| Summary: | dev-libs/boost-1.49.0: handling of destructors in intrusive_base.hpp | ||
|---|---|---|---|
| Product: | Gentoo Linux | Reporter: | Markus Peloquin <markus> |
| Component: | [OLD] Core system | Assignee: | C++ Team [disbanded] <cpp+disabled> |
| Status: | RESOLVED FIXED | ||
| Severity: | normal | CC: | arfrever.fta, dev-zero, hwoarang, SebastianLuther |
| Priority: | Normal | ||
| Version: | unspecified | ||
| Hardware: | All | ||
| OS: | Linux | ||
| See Also: | https://svn.boost.org/trac/boost/ticket/6794 | ||
| Whiteboard: | |||
| Package list: | Runtime testing required: | --- | |
| Attachments: | boost-1.49.0-gcc462-intrusive_ptr.patch | ||
|
Description
Markus Peloquin
2012-04-16 00:13:00 UTC
Oh right, the upstream bug is https://svn.boost.org/trac/boost/ticket/6794 The Boost developer seems to think gcc's issue is that a subclass of intrusive_base declares it's destructor virtual, which is fine as long as you never try to delete an intrusive_base pointer or make some smart pointer like auto_ptr<intrusive_base>. this bug is only about the issue: ALWAYS make destructors virtual, unless you really don't like OOP. So it's both a Boost bug and a GCC bug. I am so proud of myself right now :) Enough excitement for me, I need to get about six hours of sleep now. the gcc crash is bug 411691 Boost 1.50.0 includes the fix. |