[vox-tech] monitor troubles

[Jennifer Stickel]

Thanks for all your help.  I decided that I am just going to find a 
cheap video card and forget this integrated one.  It is not worth the 
hassle that it is taking.

Jennifer


Rick Moen wrote:

>[Kudos to Pete for calling your attention to /dev/agpgart.]
>
>Quoting Jennifer Stickel (jrstickel@ucdavis.edu):
>
>  
>
>>It looks like this might be my problem.  I checked and there is 64MB of
>>video RAM for this card..., 
>>    
>>
>
>Hold that thought.  We'll get back to it.
>
>  
>
>>but this is what I am getting in my XF86config.0.log file
>><big snip>
>>    
>>
>
>  
>
>>(II) I810(0): VESA VBE OEM: Intel(r)865G Graphics Chip Accelerated VGA
>>    
>>
>BIOS
>  
>
>>(II) I810(0): VESA VBE OEM Software Rev: 1.0
>>(II) I810(0): VESA VBE OEM Vendor: Intel Corporation
>>(II) I810(0): VESA VBE OEM Product: Intel(r)865G Graphics Controller
>>(II) I810(0): VESA VBE OEM Product Rev: Hardware Version 0.0
>>(II) I810(0): Integrated Graphics Chipset: Intel(R) 865G
>>(--) I810(0): Chipset: "865G"
>>(--) I810(0): Linear framebuffer at 0xE8000000
>>(--) I810(0): IO registers at addr 0xFEB80000
>>(II) I810(0): detected 892 kB stolen memory.
>>(EE) GARTInit: Unable to open /dev/agpgart (No such file or directory)
>>(WW) I810(0): /dev/agpgart is either not available, or no memory is available
>>for allocation.  Using pre-allocated memory only.
>>(WW) I810(0): VideoRAM reduced to 892 kByte (limited to available sysmem)
>>(--) I810(0): Pre-allocated VideoRAM: 892 kByte
>>(--) I810(0): VideoRAM: 892 kByte
>>    
>>
>
>
>  
>
>>I did build it as a module into the kernel and modprobe -a agpgart gives
>>
>>WARNING: Module agpgart already in kernel
>>
>>I also tried modprobe intel-agp
>>
>>This one eliminates the error message and by specifying the amount of 
>>    
>>
>                                               ^^^^^^^^^^^^^^^^^^^^^^^^
>  
>
>>video RAM in XF86config I don't get any error messages in the 
>>    
>>
>  ^^^^^^^^^^^^^^^^^^^^^^^
>  
>
>>XF86config.0.log file.  But I do get these lines
>>    
>>
>
>Hold that thought, too.  I'll be getting to my point in a minute.
>
>  
>
>>(II) I810(0): Not using mode "1024x768" (no mode of this name)
>>(II) I810(0): Increasing the scanline pitch to allow tiling mode (640 -> 
>>1024).
>>(--) I810(0): Virtual size is 640x480 (pitch 1024)
>>(**) I810(0):  Built-in mode "640x480"
>>(II) I810(0): Attempting to use 73Hz refresh for mode "640x480" (41)
>>(--) I810(0): Display dimensions: (300, 220) mm
>>(--) I810(0): DPI set to (54, 55)
>>    
>>
>
>
>OK.  Your Intel 865G is one of a series of motherboard-integrated video
>chipsets produced by Intel, over the last few years (i810, i830, i830M,
>i845, i865, i865G, i830M, i845G, i852GM, i855G, i852GM, i855GM).  All of
>those chipsets actually have _zero_ video memory.  Zip.  Nada.  Instead,
>they dip into main system RAM and use that, by way of some weird system
>facility called "GART" that's part of the AGP bus -- thus the
>/dev/agpgart device file, which is somehow used for that purpose.
>
>The kernel you're running needs to include _both_ generic agpgart
>support and the particular video chip's matching driver -- in addition
>to using a modern enough XFree86 server to be able to usefully talk to
>the kernel layers.
>
>_Before_ you "specified the amount of video RAM in XF86Config", the
>XFree86 logfile indicated that, because of problems addressing
>/dev/agpgart, only 892 kB of RAM could be used.  That indicates a
>fundamental kernel (or device file) problem.  Therefore, forcing XFree86 
>to think there's a larger amount of video RAM isn't useful.  You need to
>figure out the kernel problem (well, or at least make it go away  ;->  ).
>
>The easier portion of that investigative path involves verifying that
>/dev/agpgart exists and has sensible ownership / permissions.  Something
>like:
>
>crw-rw----    1 root     video     10, 175 2001-05-30 06:08 agpgart
>
>But remember, a device file is merely a file-like interface to a
>software or hardware structure.  The device file being present in no way
>guarantees that talking to it does anything useful.  That's the larger
>portion of the problem, and very likely the best way to address it is to
>get source code for a really recent kernel version, and craft your own
>kernel to make absolutely sure that you have all necessary kernel
>support for your video card.
>
>If that sounds like way too much trouble, I agree.  That's why I
>personally am leery of motherboard-integrated video, and would in your
>shoes go find a cheap used Matrox G400 (or such) instead:  The former's
>trait of stealing system RAM strikes me as cheesy, and the requirement
>for special kernel-level driver support seems like (if nothing else) way
>too much hassle just for generic video.
>
>  
>