| 1 |
Revision history for Perl module Test::TCP |
| 2 |
|
| 3 |
2.22 2019-10-08T08:15:34Z |
| 4 |
- Portability fixes for Win32 and non-linux #83, #87 |
| 5 |
|
| 6 |
2.21 2019-10-03T03:15:22Z |
| 7 |
- Fix race condition in check_port(UDP) #78 |
| 8 |
|
| 9 |
2.20 2019-08-03T22:47:58Z |
| 10 |
|
| 11 |
- Fix test for . not in @INC by default #58 |
| 12 |
|
| 13 |
2.19 2017-05-11T13:02:47Z |
| 14 |
|
| 15 |
commit aac1b06c1e3b0d73ca964794bab0c444d454ebcc |
| 16 |
Author: Yasuhiro Matsumoto <[email protected]> |
| 17 |
Date: Thu May 11 18:40:13 2017 +0900 |
| 18 |
|
| 19 |
don't run command when the port is empty |
| 20 |
|
| 21 |
Test::TCP do hack on Windows for checking port used. |
| 22 |
|
| 23 |
https://github.com/tokuhirom/Test-TCP/commit/0f4510d8 |
| 24 |
|
| 25 |
But this way always return 0 with succeeded with empty port. This is not |
| 26 |
expected behavior. So this change fix to call check_port if the port is |
| 27 |
not defined. |
| 28 |
|
| 29 |
fixes #60 |
| 30 |
|
| 31 |
2.18 2017-04-24T06:00:25Z |
| 32 |
- let the kernel select an empty port #59 |
| 33 |
https://github.com/tokuhirom/Test-TCP/pull/59 |
| 34 |
(i110) |
| 35 |
|
| 36 |
2.17 2016-08-18T14:42:56Z |
| 37 |
- check_port can take 'host' argument(#55) |
| 38 |
- Fix test on Windows issue(#55) |
| 39 |
|
| 40 |
2.16 2016-04-13T02:27:43Z |
| 41 |
|
| 42 |
commit 3ae73a349f70f06c60de3f635916929bc8643429 |
| 43 |
Author: Alex Vandiver <[email protected]> |
| 44 |
Date: Tue Apr 12 00:33:21 2016 -0700 |
| 45 |
|
| 46 |
Ensure that IO::Socket::IP is DESTROY'd before can_bind returns |
| 47 |
|
| 48 |
be316f1f worked around a core perl bug[1] which causes Perl < 5.24.0 |
| 49 |
to DESTROY the IO::Socket::IP object too late, causing the empty port |
| 50 |
to not actually be empty if it was used immediately. |
| 51 |
|
| 52 |
Unfortunately, this change was lost in ba745fd1, which again put the |
| 53 |
return of `can_bind` on the stack along with the actual return value, |
| 54 |
re-instating the bug. |
| 55 |
|
| 56 |
Split the statement across two statements again, and add a comment on |
| 57 |
the importance of preserving it. |
| 58 |
|
| 59 |
[1] https://rt.perl.org/Public/Bug/Display.html?id=124248 |
| 60 |
|
| 61 |
2.15 2016-03-15T00:25:52Z |
| 62 |
|
| 63 |
- Add listen_socket function and listen option for race-free operation |
| 64 |
|
| 65 |
2.14 2015-09-29T22:36:44Z |
| 66 |
|
| 67 |
- Fix race condition in t/10_oo.t(exodist) |
| 68 |
|
| 69 |
2.13 2015-07-24T02:30:17Z |
| 70 |
|
| 71 |
- check whether the OS implements IPV6_V6ONLY before using it |
| 72 |
|
| 73 |
2.12 2015-05-18T08:14:30Z |
| 74 |
|
| 75 |
- Fixed spelling mistake |
| 76 |
(Reported by gregor herrmann) |
| 77 |
|
| 78 |
2.11 2015-04-07T00:07:25Z |
| 79 |
- declare IO::Socket::IP as dependency #36 |
| 80 |
|
| 81 |
2.10 2015-04-06T19:23:43Z |
| 82 |
- ensure the test object is DESTROYed when Net::EmptyPort::empty_port exits https://rt.cpan.org/Public/Bug/Display.html?id=103299 |
| 83 |
|
| 84 |
2.09 2015-04-02T21:55:18Z |
| 85 |
- fix tests running for a long time on systems that do not support IPv6 #35 |
| 86 |
|
| 87 |
2.08 2015-04-02T04:04:33Z |
| 88 |
- add `host` argument to various functions for binding to arbitrary address (incl. IPv6) #33 |
| 89 |
- add function `Net::EmptyPort::can_bind` #34 |
| 90 |
|
| 91 |
2.07 2015-01-22T10:44:21Z |
| 92 |
|
| 93 |
Old versions of Test::SharedFork are incompatibles with the new Test::Builder: I had "missing TB2::History" errors. |
| 94 |
It would be helpful to upgrade the Test::SharedFork dependency to help the user to avoid to encounter such hard to track deep errors. |
| 95 |
(Reported by dolmen++) |
| 96 |
|
| 97 |
|
| 98 |
2.06 2014-07-01T10:01:44Z |
| 99 |
|
| 100 |
commit 8259d5eb28919bc766c8b500151d5be7e944b7f2 |
| 101 |
Author: Petr Písař <[email protected]> |
| 102 |
Date: Fri Jun 27 13:37:20 2014 +0200 |
| 103 |
|
| 104 |
Wait infinitely if max_wait is negative |
| 105 |
|
| 106 |
The t/12_pass_wait_port_options.t will fail if the server process does not |
| 107 |
start listening in max_wait limit. This can happen if the host is |
| 108 |
loaded or just if the scheduler decides to postpone the process. |
| 109 |
|
| 110 |
This patch adds possibility to wait infitely by passing a negative |
| 111 |
max_wait value to the Test::TCP object and it changes the |
| 112 |
t/12_pass_wait_port_options.t test to use this feature. |
| 113 |
|
| 114 |
https://github.com/tokuhirom/Test-TCP/issues/28 |
| 115 |
Signed-off-by: Petr Písař <[email protected]> |
| 116 |
|
| 117 |
2.05 2014-06-24T00:49:45Z |
| 118 |
|
| 119 |
- Release again with latest minil. |
| 120 |
|
| 121 |
2.04 2014-06-23T23:42:28Z |
| 122 |
|
| 123 |
- Release. |
| 124 |
|
| 125 |
2.03_02 2014-06-23T23:37:07Z |
| 126 |
|
| 127 |
- Release to CPAN. |
| 128 |
|
| 129 |
2.03_01 2014-06-23T23:34:38Z |
| 130 |
|
| 131 |
- Switch to ExtUtils::MakeMaker. |
| 132 |
|
| 133 |
2.03 2014-06-23T10:18:53Z |
| 134 |
|
| 135 |
- Re-packaging with Minilla v2.0.0-TRIAL |
| 136 |
|
| 137 |
2.02 2013-10-30T03:22:39Z |
| 138 |
|
| 139 |
- Fixed fork(2) error handling. |
| 140 |
(tokuhirom) |
| 141 |
|
| 142 |
2.01 2013-09-22T04:13:53Z |
| 143 |
|
| 144 |
[Changes from MITHALDU (Christian Walde)] |
| 145 |
- test waitport argument by running full code-chain, not with partial mocking |
| 146 |
|
| 147 |
- The partial mocks cause stuck forks on win32 at times, causing the test to |
| 148 |
hang. With the full code chain present the test runs reliably |
| 149 |
|
| 150 |
2.00 2013-06-11T04:41:43Z |
| 151 |
|
| 152 |
[INCOMPATIBLE CHANGES] |
| 153 |
- Change wait_port() API to take max_wait seconds, and wait exponentially. |
| 154 |
Dropped wait_port_retry argument introduced at 1.27. |
| 155 |
(miyagawa) |
| 156 |
|
| 157 |
1.30 2013-06-03T05:06:28Z |
| 158 |
|
| 159 |
- Increase the default sleep and retry: 0.01 sec and 1000 times |
| 160 |
(miyagawa) |
| 161 |
|
| 162 |
1.29 2013-05-29T01:12:47Z |
| 163 |
|
| 164 |
- Care the `test_tcp(port => undef)` case. |
| 165 |
https://github.com/tokuhirom/Test-TCP/issues/13 |
| 166 |
|
| 167 |
1.28 2013-05-28T11:14:27Z |
| 168 |
|
| 169 |
- Added wait_port_sleep and wait_port_retry option for |
| 170 |
Test::TCP::wait_port, Test::TCP#new, test_tcp(). |
| 171 |
(tokuhirom) |
| 172 |
|
| 173 |
1.27 2013-05-15T10:13:18Z |
| 174 |
|
| 175 |
- export wait_port from Net::EmptyPort |
| 176 |
(Pavel Shaydo) |
| 177 |
- Make default wait_port() sleeping time shorter. |
| 178 |
0.1sec => 0.0001sec. |
| 179 |
(tokuhirom) |
| 180 |
|
| 181 |
1.26 2013-03-29T08:46:20Z |
| 182 |
|
| 183 |
- Remove unused deps from cpanfile. |
| 184 |
|
| 185 |
1.25 2013-03-29T08:07:24Z |
| 186 |
|
| 187 |
- re packaging, again & again |
| 188 |
|
| 189 |
1.24 2013-03-28T02:49:32Z |
| 190 |
|
| 191 |
- re-packing, again. |
| 192 |
|
| 193 |
1.23 2013-03-27T01:37:08Z |
| 194 |
|
| 195 |
- Just re-packaging |
| 196 |
|
| 197 |
1.22 2013-03-27T01:17:38Z |
| 198 |
|
| 199 |
- Add doc for Test::TCP::CheckPort |
| 200 |
(Tatsuhiko Miyagawa) |
| 201 |
- revert b4fc237697648234c960f6714d995210d4250e42 to fix Win32 breakage |
| 202 |
Test::TCP::CheckPort must be used on win32 because it prints to STDOUT |
| 203 |
(Christian Walde) |
| 204 |
- Add UDP support to Net::EmptyPort |
| 205 |
(Brendan Byrd) |
| 206 |
|
| 207 |
1.21 2013-03-03T12:33:49 |
| 208 |
|
| 209 |
- Added Net::EmptyPort::wait_port() |
| 210 |
(tokuhirom) |
| 211 |
|
| 212 |
1.20 2013-03-02T18:30:25 |
| 213 |
|
| 214 |
- bump up version |
| 215 |
|
| 216 |
1.19 2013-02-27T14:02:13 |
| 217 |
|
| 218 |
- Split empty_port() function to Net::EmptyPort |
| 219 |
(Thomas Klausner) |
| 220 |
|
| 221 |
1.18 2012-10-25 |
| 222 |
|
| 223 |
- Added more documentation about empty_port() |
| 224 |
|
| 225 |
1.17 2012-07-27 |
| 226 |
|
| 227 |
[Changes from MITHALDU (Christian Walde)] |
| 228 |
|
| 229 |
- fix port checking on Win32 by performing it in another process |
| 230 |
|
| 231 |
- On Win32 fork is emulating by creating another thread in the same process. |
| 232 |
This leads to a possible bug/race condition when a server tries to open a |
| 233 |
port and listen on it, while in the same process a client tries to connect |
| 234 |
to the same port. This manifests by the accept call of the server failing |
| 235 |
with an error of "Bad file descriptor". |
| 236 |
|
| 237 |
This is easily fixed by having another process perform the port checking, |
| 238 |
since that will not interfere with the internals. |
| 239 |
|
| 240 |
1.16 2012-06-30 |
| 241 |
|
| 242 |
[TEST FIX] |
| 243 |
- 1.15 breaks AIX fix on 1.14. |
| 244 |
t/05_sigint.t skips on perl <= 5.8.8. |
| 245 |
|
| 246 |
1.15 2012-01-31 |
| 247 |
|
| 248 |
[TEST FIX] |
| 249 |
- ${^CHILD_ERROR_NATIVE} is only available in perl 5.8.9 and later. |
| 250 |
(kazeburo++) |
| 251 |
|
| 252 |
1.14 2011-11-29 |
| 253 |
|
| 254 |
- Fixed testing issue on AIX. |
| 255 |
https://rt.cpan.org/Public/Bug/Display.html?id=72779 |
| 256 |
(w.phillip.moore) |
| 257 |
- depend to Test::SharedFork 0.19 for better TB2 support |
| 258 |
(tokuhirom) |
| 259 |
|
| 260 |
1.13 2011-05-31 |
| 261 |
|
| 262 |
- RT#67292: Tests are blocking in Windows 7. |
| 263 |
https://rt.cpan.org/Ticket/Display.html?id=67292 |
| 264 |
- change port number range from 10000 .. 11000 to 50000 .. 60000 to respect IANA. |
| 265 |
https://rt.cpan.org/Ticket/Display.html?id=64012 |
| 266 |
|
| 267 |
1.12 2011-03-03 |
| 268 |
|
| 269 |
- workaround for win32 test fails. |
| 270 |
https://rt.cpan.org/Ticket/Display.html?id=66016 |
| 271 |
- more diagnostic messages |
| 272 |
|
| 273 |
1.11 2010-12-20 |
| 274 |
|
| 275 |
- localize $@ in Test::TCP::DESTROY |
| 276 |
|
| 277 |
1.10 2010-12-18 |
| 278 |
|
| 279 |
- added Test::TCP->stop method |
| 280 |
|
| 281 |
1.09 2010-12-18 |
| 282 |
|
| 283 |
- added document in FAQ section |
| 284 |
|
| 285 |
1.08 2010-12-16 |
| 286 |
|
| 287 |
- no feature changes |
| 288 |
|
| 289 |
1.07_01 2010-12-16 |
| 290 |
|
| 291 |
- new OO interface! |
| 292 |
|
| 293 |
1.07 2010-11-11 |
| 294 |
|
| 295 |
- allow forking in the client(lestrrat) |
| 296 |
|
| 297 |
1.06 2010-09-11 |
| 298 |
|
| 299 |
- depend to Test::SharedFork 0.14. It fixes issue with Test::Builder2. |
| 300 |
|
| 301 |
1.05 2010-09-10 |
| 302 |
|
| 303 |
- workaround for $@ issue in test case. |
| 304 |
Data::Util's method modifier clears $@. |
| 305 |
|
| 306 |
1.04 2010-08-24 |
| 307 |
|
| 308 |
- fixed local $@ issue. this happens on some version of perl5. |
| 309 |
|
| 310 |
1.03 2010-08-23 |
| 311 |
|
| 312 |
- release to cpan |
| 313 |
- fixed win32 issue(charsbar) |
| 314 |
|
| 315 |
1.02_02 2010-08-23 |
| 316 |
|
| 317 |
- use randomness on finding empty port(suggested by kazuhooku) |
| 318 |
- try to connect the port before bind(Tatsuhiko Miyagawa) |
| 319 |
|
| 320 |
1.02_01 2010-08-23 |
| 321 |
|
| 322 |
- better cleanup code by RAII pattern. |
| 323 |
https://rt.cpan.org/Ticket/Display.html?id=60657 |
| 324 |
(reported by dgl) |
| 325 |
|
| 326 |
1.02 2010-08-17 |
| 327 |
|
| 328 |
- lazy loading issue was fixed at Test::SharedFork 0.12. |
| 329 |
Depend to it. |
| 330 |
https://rt.cpan.org/Public/Bug/Display.html?id=60426 |
| 331 |
(reported by J.) |
| 332 |
|
| 333 |
1.01 2010-08-15 |
| 334 |
|
| 335 |
- remove unused deps for use_test_base(). |
| 336 |
|
| 337 |
1.00 2010-08-08 |
| 338 |
|
| 339 |
- bump up version! |
| 340 |
|
| 341 |
0.16_02 2010-02-20 |
| 342 |
|
| 343 |
- oops. packaging miss. |
| 344 |
|
| 345 |
0.16_01 2010-02-20 |
| 346 |
|
| 347 |
- Do not depend to IO::Socket::INET 1.31. |
| 348 |
Test::TCP works well with older IO, I hope. |
| 349 |
(suggested by mst) |
| 350 |
|
| 351 |
0.16 2010-01-04 |
| 352 |
|
| 353 |
- check port asap(suggested by Hideki YAMAMURA) |
| 354 |
|
| 355 |
0.15 2009-11-28 |
| 356 |
|
| 357 |
- fixed win32 issue(by charsbar++) |
| 358 |
|
| 359 |
0.14 2009-10-18 |
| 360 |
|
| 361 |
- support win32 |
| 362 |
|
| 363 |
0.13 2009-10-14 |
| 364 |
|
| 365 |
- handle sigint(reported by kazuho++) |
| 366 |
|
| 367 |
0.12 2009-10-04 |
| 368 |
|
| 369 |
- fixed broken test on solaris |
| 370 |
|
| 371 |
0.11 2009-09-17 |
| 372 |
|
| 373 |
- fixed broken test, reported by drangon3++ |
| 374 |
|
| 375 |
0.10 2009-09-17 |
| 376 |
|
| 377 |
- don't leak the process when get a SIGDIE |
| 378 |
- diag when get SIGPIPE |
| 379 |
|
| 380 |
0.09 2009-09-15 |
| 381 |
|
| 382 |
- don't leak the control by SIGTERM, some tcp server return the control by SIGTERM. |
| 383 |
(reported by kazuho++) |
| 384 |
|
| 385 |
0.08 2009-09-14 |
| 386 |
|
| 387 |
- oops. I forget to remove 'use Params::Validate' statement. |
| 388 |
|
| 389 |
0.07 2009-09-13 |
| 390 |
|
| 391 |
- diag when get a SIGABRT |
| 392 |
|
| 393 |
0.06 2009-09-08 |
| 394 |
|
| 395 |
- pass the pid to client(kazuho) |
| 396 |
- remove deps for Params::Validate(tokuhirom) |
| 397 |
|
| 398 |
0.05 2009-06-22 |
| 399 |
|
| 400 |
- updated docs |
| 401 |
- set ReuseAddr as true |
| 402 |
|
| 403 |
0.04 2009-03-25 |
| 404 |
|
| 405 |
- Test::SharedFork 0.03 has a bug. depend to 0.04 |
| 406 |
|
| 407 |
0.03 2009-03-25 |
| 408 |
|
| 409 |
- do not depend to Sub::Exporter |
| 410 |
- use Test::SharedFork |
| 411 |
|
| 412 |
0.02 2008-08-28 |
| 413 |
|
| 414 |
- fixed deps for Sub::Exporter(by yappo++) |
| 415 |
|
| 416 |
0.01 2008-08-17 |
| 417 |
- original version |
| 418 |
|