[Sdcc-user] News about the "SIGSEGV"-error i caught...

[Thorsten G. <tho...@gm...>]

Hi Bernhard, hi Johan, hi rest of the group,

i think, the "Caught SIGSEGV: 11"-problem i got, has to do
with the length of assembly code produced for a single CASE
in a SWITCH/CASE construct.

I came again in trouble with another project, where i use
also a state machine out of SWITCH/CASEs. Not so huge and
complex as the other, but the same kind of mechanism.

I can reproduce it, if i insert a single line in a CASE,
which produces 2 byte more code (e.g. P1_0 = 0;) for the
CASE. These two bytes causes the SIGSEGV-error.
Inside the CASE there are lots of computations with long
variables, so the few lines produces a lot of assembly code.

If its interesting for you, i will cut the part from the .lst-file,
so you can see where the limit for the CASE-length is.

Best regards, Thorsten

-- 
-----------------------------------------------------

Dipl.-Ing.(FH) Thorsten Godau

Voicemail/Fax +49(0)89/1488-206580
@Home mailto:tho...@gm...
@Work mailto:tho...@tz...
http://www.qsl.net/dl9sec  
http://webox.tz-mikroelektronik.de
AX25: DL9SEC@DB0GPP.#BW.DEU.EU

-----------------------------------------------------