[SOLVED] dmesg output question

[mintage]

I kept seeing dmesg in troubleshooting posts, and tried to figure out what it does or how to use it.
When I enter:

Code: Select all

dmesg | grep -i usb
dmesg | grep -i dma
dmesg | grep -i memory

nothing happens.

When I enter:

Code: Select all

dmesg | less

it outputs hundreds of lines of a really colorful repeating error with PCEi bus Correcting id=00e5 like:

Code: Select all

[23761.292460] pcieport 0000:00:1c.5: AER: Corrected error received: id=00e5
[23761.293103] pcieport 0000:00:1c.5: can't find device of ID00e5
[23761.293120] pcieport 0000:00:1c.5: AER: Corrected error received: id=00e5
[23761.293127] pcieport 0000:00:1c.5: PCIe Bus Error: severity=Corrected, type=Physical Layer, id=00e5(Receiver ID)
[23761.293129] pcieport 0000:00:1c.5:   device [8086:9d15] error status/mask=00000001/00002000
[23761.293131] pcieport 0000:00:1c.5:    [ 0] Receiver Error

Is this a normal output for dmesg?

[all41]

Code: Select all

man dmesg

[WharfRat]

mintage,

It is surprising that the three dmesg|greps doesn't give you any results, The grep of usb returns 86 lines for me

Code: Select all

[bill@XPS] ~/Downloads $ dmesg | grep -i usb|wc -l
86

I also get results from the other two

In any event you might have run into a kernel bug, According to this post if you try pci=nomsi it might eliminate the error spam.

The error reported here is very similar and suggests adding pci=noaer to your kernel command line.

Good luck

[mintage]

I updated the kernel to 4.8.0-32. I am still getting this "error spam" i believe it is called. I don't know how to add things to the kernel command line, and just what to expect. I find there is so much information on this I just spend hours trying to find something relevant and understandable.

I do understand that between the /var/log/(syslog & kern.log) i have almost 16 GB of text of the error described above.

Code: Select all

Jan  1 11:44:43 mint-HP-Notebook kernel: [  321.570468] pcieport 0000:00:1c.5: AER: Multiple Corrected error received: id=00e5
Jan  1 11:44:43 mint-HP-Notebook kernel: [  321.571124] pcieport 0000:00:1c.5: PCIe Bus Error: severity=Corrected, type=Physical Layer, id=00e5(Receiver ID)
Jan  1 11:44:43 mint-HP-Notebook kernel: [  321.571130] pcieport 0000:00:1c.5:   device [8086:9d15] error status/mask=00000001/00002000
Jan  1 11:44:43 mint-HP-Notebook kernel: [  321.571134] pcieport 0000:00:1c.5:    [ 0] Receiver Error        
Jan  1 11:44:43 mint-HP-Notebook kernel: [  321.571143] pcieport 0000:00:1c.5: AER: Corrected error received: id=00e5
Jan  1 11:44:43 mint-HP-Notebook kernel: [  321.571787] pcieport 0000:00:1c.5: can't find device of ID00e5
Jan  1 11:44:43 mint-HP-Notebook kernel: [  321.571790] pcieport 0000:00:1c.5: AER: Corrected error received: id=00e5

I would like to present this to HP to see if they would have an idea what this means, but I don't know how to tell them what the problem is. Any suggestions would be nice.

[WharfRat]

I wouldn't know how to contact HP, but to add the kernel parameter you would edit a line in /etc/default/grub then update the grub menu.

Code: Select all

gksu xed /etc/default/grub&

Look for the line GRUB_CMDLINE_LINUX_DEFAULT= and whatever the contents are add pci=nomsi at the beginning or the end inside the quotes.

I don't remember the original content, but mine looks like this: GRUB_CMDLINE_LINUX_DEFAULT="pcie_aspm=off debug=vt"

I needed the pcie_aspm=off parameter to get this thing to boot.

After you save the file run sudo update-grub in the terminal and reboot.

If that parameter isn't any help then use the same procedure to try pci=noaer

Good luck

[mintage]

The Linux community is really the only place to get Linux answers it seems. So thank you for your input. I really am up a creek here. I read in the HP forum you aren't "a real Linux user until you've compiled your own kernel". (Ironic, coming from there.) I was kinda hoping to put that off for later.

Contacted HP. they have an HP Smart Friend program that costs 35USD to set up, 11.xxUSD/month with a 70+USD early termination fee if you cancel before the anniversary date and; by the way, they "only provide support for Microsoft Software."

I expected nothing less, but I was scared of the text I found in the etc/default/grub file.
I made a grubbkup just in case.
rebooting...
result...

Code: Select all

was:
GRUB_CMDLINE_LINUX_DEFAULT="quiet splash"
is:
GRUB_CMDLINE_LINUX_DEFAULT="quiet splash pci=nomsi"

sudo grub-update

dmesg is now green and probably coherent without the multi-color repeating error. JUBILATION! Thanks for your helpful suggestion WharfRat!

[WharfRat]

I read in the HP forum you aren't "a real Linux user until you've compiled your own kernel". (Ironic, coming from there.) I was kinda hoping to put that off for later.

Well you might as well get it over with then - it's not that hard.

Just kidding

Glad to hear that the kernel parameter worked for you

[mintage]

Laughing for a good while on that one. I see that kernel building is a distant prospect for me. Thank you again.

Referring to The size of the log files in /var/log gets out of hand to clear out about 20GB of repeating errors.

[WharfRat]

First try sudo /usr/sbin/logrotate -fv /etc/logrotate.conf to force a rotatation of the logs,

It's probably the error spam that caused it to swell so since that has been addressed it should be OK now.

[Flemur]

I read in the HP forum you aren't "a real Linux user until you've compiled your own kernel". (Ironic, coming from there.) I was kinda hoping to put that off for later.

Well you might as well get it over with then - it's not that hard.
Just kidding
Glad to hear that the kernel parameter worked for you

"Compiled my first kernel - it was easy." - by Me!
viewtopic.php?f=42&t=113329