CentOS 7.6 update, lightdm fails

[TrevorH]

I don't know what comment 21 is but no, I was referring to the link from our 7.6 release notes https://wiki.centos.org/Manuals/Release ... ntOS7.1810

Users of (at least some) ATI cards may suffer from video corruption or freezes after the 7.6 update. It is reported that this can be fixed by adding a config file to change some options. You can find details of the added file in viewtopic.php?f=49&t=69014&p=289882#p289784

[pcaviede]

The fix is for a different card. My card info is:

lshw -class display
*-display UNCLAIMED
description: VGA compatible controller
product: Rage 3 [Rage XL PCI]
vendor: Advanced Micro Devices, Inc. [AMD/ATI]
physical id: 2
bus info: pci@0000:04:02.0
version: 27
width: 32 bits
clock: 33MHz
capabilities: pm vga_controller bus_master cap_list
configuration: latency=64 mingnt=8
resources: memory:df000000-dfffffff ioport:e800(size=256) memory:defff000-deffffff memory:defc0000-defdffff

So, I don't know what info to add to a conf file to create for the card (the X11 conf directory only has a keyboard config file).

Still stumped...

[pcaviede]

There may me light at the end of the tunnel. Last Jan 29, Redhat acknowledged an issue with drivers for the older ATI cards, in the 7.6 update. This is the link with the info:

https://access.redhat.com/errata/RHBA-2019:0176

The update seems to be for subscribers only, though. I can't find where to download it

[TrevorH]

I would suspect that'll be one of the 36 or 37 updates released by Redhat a day or two ago. Those are all in progress now for CentOS and will be released when ready (hopefully soon). There'll be a mail to the centos-announce mailing list when they're done.

[pcaviede]

Well, the update came out yesterday (Kernel 3.10.0-957.5.1), as well as the new ATI drv 18.1 (which some circles reports it works) and new Xorg server.

Did the update and no joy, the system still gets stuck at boot (see figure). So, no GUI, I can still access via SSH, and Web interfaces, webpage, all work fine

The Xorg server still cannot start. It's giving the same errors as the initial ones posted.

The logs report Failed to start Terminate Plymouth Boot Screen; Failed to start Light Display Manager.

Back to square one, it seems. Am I the only one left on the biosphere using this ATI card?

[desertcat]

Well, the update came out yesterday (Kernel 3.10.0-957.5.1), as well as the new ATI drv 18.1 (which some circles reports it works) and new Xorg server.

Did the update and no joy, the system still gets stuck at boot (see figure). So, no GUI, I can still access via SSH, and Web interfaces, webpage, all work fine

The Xorg server still cannot start. It's giving the same errors as the initial ones posted.

The logs report Failed to start Terminate Plymouth Boot Screen; Failed to start Light Display Manager.

Back to square one, it seems. Am I the only one left on the biosphere using this ATI card?

As just a curiosity do you have an el cheapo nvidia card floating around -- a more recent one would be better, but really *any* nvidia card would do -- and trying it in your system.

Did you try installing lightdm this way:

viewtopic.php?f=47&t=67166

We are just covering all bases. I *suspect* -- as do you -- that it is indeed you graphics card, but an el cheapo nvidia card might be a good stand-in until a fix for your card card comes out.

As to installing lightdm, I had a lot of trouble installing it -- no matter what I tried I could not get it to work, until I stumbled across the above method. It is worth a shot. If it does not work... well you have narrowed down the problem. An el cheapo nvidia card might be a quick temporary fix.

[pcaviede]

Hi:
I had seen that post. The weird thing was that lightdm was working perfectly. until the rollover to 7.6, which in essence trashed Xorg. I did try enablin gdm and disabliong lightdm, but again, no GUI.

I've been fiddling around with the Xorg.conf file following suggestions in some posts, and I've managed to get the glx, ATI modules loaded OK... but the system can't find the mach64 module. Here is the output of the Xorg.log:

(II) LoadModule: "mach64"
[ 66066.296] (WW) Warning, couldn't open module mach64
[ 66066.296] (EE) Failed to load module "mach64" (module does not exist, 0)
[ 66066.296] (EE) No drivers available.
[ 66066.296] (EE)
Fatal server error:
[ 66066.296] (EE) no screens found(EE)
[ 66066.297] (EE)
Please consult the The X.Org Foundation support
at http://wiki.x.org
for help.
[ 66066.297] (EE) Please also check the log file at "/var/log/Xorg.0.log" for additional information.
[ 66066.297] (EE)
[ 66066.297] (EE) Server terminated with error (1). Closing log file.

SO, X can't find the mach64 driver. On the other hand, I can't see why it's invoking it, since it is not in xorg.conf

I wonder is a clean reinstall of Xorg, lightdm and/or anything else would do the trick

[TrevorH]

Nothing in CentOS 7 supplies a mach64 kernel module. As far as I can see, nothing ever has. So the question then becomes where did you get your copy of mach64.ko[.xz] from that was working before?

[pcaviede]

Can't say for certain, but it may come from older Centos version, when I moved the server to Centos years ago. I'm pretty sure it was older than 7, and I've been upgrading as the new versions start coming out. So, the mach64 may still be lingering from those days. Question is: where in the configuration is it still being specified? The driver is nowhere in the system, but X is calling it when you try to run it.

[TrevorH]

No,it would have to be part of the kernel I think.

Do a `locate mach64.ko` and see what you get back.