Hi Christian,
What I have done is attempted to set up a PPC and an Intel 686 box with
pretty much identical setups by checking out breg, brat, and brag from
the cvs. In each case I've used the --prefix=/usr/local flag on the
configure
and-- although I don't know what they do, but you seem to recommend them--
the --enable-rpath and --enable-soversion flags. (I'm not sure I used these
flags when configuring brhijing.) {Incidentally, the configure and make
steps
are quite nice...}
There is an extra character (a "d") in the hijing.dat file that one gets
from the
cvs, but this is pretty obvious.
The problem that I see occurs on both the PPC and Intel boxes. It is true
that I am running RedHat 7.1 on the Intel machine and a RedHat variant
(YellowDog Linux) on the PPC.
Running diff on the breg version of hijing1.36.f and the version that
works (for me)
does not show any differences.
I have noticed that breg copiles with -O2, whereas the version that is
working just
has -O. However, I just recompiled brhijing with -O and it doesn't seem
to have
made any difference. (I have not tried to recompile the base programs.)
Regards, Steve
Christian Holm Christensen wrote:
> Hi Steve et al,
>
> On Tue, 24 Jul 2001 10:59:41 -0500
> "Stephen J. Sanders" <ssanders@ku.edu> wrote
> concerning ": Problem with breg/hijing":
>
>> Hi, I have been trying to run Hijing using the BREG code base.
>> Unfortunately, the program enters an infinite loop in the 2nd event.
>> This occurs even changing the input parameters. I went back to the
>> code Flemming has sitting around on the rcas machines and it runs
>> fine.
>
>
> The "code base" - yiecks, such a M$ term - is exactly the same in
> breg/hijing as in obsolete/hijing.
>
> What has changed, is that most of the code is put in a shared library
> and some compiler flags may be different.
>
> Which brhijing were you executing? The one in
> /afs/rhic/opt/brahms/pro/bin? If yes, on which machine?
>
>> Using gdb on the breg code, I was able to locate the loop
>> where the code is sitting: (near line 2353 in hijing1.36.f)
>>
>> 155 CALL PYTHIA
>> JJ=MINT(31)
>> IF(JJ.NE.1) GO TO 155
>>
>> It seems clear that MINT(31) is being clobbered. However, I have
>> not been able to find any coding errors, yet.
>
>
> To me, this sounds like a compiler, linker, and or machine problem.
>
> Did you try breg/hijing and old hijing on the same machine, using the
> same compiler and so on?
>
> You could try to link brhijing statically by passing the
> --disable-shared and --enable-static option to the configure script.
>
> Did you try to compare the file hijing1.36.f with the upstream same?
> It's avaliable from
>
> http://nta0.lbl.gov/~xnwang/hijing/
>
> You'll find very few differences indeed.
>
> BTW, it seems a whole new HIJING (2.03) is avaliable.
>
>> Has anyone successfully run this (breg) code? I'm wondering if I
>> might have something set up incorrectly.
>
>
> The whole point of BREG, BRAG, BROP, BRAT(2) and so on, is that you
> don't have to set up anything, except your PATH should contain the
> directory where you install everything.
>
> Yours,
>
> Christian -----------------------------------------------------------
> Holm Christensen Phone: (+45) 35 35 96 91
> Sankt Hansgade 23, 1. th. Office: (+45) 353 25 305
> DK-2200 Copenhagen N Web: www.nbi.dk/~cholm
> Denmark Email: cholm@nbi.dk
This archive was generated by hypermail 2b30 : Tue Jul 24 2001 - 13:07:03 EDT