A user is reporting getting messages of the form:
Error in <TBranch::GetBasket>: File: test_input.root at byte:0, branch:spillserver.gpserror, entry:3987, badread=0, nerrors=1, basketnumber=1
when processing a file:
$ file test_input.root
test_input.root: ROOT file Version 52000 (Compression: 1)
using the script:
#include "Riostream.h"
#include "TChain.h"
void readtest(){
TChain* chain = new TChain("ntuple"); chain->Add("test_input.root");
// Long64_t nentries = chain->GetEntriesFast();
int nent = chain->GetEntries();
Long64_t nbytes = 0, nb = 0;
for (Long64_t jentry=0; jentry<nent;jentry++) {
nb = chain->GetEntry(jentry); nbytes += nb; if (jentry%1000==0) cout<<jentry<<"/"<<nent<<endl;}
using ROOT build from SVN "trunk" as of last night. It was built on:
$ uname -a
Linux minos12.fnal.gov 2.6.9-89.0.20.ELsmp #1 SMP Tue Feb 2 20:15:12
CST 2010 i686 i686 i386 GNU/Linux
$ cat /etc/redhat-release
Scientific Linux Fermi LTS release 4.4 (Wilson)
and running on that machine is fine. But running the _exact_ same ROOT binary & libraries on:
$ uname -a
Linux minos27.fnal.gov 2.6.9-89.0.20.ELsmp #1 SMP Tue Feb 2 14:13:40
CST 2010 x86_64 x86_64 x86_64 GNU/Linux
$ cat /etc/redhat-release
Scientific Linux SL release 4.7 (Beryllium)
This implies some kind of skew due to running on a 64-bit kernel node (relative to having built on a 32-bit one). Possibly an uninitialized value somewhere?
I've tried a v5.26/00b build of ROOT (again built on 32-bit kernel) and don't have issues when run on either platform.
The script and test file can be found at:
/afs/fnal.gov/files/data/minos/d171/roottest
-robert
Robert W. Hatcher | rhatcher_at_fnal.gov 630-840-3102 FNAL CD/REX (MINOS) | MS 220, PO Box 500, Batavia IL 60510 Received on Tue Mar 23 2010 - 19:14:17 CET
This archive was generated by hypermail 2.2.0 : Wed Mar 24 2010 - 17:50:01 CET