lshtp-ish timed out

Issues related to hardware problems
someotherguy
Posts: 32
Joined: 2018/08/25 20:25:39

lshtp-ish timed out

Post by someotherguy » 2020/04/20 22:08:58

Multiple issues began when I installed a PCI SCSI interface card. Removing it has not fixed the problem(s). The system boots up, but the screen after the one showing multiple versions of Centos 7 (where you can choose which to run) is the usual "i8042 no controller found", followed by the new message "ishtp-ish timed out waiting for FW-initiated reset". This is followed by "ishtp ISH: HW failed to start" and then the usual message about ignoring an unsafe power cap. Eventually it goes into emergency mode, where I can either hit control D, whereupon I get the message "Authorization not available, check if polkit service is running or see debug message for more info", or I can enter a root password for maintenance, which gives me a root@Localhost prompt. Journalctl -u polkit.service -b yields "no entries". Systemctl status polkit.service shows it as enabled, but inactive(dead). Systemctl start polkit.service gives me the "authorization not available" message again.

Any suggestions greatly appreciated.

aks
Posts: 3073
Joined: 2014/09/20 11:22:14

Re: lshtp-ish timed out

Post by aks » 2020/04/21 06:41:20

Not really an answer.

But what is ish? https://www.kernel.org/doc/html/latest/ ... h-hid.html
Seemingly this will init your PCI, so I guess that's failing.
I think this is a kernel module - perhaps boot from media and blacklist it (assuming that hardware is not there?)

User avatar
TrevorH
Site Admin
Posts: 33202
Joined: 2009/09/24 10:40:56
Location: Brighton, UK

Re: lshtp-ish timed out

Post by TrevorH » 2020/04/21 08:43:51

Did you also add an entry to /etc/fstab that you forgot to remove? That's the most common and easiest way to enter emergency mode...
The future appears to be RHEL or Debian. I think I'm going Debian.
Info for USB installs on http://wiki.centos.org/HowTos/InstallFromUSBkey
CentOS 5 and 6 are deadest, do not use them.
Use the FAQ Luke

someotherguy
Posts: 32
Joined: 2018/08/25 20:25:39

Re: lshtp-ish timed out

Post by someotherguy » 2020/04/21 19:51:55

TrevorH wrote:
2020/04/21 08:43:51
Did you also add an entry to /etc/fstab that you forgot to remove? That's the most common and easiest way to enter emergency mode...
[root@localhost ~]# cat /etc/fstab
#several commented lines
/dev/mapper/centos/root / xfs defaults 0 0 UUID=9dedb8f1-885a-4baa-8496-a04d411d5c6b /boot xfs defaults 0 0
/dev mapper/centos-home /home xfs defaults 0 0
/dev mapper/centos-swap swap swap defaults 0 0
LABEL=datapartition /mnt/data ext 4 defaults 0 2

Anything in there look suspicious to you? It's mostly Greek to me.
In case it matters, my system is set up with as SSD card onto which the operating is installed (sda, if I recall) and a 4-disk RAID for data. I believe the /mnt/data partition refers to it.

User avatar
TrevorH
Site Admin
Posts: 33202
Joined: 2009/09/24 10:40:56
Location: Brighton, UK

Re: lshtp-ish timed out

Post by TrevorH » 2020/04/21 21:38:36

So that RAID array is not/was not hung off the PCI SCSI interface card that you installed?
The future appears to be RHEL or Debian. I think I'm going Debian.
Info for USB installs on http://wiki.centos.org/HowTos/InstallFromUSBkey
CentOS 5 and 6 are deadest, do not use them.
Use the FAQ Luke

someotherguy
Posts: 32
Joined: 2018/08/25 20:25:39

Re: lshtp-ish timed out

Post by someotherguy » 2020/04/22 03:07:07

TrevorH wrote:
2020/04/21 21:38:36
So that RAID array is not/was not hung off the PCI SCSI interface card that you installed?
No the RAID was there from the beginning, connected directly to SATA connectors on the motherboard (Supermicro X11SSA-F).
I put the SCSI interface card back in, and during boot-up I see it listed as "Adaptec AHA 2930CU V1.34.1". Looking at its specs on the Microsemi web site, it does not list Linux as a supported operating system. With it installed and me logged in in Recovery mode, running lsscsi shows the SSD as /dev/sda. the 4 HHD's as /dev/sdb to /dev/sde, and the cd/dvd as /dev/sr0. There is no mention of the SCSI interface card or the device attached to it. lspci shows it as SCSI Storage controller: Adaptec AHA2930CU (rev3). I don't really need it, but installed it more as an experiment than anything, but as I mentioned previously, removing it does not solve my problem.

My current focus is trying to enable / view debugging of ISH. As is probably obvious by now, I'm a relative newbie to Linux.

User avatar
TrevorH
Site Admin
Posts: 33202
Joined: 2009/09/24 10:40:56
Location: Brighton, UK

Re: lshtp-ish timed out

Post by TrevorH » 2020/04/22 09:06:44

Show us the console output from when it goes wrong. You may have to page up using Shift-PgUp to see the relevant error(s).
The future appears to be RHEL or Debian. I think I'm going Debian.
Info for USB installs on http://wiki.centos.org/HowTos/InstallFromUSBkey
CentOS 5 and 6 are deadest, do not use them.
Use the FAQ Luke

someotherguy
Posts: 32
Joined: 2018/08/25 20:25:39

Re: lshtp-ish timed out

Post by someotherguy » 2020/04/22 18:38:28

TrevorH wrote:
2020/04/22 09:06:44
Show us the console output from when it goes wrong. You may have to page up using Shift-PgUp to see the relevant error(s).
Not quite sure what you want here. I think the gist of what you are requesting is listed in my original message in this thread.

Once the system is finished doing its thing, the screen output says:
"Welcome to emergency mode! After logging in, type "journalctl -xb" to view system logs, "systemctl reboot" to reboot, "systemctl default" or ^D to try again to boot into default mode.
Give root password for maintenance
(Or press Control-D to continue)"

If I press control-D, the message comes up:
"Authorization not available. Check if polkit service is running or see debug message for more information" and the system apparently goes back into at least a partial reboot mode, with a progress bar at the bottom of the screen proceeding left to right with "Centos 7" at the right end of the bar. At the end of this the "welcome to emergency mode" screen as above comes up again.

If I type in the root password I get the prompt "[root @locahost ~]#"
If I then type in "journalctl -xb" I get a multiple line listing. I would post it here as an attachment, but I'm not aware of any mechanism I can use to get it from the server console to the desktop computer than I'm using here. Is there a way to perhaps copy the file to a USB stick which I could then sneaker net to this computer?

Scrolling through the journal -xb output, I come across a few lines listed in red that correspond to the transient messages showing up during the reboot process, namely:
{timestamp}localhost.localdomain kernel: i8042: no controller found
{timestamp}localhost.localdomain kernel: intel_ipc_0000:00:13.0. [ishtp-ish]: timed out waiting for FW-initiated reset.
{timestamp}localhost.localdomain kernel: intel_ipc_0000:00:13.0. [ishtp-ish]: hw start failed.
{timestamp}localhost.localdomain kernel: power_meter ACP1000D:00: Ignoring unsafe software power cap!

As mentioned previously, the first and last lines above have always displayed during normal startup, and therefore are probably irrelevant to my current problem. In contrast, the first appearance of the middle 2 lines seems to coincide with the development of my current problem

An additional red line in the journal -xb output, which I have never seen before is:
{timestamp}localhost.localdomain systemD[1]: Timed out waiting for device dev-disk-by\ x2dlabel-datapartitiondevice.device
This is followed be several lines relating to the fact that mounting the data partition on the RAID has failed.

I appreciate the time and effort you are expending to try to help me solve this problem.

someotherguy
Posts: 32
Joined: 2018/08/25 20:25:39

Re: lshtp-ish timed out

Post by someotherguy » 2020/04/22 19:51:42

aks wrote:
2020/04/21 06:41:20
Not really an answer.

But what is ish? https://www.kernel.org/doc/html/latest/ ... h-hid.html
Seemingly this will init your PCI, so I guess that's failing.
I think this is a kernel module - perhaps boot from media and blacklist it (assuming that hardware is not there?)
I looked at the link you posted, but it is way over my head. It refers multiple times to "FW" without defining what it stands for. I'd be interested in knowing since the acronym appears in my error message.

One thing I did get from the link relates to enabling debugging logs on ish. I entered "echo /sys/kernel/debug/tracing/events/intel_ish/enable cat /sys/kernel/debug/tracing/trace", then rebooted, logged on as root, and then entered "cat /sys/kernel/debug/tracing/trace", but there did not seem to be any entries in the file.

Regarding rebooting from media and black listing ish, at this point I'm willing to try just about anything, provided I can reverse it if necessary. I have a USB thumb drive that I made DOS-bootable so I could flash the motherboard BIOS, but I have no idea how to go about blacklisting a kernel module. Any chance you could point me in the right direction?

User avatar
TrevorH
Site Admin
Posts: 33202
Joined: 2009/09/24 10:40:56
Location: Brighton, UK

Re: lshtp-ish timed out

Post by TrevorH » 2020/04/22 20:07:38

When you first get to emergency mode, use Shift-PgUp to scroll back up the messages that came out before you got there. The error(s) will be listed there.
The future appears to be RHEL or Debian. I think I'm going Debian.
Info for USB installs on http://wiki.centos.org/HowTos/InstallFromUSBkey
CentOS 5 and 6 are deadest, do not use them.
Use the FAQ Luke

Post Reply