On Fri, Nov 05, 2004 at 10:44:36AM +0100, Axel Waggershauser wrote: > Hi Eric, > > On Fri, 2004-10-29 at 12:51 -0700, Eric Blossom wrote: > > On Fri, Oct 29, 2004 at 01:06:34PM +0200, Axel Waggershauser wrote: > > > another question: Have you tried accessing two devices at the same time? > > > If so, what was your total throughput per sec? > > > I haven't, but I'll give it a try over the next couple of days and let > > you know what happens. I don't expect a problem. > > Meanwhile, I tested my patched fx2_programmer on an onboard Intel > controller and reached the 32MB/sec as well. Now I am still wondering > what happens when two devices are accessed at the same time > (unfortunately I don't have two of these FX2 based cameras available at > the moment). If you are right about the controller being in general the > bottleneck, that would mean a "true kernel space" driver would not > perform any better
I wouldn't expect a kernel space driver to do any better on throughput. I'm assuming that the current strategy and driver sustains a non-zero length queue of work on the EHCI hardware at all times (I haven't checked this...). > and there might not even be any device/controller > combination with a higher throughput than those 32MB/sec available at > the moment. Has anybody any information about what is actually possible > with currently available hardware? I know of folks using the fusb code with an FX2 that get 40MB/sec. They're using the FX2 in a different mode and use different firmware. I think they're using it in a unidirectional "full-auto" mode that minimizes the work the firmware has to do. > There is of course another possibility: maybe the FX2 is the bottleneck > here. Again, I know that 40MB/sec is possible. Anyone out there with a netchip (recently bought by PLX) based device? It would be great to have data on those parts too. > So, if the total throughput with 2 devices happens to be 32MB/sec as > well, I'd conclude that there is no controller available with a higher > throughput. I would hope for a total throughput of as near to 60MB/sec > as possible :-). You'll never get 60MB/sec. The raw USB capacity is 480 Mbit/sec, but there is some minimum overhead on the order of 10%. You can work it out from the specs. IIRC, the best you can do in theory is about 54MB/sec. > In case you find the time to make this test, I'd be eager to know about > the results. I'll see what I can do. Eric ------------------------------------------------------- This SF.Net email is sponsored by: Sybase ASE Linux Express Edition - download now for FREE LinuxWorld Reader's Choice Award Winner for best database on Linux. http://ads.osdn.com/?ad_id=5588&alloc_id=12065&op=click _______________________________________________ [EMAIL PROTECTED] To unsubscribe, use the last form field at: https://lists.sourceforge.net/lists/listinfo/linux-usb-devel
