zampolli [Wed, 17 Feb 2010 10:28:24 +0000 (10:28 +0000)]
Fix for bug 62546 - when lock is on, getting an entry via AliCDBId checks whether the set run (fRun) is
inside the run range of the AliCDBId itself, not only if fRun is equal to the lower limit of
the run range. Same for the case when the cache is active.
morsch [Tue, 16 Feb 2010 19:40:17 +0000 (19:40 +0000)]
I committed changes in setting of the magnetic field
in geant4_vmc a few days ago.
While before there were hardwired the default Geant4
stepper, type of equation of motion and all tunable
parameters, user has now a possibility to change these
from a macro.
Also now the Geant4 field is created only if
user has set the field to gMC before calling gMC->Init().
In this way we avoid creating the field for applications
with no field.
This last change requires to swap the calls in AliSimulation:
gAlice->GetMCApp()->Init();
gMC->SetMagField(TGeoGlobalMagField::Instance()->GetField());
to:
gMC->SetMagField(TGeoGlobalMagField::Instance()->GetField());
gAlice->GetMCApp()->Init();
laphecet [Tue, 16 Feb 2010 10:07:13 +0000 (10:07 +0000)]
Changing once more (hopefully we get it correct this time...) the logic to trig the upload of tracker configuration) : the shuttle preprocessor is now in charge of deciding whether there was a change or not. It is the only process able to do this correctly anyway
laphecet [Mon, 15 Feb 2010 17:38:37 +0000 (17:38 +0000)]
Ped and Gain DAs update.
MUONTRKPEDda.ccx : messages coming from event decoding written in DA_PED log file
MUONTRKGAINda.cxx : List of the 11 runs needed to validate calib. DA (OnLine) displayed in the header source file
AliMUONTrackerIO.cxx : bug correction
READMAmchda.txt : README update of the DAs
richterm [Mon, 15 Feb 2010 10:37:29 +0000 (10:37 +0000)]
final version of unit test for the AliHLTDataBuffer
- allowing buffer size bigger than default page size
- implementing the SetSize for the AliHLTDataBuffer test case
richterm [Mon, 15 Feb 2010 10:34:49 +0000 (10:34 +0000)]
Important change in the AliHLTSystem memory handling, activating memory pager
- final corrections in memory pager
- dynamic page size according to requested size
- replacing current buffer handling with memory pager
This reduces the size of allocated memory for the default HLT chain by a factor of 10 and
avoids explosion of memory consumption for more advanced chains.
richterm [Mon, 15 Feb 2010 08:15:27 +0000 (08:15 +0000)]
maintenace of the extract-ddlraw macro
- new parameter added to specify max number of events
- connect to Grid if input is a remote file
- use one single publisher for all links
- skipping of CTP initialization in the HLTReconstructor in order to avoid CDB access
richterm [Mon, 15 Feb 2010 06:37:38 +0000 (06:37 +0000)]
update HLT of TRD macros (Ttheo)
- read custom parameters from CDB entry first, in order to not hard code
strings that should be read from OCDB
- use standard parameters as default
richterm [Mon, 15 Feb 2010 06:35:58 +0000 (06:35 +0000)]
update of HLTTRDClusterFinder configuration object: unify system of the input multiplier for clusterizer and tracker and
decrease the input multiplier of the clusterizer from 7 to 4 (Theo)
sgorbuno [Fri, 12 Feb 2010 15:05:23 +0000 (15:05 +0000)]
hange of the interfaces.
Now the GlobalVertexer can read tracks directly from the ITS/TPC trackers and can be called before the ESDConverter
The change is backward-compatible, the vertexer can be also called after the ESDConverter, as it was before.
The GlobalVertexerHisto component can read both the ESD and the GlobalVertexer output.
The reason for this change is very low speed of the packing/unpacking the AliESDEvent object.
The GlobalVertexer speeded up in >10 times after this change.
hdalsgaa [Fri, 12 Feb 2010 11:13:18 +0000 (11:13 +0000)]
Upgrade following the 900 GeV data. A lot of extra things added - histograms etc however core components are essentially unchanged. I have also tried to fix bug number 60424 in this update