I have two computers, each having video issues. I’ll save the nvidia issue for another thread.
I am running a Dell computer with Intel HD graphics 630 (integrated.) At random, the screen goes blank for a few seconds and when it comes back, the screen shudders and is almost unusable. My “cure” is to go to Youtube and start a random video. That ends the shudder and blanking. I’m on the x11 graphics platform using KDE. The kernel version is 6.14.6-1-default.
This computer is old (so am I) but satisfies my needs. I would like to fix this issue before I tackle the much worse nvidia issue
I don’t yet consider anything with 630 old yet. You might try installing kernel-longterm-6.12.30 to see if it changes behavior. Before installing 6.14.6 I was running 20250522 and 6.12.27 on one of my two 630s with no issues. Can you reproduce using a different user, preferably a new one? Are you sure your monitor cable remains firmly connected? What cable type are you using? Please show input/output from `inxi -GSaz’. If you have a KDE problem, I can’t help more than suggesting to log out of Plasma, and from a shell prompt login, delete ~/.cache/ before logging back into Plasma.
@Prexy I have multiple Intel CommetLake and CoffeeLake 630 GPU’s (Dell OptiPlex XE3 and 3080 Micro) here, all good. So this Dell is a desktop? If so check the monitor cable connections are seated properly…
I have not seen the problem again in the last few days. If it happens again, I will go out and get a new hdmi cable to see if that resolves the issue. In the meantime, here is the inxi ouput:
I don’t know how to log out of Plasma or how to switch to Wayland. Thank you for trying to help me. I will start another thread for the nvidia issues in my other pc when I disconnect this pc and connect the other.
In your main toolbar menu there multiple selections to quit your session. Choose the one to quit instead of shutdown or reboot. Once logged out you should be at your SDDM greeter, where you can choose Plasma session type Wayland.
Besides different kernel, 3 displays, 4 CPU outputs, DPI, & not running Plasma or SDDM, the Kaby Lake I have booted to 20250601 now does have other differences in inxi output:
I noticed:
1-i915.enable_guc=2 on my kernel cmdline. I don’t remember what it does, but once upon a time I couldn’t boot without it. Don’t concern yourself with the other options you don’t have there, but if trouble arises, you might wish to look up what it does.
2-newer gcc on mine than yours, whatever that might matter about. Maybe you should try a zypper ve to see if any installed packages aren’t right.
3-I have 2X the RAM devoted to GPU.
4-I have more of Mesa installed than you do, which shows up in yours as (KBL GT2) driver: N/A and less total information in your Vulkan section. What Vulkan actually can benefit Intel GPUs I have no idea.
5-You don’t have clinfo installed, which is probably irrelevant, but could be responsible for other differences besides its listing in Info: Tools:.
The problem has not recurred. So, I have not tried to “fix” it. I am about to start another thread concerning my pc with the nvidia chip. In the meantime, in response to the question of reliability of the hdmi cable: Both pc’s are dual boot with Windows. I booted each of them into Windows and there were no video problems. So, the cable itself is ok.
Seems like as long as it’s been around, and what it does, I’d think the average user would want i915.enable_guc=2 to be a default bootloader configuration where applicable (not AMD GPUs, not unsupported oldies). Do we know it wasn’t obsoleted in newer kernels or CPUs by kernel changes? The examples there used kernel 4.19. It’s on my cmdline, but cat /sys/kernel/debug/dri/0/i915_guc_load_status produces "No such file or directory".
@mrmazda use journalctl -b | grep -E "guc|huc" will show, also it taint’s the kernel. On newer hardware that meet guc 3 requirements it automatically enables, but no taint.
Edit cat /sys/kernel/debug/dri/0/i915_gpu_info has all the gpu details…
FWIW, I have a Dell laptop using a similar Intel HD Graphics 620 chipset. I have no such boot parameter employed. Running Plasma Wayland with no issues. For reference:
My brain’s been in a fog for the past week over chronic lack of sleep, among other problems, only modestly better today. I don’t get what I’m looking at:
# cat /sys/kernel/debug/dri/0/i915_gpu_info | grep guc
GuC firmware: i915/kbl_guc_70.1.1.bin
GuC firmware wanted: i915/kbl_guc_70.1.1.bin
has_guc_deprivilege: no
has_guc_tlb_invalidation: no
i915.enable_guc=2
i915.guc_log_level=-1
i915.guc_firmware_path=(null)
# journalctl -b --no-host | grep -E "guc"
May 28 00:53:23 kernel: Command line: root=LABEL=root08zt12 ipv6.disable=1 net.ifnames=0 noresume consoleblank=0 mitigations=auto i915.enable_guc=2 rd.hostonly=0 rd.auto=1 vga=791 video=1440x900@60
May 28 00:53:23 kernel: Kernel command line: root=LABEL=root08zt12 ipv6.disable=1 net.ifnames=0 noresume consoleblank=0 mitigations=auto i915.enable_guc=2 rd.hostonly=0 rd.auto=1 vga=791 video=1440x900@60
May 28 00:53:24 dracut-cmdline[212]: Using kernel command line parameters: root=LABEL=<filter> ipv6.disable=1 net.ifnames=0 noresume consoleblank=0 mitigations=auto i915.enable_guc=2 rd.hostonly=0 rd.auto=1 vga=791 video=1440x900@60
May 28 00:53:29 kernel: Setting dangerous option enable_guc - tainting kernel
May 28 00:53:29 kernel: i915 0000:00:02.0: [drm] GT0: GuC firmware i915/kbl_guc_70.1.1.bin version 70.1.1
# cat /sys/kernel/debug/dri/0/i915_guc_load_status
cat: /sys/kernel/debug/dri/0/i915_guc_load_status: No such file or directory
#
Setting dangerous option enable_guc - tainting kernel seems to conflict with non-existent /sys/kernel/debug/dri/0/i915_guc_load_status. Is i915.enable_guc=2 from i915_gpu_info the authority, the load status changing while its actively engaged or something?
0123456789112345678921234567893123
“guc|huc” funky " around "guc|huc"
in File Commander in HEX VIEW is
E2 80 9C 67 75 63 7C 68 75 63 E2 80 9D 20 66 75
6E 6B 79 20 22 20 61 72 6F 75 6E 64 20 22 67 75
63 7C 68 75 63 22 0A
32 characters plus EOF. Dunno if it helps. In FCL only “”" shows up, E2 (ASCII 226) 80 (ASCII 128), so not the best tool. I tried copying “ & ”, but it converted back to “”". Perhaps it happened because I copied from your message in Chromium to Konsole3 and from your message to the compose window, and in the process of all the copying and pasting I did back and forth composing, Chromium decided to do an AI conversion
“ “ “ left double quotation mark
” ” ” right double quotation mark
The only way I seem to be able to see “ or ” is right here on this forum page.
You might try installing kernel-longterm-6.12.30 to see if it changes behavior. Before installing 6.14.6 I was running 20250522 and 6.12.27 on one of my two 630s with no issues. Can you reproduce using a different user, preferably a new one? Are you sure your monitor cable remains firmly connected? What cable type are you using? Please show input/output from `inxi -GSaz’. If you have a KDE problem, I can’t help more than suggesting to log out of Plasma, and from a shell prompt login, delete ~/.cache/ before logging back into Plasma.
Maybe you should try a 
It’s loaded so if want to use those GPU hardware features it’s available.
