Hi, Sorry to repost this issue, which I posted some days ago, but there was no reaction so far. It should be easy to reproduce. I need just a short answer from somebody with a RT-CAN driver.
Thanks in advance, Best regards, Daniel Schnell. Daniel Schnell wrote: > Hi, > > Hunting down a RTCAN related bug in my application I ended up > compiling the rtcansend.c function with our make environment. > > Basicly the compilation is done like this: > > ppc_6xx-gcc -W -Wall -O2 -D_GNU_SOURCE -D_REENTRANT -D__XENO__ > -I/usr/xenomai/include -c -o rtcansend.o rtcansend.c > > Linking: > > ppc_6xx-gcc -o rtcansend rtcansend.o -L/usr/xenomai/lib -lpthread_rt > -lnative -lrtdm -lpthread -lrt > > If run, the following happens: > bash-2.05b# ./rtcansend rtcan0 -v -i0 01 interface rtcan0 s=0, > ifr_name=rtcan0 > rt_task_shadow: Device or resource busy > Cleaning up... > > If I leave away -lpthread_rt in the linking step, output is normal: > > bash-2.05b# test/rtcansend rtcan0 -v -i0 01 interface rtcan0 s=0, > ifr_name=rtcan0 <0x000> [1] 01 Cleaning up... > > I am using svn version 2256. > Am I doing here something fundamentally wrong in providing > -lpthread_rt as an additional library ? Shouldn't the native skin and > the POSIX skin just work together ? > > _______________________________________________ Xenomai-help mailing list [email protected] https://mail.gna.org/listinfo/xenomai-help
