ViewVC Help
View File | Revision Log | Show Annotations | Download File
/cvs/Coro/Changes
(Generate patch)

Comparing Coro/Changes (file contents):
Revision 1.562 by root, Tue May 10 19:55:48 2011 UTC vs.
Revision 1.657 by root, Wed Jun 22 20:24:38 2016 UTC

1Revision history for Perl extension Coro. 1Revision history for Perl extension Coro.
2 2
3TODO: should explore PerlIO::coroaio (perl leaks like hell). 3TODO: should explore PerlIO::coroaio (perl leaks like hell).
4TODO: channel->maxsize(newsize)? 4TODO: channel->maxsize(newsize)?
5TODO: http://www.microsoft.com/msj/archive/s2ce.aspx 5TODO: __GCC_HAVE_DWARF2_CFI_ASM
6TODO: swap_sv, maybe add scope_swap_sv?
7TODO: croak when async_pool tries to run canceled thread?
6 8
96.5 Wed Jun 22 22:23:50 CEST 2016
10 - swap_sv swapping order was not symmetrical, causing
11 wrong swaps when swap_sv was used multiple times on the
12 same sv in the same thread.
13 - swap_sv calls can now be undone by calling it again
14 with the same variables.
15 - swap_sv calls will now be undone in async_pool threads.
16 - split Coro::Semaphore::up/adjust into separate xs functions
17 for better error reporting, at slight codesize increase.
18 - (libcoro) arm assembly support, please test and report.
19 - adjust to PL_savestack changes in perl 5.24 (adapted from
20 the debian patch, which unfortunately gets it wrong).
21
226.49 Sat Oct 17 01:40:12 CEST 2015
23 - throwing an exception to a thread waiting in
24 Coro::Handle using Coro::EV did not stop the watchers,
25 causing the next call to fail (testcase by Martin Pritchard).
26 - bump minimum perl version to 5.10.
27
286.48 Sun Oct 4 19:03:51 CEST 2015
29 - fix memory leak when Coro::Handle uses Coro::EV internally
30 (testcase by Sten Sten).
31 - update code to libev 4 API (internally, Coro still used the
32 version 3 API calls).
33 - remove unused hv_sig.
34
356.47 Sat Jul 11 03:58:20 CEST 2015
36 - swap_svs - NVs can be in the SV head beginning with 5.20.
37
386.46 Tue Jun 30 14:36:30 CEST 2015
39 - restore portability to perls with windows fork emulation
40 (patch by Petr Písař).
41
426.45 Tue Jun 30 01:40:08 CEST 2015
43 - restore portability to perl 5.16 and below.
44
456.44 Tue Jun 30 00:41:54 CEST 2015
46 - avoid segfaulting (or worse) when tracing a canceled thread.
47 - the Event module silently broke it's public hook API, causing
48 failures on perls compiled with -Duselongdouble. Requiring
49 the latest version of Event because I don't know which release
50 changed this (it's not mentioned in the ChangeLog).
51 - allow xs level enterleave hooks via CoroAPI.
52
536.43 Thu Jun 4 15:38:14 CEST 2015
54 - use stability canary.
55 - port to stableperl-5.22.0-1.001.
56 - update libecb.
57
586.42 Wed Feb 11 20:29:52 CET 2015
59 - Coro::SemaphoreSet->try did not actually work (analyzed by
60 SATO Kentaro).
61 - upgrade libecb to be C11 compliant.
62
636.41 Sat Sep 6 22:08:46 CEST 2014
64 - restore portability to perl 5.8.x.
65 - give new Coro's a valid GvHV(PL_hintgv) - this is slow and takes
66 up some memory, but fixes "use feature" and similar modules
67 when used inside a Coro.
68 - allow zero as argument to Coro::Channel to mean the same thing
69 as no argument. this works with older versions as well,
70 but wasn't legal until now.
71 - slightly better c header file detection.
72
736.39 Mon Jun 2 00:00:08 CEST 2014
74 - work around more incompatible changes in 5.20.
75
766.38 Sun Jun 1 21:54:23 CEST 2014
77 - check that perl slots actually have enough space to hold
78 interpreter variables.
79 - untested port to perl 5.19 (64 bit tmps indices) (reported
80 by Andreas König).
81 - croak when cancel is called without a thread context.
82
836.37 Tue Mar 4 13:27:33 CET 2014
84 - *sigh*, removed leftover debugging code from debugging a
85 perl bug, of all things.
86
876.36 Tue Mar 4 07:11:59 CET 2014
88 - semaphores would not clear the destroy hook when interrupted
89 by ->throw, causing segfaults or worse.
90 - ->throw on a thread waiting for a semaphore did not acquire
91 the semaphore, but also didn't wake up other waiters,
92 possibly causing a deadlock.
93 - "FATAL: $Coro::IDLE blocked itself" will now use Carp::confess
94 to report a full stacktrace, which should help find out
95 where the actual call is.
96 - "atomically" destroy data in slf_destroy, because it is
97 the right thing to do, just in case.
98 - disable ecb.h libm dependency, to compile on ancient systems
99 or under adverse conditions.
100
1016.33 Mon Nov 18 11:26:27 CET 2013
102 - do not crash when freeing padlists with holes (in 5.18).
103 - tentative SVt_BIND 5.19 port/fix.
104
1056.32 Tue Nov 5 15:35:35 CET 2013
106 - use a new algorithm to derive padlists for perl 5.18. The old
107 one could lead to 0-pointer accesses inside perl (reported
108 by Darin McBride).
109
1106.31 Thu May 9 07:39:48 CEST 2013
111 - Coro::AIO requests would crash if the thread was ready'd
112 while the request was ongoing.
113
1146.29 Wed May 8 02:55:18 CEST 2013
115 - when an on_destroy handler destructs the coro currently being
116 destructed a perl scalar could be accessed after being freed,
117 likely causing a crash.
118
1196.28 Wed Mar 6 06:58:02 CET 2013
120 - clean remnants of existing __DIE__ and __WARN__ handlers so
121 they lose their magic and will not cause segfaults later
122 (testcase by Andrey Sagulin).
123 - improved Coro::State documentation a bit.
124 - Coro::Debug::command now flushes the output.
125 - add hack detection code for x32 abi, because the braindead slugs
126 who designed that made it look exactly like x86_64 without
127 providing proper compile time symbols to test for it. as a result,
128 this detection cannot work reliably.
129 - valgrind stack registering was broken.
130 - do not rely on Time::HiRes anymore in Coro::Debug.
131
1326.23 Fri Dec 7 23:36:37 CET 2012
133 - use experimental fiber implementation on native windows
134 perls.
135 - use sizeof (void *) as multiplication factor for stack sizes,
136 to accomodate the totally braindamaged microsoft 64 bit "os".
137 - changed verifier host from win2k-ap510-32 to win7-sp516-32/64.
138 activeperl 5.16 crashes when PerlIO_define_layer is called due
139 to some bug in the perl dll, strawberry perl at least passes
140 the testsuite.
141 - implement Coro::Handle->peeraddr/host/port, for slightly
142 improved compatibility with LWP.
143 - implement 5.17 compatibility by almost blindly applying a
144 good-looking patch by Father Chrysostomos.
145 - move stack management functions into libcoro 3.
146 - libcoro version 3 "released".
147 - support magic values as timed_io_once args.
148 - recommend AnyEvent 7+ or EV 4+, also require EV
149 version 4 or newer for Coro::EV.
150
1516.10 Tue Oct 9 01:14:27 CEST 2012
152 - updated ecb.h, it had a typo that caused it to not compile on many
153 big endian systems (reported by many people).
154 - disable memory fences in ecb.h to improve portability.
155
1566.09 Sat Oct 6 23:25:02 CEST 2012
157 - Coro::EV I/O watchers were not interruptible by exceptions
158 (Coro::State::throw) (testcase by sten).
159 - ->throw now puts threads into the ready queue, as this seems to
160 be expected by existing code, and code that doesn't cope with spurious
161 wakeups needs fixing anyway.
162 - use fd -1 in mmap.
163 - cast I32 to int in error message printf.
164 - warn about broken so-called "hardened" kernels.
165
1666.08 Fri Apr 13 12:05:47 CEST 2012
167 - be more aggressive about exiting like perl does - formerly,
168 exiting from the non-main thread would not execute END blocks.
169
1706.07 Fri Nov 11 21:21:48 CET 2011
171 - work around a bug in PerlIO (setting $SIG{__WARN__} to a PVCV).
172 - update ecb.h.
173
1746.06 Mon Aug 8 23:59:48 CEST 2011
175 - cygwin unfortunately patches the stack at runtime, so we use the pthreads
176 backend, which is an order of magnitude slower. unfortunately, cygwins
177 pthread implementation isn't very complete either, so allocate the stack
178 twice just to be sure.
179 (note: cygwin also enables mymalloc, which is NOT THREADSAFE ON WINDOWS,
180 in its ithreaded perl - best recompile cygwin and use the 'w'indows
181 backend for much better performance. also disable ithreads for
182 even better performance...).
183
1846.05 Thu Aug 4 21:36:36 CEST 2011
185 - blush, condvar values would not be propagated from send to recv anymore
186 (reported by Chip Salzenberg).
187 - use exponential increase for the readline buffer length in
188 Coro::Handle. also reduce initial allocation to 1020 from 4096 bytes.
189
1906.04 Wed Aug 3 17:15:45 CEST 2011
191 - use even more efficient and more compatible condvars for
192 compatibility to AnyEvent 6.x :)
193 - more inconsequential ecb.h updates.
194
1956.03 Wed Aug 3 11:41:30 CEST 2011
196 - change how Coro patches AnyEvent condvars for compatibility to
197 AnyEvent 6.x.
198 - update ecb.h, to no longer include <pthread.h> in case WinNT.h
199 hasn't been included.
200
2016.02 Wed Jul 13 04:35:19 CEST 2011
202 - "improve portability to Gentoo" - gentoo manages to put perl variables
203 in memory areas that are farther than 2gb apart, which the jit couldn't
204 handle and barfed. now it's just a bit slower on gentoo and similar
205 systems.
206
2076.01 Sun Jul 3 12:31:14 CEST 2011
208 - workarounds are good, but the test for whether pthreads are used
209 was not good. this one should be better.
210 - check differently whether gcc generates cfi instructions itself.
211
2126.0 Wed Jun 29 19:43:35 CEST 2011
213 - INCOMPATIBLE CHANGE: unreferenced coro objects will now be
214 destroyed and cleaned up automatically (e.g. async { schedule }).
215 - implement a JIT compiler for part of the thread switch code,
216 which gives a 50% speed improvement on threaded perls, and
217 about 4% on non-threaded perls (so threaded perls now finally
218 reach about half the speed of non-threaded perls).
219 - slightly modernise Coro::Intro, add section about rouse functions.
220 - avoid DEFSV and ERRSV, giving another 10% improvement
221 in thread switching.
222 - Coro::State->is_destroyed is now called is_zombie.
223 - implement a Coro->safe_cancel method that might fail, but
224 cancels in a "safer" way if it succeeds.
225 - add preliminary support for DEBUGGING perls.
226 - get rid of two hash-accesses when initialising a new Coro - this
227 speeds up coro creation by almost a factor of two.
7 - croak when a coro that is being cancelled tries to block 228 - croak when a coro that is being cancelled tries to block
8 (e.g. while executing a guard block), instead of crashing or 229 (e.g. while executing a guard block), instead of crashing or
9 deadlocking. 230 deadlocking.
10 - implement a Coro->safe_cancel method that might fail, but 231 - use a more robust and also faster method to identify Coro::State
11 cancels in a "safer" way if it succeeds. 232 objects - speeds up everything a bit.
12 - implement Coro->cancel in XS for a 20% speed improvement, and to 233 - implement Coro->cancel in XS for a 20% speed improvement, and to
13 be able to implement mutual cancellation. 234 be able to implement mutual cancellation.
14 - speed up context switches by a percent or two by more efficiently 235 - speed up context switches by a percent or two by more efficiently
15 allocating context stack entries. 236 allocating context stack entries.
237 - implement Coro->join and Coro->on_destroy in XS for a speedup and
238 a reduction in memory use.
16 - cancelling a coro while it itself is cancelling another coro is 239 - cancelling a coro while it itself is cancelling another coro is
17 now supported and working, instead of triggering an assertion. 240 now supported and working, instead of triggering an assertion.
18 - be a bit more crash-resistant when calling (buggy) on_destroy 241 - be a bit more crash-resistant when calling (buggy) on_destroy
19 callbacks (best effort). 242 callbacks (best effort).
20 - move on_destroy into the slf_frame, to allow extension slf 243 - move on_destroy into the slf_frame, to allow extension slf
21 functions to have destructors. 244 functions to have destructors.
245 - get rid if coro refcounting - simply crash in other interpreter
246 threads by nulling the pointers on clone.
247 - simplify warn/die hook handling when loading Coro - the convoluted
248 logic seems to be no longer neccessary.
249 - use libecb instead of our own home-grown gcc hacks.
250 - document alternatives to Coro::LWP. Please use them :)
251 - work around another mindless idiotic NEEDLESS bug in openbsd/mirbsds
252 sigaltstack. Really. wine suffers from it, erlang suffers from it,
253 and it's known since at least 2006.
22 254
235.372 Wed Feb 23 06:14:30 CET 2011 2555.372 Wed Feb 23 06:14:30 CET 2011
24 - apparently mingw doesn't provide a working gettimeofday, try to 256 - apparently mingw doesn't provide a working gettimeofday, try to
25 work around that by relying on Time::HiRes (indirectly brought to 257 work around that by relying on Time::HiRes (indirectly brought to
26 my attention by Max Maischein). 258 my attention by Max Maischein).
… …
344 - ->throw is now supported on Coro::State objects. 576 - ->throw is now supported on Coro::State objects.
345 - clean up cctx creation code a bit. 577 - clean up cctx creation code a bit.
346 - entersub is actually a UNOP, not a LOGOP (not a bugfix). 578 - entersub is actually a UNOP, not a LOGOP (not a bugfix).
347 579
3484.9 Sat Nov 8 17:45:27 CET 2008 5804.9 Sat Nov 8 17:45:27 CET 2008
349 - (libcoro) did not preserve rbp with CORO_ASM (we are getting there). 581 - (libcoro) did not preserve rbp with CORO_ASM (we are getting there).
350 - (libcoro) no longer leak threads in the experimental pthread backend, 582 - (libcoro) no longer leak threads in the experimental pthread backend,
351 also speed it up considerably. 583 also speed it up considerably.
352 - (libcoro) do not rely on makecontext passing void *'s unscathed. 584 - (libcoro) do not rely on makecontext passing void *'s unscathed.
353 - fix compiletime dependencies on libcoro in the Makefile. 585 - fix compiletime dependencies on libcoro in the Makefile.
354 - cctx_count wasn't always updated properly. 586 - cctx_count wasn't always updated properly.
… …
416 - remove debugging code related to MgPV_nolen_const, also try to 648 - remove debugging code related to MgPV_nolen_const, also try to
417 make it compile with perl 5.8.6 (yes, apple apparently loves 649 make it compile with perl 5.8.6 (yes, apple apparently loves
418 outdated software). Reported by John S. 650 outdated software). Reported by John S.
419 651
4204.744 Tue Jul 8 22:06:35 CEST 2008 6524.744 Tue Jul 8 22:06:35 CEST 2008
421 - correctly provide default DIE/WARN handlers as documented. 653 - correctly provide default DIE/WARN handlers as documented.
422 - also overwrite PL_vtbl_sigelem.svt_clear, even though current 654 - also overwrite PL_vtbl_sigelem.svt_clear, even though current
423 implementations inside perl work fine for us. 655 implementations inside perl work fine for us.
424 656
4254.743 Mon Jun 16 00:21:57 CEST 2008 6574.743 Mon Jun 16 00:21:57 CEST 2008
426 - when using Coro::EV without running EV::loop it could 658 - when using Coro::EV without running EV::loop it could
… …
448 - sprinkle "no warnings" freely over everything, also suppress 680 - sprinkle "no warnings" freely over everything, also suppress
449 warnings for some other modules. 681 warnings for some other modules.
450 - fix typo in WSAEWOULDBLOCK. 682 - fix typo in WSAEWOULDBLOCK.
451 683
4524.72 Sun May 25 05:14:36 CEST 2008 6844.72 Sun May 25 05:14:36 CEST 2008
453 - tweak META.yaml a bit, unfortunately, there is no documented way 685 - tweak META.yaml a bit, unfortunately, there is no documented way
454 to have optional dependencies with CPAN. doh :( 686 to have optional dependencies with CPAN. doh :(
455 - avoid running some tests on windows because they would fail due to 687 - avoid running some tests on windows because they would fail due to
456 perl bug (broken fork, broken pipes...). 688 perl bug (broken fork, broken pipes...).
457 - work around perl on windows bugs where perl returns undocumented 689 - work around perl on windows bugs where perl returns undocumented
458 error codes for sysread, syswrite etc. by taking advantage 690 error codes for sysread, syswrite etc. by taking advantage
… …
462 - use unix domain sockets in testsuite to work around 694 - use unix domain sockets in testsuite to work around
463 common perl implementation bugs on widows (they are emulated by 695 common perl implementation bugs on widows (they are emulated by
464 tcp sockets on windows. ugh.) 696 tcp sockets on windows. ugh.)
465 697
4664.71 Sat May 24 20:01:27 CEST 2008 6984.71 Sat May 24 20:01:27 CEST 2008
467 - fix a bug in Coro::AnyEvent ("Usage: Coro::AnyEvent::_schedule()"). 699 - fix a bug in Coro::AnyEvent ("Usage: Coro::AnyEvent::_schedule()").
468 - take advantage of async name resolution of AnyEvent::Util. 700 - take advantage of async name resolution of AnyEvent::Util.
469 - work around brutal inet_aton override in Coro::LWP. 701 - work around brutal inet_aton override in Coro::LWP.
470 - take advantage of the readyhook in Coro::EV, for smoother 702 - take advantage of the readyhook in Coro::EV, for smoother
471 scheduling. 703 scheduling.
472 704
4734.7 Sun May 11 00:32:19 CEST 2008 7054.7 Sun May 11 00:32:19 CEST 2008
474 - completely reworked the Coro manpage. 706 - completely reworked the Coro manpage.
475 - added Coro::AnyEvent, generic event loop integration. 707 - added Coro::AnyEvent, generic event loop integration.
476 - implement cancel, ready and kill commands in Coro::Debug. 708 - implement cancel, ready and kill commands in Coro::Debug.
477 - document find_coro in Coro::Debug. 709 - document find_coro in Coro::Debug.
478 - incompatible change: rename has_stack to has_cctx. 710 - incompatible change: rename has_stack to has_cctx.
479 - Coro::AIO and Coro::BDB no longer force event model detection, 711 - Coro::AIO and Coro::BDB no longer force event model detection,
480 use AnyEvent::AIO and AnyEvent::BDB. 712 use AnyEvent::AIO and AnyEvent::BDB.
… …
604 836
6054.11 Thu Oct 11 02:40:24 CEST 2007 8374.11 Thu Oct 11 02:40:24 CEST 2007
606 - port to threaded perls. 838 - port to threaded perls.
607 839
6084.1 Thu Oct 11 02:38:16 CEST 2007 8404.1 Thu Oct 11 02:38:16 CEST 2007
609 - incompatible change: $SIG{__DIE__} and $SIG{__WARN__} will now 841 - incompatible change: $SIG{__DIE__} and $SIG{__WARN__} will now
610 be local to each coro (see Coro::State). 842 be local to each coro (see Coro::State).
611 - incompatible change: for very deep reasons, cede and cede_notself 843 - incompatible change: for very deep reasons, cede and cede_notself
612 cannot return anything, so nothing will be returned. 844 cannot return anything, so nothing will be returned.
613 - possibly bring back 5.10 compatibility (untested). 845 - possibly bring back 5.10 compatibility (untested).
614 - work around stupid (and wrong) warning on 5.10 :(. 846 - work around stupid (and wrong) warning on 5.10 :(.
… …
686 but stateful semantics. 918 but stateful semantics.
687 - fixed a lot of typos in Coro.pm (patch submitted by David 919 - fixed a lot of typos in Coro.pm (patch submitted by David
688 Steinbrunner, which applied flawlessly). 920 Steinbrunner, which applied flawlessly).
689 921
6903.6 Sat Apr 14 17:13:31 CEST 2007 9223.6 Sat Apr 14 17:13:31 CEST 2007
691 - added some bugfixes to get eg/myhttpd working again. 923 - added some bugfixes to get eg/myhttpd working again.
692 - added Coro::Storable for often-cede'ing freeze/thaw. 924 - added Coro::Storable for often-cede'ing freeze/thaw.
693 - try to do a clean exit when a coroutine calls exit 925 - try to do a clean exit when a coroutine calls exit
694 (EXPERIMENTAL). 926 (EXPERIMENTAL).
695 - got rid of indirect call through _coro_init. 927 - got rid of indirect call through _coro_init.
696 - updated the partly antique examples in eg/ to 928 - updated the partly antique examples in eg/ to

Diff Legend

– Removed lines
+ Added lines
< Changed lines
> Changed lines