sharplispers / lparallel Goto Github PK

The current implementation of END-KERNEL isn't sufficient - the threads are not necessarily gone when that function returns.

I'm doing something akin to

  (let (...
        (lparallel:*kernel* (lparallel:make-kernel 4)))
    (format t "Loading...~%")
    (lparallel:pmap nil 
                    (lambda (file)
                      (import-file ...))
                    ...)
    (lparallel:end-kernel :wait t))
  (save-lisp-and-die ...)

but that (sometimes) breaks with

debugger invoked on a SB-IMPL::SAVE-WITH-MULTIPLE-THREADS-ERROR in thread
#<THREAD tid=3161136 "main thread" RUNNING {1003848003}>:
  Cannot save core with multiple threads running.

  Interactive thread (of current session):
    #<THREAD tid=3161136 "main thread" RUNNING {1003848003}>

  Other thread:
    #<THREAD "lparallel" FINISHED values: NIL {1069E88B23}>
See also:
  The SBCL Manual, Node "Saving a Core Image"

What does seem to help is to use

    (map 'nil
         #'sb-thread:join-thread
         (lparallel:end-kernel :wait t))))

but I guess that should be done in lparallel proper?!

Thanks!

Hello!
I've found out that stmx integration got broken after lparallel bt2 migration (the latest functional lparallel commit was 80fc2952a074776abd343d6b5d3ab157f0e1df7a)

Does anyone have any idea on fixing that?
Thought that simply switching bt to api v2 in stmx source code would help, but it doesn't seem to be that straightforward.

minimal reproduceable example:

(deftype unsigned-fixnum () '(integer 0 #.most-positive-fixnum))

(stmx:transactional
    (defclass account ()
      ((balance :initform 0  :initarg :balance  :type unsigned-fixnum :accessor account-balance)
       (name    :initform "" :initarg :name     :type string          :reader   account-name
                :transactional nil))))

(defun future-stm ()
  (let* ((lparallel:*kernel* (lparallel:make-kernel 4))
         (acc (make-instance 'account :name "Mario rossi" :balance 1000)))
    (dotimes (i 10) (lparallel:force
                     (lparallel:future
                       (stmx:atomic
                        (when (>= (account-balance acc) 1)
                          (decf (account-balance acc) 1)
                          t))
                       )))))

that works fine before 80fc2952a074776abd343d6b5d3ab157f0e1df7a, but fails with
(FORMAT NIL "stmx:*current-thread* contains a stale value:~& found ~S~& expecting ~S" #<BT2:THREAD "main thread" {1004AFD8C3}> #<BT2:THREAD "lparallel" {10039C5CD3}>) afterwards.

@sionescu, maybe you can help here? since you're the expert in bt obviously 🥲

As this repo is ahead of lmj/lparallel, including some bug fixes, maybe it should be added to quicklisp and/or ultralisp?

Original issue: lmj#34

Report:

The documentation string of task-handler-bind claims that it works like handler-bind but it actually only accepts symbols as condition type specifiers:

(lparallel:task-handler-bind (((and error (not undefined-function)) #'continue)) …)
=> Ill-formed binding in `task-handler-bind': ((AND ERROR (NOT UNDEFINED-FUNCTION)) #'CONTINUE)

Discussion:
none