Erik Rull wrote: > Hi all, > >> On December 19, 2013 at 1:49 PM Erik Rull <erik.r...@rdsoftware.de> wrote: >> >> >> Hi Hans, >> >> Hans de Goede wrote: >>> Hi, >>> >>> On 12/19/2013 10:51 AM, Erik Rull wrote: >>>> Hi all, >>>> >>>> I read already some threads here in the list about the "hot disconnect" and >>>> the issues that happen. >>>> >>>> I'm using the libusbx 1.0.17 and encounter the following stacktrace when >>>> removing a usb device during a transfer - the timeout is set to 50 (msec) >>>> >>>> /lib/libusb-1.0.so.0 [0xb7037f94] >>>> /lib/libusb-1.0.so.0 [0xb7037ea2] >>>> /lib/libusb-1.0.so.0:libusb_handle_events_completed() >>>> /lib/libusb-1.0.so.0:libusb_handle_events_timeout_completed() >>>> /lib/libusb-1.0.so.0 [0xb7036caa] >>>> /lib/libusb-1.0.so.0 [0xb703b27a] >>>> /lib/libc.so.6:ioctl() >>>> >>>> this stays for several seconds until I send a SEGV to the process: >>>> >>>> /lib/libusb-1.0.so.0 [0xb7037f94] >>>> /lib/libusb-1.0.so.0 [0xb7037ea2] >>>> /lib/libusb-1.0.so.0:libusb_handle_events_completed() >>>> /lib/libusb-1.0.so.0:libusb_handle_events_timeout_completed() >>>> /lib/libusb-1.0.so.0 [0xb7036caa] >>>> /lib/libusb-1.0.so.0 [0xb703b27a] >>>> /lib/libusb-1.0.so.0 [0xb7038ab6] >>>> /lib/libusb-1.0.so.0 [0xb7031fe0] >>>> /lib/libusb-1.0.so.0 [0xb7031c9c] >>>> >>>> That's it. Additonally I had a SEGV also when removing the USB device, but >>>> I wasn't able to catch the stack there, sorry. >>>> >>>> I know that the handling of the real disconnect may have several failing >>>> transfers before the disconnect is actually reported. >>>> But are there possibilities to prevent the process to stall or to SEGV? >>>> >>>> Maybe there is an easy workaround for that? >>> >>> I'm afraid that with the above info we cannot help you further, >>> because it is incomplete, can you please: >>> >>> 1) do export LIBUSB_DEBUG=4 on a terminal, and start your app from that >>> some terminal, this should produce a wealth of debugging messages, >>> please collect them all >>> >>> 2) Make sure you build libusb with debuginfo, and next time you post traces >>> make sure they include symbolic names for all involved addresses >>> >>> Thanks & Regards, >>> >>> Hans >> >> Thanks for the hints on how to get more information. I will send the output >> as soon as I got it running again. >> >> Best regards, >> >> Erik >> > > Attached the libusbx logfile (plain text, zipped), I hope this is sufficient. > It > seems to stall within an infinite loop within the library. Please ingore the > Date / Time in front of each log line. > > Best regards, > > Erik > >
Hi all, can someone have a look at the previously sent logfiles (2014-01-07)? I would just like to know whether it's a bug on my side or a libusb bug. Thanks. Best regards, Erik ------------------------------------------------------------------------------ CenturyLink Cloud: The Leader in Enterprise Cloud Services. Learn Why More Businesses Are Choosing CenturyLink Cloud For Critical Workloads, Development Environments & Everything In Between. Get a Quote or Start a Free Trial Today. http://pubads.g.doubleclick.net/gampad/clk?id=119420431&iu=/4140/ostg.clktrk _______________________________________________ libusbx-devel mailing list libusbx-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/libusbx-devel
