[lfs-support] Chroot failure

jasonpsage at jegas.com jasonpsage at jegas.com
Sun Jan 15 05:34:30 PST 2012


>Hello.
>
>I follow LFS 7.0 book.
>After Glibc-2.14.1 has been installed (chapter 6.9) chroot fails with
>"Segmentation fault" error. 
>If I rename $LFS/lib to $LFS/lib_ just before chrooting then chroot is
>completed OK. 
>Once chrooted I rename back from /lib_ to /lib being inside chroot
>environment. 
>All further tests on dummy.c (chapter 6.10 "Re-adjusting the Toolchain")
>pass OK exactlty 
>they described in the book.
>
>Does that mean I must suspend further building to achieve a normal chrooting
>?
>
>Thank you in advance.
>
>P.S. About my host system: Puppy Linux 5.11
>

I'm not the only Puppy fan out here, but I was going to use Puppy 5.21 I
think?
I found it didn't have most of the requirements, even with the
development "pup" or SFS
file loaded.

Further, you might be running into "weird behavior" due to Puppy's
layered file system.

This means some of the file system is in ram, some is on disk and some
is "layered" to
appear to be there when underneath something else is or isn't.

For me, I came to the conclusion that there were to many variables in
Puppy. I'm not saying it can't be done, I just personally ran from it.
In fact I like Slackware a lot too and found weird things appear as I
started to get into BLFS - like Python 2 was giving me a hard time, yet
the same scripts run on Cent OS 6 (32 bit version) didn't. I suppose
this is either the "required libraries" having slightly different
versions or compiled with different options, I don't know. What I do
know is CentOS 6 32bit has proved to be a better host for me.

You do have to make sure you select Ext3 file system (I recommend for
the host AND LFS 7.0) and you might notice if your first partition is a
SWAP, it moves it to /dev/sda2 when you add the / (root) partition. (Odd
I thought - just take note).

--Jason





More information about the lfs-support mailing list