[Oberon] Assumed issue in the Lola-2 compiler

[Jörg Straube]

MouseP.Lola
One might debate whether -536 is wrong. Let’s agree on: it’s unusual.
You can influence the output a little by adding a length to it:
- with lenght, LSV writes a hex value
- without length, LSV writes a decimal value
Instead of 
   shreg := ~rst -> 0FFFFFDE8H:
write
   shreg := ~rst -> 0FFFFFDE8H’32:


SPI.Lola
Yes, should be 0FFFFFFFFH’32

br
Jörg

Gesendet von Mail für Windows 10

Von: rochus.keller at bluewin.ch
Gesendet: Samstag, 13. April 2019 13:22
An: oberon at lists.inf.ethz.ch
Betreff: [Oberon] Assumed issue in the Lola-2 compiler

I updated LSC.Mod with the proposed changes and regenerated and uploaded the C++ version of the compiler. 

The generated Verilog files now look much better, see https://github.com/rochus-keller/Lolac/tree/master/Examples/RISC5_translated

There are two other issues though:
- In MouseP.v line 27 column 45 the hex value was not properly generated
- SPI.Lola has invalid syntax on line 23

I didn't do a cross-validation with another Lola-2 compiler yet so the reason for the first issue could also be my transpiler.

@ Jörg Straube: would you mind sending me the Verilog files generated with your version of the Lola-2 compiler so I could use them for cross validation, thanks.

Btw I track the changes to the Oberon version of the Lola-2 compiler here: https://github.com/rochus-keller/Lolac/tree/master/Original

Best
R.K.
--
Oberon at lists.inf.ethz.ch mailing list for ETH Oberon and related systems
https://lists.inf.ethz.ch/mailman/listinfo/oberon

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.inf.ethz.ch/pipermail/oberon/attachments/20190413/45631fc8/attachment.html>