A couple of errors during update 2.8.2-130-g87f66feaf

Here is a zip of my config directory.
 
 

Repeating request:
@spumco :

1) post the output from:
   $ linuxcnc_info
2) post a tar or zip file of the configuration directory
3) post details on the steps needed to produce the error
4) when the error occurs, execute (in a separate shell):
   $ halcmd show pin|grep active
and post the result

Also
4) Please try to duplicate in an included distribution sim
config (qtdragon?) and post detailed steps to reproduce the error.

Files attached, along with screenshot of error message.

LCNC 2.9 RIP, qtpyvcp (master), Probe Basic (py3)

Steps to reproduce:

• Start machine & home all axis
  • jogging works
• load ngc file
  • switch back to manual mode - jogging works
• run ngc file using cycle start GUI button
  • file completes
• select manual mode GUI button
  • jogging works
• run ngc file again
  • file completes
• select manual mode GUI button
  • attempt to jog => error message + jogging not responsive
• close error message
  • jogging still not responsive
• select manual mode GUI button => error message + jogging not responsive
• select auto mode GUI button => error message + jogging not responsive

As you can see from above, the error didn't happen the first time I ran the program, but it did after the second time.  Other sessions have errored out on the 1st or later attempts.

Could not reproduce on included sim QtDragon_hd.

@spumco

The ini file PBtestATC.ini contained in HAL_rev33_ATC.zip
uses:

[DISPLAY]
DISPLAY = probe_basic

The probe_basic executable is not part of the LinuxCNC
git tree or distribution and i don't have the time or
energy to build or test executables from outside projects.

Can you
1) reproduce problem using in-tree guis?
or better:
2) reproduce problem using an in-tree SIMULATION config
like qtdragon?
Ref: configs/sim/qtvcp_screens/qtdragon/qtdragon.ini

I have some ideas, but ... the hardware configs showing
the problem are complicated to test as they include
specific hardware (hm2_eth, vc-p4s) that i do not have
so simplifying to a sim config would be helpful.

Alternatively, making a modified config to test without
use of the vc-p4s component could help to isolate
the problem as it may be related to halui plus/minus
jogging implemented therein.

[don't blame for not wanting to delve in to edge-case GUI's]

I'm unable to reproduce the error in qtdragon or qtdragon_hd sims

I turned off the vc-p4s MPG component (commented out both HALFILE and POSTGUI_HALFILE) and the error is not occurring, regardless how many times I run an NGC file and switch back and forth between manual/auto modes.

I think that narrows it down, yes?

This wasn't an issue as of few months ago and I haven't changed anything in either of the vc-p4s hal files.

@spumco :
It appears you are running a self-built Run-in-place
install for the master branch at commit 8da3ed4f04 (Jun 9 2022)

I've made a test branch with a potential fix based
on the current master branch that you can test by
checking out and building the branch:
dgarr/master_testa

Example:
$ cd your_LinuxCNC_git_top_directory
$ git fetch
$ git checkout dgarr/master_testa
$ cd src
$ make && sudo make setuid
$ test_your_hardware_config

Please test and report
===========================================================

@johnnyCNC :
It appears you are running a deb install for the 2.8 branch
commit 87f66feaf (27 may 2022)

I've made a test branch dgarr/2.8testa based on current
2.8 branch that you can get from the buildbot as a 'scratch' deb

Instructions for using a scratch deb at:
buildbot.linuxcnc.org/

Please test and report