|
There
is no way that any compiler out there would fail in this respect. If it did, it
wouldn't be usable. This is a completely core tenet of the C++ language, so
there's no need to work around it.
--------------
Dean Roddey
Software Geek
Extraordinaire
Portal, Inc
[EMAIL PROTECTED]
It is guaranteed by C++ standard, not by platfom compiler
itself :(, especially for Solaris.
So, if (fCurReader) { ... } will not do much
harm.
----- Original Message -----
Sent: Saturday, April 21, 2001 4:07
AM
Subject: Re: Bug in ReaderMgr::reset ()
?
No, it is guaranteed by the
language that the deletion of a zeroed pointer
will be
legal.
--------------------------
Dean Roddey
The CIDLib C++
Frameworks
Charmed Quark Software
[EMAIL PROTECTED]
http://www.charmedquark.com
"Why
put off until tomorrow what you can
put off until the day after
tomorrow?"
----- Original Message -----
From: "Alexander Zeyliger"
<[EMAIL PROTECTED]>
To:
<[EMAIL PROTECTED]>
Sent:
Friday, April 20, 2001 5:59 PM
Subject: Bug in ReaderMgr::reset ()
?
>
>
> Hello,
>
> I've been getting
some crashes in the library (1.4.0, Solaris 8)
> in the
ReadeMgr::reset call. By looking at the code (addmittedly
> not
very thouroughly), I think there is a chance that this
> method is
called twice in a row (from XMLScan, when certain exceptions
>
happen). However, there is a couple of lines there:
>
>
delete fCurReader;
> fCurReader = 0;
>
> which may
definitely cause problems if called twice.
> Shouldn't it be if
(fCurReader) { ... } ?
>
> Thanks,
>
AZ
>
>
>
---------------------------------------------------------------------
>
To unsubscribe, e-mail: [EMAIL PROTECTED]
>
For additional commands, e-mail: [EMAIL PROTECTED]
>
---------------------------------------------------------------------
To
unsubscribe, e-mail: [EMAIL PROTECTED]
For
additional commands, e-mail: [EMAIL PROTECTED]
|
