Login

grindstone · 05-21-2024, 11:24 PM

Of course it would... Smile

Wish we knew why. Got any log info on that (dmesg?) If you want to mess-around and see if a lower res option doesn't cough, there's res info here.

Looked at config to verify vesa fb stuff is built-in kernel. This really should "just work" and it's about the simplest possible thing a person could try so we're not seeing something pretty fundamental.

Tried the 4.19 kernels yet? I'd probably try that because it's (comparatively) easy.

Also looks like we can turn on udev logging if we rebuild the initrd but...not ready for that.

You didn't mention if you tried the same/previous (no AutoAddDevices False) method and if it did the same thing--I just assumed it did and you didn't mention it because it was the same. Appreciate the new approach and the logs, BTW.

x18 · 05-22-2024, 07:11 AM

I have to restart my laptop every time cz can't ctr+alt+back (not working) so any dmesg info get replaced and I can't read them.
Var/log/messages is unexisting.
Removed Autoadddevices to use the latest driver but got same result.
Need time to try the old kernel.
Tried Tinycore before Dsl, it starts with Xvesa but that's not enough to run Gnumeric or Xmms. The latter is noisy when users switch among windows Meanwhile.

grindstone · (This post was last modified: 05-22-2024, 01:52 PM by grindstone.)

OK thanks. Hey -- is that all still with the 3 added to the boot string?

Reason for asking is that John's slimski runit /etc/sv/slimski/run file hunts for that. What I'm wondering is if adding the 3 stops the looping and just gets one crash when you do the startx, etc.

Maybe something like: vga=773 3

Edit: Ruled-out modesetting (KMS) as on option for SMI drivers -- it's not even in linus' tree. So we're back to looking at vesa and smi for enough res to run gnumeric, etc. I did a 771 by hand in vbox and it's just not a useful res to run gnumeric in fb--773 is probably the minimum useful display. Haven't tried the rc4 yet, just got it now.

One more thing: If you have (temporarily--save them aside/elsewhere) *nothing* in

/etc/X11/xorg.conf.d/*.conf

and no

/etc/X11/xorg.conf.in

and just reboot w/o the 3, does it just crash with the old previous 'screens found but not enough memory' kind of error? Or even using the 3 & skipping the -configure but using startx script (?)

Thanks for all the patience.

x18 · (This post was last modified: 05-23-2024, 10:33 AM by x18.)

I always boot with the appended 3, else I couldn't set or update anything.
Then I run startx manually.

Text mode works fine, so far I've just seen the ctrl+alt+back seems not to keep set. And the floppy drive starts and never stops if mounted.

Booted with 773 3, moved synaptic.conf and xorg.conf.in, then run startx: "not enough memory".

Code:
[   258.495] 

X.Org X Server 1.21.1.7

X Protocol Version 11, Revision 0

[   258.497] Current Operating System: Linux dsl32 5.10.188-antix.1-486-smp #1 SMP Tue Aug 1 00:54:36 EEST 2023 i686

[   258.497] Kernel command line: BOOT_IMAGE=/boot/vmlinuz-5.10.188-antix.1-486-smp root=UUID=931488a3-14e4-473c-890f-a297f04d9733 ro quiet selinux=0 3 vga=773 3

[   258.499] xorg-server 2:21.1.7-3.0nosystemd4 (https://www.devuan.org/os/community) 

[   258.500] Current version of pixman: 0.42.2

[   258.501]     Before reporting problems, check http://wiki.x.org

    to make sure that you have the latest version.

[   258.501] Markers: (--) probed, (**) from config file, (==) default setting,

    (++) from command line, (!!) notice, (II) informational,

    (WW) warning, (EE) error, (NI) not implemented, (??) unknown.

[   258.505] (==) Log file: "/var/log/Xorg.0.log", Time: Thu May 23 14:15:11 2024

[   258.564] (==) Using system config directory "/usr/share/X11/xorg.conf.d"

[   258.588] (==) No Layout section.  Using the first Screen section.

[   258.589] (==) No screen section available. Using defaults.

[   258.589] (**) |-->Screen "Default Screen Section" (0)

[   258.589] (**) |   |-->Monitor "<default monitor>"

[   258.593] (==) No monitor specified for screen "Default Screen Section".

    Using a default monitor configuration.

[   258.593] (==) Automatically adding devices

[   258.594] (==) Automatically enabling devices

[   258.594] (==) Automatically adding GPU devices

[   258.594] (==) Automatically binding GPU devices

[   258.594] (==) Max clients allowed: 256, resource mask: 0x1fffff

[   258.652] (WW) The directory "/usr/share/fonts/X11/cyrillic" does not exist.

[   258.652]     Entry deleted from font path.

[   258.654] (WW) The directory "/usr/share/fonts/X11/75dpi/" does not exist.

[   258.654]     Entry deleted from font path.

[   258.655] (WW) The directory "/usr/share/fonts/X11/75dpi" does not exist.

[   258.655]     Entry deleted from font path.

[   258.655] (==) FontPath set to:

    /usr/share/fonts/X11/misc,

    /usr/share/fonts/X11/100dpi/:unscaled,

    /usr/share/fonts/X11/Type1,

    /usr/share/fonts/X11/100dpi,

    built-ins

[   258.656] (==) ModulePath set to "/usr/lib/xorg/modules"

[   258.656] (II) The server relies on udev to provide the list of input devices.

    If no devices become available, reconfigure udev or disable AutoAddDevices.

[   258.656] (II) Loader magic: 0x6d5780

[   258.656] (II) Module ABI versions:

[   258.656]     X.Org ANSI C Emulation: 0.4

[   258.656]     X.Org Video Driver: 25.2

[   258.657]     X.Org XInput driver : 24.4

[   258.657]     X.Org Server Extension : 10.0

[   258.657] (++) using VT number 1

[   258.657] (II) seatd_libseat init

[   258.661] (II) [libseat/libseat.c:73] Seat opened with backend 'seatd'

[   258.661] (II) [libseat/backend/seatd.c:212] Enabling seat

[   258.662] (II) seatd_libseat enable

[   258.662] (II) seatd_libseat handled 2 events

[   258.762] (II) seatd_libseat client activated

[   258.786] (--) PCI:*(0@0:2:0) 126f:0910:126f:0910 rev 181, Mem @ 0xfd000000/16777216, BIOS @ 0x????????/131072

[   258.786] (II) LoadModule: "glx"

[   258.808] (II) Loading /usr/lib/xorg/modules/extensions/libglx.so

[   259.028] (II) Module glx: vendor="X.Org Foundation"

[   259.028]     compiled for 1.21.1.7, module version = 1.0.0

[   259.028]     ABI class: X.Org Server Extension, version 10.0

[   259.029] (==) Matched siliconmotion as autoconfigured driver 0

[   259.029] (==) Matched modesetting as autoconfigured driver 1

[   259.029] (==) Matched fbdev as autoconfigured driver 2

[   259.029] (==) Matched vesa as autoconfigured driver 3

[   259.029] (==) Assigned the driver to the xf86ConfigLayout

[   259.030] (II) LoadModule: "siliconmotion"

[   259.031] (II) Loading /usr/lib/xorg/modules/drivers/siliconmotion_drv.so

[   259.059] (II) Module siliconmotion: vendor="X.Org Foundation"

[   259.059]     compiled for 1.21.1.7, module version = 1.7.10

[   259.059]     Module class: X.Org Video Driver

[   259.059]     ABI class: X.Org Video Driver, version 25.2

[   259.060] (II) LoadModule: "modesetting"

[   259.061] (II) Loading /usr/lib/xorg/modules/drivers/modesetting_drv.so

[   259.116] (II) Module modesetting: vendor="X.Org Foundation"

[   259.117]     compiled for 1.21.1.7, module version = 1.21.1

[   259.117]     Module class: X.Org Video Driver

[   259.117]     ABI class: X.Org Video Driver, version 25.2

[   259.117] (II) LoadModule: "fbdev"

[   259.118] (II) Loading /usr/lib/xorg/modules/drivers/fbdev_drv.so

[   259.137] (II) Module fbdev: vendor="X.Org Foundation"

[   259.137]     compiled for 1.21.1.3, module version = 0.5.0

[   259.137]     Module class: X.Org Video Driver

[   259.137]     ABI class: X.Org Video Driver, version 25.2

[   259.138] (II) LoadModule: "vesa"

[   259.139] (II) Loading /usr/lib/xorg/modules/drivers/vesa_drv.so

[   259.148] (II) Module vesa: vendor="X.Org Foundation"

[   259.149]     compiled for 1.21.1.3, module version = 2.5.0

[   259.149]     Module class: X.Org Video Driver

[   259.149]     ABI class: X.Org Video Driver, version 25.2

[   259.149] (II) SMI: driver (version 1.7.10) for Silicon Motion Lynx chipsets: Lynx,

    LynxE, Lynx3D, LynxEM, LynxEM+, Lynx3DM, Cougar3DR, MSOC

[   259.151] (II) modesetting: Driver for Modesetting Kernel Drivers: kms

[   259.151] (II) FBDEV: driver for framebuffer: fbdev

[   259.151] (II) VESA: driver for VESA chipsets: vesa

[   259.153] (WW) Falling back to old probe method for siliconmotion

[   259.153] (--) Assigning device section with no busID to primary device

[   259.154] (--) Chipset Lynx found

[   259.154] (WW) Falling back to old probe method for modesetting

[   259.154] (EE) open /dev/dri/card0: No such file or directory

[   259.155] (WW) Falling back to old probe method for fbdev

[   259.155] (II) Loading sub module "fbdevhw"

[   259.155] (II) LoadModule: "fbdevhw"

[   259.157] (II) Loading /usr/lib/xorg/modules/libfbdevhw.so

[   259.186] (II) Module fbdevhw: vendor="X.Org Foundation"

[   259.186]     compiled for 1.21.1.7, module version = 0.0.2

[   259.187]     ABI class: X.Org Video Driver, version 25.2

[   259.188] (II) Loading sub module "vgahw"

[   259.188] (II) LoadModule: "vgahw"

[   259.190] (II) Loading /usr/lib/xorg/modules/libvgahw.so

[   259.196] (II) Module vgahw: vendor="X.Org Foundation"

[   259.197]     compiled for 1.21.1.7, module version = 0.1.0

[   259.197]     ABI class: X.Org Video Driver, version 25.2

[   259.198] (II) SMI(0): Creating default Display subsection in Screen section

    "Default Screen Section" for depth/fbbpp 24/32

[   259.198] (==) SMI(0): Depth 24, (--) framebuffer bpp 32

[   259.198] (==) SMI(0): RGB weight 888

[   259.198] (==) SMI(0): Default visual is TrueColor

[   259.199] (==) SMI(0): PCI Burst enabled

[   259.199] (==) SMI(0): PCI Retry enabled

[   259.199] (==) SMI(0): Using Hardware Cursor

[   259.199] (II) Loading sub module "int10"

[   259.199] (II) LoadModule: "int10"

[   259.201] (II) Loading /usr/lib/xorg/modules/libint10.so

[   259.223] (II) Module int10: vendor="X.Org Foundation"

[   259.223]     compiled for 1.21.1.7, module version = 1.0.0

[   259.224]     ABI class: X.Org Video Driver, version 25.2

[   259.234] (II) SMI(0): Primary V_BIOS segment is: 0xc000

[   259.235] (II) Loading sub module "vbe"

[   259.235] (II) LoadModule: "vbe"

[   259.237] (II) Loading /usr/lib/xorg/modules/libint10.so

[   259.238] (II) Module int10: vendor="X.Org Foundation"

[   259.238]     compiled for 1.21.1.7, module version = 1.0.0

[   259.238]     ABI class: X.Org Video Driver, version 25.2

[   259.239] (II) SMI(0): VESA BIOS detected

[   259.239] (II) SMI(0): VESA VBE Version 0.0

[   259.239] (II) SMI(0): VESA VBE Total Mem: 4096 kB

[   259.240] (II) SMI(0): VESA VBE OEM: Sÿ

[   259.240] (--) SMI(0): Chipset: "Lynx"

[   259.240] (==) SMI(0): Dual head disabled

[   259.240] (==) SMI(0): Using XAA acceleration architecture

[   259.241] (--) SMI(0): videoram: 4096kB

[   259.243] (II) SMI(0): Cursor Offset: 003FFC00

[   259.244] (II) SMI(0): vgaHWGetIOBase: hwp->IOBase is 0x03d0

[   259.244] (II) SMI(0): Reserved: 003FF800

[   259.245] (II) SMI(0): TFT Panel Size = 1024x768

[   259.245] (II) Loading sub module "i2c"

[   259.245] (II) LoadModule: "i2c"

[   259.245] (II) Module "i2c" already built-in

[   259.246] (II) SMI(0): I2C bus "I2C bus" initialized.

[   259.246] (II) Loading sub module "ddc"

[   259.246] (II) LoadModule: "ddc"

[   259.246] (II) Module "ddc" already built-in

[   259.247] (==) SMI(0): Using gamma correction (1.0, 1.0, 1.0)

[   259.247] (II) SMI(0): MCLK = 80.182

[   259.247] (II) SMI(0): Output LVDS has no monitor section

[   259.247] (II) SMI(0): Printing probed modes for output LVDS

[   259.248] (II) SMI(0): Modeline "(null)"x59.9   63.50  1024 1072 1176 1328  768 771 775 798 -hsync +vsync (47.8 kHz)

[   259.248] (II) SMI(0): Output LVDS connected

[   259.248] (II) SMI(0): Using exact sizes for initial modes

[   259.248] (II) SMI(0): Output LVDS using initial mode (null) +0+0

[   259.248] (EE) SMI(0): Not enough video memory for the configured screen size (1024x1024) and color depth.

[   259.250] (II) UnloadModule: "siliconmotion"

[   259.251] (II) UnloadSubModule: "int10"

[   259.251] (II) Unloading int10

[   259.251] (II) UnloadSubModule: "int10"

[   259.252] (II) Unloading int10

[   259.252] (II) UnloadSubModule: "vgahw"

[   259.252] (II) Unloading vgahw

[   259.253] (EE) Screen(s) found, but none have a usable configuration.

[   259.253] (EE) 

Fatal server error:

[   259.253] (EE) no screens found(EE) 

[   259.254] (EE) 

Please consult the The X.Org Foundation support 

     at http://wiki.x.org

 for help. 

[   259.254] (EE) Please also check the log file at "/var/log/Xorg.0.log" for additional information.

[   259.255] (EE) 

[   259.256] (II) seatd_libseat finish

[   259.311] (EE) Server terminated with error (1). Closing log file.

Found a 2014 report about missing kms driver for SMI:
"PS. xf86-video-modesetting definitely won't work, it requires a KMS driver, and there's no such driver for siliconmotion. For xf86-video-vesa, it seems it'll only work if X is configured to run as root".
https://bbs.archlinux.org/viewtopic.php?id=187477

Found a 2012 report about the buggy SMI driver. It would have hit millions of people now:
https://bugs.freedesktop.org/show_bug.cgi?id=45634

Wonder (desperately) if I can reuse the old Red Hat driver, still got the CD.

Thanks to all of you.

grindstone · (This post was last modified: 05-23-2024, 03:08 PM by grindstone.)

Thanks for the log and for the effort. I found that Arch thread, too, yesterday and I think there's another shot in there.

Is there any option in BIOS to adjust amount of VRAM? Your video has the 16MB. Whenever it's probed in X (--), it's 4M (4096), right? And in Windows it's 4M. If it were NOT probed, we could simply tell X

VideoRam 16384

in the Device section, but because X does probe it, it ignores any VideoRam entry. That makes me think it's what the BIOS says it is.

What's clear is that it's defaulting to using XAA acceleration when the SMI driver is used. If you go read about EXA, it's implemented for siliconmotion.

Option "AccelMethod" "EXA"

That doesn't solve the memory limit problem but it's another thing to try in a very simple config file (w/o running the -configure).

This is actually a good bug to read, too. Origin of 1024x1024 we saw early--it's in an X/Siliconmotion bug thread.
https://bugs.freedesktop.org/show_bug.cgi?id=21528

Between that and that same Arch thread,
https://bbs.archlinux.org/viewtopic.php?id=187477

here's a hand-hacked-up shot:

Code:
Section "ServerLayout"

    Identifier     "Default Layout"

    Screen      0  "Screen0" 0 0

    InputDevice    "Mouse0" "CorePointer"

    InputDevice    "Keyboard0" "CoreKeyboard"

EndSection

Section "Files"

    ModulePath   "/usr/lib/xorg/modules"

    FontPath     "/usr/share/fonts/X11/misc"

    FontPath     "/usr/share/fonts/X11/cyrillic"

    FontPath     "/usr/share/fonts/X11/100dpi/:unscaled"

    FontPath     "/usr/share/fonts/X11/75dpi/:unscaled"

    FontPath     "/usr/share/fonts/X11/Type1"

    FontPath     "/usr/share/fonts/X11/100dpi"

    FontPath     "/usr/share/fonts/X11/75dpi"

EndSection

Section "Module"

    Load  "fbdevhw"

    Load  "freetype"

    Load  "type1"

EndSection

Section "InputDevice"

    Identifier  "Keyboard0"

    Driver      "kbd"

    Option        "XkbModel" "pc105"

    Option        "XkbLayout" "it"

EndSection

Section "InputDevice"

    Identifier  "Mouse0"

    Driver      "mouse"

    Option        "Protocol" "auto"

    Option        "Device" "/dev/input/mice"

    Option        "ZAxisMapping" "4 5 6 7"

    Option        "Emulate3Buttons" "yes"

EndSection

Section "Device"

    Identifier    "Device0"

    Driver    "siliconmotion"

    Option    "AccelMethod"    "EXA"

    Option "HWCursor" "false"

    Option "SWCursor" "true"

    Option      "NoMTRR"

EndSection

Section "Monitor"

    Identifier    "Monitor0"

    HorizSync       31.5-57.5

    VertRefresh     55-80

    Modeline "1024x768"    75    1024 1048 1184 1328   768  771  777  806 -hsync -vsync

EndSection

Section "Screen"

    Identifier    "Screen0"

    Device    "Device0"

    DefaultDepth    16

    Monitor    "Monitor0"

    SubSection "Display"

        Virtual 1024 768

    EndSubSection

EndSection

I took out loading glx which is for network stuff and might (should?) load by default anyway. The pci bus it'll find anyway, etc. See if we can get an Xorg.log outta that. The main try is the EXA vs. XAA but the SW cursor bit is from the Arch thread, too (no 3-D makes sense). Hopefully the BIOS allows an entry exceeding 4M, though, first. Then that might work the way it was.

And honestly, perhaps to elevate it to better quality help, if you burn Plop on a CD and (even USB1.0) find the same exact problem with AntiX 23.1 runit 32-bit, it may result in increased participation at the AntiX forum. OTOH, if that works...we'll have mighty valuable feedback for John.

See what you think.

x18 · 05-24-2024, 02:41 PM

I had already tried the EXA option with no luck.

My BIOS just has the option to enable or disable cache.

See if I 'll found some time to try the old kernel this w.e.

grindstone · 05-24-2024, 02:53 PM

OK thanks--I trust the cache is enabled? Also wondering what kind of no luck? Out of memory or hard crash with backtrace etc? There's the extra virtual bit down that file for smaller rendering (less mem) and the SWcursor in that .conf file, too.

x18 · 05-25-2024, 07:38 PM

Obviously enabled.

Always hard crash but the last try. Virtual didn't work either.

Couldn't install the old kernel, older hardware can't see big flash drives nor external dvd drives plugged on a USB 1 port.
Copied the iso onto a separate partition to boot it from grub but unsuccessfully: "unknown filesystem".

Unless they resize the iso image to CD size, I think it's hard to install.
Its backward compatibility doesn't go backward enough for my laptop.

grindstone · 05-26-2024, 01:19 AM

It's hard to help without error messages ("didn't work", "couldn't install", etc). What else did you do? Did you try changing 16 depth to 8 and virtual to 800x600 if it was a memory problem and not the hard crash. Got ideas, just don't know what's happening.

And a plain old

sudo apt-get install linux-4.19.276-antix.2-486-smp linux-headers-4.19.276-antix.2-486-smp

doesn't work?

And if you still want to get AntiX, that leaves the core or base and a net install of X and all that.

Something has to work.

Nathan · 05-27-2024, 12:44 AM

I don't know if these are of any help, but:
https://www.x.org/releases/current/doc/m...nf.5.xhtml
https://www.x.org/releases/current/doc/m...on.4.xhtml

Have you tried just disabling the acceleration altogether (Option "NoAccel") ?

Login
Username:
Password:	Lost Password?
	Remember me