startx locks up with red block with green blinking cursor

[TRON21Net]

I just recently upgraded my power supply to 520 watt and now my machine no longer crashs with random strings of ASCii text. But now whenever I start X11 I get this red block at the top left corner of my screen with a green cursor blinking, but I cannot do anything since the mouse is locked out and so is the keyboard.

Here are my specs:

CPU: AMD Athlon XP 1700+ (266Mhz FSB), 2.75volt
Board: Abit NF7 v1.1 (all hardware enabled: USB, audio, etc..)
Power: 520 watt TTGI TT-520SS
RAM: 512MB PC2700 333MHz, 2.5volt
Video: PNY GeForce4 Ti 4200 64MB, 4X mode
Misc.: 2 PCI cards(firewire, ethernet), 3 HDDs(6GB, 20GB, 40GB), Bus Link DVD-ROM Drive.
OS: Redhat 8 Linux w/ kernel: 2.4.18-24.8.0


At the end of my XFree86 log file as I've attached to this forum it has this at the end:

(II) NVIDIA(0): Not using default mode "960x600" (hsync out of range)
(**) NVIDIA(0): Validated modes for display device CRT-1:
(**) NVIDIA(0): Default mode "1024x768": 78.8 MHz, 60.1 kHz, 75.1 Hz
(**) NVIDIA(0): Default mode "800x600": 56.3 MHz, 53.7 kHz, 85.1 Hz
(**) NVIDIA(0): Default mode "640x480": 36.0 MHz, 43.3 kHz, 85.0 Hz
(II) NVIDIA(0): Virtual screen size determined to be 1024 x 768
(--) NVIDIA(0): Display dimensions: (280, 220) mm
(--) NVIDIA(0): DPI set to (92, 88)
(II) Loading sub module "fb"
(II) LoadModule: "fb"
(II) Loading /usr/X11R6/lib/modules/libfb.a

Looks like loading libfb.a file is causing the crash, but I've tried messing around with the XFree86.conf file to stop it from loading, but I guess it's something required that I cannot disable.

Also I've noticed that I cannot use a higher screen depth than 16bit, so I'm stuck at 16bit or else X11 quits during boot and says my hardware does not support 32bit or 24bit depth modes. :(

[Andy Mecham]

Also I've noticed that I cannot use a higher screen depth than 16bit, so I'm stuck at 16bit or else X11 quits during boot and says my hardware does not support 32bit or 24bit depth modes. :(

32bit is expected - there's no such thing as 32bit under X.

Can you post an X log (preferably verbose - 'startx -- -logverbose 5') showing the 24bit message?

Thanks!

--andy

[TRON21Net]

Here is my X11 log file as attached to this thread in verbose 5 logging mode.



Can you post an X log (preferably verbose - 'startx -- -logverbose 5') showing the 24bit message?

Sure, here is the 24bit message and related:

(**) NVIDIA(0): Depth 24, (--) framebuffer bpp 32
(==) NVIDIA(0): RGB weight 888
(==) NVIDIA(0): Default visual is TrueColor
(==) NVIDIA(0): Using gamma correction (1.0, 1.0, 1.0)
(==) NVIDIA(0): Using HW cursor
(==) NVIDIA(0): video key set to default value of 0x101fe
(--) NVIDIA(0): Linear framebuffer at 0xD4000000
(--) NVIDIA(0): MMIO registers at 0xDC000000
(--) NVIDIA(0): Found 2 CRTCs on board
(II) NVIDIA(0): Supported display device(s): CRT-0, CRT-1, DFP-0, TV-0
(II) NVIDIA(0): Boot display device(s): CRT-1
(II) NVIDIA(0): NVIDIA GPU detected as: GeForce4 Ti 4200
(II) NVIDIA(0): Chip Architecture: 0x20
(II) NVIDIA(0): Chip Implementation: 0x25
(II) NVIDIA(0): Chip Revision: 0xa3
(--) NVIDIA(0): VideoRAM: 65536 kBytes
(II) NVIDIA(0): Connected display device(s): CRT-1
(--) NVIDIA(0): Detected TV Encoder: Philips 7104
(II) NVIDIA(0): Enabled display device(s): CRT-1
(II) NVIDIA(0): Mapping display device 0 (CRT-1) to CRTC 0
(--) NVIDIA(0): Display device CRT-1: maximum pixel clock at 8 bpp: 350 MHz
(--) NVIDIA(0): Display device CRT-1: maximum pixel clock at 16 bpp: 350 MHz
(--) NVIDIA(0): Display device CRT-1: maximum pixel clock at 32 bpp: 350 MHz

[TRON21Net]

*bump*

[bwkaz]

Your "24-bit message" doesn't contain an error, and the log file you've posted doesn't either (it's just an attempt with 16-bit depth, which doesn't really give any information). Can you post the log file from an attempt at 24-bit depth?

[TRON21Net]

hmmm, that is set at 24 bit! It even says so in the log file.

It always stops and locks up at:

(II) LoadModule: "fb"
(II) Loading /usr/X11R6/lib/modules/libfb.a

[bwkaz]

Oops, hehe, I've got too many peoples' log files laying around in my home directory. Sorry about that, I was looking at a different one.

:o

Does a strings /usr/X11R6/lib/modules/extenstions/libfb.a | grep Project (case sensitive) print anything?

The next line in my logs after the last one that you have, is "(II) Module fb: vendor="The XFree86 Project"", followed by "compiled for 4.3.0, module version = 1.0.0" (my version of X is 4.3.0). So check what strings tells you about the libfb.a file, to see if there is even a vendor string in it, and/or what the version is.

What happens if you get rid of fbdevhw, record, and ddc? (if those are all modules -- I think they are)

[TRON21Net]

Well I did this and got:


[tron@linuxbrw tron]$ strings /usr/X11R6/lib/modules/extenstions/libfb.a | grep Project
strings: /usr/X11R6/lib/modules/extenstions/libfb.a: No such file or directory


So I thought maybe you're talking about the actual path of libfb.a for my machine since it might be in different places for different distro's, so I did and got:


[tron@linuxbrw tron]$ strings /usr/X11R6/lib/modules/libfb.a | grep Project
The XFree86 Project



I've already had fbdevhw module commented out in the XFree86 config file, but apparently libfb.a isn't apart of that. Look at this part that is in bold:


(II) Loading sub module "fb"
(II) LoadModule: "fb"
(II) Loading /usr/X11R6/lib/modules/libfb.a


It's stating that libfb.a is a sub module of another module, but it doesn't say of which module it is a sub of. So maybe it's a sub module of the nvidia driver? Andy you have any thoughts on this?

[TRON21Net]

*bump*

[bwkaz]

I don't know. It might be a submodule of the nVidia driver, but ... I'm not sure. The nvidia_drv.o is the last thing in my logs that was loaded before fb, so that supports that idea.

Plus, I apparently misspelled extensions. I put an extra t in it... hehe, oops.

If you get rid of record and ddc, does that help at all?

[Andy Mecham]

It's not clear how loading libfb.a could cause your machine to hang.

A few possibilities:

1) that library is corrupted in some way, and is doing something very bad.

2) that message is a red herring - there's a problem somewhere else, and that line just happens to be where the log stops.

I'm looking into this here... i'm sorry for being slow, but it's been a bit of a busy week. :)

--andy

[MostlyCompetent]

This is very similiar to my problem, with the random ASCII Chars all over the screen and the blinking cursor. Is this somehow related to my power supply? Or the fact that we both have Athlon XPs?

To Andy: I'm at school right now, I'll try the drivers again and get the failing log for you. (When I get home)