LinuxQuestions.org
Visit Jeremy's Blog.
Go Back   LinuxQuestions.org > Forums > Linux Forums > Linux - Distributions > Slackware
User Name
Password
Slackware This Forum is for the discussion of Slackware Linux.

Notices


Reply
  Search this Thread
Old 05-26-2011, 08:15 PM   #1
chexmix
Member
 
Registered: Apr 2002
Location: Arlington, MA
Distribution: Slackware, Debian, OpenBSD
Posts: 251
Blog Entries: 16

Rep: Reputation: 25
Cannot boot -current ... have done something stupid


Hi -

I've been happily running -current on a ThinkPad T42. Today I ran slackpkg to get the latest -current goodies, created a new initrd.gz (for 2.6.38.7-smp), re-ran LILO, and rebooted ...

And was greeted with:

Code:
This kernel requires the following features not present on the CPU:
pae
Unable to boot - please use a kernel appropriate for your CPU
I did see that there were changes to the kernel config for 2.6.38.7, but honestly don't know enough about these to hazard a guess as to how -- if at all -- they might impact me. Apparently at least one of them did.

My question I guess is ... what to do? I assume this requirement will be present on all the 2.6.38.7 kernels ...

Thanks in advance for any guidance,

Glenn
 
Click here to see the post LQ members have rated as the most helpful post in this thread.
Old 05-26-2011, 10:15 PM   #2
jamesf
Member
 
Registered: Dec 2004
Location: USA
Distribution: Slackware 12, Slackware64 14.2
Posts: 236
Blog Entries: 1

Rep: Reputation: 57
From Changelog:
Quote:
These are the main configuration changes from the 2.6.37.6 kernel in 13.37:
BLK_DEV_LOOP y -> m
HIGHMEM4G y -> n
HIGHMEM64G n -> y
LOG_BUF_SHIFT 15 -> 18
M686 y -> n
MPENTIUMIII n -> y
MOUSE_PS2_ELANTECH n -> y
HIGHMEM64G requires PAE, according to Google.

If I had done that, I'd boot my old install disk, mount my one partition (everything is under /), rmdir -rf everything but the /home subdirectory, reboot, and reinstall.

Then I'd re-create my users in the same order I had created them to give them the same uid so file ownership wouldn't be a problem, reboot, and start reconfiguring.

Good luck. If you're up to rebuilding a kernel without HIGHMEM64G on another system and transporting it to your non-working system you could get by without any other changes, probably. If you choose to do that grab the huge smp config from /boot of the broken system as a starting point and try only changing HIGHMEM64G _and_ HIGHMEM4G. EDIT: This paragraph assumes you want to stay with 38.7. EDIT 2: Added mention of HIGHMEM4G, which swapped settings with HIGHMEM64G.

Last edited by jamesf; 05-26-2011 at 10:41 PM.
 
Old 05-27-2011, 02:24 AM   #3
vbisis
Member
 
Registered: Dec 2005
Distribution: Slackware
Posts: 250

Rep: Reputation: 34
You can as well bott your system using installation cd and then install default kernel for 13.37 (kernel itself to boot, kernel source, headers and i think firmware)
 
Old 05-27-2011, 04:47 AM   #4
TobiSGD
Moderator
 
Registered: Dec 2009
Location: Germany
Distribution: Whatever fits the task best
Posts: 17,148
Blog Entries: 2

Rep: Reputation: 4886Reputation: 4886Reputation: 4886Reputation: 4886Reputation: 4886Reputation: 4886Reputation: 4886Reputation: 4886Reputation: 4886Reputation: 4886Reputation: 4886
Quote:
Originally Posted by jamesf View Post
If I had done that, I'd boot my old install disk, mount my one partition (everything is under /), rmdir -rf everything but the /home subdirectory, reboot, and reinstall.
Why should anyone do a complete reinstall, just because of a wrong configured kernel? Just get a huge kernel from the 13.37 repositories/DVD, install that, configure the .38-kernel the way you need and compile it.
Absolutely no need for a complete reinstall.

@chexmix: Now you know the reason why it is a good idea to always have a known good backup-kernel.
 
Old 05-27-2011, 06:10 AM   #5
chexmix
Member
 
Registered: Apr 2002
Location: Arlington, MA
Distribution: Slackware, Debian, OpenBSD
Posts: 251

Original Poster
Blog Entries: 16

Rep: Reputation: 25
I guess my question is: so because of this configuration change, I need to essentially roll my own kernels (with HIGHMEM64G reset to 'n') if I want to continue using Slackware ...?

Not that it is hard to do, but I am ... well, surprised, I guess. I've been a Slackware user for some time, and this has never happened to me before.

Please correct me if I am wrong here.
 
Old 05-27-2011, 06:21 AM   #6
TobiSGD
Moderator
 
Registered: Dec 2009
Location: Germany
Distribution: Whatever fits the task best
Posts: 17,148
Blog Entries: 2

Rep: Reputation: 4886Reputation: 4886Reputation: 4886Reputation: 4886Reputation: 4886Reputation: 4886Reputation: 4886Reputation: 4886Reputation: 4886Reputation: 4886Reputation: 4886
Quote:
Originally Posted by chexmix View Post
I guess my question is: so because of this configuration change, I need to essentially roll my own kernels (with HIGHMEM64G reset to 'n') if I want to continue using Slackware ...?
Exactly. I think it is simply changed, because nowadays there should be far more CPUs that support PAE than those that do not. From Wikipedia:
Quote:
PAE is provided by Intel Pentium Pro and above CPUs, including all later Pentium-series processors (except the 400 MHz-bus versions of the Pentium M). It is also available on other processors with similar or more advanced versions of the same architecture, such as the AMD Athlon[dubious discuss] and later AMD processor models.
 
Old 05-27-2011, 06:37 AM   #7
chexmix
Member
 
Registered: Apr 2002
Location: Arlington, MA
Distribution: Slackware, Debian, OpenBSD
Posts: 251

Original Poster
Blog Entries: 16

Rep: Reputation: 25
Quote:
Originally Posted by TobiSGD View Post
Exactly. I think it is simply changed, because nowadays there should be far more CPUs that support PAE than those that do not. From Wikipedia:
I suppose I am surprised. It feels somewhat ... unfriendly ... to people who use Slack on older hardware. But I know the team is wise & that there must be a good reason for this ...

Would it make sense to change the default behavior of slackpkg, so that when it upgrades the kernel packages it leaves the old (working) kernel in place, just in case? I don't know, just thinkin'.
 
Old 05-27-2011, 06:41 AM   #8
TobiSGD
Moderator
 
Registered: Dec 2009
Location: Germany
Distribution: Whatever fits the task best
Posts: 17,148
Blog Entries: 2

Rep: Reputation: 4886Reputation: 4886Reputation: 4886Reputation: 4886Reputation: 4886Reputation: 4886Reputation: 4886Reputation: 4886Reputation: 4886Reputation: 4886Reputation: 4886
I don't know if slackpkg really is replacing the kernel, I think it only points the links (vmlinuz, System.map, config) to the new files. just make a new entry for the old kernel in your lilo.conf and run lilo, this way you should always have a backup kernel.
 
Old 05-27-2011, 06:49 AM   #9
GazL
LQ Veteran
 
Registered: May 2008
Posts: 6,962

Rep: Reputation: 5088Reputation: 5088Reputation: 5088Reputation: 5088Reputation: 5088Reputation: 5088Reputation: 5088Reputation: 5088Reputation: 5088Reputation: 5088Reputation: 5088
The reason might very well be that they just weren't aware of the implications and didn't know it would break some systems completely.
I wouldn't be surprised to see this change reverted.

When it comes to a choice between working everywhere or including a certain feature, Pat has usually choosen the former on the principle that those who really want that extra feature can always add it themselves.
 
Old 05-27-2011, 06:58 AM   #10
markush
Senior Member
 
Registered: Apr 2007
Location: Germany
Distribution: Slackware
Posts: 3,979

Rep: Reputation: Disabled
Quote:
Originally Posted by TobiSGD View Post
I don't know if slackpkg really is replacing the kernel, I think it only points the links (vmlinuz, System.map, config) to the new files...
Yes, this is true and this is the moment when slackpkg asks us if we want to "run lilo yet". One workaround would be to skip running lilo and configure lilo.conf in order to leave the old kernel as a backup. As far as I remember slackpkg does not remove the old kernel in the /boot directory.

Save choice here (what I do) I don't use the link vmlinuz in my lilo.conf but the name of the kernel
Code:
...
# Linux bootable partition config begins
  image = /boot/vmlinuz-huge-2.6.37.6
  root = /dev/sda5
  label = 514cKW4r3
  read-only
# Linux bootable partition config ends
# OpenBSD
other = /dev/sda1
table = /dev/sda
loader = /boot/chain.b
label = OpenBSD
# Das ist das Ende
Markus
 
Old 05-27-2011, 07:03 AM   #11
chexmix
Member
 
Registered: Apr 2002
Location: Arlington, MA
Distribution: Slackware, Debian, OpenBSD
Posts: 251

Original Poster
Blog Entries: 16

Rep: Reputation: 25
Quote:
Originally Posted by GazL View Post
The reason might very well be that they just weren't aware of the implications and didn't know it would break some systems completely.
I wouldn't be surprised to see this change reverted.

When it comes to a choice between working everywhere or including a certain feature, Pat has usually choosen the former on the principle that those who really want that extra feature can always add it themselves.
Yes. That's why it felt weird to me.

Should I write PV or is there sufficient Slack-team presence on this board (I suspect the latter, but I don't want to omit doing something if I should do it)?
 
Old 05-27-2011, 07:18 AM   #12
GazL
LQ Veteran
 
Registered: May 2008
Posts: 6,962

Rep: Reputation: 5088Reputation: 5088Reputation: 5088Reputation: 5088Reputation: 5088Reputation: 5088Reputation: 5088Reputation: 5088Reputation: 5088Reputation: 5088Reputation: 5088
He might spot it here, but I'd drop him a note anyway just to be safe.
 
Old 05-27-2011, 07:37 AM   #13
hitest
Guru
 
Registered: Mar 2004
Location: Canada
Distribution: Slackware, Debian
Posts: 7,362

Rep: Reputation: 3759Reputation: 3759Reputation: 3759Reputation: 3759Reputation: 3759Reputation: 3759Reputation: 3759Reputation: 3759Reputation: 3759Reputation: 3759Reputation: 3759
Quote:
Originally Posted by chexmix View Post
Should I write PV or is there sufficient Slack-team presence on this board (I suspect the latter, but I don't want to omit doing something if I should do it)?
I think that would be a good idea so that he is aware of this issue.
 
Old 05-27-2011, 07:52 AM   #14
Darth Vader
Senior Member
 
Registered: May 2008
Location: Romania
Distribution: DARKSTAR Linux 2008.1
Posts: 2,727

Rep: Reputation: 1247Reputation: 1247Reputation: 1247Reputation: 1247Reputation: 1247Reputation: 1247Reputation: 1247Reputation: 1247Reputation: 1247
Quote:
Originally Posted by GazL View Post
The reason might very well be that they just weren't aware of the implications and didn't know it would break some systems completely.
I wouldn't be surprised to see this change reverted.

When it comes to a choice between working everywhere or including a certain feature, Pat has usually choosen the former on the principle that those who really want that extra feature can always add it themselves.
Just to notice that on slackware-current, the non-smp kernels, kernel-huge-2.6.38.7-i486-1.txz and kernel-generic-2.6.38.7-i486-1.txz, use an 'CONFIG_HIGHMEM4G=y' option, so the support for the elder hardware is unchanged.

Talking about oldies, why OP not use the non-smp kernels? I don't think that his machine is multi-cpu/core, anyway ...
 
2 members found this post helpful.
Old 05-27-2011, 08:36 AM   #15
chexmix
Member
 
Registered: Apr 2002
Location: Arlington, MA
Distribution: Slackware, Debian, OpenBSD
Posts: 251

Original Poster
Blog Entries: 16

Rep: Reputation: 25
Quote:
Originally Posted by Darth Vader View Post
Just to notice that on slackware-current, the non-smp kernels, kernel-huge-2.6.38.7-i486-1.txz and kernel-generic-2.6.38.7-i486-1.txz, use an 'CONFIG_HIGHMEM4G=y' option, so the support for the elder hardware is unchanged.

Talking about oldies, why OP not use the non-smp kernels? I don't think that his machine is multi-cpu/core, anyway ...
That's a good question and I am afraid I don't have a very good answer for it. Somewhere along the line I got the impression that it was a good idea to run the smp kernel(s) anyway (I don't have a multi-core box to my name & never have) ... I wish I remembered where I got that impression, but I don't.
 
  


Reply


Thread Tools Search this Thread
Search this Thread:

Advanced Search

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is Off
HTML code is Off



Similar Threads
Thread Thread Starter Forum Replies Last Post
stupid grub boot-screen question ight8 Linux - Software 3 11-03-2006 03:13 AM
A stupid rm -rf no more /boot,/bin,/etc what would you do? nx5000 Debian 4 12-24-2005 06:19 PM
SUSE Boot CD stupid situation Sagittarius20 Linux - Newbie 8 09-19-2005 01:59 AM
Stupid dual boot question sumguy231 Linux - General 3 03-15-2005 07:51 PM
stupid question, boot.log bit7 Linux - Software 1 03-13-2003 09:12 AM

LinuxQuestions.org > Forums > Linux Forums > Linux - Distributions > Slackware

All times are GMT -5. The time now is 06:44 AM.

Main Menu
Advertisement
My LQ
Write for LQ
LinuxQuestions.org is looking for people interested in writing Editorials, Articles, Reviews, and more. If you'd like to contribute content, let us know.
Main Menu
Syndicate
RSS1  Latest Threads
RSS1  LQ News
Twitter: @linuxquestions
Open Source Consulting | Domain Registration