Simulating MCS-52 BASIC V1.31 that runs on a CH552T fails

[oloftangrot]

uCsim is installed as provided by sdcc on an Ubuntu 24.04 LTS machine.

When I try to simulate BAS552T.hex downloaded from https://github.com/mit41301/BASIC-52_CH552T
the BASIC interpreters seems to start and output its prompt at the connected tty (/dev/pts/0 in my case) but when I try to run a single row BASIC program nothing happens after CR and the terminal becomes unresponsive.

Mayby it is due to some XRAM(?) issue with the CH552T but as far as I can tell there is 64 k available.

The code I try to run is built for a CH552T which is an enhanced 8051 processor. I have assumed that invoking s51 without any specific processor type is good enough. If I deploy the HEX-file on real CH552T hardware it seems to run with no hassles.

Is there some fault I make that prevents the simulation to run properly?

[danieldrotos]

I've checked the code of this BASIC and I see that some 552 specific peripherals are used which are not simulated by uCsim. But the BASIC should work anyway. I think there is a problem with the line endings.

You can start uCsim with this command:

ucsim_s51 -V -Suart=0,port=5555 BAS552T.hex

and start a terminal emulator program which supports telnet and connect to localhost:5555 (putty, for example). Try to configure how to send line ending, or send CR by ^M and/or LF by ^J. In uCsim console you will get a debug message about the characters received the UART.

As I see, the CR-LF doesn't work, because after LF the BASIC continues and waits for CR.

If I send program lines (eg. 10 rem) with CR only, they are accepted, but I can not send commands (eg. list, run) with any kind of line endings, I got ERROR: BAD SYNTAX answer...

So, there can be a simulation problem, but I'm not sure.

[oloftangrot]

I will try as soon as possible. A question; is the serial input buffered? That is, if the simlation is too slow might a virtual therminal override the oldest character in a CRLF sequence? Den ons 1 apr. 2026 09:50danieldrotos ***@***.***> skrev:

…

*danieldrotos* left a comment (danieldrotos/ucsim#10) <#10 (comment)> I've checked the code of this BASIC and I see that some 552 specific peripherals are used which are not simulated by uCsim. But the BASIC should work anyway. I think there is a problem with the line endings. You can start uCsim with this command: ucsim_s51 -V -Suart=0,port=5555 BAS552T.hex and start a terminal emulator program which supports telnet and connect to localhost:5555 (putty, for example). Try to configure how to send line ending, or send CR by ^M and/or LF by ^J. In uCsim console you will get a debug message about the characters received the UART. As I see, the CR-LF doesn't work, because after LF the BASIC continues and waits for CR. If I send program lines (eg. 10 rem) with CR only, they are accepted, but I can not send commands (eg. list, run) with any kind of line endings, I got ERROR: BAD SYNTAX answer... So, there can be a simulation problem, but I'm not sure. — Reply to this email directly, view it on GitHub <#10?email_source=notifications&email_token=AC33OG4JX26IL6NQ6U5PVYL4TTC5FA5CNFSNUABFM5UWIORPF5TWS5BNNB2WEL2JONZXKZKDN5WW2ZLOOQXTIMJWHAZDEMRXHEZ2M4TFMFZW63VGMF2XI2DPOKSWK5TFNZ2KYZTPN52GK4S7MNWGSY3L#issuecomment-4168222793>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AC33OG5BHD5JU4U4RXHGUOL4TTC5FAVCNFSM6AAAAACXIHJ2B2VHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHM2DCNRYGIZDENZZGM> . You are receiving this because you authored the thread.Message ID: ***@***.***>

[danieldrotos]

Both OS and uCsim do buffering on incoming trafic. You can type ahead in terminal, uCsim will buffer it and pick chars one by one when simulation of UART counts appropriate amounts of ticks.

[oloftangrot]

So I tested with telnet on the same linux host as uCsim executes. Now I get the same result as you with a BAD SYNTAX response. I have tested ^M and ^J in the terminal and the BAD SYNTAX is only shown on CR. LF does not matter. Typing in program statements seems to work. Like 20 LET A=5 returns the ">" prompt. But LIST or RUN fails.

[danieldrotos]

Maybe the sw checks some 552 specific part which is not simulated by uCsim. Somebody should re-engineer the code to figure it out.

[oloftangrot]

I shall try to find a more generic 8051 built 1.31 and verify if it runs. Den tors 2 apr. 2026 09:08danieldrotos ***@***.***> skrev:

…

*danieldrotos* left a comment (danieldrotos/ucsim#10) <#10 (comment)> Maybe the sw checks some 552 specific part which is not simulated by uCsim. Somebody should re-engineer the code to figure it out. — Reply to this email directly, view it on GitHub <#10?email_source=notifications&email_token=AC33OG3GVZ6WIAOLF3NMVHT4TYGW5A5CNFSNUABFM5UWIORPF5TWS5BNNB2WEL2JONZXKZKDN5WW2ZLOOQXTIMJXGUYTKNBUHAY2M4TFMFZW63VGMF2XI2DPOKSWK5TFNZ2KYZTPN52GK4S7MNWGSY3L#issuecomment-4175154481>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AC33OG7JC3R3ET2PXLDI4DT4TYGW5AVCNFSM6AAAAACXIHJ2B2VHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHM2DCNZVGE2TINBYGE> . You are receiving this because you authored the thread.Message ID: ***@***.***>

[oloftangrot]

So I tested an old 1.1 version and a 1.3 version but I get no start up prompt. This might be due to auto detection of the connecting terminal baudrate. So the code waits for a SPACE character a that an the detection might not simulate correctly. Or maybe I choose the wrong processor type.

The HEX-files I tested was found here:
https://www.dos4ever.com/8031board/