This appears to be an unavoidable problem with GCC's tuple
implementation. For example:
http://stackoverflow.com/q/23374953/2434
Appears to be okay in later versions of GCC though.
Visual C++ is warning that memory can't be tracked for allocators whose
pointer types aren't actually pointers, which is a correct warning but
not relevant for our concerns, and is caused by the unit tests, not the
container implementation.
The hash and key equality functions were assigned before allocating new
buckets. If that allocation failed, then the existing elements would be
left in place - so if accessed after the exception they could be in the
wrong buckets or equivalent elements could be incorrectly grouped
together.
For some reason the unordered test results are not showing up any more.
This is odd, as when I try locally they do run. I don't know why this
is, but it's possible that it's because there isn't an actual project in
the test directory, so try adding one.
- More template use, less preprocessor use.
- Removed some of the Visual C++ 6 workarounds.
- Reduced memory use of the main object.
- Split into smaller headers.
[SVN r55878]
https://svn.boost.org/svn/boost/branches/unordered/trunk
........
r43840 | danieljames | 2008-03-24 17:25:07 +0000 (Mon, 24 Mar 2008) | 1 line
Fix a g++ warning.
........
r43844 | danieljames | 2008-03-24 17:56:28 +0000 (Mon, 24 Mar 2008) | 1 line
It's a new-ish year.
........
r43885 | danieljames | 2008-03-27 20:36:10 +0000 (Thu, 27 Mar 2008) | 1 line
The release script doesn't need to copy images and css - because that's now done in the jamfiles. Also tweak the shell script a tad bit.
........
r43890 | danieljames | 2008-03-27 23:01:40 +0000 (Thu, 27 Mar 2008) | 1 line
Starting to add a docbook bibliography.
........
r43894 | danieljames | 2008-03-27 23:24:18 +0000 (Thu, 27 Mar 2008) | 1 line
Redeclare 'data' in iterator_base to help compilers which have trouble with accessing the nested typedef.
........
[SVN r43895]
Copyright update.
Switch back to the version where the sentinel points to itself.
Remove alternative versions of swap.
Workaround a borland bug or two.
More consistent use of class/swap/template.
Avoid a few warnings.
Add a no-throw swap to the allocator for exception testing.
[SVN r3793]
Daniel James