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

Comparing Coro/Changes (file contents):
Revision 1.534 by root, Sun Apr 11 19:23:18 2010 UTC vs.
Revision 1.640 by root, Sun Jun 7 02:01:04 2015 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: unready_all
5TODO: myhttpd header parsing
6TODO: channel->maxsize(newsize)? 4TODO: channel->maxsize(newsize)?
5TODO: __GCC_HAVE_DWARF2_CFI_ASM
6TODO: swap_sv, maybe add scope_swap_sv, andallow to unswap by repeat?
7TODO: swap_sv not undone in pool
8TODO: croak when async_pool tries to run canceled thread?
7 9
8TODO: http://www.microsoft.com/msj/archive/s2ce.aspx 106.43 Thu Jun 4 15:38:14 CEST 2015
11 - use stability canary.
12 - port to stableperl-5.22.0-1.001.
13 - update libecb.
9 14
156.42 Wed Feb 11 20:29:52 CET 2015
16 - Coro::SemaphoreSet->try did not actually work (analyzed by
17 SATO Kentaro).
18 - upgrade libecb to be C11 compliant.
19
206.41 Sat Sep 6 22:08:46 CEST 2014
21 - restore portability to perl 5.8.x.
22 - give new Coro's a valid GvHV(PL_hintgv) - this is slow and takes
23 up some memory, but fixes "use feature" and similar modules
24 when used inside a Coro.
25 - allow zero as argument to Coro::Channel to mean the same thing
26 as no argument. this works with older versions as well,
27 but wasn't legal until now.
28 - slightly better c header file detection.
29
306.39 Mon Jun 2 00:00:08 CEST 2014
31 - work around more incompatible changes in 5.20.
32
336.38 Sun Jun 1 21:54:23 CEST 2014
34 - check that perl slots actually have enough space to hold
35 interpreter variables.
36 - untested port to perl 5.19 (64 bit tmps indices) (reported
37 by Andreas König).
38 - croak when cancel is called without a thread context.
39
406.37 Tue Mar 4 13:27:33 CET 2014
41 - *sigh*, removed leftover debugging code from debugging a
42 perl bug, of all things.
43
446.36 Tue Mar 4 07:11:59 CET 2014
45 - semaphores would not clear the destroy hook when interrupted
46 by ->throw, causing segfaults or worse.
47 - ->throw on a thread waiting for a semaphore did not acquire
48 the semaphore, but also didn't wake up other waiters,
49 possibly causing a deadlock.
50 - "FATAL: $Coro::IDLE blocked itself" will now use Carp::confess
51 to report a full stacktrace, which should help find out
52 where the actual call is.
53 - "atomically" destroy data in slf_destroy, because it is
54 the right thing to do, just in case.
55 - disable ecb.h libm dependency, to compile on ancient systems
56 or under adverse conditions.
57
586.33 Mon Nov 18 11:26:27 CET 2013
59 - do not crash when freeing padlists with holes (in 5.18).
60 - tentative SVt_BIND 5.19 port/fix.
61
626.32 Tue Nov 5 15:35:35 CET 2013
63 - use a new algorithm to derive padlists for perl 5.18. The old
64 one could lead to 0-pointer accesses inside perl (reported
65 by Darin McBride).
66
676.31 Thu May 9 07:39:48 CEST 2013
68 - Coro::AIO requests would crash if the thread was ready'd
69 while the request was ongoing.
70
716.29 Wed May 8 02:55:18 CEST 2013
72 - when an on_destroy handler destructs the coro currently being
73 destructed a perl scalar could be accessed after being freed,
74 likely causing a crash.
75
766.28 Wed Mar 6 06:58:02 CET 2013
77 - clean remnants of existing __DIE__ and __WARN__ handlers so
78 they lose their magic and will not cause segfaults later
79 (testcase by Andrey Sagulin).
80 - improved Coro::State documentation a bit.
81 - Coro::Debug::command now flushes the output.
82 - add hack detection code for x32 abi, because the braindead slugs
83 who designed that made it look exactly like x86_64 without
84 providing proper compile time symbols to test for it. as a result,
85 this detection cannot work reliably.
86 - valgrind stack registering was broken.
87 - do not rely on Time::HiRes anymore in Coro::Debug.
88
896.23 Fri Dec 7 23:36:37 CET 2012
90 - use experimental fiber implementation on native windows
91 perls.
92 - use sizeof (void *) as multiplication factor for stack sizes,
93 to accomodate the totally braindamaged microsoft 64 bit "os".
94 - changed verifier host from win2k-ap510-32 to win7-sp516-32/64.
95 activeperl 5.16 crashes when PerlIO_define_layer is called due
96 to some bug in the perl dll, strawberry perl at least passes
97 the testsuite.
98 - implement Coro::Handle->peeraddr/host/port, for slightly
99 improved compatibility with LWP.
100 - implement 5.17 compatibility by almost blindly applying a
101 good-looking patch by Father Chrysostomos.
102 - move stack management functions into libcoro 3.
103 - libcoro version 3 "released".
104 - support magic values as timed_io_once args.
105 - recommend AnyEvent 7+ or EV 4+, also require EV
106 version 4 or newer for Coro::EV.
107
1086.10 Tue Oct 9 01:14:27 CEST 2012
109 - updated ecb.h, it had a typo that caused it to not compile on many
110 big endian systems (reported by many people).
111 - disable memory fences in ecb.h to improve portability.
112
1136.09 Sat Oct 6 23:25:02 CEST 2012
114 - Coro::EV I/O watchers were not interruptible by exceptions
115 (Coro::State::throw) (testcase by sten).
116 - ->throw now puts threads into the ready queue, as this seems to
117 be expected by existing code, and code that doesn't cope with spurious
118 wakeups needs fixing anyway.
119 - use fd -1 in mmap.
120 - cast I32 to int in error message printf.
121 - warn about broken so-called "hardened" kernels.
122
1236.08 Fri Apr 13 12:05:47 CEST 2012
124 - be more aggressive about exiting like perl does - formerly,
125 exiting from the non-main thread would not execute END blocks.
126
1276.07 Fri Nov 11 21:21:48 CET 2011
128 - work around a bug in PerlIO (setting $SIG{__WARN__} to a PVCV).
129 - update ecb.h.
130
1316.06 Mon Aug 8 23:59:48 CEST 2011
132 - cygwin unfortunately patches the stack at runtime, so we use the pthreads
133 backend, which is an order of magnitude slower. unfortunately, cygwins
134 pthread implementation isn't very complete either, so allocate the stack
135 twice just to be sure.
136 (note: cygwin also enables mymalloc, which is NOT THREADSAFE ON WINDOWS,
137 in its ithreaded perl - best recompile cygwin and use the 'w'indows
138 backend for much better performance. also disable ithreads for
139 even better performance...).
140
1416.05 Thu Aug 4 21:36:36 CEST 2011
142 - blush, condvar values would not be propagated from send to recv anymore
143 (reported by Chip Salzenberg).
144 - use exponential increase for the readline buffer length in
145 Coro::Handle. also reduce initial allocation to 1020 from 4096 bytes.
146
1476.04 Wed Aug 3 17:15:45 CEST 2011
148 - use even more efficient and more compatible condvars for
149 compatibility to AnyEvent 6.x :)
150 - more inconsequential ecb.h updates.
151
1526.03 Wed Aug 3 11:41:30 CEST 2011
153 - change how Coro patches AnyEvent condvars for compatibility to
154 AnyEvent 6.x.
155 - update ecb.h, to no longer include <pthread.h> in case WinNT.h
156 hasn't been included.
157
1586.02 Wed Jul 13 04:35:19 CEST 2011
159 - "improve portability to Gentoo" - gentoo manages to put perl variables
160 in memory areas that are farther than 2gb apart, which the jit couldn't
161 handle and barfed. now it's just a bit slower on gentoo and similar
162 systems.
163
1646.01 Sun Jul 3 12:31:14 CEST 2011
165 - workarounds are good, but the test for whether pthreads are used
166 was not good. this one should be better.
167 - check differently whether gcc generates cfi instructions itself.
168
1696.0 Wed Jun 29 19:43:35 CEST 2011
170 - INCOMPATIBLE CHANGE: unreferenced coro objects will now be
171 destroyed and cleaned up automatically (e.g. async { schedule }).
172 - implement a JIT compiler for part of the thread switch code,
173 which gives a 50% speed improvement on threaded perls, and
174 about 4% on non-threaded perls (so threaded perls now finally
175 reach about half the speed of non-threaded perls).
176 - slightly modernise Coro::Intro, add section about rouse functions.
177 - avoid DEFSV and ERRSV, giving another 10% improvement
178 in thread switching.
179 - Coro::State->is_destroyed is now called is_zombie.
180 - implement a Coro->safe_cancel method that might fail, but
181 cancels in a "safer" way if it succeeds.
182 - add preliminary support for DEBUGGING perls.
183 - get rid of two hash-accesses when initialising a new Coro - this
184 speeds up coro creation by almost a factor of two.
185 - croak when a coro that is being cancelled tries to block
186 (e.g. while executing a guard block), instead of crashing or
187 deadlocking.
188 - use a more robust and also faster method to identify Coro::State
189 objects - speeds up everything a bit.
190 - implement Coro->cancel in XS for a 20% speed improvement, and to
191 be able to implement mutual cancellation.
192 - speed up context switches by a percent or two by more efficiently
193 allocating context stack entries.
194 - implement Coro->join and Coro->on_destroy in XS for a speedup and
195 a reduction in memory use.
196 - cancelling a coro while it itself is cancelling another coro is
197 now supported and working, instead of triggering an assertion.
198 - be a bit more crash-resistant when calling (buggy) on_destroy
199 callbacks (best effort).
200 - move on_destroy into the slf_frame, to allow extension slf
201 functions to have destructors.
202 - get rid if coro refcounting - simply crash in other interpreter
203 threads by nulling the pointers on clone.
204 - simplify warn/die hook handling when loading Coro - the convoluted
205 logic seems to be no longer neccessary.
206 - use libecb instead of our own home-grown gcc hacks.
207 - document alternatives to Coro::LWP. Please use them :)
208 - work around another mindless idiotic NEEDLESS bug in openbsd/mirbsds
209 sigaltstack. Really. wine suffers from it, erlang suffers from it,
210 and it's known since at least 2006.
211
2125.372 Wed Feb 23 06:14:30 CET 2011
213 - apparently mingw doesn't provide a working gettimeofday, try to
214 work around that by relying on Time::HiRes (indirectly brought to
215 my attention by Max Maischein).
216 - fix some portability issues when Time::HiRes was used.
217
2185.371 Mon Feb 21 14:36:08 CET 2011
219 - backport to windows process emulation code again.
220
2215.37 Sat Feb 19 07:49:44 CET 2011
222 - add a big "Coro thread life cycle" section to "man Coro".
223 - try a tentative workaround against the breakage that 5.13 has
224 introduced without depreciation period. sigh.
225 - no longer use Time::HiRes if gettimeofday is available, which
226 saves quite a lot of memory.
227
2285.36 Sun Feb 13 05:33:41 CET 2011
229 - automatically load Coro::Channel, Coro::RWLock, Coro::Semaphore,
230 Coro::SemaphoreSet, Coro::Signal and Coro::Specific on first "new"
231 method call.
232 - undocument Coro::Timer::sleep and obsolete whole module.
233 - optimise Coro::Timer::timeout memory and cpu usage.
234 - slightly updated Coro::Intro for recent changes.
235 - do not initialise PL_dirty anymore.
236
2375.25 Thu Nov 11 01:08:39 CET 2010
238 - try a different approach on netbsd - netbsd 5 finally has marginally
239 working pthreads, but still broken ucontext/sigaltstack.
240 - openbsd 4.8 finally got their act together, Coro works out of the box
241 with asm, setjmp and pthreads (no change, just informational).
242
2435.24 Sat Oct 23 11:27:12 CEST 2010
244 - port to the EV 4.0 API.
245 - work around bugs in mingw32, making strawberry perl work
246 out of the box.
247 - correctly modify Coro::AIO function prototypes
248 so that they reflect the "no optional parameters" rule.
249 - "ported" libcoro to C++.
250
2515.23 Mon May 17 18:50:42 CEST 2010
252 - be more resistant to ordering changes when initialising
253 Coro::AnyEvent, Coro::EV and Coro::Event (reported by Matthias
254 Waldorf).
255 - document that perl 5.12 deliberately removed support for cloning.
256
2575.22 Wed Apr 14 03:55:35 CEST 2010
258 - correctly return udnef on errors in Coro::Handle::read/write
259 (testcase by Marc Mims).
10 - convert Coro::Util into a "perl compatibility wrapper" - the functions 260 - convert Coro::Util into a "perl compatibility wrapper" - the functions
11 are less useful now, but are drop-in replacements for existing 261 are less useful now, but are drop-in replacements for existing
12 functions, listing better alternatives in the documentation. This also 262 functions, listing better alternatives in the documentation. This also
13 fixes a bug in Coro::LWP which naively substituted Socket::inet_aton 263 fixes a bug in Coro::LWP which naively substituted Socket::inet_aton
14 with Coro::Util::inet_aton. 264 with Coro::Util::inet_aton.
… …
283 - ->throw is now supported on Coro::State objects. 533 - ->throw is now supported on Coro::State objects.
284 - clean up cctx creation code a bit. 534 - clean up cctx creation code a bit.
285 - entersub is actually a UNOP, not a LOGOP (not a bugfix). 535 - entersub is actually a UNOP, not a LOGOP (not a bugfix).
286 536
2874.9 Sat Nov 8 17:45:27 CET 2008 5374.9 Sat Nov 8 17:45:27 CET 2008
288 - (libcoro) did not preserve rbp with CORO_ASM (we are getting there). 538 - (libcoro) did not preserve rbp with CORO_ASM (we are getting there).
289 - (libcoro) no longer leak threads in the experimental pthread backend, 539 - (libcoro) no longer leak threads in the experimental pthread backend,
290 also speed it up considerably. 540 also speed it up considerably.
291 - (libcoro) do not rely on makecontext passing void *'s unscathed. 541 - (libcoro) do not rely on makecontext passing void *'s unscathed.
292 - fix compiletime dependencies on libcoro in the Makefile. 542 - fix compiletime dependencies on libcoro in the Makefile.
293 - cctx_count wasn't always updated properly. 543 - cctx_count wasn't always updated properly.
… …
355 - remove debugging code related to MgPV_nolen_const, also try to 605 - remove debugging code related to MgPV_nolen_const, also try to
356 make it compile with perl 5.8.6 (yes, apple apparently loves 606 make it compile with perl 5.8.6 (yes, apple apparently loves
357 outdated software). Reported by John S. 607 outdated software). Reported by John S.
358 608
3594.744 Tue Jul 8 22:06:35 CEST 2008 6094.744 Tue Jul 8 22:06:35 CEST 2008
360 - correctly provide default DIE/WARN handlers as documented. 610 - correctly provide default DIE/WARN handlers as documented.
361 - also overwrite PL_vtbl_sigelem.svt_clear, even though current 611 - also overwrite PL_vtbl_sigelem.svt_clear, even though current
362 implementations inside perl work fine for us. 612 implementations inside perl work fine for us.
363 613
3644.743 Mon Jun 16 00:21:57 CEST 2008 6144.743 Mon Jun 16 00:21:57 CEST 2008
365 - when using Coro::EV without running EV::loop it could 615 - when using Coro::EV without running EV::loop it could
… …
387 - sprinkle "no warnings" freely over everything, also suppress 637 - sprinkle "no warnings" freely over everything, also suppress
388 warnings for some other modules. 638 warnings for some other modules.
389 - fix typo in WSAEWOULDBLOCK. 639 - fix typo in WSAEWOULDBLOCK.
390 640
3914.72 Sun May 25 05:14:36 CEST 2008 6414.72 Sun May 25 05:14:36 CEST 2008
392 - tweak META.yaml a bit, unfortunately, there is no documented way 642 - tweak META.yaml a bit, unfortunately, there is no documented way
393 to have optional dependencies with CPAN. doh :( 643 to have optional dependencies with CPAN. doh :(
394 - avoid running some tests on windows because they would fail due to 644 - avoid running some tests on windows because they would fail due to
395 perl bug (broken fork, broken pipes...). 645 perl bug (broken fork, broken pipes...).
396 - work around perl on windows bugs where perl returns undocumented 646 - work around perl on windows bugs where perl returns undocumented
397 error codes for sysread, syswrite etc. by taking advantage 647 error codes for sysread, syswrite etc. by taking advantage
… …
401 - use unix domain sockets in testsuite to work around 651 - use unix domain sockets in testsuite to work around
402 common perl implementation bugs on widows (they are emulated by 652 common perl implementation bugs on widows (they are emulated by
403 tcp sockets on windows. ugh.) 653 tcp sockets on windows. ugh.)
404 654
4054.71 Sat May 24 20:01:27 CEST 2008 6554.71 Sat May 24 20:01:27 CEST 2008
406 - fix a bug in Coro::AnyEvent ("Usage: Coro::AnyEvent::_schedule()"). 656 - fix a bug in Coro::AnyEvent ("Usage: Coro::AnyEvent::_schedule()").
407 - take advantage of async name resolution of AnyEvent::Util. 657 - take advantage of async name resolution of AnyEvent::Util.
408 - work around brutal inet_aton override in Coro::LWP. 658 - work around brutal inet_aton override in Coro::LWP.
409 - take advantage of the readyhook in Coro::EV, for smoother 659 - take advantage of the readyhook in Coro::EV, for smoother
410 scheduling. 660 scheduling.
411 661
4124.7 Sun May 11 00:32:19 CEST 2008 6624.7 Sun May 11 00:32:19 CEST 2008
413 - completely reworked the Coro manpage. 663 - completely reworked the Coro manpage.
414 - added Coro::AnyEvent, generic event loop integration. 664 - added Coro::AnyEvent, generic event loop integration.
415 - implement cancel, ready and kill commands in Coro::Debug. 665 - implement cancel, ready and kill commands in Coro::Debug.
416 - document find_coro in Coro::Debug. 666 - document find_coro in Coro::Debug.
417 - incompatible change: rename has_stack to has_cctx. 667 - incompatible change: rename has_stack to has_cctx.
418 - Coro::AIO and Coro::BDB no longer force event model detection, 668 - Coro::AIO and Coro::BDB no longer force event model detection,
419 use AnyEvent::AIO and AnyEvent::BDB. 669 use AnyEvent::AIO and AnyEvent::BDB.
… …
543 793
5444.11 Thu Oct 11 02:40:24 CEST 2007 7944.11 Thu Oct 11 02:40:24 CEST 2007
545 - port to threaded perls. 795 - port to threaded perls.
546 796
5474.1 Thu Oct 11 02:38:16 CEST 2007 7974.1 Thu Oct 11 02:38:16 CEST 2007
548 - incompatible change: $SIG{__DIE__} and $SIG{__WARN__} will now 798 - incompatible change: $SIG{__DIE__} and $SIG{__WARN__} will now
549 be local to each coro (see Coro::State). 799 be local to each coro (see Coro::State).
550 - incompatible change: for very deep reasons, cede and cede_notself 800 - incompatible change: for very deep reasons, cede and cede_notself
551 cannot return anything, so nothing will be returned. 801 cannot return anything, so nothing will be returned.
552 - possibly bring back 5.10 compatibility (untested). 802 - possibly bring back 5.10 compatibility (untested).
553 - work around stupid (and wrong) warning on 5.10 :(. 803 - work around stupid (and wrong) warning on 5.10 :(.
… …
625 but stateful semantics. 875 but stateful semantics.
626 - fixed a lot of typos in Coro.pm (patch submitted by David 876 - fixed a lot of typos in Coro.pm (patch submitted by David
627 Steinbrunner, which applied flawlessly). 877 Steinbrunner, which applied flawlessly).
628 878
6293.6 Sat Apr 14 17:13:31 CEST 2007 8793.6 Sat Apr 14 17:13:31 CEST 2007
630 - added some bugfixes to get eg/myhttpd working again. 880 - added some bugfixes to get eg/myhttpd working again.
631 - added Coro::Storable for often-cede'ing freeze/thaw. 881 - added Coro::Storable for often-cede'ing freeze/thaw.
632 - try to do a clean exit when a coroutine calls exit 882 - try to do a clean exit when a coroutine calls exit
633 (EXPERIMENTAL). 883 (EXPERIMENTAL).
634 - got rid of indirect call through _coro_init. 884 - got rid of indirect call through _coro_init.
635 - updated the partly antique examples in eg/ to 885 - updated the partly antique examples in eg/ to

Diff Legend

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