#14 closed defect (fixed)
Garbage in in state density when integrating several processes
| Reported by: | Juergen Reuter | Owned by: | kilian |
|---|---|---|---|
| Priority: | P4 | Milestone: | |
| Component: | core | Version: | 1.93 |
| Severity: | normal | Keywords: | wrong code |
| Cc: |
Description
When integrating a list of processes, sometimes NaNs? in expressions of the matrix element code trigger zeros in cross sections or even wrong values. Thorsten traced this down at some point to garbage appearing in the state density.
Change History (4)
comment:1 Changed 15 years ago by
| Milestone: | → v1.9x |
|---|
comment:2 Changed 15 years ago by
| Priority: | P3 → P4 |
|---|
comment:3 Changed 15 years ago by
| Resolution: | → fixed |
|---|---|
| Status: | new → closed |
Note: See
TracTickets for help on using
tickets.
This error was due to a stack overflow from the helicity state density, which was allocated as (-1:1,-1:1) while state vectors ran from -2:2 to incorporate gravitinos and gravitons. Big thanks to Christian Specki for finding this dirty bug. (tested for Linux 32 and 64 bit and MAC OS X with gfortran, g95 and ifort)