[SOLVED] Very slow startup of X (AntiX 21)

hazel · 12-02-2022, 03:45 AM

This only happens occasionally but I'd still like to know what causes it.

When X comes up initially, it's just a uniform light grey screen. Then the slim greeter screen comes up to collect login and password. Normally this happens very quickly. But twice now, it got stuck on the blank grey screen for half a minute or so. Is there an incipient problem here?

pan64 · 12-02-2022, 06:17 AM

I guess you need to check what's going on in the background. What processes are running, what about the RAM. Or probably there is a hardware issue somewhere.... Did you read the logs?

hazel · 12-02-2022, 09:24 AM

RAM is very short on that machine, about 1 GB. The only processes at this stage are daemons; I already got rid of those I think I don't need. Haven't looked at the logs yet. AntiX runs X headless so I can't just switch back to tty1 to read live output.

rokytnji · 12-02-2022, 12:47 PM

Owning older gear myself < ide drives >

Just wondering if hard drive is failing. Or maybe just needs a gentle can of air cleaning.
Living in the desert out here. Dust is a problem.

smartctl is in antiX. Here is a older readout from a post in the forum

Code:

root@antix1:/home/roland/Desktop# smartctl --health /dev/sdc1
smartctl 6.6 2017-11-05 r4594 [x86_64-linux-5.10.88-antix.1-amd64-smp] (local
build)
Copyright (C) 2002-17, Bruce Allen, Christian Franke, www.smartmontools.org
=== START OF READ SMART DATA SECTION ===
SMART overall-health self-assessment test result: FAILED!
Drive failure expected in less than 24 hours. SAVE ALL DATA.
Failed Attributes:
ID# ATTRIBUTE_NAME FLAG VALUE WORST THRESH TYPE UPDATED
WHEN_FAILED RAW_VALUE
5 Reallocated_Sector_Ct 0x0033 001 001 036 Pre-fail Always
FAILING_NOW 4284

hazel · 12-03-2022, 04:12 AM

Quote:

Originally Posted by rokytnji

Just wondering if hard drive is failing. Or maybe just needs a gentle can of air cleaning.

You may be right, my friend. For some time now I have been getting occasional disk error messages at startup, not every time but some times.

Today I did an update and everything was normal. No disk errors and no retardation of X. But I shall definitely run your smartctl command and see what happens.

hazel · 12-03-2022, 08:44 AM

So I switched it on again and got another long wait. I'm now going through the Xorg log and there are two gaps:

Code:

29.685] (II) AIGLX: Screen 0 is not DRI2 capable
73.605] (II) IGLX: Loaded and initialised swrast
73.608] (II) GLX: Initialised DRISWRAST GL provider for screen 0 
......
74.501] (II) CHROME(0): Exiting viaFPIOPadSetting.
96.449] (--) CHROME(0): Probing for a VGA monitor on I2C Bus 1

When I next get a normal startup, I'll post the corresponding bits.

PS: just ran smartctl. PASSED. And no disk errors reported on boot.

pan64 · 12-03-2022, 08:45 AM

Quote:

Originally Posted by hazel

You may be right, my friend. For some time now I have been getting occasional disk error messages at startup, not every time but some times.

Today I did an update and everything was normal. No disk errors and no retardation of X. But I shall definitely run your smartctl command and see what happens.

and again, If you had a similar problem, you should find the clues somewhere in the logs. It is definitely a good way to identify "misbehaving" hardware. But [obviously] you can run not only smartctl but any other diagnostic tool (like memchk).

rokytnji · 12-03-2022, 10:48 AM

Good thing smarctl gave your drive a thumbs up. I only mentioned this because on my motorcycle shop wireless IBM M57 desktop running AntiX 21. My boot errors are is a long list of failure to communicate to tpm chip.

Shop stays dusty so I ignore the error. Aint broke down yet. Used mostly for work orders, manuals, search for solution to a motorcycle problem, and jukebox. Also a loaner for folks that hang out.

Here is what I mean by dusty < Losing cause in my case >

https://ibb.co/vczhkrC
https://ibb.co/BBLjjXm

I'm not paranoid about the setup breaking down. I have a older spare free Dell Server from City Hall sitting under the jury table I use in my computer room in the house. If it breaks. I have a spare.

Wondering if dmesg and kernel logs are going to be of use for you. Hardware trouble shooting can be hard. At least for me.
If a power supply croaks, or capacitor cracks open. I replace the box. If harddrive, bios battery, or ram is the culprit. I can fix that.

Since this is one of your main boxes. I would just go through sata and ide cables. Depending on connector used. And re-seating the connections.
Don't cost nothing and is easy to do. Same with ram sticks and power supply connectors. Just on the motherboard. Nothing too complicated.
Reboot and cross fingers. Good luck with it.

hazel · 12-04-2022, 07:39 AM

Quote:

Originally Posted by rokytnji

Since this is one of your main boxes. I would just go through sata and ide cables. Depending on connector used. And re-seating the connections.
Don't cost nothing and is easy to do. Same with ram sticks and power supply connectors. Just on the motherboard. Nothing too complicated.
Reboot and cross fingers. Good luck with it.

I've never opened up a laptop. Wouldn't know how to! But it's hardly a "main box" anyway. It has very little memory (too little even to run FF decently) and I've always regarded it more as a toy than a serious computer. The only important thing I keep on it is the family database, but I've found out how to copy that into Slackware on my desktop machine. I found a build for the gramps software at slackbuilds.org. It's an older version than the one on AntiX, which comes from Debian Bullseye, and it won't read the same internal files, but you can create an empty table and then import the most recent data. So I'm not too worried, just curious.

hazel · 12-10-2022, 09:57 AM

I am now in a position to show you the difference between a slow and a normal X startup. Here is the slow one:

Code:

29.685] (II) AIGLX: Screen 0 is not DRI2 capable
73.605] (II) IGLX: Loaded and initialised swrast
73.608] (II) GLX: Initialised DRISWRAST GL provider for screen 0 
......
74.501] (II) CHROME(0): Exiting viaFPIOPadSetting.
96.449] (--) CHROME(0): Probing for a VGA monitor on I2C Bus 1

And a normal one looks like this:

Code:

30.466] (II) AIGLX: Screen 0 is not DRI2 capable
33.485] (II) IGLX: Loaded and initialised swrast
33.485] (II) GLX: Initrialized DRISWRAST GL provider for screen 0
...........
34.370  (II) CHROME(0): Exiting viaFPIOPadSetting.
58.582  (--) CHROME(0): Probing for a VGA monitor on I2C Bus 1

Clearly then the first (GLX) section is where the problem occurs when it occurs at all. The 12-14 second pause during the second section is normal.

Two further observations that may be of interest.
A) X always comes up in 3 stages:
1) a light grey speckled screen that looks like the visual equivalent of electronic noise
2) a dark grey smoothly shaded screen
3) the same with login boxes.
If there is a hold-up, it invariably occurs between 1) and 2). I believe that 2) is already a GLX rendered image. The lighter screen at 1) looks to me more like direct output to the video card.

B) The openchrome driver used in AntiX-21 is definitely not the same one as in AntiX-19. The latter would not work at all unless the kernel was set to boot with "iomem=relaxed". This one does not need special boot parameters. So something that was wrong has been fixed, and these occasional blips are probably the results of the fix.

rokytnji · 12-10-2022, 08:37 PM

Just food for thought. xorg and open chrome drivers have to be compatible .

Which is probably probably why you noticed antiX 19 worked better than 21 or 22.

When I had my via issues. It was because the Open Chrome driver developer had not updated his drivers to match the newer xorg.
He handles via also. I had to run vesa on my via chip.

Just wondering if taking out splash might fix your issue

Code:

GRUB_CMDLINE_LINUX_DEFAULT="quiet/splash/vga=current/loglevel=3/gfxmode=text"

I thought we were talking Desktop. Not laptop. My mistake.

pan64 · 12-11-2022, 04:32 AM

I think there should be something else in the logs (syslog, dmesg, ??) before that line (in time) which will be definitely different in the two cases.

hazel · 12-11-2022, 11:28 AM

Quote:

Originally Posted by pan64

I think there should be something else in the logs (syslog, dmesg, ??) before that line (in time) which will be definitely different in the two cases.

How would I correlate them? The Xorg logs use an elapsed time (presumably from the launch of Xorg itself) whereas syslog and messages use clock time.

pan64 · 12-12-2022, 01:26 AM

Quote:

Originally Posted by hazel

How would I correlate them? The Xorg logs use an elapsed time (presumably from the launch of Xorg itself) whereas syslog and messages use clock time.

yes, not an easy case. Remove time related info and compare. You may find probably one of them will be longer or just contains a different line. Personally I find useful the tool meld (to compare files).

hazel · 01-01-2023, 08:36 AM

I don't know if my present graphics problems are an extension of the earlier ones described here but I thought I would continue this thread in case it's all one big problem.

Now I can no longer get X to load at all via the display manager slimski. There is a slimski log but it is really a bad joke with hardly any information in it.

Code:

envlang received as: en_GB.UTF-8
slimski: waiting for X server to begin accepting connections

So I tried using startx. Unfortunately I can't get this to work as an unprivileged user, though I am a member of both the video and the tty groups, and on most distros that would be enough. On AntiX I get:

Code:

xf86EnableIOPorts: failed to set IOPL for I/O (Operation not permitted)
(EE) Fatal server error

and in the Xorg log:

Code:

[   166.637] (II) CHROME(0): Entered viaMapMMIO.
[   166.637] (--) CHROME(0): Mapping MMIO at address 0xFC000000 with
size 52 KB.
[   166.637] (EE) CHROME(0): Unable to map MMIO.
Error: Permission denied (13)
[   166.637] (II) CHROME(0): Exiting viaMapMMIO.

I can do this as root and it goes all the way to a graphical screen, so it seems the video driver still works. Curiously the log gets written in /var/log and not in /root/.local/share/xorg. Of course it's bad practice to run a desktop as root but I don't do any browsing in Littleboy. It only has 1 GB of RAM and will barely run FF. What I do use it for is my genealogical program gramps and also as a dump location for Bigboy, my main machine. But really I'm getting very tired of its eccentricities.

What do I have to do to get this to work? And, more important, why has it suddenly stopped working? It was fine a week ago apart from the occasional slowdowns. I have checked the previous update and it contained nothing relevant to X. Is the inability to use startx as a normal user due to some kernel setting that I could change using sysctl or a boot line argument?