On Fri, Oct 03, 2025 at 09:44:28AM +0700, Bagas Sanjaya wrote:
> On Thu, Oct 02, 2025 at 05:12:28PM +0900, Byungchul Park wrote:
> > This document describes the concept and APIs of dept.
> >
> > Signed-off-by: Byungchul Park <[email protected]>
> > ---
> >  Documentation/dependency/dept.txt     | 735 ++++++++++++++++++++++++++
> >  Documentation/dependency/dept_api.txt | 117 ++++
> >  2 files changed, 852 insertions(+)
> >  create mode 100644 Documentation/dependency/dept.txt
> >  create mode 100644 Documentation/dependency/dept_api.txt
> 
> What about writing dept docs in reST (like the rest of kernel documentation)?

Sorry for late reply, but sure.  I should and will.  Thank you!

> ---- >8 ----
> diff --git a/Documentation/dependency/dept.txt 
> b/Documentation/locking/dept.rst
> similarity index 92%
> rename from Documentation/dependency/dept.txt
> rename to Documentation/locking/dept.rst
> index 5dd358b96734e6..7b90a0d95f0876 100644
> --- a/Documentation/dependency/dept.txt
> +++ b/Documentation/locking/dept.rst

However, I'm not sure if dept is about locking.  dept is about general
waits e.g. wait_for_completion(), wait_event(), and so on, rather than
just waits involved in typical locking mechanisms e.g. spin lock, mutex,
and so on.

[snip]

> > +
> > +   context X            context Y
> > +
                             /*
                              * Acquired A.
                              */
> > +                        mutex_lock A

                             /*
                              * Request something that will be handled
                              * through e.g. wq, deamon or any its own
                              * way, and then do 'complete B'.
                              */
                             request_xxx_and_complete_B();
        /*
         * The request from
         * context Y has been
         * done.  So running
         * toward 'complete B'.
         */

        /*
         * Wait for A to be
         * released, but will
         * never happen.
         */
> > +   mutex_lock A <- DEADLOCK
                             /*
                              * Wait for 'complete B' to happen, but
                              * will never happen.
                              */
> > +                        wait_for_complete B <- DEADLOCK
        /*
         * Never reachable.
         */
> > +   complete B

                             /*
                              * Never reachable.
                              */
> > +                        mutex_unlock A
> > +   mutex_unlock A
> 
> Can you explain how DEPT detects deadlock on the second example above (like
> the first one being described in "How DEPT works" section)?

Sure.  I added the explanation inline above.  Don't hesitate if you have
any questions.  Thanks.

        Byungchul
> 
> Confused...
> 
> --
> An old man doll... just what I always wanted! - Clara

Reply via email to