[interprocess] Two potential bugs with b::int::vector of b::i::weak_ptr

[Gaëtan Gaumer <gaetan.gaumer@…>]

Here are two potential bugs discussed on boost-users mailing list when using b::int::vector of b::i::weak_ptr

• 1 - Erase of an element delete the wrong post

The original post : ​http://lists.boost.org/boost-users/2010/11/64591.php

This bug is reproduced by the sample programm pasted here : ​http://pastebin.com/yDCngGw4

It's still present in v1.50

• 2 - After a pushback sometimes the last element is not the one push_backed.

The original post : ​http://lists.boost.org/boost-users/2012/07/75311.php

The sample program trying to reproduce the bug : ​http://pastebin.com/rdpZA7CU

Note that this program does not reproduce the problem, as when it runs, only the priv_range_insert_new_allocation method is called. I don't know how to force the program to use priv_range_insert_expand_backwards to reproduce the problem.

A sample of a trace when the bug occurs :

In priv_range_insert_expand_backwards possibility 4/8  and  NOT do_after
After pushback in cableToUpdate vector last elem Id=50332165 added cable Id =134218245
!--> sizeBefore=8, capacityBefore=8, sizeAfter=9, capacityAfter=13
CableToUpdate[16777216260, 16777216260, 16777216260, 50332165, 67109381, 16777216260, 16777216260, 16777216260, 50332165, ]

The 4 firsts elements are wrongly duplicated at index 5-8. This vector corruption is perhaps done when the vector is resized from 5 to 8.

[Ion Gaztañaga]

(In [82810]) Fixes #7164 (first bug only)

[Ion Gaztañaga]

Thanks for the report. The first bug it's caused by a missing assignment in weak_count. I'll reopen the bug until you can confirm current trunk code does not suffer from bug #2

[Ion Gaztañaga]

No bug report received so closing the bug as fixed.