Skip to content

HTTPS clone URL

Subversion checkout URL

You can clone with HTTPS or Subversion.

Download ZIP
Commits on Jan 22, 2012
  1. @jnthn

    Don't try and calculate needed stack space - it ain't quite that simp…

    jnthn authored
    …le. Just do what all the docs do and allow a decent amount.
  2. @jnthn
  3. @moritz

    bump VERSION to 2012.01

    moritz authored
  4. @jnthn

    Helps to actually have the cstack pulled out of the cursor before we …

    jnthn authored
    …go trying to look at stuff in it.
Commits on Jan 21, 2012
  1. @jnthn

    Todo a QRegex test we passed for the wrong reasons before (it tested …

    jnthn authored
    …:, and passed not because it's implemented but because backtracking into subrules never worked.)
  2. @jnthn

    Fix !cursor_next; this unbusts all of the remaining regressed QRegex …

    jnthn authored
    …tests apart from 545, which we only passed before for the wrong reason.
  3. @jnthn
  4. @jnthn

    Fix regression of tests 23,24,25,44,45 that were broken in backtracki…

    jnthn authored
    …ng into subrules patches, which all had the same root cause.
Commits on Jan 20, 2012
  1. @jnthn
Commits on Jan 18, 2012
  1. @jnthn
  2. @jnthn
Commits on Jan 17, 2012
  1. @jnthn

    Oops, I meant this...

    jnthn authored
  2. @jnthn

    Prevent a crash when building CAPHASH if one of the cstack entries is…

    jnthn authored
    … there solely for backtracking into rather than serving as a capture.
  3. @jnthn
Commits on Jan 16, 2012
  1. @moritz

    nqp::expmod_I

    moritz authored
Commits on Jan 15, 2012
  1. @jnthn

    Merge pull request #27 from goneri/master

    jnthn authored
    fix kFreeBSD build and probably other glibc based OS like hurd
  2. @goneri

    fix: kfreebsd build

    goneri authored
    GNU/kFreeBSD uses the GNU libc whereas Android uses
    its own libc. So OS_Linux is not reliable enough to
    identify a glibc.
    
    We use __GLIBC__ to know if we are on a glibc instead of
    OS_Linux.
    
    __GLIBC__ is defined by features.hi which is pulled
    by the glibc headers. That's the reason why we can't
    check __GLIBC__ at the begining of the file.
Commits on Jan 13, 2012
  1. @jnthn
  2. @jnthn

    Add missing write barrier.

    jnthn authored
  3. @jnthn
  4. @moritz
  5. @jnthn
  6. @jnthn
  7. @moritz

    adapt tests to new bigint usage

    moritz authored
  8. @moritz

    fix signatures in NQP.pir

    moritz authored
  9. @moritz

    give most nqp_bigint_* an additional parameter for the type of the re…

    moritz authored
    …turn value
    
    Also regularize the signature of from_str to have the type object come last,
    as all other ops do.
Commits on Jan 12, 2012
  1. @jnthn

    REPRS that flatten other REPRs into themselves should delegate to the…

    jnthn authored
    …ir copy_to. Fixes Int cloning issue found by llamarider++.
Commits on Jan 9, 2012
  1. @bacek
Commits on Jan 8, 2012
  1. @moritz
Commits on Jan 6, 2012
  1. @jnthn
  2. @jnthn

    Include timestamp again for NQP SC identifier; we don't face this iss…

    jnthn authored
    …ue in Rakudo since we aren't trying to bootstrap it, so no changes there (we primarily wanted the original change in Rakudo rather than NQP anyway).
  3. @jnthn
  4. @jnthn

    Toss outdated comment.

    jnthn authored
Commits on Jan 5, 2012
  1. @jnthn

    Accessing attributes may trigger auto-vivification of a scalar contai…

    jnthn authored
    …ner with P6opaque; we were missing some write-barriers for these cases. Add them.
  2. @KrisShannon
Something went wrong with that request. Please try again.