Finalize earliest deadline scheduler
Merge request reports
Activity
requested review from @oscgu95
assigned to @simbj106
Should probably write some automatic tests as well. I have done a couple of manual tests (direct FIR and direct IIR 1 & 2) and they look fine. What do you think? @oscgu95
It is possible to aim this MR to another branch than main (the "previous" commit/MR). In that way things will become a bit easier later (like not having 34 commits in this MR, but only the "additional" ones"). I am not fully sure how it works though, but we can talk to more knowledgeable people (like Mikael and Frans).
1688 1688 The edited schedule. 1689 1689 """ 1690 1690 if not QApplication.instance(): 1691 QApplication.setAttribute(Qt.AA_EnableHighDpiScaling) From what I've found high DPI scaling is default on QT6. https://forum.qt.io/topic/141511/qt_enable_highdpi_scaling-has-no-effect/2
399 386 """The start times of the operations in the schedule.""" 400 387 return self._start_times 401 388 389 @start_times.setter 390 def start_times(self, start_times: dict[GraphID, int]) -> None: 391 if not isinstance(start_times, dict[GraphID, int]): changed this line in version 2 of the diff
added 1 commit
- 23f06d33 - added tests, did some bug fixes and added .coveragerc file for setting up...
47 73 # handle the remaining sets 48 74 for outports in prec_list[2:]: 49 75 for outport in outports: 50 76 operation = outport.operation 51 if operation.graph_id not in sched._start_times: 77 if operation.graph_id not in schedule.start_times: 52 78 op_start_time = 0 53 79 for current_input in operation.inputs: 54 if len(current_input.signals) != 1: 55 raise ValueError( 56 "Error in scheduling, dangling input port detected." 57 ) 58 if current_input.signals[0].source is None: 59 raise ValueError( 60 "Error in scheduling, signal with no source detected." 61 ) - .coveragerc 0 → 100644
added 11 commits
-
23f06d33...b8f2960e - 2 commits from branch
add-direct-form-iir-sfg-gen
- 87bf24d6 - redid schedule/scheduler interface to new standard
- c9c9e1bc - tests now working
- 4303787c - updated examples to not fail
- d5935e9d - added docstrings for scheduler
- 87268b64 - seems like its working for execution time 1
- f36057d7 - there is a bug that can occur when an op is scheduled before predecessors
- 5ec071d9 - solved previous bug, now possible to schedule for different execution times,...
- 12402efb - fixed start_times setter
- fd3ffdf6 - added tests, did some bug fixes and added .coveragerc file for setting up...
Toggle commit list-
23f06d33...b8f2960e - 2 commits from branch
added 10 commits
- f505a53c - redid schedule/scheduler interface to new standard
- 3b2d67c2 - tests now working
- 11845eb0 - updated examples to not fail
- 823bb689 - added docstrings for scheduler
- bc0326a1 - seems like its working for execution time 1
- 081823ce - there is a bug that can occur when an op is scheduled before predecessors
- d1cde303 - solved previous bug, now possible to schedule for different execution times,...
- 012aaf31 - fixed start_times setter
- 23f06d33 - added tests, did some bug fixes and added .coveragerc file for setting up...
- c45a6f72 - Merge branch 'finalize-earliest-deadline-first' of gitlab.liu.se:da/B-ASIC...
Toggle commit listadded 1 commit
- 7d6742eb - fixed bug where pyside2 was imported even though pyside6 used
added 1 commit
- 8a38ddc0 - updated compile.py to hopefully work with pyside6 and pyqt6, also recompiled
added 1 commit
- 127a5687 - recompiled with pyside6, will hopefully work with pyqt6 also
Now the pipeline actually passes when running pyside6 and pyqt6.
Edited by Simon Bjurekadded 3 commits
-
127a5687...826625f3 - 2 commits from branch
master
- 33806844 - Merge branch 'master' of gitlab.liu.se:da/B-ASIC into finalize-earliest-deadline-first
-
127a5687...826625f3 - 2 commits from branch
enabled an automatic merge when all merge checks for 8b68ff9e pass