Sebastian Hahn
32220d38c0
Ensure worker threads actually exit when it is time
...
This includes a small refactoring to use a new enum (workqueue_reply_t)
for the return values instead of just ints.
2015-08-21 10:36:53 -04:00
Nick Mathewson
f724b2e5aa
Merge remote-tracking branch 'public/bug16741_026'
2015-08-17 14:40:27 -04:00
Nick Mathewson
720a9ccb2f
Check for EINTR correctly on windows
...
(even though these are nonblocking calls and EINTR shouldn't be possible).
Also, log what error we're seing if drain_fn fails.
2015-08-07 09:12:33 -04:00
cypherpunks
b3ea3c8e2f
Switch order of unblocking threads and releasing the mutex.
...
According to POSIX, the mutex must be locked by the thread calling the signal
functions to ensure predictable scheduling behavior.
Found the issue using Helgrind which gave the warning `dubious: associated lock
is not held by any thread`.
2015-08-04 13:35:02 -04:00
cypherpunks
3b3b447f75
Fix some potential memory leaks in the thread pool code.
2015-07-21 13:57:53 -04:00
Nick Mathewson
a194385d56
Impose an upper limit on threads per threadpool.
...
Found by Coverity; Fixes CID 1268069
2015-05-28 12:24:29 -04:00
Nick Mathewson
0b46b08225
Check thread count for negative; realloc->reallocarray
...
CID 1268069
2015-02-17 08:46:11 -05:00
Nick Mathewson
5d2a23397a
Fix a few coverity "Use after NULL check" warnings
...
Also remove the unit test mocks that allowed get_options() to be
NULL; that's an invariant violation for get_options().
2015-02-16 15:40:15 -05:00
Sebastian Hahn
424edd5710
Don't leak a cond var when starting threads in a pool
2015-02-15 11:21:54 +01:00
Nick Mathewson
38b3f9a619
use the correct free fn. spotted by dgoulet
2015-01-21 14:54:38 -05:00
Nick Mathewson
3c8dabf69a
Fix up some workqueue/threading issues spotted by dgoulet.
2015-01-21 12:26:41 -05:00
Nick Mathewson
a52e549124
Update workqueue implementation to use a single queue for the work
...
Previously I used one queue per worker; now I use one queue for
everyone. The "broadcast" code is gone, replaced with an idempotent
'update' operation.
2015-01-15 11:05:22 -05:00
Nick Mathewson
051ad788e0
Incorporate some comments based on notes from dgoulet
2015-01-14 11:31:14 -05:00
Nick Mathewson
1e896214e7
Refactor cpuworker to use workqueue/threadpool code.
2015-01-14 11:23:34 -05:00
Nick Mathewson
e5f8c772f4
Test and fix workqueue_entry_cancel().
2015-01-14 11:17:46 -05:00
Nick Mathewson
c51f7c23e3
Test a little more of compat_threads.c
2015-01-14 11:05:56 -05:00
Nick Mathewson
b2db3fb462
Documentation for new workqueue and condition and locking stuff
2015-01-14 11:05:54 -05:00
Nick Mathewson
4abbf13f99
Add a way to tell all threads to do something.
2015-01-14 11:01:21 -05:00
Nick Mathewson
51bc0e7f3d
Isolate the "socketpair or a pipe" logic for alerting main thread
...
This way we can use the linux eventfd extension where available.
Using EVFILT_USER on the BSDs will be a teeny bit trickier, and will
require libevent hacking.
2015-01-14 11:01:19 -05:00
Nick Mathewson
c7eebe237d
Make pending work cancellable.
2015-01-14 10:56:27 -05:00
Nick Mathewson
a82604b526
Initial workqueue implemention, with a simple test.
...
It seems to be working, but more tuning is needed.
2015-01-14 10:56:24 -05:00
