G28 Z fails to stop stepper motion until the stop is hit a second time

[manbehindthemadness]

I am experiencing a problem when homing the z axis. When the z-end stop is hit the first time (before back off) the z-axis fails to halt motion; however, if the stop is triggered a second time it behaves as expected. This can be confirmed as a non-mechanical or electrical issue by monitoring the system log using journalctl -f.
End stop config (Z2 being the stop in question):
[Endstops]
homing_only_endstops = z2
invert_x1 = True
invert_x2 = False
invert_y1 = True
invert_y2 = False
invert_z1 = False
invert_z2 = False
end_stop_x1_stops = x_ccw, y_ccw
end_stop_y1_stops = x_ccw, y_cw
end_stop_z1_stops = z_ccw
end_stop_x2_stops =
end_stop_y2_stops =
end_stop_z2_stops =
soft_end_stop_min_x = 0.0
soft_end_stop_min_y = 0.0
soft_end_stop_min_z = 0.0
soft_end_stop_min_e = -1000.0
soft_end_stop_min_h = -1000.0
soft_end_stop_min_a = -1000.0
soft_end_stop_min_b = -1000.0
soft_end_stop_min_c = -1000.0
soft_end_stop_max_x = 0.342
soft_end_stop_max_y = 0.189
soft_end_stop_max_z = 0.350
soft_end_stop_max_e = 1000.0
soft_end_stop_max_h = 1000.0
soft_end_stop_max_a = 1000.0
soft_end_stop_max_b = 1000.0
soft_end_stop_max_c = 1000.0

It's also important to note that when I hit the probe directly I see a notification that endstop Z2 has been hit in the syslog; however, when the probe is hit during a normal non-homing movement the notification in the syslog is "physical end stop hit Y2"

[manbehindthemadness]

I have also noticed this in the syslog:

Oct 20 19:14:25 kamikaze redeem[1135]: 10-20 19:14 root INFO End Stop Z1 hit!
Oct 20 19:14:25 kamikaze systemd-logind[427]: Power key pressed.
Oct 20 19:14:25 kamikaze redeem[1135]: 10-20 19:14 root INFO End Stop Z1 released!
Oct 20 19:14:26 kamikaze redeem[1135]: 10-20 19:14 root INFO End Stop Z1 hit!
Oct 20 19:14:26 kamikaze systemd-logind[427]: Power key pressed.
Oct 20 19:14:26 kamikaze redeem[1135]: 10-20 19:14 root INFO End Stop Z1 released!
Oct 20 19:14:28 kamikaze redeem[1135]: 10-20 19:14 root INFO End Stop Z1 hit!
Oct 20 19:14:28 kamikaze systemd-logind[427]: Power key pressed.
Oct 20 19:14:28 kamikaze redeem[1135]: 10-20 19:14 root INFO End Stop Z1 released!

Unsure if this is a function of the latest Beaglebone software revision, but it seems that the Z1 endstop GPIO is being captured as the power button. Will investigate further.

[manbehindthemadness]

After extensive testing I have confirmed that on the Master and Revolve branches we have a problem reacting to the BLTouch pulse trigger during Z homing only.