@cscs you can’t set any value. the intel_pstate will override any setting and any value once it is activated ! read again !!!
Calm down.
I mean that
As I already wrote and would appreciate you reading.
Unless you mean somehow that those are ineffective, which would be news to me.
swapping the stock 250 cd/m² panel for a 400 cd/m² one
Out of curiosity, I put the old screen back in, and the problem disappears.
Strange - one would never suspect swapping a panel could cause this
From Intel specification site
Launch Date: Q1’22
Max Resolution (eDP - Integrated Flat Panel): 4096 x 2304 @ 120Hz
Pure speculation - perhaps the new screen cannot communicate correct with the GPU?
I have laptop with i7-13700H using Irix Xe Graphics where I use
GRUB_CMDLINE_LINUX_DEFAULT=" i915.enable_guc=2 "
You can get more info
modinfo i915
Since you specifically mention 5.15 as a turning point - I am thinking mitigations
Use mitigations=off
in the kernel commandline to disable all mitigations.
It is well known that mitigations comes with a performance penalty.
Looking through the Latitude 5430 Service Manual I noticed in the Power section of your BIOS there is a setting:
- Intel Speed Shift Technology
Enable or disable the Intel speed shift technology support.
The Intel Speed Shift Technology option is apparently enabled, by default…
What happens if you disable it?
(If disabling it does nothing useful, you can enable it again).
Plus, while you’re in the BIOS anyway, disable the Fastboot option (in the Pre-boot Behaviour section) if it isn’t already.
Regards.
Thank guys. What I tried from the suggestions:
intel_pstate=disable
- without any resultsGRUB_CMDLINE_LINUX_DEFAULT=
withi915.enable_guc=2
and/ormitigations=off
- without any results- turn off
Intel Speed Shift Technology
- without any results
All of these solutions I was thy with external power and with battery only
@soundofthunder @linux-aarhus Message above
This is genuinely an XY problem - where the cause is the replacement of a - possible incompatible - screen matrix which only works with kernel below 5.15.
You should really confirm that screen is genuine replacement valid for your Dell system and not some cheap OEM display.
The reason for the incompatibility is unknown and there is zero options for troubleshooting it.
And there is nothing a forum discussion can accomplish in solving this issue.
So as I see it you have 2 options
Either Use kernel 5.15 until it goes EOL
Or Build a clean kernel using linux-next from upstream kernel.org
Build it running the 5.15 kernel so you benefit from the full potential of the CPU.
There is a PKGBUILD on AUR for doing exactly that - it is a (almost) clean linux-next
Only two patches are applied (PKGBUILD - aur.git - AUR Package Repositories)
sudo pacman -Syu base-devel --needed
git clone https://aur.archlinux.org/linux-next-git
cd linux-next-git
makepkg -si
Building this - thus documenting your issue may pave the way for the kernel developers to accept there may be an issue with your specific configuration and if it is worth their effort to fix it.
I think I will choose the third option, which is to use kernel 5.15 until I buy a new laptop (in 1-2 years), and then I will reinstall the old screen in my current laptop.
I am still undecided, but I am considering the TUXEDO InfinityBook Pro 14 - Gen9 - AMD or the MacBook Pro 14. MacBooks are excellent laptops from my own experience, but I also really like Manjaro KDE for many reasons.
The kernel 5.15 is project EOL in Dec, 2026 (The Linux Kernel Archives - Releases) and when the time comes there is new hardware versions to be considered.
The kernel 5.15 is project EOL in Dec, 2026 and when the time comes there is new hardware versions to be considered.
But the end of kernel support doesn’t mean it will stop working. However, if there’s a significant reason to update the kernel and I still haven’t bought a new laptop, I will reinstall the old screen matrix.
end of kernel support doesn’t mean it will stop working
That is correct - with that perspective the system may last time indefinite
As long as one knows this - there is no reason to change that
Actually it’s likely that one will run into issues when upgrading the system whilst still running an EOL kernel series.
I had this with my mate’s machine where I’d neglected to change to a newer Kernel; worked fine for a few months’ updates but then had to chroot in to sort it out.
I guess having things like VirtualBox installed adds to the carnage, although I don’t think he has that on his.
After all said and done, it’s also possible that you bought a lemon – you mentioned that you bought the screen on eBay; this might raise a flag or two (at least, in my mind) as to how fit for purpose the item actually was.
Pure speculation - perhaps the new screen cannot communicate correct with the GPU?
This also is speculation;
Perhaps the new screen draws too much power in comparison to the old one; in combination with other modifications you’ve made this might contribute to lowering CPU frequency and decreasing overall performance.
Nonetheless, a new laptop will no doubt solve your problems; let’s hope you find a suitable one sooner rather than later.
Regards.
I think one way to get to the bottom of this is to bisect the kernel. If linux 6.1 does not work either, then there are only around 90k commits between 5.15 and 6.1, which is ideally about 17 steps kernel compilations and checks. But of course this is quite a time investment, nonetheless, if this is an easily reproducible issue (which it appears to be based on what you said), this is probably the best way to determine the cause.
Another - much easier - option could be filing an issue in the kernel bugzilla, or writing to a kernel mailing list (although I’m not sure which one…).
I tried all stable versions of kernel 6, not the RC version. Latest 6.12
Not sure if this has been discussed already, but have you tried 6.13 since?
but have you tried 6.13 since?
Yep, unfortunately without any results