Thanks for the tip, John. You were spot on. I check my kernel config and it turns out that CONFIG_PREEMPT_RT_FULL had been unset somehow. I reset that, re-compiled, installed the new kernel, and after rebooting, /sys/kernel/realtime showed up.
Cameron On Thursday, November 2, 2017 at 8:06:25 PM UTC-6, John Morris wrote: > > Here's where we got that path in the first place: > > http://www.spinics.net/lists/linux-rt-users/msg09173.html > > Perhaps things have changed in recent kernels, or perhaps you could > double-check (double-grep?) your kbuild configuration for > CONFIG_PREEMPT_RT? > > John > > > On 11/02/2017 02:39 PM, Cameron McQuinn wrote: > > In order to avoid having to deal with compiling from source, I made a > > Debian Jessie rootfs using qemu-debootstrap, and got it up and running > > on my Snickerdoodle. I installed machinekit, and tried to start it, but > > got the same error messages, only this time without the stack smashing > > error. I took a look at `src/rtapi/rtapi_compat.*` like you suggested, > > and after splitting off `kernel_is_rtpreempt()` into a standalone test c > > program, I realized that my system does not have `/sys/kernel/realtime`, > > so `kernel_is_rtpreempt()` always returns 0. I am not sure why that file > > is missing, because I'm running a kernel that I patched and compiled > myself. > > > > Are there any other reliable ways to detect the rt_preempt kernel that I > > could try? I did not have a chance to try manually setting the flavor > > environment variable yet, but I will try that and see what happens. > > > > Cameron > > > > On Thursday, November 2, 2017 at 9:57:23 AM UTC-6, John Morris wrote: > > > > > > > > On 10/20/2017 12:51 PM, Cameron McQuinn wrote: > > [...] > > > > > > I managed to compile MachineKit from source, and am trying to > > start it > > > using the 5i25-zynq configuration (I don't actually have a > > bitstream to > > > program the FPGA yet, just trying to get MK running). Starting > > from the > > > command line, MachineKit errors out because the rtapi app gets > > > terminated due to "stack smashing". Output to from the terminal > > is as > > > follows: > > > > > [...] > > > > > > Interestingly, I have to "export FLAVOR=rt-preempt" in the shell > > before > > > attempting to start machinekit, or else it tries to load > > > rtapi_app_posix, which I did not compile. > > [...] > > > > I encountered these exact same two problems a few days ago with the > > amd64 Stretch packages, but can't for the life of me recall what the > > problem finally was. I'm sorry to be so unhelpful here, but I'm > > self-flagellating my brain over it. > > > > I do remember I couldn't get a handle on the 'stack smashing' > problem, > > so I went with the flavor detection problem instead, and it took > > maybe a > > half hour to track down. That stuff is all in and around > > `src/rtapi/rtapi_compat.*`. It might have been one of those cases > > where > > there was an error somewhere along the way, but no error message and > > not > > handled. Once I found it, I knew what the problem in my environment > > was > > and easily fixed it. The 'stack smashing' thing just went away > after > > that, IIRC. > > > > Good luck, and please report back if/when you find it so it has a > > chance > > of being fixed for the next person. > > > > John > > > > -- > > website: http://www.machinekit.io blog: http://blog.machinekit.io > > github: https://github.com/machinekit > > --- > > You received this message because you are subscribed to the Google > > Groups "Machinekit" group. > > To unsubscribe from this group and stop receiving emails from it, send > > an email to [email protected] <javascript:> > > <mailto:[email protected] <javascript:>>. > > Visit this group at https://groups.google.com/group/machinekit. > > For more options, visit https://groups.google.com/d/optout. > -- website: http://www.machinekit.io blog: http://blog.machinekit.io github: https://github.com/machinekit --- You received this message because you are subscribed to the Google Groups "Machinekit" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. Visit this group at https://groups.google.com/group/machinekit. For more options, visit https://groups.google.com/d/optout.
