Hi Daniel,
Thanks for your example, I will try to do it right:). But I'm still
getting the similar core dump pointing to url.c, even after replacing
the realloc by myrealloc. Here is the core dbx stack trace:

Type 'help' for help.
reading symbolic information ...
[using memory image in core]

Segmentation fault in ConnectionKillOne at line 1678 in file "" ($t1)
couldn't read "url.c"
(dbx) where
ConnectionKillOne(data = 0x2162262c), line 1678 in "url.c"
Curl_close(data = 0x2162262c), line 206 in "url.c"
curl_easy_cleanup(0x2162262c), line 394 in "easy.c"
getMMValueSql(0x21383d9d), line 498 in "das.c"

Cant think of anything else. Thanks for your help.
-Sachin

On Tue, 2 Nov 2004 08:29:08 +0100 (CET), Daniel Stenberg
<daniel-curl_at_haxx.se> wrote:
> On Mon, 1 Nov 2004, Sachin Mahajan wrote:
>
> > I think you are right. Is there any alternative?
>
> Yes, the alternative to doing it wrong is doing it right! ;-)
>
> Like my updated example:
>
> http://curl.haxx.se/lxr/source/docs/examples/getinmemory.c
>
> > This happens when I try to upload large pdf files(not more then 600k
> > though). Is there any workaround for the cleanup(realloc)
>
> The crash in the cleanup seems to happen because the memory list is destroyed
> when free() is called.
>
> --
>
>
> Daniel Stenberg -- http://curl.haxx.se -- http://daniel.haxx.se
> Dedicated custom curl help for hire: http://haxx.se/curl.html
>
Received on 2004-11-02