Screen brightness on XPS 13 9343

This is caused by a "feature" that cannot be turned off - called adaptive brightness or dynamic contrast.

There should be a way to disable this, bu there is not to my knowledge.

Perhaps there is a BIOS A05 in the near future, with a real keyboard fix and a setting to disable adaptive brightness.

Keyboard still causing you trouble? I installed the fixes and A04 and it stopped giving me grief, but that was with the 3.13 kernel. I just got it running 3.19; no troubles yet.

Back to the brightness issue. It seems like the Intel driver can disable it on Windows, so I started investigating the i915 code in the kernel. More reading about it than reading the code; there is a lot in there. The best lead I have found is a patch for Chromium: chromium-review.googlesource.com https://chromium-review.googlesource.com/#/c/244165/

Seems like that might address the problem on some hardware, but no clue if that includes the XPS 13.

It is a really nice machine otherwise, but I'm getting close to calling it defective by design.

Keyboard still causing you trouble? I installed the fixes and A04 and it stopped giving me grief, but that was with the 3.13 kernel. I just got it running 3.19; no troubles yet.

Yes it is. I type 100+ WPM so perhaps that is a contributing factor. The keyboard sometimes works fine, as it's doing now. Other times it throws out all kinds of spurious duplicate chars. BIOS A04 made it a little  better but it didn't fix it. It definitely needs a better de-bouncing algorithm.

The keyboard is usable when it's not acting up, just not great. It's the weak point of an otherwise fantastic laptop.

Yes it is. I type 100+ WPM so perhaps that is a contributing factor. The keyboard sometimes works fine, as it's doing now. Other times it throws out all kinds of spurious duplicate chars. BIOS A04 made it a little  better but it didn't fix it. It definitely needs a better de-bouncing algorithm.

Thanks. That's actually a really helpful piece of information, which I've passed on to our support people. We haven't been able to diagnose what sort of typing has been causing this issue to persist for a small percentage of users. A really high typing speed may be it.

Also, for what it's worth, my friend Matthew Garrett looked into the issue a while back (before A03 came out). One thing I *** him mentioning was that the issue appeared to him to be an interaction between the EC behavior and Linux/X's own debounce code. (I won't say more than that because I can't remember the exact details. Maybe he'll join here and chime in.) My understanding was that basically we see a race condition, and that at least one of the EC changes (not sure if it was the EC change in A03 or A04) changed a timer to mitigate the issue. If I *** correctly, Windows doesn't have the same repeat key issue because it doesn't do anything special for keyboard debouncing, so it doesn't encounter the same repeat key issues.

Lastly, because of the out-of-sync/bad data issue with the touchpad in PS/2 mode, since the keyboard is also on the PS/2 bus, if you have the touchpad in PS/2 mode, you're more likely to see repeat keypresses. This is the case even if you use the psmouse.reset_after=0 kernel argument. So, if you're still seeing issues with repeat keys, you probably want to verify that the touchpad is in I2C mode.

Thanks for the info. So there is hope we may get another keyboard improvement.

My touchpad is running in i2c mode - I'm using kernel 3.19 and have blacklisted psmouse. The touchpad is excellent - as good as my Macbook Air, which is saying a lot.

PS - if the keyboard problem is a race condition that would explain why it's intermittent. It does happen frequently but not constantly.