Wow, I didn't know about box-cas! That seems to be a good solution in this case. Thanks!
-- Erich On Thu, 1 Sep 2016 09:37:06 -0300 Gustavo Massaccesi <gust...@oma.org.ar> wrote: > What about box-cas! ? > https://docs.racket-lang.org/reference/boxes.html#%28def._%28%28quote._~23~25kernel%29._box-cas%21%29%29 > > Something like: > ;--- > > #lang racket > (define handle-counter (box 1)) > (define (new-handle) > (define old (unbox handle-counter)) > (define new (add1 old)) > (unless (box-cas! handle-counter old new) > (new-handle)) > old) ; or new ? > ;--- > > Gustavo > > > > On Thu, Sep 1, 2016 at 6:44 AM, Erich Rast <er...@snafu.de> wrote: > > This question is also about style. Take an admittedly ugly > > counter like this > > > > ;; increasing counter for handles > > (define handle-counter 1) > > > > (define (new-handle) > > (begin0 > > handle-counter > > (set! handle-counter (add1 handle-counter)))) > > > > Various threads may call (new-handle) concurrently, so I need to > > ensure that every thread gets a unique handle. However, the handles > > need not reflect the temporal order of the calling sequence in any > > way. It's hard to debug but I believe I've seen a case in which the > > above code has assigned the same handle to concurrent threads, as > > one would expect. > > > > So basically new-handle needs to be in a critical section, right? I > > feel stupid that I don't know the answer, but how do I do this? > > > > In the docs I've found ffi/unsafe/atomic, but this seems like > > overkill and is unsafe and only intended for the FFI. > > > > Second question: Should I use a parameter or something else for this > > construct, and if so, what are the benefits of doing so? What would > > be the 'right' way of providing unique numbers in a synchronized > > way? > > > > Best, > > > > Erich > > > > -- > > You received this message because you are subscribed to the Google > > Groups "Racket Users" group. To unsubscribe from this group and > > stop receiving emails from it, send an email to > > racket-users+unsubscr...@googlegroups.com. For more options, visit > > https://groups.google.com/d/optout. -- You received this message because you are subscribed to the Google Groups "Racket Users" group. To unsubscribe from this group and stop receiving emails from it, send an email to racket-users+unsubscr...@googlegroups.com. For more options, visit https://groups.google.com/d/optout.
