We had some problems with video capture to frame buffer as well and it
turned out that having two devices attempting to write on the PCI bus at
the same time created the problem.
In our case, the bt848 video decoder generated burst writes on PCI to the
frame buffer, meanwhile, software on the processor generated other single
cycle writes and reads to the same frame buffer device (sounds like your X
case). Through some interaction, the bursts of video DMA broke up into
more shorter bursts...something about having to drain the FIFO in the PCI
bridge to get the read data back. Once the bursts got shorter, efficiency
and throughput of PCI drops and video can't keep up. We had to tune some
things to get it working. If you need more info, I'll have to ask some
people on Monday.
jeff
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/