You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Problem discovered on Linux aarch64 (was not present on armv7l). Thanks to @StephaneStH for diagnostic and fix for this problem. Our firmware monitors all threads and asserts if anyone of them stops looping. The problem is reproduced with a SRT listener in listening state while system time is set in the past: the pthread_cond_timedwait() scheduled to timeout every 1 sec. will never timeout. This is probably not critical to most systems but it shows a design flaw. The fix that will be submitted in a pull request is to use clock_gettime(CLOCK_MONOTONIC, &timeout); instead of gettimeofday(&now,0).
The text was updated successfully, but these errors were encountered:
Problem discovered on Linux aarch64 (was not present on armv7l). Thanks to @StephaneStH for diagnostic and fix for this problem. Our firmware monitors all threads and asserts if anyone of them stops looping. The problem is reproduced with a SRT listener in listening state while system time is set in the past: the pthread_cond_timedwait() scheduled to timeout every 1 sec. will never timeout. This is probably not critical to most systems but it shows a design flaw. The fix that will be submitted in a pull request is to use clock_gettime(CLOCK_MONOTONIC, &timeout); instead of gettimeofday(&now,0).
The text was updated successfully, but these errors were encountered: