[lug] whole screen (X) flashes

D. Stimits stimits at attbi.com
Sat Aug 17 19:54:41 MDT 2002


John Karns wrote:
> On Sat, 17 Aug 2002, D. Stimits said:
> 
> 
>>By any chance is the network card or modem on the same interrupt as the
>>video card? What else shares the irq of the video card? Perhaps you have
>>hardware competing.
> 
> 
> Not AFAICT, although the NIC is sharing irq 10 with the USB controller,
> the video is on irq 11.
> 
>>From /proc/interrupts,
> 
> 10:      48587          XT-PIC  i82365, eth0, usb-uhci
> 
> Another oddity: the NC driver is running as a loadable module, as is
> USB.  Usb and eth0 show up in the proc irq list, but the video irq is
> conspicuously absent (no irq 11).  <scratches head>
> 
>            CPU0
>   0:     625848          XT-PIC  timer
>   1:       9211          XT-PIC  keyboard
>   2:          0          XT-PIC  cascade
>   5:      12178          XT-PIC  Maestro3(i)
>   8:     820144          XT-PIC  rtc
>  10:      51699          XT-PIC  i82365, eth0, usb-uhci
>  12:          4          XT-PIC  PS/2 Mouse
>  13:          1          XT-PIC  fpu
>  14:      44849          XT-PIC  ide0
> 
> It *does* show up however, in /proc/pci
> 
> Bus  1, device   0, function  0:
>   VGA compatible controller: ATI Unknown device (rev 0).
>     Vendor id=1002. Device id=4d46.
>     Medium devsel.  Fast back-to-back capable.  IRQ 11.  Master Capable.
>     Prefetchable 32 bit memory at 0xe8000000 [0xe8000008].
>     I/O at 0xcc00 [0xcc01].
>     Non-prefetchable 32 bit memory at 0xfcffc000 [0xfcffc000].
> 
> My intuition is that the phenomenon is triggered by certain sequences of
> display attribute control characters generated from the html browser.  When
> it happens with lynx, if I scroll up or down far enough to get the offending
> attributes out of the view window, the flashing stops.
> 

Hmm...I recall a windows bug where certain international character set 
characters or malformed character strings would cause a failure. In fact 
some denial of service type attacks based themselves on it. This would 
not have an effect on linux though, so if your problem is both linux and 
windows, I have to go back and wonder about hardware issues, like the 
strange lack of irq 11 in /proc/interrupts. But if this shows up under 
only windows, you can probably go to updates.microsoft.com and get an 
update to fix it. If something is failing to show up in /proc/interrupts 
under linux, I would have to wonder if perhaps a bios setting for agp or 
video in general is causing this.

D. Stimits, stimits AT attbi.com




More information about the LUG mailing list