OESF Portal | OESF Forum | OESF Wiki | LinuxPDA | #planetgemini chat on matrix.org | #gemini-pda chat on Freenode | #zaurus and #alarmz chat on Freenode | ELSI (coming soon) | Ibiblio

IPB

Welcome Guest ( Log In | Register )

 
Reply to this topicStart new topic
> unable to boot debian after crash
Jeffrey Mark Sis...
post Sep 23 2018, 06:29 AM
Post #1





Group: Members
Posts: 10
Joined: 11-July 18
Member No.: 825,932



Gemini Debian is not able to reliably poweroff. (I'll explain this more below.) It often crashes when trying to poweroff. When this happens, it often is not able to reboot. It just hangs in the splash screen and doesn't display "Linux".

Through various analyses, I have come to believe that the cause of the problem is that a reboot after a crash runs fsck and that often requires user input. AFAIK, Gemini has no equivalent of the grub and VT screens from Debian on amd64 that allow answering fsck queries or booting into a recovery mode.

In my case this is exacerbated by the fact that I have a 512GB SDXC card formatted as one large ext4 partition that I mount with an entry in /etc/fstab. I have discovered that this is likely the cause of the failure to reboot because when that happens, I reinstall Debian into the main flash device, reboot without the /etc/fstab entry, and manually run fsck on the SDXC card, discovering the errors.

This issue has required me to reinstall Debian from scratch 4 times now.

My two questions:

1. Is there an analog of grub booting to a recovery mode or booting with a VT so that I can answer fsck queries?
2. How can I reliably poweroff so this doesn't happen?

I mount the SDXC card as /fs. I have /home/qobi redirected to /fs/home/qobi. And when logged in as qobi, I run twm through .xsession instead of lxqt. I could try to reduce the likelihood of this occurring by eliminating the /etc/fstab entry but this would require me, every time I poweron, to login as gemini to lxqt, mount /fs, logout from gemini, and then login as qobi. And it would require me, every time I poweroff, to logout from qobi, login as gemini to lxqt, umount /fs, then try to poweroff, suspend, or hibernate. This would be tedious.

I poweroff, suspend, and hibernate by running the lxqt variants within my environment. Sometimes some of them work, sometimes not. Even when I use them from within lxqt logged in as gemini, sometimes they work, sometimes not.

It takes me 4 hours to reinstall and reconfigure Debian the way I like it from scratch. I am currently in an unbootable state so an answer to (1) above can help save me 4 hours.
Go to the top of the page
 
+Quote Post
mithrandir
post Sep 23 2018, 12:44 PM
Post #2





Group: Members
Posts: 121
Joined: 7-January 18
Member No.: 815,997



I have not tried this out myself, but maybe the UART cable described in https://www.oesf.org/forum/index.php?showtopic=35286 can be of some help. Probably you can get console access with it.
Mith
Go to the top of the page
 
+Quote Post
jakfish
post Sep 24 2018, 02:01 PM
Post #3





Group: Members
Posts: 270
Joined: 12-September 10
Member No.: 91,772



@ Jeffrey Mark Siskind: while I do not share your reboot issues, I can confirm your shut-down problems on my own Debian setup (which is factory default). So your sd and fstab problems may be the ultimate quandary.

But your shut-down problems are not just yours.

Sorry for your hassles,
Jake
Go to the top of the page
 
+Quote Post
n9dreamer
post Mar 3 2019, 11:42 AM
Post #4





Group: Members
Posts: 4
Joined: 17-February 18
Member No.: 817,949



I have a similar problem to that mentioned here. I shutdown Debian and on reboot get as far as the Planet / Tux / Linux screen but with perpetual green and blue alternating LEDs. I don't know how to resolve other than to reflash? I'm not familiar enough with UART / console methods to attempt that. Any suggestions gratefully received smile.gif Is there any hope for new workarounds with TP3?
Go to the top of the page
 
+Quote Post
Adam Boardman
post Mar 3 2019, 02:14 PM
Post #5





Group: Members
Posts: 173
Joined: 29-December 17
Member No.: 815,489



QUOTE(n9dreamer @ Mar 3 2019, 08:42 PM) *
I shutdown Debian and on reboot get as far as the Planet / Tux / Linux screen but with perpetual green and blue alternating LEDs. I don't know how to resolve other than to reflash?

If you get the green led in the cycle then that means you have wifi on and should have the ability to ssh in? (if it wasn't on then USB login as per gemian wiki Debugging page would be the alternative).

If the device boots as far as adding the words 'Linux' to the splash screen you probably want to start by looking at /var/log/Xorg.0.log for hints as to what has gone wrong, the most common thing is a confused X11 setup.

If it is a fsck problem with a large memory card you could consider disabling the check on boot - 0 in last column of fstab and change the check on mounts count to -1 with tune2fs etc. Up to you whether your willing to take the possible risk to your data from not checking the file system though.
Go to the top of the page
 
+Quote Post
n9dreamer
post Mar 5 2019, 01:02 AM
Post #6





Group: Members
Posts: 4
Joined: 17-February 18
Member No.: 817,949



Thanks for the advice - will give it a go smile.gif
Go to the top of the page
 
+Quote Post
n9dreamer
post Mar 8 2019, 02:07 PM
Post #7





Group: Members
Posts: 4
Joined: 17-February 18
Member No.: 817,949



ssh'd in via wifi and /var/log/Xorg.0.log says it's 'Failed to load module hwcomposer' and then a message to say it can't open a display.

I haven't knowingly tweaked the X setup but have installed Qt5 which may have some EGL dependencies?

[My gemini has run out of juice at present and I have no way to get more verbose output across but will add in when possible]
Go to the top of the page
 
+Quote Post
TheKit
post Mar 9 2019, 08:31 AM
Post #8





Group: Members
Posts: 27
Joined: 19-February 18
Member No.: 818,021



QUOTE(n9dreamer @ Mar 9 2019, 01:07 AM) *
ssh'd in via wifi and /var/log/Xorg.0.log says it's 'Failed to load module hwcomposer' and then a message to say it can't open a display.

I haven't knowingly tweaked the X setup but have installed Qt5 which may have some EGL dependencies?

[My gemini has run out of juice at present and I have no way to get more verbose output across but will add in when possible]

I'm interested in full Xorg.0.log when you're able to. Also please post the output of test_egl_configs command from libhybris-tests package.
Go to the top of the page
 
+Quote Post
n9dreamer
post Mar 10 2019, 06:03 AM
Post #9





Group: Members
Posts: 4
Joined: 17-February 18
Member No.: 817,949



QUOTE(TheKit @ Mar 9 2019, 08:31 AM) *
I'm interested in full Xorg.0.log when you're able to. Also please post the output of test_egl_configs command from libhybris-tests package.


Hi,

Output from test_egl_configs:

CODE
Starting test (EGL_PLATFORM=x11)
library "libwlparser.so" not found
EGL Version 1.4
Extensions: EGL_KHR_get_all_proc_addresses EGL_ANDROID_presentation_time EGL_KHR_swap_buffers_with_damage EGL_ANDROID_create_native_client_buffer EGL_ANDROID_front_buffer_auto_refresh EGL_KHR_image EGL_KHR_image_base EGL_KHR_gl_texture_2D_image EGL_KHR_gl_texture_cubemap_image EGL_KHR_gl_renderbuffer_image EGL_KHR_fence_sync EGL_KHR_create_context EGL_KHR_config_attribs EGL_KHR_surfaceless_context EGL_EXT_create_context_robustness EGL_ANDROID_image_native_buffer EGL_KHR_wait_sync EGL_ANDROID_recordable EGL_KHR_partial_update EGL_HYBRIS_native_buffer2 EGL_HYBRIS_WL_acquire_native_buffer EGL_WL_bind_wayland_display  EGL_EXT_swap_buffers_with_damage EGL_WL_create_x11_buffer_from_image
Available configurations: 22
===== Configuration #0 =====
  EGL_RED_SIZE: 8
  EGL_GREEN_SIZE: 8
  EGL_BLUE_SIZE: 8
  EGL_ALPHA_SIZE: 8
  EGL_BUFFER_SIZE: 32
  EGL_DEPTH_SIZE: 0
  EGL_STENCIL_SIZE: 0
  EGL_CONFIG_ID: 1
  EGL_LEVEL: 0
  EGL_SAMPLE_BUFFERS: 0
  EGL_SAMPLES: 0
  EGL_MAX_PBUFFER_WIDTH: 8192
  EGL_MAX_PBUFFER_HEIGHT: 8192
  EGL_MAX_PBUFFER_PIXELS: 67108864
  EGL_MIN_SWAP_INTERVAL: 0
  EGL_MAX_SWAP_INTERVAL: 1
Segmentation fault (core dumped)


I don't know how to attach a file here so posting the text (sorry for hijacking the thread unsure.gif :

CODE
[ 14081.033] (--) Log file renamed from "/var/log/Xorg.pid-21033.log" to "/var/log/Xorg.0.log"
[ 14081.081]
X.Org X Server 1.19.2
Release Date: 2017-03-02
[ 14081.082] X Protocol Version 11, Revision 0
[ 14081.082] Build Operating System: Linux 4.9.0-8-arm64 aarch64 Debian
[ 14081.082] Current Operating System: Linux localhost 3.18.41+ #7 SMP PREEMPT Wed May 30 16:45:45 MSK 2018 aarch64
[ 14081.083] Kernel command line: console=tty0 console=ttyMT0,921600n1 root=/dev/ram vmalloc=496M slub_max_order=0 slub_debug=OFZPU androidboot.hardware=mt6797 maxcpus=5 androidboot.verifiedbootstate=green bootopt=64S3,32N2,64N2 printk.disable_uart=1 bootprof.pl_t=4932 bootprof.lk_t=2894 boot_reason=4 androidboot.serialno=AI4HHYSSO78PAUSO androidboot.bootreason=wdt_by_pass_pwk gpt=1 usb2jtag_mode=0
[ 14081.084] Build Date: 03 November 2018  03:03:37AM
[ 14081.084] xorg-server 2:1.19.2-1+deb9u5 (https://www.debian.org/support)
[ 14081.084] Current version of pixman: 0.34.0
[ 14081.085]     Before reporting problems, check http://wiki.x.org
    to make sure that you have the latest version.
[ 14081.085] Markers: (--) probed, (**) from config file, (==) default setting,
    (++) from command line, (!!) notice, (II) informational,
    (WW) warning, (EE) error, (NI) not implemented, (??) unknown.
[ 14081.087] (==) Log file: "/var/log/Xorg.0.log", Time: Sat Mar  2 00:26:28 2019
[ 14081.088] (==) Using config file: "/etc/X11/xorg.conf"
[ 14081.088] (==) Using system config directory "/usr/share/X11/xorg.conf.d"
[ 14081.096] (==) No Layout section.  Using the first Screen section.
[ 14081.097] (**) |-->Screen "Screen0" (0)
[ 14081.097] (**) |   |-->Monitor "Monitor0"
[ 14081.102] (**) |   |-->Device "MediaTek HWC"
[ 14081.103] (**) Option "OffTime" "1"
[ 14081.103] (==) Automatically adding devices
[ 14081.103] (==) Automatically enabling devices
[ 14081.104] (==) Automatically adding GPU devices
[ 14081.104] (==) Max clients allowed: 256, resource mask: 0x1fffff
[ 14081.104] (WW) The directory "/usr/share/fonts/X11/cyrillic" does not exist.
[ 14081.104]     Entry deleted from font path.
[ 14081.105] (==) FontPath set to:
    /usr/share/fonts/X11/misc,
    /usr/share/fonts/X11/100dpi/:unscaled,
    /usr/share/fonts/X11/75dpi/:unscaled,
    /usr/share/fonts/X11/Type1,
    /usr/share/fonts/X11/100dpi,
    /usr/share/fonts/X11/75dpi,
    built-ins
[ 14081.105] (==) ModulePath set to "/usr/lib/xorg/modules"
[ 14081.105] (II) The server relies on udev to provide the list of input devices.
    If no devices become available, reconfigure udev or disable AutoAddDevices.
[ 14081.106] (II) Loader magic: 0x7e5a48cdc8
[ 14081.106] (II) Module ABI versions:
[ 14081.106]     X.Org ANSI C Emulation: 0.4
[ 14081.106]     X.Org Video Driver: 23.0
[ 14081.106]     X.Org XInput driver : 24.1
[ 14081.107]     X.Org Server Extension : 10.0
[ 14081.112] (++) using VT number 7

[ 14081.113] (II) systemd-logind: logind integration requires -keeptty and -keeptty was not provided, disabling logind integration
[ 14081.113] (II) no primary bus or device found
[ 14081.113] (II) LoadModule: "glx"
[ 14081.114] (II) Loading /usr/lib/xorg/modules/extensions/libglx.so
[ 14081.119] (II) Module glx: vendor="X.Org Foundation"
[ 14081.119]     compiled for 1.19.2, module version = 1.0.0
[ 14081.119]     ABI class: X.Org Server Extension, version 10.0
[ 14081.119] (II) LoadModule: "hwcomposer"
[ 14081.121] (WW) Warning, couldn't open module hwcomposer
[ 14081.121] (II) UnloadModule: "hwcomposer"
[ 14081.121] (II) Unloading hwcomposer
[ 14081.121] (EE) Failed to load module "hwcomposer" (module does not exist, 0)
[ 14081.121] (EE) No drivers available.
[ 14081.121] (EE)
Fatal server error:
[ 14081.121] (EE) no screens found(EE)
[ 14081.121] (EE)
Please consult the The X.Org Foundation support
     at http://wiki.x.org
for help.
[ 14081.121] (EE) Please also check the log file at "/var/log/Xorg.0.log" for additional information.
[ 14081.121] (EE)
[ 14081.122] (EE) Server terminated with error (1). Closing log file.
Go to the top of the page
 
+Quote Post

Reply to this topicStart new topic
2 User(s) are reading this topic (2 Guests and 0 Anonymous Users)
0 Members:

 



RSS Lo-Fi Version Time is now: 17th November 2019 - 11:39 PM