1 #+(or (not sb-thread
) win32
) (invoke-restart 'run-tests
::skip-file
)
3 ;;; Not an exactly an "exit hang" test, but there was a different hang
4 ;;; regarding concurrent JOIN-THREAD on 1 thread.
5 ;;; Even though POSIX threads would consider this to be undefined behavior with
6 ;;; its thread abstraction, it's not undefined behavior in SBCL (for now)
7 ;;; though I do think it's slightly suspicious to depend on this.
8 (with-test (:name
:concurrent-join-thread
)
9 (let* ((other-guy (sb-thread:make-thread
#'sleep
:arguments
.2 :name
"sleepyhead"))
12 collect
(sb-thread:make-thread
#'sb-thread
:join-thread
13 :arguments other-guy
))))
14 ;; The joiners should all return
15 (mapc 'sb-thread
:join-thread joiners
)))
17 ;;; This uses the same C source file as fcb-threads.
18 ;;; This is OK in the parallel test runner because WITH-SCRATCH-FILE
19 ;;; includes the PID in the temp file name.
20 (if (probe-file "fcb-threads.so")
21 ;; Assume the test automator built this for us
22 (load-shared-object (truename "fcb-threads.so"))
23 ;; Otherwise, write into /tmp so that we never fail to rebuild
24 ;; the '.so' if it gets changed, and assume that it's OK to
25 ;; delete a mapped file (which it is for *nix).
26 (with-scratch-file (solib "so")
27 (sb-ext:run-program
"/bin/sh"
28 `("run-compiler.sh" "-sbcl-pic" "-sbcl-shared"
29 "-o" ,solib
"fcb-threads.c"))
30 (sb-alien:load-shared-object solib
)))
31 #+(and linux gc-stress
) (invoke-restart 'run-tests
::skip-file
)
32 ;;; Final test: EXIT does not lock up due to (simulated) C++ destructors
33 ;;; or free() or most anything else involved in stopping the main thread.
34 ;;; The point of the test is to mock a Lisp thread that uses foreign code
35 ;;; that uses malloc and free or equivalent from C++.
36 ;;; The behavior being tested is the effect of SB-THREAD:ABORT-THREAD on
37 ;;; a thread that happened to be just at that moment in the foreign code.
38 ;;; We can't - or don't need to - exactly replicate the behavior
39 ;;; of doing a lot of memory allocation. All we need to demonstrate is
40 ;;; that we won't interrupt a malloc() or free().
41 (defglobal *should-i-keep-going
* t
)
42 (defun mess-around-with-foreign-calls ()
43 ;; In reality the thread would not permanently own the lock, but this is the
44 ;; simplest way to simulate the random occurrence that it does own the lock
45 ;; exactly when terminated.
46 ;; So make it own the lock forever unless politely (i.e. not forcibly) terminated.
47 (alien-funcall (extern-alien "acquire_a_global_lock" (function void
)))
48 (loop (sb-thread:barrier
(:read
))
49 (unless *should-i-keep-going
* (return))
51 (format *error-output
* "~&Worker thread politely exiting~%")
52 (alien-funcall (extern-alien "release_a_global_lock" (function void
))))
54 (sb-thread:make-thread
#'mess-around-with-foreign-calls
)
58 (format t
"~&Invoked exit hook~%")
59 (setq *should-i-keep-going
* nil
)))
62 ;;; The actual code under test involved C++ destructors that are
63 ;;; interposed between our call to OS-EXIT and the OS call per se.
64 ;;; Acquiring a globally shared mutex in OS-EXIT simulates that.
67 'simulate-c
++-destructors
68 (lambda (realfun code
&key abort
)
69 (format t
"~&Enter OS-EXIT ~s ~s~%" code abort
)
70 (alien-funcall (extern-alien "acquire_a_global_lock" (function void
)))
71 (alien-funcall (extern-alien "release_a_global_lock" (function void
)))
72 (funcall realfun code
:abort abort
)))
74 ;;; Give ourselves 3 seconds to exit.
75 (alien-funcall (extern-alien "prepare_exit_test" (function void int
)) 3)
76 (setq sb-ext
:*forcibly-terminate-threads-on-exit
* nil
)