1 Revision history for indirect
3 0.31 2013-09-05 16:45 UTC
4 + Fix : [RT #88428] : no indirect in eval can trigger for direct calls
6 Thanks Graham Knop for reporting.
7 + Tst : Author tests are no longer bundled with this distribution.
8 They are only made available to authors in the git repository.
10 0.30 2013-05-16 15:55 UTC
11 + Fix : [RT #83806] : false positives with Devel::Declare
12 [RT #83839] : false positive using ? : syntax
13 Thanks Andrew Main for the patch.
14 However, please note that the reason this patch seems to fix
15 thinks has not been explained.
16 + Fix : [RT #84649] : incorrect RT link in metadata
17 Thanks Karen Etheridge for reporting.
19 0.29 2013-03-05 01:30 UTC
20 + Fix : [RT #83659] : false positives
21 Proper method calls in string-like environments (like
22 "@{[ $x->new ]}" will no longer be reported as indirect.
23 This was a regression in 0.28.
24 Thanks Andrew Main for reporting.
25 + Fix : Broken linkage on Windows with gcc 3.4, which appears in
26 particular when using ActivePerl's default compiler suite.
27 For those setups, the indirect shared library will now be
28 linked against the perl dll directly (instead of the import
31 0.28 2013-02-26 17:05 UTC
32 + Fix : [RT #83450] : newlines confuse indirect
33 Perl sometimes resets the line buffer between the object and
34 the method name (e.g. for "sort Class\n->method" outside of
35 eval), and this could cause direct method calls to be reported
37 Thanks Gianni Ceccarelli for reporting.
38 + Fix : Check functions are now replaced and restored in a thread-safe
39 manner, either by using the wrap_op_checker() function from perl
40 when it is available (starting from perl 5.16) or by taking the
41 OP_REFCNT mutex on older perls.
43 0.27 2013-01-30 19:00 UTC
44 + Fix : [RT #82562] : indirect/Devel::CallParser interaction
45 indirect has been taught to play nicely with Devel::CallParser.
46 Thanks Andrew Main for the patch.
47 + Tst : Author tests overhaul.
49 0.26 2011-10-23 14:25 UTC
50 + Add : "no indirect 'global'" enables the pragma for the whole program,
51 except for lexical scopes that "use indirect" explicitely.
52 + Chg : Passing both the 'fatal' and 'hook' options to unimport() is
53 now considered an error, and will result in an exception.
54 unimport() used to consider only the first passed option of
55 those two, and silently ignored the other.
56 + Tst : Test failures of t/41-threads-teardown.t and t/50-external.t on
57 Cygwin should have been addressed.
58 + Tst : Threads tests will not fail anymore if resources constraints
59 prevent the system from creating all the required threads.
61 0.25 2011-08-24 15:40 UTC
62 + Fix : RT #69291 is now also fixed for perl 5.8. The pragma will no
63 longer vivify the "indirect" entry in the hints hash %^H on
65 + Tst : Attempt to make t/50-external.t pass on Cygwin.
67 0.24 2011-07-17 23:15 UTC
68 + Fix : [RT #64521] : "no indirect" leaking into eval.
69 This is currently only fixed for perl 5.10 (perl 5.12 and
70 higher were never affected). It was caused by a very stupid
71 mistake of mine that was introduced in indirect version 0.23.
72 Thanks Michael G Schwern for reporting.
73 + Fix : [RT #69291] : indirect.pm breaks %^H.
74 This was caused by the same mistake as for the previous bug,
75 and as such it is also only fixed for perl 5.10 (and never
77 Thanks Andrew Main for reporting.
78 + Doc : C++ compilers are officially NOT supported.
80 0.23 2010-10-03 00:15 UTC
81 + Fix : Some indirect constructs could be incorrectly reported when
82 several modules were used in the same scope. This caused
83 t/30-scope.t to fail randomly.
84 + Tst : Threads tests are now only run on perl 5.13.4 and higher.
85 They could segfault randomly because of what seems to be an
86 internal bug of Perl, which has been addressed in 5.13.4.
87 There is also an environment variable that allows you to
88 forcefully run those tests, but it should be set only for
89 author testing and not for end users.
91 0.22 2010-08-16 16:00 UTC
92 + Add : Indirect constructs are now reported for code interpolated
93 in quote-like environments, like "${\( ... )}", "@{[ ... ]}",
94 s/pattern/ ... /e, qr/(?{ ... })/ or qr/(??{ ... })/.
95 + Add : You can now make the pragma lethal by passing anything matching
96 /^:?fatal$/i to import(), including "FATAL" and ":Fatal".
97 + Fix : [RT #60378] : segmentation fault on indirect_ck_method.
98 This caused constructs like "@{[ $obj->$meth ]}" to segfault
99 when $meth was a lexical.
100 Thanks Tokuhiro Matsuno for reporting.
102 0.21 2010-05-31 23:10 UTC
103 + Chg : perl 5.8.1 is now required (instead of 5.8.0).
104 + Fix : [RT #57699] : indirect fail with 64-bit int on 5.13.1.
105 It was actually a problem with thread destructors segfaulting
106 because they weren't called at the right time anymore.
107 Thanks Andrew Main for reporting.
108 + Tst : A few more regression tests about the scope leak bug.
110 0.20 2010-04-18 21:25 UTC
111 + Fix : [RT #50570] : "indirect" leaking into LWP.
112 Thanks Andrew Main for reporting.
113 More generally, the require propagation workaround on 5.8-5.10.0
114 has been overhauled, and other scope leaks should be fixed.
115 + Fix : Test failures with 5.12 on Windows where Strawberry Perl crashes
116 because the SystemRoot environment variable is missing.
117 + Fix : Work around Kwalitee test misfailures.
119 0.19 2009-08-28 18:40 UTC
120 + Add : The new constant I_FORKSAFE can be tested to know whether the
121 module will behave nicely when fork()ing. It's currently always
122 true except on Windows where you need perl 5.10.1 for it to be
124 + Fix : I_THREADSAFE and I_FORKSAFE ought to be true when
125 PERL_INDIRECT_PM_DISABLE is set.
126 + Fix : The pragma could still leak if you passed to the "hook" option
127 a reference to the same (named) subroutine from two different
128 require scopes. The fix also provides a better solution for
130 + Fix : Segfaults when indirect is loaded for the first time from
132 + Fix : Leaks of memory associated with the root interpreter.
133 + Opt : Less memory will be used for non-threaded perls version 5.10.0
134 and below, and for threaded perls from version 5.10.1.
136 0.18 2009-08-23 16:15 UTC
137 + Add : When set, the PERL_INDIRECT_PM_DISABLE environment variable
138 disables the pragma globally.
140 0.17 2009-07-16 12:10 UTC
141 + Fix : [RT #47902] : "no indirect" leaking again.
142 This actually turned out to be a bug in how the hook coderefs
143 were stored in the hints hash.
144 Thanks Andrew Main for reporting once again.
145 + Fix : t/80-regressions.t failing on Windows.
146 + Tst : Yet more cleanups.
148 0.16 2009-07-14 16:50 UTC
149 + Add : Indirect calls on blocks are now reported. For those, '{' is
150 passed to the hook as the object description.
151 + Add : The new indirect::msg() function publicizes the default
152 warning/exception message.
153 + Fix : [RT #47866] : Segfault with UTF-8 regexps.
154 Thanks Andrew Main for reporting.
157 0.15 2009-07-08 22:55 UTC
158 + Fix : Invalid constructs with the same method and package name were
160 + Fix : The error line number used to point to the end of the expression
161 instead of its beginning.
163 0.14 2009-06-04 21:55 UTC
164 + Fix : Prevent bogus invalid syntaxes caused by reallocated memory
165 chunks. Thanks Andrew Main for reporting with a reproducible
168 0.13 2009-05-24 18:50 UTC
169 + Add : The hook now receives the file name and the line where the
170 error happened in respectively $_[2] and $_[3].
171 + Fix : Pass mortalized copies of the method name and object to the
172 hook. This seems to fix some rare crashes.
173 + Fix : Work around a bug in perl 5.10.0 and lower. Thanks Andrew Main
174 for teaching me about this issue.
175 + Fix : Report the correct file in error messages (a regression from the
178 0.12 2009-05-03 14:30 UTC
179 + Add : You can specify the action to execute for each indirect
180 construct encountered with the new "hook => $coderef" unimport()
182 + Chg : A ptable is now used internally for the op => position mapping.
183 + Fix : The pragma is now built with thread safety features enabled
184 whenever possible (a notable exception is perl 5.8.x on Win32,
185 as something seems wrong with its context handling).
186 The new indirect::I_THREADSAFE() constant reflects this.
187 + Fix : A negation precedence nit in indirect_ck_entersub().
188 + Tst : "use/no indirect" while parsing an indirect construct.
189 + Tst : Thread safety.
191 0.11 2009-02-08 18:35 UTC
192 + Fix : Potential collisions by hashing pointers with a wrong format.
193 + Upd : Resources in META.yml.
195 0.10 2009-01-17 12:40 UTC
196 Re-release 0.09_01 as stable.
198 0.09_01 2008-12-08 17:55 UTC
199 + Fix : Invalid memory read with "${\(new Foo)}" constructs. The
200 corresponding test is turned back on.
201 + Tst : Refinements in t/30-scope.t
203 0.09 2008-12-05 20:35 UTC
204 + Add : Support for perl 5.8.
205 + Tst : Skip a test in t/10-good.t that randomly segfaults for (I guess)
206 systems stricter than linux in the way they manage their memory.
208 0.08 2008-10-22 14:45 UTC
209 + Fix : A rare edge case for package whose names are prefix of 'main'.
210 + Tst : Test $$ as variable and state variables.
212 0.07_03 2008-10-17 20:10 UTC
213 + Add : Support and tests for variables with spaces after the sigil.
214 + Upd : META.yml spec updated to 1.4.
216 0.07_02 2008-10-15 21:10 UTC
217 + Add : Support and tests for package variables.
218 + Tst : Coverage improved by removing dead code.
220 0.07_01 2008-10-15 16:00 UTC
221 + Fix : [RT #40055] : Not handling RV2SV => GV(SV) correctly, which
222 could cause 'no indirect; print' segfaults. Thanks Goro Fuji for
225 0.06 2008-10-11 16:45 UTC
227 + Tst : Test "no indirect 'anything'", "foo Class->bar", and indirect
228 uses of exec() and system().
230 0.05 2008-10-02 14:40 UTC
231 + Chg : Now the old check function is always called before storing an op
233 + Fix : Misc code and docs refinements.
235 0.04 2008-08-30 19:00 UTC
236 + Fix : Clean up the op->src hash when we're done with an entersub.
237 + Tst : No longer fork for testing. IPC::Cmd isn't required anymore.
239 0.03 2008-08-12 15:25 UTC
240 This release is kindly supported by Copenhagen Hotel Centrum WiFi.
241 + Fix : Tests used not to pass PERL5OPTS to their kids. This lead to
242 failures under CPAN. I think.
245 0.02 2008-08-11 15:55 UTC
246 + Fix : Some building failures with old gcc versions that didn't seem to
247 like the ((hint == 2) ? croak : warn)(msg) construct. I think.
248 + Rem : Unused cruft from a previous implementation.
249 + Tst : Fail more gracefully when we can't capture buffers or when the
250 child returned an error.
252 0.01 2008-08-10 20:40 UTC
253 First version, released on an unsuspecting world.