You are not logged in.
How did you guys get X working? When I run startx, it errors out with cannot stat 'nvidia'.
Edit: I got X started up, easy fix somehow.
Working:
* X
* Speaker Sound
* Trackpad
Not working that I want to work:
* Screen Backlight Control and Hotkeys
* Keyboard Backlight Control and Hotkeys
* Fan Speed Control (anyone else running hot?)
* External USB sound device for big speakers
Last edited by Laggg (2013-10-04 18:41:46)
Offline
frigaut wrote:Anyone got news about the microphone??? Still not working at my end. I guess I should take the time to contact the upstream devs. Anyone's got pointer on who/what mailing list to contact? Alsa? Pulseaudio? Kernel?
The latest linux-mainline fixes the microphone. I added the following file to /etc/modprobe.d/alsa-base.conf:
options snd-hda-intel model=mbp101
Working:
suspend
sound
mic
external display
display brightness fn key
sound fn key
wifi
thunderbolt/nic adapter
Borked:
camera - It worked on linux-3.6.6, but under linux-mainline (3.7-rc6) the uvcvideo module is missing
keyboard backlight fn keys - although echo'ing a value to /sys/devices/platform/applesmc.768/leds/smc::kbd_backlight/brightness works
thunderbolt - not working for me
BTW, I have an MBP 13" (macbookpro10,2)
How did you get the backlight brightness working? and with the hotkeys?
Offline
How did you get the backlight brightness working? and with the hotkeys?
By writing in /sys/class/backlight/gmux_backlight/brightness
I think this works only with the intel though.
Put that in a script and use you window manager to assign a key to the script execution
Archer since 03/2009 - AUR packages
Offline
Laggg wrote:How did you get the backlight brightness working? and with the hotkeys?
By writing in /sys/class/backlight/gmux_backlight/brightness
I think this works only with the intel though.
Put that in a script and use you window manager to assign a key to the script execution
that does nothing for me
I'm pretty sure im on the i915 module, since i took nouveau out of /etc/mkinit.conf
Offline
Running on the i915 here. I have Gnome installed right now, and the brightness (both backlight and keyboard) worked with no trouble using the hotkeys.
When I was running with Nouveau, I could not control brightness of the backlight at all, with any method. The symptom was as you describe, you do the step (either the echo or use the brightness key in your DE of choice) and nothing would change. In Gnome the UI would indicate lowering brightness but there was no visible change.
Offline
Running on the i915 here. I have Gnome installed right now, and the brightness (both backlight and keyboard) worked with no trouble using the hotkeys.
When I was running with Nouveau, I could not control brightness of the backlight at all, with any method. The symptom was as you describe, you do the step (either the echo or use the brightness key in your DE of choice) and nothing would change. In Gnome the UI would indicate lowering brightness but there was no visible change.
I'm pretty sure I got the backlight setting turned down using XRANDR --brightness or something like that. The keyboard backlight isn't working though. Do I need to install gnome for that?
My ideal setup is X + SubtleWM. And being able to control screen brightness and keyboard backlight. I'm using the i915, I took nouveau out of the MODULES so it only has i915 in there.
Can I get a look at anyone else's MKINITCPIO.CONF? I'm on the Macbook Pro Retina 15" and I feel like I'm missing a few things maybe.
I'm in the process of setting up MPD with NCMPCPP, using the music library from the mac. I can mount the mac partition and see all the files/folders, but ncmpcpp fails to add them to it's library. Something with permissions, maybe?
I can post any configs necessary to get my machine diagnosed. I have a feeling arch will run very well once all the hardware is accounted for.
EDIT: Also, arch is not playing nice with my Thunderbolt display either. It works and mirrors the display but if I kill X and restart it, it doesn't come back on.
EDIT: Anyone seen this? https://github.com/jantman/puppet-archl … bookretina
Last edited by Laggg (2013-10-07 19:30:39)
Offline
At least for me, nvidia still only works with bios boot. Sleep works, but not changing brightness.
Take a look at this thread: https://devtalk.nvidia.com/default/topi … th-325-15/
Someone from nvidia said it should be fixed in 331
No it's not. I'm running 331:
[ 8.568] (II) NVIDIA dlloader X Driver 331.13 Sun Sep 29 20:48:50 PDT 2013
and I still get:
[ 13.151] (EE) NVIDIA(GPU-0): Failed to tear down EVO channel
Archer since 03/2009 - AUR packages
Offline
Hi guys I have almost everything working. I was wondering if anyone hit a similar problem as me:
Whenever I boot up after LILO (BIOS emulation in use), the USB ports go dead. My mouse light shuts off and my ethernet adapter never appears in lsusb output. I see this a lot of times when I am booting: "System wakeup disabled by ACPI" (and have no idea what it means). I have tried turning off auto-suspend. I wonder if it has anything to do with rEFInd as it seems to turn on USB first, then shuts it off, then LILO turns it on again, then Linux shuts it off again (right after the repeated messages).
I am using rEFInd + LILO and I am on Gentoo.
If you have USB ports fully working, could one of you please post your kernel configuration so I can compare? zcat /proc/config.gz
I am also on kernel 3.10.7 but this should not matter that much.
Just to note, the internal devices which are USB (webcam, bluetooth, etc) are all listed in lsusb. Just not anything external. The external devices work fine in OS X.
Offline
Update:
Benoit Gschwind has found the bug in the i915 graphic driver that corrupted the video signal.
See http://lists.freedesktop.org/archives/i … 19522.html for a patch.Unfortunately, there is still a problem about the backlight not being turned on after suspend, so for me the i915 is still not usable for my daily uses. This might be related to gmux, which is going to be addressed soon I understand.
Related or not to gmux, the full advantage of using the intel graphic driver (which for me is to save power by being able to turn the nvidia graphic card off when I need it), will only be fulfilled when gmux is fixed.
hi i seem to have screen corruption issues on my oct 2013 rmbp (when i don't use nomodeset), vertical slices of the screen bunched up on the right, and funky colours
is that 'patch' something i still have to do myself manually, or is it integrated into the intel driver and i am just having a different issue (and thus should open up a new thread).
I have the iris pro only model.
thanks
Last edited by rmbpefi (2013-10-25 07:58:39)
Offline
This has been in the kernel tree for a while now (close to a year). So your problem ought to be different.
Archer since 03/2009 - AUR packages
Offline
I was wondering if anyone has had any luck with nvidia in EFI boot. I've got the same MBP 10,1, and have no luck. At work I run a 30" external monitor over the MiniDP adapter, so the onboard intel isn't really an option for me. I also do *not* want to fall back to BIOS boot, I'm quite happy with rEFInd.
Almost two months ago I did a pacman upgrade which bumped me from kernel 3.9.8-1 and nvidia 319.32-2 to kernel 3.10.10-1 and nvidia 325.15-5. X wouldn't start, so I rolled back. This morning I tried a pacman system update again, which brought in kernel 3.11.6-1-ARCH and nividia 325.15-10, and once again, X won't start with the same exact errors:
[ 7.419] (II) NVIDIA(GPU-0): Found DRM driver nvidia-drm (20130102)
[ 7.421] (II) NVIDIA(0): NVIDIA GPU GeForce GT 650M (GK107) at PCI:1:0:0 (GPU-0)
[ 7.421] (--) NVIDIA(0): Memory: 1048576 kBytes
[ 7.421] (--) NVIDIA(0): VideoBIOS: 80.07.26.04.01
[ 7.421] (II) NVIDIA(0): Detected PCI Express Link width: 16X
[ 7.431] (II) NVIDIA(GPU-0): Display (Apple Color LCD (DFP-3)) does not support NVIDIA 3D
[ 7.431] (II) NVIDIA(GPU-0): Vision stereo.
[ 7.435] (--) NVIDIA(0): Valid display device(s) on GeForce GT 650M at PCI:1:0:0
[ 7.435] (--) NVIDIA(0): DFP-0
[ 7.435] (--) NVIDIA(0): DFP-1
[ 7.435] (--) NVIDIA(0): DFP-2
[ 7.435] (--) NVIDIA(0): Apple Color LCD (DFP-3) (boot, connected)
[ 7.435] (--) NVIDIA(0): DFP-4
[ 7.435] (--) NVIDIA(0): DFP-5
[ 7.435] (--) NVIDIA(0): DFP-0: 165.0 MHz maximum pixel clock
[ 7.435] (--) NVIDIA(0): DFP-0: Internal Single Link TMDS
[ 7.435] (--) NVIDIA(0): DFP-1: 165.0 MHz maximum pixel clock
[ 7.435] (--) NVIDIA(0): DFP-1: Internal Single Link TMDS
[ 7.435] (--) NVIDIA(0): DFP-2: 165.0 MHz maximum pixel clock
[ 7.435] (--) NVIDIA(0): DFP-2: Internal Single Link TMDS
[ 7.435] (--) NVIDIA(0): Apple Color LCD (DFP-3): 480.0 MHz maximum pixel clock
[ 7.435] (--) NVIDIA(0): Apple Color LCD (DFP-3): Internal DisplayPort
[ 7.435] (--) NVIDIA(0): DFP-4: 480.0 MHz maximum pixel clock
[ 7.435] (--) NVIDIA(0): DFP-4: Internal DisplayPort
[ 7.435] (--) NVIDIA(0): DFP-5: 480.0 MHz maximum pixel clock
[ 7.435] (--) NVIDIA(0): DFP-5: Internal DisplayPort
[ 7.435] (**) NVIDIA(0): Using HorizSync/VertRefresh ranges from the EDID for display
[ 7.435] (**) NVIDIA(0): device Apple Color LCD (DFP-3) (Using EDID frequencies has
[ 7.435] (**) NVIDIA(0): been enabled on all display devices.)
[ 8.467] (==) NVIDIA(0):
[ 8.467] (==) NVIDIA(0): No modes were requested; the default mode "nvidia-auto-select"
[ 8.467] (==) NVIDIA(0): will be used as the requested mode.
[ 8.467] (==) NVIDIA(0):
[ 8.467] (II) NVIDIA(0): Validated MetaModes:
[ 8.467] (II) NVIDIA(0): "DFP-3:nvidia-auto-select{}"
[ 8.467] (II) NVIDIA(0): Virtual screen size determined to be 2880 x 1800
[ 12.475] (EE) NVIDIA(GPU-0): Failed to tear down EVO channel
[ 12.477] (--) NVIDIA(0): DPI set to (221, 217); computed from "UseEdidDpi" X config
[ 12.477] (--) NVIDIA(0): option
[ 12.477] (--) Depth 24 pixmap format is 32 bpp
[ 12.477] (II) NVIDIA: Using 3072.00 MB of virtual memory for indirect memory
[ 12.477] (II) NVIDIA: access.
[ 12.480] (II) NVIDIA(0): ACPI: failed to connect to the ACPI event daemon; the daemon
[ 12.480] (II) NVIDIA(0): may not be running or the "AcpidSocketPath" X
[ 12.480] (II) NVIDIA(0): configuration option may not be set correctly. When the
[ 12.480] (II) NVIDIA(0): ACPI event daemon is available, the NVIDIA X driver will
[ 12.480] (II) NVIDIA(0): try to use it to receive ACPI event notifications. For
[ 12.480] (II) NVIDIA(0): details, please see the "ConnectToAcpid" and
[ 12.480] (II) NVIDIA(0): "AcpidSocketPath" X configuration options in Appendix B: X
[ 12.480] (II) NVIDIA(0): Config Options in the README.
[ 16.482] (EE) NVIDIA(GPU-0): EVO Push buffer channel allocation failed
[ 16.484] (EE) NVIDIA(GPU-0): Failed to allocate EVO core DMA push buffer
[ 16.486] (EE)
Fatal server error:
[ 16.486] (EE) AddScreen/ScreenInit failed for driver 0
[ 16.486] (EE)
[ 16.486] (EE)
Please consult the The X.Org Foundation support
at http://wiki.x.org
for help.
[ 16.486] (EE) Please also check the log file at "/var/log/Xorg.0.log" for additional information.
[ 16.486] (EE)
[ 16.494] (EE) Server terminated with error (1). Closing log file.
I've tried booting with nomodeset and noapic, and that makes no difference that I can tell. My xorg.conf device section is:
Section "Device"
Identifier "Device0"
Driver "nvidia"
VendorName "NVIDIA Corporation"
Option "UseDPLib" "False"
Option "ProbeAllGpus" "False"
EndSection
I just added ProbeAllGpus False this morning, and that doesn't appear to have made a difference either.
I can confirm that the thread mentioned previously, https://devtalk.nvidia.com/default/topi … th-325-15/ seems to be the same problem I'm seeing. Per the latest post in that forum, I jumped through the hooks (pkgbuild and depdendency issues) to install nvidia-beta 331.17-1 and its dependencies from AUR, but that doesn't seem to solve anything; I'll admit that I didn't pull the xorg logs, but even with a few xorg.conf tweaks, I saw the same black-screen behavior.
I ended up rolling back to 3.9.8-1 and nvidia 319.32-1
Offline
I use EFI boot as well and I can't get nvidia working either. nouveau works, but I think the intel driver is better than it, so I don't really see the point of using it. I just use intel only.
Also, am I the only person whose wifi recently started magically working perfectly? It's been good for the past couple of days with no problems (broadcom-wl-dkms) and this is really exciting for me. (It has possibly been good for longer than that, I just wouldn't have noticed since I got used to using wired) Up until now there have been good times and bad times, but usually it would have slower speeds and drop often. There was no recent upgrade to broadcom-wl-dkms, so I'm assuming it was a recent kernel upgrade or something?
I have finally unplugged my ethernet adapter and am running purely on wifi these days!
I hope it doesn't get worse again. >_<
EDIT: Well, I wouldn't say it has been working perfectly since, there were a few good days and bad days. But overall I generally still mainly use wifi and don't need to plug in the ethernet adapter unless the wifi isn't working and I'm doing something important that needs internet at the moment.
Last edited by sokuban (2013-11-26 20:55:58)
Offline
Since a recent full system update, I've been having a bit of trouble with the keyboard driver (allegedly). Powertop reports the following
Power est. Usage Events/s Category Description
9.14 W 100.0% Device USB device: Apple Internal Keyboard / Trackpad (Apple Inc.)
1.25 W 24.4% Device Display backlight
The computer is also very hot to touch. Has anyone encountered this before and, if so, how was it resolved?
Offline
Public Service Announcement:
If anyone who uses xf86-input-mtrack's trackpad suddenly stopped working after a recent update, a recompile of mtrack fixes this.
Now the trackpad seems to work even better tho.
I haven't noticed anything funny about the keyboard or temperature for me though...
Last edited by sokuban (2014-01-16 03:49:19)
Offline
Public Service Announcement:
If anyone who uses xf86-input-mtrack's trackpad suddenly stopped working after a recent update, a recompile of mtrack fixes this.
Now the trackpad seems to work even better tho.
I haven't noticed anything funny about the keyboard or temperature for me though...
Aargh! I should have come here first, have been trying to track down this issue for the last 3 days!! Thanks for the tip
Offline
I ended up rolling back to 3.9.8-1 and nvidia 319.32-1
What is the easiest way to roll back the kernel version to 3.9.8-1?
Offline
jantman wrote:I ended up rolling back to 3.9.8-1 and nvidia 319.32-1
What is the easiest way to roll back the kernel version to 3.9.8-1?
Ummm... get the 3.9.8-1 package and install it? I feel like I either missed something or this is a trick question...
Offline
nobled wrote:jantman wrote:I ended up rolling back to 3.9.8-1 and nvidia 319.32-1
What is the easiest way to roll back the kernel version to 3.9.8-1?
Ummm... get the 3.9.8-1 package and install it? I feel like I either missed something or this is a trick question...
No, I'm just an arch noob and didn't know about the ARM. I thought I might have to compile the kernel myself. With other distros I've used it wasn't super easy to roll back to a previous kernel because it would break the dependancies of a bunch of packages, and things like that. If I read the ARM page right, it says that it's not safe to mix ARM packages and regular ones, so doesn't that mean I'd have to update my packman.conf to point at a date that included that kernel version and then do an update? Would an update even work since I'd have newer versions than what's in the repo? If I install just that kernel package without changing my packman.conf wouldn't any other packages that include kernel modules potentially no longer work? Thanks for the help.
Offline
It also looks like the 3.9.8-1 package isn't in the ARM anymore. Is there another place to get older packages?
Offline
In the past week, my wireless speed has dropped suddenly and drastically. I'm using the broadcom-wl-dkms drivers. I've tested using long downloads and testmy.net. It seems like every few seconds, it stalls out for a few seconds, then resumes. Is anybody else experiencing this? I can't connect it to one particular upgrade. Took me a while to realize that's what was happening.
Offline
@nobled:
I find that the best way to get old packages in Arch is just to compile them from source. This isn't too hard on Arch, you can just use ABS (dump all these files in a folder and run "makepkg"): https://projects.archlinux.org/svntogit … 548e485cf8
I have no idea if that is acceptable practice or not. (I suppose you would have to recompile it everytime you update something it depends on, and compile everything that depends on it whenever you want to upgrade them?) I only do this for ibus, because I need functionality that was removed from newer versions, and it's not too much of a pain for me to compile everything that depends on it because there aren't that many such things, but I can imagine this being annoying for other packages. (And also if you ever plan on going back to the current version, then you'd have to remember which packages you need to resync.) Now if only Arch had a portage like system that would do all this automatically...
I also have no idea if this would work for binary packages like nvidia.
@abruptadieu:
I haven't noticed anything in particular with the wireless speeds. But seriously, my wireless card has its good weeks and bad weeks. I'm not even sure if it's at all related to updates or not. I've just learnt to deal with it by plugging in my ethernet adaptor and regretting buying a macbook. (I should get one of those handy USB wireless plugs someday, anyone have any good recommendations?)
Also, did anyone have serious screen artifact/tarring problems since the latest mesa update? I filed a bug report here: https://bugs.archlinux.org/task/39352
No idea how to fix this, so I've just downgraded (via ABS) for now. I only downgraded the 64bit package because I couldn't compile the multilib one for some reason. The problems have gone away anyways. (I rarely use skype or acroread, the only programs that need 32bit mesa, but even when I ran them as a test there were no problems.)
EDIT: Also, I just saw this thread that says that the new kernel has optimus support. Does this concern us? (I have never gotten nvidia working in the first place, and intel works better than nouveau for me so... not even sure if there is a point, but hey.) https://bbs.archlinux.org/viewtopic.php?id=172920
Last edited by sokuban (2014-03-14 05:06:25)
Offline
EDIT: Also, I just saw this thread that says that the new kernel has optimus support. Does this concern us? (I have never gotten nvidia working in the first place, and intel works better than nouveau for me so... not even sure if there is a point, but hey.) https://bbs.archlinux.org/viewtopic.php?id=172920
Apple didn't use Optimus for it's graphics switching so anything done for optimus likely won't work on MacBooks. When running windows in bootcamp the optimus drivers don't work and there is no way to get graphics switching working, you're just stuck using the nvidia card all the time.
(see this and this from apple's support site)
Also, does anyone know if you can get multiple thunderbolt / display port monitors working using the intel driver or nouveau driver? Or can you get one external display and the laptop display to work with either driver?
Offline
OK, so I've been running Arch on two retina Macbooks pretty well (fingers crossed) the past two weeks now. Just wanted to post my experiences of this.
I turned off the nvidia card for the time being since I don't use an external monitor or projector or anything of that sort, so I'm just running the intel card. Backlight control didn't work for the nvidia card even when I switched to the open source driver. It works for the Intel card, though, even when both GPUs are on.
Found that the battery life is really good when the dGPU is off (6-7 hours), which is expected. It may be a deal-breaker if you need to use multiple monitors, though, as that requires the dGPU.
I documented the setup process here; if anyone has feedback on the process I outlined, it'd be much appreciated. The notes should be applicable to the 10,x models. I have both the 15" and 13" models. Was successful in replicating the process on the second model.
Also, anybody have trouble building broadcom-wl for 3.14.2 on the release of the latest kernel? I was having trouble earlier after running a full system update (which updates the kernel and gcc), but thanks to some comments on the AUR packages for the dkms version of that module I finally got it to build successfully. I put instructions and the relevant PKGBUILD and patch here. Hope that helps.
How have people found the performance of broadcom-wl-dkms compared to broadcom-wl? I got pretty unstable wifi from the dkms module; connection would drop off after awhile. Haven't found that to be the case on the non-dkms module, though.
Last edited by hckr (2014-05-03 00:37:51)
Offline
Has anyone ever managed to use the SD card reader in this laptop?
I bought a PNY Storedge SD card, planning to move my Linux installations there, but I am getting the dreaded "mmc0: Timeout waiting for hardware interrupt." error whenever I plug it in. This appears similar to https://bugzilla.kernel.org/show_bug.cgi?id=59311
I spent a few hours trying to find a workaround, without success. None of the commonly suggested workarounds appear to work (noapic, irqpoll or pcie_aspm=off).
Any ideas?
Edit: I should mention I've tried kernels 3.13 and 3.14 without success. Ubuntu 14.04 also fails.
Edit 2: I should also mention that the same SD card works on Mac OS X and Windows, so this is probably a driver issue, not a hardware issue.
Last edited by TheFiddler (2014-05-07 08:31:05)
The Open Toolkit: OpenGL, OpenAL, OpenCL and Vulkan for Mono/.Net.
Offline
I'm on the MacBookPro10,1 (Mid 2012 15in Retina) and I've been getting disconnected from the wifi every few minutes. I've worked around it with a cronie script that runs every minute.
#!/bin/bash
target=www.google.com
count=$( ping -c 1 $target | grep icmp* | wc -l )
if [ $count -eq 0 ]
then
sudo systemctl stop NetworkManager
sudo systemctl start NetworkManager
else
echo "Yes! Host is Alive!"
fi
I've been getting a few errors in my dmesg, the regualr inform_bss_frame errors and the invalid ROM contents are the ones that worry me. The relevant sections of the dmesg are below.
[ +0.084136] INFO @wl_cfg80211_attach : Registered CFG80211 phy
[ +0.001184] applesmc: key=493 fan=2 temp=44 index=43 acc=0 lux=2 kbd=1
[ +0.001253] lib80211_crypt: registered algorithm 'TKIP'
[ +0.000144] wlan0: Broadcom BCM4331 802.11 Hybrid Wireless Controller 6.30.223.141 (r415941)
[ +0.029604] i915 0000:00:02.0: irq 49 for MSI/MSI-X
[ +0.000014] [drm] Supports vblank timestamp caching Rev 2 (21.10.2013).
[ +0.000003] [drm] Driver supports precise vblank timestamp query.
[ +0.000010] i915 0000:00:02.0: Invalid ROM contents
[ +0.000007] [drm] failed to find VBIOS tables
[ +0.000151] vgaarb: device changed decodes: PCI:0000:00:02.0,olddecodes=io+mem,decodes=none:owns=none
[ +0.023189] systemd-udevd[210]: renamed network interface wlan0 to wlp4s0
and
[ +1.999875] ERROR @wl_inform_single_bss : cfg80211_inform_bss_frame error
[ +3.454905] ERROR @wl_inform_single_bss : cfg80211_inform_bss_frame error
[ +1.931689] fuse init (API version 7.22)
[Jun 9 17:37] ERROR @wl_inform_single_bss : cfg80211_inform_bss_frame error
[ +42.967942] ERROR @wl_inform_single_bss : cfg80211_inform_bss_frame error
[Jun 9 17:38] ERROR @wl_inform_single_bss : cfg80211_inform_bss_frame error
[Jun 9 17:40] ERROR @wl_inform_single_bss : cfg80211_inform_bss_frame error
[Jun 9 17:41] ERROR @wl_inform_single_bss : cfg80211_inform_bss_frame error
[Jun 9 17:43] ERROR @wl_inform_single_bss : cfg80211_inform_bss_frame error
[Jun 9 17:45] ERROR @wl_inform_single_bss : cfg80211_inform_bss_frame error
[Jun 9 17:47] ERROR @wl_inform_single_bss : cfg80211_inform_bss_frame error
My questions are thus:
Has anyone else experienced this or similar issues?
If so, what did you do fix it?
If you haven't had similar problems, do you have any suggestions for how fix this?
Offline