lfs/sources/glibc-2.10.1/elf' i686-lfs-linux-gnu-gcc -shared -static-libgcc -Wl,-O1 -Wl,-z,defs -Wl,-dynamic-linker=/tools/lib/ld-linux.so.2 -B/mnt/lfs/sources/glibc-build/csu/ -Wl,--version-script=/mnt/lfs/sources/glibc-build/libc.map -Wl,-soname=libc.so.6 -Wl,-z,combreloc -Wl,-z,relro -Wl,--hash-style=both -nostdlib -nostartfiles -e __libc_main -L/mnt/lfs/sources/glibc-build -L/mnt/lfs/sources/glibc-build/math -L/mnt/lfs/sources/glibc-build/elf -L/mnt/lfs/sources/glibc-build/dlfcn -L/mnt/lfs/sources/glibc-build/nss -L/mnt/lfs/sources/glibc-build/nis -L/mnt/lfs/sources/glibc-build/rt -L/mnt/lfs/sources/glibc-build/resolv -L/mnt/lfs/sources/glibc-build/crypt -L/mnt/lfs/sources/glibc-build/nptl -Wl,-rpath-link=/mnt/lfs/sources/glibc-build:/mnt/lfs/sources/glibc-build/math:/mnt/lfs/sources/glibc-build/elf:/mnt/lfs/sources/glibc-build/dlfcn:/mnt/lfs/sources/glibc-build/nss:/mnt/lfs/sources/glibc-build/nis:/mnt/lfs/sources/glibc-build/rt:/mnt/lfs/sources/glibc-build/resolv:/mnt/lfs/sources/glibc-build/crypt:/mnt/lfs/sources/glibc-build/nptl -o /mnt/lfs/sources/glibc-build/libc.so -T /mnt/lfs/sources/glibc-build/shlib.lds /mnt/lfs/sources/glibc-build/csu/abi-note.o /mnt/lfs/sources/glibc-build/elf/soinit.os /mnt/lfs/sources/glibc-build/libc_pic.os /mnt/lfs/sources/glibc-build/elf/sofini.os /mnt/lfs/sources/glibc-build/elf/interp.os /mnt/lfs/sources/glibc-build/elf/ld.so -lgcc /mnt/lfs/sources/glibc-build/libc_pic.os: In function `__libc_fork': /mnt/lfs/sources/glibc-2.10.1/posix/../nptl/sysdeps/unix/sysv/linux/i386/../fork.c:79: undefined reference to `__sync_bool_compare_and_swap_4' /mnt/lfs/sources/glibc-build/libc_pic.os: In function `__nscd_drop_map_ref': /mnt/lfs/sources/glibc-2.10.1/nscd/nscd-client.h:320: undefined reference to `__sync_fetch_and_add_4' /mnt/lfs/sources/glibc-build/libc_pic.os: In function `nscd_getpw_r': /mnt/lfs/sources/glibc-2.10.1/nscd/nscd_getpw_r.c:232: undefined reference to `__sync_fetch_and_add_4' /mnt/lfs/sources/glibc-build/libc_pic.os: In function `__nscd_drop_map_ref': /mnt/lfs/sources/glibc-2.10.1/nscd/nscd-client.h:320: undefined reference to `__sync_fetch_and_add_4' /mnt/lfs/sources/glibc-build/libc_pic.os: In function `nscd_getgr_r': /mnt/lfs/sources/glibc-2.10.1/nscd/nscd_getgr_r.c:321: undefined reference to `__sync_fetch_and_add_4' /mnt/lfs/sources/glibc-build/libc_pic.os: In function `__nscd_drop_map_ref': /mnt/lfs/sources/glibc-2.10.1/nscd/nscd-client.h:320: undefined reference to `__sync_fetch_and_add_4' /mnt/lfs/sources/glibc-build/libc_pic.os:/mnt/lfs/sources/glibc-2.10.1/nscd/nscd_gethst_r.c:413: more undefined references to `__sync_fetch_and_add_4' follow /mnt/lfs/sources/glibc-build/libc_pic.os: In function `__nscd_get_map_ref': /mnt/lfs/sources/glibc-2.10.1/nscd/nscd_helper.c:432: undefined reference to `__sync_val_compare_and_swap_4' /mnt/lfs/sources/glibc-build/libc_pic.os: In function `*__GI___libc_freeres': /mnt/lfs/sources/glibc-2.10.1/malloc/set-freeres.c:39: undefined reference to `__sync_bool_compare_and_swap_4' collect2: ld returned 1 exit status make[1]: *** [/mnt/lfs/sources/glibc-build/libc.so] Error 1 make[1]: Leaving directory `/mnt/lfs/sources/glibc-2.10.1' make: *** [all] Error 2 Not sure, you can try and rm the glibc-build directory and start fresh to see(or just make clean). Justin P. Mattock -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: Problem compiling Glibc Lfs 6.5
Giorgio Gambino wrote: *nothing to do :( the compilation makes the same output*.. *i'm thinking that may be a cpu compatibility problem.. is it possible?* 2009/8/31 Justin P. Mattock justinmatt...@gmail.com mailto:justinmatt...@gmail.com Giorgio Gambino wrote: *Thanks i've resolved this problem installing gawk but there's another..* *sorry for the banality of these problems but I am a beginner :)* make[2]: Leaving directory `/mnt/lfs/sources/glibc-2.10.1/elf' i686-lfs-linux-gnu-gcc -shared -static-libgcc -Wl,-O1 -Wl,-z,defs -Wl,-dynamic-linker=/tools/lib/ld-linux.so.2 -B/mnt/lfs/sources/glibc-build/csu/ -Wl,--version-script=/mnt/lfs/sources/glibc-build/libc.map -Wl,-soname=libc.so.6 -Wl,-z,combreloc -Wl,-z,relro -Wl,--hash-style=both -nostdlib -nostartfiles -e __libc_main -L/mnt/lfs/sources/glibc-build -L/mnt/lfs/sources/glibc-build/math -L/mnt/lfs/sources/glibc-build/elf -L/mnt/lfs/sources/glibc-build/dlfcn -L/mnt/lfs/sources/glibc-build/nss -L/mnt/lfs/sources/glibc-build/nis -L/mnt/lfs/sources/glibc-build/rt -L/mnt/lfs/sources/glibc-build/resolv -L/mnt/lfs/sources/glibc-build/crypt -L/mnt/lfs/sources/glibc-build/nptl -Wl,-rpath-link=/mnt/lfs/sources/glibc-build:/mnt/lfs/sources/glibc-build/math:/mnt/lfs/sources/glibc-build/elf:/mnt/lfs/sources/glibc-build/dlfcn:/mnt/lfs/sources/glibc-build/nss:/mnt/lfs/sources/glibc-build/nis:/mnt/lfs/sources/glibc-build/rt:/mnt/lfs/sources/glibc-build/resolv:/mnt/lfs/sources/glibc-build/crypt:/mnt/lfs/sources/glibc-build/nptl -o /mnt/lfs/sources/glibc-build/libc.so -T /mnt/lfs/sources/glibc-build/shlib.lds /mnt/lfs/sources/glibc-build/csu/abi-note.o /mnt/lfs/sources/glibc-build/elf/soinit.os /mnt/lfs/sources/glibc-build/libc_pic.os /mnt/lfs/sources/glibc-build/elf/sofini.os /mnt/lfs/sources/glibc-build/elf/interp.os /mnt/lfs/sources/glibc-build/elf/ld.so -lgcc /mnt/lfs/sources/glibc-build/libc_pic.os: In function `__libc_fork': /mnt/lfs/sources/glibc-2.10.1/posix/../nptl/sysdeps/unix/sysv/linux/i386/../fork.c:79: undefined reference to `__sync_bool_compare_and_swap_4' /mnt/lfs/sources/glibc-build/libc_pic.os: In function `__nscd_drop_map_ref': /mnt/lfs/sources/glibc-2.10.1/nscd/nscd-client.h:320: undefined reference to `__sync_fetch_and_add_4' /mnt/lfs/sources/glibc-build/libc_pic.os: In function `nscd_getpw_r': /mnt/lfs/sources/glibc-2.10.1/nscd/nscd_getpw_r.c:232: undefined reference to `__sync_fetch_and_add_4' /mnt/lfs/sources/glibc-build/libc_pic.os: In function `__nscd_drop_map_ref': /mnt/lfs/sources/glibc-2.10.1/nscd/nscd-client.h:320: undefined reference to `__sync_fetch_and_add_4' /mnt/lfs/sources/glibc-build/libc_pic.os: In function `nscd_getgr_r': /mnt/lfs/sources/glibc-2.10.1/nscd/nscd_getgr_r.c:321: undefined reference to `__sync_fetch_and_add_4' /mnt/lfs/sources/glibc-build/libc_pic.os: In function `__nscd_drop_map_ref': /mnt/lfs/sources/glibc-2.10.1/nscd/nscd-client.h:320: undefined reference to `__sync_fetch_and_add_4' /mnt/lfs/sources/glibc-build/libc_pic.os:/mnt/lfs/sources/glibc-2.10.1/nscd/nscd_gethst_r.c:413: more undefined references to `__sync_fetch_and_add_4' follow /mnt/lfs/sources/glibc-build/libc_pic.os: In function `__nscd_get_map_ref': /mnt/lfs/sources/glibc-2.10.1/nscd/nscd_helper.c:432: undefined reference to `__sync_val_compare_and_swap_4' /mnt/lfs/sources/glibc-build/libc_pic.os: In function `*__GI___libc_freeres': /mnt/lfs/sources/glibc-2.10.1/malloc/set-freeres.c:39: undefined reference to `__sync_bool_compare_and_swap_4' collect2: ld returned 1 exit status make[1]: *** [/mnt/lfs/sources/glibc-build/libc.so] Error 1 make[1]: Leaving directory `/mnt/lfs/sources/glibc-2.10.1' make: *** [all] Error 2 Not sure, you can try and rm the glibc-build directory and start fresh to see(or just make clean). Justin P. Mattock -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page Normally undefined reference means that there is an changed/outdated *.h file in /usr/include. (or somewhere else). Justin P. Mattock -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
dynamic linker problem with pure64
I think I screwed up here, when building a pure64 I used ubuntu x86_64(which is multilib) after compiling and building everything I chrooted into the directory to build the kernel. Im noticing I have some commands that work like mv, ls dmesg, and probably some others. but when I make menuconfig I receive a command not found. when I do ls -l /usr/bin/make I see: -rwxr-xr-x 1 root:root /usr/bin/make but ldd /usr/bin/make gives: not a dynamic executable (seems othere apps/libs are in the same state) My guess is during building some programs get hard wired(maybe by pkgconfig) to reading the link to lib which gets symlinked to lib64 (this is how ubuntu has it setup). is there a way to adjust this, or should I start over on a regular 686 and build a pure64 from there so the linkers get directed to /lib without having the multilib obstruction? Justin P. Mattock -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: Make a bootable LFS iso.
Mike McCarty wrote: Justin P. Mattock wrote: alright, (didn't see that while doing a quick google). There has to be something simple to do this. I myself would like to create an .iso of what I created. So If I blow up my system it a simple insert the livecd and re-install, rather than starting from the beginning. Sorry to reply so late. I've been busy with other things. What you are describing is not making an ISO image of what you have built. What you are describing is a bare metal recovery backup, which is distinctly a different thing. There are several ways to do this. A very simple way is to boot using some LiveCD (I like Knoppix) and mount your file system(s) RW. Then you can fill them from /dev/zero until they are maxed out, and dd (or whatever you use) aborts with the disc full. Then you delete the huge file. This fills your partition with all zeroes, so things compress better. After that, you can simply use gzip (or you favorite compression algorithm) to compress the entire file system, and using any of several techniques, break it up into multiple CDROM or DVDROMs. Yakup shows one way to do this. There are several others. A somewhat more sophisticated technique is detailed in http://www.charlescurley.com/Linux-Complete-Backup-and-Recovery-HOWTO.html This allows you to bypass re installation and restoral of a backup, but doesn't absolutely tie you down to bare partition images. This may be the easiest thing for you to use. However, it also pretty much presumes very little actual hardware change. There are other packages out there, some free, some not so free, which promise to do it for you. However, if you want to understand what happens and be in control of how the backup and restore take place, then try designing your own, or follow the HOWTO I gave the link to above. Mike Cool, thanks for that. The main problem I have right now is if I use fedora or ubuntu, it just takes too long to load up(about 10min). I'm looking for something that I can load up, and have boot just as fast as a regular system or close to it, this way if I need to make a small fix it's not an all day event kinda thing. but will see, Ill have a look at the documentation that you had posted. Justin P. Mattock -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: Make a bootable LFS iso.
Mike McCarty wrote: Justin P. Mattock wrote: Cool, thanks for that. Welcome. The main problem I have right now is if I use fedora or ubuntu, it just takes too long to load up(about 10min). I'm looking for something that Perhaps I didn't understand what you need. You wrote I myself would like to create an .iso of what I created. So If I blow up my system it a simple insert the livecd and re-install, rather than starting from the beginning. which sounds like you need disaster recovery. I can load up, and have boot just as fast as a regular system or close to it, this way if I need to make a small fix it's not an all day event kinda thing. Are you saying that Fedora takes 10 mins to boot on your system? I use an old 2.4GHz machine, and FC2 (old, too, BTW) boots in about 2 mins. Something doesn't sound right to me. Why does Fedora take so long to boot? yeah, for some reason fedora, and ubuntu(both x86_64 latest) take about that time(not to impede their work). I think it might be my hardware(imac9,1). Anyway, if all you want is to be able to boot a system rescue CD, several are available for download and use. I like Knoppix. I haven't liked the 6.x release, but any of the 5.x releases should work well enough for you. For that matter, the LFS 6.3 LiveCD probably has all the tools you need. All you need is something to fix the PT, maybe run fsck in repair mode, and something which can mount your FS and run an editor. You don't need complicated tools, because you aren't discussing full system recovery, just a fixup for a screwup. from time to time, Ill find myself needing to use a rescue cd. example would be pam_namespace(if I don't set it correct). but will see, Ill have a look at the documentation that you had posted. What I posted is for disaster recovery, not quick boot and fix a simple problem. That's system rescue. Mike I would probably be in the area of system rescue. (just if I forget something, and the system gets in a jam, its easy to fix). Justin P. Mattock -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: Make a bootable LFS iso.
Mike McCarty wrote: Justin P. Mattock wrote: Cool, thanks for that. Welcome. The main problem I have right now is if I use fedora or ubuntu, it just takes too long to load up(about 10min). I'm looking for something that Perhaps I didn't understand what you need. You wrote I myself would like to create an .iso of what I created. So If I blow up my system it a simple insert the livecd and re-install, rather than starting from the beginning. which sounds like you need disaster recovery. I can load up, and have boot just as fast as a regular system or close to it, this way if I need to make a small fix it's not an all day event kinda thing. Are you saying that Fedora takes 10 mins to boot on your system? I use an old 2.4GHz machine, and FC2 (old, too, BTW) boots in about 2 mins. Something doesn't sound right to me. Why does Fedora take so long to boot? Anyway, if all you want is to be able to boot a system rescue CD, several are available for download and use. I like Knoppix. I haven't liked the 6.x release, but any of the 5.x releases should work well enough for you. For that matter, the LFS 6.3 LiveCD probably has all the tools you need. All you need is something to fix the PT, maybe run fsck in repair mode, and something which can mount your FS and run an editor. You don't need complicated tools, because you aren't discussing full system recovery, just a fixup for a screwup. but will see, Ill have a look at the documentation that you had posted. What I posted is for disaster recovery, not quick boot and fix a simple problem. That's system rescue. Mike First things first, I apologize for giving the wrong time information(I shouldn't base my timing by going outside for a smoke). So after getting my timer out here are the results: (livecd's) fedora 11 x86_64 = 4:43 ubuntu karmic x86_64 = 3:49 (systems on hard drive) my home brwed LFS = 1:03 osx = 00:36 It's not the 2:00 min mark, but it also is not the 10:00 min that I regret saying. I guess it's not so bad. Justin P. Mattock -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: Make a bootable LFS iso.
Bruce Dubbs wrote: Justin P. Mattock wrote: So after getting my timer out here are the results: (livecd's) fedora 11 x86_64 = 4:43 ubuntu karmic x86_64 = 3:49 Live CDs are always slow due to disk access time. (systems on hard drive) my home brwed LFS = 1:03 osx = 00:36 I don't know what you are doing for LFS. My P4 system takes about 20 seconds. My new Core2 Duo takes about 11 seconds to a boot prompt from the grub selection. -- Bruce There's a bit of a wait from pressing the powerbutton to when refit starts, to the grub2 menu. roughly around 00:30 or so.(Ill have to time that part to make sure). Justin P. Mattock -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: cannot find bash?
hmmm. maybe /etc/grouppasswd isn't correct. normal this type of error, from what I've experienced is from not having the symlink. (other than that not sure). Justin P. Mattock -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: with new system, how to run a test boot?
Bruce Dubbs wrote: Justin P. Mattock wrote: I like that idea, so you would have let's say 3 or 4 100mb test runs setup for multi booting. I like to use 10GB for the systems. 100MB is way too small. USe the 100MB partition for /boot as you have below. so a simple scheme would be like this: (using the system I have now I would do:) /dev/sda1 /boot(containing grub) /dev/sda2/ext3 (containing a x86_32 system) /dev/sda4/ext4 ( containing a x86_64 system). /dev/sda* Seems better to do a scheme like that. I thought I said that. However sda4 should be an extended partition so you can have logical partitions too. Also you may want a swap partition that can be shared by every system too. -- Bruce Yeah you did say that, I ended up writing what I was thinking.. and not thinking as I was sending. Aside from that 10G is probably more than enough. Justin P. Mattock -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: with new system, how to run a test boot?
Nice, so it is possible to do this. I'll give this a try. Russell Stockhammer wrote: You can't boot into a sub-directory of a file system but you could do the following 1) Configure grub to boot the kernel in the /mnt/lfs directory with the current root file system as a the root directory 2) Boot grub and pass the command init=/mnt/lfs/bin/sh this will run the LFS bash shell instead of the current/host init. 3) Once the kernel has booted and you are dropped into the shell run; exec chroot /mnt/lfs exec /sbin/init. This will chroot into the /mnt/lfs system and start init as if the kernel started it at boot. NOTE: the exec is important because init -MUST- be run as PID 1. Russ Date: Sun, 16 Aug 2009 21:37:05 -0700 From: justinmatt...@gmail.com To: lfs-support@linuxfromscratch.org Subject: with new system, how to run a test boot? quick question, with a new fresh system in the /where directory is there a way to adjust grub on the host system to actually boot the new system, before moving the newly created system to / Justin P. Mattock -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page Web IM has arrived! Use Windows Live Messenger from your Hotmail inbox http://windowslive.ninemsn.com.au/article.aspx?id=823454 -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: with new system, how to run a test boot?
cool, thanks for the info. I'll have a look and see if I can do this. Robert A. Lerche wrote: There's an LFS hint describing how to boot LFS without requiring a separate partition (i.e., in the same file system as another operating system). The trick is a special pre-init program that does a chroot early in the boot process (automatically, rather than manually as Russell Stockhammer suggests). See http://www.linuxfromscratch.org/hints/downloads/files/lfs_next_to_existing_systems.txt -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: with new system, how to run a test boot?
Michael Tsang wrote: On Monday 17 August 2009 15:15:39 Russell Stockhammer wrote: You can't boot into a sub-directory of a file system but you could do the following 1) Configure grub to boot the kernel in the /mnt/lfs directory with the current root file system as a the root directory 2) Boot grub and pass the command init=/mnt/lfs/bin/sh this will run the LFS bash shell instead of the current/host init. 3) Once the kernel has booted and you are dropped into the shell run; exec chroot /mnt/lfs exec /sbin/init. This will chroot into the /mnt/lfs system and start init as if the kernel started it at boot. NOTE: the exec is important because init -MUST- be run as PID 1. You can create a stub executable placed in anywhere: exec chroot /mnt/lfs exec /sbin/init save it and pass it to the init parameter at the kernel Cool. at the moment I never had a chance to try any of these procedures out.(don't worry I will) seems 2.6.31-rc6 is stuck during boot, spent most of the day trying to locate the commit that causes this stuckage. (seems to only affect x86_64, as my other system boots fine). Justin P. Mattock -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
with new system, how to run a test boot?
quick question, with a new fresh system in the /where directory is there a way to adjust grub on the host system to actually boot the new system, before moving the newly created system to / Justin P. Mattock -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: with new system, how to run a test boot?
Bruce Dubbs wrote: Justin P. Mattock wrote: quick question, with a new fresh system in the /where directory is there a way to adjust grub on the host system to actually boot the new system, before moving the newly created system to / You don't give particulars but if you built on a separate partition, than you just set up that partition as root: title LFS 6.5 root (hd0,4) kernel /boot/linux-2.6.30.2 root=/dev/sda5 Personally, I like to set up a separate 100MB partition that is mounted as /boot and then all my builds put the kernels in the same place. Then the grub configuration looks like: itle LFS-dev-2.6.12.5-20051115 root (hd0,2) kernel /linux-2.6.12.5-20071115 root=/dev/sda5 title LFS-dev-2.6.22.5 root (hd0,2) kernel /linux-2.6.22.5 root=/dev/sda6 etc. In this case the /boot partition is sda3. -- Bruce I like that idea, so you would have let's say 3 or 4 100mb test runs setup for multi booting. so a simple scheme would be like this: (using the system I have now I would do:) /dev/sda1 /boot(containing grub) /dev/sda2/ext3 (containing a x86_32 system) /dev/sda4/ext4 ( containing a x86_64 system). /dev/sda* Seems better to do a scheme like that. Justin P. Mattock -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
any ideas where to find a patch for sysvinit: load_policy
when building sysvinit, if wanting to add SELinux support there needs to be a patch applied(or at least I think) so that the policy can load at a certain time during boot. I've tried certain patches from googling but haven't found one yet that really works. where might I look for such a patch? Justin P. Mattock -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: creating ext4 filesystem for a new lfs build
Michael Tsang wrote: On Tuesday 04 August 2009 14:06:59 Justin P. Mattock wrote: Michael Tsang wrote: On Monday 03 August 2009 20:04:16 Ken Moffat wrote: 2009/8/3 Justin P. Mattockjustinmatt...@gmail.com: On x86_64, I use pure64. But then, I'm happy to use lilo and I've no need of pre-built binaries. Certainly, building multilib is very educational (when things go wrong). ĸen I always build pure64 and always use the boot loader from the host. I don't want to use lilo because it has been deprecated for a long time in favour of GRUB. I think multilib is a waste of both disk space and manpower. Because the whole system is built from source, all the static libraries and 32-bit libraries are useless. I'm seeing something interesting here, while creating the partition table I'm noticing mke2fs -t ext4 /dev/* goes really fast, then loading the filesystem up I'm only allowed 200mb or so. Am I seeing this correctly, for the boot partition I have to create ext3 then converted to ext4 in order to use ext4 (not partitioned to only ext4) Justin P. Mattock For GRUB legacy, you need to patch the source to have ext4, build it, and run install-grub *BEFORE* you convert the /boot partition to ext4 Hey alright, after pain and anguish I finally startedX mostly everything is in /lib64 and /usr/lib64 (not sure if this a pure64 or partially pure64) overall what a pain in the a** I think I'm going to stick with just regular builds. I owe CBLFS a whole lot of thanks, without their tutorial I would of been stuck at the beginning. regards, Justin P. Mattock -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: creating ext4 filesystem for a new lfs build
I must have been tired or something, for some reason fdisk was not erasing the old partitions(even though they were gone, and had rebooted) and only reformatting the partition to the previous size that was supposed to be deleted. (under desperation ended up just using ubuntu to reformat the partitions to get it to work properly). Justin P. Mattock -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: creating ext4 filesystem for a new lfs build
Michael Tsang wrote: On Tuesday 04 August 2009 14:06:59 Justin P. Mattock wrote: Michael Tsang wrote: On Monday 03 August 2009 20:04:16 Ken Moffat wrote: 2009/8/3 Justin P. Mattockjustinmatt...@gmail.com: On x86_64, I use pure64. But then, I'm happy to use lilo and I've no need of pre-built binaries. Certainly, building multilib is very educational (when things go wrong). ĸen I always build pure64 and always use the boot loader from the host. I don't want to use lilo because it has been deprecated for a long time in favour of GRUB. I think multilib is a waste of both disk space and manpower. Because the whole system is built from source, all the static libraries and 32-bit libraries are useless. I'm seeing something interesting here, while creating the partition table I'm noticing mke2fs -t ext4 /dev/* goes really fast, then loading the filesystem up I'm only allowed 200mb or so. Am I seeing this correctly, for the boot partition I have to create ext3 then converted to ext4 in order to use ext4 (not partitioned to only ext4) Justin P. Mattock For GRUB legacy, you need to patch the source to have ext4, build it, and run install-grub *BEFORE* you convert the /boot partition to ext4 hey alright she boots-up. With not knowing what to do for grub I ended up using grub2 which seems to boot ext4 without having to partition to ext3 and converting to ext4. in dmesg the only issue I see is: EXT2-fs: sda3: couldn't mount because of unsupported optional feature (240) EXT3-fs: sda3: couldn't mount because of unsupported optional feature (244) then goes on to mounting ext4 (Anyways I'm going to see how much of a pain it is to build the xserver x86_64, bit just to see). Justin P. Mattock -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: creating ext4 filesystem for a new lfs build
Michael Tsang wrote: On Monday 03 August 2009 20:04:16 Ken Moffat wrote: 2009/8/3 Justin P. Mattockjustinmatt...@gmail.com: On x86_64, I use pure64. But then, I'm happy to use lilo and I've no need of pre-built binaries. Certainly, building multilib is very educational (when things go wrong). ĸen I always build pure64 and always use the boot loader from the host. I don't want to use lilo because it has been deprecated for a long time in favour of GRUB. I think multilib is a waste of both disk space and manpower. Because the whole system is built from source, all the static libraries and 32-bit libraries are useless. luckily I have a copy of the other lfs build, so if this crashes and burns, well that's o.k. (just curious about building x86_64 due to never building that type before). So I have the decision of using pure 64 bit libs only, or mixing 32bit and 64bit? As for the manpower part(s^it), a bit tricky and time consuming.(took all day to build most of the packages, on the lfs build seemed not as long due to not worrying where the locations of the libs go). In any case if this things boots it boots, if it boots but acts sluggish, or any oddities, Ill revert using it. Justin P. Mattock -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: creating ext4 filesystem for a new lfs build
Simon Geard wrote: On Sun, 2009-08-02 at 02:54 +0100, Ken Moffat wrote: Pah! Why do you waste space on /boot iwth a journal ? ext2 for /boot! Also, swap could be optional if you have enough real memory and don't intend to suspend to disk. Can't speak for Justin, but I have roughly half a terabyte of disk in my machine, about 80% of which is unused. I feel no particular need to worry about wasting a hundred megabytes or so... Why do I have so much unused space? Because a 500GB drive is about the smallest and cheapest disk you can find these days... Simon funny thing with that is I decided to do lfs/blfs except leave every package on the machine and not clean them etc.. after compiling all of gnome etc.. I had only used 15 gigs or so..(as I look to burn off 500 gig's realizing I didn't even scratch the surface, I guess movies and mp3's are what do it if your curious in burning off disk space). Anyways as for creating ext4 and x86_64 system, I started yesterday, and everything seems good. /lib64 and /usr/lib64 seem simple enough to have the packages go there. but hopefully It's not to tough to make it to the xserver etc.. As for the build should I have used x86_64-unkown-linux-gnu or x86_64-pc-linux-gnu Justin P. Mattock -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: creating ext4 filesystem for a new lfs build
Ken Moffat wrote: 2009/8/2 Justin P. Mattockjustinmatt...@gmail.com: Anyways as for creating ext4 and x86_64 system, I started yesterday, and everything seems good. /lib64 and /usr/lib64 seem simple enough to have the packages go there. but hopefully It's not to tough to make it to the xserver etc.. As for the build should I have used x86_64-unkown-linux-gnu or x86_64-pc-linux-gnu Justin P. Mattock -- Since you are building multilib, you really need to check what is in cblfs - for my first few multilib builds I had all sorts of weird and unrepeatable problems, partly because I never saw a need to build multiple sizes of everything. You youngsters with terrabyte disks and multiple CPUs probably don't have a problem with that ;-) In particular, imake (for xmkmf) needs to be available in both sizes. Nowadays, I've given up on multilib because for me it doesn't add enough benefits to compensate for the extended build. YMMV. ĸen Cool thanks, at the moment I'm getting the hang of /lib64 /usr/lib64 seems a bit challenging to build things like perl-5.10.0. Thanks to you guys having the cblfs available I was able to do that right.(if not perl would of been creating /usr/lib every time). All of this lib64 stuff has me wondering do I really need this? i.g. if I set the compiler flags to x86_64 will the system be x86_64 without the need for lib64's, or do I need the lib64's to work with things like firefox(prebuilt x86_64)? Seems a bit faster to build regularly without having to worry about the whole better make sure the lib is set to the right location. (anyways never have done this, so might as well). Justin P. Mattock -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
creating ext4 filesystem for a new lfs build
After playing around with fedora 11 I notice they are using ext4. So after thinking.. how difficult is this todo on lfs? from installing I noticed I needed (partition for ext4 to work). /dev/sda3 ext3 200mb for /boot /dev/sda4 ext4 * mb for / /dev/sda5 swap 200 mb swap was all that is needed for the filesystem to load the O.S. Justin P. Mattock -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: creating ext4 filesystem for a new lfs build
Ken Moffat wrote: 2009/8/2 Justin P. Mattockjustinmatt...@gmail.com: After playing around with fedora 11 I notice they are using ext4. So after thinking.. how difficult is this todo on lfs? from installing I noticed I needed (partition for ext4 to work). /dev/sda3 ext3 200mb for /boot /dev/sda4 ext4 * mb for / /dev/sda5 swap 200 mb swap Pah! Why do you waste space on /boot iwth a journal ? ext2 for /boot! Also, swap could be optional if you have enough real memory and don't intend to suspend to disk. For me, the only downside to ext4 is that it needs a *modern* kernel (I suspect that means 2.6.30 or later, but perhaps 2.6.29 was ok), which means I can't yet use it on /home which is shared with older systems. Certainly, trying to create an ext4 system for the next build from a slightly too old LFS system was not particularly pleasant - had to upgrade both e2fsprogs and util-linux-ng (as well as the kernel) before I was able to create and mount an ext4 rootfs. Now, it should be a walk in the park :-) ĸen This was what I ended up using to get fedora installed. (didn't really spend too much time adding) From doing a quick google it sounds like you can either move ext3 to ext4 using specific tools. or clear the board reformat to ext4 then fill it up with libs/apps.(seems this approach gives you the most performance) As for the /boot thing seems there is a patch for grub2 to not need ext3 to boot to ext4. for the swap I have 300mb allocated for that. (the above partition scheme was for a new machine). I think what I might try is reformatting to ext4 and then use the tutorial to create a new system. (fedora 11 is nice, but this lfs building is addictive). Only real issue I see is I need to create an x86_64 system (need to find the right cflags and such). Justin P. Mattock -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: Error when trying make the kernel
Bruce Dubbs wrote: Justin P. Mattock wrote: Justin, Trim your posts to the minimum needed to reply. -- Bruce Alright. (posting on lkml teaches me to do this) I'll write a post-it note so I don't do this for LFS. Justin P. Mattock -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: Make a bootable LFS iso.
khaled gouaich wrote: Hi Mike, Thanks for the advice. I'll look for more informations about making LiveCD. If you get some time, please can you show me the way how you made it. 2009/7/27 Mike McCartymike.mcca...@sbcglobal.net: khaled gouaich wrote: Hi all, I need help for making LFS bootable iso. I have already googled about how making an bootable iso but i have no idea how should i process -_- If some one did it yet, please i need to know all the steps. Thanks. You could read about AnyLive and get some thoughts there. I went and looked briefly, but couldn't find the information quickly on the web. I've got it on my machine, but I've got no time right now. Mike -- p=p=%c%s%c;main(){printf(p,34,p,34);};main(){printf(p,34,p,34);} Oppose globalization and One World Governments like the UN. This message made from 100% recycled bits. You have found the bank of Larn. I speak only for myself, and I am unanimous in that! -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page There's that app called mastersys or something like that. Justin P. Mattock -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: Make a bootable LFS iso.
Smartboy wrote: On Tue, Jul 28, 2009 at 7:21 AM, Justin P. Mattock justinmatt...@gmail.com mailto:justinmatt...@gmail.com wrote: khaled gouaich wrote: Hi Mike, Thanks for the advice. I'll look for more informations about making LiveCD. If you get some time, please can you show me the way how you made it. 2009/7/27 Mike McCartymike.mcca...@sbcglobal.net mailto:mike.mcca...@sbcglobal.net: khaled gouaich wrote: Hi all, I need help for making LFS bootable iso. I have already googled about how making an bootable iso but i have no idea how should i process -_- If some one did it yet, please i need to know all the steps. Thanks. You could read about AnyLive and get some thoughts there. I went and looked briefly, but couldn't find the information quickly on the web. I've got it on my machine, but I've got no time right now. Mike -- p=p=%c%s%c;main(){printf(p,34,p,34);};main(){printf(p,34,p,34);} Oppose globalization and One World Governments like the UN. This message made from 100% recycled bits. You have found the bank of Larn. I speak only for myself, and I am unanimous in that! -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page There's that app called mastersys or something like that. Justin P. Mattock -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page Remastersys is *buntu/Debian-only, which means if your distro isn't compatible with them, I don't think it will work. On Tue, Jul 28, 2009 at 7:32 AM, Robert A. Lerche r...@msbit.com mailto:r...@msbit.com wrote: I need help for making LFS bootable iso. I have already googled about how making an bootable iso but i have no idea how should i process -_- If some one did it yet, please i need to know all the steps. Thanks. I recently built a custom LFS live CD for the Open Source Digital Voting Foundation. I updated the live CD makefiles to reflect LFS 6.4 (plus a few other changes) and added an application program (Pvote) and its dependencies (Python, SDL, pygame). I made it boot to the root prompt (with an empty root password), or just run the demo (with a different /etc/inittab). You can fetch my build it sources (that is, the updated makefiles, my notes, etc.) from http://msbit.com/osdv/osdvlfs.tgz The .iso is there too (50M), but msbit.com http://msbit.com/ is on a slow link, so be patient. The OSDV Wiki moved recently and isn't quite working yet. Please feel free to email me with any questions. -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page I am going to have a look at this myself, as I wanted a 6.4 livecd, though the SVN branch doesn't want to agree. Thanks for this! alright, (didn't see that while doing a quick google). There has to be something simple to do this. I myself would like to create an .iso of what I created. So If I blow up my system it a simple insert the livecd and re-install, rather than starting from the beginning. Justin P. Mattock -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: Make a bootable LFS iso.
Smartboy wrote: On Tue, Jul 28, 2009 at 8:25 AM, Justin P. Mattock justinmatt...@gmail.com mailto:justinmatt...@gmail.com wrote: Smartboy wrote: On Tue, Jul 28, 2009 at 7:21 AM, Justin P. Mattock justinmatt...@gmail.com mailto:justinmatt...@gmail.com mailto:justinmatt...@gmail.com mailto:justinmatt...@gmail.com wrote: khaled gouaich wrote: Hi Mike, Thanks for the advice. I'll look for more informations about making LiveCD. If you get some time, please can you show me the way how you made it. 2009/7/27 Mike McCartymike.mcca...@sbcglobal.net mailto:mike.mcca...@sbcglobal.net mailto:mike.mcca...@sbcglobal.net mailto:mike.mcca...@sbcglobal.net: khaled gouaich wrote: Hi all, I need help for making LFS bootable iso. I have already googled about how making an bootable iso but i have no idea how should i process -_- If some one did it yet, please i need to know all the steps. Thanks. You could read about AnyLive and get some thoughts there. I went and looked briefly, but couldn't find the information quickly on the web. I've got it on my machine, but I've got no time right now. Mike -- p=p=%c%s%c;main(){printf(p,34,p,34);};main(){printf(p,34,p,34);} Oppose globalization and One World Governments like the UN. This message made from 100% recycled bits. You have found the bank of Larn. I speak only for myself, and I am unanimous in that! -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page There's that app called mastersys or something like that. Justin P. Mattock -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page Remastersys is *buntu/Debian-only, which means if your distro isn't compatible with them, I don't think it will work. On Tue, Jul 28, 2009 at 7:32 AM, Robert A. Lerche r...@msbit.com mailto:r...@msbit.com mailto:r...@msbit.com mailto:r...@msbit.com wrote: I need help for making LFS bootable iso. I have already googled about how making an bootable iso but i have no idea how should i process -_- If some one did it yet, please i need to know all the steps. Thanks. I recently built a custom LFS live CD for the Open Source Digital Voting Foundation. I updated the live CD makefiles to reflect LFS 6.4 (plus a few other changes) and added an application program (Pvote) and its dependencies (Python, SDL, pygame). I made it boot to the root prompt (with an empty root password), or just run the demo (with a different /etc/inittab). You can fetch my build it sources (that is, the updated makefiles, my notes, etc.) from http://msbit.com/osdv/osdvlfs.tgz The .iso is there too (50M), but msbit.com http://msbit.com http://msbit.com/ is on a slow link, so be patient. The OSDV Wiki moved recently and isn't quite working yet. Please feel free to email me with any questions. -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page I am going to have a look at this myself, as I wanted a 6.4 livecd, though the SVN branch doesn't want to agree. Thanks for this! alright, (didn't see that while doing a quick google). There has to be something simple to do this. I myself would like to create an .iso of what I created. So If I blow up my system it a simple insert the livecd and re-install, rather than starting from the beginning. Justin P. Mattock -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page The problem with installations from LiveCD is that you need to change some things for a system to run as a livecd, and change them back when you run from a normal install. If you wanted something like this, you would probably want to go with jhalfs, which automates LFS, BLFS, etc. Just edit the config files, and it will build it for you. :) At the moment I just used cp -p to an external usb stick. (this way I can always copy everything to the disk if need be). I'm wondering if loading an ubuntu livecd then installing remstersys(on the livecd) might work
Re: Error when trying make the kernel
William Immendorf wrote: 2009/7/28 peterpe...@bios.co.jp: 2.6.27.4/usr/bin/ld:warning:cannotfindentrysymbol_start /bin/sh: defaultingto0804080b8: command not found make: *** [include/config/kernel.release] Error 127 Sorry, your LFS text is a bit mangled. Try the development version, it WFM on newer hosts. William I'm guessing /bin/sh is not present. under the instructions it says to create a symlink: ln -fs /bin/sh bash or /bin/bash sh (can't remember which way it goes). Justin P. Mattock -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: Error when trying make the kernel
peter wrote: Hi Justin I tried to create a symlink as you said, but still getting exact same messages: 2.6.27.4/usr/bin/ld:warning:cannotfindentrysymbol_start /bin/sh: defaultingto0804080b8: command not found make: *** [include/config/kernel.release] Error 127 I also borrowed .config file from old one incase menuconfig didn't work well - Original Message - From: Justin P. Mattockjustinmatt...@gmail.com To: LFS Support Listlfs-support@linuxfromscratch.org Sent: Wednesday, July 29, 2009 9:49 AM Subject: Re: Error when trying make the kernel William Immendorf wrote: 2009/7/28 peterpe...@bios.co.jp: 2.6.27.4/usr/bin/ld:warning:cannotfindentrysymbol_start /bin/sh: defaultingto0804080b8: command not found make: *** [include/config/kernel.release] Error 127 Sorry, your LFS text is a bit mangled. Try the development version, it WFM on newer hosts. William I'm guessing /bin/sh is not present. under the instructions it says to create a symlink: ln -fs /bin/sh bash or /bin/bash sh (can't remember which way it goes). Justin P. Mattock -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page Not sure whats happening. you compiled everything from the tutorial? you have the kernel headers in /usr/include/* main issue I had while compiling the kernel was making sure the headers were there, and binutils was installed. are you hitting this as soon as you do: make menuconfig or this hits while compiling? Justin P. Mattock -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: Error when trying make the kernel
peter wrote: Oh that error happens while compiling, after menuconfig. that was why I was wondering if I made mistake with config settings. Is there a site I can get detail info about the kernel setting using menuconfig? Peter - Original Message - From: Justin P. Mattockjustinmatt...@gmail.com To: LFS Support Listlfs-support@linuxfromscratch.org Sent: Wednesday, July 29, 2009 12:13 PM Subject: Re: Error when trying make the kernel peter wrote: Hi Justin I tried to create a symlink as you said, but still getting exact same messages: 2.6.27.4/usr/bin/ld:warning:cannotfindentrysymbol_start /bin/sh: defaultingto0804080b8: command not found make: *** [include/config/kernel.release] Error 127 I also borrowed .config file from old one incase menuconfig didn't work well - Original Message - From: Justin P. Mattockjustinmatt...@gmail.com To: LFS Support Listlfs-support@linuxfromscratch.org Sent: Wednesday, July 29, 2009 9:49 AM Subject: Re: Error when trying make the kernel William Immendorf wrote: 2009/7/28 peterpe...@bios.co.jp: 2.6.27.4/usr/bin/ld:warning:cannotfindentrysymbol_start /bin/sh: defaultingto0804080b8: command not found make: *** [include/config/kernel.release] Error 127 Sorry, your LFS text is a bit mangled. Try the development version, it WFM on newer hosts. William I'm guessing /bin/sh is not present. under the instructions it says to create a symlink: ln -fs /bin/sh bash or /bin/bash sh (can't remember which way it goes). Justin P. Mattock -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page Not sure whats happening. you compiled everything from the tutorial? you have the kernel headers in /usr/include/* main issue I had while compiling the kernel was making sure the headers were there, and binutils was installed. are you hitting this as soon as you do: make menuconfig or this hits while compiling? Justin P. Mattock -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page Tough to say what might be happening, Especially if you followed the tutorial. Have you checked your install with bash. also maybe the chroot environment might have something to do with this error. From examining the error: 2.6.27.4/usr/bin/ld:warning:cannotfindentrysymbol_start something with libc /bin/sh: defaultingto0804080b8: command not found maybe the environment isn't set correctly. make: *** [include/config/kernel.release] Error 127 maybe make is compile wrong. but then again could be something completely different. Any ideas anybody? (I'm lost on this one!!) Justin P. Mattock -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: Glib-2.5.1 Error (Need linker with .init_array/.fini_array support.)
Chris Staub wrote: Jomon John wrote: Am building LFS using the Sources from *lfslivecd-x86-6.3-r2160.iso* with *LFS-BOOK 6.3*.My host system is Loaded with *Ubuntu 9.04 * During the configuration of *Glibc-2.5.1* in the *chapter 5.6* i got following Warnings finally a error.. *** WARNING: You should not compile GNU libc without versioning. Not using *** versioning will introduce incompatibilities so that old binaries *** will not run anymore. *** For versioning you need recent binutils (binutils-2.8.1.0.23 or newer). . . . checking for .preinit_array/.init_array/.fini_array support... no *configure: error: Need linker with .init_array/.fini_array support*. Please Help me This is usually due to an incorrect PATH. ~~~ The Complete Output of configuration is given below... l...@alexa:/mnt/lfs/glibc-build$ sudo ../glibc-2.5.1/configure --prefix=/tools \ --disable-profile --enable-add-ons \ --enable-kernel=2.6.0 --with-binutils=/tools/bin \ --without-gd --with-headers=/tools/include \ --without-selinux [sudo] password for lfs: The book does not say to use sudo, and it is not needed if you are following the directions. The glibc-build dir should be in $LFS/sources. My guess is you overlooked an instruction on page 5.1. The first Important note explains that before you do anything on a package installation page, you first unpack the source and cd into the created source dir. Therefore, you should be in the Glibc source dir when you create the build dir. Your problem is likely due to the use of sudo - when sudo elevates you to root, the necessary lfs user environment is not used. never, ever compile with sudo(golden rule).. as for this type of error, experience something similar with the xserver ended I needed to use sh ./autogen.sh to set it right before compiling. (but this might not be the case for you). hope this helps. Justin P. Mattock -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: lfs-support Digest, Vol 1747, Issue 2
whether .text pseudo-op must be used... yes checking for assembler global-symbol directive... .globl checking for .set assembler directive... no checking for assembler .type directive prefix... @ checking for .symver assembler directive... yes checking for ld --version-script... no *** WARNING: You should not compile GNU libc without versioning. Not using *** versioning will introduce incompatibilities so that old binaries *** will not run anymore. *** For versioning you need recent binutils (binutils-2.8.1.0.23 or newer). checking for .previous assembler directive... yes checking for .protected and .hidden assembler directive... yes checking whether __attribute__((visibility())) is supported... yes checking for broken __attribute__((visibility()))... no checking for broken __attribute__((alias()))... no checking whether to put _rtld_local into .sdata section... no checking for .preinit_array/.init_array/.fini_array support... no configure: error: Need linker with .init_array/.fini_array support. But when it comes to personale preference, some people still enjoy building kernel 2.4.* and having older packages. (up to you). :^) Justin P. Mattock -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: sysvinit and inittab to shutdown
Bruce Dubbs wrote: Justin P. Mattock wrote: Any ideas on why ctl+alt+del doesn't shutdown my machine? the inittab I copied was from 6.54.3 which has the entry there, but when exiting the xserver and issuing those commands, I have no movement.(also my user name is in shutdown.allow) Do you have ca:12345:ctrlaltdel:/sbin/shutdown -t1 -a -r now in /etc/inittab? -- Bruce O.k. after messing around with us.map.gz (and at one point not having the correct keys while typing!!) I ended up finding this procedure which works easy: http://humanreadable.nfshost.com/howtos/inittab_shutdown.htm Now I can shutdown the machine, without having to use sudo. Justin P. Mattock -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: Any issues upgrading LFS 6.3 - 6.5
Craig Jackson wrote: Anyone else have ideas on how we can come together and get the project moving ahead at a faster pace? I just thought of the blood bank analogy. When someone offers to donate, the blood bank tells the donor what type of blood is most needed. If there is an area of BLFS that needs more work (i.e. Gnome) then a blurb on the homepage (or a testers area) about areas that need testing focus might help. Through my series of recent mails with the Dev team I have gotten word that Gnome is in need of some work, so I plan to focus on that as far as my current testing goes. A nice TODO for the testers might free up some some developer resources to concentrate on the more advanced patching. Gnome is cool to build, didn't like building apache ant though.(seemed broken at the time). Justin P. Mattock -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
sysvinit and inittab to shutdown
Any ideas on why ctl+alt+del doesn't shutdown my machine? the inittab I copied was from 6.54.3 which has the entry there, but when exiting the xserver and issuing those commands, I have no movement.(also my user name is in shutdown.allow) Justin P. Mattock -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: sysvinit and inittab to shutdown
Bruce Dubbs wrote: Justin P. Mattock wrote: Any ideas on why ctl+alt+del doesn't shutdown my machine? the inittab I copied was from 6.54.3 which has the entry there, but when exiting the xserver and issuing those commands, I have no movement.(also my user name is in shutdown.allow) Do you have ca:12345:ctrlaltdel:/sbin/shutdown -t1 -a -r now in /etc/inittab? -- Bruce yea, except I changed it to: ca:12345:ctrlaltdel:/sbin/shutdown -a -h now (reboot doesn't work so well with 2.6.31). added my user name to /etc/shutdown.allow I think my keycodes are out of whack when exiting the xserver. Justin P. Mattock -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: sysvinit and inittab to shutdown
Bruce Dubbs wrote: Justin P. Mattock wrote: Any ideas on why ctl+alt+del doesn't shutdown my machine? the inittab I copied was from 6.54.3 which has the entry there, but when exiting the xserver and issuing those commands, I have no movement.(also my user name is in shutdown.allow) Do you have ca:12345:ctrlaltdel:/sbin/shutdown -t1 -a -r now in /etc/inittab? -- Bruce just as I thought, del key works when the xserver is started, but nothing once I exit the xserver. /etc/inputrc shows: \e[3~: delete-char which should give me del, I just don't know where that is on the keyboard . Justin P. Mattock -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: GRUB error
Russell Stockhammer wrote: What file system is hd0 using? You should be doing this from within the chroot environment. If you exited and reentered chroot, do you still have /dev mounted? I.e. is has the `mount -v --bind /dev $LFS/dev` still valid? What is the output of `ls /dev/?da*` ? The paartition (hd0,0) is ext3 This is in the chroot environment and /dev is mounted There are no /dev/?da* devices this system uses /dev/cciss/c0d0. New information... If I run the command device (hd0) /dev/cciss/c0d0 then root (hd0,0) works. Now setup (hd0) produces this error... grub setup (hd0) Checking if /boot/grub/stage1 exists... yes Checking if /boot/grub/stage2 exists... yes Checking if /boot/grub/e2fs_stage1_5 exists... yes Running embed /boot/grub/e2fs_stage1_5 (hd0)... 15 sectors are embedded. succeeded Running install /boot/grub/stage1 (hd0) (hd0)1+15 p (hd0,0)/boot/grub/stage2 /boot/grub/menu.lst... failed Error 22: No such partition grub Russ Find your next place with Ninemsn property Looking for a place to rent, share or buy this winter? http://a.ninemsn.com.au/b.aspx?URL=http%3A%2F%2Fninemsn%2Edomain%2Ecom%2Eau%2F%3Fs%5Fcid%3DFDMedia%3ANineMSN%5FHotmail%5FTagline_t=774152450_r=Domain_tagline_m=EXT Not sure how grub works,I ended up using grub2. one thing I noticed is when doing chroot to install grub I ended up having to: mount -t proc none /mnt/ mount --bind /dev /mnt/dev before being able to have a successful build, and setup of grub2. Keep in mind I haven't tried the chroot procedure in the book, nor the regular grub. Justin P. Mattock -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: making LFS
Bruce Dubbs wrote: mahesh kumar wrote: hi.., i had studied the LFS book. And i started working on that. But, one thing i didn't get that, where shd we be while compiling and running the sources? I mean, what is the path shd be? one thing i knw that, packages shd be downloaded to /mnt/lfs/sources folder, and installations shd be done at /mnt/lfs/tools directory. But, as per the book, i didn't get that where shd we initially be...! either /mnt/lfs or /mnt/lfs/tools or /mnt/lfs/source??? Section 3.1, 3rd paragraph says where to store the tarballs, but it really can be anywhere. Section 5.3, the last note labeled 'Important' says 'a cd into the created directory should be performed.' -- Bruce I never really followed the book word by word, I ended up just making a /dir in my home directory and started to fill it up with what I wanted. Justin P. Mattock -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: Strange compile error on file-4.26
Thomas Tutone wrote: --- On Sun, 7/19/09, Bruce Dubbsbruce.du...@gmail.com wrote: Two things: 1. We don't need the full compile log. Just send the last part where the error occurred. Looking at successful compiles is a waste of time and space. Sorry about that. I guess I wanted to provide all necessary context, but obviously I provided way too much. 2. Why don't you try file-5.00? I tried it, and it compiled and installed cleanly. Thanks for the suggestion. I still am curious why my original build of 4.26 did not work. For the record, the relevant part of the compile log was: make[2]: Leaving directory `/sources/file-4.26/src' Making all in magic make[2]: Entering directory `/sources/file-4.26/magic' ../src/file -C -m ../magic/Magdir lt-file: No current entry for continuation make[2]: *** [magic.mgc] Error 255 make[2]: Leaving directory `/sources/file-4.26/magic' make[1]: *** [all-recursive] Error 1 make[1]: Leaving directory `/sources/file-4.26' make: *** [all] Error 2 Cheers. Tom Something to do with the file magic Justin P. Mattock -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: pass 2 build of gcc-3.4.2 has a problem with libgcc_s.so.1
Pol Vangheluwe wrote: I was maybe not clear enough: I am not yet at chapter 6. I am doing section 5.12, so no chrooting in the game. I have some doubts about the manipulation of the specs file in section 5.8. Nothing changed when I view with -dumpspecs: lfs:/lfs/sources/bash-3.2$ gcc -dumpspecs | grep ld.so.1 -m elf32ppclinux %{!shared: %{!static: %{rdynamic:-export-dynamic} %{!dynamic-linker:-dynamic-linker %{muclibc:%{mglibc:%e-mglibc and -muclibc used together}/lib/ld-uClibc.so.0;:/lib/ld.so.1 -m elf32ppclinux %{!shared: %{!static: %{rdynamic:-export-dynamic} %{!dynamic-linker:-dynamic-linker /lib/ld.so.1}}} while I find in the specs file itself the expected path: lfs:/lfs/sources/bash-3.2$ cat /lfs/tools/bin/../lib/gcc/powerpc-unknown-linux-gnu/4.3.2/specs | grep ld.so.1 -m elf32ppclinux %{!shared: %{!static: %{rdynamic:-export-dynamic} %{!dynamic-linker:-dynamic-linker %{muclibc:%{mglibc:%e-mglibc and -muclibc used together}/lib/ld-uClibc.so.0;:/tools/lib/ld.so.1 -m elf32ppclinux %{!shared: %{!static: %{rdynamic:-export-dynamic} %{!dynamic-linker:-dynamic-linker /tools/lib/ld.so.1}}} Can this have something to do with my problem? pvg Op 16-jul-09, om 21:41 heeft Justin P. Mattock het volgende geschreven: Pol Vangheluwe wrote: But why is it using /lib/libc.so.6 (from the original distribution) instead of /lfs/tools/lib/libc.so.6? pvg Op 16-jul-09, om 16:11 heeft Justin P. Mattock het volgende geschreven: Pol Vangheluwe wrote: I am trying to install LFS, edition 6.4 on an Apple PowerPC 7200 (indeed, an old-world Mac). I installed linux from the MkLinux distribution R1 (yes, a very old distribution) and then upgraded the system to Linux 2.6 from sources. I could not upgrade glibc without breaking the system, so it is still running glibc-2.1.3. This is the only requirement for the host system I could not fulfill. lfs:/lfs/sources/gcc-build$ cat /etc/issue Linux for PowerPC. Brought to you by The MkLinux Project. Based on Red Hat Linux Red Hat Linux release 6.0 (Hedwig) Kernel 2.6.26BuiltbySoftPol on a PowerPC 601 I have a problem, building gcc-4.3.2 in pass 2: lfs:/lfs/sources/gcc-build$ ../gcc-4.3.2/configure --prefix=/tools --with-local-prefix=/tools --enable-clocale=gnu --enable-shared --enable-threads=posix --enable-__cxa_atexit --enable-languages=c,c++ --disable-libstdcxx-pch --disable-bootstrap (…) lfs:/lfs/sources/gcc-build$ make (…) # Now that we have built all the objects, we need to copy # them back to the GCC directory. Too many things (other # in-tree libraries, and DejaGNU) know about the layout # of the build tree, for now. make install-leaf DESTDIR=../.././gcc \ slibdir= libsubdir= MULTIOSDIR=. make[3]: Entering directory `/lfs/sources/gcc-build/powerpc-unknown-linux-gnu/libgcc' /bin/sh ../../../gcc-4.3.2/libgcc/../mkinstalldirs ../.././gcc /usr/local/bin/install -c -m 644 libgcc_eh.a ../.././gcc/ chmod 644 ../.././gcc/libgcc_eh.a /tools/powerpc-unknown-linux-gnu/bin/ranlib ../.././gcc/libgcc_eh.a /bin/sh ../../../gcc-4.3.2/libgcc/../mkinstalldirs ../.././gcc; /usr/local/bin/install -c -m 644 ./libgcc_s.so.1 ../.././gcc/libgcc_s.so.1; rm -f ../.././gcc/libgcc_s.so; ln -s libgcc_s.so.1 ../.././gcc/libgcc_s.so rm: /lib/libc.so.6: version `GLIBC_2.2.4' not found (required by /lfs/sources/gcc-build/./gcc/libgcc_s.so.1) ln: /lib/libc.so.6: version `GLIBC_2.2.4' not found (required by /lfs/sources/gcc-build/./gcc/libgcc_s.so.1) make[3]: *** [install-shared] Error 1 make[3]: Leaving directory `/lfs/sources/gcc-build/powerpc-unknown-linux-gnu/libgcc' make[2]: *** [all] Error 2 make[2]: Leaving directory `/lfs/sources/gcc-build/powerpc-unknown-linux-gnu/libgcc' make[1]: *** [all-target-libgcc] Error 2 make[1]: Leaving directory `/lfs/sources/gcc-build' make: *** [all] Error 2 I am surprised to see that there are still dependencies in this phase of the project on the host system... lfs:/lfs/sources/gcc-build$ locate libc.so.6 /home/ftp/lib/libc.so.6 /lfs/tools/lib/libc.so.6 /lib/libc.so.6 lfs:/lfs/sources/gcc-build$ env TERM=xterm-color OLDPWD=/home/lfs LC_ALL=POSIX LFS=/mnt/lfs PATH=/tools/bin:/usr/local/bin:/bin:/usr/bin PWD=/lfs/sources/gcc-build PS1=\u:\w\$ SHLVL=1 HOME=/home/lfs _=/usr/local/bin/env lfs:/lfs/sources/gcc-build$ locate bin/rm /bin/rm (…) /usr/local/bin/rm (…) lfs:/lfs/sources/gcc-build$ locate bin/ln /bin/ln /usr/local/bin/ln (…) pvg oops, after sending I realized you were not building libc but gcc,(it early morning here). The compiler is looking for glibc-2.2.4 So it can finish with compiling. Justin P. Mattock -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page My guess is you haven't set the chroot environment correctly, causing gcc to look for certain libraries on the host still instead
Re: pass 2 build of gcc-3.4.2 has a problem with libgcc_s.so.1
Pol Vangheluwe wrote: I try to follow the LFS 4.6 book as closely as possible. I had to apply the following deviations: - the host system has glibc-2.1.3 instead of the recommended glibc-2.2.5. I broke several times my system, by trying to upgrade glibc. Then I gave up and turned to LFS. - replaced /usr/bin by /usr/local/bin in several commands because my host system was upgraded from a 10-year old distribution to comply with the LFS recommendations and the new binaries reside in /usr/local/bin. - didn't add the echo CFLAGS += -march=i486 -mtune=native configparms in section 5.7, because my system has no Intel processor; - the name of my dynamic linker is ld.so.1 instead of ld-linux.so.2. So far, I didn't consider using a newer version then the one used in the LFS book (it takes about 30 hours to perform a bootstrap build of gcc on my poor 100Mhz PowerPC 601…) I didn't use particular CFLAGS for my hardware. I still suspect a problem with mixed-up search paths… pvg Op 18-jul-09, om 21:54 heeft Justin P. Mattock het volgende geschreven: Pol Vangheluwe wrote: I was maybe not clear enough: I am not yet at chapter 6. I am doing section 5.12, so no chrooting in the game. I have some doubts about the manipulation of the specs file in section 5.8. Nothing changed when I view with -dumpspecs: lfs:/lfs/sources/bash-3.2$ gcc -dumpspecs | grep ld.so.1 -m elf32ppclinux %{!shared: %{!static: %{rdynamic:-export-dynamic} %{!dynamic-linker:-dynamic-linker %{muclibc:%{mglibc:%e-mglibc and -muclibc used together}/lib/ld-uClibc.so.0;:*/lib/ld.so.1* -m elf32ppclinux %{!shared: %{!static: %{rdynamic:-export-dynamic} %{!dynamic-linker:-dynamic-linker */lib/ld.so.1*}}} while I find in the specs file itself the expected path: lfs:/lfs/sources/bash-3.2$ cat /lfs/tools/bin/../lib/gcc/powerpc-unknown-linux-gnu/4.3.2/specs | grep ld.so.1 -m elf32ppclinux %{!shared: %{!static: %{rdynamic:-export-dynamic} %{!dynamic-linker:-dynamic-linker %{muclibc:%{mglibc:%e-mglibc and -muclibc used together}/lib/ld-uClibc.so.0;*:/tools/lib/ld.so.1* -m elf32ppclinux %{!shared: %{!static: %{rdynamic:-export-dynamic} %{!dynamic-linker:-dynamic-linker */tools/lib/ld.so.1*}}} Can this have something to do with my problem? pvg Op 16-jul-09, om 21:41 heeft Justin P. Mattock het volgende geschreven: Pol Vangheluwe wrote: But why is it using */lib/libc.so.6* (from the original distribution) instead of */lfs/tools/lib/libc.so.6*? pvg Op 16-jul-09, om 16:11 heeft Justin P. Mattock het volgende geschreven: Pol Vangheluwe wrote: I am trying to install LFS, edition 6.4 on an Apple PowerPC 7200 (indeed, an old-world Mac). I installed linux from the MkLinux distribution R1 (yes, a very old distribution) and then upgraded the system to Linux 2.6 from sources. I could not upgrade glibc without breaking the system, so it is still running glibc-2.1.3. This is the only requirement for the host system I could not fulfill. lfs:/lfs/sources/gcc-build$ cat /etc/issue Linux for PowerPC. Brought to you by The MkLinux Project. Based on Red Hat Linux Red Hat Linux release 6.0 (Hedwig) Kernel 2.6.26BuiltbySoftPol on a PowerPC 601 I have a problem, building gcc-4.3.2 in pass 2: lfs:/lfs/sources/gcc-build$ ../gcc-4.3.2/configure --prefix=/tools --with-local-prefix=/tools --enable-clocale=gnu --enable-shared --enable-threads=posix --enable-__cxa_atexit --enable-languages=c,c++ --disable-libstdcxx-pch --disable-bootstrap (…) lfs:/lfs/sources/gcc-build$ make (…) # Now that we have built all the objects, we need to copy # them back to the GCC directory. Too many things (other # in-tree libraries, and DejaGNU) know about the layout # of the build tree, for now. make install-leaf DESTDIR=../.././gcc \ slibdir= libsubdir= MULTIOSDIR=. make[3]: Entering directory `/lfs/sources/gcc-build/powerpc-unknown-linux-gnu/libgcc' /bin/sh ../../../gcc-4.3.2/libgcc/../mkinstalldirs ../.././gcc /usr/local/bin/install -c -m 644 libgcc_eh.a ../.././gcc/ chmod 644 ../.././gcc/libgcc_eh.a /tools/powerpc-unknown-linux-gnu/bin/ranlib ../.././gcc/libgcc_eh.a /bin/sh ../../../gcc-4.3.2/libgcc/../mkinstalldirs ../.././gcc; /usr/local/bin/install -c -m 644 ./libgcc_s.so.1 ../.././gcc/libgcc_s.so.1; rm -f ../.././gcc/libgcc_s.so; ln -s libgcc_s.so.1 ../.././gcc/libgcc_s.so rm: */lib/libc.so.6*: version `GLIBC_2.2.4' not found (required by */lfs/sources/gcc-build/./gcc/libgcc_s.so.1*) ln: */lib/libc.so.6*: version `GLIBC_2.2.4' not found (required by */lfs/sources/gcc-build/./gcc/libgcc_s.so.1*) make[3]: *** [install-shared] Error 1 make[3]: Leaving directory `/lfs/sources/gcc-build/powerpc-unknown-linux-gnu/libgcc' make[2]: *** [all] Error 2 make[2]: Leaving directory `/lfs/sources/gcc-build/powerpc-unknown-linux-gnu/libgcc' make[1]: *** [all-target-libgcc] Error 2 make[1]: Leaving directory `/lfs/sources/gcc-build' make: *** [all] Error 2 I am surprised
Re: pass 2 build of gcc-3.4.2 has a problem with libgcc_s.so.1
Bruce Dubbs wrote: Justin P. Mattock wrote: [Lots o fdeleted stuff] 30 hours(sh^t) that's too long.. As for 6.4 I can't even find that one, I ended up following this: http://www.linuxfromscratch.org/lfs/view/development/ there you have libc 2.10, gcc 4.4.*(seems more up-to-date, but don't let that deter you from using older versions). I'll have to have to try this on my old dell to see if this takes that long. Justin, Please trim your posts to the minimum necessary to keep your reply in context. -- Bruce Alright, just to make sure (top post instead of bottom post) Justin P. Mattock -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: pass 2 build of gcc-3.4.2 has a problem with libgcc_s.so.1
Pol Vangheluwe wrote: I am trying to install LFS, edition 6.4 on an Apple PowerPC 7200 (indeed, an old-world Mac). I installed linux from the MkLinux distribution R1 (yes, a very old distribution) and then upgraded the system to Linux 2.6 from sources. I could not upgrade glibc without breaking the system, so it is still running glibc-2.1.3. This is the only requirement for the host system I could not fulfill. lfs:/lfs/sources/gcc-build$ cat /etc/issue Linux for PowerPC. Brought to you by The MkLinux Project. Based on Red Hat Linux Red Hat Linux release 6.0 (Hedwig) Kernel 2.6.26BuiltbySoftPol on a PowerPC 601 I have a problem, building gcc-4.3.2 in pass 2: lfs:/lfs/sources/gcc-build$ ../gcc-4.3.2/configure --prefix=/tools --with-local-prefix=/tools --enable-clocale=gnu --enable-shared --enable-threads=posix --enable-__cxa_atexit --enable-languages=c,c++ --disable-libstdcxx-pch --disable-bootstrap (…) lfs:/lfs/sources/gcc-build$ make (…) # Now that we have built all the objects, we need to copy # them back to the GCC directory. Too many things (other # in-tree libraries, and DejaGNU) know about the layout # of the build tree, for now. make install-leaf DESTDIR=../.././gcc \ slibdir= libsubdir= MULTIOSDIR=. make[3]: Entering directory `/lfs/sources/gcc-build/powerpc-unknown-linux-gnu/libgcc' /bin/sh ../../../gcc-4.3.2/libgcc/../mkinstalldirs ../.././gcc /usr/local/bin/install -c -m 644 libgcc_eh.a ../.././gcc/ chmod 644 ../.././gcc/libgcc_eh.a /tools/powerpc-unknown-linux-gnu/bin/ranlib ../.././gcc/libgcc_eh.a /bin/sh ../../../gcc-4.3.2/libgcc/../mkinstalldirs ../.././gcc; /usr/local/bin/install -c -m 644 ./libgcc_s.so.1 ../.././gcc/libgcc_s.so.1; rm -f ../.././gcc/libgcc_s.so; ln -s libgcc_s.so.1 ../.././gcc/libgcc_s.so rm: /lib/libc.so.6: version `GLIBC_2.2.4' not found (required by /lfs/sources/gcc-build/./gcc/libgcc_s.so.1) ln: /lib/libc.so.6: version `GLIBC_2.2.4' not found (required by /lfs/sources/gcc-build/./gcc/libgcc_s.so.1) make[3]: *** [install-shared] Error 1 make[3]: Leaving directory `/lfs/sources/gcc-build/powerpc-unknown-linux-gnu/libgcc' make[2]: *** [all] Error 2 make[2]: Leaving directory `/lfs/sources/gcc-build/powerpc-unknown-linux-gnu/libgcc' make[1]: *** [all-target-libgcc] Error 2 make[1]: Leaving directory `/lfs/sources/gcc-build' make: *** [all] Error 2 I am surprised to see that there are still dependencies in this phase of the project on the host system... lfs:/lfs/sources/gcc-build$ locate libc.so.6 /home/ftp/lib/libc.so.6 /lfs/tools/lib/libc.so.6 /lib/libc.so.6 lfs:/lfs/sources/gcc-build$ env TERM=xterm-color OLDPWD=/home/lfs LC_ALL=POSIX LFS=/mnt/lfs PATH=/tools/bin:/usr/local/bin:/bin:/usr/bin PWD=/lfs/sources/gcc-build PS1=\u:\w\$ SHLVL=1 HOME=/home/lfs _=/usr/local/bin/env lfs:/lfs/sources/gcc-build$ locate bin/rm /bin/rm (…) /usr/local/bin/rm (…) lfs:/lfs/sources/gcc-build$ locate bin/ln /bin/ln /usr/local/bin/ln (…) pvg Not sure what might be happening, but when looking at the web site it suggests that you build libc in a chroot environment. (my guess is the above error is due to maybe an older/newer version clash). Justin P. Mattock -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: pass 2 build of gcc-3.4.2 has a problem with libgcc_s.so.1
Pol Vangheluwe wrote: I am trying to install LFS, edition 6.4 on an Apple PowerPC 7200 (indeed, an old-world Mac). I installed linux from the MkLinux distribution R1 (yes, a very old distribution) and then upgraded the system to Linux 2.6 from sources. I could not upgrade glibc without breaking the system, so it is still running glibc-2.1.3. This is the only requirement for the host system I could not fulfill. lfs:/lfs/sources/gcc-build$ cat /etc/issue Linux for PowerPC. Brought to you by The MkLinux Project. Based on Red Hat Linux Red Hat Linux release 6.0 (Hedwig) Kernel 2.6.26BuiltbySoftPol on a PowerPC 601 I have a problem, building gcc-4.3.2 in pass 2: lfs:/lfs/sources/gcc-build$ ../gcc-4.3.2/configure --prefix=/tools --with-local-prefix=/tools --enable-clocale=gnu --enable-shared --enable-threads=posix --enable-__cxa_atexit --enable-languages=c,c++ --disable-libstdcxx-pch --disable-bootstrap (…) lfs:/lfs/sources/gcc-build$ make (…) # Now that we have built all the objects, we need to copy # them back to the GCC directory. Too many things (other # in-tree libraries, and DejaGNU) know about the layout # of the build tree, for now. make install-leaf DESTDIR=../.././gcc \ slibdir= libsubdir= MULTIOSDIR=. make[3]: Entering directory `/lfs/sources/gcc-build/powerpc-unknown-linux-gnu/libgcc' /bin/sh ../../../gcc-4.3.2/libgcc/../mkinstalldirs ../.././gcc /usr/local/bin/install -c -m 644 libgcc_eh.a ../.././gcc/ chmod 644 ../.././gcc/libgcc_eh.a /tools/powerpc-unknown-linux-gnu/bin/ranlib ../.././gcc/libgcc_eh.a /bin/sh ../../../gcc-4.3.2/libgcc/../mkinstalldirs ../.././gcc; /usr/local/bin/install -c -m 644 ./libgcc_s.so.1 ../.././gcc/libgcc_s.so.1; rm -f ../.././gcc/libgcc_s.so; ln -s libgcc_s.so.1 ../.././gcc/libgcc_s.so rm: /lib/libc.so.6: version `GLIBC_2.2.4' not found (required by /lfs/sources/gcc-build/./gcc/libgcc_s.so.1) ln: /lib/libc.so.6: version `GLIBC_2.2.4' not found (required by /lfs/sources/gcc-build/./gcc/libgcc_s.so.1) make[3]: *** [install-shared] Error 1 make[3]: Leaving directory `/lfs/sources/gcc-build/powerpc-unknown-linux-gnu/libgcc' make[2]: *** [all] Error 2 make[2]: Leaving directory `/lfs/sources/gcc-build/powerpc-unknown-linux-gnu/libgcc' make[1]: *** [all-target-libgcc] Error 2 make[1]: Leaving directory `/lfs/sources/gcc-build' make: *** [all] Error 2 I am surprised to see that there are still dependencies in this phase of the project on the host system... lfs:/lfs/sources/gcc-build$ locate libc.so.6 /home/ftp/lib/libc.so.6 /lfs/tools/lib/libc.so.6 /lib/libc.so.6 lfs:/lfs/sources/gcc-build$ env TERM=xterm-color OLDPWD=/home/lfs LC_ALL=POSIX LFS=/mnt/lfs PATH=/tools/bin:/usr/local/bin:/bin:/usr/bin PWD=/lfs/sources/gcc-build PS1=\u:\w\$ SHLVL=1 HOME=/home/lfs _=/usr/local/bin/env lfs:/lfs/sources/gcc-build$ locate bin/rm /bin/rm (…) /usr/local/bin/rm (…) lfs:/lfs/sources/gcc-build$ locate bin/ln /bin/ln /usr/local/bin/ln (…) pvg oops, after sending I realized you were not building libc but gcc,(it early morning here). The compiler is looking for glibc-2.2.4 So it can finish with compiling. Justin P. Mattock -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: pass 2 build of gcc-3.4.2 has a problem with libgcc_s.so.1
Pol Vangheluwe wrote: But why is it using /lib/libc.so.6 (from the original distribution) instead of /lfs/tools/lib/libc.so.6? pvg Op 16-jul-09, om 16:11 heeft Justin P. Mattock het volgende geschreven: Pol Vangheluwe wrote: I am trying to install LFS, edition 6.4 on an Apple PowerPC 7200 (indeed, an old-world Mac). I installed linux from the MkLinux distribution R1 (yes, a very old distribution) and then upgraded the system to Linux 2.6 from sources. I could not upgrade glibc without breaking the system, so it is still running glibc-2.1.3. This is the only requirement for the host system I could not fulfill. lfs:/lfs/sources/gcc-build$ cat /etc/issue Linux for PowerPC. Brought to you by The MkLinux Project. Based on Red Hat Linux Red Hat Linux release 6.0 (Hedwig) Kernel 2.6.26BuiltbySoftPol on a PowerPC 601 I have a problem, building gcc-4.3.2 in pass 2: lfs:/lfs/sources/gcc-build$ ../gcc-4.3.2/configure --prefix=/tools --with-local-prefix=/tools --enable-clocale=gnu --enable-shared --enable-threads=posix --enable-__cxa_atexit --enable-languages=c,c++ --disable-libstdcxx-pch --disable-bootstrap (…) lfs:/lfs/sources/gcc-build$ make (…) # Now that we have built all the objects, we need to copy # them back to the GCC directory. Too many things (other # in-tree libraries, and DejaGNU) know about the layout # of the build tree, for now. make install-leaf DESTDIR=../.././gcc \ slibdir= libsubdir= MULTIOSDIR=. make[3]: Entering directory `/lfs/sources/gcc-build/powerpc-unknown-linux-gnu/libgcc' /bin/sh ../../../gcc-4.3.2/libgcc/../mkinstalldirs ../.././gcc /usr/local/bin/install -c -m 644 libgcc_eh.a ../.././gcc/ chmod 644 ../.././gcc/libgcc_eh.a /tools/powerpc-unknown-linux-gnu/bin/ranlib ../.././gcc/libgcc_eh.a /bin/sh ../../../gcc-4.3.2/libgcc/../mkinstalldirs ../.././gcc; /usr/local/bin/install -c -m 644 ./libgcc_s.so.1 ../.././gcc/libgcc_s.so.1; rm -f ../.././gcc/libgcc_s.so; ln -s libgcc_s.so.1 ../.././gcc/libgcc_s.so rm: /lib/libc.so.6: version `GLIBC_2.2.4' not found (required by /lfs/sources/gcc-build/./gcc/libgcc_s.so.1) ln: /lib/libc.so.6: version `GLIBC_2.2.4' not found (required by /lfs/sources/gcc-build/./gcc/libgcc_s.so.1) make[3]: *** [install-shared] Error 1 make[3]: Leaving directory `/lfs/sources/gcc-build/powerpc-unknown-linux-gnu/libgcc' make[2]: *** [all] Error 2 make[2]: Leaving directory `/lfs/sources/gcc-build/powerpc-unknown-linux-gnu/libgcc' make[1]: *** [all-target-libgcc] Error 2 make[1]: Leaving directory `/lfs/sources/gcc-build' make: *** [all] Error 2 I am surprised to see that there are still dependencies in this phase of the project on the host system... lfs:/lfs/sources/gcc-build$ locate libc.so.6 /home/ftp/lib/libc.so.6 /lfs/tools/lib/libc.so.6 /lib/libc.so.6 lfs:/lfs/sources/gcc-build$ env TERM=xterm-color OLDPWD=/home/lfs LC_ALL=POSIX LFS=/mnt/lfs PATH=/tools/bin:/usr/local/bin:/bin:/usr/bin PWD=/lfs/sources/gcc-build PS1=\u:\w\$ SHLVL=1 HOME=/home/lfs _=/usr/local/bin/env lfs:/lfs/sources/gcc-build$ locate bin/rm /bin/rm (…) /usr/local/bin/rm (…) lfs:/lfs/sources/gcc-build$ locate bin/ln /bin/ln /usr/local/bin/ln (…) pvg oops, after sending I realized you were not building libc but gcc,(it early morning here). The compiler is looking for glibc-2.2.4 So it can finish with compiling. Justin P. Mattock -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page My guess is you haven't set the chroot environment correctly, causing gcc to look for certain libraries on the host still instead of /tools/* Justin P. Mattock -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: chroot problems
Russell Stockhammer wrote: Gday everyone, This may not be a LFS issue but I will ask here first since it happened with the LFS build. I am at the stage where I am chroot'ing into the temporary LFS environment but I am having problems. If I run the command: chroot $LFS /tools/bin/env -i \ HOME=/root TERM=$TERM PS1='\u:\w\$ ' \ PATH=/bin:/usr/bin:/sbin:/usr/sbin:/tools/bin \ /tools/bin/bash --login +h I get the following error; chroot: cannot run command `/tools/bin/env': No such file or directory This is very unusual since the env binary does exist at that location. To try and debug this I try running: chroot $LFS /tools/bin/bash This runs but something strange happens... I have been chroot'ed into a directory that doesn't contain anything, no files, no directories, no nothing. All I can do is run the commands built-in to the bash shell. Now this may be a problem with the chroot provided by my host system so I tried: /tools/bin/chroot / /tools/bin/bash This works perfectly fine. I am at a loss to figure out what has gone wrong does anyone else have a clue? Thanks Russ Click here to find out more POP access for Hotmail is here! http://windowslive.ninemsn.com.au/article.aspx?id=802246 you need the coreutils package. Justin P. Mattock -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: LFS 6.4 - Coreutils-6.12 - problems compiling chcon
Russell Stockhammer wrote: It sounds like you missed the --without-selinux when building glibc in Chapter 5. I am sure that I did, but just to be sure I re-ran the glibc build still no go with the coreutils build. I noticed the following during the configure script run: checking for rpmatch... yes checking selinux/flask.h presence... yes checking for selinux/flask.h... yes checking for library containing setfilecon... no checking selinux/selinux.h usability... yes checking selinux/selinux.h presence... yes checking for selinux/selinux.h... yes checking selinux/context.h usability... yes checking selinux/context.h presence... yes checking for selinux/context.h... yes checking for sig2str... no Even after deleting all the selinux header files... No luck. Any further ideas? Thanks Russ Let us help with car news, reviews and more Looking for a new car this winter? http://a.ninemsn.com.au/b.aspx?URL=http%3A%2F%2Fsecure%2Dau%2Eimrworldwide%2Ecom%2Fcgi%2Dbin%2Fa%2Fci%5F450304%2Fet%5F2%2Fcg%5F801459%2Fpi%5F1004813%2Fai%5F859641_t=762955845_r=tig_OCT07_m=EXT With your host system you probably have libselinux(and so forth), I recommend to go ahead and building everything with SELiux support, then, at the end of the day if you decide to add a policy you do, if not then you don't. (at least the system has a security infrastructure). Justin P. Mattock -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: LFS 6.4 - Coreutils-6.12 - problems compiling chcon
Bruce Dubbs wrote: Russell Stockhammer wrote: It sounds like you missed the --without-selinux when building glibc in Chapter 5. I am sure that I did, but just to be sure I re-ran the glibc build still no go with the coreutils build. I noticed the following during the configure script run: checking for rpmatch... yes checking selinux/flask.h presence... yes checking for selinux/flask.h... yes checking for library containing setfilecon... no checking selinux/selinux.h usability... yes checking selinux/selinux.h presence... yes checking for selinux/selinux.h... yes checking selinux/context.h usability... yes checking selinux/context.h presence... yes checking for selinux/context.h... yes checking for sig2str... no Even after deleting all the selinux header files... No luck. Any further ideas? Something you did early on corrupted the system. I think the easiest thing would be to start over and check for those selinux files after installing glibc in Chapter 5. I'd also double check the build of coreutils in Chapter 5 to see if selinux was built there. It shouldn't be. -- Bruce I agree, Something was f^*~d up along the way. in any case (lazy approach) do an make -i (what are you going to use chcon for anyways!!) or load the updated headers for coreutils to compile properly. Justin P. Mattock -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: LFS 6.4 - Coreutils-6.12 - problems compiling chcon
Mike McCarty wrote: Justin P. Mattock wrote: With your host system you probably have libselinux(and so forth), I recommend to go ahead and building everything with SELiux support, then, at the end of the day if you decide to add a policy you do, if not then you don't. (at least the system has a security infrastructure). Justin P. Mattock I recommend that, for a first build, NO DEVIATIONS are made from the directions in the book. Also, personally, one of the reasons I have built LFS is so I WON'T have SeLinux on my machine. Like everything, there are advantages and disadvantages to having SeLinux, even if it is not enabled. YMMV Mike Some people like it some people hate it, in any case it shouldn't be that big of an issue if something gets compiled with a header or so. (but could be wrong). My guess is using a host system that has nothing of SELinux on it at all before you start compiling all of the packages. or the opposite if you want to enable SELinux. Justin P. Mattock -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: LFS 6.4 - Coreutils-6.12 - problems compiling chcon
Rajinder Yadav wrote: On Tue, Jul 14, 2009 at 7:14 PM, Russell Stockhammerrstockham...@hotmail.com wrote: I recommend that, for a first build, NO DEVIATIONS are made from the directions in the book. I've made no deviations, I've cut and pasted command yet still I am cursed with some echo of selinux. After trying make -i things have only gotten worse there are MANY coreutil programs that are looking for the selinux library, important ones that can't just be skipped over. I am sure I've done nothing wrong following the instructions. Ok here is a solution I am trying... Using coreutils-5.97 this is before selinux existed so it doesn't even know to look for selinux so no compile errors. Now does anyone see and problems with what I am doing? Thanks Russ I managed to complete all of chapter 5 on my CentOS 5.3 box without removing selinux support (in fact I wanted it included). So you should not be having any issues with your RH box. I wish I would provide more help but don't get hung up on selinux, something else is off. I have decided to start all over with my LFS learning on another box running kubuntu 64bit. If I manage to get this completed for kicks I want to cross compile on this box (which is a lot faster) for my CentOS 5.3 32bit system. Which I am hoping will give me a deeper understanding of things =) I enjoyed building LFS (They do a nice job with the instructions) As for the make -i (apologize for the bad advice) From what it sounds like you want a SELinux free system. you might want to try the LFS livecd instead of RH due to RH having SELinux enabled from the start. (not sure if the lfs livecd has libselinux installed or not) but if it doesn't then life for you will be easier, as opposed to a system have all of the SELinux infrastructure compiled in it. Justin P. Mattock -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: LFS 6.4 - Coreutils-6.12 - problems compiling chcon
Russell Stockhammer wrote: OK thanks for everyones help by using coreutils-5.97 I mananged to get past this step without anymore hassles. Thanks everyone for your help! Russ Click Here View photos of singles in your area http://dating.ninemsn.com.au/search/search.aspx?exec=gotp=qgc=2tr=1lage=18uage=55cl=14sl=0dist=50po=1do=2trackingid=1046138r2s=1_t=773166090_r=WLM_EndText Hey alright!! Justin P. Mattock -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: Can not boot from LFS-6.4
khaled gouaich wrote: Hi, Thanks for the suggestions even if i fixed the problem. the problem was about the hd*. In the feisty i got /dev/hd* but in LFS 2.6.27 i hve to put /dev/sd*. But LFS still not recognize the /dev/sda2 (wich is /dev/hda5 in my feisty) as swap. Any idea ?? 2009/7/7 Mike McCartymike.mcca...@sbcglobal.net: khaled gouaich wrote: Hi all, I have a problem with bootig my lfs-6.4, here the out put : -VFS: cannot open root device hda3 or unknown-block (2-0) Please append a correct root=; boot option .. kernel panic-not syncing : VFS :unable to mount root fs on unknown-block (2,0) The kernel is looking for a file system to mount, and can't find it. The most likely cause for this is not building drivers compatible with your actual disc interface hardware into the kernel. Boot your host system, and use lspci or something similar to find out what your hardware chips are (or look inside your machine) and then check your kernel configuration and make sure you include drivers for those chips, and build them in (not make them loadable modules). Another possibility is that you don't have the file system driver built in. [...] Mike -- p=p=%c%s%c;main(){printf(p,34,p,34);};main(){printf(p,34,p,34);} Oppose globalization and One World Governments like the UN. This message made from 100% recycled bits. You have found the bank of Larn. I speak only for myself, and I am unanimous in that! -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page Im guessing your using the old ati-piix verses the new ati-piix(hda to sda) how are the entries in /dev/* I ended up just copying /dev over to the new partition(too lazy to mknod) Justin P. Mattock -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: confused at chapter 5.6
Rajinder Yadav wrote: Hi Justin, after Michael Tsang feedback on where to put mpfr and gmp, I went to the GCC build page and it states that if the sources are put under the gcc source, they will get built when gcc is built: Ref: http://gcc.gnu.org/install/prerequisites.html Multiple Precision Library (GMP) version 4.2 (or later) Necessary to build GCC. If you do not have it installed in your library search path, you will have to configure with the --with-gmp configure option. See also --with-gmp-lib and --with-gmp-include. Alternatively, if a GMP source distribution is found in a subdirectory of your GCC sources named gmp, it will be built together with GCC. Kind Regards, Rajinder Yadav --- On Sun, 7/5/09, Justin P. Mattockjustinmatt...@gmail.com wrote: From: Justin P. Mattockjustinmatt...@gmail.com Subject: Re: confused at chapter 5.6 To: LFS Support Listlfs-support@linuxfromscratch.org Received: Sunday, July 5, 2009, 12:25 AM Rajinder Yadav wrote: Hi Justin, thanks that cleared things up and I was able to proceed. Also, I was able to build and install GCC without any issues but I didn't build the supporting packages (mpfr and gmp) and I don't have the devel package for them installed. I guess it would be safe to rebuild GCC =) after building mpfr and gmp! Kind Regards, Rajinder Yadav --- On Sat, 7/4/09, Justin P. Mattockjustinmatt...@gmail.com wrote: From: Justin P. Mattockjustinmatt...@gmail.com Subject: Re: confused at chapter 5.6 To: LFS Support Listlfs-support@linuxfromscratch.org Received: Saturday, July 4, 2009, 11:04 PM Rajinder Yadav wrote: I am confused at what needs to be done in chapter 5.6. Linux-2.6.27.4 API Headers. do I need to unzip the linux kernel at this point? which direction is this command executed from? where is 'headers_check' defined for 'make headers_check' step? Also in the prior step for building gcc, were we just to unzip and rename files mpfr and gmp without building these packages? Kind Regards, Rajinder Yadav __ The new Internet Explorer® 8 - Faster, safer, easier. Optimized for Yahoo! Get it Now for Free! at http://downloads.yahoo.com/ca/internetexplorer/ Well firstly you would have untar the tar ball from kernel.org, then move into the new kernel's source tree to execute those commands. or if your distribution already has a kernel move to it's source tree to execute those commands. mpfr and gmp need to be built and installed before compiling gcc(or use the hosts systems gmp/mpfr i.g. apt-get install * if your on debian/ubuntu). Justin P. Mattock -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page __ Be smarter than spam. See how smart SpamGuard is at giving junk email the boot with the All-new Yahoo! Mail. Click on Options in Mail and switch to New Mail today or register for free at http://mail.yahoo.ca Whatever distro you have probably has those headers and libs, but once you boot into the new system and decide to compile gcc, you probably will need those. (but could be wrong) As for compiling mpfr/gmp, Ive heard about putting those into the gcc source tree, but never did that approach I just compiled separately and then told gcc where they are located. Justin P. Mattock -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page __ Looking for the perfect gift? Give the gift of Flickr! http://www.flickr.com/gift/ Cool thanks for that, makes things easier. Justin P. Mattock -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: SELinux
Michael Tsang wrote: On Sunday 05 July 2009 16:43:10 Rajinder Yadav wrote: I noticed one of the switch passed when building glibc is --without-selinux I assume the resulting LFS system, will it be missing Security Enhancement? If so what is required to build a SE LFS system. I assume SE is part of the Linux kernel and thus needs to be built as part of the Linux kernel? Kind Regards, Rajinder Yadav __ The new Internet Explorer® 8 - Faster, safer, easier. Optimized for Yahoo! Get it Now for Free! at http://downloads.yahoo.com/ca/internetexplorer/ In which ver. and ch. does the book told you to do so? The book does not told me to do so (I am using SVN-20090629). Also, you need to enable SELinux in the kernel to make it works. Michael Tsang Even under hlfs there's not much about SELinux. If you wanted SELinux I would imagine you would start by having all of the security headers/libs(libselinux,libaudit,libattr,etc..) before building libc, then after libc making sure all other apps/libs that give an SElinux switch are turned on. Then once thats done grab refpolicy and start locking down your system. Justin P. Mattock -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: SELinux
William Immendorf wrote: On Sun, Jul 5, 2009 at 3:43 AM, Rajinder Yadavdev...@ymail.com wrote: The new Internet Explorer® 8 - Faster, safer, easier. Optimized for Yahoo! Get it Now for Free! at http://downloads.yahoo.com/ca/internetexplorer/ Althrought this is unleated, I ROFL'D at the above line. I don't think Yahoo knows that this is a Linux mailing list... William I was wondering were/what that was. Justin P. Mattock -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: SELinux
Rajinder Yadav wrote: --- On Sun, 7/5/09, Justin P. Mattockjustinmatt...@gmail.com wrote: From: Justin P. Mattockjustinmatt...@gmail.com Subject: Re: SELinux To: LFS Support Listlfs-support@linuxfromscratch.org Received: Sunday, July 5, 2009, 10:46 AM Michael Tsang wrote: On Sunday 05 July 2009 16:43:10 Rajinder Yadav wrote: I noticed one of the switch passed when building glibc is --without-selinux I assume the resulting LFS system, will it be missing Security Enhancement? If so what is required to build a SE LFS system. I assume SE is part of the Linux kernel and thus needs to be built as part of the Linux kernel? Kind Regards, Rajinder Yadav __ The new Internet Explorer® 8 - Faster, safer, easier. Optimized for Yahoo! Get it Now for Free! at http://downloads.yahoo.com/ca/internetexplorer/ In which ver. and ch. does the book told you to do so? The book does not told me to do so (I am using SVN-20090629). Also, you need to enable SELinux in the kernel to make it works. Michael Tsang Even under hlfs there's not much about SELinux. If you wanted SELinux I would imagine you would start by having all of the security headers/libs(libselinux,libaudit,libattr,etc..) before building libc, then after libc making sure all other apps/libs that give an SElinux switch are turned on. Then once thats done grab refpolicy and start locking down your system. Justin P. Mattock -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page Hi Justin, thank you for the pointers, SeLinux LFS sounds like a plan down the road, something else to get into and learn =) Kind Regards, Rajinder Yadav __ Looking for the perfect gift? Give the gift of Flickr! http://www.flickr.com/gift/ Well, If you need to its possible, I myself run an LFS system with an SELinux policy running(latest refpolicy from tresys). just remember that you need to tell your packages early in the build process that you want to enable SELinux, or for example coreutils ls -Z won't show the contexts) Justin P. Mattock -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: confused at chapter 5.6
Rajinder Yadav wrote: I am confused at what needs to be done in chapter 5.6. Linux-2.6.27.4 API Headers. do I need to unzip the linux kernel at this point? which direction is this command executed from? where is 'headers_check' defined for 'make headers_check' step? Also in the prior step for building gcc, were we just to unzip and rename files mpfr and gmp without building these packages? Kind Regards, Rajinder Yadav __ The new Internet Explorer® 8 - Faster, safer, easier. Optimized for Yahoo! Get it Now for Free! at http://downloads.yahoo.com/ca/internetexplorer/ Well firstly you would have untar the tar ball from kernel.org, then move into the new kernel's source tree to execute those commands. or if your distribution already has a kernel move to it's source tree to execute those commands. mpfr and gmp need to be built and installed before compiling gcc(or use the hosts systems gmp/mpfr i.g. apt-get install * if your on debian/ubuntu). Justin P. Mattock -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: confused at chapter 5.6
Rajinder Yadav wrote: Hi Justin, thanks that cleared things up and I was able to proceed. Also, I was able to build and install GCC without any issues but I didn't build the supporting packages (mpfr and gmp) and I don't have the devel package for them installed. I guess it would be safe to rebuild GCC =) after building mpfr and gmp! Kind Regards, Rajinder Yadav --- On Sat, 7/4/09, Justin P. Mattockjustinmatt...@gmail.com wrote: From: Justin P. Mattockjustinmatt...@gmail.com Subject: Re: confused at chapter 5.6 To: LFS Support Listlfs-support@linuxfromscratch.org Received: Saturday, July 4, 2009, 11:04 PM Rajinder Yadav wrote: I am confused at what needs to be done in chapter 5.6. Linux-2.6.27.4 API Headers. do I need to unzip the linux kernel at this point? which direction is this command executed from? where is 'headers_check' defined for 'make headers_check' step? Also in the prior step for building gcc, were we just to unzip and rename files mpfr and gmp without building these packages? Kind Regards, Rajinder Yadav __ The new Internet Explorer® 8 - Faster, safer, easier. Optimized for Yahoo! Get it Now for Free! at http://downloads.yahoo.com/ca/internetexplorer/ Well firstly you would have untar the tar ball from kernel.org, then move into the new kernel's source tree to execute those commands. or if your distribution already has a kernel move to it's source tree to execute those commands. mpfr and gmp need to be built and installed before compiling gcc(or use the hosts systems gmp/mpfr i.g. apt-get install * if your on debian/ubuntu). Justin P. Mattock -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page __ Be smarter than spam. See how smart SpamGuard is at giving junk email the boot with the All-new Yahoo! Mail. Click on Options in Mail and switch to New Mail today or register for free at http://mail.yahoo.ca Whatever distro you have probably has those headers and libs, but once you boot into the new system and decide to compile gcc, you probably will need those. (but could be wrong) As for compiling mpfr/gmp, Ive heard about putting those into the gcc source tree, but never did that approach I just compiled separately and then told gcc where they are located. Justin P. Mattock -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: Chapter 5.6: Linux API Headers won't build
Smartboy wrote: I have checked, and went over the build exactly how the book says. Yet, when I try to compile the Linux API headers, I get stuck at make headers_check with the error message: scripts/unifdef.c:209: error: conflicting types for 'getline' /usr/include/stdio.h:651: error: previous declaration of 'getline' was here I did a google search, and it seems that a patch was made for this, but for more recent kernels. You can find the patch here (http://lkml.indiana.edu/hypermail/linux/kernel/0903.0/01708.html), but if you try to apply it, you should get errors. Anyone else experience this? Any solutions? update your kernel, or add this commit: http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=d15bd1067b1fcb2b7250d22bc0c7c7fea0b759f7 regards, Justin P. Mattock -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page