You are not logged in.

#226 2018-05-24 06:21:51

seiichiro0185
Member
From: Leipzig/Germany
Registered: 2009-04-09
Posts: 226
Website

Re: Dell XPS 13 Late 2016 9360 Kaby Lake 7th Gen

I actually don't see any of these Issues on my 9360. Performance is the same with both internal and external Displays (external display is a Displayport over USB-C capable Lenovo P24H). Only difference is that the fan works a lot more when plugged in, pobably due to having to drive a higher resolution display and additional heat created by the batter charging. I don't use bluetooth very often, but until now it did work fine when I needed it (mostly using a bluetooth mouse). I'm on a up to date Arch with Kernel 4.16.10 as well. Bios / TB Firmware are up to date as well.

Last edited by seiichiro0185 (2018-05-24 06:23:21)


My System: Dell XPS 13 | i7-7560U | 16GB RAM | 512GB SSD | FHD Screen | Arch Linux
My Workstation/Server: Supermicro X11SSZ-F | Xeon E3-1245 v6 | 64GB RAM | 1TB SSD Raid 1 + 6TB HDD ZFS Raid Z1 | Proxmox VE
My Stuff at Github: github
My Homepage: Seiichiros HP

Offline

#227 2018-05-24 18:03:08

Asriel
Member
Registered: 2016-11-28
Posts: 37

Re: Dell XPS 13 Late 2016 9360 Kaby Lake 7th Gen

What I have noticed is the external USB-C-HDMI display wakes up thunderbolt controller, and from that point CPU is not going into deep power saving states anymore, and thunderbolt controller also eat a lot. Whithout external monitor powerstat shows 1.6-1.8 watts idling, but with the USB-C connecting it never goes below 6 watts. So it might be simple thermal throttling. Anyway it's just a guess - I have no performance issues but I've repasted my 9360 and add thermal rubber between CPU cooler and lower metal plate - so my CPU does mot throttle even at heavy compilation processs.

Offline

#228 2018-06-02 15:55:21

NomadicAgenda
Member
Registered: 2018-06-02
Posts: 1

Re: Dell XPS 13 Late 2016 9360 Kaby Lake 7th Gen

I just updated my  9360 and now I'm experiencing WiFi issues. The connection will drop at random, and sometimes restarting the WiFi adapter will sometimes induce a kernel panic. I just started troubleshooting but thought I'd check here to see if anybody else is seeing this new issue.

Offline

#229 2018-06-20 04:29:20

noicie
Member
Registered: 2018-06-20
Posts: 1

Re: Dell XPS 13 Late 2016 9360 Kaby Lake 7th Gen

NomadicAgenda wrote:

I just updated my  9360 and now I'm experiencing WiFi issues. The connection will drop at random, and sometimes restarting the WiFi adapter will sometimes induce a kernel panic. I just started troubleshooting but thought I'd check here to see if anybody else is seeing this new issue.

Can confirm. This seems to happen under heavy load

Offline

#230 2018-06-20 18:01:06

clownfish
Member
Registered: 2006-09-04
Posts: 72

Re: Dell XPS 13 Late 2016 9360 Kaby Lake 7th Gen

clownfish wrote:

I'm on a 9360 since about one and a half year. Used Xubuntu 17.04, 17.10 and now after the update to 18.04 I swapped that crap for Arch, hoping that here i have no bluetooth and performance issues (and other stupidities introduced in 18.04). The chance for the integrated bluetooth to work changed from 90% to 10% (now using an external adapter which until now always worked). I'm OK with that. But what got also worse, was the graphics over USB-C/Thunderbolt. When I use the device without a display attached, it's smooth. Using a USB-C to HDMI cable or a USB-C dock with HDMI port, it turns into a slow beast. Browsing in chrome or firefox is slow (especially when i open webapps which utilize a lot of javascript, which now is nearly every website). Terminal lags (not talking about seconds when typing, but it's just not smooth). The Jetbrains IDEs are slow. Switching folders in a filemanager like thunar is slow, especially when browsing network shares. All this goes away if i unplug the display and work on the device itself (except the bluetooth issue). I tried all the things mentioned in the wiki article, updated BIOS and Thunderbolt firmware, but it's always the same.

My kernel is 4.16.10-1. And I'm using i3 without a compositor or anything else which demands graphics power. I had this on Xubuntu too, but just on about one of ten starts, sometimes it needed one or two reboots, sometimes just plugging the USB-C out and in again was enough to fix it. On arch, I didn't find a way to solve this (did a lot of useless reboots).

What helps a bit is, changing the cpu frequency scaling governor from powersave to performance. When I do this, frequencies go up to 2-3GHz (only this chrome window open and the terminal which shows the frequencies), when using powersave governor, they are most of the time in three-digit-ranges. But also using the performance governor, it's not like without display. Lags are not that bad, but they are there. So the performance governor is no solution.

Does anybody have similar issues?

Solved this. Pretty stupid. To have an absolute minimum of cables, I have the display attached by USB-C, and a USB-C charger with power delivery attached to the USB-C/HDMI cable. So I have just one cable attached. I used a 30W charger, the BIOS tells you on boot, that this charger is underpowered and that it's not sure if it charges correctly. As I had no issues with charging, I disabled this warning. I never ran out of power, and I charges a bit slower than the original, but in terms of charging it's just fine. I started watching CPU load, usage and clock. When I stress the CPU (e.g. a website with WebGL or some animations) the device became slow and laggy. The CPU usage then is up to nearly or at 100%, but the clock is down in three digit numbers! It even shows 1xx or 2xx values, which shouldn't be possible, as lscpu states "CPU min MHz: 400.0000". If I disconnect the USB-C/HDMI, and therefore Power delivery, the clock goes up as one would expect it, the usage and load numbers calm down, and also the clock reacts as soon as the CPU stress is done. I always suspected some kind of USB-C quirks, that is mentioned for the XPS13 with different issues. But I never thought about the charger, because it's hidden under the table, and attached to the HDMI part of the cable, which is behind the display. Out of sight I just didn't recognize that I also disconnect power when I disconnect the display over USB-C. Today I tried another charger of a colleague, 60W - all issues gone instantly. Not even a reboot or reconnect of the USB-C was needed (e.g. for pulling the scaling governor out of his dreams), it just works with the 60W charger. Now I use the original charger, until I get a new USB-C Power delivery charger with at least the 45W, but I think I will just buy the 60W charger that my colleague uses, as this one is approved to work (if someone's interested: Inateck 60W USB-C, this one: https://www.inateck.com/uk/chargers-cab … black.html).

I have/had some rare small Bluetooth and display issues too, nothing that is as annoying as the lags, but the issues are there. I'm curios, If they are gone now too. In "laptop mode" I don't have a display nor I use bluetooth, so I don't know if they only occoured because of power/USB-C connected. At least today I didn't have them, but they only occour 2 or 3 times per week - I'll report back, if they are there still or are gone too with the original charger.

EDIT: Corrected the numbers, mine that caused the issues was a 30W charger, the one from my colleague a 60W, the original is 45W.

Last edited by clownfish (2018-06-20 18:51:28)

Offline

#231 2018-06-23 23:31:34

jshap70
Member
Registered: 2017-12-14
Posts: 10
Website

Re: Dell XPS 13 Late 2016 9360 Kaby Lake 7th Gen

Do people generally use  the `xf86-video-intel` i915 driver or do they use the modesetting one? I was using modesetting but switched to `xf86-video-intel` so I could use `Option "AccelMethod" "sna"` to get rid of that silly xorg-1.20 artifacting bug, but from what I can tell the screen-tearing is noticeably worse, even with `Option "TearFree" "true"` set.   
Just curious if this is anyone else's experience too.

Offline

#232 2018-06-25 04:50:50

seiichiro0185
Member
From: Leipzig/Germany
Registered: 2009-04-09
Posts: 226
Website

Re: Dell XPS 13 Late 2016 9360 Kaby Lake 7th Gen

I use the mopdesetting driver on all my Systems with Intel Graphics, including the XPS. So far did not see any artifacting or other issues (other than the well known screen flickering when using the "wrong" i915 module options).


My System: Dell XPS 13 | i7-7560U | 16GB RAM | 512GB SSD | FHD Screen | Arch Linux
My Workstation/Server: Supermicro X11SSZ-F | Xeon E3-1245 v6 | 64GB RAM | 1TB SSD Raid 1 + 6TB HDD ZFS Raid Z1 | Proxmox VE
My Stuff at Github: github
My Homepage: Seiichiros HP

Offline

#233 2018-06-26 00:34:06

CarbonChauvinist
Member
Registered: 2012-06-16
Posts: 70

Re: Dell XPS 13 Late 2016 9360 Kaby Lake 7th Gen

seiichiro0185 wrote:

... (other than the well known screen flickering when using the "wrong" i915 module options).

Can you expound on this please?

Offline

#234 2018-08-05 20:05:29

Espionage724
Member
From: Pennsylvania, United States
Registered: 2013-09-12
Posts: 19
Website

Re: Dell XPS 13 Late 2016 9360 Kaby Lake 7th Gen

jshap70 wrote:

Do people generally use  the `xf86-video-intel` i915 driver or do they use the modesetting one?

xf86-video-intel (and equivalent for other distros); I've had no problems with it across multiple computers (including my XPS 13 9360) and as I understand it's better optimized. With that said though, I've had no problems with modesetting either.

I just got one of these laptops the other day:

1. Is suspend broken for anyone else? If I suspend the laptop with out-the-box settings, it won't wake up. If I use mem_sleep_default=s2idle, suspend and wake works fine
2. Has anyone tried different i915.enable_psr options? With 1, I believe the sys report for it said something non-ideal, and there was a bit of flickering; with 2 though, everything seems fine

Last edited by Espionage724 (2018-08-05 20:09:22)

Offline

Board footer

Powered by FluxBB