LinuxQuestions.org
Latest LQ Deal: Latest LQ Deals
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 10-18-2019, 01:57 PM   #31
abga
Senior Member
 
Registered: Jul 2017
Location: EU
Distribution: Slackware
Posts: 1,275

Rep: Reputation: 673Reputation: 673Reputation: 673Reputation: 673Reputation: 673Reputation: 673

Happy to hear that you've found the right parameter for making your system stable.

For the standard Slackware installation the following lilo append line should then be used to instruct the kernel to "talk the same acpi language(OSI)" as the BIOS acpi, for systems that aren't stable otherwise:
Code:
append="vt.default_utf8=0 acpi_osi=Linux"
Wondering where did you get the "utf8xx" ?
https://docs.slackware.com/talk:slac...vtdefault_utf8
 
Old 10-21-2019, 04:41 AM   #32
Twigster
Member
 
Registered: Oct 2019
Location: France
Distribution: Slackware64 14.2
Posts: 34

Original Poster
Rep: Reputation: Disabled
Speaking of docs, is there a place to write about this kind of issues? I could only find a EEE PC-related howto.

Maybe the Slackware etiquette requires asking someplace else when it comes to editing documentation?
 
Old 10-21-2019, 09:13 AM   #33
Twigster
Member
 
Registered: Oct 2019
Location: France
Distribution: Slackware64 14.2
Posts: 34

Original Poster
Rep: Reputation: Disabled
Ok, tough luck, the computer hanged again after a whole weekend of stability. I also get some acpi-related errors :
Code:
 [    0.217422] ACPI: Added _OSI(Module Device)
 [    0.217507] ACPI: Added _OSI(Processor Device)
 [    0.217592] ACPI: Added _OSI(3.0 _SCP Extensions)
 [    0.217677] ACPI: Added _OSI(Processor Aggregator Device)
 [    0.217764] ACPI: Added _OSI(Linux)
 [    0.233828] ACPI: Interpreter enabled
 [    0.233934] ACPI: (supports S0 S3 S4 S5)
 [    0.234003] ACPI: Using IOAPIC for interrupt routing
 [    0.234122] PCI: Using host bridge windows from ACPI; if necessary, use "pci=nocrs" and report a bug
 [    0.280537] acpi LNXIOBAY:00: ACPI dock station (docks/bays count: 1)
 [    0.285125] acpi PNP0C15:00: ACPI dock station (docks/bays count: 2)
 [    0.293545] ACPI: PCI Root Bridge [PCI0] (domain 0000 [bus 00-ff])
 [    0.293642] acpi PNP0A03:00: _OSC: OS supports [ExtendedConfig ASPM ClockPM Segments MSI]
 [    0.293782] acpi PNP0A03:00: _OSC failed (AE_NOT_FOUND); disabling ASPM
 [    0.293877] acpi PNP0A03:00: [Firmware Info]: MMCONFIG for domain 0000 [bus 00-3f] only partially covers this bridge
 [    0.065000] tsc: Marking TSC unstable due to check_tsc_sync_source failed
Reading here tells me that i might go for acpi_osi='Windows XP' but i'm puzzled about acpi_osi!= and acpi_rev_override. And I do not know how my kernel decides to add all those other OSI strings before the one I told it to add.

EDIT : i'm not even sure i set the parameter correctly :
Code:
[    0.216766] ACPI: Added _OSI('Windows)
root@~# cat /etc/lilo.conf | grep append
append="vt.default_utf8=0 acpi_osi='Windows 2001'"
With the system in that state, I can reproduce the crash doing the same steps as this morning.
With acpi=off, no crash with the same modus operandi.

EDIT2 : this documentation seems to say that acpi_osi=Linux is bad practice?

Last edited by Twigster; 10-21-2019 at 10:32 AM.
 
Old 10-21-2019, 11:53 AM   #34
abga
Senior Member
 
Registered: Jul 2017
Location: EU
Distribution: Slackware
Posts: 1,275

Rep: Reputation: 673Reputation: 673Reputation: 673Reputation: 673Reputation: 673Reputation: 673
Quote:
Originally Posted by Twigster View Post

Reading here tells me that i might go for acpi_osi='Windows XP' but i'm puzzled about acpi_osi!= and acpi_rev_override. And I do not know how my kernel decides to add all those other OSI strings before the one I told it to add.
The manjaro article is well composed and contains all the "guess-work" acpi_osi= parameters you could try. And it's acpi_osi=! and not acpi_osi!= to disable all the built-in OS vendor strings.
Here you have a short explanation about the OSI strings (first answer):
https://unix.stackexchange.com/quest...ight-vendor-do

Quote:
Originally Posted by Twigster View Post
EDIT2 : this documentation seems to say that acpi_osi=Linux is bad practice?
I saved the kernel doc link you provided for my reply, meanwhile you already found it yourself. No, acpi_osi=Linux is not bad practice, it should be the desired option for Linux compatibility, that's if the vendor implemented&tested Linux in their acpi design. It doesn't look to be the case for your system and I'd suggest to try acpi_osi=Windows XP
I guess you need to use quotes for the proper syntax, like acpi_osi="Windows XP" (kernel doc) or acpi_osi='Windows XP' (manjaro article).
 
Old 10-22-2019, 05:12 AM   #35
Twigster
Member
 
Registered: Oct 2019
Location: France
Distribution: Slackware64 14.2
Posts: 34

Original Poster
Rep: Reputation: Disabled
The LILO Doc has to be the least clear one on character escaping (nothing at all about quotes)

I found by trial and error that the quotes themselves need to be escaped and not the space within, but I run into yet another issue :
Code:
root@:~# cat /etc/lilo.conf | grep append
append="vt.default_utf8=0 acpi_osi=\"Windows XP\""

root@lagann:~# dmesg | grep OSI
[    0.216428] ACPI: Added _OSI(Module Device)
[    0.216514] ACPI: Added _OSI(Processor Device)
[    0.216598] ACPI: Added _OSI(3.0 _SCP Extensions)
[    0.216684] ACPI: Added _OSI(Processor Aggregator Device)
[    0.216771] ACPI: Added _OSI(Windows XP)
This works, but the manjaro reference tells me it's not the correct _OSI string. No big deal, right?

Code:
cat /etc/lilo.conf | grep append
append="vt.default_utf8=0 acpi_osi=\"Windows 2001\""

dmesg | grep OSI 
[    0.217429] ACPI: Added _OSI(Module Device)
[    0.217514] ACPI: Added _OSI(Processor Device)
[    0.217599] ACPI: Added _OSI(3.0 _SCP Extensions)
[    0.217683] ACPI: Added _OSI(Processor Aggregator Device)
root@:
What the hell?
From /var/log/messages:
Code:
Oct 22 10:54:20 lagann kernel: [    0.000000] Command line: BOOT_IMAGE=Linux ro root=802 vt.default_utf8=0 acpi_osi="Windows 2001"
Where did my OSI string go?

Last edited by Twigster; 10-22-2019 at 09:57 AM.
 
Old 10-22-2019, 11:41 AM   #36
Richard Cranium
Senior Member
 
Registered: Apr 2009
Location: Carrollton, Texas
Distribution: Slackware64 14.2
Posts: 3,505

Rep: Reputation: 1836Reputation: 1836Reputation: 1836Reputation: 1836Reputation: 1836Reputation: 1836Reputation: 1836Reputation: 1836Reputation: 1836Reputation: 1836Reputation: 1836
Quote:
Originally Posted by Twigster View Post
Where did my OSI string go?
I suspect that your BIOS didn't respond to that OSI function, so the kernel couldn't add it.
 
1 members found this post helpful.
Old 10-22-2019, 01:02 PM   #37
abga
Senior Member
 
Registered: Jul 2017
Location: EU
Distribution: Slackware
Posts: 1,275

Rep: Reputation: 673Reputation: 673Reputation: 673Reputation: 673Reputation: 673Reputation: 673
Quote:
Originally Posted by Twigster View Post
The LILO Doc has to be the least clear one on character escaping (nothing at all about quotes)

I found by trial and error that the quotes themselves need to be escaped and not the space within, but I run into yet another issue :
Code:
root@:~# cat /etc/lilo.conf | grep append
append="vt.default_utf8=0 acpi_osi=\"Windows XP\""

root@lagann:~# dmesg | grep OSI
[    0.216428] ACPI: Added _OSI(Module Device)
[    0.216514] ACPI: Added _OSI(Processor Device)
[    0.216598] ACPI: Added _OSI(3.0 _SCP Extensions)
[    0.216684] ACPI: Added _OSI(Processor Aggregator Device)
[    0.216771] ACPI: Added _OSI(Windows XP)
This works, but the manjaro reference tells me it's not the correct _OSI string. No big deal, right?
I had no experience with space divided kernel boot parameters before and couldn't find a proper lilo doc source (missed the manual you provided), thus, considered that the quotes would do the job. With your experience now I also learned something. Thanks!

Where does the manjaro reference specify that the _OSI string isn't correct?
It only states:
Quote:
Unlike the Windows kernel, the Linux kernel can determine what “power calls” are sent to the bios by the use of these OSI Strings in the acpi kernel parameter. By adding these parameters, you can basically tell Linux to “mimic” the acpi calls sent by another OS.

When you purchased your laptop, it probably came with Windows pre-installed. The version of Windows that was installed is your first clue in figuring out which OSI String your acpi kernel parameter needs.
If you have a sticker on your laptop, displaying: "Designed for Windows XXX", then the Windows XXX version is the one the system manufacturer designed & tested the acpi implementation for (for example, the sticker might contain both XP & 2000). I suggested to use Windows XP because I recall the D620 I was using came with Windows XP preloaded and had that M$ sticker glued on it.

EDIT >
Actually the manjaro article lists valid OSI Strings in the table they present and for Windows XP the OSI String should be "Windows 2001".
I wanted to learn what the valid strings are and took a look in the kernel source code. Not really successful in identifying the valid list but found some references:
- an older patch with the rationale: "_OSI handling code grows giant and it's time to move them into one file"
https://lore.kernel.org/patchwork/patch/673265/
- and the only list of "Windows XXXX" strings I could find ATM in the source of 4.4.197 is contained in the blacklist.c file, however "Windows 2001" is not present.
https://git.kernel.org/pub/scm/linux...t.c?h=v4.4.197
- if you have more time, you could further inspect the source files from:
https://git.kernel.org/pub/scm/linux...cpi?h=v4.4.197
< EDIT

Should you fail to get your system stable with acpi_osi="Windows XXXX" and don't really care about the acpi PM calls, then try using acpi_osi=!
If your system is recognizing the video adapter and becomes stable, then you've got a solution ... finally

Last edited by abga; 10-22-2019 at 02:41 PM. Reason: EDIT
 
Old 10-22-2019, 03:06 PM   #38
enorbet
Senior Member
 
Registered: Jun 2003
Location: Virginia
Distribution: Slackware = Main OpSys for decades while testing others to keep up
Posts: 2,483

Rep: Reputation: 2528Reputation: 2528Reputation: 2528Reputation: 2528Reputation: 2528Reputation: 2528Reputation: 2528Reputation: 2528Reputation: 2528Reputation: 2528Reputation: 2528
Maybe it is just because we are still at the beginning of the digital age but there is a big difference between it and other electronics fields. For example I have some audiophile quality old tube gear from the 1950s that not only still works but is STILL audiophile in quality, yet recently (and much against my skinflint instincts) I tossed out several Socket 7 and Socket 370 laptops and PCs. They just have far too many restrictions and lacking features to effectively run modern software, especially web-related software. Now the oldest gear I have is a single core AMD FX-57 box and a Core 2 Duo T-60 Thinkpad laptop. The laptop cost me $90.00 USD used and I spent another 50 bucks on a high speed, large storage Seagate hard drive.

I actually sought out the T-60 because it's release was prior to the security-compromising inclusion of MCH. The version I got also has the benefit of a high performance nVidia Quadro graphics system. It is only a very short time newer than the Dell D620 but I didn't have to jump through any hoops, didn't have to research to discover any compromises or losses to get even a vanilla kernel working. The sole addon was simply the firmware blob for the wireless chip. which I rarely use.

Conclusive Suggestion and TLDR - It might just be time to upgrade your hardware, even if only a little bit.
 
Old 10-22-2019, 05:20 PM   #39
Twigster
Member
 
Registered: Oct 2019
Location: France
Distribution: Slackware64 14.2
Posts: 34

Original Poster
Rep: Reputation: Disabled
Quote:
I had no experience with space divided kernel boot parameters before and couldn't find a proper lilo doc source (missed the manual you provided), thus, considered that the quotes would do the job. With your experience now I also learned something. Thanks!
No worries, I just wish the manual had an example of quote escaping, or a sentence mentioning it =)

Quote:
If you have a sticker on your laptop, displaying: "Designed for Windows XXX", then the Windows XXX version is the one the system manufacturer designed & tested the acpi implementation for (for example, the sticker might contain both XP & 2000). I suggested to use Windows XP because I recall the D620 I was using came with Windows XP preloaded and had that M$ sticker glued on it.
You are correct, and I do have a Windows XP sticker on it.


Quote:
Should you fail to get your system stable with acpi_osi="Windows XXXX" and don't really care about the acpi PM calls, then try using acpi_osi=!
So If I do acpi_osi=!, the kernel says it's not Linux, nor Windows?
Not sure what I need to do; putting in the append acpi_osi=! or acpi_osi=!* didnt solve the freeze
Quote:
Maybe it is just because we are still at the beginning of the digital age but there is a big difference between it and other electronics fields. For example I have some audiophile quality old tube gear from the 1950s that not only still works but is STILL audiophile in quality, yet recently (and much against my skinflint instincts) I tossed out several Socket 7 and Socket 370 laptops and PCs. They just have far too many restrictions and lacking features to effectively run modern software, especially web-related software. Now the oldest gear I have is a single core AMD FX-57 box and a Core 2 Duo T-60 Thinkpad laptop. The laptop cost me $90.00 USD used and I spent another 50 bucks on a high speed, large storage Seagate hard drive.

I actually sought out the T-60 because it's release was prior to the security-compromising inclusion of MCH. The version I got also has the benefit of a high performance nVidia Quadro graphics system. It is only a very short time newer than the Dell D620 but I didn't have to jump through any hoops, didn't have to research to discover any compromises or losses to get even a vanilla kernel working. The sole addon was simply the firmware blob for the wireless chip. which I rarely use.

Conclusive Suggestion and TLDR - It might just be time to upgrade your hardware, even if only a little bit.
When I registered here, I went out of my usual ghost reading ways and took this investigation as a hobby because I have spare time. It is only because Steam dropped Windows XP support this year that I swapped to Linux. What makes this problem interesting is the digging and learning along the way. Good for you that you didn't have any hiccups with your T-60.

Last edited by Twigster; 10-22-2019 at 05:54 PM.
 
Old 10-22-2019, 05:58 PM   #40
abga
Senior Member
 
Registered: Jul 2017
Location: EU
Distribution: Slackware
Posts: 1,275

Rep: Reputation: 673Reputation: 673Reputation: 673Reputation: 673Reputation: 673Reputation: 673
Quote:
Originally Posted by Twigster View Post
So If I do acpi_osi=!, the kernel says it's not Linux, nor Windows?
Not sure what I need to do; simply putting in the append acpi_osi=! made dmesg print "disabling all OSI vendor strings" and then it still added Module Device, Processor, etc, and still crashed.
That's interesting, so disabling all specific built-in OS vendor strings (& functions/workarounds) still makes the system unstable.
You could go even further and try disabling all strings:
https://www.kernel.org/doc/Documenta...parameters.txt
acpi_osi= # disable all strings

In #35 you stated that you tried both acpi_osi="Windows XP" & acpi_osi="Windows 2001" and that "it worked". Was your system stable? If positive, then use them (one of them), end of story.
You mentioned acpi_osi="Windows 2001" was not showing up in the kernel log and I couldn't find "Windows 2001" in the kernel source files (blacklist.c) I presented the link for in my post #37 (nor "Windows XP"), but entries starting with "Windows 2006" - which according to the manjaro article should correspond for Windows Vista. This is why I suggested to further inspect the source files and look for valid parameters.
 
Old 10-22-2019, 05:59 PM   #41
Twigster
Member
 
Registered: Oct 2019
Location: France
Distribution: Slackware64 14.2
Posts: 34

Original Poster
Rep: Reputation: Disabled
Quote:
Originally Posted by Richard Cranium View Post
I suspect that your BIOS didn't respond to that OSI function, so the kernel couldn't add it.
Here is the microsoft source used in the manjaro link I got :
https://docs.microsoft.com/en-us/win...pi/winacpi-osi

What I deduce from this is that for the hardware to support Windows XP, It has to know the OSI string 'Windows 2001'. I find it hard to believe that there is no way in the Linux kernel to tell it to behave like a (old) widespread OS like Windows XP.

Last edited by Twigster; 10-22-2019 at 06:01 PM.
 
Old 10-22-2019, 07:01 PM   #42
Richard Cranium
Senior Member
 
Registered: Apr 2009
Location: Carrollton, Texas
Distribution: Slackware64 14.2
Posts: 3,505

Rep: Reputation: 1836Reputation: 1836Reputation: 1836Reputation: 1836Reputation: 1836Reputation: 1836Reputation: 1836Reputation: 1836Reputation: 1836Reputation: 1836Reputation: 1836
Quote:
Originally Posted by Twigster View Post
Here is the microsoft source used in the manjaro link I got :
https://docs.microsoft.com/en-us/win...pi/winacpi-osi

What I deduce from this is that for the hardware to support Windows XP, It has to know the OSI string 'Windows 2001'. I find it hard to believe that there is no way in the Linux kernel to tell it to behave like a (old) widespread OS like Windows XP.
Was that your log showing...

Code:
[    0.216771] ACPI: Added _OSI(Windows XP)
...because it would appear to me that "Windows XP" is the string the BIOS expects. I most certainly could be wrong.
 
Old 10-22-2019, 07:02 PM   #43
enorbet
Senior Member
 
Registered: Jun 2003
Location: Virginia
Distribution: Slackware = Main OpSys for decades while testing others to keep up
Posts: 2,483

Rep: Reputation: 2528Reputation: 2528Reputation: 2528Reputation: 2528Reputation: 2528Reputation: 2528Reputation: 2528Reputation: 2528Reputation: 2528Reputation: 2528Reputation: 2528
Quote:
Originally Posted by Twigster View Post
When I registered here, I went out of my usual ghost reading ways and took this investigation as a hobby because I have spare time. It is only because Steam dropped Windows XP support this year that I swapped to Linux. What makes this problem interesting is the digging and learning along the way. Good for you that you didn't have any hiccups with your T-60.
My post wasn't to toot my own horn but merely an FYI that if you found altering acpi functionality to be a serious loss I thought you might like to know it wouldn't be expensive to solve that. Best Wishes, brother.
 
Old 10-22-2019, 07:27 PM   #44
Twigster
Member
 
Registered: Oct 2019
Location: France
Distribution: Slackware64 14.2
Posts: 34

Original Poster
Rep: Reputation: Disabled
Quote:
Originally Posted by enorbet View Post
My post wasn't to toot my own horn but merely an FYI that if you found altering acpi functionality to be a serious loss I thought you might like to know it wouldn't be expensive to solve that. Best Wishes, brother.
I hear you, I bought the laptop for 50€ anyway, not much cheaper than yours =)
Quote:
Was that your log showing...

Code:
[ 0.216771] ACPI: Added _OSI(Windows XP)
...because it would appear to me that "Windows XP" is the string the BIOS expects. I most certainly could be wrong.
Oh, so I can tell the kernel to add any OSI string I like, but I will get a confirmation message only when the string exists?

Quote:
In #35 you stated that you tried both acpi_osi="Windows XP" & acpi_osi="Windows 2001" and that "it worked". Was your system stable? If positive, then use them (one of them), end of story.
I realize my sentence wasn't clear. What I meant was that I could see in dmesg that the kernel accepted the OSI string "Windows XP", however I could still produce the problematic freeze behaviour.

Last tests I've done :
Code:
:~# cat /etc/lilo.conf | grep append
append="vt.default_utf8=0 acpi_osi=! acpi_osi=\"Windows XP\""

~# dmesg | grep OSI
[    0.216423] ACPI: Disabled all _OSI OS vendors
[    0.216510] ACPI: Added _OSI(Module Device)
[    0.216593] ACPI: Added _OSI(Processor Device)
[    0.216678] ACPI: Added _OSI(3.0 _SCP Extensions)
[    0.216764] ACPI: Added _OSI(Processor Aggregator Device)
[    0.216851] ACPI: Added _OSI(Windows XP)
..Still freezes

Code:
~# dmesg | grep acpi_osi
[    0.000000] Command line: BOOT_IMAGE=Linux ro root=802 vt.default_utf8=0 acpi_osi=
[    0.000000] Kernel command line: BOOT_IMAGE=Linux ro root=802 vt.default_utf8=0 acpi_osi=
root@lagann:~# dmesg | grep OSI
[    0.000000] ACPI: _OSI method disabled
[    0.216423] ACPI: Added _OSI(Module Device)
[    0.216508] ACPI: Added _OSI(Processor Device)
[    0.216593] ACPI: Added _OSI(3.0 _SCP Extensions)
[    0.216679] ACPI: Added _OSI(Processor Aggregator Device)
.. Still freezes. Computer behaviour changed slightly though, much slower to boot/much more sluggish when displaying pages and pop ups. It boots so slow that I can read all the text messages before X starts. I've noticed some other ACPI messages in dmesg, but I'm not 100% sure they weren't there before. Here's the dmesg log (where you can see how slow it starts, even copypasting it took 2seconds). I've seen some acpi stuff, but I won't investigate tonight.

Last edited by Twigster; 10-22-2019 at 08:01 PM. Reason: I'm editing a million times as I collect information + I post on two computers
 
Old 10-22-2019, 08:11 PM   #45
Richard Cranium
Senior Member
 
Registered: Apr 2009
Location: Carrollton, Texas
Distribution: Slackware64 14.2
Posts: 3,505

Rep: Reputation: 1836Reputation: 1836Reputation: 1836Reputation: 1836Reputation: 1836Reputation: 1836Reputation: 1836Reputation: 1836Reputation: 1836Reputation: 1836Reputation: 1836
Quote:
Originally Posted by Twigster View Post
Oh, so I can tell the kernel to add any OSI string I like, but I will get a confirmation message only when the string exists?
That's what it looks like to me, but I haven't checked the kernel code. You'd think there would be a "I don't know what you're talking about" type of error message when you do provide a string that doesn't work, but I guess this way works as well. \_(ツ)_/
 
  


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
Problems with HP 3500c scanner and Dell D620 laptop runing Ubuntu 9.10 wa3fkg Linux - Laptop and Netbook 1 12-03-2009 10:37 AM
SuSE on Dell D620 laptop cvzyl SUSE / openSUSE 4 04-08-2008 07:53 AM
FC5 and Laptop Dell Latitude D620 zillah Linux - General 3 11-07-2006 12:51 AM
Dell d620 graphics support bagpussnz Linux - Laptop and Netbook 2 05-10-2006 04:04 PM

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

All times are GMT -5. The time now is 07:59 PM.

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
Facebook: linuxquestions Google+: linuxquestions
Open Source Consulting | Domain Registration