LinuxQuestions.org
Review your favorite Linux distribution.
Go Back   LinuxQuestions.org > Forums > Linux Forums > Linux - Distributions > Debian
User Name
Password
Debian This forum is for the discussion of Debian Linux.

Notices

Reply
 
Search this Thread
Old 03-22-2011, 06:15 PM   #1
xaos5
Member
 
Registered: Dec 2004
Distribution: debian and slackware
Posts: 217

Rep: Reputation: 31
Figuring out what the kernel is doing (modprobe r8169 freezes computer)?


I have a hardware configuration that causes the kernel to get 'stuck' on boot.

Whenever I place my pcie 1x card into the 16x slot it does this (I know the card works I've tested it in the 1x slot and with the second 1x network card I have).

How can I tell what the kernel is doing?

I've tried removing the 'quiet' option in grub and played with all of these settings:

noacpi nolacpi noapic=off pci=[basically every option]
only option that seemed to work was pci=off but it couldn't find my hard drive anymore to continue booting.

EDIT: GO HERE FOR LSPCI INFO and current problem.

Extra info:
I have 5 slots on the Motherboard 1 pcie-1x, 1 pcie-16x, 3 pci.

Last edited by xaos5; 04-04-2011 at 11:58 PM.
 
Old 03-23-2011, 03:03 PM   #2
business_kid
Guru
 
Registered: Jan 2006
Location: Ireland
Distribution: Slackware & Android
Posts: 6,170

Rep: Reputation: 526Reputation: 526Reputation: 526Reputation: 526Reputation: 526Reputation: 526
pcie-1.x and pcie-whatever-it-is-now probably have different voltage requirements. The higher the speed, the lower the voltage (Within limits). Also, I have found m/b slots to be much less backwards compatible than for example, ide cables. I would nopt expect your arrangement to work, and would start worrying about voltage feeding back into lower supply lines, which is not a good thing.
 
Old 03-27-2011, 11:03 PM   #3
xaos5
Member
 
Registered: Dec 2004
Distribution: debian and slackware
Posts: 217

Original Poster
Rep: Reputation: 31
PCI Express v1.0 and v2.0 are backwards compatible with each other. Voltage doesn't work like that either. If you over-volt a device (PCI-express card in this case), nothing should happen but the exploding chip could possibly cause voltage to flow through ground destroying everything else on that rail (I can only assume PC ICs aren't tolerant of such things). PCI Express 16x slots are suppose to be compatible with cards all the way down to 1x (like in my case) so hardware compatibility isn't any issue here unless its a chipset firmware limitation.

The kernel loads but it gets stuck at a certain part, I'd like to figure out what part that is.
 
Old 03-28-2011, 02:59 AM   #4
business_kid
Guru
 
Registered: Jan 2006
Location: Ireland
Distribution: Slackware & Android
Posts: 6,170

Rep: Reputation: 526Reputation: 526Reputation: 526Reputation: 526Reputation: 526Reputation: 526
Voltage does work like that - within limits. 8088s had a 5V core. Todays cpus have a 1.xV core.
Pci-1.x has 5V rating. Pci-2.x is 3.3V. Agp-4.x/*x is lower again (1.5V or something?). In fact a lot of tolerance is built in to the circuitry. Most interface points are open collector, and tolerant of higher voltage. You get to go faster in electronics by getting smaller components. That means lower voltages. When I started, there was 2.0micron chip fab. It's come down to 0.45microns, and you can be sure someone is building smaller, or already has smaller. That level of precision is called into question by background stuff, so fab facilities are extremely careful about thatĚ

But this matters because a bus may attempt to revert to pcie-1.x and not do so successfully.
 
Old 03-28-2011, 03:44 AM   #5
JZL240I-U
Senior Member
 
Registered: Apr 2003
Location: Germany
Distribution: openSuSE 13.1 / 12.3_64-KDE, Ubuntu 14.04, Fedora 20, Mint 17, Chakra
Posts: 3,644

Rep: Reputation: Disabled
You might try the function keys. <Ctrl><Alt><Fn> with n=9 shows kernel messages on my system. If that doesn't help boot from a live CD and dig into the files in /var/log on your hard disk.
 
Old 03-28-2011, 09:59 AM   #6
xaos5
Member
 
Registered: Dec 2004
Distribution: debian and slackware
Posts: 217

Original Poster
Rep: Reputation: 31
Quote:
Originally Posted by business_kid View Post
Voltage does work like that - within limits. 8088s had a 5V core. Todays cpus have a 1.xV core.
Pci-1.x has 5V rating. Pci-2.x is 3.3V. Agp-4.x/*x is lower again (1.5V or something?). In fact a lot of tolerance is built in to the circuitry. Most interface points are open collector, and tolerant of higher voltage. You get to go faster in electronics by getting smaller components. That means lower voltages. When I started, there was 2.0micron chip fab. It's come down to 0.45microns, and you can be sure someone is building smaller, or already has smaller. That level of precision is called into question by background stuff, so fab facilities are extremely careful about thatĚ

But this matters because a bus may attempt to revert to pcie-1.x and not do so successfully.
If your talking about the actual circuity on the card then yes I can see that running lower voltage, But as for the voltage going to the card its 12v for the supply (and card detect) and 3.3v for the SMBUS (I2C). See http://en.wikipedia.org/wiki/Pci_express#Pinout

Most ICs don't have a tolerance much higher than their supply voltage. If you supply more than its absolute maximum, its going to go boom.


Quote:
Originally Posted by JZL240I-U
You might try the function keys. <Ctrl><Alt><Fn> with n=9 shows kernel messages on my system. If that doesn't help boot from a live CD and dig into the files in /var/log on your hard disk.
I'll have to try this when I get home, do you know which /var/log/ to look at? I believe I already tried dmesg without any luck.
 
Old 03-28-2011, 10:05 AM   #7
JZL240I-U
Senior Member
 
Registered: Apr 2003
Location: Germany
Distribution: openSuSE 13.1 / 12.3_64-KDE, Ubuntu 14.04, Fedora 20, Mint 17, Chakra
Posts: 3,644

Rep: Reputation: Disabled
No. I just look at several of the newest from the latest booting. I always assume that there is something noteworthy to tell when the system logs something while I'm having problems .
 
Old 03-28-2011, 10:06 AM   #8
xaos5
Member
 
Registered: Dec 2004
Distribution: debian and slackware
Posts: 217

Original Poster
Rep: Reputation: 31
Quote:
Originally Posted by JZL240I-U View Post
No. I just look at several of the newest from the latest booting. I always assume that there is something noteworthy to tell when the system logs something while I'm having problems .
That's a good idea, never thought about sorting the logs by last modified!
 
Old 03-28-2011, 09:06 PM   #9
xaos5
Member
 
Registered: Dec 2004
Distribution: debian and slackware
Posts: 217

Original Poster
Rep: Reputation: 31
I've started looking through the logs and found this pertaining to the first Network card:
Code:
Mar 28 21:44:12 debian kernel: [    0.775810] r8169 0000:02:00.0: region #2 not an MMIO resource, aborting
Mar 28 21:44:12 debian kernel: [    0.776209] r8169 0000:02:00.0: PCI INT ? disabled
This happens when I boot with a network card in the 16x slot. The 1x slot network card doesn't work until I unplug it, boot, shutdown, plug it in, boot. Not really sure what to make out of this error though. It looks like the kernel doesn't get a chance to log anything to files before this happens...

lspci -vxx looks like this:
Code:
02:00.0 Unassigned class [ffff]: Realtek Semiconductor Co., Ltd. RTL8111/8168B PCI Express Gigabit Ethernet controller (rev ff) (prog-if ff)
	!!! Unknown header type 7f
00: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff
10: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff
20: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff
30: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff
 
Old 03-28-2011, 09:47 PM   #10
xaos5
Member
 
Registered: Dec 2004
Distribution: debian and slackware
Posts: 217

Original Poster
Rep: Reputation: 31
I've confirmed that it doesn't log anything when it I have that card in there as there isn't anything in /var/log/messages to that time.

After grub the it sits on a loading, please wait... message and never continues. I'm lost on what to do from here.
 
Old 03-29-2011, 02:55 AM   #11
business_kid
Guru
 
Registered: Jan 2006
Location: Ireland
Distribution: Slackware & Android
Posts: 6,170

Rep: Reputation: 526Reputation: 526Reputation: 526Reputation: 526Reputation: 526Reputation: 526
This sounds more like a bios thing.

Grub loading please wait... is before the kernel gets loaded. Ditto the shutdown requirement because the bios feels the basic devices (arms & legs of pc) on bootup.

Check the boot order, and take the network card out of it. Perhaps try configuring interrupts. In post #2 I did remark about m/b slots being less backwards compatible. This may be a case of it.
 
Old 03-29-2011, 09:08 PM   #12
xaos5
Member
 
Registered: Dec 2004
Distribution: debian and slackware
Posts: 217

Original Poster
Rep: Reputation: 31
I think I'm stuck trying to run a debugger on the kernel unless I can get it to log earlier in the boot process. Removing 'quiet' from the grub lines does show the kernel does load and do some stuff but it seems to get 'stuck'.
 
Old 03-30-2011, 03:07 AM   #13
business_kid
Guru
 
Registered: Jan 2006
Location: Ireland
Distribution: Slackware & Android
Posts: 6,170

Rep: Reputation: 526Reputation: 526Reputation: 526Reputation: 526Reputation: 526Reputation: 526
Quote:
Originally Posted by xaos5 View Post
Removing 'quiet' from the grub lines does show the kernel does load and do some stuff but it seems to get 'stuck'.
Excellent. Tell us where it gets stuck and we can make constructive comment.
 
Old 04-02-2011, 01:04 PM   #14
xaos5
Member
 
Registered: Dec 2004
Distribution: debian and slackware
Posts: 217

Original Poster
Rep: Reputation: 31
Quote:
Originally Posted by business_kid View Post
Excellent. Tell us where it gets stuck and we can make constructive comment.
Is there a way to output that to something beside trying to type it all manually here? Its does go through quite a few lines before it gets 'stuck'. I'll copy down the last few lines in a bit.
 
Old 04-02-2011, 01:15 PM   #15
xaos5
Member
 
Registered: Dec 2004
Distribution: debian and slackware
Posts: 217

Original Poster
Rep: Reputation: 31
Alright looks like it gets stuck after initializing the onboard network card.

Code:
[  0.753080] r8169 Gigabit Ethernet driver 2.3LK-NAPI loaded
[  0.753136] r8169 0000:02:00.0: PCI INT A -> GSI 18 (level, low) -> IRQ 18
[  0.753704] eth0: RTL8168d/8111d at 0xffffc90000c6e000, 00:0a:cd:1c:1f:25, XID 081000c0 IRQ 26
_
(Last line is a blinking cursor).
 
  


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
Need help figuring out how to get my network going. dragnrebrn1 Linux - Networking 2 02-07-2009 01:12 PM
Figuring Our Freeze Issues revof11 Linux - Hardware 1 05-11-2007 05:33 PM
Figuring out which shell? daihard Linux - Software 8 03-22-2006 08:47 PM
Having problems figuring out these commands Dandy Linux - Newbie 2 04-18-2004 10:21 PM
figuring out what ports are open joesbox Linux - Security 7 03-18-2003 04:44 PM


All times are GMT -5. The time now is 08:43 AM.

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