mcuee

> > [@Mr-Bossman](https://github.com/Mr-Bossman) > > Please test with latest git head to see if the issue is still present on your system or not. Thanks. > > Yes my head...

> [@mcuee](https://github.com/mcuee) can you also reproduce the segfault with this timeout value? Not yet with built-in xusb example. ``` mcuee@Macmini libusb/libusb - (master)> git diff diff --git a/libusb/os/darwin_usb.c b/libusb/os/darwin_usb.c index...

I cannot reproduce the issue with the test prorgam either -- no crash after many runs. Maybe the timeout value is too long for my devices. I have configured and...

> Sorry im on my phone otherwise I would share a commit. Have you been able to recreate the issue on 1.0.28 so without [9cef804](https://github.com/libusb/libusb/commit/9cef804b2454a2226f5fa5db79a7e9aa8a45d4d4) ? You need to run...

> I did something like this > > [Mr-Bossman@bff821d](https://github.com/Mr-Bossman/libusb/commit/bff821d15ea8acc1326977e8880a313cbd25a9f7) Thanks. I can reproduce the issue with your patch. ``` cuee@Macmini libusb/libusb - (master)> sudo ./examples/listdevs Password: darwin/reenumerate_device: New DARWIN_REENUMERATE_TIMEOUT_US should...

> Also running with `-fsanitize=thread` shows that the assert fails in `libusb_unref_device` when called from the last call of it in `process_new_device` Here is the run log I have. ```...

Interestingly last time we had a similar crashing issue as well. * https://github.com/libusb/libusb/issues/1386 Fixed in commit https://github.com/libusb/libusb/commit/9401e6c7d9b45234f14075ccfcaaeccc48fadbe4.

@Mr-Bossman Thanks for your patch, now I can also easily reproduct the issue with unmodified xusb example. Rough 4 or 5 runs will get one crash. ``` mcuee@Macmini libusb/libusb -...

Just FYI, I cannot reproduce the issue with xusb example with libusb-1.0.27 and 1.0.28 release. Obviously, these two releases will not have the changes to `libusb/os/darwin_usb.c`.

@hjelmn and @osy Just wondering if you can take a look. Thanks.