7i73 LCD shows rubisch

Yeah, not only is the data screwy, the write order is backwards so the data may or
may not be sent properly out the serial port

Ok, i understand that now i will wait.
Still some things to do in my application anyhow. Plus rewiring my old lady to trsnsfer from 7i43 to new 7i76e/7i73 combo.

Thx guys, honestly as one who is in progressional life used to work with lots of suppliers ad programmers. All your support is top knodge !!!!!!

Good night
Mike

can someone please try if this fixes the issue?
github.com/rene-dev/linuxcnc/tree/sserial-race-fix

[strike]I was wondering if that was the answer, or whether we needed to keep the setup first and tag a separate DoIt write on the end.[/strike]

I was thinking that the problem was here:
github.com/rene-dev/linuxcnc/blob/4f7110...tmot2/sserial.c#L968

ie, the DoIt is added to the TRAM list by the setup_channel function, not at the end of the setup_remotes function.
(They are called in this sequence here)
github.com/rene-dev/linuxcnc/blob/4f7110...tmot2/sserial.c#L802

Im not sure I understand what you see as the issue.
My change sets the cs with the do-it after setting the user registers.
that could be an issue if the fpga reads the user register while its written by the driver.
I verified the change of the order by looking at the packets.
now that I think about this issue, that doesnt explain why there is rubbish in the dump.
the data should be there, just in the wrong place.

It has been a long time since I looked at this, but I thought that the "DoIt" command was in the command register, not the CS register.