feedc0de/boost_container_hash
1
0
Fork 0
forked from boostorg/container_hash
Code Pull Requests Activity

Merge branch 'develop'

Browse Source
This commit is contained in:
Daniel James
2017-10-27 11:06:13 +01:00
parent 7a6c0f20d5 d803b3bdc0
commit 7f3a063064
2 changed files with 91 additions and 1 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
include/boost/functional/hash/detail/hash_float.hpp
View File

@@ -241,7 +241,7 @@ namespace boost
template <class T> template <class T>
inline bool is_zero(T v) inline bool is_zero(T v)
{ {
#if !defined(__GNUC__) #if !defined(__GNUC__) && !defined(__clang__)
return v == 0; return v == 0;
#else #else
// GCC's '-Wfloat-equal' will complain about comparing // GCC's '-Wfloat-equal' will complain about comparing

90
meta/explicit-failures-markup.xml Normal file
View File

@@ -0,0 +1,90 @@
<?xml version="1.0" encoding="utf-8"?>
<explicit-failures-markup>
<!-- functional/hash -->
<library name="functional/hash">
<mark-expected-failures>
<test name="hash_value_array_test"/>
<toolset name="msvc-6.5*"/>
<toolset name="msvc-7.0*"/>
<note author="Daniel James">
hash_value is not overloaded for arrays for older versions
of Visual C++. There is a work around so that
boost::hash&lt;T[N]&gt;, boost::hash_combine and boost::hash_range
work.
</note>
</mark-expected-failures>
<mark-expected-failures>
<test name="hash_function_pointer_test"/>
<toolset name="msvc-6.5*"/>
<toolset name="msvc-7.0*"/>
<note refid="2" author="Daniel James"/>
</mark-expected-failures>
<mark-expected-failures>
<test name="hash_function_pointer_test"/>
<toolset name="sun-5.7"/>
<toolset name="sun-5.8"/>
<toolset name="sun-5.9"/>
<note author="Daniel James">
On these compilers the wrong overload of hash_value is called
when the argument is a hash function pointer. So calling
hash_value doesn't work but boost::hash does work (and it's
recommended that user never call hash_value directly so this
shouldn't be a problem).
</note>
</mark-expected-failures>
<mark-expected-failures>
<test name="hash_long_double_test"/>
<toolset name="gcc-3.4.3_sunos"/>
<toolset name="*pa_risc"/>
<note author="Daniel James">
This platform has poor support for <code>long double</code> so
the hash function perform poorly for values out of the range
of <code>double</code> or if they differ at a greater precision
that <code>double</code> is capable of representing.
</note>
</mark-expected-failures>
<mark-expected-failures>
<test name="point" />
<test name="books" />
<toolset name="msvc-6.5*"/>
<toolset name="msvc-7.0*"/>
<note author="Daniel James">
These examples only work on compilers with support for ADL.
It is possible to work around this, but I wanted to keep the
example code as clean as possible.
</note>
</mark-expected-failures>
<mark-expected-failures>
<test name="point" />
<toolset name="borland-*"/>
<note author="Daniel James">
It appears that Borland doesn't find friend functions defined
in a class by ADL. This is easily fixed but this example is
meant to show the typical way of customising boost::hash, not
the portable way.
</note>
</mark-expected-failures>
<mark-expected-failures>
<test name="hash_global_namespace_test" />
<toolset name="borland-*"/>
<note author="Daniel James">
The test demonstrates a Borland bug - functions that aren't
in a namespace don't appear to be found by ADL.
</note>
</mark-expected-failures>
<mark-expected-failures>
<test name="container_fwd_gcc_debug"/>
<toolset name="darwin-4.2"/>
<note author="Daniel James">
Debug containers aren't supported on Apple's version of gcc 4.2.
</note>
</mark-expected-failures>
</library>
</explicit-failures-markup>