#linuxcnc-devel Logs

Feb 11 2023

02:05 PM rmu: so i'm tinkering around with a self-written smart serial communication, connected to a 7c80. i got it to answer to mesaflash without errors and it kinda works with hm2 in linuxcnc, but process data discovery somehow hangs.
02:13 PM rmu: hm2 gives these errors https://paste.debian.net/1270416/
02:13 PM rmu: on serial trace I see that discovery RPC is send and answered, but then it pauses 100ms and begins again. whereas a 7i90, after discovery RPC, the TOC is queried via 0x4c "L" command. anybody has any hints how to debug this further? thx
02:18 PM pcw---home: sounds like the sserial remote didn't respond to some request during discovery
02:32 PM pcw---home: Can you check the SSLBP state?
03:03 PM rmu: up to the discovery rpc everthing looks identical (modulo timing). how do i check SSLBP state?
03:04 PM pcw---home: I think LinuxCNC reports it
03:06 PM rmu: after the pasted error there i sometimes get a message of a realtime delay but that is all.
03:08 PM rmu: and process data exchange is running with 1kHz
03:11 PM pcw---home: The error suggests a read byte operation timed out during discovery
03:17 PM rmu: i transcribed the trace from serial comms http://paste.debian.net/1270434/
03:22 PM rmu: thank you for looking at it. I will quit for today and continue tomorrow with a fresh set of eyes, it's late around here.
03:27 PM pcw---home: maybe something sslbp did not like about the discovery data
03:28 PM pcw---home: Let me look at the source
05:38 PM rmu: pcw---home: I think i found it. my transmit uart was not enabled before receiving first LBP cookie, and I think transmit line was not in proper idle level. 2 cookies were exchanged with a time difference of about a millisecond, I suppose the first answer is misinterpreted and probably messes something up in SSLBP. changed it and now the TOCs and PD data is queried ans transmitted.
05:46 PM pcw---home: Glad you found it. I don't think SSLBP is confused as answer without a proper CRC would be deleted
05:47 PM pcw---home: The repeat loop at discovery indicates that SSLBP didn't like the response (either a CRC or timeout error and it then retries)
05:49 PM pcw---home: the first 1 ms indicates a timeout/retry on the cookie