[#30589] [Bug #3391] Use single exclamation mark instead of double exclamation mark for IRB — Diego Viola <redmine@...>

Bug #3391: Use single exclamation mark instead of double exclamation mark for IRB

10 messages 2010/06/04

[#30672] [Bug #3411] Time.local 1916,5,1 #=> 1916-04-30 23:00:00 +0100 — Benoit Daloze <redmine@...>

Bug #3411: Time.local 1916,5,1 #=> 1916-04-30 23:00:00 +0100

12 messages 2010/06/08

[#30699] [Bug #3419] 1.9.2-preview3 possible bug with Rails 3 active_record sqlite_adapter — Joe Sak <redmine@...>

Bug #3419: 1.9.2-preview3 possible bug with Rails 3 active_record sqlite_adapter

9 messages 2010/06/09

[#30734] [Bug #3428] ri outputs ansi escape sequences even when stdout is not a tty — caleb clausen <redmine@...>

Bug #3428: ri outputs ansi escape sequences even when stdout is not a tty

11 messages 2010/06/11

[#30756] [Feature #3436] Spawn the timer thread lazily — Maximilian Gass <redmine@...>

Feature #3436: Spawn the timer thread lazily

15 messages 2010/06/13
[#32686] [Ruby 1.9-Feature#3436] Spawn the timer thread lazily — Mark Somerville <redmine@...> 2010/10/04

Issue #3436 has been updated by Mark Somerville.

[ruby-core:30931] Re: [RubyInstaller] undefined reference to 'rb_encdb_declare'; ruby-1.9.2-preview3 64-bit on Windows

From: Chuck Remes <cremes.devlist@...>
Date: 2010-06-27 20:04:10 UTC
List: ruby-core #30931
On Jun 27, 2010, at 1:56 PM, Luis Lavena wrote:

> On Sun, Jun 27, 2010 at 2:36 PM, Chuck Remes <cremes.devlist@mac.com> wrote:
>> [cross-posted to rubyinstaller ML]
>> 
>> I'm working on getting ruby-1.9.2-preview3 to compile under mingw-w64 (64-bit) on Windows 7 x64. So far I have only had to make one minor change to the sources (an additional check in win32.h). When I get a complete and working build, I'll be sure to send patches upstream.
>> 
>> Right now I am stuck trying to get the encodings to compile via the generated enc.mk makefile. I have a complete list of the make output here along with the generated enc.mk file:
>> 
>> http://gist.github.com/455064
>> 
>> It can't resolve the references to rb_encdb_declare, rb_encdb_alias and rb_encdb_replicate which are in the encoding.o object file. I thought this was probably a problem with the enc.mk file but it looks okay to me. I diff'ed it against the same file for an OSX install and the differences looked reasonable based on the different platforms.
>> 
>> Does anyone have any idea how I can get ld to use encoding.o to resolve those symbols when building the encodings? I've been banging my head on this for a few hours now and cannot fix it myself. I can provide additional info (config.log snippets, etc.) if you need more information.
>> 
> 
> can you check if the shared library (libmsvcrt-ruby191.dll.a) really
> contain the exported symbols? (There is no export-all displayed in
> line 95..97 of make output)

Luis, thank you for the hint. None of the symbols were being exported into the dll.

I am currently looking at the win32/mkexports.rb file. It is trying to open the libmsvcrt-ruby191-static.a file and read the symbols from it. When I run "strings" against that library it lists all of the rb_* names plus a few more.

Working through the mkexports.rb file, the #objdump method doesn't output anything.

  def each_export(objs)
    objdump(objs) do |l|
      next if /@.*@/ =~ l
      yield $2, !$1 if /\s(?:(T)|[[:upper:]])\s_((?!Init_|.*_threadptr_|DllMain@).*)$/ =~ l
    end
  end

I am printing out "l" in this case and see that it is a triplet of data separated by spaces.

Example:

0000000000004860 T rb_ascii8bit_encindex\n

That regex is not matching that data at all, so all of the lines are rejected and nothing gets exported.

Luis, if you have that static library on your machine can you give me a sample of the input it is providing to this method? I can then work on crafting a new regex to filter out the unnecessary symbols.

cr


In This Thread