| View previous topic :: View next topic |
| Author |
Message |
gilhad
n00b
Joined: 18 Jan 2007
Posts: 15
|
 Posted: Sat Apr 27, 2024 6:11 am Post subject: Blank screen after boot - Aspire One D270 |
 |
|
Hello,
I have problem to set kernel/boot/grapfic card correctly.
After boot I can see some messages, then messages with [OK] on end, then screen blanks and change resolution, probabely logo with tuxes try to display (big blank rectanle on top right of screen) then more messages with [OK] on right (like three lines or so) and then screen blanks again and stand blank.
I can login and run commands (blindly), I can startx and the X server probably starts and music start playing (so it probably works), but screen is blank.
If I re-boot with kernel parameter nomodeset, the boot looks similar, but no change of resolution and no logo, it goes in text up to login. I cannot start X server, and the text screen works chaotically, the text is sometime overridden by boot messages, lines scroll on top only and such.
But if I re-boot after that without nomodeset after this, the boot goes right, the logo is still just white rectangle, but everything after is right, no garbled text and I can start Xserver and it works normally.
---
The netbook is Aspire One D270 with Intel(R) Atom(TM) CPU N2800 @ 1.86GHz.
| Code: | uname -a
Linux Mahoro 6.6.21-gentoo-x86_64 #26 SMP PREEMPT_DYNAMIC Sat Apr 27 07:11:14 CEST 2024 x86_64 Intel(R) Atom(TM) CPU N2800 @ 1.86GHz GenuineIntel GNU/Linux
|
It worked many years on 32-bit system (even with tuxes logo good), now I tried to install 64 bit and the problems started.
It have the cursed PowerVR
| Code: | 00:02.0 VGA compatible controller: Intel Corporation Atom Processor D2xxx/N2xxx Integrated Graphics Controller (rev 09)
Subsystem: Acer Incorporated [ALI] Atom Processor D2xxx/N2xxx Integrated Graphics Controller
Kernel driver in use: gma500
Kernel modules: gma500_gfx |
I use legacy grub with the lines
| Code: | title Mahoro/Gentoo Linux 2024 acpi_osi="Linux"
root (hd0,0)
kernel /boot/Mahoro/bzImage.2024 root=/dev/sda3 video=inteldrmfb:mode_option=1280x1024-24 net.ifnames=0 raid=noautodetect resume=/dev/sda11 acpi_osi="Linux"
title Mahoro/Gentoo Linux 2024 nomodeset
root (hd0,0)
kernel /boot/Mahoro/bzImage.2024 root=/dev/sda3 nomodeset video=inteldrmfb:mode_option=1280x1024-24 net.ifnames=0 raid=noautodetect resume=/dev/sda11
|
---
I tried to search help, but had no luck, all I found was outdated for kernels 2.xx and 3.xx, not near 6.6.21, I tried the acpi_osi="Linux" recommended on https://wiki.gentoo.org/wiki/Intel, but it not solve it.
Thanks for any comments |
|
| Back to top |
|
 |
NeddySeagoon
Administrator
Joined: 05 Jul 2003
Posts: 54298
Location: 56N 3W
|
| Posted: Sat Apr 27, 2024 3:08 pm Post subject: |
|
|
gilhad,
This usually happens when the built into the kernel console driver, give way to the DRM console driver later in the boot process but the DRM console driver is broken for some reason.
We need to see your kernel .config file and your entire output.
Use wgetpaste to put your kernel .config onto a pastebin site.
_________________
Regards,
NeddySeagoon
Computer users fall into two groups:-
those that do backups
those that have never had a hard drive fail. |
|
| Back to top |
|
|
gilhad
n00b
Joined: 18 Jan 2007
Posts: 15
|
| Posted: Sun Apr 28, 2024 2:00 am Post subject: |
|
|
Thank you
Here is my lspci -nnk
and my actual /usr/src/linux/.config
I hope, I posted it right
(There may be many settings, which I set by mistake, misunderstanding or just inherited from somewhere else, I am not kernel hacker. I am open to suggestions  ) |
|
| Back to top |
|
|
NeddySeagoon
Administrator
Joined: 05 Jul 2003
Posts: 54298
Location: 56N 3W
|
| Posted: Sun Apr 28, 2024 9:02 am Post subject: |
|
|
gilhad,
You have exactly one video card whick keeps things simple.
| Code: | 00:02.0 VGA compatible controller [0300]: Intel Corporation Atom Processor D2xxx/N2xxx Integrated Graphics Controller [8086:0be2] (rev 09)
Subsystem: Acer Incorporated [ALI] Atom Processor D2xxx/N2xxx Integrated Graphics Controller [1025:061f]
Kernel driver in use: gma500
Kernel modules: gma500_gfx |
Its using the gma500 driver.
From your kernel .config
is built into the kernel. From your opening post, that works.
The wheel comes off when the kernel switches from the vesafb driver to the gma500 driver.
Turn off | Code: | | CONFIG_VIDEO_NOMODESET=y |
There is a kernel command line parameter for that. Its only intended as a get you going asd should never be set by default.
It prevents all the modesetting video drivers from working, and that's most of them.
xorg-drivers suppornts the following | Code: | | VIDEO_CARDS="amdgpu ast dummy fbdev freedreno geode i915 intel mga nouveau nvidia omap qxl r128 radeon radeonsi siliconmotion tegra vc4 vesa via virtualbox vmware"} |
There is no drivel listed for your chipset.
That means you have a choice of fbdev or modesetting ... but modesetting support is off in you kernel.
fbdev only uses your GPU RAM as the pixel buffer. The CPU does all the hard work. If plots every pixel. You need to add this to your VIDEO_CARDS and make a xorg.conf fragment if you want to try it.
modesetting will use mesa to try to give you some acceleration. That is free with Xorg and is probably the least worst driver.
Your chipset is not supported by any of the Intel drivers.
At one time, there was the EMGD Driver but https://community.intel.com/t5/Embedded-Intel-Atom-Processors/EMGD-GMA500-Poulsbo-on-GENTOO-Linux/td-p/254461 and that was 12 years ago.
_________________
Regards,
NeddySeagoon
Computer users fall into two groups:-
those that do backups
those that have never had a hard drive fail. |
|
| Back to top |
|
|
gilhad
n00b
Joined: 18 Jan 2007
Posts: 15
|
| Posted: Sun Apr 28, 2024 12:26 pm Post subject: |
|
|
I tried to find VIDEO_NOMODESET in make menuconfig, but it looks like it is not in the menu, but it is just some automatic variable (There is no number to press to get there, there is no menu structure)
| Code: | Symbol: VIDEO_NOMODESET [=y]
Type : bool
Defined at drivers/video/Kconfig:24
Selected by [y]:
- DRM [=y] && HAS_IOMEM [=y] && (AGP [=n] || AGP [=n]=n) && !EMULATED_CMPXCHG && HAS_DMA [=y]
Selected by [n]:
- FB_CIRRUS [=n] && HAS_IOMEM [=y] && FB [=y] && (ZORRO || PCI [=y])
- FB_PM2 [=n] && HAS_IOMEM [=y] && FB [=y] && (AMIGA && BROKEN [=n] || PCI [=y])
- FB_CYBER2000 [=n] && HAS_IOMEM [=y] && FB [=y] && PCI [=y] && (BROKEN [=n] || !SPARC64)
- FB_CT65550 [=n] && HAS_IOMEM [=y] && FB [=y]=y && PPC32 && PCI [=y]
- FB_ASILIANT [=n] && HAS_IOMEM [=y] && FB [=y]=y && PCI [=y]
- FB_IMSTT [=n] && HAS_IOMEM [=y] && FB [=y]=y && PCI [=y]
- FB_TGA [=n] && HAS_IOMEM [=y] && FB [=y] && (PCI [=y] || TC) && (ALPHA || TC)
- FB_XVR500 [=n] && HAS_IOMEM [=y] && FB [=y]=y && PCI [=y] && SPARC64
- FB_XVR2500 [=n] && HAS_IOMEM [=y] && FB [=y]=y && PCI [=y] && SPARC64
- FB_PVR2 [=n] && FB [=y] && HAS_IOMEM [=y] && (SH_DREAMCAST || COMPILE_TEST [=n])
- FB_NVIDIA [=n] && HAS_IOMEM [=y] && FB [=y] && PCI [=y]
- FB_RIVA [=n] && HAS_IOMEM [=y] && FB [=y] && PCI [=y]
- FB_I740 [=n] && HAS_IOMEM [=y] && FB [=y] && PCI [=y]
- FB_I810 [=n] && HAS_IOMEM [=y] && FB [=y] && PCI [=y] && X86_32 [=n] && AGP_INTEL [=n]
- FB_LE80578 [=n] && HAS_IOMEM [=y] && FB [=y] && PCI [=y] && X86 [=y]
- FB_INTEL [=n] && HAS_IOMEM [=y] && FB [=y] && PCI [=y] && X86 [=y] && AGP_INTEL [=n] && EXPERT [=n] && !DRM_I915 [=y]
- FB_MATROX [=n] && HAS_IOMEM [=y] && FB [=y] && PCI [=y]
- FB_RADEON [=n] && HAS_IOMEM [=y] && FB [=y] && PCI [=y]
- FB_ATY128 [=n] && HAS_IOMEM [=y] && FB [=y] && PCI [=y]
- FB_ATY [=n] && HAS_IOMEM [=y] && FB [=y] && !SPARC32
- FB_S3 [=n] && HAS_IOMEM [=y] && FB [=y] && PCI [=y]
- FB_SAVAGE [=n] && HAS_IOMEM [=y] && FB [=y] && PCI [=y]
- FB_SIS [=n] && HAS_IOMEM [=y] && FB [=y] && PCI [=y]
- FB_VIA [=n] && HAS_IOMEM [=y] && FB [=y] && PCI [=y] && GPIOLIB [=n] && I2C [=y] && (X86 [=y] || COMPILE_TEST [=n])
- FB_NEOMAGIC [=n] && HAS_IOMEM [=y] && FB [=y] && PCI [=y]
- FB_KYRO [=n] && HAS_IOMEM [=y] && FB [=y] && PCI [=y]
- FB_3DFX [=n] && HAS_IOMEM [=y] && FB [=y] && PCI [=y]
- FB_VOODOO1 [=n] && HAS_IOMEM [=y] && FB [=y] && PCI [=y] && FB_DEVICE [=y]
- FB_VT8623 [=n] && HAS_IOMEM [=y] && FB [=y] && PCI [=y]
- FB_TRIDENT [=n] && HAS_IOMEM [=y] && FB [=y] && PCI [=y]
- FB_ARK [=n] && HAS_IOMEM [=y] && FB [=y] && PCI [=y]
- FB_PM3 [=n] && HAS_IOMEM [=y] && FB [=y] && PCI [=y]
- FB_CARMINE [=n] && HAS_IOMEM [=y] && FB [=y] && PCI [=y]
- FB_GEODE_LX [=n] && HAS_IOMEM [=y] && FB [=y] && FB_GEODE [=n]
- FB_GEODE_GX [=n] && HAS_IOMEM [=y] && FB [=y] && FB_GEODE [=n]
- FB_GEODE_GX1 [=n] && HAS_IOMEM [=y] && FB [=y] && FB_GEODE [=n]
- FB_IBM_GXT4500 [=n] && HAS_IOMEM [=y] && FB [=y]
- FB_MB862XX [=n] && HAS_IOMEM [=y] && FB [=y] && (PCI [=y] || OF [=n] && PPC)
- FB_HYPERV [=n] && HAS_IOMEM [=y] && FB [=y] && HYPERV [=n]
- FB_SM712 [=n] && HAS_IOMEM [=y] && FB [=y] && PCI [=y]
- FB_SM750 [=n] && STAGING [=y] && FB [=y] && PCI [=y] && HAS_IOPORT [=y]
|
|
|
| Back to top |
|
|
NeddySeagoon
Administrator
Joined: 05 Jul 2003
Posts: 54298
Location: 56N 3W
|
| Posted: Sun Apr 28, 2024 12:59 pm Post subject: |
|
|
gilhad,
Sorry about that.
We need your dmesg from a broken boot then.
As you say it appears to all work, just no console, there are two ways.
1. log in with ssh.
2. Save dmesg to a file and post it later.
This can be automated with the following
Make file | Code: | | /etc/local.d/save_dmesg.start |
with the following content.
| Code: | #!/bin/bash
dmesg > /dmesg.txt |
| Code: | | chmod +x /etc/local.d/save_dmesg.start |
When the local service starts, all the scripts ending in .start will run.
It will save a file called dmesg.txt in the root filesystem, which makes it easy to find.
Pastebin that file from a liveCD or another boot
_________________
Regards,
NeddySeagoon
Computer users fall into two groups:-
those that do backups
those that have never had a hard drive fail. |
|
| Back to top |
|
|
grknight
Retired Dev
Joined: 20 Feb 2015
Posts: 1680
|
| Posted: Sun Apr 28, 2024 1:24 pm Post subject: |
|
|
| gilhad wrote: | I tried to find VIDEO_NOMODESET in make menuconfig, but it looks like it is not in the menu, but it is just some automatic variable (There is no number to press to get there, there is no menu structure) | Code: | Symbol: VIDEO_NOMODESET [=y]
Type : bool
Defined at drivers/video/Kconfig:24
Selected by [y]:
- DRM [=y] && HAS_IOMEM [=y] && (AGP [=n] || AGP [=n]=n) && !EMULATED_CMPXCHG && HAS_DMA [=y]
|
|
Turn on CONFIG_AGP to make this go away. Most people with graphics want this anyway. |
|
| Back to top |
|
|
gilhad
n00b
Joined: 18 Jan 2007
Posts: 15
|
| Posted: Sun Apr 28, 2024 4:44 pm Post subject: |
|
|
Here is dmesg -T from broken boot.
I have that problematic netbook sit on my table near my main PC. I configured the netbook to start net.eth0 and sshd during boot already, so I can ssh there and compile kernel (or do anything else) on it, even when the LCD is blank.
This is seen on the last line | Code: | | [Sun Apr 28 02:48:35 2024] elogind-daemon[1068]: New session c1 of user root. |
/dev/sda3 is root (/) partition
/dev/sda5 is /home partition
/dev/sda11 is swap currently, suspend/resume is not configured yet.
(It is old netbook so there is lot of partitions full with some old stuff. The 3., 5. and 11. was formated as ext4, ext3 and swap before instalation)
---
Should I try to set the CONFIG_AGP as grknight suggest, or should I keep it intact for your testing session? |
|
| Back to top |
|
|
NeddySeagoon
Administrator
Joined: 05 Jul 2003
Posts: 54298
Location: 56N 3W
|
| Posted: Sun Apr 28, 2024 4:58 pm Post subject: |
|
|
gilhad,
| Code: | | video=inteldrmfb:mode_option=1280x1024-24 |
on the kernel command line cannot work.
It goes wrong
| Code: | [Sun Apr 28 02:47:47 2024] [drm] Initialized gma500 1.0.0 20140314 for 0000:00:02.0 on minor 0
[Sun Apr 28 02:47:47 2024] ath: phy0: ASPM enabled: 0x42
[Sun Apr 28 02:47:47 2024] ath: EEPROM regdomain: 0x6c
[Sun Apr 28 02:47:47 2024] ath: EEPROM indicates we should expect a direct regpair map
[Sun Apr 28 02:47:47 2024] ath: Country alpha2 being used: 00
[Sun Apr 28 02:47:47 2024] ath: Regpair used: 0x6c
[Sun Apr 28 02:47:47 2024] ieee80211 phy0: Selected rate control algorithm 'minstrel_ht'
[Sun Apr 28 02:47:47 2024] ieee80211 phy0: Atheros AR9485 Rev:1 mem=0x(____ptrval____), irq=17
[Sun Apr 28 02:47:47 2024] usb 1-3: new high-speed USB device number 2 using ehci-pci
[Sun Apr 28 02:47:47 2024] acer_wmi: Acer Laptop ACPI-WMI Extras
[Sun Apr 28 02:47:47 2024] acer_wmi: Function bitmap for Communication Button: 0x801
[Sun Apr 28 02:47:48 2024] input: Acer WMI hotkeys as /devices/virtual/input/input18
[Sun Apr 28 02:47:48 2024] fbcon: gma500drmfb (fb0) is primary device
...
[Sun Apr 28 02:47:48 2024] Console: switching to colour frame buffer device 128x37 |
Here, when the console switches to gma500drmfb
There is no missing firmware or dmesg would tell you all about it.
First, remove that video=inteldrmfb .... You are not using the i915 kernel driver. Then reboot.
If that works and its not to your taste, add the entry back with gma500drmfb in place of inteldrmfb so that the settings apply to the driver you are using.
_________________
Regards,
NeddySeagoon
Computer users fall into two groups:-
those that do backups
those that have never had a hard drive fail. |
|
| Back to top |
|
|
gilhad
n00b
Joined: 18 Jan 2007
Posts: 15
|
|
| Back to top |
|
|
NeddySeagoon
Administrator
Joined: 05 Jul 2003
Posts: 54298
Location: 56N 3W
|
| Posted: Sun Apr 28, 2024 6:04 pm Post subject: |
|
|
gilhad.
The one that worked still says that its using the gma500drmfb driver, which is what you want.
| mesg-2024.04.28-19:17-nodrmfb-videook: |
[Sun Apr 28 19:16:22 2024] fbcon: gma500drmfb (fb0) is primary device
...
[Sun Apr 28 19:16:23 2024] Console: switching to colour frame buffer device 128x37
[Sun Apr 28 19:16:23 2024] gma500 0000:00:02.0: [drm] fb0: gma500drmfb frame buffer device |
The bad one, a few minuets later, says the same thing.
| Code: | [Sun Apr 28 19:20:30 2024] fbcon: gma500drmfb (fb0) is primary device
...
[Sun Apr 28 19:20:31 2024] Console: switching to colour frame buffer device 128x37
[Sun Apr 28 19:20:31 2024] gma500 0000:00:02.0: [drm] fb0: gma500drmfb frame buffer device. |
Your diff confirms that the bits that matter are the same.
Go back to booted without nomodeset - it was OK dmesg-2024.04.28-19:17-nodrmfb-videook and have a look at your Xorg.0.log file.
What video driver is Xorg using?
_________________
Regards,
NeddySeagoon
Computer users fall into two groups:-
those that do backups
those that have never had a hard drive fail. |
|
| Back to top |
|
|
gilhad
n00b
Joined: 18 Jan 2007
Posts: 15
|
| Posted: Sun Apr 28, 2024 10:53 pm Post subject: |
|
|
I have no Xorg log from that boot. I boot to text login, and later I may (or may not) start Xorg manually.
In this testing I usually do not start it, as the screen blanks way before first login prompt (somewhere around checking root partition before mounting it read/write), and every time I was able boot without blank screen (boot OK) and tried to startx it worked just fine.
(Also when I started Xorg with blank screen, it started successfully and I was able to hear music - I have music player started in .fluxbox/startup between other widgets for my user account. )
Anyway, I used proven sequence (reboot with nomodeset, login as root, run halt command, after netbook turns off, start it (without nomodeset)), loged in as user, run startx, Xorg started fine.
Here is /home/gilhad/.local/share/xorg/Xorg.0.log
dmesq looks similar to previous OK dmesq with different timing of some opperations, different BogoMIPs and useconds, here it is |
|
| Back to top |
|
|
NeddySeagoon
Administrator
Joined: 05 Jul 2003
Posts: 54298
Location: 56N 3W
|
| Posted: Mon Apr 29, 2024 9:56 am Post subject: |
|
|
gilhad,
| Code: | [ 50.655] (II) modeset(0): First detailed timing is preferred mode
[ 50.655] (II) modeset(0): redX: 0.590 redY: 0.345 greenX: 0.325 greenY: 0.540
[ 50.655] (II) modeset(0): blueX: 0.150 blueY: 0.145 whiteX: 0.313 whiteY: 0.329
[ 50.655] (II) modeset(0): Manufacturer's mask: 0
[ 50.655] (II) modeset(0): Supported detailed timing:
[ 50.655] (II) modeset(0): clock: 49.8 MHz Image Size: 222 x 125 mm
[ 50.655] (II) modeset(0): h_active: 1024 h_sync: 1072 h_sync_end 1104 h_blank_end 1338 h_border: 0
[ 50.655] (II) modeset(0): v_active: 600 v_sync: 602 v_sync_end 608 v_blanking: 620 v_border: 0 |
That tells that Xorg is using the modeset driver, which is what you want, and generating an image of 1024x600 which is the best of all the
| Code: | [ 50.656] (II) modeset(0): Printing probed modes for output LVDS-1
[ 50.656] (II) modeset(0): Modeline "1024x600"x60.0 49.80 1024 1072 1104 1338 600 602 608 620 -hsync -vsync (37.2 kHz eP) |
That suggests that the video mode you tried to select with | Code: | | video=inteldrmfb:mode_option=1280x1024-24 |
at 1280x1024 is not supported by your hardware.
In other words, that's a s good as it gets unless you connect an external display.
_________________
Regards,
NeddySeagoon
Computer users fall into two groups:-
those that do backups
those that have never had a hard drive fail. |
|
| Back to top |
|
|
|
|
You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum
|
|
Kernel & Hardware
