I can confirm that doing the device reset on exit corrects this issue. One additional note, it is probably impossible to protect everything with a try: expect: for a KeyboardInterrupt. This exception can happen more or less at random anywhere or anytime in your program when the user does a ^C.
The only way to have a stable solution is to install a signal handler for SIGINT. Then implement a graceful exit strategy that can do the device reset at the correct time as you bring down the program. See: https://docs.python.org/3/library/signal.html If a signal handler raises an exception, the exception will be propagated > to the main thread and may be raised after any bytecode instruction. Most > notably, a KeyboardInterrupt may appear at any point during execution. Most > Python code, including the standard library, cannot be made robust against > this, and so a KeyboardInterrupt (or any other exception resulting from a > signal handler) may on rare occasions put the program in an unexpected > state. > For many programs, especially those that merely want to exit on > KeyboardInterrupt, this is not a problem, but applications that are complex > or require high reliability should avoid raising exceptions from signal > handlers. They should also avoid catching KeyboardInterrupt as a means of > gracefully shutting down. Instead, they should install their own SIGINT > handler. On Sat, Apr 23, 2022 at 11:30 PM stuart lynne <stuart.ly...@gmail.com> wrote: > The device reset seems to help. > > When properly exited it went about eight restarts correctly. > > Interrupted (^C) exit I think misses the device reset in some cases (I can > fix that) and eventually it got to the FUBAR state. > > Once in that state opening and closing with reset does not fix it. > Physical replug necessary. > > I'll clean up the exception handling so that breaking to exit goes through > the proper closing sequence and see if that helps. > > > Thanks! > > > On Sat, Apr 23, 2022 at 8:45 PM Xiaofan Chen <xiaof...@gmail.com> wrote: > >> On Sun, Apr 24, 2022 at 8:43 AM stuart lynne <stuart.ly...@gmail.com> >> wrote: >> > >> > I am seeing a consistent issue with Garmin Ant+ dongles under Windows >> with libusb0. >> > >> > They work the first time my application runs, but fail on the second >> attempt. >> > >> > I have verified the same behaviour across three Garmin Dongles. I have >> verified >> > that the Cycplus Dongle works correctly. >> > >> > To maintain compatibility with the other applications running (Zwift >> and PerfPro >> > CT Smart) which rely on libusb0 I need to use that backend. >> > >> > Those applications do not have any problems with the Garmin (or Cycplus) >> > dongles, using them multiple times without a problem. >> > >> > Is it possible that I need to do something on exit that might cause >> this? >> > >> > Is there anything I can do to remediate this when seen? >> > >> >> Does it make a difference if you use the libusb-1.0 backend? This is not >> a solution but rather a troubleshooting step. You can keep your driver >> as it is, but switch to libusb-1.0 backend. You need to have libusbk.dll >> and libusb-1.0.dll in place to make that work. >> >> I understand that you may have to keep with the libusb0.sys based >> driver provided >> by Garmin (I used to have one such USB dongle for my old Garmin Vivofit >> 1). >> >> The other potential work-around is to add a device reset at exit. Usually >> this kind of issue is caused by device firmware bug (eg: USB data toggle >> issues) and the hope is that the device reset can put the firmware to a >> good >> state. >> >> >> >> -- >> Xiaofan >> >> >> _______________________________________________ >> pyusb-users mailing list >> pyusb-users@lists.sourceforge.net >> https://lists.sourceforge.net/lists/listinfo/pyusb-users >> > > > -- > __________O___________ > _______-\<,____________ > _____(_)/_(_)___________ > _________________________ > Stuart_Lynne____<stuart.ly...@gmail.com > >____604-518-1749(m)__604-461-7532(h) > -- __________O___________ _______-\<,____________ _____(_)/_(_)___________ _________________________ Stuart_Lynne____<stuart.ly...@gmail.com>____604-518-1749(m)__604-461-7532(h)
_______________________________________________ pyusb-users mailing list pyusb-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/pyusb-users
