Paul Davis 18f465393b remove use of a boost::fast_pool_allocator without a mutex for Evoral::ControlList. this was not thread-safe and evidence from googling around suggests that it is actually not faster for this sort of use than std::alloc. removing this fixes bizarre errors caused by multithreaded use of the singleton pool that backs the allocator, which started to become more clear during experiments on 128 track sessions ... git-svn-id: svn://localhost/ardour2/branches/3.0@13672 d708f5d6-7413-0410-9779-e7cbd77b26cf		2012-12-17 18:41:58 +00:00
..
evoral
src
test
run-tests.sh
wscript