moved TCScheduling Service to use timeval internally #735

Open
mohr wants to merge 7 commits from mohr/timetag_fix into development
Owner

Compiler was warning about comparisons between signed and unsigned so I just used timeval internally to be safe (and exact).

In the future we could think about moving to 64 bit in the TC as well...

Compiler was warning about comparisons between signed and unsigned so I just used timeval internally to be safe (and exact). In the future we could think about moving to 64 bit in the TC as well...
mohr added 1 commit 2023-02-07 14:30:27 +01:00
moved TCScheduling Service to use timeval internally
Some checks failed
fsfw/fsfw/pipeline/head There was a failure building this commit
fsfw/fsfw/pipeline/pr-development There was a failure building this commit
61620bfce7
mohr added 1 commit 2023-02-09 12:07:52 +01:00
Merge branch 'development' into mohr/timetag_fix
Some checks failed
fsfw/fsfw/pipeline/pr-development There was a failure building this commit
9a09cadb4f
mohr added 1 commit 2023-02-09 13:20:39 +01:00
missed one
Some checks failed
fsfw/fsfw/pipeline/pr-development There was a failure building this commit
319199cb22
muellerr requested changes 2023-02-10 13:59:43 +01:00
@ -163,1 +168,3 @@
if (timestamp - tNow.tv_sec <= RELEASE_TIME_MARGIN_SECONDS) {
timeval timeDifference = scheduledTime - tNow;
// round subseconds up
if (timeDifference.tv_sec + 1 <= RELEASE_TIME_MARGIN_SECONDS) {
Owner

Is timeval guaranteed to be signed?
This also rounds up when subsecond micros are 0, is this intended?

Is timeval guaranteed to be signed? This also rounds up when subsecond micros are 0, is this intended?
Author
Owner
  1. Yes, timeval is using time_t in unixoides and long in windows.
  2. Admittedly I was lazy, I will change it so that 0 does not round up.
1) Yes, timeval is using time_t in unixoides and long in windows. 2) Admittedly I was lazy, I will change it so that 0 does not round up.
Owner

Can we use timestamp <= tNow.tv_sec + RELEASE_TIME_MARGIN_SECONDS compare logic instead?

That is a bit more readable in my opinion.

Can we use `timestamp <= tNow.tv_sec + RELEASE_TIME_MARGIN_SECONDS` compare logic instead? That is a bit more readable in my opinion.
Author
Owner

Changed rounding to be correct.
The unreadable comparison is removed with this PR.

Changed rounding to be correct. The unreadable comparison is removed with this PR.
mohr added 3 commits 2023-05-08 15:16:18 +02:00
CHANGELOG
Some checks failed
fsfw/fsfw/pipeline/pr-development There was a failure building this commit
5a35f8e50b
mohr added 1 commit 2023-05-08 15:30:00 +02:00
Merge branch 'development' into mohr/timetag_fix
Some checks failed
fsfw/fsfw/pipeline/pr-development There was a failure building this commit
027d309a7a
muellerr added this to the vNext milestone 2023-09-01 12:21:48 +02:00
Some checks failed
fsfw/fsfw/pipeline/pr-development There was a failure building this commit
This pull request has changes conflicting with the target branch.
  • CHANGELOG.md

Checkout

From your project repository, check out a new branch and test the changes.
git fetch -u origin mohr/timetag_fix:mohr/timetag_fix
git checkout mohr/timetag_fix
Sign in to join this conversation.
No description provided.