gmane.linux.lfs.support

Re: 8.2. Creating the /etc/fstab File

Bruce Dubbs — 2013-05-21T17:36:12

The statement is still valid for for both ext3 and ext4.  Using 
barrier=1 may also affect disk speed, but I suspect that's more 
theoretical than noticeable.

There are other options that may be useful that we don't mention.  For 
example noatime,discard,data=writeback is useful for ssd drives.

Perhaps a more general statement about mount options would be more 
appropriate than a mention of just barrier.

   -- Bruce

8.2. Creating the /etc/fstab File

Baho Utot — 2013-05-21T16:02:55

I found this error in the book....

It is possible to make the ext3 filesystem reliable across power 
failures for some hard disk types. To do this, add the barrier=1 mount 
option to the appropriate entry in /etc/fstab. To check if the disk 
drive supports this option, run hdparm on the applicable disk drive. For 
example, if:


The ext3 should be ext4 if this is the default for LFS should it not?

Re: Q: re. LFS 7.3 errata

Jeremy Henty — 2013-05-20T19:25:28

Pierre Labastie wrote:

Thanks for clarifying.   I just finished LFS 7.3 chapter  5 and it all
looks good so far!

Regards,

Jeremy Henty

Re: LFS-SVN ext4 partition

Baho Utot — 2013-05-20T10:42:32

OK thanks I will try that

Re: LFS-SVN ext4 partition

Markku Pesonen — 2013-05-20T06:12:37

Unless you are using an initramfs, Linux doesn't know what filesystem
you are using on your root partition. Linux tries to mount it using all
built-in filesystems available, listed in /proc/filesystems, in that
order. If you have ext2 and ext3 built-in, Linux tries those before
using ext4, and it looks something like this:
EXT3-fs (sda3): error: couldn't mount because of unsupported optional
features (240)
EXT2-fs (sda3): error: couldn't mount because of unsupported optional
features (240)
EXT4-fs (sda3): mounted filesystem with ordered data mode. Opts: (null)

I believe this is harmless. If you really want, you can get rid of those
errors by using the rootfstype=ext4 kernel parameter.

Re: LFS-SVN ext4 partition

Baho Utot — 2013-05-20T01:34:35

Yes that is indeed what I am seeing also, I am using SATA drives,  I did 
know about the EXT4_FS mounting EXT2 and EXT3.  Maybe the fix is to drop 
the EXT[2,3]_FS and only use EXT4_FS.


I'll have to research my kernel config, although I am using the one from 
slackware with PAE as I am building i686 right now.   I don't know what 
they have set for EXTx_FS.

I will get to that in the morning.

Re: LFS-SVN ext4 partition

Bruce Dubbs — 2013-05-20T01:11:53

I don't see that right now, but I'm in between systems.  Right now I'm 
using a 3.8.3 kernel with:

# CONFIG_EXT2_FS is not set
CONFIG_EXT3_FS=y
# CONFIG_EXT3_DEFAULTS_TO_ORDERED is not set
CONFIG_EXT3_FS_XATTR=y
CONFIG_EXT3_FS_POSIX_ACL=y
CONFIG_EXT3_FS_SECURITY=y
CONFIG_EXT4_FS=y
CONFIG_EXT4_USE_FOR_EXT23=y
# CONFIG_EXT4_FS_POSIX_ACL is not set
# CONFIG_EXT4_FS_SECURITY is not set
# CONFIG_EXT4_DEBUG is not set

But I use ext2 for /boot, ext3 for /, and ext4 for /mnt/lfs.  I also 
have one partition mounted as reiserfs.

   -- Bruce

Re: LFS-SVN ext4 partition

Ken Moffat — 2013-05-20T00:22:42

 sort of.  ISTR there are config option(s) to mount ext{2,3} using
ext4.  On at least one of my machines I see two such messages, but
the first is for ext3 and the second for ext2 :

[    1.765843] sd 0:0:0:0: [sda] Attached SCSI disk
[    1.792078] EXT3-fs (sda6): error: couldn't mount because of unsupported optional features (240)
[    1.795575] EXT2-fs (sda6): error: couldn't mount because of unsupported optional features (240)
[    1.818698] EXT4-fs (sda6): mounted filesystem with ordered data mode. Opts: (null)
[    1.821899] VFS: Mounted root (ext4 filesystem) readonly on device 8:6.

 I can't remember when this started, and google is useless (loads of
references to converting *to* ext4, or for an obsolete config
option), but at some time the ext4 driver became able to mount ext3
and ext2 - if those were enabled.

ĸen

LFS-SVN ext4 partition

Baho Utot — 2013-05-19T21:22:45

I have always in the past used a ext3 partition on my LFS systems. I 
have changed to ext4 and on boot I get errors that are saying than it 
had problems mounting the partition due to unsupported  options, it has 
EXT3-fs in the first two error messages, the third message says it uses 
EXT4-fs and it then mounts without error.
I belive  it is coming from the kernel as it has a kernal time of [ 
1.568232].

Is this normal for the boot to try mount the root partition with EXT3-fs 
twice before then using EXT4-fs?

Re: Q: re. LFS 7.3 errata

Pierre Labastie — 2013-05-19T17:26:18

Le 19/05/2013 15:31, Jeremy Henty a écrit :


It is "--with-sysroot" and nothing more. That changes the behavior
of ld in some circumstances, which are met when building 'check'.

You may want to read http://article.gmane.org/gmane.linux.lfs.devel/13809
and http://article.gmane.org/gmane.linux.lfs.devel/13812, where
I tried to explain the use of that switch.

Regards
Pierre

Re: Q: re. LFS 7.3 errata

William Harrington — 2013-05-19T13:54:51

On May 19, 2013, at 8:31 AM, Jeremy Henty wrote:


With your host you will be okay if you follow it or not.

If you are using a newer multiarch host, you most likely will not be  
okay.

Bruce stated it doesn't matter what your host is, if you follow the  
errata, you will be okay.
Follow the errata regardless of the host.

Sincerely,

William Harrington-- 
http://linuxfromscratch.org/mailman/listinfo/lfs-support
FAQ: http://www.linuxfromscratch.org/lfs/faq.html
Unsubscribe: See the above information page

Re: Q: re. LFS 7.3 errata

Jeremy Henty — 2013-05-19T13:31:39

Sorry, I am confused.   Do you mean I will be  OK ignoring the erratum
or I will  be OK if I  follow it?  And does  "add --with-sysroot" mean
literally  to add  "--with-sysroot" rather  than "--with-sysroot=$LFS"
(as in pass 1).

Regards

Jeremy Henty

Re: LFS7.3: zlib-1.2.7 download link is not available anymore

Armin K. — 2013-05-18T08:00:42

You can always get all packages from here

http://anduin.linuxfromscratch.org/sources/LFS/lfs-packages/7.3/

or here

http://ftp.lfs-matrix.net/pub/lfs/lfs-packages/7.3/

or any other mirror listed here:

http://www.linuxfromscratch.org/lfs/download.html (end of the page)

Re: LFS7.3: zlib-1.2.7 download link is not availableanymore

Jon Franklin — 2013-05-18T07:53:14

Because you are GFWed.
http://superb-dca2.dl.sourceforge.net/project/libpng/zlib/1.2.7/zlib-1.2.7.tar.bz2
This must be available.
On Sat, May 18, 2013 at 10:41 AM, Chunmeng Zhou <chunmengzhou< at >gmail.com> wrote:

LFS7.3: zlib-1.2.7 download link is not availableanymore

Chunmeng Zhou — 2013-05-18T02:41:39

The zlib 1.2.7 download link (http://www.zlib.net/zlib-1.2.7.tar.bz2) is not
available anymore.

You can download it from sourceforge:
wget http://prdownloads.sourceforge.net/libpng/zlib-1.2.7.tar.bz2

It seems the developer released the latest zlib on Aprial 28th and moved all
the previous version to sourceforge.



Chun

Re: Rename everything lfs/LFS -> unim/UNIM - chapter 6.61. Udev-197, LFS 7.3

Baho Utot — 2013-05-17T21:45:44

You just did it again "Top Posted" a little google.com work will explain 
it to you

The user group names you changed/used are not known in the chroot, any 
way as root it doesn't matter.

Also noticed how I trimmed this post?

Re: Rename everything lfs/LFS -> unim/UNIM - chapter 6.61. Udev-197, LFS 7.3

Sandy Widianto — 2013-05-17T21:29:07

I don't understand about "top post" I just replied it. Now I will continue
for compilations, I want everything goes smoothly. Thank you Mr. Bruce.
Sandy Widianto

On Sat, May 18, 2013 at 4:12 AM, Bruce Dubbs <bruce.dubbs< at >gmail.com> wrote:

Re: Rename everything lfs/LFS -> unim/UNIM - chapter 6.61. Udev-197, LFS 7.3

Bruce Dubbs — 2013-05-17T21:12:37

Inside of chroot, your unim/UNIM constructs are not visible.

No.  There is no need to edit anything from udev-lfs-197-2.tar.bz2.

Please do not top post and trim replies.

   -- Bruce

Re: Rename everything lfs/LFS -> unim/UNIM - chapter 6.61. Udev-197, LFS 7.3

Sandy Widianto — 2013-05-17T20:36:25

I mean should I extract udev-lfs-197-2.tar.bz2 and edit the files inside
the package to rename lfs/LFS -> unim/UNIM if I found it to compile?
Thank you for your kind words and support Mr. Bruce.


On Sat, May 18, 2013 at 3:19 AM, Bruce Dubbs <bruce.dubbs< at >gmail.com> wrote:

Re: Rename everything lfs/LFS -> unim/UNIM - chapter 6.61. Udev-197, LFS 7.3

Bruce Dubbs — 2013-05-17T20:19:06

I'm not sure I understand your question.  You did not follow the book's 
recommendation and use the name LFS for the target partition, but that's 
OK, even if a little confusing to others.  It's your distro.  However, 
once you enter chroot, everything is done as root, no changes to user or 
location need to be made.  Just follow the book.

   -- Bruce

Rename everything lfs/LFS -> unim/UNIM - chapter6.61. Udev-197, LFS 7.3

Sandy Widianto — 2013-05-17T20:03:51

Because of my project and my learning purpose, I renamed everything lfs/LFS
to unim/UNIM.
user/group:
lfs -> unim
on commands for everything, examples:
---------------
mkdir -v $UNIM/sources
chmod -v a+wt $UNIM/sources
mkdir -v $UNIM/tools
ln -sv $UNIM/tools /
groupadd unim
useradd -s /bin/bash -g unim -m -k /dev/null unim
passwd unim
chown -v unim $UNIM/tools
chown -v unim $UNIM/sources

cat > ~/.bashrc << "EOF"
set +h
umask 022
UNIM=/mnt/unim
LC_ALL=POSIX
UNIM_TGT=$(uname -m)-unim-linux-gnu
PATH=/tools/bin:/bin:/usr/bin
export UNIM LC_ALL UNIM_TGT PATH
EOF
-------------
mkdir -v $UNIM/{dev,proc,sys}
mknod -m 600 $UNIM/dev/console c 5 1
mknod -m 666 $UNIM/dev/null c 1 3
mount -v --bind /dev $UNIM/dev
mount -vt devpts devpts $UNIM/dev/pts
mount -vt proc proc $UNIM/proc
mount -vt sysfs sysfs $UNIM/sys

if [ -h $UNIM/dev/shm ]; then
  link=$(readlink $UNIM/dev/shm)
  mkdir -p $UNIM/$link
  mount -vt tmpfs shm $UNIM/$link
  unset link
else
  mount -vt tmpfs shm $UNIM/dev/shm
fi

-------------
chroot "$UNIM" /to