🚀 go-pugleaf

RetroBBS NetNews Server

Inspired by RockSolid Light RIP Retro Guy

Thread View: comp.sys.ibm.pc.programmer
4 messages
4 total messages Started by Lars Sigebo Fri, 04 Sep 1992 21:35
Group DGROUP exceeds 64K
#475
Author: Lars Sigebo
Date: Fri, 04 Sep 1992 21:35
15 lines
952 bytes
I'm using Borland C++ version 3.0 in large memory model and are trying to 
link a couple of modules. I got this message from TLINK:

     Error: Group DGROUP exceeds 64K

Tools and Utilities Guide says: 
A group exceeded 64K bytes when the segments of the group wwere combined.

Any hint or workaround to solve my problem are appreciated.

Thanks in advance,
    Lars
--------------------------
Lars.Sigebo@ldc.lu.se
Lund University Computing Center, Sweden
Re: Group DGROUP exceeds 64K
#476
Author: Ron Dippold
Date: Sat, 05 Sep 1992 03:51
27 lines
1174 bytes
lars_s@gemini.ldc.lu.se (Lars Sigebo) writes:
>I'm using Borland C++ version 3.0 in large memory model and are trying to 
>link a couple of modules. I got this message from TLINK:

>     Error: Group DGROUP exceeds 64K

Move your "static" declarations in functions out to the top level of
the file, from:

main()
{
	static int big[30000];
	/* lotsa stuff */
}

to 

static int big[30000];
main()
{
	/* lotsa stuff */
}

The same thing happens under Microsloth C.  Pain in the ass, and nasty
for scoping, but Intel real mode is Intel real mode...
-- 
Q: Why did the tachyon cross the road?  A: Because it was on the other side.
Re: Group DGROUP exceeds 64K
#478
Author: Paul Penrod
Date: Sat, 05 Sep 1992 18:06
49 lines
2233 bytes
In article <rdippold.715665099@cancun> rdippold@cancun.qualcomm.com (Ron Dippold) writes:
>lars_s@gemini.ldc.lu.se (Lars Sigebo) writes:
>>I'm using Borland C++ version 3.0 in large memory model and are trying to 
>>link a couple of modules. I got this message from TLINK:
>
>>     Error: Group DGROUP exceeds 64K
>
>Move your "static" declarations in functions out to the top level of
>the file, from:
>
>main()
>{
>	static int big[30000];
>	/* lotsa stuff */
>}
>
>to 
>
>static int big[30000];
>main()
>{
>	/* lotsa stuff */
>}
>
>The same thing happens under Microsloth C.  Pain in the ass, and nasty
>for scoping, but Intel real mode is Intel real mode...
>-- 
>Q: Why did the tachyon cross the road?  A: Because it was on the other side.

One other suggestion is to declare any large structures of arrays
to be far. This will force them into their own data segment and
allow the Large Model to link without incident. Under Large Model,
far pointers are the rule, but they only operate 16 bit offsets, so
the largest structure or declaration(s) you can have must not total
more than 64K per segment . Also be aware that storage space is required for
arrays and structures so you can not legally declare something
like:

char foo[65535];

because of the storage space required to declare and utilize the
array at runtime. If you need to have larger structures, then use
Huge Model, which uses true 32 bit (normalized) pointers.

-- 
--------------------------------------------------------------------
            Bureaucracy: noun, plural - Bureaucracies.
         The process of turning energy into solid waste.
---------------------------------------------------------------------
Re: Group DGROUP exceeds 64K
#479
Author: Matthew Smith
Date: Wed, 09 Sep 1992 16:54
29 lines
1494 bytes
In article <1992Sep4.213550.13151@pollux.lu.se> lars_s@gemini.ldc.lu.se writes:
>I'm using Borland C++ version 3.0 in large memory model and are trying to 
>link a couple of modules. I got this message from TLINK:
>
>     Error: Group DGROUP exceeds 64K
>
>Tools and Utilities Guide says: 
>A group exceeded 64K bytes when the segments of the group wwere combined.
>
>Any hint or workaround to solve my problem are appreciated.

Are you using C++ specific code?  If so, you can use the -Vf switch for BCC,
or look for a switch in the IDE which moves the C++ virtual tables out of 
the data segment.  

Basically, this error means that your global variable area has exceeded 64k.
Evidently, Borland puts their virtual tables (or at least some information 
concerning them) into this segment.  

>Thanks in advance,
>    Lars
>--------------------------
>Lars.Sigebo@ldc.lu.se
>Lund University Computing Center, Sweden


Matt Smith
msmith@peruvian.utah.edu
Student - University of Utah Department of Computer Science
Thread Navigation

This is a paginated view of messages in the thread with full content displayed inline.

Messages are displayed in chronological order, with the original post highlighted in green.

Use pagination controls to navigate through all messages in large threads.

Back to All Threads