I've modified the USBSimple example to open interrupt pipes instead
of bulk transfer pipes. In doing this, I've managed to figure out a
reasonable 7-byte output pipe format that causes the desired LED
changes on my device.
However, the LEDs don't update immediately. In my testing, I noticed
that I would see the updated LEDs only on the *second* run of the
program. After a little debugging, I discovered the "truism":
-> After a WritePipe or WritePipeAsync (and runloop run to
completion), the data written to the device doesn't seem to get there
until the *next* time I call SetConfiguration on the device.
I'm pretty sure it must be a side-effect of the SetConfiguration call
that is causing the data to "flush" or whatever. But does anybody
have any tips for how I might proceed? I thought I would try to
remote-debug the system and step into the SetConfiguration call, to
see which call in the kernel corresponds with the flush. But that
turned out to be more difficult than I was thinking. Before I hunker
down again to get that debugging environment working as expected, I
thought I might get lucky and gather some good leads here.
Is it typical that I should have to do something else after the
WritePipe() call to make sure that the data gets pushed out? I would
think that whatever "starting over" effect that calling
SetConfiguration has would be roughly equivalent to the behavior of
closing down the interface and device when the application quits. But
even then the data doesn't seem to arrive until the next launch.
I suppose this is probably one of those famous "check the bus
analyzer" type questions, but unfortunately I don't have access to
one, and probably won't given the budget of this project.
Thanks for any leads you can provide!
Daniel
_______________________________________________
Do not post admin requests to the list. They will be ignored.
Usb mailing list (email@hidden)
Help/Unsubscribe/Update your Subscription:
http://lists.apple.com/mailman/options/usb/email@hidden
