You are not logged in.
We found the real issue here, the culprit is the e1000e module, put that module in blacklist and the problem should go away.
Thanks to the Intel people, they are awesome for helping. Thanks everyone for helping on this .
You can follow our progress here:
https://bugzilla.kernel.org/show_bug.cgi?id=33872P.S.: Hopefully they will fix that e1000e driver now.
Wow, thanks to you too Diego.
Offline
diegoviola wrote:We found the real issue here, the culprit is the e1000e module, put that module in blacklist and the problem should go away.
Thanks to the Intel people, they are awesome for helping. Thanks everyone for helping on this .
You can follow our progress here:
https://bugzilla.kernel.org/show_bug.cgi?id=33872P.S.: Hopefully they will fix that e1000e driver now.
Wow, thanks to you too Diego.
I found that I can't disable e1000e module because it doesn't even loaded, it's not listed in /proc/modules. So maybe I will try to experiment with other modules.
Offline
Deleted.
Last edited by diegoviola (2023-04-17 22:37:38)
Offline
I found that I can't disable e1000e module because it doesn't even loaded, it's not listed in /proc/modules. So maybe I will try to experiment with other modules.
Do you have an intel ethernet controller? What happens if you rmmod your ethernet driver?
Thanks,
Ryan
Offline
Do you have an intel ethernet controller? What happens if you rmmod your ethernet driver?
Thanks,
Ryan
I don't have intel ethernet controller, lspci shows this:
07:00.5 Ethernet controller: JMicron Technology Corp. JMC250 PCI Express Gigabit Ethernet Controller (rev 3)
laptop-mode was disabled and strangely, I failed to reproduce to bug (the system shuts down properly, lucky?). And after I re-enable laptop-mode, then I can reproduce the bug. So I rrmod the jme module (I googled for the module name of JMicron JMC250), but that doesn't fix the problem.
Offline
Deleted.
Last edited by diegoviola (2023-04-17 22:37:25)
Offline
anezch wrote:measure wrote:Do you have an intel ethernet controller? What happens if you rmmod your ethernet driver?
Thanks,
RyanI don't have intel ethernet controller, lspci shows this:
07:00.5 Ethernet controller: JMicron Technology Corp. JMC250 PCI Express Gigabit Ethernet Controller (rev 3)
laptop-mode was disabled and strangely, I failed to reproduce to bug (the system shuts down properly, lucky?). And after I re-enable laptop-mode, then I can reproduce the bug. So I rrmod the jme module (I googled for the module name of JMicron JMC250), but that doesn't fix the problem.
Have you tried blacklisting that driver, so it doesn't load the next time you boot your machine, not just rmmod'ing it? Please try that.
I added 'blacklist jme' to /etc/modprobe.d/modprobe.conf but my laptop still failed to poweroff
Offline
diegoviola wrote:anezch wrote:I don't have intel ethernet controller, lspci shows this:
07:00.5 Ethernet controller: JMicron Technology Corp. JMC250 PCI Express Gigabit Ethernet Controller (rev 3)
laptop-mode was disabled and strangely, I failed to reproduce to bug (the system shuts down properly, lucky?). And after I re-enable laptop-mode, then I can reproduce the bug. So I rrmod the jme module (I googled for the module name of JMicron JMC250), but that doesn't fix the problem.
Have you tried blacklisting that driver, so it doesn't load the next time you boot your machine, not just rmmod'ing it? Please try that.
I added 'blacklist jme' to /etc/modprobe.d/modprobe.conf but my laptop still failed to poweroff
At last! After trying several modules, now I am sure that ehci_hcd module caused the problem. After adding 'rmmod ehci_hcd' to /etc/rc.local.shutdown, my laptop now shuts down and reboots properly.
Offline
Deleted.
Last edited by diegoviola (2023-04-17 22:37:14)
Offline
Deleted.
Last edited by diegoviola (2023-04-17 22:36:17)
Offline
Strange, blacklisting the e1000e module makes my laptop shutdown fine, but I don't think this is a definitive fix, just a workaround.
I'm starting to think this is a problem with the Linux ACPI subsystem. I was expecting that a shutdown trace (which I already provided on the bugtracker) would help to detect issues. *sigh*
Could you tell me how to make the trace? I will make my shutdown trace and send it to the bugtracker. Yeah, I think there are some general issues that make some module prevent the system from halting. When you mention the Linux ACPI subsystem, I started to think the same. My laptop is about 10 C hotter after resumed from suspend even when the CPUs are idle.
Offline
Deleted.
Last edited by diegoviola (2023-04-17 22:36:07)
Offline
Sure. You can make the trace by passing the parameter 'initcall_debug' when you boot the kernel (in your bootloader). Then 'echo 8 > /proc/sys/kernel/printk' as root. Then shutdown, and it should print the trace in the shutdown state.
I'm not sure how to send that information to a text file tough, maybe you will have to take a picture of your screen and attach it into the tracker.
The trace should be similar to this: https://bugzilla.kernel.org/attachment.cgi?id=60532
Thanks.
Thanks Diego, I'll make it later because my phone cam is broken.
Offline
Deleted.
Last edited by diegoviola (2023-04-17 22:35:57)
Offline
Hopefully with patience we'll find the root of the cause of this problem and fix it once and for all.
Diego, it is clear that you have the best of intentions and only wish to help in determining the real problem. Some of your comments on the kernel bug tracker, however, are a bit counter productive. In particular, I am referring to Comment #64. I suggest reading http://www.catb.org/~esr/faqs/smart-questions.html, a link I found from ewaller's signature a while ago. After having read it, whenever I think about posting, it is always in the back of my mind.
For example, is this issue actually that annoying? I can imagine that you're frustrated because you've worked fairly hard to help in identifying it. How often do you reboot/poweroff your computer though? I generally only reboot when I upgrade my kernel (although I've rebooted quite a bit in testing this issue).
anezch has quite different hardware from many of the people who have posted here. It is not clear that he is experiencing the same issue we are, which does seem to be related to an intel driver. I can say I tested my ehci_hcd drivers the first time I saw the message, as IRQ #19 is the number for the USB requests, but this unfortunately had no effect on my system. I'm not saying that these issues aren't related, I'm just saying it's not clear that they are. The same symptom does not imply the same cause.
I really do write this to commend your hard work, but also to help you step back for a second.
Thanks,
Ryan
Offline
I tried blacklisting that module but it didn't work for me. Still stuck as "poweroff"
Tried the scripts in this thread as well, still stuck. Anyway around this? I have an Asus u52f.
EDIT: putting rmmod ehci_hcd in my rc.local.shutdown fixed it for me.
Last edited by bwat47 (2011-06-06 17:05:51)
Offline
Deleted.
Last edited by diegoviola (2023-04-17 22:35:46)
Offline
Deleted.
Last edited by diegoviola (2023-04-17 22:35:33)
Offline
[...]
For example, is this issue actually that annoying? I can imagine that you're frustrated because you've worked fairly hard to help in identifying it. How often do you reboot/poweroff your computer though? I generally only reboot when I upgrade my kernel (although I've rebooted quite a bit in testing this issue).
[...]
Imagine a script which initiates a shutdown because of hardware temperature problems but the computer fails to power off. I think this is very annoying.
Offline
anezch has quite different hardware from many of the people who have posted here. It is not clear that he is experiencing the same issue we are, which does seem to be related to an intel driver. I can say I tested my ehci_hcd drivers the first time I saw the message, as IRQ #19 is the number for the USB requests, but this unfortunately had no effect on my system. I'm not saying that these issues aren't related, I'm just saying it's not clear that they are. The same symptom does not imply the same cause.
Ryan,
Yes you are right, it appears that my problem is different than Diego's. Shaohua has confirmed this in the bug report.
Diego,
Please apologize me for a bunch of irrelevant posts to your report. I guess I was too careless and didn't think enough before posting them...
Thanks,
Agustianes
Offline
Deleted.
Last edited by diegoviola (2023-04-17 22:35:22)
Offline
I can confirm that adding rmmod ehci_hcd to /etc/rc.local.shutdown solves the shutdown problem for me at a Asus U35JC.
Offline
Deleted.
Last edited by diegoviola (2023-04-17 22:35:04)
Offline
I can confirm that adding rmmod ehci_hcd to /etc/rc.local.shutdown solves the shutdown problem for me at a Asus U35JC.
It was working for me too for about 3 days (approximately 6 poweroffs), but suddenly it hangs again. rmmod ehci_hcd in /etc/rc.local.shutdown doesn't really fix the problem, it just minimizing the chance of this bug to occur.
Now I have to do it manually from shell/terminal before poweroff to make sure there are enough interval between rrmod and acpi shutdown.
Offline
It was working for me too for about 3 days (approximately 6 poweroffs), but suddenly it hangs again. rmmod ehci_hcd in /etc/rc.local.shutdown doesn't really fix the problem, it just minimizing the chance of this bug to occur.
Now I have to do it manually from shell/terminal before poweroff to make sure there are enough interval between rrmod and acpi shutdown.
Ok that might be the case. I dont know in what order rc.local.shutdown enters the shutdown progress so maybe a little sleep doesnt help. Personally I would just ad some sleep in my shutdown alias but thats more of a work around.
Offline