[rohtang]

Tofe wrote:

For the edge swipes of the touchpad, I’ve found a workaround in Gnome Shell: just unmap the corresponding keyboard shortcuts (in my case, mainly Win+d to show the desktop for the top edge). The touchpad edge swipes are directly sent as a keyboard event to the system, as if it was done with the keyboard, so it’s not possible to filter out just the touchpad, unfortunately. I guess it’s the same for Windows 10.

 

I guess one could write a patch for the driver to just swallow the edge events, I’ll look in to it once my vacation starts.

 

Also, nice find on the sensor, now I can finally write that auto-rotate daemon.

[rohtang]

Tofe wrote:

Unfortunately, touchscreen events are handled per-app. So gnome-shell multitouch gestures work well, but I still can’t scroll with two fingers in Terminal or Firefox. I do find it quite absurd, because such gestures work flawlessly with the touchpad, and everywhere…

FWIW, Chrome/Chromium works with multitouch. I wish there was 1)  a uniform way of handling multitouch inputdev and 2) window managers embedding proper support for multitouch.

 

 

[rohtang]

Not fixed as of yet in latest 4.6-trunk.

FWIW, I’ve been running linux on the thing for a while now, with no ill side effects. Except slightly worse battery performance compared to w10.

Intels cautionary clause regarding long term negative impact on the SoC due to not entering lower power states isn’t specific to skylake, btw.

[rohtang]

Hey,

 

It’s firmly stuck in the PC2 package state – not sure what’s preventing it from going lower in to one of the lower states. Running 4.6 and I’ve updated the BIOS (and, implicitly, microcode). It’s probably some device that’s preventing deeper sleep, I’ve yet to determine the cause though.

          Package   |             Core    |            CPU 0       CPU 2
                    |                     | C0 active   5.1%        5.3%
                    |                     | POLL        0.1%    3.8 ms  0.0%
                    |                     | C1E-SKL     6.9%    1.9 ms  1.6%
C2 (pc2)   51.6%    |                     |
C3 (pc3)    0.0%    | C3 (cc3)    2.7%    | C3-SKL      2.3%    1.5 ms  0.6%
C6 (pc6)    0.0%    | C6 (cc6)    4.6%    | C6-SKL      4.6%    1.3 ms  2.0%
C7 (pc7)    0.0%    | C7 (cc7)   61.5%    | C7s-SKL     0.0%    0.3 ms  0.0%
C8 (pc8)    0.0%    |                     | C8-SKL     21.2%    2.9 ms 31.1%
C9 (pc9)    0.0%    |                     | C9-SKL      0.0%    0.0 ms  0.0%
C10 (pc10)  0.0%    |                     | C10-SKL    54.0%   10.3 ms 52.2%

                    |             Core    |            CPU 1       CPU 3
                    |                     | C0 active   3.4%        3.7%
                    |                     | POLL        0.0%    0.0 ms  0.2%
                    |                     | C1E-SKL     3.3%    0.8 ms  6.9%
                    |                     |
                    | C3 (cc3)    0.7%    | C3-SKL      0.2%    0.3 ms  0.7%
                    | C6 (cc6)    7.2%    | C6-SKL      4.4%    0.9 ms  5.7%

[rohtang]

Provided that your powerbank can provide 12V at 2.5A, then yes, an USB -> DC adapter will work just fine for charging.

 

…However, why not just pick up an USB-C cable for a quid or two and use that to plug in to your powerbank?

It does seem, however, that power banks that can support [email protected] are exceedingly rare. Mine can do [email protected], but that doesn’t really cut it.

[rohtang]

Cheers, glad to be of assistance.

Regarding battery life – I find it acceptable and it’s in no way catastrophic. I still get a good 6-7 hours at ~50% back light – running build chains, watching videos along with web browsing. Hopefully an upcoming kernel will be able to make the CPU enter the higher sleep states…

[Author]

Posts