Re: Glibc make error
On Wed, 2010-03-31 at 12:05 -0400, Kaleb Hosie wrote: > I'm having some issues with installing glibc while following the instructions > on chapter 5.7. > > >From what I understand from the instructions, I made the file: > /mnt/lfs/sources/glibc-build/configparms > > In the configparms file, I should have the following text: > case `uname -m` in > i?86) echo "CFLAGS += -march=i486 -mtune=native" > configparms ;; > esac ... > Have I understood the instructions correctly? Thanks for your help! You've misunderstood it - the lines are a command to be run to create the configparms file, not the contents to be added to it. Basically it means that if your system matches i?86 (i.e it's an x86 processor in 32-bit mode), write the string "CFLAGS += -march=i486 -mtune=native" to the configparms file - otherwise do nothing. Simon. signature.asc Description: This is a digitally signed message part -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: LFS 6.6 X86_64
On Sat, 2010-04-03 at 00:36 -0400, stosss wrote: > Any one have any suggestions? > > The LFS x86_64 6.3 r2160 LiveCD boots but won't read the SATA discs. > > I tried the Fedora LiveCD and it won't even boot. Neither will the > Gentoo x86_64 or Debian x86_64 install CDs I've generally found the Ubuntu LiveCDs (actually flash drives) to work pretty well, including for x86_64... Simon. signature.asc Description: This is a digitally signed message part -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: Stuck on Section 5.6
On Tue, 2010-04-13 at 08:52 -0400, duck wilson wrote: > According to the book my next step is to do "make mrproper" which has > gotten me stumped. Do I do this make in $LFS/sources or do I need to > extract the Linux kernel files and cd to the kernel directory and > execute the commands in Section 5.6 in the kernel directory or > something else entirely? Same as for every other package in the book, as explained under General Compilation Instructions. It's always the same routine - each page of the book assumes that you've already extracted the relevant package, and changed into the resulting directory. Simon. signature.asc Description: This is a digitally signed message part -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: glibc compiling problem in lfs 6.6
On Tue, 2010-04-13 at 13:38 -0500, Jordan Peters wrote: > afte binutils is done with the make install i try testing it like it > says in the book. i am told it cannot run compiled c program. is > x86_64 32 bit? You need to give us a bit more detail for us to help. The *exact* error copied into an email, not just your own summary of what it said. It'd also help to know what distro/version you're building LFS from, and whether it's 32-bit (x86) or 64-bit (x86_64)... Simon. signature.asc Description: This is a digitally signed message part -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: Auto mounting USB thumb drives and CD/DVD
On Tue, 2010-04-13 at 09:35 -0400, linux fan wrote: > I stopped the haldeamon and started up kde 4.4.2 (pretty new stuff). > Then I started dragonplayer and it immediately told me that my audio > playback device does not work. > > I guess "deprecated" is not synonymous with "you don't need it". No, it's not. 'Deprecated' generally means that the developers aren't maintaining it anymore, and that packages requiring it should migrate to something else. It doesn't at all imply that nothing still uses it, only that doing so should be considered a bug. That said, we're a long way into the process of removing those dependencies. I don't know how close KDE is to dropping it, but as I said before, you can build a current Xorg and Gnome without it and without any loss of functionality. Simon. signature.asc Description: This is a digitally signed message part -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: Is it supported if I want to create a non-standard file structure?
On Thu, 2010-04-15 at 17:26 -0700, Jordan Peters wrote: > I'm going through the LFS 6.6 book with ubuntu 9.10 as the host and > I'm wondering if there are any tutorials anywhere to help if I want to > try creating a custom filesystem setup. Like having /apps instead > of /bin, /users instead of /home, etc. > I'm aware I'll have to edit every makefile I come across, but I > figured this is an experiment already, might as well have more fun > with it and make it more "mine". I think you'll find you're completely on your own with such a setup. Try, by all means, but when you run into difficulties, you'll struggle to find people who can help... For what it's work, having /users instead of /home wouldn't be a big deal - nothing actually cares where a user's home directory is. Nor is it quite a matter of editing every makefile - a lot should be possible just from passing --prefix, --bindir, etc to configure scripts. But expect a *lot* of trouble with the basic toolchain - gcc, glibc, binutils. Those three are the critical pieces that make everything else work - if your libraries aren't where the linker is looking for them, you're not going to get far until you work out how to change the linker. You'll also find problems from packages that simply assume you've followed the conventions - for example, scripts with #!/bin/sh hardcoded at the top. Simon. signature.asc Description: This is a digitally signed message part -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Gcc 4.5.0
Anyone here had tried building LFS with the newly released gcc 4.5? Just tried, using the same basic instructions as for current LFS (barring the addition of MPC to GMP and MPFR), but get the following partway through the 'make' step: checking for library containing strerror... configure: error: Link tests are not allowed after GCC_NO_EXECUTABLES. make[1]: *** [configure-target-libiberty] Error 1 make[1]: Leaving directory `/mnt/lfs/sources/gcc-build' make: *** [all] Error 2 I see a fair amount of history on gcc mailing lists with previous versions, but no clear answers... Simon. signature.asc Description: This is a digitally signed message part -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: Gcc 4.5.0
On Fri, 2010-04-16 at 23:45 -0500, Bruce Dubbs wrote: > Simon Geard wrote: > > Anyone here had tried building LFS with the newly released gcc 4.5? Just > > tried, using the same basic instructions as for current LFS (barring the > > addition of MPC to GMP and MPFR), but get the following partway through > > the 'make' step: > > > > checking for library containing strerror... configure: error: Link tests > > are not allowed after GCC_NO_EXECUTABLES. > > make[1]: *** [configure-target-libiberty] Error 1 > > make[1]: Leaving directory `/mnt/lfs/sources/gcc-build' > > make: *** [all] Error 2 > > > > I see a fair amount of history on gcc mailing lists with previous > > versions, but no clear answers... > > Simon, read the last two days posts of -dev. Just read them, so I see others are already looking at the matter. Yet you appear to have had no problem with gcc pass1? The instructions I've run are identical to those you've posted below, since barring MPC, they appear unchanged from 4.4.3... http://linuxfromscratch.org/pipermail/lfs-dev/2010-April/063753.html Simon. signature.asc Description: This is a digitally signed message part -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: NetworkManager working on Linux
On Wed, 2010-04-21 at 11:40 -0700, James Richard Tyrer wrote: > On 04/21/10 03:55, Simon Geard wrote: > > On Tue, 2010-04-20 at 14:57 -0700, James Richard Tyrer wrote: > >> As I mentioned elsewhere, I had an issue building it, but I now have > >> version 0.8 of it and the applet installed. > >> > >> To work, you need to add a basic configuration file: > >> > >> > >> /etc/NetworkManager/nm-system-settings.conf > > > > The file's not actually required - NM will run just fine without it, > > Do you have version 0.8 installed? Previous versions didn't seem to > need it, but I couldn't get them to work. :-| But, perhaps it isn't > always needed, although the documentation at GNOME Live states that it is. Yes, using 0.8, but I *do* have a config file now - added it at some point to make the warning go away. You could be right though, 0.8 might not start without it... > > though it's probably necessary in order to define system connections > > (i.e ones not tied to a logged-in user). > > It is necessary to define which plugin(s) you are using. And, I have > previously had issues with NetworkManager changing my hostname. The latter isn't a problem under recent versions. You're right, it used to reset to 'localhost', but at some point (one of the pre-0.8 snapshots, maybe) they've changed that logic. You only need to specify a plugin to set up system connections (or to set the hostname if not using LFS to do it). If you just use NM to connect to wireless while logged in, it'll work quite happily without any configuration... That said, my next LFS build will probably patch out all the network support scripts, and rely on NM to do everything... no point in having *two* things handling network config... Simon. Simon signature.asc Description: This is a digitally signed message part -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: Make error in glibc
On Tue, 2010-05-04 at 07:54 +0530, gaurav k wrote: > I have not deviated from the book at all, the only possible difference > might be any differences in the host system requirement. For > example, /bin/sh is not a symlink to bash, but I've only used bash > throughout. That *is* a deviation - you might be typing your commands into a bash shell personally, but scripts you run from there will be using /bin/sh. In theory programs using /bin/sh shouldn't use bash-specific features, but in practice, many do. Simon. signature.asc Description: This is a digitally signed message part -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: Make error in glibc
On Wed, 2010-05-05 at 12:05 -0400, Chris Staub wrote: > Then those that do should be considered buggy and must be fixed. > Personally I think that requiring /bin/sh -> bash is pointless, and that > if there are any packages that have /bin/sh while using Bash features > those packages should be corrected...but then I seem to be alone on this... Not at all alone - I personally think bash should automatically disable all bash-specific features if invoked as 'sh', precisely to avoid such problems. I've been bitten by this kind of thing a few times, learning the hard way that on some platforms (Solaris in particular), sh really is the classic Bourne shell... Simon. signature.asc Description: This is a digitally signed message part -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: Make error in glibc
On Wed, 2010-05-05 at 22:12 +0100, Ken Moffat wrote: > In the real world, I believe dash has improved a > lot once they started using it as /bin/sh, but it's > still sufficently different to cause problems, e.g. > in the kernel's build scripts. That, and most package maintainers have been happy to fix those problems - either by removing bash syntax, or by explicitly using /bin/bash instead. Simon. signature.asc Description: This is a digitally signed message part -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: How to create a list of md5 for every package
On Fri, 2010-05-07 at 21:37 +1200, jumbophut wrote: > On Fri, May 7, 2010 at 6:08 PM, xinglp wrote: > > like this > > a6de1cc6434cd64038b0a0ae4e252b33 autoconf-2.65.tar.bz2 > > c2972c4d9b3e29c03d5f2af86249876f automake-1.11.1.tar.bz2 > > 9800d8724815fd84994d9be65ab5e7b8 bash-4.1.tar.gz > > just like it `make` create wget-list > > > Or perhaps you meant: > > cd sources # the directory with all your package tarballs in it > for i in *; do echo $(md5sum $i) $i; done Unnecessarily complicated - just run "md5sum *" in the directory... Simon. signature.asc Description: This is a digitally signed message part -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: What works for me (LFS-6.6, x86_64)
On Mon, 2010-05-10 at 18:11 +0200, Dr.-Ing. Edgar Alwers wrote: > This is a sentence I clearly understand. Simon, I apologize for beeing > a little sarcastic instead of trying to contribute with a better text. > The problem is, I have had bad experiences with some developers. As > "Linux fan" pointed, "Little distros like lfs "don't get no respect" " I don't think it's that they don't get respect, they're just overlooked - remember that much of the development is done by developers employed by the big distros, by Redhat, Novell, Canonical. If people want them to consider the needs of other distros, those people need to be prepared to put in some work of their own. To use the PK case again, I'm sure the developers would be happy to support systems that don't use PAM - *if* someone provides a patch to do so, and will commit to maintaining it for the foreseeable future. The latter being critical - I'd certainly expect them to reject such a patch if nobody was willing to own it. Simon. signature.asc Description: This is a digitally signed message part -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: LFS for old Compaq LTE 5300 laptop
On Wed, 2010-05-19 at 00:12 -0500, Mike McCarty wrote: > Leho PƤrnapuu wrote: > > Hello, > > > > I have one old Laptop: Compaq LTE 5300. > > Something to watch for with Compaq computers: They > don't like to boot other than the pre installed > OS. Some of them will complain that you have a corrupted > MBR and enter some sort of recovery mode if they > have other than the pre installed OS on them. Many > of them have the BIOS in a special disc sector > or partition. If you inadvertently overwrite it, > then you may wind up with a mostly unusable system. What time period are you talking about for these comments? For it to be a 133MHz machine, we're talking about a machine from 1997 or so - a good thirteen years ago. Simon. signature.asc Description: This is a digitally signed message part -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: Nuisance (Was: Nusience)
On Wed, 2010-05-26 at 01:05 -0700, Johnneylee Rollins wrote: > You might want to master killfiles. > Just add the user to them. Also, gmail as a rather easy to manage > ignore feature. Blocking the user isn't the answer, since the emails appear to come from the LinkedIn member's email address, not from LinkedIn itself. But certainly, blocking them on some suitable header seems entirely reasonable. Simon. signature.asc Description: This is a digitally signed message part -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: chapter 6.16: configure: error: in `/sources/gcc-build/i686-pc-linux-gnu/libgcc':, ... C preprocessor "/lib/cpp" fails sanity check
On Tue, 2010-05-25 at 15:48 -0500, Bruce Dubbs wrote: > > just to bother, the book mention something about that somewhere?, do I > > miss something... > > 5.2 Important Note #1. > >-- Bruce Out of curiosity, do you have any idea what proportion of questions on this list are due to people not reading, or forgetting the contents of either that note, or the one about where instructions should be run from. I can't really see how things can be improved (short of blinking red text on *every* page), but it does seem like half the problems we deal with are from people not following it. Are people not even reading that entire "General Compilation Instructions" page, and skipping straight to the packages? Simon. signature.asc Description: This is a digitally signed message part -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: LFS-6.6, Stage2, glibc, nscd.c:442
On Mon, 2010-05-31 at 09:35 -0700, Paul Rogers wrote: > On Mon, 2010-05-31 at 00:36 +0100, Ken Moffat wrote: > > What you are overlooking is that "doing it my way" comes with "when > > it breaks, I get to keep all the pieces". > > What a curious thing to write in a SUPPORT forum of a LINUX distribution > with the motto, "YOUR DISTRO, YOUR RULES." "Your distro, your rules" - but going along with that, there's the unstated "your rules, your problems". There's nothing wrong with exploring uncharted waters, but if you sail into the middle of "here be dragons", you can't expect to stop and ask for directions... Simon. signature.asc Description: This is a digitally signed message part -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: LFS-6.6, Stage2, glibc, nscd.c:442
On Mon, 2010-05-31 at 19:11 -0400, x2...@lycos.com wrote: > So Paul, what i really need to say is "that if it breaks, you get to > keep the pieces" but maybe, more important, "if it doesn't break, > you still get to keep the pieces". Well yes, that's true. But then, LFS is a natural path for those of us who learn by taking things apart and trying to put them back together again. Having it still actually working afterwards was always a bonus... :) Simon. signature.asc Description: This is a digitally signed message part -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: perl, gdbm and perl's obsequious help
On Thu, 2010-06-03 at 02:39 -0400, Neal Murphy wrote: > On Thursday 03 June 2010 00:30:58 Chris Staub wrote: > > I just built Perl in Chapter 5, and I do have gdbm installed (this is on > > a couple-week-old LFS svn host system). I get " NOT found." > > during Configure, and it doesn't look like anything in Perl tries to > > link to libgdbm. > > Dig deep. Perl's configure, trying to be ever-so-helpful, finds it > (libgdbm.*) > on the host (your LFS SVN host) and configures support for the package. Curious... like Chris, I routinely build LFS from a host with gdbm installed (i.e LFS itself), and never observed any problems relating to perl. Indeed, even on completed LFS install, /usr/bin/perl isn't linked against gdbm, and runs fine if I remove the libgdbm libraries. Can you be more specific about the problems you're seeing? Does the perl executable fail to run at all, unable to link to libgdbm.so? Or is it something less obvious? Simon. signature.asc Description: This is a digitally signed message part -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: LFS-6.6, Stage2, glibc, nscd.c:442
On Wed, 2010-06-02 at 17:05 -0700, Paul Rogers wrote: > And if that leads some of us to scoff at the idea of dragons, it has, in > fact, done the job. Oh, I don't scoff at the idea of dragons - I just don't recommend stopping and asking them for directions. :) Simon. signature.asc Description: This is a digitally signed message part -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: the issue of environment value $PS1 set under Ubuntu 10.04
On Sun, 2010-06-06 at 17:56 +0800, littlebat wrote: > Base on my knowledge, I found the env command can't set like "set +h" > or "umask 022" those are set in the file .bashrc. I don't know of any way of setting umask that way, but any option you can pass to "set" can also be passed to "bash" itself, i.e "bash +h" to run a shell with path caching disabled. Simon. signature.asc Description: This is a digitally signed message part -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: Some LFS 6.5 -> 6.6 Inconsistencies
On Mon, 2010-06-07 at 17:45 -0500, Bruce Dubbs wrote: > That's not so easy. We create the book with Docbook xml. I don't know > of a way to to do that. We do have both a change log and a "What's new" > section. We are trying to do a new release every 6 months. If you look > at "What's new", almost every package changes each time, but most use > the instructions unchanged. Yeah, I generally work off the Changelog page when updating the scripts I build everything with, working up the list from whatever date I last built against. Works for me - Changelog open in one tab, and the index in a second one. Simon. signature.asc Description: This is a digitally signed message part -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: HSR's
On Tue, 2010-06-08 at 19:54 -0400, Neal Murphy wrote: > So, in the case that spawned this thread, the used could have continued to > use > his shell, albeit vey carefully and judiciously, until he exitted that shell. > Even if he performed an 'rm -rf /', his shell would continue to run and he > would continue to be able to use shell built-ins until he exitted that shell. > At that point, only the root directory, '.' and '..' would be accessible. Not necessarily - it depends on what else is happening on the system at the same time. A few months back, I accidentally removed part of udev, and the system died almost instantly - I never found out exactly what went wrong, but speculate that something caused /dev to go missing, which locked up X... That was fun... ended up running off a Ubuntu live-boot for a week until I could get a new LFS build done... :( Simon. signature.asc Description: This is a digitally signed message part -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: live and learn
On Wed, 2010-06-09 at 12:00 -0400, linux fan wrote: > Speaking of backups and rescue disks, I am using an "rsync snapshots" > style of backup. It does not compress, yet it can save multiple > "copies" of a system at different points in time using remarkable > little disk space. Any "copy" can be rsynced to any mount point and so > I can recover or load up any system "copy" in around 20 minutes. I quite like an idea Fedora are working on - if installed to a btrfs partition, use it's snapshot support to provide an easy rollback option when installing updates. I've been meaning to try that approach myself, since I've trashed more than a few systems while trying to update them. Simon. signature.asc Description: This is a digitally signed message part -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: live and learn
On Thu, 2010-06-10 at 03:55 -0500, Mike McCarty wrote: > Simon Geard wrote: > > I quite like an idea Fedora are working on - if installed to a btrfs > > partition, use it's snapshot support to provide an easy rollback option > > when installing updates. > > > > I've been meaning to try that approach myself, since I've trashed more > > than a few systems while trying to update them. > > Hey, if you find out more, please do post information here! > That sounds like a very attractive possibility. The concept is simple enough - create a snapshot before doing anything hazardous, and provide a mechanism for mounting a specific snapshot at boot time. Here's a link to the Fedora proposal... https://fedoraproject.org/wiki/Features/SystemRollbackWithBtrfs It seems this feature is partially implemented in F13 - the automatic snapshots are enabled, though the UI is missing. On an LFS, there's no automatic update process, so I figure I'd just create periodic snapshots, say a weekly cron job, or maybe just manually. And then configure an entry in grub to boot from the previous snapshot. The concept is simple enough, I just haven't had a good chance to try it out... Simon. signature.asc Description: This is a digitally signed message part -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: live and learn
On Fri, 2010-06-11 at 13:16 -0500, Mike McCarty wrote: > Use a red colored prompt when running with root authorization. Oh yes, I can't agree with that one enough. Not that you want to accidentally run "rm -rf ~" as *any* user, but the red root prompt is a useful reminder that you're playing with a loaded gun. Simon. signature.asc Description: This is a digitally signed message part -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: Linux From Scratch (Version SVN-20100529) - 6.9. Glibc-2.11.2
On Sat, 2010-06-12 at 02:55 +, Face Man wrote: > checking whether ld is GNU ld... no > checking for > /tools/lib/gcc/x86_64-unknown-linux-gnu/4.5.0/../../../../x86_64-unknown-linux-gnu/bin/as... > > /tools/lib/gcc/x86_64-unknown-linux-gnu/4.5.0/../../../../x86_64-unknown-linux-gnu/bin/as > checking version of > /tools/lib/gcc/x86_64-unknown-linux-gnu/4.5.0/../../../../x86_64-unknown-linux-gnu/bin/as... > 2.20.1.20100303, ok > checking for ld... ld > checking version of ld... v. ?.??, bad This looks dodgy - something has clearly gone wrong at an earlier point in the build. ld is part of binutils, so the problem was probably with the installation of that package back in chapter 5. Does /tools/bin/ld even exist? Simon. signature.asc Description: This is a digitally signed message part -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: LFS (Version SVN-20100529) - 5.19. File-5.04
On Mon, 2010-06-14 at 13:43 +0300, Face wrote: > On Mon, Jun 14, 2010 at 1:25 PM, littlebat wrote: > > It seems you have missed something, I remember we should always use > > non-root user (lfs) to compile LFS in chapter 5 at least. > > > when i add lfs user i get permissions problems This probably means that you've either missed the commands that give 'lfs' write access to $LFS/tools and $LFS/sources (section 4.3), or you're trying to extract tarballs or install packages to paths outside of those two directories (see directions at bottom of section 5.3). An observation - doing the chapter 5 build as root isn't going to cause problems if everything else is done exactly right. But it has downsides if you *don't* get everything else exactly right. The limited 'lfs' user can write to just those two directories, and mistyped commands will fail with permission errors. Running as root, the same mistyped command may appear to succeed, but cause problems later - it may also damage the host system, since it has permissions to mess with /bin, /etc... Simon. signature.asc Description: This is a digitally signed message part -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: priority in install of packages
On Tue, 2010-06-15 at 12:23 -0400, Neal Murphy wrote: > To be absolutely clear, and to directly answer the question, yes, it is a > problem because Expect will fail to compile if the TCL programs and libraries > are not found. Or worse, will find the chapter 5 versions of those programs & libraries in /tools, hardcode references to them, and break one /tools is deleted in the future. Simon. signature.asc Description: This is a digitally signed message part -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: priority in install of packages
On Tue, 2010-06-15 at 21:45 +0430, Mohsen Pahlevanzadeh wrote: > In appendex (Dependency section), we have 2 part for each package: The appendix is for information only. When it comes to installing packages, follow the *exact* instructions in the main book, in the *exact* order they're in. Simon. signature.asc Description: This is a digitally signed message part -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: Problem installing the "nouveau" driver
On Fri, 2010-06-18 at 21:59 +0100, Andrew Benton wrote: > No, you're not helping. His video card need some firmware to work with > the nouveau driver. He needs to either put the firmware in /lib/firmware > or compile it into the kernel. If he read the syslog from the failed > boot he would see the kernel complaining about the lack of firmware. > Without the firmware his video card can't bring up the framebuffer. Your advice is out of date - it was valid when Nouveau first entered the kernel, but the 2.6.34 kernel Alex is running does not use external firmware. Simon. signature.asc Description: This is a digitally signed message part -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: Problem installing the "nouveau" driver
On Fri, 2010-06-18 at 17:35 +0100, Andrew Benton wrote: > On 18/06/10 02:29, al...@verizon.net wrote: > > Hello, > > > > i686-pc-linux-gnu, 2.6.34, udev-156, (B)LFS > > NVidia GeForce 8300GS > > > > I think the 8000 series needs extra firmware. My wife's computer has a > 8400GS which needs nv86.ctxprog and nv86.ctxvals. I compile them into > the kernel with Not as of the 2.6.34 kernel - my 8500GT card works fine under that kernel. It may be a factor that I have support built-in rather than as a module - certainly, I've not seen the problems Alex reports. FYI, I've been trying out the full Nouveau 3D (Gallium) driver lately - it's easy enough to do, if you're happy to build libdrm and Mesa from git. It works basically, enough to run some OpenGL games, though clearly has some missing functionality - e.g the Spring RTS engine is playable, but with some terrain textures not rendering... Simon. signature.asc Description: This is a digitally signed message part -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: Problem installing the "nouveau" driver
On Sun, 2010-06-20 at 18:04 -0500, al...@verizon.net wrote: > I bzip2'ed the file, so let's see if 11702 bytes can now > sneak below the 50 KB limit :). > > BTW, 'config' is for a system of this nature: > > ASUS P5E-VM HDMI, intel Core2Duo E8400 @ 3.00GHz. > G.SKILL 4GB (2 x 2GB) DDR2 1000 (PC2 8000). > NVIDIA GeForce 8300GS 512MB 256-bit GDDR2 > PCIe-2.0x16 BIOS 60.86.45.00.26, > Samsung SyncMaster 2494 60Hz, DVI connection. Your system and config seems broadly similar to mine, so I don't see an obvious reason why mine works and your doesn't. The only thing that particularly stands out is that you have all the DRM stuff built as modules, where I have it built-in. If you've not already done so, perhaps that'd be worth trying? Simon. signature.asc Description: This is a digitally signed message part -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: urgent
On Sun, 2010-06-20 at 22:32 +0100, Andrew Benton wrote: > Err, I think you'll find that learning how to "design and implement my > own operating system" is more than a bit of homework. It would take > hundreds of gifted people more than a decade to achieve such a thing. Not necessarily - I gather building your own OS is a relatively common advanced project in Comp Sci courses. Not a production-ready system like Linux, of course, but a minimal system one step up from a bootloader, capable of loading a binary from disk and running it. Simon. signature.asc Description: This is a digitally signed message part -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: Re: Problem installing the "nouveau" driver
On Mon, 2010-06-21 at 11:21 -0500, al...@verizon.net wrote: > I'm attaching the compressed copy of the new 'config'. I notice CONFIG_DRM_I915 is set, as a module. Shouldn't matter in theory since you presumably don't have Intel graphics hardware as well as NVidia, but probably worth disabling it anyway. Also, you have CONFIG_FIRMWARE_EDID=y, which isn't set on my system. Might be relevant, since while the docs say it should be safe to enable, they note: "If you experience extremely long delays while booting before you get something on your display, try setting this to N." So, probably worth turning it off, see if that makes a difference. From the docs, it sounds like it's an option you enable to work around problem hardware, isn't needed by the majority, and might potentially be causing problems. Simon. signature.asc Description: This is a digitally signed message part -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: Problem installing the "nouveau" driver
On Tue, 2010-06-22 at 10:49 -0500, al...@verizon.net wrote: > How come the original NVIDIA driver, after all these ugly > and hostile parameter settings I threw at it, it was left still > standing at the end of the day? The one thing the nVidia driver is supposedly intolerant of is framebuffer drivers - I'd guess it won't mix well with using Nouveau to get a high-res console. Barring that, it uses so little of the standard graphics stack that as long as nothing actively interferes, it's hard to misconfigure. Simon. signature.asc Description: This is a digitally signed message part -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: Problem installing the "nouveau" driver
On Mon, 2010-06-28 at 13:28 -0500, al...@verizon.net wrote: > Whether I have "fbcon" in the kernel or as a module (with its > convenient font, 80x25 full wide-screen) I cannot (or I > don't know how to) change the 240x67 resolution back to > what I have always used and preferred, 80x25 full wide-screen. > The original NVIDIA driver does not "touch" my text screen > so this switch to 240x67 has caught me by surprise. Right, because the NVIDIA driver is an X driver only - it has no effect when the X server isn't running. The piece of Nouveau included in the kernel isn't an X driver - it's a kernel framebuffer driver, meaning that it's not limited to X. The issue with the modes is because "text resolutions" like your favoured 80x25 are provided by the BIOS, the lowest-common-denominator interface that goes back to the stone age. Once Nouveau loads, it's talking to the video card natively, and those BIOS modes just aren't applicable anymore - all that matters is pixels, 1920x1080 of them in your case. The right fonts can emulate that 80x25, but nothing will actually put you in that mode until the BIOS is driving the video card again. What exactly is your problem with running in the high-res mode? If it's just that the font size is too small, I'd have to repeat Stephane's suggestion of selecting a different console font (somewhere in the kernel config, I think). Simon. signature.asc Description: This is a digitally signed message part -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: Problem installing the "nouveau" driver
On Tue, 2010-06-29 at 10:25 -0500, al...@verizon.net wrote: > With "nouveau", my boot-up goes through these basic steps: > > 1. The original/"regular" console sequence (80x25) At this point, the display is under BIOS control, running in the default text mode - the same as for GRUB, and (depending on your hardware), the BIOS setup screen. Emphasis on the BIOS bit - no video driver involved, this is entirely between the BIOS chip on the motherboard, and a counterpart on the video card. > 2. Nouveau is loaded by UDEV > 3. At this point, the console goes blank for a sec or so. > The preceding messages are wiped out. At this point, Nouveau takes over the video card from the BIOS - think of it as rebooting the GPU. Nouveau can make far more effective use of the card, but it does so very differently. > 4. The remainder of the boot-up sequence proceeds and stays > in 240x67 all the way to the prompt (and beyond). > 5. I can never change the 240x67 resolution of the console > text mode. The key thing is that there's no such thing as a 240x67 resolution. What you have is a combination of 1920x1080 pixels, and a font size that just happens to fit 240 columns and 67 lines of text - no different from running a full-screen terminal window under X. So if you want 80x25 under Nouveau (or any other framebuffer driver), you need the kernel to use the right fonts to achieve that. The 80x25 mode you boot in is special in that regard, because that combination of pixels and fonts is implemented in the BIOS itself, as a standard feature going back 20 years. And being provided by the BIOS, it's not available if something else is driving the hardware. An aside: One thing you might have noticed under the NVIDIA binaries. If you switch from X to a console, it switches back to 80x25, since those drivers are X only. When you switch back to X, they take over again, and have to work out what state the hardware is in - occasionally you see corruption (usually in OpenGL apps), where the drivers haven't gotten it right. I've once seen black and white patterns reminiscent of a text console appearing in textures, for example, as if the BIOS had put something into video memory, and NVIDIA hadn't cleared it out again. It's a driver bug when it happens, of course, but I mention it to illustrate the conflict between BIOS and OS driving the hardware. > BTW, the _exact_ steps of how to set a font in kernel which > would take effect after "nouveau" is up, will be highly > appreciated. I've not tried it myself, but take a look in menuconfig under "Device Drivers" -> "Graphics Support" -> "Console display driver support" -> "Select compiled-in fonts". I think that if you pick something like "VGA 8x16", you'll get something like what you want. But like I said, I've not tried it. Simon. signature.asc Description: This is a digitally signed message part -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: Problem installing the "nouveau" driver
On Wed, 2010-06-30 at 13:25 -0500, Mike McCarty wrote: > Simon Geard wrote: > > On Tue, 2010-06-29 at 10:25 -0500, al...@verizon.net wrote: > > You were ok up to here... A combination of trying to simplify, and trying to remember stuff I last played with in college, back when accelerated graphics add-on cards were a rarity... :) > That font is implemented in ROM inside the graphics controller, not > the BIOS. To get access to it, one simply has to command the controller > to use it. If one does that, the boot screen content usually magically > reappears, unless one also switches to a different RAM page for display, > or if some frame buffer mode was used during boot. Ok, so that stuff is on the video card, not the motherboard? Nevertheless, the point was more to do with the two ways of talking to that controller - either by treating it as an modern NVidia card (as Nouveau does), or as an antique VGA chip (as the BIOS does). And the latter isn't consistent with Nouveau's purpose in providing the former. Simon. signature.asc Description: This is a digitally signed message part -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: Problem installing the "nouveau" driver
On Wed, 2010-06-30 at 20:46 -0500, al...@verizon.net wrote: > Simon, Trent, Mike: > > Thank you very much for the rich, detailed, interesting > theories and explanations about video drivers, booting, > X, fonts, pixels, my boot-up component steps, etc. > All this knowledge and experience is obviously worth a > thread of its own, something I'd be following keenly. I'm not providing that information just for your curiosity, though, nor (entirely) out of a propensity to lecture. I'm providing those explanations to help you understand what you're dealing with, and why what you're asking for isn't as simple as just "switch back to 80x25". Ultimately, you've got two options. One, use Nouveau's framebuffer console, and emulate 80x25 through appropriate font selection - I can't help you any further on that, because I don't know. Or two, don't use Nouveau - I know that's not much help if you want to use X, but that's how it works. That said, one thing you might try is unloading the Nouveau kernel module. If that successfully returns you to 80x25, you could probably blacklist the driver to prevent it from loading on bootup, and write a wrapper script around 'startx' to rmmod the module once you're done with X. No idea if it'd work, but it's certainly worth a try... Simon. signature.asc Description: This is a digitally signed message part -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: Problem installing the "nouveau" driver
On Thu, 2010-07-01 at 17:59 -0500, Mike McCarty wrote: > Apparently, it's the stated non-goal to support switching back > by the authors (or at least owners) of the driver. > So, unless he's willing to get the source for the driver, and > rewrite portions of it himself, there's not going to be a way > to "go back". I don't think it's a matter of Nouveau specifically, as the kernel graphics infrastructure in general. Having any such driver loaded means using the graphics-mode console. That's the impression I got from Stephane's comments to Alex, and it makes sense to me that things work that way. Simon. signature.asc Description: This is a digitally signed message part -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: LFS (Version SVN-20100529) - 5.9. Binutils-2.20.1 - Pass 2
On Thu, 2010-07-01 at 22:39 -0400, Neal Murphy wrote: > Sometimes, deleting everything and starting anew is all you *can* do. More than that, it's often easier too. Less effort to start again paying closer attention to the instructions, than to spend time trying to work out how to fix the results of a mistake. Simon. signature.asc Description: This is a digitally signed message part -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: LFS (Version SVN-20100529) - 5.9. Binutils-2.20.1 - Pass 2
On Sat, 2010-07-03 at 02:19 +0300, Face wrote: > Neal, > > thanks alot for going through the script I appreciated. > > well, I do have error detection i disable it on the script i attached. > its like this That's not error detection - that's just logging stderr. The biggest problem with scripted builds is making sure the script stops immediately when an error occurs, instead of ignoring it and continuing. If "configure" fails, don't run "make". If "make" fails, don't run "make install". And if *anything* fails, don't just move onto the next package. That's experience talking, I might add. My LFS builds are almost always scripted, and more than once, I've investigated a compile problem that's turned out to be the result of an earlier package failing unnoticed. Wastes a lot of time, that way. Simon. signature.asc Description: This is a digitally signed message part -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: LFS (Version SVN-20100529) - 5.9. Binutils-2.20.1 - Pass 2
On Sat, 2010-07-03 at 03:40 +0300, Face wrote: > it has been over 2 months since my first try i think. going through > the book manually over and over, I had many mistypes which force me to > start over. Therefore, I put them in a script so i can avoid mistypes. To start with, copy-paste from the book is generally a better solution to mistyping than scripting - it means you're running them one by one just as if you were typing them by hand, so you can easily notice if something goes wrong. Scripted builds are fine, but they take quite a bit of work to get right. If you're not able to run through the process manually, I suggest trying to automate that process will be just as hard, if not more so. Simon. signature.asc Description: This is a digitally signed message part -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: Problem installing the "nouveau" driver
On Fri, 2010-07-02 at 11:14 -0500, Mike McCarty wrote: > OTOH, my version (with X and GNOME) seems to switch to 80x25 very > easily when I press CTRL-ALT-F1 through CTRL-ALT-F6, so the kernel > must support _something_ along those lines. You're using the Nouveau driver? The X-only drivers like NVIDIA behave that way, but my understanding is that newer open drivers like Nouveau are a different story. Something to do with the kernel modesetting work that's been going on, I guess... Simon. signature.asc Description: This is a digitally signed message part -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: LFS (Version SVN-20100529) - 5.9. Binutils-2.20.1 - Pass 2
On Wed, 2010-07-14 at 13:37 -0500, Mike McCarty wrote: > Simon Geard wrote: > > [...] > > > That's experience talking, I might add. My LFS builds are almost always > > scripted, and more than once, I've investigated a compile problem that's > > Is there a reason you prefer a script to an automated package like > JHALFS? Mostly because those scripts pre-date JHALFS - I've been using them for many years now. Also, because I enjoy doing things my own way. Simon. signature.asc Description: This is a digitally signed message part -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: Can i build lfs without a dedicated partition ?
On Thu, 2010-07-08 at 22:32 +0530, $reeHari wrote: > Hi , > I just started with lfs using book version 6.6 . I had previously > started with book ver 6.3 and in dat , author said it is okey to build > lfs without a partition , in a directory . But in the new version of > the book , no such thing is mentioned . So somebody advice me on > this . Do i really need a dedicated partition for the build or can i > build it in a directory as said in ver 6.3 ? If a dedicated partition > is essential , can i use a usb stick ? *Building* without a separate partition is easy enough - a blank partition mounted to /mnt/lfs is much the same as an empty /mnt/lfs with nothing mounted to it. However, that will leave you with an LFS system in a subdirectory - not something you can actually boot. But you could then copy that directory to a new location, make any config changes to suite that location, etc. And yes, you can build it on a USB stick quite happily, as if it were a normal hard disk. The catch is whether the BIOS can boot from it, but that shouldn't be a problem unless you're using very old hardware... Simon. signature.asc Description: This is a digitally signed message part -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: LFS (Version SVN-20100529) - 5.9. Binutils-2.20.1 - Pass 2
On Fri, 2010-07-16 at 23:44 -0500, Bruce Dubbs wrote: > That's a good reason. We each have our preferences. I like jhalfs for > lfs and scripts for blfs. Scripts for BLFS is a good challenge too, since 1) there are a lot more packages (and flavours of build) involved, and 2), it's not all a linear progression like LFS, there's dependencies to worry about. My current approach involves hundreds of small 'spec' files, and an over-complicated Python script that generates a Makefile that performs the actual installations. Needs a bit of work, though, since the current scripts can't really handle parallel versioning very well, which I need for the Gtk+/Gnome 2 -> 3 migration... Simon. signature.asc Description: This is a digitally signed message part -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: Hit a Speed Bumps with Grub 2
On Wed, 2010-07-21 at 15:06 -0500, Bruce Dubbs wrote: > Good point Trent. I do recommend the use of a separate 100M partition > for /boot for all OSes installed. It avoids that kind of problem. Only for Linux. Not much help for Windows, which Dan was asking about. I don't think I'd use such a partition between Linux distros either - LFS won't clobber files from other distros, but I'm not sure I'd trust Fedora or Ubuntu to do the same. Simon. signature.asc Description: This is a digitally signed message part -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: Networking Is My Weakest Area--A Really Basic Question
On Wed, 2010-07-21 at 10:46 -0500, Dan McGhee wrote: > My biggest conceptual error was that I thought all the static stuff got > used only with an ISP and their stuff is all on the router. Nope - the router will be configured to talk to the ISP, but your computer still needs to know how to talk to the router. If you use a DHCP client, the computer gets that information from the DHCP server (in a home setup, that's the router itself). Otherwise, you get to set it up yourself. Unsurprisingly, most systems will use dhcp by default. But LFS, being what it is, leaves that up to the user. Simon. signature.asc Description: This is a digitally signed message part -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: 5.7.1 Installation of Glibc-2.11.1
On Sun, 2010-07-25 at 00:08 -0700, Clyde Reed wrote: > bash: ./version-check.sh: /bin/bash^M: bad interpreter: No such file > or directory Did you create the script with a Windows text editor such as Notepad? The error above occurs when running a script that's been saved with Windows-style end-of-line characters - the ^M above is a carriage-return character, which Windows uses along with a line-feed to indicate end-of-line. On Unix, only the LF is used, and the CR becomes part of the preceeding line. Simon. signature.asc Description: This is a digitally signed message part -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: Working With the Boot Scripts to Bring Up Network
On Sat, 2010-07-24 at 12:39 -0500, Dan McGhee wrote: > While I like to do these sorts of things manually a couple of times, I'd > like it to me "seamless" after I get more of my system built. To this > end, Simon mentioned Network Manager and Neal suggested wicd as a > possibility. Ultimately, I'll probably go with wicd. As a third option, there's also the unfortunately named ConnMan, though I don't know much about it. Simon. signature.asc Description: This is a digitally signed message part -- http://linuxfromscratch.org/mailman/listinfo/blfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
Re: GCC Pass 2 Error - LFS-svn
On Fri, 2010-09-03 at 00:15 +0200, Martin ZajĆc wrote: > configure: error: Building GCC requires GMP 4.2+, MPFR 2.3.1+ and MPC 0.8.0+. See comments below... > /mnt/lfs/scripts/chap5/gccP2: line 140: --disable-libstdcxx-pch: > command not found That one looks like a typo - it's trying to run one of the parameters to configure as a separate command, which means you've probably missed a '\' at the end of a line, or something like that. > I have all folders on right place. > . > āāā gcc-4.5.1 > āāā gcc-4.5.1-BUILD > āāā gmp > āāā mpc > āāā mpfr > No, you don't. Remember, *all* instructions for a package are assumed to be run from inside the extracted directory for that package, which in this case is gcc-4.5.1. The gmp, mpc, and mpfr directories should be inside that one, as they will be if you're in that directory when you extract them. Simon. signature.asc Description: This is a digitally signed message part -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: LFS-svn chap 6.7 (gcc: No such file or directory)
On Sat, 2010-09-04 at 20:05 +0100, Ken Moffat wrote: > Sometimes people accidentally miss one of the packages, > or build it but fail to install it. Or more commonly with scripted builds, the script fails to recognise that a command has failed, and continues with the next step. Scripted builds *must* detect errors and stop immediately - otherwise you just get obscure problems later in the build, where it's much harder to find the original cause. Simon. signature.asc Description: This is a digitally signed message part -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: LFS-svn chap 6.7 (gcc: No such file or directory)
On Sun, 2010-09-05 at 13:22 +0200, Martin ZajĆc wrote: > And errors in configure, make, make install scripts detecting > automatically and stop build. Are you sure? Until recently, that's what I *thought* my scripts were doing, until one of the aforementioned "obscure problems" convinced me otherwise. Turned out the error detection had been broken by a change I'd made to the scrips a few months earlier... Simon signature.asc Description: This is a digitally signed message part -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: "auto" in /etc/fstab
On Thu, 2010-09-30 at 16:18 -0500, rhubarb...@poetworld.net wrote: > Is there a disadvantage of using auto for file system type in the > /etc/fstab file? I have three non-swap partitions. Why not list them > all as auto? I may be wrong, but I *think* auto works only for filesystems that are registered at the time you try to mount them - in other words, if your filesystem support is built as a loadable module, it may not work. I remember having issues of this kind while playing with initramfs support ages ago... Simon. signature.asc Description: This is a digitally signed message part -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: LFS Init scripts?
On Thu, 2010-09-30 at 13:32 -0700, Mark Knecht wrote: > Am I fooling myself in thinking that for a very simple hardware > system, maybe just an EXT2 boot, EXT3 / and a swap partition, that for > this specific machine the init scripts might be reduced to something > like 10-20 bash commands which get me to the bash command line where > as root I could use the system? Yeah, could be. Look at what the scripts in /etc/rc.d/rcsysinit.d are doing, but the minimal list is going to look something like - * mount /proc and /sys * start udev to populate /dev (see /etc/rc.d/init.d/udev) * remount / read-write (kernel mounts it read-only before init starts) * start swap * bring up localhost network There's other stuff done by LFS, but I think that list is probably enough to get you to a bash prompt. Actually, the last two aren't critical for boot either, but you likely want them anyway... Simon. signature.asc Description: This is a digitally signed message part -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: SVN 20100923 - gcc-4.5.1, pass 1, won't compile
On Sat, 2010-10-09 at 23:15 +0100, Andrew Benton wrote: > Also, you may need to tweak your glibc check to work on systems that > don't have executable shared libraries. On my systems almost all shared > libraries (including /lib/libc.so.6) have premissions 644 so I get: Out of curiosity, why do you do things that way? It's certainly unusual; indeed, I always assumed .so files were required to be executable - otherwise why would every single distribution and installer make them so? Simon. signature.asc Description: This is a digitally signed message part -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: SVN 20100923 - gcc-4.5.1, pass 1, won't compile
On Sun, 2010-10-10 at 11:57 +0100, Andrew Benton wrote: > On Sun, 10 Oct 2010 21:09:38 +1300 > Simon Geard wrote: > > > On Sat, 2010-10-09 at 23:15 +0100, Andrew Benton wrote: > > > Also, you may need to tweak your glibc check to work on systems that > > > don't have executable shared libraries. On my systems almost all shared > > > libraries (including /lib/libc.so.6) have permissions 644 so I get: > > > > Out of curiosity, why do you do things that way? It's certainly unusual; > > indeed, I always assumed .so files were required to be executable - > > otherwise why would every single distribution and installer make them > > so? > > > > I got the idea from Ubunut, where most of the .so files are 644. Huh... so they do... never noticed that before. Ok, forget that "every single distribution" comment then... I guess some non-Linux (or non-glibc) systems must require (or have required) shared libraries to be executable, so that it's become a convention even where it's not necessary. I should try the experiment on some of the assorted AIX and HP-UX servers at work... Simon. signature.asc Description: This is a digitally signed message part -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: grub installation error at chapter-8.4
On Wed, 2010-10-13 at 07:35 +0530, Madhan M wrote: > On 13 October 2010 07:26, Bruce Dubbs wrote: > Yes, but is /dev populated when you run grub-install? > Thanks for your reply Mr.Bruce. When i issue the grub-install > command,there is no further progress. The terminal is freezed. No > change. Hi Madhan. We understand that - however, that's not what Bruce asked. The question is whether /dev is populated at the time you run grub-install. If grub-install cannot see the device it's trying to install to, it's not going to work. Simon. signature.asc Description: This is a digitally signed message part -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: Book-5.8_Adjusting the ToolChain (v6.7)
On Fri, 2010-10-22 at 08:59 -0700, Nolan wrote: > Can I change the directory and file ownership to "lfs" with a chown, or > is the fact that the programs were compiled as > "root" have a detrimental effect? The problem isn't that you're running as root - it's that you found that you needed to run as root. That's a sign that you've done something wrong somewhere in the instructions, since the 'lfs' user should have permission to write to anywhere it *should* try to write to. Now, that could simply be that you've not run the commands giving 'lfs' permission to write to the necessary locations - in which case, running as root was harmless. But it could also be that you've mistyped one of the build steps, and the 'lfs' user is trying to write to locations it *shouldn't* be able to write to - for example, you ran something with --prefix=/usr instead of /tools. In that case, running as root is potentially disastrous, since you might be overwriting files belonging to the host distribution, rendering it unbootable. Simon. signature.asc Description: This is a digitally signed message part -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: Book-5.8_Adjusting the ToolChain (v6.7)
On Mon, 2010-10-25 at 22:14 -0700, lfs wrote: > So the question is, Should I create the gcc-build directory and *then* > untar gmp,mpfr, mpc, from that(gcc-build) directory No. Remember the rule - before starting on a particular page, extract it's tarball, and cd into the directory. So, when extracting gmp, mpfr, and mpc, you should be in the GCC source directory (gcc-4.5.1). The idea is that you should then have 'gmp', 'mpfr' and 'mpc' directories within the main GCC source, and the reason is that they'll be built by GCC automatically. Unfortunately, the book doesn't really explain this, with the result that people seem to improvise this part, instead of following the instructions to the letter. Editors, can we perhaps add some extra text to the GCC Pass 1 instructions (and perhaps a short reminder to Pass 2 as well)? Right now, all the book has to say is: "GCC now requires the GMP, MPFR and MPC packages. As these packages may not be included in your host distribution, they will be built with GCC" Can we go into a bit more detail about that "build with GCC" bit? Extracting seemingly unrelated tarballs into the GCC source tree *is* an unusual thing to do, unless you're aware that the GCC build will magically build those packages if present. A short explanation of how this works, and why we do it would 1) improve the educational aspects of the book, and 2) discourage people from improvising because the instructions seem wrong. Simon. signature.asc Description: This is a digitally signed message part -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: GCC Build Instructions (was Re: Book-5.8_Adjusting the ...)
On Tue, 2010-10-26 at 15:54 -0500, Bruce Dubbs wrote: > If you are referring to the tar commands in section 5.5.1, what is not > clear? There is no chdir command before the tar commands. Section 5.3 > clearly says: > > b. Change to the directory created when the package was extracted. > c. Follow the book's instructions for building the package. > > I am not clear about what is unclear. Bruce, I believe the problem is that these specific instructions are *unusual*. Given that we install these three packages separately later in the book, an instruction to extract them into the gcc source directory seems wrong, and should have an explanation as to why it's the correct thing to do. I don't mean a clarification of the instructions - I mean a statement of why the instructions are there at all. Consider, we have a short explanation for each of the options passed to configure, and why we have to fix a missing symlink after the install. But we don't give anything for the tar commands, beyond a vague comment about "will be built with GCC". Why do we extract these files into this directory, then apparently do nothing with them? My thinking here is that our readers are smart people. And smart people don't just follow the instructions blindly - they see odd stuff like this, and wonder if it's an error in the book, wonder if those packages should have been installed 'properly' like in Chapter 6... Simon. signature.asc Description: This is a digitally signed message part -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: GCC Build Instructions (was Re: Book-5.8_Adjusting the ...)
On Tue, 2010-10-26 at 21:23 -0500, Bruce Dubbs wrote: > Yes. We've had enough posts of people reading too much into these > instructions. I've added some clarifying text. Let's see if that is > enough. For me, I think the important detail is that we extract these packages here, so that the GCC build will automatically pick them up. I think that's clear in the new text. Can we maybe echo the reminder again in Pass 2, e.g replacing: "As in the first build of GCC it requires the GMP, MPFR and MPC packages. Unpack the tarballs and move them into the required directory names:" with: "As in the first build of GCC it requires the GMP, MPFR and MPC packages to be extracted into the GCC source directory. Unpack the tarballs and move them into the required directory names: Simon. signature.asc Description: This is a digitally signed message part -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: GCC Build Instructions (was Re: Book-5.8_Adjusting the ...)
On Fri, 2010-10-29 at 10:14 -0500, Bruce Dubbs wrote: > We could, but we refer to the first build. I wouldn't think that this > change is necessary. However, I didn't think the change to pass 1 was > unclear either. The change already made was to prevent users from > assuming things that were not stated. I'd think that if they would > recognize that this was the same. > > Let's wait to see if we get any questions on this. Fair enough. I don't recall that we've ever had much problem with this in pass2 before anyway - I suppose that you're right, that the instructions are familiar by the time they reach that point. Simon. signature.asc Description: This is a digitally signed message part -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: boot 6.7 troubles
On Sat, 2010-10-30 at 18:11 -0600, John Mitchell wrote: > Thanks Andy and Ken. > > I suspect that some combination of modules and perhaps a missing > initrd.img file is my issue. > You don't need an initrd, unless you've created some fairly exotic setup - distros use them because they give a lot more flexibility around mounting the root filesystem, allowing for stuff like RAID / LVM setups, encrypted filesystems, or dealing with the case where a single kernel image needs to work on every possible piece of hardware and filesystem type. What *you* need is probably just a straightforward config that supports your hardware. In particular, it needs the drivers for the disk controller and for the root filesystem - and it needs those drivers as built-ins, not as modules. Running lsmod on Fedora can probably tell you what driver it's using for the controller - if it's a fairly modern SATA-based setup, it's likely to be the standard 'ahci', but on an older machine it might be something chipset-specific (which in my case is 'sata_nv'). Getting those two bits right should give you a booting system, at least. With that working, you can then go looking at the other config options to support the rest of your hardware - sound, network, etc - and anything else you might need or want. Simon signature.asc Description: This is a digitally signed message part -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: binutils "make" ... no go
On Wed, 2010-11-24 at 20:40 -0600, Bruce Dubbs wrote: > Your host is missing required packages. Run the script in section vii. Just a reminder, Bruce - pages are only numbered on the page itself, not in the table of contents. Referring to the "Host System Requirements" page is somewhat clearer than "Section VII" for those who don't know the book by heart. Simon. signature.asc Description: This is a digitally signed message part -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: make mrproper
On Sat, 2010-11-27 at 12:19 -0600, robert wrote: > r...@acer1:/usr/src/linux-headers-2.6.32-24-generic# make mrproper Ok, start from here. There's no way you should have ended up with a directory called "linux-headers-2.6.32-24-generic" by following directions in the book - there's no source package of that name, nor any package containing a directory of that name. So, where did it come from? The name sounds like some kind of distro-specific package... Simon. signature.asc Description: This is a digitally signed message part -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: make mrproper redux
On Mon, 2010-11-29 at 13:11 -0600, robert wrote: > What has gone wrong? Am I supposed to cd /sources/linux-2.6.36.1 and > then $make mrproper? Yes - remember the "General Compilation Instructions": 3. For each package: a. Using the tar program, extract the package to be built. b. Change to the directory created when the package was extracted. c. Follow the book's instructions for building the package. These instructions are important. Simon. signature.asc Description: This is a digitally signed message part -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: glibc compilation ok on Fedora 14
On Sun, 2010-12-05 at 11:45 -0500, Neal Murphy wrote: > On Sunday 05 December 2010 10:55:13 Rosario Turco wrote: > > ../glibc-2.12.1/configure - prefix = / tools \ > > - Host = $ LFS_TGT - build =$(../ glibc-2.12.1/scripts/config.guess) \ > > - Disable-profile - enable-add-ons \ > > - Enable-kernel = 2.6.22.5 - with-headers = / tools / include \ > > libc_cv_forced_unwind libc_cv_c_cleanup = yes = yes > > libc_cv_visibility_attribute=yes libc_cv_broken_visibility_attribute = no > > I surely hope the embedded spaces and lost hyphens are a result of language > translation. Most likely a 'helpful' email program. The extra spaces are a novelty, but I know Outlook is a terrible offender when it comes to capitalising sentences for you, and for replacing paired dashes with em-dash (which then gets converted to a single-dash for ASCII encoding). Maybe it's useful for business use, but it's awful for technical content. signature.asc Description: This is a digitally signed message part -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: IFS6.7-Sec6.22 -- No +h for shell when running tests?
On Wed, 2010-12-08 at 14:55 -0700, Dax Mickelson wrote: > It seems we've been so careful to ensure no hashing occurs so far in > the build that I'm curious as to why the test 'su-tools nobody > -s /bin/bash -c "make RUN_EXPENSIVE_TEST=yes check" doesn't have some > sort of '+h' switch too. > > Obviously I need serious schooling. :-P Because the point of avoiding hashing is to ensure that when a new program is installed into /bin or /usr/bin, it will immediately be used in preference to one in /tools/bin. Otherwise, it will remember that since 'ls' was '/tools/bin/ls' last time, it should keep using that one even though '/bin/ls' has appeared and is higher up the PATH. In this case, 1) a new shell is being started to run a single command before exiting, and 2) nothing is being installed during this step. For both reasons, hashing is irrelevant. Simon. signature.asc Description: This is a digitally signed message part -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: Stumbling across the finish line
On Sun, 2010-12-12 at 01:22 -0500, Neal Murphy wrote: > If that isn't it, it may be that LFS detects alien usage of swap (by Ubuntu). > The warning may be related to suspending/hibernating Ubuntu. Along those lines, mounting a partition simultaneously in both a VM and it's host is a *great* way to get filesystem corruption. signature.asc Description: This is a digitally signed message part -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: Kernel module programming
On Thu, 2010-12-16 at 12:57 -0700, Daryl Lee wrote: > The rest, I have no clue about. Is there someone on this list who can point > me in the right direction? Is there a better forum to raise the question? Since you're working with the Linux Kernel Programming Guide book, the obvious answer would be to contact the author(s) of that book - one would assume they know more than most people about the subject, and has an interest in spreading that knowledge. Simon. signature.asc Description: This is a digitally signed message part -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: tar-1.25 got a tarball file lager than the dir
On Thu, 2010-12-23 at 07:44 -0600, William Immendorf wrote: > On Thu, Dec 23, 2010 at 12:38 AM, xinglp wrote: > > sorry about my Egnrish ;-) > ... what did I tell you about the fact that this thread DOES NOT > BELONG ON THIS LIST!!! No need to be rude about it... the world isn't going to end just because someone asked a question on the wrong mailing list... Simon. signature.asc Description: This is a digitally signed message part -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: Copy and Paste
On Thu, 2010-12-30 at 20:32 -0600, Bruce Dubbs wrote: > One of the problems with a boot CD is that you will need to do that > every time you boot. What would be better is to just do a minimal > install of a distro like Ubuntu or Fedora and use that. Doing an actual install seems excessive - just use the Ubuntu or Fedora images off a flash drive. Ubuntu at least (and I assume also Fedora) can persist config changes like that between boots... Simon. signature.asc Description: This is a digitally signed message part -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: Udev-165 (apparently) System Crash
On Wed, 2011-01-05 at 20:12 -0600, al...@verizon.net wrote: > QUESTION > Is there a simple way of analyzing the crash, maybe with help of > the specialists here, as opposed to formally reporting it? Never tried, myself, but udevd supports a couple of options that might help, --debug and --debug-trace. You might try adding those to the command in /etc/rc.d/init.d/udev. Also, if you mount the partition from another system, you can look at the logs in /var/log, and see if anything looks relevant. Finally, if the system is completely frozen, it may be hardware related - either an actual hardware problem, or broken software talking to hardware. So, you could try removing/disconnecting any unnecessary devices, and see if the problem goes away. Simon. signature.asc Description: This is a digitally signed message part -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: Udev-165 (apparently) System Crash
On Thu, 2011-01-06 at 18:36 -0600, al...@verizon.net wrote: > The important thing for me is you appear to support me, even if > indirectly, in my attempt to avoid kdump in my troubleshoot, if at all > possible :) Happy to help. Simon. signature.asc Description: This is a digitally signed message part -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
RE: question about gcc
On Mon, 2011-01-10 at 10:14 -0600, mike cutie and maia wrote: > the book didn't tell me to put them in the source dir You didn't say which version of the book you're using, but the development version *does* actually tell you that... From: http://www.linuxfromscratch.org/lfs/view/development/chapter05/gcc-pass1.html "GCC now requires the GMP, MPFR and MPC packages. As these packages may not be included in your host distribution, they will be built with GCC. Unpack each package into the GCC source directory and rename the resulting directories so the GCC build procedures will automatically use them:" Simon. signature.asc Description: This is a digitally signed message part -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: questions about modifying the temp tool chain
On Wed, 2011-01-12 at 09:15 -0500, Neal Murphy wrote: > While the Book may be technically correct, it isn't necessarily friendly. To > be sure, I'm not suggesting the book be rewritten or reformatted just to > coddle people through the entire process. But perhaps a checkbox that shows > or > hides the 'training wheels' would suffice. Conversely, the downside to making it *too* easy is that people think they can turn their brain off and just copy-paste their way from start to finish. And while it may sound elitist, LFS isn't for those people - it needs a significant amount of knowledge and skill, and an ongoing willingness to invest time and effort in learning. I'm all for making sure the book is clear and unambiguous. But to use your analogy of training wheels, I think they're a bad idea, giving people an unjustified sense of competence because it looks easy. It's *not* easy, and if people are in over their heads, they need to realise that quickly. Simon. signature.asc Description: This is a digitally signed message part -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: questions about modifying the temp tool chain
On Thu, 2011-01-13 at 10:40 -0600, Bruce Dubbs wrote: > I'm glad gcc comes early because that's where users seem to misapply the > instructions a lot. Yeah, it's that bit with extracting the other packages into the GCC directory - it throws both the people who aren't paying attention to the general instructions, and the people overthinking things. That's why I was advocating that bit that went into the development book - it's not just a reminder for the former, it's about explaining the reasoning for the latter. Simon. signature.asc Description: This is a digitally signed message part -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: Ruminations on Udev, null and console
On Tue, 2011-02-22 at 22:38 -0500, Neal Murphy wrote: > [rant] Digressing a little, I have a bone to pick with Kay Seivers and Greg K- > H. Their attitude is that only symlinks should be used in /dev and created > from udev rules (for the most part), and those symlinks may only be named > with > character sequences that are incomprehensible to ordinary humans.[/rant] Are you unaware then, that udev provides the /dev/disks/by-label directory, which contains volume labels as symlinks to the kernel-named devices? For example, my fstab file doesn't reference /dev/sda6 for the /home partition - it uses /dev/disk/by-label/home instead. Likewise, /boot is mounted as /dev/disk/by-label/boot and the swap partition as /dev/disk/by-label/swap. Simon. signature.asc Description: This is a digitally signed message part -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: Ruminations on Udev, null and console
On Tue, 2011-02-22 at 18:44 -0600, al...@verizon.net wrote: > Unaware, many people still carry a lot of old stuff (including > "console" and "null") on their "metal" /dev. Not accidental. You should re-read "Preparing Virtual Kernel File Systems" in the LFS book, specifically section 6.2.1 in the current development version, which I quote here for reference: 6.2.1. Creating Initial Device Nodes When the kernel boots the system, it requires the presence of a few device nodes, in particular the console and null devices. The device nodes will be created on the hard disk so that they are available before udevd has been started, and additionally when Linux is started with init=/bin/bash. Create the devices by running the following commands: mknod -m 600 $LFS/dev/console c 5 1 mknod -m 666 $LFS/dev/null c 1 3 In short, those nodes are supposed to be there in /dev on the root partition. If you don't have them, you've missed a step in the build. Simon. signature.asc Description: This is a digitally signed message part -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: Ruminations on Udev, null and console
On Wed, 2011-02-23 at 14:41 -0500, Neal Murphy wrote: > On Wednesday 23 February 2011 04:54:10 Simon Geard wrote: > > Are you unaware then, that udev provides the /dev/disks/by-label > > directory, which contains volume labels as symlinks to the kernel-named > > devices? For example, my fstab file doesn't reference /dev/sda6 for > > the /home partition - it uses /dev/disk/by-label/home instead. > > > > Likewise, /boot is mounted as /dev/disk/by-label/boot and the swap > > partition as /dev/disk/by-label/swap. > > You must be superhuman, able to memorize names like 485cc6bb- > ef7f-44ca-91c4-55d884c4083e, 4e23b7e0-af66-45b3-b0c2-a5af31c14c1b, > 5e34b5db-1fa8-49e7-9250-fb95ea301579, 73B115E331E98B57, > 74fbc4cd-784e-49fc-8d24-ca51623ba15c, 965800A05800816B, bbde5991- > c72e-424f-97eb-00acfba10aff, f3ab5997-5ecd-4486-ab86-4fd172d714ea, which > parritions they are and which HDs they belong to. Huh? Who said anything about names like that? Sure, you can mount by uuid if you really want, but I said by *label*. It doesn't seem particularly superhuman to remember that /home is found at /dev/disks/by-label/home. Simon. signature.asc Description: This is a digitally signed message part -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: Ruminations on Udev, null and console
On Wed, 2011-02-23 at 11:24 -0600, Bruce Dubbs wrote: > What am I missing? My latest only has > > $ ls /dev/disk > by-id by-path by-uuid Odd - I see those, and by-label too. My main desktop is a little old now (about ten months old, running udev 151), but I can't see them dropping that feature. I'll have to try on an SVN build... > What do you get when you do something like 'df -h'? That shows the real device, /dev/sda1, so I admit the symlink approach isn't perfect. I mention it mostly in reply to Neal's assertion that "symlinks may only be named with character sequences that are incomprehensible to ordinary humans". Using the human-assigned volume labels seems fairly friendly to me... > Also, I like to see how my disk is organized. I will do 'fdisk -l > /dev/sda' How would you do something like this? Given that I only have the one disk, it's not too hard to remember that it's /dev/sda. That said, Gnome Disk Utility provides a very good GUI for that sort of thing - browse all storage devices, show partitioning, format voumes, etc... Simon. signature.asc Description: This is a digitally signed message part -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: Ruminations on Udev, null and console
On Wed, 2011-02-23 at 12:24 -0600, al...@verizon.net wrote: > BTW, there is a [rant][/rant] I'm skipping, although the idea > is not the first time to come up, where people "grow" from LFS > to BLFS and become confused and/or _misunderstood_ about "falling > behind" on the latest and greatest developments in LFS > (my LFS system was built early 2005). I'm guessing you don't run a 'modern' desktop then? It's pretty much impossible to build and run a modern Gnome desktop without using udev - mostly because udev these days isn't just about populating /dev nodes, it's also how you identify what hardware is present. Simon. signature.asc Description: This is a digitally signed message part -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: Ruminations on Udev, null and console
On Fri, 2011-02-25 at 22:11 +1300, Simon Geard wrote: > On Wed, 2011-02-23 at 11:24 -0600, Bruce Dubbs wrote: > > What am I missing? My latest only has > > > > $ ls /dev/disk > > by-id by-path by-uuid > > Odd - I see those, and by-label too. My main desktop is a little old > now (about ten months old, running udev 151), but I can't see them > dropping that feature. I'll have to try on an SVN build... Just checked, the by-label directory is also present on my SVN build running udev 166. I wonder if this is provided by the 'extras' which we disable in LFS due to dependencies, but which are present on my system due to reinstalling udev during BLFS. Simon. signature.asc Description: This is a digitally signed message part -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: Ruminations on Udev, null and console
On Fri, 2011-02-25 at 13:21 -0600, al...@verizon.net wrote: > 2. _Strictly_ for the record and completeness: > I do no use Gnome. Never have. > I use KDE (3.5.10) and Fluxbox (1.1.1). > However, IMHO this is absolutely irrelevant as far as the thread > subject is concerned. > Unfortunately, the above admission might trigger another unrelated > debate :( I'm resigned to it, though; you win a few lose a few. Not meaning to start an argument on that basis. But by 'modern' desktops, I meant recent versions of Gnome or KDE, compared to more lightweight setups. As to the relevance to the thread, I know I didn't make that clear, but my thinking was that device names really aren't very important to someone using one of those desktops. My DVD drive may be called 'sr0', but more importantly, it's called "CD/DVD drive" on my desktop. And if I plug in a USB drive, I really have no idea what the device node is called - all that matters is that it's automatically mounted to somewhere under /media and a file manager window opened on it. For the same reason, recent kernel announcements about changing the network device naming don't exactly bother me - as far as I'm concerned, I have two network devices, called "Wired" and "Wireless" in Network Manager. The kernel calls them eth0 and wlan0, but if a patch were to rename them 'frog' and 'fish', I'd be unlikely to notice. As to the rest, it may be I've misunderstood your original post - the confusion was over the claim that people are "unaware" they they have those nodes on the "metal" /dev. Given these nodes are explicitly created with a clear explanation as to why, it seemed to me that you must have overlooked this section in the book. Simon. signature.asc Description: This is a digitally signed message part -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: Ruminations on Udev, null and console
On Fri, 2011-02-25 at 22:12 -0600, Bruce Dubbs wrote: > I suspect you have to add -L volume-label to mke2fs (or other partition > formatting tool) for at least one partition for by-label to be created. Ah, yes, that makes sense - if the directory is created by the presence of symlinks within it, then it won't be present unless you have labeled volumes present. My root partitions don't have labels, for example, and so aren't listed... For interest, the directory looks something like: $ ll /dev/disk/by-label/ total 0 lrwxrwxrwx 1 root root 10 Feb 28 18:07 boot -> ../../sda1 lrwxrwxrwx 1 root root 10 Feb 28 18:07 home -> ../../sda6 lrwxrwxrwx 1 root root 10 Feb 28 18:07 oldboot -> ../../sdb1 lrwxrwxrwx 1 root root 10 Feb 28 18:07 oldhome -> ../../sdb7 lrwxrwxrwx 1 root root 10 Feb 28 18:07 swap -> ../../sda2 Simon. signature.asc Description: This is a digitally signed message part -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: gcc test failures
On Sun, 2011-03-13 at 20:48 -0700, Alex Bosworth wrote: > I were making use of "-j2" make option, I wonder if would make a > difference if I set it to "-j1" and lower the priority for make > (perhaps set it to +20 ?) and see if that helps to keep the heat > low !? Changing the priority doesn't really do anything to reduce load on the machine - it just means that those processes will be treated as less important when competing for CPU time. If there's nothing else running, there's no competition, so they'll still get all the CPU time they can. However, changing from -j2 to -j1 (or removing it entirely) may help, since only one process will be wanting CPU time at once. Simon. signature.asc Description: This is a digitally signed message part -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: Incompatibility of udev and /usr
On Wed, 2011-04-13 at 21:04 -0500, Mike McCarty wrote: > There is an incompatibility with using udev and /usr being a > separate file system, which users of LFS need to be aware of. > It is presently not possible, in general, to use udev and have > /usr be a separately mounted file system. This is something to > consider when planning the layout of the disc drives. The current > implementation of udev is incompatible with the File System Hierarchy > Standard. Yes, there's been a bit of discussion of this among the distributions of late. Here's a couple of the links I've read on the subject... http://freedesktop.org/wiki/Software/systemd/separate-usr-is-broken http://lists.debian.org/debian-devel/2009/05/msg00075.html http://lists.debian.org/debian-devel/2011/01/msg00152.html While not universal, there seems to be a growing feeling that having a separate /usr partition serves no useful purpose these days. The third of those links gives a pretty good summary of that viewpoint. As to compatibility with the FHS, distros seem inclined to ignore the spec, on the basis that it's not being updated, and no longer reflects reality (e.g no mention of /sys). Another discussion on that subject: http://lists.debian.org/debian-devel/2009/02/msg00395.html Simon. signature.asc Description: This is a digitally signed message part -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: Incompatibility of udev and /usr
On Thu, 2011-04-14 at 19:55 +1200, Simon Geard wrote: > While not universal, there seems to be a growing feeling that having a > separate /usr partition serves no useful purpose these days. The third > of those links gives a pretty good summary of that viewpoint. Actually, some of them are going even further, advocating the removal of /usr entirely, using --prefix=/ for everything. Not seeing much support for that, though... Simon. signature.asc Description: This is a digitally signed message part -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: Incompatibility of udev and /usr
On Thu, 2011-04-14 at 09:27 -0500, Mike McCarty wrote: > Simon Geard wrote: > > [...] > > > While not universal, there seems to be a growing feeling that having a > > separate /usr partition serves no useful purpose these days. The third > > of those links gives a pretty good summary of that viewpoint. > > Well, I also have read this argument, and it cuts no water > with me. Fair enough. Having a separate /usr partition makes little sense for me given my setup (no NFS, no encrypted FS, etc) but it certainly sounds like some people have uses for it. The Debian list does raise an interesting question of how package management handles an NFS-mounted /usr, though. While solvable, it's hard to see any distro doing a good job of that out of the box... > > As to compatibility with the FHS, distros seem inclined to ignore the > > spec, on the basis that it's not being updated, and no longer reflects > > reality (e.g no mention of /sys). Another discussion on that subject: > > > > http://lists.debian.org/debian-devel/2009/02/msg00395.html > > Interesting. I was unaware of that. Thanks! The subject has come up again recently with respect to the creation of a /run mountpoint (essentially moving /var/run to /) - it's been raised with FHS, but with no response. Fundamentally, I think the LSB / FHS is somewhat redundant - because as a practical matter, the real standard is whatever the big distros do. If Debian, Ubuntu, Fedora and SuSe all decide this new /run thing is a good idea, does it really matter what the standard says on the subject? Simon. signature.asc Description: This is a digitally signed message part -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: Incompatibility of udev and /usr
On Fri, 2011-04-15 at 10:08 -0500, Bruce Dubbs wrote: > It seems to me that the reason for an NFS-mounted /usr is to be able to > update all systems at once. At one time, it was to conserve disk space > too, but the cost of disk is so cheap now, it has become a non-factor. Somewhat redundant for update purposes too, I think. We run a cluster of Red Hat boxes at work, and just use cfengine to push updates out to them as needed. That's actually preferable for us, since it's easy to temporarily prevent updates to an individual machine if you want to test something new without affecting anyone else. > That seems to be a practical approach. I do read the LSB mailing list > and there are some discussions about being a trailing standard and not a > leading standard, but I have lowered my opinion of LSB overall. It > would appear to me that the LSB exists primarily to allow proprietary > software to be installed on multiple distributions. I wonder how many such developers actually care about supporting multiple distros? I guess it matters for desktop-oriented software, but speaking as an enterprise-app developer, it's not at all important to us. We certify whatever our customers want, which when it comes to Linux, means Red Hat. > On the other hand, I see that Ubuntu and other distributions are using > upstart instead of sysvinit. Perhaps I'm set in my ways, but I don't > really see an advantage. Perhaps it speeds up the boot process when you > have hundreds of drivers built as modules and need to test them all each > time you boot, but that is generally not an issue for LFS users. Ubuntu is using upstart, Fedora is now using systemd, not sure what others are doing. But from what I understand, the concept for both is similar - to replace sysvinit with something that a) deals with dependencies, b) can start stuff in parallel, and c) provides consistency across distros to aid 3rd-party developers. Although until the distros agree on *one* sysvinit replacement, the consistency argument is somewhat moot. I agree, not all that necessary on a base LFS system. But I've read a bit about systemd, which makes use of newer Linux features like cgroups and namespaces, and it's interesting. Maybe not essential to me, but interesting. Simon. signature.asc Description: This is a digitally signed message part -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: echo $LFS set
On Sat, 2011-04-16 at 15:09 +0530, janu mam wrote: > 6.4. Entering the Chroot Environment (page no-79) > - > see third paragaph > >From this point on, there is no need to use the LFS variable anymore, > because all work will be restricted to the LFS > file system. This is because the Bash shell is told that $LFS is now > the root (/) directory. > > root[ ~ [#chroot "$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 > result is I have no name!:/# this is right according to book > > I have no name!:/#echo $LFS > no (any) output > this is continuing till end of chapter 6,you are saying that it should > set (echo $LFS)) You're misunderstanding that third paragraph. Once you enter the chroot environment, the $LFS directory becomes the root directory for all further commands, and there's no longer any need for the variable. While in that environment, the variable shouldn't be set. Simon. signature.asc Description: This is a digitally signed message part -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: Incompatibility of udev and /usr
On Sat, 2011-04-16 at 13:29 -0500, DJ Lucas wrote: > On 04/13/2011 09:04 PM, Mike McCarty wrote: > > There is an incompatibility with using udev and /usr being a > > separate file system, which users of LFS need to be aware of. > > It is presently not possible, in general, to use udev and have > > /usr be a separately mounted file system. This is something to > > consider when planning the layout of the disc drives. The current > > implementation of udev is incompatible with the File System Hierarchy > > Standard. > > This is incorrect. udev is perfectly FHS compliant as installed in LFS > and provides only minimal challenges to make it so in BLFS. > > dj [ glibc-build ]$ ldd /lib/udev/* 2>/dev/null | grep usr > libusb-0.1.so.4 => /usr/lib/libusb-0.1.so.4 (0x7f1f8534a000) > libusb-1.0.so.0 => /usr/lib/libusb-1.0.so.0 (0x7f1f849af000) > libusb-0.1.so.4 => /usr/lib/libusb-0.1.so.4 (0x7f37725be000) > libusb-1.0.so.0 => /usr/lib/libusb-1.0.so.0 (0x7f3771e2b000) > libglib-2.0.so.0 => /usr/lib/libglib-2.0.so.0 (0x7fcb0a3ce000) > libdevmapper.so.1.02 => /usr/lib/libdevmapper.so.1.02 > (0x7fbc22d53000) > libglib-2.0.so.0 => /usr/lib/libglib-2.0.so.0 (0x7fbc22a6c000) > libglib-2.0.so.0 => /usr/lib/libglib-2.0.so.0 (0x7f5455587000) > libparted.so.0 => /usr/lib/libparted.so.0 (0x7f54550fa000) > libdevmapper.so.1.02 => /usr/lib/libdevmapper.so.1.02 > (0x7f545452d000) > libatasmart.so.4 => /usr/lib/libatasmart.so.4 (0x7f4d23ac3000) > libglib-2.0.so.0 => /usr/lib/libglib-2.0.so.0 (0x7f2631959000) > dj [ glibc-build ]$ My understanding is that the problem isn't with the location of libraries - it's with the location of data under /usr/share. Stuff like the pci.ids and usb.ids files, which are apparently required for some of the udev rules. Those files could presumably be moved to somewhere under /, but there's no obvious place to put them, no /share directory... Simonl signature.asc Description: This is a digitally signed message part -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: Incompatibility of udev and /usr
On Sun, 2011-04-17 at 00:26 -0500, DJ Lucas wrote: > Ahh...lightbulb. This is why we currently have the udev-retry in our > bootscripts. Are the ids files accessed directly by external programs or > by the utility libraries/programs? Provide a common library to access > the files (if not done already) and install into the root and place the > ids files into the libexecdir, problem solved. I don't pretend to follow the details - I'm going mostly by the statements made by systemd-developer Lennart Poettering on the subject, in response to some of the more recent arguments on the subject: http://freedesktop.org/wiki/Software/systemd/separate-usr-is-broken Looking on my own system, examples of offending rules seem to be things like 78-sound-card.rules, which uses the *descriptions* associated with USB device IDs to classify whether a device is a headphone, microphone, speaker, etc. A bit hacky, perhaps (and that file admits as much), but I suppose it's the only way to deal with some hardware. As to why that's an issue at boot time, I *assume* that if the id files aren't present when the hardware is detected on boot, the device won't be correctly recognised. It won't break the boot, but it will cause problems when the user tries using their USB speakers or whatever. Simon. signature.asc Description: This is a digitally signed message part -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: Incompatibility of udev and /usr
On Sun, 2011-04-17 at 00:56 -0400, Neal Murphy wrote:
> Gently illuminating a point that does not necessarily invalidate your whole
> argument: pci.ids and the IDs used in drivers are not necessarily in sync.
> Udev should depend solely on the IDs compiled into the drivers; it can easily
> obtain this info from the driver itself. Udev v165 seems to care about
> pci.ids
> only in config.h*, configure, and configure.ac; pci.ids doesn't seem to be
> referenced anywhere else in the source. Neither 'pci.ids' nor #defines using
> the file seem to be used in any code, which makes sense since a driver should
> not claim any device that does not match its compiled-in IDs. I would expect
> usb.ids to be treated similarly. I wouldn't expect great changes within
> several releases of udev.
As I understand it, it's not about drivers; at least, not the kernel
variety. The only example I can see in my installation is the file:
/lib/udev/rules.d/78-sound-card.rules
And from what I can see, they're using the descriptions from the usb.ids
to recognise arbitrary devices as being USB-connected speakers,
headsets, microphones, etc. Rules like:
---
# Matching on the model strings is a bit ugly, I admit
ENV{ID_MODEL}=="*[Ss]peaker*", ENV{SOUND_FORM_FACTOR}="speaker",
GOTO="sound_end"
ENV{ID_MODEL_FROM_DATABASE}=="*[Ss]peaker*",
ENV{SOUND_FORM_FACTOR}="speaker", GOTO="sound_end"
ENV{ID_MODEL}=="*[Hh]eadphone*", ENV{SOUND_FORM_FACTOR}="headphone",
GOTO="sound_end"
ENV{ID_MODEL_FROM_DATABASE}=="*[Hh]eadphone*",
ENV{SOUND_FORM_FACTOR}="headphone", GOTO="sound_end"
---
I guess the issue is that a lot of USB devices aren't handled by kernel
drivers matching on known device ids - they're handled in userspace. And
udev is the bit providing information so that userspace can tell which
devices it cares about.
Simon.
signature.asc
Description: This is a digitally signed message part
--
http://linuxfromscratch.org/mailman/listinfo/lfs-support
FAQ: http://www.linuxfromscratch.org/lfs/faq.html
Unsubscribe: See the above information page
Re: Incompatibility of udev and /usr
On Sun, 2011-04-17 at 12:33 +0100, Andrew Benton wrote: > I think that the idea of doing away with /usr and just installing > everything into / is interesting as it would simplify the structure of > the directories. I also think that making the change would be like > poking yourself in the eye with a stick; painful and with no obvious > benefit. There would be a million things that would need patching > because they were looking for /usr/bin/perl or whatever. With no strong > reason to make the change it's difficult to summon the energy to do the > work. Yes, that was my thought too. It *would* simplify the directory structures, and in most cases wouldn't be too hard to do, just passing --prefix=/ to everything. But really, I wouldn't be gaining anything for the effort... Simon. signature.asc Description: This is a digitally signed message part -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
Re: Incompatibility of udev and /usr
On Sun, 2011-04-17 at 13:41 -0500, DJ Lucas wrote: > On 04/17/2011 01:31 PM, DJ Lucas wrote: > > > > > Anyway, udev starts 4th in the startup scripts, it runs across a uevent > > that uses a rule found in /usr, and it fails to create the device node. > > Errit creates the device node, but fails to run whatever program in > /usr that is required to make it work with the system correctly. Yes, that sounds right. In the audio example, the device will be detected and a /dev node created, but because the rule that tags it as a "speaker" didn't work, it then won't be recognised by some userspace application (pulse-audio / gstreamer / alsa?), and the user won't be able to use their USB-connected speakers until they unplug and reconnect them... Simon. signature.asc Description: This is a digitally signed message part -- http://linuxfromscratch.org/mailman/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/lfs/faq.html Unsubscribe: See the above information page
