Gentoo Websites Logo
Go to: Gentoo Home Documentation Forums Lists Bugs Planet Store Wiki Get Gentoo!
Bug 12463 - 1.4_rc1 Installation issues on ASUS L3800C notebook
Summary: 1.4_rc1 Installation issues on ASUS L3800C notebook
Status: RESOLVED INVALID
Alias: None
Product: Gentoo Release Media
Classification: Unclassified
Component: Everything (show other bugs)
Hardware: x86 All
: High normal
Assignee: Daniel Robbins (RETIRED)
URL:
Whiteboard:
Keywords:
Depends on:
Blocks:
 
Reported: 2002-12-19 20:12 UTC by Martin Mokrejš
Modified: 2003-02-04 19:42 UTC (History)
0 users

See Also:
Package list:
Runtime testing required: ---


Attachments

Note You need to log in before you can comment on or make changes to this bug.
Description Martin Mokrejš 2002-12-19 20:12:32 UTC
Hi,
  I have couple of problems or hints at least. The kernel distributed at
http://www.ibiblio.org/pub/Linux/distributions/gentoo/releases/1.4_rc1/x86/livecd-basic-1.4_rc1-r2.iso
seems to be old? Is there anything never, so is 1.4 version out finally?

  After I booted this, the kernel crashes after inserting "Compaq SMART2 Driver
(2.4.21). Fortunately, it was just some modprobe, so the system continues and
give sme login prompt. ;)

  This notebook should have Intel Solano-M network card, but Realtek 8139cp is
used. After I've rebooted with new kernel, I cannot load 3139cp module, but
8139too module is loadable and works well (linux-2.4.19-gentoo-r10 kernel).

ping(1) to some hosts reports some weird errors about some bytes inside a packet
having wrong values. But in principle the card works. Sorry, I cannot reproduce
the problem right now, but I've seen this twice after fresh bootups.


Few notes regarding teh installation docs:

1. The installation document should lead user to use 
  tunefs -i 0 -c 0 /dev/hdaBOOT
and tunefs(1) should be provided.

2. the network setup should be run after partitioning and reboot. It's useless
to configure the network, make partition and reboot to make the surely visible,
right?

3. I thing the network broadcast address isn't necessary, as long as the user
supplies interface IP address/netmask.

4. at the step of mounting newly created filesystems, I think there should be
mkdir /mnr/gentoo/boot
command issued after the root+usr+var partition has been mounted. So the
mountpoint is created on that filesystem. At least that's what I've done.

5. emerge sync should handle case when DNS resolving does not help. For example
as user did not copy the /etc/resolv.conf over yet. :(

6. in the new chrooted environment, is it possible to have vi(1) too, not only
nano? BTW: From the chrooted environment df(1) command does not work anymore! If
I login through another virtual console to non-chrooted environment, it works as
expected.

7. it seems during the bootstrap step and stage2 too, my CFLAGS from
/etc/make.conf haven't been accepted, as for example gcc was compiled with -O2
and not -O3 etc. Is it possible? ). Could bootstrap.sh accept ctrl+z so that
shell would really suspend a job? The reason why I ask is that I've seen some
errors (error 404 from wget that some package wasn't found, I think it was
portage* package, and ctrl+z made the system stop processing, but did not give
me a shell prompt)... Is it the running shell which does not support jobs
suspension? I guess there's a plenty of space on the cdrom to contain bash with
this feature enabled.

As I got to stage3 I can say now, the /etc/make.conf does not contain my
changes? Is it possible it got automatically overwritten?

And I've realized just before the end of installation, that eror 404 is handled
and the package is downloaded from another host! Good jobs, just mention this in
the documentation!!!

Can one safely interrupt bootstrap.sh and restart, or is
rm -f /mnt/gentoo/var/tmp/portage/*
sufficient?

Do I understand right that the portage is on the livecd itself, so older version
then is the one on the ftp server? Could be mentioned in the docs how to update
portage before bootsrapping, so that I will not face same installation errors?

8. "emerge -p system" tells me now on one line

[ebuild    UD] sys-apps/portage-2.0.45-r5 [2.0.46-r2]
*** Portage will stop merging at this point and reload ....

What's that? So the portage package won't get downgraded, right?
Why does it have to reload itself then? I guess in the installtion manual should
be explained what might go on here, so how how to interpret the output.
For example, how could one delete an entry from the list of packages which will
be installed? Note about "emerge --help" doesn't help, sorry.

After I got  to the stage3 I can tell you, it downgraded and in the "emerge -u
world" suggests to upgrade back.;)

I got left with some un-upgraded config files, but at stage 3 it turned out that
CVS version number was higher by those original config files, so I just had to
delete those _cfg* files. Anyway, note about
etc-update program would be nice, probably in the "emerge --help" output too.
;-)

9. Nowhere in the beginning of the installtion document is clearly explained,
that one needs network connection online. Why couldn't we download the
distfiles/ directory and burn on a cdrom at least?

10. during stage2 I've seen errors that mount(1) was missing. Is anywhere logged
output from portage?

11. I'd suggest to mention requirement for tmpfs and devfs more in the beggining
of install manual, probably in the introduction and expalining why. I guess it's
related to the /boot extra partition?

12. cron stuff
the link to
http://lists.gentoo.org/pipermail/gentoo-announce/2002-April/000151.html
does not work

13. In final steps you meantion to set root password. Actually, the install
process already told user at the very beginning to set password immediately, as
the user is connected to network. So I think this note should be moved in the
docs to very beginning.

No, sorry, at that step I've changed password in the tmpfs, which is lost after
reboot. So here (final step) it has to be stated clearly to the user that he has
to set password again. This password will remain after reboot.

14. I see that fsck.reiserfs called by some init.d script is missing.

15. If I recompile kernel, do I have to rerun somehow grub like "lilo -r"?
Grub actually gives me on boot only a prompts. it doesn't boot straight into
linux? How can I adjust this?

16. the USE index page seems to be incomplete, right? Where's gdb, procmail,
ghostscript, gv, mc, libpng, jpeg, libmng, tiff? "kerberos" refers to which
kerberos distribution? What is "pic"? See
http://www.gentoo.org/dyn/use-index.xml

Further, I did not get an idea where actually user selects which packages will
be installed. During the installation process I've realized, but I think it
should be more clearly stated at the beginning.

17. You do not mention at all how is one upgrading the system. Is it just by
running "emerge -u world" or "emerge -u"?


I wish you good luck and thanks for interresting product.
Martin
Comment 1 Daniel Robbins (RETIRED) gentoo-dev 2002-12-20 14:51:32 UTC
Try http://www.gentoo.org/dyn/experimental/livecd/livecd-basic-x86-
2002121700.iso.
It should work. (stage1 tarball in /mnt/cdroot/gentoo if you're interested.)
Please post your results here.
Comment 2 Daniel Robbins (RETIRED) gentoo-dev 2002-12-22 01:20:44 UTC
This bug report is way too long. You should split it into individual bug 
reports for each particular issue so that they can be assigned to various 
people depending on the type of report.