www.delorie.com/archives/browse.cgi   search  
Mail Archives: djgpp/1996/04/29/10:19:53

Message-ID: <31854B53.1C71@goliat.eik.bme.hu>
Date: Mon, 29 Apr 1996 16:05:55 -0700
From: "Dr. Andras Solyom" <solyom AT eik DOT bme DOT hu>
Organization: Dept. Atomic Phys. Technical University of Budapest
MIME-Version: 1.0
To: djgpp AT delorie DOT com
Subject: V2 vs. 1.12m5

I have written a chain of programs for processing some very big (>300M) structured data 
files using DJGPP V1.12. I had to modify them to incorporate some new features right 
now and decided to use V2. After the modifications the third program, which cuts up the 
files to words and creates an index has died with a SIGSEV error in malloc. I know from 
this forum that CWSDPMI is the cause of this error, because I am allocating and 
reallocating a lot for the hash table. So I switched back to version 1.12maint5 and 
discovered that the new (modified) programs that were running  without any error when I 
was using V2 now exit with segmentation fault except this third program which runs just 
fine with 1.12 and faults with V1.12. 

So now I am very troubled, because I cannot rewrite malloc and the virtual memory 
handling for V2, just because CWSDPMI has some errors (I need the virtual memory, 
because my machine has only 64 Mbyte RAM, while the data which I keep in the hash table 
are about 150 Mbyte...), and I cannot count on V1.12 either it seems. It is of course 
plausible, that my source  has some errors which appears in V1.12 and some other errors 
which appears only with V2.0, but why can't I see the compiler issuing any error or at 
least warning messages ? I am using -O3 and -Wall.

So my question really is this: What are the differences between the compilers in V1.12 
and in V2 that can cause different behaviour with the same source? When can I expect a 
corrected CWSDPMI, with which I can allocate and deallocate as many times I like?

-- 
**********************************************************************************
*             ,   ,                                                              *
*     Dr. Andras Solyom                       e-mail:solyom AT goliat DOT eik DOT bme DOT hu    *
*   Techn. Univ. Budapest                     Phone: (36 1) 463 1580             *
*   Dept. Atomic Physics                      Fax:   (36 1) 463 4194             *
*   H-1111 Budafoki ut 8.                                                        *
**********************************************************************************

- Raw text -


  webmaster     delorie software   privacy  
  Copyright © 2019   by DJ Delorie     Updated Jul 2019