Re: [RTnet-users] WAITONCYCLE_EX cycle number problem

2006-08-10 Thread Jan Kiszka
matrix_df hotmail wrote: > Hello > >> Both lines above are essential. The first one makes sure your >> application's memory will remain physically present and not swapped away >> e.g. The second one turns the main thread of this program into a Xenomai >> thread, and this is required to block on an

Re: [RTnet-users] WAITONCYCLE_EX cycle number problem

2006-08-10 Thread matrix_df hotmail
RE) in main() and rt_task_shadow(&task, "waitonsync", 1, 0) in xmit_msg(void *arg) ? Jochen >From: Jan Kiszka <[EMAIL PROTECTED]> >To: [EMAIL PROTECTED] >CC: rtnet-users@lists.sourceforge.net >Subject: Re: [RTnet-users] WAITONCYCLE_EX cycle number problem

Re: [RTnet-users] WAITONCYCLE_EX cycle number problem

2006-08-10 Thread Jan Kiszka
matrix_df hotmail wrote: > Hello > >> Your build must have become inconsistent in some way. Are you sure that >> the latest RTnet modules are installed on the master, not accidentally >> some older ones? > > I am sure, but I will double check. > > When I use your test program it works: > >

Re: [RTnet-users] WAITONCYCLE_EX cycle number problem

2006-08-10 Thread matrix_df hotmail
--- BUT when I remove the lines with mlockall(MCL_CURRENT | MCL_FUTURE); and rt_task_shadow(&task, "waitonsync", 1, 0); it FAILS with error=-13: Jochen >From: Jan Kiszka <[EMAIL PROTECTED]> >To: [EMAIL PROTECTED] &g

Re: [RTnet-users] WAITONCYCLE_EX cycle number problem

2006-08-10 Thread Jan Kiszka
matrix_df hotmail wrote: > Hello > > I recompiled. (make clean, make) > But I have installed the new svn-version only on master. Is that a problem? Nope. > > Now I removed some lines and I can say that neither WAITONCYCLE nor > WAITONCYCLE_EX return. They just wait and hang. > >//ret = rt_

Re: [RTnet-users] WAITONCYCLE_EX cycle number problem

2006-08-10 Thread matrix_df hotmail
Hello I recompiled. (make clean, make) But I have installed the new svn-version only on master. Is that a problem? Now I removed some lines and I can say that neither WAITONCYCLE nor WAITONCYCLE_EX return. They just wait and hang. //ret = rt_dev_ioctl(fdtdma, RTMAC_RTIOC_WAITONCYCLE, TDMA_W

Re: [RTnet-users] WAITONCYCLE_EX cycle number problem

2006-08-10 Thread Jan Kiszka
matrix_df hotmail wrote: > Hello > > With new rtnet-0.9.4-svn (10.8) > neither WAITONCYCLE nor WAITONCYCLE_EX waits on the cycle and just > return !! > (tested on master) > > RTMAC_RTIOC_WAITONCYCLE, TDMA_WAIT_ON_SYNC > As I said, I broke the ABI by redefining the numeric value of TDMA_WAIT_ON

Re: [RTnet-users] WAITONCYCLE_EX cycle number problem

2006-08-10 Thread matrix_df hotmail
Hello With new rtnet-0.9.4-svn (10.8) neither WAITONCYCLE nor WAITONCYCLE_EX waits on the cycle and just return !! (tested on master) RTMAC_RTIOC_WAITONCYCLE, TDMA_WAIT_ON_SYNC Jochen >From: Jan Kiszka <[EMAIL PROTECTED]> >To: [EMAIL PROTECTED] >CC: rtnet-users@lists.sourceforge.net >Subjec

Re: [RTnet-users] WAITONCYCLE_EX cycle number problem

2006-08-10 Thread matrix_df hotmail
Hello With new rtnet-0.9.4-svn (10.8) neither WAITONCYCLE nor WAITONCYCLE_EX waits on the cycle and just return !! (tested on master) Jochen >From: Jan Kiszka <[EMAIL PROTECTED]> >To: [EMAIL PROTECTED] >CC: rtnet-users@lists.sourceforge.net >Subject: Re: WAITONCYCLE_EX cycle number problem >

Re: [RTnet-users] WAITONCYCLE_EX cycle number problem

2006-08-09 Thread Jan Kiszka
matrix_df hotmail wrote: > Hallo Jan > > rtnet-0.9.4-svn + xenomai 2.2.0 > > WAITONCYCLE seems to work now on master. > > Now I have a problem with WAITONCYCLE_EX. > I always get cycle number 0 returned after executing the ioctl function. Yep, you managed to find another piece of yet-unused RTn

[RTnet-users] WAITONCYCLE_EX cycle number problem

2006-08-09 Thread matrix_df hotmail
Hallo Jan rtnet-0.9.4-svn + xenomai 2.2.0 WAITONCYCLE seems to work now on master. Now I have a problem with WAITONCYCLE_EX. I always get cycle number 0 returned after executing the ioctl function. waitinfo.type = TDMA_WAIT_ON_SYNC; waitinfo.ext_size = 0; ret = rt_dev_i