Skip to content

HTTPS clone URL

Subversion checkout URL

You can clone with HTTPS or Subversion.

Download ZIP
Commits on Feb 4, 2012
  1. John Firebaugh

    Don't use char where an unsigned type is needed

    jfirebaugh authored
    With char, the C cast to native_int produces a negative integer
    in some situations.
    
    Fixes #1513.
Commits on Feb 3, 2012
  1. Evan Phoenix
  2. Evan Phoenix

    Fix #1 of the JIT stalling GC runs

    evanphx authored
    This allows us to generate the IR while being gc dependent, but then
    switch independent while running LLVM's opt passes and generating
    machine code.
  3. Dirkjan Bussink
  4. Dirkjan Bussink
  5. Dirkjan Bussink

    Remove define so we don't use a global datastructure in Oniguruma

    dbussink authored
    Usage of this shared global datastructure is a big problema. We
    should never ever use this since it is fundamentally broken and
    was causing spurious crashes during compilation of regular expressions.
  6. John Firebaugh
  7. John Firebaugh

    Move coercion into parse_mode

    jfirebaugh authored
    Fixes 'IO.popen coerces mode argument with #to_str'
    
    I had to remove the interpolation of str_mode into the exception
    string, but that was buggy anyway, because it was uninitialized
    in the case that mode was an Integer to begin with.
  8. John Firebaugh
  9. John Firebaugh

    Clean up IO.popen specs

    jfirebaugh authored
    * Use finer grained examples
    * Close io consistently in after block
    * Use have_data
  10. John Firebaugh

    Add myself to AUTHORS

    jfirebaugh authored
  11. Jesse Cooke

    Merge pull request #1542 from leocassarani/patch-1

    jc00ke authored
    Delete misplaced ".should" in DateTime spec
  12. Leo Cassarani
Commits on Feb 2, 2012
  1. Luis Lavena

    Minor correction to typo

    luislavena authored
  2. Evan Phoenix
  3. Evan Phoenix
Commits on Feb 1, 2012
  1. Brian Shirai
  2. Dirkjan Bussink
  3. Dirkjan Bussink
  4. Dirkjan Bussink

    Add spec for having spaces before the block arguments

    dbussink authored
    This was causing a crash in the Rubinius compiler.
Commits on Jan 31, 2012
  1. Evan Phoenix
  2. Evan Phoenix

    Fix 2 IO gc issues

    evanphx authored
    Both caused objects to allocated in the mature gen far too often,
    causing many more full collections that was needed.
  3. Dirkjan Bussink
  4. Dirkjan Bussink
  5. Evan Phoenix
  6. Evan Phoenix
Commits on Jan 30, 2012
  1. Dirkjan Bussink
  2. Dirkjan Bussink

    Use Hash.allocate instead of Hash.new

    dbussink authored
    MRI also doesn't call the constructor in this case.
    
    Fixes #1536
  3. Dirkjan Bussink
  4. Dirkjan Bussink

    Increase the JIT thresholds

    dbussink authored
    I've investigated this too and I can confirm the numbers of
    @hosiawak as decribed in #1426. I've also made the threshold
    of when not to include methods into the inlining anumore an
    option.
    
    Fixes #1426
Commits on Jan 29, 2012
  1. Dirkjan Bussink
  2. Dirkjan Bussink

    Fix potential OpenSSL segfault

    dbussink authored
    This was only recently backported to the 1.8.7 OpenSSL extension
    in MRI but was already fixed long ago in 1.9.x. Since the backport
    to 1.8.7 only happened after we last upgraded the 1.8.7 version,
    we need to add this fix too.
    
    See http://bugs.ruby-lang.org/issues/show/3000 and
    http://svn.ruby-lang.org/cgi-bin/viewvc.cgi?view=revision&revision=33999
    
    Fixes #1514 and #1509
Commits on Jan 28, 2012
  1. Dirkjan Bussink

    Use nil<>() properly

    dbussink authored
  2. Dirkjan Bussink

    Initialize inliners_ to 0

    dbussink authored
  3. Dirkjan Bussink

    Register Inliners properly as a CodeResource so it can be cleaned up

    dbussink authored
    This means we also need to mark Inliners which now happens in the
    mark_inliners function used by various mark functions for different
    Executable's.
    
    This also showed that NativeFunction wasn't properly marking it's
    inliners which probably never led to any real problems, but with
    the change to track the Inliners properly, this did expose this
    lingering issue.
    
    Also removed some redundant checks since we just asserted in the
    line above that inliners_ isn't null.
Something went wrong with that request. Please try again.