You are not logged in.
Alll known major issues are fixed. We are on par for support with regular kernel as of 4.4-rc3. So we aren't bound to any specific patches at least for FHD.
Offline
Hi All,
I compiled 4.4-rc1 (first time I've compiled a kernal) and managed to boot up and load the intel video card. However I'm having trouble with wifi. My machine has the broadcom chip installed so I installed broadcom-wl-dkms from Aur then as instructed by the install output ran:
$ rmmod b43 b43legacy ssb bcm43xx brcm80211 brcmfmac bccmsmac bcma wl
$ modprobe wl
However I now get a black screen after grub has loaded.
I'm also unable to use a wifi dongle which I had been using successfully while the broadcom wifi card has been offline.
Any ideas?
Offline
Any ideas?
install from aur: linux-firmware-git https://aur.archlinux.org/packages/linux-firmware-git/ and a kernel linux-mainline https://aur.archlinux.org/packages/linux-mainline/ (4.4-rc3) or linux-bcm4350 https://aur.archlinux.org/packages/linux-bcm4350/ (4.3)
Offline
Cheers for that.
So you don't recommend linux-git from Aur?
Whats the difference to mainline?
Sorry if they are silly questions, kernals are all a bit new to me!
Tom
Last edited by anthillsocial (2015-12-02 09:58:59)
Offline
Cheers for that.
So you don't recommend linux-git from Aur?
Whats the difference to mainline?
Sorry if they are silly questions, kernals are all a bit new to me!
Tom
no I do not recommend to use git version, I do not recommend to use the mainline without a real requirement also, it is better to use stable kernels. but 4.4 has full support for Dell XPS 9350 , thats why I suggested you to select between 2 kernels, one is stable with patches for the broadcom module ported from 4.4 kernel (they just add the device id to be recognized by kernel) or mainline, it is up to you to make that decision. also you need a blob for the broadcom module, it is available now in the linux-firmware-git AUR package, and will be available in linux-firmware on the next release. broadcom-wl-dkms package doesn't support our wireless card and should be remove. also you should visit https://wiki.archlinux.org/index.php/Dell_XPS_13_(2016) link there you can find more information about your laptop.
Last edited by ksim (2015-12-02 13:51:31)
Offline
For any night time users. Download xcalib and set
xcalib -co 40 -a
Our screen is just way too bright sometimes.
Offline
vilavil wrote:I compiled NVME into kernel so that the SSD gets mounted early
you should not do that in theory because loaded modules should be detected on initramfs build and nvme should be added automatically to initramfs image
in the booting process and it can find potential firmware files (like for i915) on disk and you don't have to bother including them in initramfs
Again, mkinicpio does that automatically if you have i915 in MODULES section, no one should be bothered
ᐅ lsinitcpio /boot/initramfs-linux-bcm4350.img | grep i915 ./usr/lib/firmware/i915 ./usr/lib/firmware/i915/skl_dmc_ver1.bin ./usr/lib/modules/4.3.0-2-bcm4350/kernel/i915.ko ./etc/modprobe.d/i915.conf ᐅ lsinitcpio /boot/initramfs-linux-mainline.img | grep i915 ./usr/lib/firmware/i915 ./usr/lib/firmware/i915/skl_guc_ver4.bin ./usr/lib/firmware/i915/skl_dmc_ver1.bin ./usr/lib/firmware/i915/bxt_dmc_ver1.bin ./usr/lib/modules/4.4.0-rc3-mainline/kernel/i915.ko ./etc/modprobe.d/i915.conf
(the firmware goes to /lib/firmware/i915/ like the included bash script does for you)
this firmwares already in linux-firmware package, the same latest version.
I am using linux-mainline without any patches and without any manual modifications. only i915 included in MODULES in mkinitcpio.
Alright, thanks for the tips. Is there a practical downside why I should not compile those into kernel or is it more like adding them as modules is the proper way to do?
For those concerned about battery life, there are some very good (upcoming) news. A new BIOS update will be released mid-December boosting battery life, fixing the NVME issues ( http://www.laptopmag.com/articles/dell- … y-life-fix ). The article claims that the SSD was preventing the CPU from entering higher power states but analyzing power states with powertop shows that 98% of the time, the CPU was in the highest power state. Furthermore it seems that the overall CPU + GPU package was only taking around 2W out of a total 6W so the issue is probably rather the SSD. I'm guessing they activated ASPM or something similar to allow the SSD to go into idle mode.
Offline
Alright, thanks for the tips. Is there a practical downside why I should not compile those into kernel or is it more like adding them as modules is the proper way to do?
Only downside is that you have to compile your own version of kernel every time.
Offline
@vilavil, I will now try to forget about this news as I wait for it to appear for download.
Offline
@frank Thanks for the brightness command. You are saving my eyes at night.
Offline
@sheep, np, I prefer 70 than 40 but 40 is for when wife is already sleeping She can't stand any light...
70 is good for allround setting. I put that in my .xinitrc (you can put it in whatever autostart you have setup).
xcalib -c
clears it, so if you want to reset it back to default.
Last edited by frank604 (2015-12-03 20:13:14)
Offline
What is the wireless performance like with patches/4.4? Is the broadcom wireless in all variants of this model now and has anyone replaced it with an Intel?
Been waiting to get one of these and I'm excited that everything seems to be fixed/workable.
Offline
What is the wireless performance like with patches/4.4?
Great.
Is the broadcom wireless in all variants of this model now
Yes.
and has anyone replaced it with an Intel?
I would think some have. For sure I've read people on other distros doing so instead of patching or waiting for 4.4
Been waiting to get one of these and I'm excited that everything seems to be fixed/workable.
I think we are at a good place. Much less headache now compared to when it first arrived. This device features many new tech (nvme and skylake) which took some time for support to fully kick in. We are still waiting for an nvme fix in 1.17 as per the news link in a few posts above. I think there is greater potential for battery savings coming soon (don't let me down Dell!!!). Everything else working like a charm. Heck, I'm happy just resting my palms on this laptop. Very svelte.
Last edited by frank604 (2015-12-04 09:09:43)
Offline
I think there is greater potential for battery savings coming soon (don't let me down Dell!!!). Everything else working like a charm. Heck, I'm happy just resting my palms on this laptop. Very svelte.
Yep, it's definitely the most comfortable laptop I have owned so far.
Here's a report ( thanks to @bogenpirat for the heads up : https://www.reddit.com/r/Dell/comments/ … me/cxlnceu ) of a 128GB non-NVME model owner who gets 3W idling with wifi and 30% brightness, which means that if the folks at Dell do their jobs properly we can expect a pretty sweet pre-christmas present
Offline
Hi,
First thank you all for your work done in order to get everything running. I wouldn't have been able to set it all up that easy.
Now everything is working quite good since around a week. The only problem I'm facing is that if the notebook was hibernating it freezes when I wan't to log in. I can type the password and after clicking enter everything freezes. Sometimes the login succeeds after around 30 seconds but most of the times it won't succeed at all (after two minutes at least...). I'm using Gnome.
Anyone else having similar problems?
Offline
Hi ,
Can you share you powertop output with different screen brightness ? Powertop seems stuck at 4W for me, even if I reduce or increase the brightness.
Offline
@soxyl, this isn't a dell xps 13 9350 specific issue. Please feel free to open a thread in "Newbie Corner" providing logs (dmesg? xorg?) and details of your setup (gdm? sddm? ??).
@sheeponmeth, you are pressing 'r' to refresh yes? mine changes fro 4-6 on lowest brightness and 7-8 on highest brightness.
Last edited by frank604 (2015-12-05 17:06:55)
Offline
After running powertop --calibrate brcmfmac (wifi) shows up at around 2.5W-3W. Is this normal? It sounds like way too much when there is 1pkts/s
With backlight at 50% the total power usage at idle is around 8W.
Running wpa_supplicant directly (no difference with NetworkManager)
4.4-rc3, firmware-git.
Offline
@frank, I though that tab was also refreshing , but I am finding the same number as you.
@condensator is the power management on for the wifi card ? Look at the output of iwconfig.
I have change the wifi card myself for a Intel 8260, I am not able to reply.
Offline
@frank, I though that tab was also refreshing , but I am finding the same number as you.
@condensator is the power management on for the wifi card ? Look at the output of iwconfig.
I have change the wifi card myself for a Intel 8260, I am not able to reply.
It is on. Weird. Perhaps powertop is reporting it incorrectly.
Offline
@soxyl, this isn't a dell xps 13 9350 specific issue. Please feel free to open a thread in "Newbie Corner" providing logs (dmesg? xorg?) and details of your setup (gdm? sddm? ??).
Thanks, Frank. I'll search my way through the logs and open an thread if necessary.
Offline
SheepOnMeth wrote:@frank, I though that tab was also refreshing , but I am finding the same number as you.
@condensator is the power management on for the wifi card ? Look at the output of iwconfig.
I have change the wifi card myself for a Intel 8260, I am not able to reply.
It is on. Weird. Perhaps powertop is reporting it incorrectly.
Powetop seems to have trouble estimating the consumption of individual components. When my backlight is at 50% it estimates that backlight display uses 6.5W, when total consumption is 6.5W.
For those really impatient about the new BIOS, here's a Python 3 script notifying you when it's online
import urllib.request
import re
import json
import time
while True:
with urllib.request.urlopen("http://www.dell.com/support/home/us/en/19/product-support/product/xps-13-9350-laptop/drivers") as web:
webstring = web.read().decode('utf-8')
p = re.compile('var masterDriversData = (.*?\});', re.MULTILINE)
m = p.search(webstring)
extract = m.group(1)
driverdict = json.loads(extract)['GroupItem']
biosdict = filter(lambda x : x['GroupItemName'] == 'BIOS', driverdict)
for bios in biosdict:
if bios['Drivers'][0]['VendorVer'] != '1.0.5':
print('Victory!!!') #Replace this with a line sending you a text / ordering you a pizza
time.sleep(121)
Offline
did someone experiencing same issue
╰─ cat /sys/class/power_supply/BAT0/status
Unknown
xps13@late2015x16gbx512g
Offline
did someone experiencing same issue
╰─ cat /sys/class/power_supply/BAT0/status Unknown
What kernel are you running? Sometimes it takes a while when plugging the AC cord in/out to update the battery status. I do not have this issue and I've tested on kernel 4.3, 4.4-rc3 and 4.4-rc4. Try going through the steps in Power Management
Offline
Just thought to share my experience.
I am now running on 4.4-rc4 as on 4.3 and 4.4-rc3 I had awful boot experience. I was greeted with a blank screen 3 out of 4 times. On 4.4-rc4 I have no such issue and everything seems buttery smooth... well, almost. Once in a while, I have an issue shutting down. My guess is that a USB device prevents shutdown, but I have not been able to determine this exactly. I also want to point out that encryption with dm-crypt runs fine and very fast on top of that. It is a requirement with me, so this is excellent. Idle power is currently sitting in the range of 6.5-7.0 W with pcie_aspm=force i195.enable_fbc=1 i915.enable_rc6=7.
The only current issues that bother me are concerned with the audio and USB-C port. First, the audio does not seems to configure the headphone channels properly and I end up with background volume being overly loud and foreground very faint. So if I watch video, I hear the music and certain effects, but speech is very faint and seems like it comes from afar. There is no static, so I guess at least the driver does at least something well.
For the USB-C port, things are a bit difficult for me to diagnose because I don't have very many devices to plug into it. But for now, my Nexus 6P can only charge, no communication is established whatsoever. The Nexus connects fine to the other USB ports. Not sure how to test it further. I am planning on getting the USB-C hub that is coming out in Jan, so I will be able to update you guys if it works. A Dell representative assured me, the hub is Linux compatible, so we'll see.
Any comments are welcome.
Offline