Since updates Saturday night I have been having hard lock ups. It looks like Firefox is the trigger.
Lenovo X140e notebook
F35 current with all updates as of Saturday night.
Xfce desktop
It was really bad at first. I finally managed to kill Firefox auto load pages by in a non-graphics session, watching what top was reporting going on in the graphics session and killing Firefox.
Then starting Firefox and closing all my windows as they open. Finally got it to start up with nothing to open. That was this morning. Things seemed to be working OK until now when I got to my Hilton hotel for the night and going through their miserable portal that FIrefox would not accept their cert, and then it locked again.
Just booted. Updated again and only got added:
Upgraded: alsa-sof-firmware-2.1.1-1.fc35.noarch mariadb-connector-c-3.2.7-1.fc35.x86_64 mariadb-connector-c-config-3.2.7-1.fc35.noarch pipewire-0.3.52-4.fc35.x86_64 pipewire-alsa-0.3.52-4.fc35.x86_64 pipewire-gstreamer-0.3.52-4.fc35.x86_64 pipewire-jack-audio-connection-kit-0.3.52-4.fc35.x86_64 pipewire-libs-0.3.52-4.fc35.x86_64 pipewire-pulseaudio-0.3.52-4.fc35.x86_64 pipewire-utils-0.3.52-4.fc35.x86_64 Skipped: python3-reportlab-3.6.10-1.fc35.x86_64
I can't tell where, in /var/log/messages the system locked and then I rebooted. I did not note the times closely.
Anyone hear anything like this? So far can't find anything in bugs.
And I should add that the cursor still moves around but no response to mouse clicks or keystrokes. And video is frozen.
On 6/27/22 00:47, Robert Moskowitz wrote:
Since updates Saturday night I have been having hard lock ups. It looks like Firefox is the trigger.
Lenovo X140e notebook
F35 current with all updates as of Saturday night.
Xfce desktop
It was really bad at first. I finally managed to kill Firefox auto load pages by in a non-graphics session, watching what top was reporting going on in the graphics session and killing Firefox.
Then starting Firefox and closing all my windows as they open. Finally got it to start up with nothing to open. That was this morning. Things seemed to be working OK until now when I got to my Hilton hotel for the night and going through their miserable portal that FIrefox would not accept their cert, and then it locked again.
Just booted. Updated again and only got added:
Upgraded: alsa-sof-firmware-2.1.1-1.fc35.noarch mariadb-connector-c-3.2.7-1.fc35.x86_64 mariadb-connector-c-config-3.2.7-1.fc35.noarch pipewire-0.3.52-4.fc35.x86_64 pipewire-alsa-0.3.52-4.fc35.x86_64 pipewire-gstreamer-0.3.52-4.fc35.x86_64 pipewire-jack-audio-connection-kit-0.3.52-4.fc35.x86_64 pipewire-libs-0.3.52-4.fc35.x86_64 pipewire-pulseaudio-0.3.52-4.fc35.x86_64 pipewire-utils-0.3.52-4.fc35.x86_64 Skipped: python3-reportlab-3.6.10-1.fc35.x86_64
I can't tell where, in /var/log/messages the system locked and then I rebooted. I did not note the times closely.
Anyone hear anything like this? So far can't find anything in bugs.
users mailing list -- users@lists.fedoraproject.org To unsubscribe send an email to users-leave@lists.fedoraproject.org Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org Do not reply to spam on the list, report it: https://pagure.io/fedora-infrastructure
On Mon, 27 Jun 2022 08:53:48 -0400 Robert Moskowitz wrote:
And I should add that the cursor still moves around but no response to mouse clicks or keystrokes. And video is frozen.
Every time I install a new version of fedora, I give the nouveau drivers a chance to see if they have improved, and within a week (usually within a couple of hours), the video freezes up and the system is hosed.
I then switch to the rpmfusion binary nvidia drivers and the problem vanishes (till the next new version of fedora :-).
I don't know if you have an nvidia card or are using the nouveau drivers, but if you are I'd switch to the nvidia drivers to see if that fixes it.
I forgot the command to see what my video driver is, but I recall it is not nvidia.
On 6/27/22 09:40, Tom Horsley wrote:
On Mon, 27 Jun 2022 08:53:48 -0400 Robert Moskowitz wrote:
And I should add that the cursor still moves around but no response to mouse clicks or keystrokes. And video is frozen.
Every time I install a new version of fedora, I give the nouveau drivers a chance to see if they have improved, and within a week (usually within a couple of hours), the video freezes up and the system is hosed.
I then switch to the rpmfusion binary nvidia drivers and the problem vanishes (till the next new version of fedora :-).
I don't know if you have an nvidia card or are using the nouveau drivers, but if you are I'd switch to the nvidia drivers to see if that fixes it. _______________________________________________ users mailing list -- users@lists.fedoraproject.org To unsubscribe send an email to users-leave@lists.fedoraproject.org Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org Do not reply to spam on the list, report it: https://pagure.io/fedora-infrastructure
On Mon, 27 Jun 2022 08:53:48 -0400 Robert Moskowitz rgm@htt-consult.com wrote:
And I should add that the cursor still moves around but no response to mouse clicks or keystrokes. And video is frozen.
On 6/27/22 00:47, Robert Moskowitz wrote:
Since updates Saturday night I have been having hard lock ups. It looks like Firefox is the trigger.
Lenovo X140e notebook
F35 current with all updates as of Saturday night.
Xfce desktop
It was really bad at first. I finally managed to kill Firefox auto load pages by in a non-graphics session, watching what top was reporting going on in the graphics session and killing Firefox.
Then starting Firefox and closing all my windows as they open. Finally got it to start up with nothing to open. That was this morning. Things seemed to be working OK until now when I got to my Hilton hotel for the night and going through their miserable portal that FIrefox would not accept their cert, and then it locked again.
[snip]
Just booted. Updated again and only got added: Anyone hear anything like this? So far can't find anything in bugs.
I recall seeing something like this on a site that had a bunch of javascript video players on a single page. Click on the page and every video activated at once. Firefox went crazy. When I would look in top I would see all kinds of threads demanding CPU and the parent thread pegging all the cores. I haven't seen it for a while, so either I haven't visited a page like that again, or the page got fixed, or firefox put in some kind of governor to limit the amout of CPU that a single page can use.
Might not be your problem, but another datapoint for you.
On 6/27/22 10:17, stan via users wrote:
On Mon, 27 Jun 2022 08:53:48 -0400 Robert Moskowitz rgm@htt-consult.com wrote:
And I should add that the cursor still moves around but no response to mouse clicks or keystrokes. And video is frozen.
On 6/27/22 00:47, Robert Moskowitz wrote:
Since updates Saturday night I have been having hard lock ups. It looks like Firefox is the trigger.
Lenovo X140e notebook
F35 current with all updates as of Saturday night.
Xfce desktop
It was really bad at first. I finally managed to kill Firefox auto load pages by in a non-graphics session, watching what top was reporting going on in the graphics session and killing Firefox.
Then starting Firefox and closing all my windows as they open. Finally got it to start up with nothing to open. That was this morning. Things seemed to be working OK until now when I got to my Hilton hotel for the night and going through their miserable portal that FIrefox would not accept their cert, and then it locked again.
[snip]
Just booted. Updated again and only got added: Anyone hear anything like this? So far can't find anything in bugs.
I recall seeing something like this on a site that had a bunch of javascript video players on a single page. Click on the page and every video activated at once. Firefox went crazy. When I would look in top I would see all kinds of threads demanding CPU and the parent thread pegging all the cores. I haven't seen it for a while, so either I haven't visited a page like that again, or the page got fixed, or firefox put in some kind of governor to limit the amout of CPU that a single page can use.
Yes. I have seen Firefox eat up so much memory, that I cannot suspend to disk, the system just hangs itself. If I have been working for a while and zram is being used, I close Firefox before suspending.
This is worst. Something is locking up everything and it started with an update, when I had not updated for about 2 weeks.
Might not be your problem, but another datapoint for you. _______________________________________________ users mailing list -- users@lists.fedoraproject.org To unsubscribe send an email to users-leave@lists.fedoraproject.org Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org Do not reply to spam on the list, report it: https://pagure.io/fedora-infrastructure
How did you reboot the machine?
Cursor still moving would seem to indicate paging possibly, or simply userspace stopping.
If you have another machine ssh into it and run "top" before trying what crashes and see if top stops or keeps responding. if it keeps responding, then what it says should be useful.
If ssh does not respond then see if you can ping it (if ping works this confirms paging and/or user space deadlock).
If I leave weather channel open (on either windows or linux) it will eventually use a significant amount of ram with firefox, I have not tried it with chrome. There is a pretty good chance that other web pages can do the same thing.
On Sun, Jun 26, 2022 at 11:48 PM Robert Moskowitz rgm@htt-consult.com wrote:
Since updates Saturday night I have been having hard lock ups. It looks like Firefox is the trigger.
Lenovo X140e notebook
F35 current with all updates as of Saturday night.
Xfce desktop
It was really bad at first. I finally managed to kill Firefox auto load pages by in a non-graphics session, watching what top was reporting going on in the graphics session and killing Firefox.
Then starting Firefox and closing all my windows as they open. Finally got it to start up with nothing to open. That was this morning. Things seemed to be working OK until now when I got to my Hilton hotel for the night and going through their miserable portal that FIrefox would not accept their cert, and then it locked again.
Just booted. Updated again and only got added:
Upgraded: alsa-sof-firmware-2.1.1-1.fc35.noarch mariadb-connector-c-3.2.7-1.fc35.x86_64 mariadb-connector-c-config-3.2.7-1.fc35.noarch pipewire-0.3.52-4.fc35.x86_64 pipewire-alsa-0.3.52-4.fc35.x86_64 pipewire-gstreamer-0.3.52-4.fc35.x86_64 pipewire-jack-audio-connection-kit-0.3.52-4.fc35.x86_64 pipewire-libs-0.3.52-4.fc35.x86_64 pipewire-pulseaudio-0.3.52-4.fc35.x86_64 pipewire-utils-0.3.52-4.fc35.x86_64 Skipped: python3-reportlab-3.6.10-1.fc35.x86_64
I can't tell where, in /var/log/messages the system locked and then I rebooted. I did not note the times closely.
Anyone hear anything like this? So far can't find anything in bugs.
users mailing list -- users@lists.fedoraproject.org To unsubscribe send an email to users-leave@lists.fedoraproject.org Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org Do not reply to spam on the list, report it: https://pagure.io/fedora-infrastructure
Sounds interesting. Will try this if nothing else gets fixed when I get home Friday.
On 6/27/22 10:44, Roger Heflin wrote:
How did you reboot the machine?
Cursor still moving would seem to indicate paging possibly, or simply userspace stopping.
If you have another machine ssh into it and run "top" before trying what crashes and see if top stops or keeps responding. if it keeps responding, then what it says should be useful.
If ssh does not respond then see if you can ping it (if ping works this confirms paging and/or user space deadlock).
If I leave weather channel open (on either windows or linux) it will eventually use a significant amount of ram with firefox, I have not tried it with chrome. There is a pretty good chance that other web pages can do the same thing.
On Sun, Jun 26, 2022 at 11:48 PM Robert Moskowitz rgm@htt-consult.com wrote:
Since updates Saturday night I have been having hard lock ups. It looks like Firefox is the trigger. Lenovo X140e notebook F35 current with all updates as of Saturday night. Xfce desktop It was really bad at first. I finally managed to kill Firefox auto load pages by in a non-graphics session, watching what top was reporting going on in the graphics session and killing Firefox. Then starting Firefox and closing all my windows as they open. Finally got it to start up with nothing to open. That was this morning. Things seemed to be working OK until now when I got to my Hilton hotel for the night and going through their miserable portal that FIrefox would not accept their cert, and then it locked again. Just booted. Updated again and only got added: Upgraded: alsa-sof-firmware-2.1.1-1.fc35.noarch mariadb-connector-c-3.2.7-1.fc35.x86_64 mariadb-connector-c-config-3.2.7-1.fc35.noarch pipewire-0.3.52-4.fc35.x86_64 pipewire-alsa-0.3.52-4.fc35.x86_64 pipewire-gstreamer-0.3.52-4.fc35.x86_64 pipewire-jack-audio-connection-kit-0.3.52-4.fc35.x86_64 pipewire-libs-0.3.52-4.fc35.x86_64 pipewire-pulseaudio-0.3.52-4.fc35.x86_64 pipewire-utils-0.3.52-4.fc35.x86_64 Skipped: python3-reportlab-3.6.10-1.fc35.x86_64 I can't tell where, in /var/log/messages the system locked and then I rebooted. I did not note the times closely. Anyone hear anything like this? So far can't find anything in bugs. _______________________________________________ users mailing list -- users@lists.fedoraproject.org To unsubscribe send an email to users-leave@lists.fedoraproject.org Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org Do not reply to spam on the list, report it: https://pagure.io/fedora-infrastructure
users mailing list --users@lists.fedoraproject.org To unsubscribe send an email tousers-leave@lists.fedoraproject.org Fedora Code of Conduct:https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines:https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives:https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org Do not reply to spam on the list, report it:https://pagure.io/fedora-infrastructure
Happened again. I submitted a bug:
https://bugzilla.redhat.com/show_bug.cgi?id=2101629
On 6/27/22 11:35, Robert Moskowitz wrote:
Sounds interesting. Will try this if nothing else gets fixed when I get home Friday.
On 6/27/22 10:44, Roger Heflin wrote:
How did you reboot the machine?
Cursor still moving would seem to indicate paging possibly, or simply userspace stopping.
If you have another machine ssh into it and run "top" before trying what crashes and see if top stops or keeps responding. if it keeps responding, then what it says should be useful.
If ssh does not respond then see if you can ping it (if ping works this confirms paging and/or user space deadlock).
If I leave weather channel open (on either windows or linux) it will eventually use a significant amount of ram with firefox, I have not tried it with chrome. There is a pretty good chance that other web pages can do the same thing.
On Sun, Jun 26, 2022 at 11:48 PM Robert Moskowitz rgm@htt-consult.com wrote:
Since updates Saturday night I have been having hard lock ups. It looks like Firefox is the trigger. Lenovo X140e notebook F35 current with all updates as of Saturday night. Xfce desktop It was really bad at first. I finally managed to kill Firefox auto load pages by in a non-graphics session, watching what top was reporting going on in the graphics session and killing Firefox. Then starting Firefox and closing all my windows as they open. Finally got it to start up with nothing to open. That was this morning. Things seemed to be working OK until now when I got to my Hilton hotel for the night and going through their miserable portal that FIrefox would not accept their cert, and then it locked again. Just booted. Updated again and only got added: Upgraded: alsa-sof-firmware-2.1.1-1.fc35.noarch mariadb-connector-c-3.2.7-1.fc35.x86_64 mariadb-connector-c-config-3.2.7-1.fc35.noarch pipewire-0.3.52-4.fc35.x86_64 pipewire-alsa-0.3.52-4.fc35.x86_64 pipewire-gstreamer-0.3.52-4.fc35.x86_64 pipewire-jack-audio-connection-kit-0.3.52-4.fc35.x86_64 pipewire-libs-0.3.52-4.fc35.x86_64 pipewire-pulseaudio-0.3.52-4.fc35.x86_64 pipewire-utils-0.3.52-4.fc35.x86_64 Skipped: python3-reportlab-3.6.10-1.fc35.x86_64 I can't tell where, in /var/log/messages the system locked and then I rebooted. I did not note the times closely. Anyone hear anything like this? So far can't find anything in bugs. _______________________________________________ users mailing list -- users@lists.fedoraproject.org To unsubscribe send an email to users-leave@lists.fedoraproject.org Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org Do not reply to spam on the list, report it: https://pagure.io/fedora-infrastructure
users mailing list --users@lists.fedoraproject.org To unsubscribe send an email tousers-leave@lists.fedoraproject.org Fedora Code of Conduct:https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines:https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives:https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org Do not reply to spam on the list, report it:https://pagure.io/fedora-infrastructure
users mailing list --users@lists.fedoraproject.org To unsubscribe send an email tousers-leave@lists.fedoraproject.org Fedora Code of Conduct:https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines:https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives:https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org Do not reply to spam on the list, report it:https://pagure.io/fedora-infrastructure
On 6/27/22 19:20, Robert Moskowitz wrote:
Happened again. I submitted a bug:
If you think it's the kernel, did you try booting the previous kernel to see if it still happens?
On 6/28/22 02:27, Samuel Sieb wrote:
On 6/27/22 19:20, Robert Moskowitz wrote:
Happened again. I submitted a bug:
If you think it's the kernel, did you try booting the previous kernel to see if it still happens?
Dah.....
Next time! I am in meetings this week and only will reboot if I have to.
Like another crash...
My mind is so focused on the work I am doing and the writing, that I am forgetting the obvious things to do.
Sigh. And thanks.
Just locked again.
Now running with kernel: 15.17.12.200
On 6/28/22 06:18, Robert Moskowitz wrote:
On 6/28/22 02:27, Samuel Sieb wrote:
On 6/27/22 19:20, Robert Moskowitz wrote:
Happened again. I submitted a bug:
If you think it's the kernel, did you try booting the previous kernel to see if it still happens?
Dah.....
Next time! I am in meetings this week and only will reboot if I have to.
Like another crash...
My mind is so focused on the work I am doing and the writing, that I am forgetting the obvious things to do.
Sigh. And thanks. _______________________________________________ users mailing list -- users@lists.fedoraproject.org To unsubscribe send an email to users-leave@lists.fedoraproject.org Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org Do not reply to spam on the list, report it: https://pagure.io/fedora-infrastructure
Samuel,
Well thanks. It is NOT the kernel as I am still locking up. But it is something else related to that update, as I did not get these lockups before. I have updated the bug report and sent a number of related selinux error details and messages
My video is radeon. That is showing in the error messages in /var/log/messages I included in the bug report.
Hopefully this gets solved soon. I can't go to a number of web sites like in the past without the lockups.
Sigh.
On 6/28/22 07:44, Robert Moskowitz wrote:
Just locked again.
Now running with kernel: 15.17.12.200
On 6/28/22 06:18, Robert Moskowitz wrote:
On 6/28/22 02:27, Samuel Sieb wrote:
On 6/27/22 19:20, Robert Moskowitz wrote:
Happened again. I submitted a bug:
If you think it's the kernel, did you try booting the previous kernel to see if it still happens?
Dah.....
Next time! I am in meetings this week and only will reboot if I have to.
Like another crash...
My mind is so focused on the work I am doing and the writing, that I am forgetting the obvious things to do.
Sigh. And thanks. _______________________________________________ users mailing list -- users@lists.fedoraproject.org To unsubscribe send an email to users-leave@lists.fedoraproject.org Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org Do not reply to spam on the list, report it: https://pagure.io/fedora-infrastructure
users mailing list -- users@lists.fedoraproject.org To unsubscribe send an email to users-leave@lists.fedoraproject.org Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org Do not reply to spam on the list, report it: https://pagure.io/fedora-infrastructure
On Tue, Jun 28, 2022 at 8:40 PM Robert Moskowitz rgm@htt-consult.com wrote:
Samuel,
Well thanks. It is NOT the kernel as I am still locking up. But it is something else related to that update, as I did not get these lockups before. I have updated the bug report and sent a number of related selinux error details and messages
My video is radeon. That is showing in the error messages in /var/log/messages I included in the bug report.
https://bugzilla.redhat.com/show_bug.cgi?id=2101629 has:ring test failures.
https://docs.oracle.com/cd/E50245_01/E63729/html/vmrns-radeon-ring-test-fail... gives::
*Workaround:* Ensure that the radeon module is loaded with the modeset parameter set to zero. To do this, add the following line in /etc/modprobe.conf:
options radeon modeset=0
Hopefully this gets solved soon. I can't go to a number of web sites like in the past without the lockups.
It could be a hardware failure. Laptops over 5 years old are not to be trusted.
Sigh.
On 6/28/22 07:44, Robert Moskowitz wrote:
Just locked again.
Now running with kernel: 15.17.12.200
On 6/28/22 06:18, Robert Moskowitz wrote:
On 6/28/22 02:27, Samuel Sieb wrote:
On 6/27/22 19:20, Robert Moskowitz wrote:
Happened again. I submitted a bug:
If you think it's the kernel, did you try booting the previous kernel to see if it still happens?
Dah.....
Next time! I am in meetings this week and only will reboot if I have to.
Like another crash...
My mind is so focused on the work I am doing and the writing, that I am forgetting the obvious things to do.
Sigh. And thanks. _______________________________________________ users mailing list -- users@lists.fedoraproject.org To unsubscribe send an email to users-leave@lists.fedoraproject.org Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives:
https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org
Do not reply to spam on the list, report it: https://pagure.io/fedora-infrastructure
users mailing list -- users@lists.fedoraproject.org To unsubscribe send an email to users-leave@lists.fedoraproject.org Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives:
https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org
Do not reply to spam on the list, report it: https://pagure.io/fedora-infrastructure
users mailing list -- users@lists.fedoraproject.org To unsubscribe send an email to users-leave@lists.fedoraproject.org Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org Do not reply to spam on the list, report it: https://pagure.io/fedora-infrastructure
On 6/28/22 20:52, George N. White III wrote:
On Tue, Jun 28, 2022 at 8:40 PM Robert Moskowitz rgm@htt-consult.com wrote:
Samuel, Well thanks. It is NOT the kernel as I am still locking up. But it is something else related to that update, as I did not get these lockups before. I have updated the bug report and sent a number of related selinux error details and messages My video is radeon. That is showing in the error messages in /var/log/messages I included in the bug report.
https://bugzilla.redhat.com/show_bug.cgi?id=2101629 has:ring test failures.
https://docs.oracle.com/cd/E50245_01/E63729/html/vmrns-radeon-ring-test-fail... gives::
*Workaround:* Ensure that the radeon module is loaded with the modeset parameter set to zero. To do this, add the following line in |/etc/modprobe.conf|:
options radeon modeset=0
Only way to get this operational is reboot? I have this foggy memory about someway to force modprobe changes?
Hopefully this gets solved soon. I can't go to a number of web sites like in the past without the lockups.
It could be a hardware failure. Laptops over 5 years old are not to be trusted.
I was kind of wondering about that. At home I have another x140e that has my old F31 install. I am considering swapping drives (and memory) and seeing how this install behaves on the other notebook.
Sigh. On 6/28/22 07:44, Robert Moskowitz wrote: > Just locked again. > > Now running with kernel: 15.17.12.200 > > On 6/28/22 06:18, Robert Moskowitz wrote: >> >> >> On 6/28/22 02:27, Samuel Sieb wrote: >>> On 6/27/22 19:20, Robert Moskowitz wrote: >>>> Happened again. I submitted a bug: >>>> >>>> https://bugzilla.redhat.com/show_bug.cgi?id=2101629 >>> >>> If you think it's the kernel, did you try booting the previous >>> kernel to see if it still happens? >> >> >> Dah..... >> >> Next time! I am in meetings this week and only will reboot if I have >> to. >> >> Like another crash... >> >> My mind is so focused on the work I am doing and the writing, that I >> am forgetting the obvious things to do. >> >> Sigh. And thanks. >> _______________________________________________ >> users mailing list -- users@lists.fedoraproject.org >> To unsubscribe send an email to users-leave@lists.fedoraproject.org >> Fedora Code of Conduct: >> https://docs.fedoraproject.org/en-US/project/code-of-conduct/ >> List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines >> List Archives: >> https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org >> Do not reply to spam on the list, report it: >> https://pagure.io/fedora-infrastructure > _______________________________________________ > users mailing list -- users@lists.fedoraproject.org > To unsubscribe send an email to users-leave@lists.fedoraproject.org > Fedora Code of Conduct: > https://docs.fedoraproject.org/en-US/project/code-of-conduct/ > List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines > List Archives: > https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org > Do not reply to spam on the list, report it: > https://pagure.io/fedora-infrastructure _______________________________________________ users mailing list -- users@lists.fedoraproject.org To unsubscribe send an email to users-leave@lists.fedoraproject.org Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org Do not reply to spam on the list, report it: https://pagure.io/fedora-infrastructure
-- George N. White III
users mailing list --users@lists.fedoraproject.org To unsubscribe send an email tousers-leave@lists.fedoraproject.org Fedora Code of Conduct:https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines:https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives:https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org Do not reply to spam on the list, report it:https://pagure.io/fedora-infrastructure
On 6/28/22 20:52, George N. White III wrote:
On Tue, Jun 28, 2022 at 8:40 PM Robert Moskowitz rgm@htt-consult.com wrote:
Samuel, Well thanks. It is NOT the kernel as I am still locking up. But it is something else related to that update, as I did not get these lockups before. I have updated the bug report and sent a number of related selinux error details and messages My video is radeon. That is showing in the error messages in /var/log/messages I included in the bug report.
https://bugzilla.redhat.com/show_bug.cgi?id=2101629 has:ring test failures.
https://docs.oracle.com/cd/E50245_01/E63729/html/vmrns-radeon-ring-test-fail... gives::
*Workaround:* Ensure that the radeon module is loaded with the modeset parameter set to zero. To do this, add the following line in |/etc/modprobe.conf|:
options radeon modeset=0
This does not change anything. I crashed and rebooted with this line in modprobe.conf.
Got into firefox and cnn and lock.
Restarted.
Had nothing really running (no thunderbirds) got firefox going. lock.
I looked to see if chrome is available to try it, but don't see it in the repo.
Any other browser?
Of course it should not be the browser, as all that is happening is some function is being called that crashes the radeon driver.
Hopefully this gets solved soon. I can't go to a number of web sites like in the past without the lockups.
It could be a hardware failure. Laptops over 5 years old are not to be trusted.
Sigh. On 6/28/22 07:44, Robert Moskowitz wrote: > Just locked again. > > Now running with kernel: 15.17.12.200 > > On 6/28/22 06:18, Robert Moskowitz wrote: >> >> >> On 6/28/22 02:27, Samuel Sieb wrote: >>> On 6/27/22 19:20, Robert Moskowitz wrote: >>>> Happened again. I submitted a bug: >>>> >>>> https://bugzilla.redhat.com/show_bug.cgi?id=2101629 >>> >>> If you think it's the kernel, did you try booting the previous >>> kernel to see if it still happens? >> >> >> Dah..... >> >> Next time! I am in meetings this week and only will reboot if I have >> to. >> >> Like another crash... >> >> My mind is so focused on the work I am doing and the writing, that I >> am forgetting the obvious things to do. >> >> Sigh. And thanks. >> _______________________________________________ >> users mailing list -- users@lists.fedoraproject.org >> To unsubscribe send an email to users-leave@lists.fedoraproject.org >> Fedora Code of Conduct: >> https://docs.fedoraproject.org/en-US/project/code-of-conduct/ >> List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines >> List Archives: >> https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org >> Do not reply to spam on the list, report it: >> https://pagure.io/fedora-infrastructure > _______________________________________________ > users mailing list -- users@lists.fedoraproject.org > To unsubscribe send an email to users-leave@lists.fedoraproject.org > Fedora Code of Conduct: > https://docs.fedoraproject.org/en-US/project/code-of-conduct/ > List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines > List Archives: > https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org > Do not reply to spam on the list, report it: > https://pagure.io/fedora-infrastructure _______________________________________________ users mailing list -- users@lists.fedoraproject.org To unsubscribe send an email to users-leave@lists.fedoraproject.org Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org Do not reply to spam on the list, report it: https://pagure.io/fedora-infrastructure
-- George N. White III
users mailing list --users@lists.fedoraproject.org To unsubscribe send an email tousers-leave@lists.fedoraproject.org Fedora Code of Conduct:https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines:https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives:https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org Do not reply to spam on the list, report it:https://pagure.io/fedora-infrastructure
What kind of video card/builtin do you have? (lspci | grep -i vga)
Only really old amd/ati cards use the radeon driver. The newer cards use amdgpu. And your dmesg lists your bios as being a 2020 bios release and that would indicate the hardware you are using is new so makes me wonder why you have a radeon driver in use.
On Wed, Jun 29, 2022 at 6:08 PM Robert Moskowitz rgm@htt-consult.com wrote:
On 6/28/22 20:52, George N. White III wrote:
On Tue, Jun 28, 2022 at 8:40 PM Robert Moskowitz rgm@htt-consult.com wrote:
Samuel,
Well thanks. It is NOT the kernel as I am still locking up. But it is something else related to that update, as I did not get these lockups before. I have updated the bug report and sent a number of related selinux error details and messages
My video is radeon. That is showing in the error messages in /var/log/messages I included in the bug report.
https://bugzilla.redhat.com/show_bug.cgi?id=2101629 has:ring test failures.
https://docs.oracle.com/cd/E50245_01/E63729/html/vmrns-radeon-ring-test-fail... gives::
Workaround: Ensure that the radeon module is loaded with the modeset parameter set to zero. To do this, add the following line in /etc/modprobe.conf:
options radeon modeset=0
This does not change anything. I crashed and rebooted with this line in modprobe.conf.
Got into firefox and cnn and lock.
Restarted.
Had nothing really running (no thunderbirds) got firefox going. lock.
I looked to see if chrome is available to try it, but don't see it in the repo.
Any other browser?
Of course it should not be the browser, as all that is happening is some function is being called that crashes the radeon driver.
Hopefully this gets solved soon. I can't go to a number of web sites like in the past without the lockups.
It could be a hardware failure. Laptops over 5 years old are not to be trusted.
Sigh.
On 6/28/22 07:44, Robert Moskowitz wrote:
Just locked again.
Now running with kernel: 15.17.12.200
On 6/28/22 06:18, Robert Moskowitz wrote:
On 6/28/22 02:27, Samuel Sieb wrote:
On 6/27/22 19:20, Robert Moskowitz wrote:
Happened again. I submitted a bug:
If you think it's the kernel, did you try booting the previous kernel to see if it still happens?
Dah.....
Next time! I am in meetings this week and only will reboot if I have to.
Like another crash...
My mind is so focused on the work I am doing and the writing, that I am forgetting the obvious things to do.
Sigh. And thanks. _______________________________________________ users mailing list -- users@lists.fedoraproject.org To unsubscribe send an email to users-leave@lists.fedoraproject.org Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org Do not reply to spam on the list, report it: https://pagure.io/fedora-infrastructure
users mailing list -- users@lists.fedoraproject.org To unsubscribe send an email to users-leave@lists.fedoraproject.org Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org Do not reply to spam on the list, report it: https://pagure.io/fedora-infrastructure
users mailing list -- users@lists.fedoraproject.org To unsubscribe send an email to users-leave@lists.fedoraproject.org Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org Do not reply to spam on the list, report it: https://pagure.io/fedora-infrastructure
-- George N. White III
users mailing list -- users@lists.fedoraproject.org To unsubscribe send an email to users-leave@lists.fedoraproject.org Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org Do not reply to spam on the list, report it: https://pagure.io/fedora-infrastructure
users mailing list -- users@lists.fedoraproject.org To unsubscribe send an email to users-leave@lists.fedoraproject.org Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org Do not reply to spam on the list, report it: https://pagure.io/fedora-infrastructure
lspci -nnk | egrep -i "VGA|DISPLAY" -A3
and see if the card reports amdgpu could also work for the card. I am going to guess amdgpu can be used, my 2010 vintage FM2 amd can apparently use either driver also.
On Wed, Jun 29, 2022 at 6:49 PM Roger Heflin rogerheflin@gmail.com wrote:
What kind of video card/builtin do you have? (lspci | grep -i vga)
Only really old amd/ati cards use the radeon driver. The newer cards use amdgpu. And your dmesg lists your bios as being a 2020 bios release and that would indicate the hardware you are using is new so makes me wonder why you have a radeon driver in use.
On Wed, Jun 29, 2022 at 6:08 PM Robert Moskowitz rgm@htt-consult.com wrote:
On 6/28/22 20:52, George N. White III wrote:
On Tue, Jun 28, 2022 at 8:40 PM Robert Moskowitz rgm@htt-consult.com wrote:
Samuel,
Well thanks. It is NOT the kernel as I am still locking up. But it is something else related to that update, as I did not get these lockups before. I have updated the bug report and sent a number of related selinux error details and messages
My video is radeon. That is showing in the error messages in /var/log/messages I included in the bug report.
https://bugzilla.redhat.com/show_bug.cgi?id=2101629 has:ring test failures.
https://docs.oracle.com/cd/E50245_01/E63729/html/vmrns-radeon-ring-test-fail... gives::
Workaround: Ensure that the radeon module is loaded with the modeset parameter set to zero. To do this, add the following line in /etc/modprobe.conf:
options radeon modeset=0
This does not change anything. I crashed and rebooted with this line in modprobe.conf.
Got into firefox and cnn and lock.
Restarted.
Had nothing really running (no thunderbirds) got firefox going. lock.
I looked to see if chrome is available to try it, but don't see it in the repo.
Any other browser?
Of course it should not be the browser, as all that is happening is some function is being called that crashes the radeon driver.
Hopefully this gets solved soon. I can't go to a number of web sites like in the past without the lockups.
It could be a hardware failure. Laptops over 5 years old are not to be trusted.
Sigh.
On 6/28/22 07:44, Robert Moskowitz wrote:
Just locked again.
Now running with kernel: 15.17.12.200
On 6/28/22 06:18, Robert Moskowitz wrote:
On 6/28/22 02:27, Samuel Sieb wrote:
On 6/27/22 19:20, Robert Moskowitz wrote: > Happened again. I submitted a bug: > > https://bugzilla.redhat.com/show_bug.cgi?id=2101629
If you think it's the kernel, did you try booting the previous kernel to see if it still happens?
Dah.....
Next time! I am in meetings this week and only will reboot if I have to.
Like another crash...
My mind is so focused on the work I am doing and the writing, that I am forgetting the obvious things to do.
Sigh. And thanks. _______________________________________________ users mailing list -- users@lists.fedoraproject.org To unsubscribe send an email to users-leave@lists.fedoraproject.org Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org Do not reply to spam on the list, report it: https://pagure.io/fedora-infrastructure
users mailing list -- users@lists.fedoraproject.org To unsubscribe send an email to users-leave@lists.fedoraproject.org Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org Do not reply to spam on the list, report it: https://pagure.io/fedora-infrastructure
users mailing list -- users@lists.fedoraproject.org To unsubscribe send an email to users-leave@lists.fedoraproject.org Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org Do not reply to spam on the list, report it: https://pagure.io/fedora-infrastructure
-- George N. White III
users mailing list -- users@lists.fedoraproject.org To unsubscribe send an email to users-leave@lists.fedoraproject.org Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org Do not reply to spam on the list, report it: https://pagure.io/fedora-infrastructure
users mailing list -- users@lists.fedoraproject.org To unsubscribe send an email to users-leave@lists.fedoraproject.org Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org Do not reply to spam on the list, report it: https://pagure.io/fedora-infrastructure
On 6/29/22 20:13, Roger Heflin wrote:
lspci -nnk | egrep -i "VGA|DISPLAY" -A3
lspci -nnk | egrep -i "VGA|DISPLAY" -A3 00:01.0 VGA compatible controller [0300]: Advanced Micro Devices, Inc. [AMD/ATI] Kabini [Radeon HD 8330] [1002:9832] Subsystem: Lenovo Device [17aa:2219] Kernel driver in use: radeon Kernel modules: radeon, amdgpu
and see if the card reports amdgpu could also work for the card. I am going to guess amdgpu can be used, my 2010 vintage FM2 amd can apparently use either driver also.
So how do I switch to the amdgpu driver?
On Wed, Jun 29, 2022 at 6:49 PM Roger Heflin rogerheflin@gmail.com wrote:
What kind of video card/builtin do you have? (lspci | grep -i vga)
Only really old amd/ati cards use the radeon driver. The newer cards use amdgpu. And your dmesg lists your bios as being a 2020 bios release and that would indicate the hardware you are using is new so makes me wonder why you have a radeon driver in use.
On Wed, Jun 29, 2022 at 6:08 PM Robert Moskowitz rgm@htt-consult.com wrote:
On 6/28/22 20:52, George N. White III wrote:
On Tue, Jun 28, 2022 at 8:40 PM Robert Moskowitz rgm@htt-consult.com wrote:
Samuel,
Well thanks. It is NOT the kernel as I am still locking up. But it is something else related to that update, as I did not get these lockups before. I have updated the bug report and sent a number of related selinux error details and messages
My video is radeon. That is showing in the error messages in /var/log/messages I included in the bug report.
https://bugzilla.redhat.com/show_bug.cgi?id=2101629 has:ring test failures.
https://docs.oracle.com/cd/E50245_01/E63729/html/vmrns-radeon-ring-test-fail... gives::
Workaround: Ensure that the radeon module is loaded with the modeset parameter set to zero. To do this, add the following line in /etc/modprobe.conf:
options radeon modeset=0
This does not change anything. I crashed and rebooted with this line in modprobe.conf.
Got into firefox and cnn and lock.
Restarted.
Had nothing really running (no thunderbirds) got firefox going. lock.
I looked to see if chrome is available to try it, but don't see it in the repo.
Any other browser?
Of course it should not be the browser, as all that is happening is some function is being called that crashes the radeon driver.
Hopefully this gets solved soon. I can't go to a number of web sites like in the past without the lockups.
It could be a hardware failure. Laptops over 5 years old are not to be trusted.
Sigh.
On 6/28/22 07:44, Robert Moskowitz wrote:
Just locked again.
Now running with kernel: 15.17.12.200
On 6/28/22 06:18, Robert Moskowitz wrote:
On 6/28/22 02:27, Samuel Sieb wrote: > On 6/27/22 19:20, Robert Moskowitz wrote: >> Happened again. I submitted a bug: >> >> https://bugzilla.redhat.com/show_bug.cgi?id=2101629 > If you think it's the kernel, did you try booting the previous > kernel to see if it still happens?
Dah.....
Next time! I am in meetings this week and only will reboot if I have to.
Like another crash...
My mind is so focused on the work I am doing and the writing, that I am forgetting the obvious things to do.
Sigh. And thanks. _______________________________________________ users mailing list -- users@lists.fedoraproject.org To unsubscribe send an email to users-leave@lists.fedoraproject.org Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org Do not reply to spam on the list, report it: https://pagure.io/fedora-infrastructure
users mailing list -- users@lists.fedoraproject.org To unsubscribe send an email to users-leave@lists.fedoraproject.org Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org Do not reply to spam on the list, report it: https://pagure.io/fedora-infrastructure
users mailing list -- users@lists.fedoraproject.org To unsubscribe send an email to users-leave@lists.fedoraproject.org Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org Do not reply to spam on the list, report it: https://pagure.io/fedora-infrastructure
-- George N. White III
users mailing list -- users@lists.fedoraproject.org To unsubscribe send an email to users-leave@lists.fedoraproject.org Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org Do not reply to spam on the list, report it: https://pagure.io/fedora-infrastructure
users mailing list -- users@lists.fedoraproject.org To unsubscribe send an email to users-leave@lists.fedoraproject.org Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org Do not reply to spam on the list, report it: https://pagure.io/fedora-infrastructure
users mailing list -- users@lists.fedoraproject.org To unsubscribe send an email to users-leave@lists.fedoraproject.org Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org Do not reply to spam on the list, report it: https://pagure.io/fedora-infrastructure
The notes I found say to create a file in /etc/modprobe.d
And say call it anything but something like enable_amdgpu_disable_radeon.conf and put this in it: options amdgpu si_support=1 options amdgpu cik_support=1 options radeon si_support=0 options radeon cik_support=0
and then do a "dracut --force" and reboot.
Make sure there is not a new kernel, as the dracut will built the current booted kernel and not the new/other kernels. So make sure you boot back into the same kernel.
adding --regenerate-all on the dracut command line will cause it to rebuild all kernel initramfs, I would not do that until you determine this one works better than the prior one.
si/cik are AMD/ATI graphics cards, The SI and CIK families can use either driver, default enable is radeon, disable on amdgpu.
I have not rebooted my machine that is using the radeon driver yet (it is a main server so it does not get booted often), so I don't know if that is all that is needed for it to work.
On Wed, Jun 29, 2022 at 7:38 PM Robert Moskowitz rgm@htt-consult.com wrote:
On 6/29/22 20:13, Roger Heflin wrote:
lspci -nnk | egrep -i "VGA|DISPLAY" -A3
lspci -nnk | egrep -i "VGA|DISPLAY" -A3 00:01.0 VGA compatible controller [0300]: Advanced Micro Devices, Inc. [AMD/ATI] Kabini [Radeon HD 8330] [1002:9832] Subsystem: Lenovo Device [17aa:2219] Kernel driver in use: radeon Kernel modules: radeon, amdgpu
and see if the card reports amdgpu could also work for the card. I am going to guess amdgpu can be used, my 2010 vintage FM2 amd can apparently use either driver also.
So how do I switch to the amdgpu driver?
On Wed, Jun 29, 2022 at 6:49 PM Roger Heflin rogerheflin@gmail.com wrote:
What kind of video card/builtin do you have? (lspci | grep -i vga)
Only really old amd/ati cards use the radeon driver. The newer cards use amdgpu. And your dmesg lists your bios as being a 2020 bios release and that would indicate the hardware you are using is new so makes me wonder why you have a radeon driver in use.
On Wed, Jun 29, 2022 at 6:08 PM Robert Moskowitz rgm@htt-consult.com wrote:
On 6/28/22 20:52, George N. White III wrote:
On Tue, Jun 28, 2022 at 8:40 PM Robert Moskowitz rgm@htt-consult.com wrote:
Samuel,
Well thanks. It is NOT the kernel as I am still locking up. But it is something else related to that update, as I did not get these lockups before. I have updated the bug report and sent a number of related selinux error details and messages
My video is radeon. That is showing in the error messages in /var/log/messages I included in the bug report.
https://bugzilla.redhat.com/show_bug.cgi?id=2101629 has:ring test failures.
https://docs.oracle.com/cd/E50245_01/E63729/html/vmrns-radeon-ring-test-fail... gives::
Workaround: Ensure that the radeon module is loaded with the modeset parameter set to zero. To do this, add the following line in /etc/modprobe.conf:
options radeon modeset=0
This does not change anything. I crashed and rebooted with this line in modprobe.conf.
Got into firefox and cnn and lock.
Restarted.
Had nothing really running (no thunderbirds) got firefox going. lock.
I looked to see if chrome is available to try it, but don't see it in the repo.
Any other browser?
Of course it should not be the browser, as all that is happening is some function is being called that crashes the radeon driver.
Hopefully this gets solved soon. I can't go to a number of web sites like in the past without the lockups.
It could be a hardware failure. Laptops over 5 years old are not to be trusted.
Sigh.
On 6/28/22 07:44, Robert Moskowitz wrote:
Just locked again.
Now running with kernel: 15.17.12.200
On 6/28/22 06:18, Robert Moskowitz wrote: > > On 6/28/22 02:27, Samuel Sieb wrote: >> On 6/27/22 19:20, Robert Moskowitz wrote: >>> Happened again. I submitted a bug: >>> >>> https://bugzilla.redhat.com/show_bug.cgi?id=2101629 >> If you think it's the kernel, did you try booting the previous >> kernel to see if it still happens? > > Dah..... > > Next time! I am in meetings this week and only will reboot if I have > to. > > Like another crash... > > My mind is so focused on the work I am doing and the writing, that I > am forgetting the obvious things to do. > > Sigh. And thanks. > _______________________________________________ > users mailing list -- users@lists.fedoraproject.org > To unsubscribe send an email to users-leave@lists.fedoraproject.org > Fedora Code of Conduct: > https://docs.fedoraproject.org/en-US/project/code-of-conduct/ > List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines > List Archives: > https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org > Do not reply to spam on the list, report it: > https://pagure.io/fedora-infrastructure _______________________________________________ users mailing list -- users@lists.fedoraproject.org To unsubscribe send an email to users-leave@lists.fedoraproject.org Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org Do not reply to spam on the list, report it: https://pagure.io/fedora-infrastructure
users mailing list -- users@lists.fedoraproject.org To unsubscribe send an email to users-leave@lists.fedoraproject.org Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org Do not reply to spam on the list, report it: https://pagure.io/fedora-infrastructure
-- George N. White III
users mailing list -- users@lists.fedoraproject.org To unsubscribe send an email to users-leave@lists.fedoraproject.org Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org Do not reply to spam on the list, report it: https://pagure.io/fedora-infrastructure
users mailing list -- users@lists.fedoraproject.org To unsubscribe send an email to users-leave@lists.fedoraproject.org Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org Do not reply to spam on the list, report it: https://pagure.io/fedora-infrastructure
users mailing list -- users@lists.fedoraproject.org To unsubscribe send an email to users-leave@lists.fedoraproject.org Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org Do not reply to spam on the list, report it: https://pagure.io/fedora-infrastructure
On 6/29/22 18:25, Roger Heflin wrote:
The notes I found say to create a file in /etc/modprobe.d
And say call it anything but something like enable_amdgpu_disable_radeon.conf and put this in it: options amdgpu si_support=1 options amdgpu cik_support=1 options radeon si_support=0 options radeon cik_support=0
and then do a "dracut --force" and reboot.
The other option is to make a file called radeon-blacklist.conf and put in: blacklist radeon
The filenames don't actually matter as long as they end in ".conf".
That will not work.
If I am reading the module options right (and other peoples comments), amdgpu will not manage a si/cik chipset unless the correct _support is set to 1.
On Thu, Jun 30, 2022 at 12:16 AM Samuel Sieb samuel@sieb.net wrote:
On 6/29/22 18:25, Roger Heflin wrote:
The notes I found say to create a file in /etc/modprobe.d
And say call it anything but something like enable_amdgpu_disable_radeon.conf and put this in it: options amdgpu si_support=1 options amdgpu cik_support=1 options radeon si_support=0 options radeon cik_support=0
and then do a "dracut --force" and reboot.
The other option is to make a file called radeon-blacklist.conf and put in: blacklist radeon
The filenames don't actually matter as long as they end in ".conf". _______________________________________________ users mailing list -- users@lists.fedoraproject.org To unsubscribe send an email to users-leave@lists.fedoraproject.org Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org Do not reply to spam on the list, report it: https://pagure.io/fedora-infrastructure
Is it correct to assume that I don't have to install anything more to get the amdgpu driver?
On 6/30/22 06:22, Roger Heflin wrote:
That will not work.
If I am reading the module options right (and other peoples comments), amdgpu will not manage a si/cik chipset unless the correct _support is set to 1.
On Thu, Jun 30, 2022 at 12:16 AM Samuel Sieb samuel@sieb.net wrote:
On 6/29/22 18:25, Roger Heflin wrote:
The notes I found say to create a file in /etc/modprobe.d
And say call it anything but something like enable_amdgpu_disable_radeon.conf and put this in it: options amdgpu si_support=1 options amdgpu cik_support=1 options radeon si_support=0 options radeon cik_support=0
and then do a "dracut --force" and reboot.
The other option is to make a file called radeon-blacklist.conf and put in: blacklist radeon
The filenames don't actually matter as long as they end in ".conf". _______________________________________________ users mailing list -- users@lists.fedoraproject.org To unsubscribe send an email to users-leave@lists.fedoraproject.org Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org Do not reply to spam on the list, report it: https://pagure.io/fedora-infrastructure
users mailing list -- users@lists.fedoraproject.org To unsubscribe send an email to users-leave@lists.fedoraproject.org Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org Do not reply to spam on the list, report it: https://pagure.io/fedora-infrastructure
I do not believe anything else is needed. Where dmesg said "radeon:", it should now say "amdgpu:". amdgpu seems to be the replacement and/or currently developed driver. And the recent non-ryzen AMD laptops are using an older AMD cpu + ATI/AMD build-in video, that originally were blocked from using the (at the time - ) experimental amdgpu driver.
It seems pretty likely that there is code from the radeon driver in the amdgpu driver, a lot of the messages from it look similar/identical, and on my ryzen system that default uses amdgpu.
On Thu, Jun 30, 2022 at 6:21 AM Robert Moskowitz rgm@htt-consult.com wrote:
Is it correct to assume that I don't have to install anything more to get the amdgpu driver?
On 6/30/22 06:22, Roger Heflin wrote:
That will not work.
If I am reading the module options right (and other peoples comments), amdgpu will not manage a si/cik chipset unless the correct _support is set to 1.
On Thu, Jun 30, 2022 at 12:16 AM Samuel Sieb samuel@sieb.net wrote:
On 6/29/22 18:25, Roger Heflin wrote:
The notes I found say to create a file in /etc/modprobe.d
And say call it anything but something like enable_amdgpu_disable_radeon.conf and put this in it: options amdgpu si_support=1 options amdgpu cik_support=1 options radeon si_support=0 options radeon cik_support=0
and then do a "dracut --force" and reboot.
The other option is to make a file called radeon-blacklist.conf and put in: blacklist radeon
The filenames don't actually matter as long as they end in ".conf". _______________________________________________ users mailing list -- users@lists.fedoraproject.org To unsubscribe send an email to users-leave@lists.fedoraproject.org Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org Do not reply to spam on the list, report it: https://pagure.io/fedora-infrastructure
users mailing list -- users@lists.fedoraproject.org To unsubscribe send an email to users-leave@lists.fedoraproject.org Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org Do not reply to spam on the list, report it: https://pagure.io/fedora-infrastructure
Finally did this:
I did not make the change when I got back to my office/home, and things worked fine into mid-sunday when it locked again. Did a dnf update from a character session after power cycle and new kernel, firefox, and linuxfirmware. So decided to see how it would go.
And I could work with cnn and youtube and other video stuff that was locking me up. Plus other apps, so I thought maybe things were fixed....
But twice my session aborted without a system lockup and I had to log back in and restart everything.
Then I got a lock up again, so I power cycled and hay! New kernel. But I still did as you instructed below, making the file and running dracut. Now on reboot:
# lspci -nnk | egrep -i "VGA|DISPLAY" -A3 00:01.0 VGA compatible controller [0300]: Advanced Micro Devices, Inc. [AMD/ATI] Kabini [Radeon HD 8330] [1002:9832] Subsystem: Lenovo Device [17aa:2219] Kernel driver in use: radeon Kernel modules: radeon, amdgpu
which is the same as before:
It still says using radeon driver?
Should I try your radeon_blacklist.conf ?
On 6/30/22 08:04, Roger Heflin wrote:
I do not believe anything else is needed. Where dmesg said "radeon:", it should now say "amdgpu:". amdgpu seems to be the replacement and/or currently developed driver. And the recent non-ryzen AMD laptops are using an older AMD cpu + ATI/AMD build-in video, that originally were blocked from using the (at the time - ) experimental amdgpu driver.
It seems pretty likely that there is code from the radeon driver in the amdgpu driver, a lot of the messages from it look similar/identical, and on my ryzen system that default uses amdgpu.
On Thu, Jun 30, 2022 at 6:21 AM Robert Moskowitz rgm@htt-consult.com wrote:
Is it correct to assume that I don't have to install anything more to get the amdgpu driver?
On 6/30/22 06:22, Roger Heflin wrote:
That will not work.
If I am reading the module options right (and other peoples comments), amdgpu will not manage a si/cik chipset unless the correct _support is set to 1.
On Thu, Jun 30, 2022 at 12:16 AM Samuel Sieb samuel@sieb.net wrote:
On 6/29/22 18:25, Roger Heflin wrote:
The notes I found say to create a file in /etc/modprobe.d
And say call it anything but something like enable_amdgpu_disable_radeon.conf and put this in it: options amdgpu si_support=1 options amdgpu cik_support=1 options radeon si_support=0 options radeon cik_support=0
and then do a "dracut --force" and reboot.
The other option is to make a file called radeon-blacklist.conf and put in: blacklist radeon
The filenames don't actually matter as long as they end in ".conf". _______________________________________________ users mailing list -- users@lists.fedoraproject.org To unsubscribe send an email to users-leave@lists.fedoraproject.org Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org Do not reply to spam on the list, report it: https://pagure.io/fedora-infrastructure
users mailing list -- users@lists.fedoraproject.org To unsubscribe send an email to users-leave@lists.fedoraproject.org Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org Do not reply to spam on the list, report it: https://pagure.io/fedora-infrastructure
try this and see what it shows:
grep -v adfasd /sys/module/*/parameters/si_support /sys/module/*/parameters/cik_support
to verify that radeon is disabled and amdgpu is enabled.
I tried on mine, an my amd gpu is a really old one while claiming to be supported by amdgpu, acts like it is not.
On Mon, Jul 4, 2022 at 10:39 AM Robert Moskowitz rgm@htt-consult.com wrote:
Finally did this:
I did not make the change when I got back to my office/home, and things worked fine into mid-sunday when it locked again. Did a dnf update from a character session after power cycle and new kernel, firefox, and linuxfirmware. So decided to see how it would go.
And I could work with cnn and youtube and other video stuff that was locking me up. Plus other apps, so I thought maybe things were fixed....
But twice my session aborted without a system lockup and I had to log back in and restart everything.
Then I got a lock up again, so I power cycled and hay! New kernel. But I still did as you instructed below, making the file and running dracut. Now on reboot:
# lspci -nnk | egrep -i "VGA|DISPLAY" -A3 00:01.0 VGA compatible controller [0300]: Advanced Micro Devices, Inc. [AMD/ATI] Kabini [Radeon HD 8330] [1002:9832] Subsystem: Lenovo Device [17aa:2219] Kernel driver in use: radeon Kernel modules: radeon, amdgpu
which is the same as before:
It still says using radeon driver?
Should I try your radeon_blacklist.conf ?
On 6/30/22 08:04, Roger Heflin wrote:
I do not believe anything else is needed. Where dmesg said "radeon:", it should now say "amdgpu:". amdgpu seems to be the replacement and/or currently developed driver. And the recent non-ryzen AMD laptops are using an older AMD cpu + ATI/AMD build-in video, that originally were blocked from using the (at the time - ) experimental amdgpu driver.
It seems pretty likely that there is code from the radeon driver in the amdgpu driver, a lot of the messages from it look similar/identical, and on my ryzen system that default uses amdgpu.
On Thu, Jun 30, 2022 at 6:21 AM Robert Moskowitz rgm@htt-consult.com wrote:
Is it correct to assume that I don't have to install anything more to get the amdgpu driver?
On 6/30/22 06:22, Roger Heflin wrote:
That will not work.
If I am reading the module options right (and other peoples comments), amdgpu will not manage a si/cik chipset unless the correct _support is set to 1.
On Thu, Jun 30, 2022 at 12:16 AM Samuel Sieb samuel@sieb.net wrote:
On 6/29/22 18:25, Roger Heflin wrote:
The notes I found say to create a file in /etc/modprobe.d
And say call it anything but something like enable_amdgpu_disable_radeon.conf and put this in it: options amdgpu si_support=1 options amdgpu cik_support=1 options radeon si_support=0 options radeon cik_support=0
and then do a "dracut --force" and reboot.
The other option is to make a file called radeon-blacklist.conf and put in: blacklist radeon
The filenames don't actually matter as long as they end in ".conf". _______________________________________________ users mailing list -- users@lists.fedoraproject.org To unsubscribe send an email to users-leave@lists.fedoraproject.org Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org Do not reply to spam on the list, report it: https://pagure.io/fedora-infrastructure
users mailing list -- users@lists.fedoraproject.org To unsubscribe send an email to users-leave@lists.fedoraproject.org Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org Do not reply to spam on the list, report it: https://pagure.io/fedora-infrastructure
On 7/4/22 15:19, Roger Heflin wrote:
try this and see what it shows:
grep -v adfasd /sys/module/*/parameters/si_support
/sys/module/amdgpu/parameters/si_support:0 /sys/module/radeon/parameters/si_support:1
/sys/module/*/parameters/cik_support
As root:
-bash: /sys/module/amdgpu/parameters/cik_support: Permission denied
And I checked:
# cat /etc/modprobe.d/enable_amdgpu_disable_radeon.conf options amdgpu si_support=1 options amdgpu cik_support=1 options radeon si_support=0 options radeon cik_support=0
to verify that radeon is disabled and amdgpu is enabled.
I tried on mine, an my amd gpu is a really old one while claiming to be supported by amdgpu, acts like it is not.
On Mon, Jul 4, 2022 at 10:39 AM Robert Moskowitz rgm@htt-consult.com wrote:
Finally did this:
I did not make the change when I got back to my office/home, and things worked fine into mid-sunday when it locked again. Did a dnf update from a character session after power cycle and new kernel, firefox, and linuxfirmware. So decided to see how it would go.
And I could work with cnn and youtube and other video stuff that was locking me up. Plus other apps, so I thought maybe things were fixed....
But twice my session aborted without a system lockup and I had to log back in and restart everything.
Then I got a lock up again, so I power cycled and hay! New kernel. But I still did as you instructed below, making the file and running dracut. Now on reboot:
# lspci -nnk | egrep -i "VGA|DISPLAY" -A3 00:01.0 VGA compatible controller [0300]: Advanced Micro Devices, Inc. [AMD/ATI] Kabini [Radeon HD 8330] [1002:9832] Subsystem: Lenovo Device [17aa:2219] Kernel driver in use: radeon Kernel modules: radeon, amdgpu
which is the same as before:
It still says using radeon driver?
Should I try your radeon_blacklist.conf ?
On 6/30/22 08:04, Roger Heflin wrote:
I do not believe anything else is needed. Where dmesg said "radeon:", it should now say "amdgpu:". amdgpu seems to be the replacement and/or currently developed driver. And the recent non-ryzen AMD laptops are using an older AMD cpu + ATI/AMD build-in video, that originally were blocked from using the (at the time - ) experimental amdgpu driver.
It seems pretty likely that there is code from the radeon driver in the amdgpu driver, a lot of the messages from it look similar/identical, and on my ryzen system that default uses amdgpu.
On Thu, Jun 30, 2022 at 6:21 AM Robert Moskowitz rgm@htt-consult.com wrote:
Is it correct to assume that I don't have to install anything more to get the amdgpu driver?
On 6/30/22 06:22, Roger Heflin wrote:
That will not work.
If I am reading the module options right (and other peoples comments), amdgpu will not manage a si/cik chipset unless the correct _support is set to 1.
On Thu, Jun 30, 2022 at 12:16 AM Samuel Sieb samuel@sieb.net wrote:
On 6/29/22 18:25, Roger Heflin wrote: > The notes I found say to create a file in /etc/modprobe.d > > And say call it anything but something like > enable_amdgpu_disable_radeon.conf and put this in it: > options amdgpu si_support=1 > options amdgpu cik_support=1 > options radeon si_support=0 > options radeon cik_support=0 > > and then do a "dracut --force" and reboot. The other option is to make a file called radeon-blacklist.conf and put in: blacklist radeon
The filenames don't actually matter as long as they end in ".conf". _______________________________________________ users mailing list -- users@lists.fedoraproject.org To unsubscribe send an email to users-leave@lists.fedoraproject.org Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org Do not reply to spam on the list, report it: https://pagure.io/fedora-infrastructure
users mailing list -- users@lists.fedoraproject.org To unsubscribe send an email to users-leave@lists.fedoraproject.org Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org Do not reply to spam on the list, report it: https://pagure.io/fedora-infrastructure
the 2nd entry needs to be on the same grep line.
But based on the output the options in that file aren't being used.
On mine both of the options are set to one on amdgpu, and 0 on radeon.
do lsinitrd | grep enable_amdgpu_disable_radeon
you should see the file being included in the output.
If it is not, try running "dracut -f" and reboot.
On Mon, Jul 4, 2022 at 3:13 PM Robert Moskowitz rgm@htt-consult.com wrote:
On 7/4/22 15:19, Roger Heflin wrote:
try this and see what it shows:
grep -v adfasd /sys/module/*/parameters/si_support
/sys/module/amdgpu/parameters/si_support:0 /sys/module/radeon/parameters/si_support:1
/sys/module/*/parameters/cik_support
As root:
-bash: /sys/module/amdgpu/parameters/cik_support: Permission denied
And I checked:
# cat /etc/modprobe.d/enable_amdgpu_disable_radeon.conf options amdgpu si_support=1 options amdgpu cik_support=1 options radeon si_support=0 options radeon cik_support=0
to verify that radeon is disabled and amdgpu is enabled.
I tried on mine, an my amd gpu is a really old one while claiming to be supported by amdgpu, acts like it is not.
On Mon, Jul 4, 2022 at 10:39 AM Robert Moskowitz rgm@htt-consult.com wrote:
Finally did this:
I did not make the change when I got back to my office/home, and things worked fine into mid-sunday when it locked again. Did a dnf update from a character session after power cycle and new kernel, firefox, and linuxfirmware. So decided to see how it would go.
And I could work with cnn and youtube and other video stuff that was locking me up. Plus other apps, so I thought maybe things were fixed....
But twice my session aborted without a system lockup and I had to log back in and restart everything.
Then I got a lock up again, so I power cycled and hay! New kernel. But I still did as you instructed below, making the file and running dracut. Now on reboot:
# lspci -nnk | egrep -i "VGA|DISPLAY" -A3 00:01.0 VGA compatible controller [0300]: Advanced Micro Devices, Inc. [AMD/ATI] Kabini [Radeon HD 8330] [1002:9832] Subsystem: Lenovo Device [17aa:2219] Kernel driver in use: radeon Kernel modules: radeon, amdgpu
which is the same as before:
It still says using radeon driver?
Should I try your radeon_blacklist.conf ?
On 6/30/22 08:04, Roger Heflin wrote:
I do not believe anything else is needed. Where dmesg said "radeon:", it should now say "amdgpu:". amdgpu seems to be the replacement and/or currently developed driver. And the recent non-ryzen AMD laptops are using an older AMD cpu + ATI/AMD build-in video, that originally were blocked from using the (at the time - ) experimental amdgpu driver.
It seems pretty likely that there is code from the radeon driver in the amdgpu driver, a lot of the messages from it look similar/identical, and on my ryzen system that default uses amdgpu.
On Thu, Jun 30, 2022 at 6:21 AM Robert Moskowitz rgm@htt-consult.com wrote:
Is it correct to assume that I don't have to install anything more to get the amdgpu driver?
On 6/30/22 06:22, Roger Heflin wrote:
That will not work.
If I am reading the module options right (and other peoples comments), amdgpu will not manage a si/cik chipset unless the correct _support is set to 1.
On Thu, Jun 30, 2022 at 12:16 AM Samuel Sieb samuel@sieb.net wrote: > On 6/29/22 18:25, Roger Heflin wrote: >> The notes I found say to create a file in /etc/modprobe.d >> >> And say call it anything but something like >> enable_amdgpu_disable_radeon.conf and put this in it: >> options amdgpu si_support=1 >> options amdgpu cik_support=1 >> options radeon si_support=0 >> options radeon cik_support=0 >> >> and then do a "dracut --force" and reboot. > The other option is to make a file called radeon-blacklist.conf and put in: > blacklist radeon > > The filenames don't actually matter as long as they end in ".conf". > _______________________________________________ > users mailing list -- users@lists.fedoraproject.org > To unsubscribe send an email to users-leave@lists.fedoraproject.org > Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ > List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines > List Archives: https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org > Do not reply to spam on the list, report it: https://pagure.io/fedora-infrastructure _______________________________________________ users mailing list -- users@lists.fedoraproject.org To unsubscribe send an email to users-leave@lists.fedoraproject.org Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org Do not reply to spam on the list, report it: https://pagure.io/fedora-infrastructure
On 7/4/22 16:24, Roger Heflin wrote:
the 2nd entry needs to be on the same grep line.
But based on the output the options in that file aren't being used.
On mine both of the options are set to one on amdgpu, and 0 on radeon.
do lsinitrd | grep enable_amdgpu_disable_radeon
you should see the file being included in the output.
# lsinitrd | grep enable_amdgpu_disable_radeon #
Nothing found.
# dracut -f #
then:
# lsinitrd | grep enable_amdgpu_disable_radeon -rw-r--r-- 1 root root 114 Oct 28 2021 etc/modprobe.d/enable_amdgpu_disable_radeon.conf
So maybe this time....
Rebooting now.
If it is not, try running "dracut -f" and reboot.
On Mon, Jul 4, 2022 at 3:13 PM Robert Moskowitz rgm@htt-consult.com wrote:
On 7/4/22 15:19, Roger Heflin wrote:
try this and see what it shows:
grep -v adfasd /sys/module/*/parameters/si_support
/sys/module/amdgpu/parameters/si_support:0 /sys/module/radeon/parameters/si_support:1
/sys/module/*/parameters/cik_support
As root:
-bash: /sys/module/amdgpu/parameters/cik_support: Permission denied
And I checked:
# cat /etc/modprobe.d/enable_amdgpu_disable_radeon.conf options amdgpu si_support=1 options amdgpu cik_support=1 options radeon si_support=0 options radeon cik_support=0
to verify that radeon is disabled and amdgpu is enabled.
I tried on mine, an my amd gpu is a really old one while claiming to be supported by amdgpu, acts like it is not.
On Mon, Jul 4, 2022 at 10:39 AM Robert Moskowitz rgm@htt-consult.com wrote:
Finally did this:
I did not make the change when I got back to my office/home, and things worked fine into mid-sunday when it locked again. Did a dnf update from a character session after power cycle and new kernel, firefox, and linuxfirmware. So decided to see how it would go.
And I could work with cnn and youtube and other video stuff that was locking me up. Plus other apps, so I thought maybe things were fixed....
But twice my session aborted without a system lockup and I had to log back in and restart everything.
Then I got a lock up again, so I power cycled and hay! New kernel. But I still did as you instructed below, making the file and running dracut. Now on reboot:
# lspci -nnk | egrep -i "VGA|DISPLAY" -A3 00:01.0 VGA compatible controller [0300]: Advanced Micro Devices, Inc. [AMD/ATI] Kabini [Radeon HD 8330] [1002:9832] Subsystem: Lenovo Device [17aa:2219] Kernel driver in use: radeon Kernel modules: radeon, amdgpu
which is the same as before:
It still says using radeon driver?
Should I try your radeon_blacklist.conf ?
On 6/30/22 08:04, Roger Heflin wrote:
I do not believe anything else is needed. Where dmesg said "radeon:", it should now say "amdgpu:". amdgpu seems to be the replacement and/or currently developed driver. And the recent non-ryzen AMD laptops are using an older AMD cpu + ATI/AMD build-in video, that originally were blocked from using the (at the time - ) experimental amdgpu driver.
It seems pretty likely that there is code from the radeon driver in the amdgpu driver, a lot of the messages from it look similar/identical, and on my ryzen system that default uses amdgpu.
On Thu, Jun 30, 2022 at 6:21 AM Robert Moskowitz rgm@htt-consult.com wrote:
Is it correct to assume that I don't have to install anything more to get the amdgpu driver?
On 6/30/22 06:22, Roger Heflin wrote: > That will not work. > > If I am reading the module options right (and other peoples comments), > amdgpu will not manage a si/cik chipset unless the correct _support is > set to 1. > > On Thu, Jun 30, 2022 at 12:16 AM Samuel Sieb samuel@sieb.net wrote: >> On 6/29/22 18:25, Roger Heflin wrote: >>> The notes I found say to create a file in /etc/modprobe.d >>> >>> And say call it anything but something like >>> enable_amdgpu_disable_radeon.conf and put this in it: >>> options amdgpu si_support=1 >>> options amdgpu cik_support=1 >>> options radeon si_support=0 >>> options radeon cik_support=0 >>> >>> and then do a "dracut --force" and reboot. >> The other option is to make a file called radeon-blacklist.conf and put in: >> blacklist radeon >> >> The filenames don't actually matter as long as they end in ".conf". >> _______________________________________________ >> users mailing list -- users@lists.fedoraproject.org >> To unsubscribe send an email to users-leave@lists.fedoraproject.org >> Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ >> List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines >> List Archives: https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org >> Do not reply to spam on the list, report it: https://pagure.io/fedora-infrastructure > _______________________________________________ > users mailing list -- users@lists.fedoraproject.org > To unsubscribe send an email to users-leave@lists.fedoraproject.org > Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ > List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines > List Archives: https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org > Do not reply to spam on the list, report it: https://pagure.io/fedora-infrastructure
Looks like it took this time:
# grep -v adfasd /sys/module/*/parameters/si_support /sys/module/*/parameters/cik_support /sys/module/amdgpu/parameters/si_support:1 /sys/module/radeon/parameters/si_support:0 /sys/module/amdgpu/parameters/cik_support:1 /sys/module/radeon/parameters/cik_support:0
# lspci -nnk | egrep -i "VGA|DISPLAY" -A3 00:01.0 VGA compatible controller [0300]: Advanced Micro Devices, Inc. [AMD/ATI] Kabini [Radeon HD 8330] [1002:9832] Subsystem: Lenovo Device [17aa:2219] Kernel driver in use: amdgpu Kernel modules: radeon, amdgpu
In your first instructions you had "dracut --force"
This time you had "dracut -f"
Don't know if that was the difference or something else.
But now I have to put this into my installation instructions.
First we will see how things run for the week.
Thanks!
On 7/4/22 17:24, Robert Moskowitz wrote:
On 7/4/22 16:24, Roger Heflin wrote:
the 2nd entry needs to be on the same grep line.
But based on the output the options in that file aren't being used.
On mine both of the options are set to one on amdgpu, and 0 on radeon.
do lsinitrd | grep enable_amdgpu_disable_radeon
you should see the file being included in the output.
# lsinitrd | grep enable_amdgpu_disable_radeon #
Nothing found.
# dracut -f #
then:
# lsinitrd | grep enable_amdgpu_disable_radeon -rw-r--r-- 1 root root 114 Oct 28 2021 etc/modprobe.d/enable_amdgpu_disable_radeon.conf
So maybe this time....
Rebooting now.
If it is not, try running "dracut -f" and reboot.
On Mon, Jul 4, 2022 at 3:13 PM Robert Moskowitz rgm@htt-consult.com wrote:
On 7/4/22 15:19, Roger Heflin wrote:
try this and see what it shows:
grep -v adfasd /sys/module/*/parameters/si_support
/sys/module/amdgpu/parameters/si_support:0 /sys/module/radeon/parameters/si_support:1
/sys/module/*/parameters/cik_support
As root:
-bash: /sys/module/amdgpu/parameters/cik_support: Permission denied
And I checked:
# cat /etc/modprobe.d/enable_amdgpu_disable_radeon.conf options amdgpu si_support=1 options amdgpu cik_support=1 options radeon si_support=0 options radeon cik_support=0
to verify that radeon is disabled and amdgpu is enabled.
I tried on mine, an my amd gpu is a really old one while claiming to be supported by amdgpu, acts like it is not.
On Mon, Jul 4, 2022 at 10:39 AM Robert Moskowitz rgm@htt-consult.com wrote:
Finally did this:
I did not make the change when I got back to my office/home, and things worked fine into mid-sunday when it locked again. Did a dnf update from a character session after power cycle and new kernel, firefox, and linuxfirmware. So decided to see how it would go.
And I could work with cnn and youtube and other video stuff that was locking me up. Plus other apps, so I thought maybe things were fixed....
But twice my session aborted without a system lockup and I had to log back in and restart everything.
Then I got a lock up again, so I power cycled and hay! New kernel. But I still did as you instructed below, making the file and running dracut. Now on reboot:
# lspci -nnk | egrep -i "VGA|DISPLAY" -A3 00:01.0 VGA compatible controller [0300]: Advanced Micro Devices, Inc. [AMD/ATI] Kabini [Radeon HD 8330] [1002:9832] Subsystem: Lenovo Device [17aa:2219] Kernel driver in use: radeon Kernel modules: radeon, amdgpu
which is the same as before:
It still says using radeon driver?
Should I try your radeon_blacklist.conf ?
On 6/30/22 08:04, Roger Heflin wrote:
I do not believe anything else is needed. Where dmesg said "radeon:", it should now say "amdgpu:". amdgpu seems to be the replacement and/or currently developed driver. And the recent non-ryzen AMD laptops are using an older AMD cpu + ATI/AMD build-in video, that originally were blocked from using the (at the time - ) experimental amdgpu driver.
It seems pretty likely that there is code from the radeon driver in the amdgpu driver, a lot of the messages from it look similar/identical, and on my ryzen system that default uses amdgpu.
On Thu, Jun 30, 2022 at 6:21 AM Robert Moskowitz rgm@htt-consult.com wrote: > Is it correct to assume that I don't have to install anything > more to > get the amdgpu driver? > > On 6/30/22 06:22, Roger Heflin wrote: >> That will not work. >> >> If I am reading the module options right (and other peoples >> comments), >> amdgpu will not manage a si/cik chipset unless the correct >> _support is >> set to 1. >> >> On Thu, Jun 30, 2022 at 12:16 AM Samuel Sieb samuel@sieb.net >> wrote: >>> On 6/29/22 18:25, Roger Heflin wrote: >>>> The notes I found say to create a file in /etc/modprobe.d >>>> >>>> And say call it anything but something like >>>> enable_amdgpu_disable_radeon.conf and put this in it: >>>> options amdgpu si_support=1 >>>> options amdgpu cik_support=1 >>>> options radeon si_support=0 >>>> options radeon cik_support=0 >>>> >>>> and then do a "dracut --force" and reboot. >>> The other option is to make a file called >>> radeon-blacklist.conf and put in: >>> blacklist radeon >>> >>> The filenames don't actually matter as long as they end in >>> ".conf". >>> _______________________________________________ >>> users mailing list -- users@lists.fedoraproject.org >>> To unsubscribe send an email to >>> users-leave@lists.fedoraproject.org >>> Fedora Code of Conduct: >>> https://docs.fedoraproject.org/en-US/project/code-of-conduct/ >>> List Guidelines: >>> https://fedoraproject.org/wiki/Mailing_list_guidelines >>> List Archives: >>> https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org >>> Do not reply to spam on the list, report it: >>> https://pagure.io/fedora-infrastructure >> _______________________________________________ >> users mailing list -- users@lists.fedoraproject.org >> To unsubscribe send an email to >> users-leave@lists.fedoraproject.org >> Fedora Code of Conduct: >> https://docs.fedoraproject.org/en-US/project/code-of-conduct/ >> List Guidelines: >> https://fedoraproject.org/wiki/Mailing_list_guidelines >> List Archives: >> https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org >> Do not reply to spam on the list, report it: >> https://pagure.io/fedora-infrastructure
users mailing list -- users@lists.fedoraproject.org To unsubscribe send an email to users-leave@lists.fedoraproject.org Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org Do not reply to spam on the list, report it: https://pagure.io/fedora-infrastructure
Once the conf file is in place any upgrades will keep this so long as you don't do a clean instat..
-f and --force are the same, so I would suspect you were booted into an older/different kernel and it rebuilt the initramfs for that older kernel but this kernel had been installed earlier and dracut by default only rebuilds the current kernels initrd and not all kernels unless another option is added.
Hopefully since this is a newer driver for the same card that is still being developed this one actually works.
On Mon, Jul 4, 2022 at 4:33 PM Robert Moskowitz rgm@htt-consult.com wrote:
Looks like it took this time:
# grep -v adfasd /sys/module/*/parameters/si_support /sys/module/*/parameters/cik_support /sys/module/amdgpu/parameters/si_support:1 /sys/module/radeon/parameters/si_support:0 /sys/module/amdgpu/parameters/cik_support:1 /sys/module/radeon/parameters/cik_support:0
# lspci -nnk | egrep -i "VGA|DISPLAY" -A3 00:01.0 VGA compatible controller [0300]: Advanced Micro Devices, Inc. [AMD/ATI] Kabini [Radeon HD 8330] [1002:9832] Subsystem: Lenovo Device [17aa:2219] Kernel driver in use: amdgpu Kernel modules: radeon, amdgpu
In your first instructions you had "dracut --force"
This time you had "dracut -f"
Don't know if that was the difference or something else.
But now I have to put this into my installation instructions.
First we will see how things run for the week.
Thanks!
On 7/4/22 17:24, Robert Moskowitz wrote:
On 7/4/22 16:24, Roger Heflin wrote:
the 2nd entry needs to be on the same grep line.
But based on the output the options in that file aren't being used.
On mine both of the options are set to one on amdgpu, and 0 on radeon.
do lsinitrd | grep enable_amdgpu_disable_radeon
you should see the file being included in the output.
# lsinitrd | grep enable_amdgpu_disable_radeon #
Nothing found.
# dracut -f #
then:
# lsinitrd | grep enable_amdgpu_disable_radeon -rw-r--r-- 1 root root 114 Oct 28 2021 etc/modprobe.d/enable_amdgpu_disable_radeon.conf
So maybe this time....
Rebooting now.
If it is not, try running "dracut -f" and reboot.
On Mon, Jul 4, 2022 at 3:13 PM Robert Moskowitz rgm@htt-consult.com wrote:
On 7/4/22 15:19, Roger Heflin wrote:
try this and see what it shows:
grep -v adfasd /sys/module/*/parameters/si_support
/sys/module/amdgpu/parameters/si_support:0 /sys/module/radeon/parameters/si_support:1
/sys/module/*/parameters/cik_support
As root:
-bash: /sys/module/amdgpu/parameters/cik_support: Permission denied
And I checked:
# cat /etc/modprobe.d/enable_amdgpu_disable_radeon.conf options amdgpu si_support=1 options amdgpu cik_support=1 options radeon si_support=0 options radeon cik_support=0
to verify that radeon is disabled and amdgpu is enabled.
I tried on mine, an my amd gpu is a really old one while claiming to be supported by amdgpu, acts like it is not.
On Mon, Jul 4, 2022 at 10:39 AM Robert Moskowitz rgm@htt-consult.com wrote:
Finally did this:
I did not make the change when I got back to my office/home, and things worked fine into mid-sunday when it locked again. Did a dnf update from a character session after power cycle and new kernel, firefox, and linuxfirmware. So decided to see how it would go.
And I could work with cnn and youtube and other video stuff that was locking me up. Plus other apps, so I thought maybe things were fixed....
But twice my session aborted without a system lockup and I had to log back in and restart everything.
Then I got a lock up again, so I power cycled and hay! New kernel. But I still did as you instructed below, making the file and running dracut. Now on reboot:
# lspci -nnk | egrep -i "VGA|DISPLAY" -A3 00:01.0 VGA compatible controller [0300]: Advanced Micro Devices, Inc. [AMD/ATI] Kabini [Radeon HD 8330] [1002:9832] Subsystem: Lenovo Device [17aa:2219] Kernel driver in use: radeon Kernel modules: radeon, amdgpu
which is the same as before:
It still says using radeon driver?
Should I try your radeon_blacklist.conf ?
On 6/30/22 08:04, Roger Heflin wrote: > I do not believe anything else is needed. Where dmesg said > "radeon:", it should now say "amdgpu:". amdgpu seems to be the > replacement and/or currently developed driver. And the recent > non-ryzen AMD laptops are using an older AMD cpu + ATI/AMD build-in > video, that originally were blocked from using the (at the time - ) > experimental amdgpu driver. > > It seems pretty likely that there is code from the radeon driver in > the amdgpu driver, a lot of the messages from it look > similar/identical, and on my ryzen system that default uses amdgpu. > > On Thu, Jun 30, 2022 at 6:21 AM Robert Moskowitz > rgm@htt-consult.com wrote: >> Is it correct to assume that I don't have to install anything >> more to >> get the amdgpu driver? >> >> On 6/30/22 06:22, Roger Heflin wrote: >>> That will not work. >>> >>> If I am reading the module options right (and other peoples >>> comments), >>> amdgpu will not manage a si/cik chipset unless the correct >>> _support is >>> set to 1. >>> >>> On Thu, Jun 30, 2022 at 12:16 AM Samuel Sieb samuel@sieb.net >>> wrote: >>>> On 6/29/22 18:25, Roger Heflin wrote: >>>>> The notes I found say to create a file in /etc/modprobe.d >>>>> >>>>> And say call it anything but something like >>>>> enable_amdgpu_disable_radeon.conf and put this in it: >>>>> options amdgpu si_support=1 >>>>> options amdgpu cik_support=1 >>>>> options radeon si_support=0 >>>>> options radeon cik_support=0 >>>>> >>>>> and then do a "dracut --force" and reboot. >>>> The other option is to make a file called >>>> radeon-blacklist.conf and put in: >>>> blacklist radeon >>>> >>>> The filenames don't actually matter as long as they end in >>>> ".conf". >>>> _______________________________________________ >>>> users mailing list -- users@lists.fedoraproject.org >>>> To unsubscribe send an email to >>>> users-leave@lists.fedoraproject.org >>>> Fedora Code of Conduct: >>>> https://docs.fedoraproject.org/en-US/project/code-of-conduct/ >>>> List Guidelines: >>>> https://fedoraproject.org/wiki/Mailing_list_guidelines >>>> List Archives: >>>> https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org >>>> Do not reply to spam on the list, report it: >>>> https://pagure.io/fedora-infrastructure >>> _______________________________________________ >>> users mailing list -- users@lists.fedoraproject.org >>> To unsubscribe send an email to >>> users-leave@lists.fedoraproject.org >>> Fedora Code of Conduct: >>> https://docs.fedoraproject.org/en-US/project/code-of-conduct/ >>> List Guidelines: >>> https://fedoraproject.org/wiki/Mailing_list_guidelines >>> List Archives: >>> https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org >>> Do not reply to spam on the list, report it: >>> https://pagure.io/fedora-infrastructure
users mailing list -- users@lists.fedoraproject.org To unsubscribe send an email to users-leave@lists.fedoraproject.org Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org Do not reply to spam on the list, report it: https://pagure.io/fedora-infrastructure
On 7/4/22 19:51, Roger Heflin wrote:
Once the conf file is in place any upgrades will keep this so long as you don't do a clean instat..
-f and --force are the same, so I would suspect you were booted into an older/different kernel and it rebuilt the initramfs for that older kernel but this kernel had been installed earlier and dracut by default only rebuilds the current kernels initrd and not all kernels unless another option is added.
Yes, that would have been the sequence. After I ran dracut the first time, I did boot up in a new kernel.
Hopefully since this is a newer driver for the same card that is still being developed this one actually works.
And I will have to put this into my install procedures for when I build my next system, I will get this done. I don't do upgrades, I take the old system (right now the one with F31 installed), install the new release (will probably wait for F37) then move everything over. So I have to maintain decent notes about what I have to do each time!
On Mon, Jul 4, 2022 at 4:33 PM Robert Moskowitz rgm@htt-consult.com wrote:
Looks like it took this time:
# grep -v adfasd /sys/module/*/parameters/si_support /sys/module/*/parameters/cik_support /sys/module/amdgpu/parameters/si_support:1 /sys/module/radeon/parameters/si_support:0 /sys/module/amdgpu/parameters/cik_support:1 /sys/module/radeon/parameters/cik_support:0
# lspci -nnk | egrep -i "VGA|DISPLAY" -A3 00:01.0 VGA compatible controller [0300]: Advanced Micro Devices, Inc. [AMD/ATI] Kabini [Radeon HD 8330] [1002:9832] Subsystem: Lenovo Device [17aa:2219] Kernel driver in use: amdgpu Kernel modules: radeon, amdgpu
In your first instructions you had "dracut --force"
This time you had "dracut -f"
Don't know if that was the difference or something else.
But now I have to put this into my installation instructions.
First we will see how things run for the week.
Thanks!
On 7/4/22 17:24, Robert Moskowitz wrote:
On 7/4/22 16:24, Roger Heflin wrote:
the 2nd entry needs to be on the same grep line.
But based on the output the options in that file aren't being used.
On mine both of the options are set to one on amdgpu, and 0 on radeon.
do lsinitrd | grep enable_amdgpu_disable_radeon
you should see the file being included in the output.
# lsinitrd | grep enable_amdgpu_disable_radeon #
Nothing found.
# dracut -f #
then:
# lsinitrd | grep enable_amdgpu_disable_radeon -rw-r--r-- 1 root root 114 Oct 28 2021 etc/modprobe.d/enable_amdgpu_disable_radeon.conf
So maybe this time....
Rebooting now.
If it is not, try running "dracut -f" and reboot.
On Mon, Jul 4, 2022 at 3:13 PM Robert Moskowitz rgm@htt-consult.com wrote:
On 7/4/22 15:19, Roger Heflin wrote:
try this and see what it shows:
grep -v adfasd /sys/module/*/parameters/si_support
/sys/module/amdgpu/parameters/si_support:0 /sys/module/radeon/parameters/si_support:1
/sys/module/*/parameters/cik_support
As root:
-bash: /sys/module/amdgpu/parameters/cik_support: Permission denied
And I checked:
# cat /etc/modprobe.d/enable_amdgpu_disable_radeon.conf options amdgpu si_support=1 options amdgpu cik_support=1 options radeon si_support=0 options radeon cik_support=0
to verify that radeon is disabled and amdgpu is enabled.
I tried on mine, an my amd gpu is a really old one while claiming to be supported by amdgpu, acts like it is not.
On Mon, Jul 4, 2022 at 10:39 AM Robert Moskowitz rgm@htt-consult.com wrote: > Finally did this: > > I did not make the change when I got back to my office/home, and > things > worked fine into mid-sunday when it locked again. Did a dnf > update from > a character session after power cycle and new kernel, firefox, and > linuxfirmware. So decided to see how it would go. > > And I could work with cnn and youtube and other video stuff that was > locking me up. Plus other apps, so I thought maybe things were > fixed.... > > But twice my session aborted without a system lockup and I had to log > back in and restart everything. > > Then I got a lock up again, so I power cycled and hay! New kernel. > But > I still did as you instructed below, making the file and running > dracut. Now on reboot: > > # lspci -nnk | egrep -i "VGA|DISPLAY" -A3 > 00:01.0 VGA compatible controller [0300]: Advanced Micro Devices, > Inc. > [AMD/ATI] Kabini [Radeon HD 8330] [1002:9832] > Subsystem: Lenovo Device [17aa:2219] > Kernel driver in use: radeon > Kernel modules: radeon, amdgpu > > which is the same as before: > > It still says using radeon driver? > > Should I try your radeon_blacklist.conf ? > > On 6/30/22 08:04, Roger Heflin wrote: >> I do not believe anything else is needed. Where dmesg said >> "radeon:", it should now say "amdgpu:". amdgpu seems to be the >> replacement and/or currently developed driver. And the recent >> non-ryzen AMD laptops are using an older AMD cpu + ATI/AMD build-in >> video, that originally were blocked from using the (at the time - ) >> experimental amdgpu driver. >> >> It seems pretty likely that there is code from the radeon driver in >> the amdgpu driver, a lot of the messages from it look >> similar/identical, and on my ryzen system that default uses amdgpu. >> >> On Thu, Jun 30, 2022 at 6:21 AM Robert Moskowitz >> rgm@htt-consult.com wrote: >>> Is it correct to assume that I don't have to install anything >>> more to >>> get the amdgpu driver? >>> >>> On 6/30/22 06:22, Roger Heflin wrote: >>>> That will not work. >>>> >>>> If I am reading the module options right (and other peoples >>>> comments), >>>> amdgpu will not manage a si/cik chipset unless the correct >>>> _support is >>>> set to 1. >>>> >>>> On Thu, Jun 30, 2022 at 12:16 AM Samuel Sieb samuel@sieb.net >>>> wrote: >>>>> On 6/29/22 18:25, Roger Heflin wrote: >>>>>> The notes I found say to create a file in /etc/modprobe.d >>>>>> >>>>>> And say call it anything but something like >>>>>> enable_amdgpu_disable_radeon.conf and put this in it: >>>>>> options amdgpu si_support=1 >>>>>> options amdgpu cik_support=1 >>>>>> options radeon si_support=0 >>>>>> options radeon cik_support=0 >>>>>> >>>>>> and then do a "dracut --force" and reboot. >>>>> The other option is to make a file called >>>>> radeon-blacklist.conf and put in: >>>>> blacklist radeon >>>>> >>>>> The filenames don't actually matter as long as they end in >>>>> ".conf". >>>>> _______________________________________________ >>>>> users mailing list -- users@lists.fedoraproject.org >>>>> To unsubscribe send an email to >>>>> users-leave@lists.fedoraproject.org >>>>> Fedora Code of Conduct: >>>>> https://docs.fedoraproject.org/en-US/project/code-of-conduct/ >>>>> List Guidelines: >>>>> https://fedoraproject.org/wiki/Mailing_list_guidelines >>>>> List Archives: >>>>> https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org >>>>> Do not reply to spam on the list, report it: >>>>> https://pagure.io/fedora-infrastructure >>>> _______________________________________________ >>>> users mailing list -- users@lists.fedoraproject.org >>>> To unsubscribe send an email to >>>> users-leave@lists.fedoraproject.org >>>> Fedora Code of Conduct: >>>> https://docs.fedoraproject.org/en-US/project/code-of-conduct/ >>>> List Guidelines: >>>> https://fedoraproject.org/wiki/Mailing_list_guidelines >>>> List Archives: >>>> https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org >>>> Do not reply to spam on the list, report it: >>>> https://pagure.io/fedora-infrastructure
users mailing list -- users@lists.fedoraproject.org To unsubscribe send an email to users-leave@lists.fedoraproject.org Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org Do not reply to spam on the list, report it: https://pagure.io/fedora-infrastructure
users mailing list -- users@lists.fedoraproject.org To unsubscribe send an email to users-leave@lists.fedoraproject.org Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org Do not reply to spam on the list, report it: https://pagure.io/fedora-infrastructure
Just want to report in that thanks to Roger, I am running smoothly with the amdgpu driver and no lockups for over a week and 1 kernel update.
I closed the bug I had filed and closing this thread.
Take care all, and thanks!
On 7/4/22 19:51, Roger Heflin wrote:
Once the conf file is in place any upgrades will keep this so long as you don't do a clean instat..
-f and --force are the same, so I would suspect you were booted into an older/different kernel and it rebuilt the initramfs for that older kernel but this kernel had been installed earlier and dracut by default only rebuilds the current kernels initrd and not all kernels unless another option is added.
Hopefully since this is a newer driver for the same card that is still being developed this one actually works.
On Mon, Jul 4, 2022 at 4:33 PM Robert Moskowitz rgm@htt-consult.com wrote:
Looks like it took this time:
# grep -v adfasd /sys/module/*/parameters/si_support /sys/module/*/parameters/cik_support /sys/module/amdgpu/parameters/si_support:1 /sys/module/radeon/parameters/si_support:0 /sys/module/amdgpu/parameters/cik_support:1 /sys/module/radeon/parameters/cik_support:0
# lspci -nnk | egrep -i "VGA|DISPLAY" -A3 00:01.0 VGA compatible controller [0300]: Advanced Micro Devices, Inc. [AMD/ATI] Kabini [Radeon HD 8330] [1002:9832] Subsystem: Lenovo Device [17aa:2219] Kernel driver in use: amdgpu Kernel modules: radeon, amdgpu
In your first instructions you had "dracut --force"
This time you had "dracut -f"
Don't know if that was the difference or something else.
But now I have to put this into my installation instructions.
First we will see how things run for the week.
Thanks!
On 7/4/22 17:24, Robert Moskowitz wrote:
On 7/4/22 16:24, Roger Heflin wrote:
the 2nd entry needs to be on the same grep line.
But based on the output the options in that file aren't being used.
On mine both of the options are set to one on amdgpu, and 0 on radeon.
do lsinitrd | grep enable_amdgpu_disable_radeon
you should see the file being included in the output.
# lsinitrd | grep enable_amdgpu_disable_radeon #
Nothing found.
# dracut -f #
then:
# lsinitrd | grep enable_amdgpu_disable_radeon -rw-r--r-- 1 root root 114 Oct 28 2021 etc/modprobe.d/enable_amdgpu_disable_radeon.conf
So maybe this time....
Rebooting now.
If it is not, try running "dracut -f" and reboot.
On Mon, Jul 4, 2022 at 3:13 PM Robert Moskowitz rgm@htt-consult.com wrote:
On 7/4/22 15:19, Roger Heflin wrote:
try this and see what it shows:
grep -v adfasd /sys/module/*/parameters/si_support
/sys/module/amdgpu/parameters/si_support:0 /sys/module/radeon/parameters/si_support:1
/sys/module/*/parameters/cik_support
As root:
-bash: /sys/module/amdgpu/parameters/cik_support: Permission denied
And I checked:
# cat /etc/modprobe.d/enable_amdgpu_disable_radeon.conf options amdgpu si_support=1 options amdgpu cik_support=1 options radeon si_support=0 options radeon cik_support=0
to verify that radeon is disabled and amdgpu is enabled.
I tried on mine, an my amd gpu is a really old one while claiming to be supported by amdgpu, acts like it is not.
On Mon, Jul 4, 2022 at 10:39 AM Robert Moskowitz rgm@htt-consult.com wrote: > Finally did this: > > I did not make the change when I got back to my office/home, and > things > worked fine into mid-sunday when it locked again. Did a dnf > update from > a character session after power cycle and new kernel, firefox, and > linuxfirmware. So decided to see how it would go. > > And I could work with cnn and youtube and other video stuff that was > locking me up. Plus other apps, so I thought maybe things were > fixed.... > > But twice my session aborted without a system lockup and I had to log > back in and restart everything. > > Then I got a lock up again, so I power cycled and hay! New kernel. > But > I still did as you instructed below, making the file and running > dracut. Now on reboot: > > # lspci -nnk | egrep -i "VGA|DISPLAY" -A3 > 00:01.0 VGA compatible controller [0300]: Advanced Micro Devices, > Inc. > [AMD/ATI] Kabini [Radeon HD 8330] [1002:9832] > Subsystem: Lenovo Device [17aa:2219] > Kernel driver in use: radeon > Kernel modules: radeon, amdgpu > > which is the same as before: > > It still says using radeon driver? > > Should I try your radeon_blacklist.conf ? > > On 6/30/22 08:04, Roger Heflin wrote: >> I do not believe anything else is needed. Where dmesg said >> "radeon:", it should now say "amdgpu:". amdgpu seems to be the >> replacement and/or currently developed driver. And the recent >> non-ryzen AMD laptops are using an older AMD cpu + ATI/AMD build-in >> video, that originally were blocked from using the (at the time - ) >> experimental amdgpu driver. >> >> It seems pretty likely that there is code from the radeon driver in >> the amdgpu driver, a lot of the messages from it look >> similar/identical, and on my ryzen system that default uses amdgpu. >> >> On Thu, Jun 30, 2022 at 6:21 AM Robert Moskowitz >> rgm@htt-consult.com wrote: >>> Is it correct to assume that I don't have to install anything >>> more to >>> get the amdgpu driver? >>> >>> On 6/30/22 06:22, Roger Heflin wrote: >>>> That will not work. >>>> >>>> If I am reading the module options right (and other peoples >>>> comments), >>>> amdgpu will not manage a si/cik chipset unless the correct >>>> _support is >>>> set to 1. >>>> >>>> On Thu, Jun 30, 2022 at 12:16 AM Samuel Sieb samuel@sieb.net >>>> wrote: >>>>> On 6/29/22 18:25, Roger Heflin wrote: >>>>>> The notes I found say to create a file in /etc/modprobe.d >>>>>> >>>>>> And say call it anything but something like >>>>>> enable_amdgpu_disable_radeon.conf and put this in it: >>>>>> options amdgpu si_support=1 >>>>>> options amdgpu cik_support=1 >>>>>> options radeon si_support=0 >>>>>> options radeon cik_support=0 >>>>>> >>>>>> and then do a "dracut --force" and reboot. >>>>> The other option is to make a file called >>>>> radeon-blacklist.conf and put in: >>>>> blacklist radeon >>>>> >>>>> The filenames don't actually matter as long as they end in >>>>> ".conf". >>>>> _______________________________________________ >>>>> users mailing list -- users@lists.fedoraproject.org >>>>> To unsubscribe send an email to >>>>> users-leave@lists.fedoraproject.org >>>>> Fedora Code of Conduct: >>>>> https://docs.fedoraproject.org/en-US/project/code-of-conduct/ >>>>> List Guidelines: >>>>> https://fedoraproject.org/wiki/Mailing_list_guidelines >>>>> List Archives: >>>>> https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org >>>>> Do not reply to spam on the list, report it: >>>>> https://pagure.io/fedora-infrastructure >>>> _______________________________________________ >>>> users mailing list -- users@lists.fedoraproject.org >>>> To unsubscribe send an email to >>>> users-leave@lists.fedoraproject.org >>>> Fedora Code of Conduct: >>>> https://docs.fedoraproject.org/en-US/project/code-of-conduct/ >>>> List Guidelines: >>>> https://fedoraproject.org/wiki/Mailing_list_guidelines >>>> List Archives: >>>> https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org >>>> Do not reply to spam on the list, report it: >>>> https://pagure.io/fedora-infrastructure
users mailing list -- users@lists.fedoraproject.org To unsubscribe send an email to users-leave@lists.fedoraproject.org Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org Do not reply to spam on the list, report it: https://pagure.io/fedora-infrastructure
users mailing list -- users@lists.fedoraproject.org To unsubscribe send an email to users-leave@lists.fedoraproject.org Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org Do not reply to spam on the list, report it: https://pagure.io/fedora-infrastructure
On 6/29/22 19:49, Roger Heflin wrote:
What kind of video card/builtin do you have? (lspci | grep -i vga)
lspci | grep -i vga 00:01.0 VGA compatible controller: Advanced Micro Devices, Inc. [AMD/ATI] Kabini [Radeon HD 8330]
Only really old amd/ati cards use the radeon driver. The newer cards use amdgpu. And your dmesg lists your bios as being a 2020 bios release and that would indicate the hardware you are using is new so makes me wonder why you have a radeon driver in use.
I got the latest bios from Lenovo before I did the F35 install back in Jul '21.
On Wed, Jun 29, 2022 at 6:08 PM Robert Moskowitz rgm@htt-consult.com wrote:
On 6/28/22 20:52, George N. White III wrote:
On Tue, Jun 28, 2022 at 8:40 PM Robert Moskowitz rgm@htt-consult.com wrote:
Samuel,
Well thanks. It is NOT the kernel as I am still locking up. But it is something else related to that update, as I did not get these lockups before. I have updated the bug report and sent a number of related selinux error details and messages
My video is radeon. That is showing in the error messages in /var/log/messages I included in the bug report.
https://bugzilla.redhat.com/show_bug.cgi?id=2101629 has:ring test failures.
https://docs.oracle.com/cd/E50245_01/E63729/html/vmrns-radeon-ring-test-fail... gives::
Workaround: Ensure that the radeon module is loaded with the modeset parameter set to zero. To do this, add the following line in /etc/modprobe.conf:
options radeon modeset=0
This does not change anything. I crashed and rebooted with this line in modprobe.conf.
Got into firefox and cnn and lock.
Restarted.
Had nothing really running (no thunderbirds) got firefox going. lock.
I looked to see if chrome is available to try it, but don't see it in the repo.
Any other browser?
Of course it should not be the browser, as all that is happening is some function is being called that crashes the radeon driver.
Hopefully this gets solved soon. I can't go to a number of web sites like in the past without the lockups.
It could be a hardware failure. Laptops over 5 years old are not to be trusted.
Sigh.
On 6/28/22 07:44, Robert Moskowitz wrote:
Just locked again.
Now running with kernel: 15.17.12.200
On 6/28/22 06:18, Robert Moskowitz wrote:
On 6/28/22 02:27, Samuel Sieb wrote:
On 6/27/22 19:20, Robert Moskowitz wrote: > Happened again. I submitted a bug: > > https://bugzilla.redhat.com/show_bug.cgi?id=2101629 If you think it's the kernel, did you try booting the previous kernel to see if it still happens?
Dah.....
Next time! I am in meetings this week and only will reboot if I have to.
Like another crash...
My mind is so focused on the work I am doing and the writing, that I am forgetting the obvious things to do.
Sigh. And thanks. _______________________________________________ users mailing list -- users@lists.fedoraproject.org To unsubscribe send an email to users-leave@lists.fedoraproject.org Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org Do not reply to spam on the list, report it: https://pagure.io/fedora-infrastructure
users mailing list -- users@lists.fedoraproject.org To unsubscribe send an email to users-leave@lists.fedoraproject.org Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org Do not reply to spam on the list, report it: https://pagure.io/fedora-infrastructure
users mailing list -- users@lists.fedoraproject.org To unsubscribe send an email to users-leave@lists.fedoraproject.org Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org Do not reply to spam on the list, report it: https://pagure.io/fedora-infrastructure
-- George N. White III
users mailing list -- users@lists.fedoraproject.org To unsubscribe send an email to users-leave@lists.fedoraproject.org Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org Do not reply to spam on the list, report it: https://pagure.io/fedora-infrastructure
users mailing list -- users@lists.fedoraproject.org To unsubscribe send an email to users-leave@lists.fedoraproject.org Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org Do not reply to spam on the list, report it: https://pagure.io/fedora-infrastructure
users mailing list -- users@lists.fedoraproject.org To unsubscribe send an email to users-leave@lists.fedoraproject.org Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org Do not reply to spam on the list, report it: https://pagure.io/fedora-infrastructure
On 6/28/22 16:38, Robert Moskowitz wrote:
Well thanks. It is NOT the kernel as I am still locking up. But it is something else related to that update, as I did not get these lockups before. I have updated the bug report and sent a number of related selinux error details and messages
The selinux error is related to the coredump. If you look in the journal, you should see what's crashing. Also, I would have expected abrt to popup a notification. Try running the "Problem Reporting" application and see if it has anything listed.
On 6/28/22 21:06, Samuel Sieb wrote:
On 6/28/22 16:38, Robert Moskowitz wrote:
Well thanks. It is NOT the kernel as I am still locking up. But it is something else related to that update, as I did not get these lockups before. I have updated the bug report and sent a number of related selinux error details and messages
The selinux error is related to the coredump. If you look in the journal, you should see what's crashing. Also, I would have expected abrt to popup a notification. Try running the "Problem Reporting" application and see if it has anything listed.
Can't see a way to copy and paste info from Problem Reporting, so got to retype it:
xorg-x11-drv-ati System failure
Name: Kernel-core Ver: 5.17.12-200 Reported: Cannot be reported
The backtrace does not contain enough meaningful function frames to be reported...
There are lots of this error going around, going back >10 years.
Thunderbird is what is crashing, see if reverting thunderbird to the prior version makes the crash go away.
In libraries and/or system calls there are often bugs that fail to block bad/invalid call parameters provided by an application that can leave a system in a bad state. Since this error has been reported many times but seemingly never fixed (in the kernel) then I would suspect that it is one of those sorts of bugs. IE thunderbird has a bug in its graphics calls that tells the kernel to do something with completely wrong call parameters and it leaves the module/kernel in a bad state and causes the crash. The bug is probably that the kernel does not return an error to thunderbird and attempts to handle the invalid request and blows the system up. It is also possible that thunderbird makes the graphics calls and died in the middle of the call chain and that leaves the system in a bad state(kernel bug is that this is not gracefully handled) and causes the crash/deadlock.
The stall could be either an invalid request causing the ring to internally crash and stall, or a request that was started but never finished (both being caused by the application doing something invalid and/or crashing).
The kernel bug is on a stall it needs to cleanup and/or the traceback is the cleanup code trying to run but said cleanup code that rarely gets executed does not actually work all of the time (most of the bugs I had to find methods to work around and/or involve a vendor to fix involve the code paths around errors getting triggered and the cleanup/errors handling simply not working right and/or having bad bugs).
On Wed, Jun 29, 2022 at 5:59 AM Robert Moskowitz rgm@htt-consult.com wrote:
On 6/28/22 21:06, Samuel Sieb wrote:
On 6/28/22 16:38, Robert Moskowitz wrote:
Well thanks. It is NOT the kernel as I am still locking up. But it is something else related to that update, as I did not get these lockups before. I have updated the bug report and sent a number of related selinux error details and messages
The selinux error is related to the coredump. If you look in the journal, you should see what's crashing. Also, I would have expected abrt to popup a notification. Try running the "Problem Reporting" application and see if it has anything listed.
Can't see a way to copy and paste info from Problem Reporting, so got to retype it:
xorg-x11-drv-ati System failure
Name: Kernel-core Ver: 5.17.12-200 Reported: Cannot be reported
The backtrace does not contain enough meaningful function frames to be reported...
users mailing list -- users@lists.fedoraproject.org To unsubscribe send an email to users-leave@lists.fedoraproject.org Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org Do not reply to spam on the list, report it: https://pagure.io/fedora-infrastructure
I am not entirely sure that Thuderbird is my problem.
I just had to get on $M Meetings and when the app came up, the app window was white. Nothing. I am going to try and figure this out later, but Meetings is not not working, when it did last week, then something really changed.
On 6/29/22 07:58, Roger Heflin wrote:
There are lots of this error going around, going back >10 years.
Thunderbird is what is crashing, see if reverting thunderbird to the prior version makes the crash go away.
In libraries and/or system calls there are often bugs that fail to block bad/invalid call parameters provided by an application that can leave a system in a bad state. Since this error has been reported many times but seemingly never fixed (in the kernel) then I would suspect that it is one of those sorts of bugs. IE thunderbird has a bug in its graphics calls that tells the kernel to do something with completely wrong call parameters and it leaves the module/kernel in a bad state and causes the crash. The bug is probably that the kernel does not return an error to thunderbird and attempts to handle the invalid request and blows the system up. It is also possible that thunderbird makes the graphics calls and died in the middle of the call chain and that leaves the system in a bad state(kernel bug is that this is not gracefully handled) and causes the crash/deadlock.
The stall could be either an invalid request causing the ring to internally crash and stall, or a request that was started but never finished (both being caused by the application doing something invalid and/or crashing).
The kernel bug is on a stall it needs to cleanup and/or the traceback is the cleanup code trying to run but said cleanup code that rarely gets executed does not actually work all of the time (most of the bugs I had to find methods to work around and/or involve a vendor to fix involve the code paths around errors getting triggered and the cleanup/errors handling simply not working right and/or having bad bugs).
On Wed, Jun 29, 2022 at 5:59 AM Robert Moskowitz rgm@htt-consult.com wrote:
On 6/28/22 21:06, Samuel Sieb wrote:
On 6/28/22 16:38, Robert Moskowitz wrote:
Well thanks. It is NOT the kernel as I am still locking up. But it is something else related to that update, as I did not get these lockups before. I have updated the bug report and sent a number of related selinux error details and messages
The selinux error is related to the coredump. If you look in the journal, you should see what's crashing. Also, I would have expected abrt to popup a notification. Try running the "Problem Reporting" application and see if it has anything listed.
Can't see a way to copy and paste info from Problem Reporting, so got to retype it:
xorg-x11-drv-ati System failure
Name: Kernel-core Ver: 5.17.12-200 Reported: Cannot be reported
The backtrace does not contain enough meaningful function frames to be reported...
users mailing list -- users@lists.fedoraproject.org To unsubscribe send an email to users-leave@lists.fedoraproject.org Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org Do not reply to spam on the list, report it: https://pagure.io/fedora-infrastructure
users mailing list -- users@lists.fedoraproject.org To unsubscribe send an email to users-leave@lists.fedoraproject.org Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org Do not reply to spam on the list, report it: https://pagure.io/fedora-infrastructure
Just crashed again.
Restarted. Had nothing running but Firefox. No problem.
Went to cnn.com and waited maybe 1min and the things were locked.
Whatever cnn is doing running things triggers the lockup. No Thunderbirds running around. QEMU is up, but image is not running.
OpenOffice is sitting at recover dialog from last crash.
But start cnn, wait, and lock.
:(
On 6/29/22 09:36, Robert Moskowitz wrote:
I am not entirely sure that Thuderbird is my problem.
I just had to get on $M Meetings and when the app came up, the app window was white. Nothing. I am going to try and figure this out later, but Meetings is not not working, when it did last week, then something really changed.
On 6/29/22 07:58, Roger Heflin wrote:
There are lots of this error going around, going back >10 years.
Thunderbird is what is crashing, see if reverting thunderbird to the prior version makes the crash go away.
In libraries and/or system calls there are often bugs that fail to block bad/invalid call parameters provided by an application that can leave a system in a bad state. Since this error has been reported many times but seemingly never fixed (in the kernel) then I would suspect that it is one of those sorts of bugs. IE thunderbird has a bug in its graphics calls that tells the kernel to do something with completely wrong call parameters and it leaves the module/kernel in a bad state and causes the crash. The bug is probably that the kernel does not return an error to thunderbird and attempts to handle the invalid request and blows the system up. It is also possible that thunderbird makes the graphics calls and died in the middle of the call chain and that leaves the system in a bad state(kernel bug is that this is not gracefully handled) and causes the crash/deadlock.
The stall could be either an invalid request causing the ring to internally crash and stall, or a request that was started but never finished (both being caused by the application doing something invalid and/or crashing).
The kernel bug is on a stall it needs to cleanup and/or the traceback is the cleanup code trying to run but said cleanup code that rarely gets executed does not actually work all of the time (most of the bugs I had to find methods to work around and/or involve a vendor to fix involve the code paths around errors getting triggered and the cleanup/errors handling simply not working right and/or having bad bugs).
On Wed, Jun 29, 2022 at 5:59 AM Robert Moskowitz rgm@htt-consult.com wrote:
On 6/28/22 21:06, Samuel Sieb wrote:
On 6/28/22 16:38, Robert Moskowitz wrote:
Well thanks. It is NOT the kernel as I am still locking up. But it is something else related to that update, as I did not get these lockups before. I have updated the bug report and sent a number of related selinux error details and messages
The selinux error is related to the coredump. If you look in the journal, you should see what's crashing. Also, I would have expected abrt to popup a notification. Try running the "Problem Reporting" application and see if it has anything listed.
Can't see a way to copy and paste info from Problem Reporting, so got to retype it:
xorg-x11-drv-ati System failure
Name: Kernel-core Ver: 5.17.12-200 Reported: Cannot be reported
The backtrace does not contain enough meaningful function frames to be reported...
users mailing list -- users@lists.fedoraproject.org To unsubscribe send an email to users-leave@lists.fedoraproject.org Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org Do not reply to spam on the list, report it: https://pagure.io/fedora-infrastructure
users mailing list -- users@lists.fedoraproject.org To unsubscribe send an email to users-leave@lists.fedoraproject.org Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org Do not reply to spam on the list, report it: https://pagure.io/fedora-infrastructure
users mailing list -- users@lists.fedoraproject.org To unsubscribe send an email to users-leave@lists.fedoraproject.org Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org Do not reply to spam on the list, report it: https://pagure.io/fedora-infrastructure
What version of Firefox are you using? There was a version a couple weeks ago that snuck passed updates-testing into stable that has a nasty memory leak. I ran into it with oomd killing it off, so I masked oomd. Haha, bad idea, the memory leak locked the whole setup, had to force power off (I gave it a few minutes, which is really too long). I think the FF bug got fixed but maybe not always? So as it turns out oomd is doing the right thing in this case, just that we're not getting a desktop notification about what happened.
-- Chris Murphy
On Wed, 2022-06-29 at 19:04 -0400, Robert Moskowitz wrote:
Restarted. Had nothing running but Firefox. No problem.
Went to cnn.com and waited maybe 1min and the things were locked.
Whatever cnn is doing running things triggers the lockup. No Thunderbirds running around. QEMU is up, but image is not running.
OpenOffice is sitting at recover dialog from last crash.
But start cnn, wait, and lock.
You might like the NoScript plugin for Firefox.
On Mon, 2022-06-27 at 00:47 -0400, Robert Moskowitz wrote:
Since updates Saturday night I have been having hard lock ups. It looks like Firefox is the trigger.
Do you have lots of tabs open? Any add-ons/extensions? Do you go through its settings and check what all the preferences are?
I go through the settings, and turn off things that cause traffic without my say-so. There were Firefox data collection options preselected, for instance. There may be pre-fetching (following links on a page that you haven't yet clicked, expecting that you might), but that's hidden in the about:config. The various "warn you about deceptive things" often work by sending data about every link you follow to some external database before loading the site (rather than it checking with one stored on your computer).
Just this last week I had to debug a Mac that kept locking up hard, you had just a few seconds each boot to try something to see what was going wrong. After a few hours of my time being wasted by it, not helped by everything being hidden where you can't see it, touch it, nor remove it , I finally managed to find the fault and eradicate it: Google Chrome installs a hidden Google software updater, and it had wedged hard. I still can't believe how crap Macs are at multitasking. One thing takes control and won't let go, it doesn't even have to be something CPU intensive. Next to impossible to browse another disk drive during a file copying process.
users@lists.stg.fedoraproject.org