From: Run Paint Run Run Date: 2010-06-02T16:07:37+09:00 Subject: [ruby-core:30555] Re: [Ann] Contribution wanted: identify tickets for 1.9.2 release > For example, some may be registered in "Ruby 1.9.1" subproject, even > though it is reproducible in trunk. > But I can't afford to check all tickets of all subproject. I've checked all open bug reports for 1.9.1 and summarised them below. I found one segfault reproducible on trunk (#1934), and a couple that can be closed, but nothing else, I'm afraid. * Bug #1573 [ruby-core:23717]: $0 behaves unexpectedly A patch was posted, OK'd by matz, and is waiting to be applied. * Bug #1569 [ruby-core:23703]: gem update --system crashes ruby with a segmentation fault. The report concerns a Windows bug, which I can't reproduce as I don't use that "OS". However, `gem update --system` is such a common code path, that if it caused a segfault reproducibly we surely would have seen additional reports. Further, the report concerns XP which is now EoL'd... This can be closed. * Bug #1765 [ruby-core:24271]: rdoc runs out of memory on Windows XP with 4GB RAM Another XP report. I can't reproduce it under Linux with trunk; in my case the given command was CPU bound, while the memory usage was quite reasonable for Ruby. * Bug #1764 [ruby-core:24270]: rdoc for JSON comes out garbled in Ruby 1.9.1 Neither Eric nor I can reproduce this with trunk. Given the age of the ticket and the lack of reproducible test cases, I suspect this can be closed. * Bug #1615 [ruby-dev:38598]: ������������������������������������������CSV������������������������������������������������������������Segmentation Fault������������ Again, there appears to be no test case here and the reporter was using XP. Short of debugging via telepathy, I don't see how this ticket is actionable. * Bug #1544 [ruby-core:23642]: rb_io_write bug? It's unclear whether this affects trunk as we lack means to reproduce. The reporter is responsive, so this is correctly tagged Feedback. * Bug #2718 [ruby-core:28089]: DRb mixing up function return values between PIDs after fork() This does have a test case. I can't reproduce the errant behaviour on trunk, but do see it on 1.8.7 (p249). I don't have any other 1.9 builds installed, so can't bisect further. This would benefit from the attention of somebody familiar with DRb so we can confirm (a) this represents a bug and (b) which versions it affects. * Bug #2714 [ruby-core:28066]: Interpreter deadlock when using ::Socket.pair This is actually a cygwin report, and as Yusuke reported in [ruby-core:30391] that platform lacks a maintainer. If this ticket is to be kept open, it should probably be downgraded to Low, for the same reason Yusuke gave in the aforementioned message. * Bug # 2701 [ruby-dev: 40279]: rubyspec: BigDecimal # div returns NaN if divided by zero and no precision given ERROR This appears to be a spec question. Presumably for Yugui to decide? * Bug #2638 [ruby-core:27757]: ruby-1.9.1-p37[68] build on aix5.3 with gcc-4.2 failed to run for me because it ignores where libgcc is located. AIX config issue. The reporter seems to be saying that he no longer has access to an AIX box, in which case its resolution depends on whether Yutaka Kanemoto can work with the information given. * Bug #2630 [ruby-core:27699]: generator.rb missing from 1.9.1 standard library? I see no reason for this ticket to remain open: the removal of the library was intentional. * Bug #2596 [ruby-core:27558]: Openssl Segmentation Faults on Red Hat Enterprise Linux 5.3 Unless somebody with RHEL is willing to debug, this report is probably hopeless. * Bug #2570 [ruby-core:27468]: ruby never stops executing when installing fxruby-1.6.20-x86-mingw32.gem This is a MinGW issue, so presumably should be assigned to nobu. (It's currently unassigned). * Bug #2528 [ruby-core:27310]: readline bug A potential segfault on PowerPC, reported last year. Been waiting for feedback for over six months, and there appears to be no maintainer for the platform. This should be closed. * Bug #2399 [ruby-core:26898]: pthread_mutex_lock: 22 Intermittent segfault reported on Darwin. No test case. If this is to remain open it needs a user of that platform to attempt debugging on the basis of a two-sentence description. * Bug #2391 [ruby-core:26858]: in CentOS after some times Ruby 1.9.1 get "segmentation fault" This has been assigned High status since last year, which is presumably incorrect. The good news is that purportedly a simple `gem list` will trigger it; the bad news is no other user on this platform has offered to reproduce. If this is to remain open, its status needs downgrading. * Bug #2383 [ruby-core:26813]: windows - calling FreeEnvironmentStrings on envarea invalidates memory in use The platform maintainer has responded, but I don't understand the conclusion or why it was orphaned. If this does represent a bug, it should probably be triaged. * Bug #2358 [ruby-core:26693]: Segmentation fault using dl/import MinGW issue. Seemingly has a testcase, but nobody appears to have tried it. * Bug #2339 [ruby-core:26554]: DRb leaves connections half-open A good report complete with patch. Needs the DRb maintainer to respond. * Bug #2309 [ruby-core:26422]: Segfault, repeatable testcase, possibly related to [2295], [2019], [1937]. This needs attention. The reporter was asked for further information, which he provided, but no follow-up occurred. I suspect this should be triaged at least as it is currently unassigned. * Bug #2295 [ruby-core:26371]: segmentation faults Again, the reporter followed-up on questions, but no action was taken. Given that the issue is intermittent and sans test case, I suspect this will have to be closed, but a review of the data would seem prudent. To this end, the next step is presumably to assign someone the ticket. * Bug #2235 [ruby-core:26168]: Segmentation fault with PTY.spawn Clear report with simple test case. I can't reproduce it on trunk or an ancient version of 1.9 I have, but Naruse can reproduce it with p376. If narrowing down the affected versions would help, I'll give it a try. * Bug #2019 [ruby-core:25203]: Segmentation fault in ruby 1.9.1-p243 on second request Reported Rails crash with no hint of reproduction. Awaiting feedback since last year. This can be closed. * Bug #1942 [ruby-core:24935]: irb not running after installation This appears to be a readline issue. A trivial patch is suggested in the comments with a report that it fixes the issue. A quick appraisal suggests that the same code exists in trunk, so it would be useful if the patch was reviewed. * Bug #1940 [ruby-core:24927]: Segmentation fault on TestFiber#test_many_fibers_with_threads (make check) Claims that the core test suite segfaults on MinGW. Has been designated as High since last year. This should be easy to test for somebody with access to the platform. * Bug #1937 [ruby-core:24919]: ruby 1.9.1p243 (2009-07-16 revision 24175) [x86_64-linux] seg.fault Yet another ActiveRecord segfault with 1.9 reported last year. No method of reproduction is given. This needs closing. * Bug #1934 [ruby-core:24894]: Segmentation fault I can reproduce this on trunk. I've updated the ticket. * Bug #1907 [ruby-core:24812]: ssl_timeout cannot be set in HTTPS request A patch is provided and has been OK'd. Needs applying. * Bug #1856 [ruby-core:24664]: [BUG] Segmentation fault in 1.9.1p149 Has testcase. Neither Yusuke nor I can reproduce it on trunk. * Bug #1773 [ruby-core:24321]: Gem path doesn't honor user gem? AFAICT, the original bug has been fixed and the ticket now constitutes a feature request: teach gem_prelude to read ~/.gemrc. If so, perhaps the ticket could be reclassified. * Bug #1767 [ruby-core:24297]: cgi/session/pstore generating filenames with to less randomness This reads like a feature request to me. In any case, this ticket was opened last year but has yet to be reviewed by its maintainer. If this is indeed judged a security issue, its priority should be increased; otherwise I suspect it would be more appropriate for the Features tracker. * Bug #3373 [ruby-core:30536]: Segmentation fault when trying access project via Mozilla Firefox In which I realise that matz's patience far exceeds mine... It appears that a segfault with the ActiveRecord PostgreSQL driver is being reported. If I can find the book, I'll see what I can infer. * Bug #3215 [ruby-core:29845]: Windows: IO.popen returns "handle is invalid" Possible bug on mswin. The Windows maintainer is looking at it. * Bug #3353 [ruby-core:30455]: Building 32-bit only on Snow Leopard (dylib failed to build) autoconf issue on the Mac. No comments yet; needs triage? * Bug #3349 [ruby-core:30445]: [BUG] cfp consistency error - send Awaiting feedback. * Bug #3293 [ruby-core:30239]: fails to build with gcc 4.4.3 Reported on Debian/Lucid. I haven't built 1.9.1 recently, but trunk builds on Lucid with GCC 4.4.3. The 1.9.1 aspect is being tracked downstream. * Bug #3292 [ruby-core:30238]: Fail 2 IO tests during make test for ruby-1.9.1-p376 with Cygwin Prioritised as Low due to lack of maintainer. * Bug #3276 [ruby-core:30155]: webrick/server.rb:90: [BUG] Segmentation fault Awaiting steps to reproduce... * Bug #3268 [ruby-core:30139]: Time.parse is not restoring the timezone properly on Windows MinGW bug. Assign to nobu? * Bug #3243 [ruby-dev:41138]: Can't build Ruby after r26506 I assume Yuki is dealing with this. If more data is needed, I'm happy to help. * Bug #3212 [ruby-core:29835]: ConditionVariable may become inconsistent for interrupted threads Unassigned, but Yusuke is actively working on it. * Bug #3201 [ruby-core:29788]: Backport request for r25932 to 1.9.1 branch (gem_prelude.rb) Misfiled backport request; needs moving to the right tracker. * Bug #3194 [ruby-core:29753]: Segmentation fault in rb_hash_tbl Unassigned, but nobu's working on it. * Bug #3193 [ruby-core:29750]: uint is not parsed in dl/cparser.rb Presumably trivial, but needs triage. * Bug #3190 [ruby-core:29739]: Segmentation fault gzip/tar Unassigned, but feedback has been requested; the status needs updating. * Bug #3189 [ruby-core:29738]: [BUG] Segmentation fault Ditto. * Bug #3188 [ruby-core:29737]: remove_method.rb:3: [BUG] Segmentation fault Yet another ActiveRecord segfault on the Mac... The ticket hasn't been responded to, but presumably more information is required. * Bug #3178 [ruby-core:29649]: Fileutils#rmdir does rescue Dir Errors without raising new No responses yet. Looks to be more of a spec issue. FileUtils doesn't have a maintainer, so this needs triage. * Bug #3174 [ruby-core:29621]: segmentation fault at running Array method each Neither tenderlove nor the reporter can reproduce it on 1.9.2. * Bug #3138 [ruby-core:29477]: Segmentation fault Rails segfault #762728. Feedback was requested and has been received. Needs triage. * Bug #3057 [ruby-core:29159]: Segmentation fault: object allocation during garbage collection phase Awaiting feedback. * Bug #3006 [ruby-core:28927]: Compiling ext/etc problems (windows) Reporter claims issue to be resolved. This can probably be closed. * Bug #2791 [ruby-core:28333]: Segmentation Fault in 1.9.1.p376 Awaiting feedback. * Bug #2783 [ruby-core:28315]: Segmentation Fault (using libxml-ruby) MinGW issue. Awaiting feedback. * Bug #2780 [ruby-core:28298]: "Handle is invalid" when accessing the file system from different threads (win32) MsWin issue. Feedback requested; status needs updating. * Bug #2738 [ruby-core:28149]: Segmentation fault - rake test Reporter has resolved the problem so asked for the ticket to be closed.