10.08.22
Gemini version available ♊︎Hostile Microsoft-backed Media Blames “Linux” for Hardware-destroying Intel Driver Bug
Reprinted with permission from Ryan
The hostile Microsoft-backed tech media has been blaming “Linux” for trouble that Intel caused which could destroy Intel hardware.
It all started last week when Intel asked Greg KH, the stable Linux kernel series branch maintainer, to bring in a patch series to better handle display power saving.
The problem was that they only marked part of the patch series for backport and what ended up in Linux 5.19.12 was a half-broken mess that was entirely the fault of Intel.
Users on bleeding edge GNU/Linux distributions, which included this author, brought in Linux 5.19.12 thinking nothing of it.
Fortunately, the problem (flickering display that may cause hardware damage) did not appear to happen to my Gen11 Tiger Lake laptop with Intel IRIS Xe graphics.
I did start experiencing some weird glitches across the system though which freaked me out wondering if they were hardware failure.
My left Ctrl key started working intermittently and my laptop kept forgetting it was plugged in and running on the battery.
Bringing in Linux 5.19.13, which reverted the changes, seems to have resolved whatever the hell was going on. So perhaps Intel caused other problems too.
Not like they’re ever going to admit that.
Intel’s products are fairly buggy.
My other system is a “Skylake” laptop, and that was the buggiest damned thing ever.
My favorite moment on that system was when I had to version lock myself to an outdated kernel in Fedora because Intel decided to turn off the power management to the GPU to try to fix a low impact security hole, and then after leaving it broken for months, gave up trying to fix the hole and turned power management back on.
Whatever Intel wants, Intel gets.
Then they get the “tech media” to “Just Blame Linux” after Intel’s “engineers” screw everything up.
In a way, they are kind of right. -drivers subsystem people, I say “people” but it’s mostly hardware companies and Microsoft, get away with murder, and it’s been like that for a while.
Linus takes the view of giving them a wide berth because if they screw up they just make a big damned problem for their users.
While that may be sort of true, it just ends up giving “Linux” more PR damage when all of the barely informed “press”, who mostly just know how to take Microsoft money to slander Free Software (where everything is also “Linux” you know, even if it isn’t), run the story.
The environment Linus created by allowing hardware companies to do whatever they want in the -drivers tree to get more drivers and marketshare is creating a quality control disaster.
OpenBSD insists on code quality audits, and that slows things down and results in missing drivers. Then people don’t use it, because “Linux has drivers”.
Then shit like this happens, with the drivers.
If you’re wondering how Intel can be so clumsy and stupid that they stopped halfway into their patch series and told Greg KH to backport it, so am I.
Maybe their buddies at Microsoft needed a fresh anti-Linux kerfuffle. That’s just conjecture, though.
Maybe they really are that stupid.
Most of their product bugs affect Windows too, not that you’d notice. Windows is already such a piece of shit you’d be hard pressed to tell where the software problems end and hardware problems begin.
Unlike Windows, GNU/Linux is generally so reliable that you notice when there’s a hardware problem and it really grabs your attention.
Basically the only reason I bought Intel again is extreme cost cutting to try to make up for their customers fleeing to AMD. But is that a longterm strategy? I don’t think that it is.
And it’s not like you can trust AMD doesn’t have bugs. They do. They’re just not as severe.
Before Apple switched to ARM processors, they started putting enablement code in MacOS for a potential switch to AMD x86, which would have been far less painful for their users, but you know Apple….. █
ᶃ Gemini Space
This post is also available in 
Content is available under CC-BY-SA