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: channel->maxsize(newsize)? |
4 | TODO: channel->maxsize(newsize)? |
5 | TODO: __GCC_HAVE_DWARF2_CFI_ASM |
5 | TODO: __GCC_HAVE_DWARF2_CFI_ASM |
6 | TODO: swap_sv, maybe add scope_swap_sv, andallow to unswap by repeat? |
6 | TODO: swap_sv, maybe add scope_swap_sv? |
7 | TODO: swap_sv not undone in pool |
7 | TODO: croak when async_pool tries to run canceled thread? |
|
|
8 | TODO: how to debug safe_cancel failure (vebrose mode? show how to make backtrace?) |
|
|
9 | TODO: scalar context for Coro::AIO wrappers? |
8 | |
10 | |
9 | Program terminated with signal 11, Segmentation fault. |
11 | TODO: State.xs could short-circuit some macros such as CORO_EXECUTE_SLF |
10 | #0 coro_semaphore_adjust (av=0x0, adjust=0) at State.xs:2915 |
12 | TODO: use frame->destroy to improve Coro::EV |
11 | 2915 State.xs: No such file or directory. |
|
|
12 | (gdb) bt |
|
|
13 | #0 coro_semaphore_adjust (av=0x0, adjust=0) at State.xs:2915 |
|
|
14 | #1 0x00007f638f0c0703 in slf_destroy (coro=0x5957180) at State.xs:2605 |
|
|
15 | #2 coro_state_destroy (coro=0x5957180) at State.xs:1712 |
|
|
16 | #3 0x00007f638f0c150c in coro_state_destroy (coro=<optimized out>) at State.xs:1709 |
|
|
17 | #4 XS_Coro__destroy (cv=<optimized out>) at State.xs:3853 |
|
|
18 | #5 0x00000000004b572d in Perl_pp_entersub () at pp_hot.c:2881 |
|
|
19 | #6 0x00000000004adfe3 in Perl_runops_standard () at run.c:42 |
|
|
20 | #7 0x000000000044335f in S_run_body (oldscope=<optimized out>) at perl.c:2462 |
|
|
21 | #8 perl_run (my_perl=<optimized out>) at perl.c:2383 |
|
|
22 | #9 0x00007f638f0ba3a1 in cctx_run (arg=<optimized out>) at State.xs:1478 |
|
|
23 | #10 0x00007f638f0b74d5 in coro_init () at libcoro/coro.c:96 |
|
|
24 | (gdb) up |
|
|
25 | #1 0x00007f638f0c0703 in slf_destroy (coro=0x5957180) at State.xs:2605 |
|
|
26 | 2605 in State.xs |
|
|
27 | (gdb) p coro->slf_frame |
|
|
28 | $1 = { |
|
|
29 | prepare = 0x7f638f0baf60 <prepare_schedule>, |
|
|
30 | check = 0x7f638f0b6e70 <slf_check_repeat>, |
|
|
31 | data = 0x0, |
|
|
32 | destroy = 0x7f638f0bb820 <coro_semaphore_destroy> |
|
|
33 | |
13 | |
|
|
14 | TODO: concurrent ->recv on same condvar |
|
|
15 | sub Coro::AnyEvent::CondVar::_send { |
|
|
16 | if ($_[0]{_ae_coro}) { |
|
|
17 | $_->ready |
|
|
18 | for values %{ delete $_[0]{_ae_coro} } |
|
|
19 | } |
|
|
20 | }; |
34 | |
21 | |
35 | TODO: warnhook reports from wrong line number now, when arg is Math::GMP number. |
22 | sub Coro::AnyEvent::CondVar::_wait { |
|
|
23 | until ($_[0]{_ae_sent}) { |
|
|
24 | $AnyEvent::CondVar::Base::WAITING = 0; # avoid recursive check by AnyEvent |
|
|
25 | local $_[0]{_ae_coro}{$Coro::current+0} = $Coro::current; |
|
|
26 | Coro::schedule; |
|
|
27 | } |
|
|
28 | }; |
|
|
29 | |
|
|
30 | - potential fix for an interetsing type of recursive coro |
|
|
31 | destruction (testcase by Christopher J. Madsen). |
|
|
32 | |
|
|
33 | 6.57 Wed Jul 29 15:08:44 CEST 2020 |
|
|
34 | - override AnyEvent condvar methods "properly", as AnyEvent |
|
|
35 | has support for pluggable condvars since 2011. This also |
|
|
36 | avoids problems with recursive waits by avoiding to mix |
|
|
37 | impl methods with coro code. |
|
|
38 | - allow rouse_wait to be called in a different thread. |
|
|
39 | - slightly clarify rouse_cb/rouse_wait documentation. |
|
|
40 | - slightly improve rouse callback performance and memory usage. |
|
|
41 | - try to report correct calller when throwing exception in |
|
|
42 | on $condvar->recv, not a useless line in Coro::AnyEvent. |
|
|
43 | |
|
|
44 | 6.56 Thu Feb 20 19:40:07 CET 2020 |
|
|
45 | - Coro::EV::time*once functions used ev_once, which is not cancelable. |
|
|
46 | This could cause problems due to I/O watchers not being stopped up |
|
|
47 | on exceptions/cancellation. |
|
|
48 | |
|
|
49 | 6.55 Fri Jul 19 17:28:09 CEST 2019 |
|
|
50 | - Coro::RWLock woke up only one reader instead of all after a wrlock |
|
|
51 | (analyzed and patch by Maxim Skorobogatov). |
|
|
52 | - slightly improved documentation of Coro::RWLock and RWLock testcase. |
|
|
53 | - quickly (re)-ported to minix 3.3 before minix crashed again. |
|
|
54 | |
|
|
55 | 6.54 Sun Dec 16 11:06:07 CET 2018 |
|
|
56 | - reinstate Coro::Select::select prototype - it breaks more than it fixes. |
|
|
57 | |
|
|
58 | 6.53 Sun Dec 16 10:33:15 CET 2018 |
|
|
59 | - the autogenerated prototypes of Coro::AIO now add ";" at the end |
|
|
60 | to avoid generating named unary operators and messing up parsing |
|
|
61 | (mostly affects aio_mlockall 4 | 8 which was parsed as |
|
|
62 | (aio_mlockall 4) | 8). |
|
|
63 | - do not use a prototype for select in Coro::Select, as CORE::select also |
|
|
64 | doesn't have one. |
|
|
65 | |
|
|
66 | 6.52 Tue Aug 14 18:49:43 CEST 2018 |
|
|
67 | - libcoro update: allow sharing of coro threads using the pthreads |
|
|
68 | backend among different pthreads, as is required by Coro::Multicore. |
|
|
69 | - always allow ->call and ->eval when the coro is the current coroutine. this |
|
|
70 | makes it possible, among other things, to call these on $Coro::main. |
|
|
71 | - minor code cleanups. |
|
|
72 | |
|
|
73 | 6.514 Thu Aug 31 18:25:31 CEST 2017 |
|
|
74 | - safe_cancel no longer croaks when called on already-destroyed |
|
|
75 | thread, but instead "just works". |
|
|
76 | |
|
|
77 | 6.513 Sat Jul 15 01:18:51 CEST 2017 |
|
|
78 | - experimental 5.22 comaptibility changes (PadlistNAMES). |
|
|
79 | - removed attributes example and made sure the other examples |
|
|
80 | still work. |
|
|
81 | - fix some statevar types for newer perls, should be of no consequence. |
|
|
82 | |
|
|
83 | 6.512 Fri Jul 14 05:03:17 CEST 2017 |
|
|
84 | - seems to pass test suite down till 5.8.9, so make this the |
|
|
85 | minimum requirement in Makefile.PL. |
|
|
86 | - experimental 5.24 compatibility changes (SUB_ARGARRAY). |
|
|
87 | - experimental 5.26 compatibility changes (PL_comppad -> PADOFFSET). |
|
|
88 | - disable FORTIFY_SOURCE to avoid broken (but well-intentioned) |
|
|
89 | fortified longjmp variants in some libcs. |
|
|
90 | - re-try JIT allocation without PROT_EXEC, to hopefully improve |
|
|
91 | portability (still doesn't seem to work around broken selinux |
|
|
92 | mmap everywhere). |
|
|
93 | - speling fixes (Xavier Guimard). |
|
|
94 | |
|
|
95 | 6.511 Sun Jun 26 23:44:50 CEST 2016 |
|
|
96 | - make it compile with threaded perls again. |
|
|
97 | - simplify and speed up __DIE__ and __WARN__ handling. |
|
|
98 | |
|
|
99 | 6.51 Sat Jun 25 20:57:20 CEST 2016 |
|
|
100 | - 6.5 release didn't compile on most older perls due to |
|
|
101 | a logic error enabling experimental 5.24 code for all versions. |
|
|
102 | - since various broken or deficient patches float around |
|
|
103 | to work around the vtbl API breakage in perl, let's provide |
|
|
104 | our own workaround, which seems to be much less invasive |
|
|
105 | and more compatible than the approaches seen so far. YMMV. |
|
|
106 | - work around assertion failure in perl_destruct on 5.24 |
|
|
107 | (analyzed by Dave Mitchell). |
|
|
108 | |
|
|
109 | 6.5 Wed Jun 22 22:23:50 CEST 2016 |
|
|
110 | - swap_sv swapping order was not symmetrical, causing |
|
|
111 | wrong swaps when swap_sv was used multiple times on the |
|
|
112 | same sv in the same thread. |
|
|
113 | - swap_sv calls can now be undone by calling it again |
|
|
114 | with the same variables. |
|
|
115 | - swap_sv calls will now be undone in async_pool threads. |
|
|
116 | - split Coro::Semaphore::up/adjust into separate xs functions |
|
|
117 | for better error reporting, at slight codesize increase. |
|
|
118 | - (libcoro) arm assembly support, please test and report. |
|
|
119 | - adjust to PL_savestack changes in perl 5.24 (adapted from |
|
|
120 | the debian patch, which unfortunately gets it wrong). |
|
|
121 | |
|
|
122 | 6.49 Sat Oct 17 01:40:12 CEST 2015 |
|
|
123 | - throwing an exception to a thread waiting in |
|
|
124 | Coro::Handle using Coro::EV did not stop the watchers, |
|
|
125 | causing the next call to fail (testcase by Martin Pritchard). |
|
|
126 | - bump minimum perl version to 5.10. |
|
|
127 | |
|
|
128 | 6.48 Sun Oct 4 19:03:51 CEST 2015 |
|
|
129 | - fix memory leak when Coro::Handle uses Coro::EV internally |
|
|
130 | (testcase by Sten Sten). |
|
|
131 | - update code to libev 4 API (internally, Coro still used the |
|
|
132 | version 3 API calls). |
|
|
133 | - remove unused hv_sig. |
|
|
134 | |
|
|
135 | 6.47 Sat Jul 11 03:58:20 CEST 2015 |
|
|
136 | - swap_svs - NVs can be in the SV head beginning with 5.20. |
|
|
137 | |
|
|
138 | 6.46 Tue Jun 30 14:36:30 CEST 2015 |
|
|
139 | - restore portability to perls with windows fork emulation |
|
|
140 | (patch by Petr Písař). |
|
|
141 | |
|
|
142 | 6.45 Tue Jun 30 01:40:08 CEST 2015 |
|
|
143 | - restore portability to perl 5.16 and below. |
|
|
144 | |
|
|
145 | 6.44 Tue Jun 30 00:41:54 CEST 2015 |
|
|
146 | - avoid segfaulting (or worse) when tracing a canceled thread. |
|
|
147 | - the Event module silently broke it's public hook API, causing |
|
|
148 | failures on perls compiled with -Duselongdouble. Requiring |
|
|
149 | the latest version of Event because I don't know which release |
|
|
150 | changed this (it's not mentioned in the ChangeLog). |
|
|
151 | - allow xs level enterleave hooks via CoroAPI. |
|
|
152 | |
|
|
153 | 6.43 Thu Jun 4 15:38:14 CEST 2015 |
|
|
154 | - use stability canary. |
|
|
155 | - port to stableperl-5.22.0-1.001. |
|
|
156 | - update libecb. |
|
|
157 | |
|
|
158 | 6.42 Wed Feb 11 20:29:52 CET 2015 |
|
|
159 | - Coro::SemaphoreSet->try did not actually work (analyzed by |
|
|
160 | SATO Kentaro). |
|
|
161 | - upgrade libecb to be C11 compliant. |
|
|
162 | |
|
|
163 | 6.41 Sat Sep 6 22:08:46 CEST 2014 |
|
|
164 | - restore portability to perl 5.8.x. |
|
|
165 | - give new Coro's a valid GvHV(PL_hintgv) - this is slow and takes |
|
|
166 | up some memory, but fixes "use feature" and similar modules |
|
|
167 | when used inside a Coro. |
|
|
168 | - allow zero as argument to Coro::Channel to mean the same thing |
|
|
169 | as no argument. this works with older versions as well, |
|
|
170 | but wasn't legal until now. |
|
|
171 | - slightly better c header file detection. |
|
|
172 | |
|
|
173 | 6.39 Mon Jun 2 00:00:08 CEST 2014 |
|
|
174 | - work around more incompatible changes in 5.20. |
|
|
175 | |
|
|
176 | 6.38 Sun Jun 1 21:54:23 CEST 2014 |
|
|
177 | - check that perl slots actually have enough space to hold |
|
|
178 | interpreter variables. |
|
|
179 | - untested port to perl 5.19 (64 bit tmps indices) (reported |
|
|
180 | by Andreas König). |
|
|
181 | - croak when cancel is called without a thread context. |
|
|
182 | |
|
|
183 | 6.37 Tue Mar 4 13:27:33 CET 2014 |
|
|
184 | - *sigh*, removed leftover debugging code from debugging a |
|
|
185 | perl bug, of all things. |
|
|
186 | |
|
|
187 | 6.36 Tue Mar 4 07:11:59 CET 2014 |
|
|
188 | - semaphores would not clear the destroy hook when interrupted |
|
|
189 | by ->throw, causing segfaults or worse. |
|
|
190 | - ->throw on a thread waiting for a semaphore did not acquire |
|
|
191 | the semaphore, but also didn't wake up other waiters, |
|
|
192 | possibly causing a deadlock. |
36 | - FATAL: $Coro::IDLE blocked itself will now use Carp::confess |
193 | - "FATAL: $Coro::IDLE blocked itself" will now use Carp::confess |
37 | to report a full stacktrace, which should help find out |
194 | to report a full stacktrace, which should help find out |
38 | where the actual call is. |
195 | where the actual call is. |
39 | - "atomically" destroy data in slf_destroy, because it is |
196 | - "atomically" destroy data in slf_destroy, because it is |
40 | the right thing to do, just in case. |
197 | the right thing to do, just in case. |
41 | - disable ecb.h libm dependency, to compile on ancient systems |
198 | - disable ecb.h libm dependency, to compile on ancient systems |
… | |
… | |
147 | handle and barfed. now it's just a bit slower on gentoo and similar |
304 | handle and barfed. now it's just a bit slower on gentoo and similar |
148 | systems. |
305 | systems. |
149 | |
306 | |
150 | 6.01 Sun Jul 3 12:31:14 CEST 2011 |
307 | 6.01 Sun Jul 3 12:31:14 CEST 2011 |
151 | - workarounds are good, but the test for whether pthreads are used |
308 | - workarounds are good, but the test for whether pthreads are used |
152 | was not good. thisone should be better. |
309 | was not good. this one should be better. |
153 | - check differently whether gcc generates cfi instructions itself. |
310 | - check differently whether gcc generates cfi instructions itself. |
154 | |
311 | |
155 | 6.0 Wed Jun 29 19:43:35 CEST 2011 |
312 | 6.0 Wed Jun 29 19:43:35 CEST 2011 |
156 | - INCOMPATIBLE CHANGE: unreferenced coro objects will now be |
313 | - INCOMPATIBLE CHANGE: unreferenced coro objects will now be |
157 | destroyed and cleaned up automatically (e.g. async { schedule }). |
314 | destroyed and cleaned up automatically (e.g. async { schedule }). |
158 | - implement a JIT compiler for part of the thread switch code, |
315 | - implement a JIT compiler for part of the thread switch code, |
159 | which gives a 50% speed improvement on threaded perls, and |
316 | which gives a 50% speed improvement on threaded perls, and |
160 | about 4% on non-thraeded perls (so threaded perls now finally |
317 | about 4% on non-threaded perls (so threaded perls now finally |
161 | reach about half the speed of non-threaded perls). |
318 | reach about half the speed of non-threaded perls). |
162 | - slightly modernise Coro::Intro, add section about rouse functions. |
319 | - slightly modernise Coro::Intro, add section about rouse functions. |
163 | - avoid DEFSV and ERRSV, giving another 10% improvement |
320 | - avoid DEFSV and ERRSV, giving another 10% improvement |
164 | in thread switching. |
321 | in thread switching. |
165 | - Coro::State->is_destroyed is now called is_zombie. |
322 | - Coro::State->is_destroyed is now called is_zombie. |