Messages in this thread | | | Date | Mon, 3 May 1999 00:54:29 +0200 (CEST) | From | Andrea Arcangeli <> | Subject | Re: question about scheduler 2.0 -> 2.2 |
| |
On Sun, 2 May 1999, R. Westbrock wrote:
>1. former member timeout in task_struct in 2.1 seems to be an extern >parameter to calls now >2. there has been added a function schedule_timeout
Some time ago I noticed that tsk->timeout was misused in the kernel because we was supposing that tsk->timeout == 0 would mean no-timeout, while instead it means that jiffy is near to wrap around.
This was the main reason why I implemented schedule_timeout() (don't forget that it's been Ingo the first who mentioned the schedule_timeout idea).
The nice side effect is that we taken out from schedule() some code from a fast path and now the additional code is executed only on demand.
We also removed a unsigned long from the task struct and now we put it on the stack only when needed.
Also the code generated by schedule_timeout() is cleaner according to me (compared to the old tsk->timeout field).
I posted all the relevant information about schedule_timeout() on the list at the time of the kenrel inclusion.
Andrea Arcangeli
- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu Please read the FAQ at http://www.tux.org/lkml/
| |