Another radeon/amdgpu problem with an early-GCN based APU

English Hardware
#1

I’ve read a few of the threads here on problems that are cropping up around AMD graphics adapters that use early GCN versions and have tried to follow the advice given in them but I’ve been unable to solve my problem so far.

The hardware that’s causing my problem is an AMD APU. There is no discrete graphics card installed in the system.

I’m running Leap 15, not Tumbleweed as many of the other people who have reported problems are.

Initial symptoms were a boot loop that’s been getting worse/more persistent for the last few months (could there there a timing/scheduling issue at play here too?). After selecting OpenSuSE from the grub boot menu, the system would reboot within a second or two. A few months ago, after a cold start, it would reboot two or three times like this before booting properly and then running without incident until shutdown. Recently the boot loops have become longer (more reboots) before finally running. Now (yesterday) the boot loop appears to be eternal.

I’ve run memtest, smartctl, changed the SATA cable to the only hard drive, checked the PSU and all have passed the tests. I cannot find any physical problem with the system.

I’ve checked journalctl for error messages from the boot loop, but it appears that the reboot happens before anything gets logged. journalctl only has knowledge of the successful boots.

BUT… As soon as I add modprobe.blacklist=radeon to the grub boot options, the computer boots all the way to a kde desktop without any trouble at all. No boot loops at all, just reliable booting to kde every time I ask it to. Albeit booting a kde desktop limited to glorious 1024x768 resolution.

here’s some pertinent info…

Kernel:

rfl@linux-rmlm:/etc/modprobe.d> uname -r
4.12.14-lp150.12.25-default

CPU:

rfl@linux-rmlm:/etc/modprobe.d> cat /proc/cpuinfo 
model name      : AMD A8-6500T APU with Radeon(tm) HD Graphics

GPU (integrated with CPU):

sudo lspci -nnkv -s 00:01.0                
00:01.0 VGA compatible controller [0300]: Advanced Micro Devices, Inc. [AMD/ATI] Richland [Radeon HD 8550D] [1002:999d] (prog-if 00 [VGA contro
ller])
        Subsystem: Micro-Star International Co., Ltd. [MSI] Device [1462:7721]
        Flags: bus master, fast devsel, latency 0, IRQ 10
        Memory at c0000000 (32-bit, prefetchable) [size=256]
        I/O ports at f000 [size=256]
        Memory at feb00000 (32-bit, non-prefetchable) [size=256]
        [virtual] Expansion ROM at 000c0000 [disabled] [size=128]
        Capabilities: [50] Power Management version 3
        Capabilities: [58] Express Root Complex Integrated Endpoint, MSI 00
        Capabilities: [a0] MSI: Enable- Count=1/1 Maskable- 64bit+
        Capabilities: [100] Vendor Specific Information: ID=0001 Rev=1 Len=010 <?>
        Kernel modules: radeon

Installed drivers:

rfl@linux-rmlm:/etc/modprobe.d> zypper search amdgpu
Repository 'Packman' is out-of-date. You can run 'zypper refresh' as root to update it.
Loading repository data...
Reading installed packages...

S  | Name                           | Summary                                                          | Type    
---+--------------------------------+------------------------------------------------------------------+--------
i  | libdrm_amdgpu1                 | Userspace interface for Kernel DRM services for AMD Radeon chips | package
i  | libdrm_amdgpu1-32bit           | Userspace interface for Kernel DRM services for AMD Radeon chips | package
   | libdrm_amdgpu1-32bit-debuginfo | Debug information for package libdrm_amdgpu1                     | package
   | libdrm_amdgpu1-debuginfo       | Debug information for package libdrm_amdgpu1                     | package
i+ | xf86-video-amdgpu              | AMDGPU video driver for the Xorg X server                        | package
   | xf86-video-amdgpu-debuginfo    | Debug information for package xf86-video-amdgpu                  | package
   | xf86-video-amdgpu-debugsource  | Debug sources for package xf86-video-amdgpu                      | package
rfl@linux-rmlm:/etc/modprobe.d> zypper search radeon
Repository 'Packman' is out-of-date. You can run 'zypper refresh' as root to update it.
Loading repository data...
Reading installed packages...

S | Name                              | Summary                                                          | Type    
--+-----------------------------------+------------------------------------------------------------------+--------
i | libdrm_radeon1                    | Userspace interface for Kernel DRM services for AMD Radeon chips | package
i | libdrm_radeon1-32bit              | Userspace interface for Kernel DRM services for AMD Radeon chips | package
  | libdrm_radeon1-32bit-debuginfo    | Debug information for package libdrm_radeon1                     | package
  | libdrm_radeon1-debuginfo          | Debug information for package libdrm_radeon1                     | package
i | libvdpau_radeonsi                 | XVMC state tracker for radeonsi                                  | package
  | libvdpau_radeonsi-32bit           | XVMC state tracker for radeonsi                                  | package
  | libvdpau_radeonsi-32bit-debuginfo | Debug information for package libvdpau_radeonsi                  | package
  | libvdpau_radeonsi-debuginfo       | Debug information for package libvdpau_radeonsi                  | package
i | libvulkan_radeon                  | Mesa vulkan driver for AMD GPU                                   | package
  | libvulkan_radeon-32bit            | Mesa vulkan driver for AMD GPU                                   | package
  | libvulkan_radeon-32bit-debuginfo  | Debug information for package libvulkan_radeon                   | package
  | libvulkan_radeon-debuginfo        | Debug information for package libvulkan_radeon                   | package
  | radeontop                         | View Radeon GPU utilization                                      | package
  | radeontop-lang                    | Translations for package radeontop                               | package

Here’s where I stand now. I’ve got a low-res kde desktop running again and no boot problems. But I don’t have anything other than a low-res last-resort graphics card driver, whatever that is.

The radeon kernel module has been blacklisted by way of /etc/default/grub to solve the boot loop problem (as adding it to /etc/modprobe.d/50-blacklist.conf did not work) and is not currently loaded according to lsmod.

The amdgpu module which was seemingly always installed but never used as it didn’t appear in lsinitrd until after I added it today to the initramfs by running mkinitrd after a ‘modprobe amdgpu’). Now it appears in lsinitrd but not in lsmod. After boot, I can modprobe amdgpu to load it… but nothing uses it and it doesn’t survive as a loaded module between reboots.

In /etc/default/grub I have added

GRUB_CMDLINE_LINUX="radeon.si_support=0 radeon.cik_support=0 modprobe.blacklist=radeon amdgpu.si_support=1 amdgpu.cik_support=1 amdgpu.exp_hw_support=1"

and run grub2-mkconfig -o /boot/grub2/grub.cfg to enable the changes

…and I’m stuck as what to do next…

I don’t actually mind whether radeon or amdgpu is used - just as long as one of them works! And I’m also rather confused why radeon stopped working? Presumably a kernel update broke it, but why?

Any tips?
thanks, Richard

[/size][/size][/size][/size]

#2

Hi and welcome to the Forum :slight_smile:
Unfortunately your GPU is radeon only…

I would imagine the fact that xf86-video-amdgpu is installed could cause issues, so would suggest you remove that package, roll back you boot kernel options to not have the amdgpu/radeon entries. The blacklisting radeon would not have allowed it to load, hence your issues.

Rebuild initrd and grub and see how that goes.

#3

UPDATE: SOLVED but questions remain!

Ok this was simple - instead of approaching the problem from the point of view that the old radeon module was a lost cause and recent amdgpu support for earlier GCN products was the new solution, I thought - what do I have to do to get radeon working again? And it was easy…

What I did was remove the blacklist for the radeon kernel module and updated /etc/default/grub to explicitly turn ON the following options

GRUB_CMDLINE_LINUX="radeon.si_support=1 radeon.cik_support=1"

and everything works as it should. No boot loops, high-res desktop etc…

Now what I’d like to know is why did the radeon module work for so long WITHOUT explicitly specifying these options in grub (albeit with intermittent boot failures) and why is it suddenly required now?

#4

Could it be that you weren’t using the radeon X driver? With a very similar GPU to yours, I’m not:

# grep RETT /etc/os-release
PRETTY_NAME="openSUSE Leap 15.0"
# uname -a
Linux hpg33 4.12.14-lp150.12.25-default #1 SMP Thu Nov 1 06:14:23 UTC 2018 (3fcf457) x86_64 x86_64 x86_64 GNU/Linux
# cat /proc/cmdline
root=LABEL=os150sd32 ipv6.disable=1 net.ifnames=0 noresume vga=791 video=1024x768@60 video=1440x900@60 3
# xrandr | egrep 'onnect|creen|\*' | grep -v disconn | sort -r
Screen 0: minimum 320 x 200, current 1920 x 1200, maximum 16384 x 16384
DP-1 connected primary 1920x1200+0+0 (normal left inverted right x axis y axis) 519mm x 324mm
   1920x1200     59.95*+
# inxi -Gxx
Graphics:  Device-1: Advanced Micro Devices [AMD/ATI] Oland [Radeon HD 8570 / R7 240/340 OEM] vendor: Dell driver: radeon
           v: kernel bus ID: 04:00.0 chip ID: 1002:6611
           Display: server: X.Org 1.19.6 driver: modesetting unloaded: fbdev,vesa alternate: ati resolution: 1920x1200~60Hz
           OpenGL: renderer: AMD OLAND (DRM 2.50.0 / 4.12.14-lp150.12.25-default LLVM 5.0.1) v: 4.5 Mesa 18.0.2 compat-v: 3.0
           direct render: Yes
# ls -l /etc/X11/xorg.conf
ls: cannot access '/etc/X11/xorg.conf': No such file or directory
#5

Hi
Similar but yours is a Southern Island card (GCN radeon or amdgpu), the OP’s card is Northern Island and radeon only.

You can check via modprobe amdgpu |grep 6111 (in your case), OP’s is 999D…

#6

I was focusing on X, not kernel. Is my 6779 not Northern Island too?

# rpm -qa | grep xf86-video
xf86-video-vesa-2.4.0-lp150.1.1.x86_64
xf86-video-fbdev-0.4.4-lp150.1.7.x86_64
# grep RETT /etc/os-release
PRETTY_NAME="openSUSE Leap 15.0"
# uname -a
Linux gx78b 4.12.14-lp150.12.25-default #1 SMP Thu Nov 1 06:14:23 UTC 2018 (3fcf457) x86_64 x86_64 x86_64 GNU/Linux
# cat /proc/cmdline
root=/dev/sda11 noresume net.ifnames=0 ipv6.disable=1 vga=791 video=1024x768@60 video=1440x900@60 3
# xrandr | egrep 'onnect|creen|\*' | grep -v disconn | sort -r
Screen 0: minimum 320 x 200, current 1920 x 1200, maximum 16384 x 16384
DP-1 connected primary 1920x1200+0+0 (normal left inverted right x axis y axis) 519mm x 324mm
   1920x1200     59.95*+
# inxi -Gxx
Graphics:  Device-1: Advanced Micro Devices [AMD/ATI] Caicos [Radeon HD 6450/7450/8450 / R5 230 OEM] vendor: Dell
           driver: radeon v: kernel bus ID: 01:00.0 chip ID: 1002:6779
           Display: server: X.Org 1.19.6 driver: modesetting unloaded: fbdev,vesa alternate: ati resolution: 1920x1200~60Hz
           OpenGL: renderer: AMD CAICOS (DRM 2.50.0 / 4.12.14-lp150.12.25-default LLVM 5.0.1) v: 3.3 Mesa 18.0.2 compat-v: 3.0
           direct render: Yes

Where are you looking to figure out which “island” a chip ID belongs to?

#7

Hi
Yes, Caicos (Engineering Name) is Northern Island…

First stop for me is here;
https://www.x.org/wiki/RadeonFeature/
Else GCN search …

Then check the PCI ID as sometimes the engineering name can be mis-reported.

#8

That I hadn’t known about. It says

For an authoritative list of marketing to family relations, please see this link.
which is where I started, and where Richland (from OP’s lspci output) was (and remains) conspicuously absent.

Else GCN search …
(Not a search I’ve not that I can recall having thought to try explicitly…)

Then check the PCI ID as sometimes the engineering name can be mis-reported.
I then went to https://pci-ids.ucw.cz/read/PC/1002 which is where I stopped.

The xorg wiki makes it clear which kernel driver is applicable, but not that I can tell which X driver. That brings me back to your #2’s (& #5’s) “radeon only” reference in apparent conflict with my demonstration that in X a Southern Islands can run on modesetting and thus not be limited to “radeon”. That in turn leaves open my question in #4.

#9

Hi
I meant radeon vs amdgpu in my references, as in this card will work with only it and not the amdgpu :wink:

For me, a check of the PCI ID and the driver via modinfo command is the definitive way as time goes by cards get added in. Then if you see both drivers present in the output from lspci, it’s time to determine, do I want to switch to amdgpu, which (if any) boot option eg si or cik, then making sure xf86-video-amdgpu is installed, after that should be good to go…

So far above has worked for Mullins, Kabani systems I have, Carrizo/Topaz dual gpu setup is amdgpu from the get go.

#10

Well, that didn’t last long!

Three days of working flawlessly and then abruptly back to an eternal boot loop this morning :frowning:

the options

GRUB_CMDLINE_LINUX="radeon.si_support=1 radeon.cik_support=1"

no longer solve the problem and I have add modprobe.blacklist=radeon OR nomodeset in order to get a low-res desktop

I’ve now removed the package xf86-video-amdgpu as suggested by malcolmlewis and run mkinitrd again. It has not helped.

Here’s the output from the last successful boot to a high-res desktop that relates to radeon and drm

 richard@linux-rmlm:~> sudo journalctl --boot -5 | grep -E "radeon|drm"Dec 13 22:56:15 linux-rmlm kernel: Command line: BOOT_IMAGE=/boot/vmlinuz-4.12.14-lp150.12.25-default root=UUID=838e1269-4ed8-4936-a80d-a8edc857e40a radeon.si_support=1 radeon.cik_support=1 resume=/dev/sda3 splash=silent quiet showopts
Dec 13 22:56:15 linux-rmlm kernel: Kernel command line: BOOT_IMAGE=/boot/vmlinuz-4.12.14-lp150.12.25-default root=UUID=838e1269-4ed8-4936-a80d-a8edc857e40a radeon.si_support=1 radeon.cik_support=1 resume=/dev/sda3 splash=silent quiet showopts
Dec 13 22:56:15 linux-rmlm dracut-cmdline[172]: Using kernel command line parameters: resume=UUID=18dde737-8f2e-4cb4-bdbe-6812d65393f9 root=UUID=838e1269-4ed8-4936-a80d-a8edc857e40a rootfstype=btrfs rootflags=rw,relatime,nossd,space_cache,subvolid=446,subvol=/@/.snapshots/135/snapshot,subvol=@/.snapshots/135/snapshot BOOT_IMAGE=/boot/vmlinuz-4.12.14-lp150.12.25-default root=UUID=838e1269-4ed8-4936-a80d-a8edc857e40a radeon.si_support=1 radeon.cik_support=1 resume=/dev/sda3 splash=silent quiet showopts
Dec 13 22:56:17 linux-rmlm kernel: radeon: unknown parameter 'si_support' ignored
Dec 13 22:56:17 linux-rmlm kernel: radeon: unknown parameter 'cik_support' ignored
Dec 13 22:56:17 linux-rmlm kernel: [drm] radeon kernel modesetting enabled.
Dec 13 22:56:17 linux-rmlm kernel: fb: switching to radeondrmfb from EFI VGA
Dec 13 22:56:17 linux-rmlm kernel: [drm] initializing kernel modesetting (ARUBA 0x1002:0x999D 0x1462:0x7721 0x00).
Dec 13 22:56:17 linux-rmlm kernel: radeon 0000:00:01.0: VRAM: 768M 0x0000000000000000 - 0x000000002FFFFFFF (768M used)
Dec 13 22:56:17 linux-rmlm kernel: radeon 0000:00:01.0: GTT: 1024M 0x0000000030000000 - 0x000000006FFFFFFF
Dec 13 22:56:17 linux-rmlm kernel: [drm] Detected VRAM RAM=768M, BAR=256M
Dec 13 22:56:17 linux-rmlm kernel: [drm] RAM width 64bits DDR
Dec 13 22:56:17 linux-rmlm kernel: [drm] radeon: 768M of VRAM memory ready
Dec 13 22:56:17 linux-rmlm kernel: [drm] radeon: 1024M of GTT memory ready.
Dec 13 22:56:17 linux-rmlm kernel: [drm] Loading ARUBA Microcode
Dec 13 22:56:17 linux-rmlm kernel: [drm] Internal thermal controller without fan control
Dec 13 22:56:17 linux-rmlm kernel: [drm] radeon: dpm initialized
Dec 13 22:56:17 linux-rmlm kernel: [drm] Found VCE firmware/feedback version 50.0.1 / 17!
Dec 13 22:56:17 linux-rmlm kernel: [drm] GART: num cpu pages 262144, num gpu pages 262144
Dec 13 22:56:17 linux-rmlm kernel: [drm] PCIE GART of 1024M enabled (table at 0x00000000001D6000).
Dec 13 22:56:17 linux-rmlm kernel: radeon 0000:00:01.0: WB enabled
Dec 13 22:56:17 linux-rmlm kernel: radeon 0000:00:01.0: fence driver on ring 0 use gpu addr 0x0000000030000c00 and cpu addr 0xffff880235857c00
Dec 13 22:56:17 linux-rmlm kernel: radeon 0000:00:01.0: fence driver on ring 5 use gpu addr 0x0000000000075a18 and cpu addr 0xffffc90001c35a18
Dec 13 22:56:17 linux-rmlm kernel: radeon 0000:00:01.0: fence driver on ring 6 use gpu addr 0x0000000030000c18 and cpu addr 0xffff880235857c18
Dec 13 22:56:17 linux-rmlm kernel: radeon 0000:00:01.0: fence driver on ring 7 use gpu addr 0x0000000030000c1c and cpu addr 0xffff880235857c1c
Dec 13 22:56:17 linux-rmlm kernel: radeon 0000:00:01.0: fence driver on ring 1 use gpu addr 0x0000000030000c04 and cpu addr 0xffff880235857c04
Dec 13 22:56:17 linux-rmlm kernel: radeon 0000:00:01.0: fence driver on ring 2 use gpu addr 0x0000000030000c08 and cpu addr 0xffff880235857c08
Dec 13 22:56:17 linux-rmlm kernel: radeon 0000:00:01.0: fence driver on ring 3 use gpu addr 0x0000000030000c0c and cpu addr 0xffff880235857c0c
Dec 13 22:56:17 linux-rmlm kernel: radeon 0000:00:01.0: fence driver on ring 4 use gpu addr 0x0000000030000c10 and cpu addr 0xffff880235857c10
Dec 13 22:56:17 linux-rmlm kernel: [drm] Supports vblank timestamp caching Rev 2 (21.10.2013).
Dec 13 22:56:17 linux-rmlm kernel: [drm] Driver supports precise vblank timestamp query.
Dec 13 22:56:17 linux-rmlm kernel: radeon 0000:00:01.0: radeon: MSI limited to 32-bit
Dec 13 22:56:17 linux-rmlm kernel: radeon 0000:00:01.0: radeon: using MSI.
Dec 13 22:56:17 linux-rmlm kernel: [drm] radeon: irq initialized.
Dec 13 22:56:18 linux-rmlm kernel: [drm] ring test on 0 succeeded in 2 usecs
Dec 13 22:56:18 linux-rmlm kernel: [drm] ring test on 3 succeeded in 3 usecs
Dec 13 22:56:18 linux-rmlm kernel: [drm] ring test on 4 succeeded in 3 usecs
Dec 13 22:56:18 linux-rmlm kernel: [drm] ring test on 5 succeeded in 2 usecs
Dec 13 22:56:18 linux-rmlm kernel: [drm] UVD initialized successfully.
Dec 13 22:56:18 linux-rmlm kernel: [drm] ring test on 6 succeeded in 17 usecs
Dec 13 22:56:18 linux-rmlm kernel: [drm] ring test on 7 succeeded in 3 usecs
Dec 13 22:56:18 linux-rmlm kernel: [drm] VCE initialized successfully.
Dec 13 22:56:18 linux-rmlm kernel: [drm] ib test on ring 0 succeeded in 0 usecs
Dec 13 22:56:18 linux-rmlm kernel: [drm] ib test on ring 3 succeeded in 0 usecs
Dec 13 22:56:18 linux-rmlm kernel: [drm] ib test on ring 4 succeeded in 0 usecs
Dec 13 22:56:18 linux-rmlm kernel: [drm] ib test on ring 5 succeeded
Dec 13 22:56:19 linux-rmlm kernel: [drm] ib test on ring 6 succeeded
Dec 13 22:56:19 linux-rmlm kernel: [drm] ib test on ring 7 succeeded
Dec 13 22:56:19 linux-rmlm kernel: [drm] Radeon Display Connectors
Dec 13 22:56:19 linux-rmlm kernel: [drm] Connector 0:
Dec 13 22:56:19 linux-rmlm kernel: [drm]   HDMI-A-1
Dec 13 22:56:19 linux-rmlm kernel: [drm]   HPD1
Dec 13 22:56:19 linux-rmlm kernel: [drm]   DDC: 0x6530 0x6530 0x6534 0x6534 0x6538 0x6538 0x653c 0x653c
Dec 13 22:56:19 linux-rmlm kernel: [drm]   Encoders:
Dec 13 22:56:19 linux-rmlm kernel: [drm]     DFP1: INTERNAL_UNIPHY2
Dec 13 22:56:19 linux-rmlm kernel: [drm] Connector 1:
Dec 13 22:56:19 linux-rmlm kernel: [drm]   VGA-1
Dec 13 22:56:19 linux-rmlm kernel: [drm]   HPD2
Dec 13 22:56:19 linux-rmlm kernel: [drm]   DDC: 0x6540 0x6540 0x6544 0x6544 0x6548 0x6548 0x654c 0x654c
Dec 13 22:56:19 linux-rmlm kernel: [drm]   Encoders:
Dec 13 22:56:19 linux-rmlm kernel: [drm]     CRT1: INTERNAL_UNIPHY2
Dec 13 22:56:19 linux-rmlm kernel: [drm]     CRT1: NUTMEG
Dec 13 22:56:19 linux-rmlm kernel: [drm] Connector 2:
Dec 13 22:56:19 linux-rmlm kernel: [drm]   DVI-D-1
Dec 13 22:56:19 linux-rmlm kernel: [drm]   HPD3
Dec 13 22:56:19 linux-rmlm kernel: [drm]   DDC: 0x6550 0x6550 0x6554 0x6554 0x6558 0x6558 0x655c 0x655c
Dec 13 22:56:19 linux-rmlm kernel: [drm]   Encoders:
Dec 13 22:56:19 linux-rmlm kernel: [drm]     DFP2: INTERNAL_UNIPHY
Dec 13 22:56:20 linux-rmlm kernel: [drm] fb mappable at 0xC03DA000
Dec 13 22:56:20 linux-rmlm kernel: [drm] vram apper at 0xC0000000
Dec 13 22:56:20 linux-rmlm kernel: [drm] size 8294400
Dec 13 22:56:20 linux-rmlm kernel: [drm] fb depth is 24
Dec 13 22:56:20 linux-rmlm kernel: [drm]    pitch is 7680
Dec 13 22:56:20 linux-rmlm kernel: fbcon: radeondrmfb (fb0) is primary device
Dec 13 22:56:20 linux-rmlm kernel: radeon 0000:00:01.0: fb0: radeondrmfb frame buffer device
Dec 13 22:56:20 linux-rmlm kernel: [drm] Initialized radeon 2.50.0 20080528 for 0000:00:01.0 on minor 0

The most interesting lines to me are the two early on that say the unrecognised parameters si_support and cik_support are being ignored! Yet the system would not boot at that time unless those “unrecognised” parameters were present ? :’(

Now here’s the same output for the current boot that’s only made possible by the use of nomodeset

richard@linux-rmlm:~> sudo journalctl --boot -0 | grep -E "radeon|drm"Dec 14 14:34:06 linux-rmlm kernel: [drm] VGACON disable radeon kernel modesetting.
Dec 14 14:34:06 linux-rmlm kernel: [drm:radeon_init [radeon]] *ERROR* No UMS support in radeon module!
Dec 14 14:34:08 linux-rmlm kernel: [drm] VGACON disable radeon kernel modesetting.
Dec 14 14:34:08 linux-rmlm kernel: [drm:radeon_init [radeon]] *ERROR* No UMS support in radeon module!

So, nothing to see there then…

I still have

richard@linux-rmlm:~> zypper search -i amdgpu                
Loading repository data...
Reading installed packages...

S | Name                 | Summary                                                          | Type    
--+----------------------+------------------------------------------------------------------+--------
i | libdrm_amdgpu1       | Userspace interface for Kernel DRM services for AMD Radeon chips | package
i | libdrm_amdgpu1-32bit | Userspace interface for Kernel DRM services for AMD Radeon chips | package

[FONT=arial]installed as zypper is proposing a “…nuke the entire site from orbit…” approach to their removal which I’m hesitant to go ahead with.
[/FONT]

richard@linux-rmlm:~> sudo zypper remove libdrm_amdgpu1[sudo] password for root: 
Loading repository data...
Warning: No repositories defined. Operating only with the installed resolvables. Nothing can be installed.
Reading installed packages...
Resolving package dependencies...


The following 26 packages are going to be REMOVED:
  kcm_tablet kcm_tablet-lang libdrm_amdgpu1 libvdpau_r300 libvdpau_r600 libvdpau_radeonsi libvulkan_radeon Mesa Mesa-dri Mesa-gallium Mesa-libGL1 Mesa-libva xf86-input-evdev xf86-input-joystick xf86-input-keyboard xf86-input-libinput xf86-input-mouse xf86-input-void
  xf86-input-wacom xf86-video-ati xf86-video-fbdev xf86-video-mach64 xf86-video-r128 xf86-video-vesa xorg-x11-driver-video xorg-x11-server


26 packages to remove.

Thanks again to malcolmlewis for the https://www.x.org/wiki/RadeonFeature/ link which is now bookmarked for future reference as it’s the best attempt I’ve ever seen at reconciling AMD’s marketing names for parts with their equivalent engineering names. It still leaves me somewhat confused about my own APU as while there is one mention of Richland on the page it is associated with HD6/7xxx GPUs, and there is zero mention of the HD8xxx series of GPUs which my APU has anywhere on the page!

To be continued… I’m following some of the other suggestions given in earlier posts now to gather more information and attempt to fix the problem again…

#11

Hi
The cik and si options are gone… if you check the output from your log it detects VCE, so try radeon.vce=1 and remove the other options.

To see all the kernel module parameters (and for researching) via;


/sbin/modinfo radeon

To see what parameters and settings on a kernel module use;


systool -vm radeon

Then look in the “Parameters:” output, then can look up what it all means… :wink:

#12

Ok, some more information - (still only booted thanks to the use of nomodeset)

richard@linux-rmlm:~> inxi -GxxxGraphics:  Card: Advanced Micro Devices [AMD/ATI] Richland [Radeon HD 8550D] bus-ID: 00:01.0 chip-ID: 1002:999d
           Display Server: X.org 1.19.6 drivers: fbdev,ati (unloaded: modesetting,vesa,radeon)
           tty size: 271x70 Advanced Data: N/A out of X

and

richard@linux-rmlm:~> sudo lspci -nnkvv -s 00:01.0
00:01.0 VGA compatible controller [0300]: Advanced Micro Devices, Inc. [AMD/ATI] Richland [Radeon HD 8550D] [1002:999d] (prog-if 00 [VGA controller])
        Subsystem: Micro-Star International Co., Ltd. [MSI] Device [1462:7721]
        Control: I/O+ Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR- FastB2B- DisINTx-
        Status: Cap+ 66MHz- UDF- FastB2B- ParErr- DEVSEL=fast >TAbort- <TAbort- <MAbort- >SERR- <PERR- INTx-
        Latency: 0, Cache Line Size: 64 bytes
        Interrupt: pin A routed to IRQ 10
        Region 0: Memory at c0000000 (32-bit, prefetchable) [size=256]
        Region 1: I/O ports at f000 [size=256]
        Region 2: Memory at feb00000 (32-bit, non-prefetchable) [size=256]
        [virtual] Expansion ROM at 000c0000 [disabled] [size=128]
        Capabilities: [50] Power Management version 3
                Flags: PMEClk- DSI- D1+ D2+ AuxCurrent=0mA PME(D0-,D1-,D2-,D3hot-,D3cold-)
                Status: D0 NoSoftRst- PME-Enable- DSel=0 DScale=0 PME-
        Capabilities: [58] Express (v2) Root Complex Integrated Endpoint, MSI 00
                DevCap: MaxPayload 128 bytes, PhantFunc 0
                        ExtTag+ RBE+
                DevCtl: Report errors: Correctable- Non-Fatal- Fatal- Unsupported-
                        RlxdOrd+ ExtTag+ PhantFunc- AuxPwr- NoSnoop+
                        MaxPayload 128 bytes, MaxReadReq 128 bytes
                DevSta: CorrErr- UncorrErr- FatalErr- UnsuppReq- AuxPwr- TransPend-
                DevCap2: Completion Timeout: Not Supported, TimeoutDis-, LTR-, OBFF Not Supported
                         AtomicOpsCap: 32bit- 64bit- 128bitCAS-
                DevCtl2: Completion Timeout: 50us to 50ms, TimeoutDis-, LTR-, OBFF Disabled
                         AtomicOpsCtl: ReqEn-
        Capabilities: [a0] MSI: Enable- Count=1/1 Maskable- 64bit+
                Address: 0000000000000000  Data: 0000
        Capabilities: [100 v1] Vendor Specific Information: ID=0001 Rev=1 Len=010 <?>
        Kernel modules: radeon

richard@linux-rmlm:~> sudo modinfo radeon | grep 999Dalias:          pci:v00001002d0000999Dsv*sd*bc*sc*i*

richard@linux-rmlm:~> sudo modinfo radeon | egrep -i "trinity|aruba|richland"firmware:       radeon/ARUBA_rlc.bin
firmware:       radeon/ARUBA_me.bin
firmware:       radeon/ARUBA_pfp.bin

And I’m at a loss as to what to do next[/size][/size][/size][/size]

#13

I’ll try those suggestions in a minute malcolm - I was typing my last message when you replied

#14

systool fails to provide any information on radeon but can I presume that modinfo is giving a list of the same options?

richard@linux-rmlm:~> sudo systool -vm radeon
Error opening module radeon
richard@linux-rmlm:~> sudo modinfo radeon | grep parmparm:           no_wb:Disable AGP writeback for scratch registers (int)
parm:           modeset:Disable/Enable modesetting (int)
parm:           dynclks:Disable/Enable dynamic clocks (int)
parm:           r4xx_atom:Enable ATOMBIOS modesetting for R4xx (int)
parm:           vramlimit:Restrict VRAM for testing, in megabytes (int)
parm:           agpmode:AGP Mode (-1 == PCI) (int)
parm:           gartsize:Size of PCIE/IGP gart to setup in megabytes (32, 64, etc., -1 = auto) (int)
parm:           benchmark:Run benchmark (int)
parm:           test:Run tests (int)
parm:           connector_table:Force connector table (int)
parm:           tv:TV enable (0 = disable) (int)
parm:           audio:Audio enable (-1 = auto, 0 = disable, 1 = enable) (int)
parm:           disp_priority:Display Priority (0 = auto, 1 = normal, 2 = high) (int)
parm:           hw_i2c:hw i2c engine enable (0 = disable) (int)
parm:           pcie_gen2:PCIE Gen2 mode (-1 = auto, 0 = disable, 1 = enable) (int)
parm:           msi:MSI support (1 = enable, 0 = disable, -1 = auto) (int)
parm:           lockup_timeout:GPU lockup timeout in ms (default 10000 = 10 seconds, 0 = disable) (int)
parm:           fastfb:Direct FB access for IGP chips (0 = disable, 1 = enable) (int)
parm:           dpm:DPM support (1 = enable, 0 = disable, -1 = auto) (int)
parm:           aspm:ASPM support (1 = enable, 0 = disable, -1 = auto) (int)
parm:           runpm:PX runtime pm (1 = force enable, 0 = disable, -1 = PX only default) (int)
parm:           hard_reset:PCI config reset (1 = force enable, 0 = disable (default)) (int)
parm:           vm_size:VM address space size in gigabytes (default 4GB) (int)
parm:           vm_block_size:VM page table size in bits (default depending on vm_size) (int)
parm:           deep_color:Deep Color support (1 = enable, 0 = disable (default)) (int)
parm:           use_pflipirq:Pflip irqs for pageflip completion (0 = disable, 1 = as fallback, 2 = exclusive (default)) (int)
parm:           bapm:BAPM support (1 = enable, 0 = disable, -1 = auto) (int)
parm:           backlight:backlight support (1 = enable, 0 = disable, -1 = auto) (int)
parm:           auxch:Use native auxch experimental support (1 = enable, 0 = disable, -1 = auto) (int)
parm:           mst:DisplayPort MST experimental support (1 = enable, 0 = disable) (int)
parm:           uvd:uvd enable/disable uvd support (1 = enable, 0 = disable) (int)
parm:           vce:vce enable/disable vce support (1 = enable, 0 = disable) (int)

Ok, now I’ve tried with radeon.vce=1 and the result is a boot loop… but with radeon.vce=0 the system boots normally to a high-res kde desktop! Behaviour is so far consistent between reboots. Thanks again malcolm! I could have googled/researched all the options given and wold never have guessed that it would be disabling VCE that would solve my problem…

#15

Hi
So if you check the info on the Radeon Feature page, you have had to disable the Video Compression Engine, maybe it needs a bug report…?

#16

Yes, I was curious about that - having to disable VCE on a chip that allegedly supports it (albeit only version 1.0) did seem a little odd. I’ll go and open a bug report for it.

#17

I just wanted to post this message to say thank you and add that the same behaviour can be observed on at least one other CPU.

I just installed leap 15 on a relative’s pc which has a AMD A4-5300 CPU with a Radeon HD 7480D (and no discrete GPU) and ran into similar issues, ie boot loops with the default kernel parameters and successful boot but low res only when adding nomodeset. I find everything related to displays to be very obscure (no pun intended), but the option radeon.vce=0 did the trick for me too, and I now have a proper resolution set, along with a graphical loading screen before the login prompt.

So once again, thanks guys.