Howdy: Just released dmalloc version 4.8.0 available via the dmalloc home page: http://dmalloc.com/releases/ This version handles a couple of issues and...
Gray Watson
gray.egroups@...
Nov 10, 2000 6:47 pm
214
Hi, i'm trying to find a memory leak in a parallel program, that uses MPI as message passing library. I'm using the LAM MPI implementation. For the first...
juergen@...
Nov 14, 2000 2:33 pm
215
Howdy: Just released dmalloc version 4.8.1 available via the dmalloc home page: http://dmalloc.com/releases/ This is a patch release to fix problems with...
Gray Watson
gray.egroups@...
Nov 15, 2000 4:03 pm
216
Hi List, I'm considering to use dmalloc and gdb to catch memory access using i386 memory protection features. Is there a know and convenient way to do it...
Vadim Fedukovich
vadim@...
Nov 16, 2000 2:24 pm
217
I changed the way the log file was written to and made the code opne the log file with append each time it wants to write to it, log entries also have the...
James Hawtin
oolon@...
Nov 20, 2000 6:18 pm
218
... Hi, thanks for changing this. In the meantime I modified my program in a way that it changes the logfile name in the environment variable DMALLOC_OPTIONS...
Juergen Wakunda
juergen@...
Nov 20, 2000 6:51 pm
221
... I'm planning in 4.9.0 to have the following expanded in the dmalloc logfile: %p - process-pid %h - hostname I'd like the do the following as well if I can...
Gray Watson
gray.egroups@...
Nov 20, 2000 9:17 pm
222
All, Dmalloc generates some return addresses to log file which gdb is telling there is no corresponding line number. I don't know why? My exe uses couple of...
ragunathd@...
Nov 21, 2000 4:23 am
223
All, I would like to have a feature in dmalloc. Let us say, i have the following function structure, getCustomerName() calls appMalloc() calls, /* Memory...
ragunathd@...
Nov 21, 2000 4:45 am
224
Ok here is the log patch again, its against 4.7.1 I have not updated it to
4.8.1 yet.
James
diff -ru dmalloc-4.7.1.ref/dmalloc.h.3 dmalloc-4.7.1/dmalloc.h.3 ...
James Hawtin
oolon@...
Nov 21, 2000 10:04 am
227
Thats kind of really difficult to do..... and quite easy.... Its really difficult to do generically, however is easy for your to do with your version..... To...
James Hawtin
oolon@...
Nov 21, 2000 10:47 am
229
I notice again the C++ debugging for dmalloc has not been enhanced. We
have had discussions about C++, and I have posted a few patches mostly the
same rework...
James Hawtin
oolon@...
Nov 22, 2000 5:39 pm
230
Hi all, I've just started using dmalloc with a program and came across a strange problem which I don't understand. It would appear I'm doing something wrong...
Raymond Wan
rwan@...
Nov 24, 2000 4:49 am
231
Check your dmalloc options one of them causes an "abort" on errors this would cause your program to jump into the memory freeing section.... Use this line to...
James Hawtin
oolon@...
Nov 24, 2000 11:12 am
232
... Register Chinese Character Domain Name- ONLY US$15/year! 100 Chinese Domain Names are registered every minute! Time is clicking, what are you waiting for? ...
hannibal
hannibal@...
Nov 27, 2000 3:57 am
233
... This is incorrect. abort() on most unix operating systems at least will bring down the program immediately without freeing memory. ... gray.reply@......
Gray Watson
gray.egroups@...
Nov 27, 2000 2:58 pm
234
... Warning bells sound. I bet that you are assuming that memory gotten from a call to malloc() has 0s in it. The Dmalloc library writes 0xc5 into allocated...
Gray Watson
gray.egroups@...
Nov 27, 2000 3:03 pm
235
%% Gray Watson <gray.egroups@...> writes: gw> The Dmalloc library writes 0xc5 into allocated memory when the gw> alloc-blank token is enabled. gw>...
Paul D. Smith
psmith@...
Nov 27, 2000 3:27 pm
236
All, What does this error mean? 975434936: 32604: free: pointer is not on block boundary: pointer '0xdd0164' from 'ra=0x1e7828' 975434936: 32604: ERROR:...
Ragunath Devarasu
ragunathd@...
Nov 28, 2000 6:46 pm
237
Pointers normally are aligned to 4 or is it 8 byte bounaries that one was obviously corrupt so dmalloc ignored it. ra=0x1e7828 was the location in memory it...
James Hawtin
oolon@...
Nov 28, 2000 6:48 pm
238
Hi Gray and everyone, Thank you all for your replies! Given the vague description of my problem, I'm surprised you were all able to spot my mistake. ... with...
Raymond Wan
rwan@...
Nov 29, 2000 6:21 am
239
Ok Dmalloc as it stands does not decode the file positions of C++ That
might explain why the pointer is comming out looking as it does...
However it here is a...
James Hawtin
oolon@...
Nov 29, 2000 9:57 am
240
To All, When I turn on the flag to check the heap I get the following error message after many (>50) successful allocations: ERROR: _chunk_check: inconsistency...
Phillip Soltan
psoltan@...
Nov 29, 2000 8:48 pm
241
Hi, I'm using g++ 2.9.5.1 with dmalloc 4.8.1 on Solaris 2.6. I'm seeing a strange problem when using the STL (stream and string). The following program: ...
Alain Champenois
achampen@...
Nov 30, 2000 2:43 am
243
... Common issue. Dmalloc is picky about freeing NULL pointers. Feel free to set the following in the settings.dist file to change the default behavior. To...
Gray Watson
gray.egroups@...
Nov 30, 2000 4:09 pm
244
... This has been reported before and may be a bug. I thought however that it had been fixed in version 4.7.2. If you are running the latest can you email to...
Gray Watson
gray.egroups@...
Nov 30, 2000 9:59 pm
245
To All, I discovered the reason for the problem I was having. I was calling malloc from an interrupt service routine and malloc is apparently not reentrant....
Phillip Soltan
psoltan@...
Dec 1, 2000 7:26 pm
246
Hi all, I have a question about memory but first, I should apologize because the question seems unrelated to dmalloc...but it could be. :) Basically, I have a...
Raymond Wan
rwan@...
Dec 4, 2000 11:47 pm
247
Newbie question, I'm compiling a simple c++ program with g++. Included the dmalloc.h file and linked with the library. I deliberately introduced memory errors...
david-m@...
Dec 10, 2000 9:45 am
250
... report ... Hi, I had a similar problem but with C. I knew it was a problem with my own code, but it was being reported in a function/line number indicated...