Gerhard Hintermayer wrote:
> ljb <lbayuk@mindspring.com> wrote in message
>
>>[...]
>>(This is for PostgreSQL-7.3.1)
>>
>>--- src/interfaces/libpgtcl/pgtclId.c.bak 2002-10-17 10:53:32.000000000 -0400
>>+++ src/interfaces/libpgtcl/pgtclId.c 2003-01-07 21:51:53.000000000 -0500
>>@@ -300,12 +300,10 @@
>> * small) amount of memory taken for the channel state representation.
>> * Note we are not leaking a socket, since libpq closed that already.
>> */
>>-#ifdef NOT_USED
>> #if TCL_MAJOR_VERSION >= 8
>>- if (connid->notifier_channel != NULL)
>>+ if (connid->notifier_channel != NULL && interp != NULL)
>> Tcl_UnregisterChannel(NULL, connid->notifier_channel);
>> #endif
>>-#endif
>>
>> /*
>> * We must use Tcl_EventuallyFree because we don't want the connid
>
>
> Your patch does indeed fix the memory leakage but seems to introduce
> another problem. Every once in a while I do get segfaults, when I
> enter the event loop after pg_disconnect-ing (well, it looks like that
> to me). I have to say, that this is the only script, where the error
> occurs. All others (~500) work with no problem at all. I was trying to
> reproduce that in a short script, but - no way. I am not even sure, if
> this is postgreSQL related. I did build a static wish with
> --enable-symbols=all and --enable-memory to track the error down. Hope
> I'll see more then.
>
> Gerhard
I found out (actually it was Joe English from TCT), that the segfault was due to
an error in tk (Bug #671330). So I see no reason, why the above patch should'nt
be applied to further releases. I have running lot's of scripts with the above
patch and not having any problem with both memory leaks or segfaults.
Gerhard
--
Gerhard Hintermayer
http://www.inode.at/g.hintermayer
