Does GCC compile option ---with-local-prefix still work?
Hi all, The configuration switch --with-local-prefix seems to be redundant in GCC package. In the LFS book, it explains as follows: The purpose of this switch is to remove /usr/local/include from gcc's include search patch. But after searching through the whole config script, it seems this option is not processed. Is there something I missed? Regards, William Zhou -- http://linuxfromscratch.org/mailman/listinfo/lfs-dev FAQ: http://www.linuxfromscratch.org/faq/ Unsubscribe: See the above information page
Re: Binutils-2.16.1-pass1
# Adjusting the Toolchain # Tcl-8.4.12 # Expect-5.43.0 # DejaGNU-1.4.4 # GCC-4.0.2 - Pass 2 # Binutils-2.16.1 - Pass 2 Between toolchain adjustment and the 2nd pass of binutils, there are five packages including binutils itself get complied. If the ld were not replaced, it would search libraries from /lib:/usr/lib:/usr/local/lib instead of /tools/lib. I am not very sure what will happen if we don't do this adjustment, but it certainly makes gcc pass2 and binutils pass2 not as "clean" as possbile. William Bob Winckelmans wrote: 5.3. Binutils-2.16.1 - Pass 1 At the bottom, the Next, prepare the linker for the “Adjusting” phase later on: make -C ld clean make -C ld LIB_PATH=/tools/lib cp -v ld/ld-new /tools/bin is no longer necessary I believe, Kind regards, Bob -- http://linuxfromscratch.org/mailman/listinfo/lfs-dev FAQ: http://www.linuxfromscratch.org/faq/ Unsubscribe: See the above information page
Re: Does GCC compile option ---with-local-prefix still work?
Thanks for the hint, I will have a look at it. William Zhou Dan Nicholson wrote: On 2/21/06, William Zhou <[EMAIL PROTECTED]> wrote: The configuration switch --with-local-prefix seems to be redundant in GCC package. But after searching through the whole config script, it seems this option is not processed. Look at gcc/configure and gcc/Makefile.in. It's there. As William said, running `make configure-host' should run all the configure scripts from the subdirectories immediately. Actually, that doesn't seem to be totally true, but you get the idea. -- Dan -- http://linuxfromscratch.org/mailman/listinfo/lfs-dev FAQ: http://www.linuxfromscratch.org/faq/ Unsubscribe: See the above information page
Re: /tools/bin/env: no such file or directory
Hi, GCC is installed beforehand and the specs file should be something like /tools/lib/gcc/.../specs instead of /usr/lib/gcc/***/specs. Make sure you have a correct PATH set. William Zhou Dominic Ringuet wrote: Simply reporting so nobody else wastes time on this. May be it could be added to the FAQ that describes this problem. if '/usr/lib/gcc/i686-pc-linux-gnu/specs' exists on the host system, the binaries compiled in chapter 5, even with the specs patch properly applied in gcc-pass2, will be linked to '/lib/ld-linux.so.2' and chroot will fail at the beginning of chapter 6 giving: '/tools/bin/env: no such file or directory' Simply ensure no host specs overrides the build in chapter 5. Dominic. -- http://linuxfromscratch.org/mailman/listinfo/lfs-dev FAQ: http://www.linuxfromscratch.org/faq/ Unsubscribe: See the above information page
udev branch. package udev.
Hello, In the udev package, is the following sentence still true for kernel linux-2.6.16.1? "Create some rules that work around broken sysfs attribute creation timing in linux-2.6.15:" BTW, in the section Important, the last sentence got a word misspelled. It is "to aid" not "to aide". Thanks for the great work. William Zhou -- http://linuxfromscratch.org/mailman/listinfo/lfs-dev FAQ: http://www.linuxfromscratch.org/faq/ Unsubscribe: See the above information page
Re: Adding newer headers to llh (Was Re: merging udev_update branch)
Jim Gifford wrote: Another option here is to use the headers package I've been working with a lot of people. It compiles a base LFS and CLFS with no issues at all.http://ftp.jg555.com/headers/linux-headers-2.6.16.2.tar.bz2, or roll your own by using http://ftp.jg555.com/headers/headers. Hi, In the beginning of the script, it mentioned the process of copying over the asm-generic directory, why not includes it in the script? With the generated headers, I successful built a CLFS x86_64 pure without any problem. Thanks for the great work. William -- http://linuxfromscratch.org/mailman/listinfo/lfs-dev FAQ: http://www.linuxfromscratch.org/faq/ Unsubscribe: See the above information page
Re: udev branch. package udev.
Alexander E. Patrakov wrote:
Archaic wrote:
On Wed, Apr 05, 2006 at 01:57:12PM +0100, William Zhou wrote:
"Create some rules that work around broken sysfs attribute creation
timing in linux-2.6.15:"
This is still in. Either it needs to be pulled, or the version needs to
refer to the entity. Alex?
It needs to be pulled, but there's another breakage reported:
http://marc.theaimsgroup.com/?l=linux-hotplug-devel&m=114460333128844&w=2
Apply this kernel patch:
http://kernel.org/git/?p=linux/kernel/git/torvalds/linux-2.6.git;a=blobdiff;h=2731570eba5b35a21c311dd587057c39805082f1;hp=dfb62998866ae2e298139164a85ec0757b7f3fc7;hb=9469d458b90bfb9117cbb488cfa645d94c3921b1;f=net/core/dev.c
[snip]
Rediffed against linux-2.6.16.5.
--- orig/net/core/dev.c
+++ new/net/core/dev.c
@@ -2932,11 +2932,11 @@
switch(dev->reg_state) {
case NETREG_REGISTERING:
+ dev->reg_state = NETREG_REGISTERED;
err = netdev_register_sysfs(dev);
if (err)
printk(KERN_ERR "%s: failed sysfs registration
(%d)\n",
dev->name, err);
- dev->reg_state = NETREG_REGISTERED;
break;
case NETREG_UNREGISTERING:
--
http://linuxfromscratch.org/mailman/listinfo/lfs-dev
FAQ: http://www.linuxfromscratch.org/faq/
Unsubscribe: See the above information page
Re: My take on directory creation
Randy McMurchy wrote: [snip] And, BTW, mkdir -p is wrong. The install command with a -mXXX parameter is a much, much better example of how to do things correctly. The install command is only needed if you need to specify the owner and the permission in one command, which is usually used in package installation. In LFS, the user is root:root and the umask is defined as 022. Everything is ready. IMHO, using install is unnecessary for a directory set up. William Zhou -- http://linuxfromscratch.org/mailman/listinfo/lfs-dev FAQ: http://www.linuxfromscratch.org/faq/ Unsubscribe: See the above information page
Re: Reinstalling Readline
M.Canales.es wrote: El Domingo, 16 de Abril de 2006 01:08, M.Canales.es escribi�: A similar issue when reinstalling Module-Init-Tools. http://wiki.linuxfromscratch.org/lfs/ticket/1771 execute "make moveold" before "make install" does the job. William -- http://linuxfromscratch.org/mailman/listinfo/lfs-dev FAQ: http://www.linuxfromscratch.org/faq/ Unsubscribe: See the above information page
Re: LFS vs BLFS -- udev rules
Jim Gifford wrote: In CLFS, we create all the users and groups, this will solve all issues if LFS will follow. It will also simplify the BLFS instructions to adding users to the appropriate groups. It's a plus for all, eliminates so of tedious work the BLFS has to do managing users/groups and gives everyone a full compliment of users/groups to start with. I prefer the CLFS way and I don't have to worry about the user/group management any more. It is nice to include some common software users in the LFS/CLFS, like svn, apache, samba, clamav, distccd. etc. William Zhou -- http://linuxfromscratch.org/mailman/listinfo/lfs-dev FAQ: http://www.linuxfromscratch.org/faq/ Unsubscribe: See the above information page
Re: LFS vs BLFS -- udev rules
Archaic wrote: On Fri, Apr 21, 2006 at 08:55:25PM +0100, William Zhou wrote: I prefer the CLFS way and I don't have to worry about the user/group management any more. Which is precisely why I don't like it. You should have to worry about it if the goal is education. Then you can devise any number of methods to suit your desires/needs and BLFS would be the guide to that. Fine. Here is an another suggestion. We started a long list, but we comment out those ones that are not used in LFS. Whenever BLFS needs it, it can change it using sed or create one using useradd. For those who don't need the education, simply change the file to whatever they need. For those who want the education, they can still follow the book and create the users/groups as the build goes. No matter which way they choose, we have an unified UID/GID combination. William Zhou -- http://linuxfromscratch.org/mailman/listinfo/lfs-dev FAQ: http://www.linuxfromscratch.org/faq/ Unsubscribe: See the above information page
Re: LFS vs BLFS -- udev rules
Jim Gifford wrote: We should be providing all the users and groups and let the people choose what they want to remove. Instead of just giving them the bare minimum. We need to provide a fully functional system, not a half-baked one. Exactly. We had a same point. :). William Zhou -- http://linuxfromscratch.org/mailman/listinfo/lfs-dev FAQ: http://www.linuxfromscratch.org/faq/ Unsubscribe: See the above information page
Re: New server details
M.Canales.es wrote: El S醔ado, 29 de Abril de 2006 18:40, Gerard Beekmans escribi�: Intel Pentium D @ 3.0 GHz 2 GB RAM Dual 160 GB SATA hard drives A very nice beast ;-) With Xen, it runs Windows too. (Though we won't do it ,,do we?) William -- http://linuxfromscratch.org/mailman/listinfo/lfs-dev FAQ: http://www.linuxfromscratch.org/faq/ Unsubscribe: See the above information page
Re: Glibc will fail to compile in chapter 6
Ismael Luceno wrote: I noticed that in the chapter 6, glibc-2.3.6 will fail to compile, because the gcc specs patch is preventing glibc from including the kernel headers at /usr/include, adding the option --with-headers should solve the problem. You must have applied the wrong patch then. William -- http://linuxfromscratch.org/mailman/listinfo/lfs-dev FAQ: http://www.linuxfromscratch.org/faq/ Unsubscribe: See the above information page
/var/log/btmp permission problem.
Hi, I got time and read the log files. I found out that sshd keeps complianting about the permission of file /var/log/btmp. The message read as "Excess permission or bad ownership on file /var/log/btmp." After changing to 640, it stops complianting. The LFS book does not specify it clearly and leave it as default(644). Regards, William-- http://linuxfromscratch.org/mailman/listinfo/lfs-dev FAQ: http://www.linuxfromscratch.org/faq/ Unsubscribe: See the above information page
lib lib32 lib64 in LFS 7 x86_64_multilib
http://www.linuxfromscratch.org/lfs/view/cross-lfs/x86_64-64/ Hi, I finished this build several days ago and started BLFS. However, the /lib is actually for lib32 instead of lib64. I believe the system tends to run in 64 bit. So all the libraries in BLFS have to go to /lib64. This is not as easy as it sounds. By specifying --libdir=/usr/lib64 solves majority of the problem. But there are still plenty of them which hardcodes the path into the program. To make this worse, library search is another problem because most of them searches /usr/lib and /lib. I don't want to sed all of them. It takes time. Why does the book use /lib32+/lib(symlink to lib64) instead? Thanks very much. William -- http://linuxfromscratch.org/mailman/listinfo/lfs-dev FAQ: http://www.linuxfromscratch.org/faq/ Unsubscribe: See the above information page
User lfs is more than optional.
I have been using LFS for more than a year's time and it is great. One of my friend started LFS several days ago and got an error when adjusting the toolchain( 5.7 ). The problem was that the gcc specs path was pointed to the host's one. It took me me a while to figure out that he ignored the creation of user lfs and thus the ~/.bashrc is not created. The PATH enviourment does not even includes /tools/bin. Most of us follows the book's recommendation and never had this problem. But I believe the creation of LFS should be stated to be mandatory, or at least make this clear. 1. GCC specs path. Different gcc yields different specs files. (obviously). 2. set +h search for commands every time 3. Get rid of useless environment variables using the method in ~/.bash_profile. ( exec env -i HOME=$HOME TERM=$TERM PS1='...' /bin/bash ) 4. LFS environment variable. (Not so important). Regards. William -- http://linuxfromscratch.org/mailman/listinfo/lfs-dev FAQ: http://www.linuxfromscratch.org/faq/ Unsubscribe: See the above information page
User lfs is more than optional.
> > And lose the excellent lesson your friend learned from this? :) We > cannot be liable for people ignoring large chunks of the book. I agreed with you. In 6.5, Creating directories. You don't have to have the /usr/local/game. If it is only a recommendation, it should not affect the LFS itself. However, this one does and might cause some other influence in the later stage. (BTW, How do I reply to a specif thread instead of creating a new one? I tried to figure it out but I can't. Thanks.) William -- http://linuxfromscratch.org/mailman/listinfo/lfs-dev FAQ: http://www.linuxfromscratch.org/faq/ Unsubscribe: See the above information page
