On Thu, 2016-01-14 at 12:20 +0000, Dean Jenkins wrote: > Unwanted blocking occurs when read_descriptors() is getting the > descriptor information from the allocated rawdescriptors structure of a > USB hub but a device connected to the hub is failing and the device is > being recovered. Quite possibly so. It is a separate issue. I'd be happy to provide a patch. But it is not the only read that might block. The rest is also annoying. An uninterruptible sleep on an event that can take so long is no good. > This patch helps in the case where the userland application is being > terminated by Ctrl-C whilst a failing USB device is being recovered due > to communication failure which causes blocking to get the USB hub > descriptors information from memory. Yes, that is the patch's point. > Have you checked that the calling sysfs layer can handle the returned > error code of -EINTR ? What code handles that ? -EINTR is supposed to go to user space. > Note that read_descriptors() gets called multiple times (I have seen 3 > calls) per a single sysfs read. OK, but what does that change? Regards Oliver -- To unsubscribe from this list: send the line "unsubscribe linux-usb" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
