| Ticket Hash: | 9843cab968b549c3a485c6f9eaf5b487bd0601eb | ||
| Title: | LED tests in pidp8i-test broken under the incandescent lamp simulator | ||
| Status: | Closed | Type: | Code Defect |
| Severity: | Cosmetic | Priority: | High |
| Subsystem: | front panel | Resolution: | Fixed |
| Last Modified: |
2016-12-07 18:13:34 8.42 years ago |
Created: |
2016-12-04 09:15:08 8.43 years ago |
| Version Found In: | |||
| User Comments: | ||||
|
tangent added on 2016-12-04 09:15:08:
This test program manipulates the ledstatus bitfield directly to change LED states, but the incandescent lamp simulation patch makes the GPIO module get its LED state values a different way, so that the test program ends up writing to a bitfield that the GPIO module completely ignores. We could rework the lamp simulator to make use of the ledstatus bitfield again. It could do something like the switch debouncer, using this exported bitfield only as a convenient untruth masking the ugly details underneath. When a bit in this bitfield changes, the lamp simulator would merely take note of it and update its simulation accordingly. This would consolidate more of the differences between the two branches into the GPIO module, where it is more easily maintained. As a side benefit, other users of the GPIO module (like our planned Deeper Thought 2 integration) would also be able to use the lamp simulation. tangent added on 2016-12-07 18:13:34: Fixed in [7a21d3f627] | ||||