[Search for users] [Overall Top Noters] [List of all Conferences] [Download this site]

Conference clt::dec_pascal_bugs

Title:DEC Pascal Bug Reports
Notice:New kit announcement in TURRIS::Pascal conference
Moderator:TLE::GARRISON
Created:Wed Sep 09 1992
Last Modified:Fri May 30 1997
Last Successful Update:Fri Jun 06 1997
Number of topics:838
Total number of notes:3659

824.0. "problem with packed array of boolean and OPTIMIZE!" by VNABRW::KARTNER_M (HOUSTON, we have a problem) Tue Jan 28 1997 08:38

    Hi!
    
    We have got the following psacal problem with the /optimize option
    under 
    
    	Alpha Pascal 5.5 ECO2 V5.5-55-329E   VMS V6.2
    	VAX   Pascal 5.5      PASCAL V5.5-54 VMS V6.2
    
    
    Example:
    
    [OPTIMIZE] PROGRAM OPT( OUTPUT );
    PROCEDURE FILL;
    VAR
        ARR : PACKED ARRAY [1..64] OF BOOLEAN VALUE ZERO;
        II  : INTEGER VALUE 0;
    BEGIN
        FOR II := 1 TO 21 DO
        BEGIN
              ARR [II*3] := TRUE;
              WRITELN( II*3, ARR [II*3] );
        END;
    END;
    BEGIN FILL; END.
    
    TCUBF1::MM:000000 >pascal pasbug
    TCUBF1::MM:000000 >link pasbug
    
    TCUBF1::MM:000000 >run pasbug
             3  TRUE
             6  TRUE
             9  TRUE
            12  TRUE
            15  TRUE
            18  TRUE
            21  TRUE
            24  TRUE
            27  TRUE
            30  TRUE
            33 FALSE
            36 FALSE
            39 FALSE
            42 FALSE
            45 FALSE
            48 FALSE
            51 FALSE
            54 FALSE
            57 FALSE
            60 FALSE
            63 FALSE
    
    The results >= 33 are obviously wrong !!
    
    Is this a known BUG ?
    
    							thanks
    							Michael
T.RTitleUserPersonal
Name		DateLines
824.1Looks like a bug in the Alpha_VMS compiler...at leastTLE::GARRISONMegTue Jan 28 1997 10:048
Hi Michael,

We haven't seen this one before.  I was able to reproduce it on 
Alpha VMS but not VAX VMS.  Are you seeing the bug on both platforms?

Thanks,

Meg
824.2a little more infoTLE::GARRISONMegTue Jan 28 1997 10:071
BTW, this starts failing at opt level 1.
824.3Problemstatus ?ATZIS2::KARTNER_MHOUSTON, we have a problemTue Feb 18 1997 03:488
    Hi!
    
    What's the status of the problem? Will there be a patch?
    Will it be fixed in a future release?
    
    								thanks
    								Michael
824.4TLE::REAGANAll of this chaos makes perfect senseTue Feb 18 1997 07:596
    It is on our list.  We have not looked at it yet, but it is near
    the top.  Once we find and fix the problem, we will provide a network
    kit.  The fix will also be included in the next regular release after
    the network kit is created.
    
    				-John
824.5TLE::REAGANAll of this chaos makes perfect senseWed Feb 19 1997 09:267
    I've looked at this and determined that it is a bug in the code
    generator.  I've forwarded the bug report off to the GEM group
    and when a corrected code generator appears, we'll make a network
    kit.
    
    				-John
824.6GEM_BUGS #5991TLE::GARRISONMegWed Feb 19 1997 13:270
824.7TLE::REAGANAll of this chaos makes perfect senseFri Feb 28 1997 11:126
    GEM has fixed the bug, but it won't appear until we move the next
    baselevel of GEM.  We hope to have a field test compiler sometime
    in late March/early April with a release in August.  Would the
    customer like to help us field test the compiler?
    
    				-John
824.8Customer very interested in fieldtestingATZIS1::KARTNER_MHOUSTON, we have a problemTue Mar 18 1997 03:518
    Hi!
    
    Sorry for the long delay, but the problem owner at the customer site
    was out of down. They are very interested into the pascal field test
    version. Please let me know how to proceed
    
    								thanks
    								Michael
824.9TLE::REAGANAll of this chaos makes perfect senseTue Mar 18 1997 10:587
    OK, can you mail me some customer info?  Name, surface address, email
    address, etc.  Thanks.
    
    As for "when" we'll have something to test, some things have come up
    and we probably won't be testing for another month or so.
    
    				-John
824.10HAve you contacted the customer ?ATZIS1::KARTNER_MHOUSTON, we have a problemTue Apr 08 1997 04:509
    Hi!
    
    HAve you contacted the customer ?
    May I close my call?
    
    								thanks
    								Michael
824.11TLE::REAGANAll of this chaos makes perfect senseTue Apr 08 1997 14:037
    No, I haven't yet.  We're just figured out when we're going to do 
    the next release.  I hope to get the name submitted this week.
    
    As for "may I close my call", thats up to you, right? :-)
    
    				-John
824.12Off course you are right !ATZIS1::KARTNER_MHOUSTON, we have a problemWed Apr 09 1997 07:4710
    Hi John!
    
    " As for "may I close my call", thats up to you, right? :-) "
    
    This statement is correct basically. But if I am shure that the
    connection between you and the customer is established I'll
    forget the case and close the call ;->
    
    								thanks
    								Michael
824.13TLE::REAGANAll of this chaos makes perfect senseFri May 30 1997 17:003
    The T5.6-58 kit just announced contains the fix to this problem.
    
    				-John