no graphics for Debian buster on Dell Inspiron
Hi, after nearly 20 years of happiness with Debian distros I am struggling to get Debian buster working on a Dell Inspiron 14 2-in-1. The installer completes, but I have no working wifi and no working graphics. Let's concentrate on the graphics for now, since it is more urgent. The laptop has an i7-10510U which should come with an embedded UHD 620 graphics and also a discrete nVidia MX230 GPU. I have tried following different ideas found on the web but with no luck. My guess is that the embedded graphics is not recognized by the driver and that MX230 does not work with nouveau, but here are some details. The following tests have been done on a fresh install (kde task selected in the installer with only firmware-nonfree, firmware-iwlwifi added afterwards). First hist: I have only /dev/dri/card0 and /dev/dri/renderD128 and they both seem related to the nVidia card; I do not seem to have any /dev/dri entry for the Intel graphics. Starting X server does not work and /var/log/Xorg.0.log contains [ 9.948] (II) xfree86: Adding drm device (/dev/dri/card0) [ 9.950] (--) PCI:*(0@0:2:0) 8086:9b41:1028:095d rev 2, Mem @ 0x60/16777216, 0x40/268435456, I/O @ 0x4000/64, BIOS @ 0x/131072 [ 9.950] (--) PCI: (1@0:0:0) 10de:1d11:1028:095d rev 161, Mem @ 0x8200/16777216, 0x7000/268435456, 0x8000/33554432, I/O @ 0x3000/128 and then loads the modesetting, fbdev, vesa, fbdevhw modules and finally fais with (EE) Cannot run in framebuffer mode. Please specify busIDs?? for all framebuffer devices If I start X with a configuration file containing only Section "Device" ?? Identifier "Card0" ?? Driver "intel" BusID "PCI:0:2:0" EndSection then it fails with [ 281.845] (II) Module intel: vendor="X.Org Foundation" [ 281.845] ?? compiled for 1.20.1, module version = 2.99.917 [ 281.845] ?? Module class: X.Org Video Driver [ 281.845] ?? ABI class: X.Org Video Driver, version 24.0 [ 281.845] (II) intel: Driver for Intel(R) Integrated Graphics Chipsets: ?? i810, i810-dc100, i810e, i815, i830M, 845G, 854, 852GM/855GM, 865G, ?? 915G, E7221 (i915), 915GM, 945G, 945GM, 945GME, Pineview GM, ?? Pineview G, 965G, G35, 965Q, 946GZ, 965GM, 965GME/GLE, G33, Q35, Q33, ?? GM45, 4 Series, G45/G43, Q45/Q43, G41, B43 [ 281.845] (II) intel: Driver for Intel(R) HD Graphics [ 281.845] (II) intel: Driver for Intel(R) Iris(TM) Graphics [ 281.845] (II) intel: Driver for Intel(R) Iris(TM) Pro Graphics [ 283.907] intel: waited 2020 ms for i915.ko driver to load [ 285.942] intel: waited 2020 ms for i915.ko driver to load [ 285.948] (EE) No devices detected. (Note that the i915 is loaded, but upon loading it does not create a /dev/dri entry, so I guess that somethings wrong with it...) Trying to force the use of the discrete graphics with a conffile with only Section "Device" ?? Identifier "Card1" ?? Driver "nouveau" ?? BusID "PCI:1:0:0" EndSection instead results in a segfault: [ 716.577] (II) LoadModule: "nouveau" [ 716.577] (II) Loading /usr/lib/xorg/modules/drivers/nouveau_drv.so [ 716.596] (II) Module nouveau: vendor="X.Org Foundation" [ 716.596] ?? compiled for 1.20.3, module version = 1.0.16 [ 716.596] ?? Module class: X.Org Video Driver [ 716.596] ?? ABI class: X.Org Video Driver, version 24.0 [ 716.596] (II) NOUVEAU driver Date: Mon Jan 28 23:25:58 2019 -0500 [ 716.596] (II) NOUVEAU driver for NVIDIA chipset families : [ 716.596] ?? RIVA TNT?? (NV04) [ 716.596] ?? RIVA TNT2 (NV05) [ 716.596] ?? GeForce 256 (NV10) [ 716.596] ?? GeForce 2 (NV11, NV15) [ 716.596] ?? GeForce 4MX (NV17, NV18) [ 716.596] ?? GeForce 3 (NV20) [ 716.596] ?? GeForce 4Ti (NV25, NV28) [ 716.596] ?? GeForce FX?? (NV3x) [ 716.596] ?? GeForce 6 (NV4x) [ 716.596] ?? GeForce 7 (G7x) [ 716.596] ?? GeForce 8 (G8x) [ 716.596] ?? GeForce 9 (G9x) [ 716.596] ?? GeForce GTX 2xx/3xx (GT2xx) [ 716.596] ?? GeForce GTX 4xx/5xx (GFxxx) [ 716.597] ?? GeForce GTX 6xx/7xx (GKxxx) [ 716.597] ?? GeForce GTX 9xx (GMxxx) [ 716.597] ?? GeForce GTX 10xx?? (GPxxx) [ 716.602] (II) [drm] nouveau interface version: 1.3.1 [ 716.603] (II) Loading sub module "dri2" [ 716.603] (II) LoadModule: "dri2" [ 716.603] (II) Module "dri2" already built-in [ 716.603] (--) NOUVEAU(0): Chipset: "NVIDIA NV138" [ 716.603
Re: no graphics for Debian buster on Dell Inspiron
On 16/11/19 23:08, eric wrote: On 11/16/19 9:22 AM, Matteo Semplice wrote: Hi, after nearly 20 years of happiness with Debian distros I am struggling to get Debian buster working on a Dell Inspiron 14 2-in-1. The installer completes, but I have no working wifi and no working graphics. Let's concentrate on the graphics for now, since it is more urgent. The laptop has an i7-10510U which should come with an embedded UHD 620 graphics and also a discrete nVidia MX230 GPU. I have tried following different ideas found on the web but with no luck. My guess is that the embedded graphics is not recognized by the driver and that MX230 does not work with nouveau, but here are some details. The following tests have been done on a fresh install (kde task selected in the installer with only firmware-nonfree, firmware-iwlwifi added afterwards). [...] On a side note, I have tried to upgrade the kernel to the 5.2 in buster-backports but both bop.2 and bop.3 packages result in unbootable kernels: the boot process just hangs somewhere. I have uninstalled them, but if you think that a newer kernel may help I can try to reinstall and boot it without the quiet option to see what goes wrong. Thanks in advance for any suggestion. Best ?? Matteo I have little experience in this regard but do have a cybertron laptop with the hybrid intel/nvidia graphics. There is a lot of confusing information on the web.?? I would start with blacklisting the nouveau driver and removing all the nvidia drivers to see if you can get just the intel graphics to work. Regards, Eric Hi, ??thanks for the tip! You've put me in the right direction and I've also done further research: in the end it seems that it is the i7-10510U cpu that is not yet fully supported. At present, I can boot 5.2 and 5.3 kernels from buster-backports and sid provided that I blacklist the intel_lpss_pci and the nouveau module. (After boot I have tried to modprobe nouveau and it seems ok, but I have not tested much as for now I just need to have the machine up and running with just any graphics user interface). It's not clear to me if better support for my Comet Lake i7-10510U will be granted by the same Ice Lake patches expected in kernel 5.4 or not, but I guess that I just need to wait few more days. For reference, I just report here the link to this reddit about installing on XPS 13 (which has an Ice Lake cpu) https://old.reddit.com/r/Dell/comments/cx0fkc/xps_13_2_in_1_7390_linux_boot_attempt/ I have taken from there the idea to blacklist the lpss driver; I'll test some of the other tips given there and report back. Regards, ?? Matteo
installation report for Debian 10 (buster) on a Dell Inspiron 14 2-in-1
Dear all, as promised, here comes an installation report of Debian Buster on a Dell Inspirion 14 2-in-1 laptop in the hope that it may be useful to someone else. (By the way, there used to be a website that collected these installation reports, but I cannot find it any more... Any ideas of where I could publish and keep updated this report?) The main problem turned out to be a lack of support for the i7-10510U by kernels prior to version 5.4, so currently I see no shorter route than mine: I have installed normally, booted into console and compiled a new kernel from source. Of course the situation may get better when newer kernels will become available in the Debian repositories. * first the SSD drive must be set in AHCI mode (as opposed to RAID) in the UEFI-BIOS. Unfortunately, Windows on the first boot sets it to RAID and then loads only the drivers for RAID: if one simply changes the setting in UEFI, Windows will not be able to boot any more. One possible procedure is to set Windows to boot in safe mode, change the UEFI setting and then reset the Windows boot. In this way, Windows will find the disk in AHCI mode when booting in safe mode and will manage to boot and set itself for loading AHCI drivers, so that normal boots will succed afterwards. The following links illustrate possible procedures to achieve this. (Note: I wonder if entering UEFI before the first boot into Windows and change the setting there might avoid this hassle altogether) http://triplescomputers.com/blog/uncategorized/solution-switch-windows-10-from-raidide-to-ahci-operation/ https://samnicholls.net/2016/01/14/how-to-switch-sata-raid-to-ahci-windows-10-xps-13/ Of course if you wish to keep Windows installed (useful e.g. to test the hardware...) you'll need to shrink the Windows partition before installing Debian. * remeber that right after the install (Nov 2019), you'll have no graphics system and no wifi. It is thus useful and almost mandatory to have an USB ethernet adapter, a supported USB WIFI adapter or even just an USB cable to use the tethering function of your mobile... * the installer will want to install the firmware-iwlwifi package. It could be useful to have it on a USB stick (in the root directory of the stick), but for me the WiFi adapter did not work out-of-the-box, so this is kind of useless with the current version (Nov 2019) of kernel/iwlwifi module and firmware. Nevertheless it is worth a try, in case newer versions bacame available in the Debian repos. * fire up the Debian 10 installer and follow the usual steps to install your favourite tasks and finish it off by installing grub. * Right after installation the pc will reboot into a terminal or may not even complete the boot sequence, but blacklisting intel_lpss_pci should do the trick: access the GRUB menu, highlight the "Adavanced options" and press "e" to edit the kernel command line, remove "quiet" and add "module_blacklist=intellpss_pci", then press F10 to boot. I had also success by booting in single core mode (kernel option "nosmp"). In order to fix these problems you'll need a recent enough kernel and a patched intel_lpsss_pci driver. I have observed different behaviours with kernels from different linux-image-amd64 packages (versions from buster, buster-backports, experimental), so your milage may vary here, but in general it sohuld be getting better and better when new kernel versions become available in Debian * wifi will most likely not be working at this point, so you'll need an alternative network connection for the following steps * fix support for the i7-10510U and Intel UHD-620 graphics I have found very useful the discussions and instructions at https://www.reddit.com/r/Dell/comments/cx0fkc/xps_13_2_in_1_7390_linux_boot_attempt/. In orderￜ to have graphics working correctly, after install you need to upgrade the kernel to at least version 5.4 (from buster-backports, experimental or compiling it from the sources at linux.org). I have mainly followed the procedure at section 4.6 of the Debian kernel handbook at https://kernel-team.pages.debian.net/kernel-handbook/ch-common-tasks.html#s-common-official. (Note: I have actually cloned a git repo for the kernel and not the tarballs since I foresee having to track the most recent kernels for an up-to-date support of all the hardware of this machine) At the time of writing (kernel 5.4-rc8), you'll need to apply a patch to the intel_lpss_pci module.ￜ The author of the above reddit has also set up a git repo with useful code and patches. Have a look at the script opt/scripts/build-custom-kernel.sh from https://gitlab.com/emrose/xps13-7390_debian.git, but keep in mind that it targets an XPS model and an older kernel version. On the Inspiron 14 2-in-1 I had to apply only the lpss.patch in the directory opt/patches of the git repo and the other ones do not seem to be needed. * nVidia MX230 support I have currently blacklist
Messy display on Satellite 110CS
Dear all,
I am running Linux2.1 on a Toshiba Satellite 110CS since 2 months
ago. (with Xfree and xdm).
Every now and then (say once every 2 weeks) the display in X would
become blurry, as if the orizontal lines were not aligned any more (i.e.
display shifted horizontally, deforming the characters & so on); when it
happened the character terminals used to be still fine and switching a
couple of times to&from (like Ctrl-Alt-F1, Ctrl-Alt-F7, Ctrl-ALtF1,
Ctrl-Alt-F7) would bring my graphics display back in order. It was fine
under Windows95 (the few times I booted into it). I was meaning to look
into the problem as soon as I had some spare time, but...
yesterday night the thing happened again and this time also the
charater display is affected and it happens also under Windows. Now I can
just read enough to log in and give the "halt" command (or to guess where
the Shutdwon button is when in Win95).
Any ideas anyone?
I am not sure what information is relevant in this case. I'll give
you all what I remember on top of my head and could be related.
Xfree86 config. Refresh rates are an interpolation between the
ones given in 2 different config found on the internet: one was specifying
very narrow ranges, the other very wide, but neither of them would work.
The computer is permenently connected to an ethernet socket and
has been running ssh, telnet and Netscape 90% of the time it has been
used. I tried TeX&Xdvi just to check that the installation went fine. I
usually bring them up from the server in my department through the ssh
channel.
Thanks for any suggestion.
matteo
---
Matteo Semplice
Wadham College
Oxford OX1 3PN
U.K.
Re: Messy display on Satellite 110CS
I really mean whatever was available 2 months ago as stable version on debian ftp server. Kernel version should appear at boot but, as I said, the diplay now is unreadable also in text mode... Anyway, would it make a difference what kernel I am running? (I am afraid the problem looks more hardware-related since seems to be operative-system independent now. Linux might have helped in creating it, but I do not think it is the cause now). matteo On Wed, 29 Mar 2000 [EMAIL PROTECTED] wrote: > > wadh1000> I am running Linux2.1 on a Toshiba Satellite 110CS > wadh1000> since 2 months ago. (with Xfree and xdm). > > by "Linux2.1", do you mean kernel version 2.1? if so, why are you running > an old development kernel? > > > -- > To UNSUBSCRIBE, email to [EMAIL PROTECTED] > with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED] > > - Matteo Semplice Wadham College Oxford OX1 3PN U.K.
Toshiba laptops.
Dear all, I saw a few qestions abot specific models of Toshiba laptops comeing on the list recently. I am afraid I cannot be of any help diretctly, but Toshiba (although sadly does not support Linux directly) hosts a mailing list on which a great amount of discussion about running Linux on specific models is going on. More info at http://www.tce.co.jp/linux/ matteo PS while I am on the subject, also the Volunteer Support Databse is worth checking. That is at: http://www.cs.utexas.edu/users/kharker/linux-laptop/volunteer.html - Matteo Semplice Wadham College Oxford OX1 3PN U.K.
Re: toshiba 1800
Domenic, even if Toshiba does not officially support Linux, they provide some info at http://www.tce.co.jp/linux There you'll find also info on a mailing list specific on Linux-on-Toshiba issues and that is read by a lot of Toshiba users and hackers... Good luck, matteo PS I am running Debian potato on a Satellite 110CS. I didn't check the specs of the 1800, but if it turns out that I can be of any help, feel free to email me. ------- Matteo Semplice Wadham College Oxford OX1 3PN U.K.
Re: APM suspend/resume problem on Toshiba Satellite 2800DVD-C302
Mike, "apm --suspend" works fine on my (ancient) Satellite110CS with kernel 2.2.17 and CONFIG_APM=y # CONFIG_APM_IGNORE_USER_SUSPEND is not set CONFIG_APM_DO_ENABLE=y CONFIG_APM_CPU_IDLE=y CONFIG_APM_DISPLAY_BLANK=y # CONFIG_APM_IGNORE_SUSPEND_BOUNCE is not set CONFIG_APM_RTC_IS_GMT=y # CONFIG_APM_ALLOW_INTS is not set # CONFIG_APM_REAL_MODE_POWER_OFF is not set # CONFIG_TOSHIBA is not set (Don't ask why CONFIG_TOSHIBA is not set: clearly I wasn't very awake when I last configured...) Only I have to exit my X session before suspending, o'wise it'll come up at another resolution and it would be back to normal only after exiting/reentering X. So, I have no clue to why it isn't working on your model, but I just wanted to point you to http://newsletter.toshiba-tro.de/main/index.html and http://linux.toshiba-dme.co.jp/linux/ where you may find info for your model or browse the toshiba-linux mailing list logs) matteo On Wed, 29 Aug 2001, Mike Alborn wrote: > Hi, > This might have been posted before, but I couldn't find it in the > archives... > I'm having trouble getting apm to work on my laptop. apm -s (suspend) > works (although slightly differently than Windows), and when I resume, the > following occurs: > > * the display remains blank > * the network is down > * the system responds to the keyboard > > By restarting the interface, I can get the network to respond, so I am > able to ssh in from another machine and take a look. I can manually activate > the > backlight, but the screen is displaying patches of multicoloured vertical > bars. > It seems that I can make a particular device work by resetting it (pcmcia, > sound, etc.,) but I have not found a way to reset the display. > BTW, I am suspending from the console. > I've played with the different kernel options for the apm module, but > they seem to be geared toward working around a machine that totally hangs; in > any case, I have had no luck changing these settings. > > > # uname -r > 2.4.5-586 (this is the debian kernel-image-2.4.5-586 package) > > > apm options: > CONFIG_APM=m > CONFIG_APM_IGNORE_USER_SUSPEND is not set > CONFIG_APM_DO_ENABLE is not set > CONFIG_APM_CPU_IDLE is not set > CONFIG_APM_DISPLAY_BLANK is not set > CONFIG_APM_RTC_IS_GMT=y > CONFIG_APM_ALLOW_INTS is not set > CONFIG_APM_REAL_MODE_POWER_OFF is not set > > Does anyone know of a way to fix this problem? I can provide more > information on the system, if needed. > > -Mike Alborn > [EMAIL PROTECTED] > http://odoitau.dyn.dhs.org > > -- > Don't shout for help at night. You might wake your neighbors. > -- Stanislaw J. Lem, "Unkempt Thoughts" > --- Matteo Semplice Wadham College Oxford OX1 3PN U.K.
Re: PCMCIA / NIC /Cable on Inspiron 8000
On Mon, 23 Jul 2001, noah wrote: > The system no longer hangs, I get a beep when the PCMCIA lines go by > during boot, and the "10/100" light on the dongle (the ethernet to > PCMCIA adapter thingy) lights up. "dmesg" reveals: Hmmm... Is that an high pitch beep? PCMCIA-HOWTO says that you should get 2 high pitch beeps: one for card recognition and a second one for card configuration. Section 3.9 in the HOWTO describe the "only 1 high beep" situation: does that apply to you? matteo --- Matteo Semplice Wadham College Oxford OX1 3PN U.K. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: toshiba 1800
Domenic, even if Toshiba does not officially support Linux, they provide some info at http://www.tce.co.jp/linux There you'll find also info on a mailing list specific on Linux-on-Toshiba issues and that is read by a lot of Toshiba users and hackers... Good luck, matteo PS I am running Debian potato on a Satellite 110CS. I didn't check the specs of the 1800, but if it turns out that I can be of any help, feel free to email me. ------- Matteo Semplice Wadham College Oxford OX1 3PN U.K. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: APM suspend/resume problem on Toshiba Satellite 2800DVD-C302
Mike, "apm --suspend" works fine on my (ancient) Satellite110CS with kernel 2.2.17 and CONFIG_APM=y # CONFIG_APM_IGNORE_USER_SUSPEND is not set CONFIG_APM_DO_ENABLE=y CONFIG_APM_CPU_IDLE=y CONFIG_APM_DISPLAY_BLANK=y # CONFIG_APM_IGNORE_SUSPEND_BOUNCE is not set CONFIG_APM_RTC_IS_GMT=y # CONFIG_APM_ALLOW_INTS is not set # CONFIG_APM_REAL_MODE_POWER_OFF is not set # CONFIG_TOSHIBA is not set (Don't ask why CONFIG_TOSHIBA is not set: clearly I wasn't very awake when I last configured...) Only I have to exit my X session before suspending, o'wise it'll come up at another resolution and it would be back to normal only after exiting/reentering X. So, I have no clue to why it isn't working on your model, but I just wanted to point you to http://newsletter.toshiba-tro.de/main/index.html and http://linux.toshiba-dme.co.jp/linux/ where you may find info for your model or browse the toshiba-linux mailing list logs) matteo On Wed, 29 Aug 2001, Mike Alborn wrote: > Hi, > This might have been posted before, but I couldn't find it in the > archives... > I'm having trouble getting apm to work on my laptop. apm -s (suspend) > works (although slightly differently than Windows), and when I resume, the > following occurs: > > * the display remains blank > * the network is down > * the system responds to the keyboard > > By restarting the interface, I can get the network to respond, so I am > able to ssh in from another machine and take a look. I can manually activate the > backlight, but the screen is displaying patches of multicoloured vertical bars. > It seems that I can make a particular device work by resetting it (pcmcia, > sound, etc.,) but I have not found a way to reset the display. > BTW, I am suspending from the console. > I've played with the different kernel options for the apm module, but > they seem to be geared toward working around a machine that totally hangs; in > any case, I have had no luck changing these settings. > > > # uname -r > 2.4.5-586 (this is the debian kernel-image-2.4.5-586 package) > > > apm options: > CONFIG_APM=m > CONFIG_APM_IGNORE_USER_SUSPEND is not set > CONFIG_APM_DO_ENABLE is not set > CONFIG_APM_CPU_IDLE is not set > CONFIG_APM_DISPLAY_BLANK is not set > CONFIG_APM_RTC_IS_GMT=y > CONFIG_APM_ALLOW_INTS is not set > CONFIG_APM_REAL_MODE_POWER_OFF is not set > > Does anyone know of a way to fix this problem? I can provide more > information on the system, if needed. > > -Mike Alborn > [EMAIL PROTECTED] > http://odoitau.dyn.dhs.org > > -- > Don't shout for help at night. You might wake your neighbors. > -- Stanislaw J. Lem, "Unkempt Thoughts" > --- Matteo Semplice Wadham College Oxford OX1 3PN U.K. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
