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

Conference vaxaxp::vmsnotes

Title:VAX and Alpha VMS
Notice:This is a new VMSnotes, please read note 2.1
Moderator:VAXAXP::BERNARDO
Created:Wed Jan 22 1997
Last Modified:Fri Jun 06 1997
Last Successful Update:Fri Jun 06 1997
Number of topics:703
Total number of notes:3722

528.0. "Linker's internal stack" by EWBV37::KIMURA_Y (Yuriko, MCS/CSC/KC2) Fri Apr 25 1997 06:12

Hello,

Could someone explain to me what's the meaning of "internal stack"
of the Linker?

When my customer links a program, she gets the following warning message;

   %LINK-W-INTSTKOV, linker internal stack of 25 overflowed
                     in module 'module-name' file 'file-name'

The warning occured only once.
The program is written by PASCAL V5.4 and compiled and linked under
OpenVMS (Alpha) V6.2.

Here is the Explanation & User Action from help messages;

   Explanation:  The linker's internal stack overflowed because of a complex
                 calculation.

   User Action:  Submit a Software Performance Report (SPR).

She doesn't use complex data.
Since she is not able to supply the program yet, she can't submit a SPR.
But, she would like to know at least the fully explanation of the message.

Regards,

Yuriko
T.RTitleUserPersonal
Name
DateLines
528.1Not that kind of complexWIBBIN::NOYCEPulling weeds, pickin' stonesFri Apr 25 1997 11:4216
The message is saying that the compiler asked the linker to compute something
(perhaps to use as a base address, but more likely to store into memory).
The way you do that in the VMS object language is with commands that push
values onto a stack, and then perform arithmetic on those values.  The message
says that the stack overflowed -- that the compiler generated commands that
pushed 25 more things onto the stack than it has popped so far.

This is probably either a linker or a compiler bug, and should also be reported
to the Pascal project (TURRIS::PASCAL or CLT::DEC_PASCAL_BUGS).  Perhaps PASCAL
is generating an unusually complicated expression for the linker to evaluate,
but more likely it has lost track of what is on the stack (analogous to a memory
leak).  The developers will almost certainly need an example of source code that
provokes this problem.  I wouldn't be surprised if the problem goes away if you
try to cut it down.  While they would surely prefer a small example, a big one
is better than nothing.

528.2AUSS::GARSONDECcharity Program OfficeMon Apr 28 1997 04:294
    re .0
    
    You should also QAR the "Explanation" text. The use of the word
    "complex" may confuse others.
528.3EWBV37::KIMURA_YYuriko, MCS/CSC/KC2Mon Apr 28 1997 08:2218
re .1
>but more likely it has lost track of what is on the stack(analogous to a memory
>leak). The developers will almost certainly need an example of source code that

The customer could afterwards re-compile and re-link with no-warnings.
The program is so big with many modules that she has decided not to worry
about that message and I can't submit a SPR about it accordingly.


re .2
>    You should also QAR the "Explanation" text. The use of the word
>    "complex" may confuse others.

In truth, I thought "complex" as (mathmatical) data type.
I'll QAR it certainly.

Thanks!
Yuriko
528.4UTRTSC::utoras-198-48-118.uto.dec.com::JurVanDerBurgChange mode to Panic!Mon Apr 28 1997 10:305
If the link gave this error once, and then repeating the command made it
disappear then you may have a hardware problem. You can check the errorlogs.

Jur.

528.5EWBV37::KIMURA_YYuriko, MCS/CSC/KC2Thu May 01 1997 07:289
re .4
>If the link gave this error once, and then repeating the command made it
>disappear then you may have a hardware problem. You can check the errorlogs.

The customer got the message only once and it had never been seen since then.
And so, I had her check the errorlog and there were no errors.

Thanks, 
Yuriko