From: Groepaz (groepaz_at_gmx.net)
Date: 2003-06-21 03:01:30
On Friday 20 June 2003 20:07, Ullrich von Bassewitz wrote: > On Fri, Jun 20, 2003 at 07:54:06PM +0200, Groepaz wrote: > > eh damn...i gotta rephrase... ofcoz i mean "if the *expected* size is > > known" ... ie all immediate loads expect 8bit (or 16bit in 65816 16bit > > mode) thus the value can be savely truncated to that size. > > As I've written before: ca65 will only handle numerical expressions without > attributes. So it cannot distinguish between the values -1 and $FFFFFFFF > (for up to 32 bit host machines). While truncating > > val = -1 > lda #val > > is acceptable, > > val = $FFFFFFFF > lda #val > > is not, because it may hide errors in the code. maybe then issue a warning whenever such conversion happens....and add a switch to hide the warning? (i know its not quite nice :=P but it would really help with both oldish source and like someone else mentioned with stuff generated by older IDA versions) as for the pei instruction.... it looks like its completely broken :=P ie, i cant even come up with a one line test prog that would assemble correctly (illegal addressmode all the time)... afaik it should take one 8bit arg in braces (zp-indirect). gpz ---------------------------------------------------------------------- To unsubscribe from the list send mail to majordomo_at_musoftware.de with the string "unsubscribe cc65" in the body(!) of the mail.
This archive was generated by hypermail 2.1.3 : 2003-06-21 03:05:16 CEST