The current code has the mcu report a trsync_state message every 10ms
and expects a time extension within 25ms. However, this means that if
a single mcu->host report is lost then 20ms would elapse until the
next report, which would allow for only a 5ms round-trip time before a
timeout error is reported.
Increase the trsync_state timing so that a message is sent every
7.5ms. This increases the total number of messages per second sent
from mcu to host to 133 (from 100). With this change, a single lost
message would still allow for up to a 10ms round-trip time.
Signed-off-by: Kevin O'Connor <kevin@koconnor.net>
When multiple MCUs are involved in homing, stagger the scheduling of
the trsync_state report messages from each mcu. Staggering helps
spread the bandwidth, helps reduce locking contention in the host, and
reduces the chance that intermittent latency could result in a
communication timeout.
Signed-off-by: Kevin O'Connor <kevin@koconnor.net>
The RetryAsyncCommand code needs to ensure that any response messages
are not from a previous (unrelated) query. To do that it compares the
'#sent_time' from potential responses to ensure they are not from a
previous session. However, if there are any low-level serial
retransmits then the low-level code sets the '#sent_time' to zero (to
indicate that the query send time is not strictly known). That could
result in a valid response not being accepted by RetryAsyncCommand.
If a low-level connection is experiencing a small amount of periodic
retransmits it could result in multiple high-level retry attempts
failing to the point that there is a user-facing error. This could
result in "Timeout on wait for 'tmcuart_response' response" errors.
Fix by accepting responses even if there is a low-level retransmit
once the code can confirm that there can be no previous query still in
progress.
Signed-off-by: Kevin O'Connor <kevin@koconnor.net>
Use mcu.lookup_command().get_command_tag() instead of
mcu.lookup_command_tag(). This improves error reporting on a protocol
mismatch. It also enables support for a msgtag that is negative.
Signed-off-by: Kevin O'Connor <kevin@koconnor.net>
An mcu device acting as an "mcu bridge" should only be reset after
other normal devices are reset - otherwise the bridge wont be able to
pass along the reset message to the downstream mcus.
Signed-off-by: Kevin O'Connor <kevin@koconnor.net>
The mcu can only enforce the max duration if the start value is the
same as the shutdown value, so verify that before configuing the mcu.
Signed-off-by: Kevin O'Connor <kevin@koconnor.net>
Commit 9ecddd1b fixed a serious error (introduced in 7d95a004). Force
a protocol version change so users are aware of the need to update the
mcu code.
Signed-off-by: Kevin O'Connor <kevin@koconnor.net>
Support endstops and probes attached to a different micro-controller
than their associated steppers.
Signed-off-by: Kevin O'Connor <kevin@koconnor.net>
Try to keep the host mcu_position synchronized with the
micro-controller by querying during startup and after every homing
event.
Signed-off-by: Kevin O'Connor <kevin@koconnor.net>
Set the reqclock on trsync commands to ensure those commands do not
cause 32-bit timer rollover errors. Otherwise, it may result in
"Timer too close" errors.
Signed-off-by: Kevin O'Connor <kevin@koconnor.net>
Separate out trsync handling from endstop handling in mcu.py code.
Enable the low-level trdispatch C code.
This is in preparation for multi-mcu homing.
Signed-off-by: Kevin O'Connor <kevin@koconnor.net>
In practice, these errors are usually due to problems on the host
computer instead of problems with communication.
Signed-off-by: Kevin O'Connor <kevin@koconnor.net>
The tmc2130 (and tmc5160) will respond back with the value written
during the next SPI command. Use this feature to verify that the
value written matches the value sent.
Signed-off-by: Kevin O'Connor <kevin@koconnor.net>
Don't try to detect a "real serial port" in serialhdl.py. Instead,
have the callers invoke either connect_uart(), connect_file(), or
connect_pipe().
Signed-off-by: Kevin O'Connor <kevin@koconnor.net>
Raise a printer.command_error() instead of an mcu.error() if a query
command does not succeed. That error is less likely to result in an
unhandled exception error.
Signed-off-by: Kevin O'Connor <kevin@koconnor.net>
Some internal code treats the message ids as encoded "variable length
quantities", while other internal code assumes the message id is
always one byte long. Continue using this scheme, but convert the VLQ
users to use the name "msgtag" while the 1-byte users use "msgid".
Increase the number of available msgids from 96 to 127 - the higher
values get encoded as negative "msgtags".
Signed-off-by: Kevin O'Connor <kevin@koconnor.net>
Update callers to only use exported methods of the msgproto objects.
This makes it easier to make internal changes to the code.
Signed-off-by: Kevin O'Connor <kevin@koconnor.net>
The digital output commands implement a subset of the software pwm
commands. Change the host code to just use the software pwm commands
and simplify the micro-controller code.
Signed-off-by: Kevin O'Connor <kevin@koconnor.net>
If the mcu config section is omitted, it leads to confusing error
messages. Don't default the serial config option to /dev/ttyS0 to
improve the error reporting.
Signed-off-by: Kevin O'Connor <kevin@koconnor.net>
Some RPi0 users were reporting "timer too close" errors due to pwm pin
resets - increase the time the host has to complete the startup to
avoid this.
Signed-off-by: Kevin O'Connor <kevin@koconnor.net>
Kevin O'Connor