| View previous topic :: View next topic |
| Author |
Message |
loupan
Joined: 22 Oct 2007
Posts: 21
|
|
 Posted: Thu Mar 18, 2010 7:54 am |
 |
|
I am seeing something very similar - see post:
http://www.ccsinfo.com/forum/viewtopic.php?t=41396&highlight=
Am I missing something? This seems like a serious problem.
Assume that I am writing software to control the cell gates in a prison.
I have two main routines:
| Code: | lockdown_riot_in_progress () { // lock all cells immdeiately
fire_detected_open_all_cells() { // let them all out |
Couldn't it be a problem if the correct addresses are not getting generated!
And as in your problem, adding code or moving things around fixes one problem but at the expense of moving the problem someplace else. |
|
 |
monsters_inc
Joined: 18 Jan 2010
Posts: 14
|
|
| Posted: Sat Mar 20, 2010 12:48 am |
|
|
Hi, loupan
In my case this problem exist only once and for now I search generated listing for address 0x20C. |
|
|
loupan
Joined: 22 Oct 2007
Posts: 21
|
|
| Posted: Sun Mar 21, 2010 9:45 am |
|
|
Hi Monsters_inc.
I too have gotten to the point where I check the listing to make sure that all CALL and RCALL addresses are legal values that have corresponding entries in the symbol table. I have written some UNIX utilities to help automate this.
This is extremely time consuming and frustrating. |
|
|
FvM
Joined: 27 Aug 2008
Posts: 2337
Location: Germany
|
| Branch address bug still pending? |
| Posted: Sat Jun 04, 2011 9:37 am |
|
|
Hello,
do you have any indications, that the reported problem of
PCD sometimes generates arbitrary wrong branch target addresses
has been fixed in a specific compiler version? Or did you observe it in recent PCD versions?
I now have a case, where the bug can be reproduced with V4.099, but I'm not sure if it's gone with newer compilers.
As the problem is not listed in the "recent changes", a pessimistic assumption would be that the bug has never been recognized by CCS. Fortunately (?), many bugs have been silently removed.
Best regards,
Frank
P.S.: I just realized, that loupan reported wrong addresses for the PCH compiler. So it's either a different bug, or a very general one, affecting even different hardware platforms. |
|
|
|
FAQ 
Search
Register
Profile 
Log in to check your private messages 
Log in
