vault backup: 2025-04-30 18:03:22
This commit is contained in:
parent
82e6f69469
commit
4ff6e7d431
1 changed files with 7 additions and 0 deletions
|
@ -145,3 +145,10 @@ LTS for $n=2$:
|
||||||
|
|
||||||
We can try to implement the scheduler in the following way:
|
We can try to implement the scheduler in the following way:
|
||||||
$$A_{i}=a_{i}.B_{i} \quad B_{i}=\bar{c}_{(i \space mod \space n)+1}.C_{i} \quad C_{i}=b_{i}.D_{i} \quad D_{i}=c_{i}.A_{i}$$
|
$$A_{i}=a_{i}.B_{i} \quad B_{i}=\bar{c}_{(i \space mod \space n)+1}.C_{i} \quad C_{i}=b_{i}.D_{i} \quad D_{i}=c_{i}.A_{i}$$
|
||||||
|
- actions of kind $\bar{c}$ are needed to signal to the next process (i.e., with the next index) that it can start working
|
||||||
|
- actions of kind $c$ are needed to receive from the previous process such a signal
|
||||||
|
- such actions implement a token ring; the token is initially given to the first process: $$S=(A_{1}|D_{2}|\dots|D_{n})\setminus _{\{ c_{1}\dots cn \}}$$
|
||||||
|
|
||||||
|
Is the implementation correct? Or, in other words, $S ≈ S_{1,∅}$?
|
||||||
|
No (ci sono rimasto male anche io).
|
||||||
|
|
||||||
|
|
Loading…
Add table
Add a link
Reference in a new issue