2013/3/15 Wander Lairson Costa <wander.lair...@gmail.com>: > 2013/3/14 Xiaofan Chen <xiaof...@gmail.com>: >> On Thu, Mar 14, 2013 at 11:07 PM, Xiaofan Chen <xiaof...@gmail.com> wrote: >>> On Sun, Mar 3, 2013 at 6:05 PM, Xiaofan Chen <xiaof...@gmail.com> wrote: >>>> On Sun, Mar 3, 2013 at 6:55 AM, Wander Lairson Costa >>>> <wander.lair...@gmail.com> wrote: >>>>> 2013/3/2 Xiaofan Chen <xiaof...@gmail.com>: >>>>>>> [1] https://github.com/walac/isochronous >>>>>>> >>>>>>> [2] https://github.com/walac/bmfw >>>>>> >>>>>> I am trying to build the isochronous test program at [1] under Mac OS >>>>>> X but it failed. How to sort out the issue with strchrnul()? >>>>> >>>>> I just commited a patch that should fix this. Please pull and trying >>>>> to build again. >>>> >>>> Thanks. Now it is okay. >>> >>> Then I tested your test program under Mac OS X with latest libusbx.git >>> and your benchmark firmware with a Microchip PIC18F87J50 USB PIM, >>> the results are a bit strange. >>> >> >> On the other hand, I have not really been able to get anything isochronous >> working under Mac OS X with libusbx. I was using libusbdotnet as the >> test program and the benchmark firmware or the Cypress Cystream >> firmware which I tested under Windows using libusbK's kbench since >> libusbx does not support isochronous transfer under Windows yet. >> >> The issue could be with libusbdotnet under Mac OS X, or with >> libusbx under Mac OS X. For example, issue #99 was there >> for a long time. So it seems to me the isoc codes are not well >> tested. >> >> Have you been able to get the benchmark firmware to work >> with the test program under Linux? There could be potential >> issues with the benchmark firmware with regard to isoc >> as well. The Cystream firmware I am using is a known-good >> firmware though. >> > > Indeed the firmware seems to have problems yet. I tested it under > Linux, as far as I remember, the problems appear when I tried non > power of two transfer sizes. Please, give me this weekend so I can run > the tests again. > >
Ok, I ran the tests again in Ubuntu 12.10 with HEAD libusbx. I problem I am getting is trying to loop less than 32 bytes (a packet size), which I read returns error. Not sure if it is a firmware bug or a limitation of isochronous transfer by itself. The "0 zero bytes" read problem I see in your code probably is due to firmware internal buffer size. BTW, I fixed the warnings you were getting. -- Best Regards, Wander Lairson Costa ------------------------------------------------------------------------------ Everyone hates slow websites. So do we. Make your web apps faster with AppDynamics Download AppDynamics Lite for free today: http://p.sf.net/sfu/appdyn_d2d_mar _______________________________________________ libusbx-devel mailing list libusbx-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/libusbx-devel
