Email: Password: Remember Me | Create Account (Free)

Back to Subject List

Old thread has been locked -- no new posts accepted in this thread
???
12/19/09 11:55
Read: times


 		 #171758 - Yes it does (sometimes)
Responding to: ???'s previous message
Andy Neil said:
Bernd Nollmeyer said:
((void (code *) (void)) 0x0000) (); // ljmp 0000H

Does it really give an LJMP rather than an LCALL ?

Yes it does but only if the statement following the LCALL is a RET instruction.

(that optimization is implemented with a peephole rule near the end of file
http://sdcc.svn.sourceforge.net/viewvc/sd...iew=markup
in rule 400.a .. 400.c)

List of 33 messages in thread
TopicAuthorDate
ljmp in C            01/01/70 00:00      
   DEAD WRONG in assembler            01/01/70 00:00      
      It would be better to describe what you need to achieve            01/01/70 00:00      
         explanation            01/01/70 00:00      
            if I understand correctly what you want...            01/01/70 00:00      
               this sort of job is best done in ssembler            01/01/70 00:00      
               re: if I understand correctly what you want...            01/01/70 00:00      
            Function pointer            01/01/70 00:00      
            not a reset            01/01/70 00:00      
               pardon?            01/01/70 00:00      
                  jump to zero and reset            01/01/70 00:00      
                     a watchdog reset does not necessarily perform a "full" reset            01/01/70 00:00      
                        ljmp 0000H tested with SDCC            01/01/70 00:00      
                           That's exactly what OP needed, not "reset"            01/01/70 00:00      
                           TOTALLY ABSOLUTELY WRONG!!            01/01/70 00:00      
                              EA=0            01/01/70 00:00      
                                 it depends            01/01/70 00:00      
                                    which leads me to            01/01/70 00:00      
                                    you prove my point            01/01/70 00:00      
                                       re: you prove my point            01/01/70 00:00      
                                          it will not do what you expect RELIABLY!!            01/01/70 00:00      
                                 I do take care of re-initializing            01/01/70 00:00      
                                    double, pls ignore            01/01/70 00:00      
                                    you don't get it, do you            01/01/70 00:00      
                           Does it really?            01/01/70 00:00      
                              Yes it does (sometimes)            01/01/70 00:00      
                                 Tail-end recursion            01/01/70 00:00      
                                 optional?            01/01/70 00:00      
                                    it's the default            01/01/70 00:00      
                                       ljmp in C            01/01/70 00:00      
                                          Have you confirmed that?            01/01/70 00:00      
                                             call != jump            01/01/70 00:00      
   I've seen some.            01/01/70 00:00      

Back to Subject List