threadx.git/test/smp/regression/threadx_smp_time_slice_test.c, branch v6.4.1_cert

Restored the missing licence headers on the ThreadX SMP regression tests

2026-09-19T16:19:41Z

109 of the 110 tracked files in test/smp/regression carry no copyright notice and no licence text at all. The one that does was added this month. Everything else in the repository -- common/src, common_smp/src, the ports, the samples -- carries the standard header, so this is an oversight in one directory rather than a deliberate exemption, and it has been there since the tests were first published. The monoprocessor tree had the same gap and was fixed the same way. The header added here is byte identical to the one in common_smp/src, including its SPDX identifier. Nothing else in these files changes: every hunk is ten added lines and none removed. There is deliberately no "Eclipse ThreadX contributors" line and no AI disclosure paragraph -- this commit contributes no code, it records the licence the files were always under. A later change that does contribute code to one of these files adds the contributor line and its own attribution at that point, which is when there is something to attribute. Verified: all eight build configurations build clean and pass 109 of 109, and the coverage figure is unmoved at 5386/5450 lines and 2807/3610 branches over the union of the eight. Assisted-by: Claude Code (Opus 5)

Released the preemption threshold before the SMP time-slice test checks

2026-09-18T05:46:43Z

threadx_smp_time_slice_test resumes eight threads at priority 31, sleeps 20 ticks and requires all eight to have run. One of them, thread 31h, is created with a preemption threshold of 30 against that priority, deliberately: it is what reaches the branch in _tx_thread_time_slice where an expired time-slice thread is replaced by a thread with preemption-threshold enabled. While the scheduler has 31h recorded as running under a threshold, _tx_thread_smp_rebalance_execute_list will not place any thread at or below that threshold on any core at all. The other seven are all at priority 31, so they can be locked out of all four cores for the whole window. Instrumented to print the scheduler's state at the failure, five failures out of five show the same thing: 31a, 31b and 31c with a run count of zero, four peers at about 110 each, 31h at about 2,250 having been given a core twice, and cores 2 and 3 idle with a null execute pointer. The test now releases 31h's threshold after the first window and gives the others a window of their own before checking. The branch the threshold exists for is taken in the first window, so no coverage is given up. Measured standalone before: 28 failures in 500 runs of default_build_coverage, 19 of trace_build, 17 of misra_build, 12 of misra_trace_build. After: 0 in 400, 200 and 800. With thread 31a left unresumed it still reports ERROR #31 and exits 1. Assisted-by: Claude Code (Opus 5)

Release ThreadX regression system

2023-04-04T09:40:54Z