 |
The key was to use -ffloat-store if I want predictable results from optimized code. Hooray. See http://developer.apple.com/documentation/DeveloperTools/gcc-3.3/gcc/Optimize-Options.html |
- Anonymous
| ID | Project | Category | View Status | Date Submitted | Last Update |
|---|---|---|---|---|---|
| 0001433 | Eressea | General | public | 2008-05-25 00:20 | 2008-08-21 12:45 |
| Reporter | Enno | Assigned To | Enno | ||
| Priority | normal | Severity | minor | Reproducibility | have not tried |
| Status | closed | Resolution | fixed | ||
| Summary | 0001433: When compiling with -O3, there is a differnce in the herb-rotting code | ||||
| Description | During turn 571, there is a change in the amount of times the random number generator gets called during the ifdef HERBS_ROT block in laws.c Release build:
| ||||
| Additional Information | Optimization flags: -O3 -fno-strict-aliasing or even just -Os This does not happen when compiling with -O0 instead | ||||
| Tags | No tags attached. | ||||
| Faction | |||||
| Game | |||||
| Report | |||||
|
|
The key was to use -ffloat-store if I want predictable results from optimized code. Hooray. See http://developer.apple.com/documentation/DeveloperTools/gcc-3.3/gcc/Optimize-Options.html |
| Date Modified | Username | Field | Change |
|---|---|---|---|
| 2008-05-25 00:20 | Enno | New Issue | |
| 2008-05-25 00:20 | Enno | Status | new => assigned |
| 2008-05-25 00:20 | Enno | Assigned To | => Enno |
| 2008-05-25 00:44 | Enno | Additional Information Updated | |
| 2008-05-26 08:54 | Enno | Status | assigned => resolved |
| 2008-05-26 08:54 | Enno | Fixed in Version | => 573 |
| 2008-05-26 08:54 | Enno | Resolution | open => fixed |
| 2008-05-26 08:54 | Enno | Note Added: 0003509 | |
| 2008-08-21 12:45 | Xolgrim | Status | resolved => closed |