Allow PST insertion in any order #46

New Issue

muellerr · 2020-04-19T21:25:42+02:00

muellerr commented

2020-04-19 21:25:42 +02:00

Right now, device handler slots must be inserted in the correct order.
Using a std::multiset instead of a std::list, insertion in any order would be possible. I think this would be a nice feature because it is convenient when the PST is not generated automatically and I suspect the overhead will be marginal (I can measure it if someone has concerns). The list stored pointers, but I think for a STL container its ok to just store the class entry, which also simplifies the syntax for iteration over the list.

The < operator was overloaded for the FixedSequenceSlot to allow this.

See pull request #24 , which also proposes a possible fix for the FreeRTOS sanity check error for slotTime = 0

Label: Enhacement

Changes to existing code:
There should be no changes, and PST entries should be able to be inserted in any order for FreeRTOS

Right now, device handler slots must be inserted in the correct order. Using a std::multiset instead of a std::list, insertion in any order would be possible. I think this would be a nice feature because it is convenient when the PST is not generated automatically and I suspect the overhead will be marginal (I can measure it if someone has concerns). The list stored pointers, but I think for a STL container its ok to just store the class entry, which also simplifies the syntax for iteration over the list. The < operator was overloaded for the FixedSequenceSlot to allow this. See pull request #24 , which also proposes a possible fix for the FreeRTOS sanity check error for slotTime = 0 Label: Enhacement Changes to existing code: There should be no changes, and PST entries should be able to be inserted in any order for FreeRTOS

muellerr referenced a pull request that will close this issue

2020-04-20 14:34:36 +02:00

all fixed slot sequence improvenements #24

muellerr added the

feature

label 2020-04-22 19:10:13 +02:00

gaisser closed this issue

2020-07-02 14:28:54 +02:00

Sign in to join this conversation.