I enabled amdgpu using the manual you sent but I saw something strange. when I rebooted my system cpu temperature was getting very high. I disabled that and enabled radeon again.
And I checked some logs that you can see below:
amirali@amiralihost:~$ cat ~/.local/share/xorg/Xorg.0.log | grep -iE "radeon|EE|WW"
[ 58.361] Current Operating System: Linux amiralihost 6.15.9-201.fc42.x86_64 #1 SMP PREEMPT_DYNAMIC Sat Aug 2 11:37:34 UTC 2025 x86_64
(WW) warning, (EE) error, (NI) not implemented, (??) unknown.
[ 58.363] (==) No Layout section. Using the first Screen section.
[ 58.363] (==) No screen section available. Using defaults.
[ 58.363] (**) |-->Screen "Default Screen Section" (0)
[ 58.363] (==) No monitor specified for screen "Default Screen Section".
[ 58.367] (II) systemd-logind: took control of session /org/freedesktop/login1/session/_31
[ 58.369] (II) xfree86: Adding drm device (/dev/dri/card2)
[ 58.371] (II) xfree86: Adding drm device (/dev/dri/card1)
[ 58.387] (WW) Warning, couldn't open module ati
[ 58.387] (EE) Failed to load module "ati" (module does not exist, 0)
[ 58.390] (WW) Warning, couldn't open module fbdev
[ 58.390] (EE) Failed to load module "fbdev" (module does not exist, 0)
[ 58.390] (WW) Warning, couldn't open module vesa
[ 58.390] (EE) Failed to load module "vesa" (module does not exist, 0)
[ 58.390] (WW) VGA arbiter: cannot open kernel arbiter, no multi-card support
[ 58.390] (II) modeset(0): Creating default Display subsection in Screen section
"Default Screen Section" for depth/fbbpp 24/32
[ 58.971] (II) modeset(0): Year: 2013 Week: 49
[ 58.971] (II) modeset(0): redX: 0.580 redY: 0.335 greenX: 0.330 greenY: 0.565
[ 59.021] (II) modeset(G0): glamor X acceleration enabled on OLAND (radeonsi, , ACO, DRM 2.50, 6.15.9-201.fc42.x86_64)
[ 59.144] (II) modeset(G0): [DRI2] DRI driver: radeonsi
[ 59.144] (II) modeset(G0): [DRI2] VDPAU driver: radeonsi
[ 59.148] (II) Initializing extension MIT-SCREEN-SAVER
[ 59.193] (II) GLX: Initialized DRI2 GL provider for screen 0
[ 59.193] (II) Initializing extension XFree86-VidModeExtension
[ 59.193] (II) Initializing extension XFree86-DGA
[ 59.195] (II) modeset(0): Setting screen physical size to 361 x 203
[ 59.376] (II) This device may have been added with another device file.
[ 59.377] (II) config/udev: Adding input device Sleep Button (/dev/input/event1)
[ 59.377] (**) Sleep Button: Applying InputClass "libinput keyboard catchall"
[ 59.377] (**) Sleep Button: Applying InputClass "system-keyboard"
[ 59.377] (II) Using input driver 'libinput' for 'Sleep Button'
[ 59.378] (**) Sleep Button: always reports core events
[ 59.380] (II) event1 - Sleep Button: is tagged by udev as: Keyboard
[ 59.381] (II) event1 - Sleep Button: device is a keyboard
[ 59.381] (II) event1 - Sleep Button: device removed
[ 59.381] (II) XINPUT: Adding extended input device "Sleep Button" (type: KEYBOARD, id 10)
[ 59.383] (II) event1 - Sleep Button: is tagged by udev as: Keyboard
[ 59.383] (II) event1 - Sleep Button: device is a keyboard
[ 59.385] (II) This device may have been added with another device file.
[ 59.386] (II) This device may have been added with another device file.
[ 59.387] (II) This device may have been added with another device file.
[ 59.401] (II) This device may have been added with another device file.
[ 59.412] (II) This device may have been added with another device file.
[ 59.448] (II) This device may have been added with another device file.
[ 59.449] (II) This device may have been added with another device file.
[ 100.120] (II) event1 - Sleep Button: device removed
[ 101.949] (WW) xf86CloseConsole: KDSETMODE failed: Input/output error
[ 101.949] (WW) xf86CloseConsole: VT_GETMODE failed: Input/output error
[ 101.949] (WW) xf86OpenConsole: VT_GETSTATE failed: Input/output error
amirali@amiralihost:~$ dmesg | grep -i radeon
dmesg: read kernel buffer failed: Operation not permitted
amirali@amiralihost:~$ sudo dmesg | grep -i radeon
[ 6.746876] [drm] radeon kernel modesetting enabled.
[ 6.747285] radeon 0000:04:00.0: enabling device (0000 -> 0003)
[ 6.825266] radeon 0000:04:00.0: VRAM: 2048M 0x0000000000000000 - 0x000000007FFFFFFF (2048M used)
[ 6.825273] radeon 0000:04:00.0: GTT: 2048M 0x0000000080000000 - 0x00000000FFFFFFFF
[ 6.825413] [drm] radeon: 2048M of VRAM memory ready
[ 6.825425] [drm] radeon: 2048M of GTT memory ready.
[ 6.843684] UBSAN: array-index-out-of-bounds in drivers/gpu/drm/radeon/radeon_atombios.c:2720:34
[ 6.843733] radeon_atombios_parse_power_table_6+0x3bc/0x3e0 [radeon]
[ 6.843886] radeon_atombios_get_power_modes+0x220/0x240 [radeon]
[ 6.843997] radeon_pm_init_dpm+0x77/0x240 [radeon]
[ 6.844134] si_init+0x223/0x470 [radeon]
[ 6.844277] radeon_device_init+0x484/0xa60 [radeon]
[ 6.844380] radeon_driver_load_kms+0xa6/0x230 [radeon]
[ 6.844478] radeon_pci_probe+0xef/0x1b0 [radeon]
[ 6.844630] ? __pfx_radeon_module_init+0x10/0x10 [radeon]
[ 6.845092] UBSAN: array-index-out-of-bounds in drivers/gpu/drm/radeon/si_dpm.c:6824:32
[ 6.845137] si_parse_power_table+0x3d6/0x420 [radeon]
[ 6.845354] si_dpm_init+0x226/0x6a0 [radeon]
[ 6.845501] radeon_pm_init_dpm+0xe3/0x240 [radeon]
[ 6.845635] si_init+0x223/0x470 [radeon]
[ 6.845776] radeon_device_init+0x484/0xa60 [radeon]
[ 6.845877] radeon_driver_load_kms+0xa6/0x230 [radeon]
[ 6.845986] radeon_pci_probe+0xef/0x1b0 [radeon]
[ 6.846136] ? __pfx_radeon_module_init+0x10/0x10 [radeon]
[ 6.855905] [drm] radeon: dpm initialized
[ 6.891656] radeon 0000:04:00.0: WB enabled
[ 6.891662] radeon 0000:04:00.0: fence driver on ring 0 uses gpu addr 0x0000000080000c00
[ 6.891667] radeon 0000:04:00.0: fence driver on ring 1 uses gpu addr 0x0000000080000c04
[ 6.891671] radeon 0000:04:00.0: fence driver on ring 2 uses gpu addr 0x0000000080000c08
[ 6.891675] radeon 0000:04:00.0: fence driver on ring 3 uses gpu addr 0x0000000080000c0c
[ 6.891679] radeon 0000:04:00.0: fence driver on ring 4 uses gpu addr 0x0000000080000c10
[ 6.892498] radeon 0000:04:00.0: fence driver on ring 5 uses gpu addr 0x0000000000075a18
[ 6.892776] radeon 0000:04:00.0: radeon: MSI limited to 32-bit
[ 6.892867] radeon 0000:04:00.0: radeon: using MSI.
[ 6.892920] [drm] radeon: irq initialized.
[ 7.752806] radeon 0000:04:00.0: vgaarb: VGA decodes changed: olddecodes=io+mem,decodes=none:owns=none
[ 7.930165] [drm] Radeon Display Connectors
[ 7.934390] [drm] Initialized radeon 2.50.0 for 0000:04:00.0 on minor 1
[ 7.934663] radeon 0000:04:00.0: [drm] Cannot find any crtc or sizes
[ 33.361999] radeon 0000:04:00.0: WB enabled
[ 33.362006] radeon 0000:04:00.0: fence driver on ring 0 uses gpu addr 0x0000000080000c00
[ 33.362012] radeon 0000:04:00.0: fence driver on ring 1 uses gpu addr 0x0000000080000c04
[ 33.362017] radeon 0000:04:00.0: fence driver on ring 2 uses gpu addr 0x0000000080000c08
[ 33.362021] radeon 0000:04:00.0: fence driver on ring 3 uses gpu addr 0x0000000080000c0c
[ 33.362025] radeon 0000:04:00.0: fence driver on ring 4 uses gpu addr 0x0000000080000c10
[ 33.362852] radeon 0000:04:00.0: fence driver on ring 5 uses gpu addr 0x0000000000075a18
[ 44.433022] radeon 0000:04:00.0: WB enabled
[ 44.433026] radeon 0000:04:00.0: fence driver on ring 0 uses gpu addr 0x0000000080000c00
[ 44.433030] radeon 0000:04:00.0: fence driver on ring 1 uses gpu addr 0x0000000080000c04
[ 44.433032] radeon 0000:04:00.0: fence driver on ring 2 uses gpu addr 0x0000000080000c08
[ 44.433034] radeon 0000:04:00.0: fence driver on ring 3 uses gpu addr 0x0000000080000c0c
[ 44.433036] radeon 0000:04:00.0: fence driver on ring 4 uses gpu addr 0x0000000080000c10
[ 44.433845] radeon 0000:04:00.0: fence driver on ring 5 uses gpu addr 0x0000000000075a18
[ 72.624472] radeon 0000:04:00.0: WB enabled
[ 72.624479] radeon 0000:04:00.0: fence driver on ring 0 uses gpu addr 0x0000000080000c00
[ 72.624486] radeon 0000:04:00.0: fence driver on ring 1 uses gpu addr 0x0000000080000c04
[ 72.624490] radeon 0000:04:00.0: fence driver on ring 2 uses gpu addr 0x0000000080000c08
[ 72.624494] radeon 0000:04:00.0: fence driver on ring 3 uses gpu addr 0x0000000080000c0c
[ 72.624498] radeon 0000:04:00.0: fence driver on ring 4 uses gpu addr 0x0000000080000c10
[ 72.625307] radeon 0000:04:00.0: fence driver on ring 5 uses gpu addr 0x0000000000075a18
amirali@amiralihost:~$ lspci -k -d ::03xx
00:02.0 VGA compatible controller: Intel Corporation Haswell-ULT Integrated Graphics Controller (rev 0b)
Subsystem: ASUSTeK Computer Inc. Device 192d
Kernel driver in use: i915
Kernel modules: i915
04:00.0 VGA compatible controller: Advanced Micro Devices, Inc. [AMD/ATI] Mars LE [Radeon HD 8530M / R5 M240]
Subsystem: ASUSTeK Computer Inc. Device 192d
Kernel driver in use: radeon
Kernel modules: radeon, amdgpu
amirali@amiralihost:~$
Hmmm, maybe âamdgpuâ does not work well with your system and causes high CPU load. Besides the issue with high temperatures, did your system then recognize the GPU with the âamdgpuâ driver active? If not, there is probably no point in looking into it further.
Oh, one more thing about the high CPU temperature: considering that the GPU is a 10+ year old model and the logs reference an Intel Haswell generation CPU (similarly old), how old is the laptop? The reason I am asking is that common thermal pastes tend to dry out after a few years and, as a result, no longer properly conduct heat from the chips to the heatsinks anymore. How are the temperatures when you put some other load on the system? While this isnât directly related to your driver issue, you might also want to re-apply fresh thermal paste at some point (or take the laptop to a shop to do it for you, laptop chips have exposed dies and are kind of fragile).
Yes, it recongnized it and when I checked kernel modules I saw this:
Kernel driver in use: amdgpu
And there is no problem with the thermal paste, I changed them a few months ago. And when I had not changed the driver, I had no problems. The problem is definitely with the driver.
Your GPU is one of the very early GCN cards. It should be well supported by amdgpu.
You can install and check thermal sensors using the lm_sensors package.
sudo dnf install lm_sensors followed by sudo sensors-detect. When running through the configuration select yes for everything. Then any time you want to check system temps, you just need to type sensors.
How high were the temps getting? Laptops traditionally run hotter than desktops, simply due to their packaging format. I have used every thing from 11 inch Gigabyte X11âs that ran in the 50s (Celsius) just running Linux, and would run all the way into the low to mid 80s when some heavy background process was running. In other laptop systems it barely made a change because the cooling system was better.
Processors and GPUs have thermal limit settings where they will ramp down their performance when itâs getting dangerously hot, or they will shut themselves down entirely for protection if thereâs some kind of massive malfunction that causes extreme temperatures.
The R5 doesnât support amdgpu fully. The R7 does (did).
Legacy GPUs (like the R5 M230) use GCN 1.0 architecture and primarily rely on the radeon driver
Modern GPUs (including many R7 series cards) use newer architectures and prefer the amdgpu driver
While some newer R5 models do support amdgpu, the specific model youâre asking about (R5 M230) uses the older architecture and requires the radeon driver
. This explains why some R5 series cards support amdgpu while others donât - it depends entirely on their underlying architecture generation.
I wish theyâd make a feature matrix for amdgpu like they did for the Radeon driver. Would be so much easier to browse than trying to remember what is and isnât supported via issue trackers and update notes. Just a one stop shop where you could point people to and say âthis is whatâs supported on your hardwareâŚâ
It was not a normal temperature. It was above 70 degrees Celsius. While in normal mode the temperature of my laptop does not exceed 50
When I activated amdgpu for my hardware, I tried to test it using vkcube --gpu_number but only intel and llvmpipe GPUs worked not AMD.
Is there any way to make radeon driver work? When I installed it the operating system had selected Radeon driver by default. Maybe thereâs a way to use radeon for it.
How can I know which of GPU 2 and GPU 3 labels corresponds to which hardware?
When I execute vkcube --gpu_number 0 It uses Intel GPU and the usage of GPU 3 label get high in the graph. And when I execute vkcube --gpu_number 1 it uses llvmpipe (CPU) the usage of GPU 3 rises again.
It seems GPU 2 is AMD because OS has been detected that, Even though it seems like it canât be used. Right?
Iâm still not sure if AMD on radeon works or not. the output of this command was intel. maybe AMD will be activated if I open a program with a heavy graphic. Does anyone know a way to recognized AMD gpu will be used or not? How do I make sure it is actived?
Do you know of a way to force a program to run on ŮAMD? Or any other way to find out that it works or not? Any terminal commands?
Ok. It maybe work. Actually I didnât see the name of my GPU in Info center and I was unable to test it using vkcube and glxgears. So thought it has a problem.
I think itâs a hole in the driver. Iâve had a similar problem. I have two nominally identical Dell OEM R7 350 cards, both 4GB. Youâd think theyâd be the same but one has the TURKS chip and the other has the OLAND chip.
One (I believe itâs the TURKS chip, it doesnât want to tell me) is in my PCLinuxOS/KDE system and works fine at any resolution. Itâs presently running at 1600x1200.
The other (OLAND, if I didnât get them mixed up) I had to eject from my Fedora and OpenMandriva (both KDE) systems, as it would work but only at 800x600, yet the driver appears to ID the card and load normally.
The card works fine in Windows, at any resolution.
Also affected, an R7-250 2GB, I believe this one was also OLAND.
After much thrashing around I concluded that there is a gap in the linux driver support for this particular Radeon chip, or some subset of it, and it simply is not supported.
Solution: find another GPU (or in the one case, just use the onboard Intel GPU, which works fine).
Someday I need to sort out all of these (very common OEM vidcards for business workstations) and submit a bug report.
My AMD GPUâs usage is 0% on windows. Because it is for heavy rendering and both windows and linux do not use it much and for every program. I found that when the gpu which it is set for vkcube is llvmpipe or intel just one of the gpus in kde system monitor was working and the usage of the one other was never raising up. I think the idle gpu in system monitor is my amd radeon. my problem is I cannot test the radeon gpu by changing the value of gpu_number parameter in vkcube. may be it starts working when executing a heavy graphical program. I do not know. the output of the command indicated that kernel has detected my gpu and it had chose the radeon driver for that.
