Thanks, Sebastian. That configure option did it. I didn't realize it was defaulting to uspace.
Now I can debug the kins. Rtapi_print is giving me some strange formatting. -- Ralph ________________________________________ From: Sebastian Kuzminsky [[email protected]] Sent: Wednesday, May 10, 2017 12:54 PM To: Enhanced Machine Controller (EMC) Subject: Re: [Emc-users] crashing under 2.8.0-pre1 On 05/10/2017 12:16 PM, Ralph Stirling wrote: > I have compiled 2.8.0-pre1 from source for a non-trivkins > project. I am using a Mesa 7I43 board. When I run linuxcnc > with software stepgens (no hostmot2), all is well. If I create > a plain-vanilla Mesa-based stepper configuration with pncconf, > the computer crashes during startup of linuxcnc and requires > power cycling. > > Kernel: 3.4-9-rtai-686-pae > Distro: Debian 7.11 (wheezy) > > If I do a "test axis" during pncconf, I get this error: > > Traceback (most recent call last): > File "/home/engrtech/linuxcnc-dev/lib/python/pncconf/pages.py", line 1129, > in > on_zaxistune_clicked > def on_zaxistune_clicked(self, *args): self.t.tune_axis('z') > File "/home/engrtech/linuxcnc-dev/lib/python/pncconf/tests.py", line 463, > in t > une_axis > if not self.a.check_for_rt(): > File "/home/engrtech/linuxcnc-dev/bin/pncconf", line 1640, in check_for_rt > elif hal.is_rt and not hal.kernel_version == actual_kernel: > AttributeError: 'module' object has no attribute 'kernel_version' > > When I actually run linuxcnc from the configuration made by pncconf, > I get a kernel panic: > > BUG: unable to handle kernel NULL pointer dereference at 00000508 > IP: [<f91e6328>] handle_lxrt_request+0x58/0xaf0 [rtai_sched] > *pde = 00000000 > Oops: 0000 [#1] PREEMPT SMP > > followed by more stack trace stuff. Looks like you built linuxcnc from source and configured --with-realtime=uspace instead of --with-realtime=/usr/realtime-3.4-9-rtai-686-pae. I'm able to reproduce the "no attribute 'kernel_version'" by configuring for uspace, but I don't have a 7i43 hooked up at the moment and without it I'm not able to reproduce the crash. > Linux 2.7.8 with hostmot2 runs fine, as does 2.8.0-pre1 with software stepgen. > Is this a problem of my making (and hence my solving), or is it a bug? I'm > wondering why I see mention of PREEMPT in the kernel panic when I am > running an RTAI kernel. Do I need to compile 2.8.0-pre1 differently? Our 3.4-9-rtai-686-pae RTAI kernel has the normal Linux PREEMPT feature enabled (this is different than the RT-PREEMPT feature, and not as realtimey), in addition to the RTAI feature. -- Sebastian Kuzminsky ------------------------------------------------------------------------------ Check out the vibrant tech community on one of the world's most engaging tech sites, Slashdot.org! http://sdm.link/slashdot _______________________________________________ Emc-users mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/emc-users ------------------------------------------------------------------------------ Check out the vibrant tech community on one of the world's most engaging tech sites, Slashdot.org! http://sdm.link/slashdot _______________________________________________ Emc-users mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/emc-users
