Error at "triggering uevents," unless I pass radeon.modeset=0. (iMac)

Clqud · 2021-09-06 13:29:05

Hi everyone!

The title pretty much says it all; I cannot boot successfully into Arch unless I pass either nomodeset or radeon.modeset=0, otherwise my screen goes black. I have looked into other forums regarding this issue, however none of the solutions I found have worked for me. Thus, as a last resort, I am coming here.

If you need any logs, just let me know and I will send them!

The iMac I am running is from Mid-2011. (As you all probably know, Macs can run PRIME, but currently that it not my concern, I just want to get out of the black screen.)

Thanks!

seth · 2021-09-06 14:12:26

Can you ssh into the "black screen" boot (or log in blindly)?
Boot w/o nomodeset, try to reboot as clean as possible (ctrl+alt+del?), boot w/ nomodeset and dump

lspci
sudo journalctl -b -1

You can use the tip in the 1st link below to pipe the output into ix.io.

Clqud · 2021-09-06 20:12:47

Sorry for the late reply!

I have tried to start up blindly without success. Unfortunately when I try to reboot, my only option is to force reboot, otherwise it does not respond.

Here is what I got for logs:

lspci: https://termbin.com/rx5s

journalctl: https://termbin.com/b792

seth · 2021-09-06 21:32:39

The journal unfortunately doesn't hold any information about the used/tested GPU(s) but:

00:02.0 Display controller: Intel Corporation 2nd Generation Core Processor Family Integrated Graphics Controller (rev 09)
…
01:00.0 VGA compatible controller: Advanced Micro Devices, Inc. [AMD/ATI] Whistler [Radeon HD 6730M/6770M/7690M XT]

You've two VGA devices - which one is/are your output/s wired to?
Is this supposed to be a hybrid system?

What if you blacklist radeon?
https://wiki.archlinux.org/title/Kernel … and_line_2

Clqud · 2021-09-06 21:57:45

I believe iMacs run hybrid (Integrated Graphics for small tasks, standalone GPU for more intense tasks), so yes. I do not mind which is used at the moment. As for checking the hardware, that would mean opening the system (again), which I would like to avoid if possible haha.

When I blacklist radeon, I get the same result as nomodeset or radeon.modeset=0, which is the following:

"xf86EnableIOPorts: failed to set IOPL for I/O (Operation not permitted) (II) [KMS] drm report modesetting isn't supported. vesa: Refusing to run on UEFI (EE) Fatal server error: (EE) Cannot run in framebuffer mode. Please specify busIDs for all framebuffer devices."

It ends by saying it is "unable to connect to X server: Connection refused."

Last edited by Clqud (2021-09-06 21:59:03)

seth · 2021-09-06 22:02:06

Does

lsmod | grep i915

indicate the i915 module is loaded?
Try adding it (i915) forhttps://wiki.archlinux.org/title/Kernel_mode_setting#Early_KMS_start

Edit: don't forget to rebuild the initramfs.

Last edited by seth (2021-09-06 22:02:33)

Clqud · 2021-09-06 22:19:45

Yeah, i915 is loaded, and still, with Early KMS and a rebuilt initramfs I am getting the same errors.

seth · 2021-09-06 22:21:19

Can you please paste the complete X11 log?

Clqud · 2021-09-06 22:23:49

Here you go,

https://termbin.com/w2y1

Let me know if you need anything else.

Clqud · 2021-09-06 22:25:49

Just now looking, at a glance it seems to still be trying to load the radeon module... Any idea how I should go about changing that?

seth · 2021-09-06 22:36:19

That's the X11 driver, not the kernel module.

Remove xf86-video-ati and xf86-video-intel and xf86-video-vesa (all xf86-video-* packages), remove the blacklisting and all nomodeset'ing efforts and reboot.
If that doesn't help, try to ignore the radeon GPU (X11 currently tries to default to it)

/etc/X11/xorg.conf

Section "ServerLayout"
        Identifier "layout"
        Screen 0 "intel"
        Inactive "radeon"
EndSection

Section "Device"
        Identifier "radeon"
        Driver "modesetting"
        BusID "PCI:1:0:0"
EndSection

Section "Screen"
        Identifier "radeon"
        Device "radeon"
EndSection

Section "Device"
        Identifier "intel"
        Driver "modesetting"
        BusID "PCI:0:2:0"
EndSection

Section "Screen"
        Identifier "intel"
        Device "intel"
EndSection

Clqud · 2021-09-07 00:29:04

Oh my bad, I misread what you asked for. When I look for /etc/X11/xorg.conf, the file does not exist, all I have is xorg.conf.d.

I also tried doing what you said about removing the drivers, and I am able to boot! However, I am trying to boot into a GUI, so I am still getting an error. The good news is that I am no longer getting KMS errors, I am only getting the "Cannot run in framebuffer mode." errors.

Last edited by Clqud (2021-09-07 00:38:16)

seth · 2021-09-07 05:50:34

When I look for /etc/X11/xorg.conf, the file does not exist, all I have is xorg.conf.d.

https://wiki.archlinux.org/title/Xorg#Configuration
You'll my personal - rare - allowance to generate a static server layout in this particular case ;-)

Before that however, please post the xorg log for the latest attempts to boot a GUI.

Clqud · 2021-09-07 13:34:23

I am afraid I am going to send the wrong log again... but here is what I *think* you are looking for.

/var/log/Xorg.0.log: https://termbin.com/ubar

After I configure the static xorg config, I am assuming you want me to fill it with what you sent in post.

EDIT: I was able to get the Intel driver installed and running as long as I removed the AMD and vesa drivers.

Last edited by Clqud (2021-09-07 13:36:34)

seth · 2021-09-07 15:02:27

The reason for the error in the last log is that there's still some configlet hanging around that references the intel driver.
Therefore re-installing xf86-video-intel fixed that bit and if radeon is happy w/ the modesetting driver (please post the last and successful? xorg log for a final review), that's ok.

Otherwise

grep -r intel /{etc,usr/share}/X11/xorg.conf*

Clqud · 2021-09-07 15:15:12

Let me give a quick update of where I am at:

When I run startx, I do not see any errors, my screen just freezes. I am able to boot into console successfully, which is an improvement, but again, once I try to run startx my screen freezes. However, I am still able to power off the machine without force shutting down. So that is a new development as well.

This is currently my xorg log: https://termbin.com/sxys

seth · 2021-09-07 15:58:26

[    34.398] (==) Using config file: "/etc/X11/xorg.conf"
[    34.399] (==) Using system config directory "/usr/share/X11/xorg.conf.d"
[    34.435] (==) ServerLayout "X.org Configured"
[    34.435] (**) |-->Screen "Screen0" (0)
[    34.435] (**) |   |-->Monitor "Monitor0"
[    34.435] (**) |   |-->Device "Card0"
[    34.435] (**) |-->Screen "Screen1" (1)
[    34.435] (**) |   |-->Monitor "Monitor1"
[    34.436] (**) |   |-->Device "Card1"
[    34.436] (**) |-->Input Device "Mouse0"
[    34.436] (**) |-->Input Device "Keyboard0"

The special permit doesn't extend to "xorg -configure" generated stuff

[    36.272] (II) Server terminated successfully (0). Closing log file.

That server starts, ignores the radeon GPU (I assume you injected the configuration from #11?) and terminates immediately what means your xinitrc doesn't run anything meaningful, so please post that as well as the present /etc/X11/xorg.conf
If you don't have an ~/.xinitc, you're using the system default which runs twm and xterm and since you likely won't have installed those: "nothing".

Clqud · 2021-09-11 00:05:49

UPDATE: SOLVED

I was able to dig around some more, and was able to find this article! https://wiki.archlinux.org/title/IMac_( … he_KMS_bug

I do not know how I missed it in the first place, but it solved my issue.

seth · 2021-09-11 06:16:57

Oh, nice - the output is bound to the wrong crtc

Please always remember to mark resolved threads by editing your initial posts subject - so others will know that there's no task left, but maybe a solution to find.
Thanks.

Arch Linux

#1 2021-09-06 13:29:05

Error at "triggering uevents," unless I pass radeon.modeset=0. (iMac)

#2 2021-09-06 14:12:26

Re: Error at "triggering uevents," unless I pass radeon.modeset=0. (iMac)

#3 2021-09-06 20:12:47

Re: Error at "triggering uevents," unless I pass radeon.modeset=0. (iMac)

#4 2021-09-06 21:32:39

Re: Error at "triggering uevents," unless I pass radeon.modeset=0. (iMac)

#5 2021-09-06 21:57:45

Re: Error at "triggering uevents," unless I pass radeon.modeset=0. (iMac)

#6 2021-09-06 22:02:06

Re: Error at "triggering uevents," unless I pass radeon.modeset=0. (iMac)

#7 2021-09-06 22:19:45

Re: Error at "triggering uevents," unless I pass radeon.modeset=0. (iMac)

#8 2021-09-06 22:21:19

Re: Error at "triggering uevents," unless I pass radeon.modeset=0. (iMac)

#9 2021-09-06 22:23:49

Re: Error at "triggering uevents," unless I pass radeon.modeset=0. (iMac)

#10 2021-09-06 22:25:49

Re: Error at "triggering uevents," unless I pass radeon.modeset=0. (iMac)

#11 2021-09-06 22:36:19

Re: Error at "triggering uevents," unless I pass radeon.modeset=0. (iMac)

#12 2021-09-07 00:29:04

Re: Error at "triggering uevents," unless I pass radeon.modeset=0. (iMac)

#13 2021-09-07 05:50:34

Re: Error at "triggering uevents," unless I pass radeon.modeset=0. (iMac)

#14 2021-09-07 13:34:23

Re: Error at "triggering uevents," unless I pass radeon.modeset=0. (iMac)

#15 2021-09-07 15:02:27

Re: Error at "triggering uevents," unless I pass radeon.modeset=0. (iMac)

#16 2021-09-07 15:15:12

Re: Error at "triggering uevents," unless I pass radeon.modeset=0. (iMac)

#17 2021-09-07 15:58:26

Re: Error at "triggering uevents," unless I pass radeon.modeset=0. (iMac)

#18 2021-09-11 00:05:49

Re: Error at "triggering uevents," unless I pass radeon.modeset=0. (iMac)

#19 2021-09-11 06:16:57

Re: Error at "triggering uevents," unless I pass radeon.modeset=0. (iMac)

Board footer