LinuxQuestions.org
Visit the LQ Articles and Editorials section
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 03-13-2013, 09:52 PM   #1
michaelslack
Member
 
Registered: Feb 2013
Location: Sydney
Distribution: slackware
Posts: 83

Rep: Reputation: 33
Are new -current kernels samsung/uefi safe?


A few weeks ago in another post relating to booting via uefi on samsung laptops, Pat suggested waiting for the new kernels:

Quote:
Originally Posted by volkerdi View Post
I'd suggest waiting until the next kernels come along in -current since those will be Samsung UEFI safe, and then trying to use those along with elilo. If that works, you can look into a boot manager like rEFIt or rEFInd for dual booting.
I seem to remember while reading around at that time somewhere it said ``the linux kernel will be samsung-safe at 3.8'' or similar (as luck would have it I can't find that reference now that I want to share it).

Now, in the last -current update we have new kernels and new uefi-booting support. However, Pat also says in the ChangeLog:

For this kernel update I decided to go with 3.7.10. Yeah, the 3.7 series
is EOL, but I've heard about some broken drivers in 3.8.x that make me
hesitate to push forward.


So I'm suspecting that maybe a few weeks ago Pat was planning to upgrade the kernels to 3.8, but found these other driver issues and so changed his mind and went 3.7.10 instead.

The $64 question is: are these new kernels samsung-safe or not? If I hadn't read about 3.8 being the first samsung-safe kernels I would have gone ahead with these new kernels and tried booting my samsung using uefi. But the possibility of bricking has made me very cautious.

Thanks in advance,

Michael
 
Old 03-13-2013, 10:09 PM   #2
volkerdi
Slackware Maintainer
 
Registered: Dec 2002
Location: Minnesota
Distribution: Slackware! :-)
Posts: 867

Rep: Reputation: 1750Reputation: 1750Reputation: 1750Reputation: 1750Reputation: 1750Reputation: 1750Reputation: 1750Reputation: 1750Reputation: 1750Reputation: 1750Reputation: 1750
The fixes went in with version 3.7.6, however the bug in the firmware remains. It's probably safe, but there are no guarantees.
 
1 members found this post helpful.
Old 03-14-2013, 01:13 AM   #3
comet.berkeley
Member
 
Registered: Dec 2009
Location: California
Distribution: Slackware current
Posts: 152

Rep: Reputation: Disabled
It looks like the same "samsung-laptop" fixes by Matt Fleming were applied to the older 3.2.38 and 3.4.31 kernels as well.

https://www.kernel.org/pub/linux/ker...angeLog-3.2.38

https://www.kernel.org/pub/linux/ker...angeLog-3.4.31

https://www.kernel.org/pub/linux/ker...hangeLog-3.7.6

But as Patrick said:
Quote:
Originally Posted by volkerdi View Post
... the bug in the firmware remains. It's probably safe, but there are no guarantees.
 
Old 03-14-2013, 01:23 AM   #4
michaelslack
Member
 
Registered: Feb 2013
Location: Sydney
Distribution: slackware
Posts: 83

Original Poster
Rep: Reputation: 33
Thanks again for the quick reply Pat. It prompted me to look around a bit and after reading Matthew Garrett's journal entry it appears that the safest thing to do is not boot samsung devices using uefi at all.

This more or less returns me to my earlier question: is there a way to change a samsung (or any other) uefi-booting device into a non-uefi/csm/legacy booting device without throwing away the custom-samsung-windows 8 installation?

You suggested in your earlier response that you didn't think it was possible to convert from GPT to MBR without wiping the whole disk. Another option is to just create recovery media in case I want to put windows 8 back onto it at a later date. I've made a factory image but I don't know if that alone is enough to re-install windows onto a wiped disk. I think it might need the hidden windows recovery environment (RE) partition there or something to work.

Anyway this is now getting off-topic so I might post a separate question.

Thanks again,

Michael

Last edited by michaelslack; 03-14-2013 at 01:29 AM.
 
Old 03-14-2013, 01:56 AM   #5
volkerdi
Slackware Maintainer
 
Registered: Dec 2002
Location: Minnesota
Distribution: Slackware! :-)
Posts: 867

Rep: Reputation: 1750Reputation: 1750Reputation: 1750Reputation: 1750Reputation: 1750Reputation: 1750Reputation: 1750Reputation: 1750Reputation: 1750Reputation: 1750Reputation: 1750
Just a bit more information... the samsung-laptop.ko kernel module is not present on the installer, so the installer cannot trigger this bug. If you're worried about it, you can always delete /mnt/lib/modules/3.7.10/kernel/drivers/platform/x86/samsung-laptop.ko (also in 3.7.10-smp for 32-bit) after installing, but before setting up elilo to boot the main system. I don't think there's any way you'd have a problem then. Probably you wouldn't anyway, but that's how to make sure.
 
1 members found this post helpful.
Old 03-21-2013, 08:02 PM   #6
comet.berkeley
Member
 
Registered: Dec 2009
Location: California
Distribution: Slackware current
Posts: 152

Rep: Reputation: Disabled
More UEFI Samsung news from Matthew Garrett

More Samsung UEFI news from Matthew Garrett:

UEFI on Samsung notebooks: Half full is almost broken
In a blog post describing the use of UEFI variables for debugging purposes, Matthew Garrett mentions that the memory for UEFI variables being filled up by more than fifty per cent is thought to be the reason why Samsung notebooks will no longer boot and may require repair in certain conditions for example after starting some Linux distributions with UEFI, or after executing a Windows test program that stores information in the UEFI firmware...

http://www.h-online.com/open/news/it...n-1827493.html

Using pstore to debug awkward kernel crashes
The problem with Samsung laptops bricking themselves turned out to be down to the UEFI variable store becoming more than 50% full and Samsung's firmware being dreadful, but the trigger was us writing a crash dump to the nvram...

http://mjg59.dreamwidth.org/23554.html
 
  


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
[SOLVED] samsung/gdisk/uefi booting anxiety michaelslack Slackware 5 07-02-2013 11:15 PM
[SOLVED] No Boot on UEFI box (Samsung 350V) business_kid Linux - Hardware 26 05-29-2013 04:34 AM
LXer: Samsung UEFI bug definitely not fixed LXer Syndicated Linux News 1 02-08-2013 01:12 PM
LXer: UEFI-enabled Samsung laptops get bricked with Linux LXer Syndicated Linux News 0 01-31-2013 08:00 AM


All times are GMT -5. The time now is 11:37 PM.

Main Menu
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
identi.ca: @linuxquestions
Facebook: linuxquestions Google+: linuxquestions
Open Source Consulting | Domain Registration