I have to agree with Alan here. I've made a driver that also handled
a video capture device, which could do 60 frames per second.
If the CPU would do a memcpy, (with Linux' efficient memcpy
implementation), then the memory would be busy enough that the chips'
FIFO would overrun.
This was most noticable with two cards running at the same time, but
could also be provoked with just one card. It depends a bit on the
motherboard too (the bus arbiter should give more priority to the
PCI device, and let the CPU just take the rest. I didn't have enough
money to buy a PCI bus analyser to look into it fully. I expect
more modern chipsets to be better able to handle this...)
Yes, a back-of-the-envelope calculation will show that it is Completely
impossible for the PCI bus not to be able to handle
60 * 512 * 240 * 4 = 29 Mb/s
1/s pix/line lines bytes/pixel = bytes/sec
becuase everything says 133Mbyte per seconde for PCI bus, but in
practise it is quite hard to get the 30Mb per second to run
faultlessly.
What chip are you using as a video/pci interface? ZR36120?
Roger.
-- | Most people would die sooner than think.... | R.E.Wolff@BitWizard.nl | in fact, most do. -- Bertrand Russsell | phone: +31-15-2137555 We write Linux device drivers for any device you may have! fax: ..-2138217- 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/