1. 04 Mar, 2015 2 commits
  2. 15 Sep, 2014 4 commits
  3. 26 Jul, 2014 5 commits
  4. 25 Jul, 2014 1 commit
  5. 23 Jul, 2014 1 commit
  6. 22 Jul, 2014 1 commit
  7. 21 Jul, 2014 2 commits
  8. 19 Jul, 2014 2 commits
  9. 16 Jul, 2014 2 commits
    • Adhemerval Zanella's avatar
      libdfp: Enable BID for 32bits · 790509ba
      Adhemerval Zanella authored
      This patch adds BID support for 32 bits. BID current relies on 128-bit
      type support, which is supplied by GCC/clang extension __uint128_t.
      Some architectures, however, does not support this extension
      (x86 32 bits, for instance).
      
      In these case the mini-gmp implementation from libgmp is used to supply
      a multiprecision math library (set to emulate a 128-bit int type).
      Currently it is included direct in libdecnumber/decimal128.c object and
      thus used in other parts of libdfp through libdecnumber.a static
      library.
      790509ba
    • Adhemerval Zanella's avatar
      fc206dcc
  10. 24 Jun, 2014 1 commit
  11. 13 Jun, 2014 6 commits
  12. 05 Jun, 2014 1 commit
    • Adhemerval Zanella's avatar
      libdfp: Always link tests with libdfp · f6e34a62
      Adhemerval Zanella authored
      This patch adds -Wl,--no-as-needed linker option when building tests to
      force linker always use libdfp, even for cases where tests do not
      actually uses libdfp symbols (test-numdigits for instance).
      
      This is required because some tests will try to printf decimal floating
      point and it requires proper initialization done by libdfp.so constructor.
      (specially when building with fortify options).
      f6e34a62
  13. 04 Jun, 2014 12 commits