Hi.
After installing amdgpu v19.50 on CentOS 8.1, my gdm crashes, so that I can't use GUI.
The below is the specification of my desktop environment.
------
CPU: Intel(R) Core(TM) i3-8100 CPU @ 3.60GHz
Graphics: Radeon RX 570 + Intel onboard graphics
(Using dual monitors)
Motherboard: Asrock Z370M Pro4
Install command:
sudo ./amdgpu-install --pro --opencl=legacy,pal -y
------
If amdgpu v19.50 driver is not installed, both monitors display well.
But, if it is installed, gdm seems to be crashed.
This is the log in /var/log/messages about it.
I omit the timestamp and process info in each line of the logs such as "Jan 18 20:52:53 localhost /usr/libexec/gdm-x-session[5066]:"
(EE)
(EE) Backtrace:
(EE) 0: /usr/libexec/Xorg (xorg_backtrace+0x7d) [0x563adbae9f2d]
(EE) 1: /usr/libexec/Xorg (0x563adb932000+0x1bbcdd) [0x563adbaedcdd]
(EE) 2: /lib64/libpthread.so.0 (0x7f729a006000+0x12dd0) [0x7f729a018dd0]
(EE) 3: /opt/amdgpu/lib64/libdrm_amdgpu.so.1 (amdgpu_query_info+0x4) [0x7f72958ce384]
(EE) 4: /opt/amdgpu/lib64/dri/amdgpu_dri.so (0x7f7290788000+0x163563c) [0x7f7291dbd63c]
(EE) 5: /opt/amdgpu/lib64/dri/amdgpu_dri.so (0x7f7290788000+0x1637303) [0x7f7291dbf303]
(EE) 6: /opt/amdgpu/lib64/dri/amdgpu_dri.so (0x7f7290788000+0x13ea738) [0x7f7291b72738]
(EE) 7: /opt/amdgpu/lib64/dri/amdgpu_dri.so (0x7f7290788000+0x1415887) [0x7f7291b9d887]
(EE) 8: /opt/amdgpu/lib64/dri/amdgpu_dri.so (0x7f7290788000+0xc52195) [0x7f72913da195]
(EE) 9: /opt/amdgpu/lib64/dri/amdgpu_dri.so (0x7f7290788000+0x142bc4c) [0x7f7291bb3c4c]
(EE) 10: /opt/amdgpu/lib64/dri/amdgpu_dri.so (0x7f7290788000+0x163934b) [0x7f7291dc134b]
(EE) 11: /opt/amdgpu/lib64/dri/amdgpu_dri.so (0x7f7290788000+0x12f90ee) [0x7f7291a810ee]
(EE) 12: /opt/amdgpu/lib64/dri/amdgpu_dri.so (0x7f7290788000+0x12f95f6) [0x7f7291a815f6]
(EE) 13: /opt/amdgpu/lib64/dri/amdgpu_dri.so (eglInitialize+0xa9) [0x7f7291a84a89]
(EE) 14: /usr/lib64/xorg/modules/libglamoregl.so (glamor_egl_init+0xa3) [0x7f728fbd4f53]
(EE) 15: /usr/lib64/xorg/modules/drivers/modesetting_drv.so (0x7f72947b5000+0xbc5a) [0x7f72947c0c5a]
(EE) 16: /usr/libexec/Xorg (InitOutput+0xb0e) [0x563adb9cfc2e]
(EE) 17: /usr/libexec/Xorg (0x563adb932000+0x60007) [0x563adb992007]
(EE) 18: /lib64/libc.so.6 (__libc_start_main+0xf3) [0x7f7299c66873]
(EE) 19: /usr/libexec/Xorg (_start+0x2e) [0x563adb97bbbe]
(EE)
(EE) Segmentation fault at address 0x10
(EE)
I tried to find the difference between amdgpu-installed case and amdgpu-NOT-installed case.
In amdgpu-installed dmesg log, the below log is a suspected part:
[ 3.138856] i915 0000:00:02.0: enabling device (0000 -> 0003)
[ 3.139487] checking generic (c0000000 8ca000) vs hw (b0000000 10000000)
[ 3.140209] [drm] Supports vblank timestamp caching Rev 2 (21.10.2013).
[ 3.140209] [drm] Driver supports precise vblank timestamp query.
[ 3.141733] i915 0000:00:02.0: Direct firmware load for i915/kbl_dmc_ver1_04.bin failed with error -2
[ 3.141734] i915 0000:00:02.0: Failed to load DMC firmware i915/kbl_dmc_ver1_04.bin. Disabling runtime power management.
[ 3.141735] i915 0000:00:02.0: DMC firmware homepage: https://git.kernel.org/pub/scm/linux/kernel/git/firmware/linux-firmware.git/tree/i915
[ 3.143682] [drm] Initialized i915 1.6.0 20190207 for 0000:00:02.0 on minor 0
On the other hand, amdgpu-NOT-installed dmesg log is the below:
[ 3.090385] i915 0000:00:02.0: enabling device (0000 -> 0003)
[ 3.090978] checking generic (c0000000 8ca000) vs hw (b0000000 10000000)
[ 3.091855] [drm] Supports vblank timestamp caching Rev 2 (21.10.2013).
[ 3.091855] [drm] Driver supports precise vblank timestamp query.
[ 3.092821] [drm] Finished loading DMC firmware i915/kbl_dmc_ver1_04.bin (v1.4)
[ 3.103428] [drm] Initialized i915 1.6.0 20190207 for 0000:00:02.0 on minor 0
I don't know why i915/kbl_dmc_ver1_04.bin loading is failed if amdgpu v19.50 is installed.
How can I solve it?
I attached the failed log. (amdgpu installed)
