Bugzilla Bug 231121

Installation goes as usual, but when I go to reboot, I get this message:

Grub Loading Please Wait

Error 17

I tried to install a second time, but I got the same results. I checked
/var/log and no install log was created this time

Reproducible: Always

Steps to Reproduce:
1.Do a clean install
2.Finish the install, then reboot
3.

I should also note that I installed Beta2 on the same computer using the exact
same options and grub installed perfectly fine (although there were other
issues).

*** This bug has been marked as a duplicate of bug 230998 ***

How is this a duplicate when I'm using 2008.0r1.

Read the version please.

Sorry, misfire :P This is a completely different issue.

What hardware is this?

That error usually means that it found the partition but it was of a type that
wasn't recognized. Can you attach the output of 'fdisk -l <your drive>' from
the LiveCD?

Athlon 64 5000+ BE
Gigabyte GA-MA770-DS3
Seagate Barracuda 7200.11 750GB SATA II Hard Drive
Diamond Viper Radeon HD 3870
Samsung 22x SATA DVD Burner, Asus SATA DVD-Rom Drive


My Harddive has 4 partitions:

sda1 & sda 2 are NTFS.
sda 3 is linux swap and sda4 is the root partition.

Here's a brief history:

1) I tried to install Gentoo 2008.0 beta2, but I canceled before it ended
2) I installed Ubuntu 8.04.

3) I found out that the beta2 installer wasn't broken,the GUI just didn't
update, so I installed it again - this time from the console. It installed fine
(although I'm sure you've heard about the lib corruption issues).

4) I attempted to install 2008.0 final, but I came across the bug I reported
yesterday.

5) I attempted to install 2008.0r1, but when I reboot I get the error that I
reported.

6) About 30 mins ago I gave the install one last try. This try, I switched the
partitions back (root back to sda4 and swap back to sda3), but it still doesn't
boot.This time, I get Error 5 instead of Error 17.


Probably a little more information than you wanted, but hey, it won't kill
anyone ;P  Here's what happens when I do what you ask:


gentoo@livecd ~ $ sudo mount /dev/sda4 /
gentoo@livecd ~ $ fdisk -l /dev/sda
bash: fdisk: command not found


I should also mention that my /boot folder only shows 3 items: the initramfs,
the kernel, and the System.map. Is it supposed to be like that?

(In reply to comment #5)
> gentoo@livecd ~ $ sudo mount /dev/sda4 /

You just mounted your /dev/sda4 over the top of the LiveCD root. That doesn't
do anything useful :P Try mounting it at /mnt/gentoo.

> gentoo@livecd ~ $ fdisk -l /dev/sda
> bash: fdisk: command not found

It needs to be run as root (or with sudo).

Bah... left off the sudo on the second command.. sorry about that. Here we go:



gentoo@livecd ~ $ sudo mount /dev/sda4 /mnt/gentoo
gentoo@livecd ~ $ sudo fdisk -l /dev/sda

Disk /dev/sda: 750.1 GB, 750155292160 bytes
255 heads, 63 sectors/track, 91201 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes
Disk identifier: 0x0005d793

   Device Boot      Start         End      Blocks   Id  System
/dev/sda1               1       32635   262140606    7  HPFS/NTFS
/dev/sda2           32636       65270   262140637+   7  HPFS/NTFS
/dev/sda3           65271       65531     2096482+  82  Linux swap / Solaris
/dev/sda4           65532       91201   206194275   83  Linux

I noticed that I made a mistake. Grub was outputting Error 2, not Error 5.
Sorry for any time that I may have wasted.

Anyways, at this point I've figured that my partition table got screwed
somehow, possibly by the first version of 2008.0.

Don't worry about it, I'll read a bit more and see if I can get it straightened
out myself. I'm sure you're quite busy with more important things to do.

Thanks for the time, and thanks for the help.

As an addendum to this bug, the 2008.0-r1 release has an issue where the grub
bootsplash image is missing.  We are working on it.

I enjoy a multi-disk multi-boot workstation:
sda1:  Master_Boot
sdc1:  Gentoo /boot
sdc3:  Gentoo /

Installing from 2008.0-r1 liveCD for amd64,
after successful installation, reboot didn't work.

Booted SystemRescueCD, I was able to install grub
onto (hd2) as well as (hd2,0).

Booting (hd2) via BIOS from sdc MBR loads stage 1.5, then yields "Error 2".

Chainloading into (hd2,0) enters into grub menu.
Setting root to (hd2,0) and commanding "boot": grub requests kernel to be
specified.

(In reply to comment #10)
Addendum:
Re-Booting SystemRescueCD and deleting the splash entry in grub.conf did _not_
solve the problem.

(In reply to comment #10)
Addendum:
specifying "kenel (hd2,0)/kernel-genkernel-..." directly in grub
yields "Error 2: Bad file or directory type"

(In reply to comment #10)
Addendum:
Re-Booting SystemRescueCD, chroot-ing into the newly installed gentoo,
commanding "grub" yields:
"grub: error while loading shared libraries:
 libncurses.so.5: cannot open shared object file:
 No such file or directory"

(In reply to comment #10)

SOLVED:

- chroot into newly installed gentoo
- emerge --sync

- emerge -av gentoolkit
- checked with revdep-rebuild -a : o.k.

- emerge -av grub
- - note hint at end of install:
- - install grub files via "--config"
- reboot : initially works - BUT:

More errors to follow: 
"network interface eth0 does not exist ..."

Thus: Built my own custom kernel ... adapted grub.conf : boot o.k.

Folks, for any beginner with Gentoo, this is _not_ really funny.

So far, you're the only one to report this issue. Perhaps it's just an oddity
with your setup.

(In reply to comment #15)

> Perhaps it's just an oddity with your setup.

ASUS M2N-SLI Deluxe, AMD 64 X2 5200,4 x S-ATA -
what do you call "odd" with such an ordinary setup?

The only "oddity" is that I don't want Gentoo to be installed 
onto the first harddisk.

> So far, you're the only one to report this issue.

Do you have any positive report of installing into another disk than the first?

I tried again - this time, onto /dev/sda2 ; same negative result.

Thereby I tried really hard to discover any option to explicitly tell the
installer where to setup grub. I could not find any.

If there is one, I would greatly appreciate a hint.

I have encountered the same error, and this has left my machine unbootable (as
a temporary measure, I have installed grml to have a grub).

I am using 2008.0 live CD

Everything goes fine until after portage installation. Then I get an error
saying there is no kernel in boot (nothing named kernel*, or something like
that). But installer goes on and then tries to install grub, which then fails
with some other error.

I am trying some magic right now. If it works, I will report.

That's not a similar error. That is a completely different problem that was
fixed with the 2008.0-r1 LiveCDs.

Oh! Sorry :)  Will open a new bug report soon...

Why would you open a new bug report when Andrew said that it was FIXED with the
-r1 CDs?  *grin*

Ok, update: see http://forums.gentoo.org/viewtopic-t-700523-highlight-.html

I am waiting for a confirmation before I open this bug.

(In reply to comment #20)

> ... it was FIXED with the -r1 CDs?

I doubt your
>   *grin*
is a constructive help.

Please, cf. my comment 14.

My questions in comment 16 are still unanswered.

Yours respectfully
Manfred

In response to comment 14: If the kernel doesn't have your network device, i'm
not sure what there is to do except rebuild your kernel.  It is not the fault
of the installer.
In response to comment 16: gli-dialog in advanced mode can let you specify the
drive and/or MBR settings for the bootloader.  don't underestimate your BIOS's
ability to choose which drive boots first.

(In reply to comment #23)
> In response to comment 14: If the kernel doesn't have your network device, i'm
> not sure what there is to do except rebuild your kernel.  It is not the fault
> of the installer.

Booting the liveCD, I had no problems to reach the net - 
even via default DHCP ...

As reported, the problem arouse only _after_ reboot.

To me, this _is_ an issue. Or, more precise, not to me, as pointed out above,
but for the intended user group being the goal of a LiveCD + graphical
Installer.

> In response to comment 16: gli-dialog in advanced mode can let you specify the
> drive and/or MBR settings for the bootloader.

Thanks a lot for your valuable hint -
I will try to find that 'advanced mode' and will report again.

> don't underestimate your BIOS's ability to choose which drive boots first.

I do exploit that every other day ...
Sorry, I didn't get what you wanted to point out with this hint
and what it has to do with this bug report ?

(In reply to comment #23)
> ... gli-dialog in advanced mode ...

I re-tried, but until "installing grub ..." I could not identify this option.

Would you be so kind to exactly define where to find it,
so that I can re-test as promised?
Thanks a lot in advance.

(My unanswered question from comment #16)

PING  @  Chris Gianelloni , Preston Cody :

> Do you have any positive report of installing into another disk than the first?

I re-installed again,
this time preparing the /boot partition with mkfs.ext2 -I 128
in order to rule out the 128bit / 256bit hassle 
introduced with new grub versions.
Still:
- chainload from /dev/sdb1: Error 13
  ==> no bootloader installed here
- BIOS - boot from /dev/sdb MBR: grub starts, but Error 15: file not found
  ==> installer installed grub here, but ...

although:

mammut mnt # mkdir sdb1
mammut mnt # mount /dev/sdb1 sdb1
mammut mnt # cd sdb1
mammut sdb1 # ll
insgesamt 26M
lrwxrwxrwx 1 root root    1 21. Jul 2008  boot -> .
drwxr-xr-x 2 root root 4,0K 21. Jul 2008  grub
-rw-r--r-- 1 root root  22M 21. Jul 2008 
initramfs-genkernel-x86_64-2.6.24-gentoo-r7
-rw-r--r-- 1 root root 2,4M 21. Jul 2008 
kernel-genkernel-x86_64-2.6.24-gentoo-r7
drwx------ 2 root root  16K 21. Jul 08:27 lost+found
-rw-r--r-- 1 root root 1,4M 21. Jul 2008 
System.map-genkernel-x86_64-2.6.24-gentoo-r7

And again: I coulkd not identify _any_ option to switch into "gli-dialog in
advanced mode".

PING @ Preston Cody (c.f. Comment  #23) :

> Would you be so kind to exactly define where to find it ... ?

There are more errors:
- selecting keyboard map /= default does not work; e.g. selecting "de":
- - the selection boxes still expect an american kbd,
- - the terminal as well, ...
- there is no hint how to enter "root" mode ( sudo su -)
- starting the Browser has problems, only works at second click at firefox
icon,
- - then results into calling a non-existing page ...
- the installer refuses to (re-) install into fs with any data in it,
- - but no option to (re-) format the fs is available;
- - people have to be knowledgeable enough to enter sh and mkfs ...
but these are really secondary, compared to the fact that the newly installed
system not come ing up at all.

As long as people in charge continue preferring to bury their head in the sand,
wasting effort into pleading "this bug doesn't exist" instead of helping
investigate to _solve_ it, for myself I will retreat to stick to manual install
(i.e. SystemRescueCD, Gentoo Handbook and Funtoo stages), and advice any
interested linux novice to enjoy his start with a distribution that just
installs.

I'm really very, very sorry.

If - one day - the RelEng team is prepared to value the setup of a team of test
people and documented (non-mainstream) test setups _and_ is prepared to take
their responses seriously, I will happily return and volunteer.

Kind regards
Manfred

(ADDENDUM to comment #27)

I disconnected my first S-ATA,
such the former /dev/sdb became the first hard disk.

Re-Installing: again entering into "grub >".

---

Re-Installing manually from an existing Gentoo (2007) installation,

chrooting,

env-update yielded 

# env-update
Traceback (most recent call last):
  File "/usr/sbin/env-update", line 27, in ?
    import portage
  File "/usr/lib64/portage/pym/portage.py", line 10043, in ?
    init_legacy_globals()
  File "/usr/lib64/portage/pym/portage.py", line 9999, in init_legacy_globals
    db = create_trees(**kwargs)
  File "/usr/lib64/portage/pym/portage.py", line 9938, in create_trees
    config_incrementals=portage_const.INCREMENTALS)
  File "/usr/lib64/portage/pym/portage.py", line 1380, in __init__
    allow_sourcing=True)
  File "/usr/lib64/portage/pym/portage_util.py", line 356, in getconfig
    raise portage_exception.ParseError(str(e)+" in "+mycfg)
portage_exception.ParseError: [Errno 2] No such file or directory:
'/usr/portage/local/layman/make.conf' in /etc/make.conf

(ADDENDUM to comment #28)

> Re-Installing manually from an existing Gentoo (2007) installation,

exactly as in "Gentoo Handbook fopr amd64", u8sing official stage-3 and
portage-latest of today.

(In reply to comment #28)
> '/usr/portage/local/layman/make.conf' in /etc/make.conf

Yes, this is what happens when you blindly copy the make.conf from another box
and don't pay attention to what's actually in it. This is completely your
fault.

(In reply to comment #30)

Damned - overlooked it - I feel embarrassed - shortage of sleep ... Sorry.

Nevertheless, my main concern (and of this bug) is with the _first_ part of
Comment  #28.

 LivCD 2008-r1 i686 and amd64  LiveCD doesn't Install Grub Properly, use
install-dialog.
 Tried SystemRescueCD as it was written here earlier:
- chroot into newly installed gentoo
- emerge --sync

- emerge -av gentoolkit
- checked with revdep-rebuild -a : o.k.

- emerge -av grub
- - note hint at end of install:
- - install grub files via "--config"
- reboot : initially works - BUT:

More errors to follow: 
"network interface eth0 does not exist ..."
Result same. Tried different m/b and hdd.
M/b: Asus P4P800SE (i686), GA-MA790FX-DS5 (amd64)
HDD: Seagate IDE/Sata different volumes

The installer is deprecated.