Hi Claire,
I am afraid that this has nothing to do with a memory leak.
This message appears when you read a file with an inconsistent version
of a class. You have added a new data member or removed one and you did not
update your Streamer function accordingly. In this case when reading the I/O
buffer will be corrupted. We recommend using the CheckByteCount facility
in Root to have a better localisation of the problem. When using
teh CheckByteCount facility, if something is wrong in one class,
the corresponding Streamer function will tell you where it is wrong.
I suggest you check the versions of your classes on all the machines where you
run.
- CheckByteCount
==============
All Root Streamers include now the CheckByteCount facility.
All LinkDef files modified accordingly.
For users intending to use the same facility, modify your Linkdef file
adding a + character at the end of the class name, eg:
#pragma link C++ class TLeafB+;
The code generated for Streamers with the bytecount facility is backward
compatible with Streamers not including this code.
Rene Brun
Claire Gwenlan wrote:
>
> Hi Everybody,
>
> I was running a macro that I've run before, so I know it should work, and
> suddenly, after it read my file, I got the error...
>
> Error in <TBuffer::ReadObject>: object tag too large, I/O buffer corrupted
> Fatal in <operator new>: storage exhausted
> aborting
>
> I then tried running it on an entirely differnet computer that I don't use
> much, and things worked OK.
>
> Is this a memory leak problem? I'm not really used to C++ so, until a
> while ago I didn't know anything about memory leaks. However, I heard that
> root has automatic destructors that, when you quit root, whil delete all
> the pointers and things so memory leak is not too much of a problem. Is
> this correct or not?
>
> If I do have a memory leak, what can I do to sort it out?
>
> Thanks, Claire
>
> __________________________________
> Claire Gwenlan
> Department of Physics & Astronomy
> University College London
> Gower Street
> LONDON
> WC1E 6BT
>
> Tel: (0171) 419 3454
> _________________________________
This archive was generated by hypermail 2b29 : Tue Jan 02 2001 - 11:50:26 MET