Hi Peter, This was clearly part of my problem. Thanks! Unfortunately, it
looks like there may be more to it. I am now failing with:
Error in <TObjArray::At>: index 231652999 out of bounds (size: 4, this:
0x7ffe5f68)
Doing a backtrace with gdb I get:
Program received signal SIGSEGV, Segmentation fault.
BrTPCCluster::GetID (this=0x0)
at /home/sanders/Brahms/brat/include/BrTPCCluster.h:52
52 Int_t GetID() const {return fID;}
(gdb) backtrace
#0 BrTPCCluster::GetID (this=0x0)
at /home/sanders/Brahms/brat/include/BrTPCCluster.h:52
#1 0xe8b0d78 in BrTPCClusterFinder::FillSubClusters (this=0x10a02ba0,
npeaks=2, SubClusters=0x7ffe5f68) at BrTPCClusterFinder.cxx:898
#2 0xe8af888 in BrTPCClusterFinder::FindSubClusters (this=0x10a02ba0,
primaryCluster=0x1211dc48, subClusters=0x7ffe5f68)
at BrTPCClusterFinder.cxx:624
#3 0xe8af158 in BrTPCClusterFinder::DeconvoluteClusters (this=0x10a02ba0)
at BrTPCClusterFinder.cxx:527
#4 0xe8aca14 in BrTPCClusterFinder::Entry (this=0x10a02ba0)
at BrTPCClusterFinder.cxx:120
#5 0xe8ab884 in BrClusterFinderBase::Event (this=0x10a02ba0,
inputnode=0x120fbf90) at BrClusterFinderBase.cxx:190
#6 0xe2b6308 in MultReplay::Replay (this=0x109d3080, fnameIn=0x1210f080 "",
runnum=2336, seq0=3, numSeqTot=1, maxevnt=10000000)
at src/multReplay.cxx:285
I'm afraid I might be too far ahead of the pack switching to root 3....Here I
thought I was just catching up!
I should also correct an error in my initial posting: I found the 3.00.06
version of root on afs. When I indicated earlier that it wasn't available on
the rcas machines, I had been looking in /brahms/u/brahmslib. I also now see
that I was trying, and failing, to build brat against 2.24.04. I will try to
build against 2.25.03--hopefully this will fix my immediate problems...
...Steve
Peter H. L. Christiansen" wrote:
> Hi Steve
>
> I don't understand exactly what is wrong, but the idea with this part of
> the code was that I look in the BrDataTable where the sequences are and I
> want to find a particular sequence, but the pointer adresses are different
> because the clusters own the sequences. I use the root method :
> http://root.cern.ch/root/html/src/TCollection.cxx.html#TCollection:FindObject
> (Not the first but the second one) and overloads IsEqual() for the
> BrTPCSequence.
>
> This used to work fine for me, but maybe something changed in root 3,
> and it did.
>
> I looked at the root home page and found that :
>
> Constness
> =========
>
> We have modified the signature of many methods (in particular in the
> TObject base class) to introduce more constness in the system.
> You must change your code if your classes derive from TObject and use
> one of the modified functions such as ls(), Print(), Compare(), Hash(),
> etc. The modified TObject methods are:
>
> virtual TObject *Clone(const char *newname="") const;
> virtual Int_t Compare(const TObject *obj) const;
> virtual void DrawClass() const;
> virtual TObject *DrawClone(Option_t *option="") const;
> virtual void Dump() const;
> virtual TObject *FindObject(const TObject *obj) const;
> virtual char *GetObjectInfo(Int_t px, Int_t py) const;
> virtual ULong_t Hash() const;
> virtual void Inspect() const;
> virtual Bool_t IsEqual(const TObject *obj) const;
> virtual void ls(Option_t *option="") const;
> virtual void Print(Option_t *option="") const;
>
> This is the only source of backward incompatibility and we urge you to
> carefully check your code. Some compilers will warn you about your old
> "Hash()" hiding the new "Hash() const" but, e.g. not g++. So if for
> example
> none of your objects are found anymore in containers than it is very
> likely
> you have to fix IsEqual(), Compare() and/or Hash().
>
> So Steve I think you should make IsEqual const (just modify it in the
> header) and check it in if it works - do we need both versions ? - so
> maybe wait with the check in.
>
> Cheers
> Peter
>
> --
> :-) --------------------------------- )-:
> |Peter H L Christiansen aka PAN @ NBI |
> |EMAIL : pchristi@nbi.dk |
> |OFFICE : Tb1 @ NBI |
> |PHONE : 353 25269 |
> |SNAIL : Hans Tavsensgade 35, 4th |
> |PHONE : 35 349336 |
> :-D --------------------------------- \-:
This archive was generated by hypermail 2b29 : Thu Mar 29 2001 - 17:29:57 EST