'Run from here' not working as expected in Axis on 2.9.1

[spacestate1]

Here are the steps I follow to reproduce the issue:

1. Open a gcode file in Axis
2. Get XYZ homed
3. Scroll through the gcode in the code window to find an arbitrary point, right click and select 'Run from here'

This is what I expected to happen:

The machine re-positions the spindle and runs from the selected line of gcode

This is what happened instead:

The code runs from the beginning of the gcode file

It worked properly before this:

I noticed the behavior start when I switched from LinuxCNC 2.8.4 to 2.9.1
The same behavior is seen in 2.9.1 Axis_mm in sim running inside a virtual machine

Information about my hardware and software:

• I am using this user interface AXIS
• I am using this interface hardware MESA 7i76 with a 7i92 interface card.
• The issue is also present in sim Axis_mm

[samcoinc]

Are you selecting the line before 'right clicking'? If not - I think it starts the program from the beginning.. sam

…

On Wed, Dec 6, 2023 at 12:21 PM Collin McRann ***@***.***> wrote: Here are the steps I follow to reproduce the issue: 1. Open a gcode file in Axis 2. Get XYZ homed 3. Scroll through the gcode in the code window to find an arbitrary point, right click and select 'Run from here' This is what I expected to happen: The machine re-positions the spindle and runs from the selected line of gcode This is what happened instead: The code runs from the beginning of the gcode file It worked properly before this: I noticed the behavior start when I switched from LinuxCNC 2.8.4 to 2.9.1 The same behavior is seen in 2.9.1 Axis_mm in sim running inside a virtual machine Information about my hardware and software: - I am using this user interface AXIS - I am using this interface hardware MESA 7i76 with a 7i92 interface card. - The issue is also present in sim Axis_mm — Reply to this email directly, view it on GitHub <#2771>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AEHRGQX74GXFSIRX4WYUUHTYICZRJAVCNFSM6AAAAABAJZ4RWKVHI2DSMVQWIX3LMV43ASLTON2WKOZSGAZDSMJTG42TANQ> . You are receiving this because you are subscribed to this thread.Message ID: ***@***.***>

[spacestate1]

I see the same behavior if I select the line and right click 'Run from here' or if I just right click and hit 'Run from here' on the line. it goes back to the beginning.

[phillc54]

It appears to be caused by the file being reloaded and having the program_start_line reset from commit bc9a79d

[Sigma1912]

Yes, if iI recall correctly, the idea there was for the code to be reloaded when the user clicks on the new 'Re-Check' button on the limit violations popup and not force a reload when running the file.

def run_warn():
    warnings = []
    if o.canon:
        machine_limit_min, machine_limit_max = soft_limits()
        for i in range(3): # Does not enforce angle limits
            if not(s.axis_mask & (1<<i)): continue
            if o.canon.min_extents_notool[i] < machine_limit_min[i]:
                warnings.append(_("Program exceeds machine minimum on axis %s")
                    % "XYZABCUVW"[i])
            if o.canon.max_extents_notool[i] > machine_limit_max[i]:
                warnings.append(_("Program exceeds machine maximum on axis %s")
                    % "XYZABCUVW"[i])
    if warnings:
        text = "\n".join(warnings)
        return int(root_window.tk.call("nf_dialog", ".error",
            _("Program exceeds machine limits"),
            text,
            "warning",
            1, _("Re-Check"), _("Run Anyway"), _("Cancel"))) + 1
    return 0

[Sigma1912]

Ah, no. I just retested the issue.
The point was:
If you change the tool after loading the gcode then the limit check does not reflect the current tool but the tool at the time when the code was loaded.

[Sigma1912]

Maybe we could check the value of 'program_start_line':

    def task_run(*event):
        global program_start_line, program_start_line_last
        if program_start_line == 0:
            res = 1
            while res == 1:
                res = run_warn()
                if res == 2: break
                if res == 3: return
                print("reload file")
                reload_file()
...

[andypugh]

It appears to be caused by the file being reloaded and having the program_start_line reset from commit bc9a79d

That change should only reload the file if the "Re-check" buttton is pressed. Is there something wrong with my logic?

[phillc54]

TBH I did not notice a recheck button. I ran the axis_mm sim and just did a run from line and the file reloaded.

[andypugh]

The change referenced ( bc9a79d ) should only be activated in cases where the limit check fails, and then the re-check button is pressed.

Is it possible for someone to run a git bisect to see when the behaviour changed? (or just choose a point in the timeline prior to my change, to make sure it's my fault)

[rmu75]

run_warn return 0 by default (i.e. the non-warning case) and that triggers a reload. Also there is #2490

[phillc54]

I can have a look tomorrow morning if no one else has by then.
I did comment out the reload_file() line in the while res block and run from line worked correctly.

[rmu75]

add "if res == 0: break" or replace "if res == 2: break" with "if res in [0, 2]: break" or something like that.

[andypugh]

And this, kids, is why I shouldn't be allowed anywhere near software.

Bother! It's an easy fix but I think it means a 2.9.2 release, and that's rather more work than the fix.

[spacestate1]

added "if res == 0: break", and it seems to fix the issue on my end.

[phillc54]

Confirming that this issue first appears in bc9a79d.
Roberts suggestion of if res in [0, 2]: break does seem to fix it for me.

[andypugh]

OK, suggested fix tested and implemented.