[Gate-users] Re: gate v.3.0.0 & optical tracking

[Sebastien JAN]

Hi,

Did you observe this "optical tracking" problem with geant4.8.0.p01 ?

I need to check that with the team which developed the optical tracking 
package.

I will be back asap

Cheers
Seb

Martin L. Purschke wrote:
> Krishnamoorthy, Srilalan wrote:
> 
>>Hello Sebastien,
>> 
>>This is Srilalan here. At the gate workshop I had informed you about a possible bug with gate v.3.0.0. 
>>While installing gatev3.0.0 we accidentally happened to install geant4.8.1.p01 instead of the recommended geant4.8.0.p01. The code compiled successfully and the benchmarks turned as per the manual. But with the optical photon tracking enabled and visualization turned on the system automatically reboots. The reboot in a couple of instances were instantaneous, while in some instances the simulation runs for a few minutes before rebooting. With visualization disabled though the simulation runs the whole length without crashing. Just want to reiterate that the problem is present only with the optical photon tracking enabled. Turning 'on' the visualization does not impede the PET and SPECT benchmarks. Here is a summary of the other components we installed.
>> 
>>OS = Red Hat Scientific Linux (release 3.0.5, Linux - 2.4.21-32.0.1.EL)
>>gcc version 3.2.3
>>CLHEP 1.9.2.3
>>geant4.8.1.p01
>>root_v5.12
>> 
>>We know that currently geant4.8.1.p01 is not supported, but thought that this could be of help since in the coming weeks you are planning on releasing a newer version of gate that is compatible with geant4.8.1.p01.
>> 
>>Thanks,
>>Srilalan 
> 
> 
> Bonjour Sebastien,
> 
> let me clarify, what happens is that the gate-X11 interaction causes the
> X server to crash (the machine survives but you lose all windows,
> obviously). I have not done a systematic test, but it also happens to a
> remote X display if the X traffic is tunneled through ssh. I assume that
> there is some malformed X11 directive sent that is not handled by the
> remote X server. And it seems to be unaffected by the particular version
> of X - my laptop (where I can reproduce the crash) runs 7.1.1, others
> some 6.8.x version.
> 
> I'm currently in Shanghai at a conference and not quite set up to
> capture the particular X11 error message. I'll be back next week.
> 
> Srilalan:
> 
> when you get a chance, log in to ratcap2 as root, and type
> 
> telinit 3
> 
> that kills X and you get a console prompt. Log in as gateuser on the
> black console, and type "startx". That gives you a one-shot X session.
> Start up Gate normally to produce the error. When X crashes, you are
> back at the black console (that's what we wanted). In this mode, the X
> is not restarted and this allows us to look at the
> /var/log/XFree86.0.log file that else gets overwritten by the new X
> server that normally starts right up again.
> 
> Please make a copy of that /var/log/XFree86.0.log file in a safe place
> so we can look at what X said why it crashed.
> 
> When you are done, become root again and type telinit 5 and all is back
> to normal.
> 
> Cheers,
> 	Martin
> 
> 
>