| 1 | Revision history for Perl extension Coro. |
1 | Revision history for Perl extension Coro. |
| 2 | |
2 | |
| 3 | TODO: should explore PerlIO::coroaio (perl leaks like hell). |
3 | TODO: should explore PerlIO::coroaio (perl leaks like hell). |
| 4 | TODO: unready_all |
|
|
| 5 | TODO: myhttpd header parsing |
|
|
| 6 | TODO: channel->maxsize(newsize)? |
4 | TODO: channel->maxsize(newsize)? |
| 7 | TODO: http://www.microsoft.com/msj/archive/s2ce.aspx |
5 | TODO: __GCC_HAVE_DWARF2_CFI_ASM |
|
|
6 | TODO: swap_sv, maybe add scope_swap_sv, andallow to unswap by repeat? |
|
|
7 | TODO: swap_sv not undone in pool |
|
|
8 | |
|
|
9 | - croak when cancel is called without a thread context. |
|
|
10 | |
|
|
11 | 6.37 Tue Mar 4 13:27:33 CET 2014 |
|
|
12 | - *sigh*, removed leftover debugging code from debugging a |
|
|
13 | perl bug, of all things. |
|
|
14 | |
|
|
15 | 6.36 Tue Mar 4 07:11:59 CET 2014 |
|
|
16 | - semaphores would not clear the destroy hook when interrupted |
|
|
17 | by ->throw, causing segfaults or worse. |
|
|
18 | - ->throw on a thread waiting for a semaphore did not acquire |
|
|
19 | the semaphore, but also didn't wake up other waiters, |
|
|
20 | possibly causing a deadlock. |
|
|
21 | - "FATAL: $Coro::IDLE blocked itself" will now use Carp::confess |
|
|
22 | to report a full stacktrace, which should help find out |
|
|
23 | where the actual call is. |
|
|
24 | - "atomically" destroy data in slf_destroy, because it is |
|
|
25 | the right thing to do, just in case. |
|
|
26 | - disable ecb.h libm dependency, to compile on ancient systems |
|
|
27 | or under adverse conditions. |
|
|
28 | |
|
|
29 | 6.33 Mon Nov 18 11:26:27 CET 2013 |
|
|
30 | - do not crash when freeing padlists with holes (in 5.18). |
|
|
31 | - tentative SVt_BIND 5.19 port/fix. |
|
|
32 | |
|
|
33 | 6.32 Tue Nov 5 15:35:35 CET 2013 |
|
|
34 | - use a new algorithm to derive padlists for perl 5.18. The old |
|
|
35 | one could lead to 0-pointer accesses inside perl (reported |
|
|
36 | by Darin McBride). |
|
|
37 | |
|
|
38 | 6.31 Thu May 9 07:39:48 CEST 2013 |
|
|
39 | - Coro::AIO requests would crash if the thread was ready'd |
|
|
40 | while the request was ongoing. |
|
|
41 | |
|
|
42 | 6.29 Wed May 8 02:55:18 CEST 2013 |
|
|
43 | - when an on_destroy handler destructs the coro currently being |
|
|
44 | destructed a perl scalar could be accessed after being freed, |
|
|
45 | likely causing a crash. |
|
|
46 | |
|
|
47 | 6.28 Wed Mar 6 06:58:02 CET 2013 |
|
|
48 | - clean remnants of existing __DIE__ and __WARN__ handlers so |
|
|
49 | they lose their magic and will not cause segfaults later |
|
|
50 | (testcase by Andrey Sagulin). |
|
|
51 | - improved Coro::State documentation a bit. |
|
|
52 | - Coro::Debug::command now flushes the output. |
|
|
53 | - add hack detection code for x32 abi, because the braindead slugs |
|
|
54 | who designed that made it look exactly like x86_64 without |
|
|
55 | providing proper compile time symbols to test for it. as a result, |
|
|
56 | this detection cannot work reliably. |
|
|
57 | - valgrind stack registering was broken. |
|
|
58 | - do not rely on Time::HiRes anymore in Coro::Debug. |
|
|
59 | |
|
|
60 | 6.23 Fri Dec 7 23:36:37 CET 2012 |
|
|
61 | - use experimental fiber implementation on native windows |
|
|
62 | perls. |
|
|
63 | - use sizeof (void *) as multiplication factor for stack sizes, |
|
|
64 | to accomodate the totally braindamaged microsoft 64 bit "os". |
|
|
65 | - changed verifier host from win2k-ap510-32 to win7-sp516-32/64. |
|
|
66 | activeperl 5.16 crashes when PerlIO_define_layer is called due |
|
|
67 | to some bug in the perl dll, strawberry perl at least passes |
|
|
68 | the testsuite. |
|
|
69 | - implement Coro::Handle->peeraddr/host/port, for slightly |
|
|
70 | improved compatibility with LWP. |
|
|
71 | - implement 5.17 compatibility by almost blindly applying a |
|
|
72 | good-looking patch by Father Chrysostomos. |
|
|
73 | - move stack management functions into libcoro 3. |
|
|
74 | - libcoro version 3 "released". |
|
|
75 | - support magic values as timed_io_once args. |
|
|
76 | - recommend AnyEvent 7+ or EV 4+, also require EV |
|
|
77 | version 4 or newer for Coro::EV. |
|
|
78 | |
|
|
79 | 6.10 Tue Oct 9 01:14:27 CEST 2012 |
|
|
80 | - updated ecb.h, it had a typo that caused it to not compile on many |
|
|
81 | big endian systems (reported by many people). |
|
|
82 | - disable memory fences in ecb.h to improve portability. |
|
|
83 | |
|
|
84 | 6.09 Sat Oct 6 23:25:02 CEST 2012 |
|
|
85 | - Coro::EV I/O watchers were not interruptible by exceptions |
|
|
86 | (Coro::State::throw) (testcase by sten). |
|
|
87 | - ->throw now puts threads into the ready queue, as this seems to |
|
|
88 | be expected by existing code, and code that doesn't cope with spurious |
|
|
89 | wakeups needs fixing anyway. |
|
|
90 | - use fd -1 in mmap. |
|
|
91 | - cast I32 to int in error message printf. |
|
|
92 | - warn about broken so-called "hardened" kernels. |
|
|
93 | |
|
|
94 | 6.08 Fri Apr 13 12:05:47 CEST 2012 |
|
|
95 | - be more aggressive about exiting like perl does - formerly, |
|
|
96 | exiting from the non-main thread would not execute END blocks. |
|
|
97 | |
|
|
98 | 6.07 Fri Nov 11 21:21:48 CET 2011 |
|
|
99 | - work around a bug in PerlIO (setting $SIG{__WARN__} to a PVCV). |
|
|
100 | - update ecb.h. |
|
|
101 | |
|
|
102 | 6.06 Mon Aug 8 23:59:48 CEST 2011 |
|
|
103 | - cygwin unfortunately patches the stack at runtime, so we use the pthreads |
|
|
104 | backend, which is an order of magnitude slower. unfortunately, cygwins |
|
|
105 | pthread implementation isn't very complete either, so allocate the stack |
|
|
106 | twice just to be sure. |
|
|
107 | (note: cygwin also enables mymalloc, which is NOT THREADSAFE ON WINDOWS, |
|
|
108 | in its ithreaded perl - best recompile cygwin and use the 'w'indows |
|
|
109 | backend for much better performance. also disable ithreads for |
|
|
110 | even better performance...). |
|
|
111 | |
|
|
112 | 6.05 Thu Aug 4 21:36:36 CEST 2011 |
|
|
113 | - blush, condvar values would not be propagated from send to recv anymore |
|
|
114 | (reported by Chip Salzenberg). |
|
|
115 | - use exponential increase for the readline buffer length in |
|
|
116 | Coro::Handle. also reduce initial allocation to 1020 from 4096 bytes. |
|
|
117 | |
|
|
118 | 6.04 Wed Aug 3 17:15:45 CEST 2011 |
|
|
119 | - use even more efficient and more compatible condvars for |
|
|
120 | compatibility to AnyEvent 6.x :) |
|
|
121 | - more inconsequential ecb.h updates. |
|
|
122 | |
|
|
123 | 6.03 Wed Aug 3 11:41:30 CEST 2011 |
|
|
124 | - change how Coro patches AnyEvent condvars for compatibility to |
|
|
125 | AnyEvent 6.x. |
|
|
126 | - update ecb.h, to no longer include <pthread.h> in case WinNT.h |
|
|
127 | hasn't been included. |
|
|
128 | |
|
|
129 | 6.02 Wed Jul 13 04:35:19 CEST 2011 |
|
|
130 | - "improve portability to Gentoo" - gentoo manages to put perl variables |
|
|
131 | in memory areas that are farther than 2gb apart, which the jit couldn't |
|
|
132 | handle and barfed. now it's just a bit slower on gentoo and similar |
|
|
133 | systems. |
|
|
134 | |
|
|
135 | 6.01 Sun Jul 3 12:31:14 CEST 2011 |
|
|
136 | - workarounds are good, but the test for whether pthreads are used |
|
|
137 | was not good. thisone should be better. |
|
|
138 | - check differently whether gcc generates cfi instructions itself. |
|
|
139 | |
|
|
140 | 6.0 Wed Jun 29 19:43:35 CEST 2011 |
|
|
141 | - INCOMPATIBLE CHANGE: unreferenced coro objects will now be |
|
|
142 | destroyed and cleaned up automatically (e.g. async { schedule }). |
|
|
143 | - implement a JIT compiler for part of the thread switch code, |
|
|
144 | which gives a 50% speed improvement on threaded perls, and |
|
|
145 | about 4% on non-thraeded perls (so threaded perls now finally |
|
|
146 | reach about half the speed of non-threaded perls). |
|
|
147 | - slightly modernise Coro::Intro, add section about rouse functions. |
|
|
148 | - avoid DEFSV and ERRSV, giving another 10% improvement |
|
|
149 | in thread switching. |
|
|
150 | - Coro::State->is_destroyed is now called is_zombie. |
|
|
151 | - implement a Coro->safe_cancel method that might fail, but |
|
|
152 | cancels in a "safer" way if it succeeds. |
|
|
153 | - add preliminary support for DEBUGGING perls. |
|
|
154 | - get rid of two hash-accesses when initialising a new Coro - this |
|
|
155 | speeds up coro creation by almost a factor of two. |
|
|
156 | - croak when a coro that is being cancelled tries to block |
|
|
157 | (e.g. while executing a guard block), instead of crashing or |
|
|
158 | deadlocking. |
|
|
159 | - use a more robust and also faster method to identify Coro::State |
|
|
160 | objects - speeds up everything a bit. |
|
|
161 | - implement Coro->cancel in XS for a 20% speed improvement, and to |
|
|
162 | be able to implement mutual cancellation. |
|
|
163 | - speed up context switches by a percent or two by more efficiently |
|
|
164 | allocating context stack entries. |
|
|
165 | - implement Coro->join and Coro->on_destroy in XS for a speedup and |
|
|
166 | a reduction in memory use. |
|
|
167 | - cancelling a coro while it itself is cancelling another coro is |
|
|
168 | now supported and working, instead of triggering an assertion. |
|
|
169 | - be a bit more crash-resistant when calling (buggy) on_destroy |
|
|
170 | callbacks (best effort). |
|
|
171 | - move on_destroy into the slf_frame, to allow extension slf |
|
|
172 | functions to have destructors. |
|
|
173 | - get rid if coro refcounting - simply crash in other interpreter |
|
|
174 | threads by nulling the pointers on clone. |
|
|
175 | - simplify warn/die hook handling when loading Coro - the convoluted |
|
|
176 | logic seems to be no longer neccessary. |
|
|
177 | - use libecb instead of our own home-grown gcc hacks. |
|
|
178 | - document alternatives to Coro::LWP. Please use them :) |
|
|
179 | - work around another mindless idiotic NEEDLESS bug in openbsd/mirbsds |
|
|
180 | sigaltstack. Really. wine suffers from it, erlang suffers from it, |
|
|
181 | and it's known since at least 2006. |
|
|
182 | |
|
|
183 | 5.372 Wed Feb 23 06:14:30 CET 2011 |
|
|
184 | - apparently mingw doesn't provide a working gettimeofday, try to |
|
|
185 | work around that by relying on Time::HiRes (indirectly brought to |
|
|
186 | my attention by Max Maischein). |
|
|
187 | - fix some portability issues when Time::HiRes was used. |
|
|
188 | |
|
|
189 | 5.371 Mon Feb 21 14:36:08 CET 2011 |
|
|
190 | - backport to windows process emulation code again. |
|
|
191 | |
|
|
192 | 5.37 Sat Feb 19 07:49:44 CET 2011 |
|
|
193 | - add a big "Coro thread life cycle" section to "man Coro". |
|
|
194 | - try a tentative workaround against the breakage that 5.13 has |
|
|
195 | introduced without depreciation period. sigh. |
|
|
196 | - no longer use Time::HiRes if gettimeofday is available, which |
|
|
197 | saves quite a lot of memory. |
|
|
198 | |
|
|
199 | 5.36 Sun Feb 13 05:33:41 CET 2011 |
|
|
200 | - automatically load Coro::Channel, Coro::RWLock, Coro::Semaphore, |
|
|
201 | Coro::SemaphoreSet, Coro::Signal and Coro::Specific on first "new" |
|
|
202 | method call. |
|
|
203 | - undocument Coro::Timer::sleep and obsolete whole module. |
|
|
204 | - optimise Coro::Timer::timeout memory and cpu usage. |
|
|
205 | - slightly updated Coro::Intro for recent changes. |
|
|
206 | - do not initialise PL_dirty anymore. |
|
|
207 | |
|
|
208 | 5.25 Thu Nov 11 01:08:39 CET 2010 |
|
|
209 | - try a different approach on netbsd - netbsd 5 finally has marginally |
|
|
210 | working pthreads, but still broken ucontext/sigaltstack. |
|
|
211 | - openbsd 4.8 finally got their act together, Coro works out of the box |
|
|
212 | with asm, setjmp and pthreads (no change, just informational). |
| 8 | |
213 | |
| 9 | 5.24 Sat Oct 23 11:27:12 CEST 2010 |
214 | 5.24 Sat Oct 23 11:27:12 CEST 2010 |
| 10 | - port to the EV 4.0 API. |
215 | - port to the EV 4.0 API. |
| 11 | - work around bugs in mingw32, making strawberry perl work |
216 | - work around bugs in mingw32, making strawberry perl work |
| 12 | out of the box. |
217 | out of the box. |
| … | |
… | |
| 299 | - ->throw is now supported on Coro::State objects. |
504 | - ->throw is now supported on Coro::State objects. |
| 300 | - clean up cctx creation code a bit. |
505 | - clean up cctx creation code a bit. |
| 301 | - entersub is actually a UNOP, not a LOGOP (not a bugfix). |
506 | - entersub is actually a UNOP, not a LOGOP (not a bugfix). |
| 302 | |
507 | |
| 303 | 4.9 Sat Nov 8 17:45:27 CET 2008 |
508 | 4.9 Sat Nov 8 17:45:27 CET 2008 |
| 304 | - (libcoro) did not preserve rbp with CORO_ASM (we are getting there). |
509 | - (libcoro) did not preserve rbp with CORO_ASM (we are getting there). |
| 305 | - (libcoro) no longer leak threads in the experimental pthread backend, |
510 | - (libcoro) no longer leak threads in the experimental pthread backend, |
| 306 | also speed it up considerably. |
511 | also speed it up considerably. |
| 307 | - (libcoro) do not rely on makecontext passing void *'s unscathed. |
512 | - (libcoro) do not rely on makecontext passing void *'s unscathed. |
| 308 | - fix compiletime dependencies on libcoro in the Makefile. |
513 | - fix compiletime dependencies on libcoro in the Makefile. |
| 309 | - cctx_count wasn't always updated properly. |
514 | - cctx_count wasn't always updated properly. |
| … | |
… | |
| 371 | - remove debugging code related to MgPV_nolen_const, also try to |
576 | - remove debugging code related to MgPV_nolen_const, also try to |
| 372 | make it compile with perl 5.8.6 (yes, apple apparently loves |
577 | make it compile with perl 5.8.6 (yes, apple apparently loves |
| 373 | outdated software). Reported by John S. |
578 | outdated software). Reported by John S. |
| 374 | |
579 | |
| 375 | 4.744 Tue Jul 8 22:06:35 CEST 2008 |
580 | 4.744 Tue Jul 8 22:06:35 CEST 2008 |
| 376 | - correctly provide default DIE/WARN handlers as documented. |
581 | - correctly provide default DIE/WARN handlers as documented. |
| 377 | - also overwrite PL_vtbl_sigelem.svt_clear, even though current |
582 | - also overwrite PL_vtbl_sigelem.svt_clear, even though current |
| 378 | implementations inside perl work fine for us. |
583 | implementations inside perl work fine for us. |
| 379 | |
584 | |
| 380 | 4.743 Mon Jun 16 00:21:57 CEST 2008 |
585 | 4.743 Mon Jun 16 00:21:57 CEST 2008 |
| 381 | - when using Coro::EV without running EV::loop it could |
586 | - when using Coro::EV without running EV::loop it could |
| … | |
… | |
| 403 | - sprinkle "no warnings" freely over everything, also suppress |
608 | - sprinkle "no warnings" freely over everything, also suppress |
| 404 | warnings for some other modules. |
609 | warnings for some other modules. |
| 405 | - fix typo in WSAEWOULDBLOCK. |
610 | - fix typo in WSAEWOULDBLOCK. |
| 406 | |
611 | |
| 407 | 4.72 Sun May 25 05:14:36 CEST 2008 |
612 | 4.72 Sun May 25 05:14:36 CEST 2008 |
| 408 | - tweak META.yaml a bit, unfortunately, there is no documented way |
613 | - tweak META.yaml a bit, unfortunately, there is no documented way |
| 409 | to have optional dependencies with CPAN. doh :( |
614 | to have optional dependencies with CPAN. doh :( |
| 410 | - avoid running some tests on windows because they would fail due to |
615 | - avoid running some tests on windows because they would fail due to |
| 411 | perl bug (broken fork, broken pipes...). |
616 | perl bug (broken fork, broken pipes...). |
| 412 | - work around perl on windows bugs where perl returns undocumented |
617 | - work around perl on windows bugs where perl returns undocumented |
| 413 | error codes for sysread, syswrite etc. by taking advantage |
618 | error codes for sysread, syswrite etc. by taking advantage |
| … | |
… | |
| 417 | - use unix domain sockets in testsuite to work around |
622 | - use unix domain sockets in testsuite to work around |
| 418 | common perl implementation bugs on widows (they are emulated by |
623 | common perl implementation bugs on widows (they are emulated by |
| 419 | tcp sockets on windows. ugh.) |
624 | tcp sockets on windows. ugh.) |
| 420 | |
625 | |
| 421 | 4.71 Sat May 24 20:01:27 CEST 2008 |
626 | 4.71 Sat May 24 20:01:27 CEST 2008 |
| 422 | - fix a bug in Coro::AnyEvent ("Usage: Coro::AnyEvent::_schedule()"). |
627 | - fix a bug in Coro::AnyEvent ("Usage: Coro::AnyEvent::_schedule()"). |
| 423 | - take advantage of async name resolution of AnyEvent::Util. |
628 | - take advantage of async name resolution of AnyEvent::Util. |
| 424 | - work around brutal inet_aton override in Coro::LWP. |
629 | - work around brutal inet_aton override in Coro::LWP. |
| 425 | - take advantage of the readyhook in Coro::EV, for smoother |
630 | - take advantage of the readyhook in Coro::EV, for smoother |
| 426 | scheduling. |
631 | scheduling. |
| 427 | |
632 | |
| 428 | 4.7 Sun May 11 00:32:19 CEST 2008 |
633 | 4.7 Sun May 11 00:32:19 CEST 2008 |
| 429 | - completely reworked the Coro manpage. |
634 | - completely reworked the Coro manpage. |
| 430 | - added Coro::AnyEvent, generic event loop integration. |
635 | - added Coro::AnyEvent, generic event loop integration. |
| 431 | - implement cancel, ready and kill commands in Coro::Debug. |
636 | - implement cancel, ready and kill commands in Coro::Debug. |
| 432 | - document find_coro in Coro::Debug. |
637 | - document find_coro in Coro::Debug. |
| 433 | - incompatible change: rename has_stack to has_cctx. |
638 | - incompatible change: rename has_stack to has_cctx. |
| 434 | - Coro::AIO and Coro::BDB no longer force event model detection, |
639 | - Coro::AIO and Coro::BDB no longer force event model detection, |
| 435 | use AnyEvent::AIO and AnyEvent::BDB. |
640 | use AnyEvent::AIO and AnyEvent::BDB. |
| … | |
… | |
| 559 | |
764 | |
| 560 | 4.11 Thu Oct 11 02:40:24 CEST 2007 |
765 | 4.11 Thu Oct 11 02:40:24 CEST 2007 |
| 561 | - port to threaded perls. |
766 | - port to threaded perls. |
| 562 | |
767 | |
| 563 | 4.1 Thu Oct 11 02:38:16 CEST 2007 |
768 | 4.1 Thu Oct 11 02:38:16 CEST 2007 |
| 564 | - incompatible change: $SIG{__DIE__} and $SIG{__WARN__} will now |
769 | - incompatible change: $SIG{__DIE__} and $SIG{__WARN__} will now |
| 565 | be local to each coro (see Coro::State). |
770 | be local to each coro (see Coro::State). |
| 566 | - incompatible change: for very deep reasons, cede and cede_notself |
771 | - incompatible change: for very deep reasons, cede and cede_notself |
| 567 | cannot return anything, so nothing will be returned. |
772 | cannot return anything, so nothing will be returned. |
| 568 | - possibly bring back 5.10 compatibility (untested). |
773 | - possibly bring back 5.10 compatibility (untested). |
| 569 | - work around stupid (and wrong) warning on 5.10 :(. |
774 | - work around stupid (and wrong) warning on 5.10 :(. |
| … | |
… | |
| 641 | but stateful semantics. |
846 | but stateful semantics. |
| 642 | - fixed a lot of typos in Coro.pm (patch submitted by David |
847 | - fixed a lot of typos in Coro.pm (patch submitted by David |
| 643 | Steinbrunner, which applied flawlessly). |
848 | Steinbrunner, which applied flawlessly). |
| 644 | |
849 | |
| 645 | 3.6 Sat Apr 14 17:13:31 CEST 2007 |
850 | 3.6 Sat Apr 14 17:13:31 CEST 2007 |
| 646 | - added some bugfixes to get eg/myhttpd working again. |
851 | - added some bugfixes to get eg/myhttpd working again. |
| 647 | - added Coro::Storable for often-cede'ing freeze/thaw. |
852 | - added Coro::Storable for often-cede'ing freeze/thaw. |
| 648 | - try to do a clean exit when a coroutine calls exit |
853 | - try to do a clean exit when a coroutine calls exit |
| 649 | (EXPERIMENTAL). |
854 | (EXPERIMENTAL). |
| 650 | - got rid of indirect call through _coro_init. |
855 | - got rid of indirect call through _coro_init. |
| 651 | - updated the partly antique examples in eg/ to |
856 | - updated the partly antique examples in eg/ to |
