Re: [ros-svn] [hbirr] 16758: Lock the handle table if we trying to get a pointer from a handle.

[Thomas Weidenmueller]

[e-mail] wrote:
> Lock the handle table if we trying to get a pointer from a handle.

These changes are incorrect. Please revert them.

Actually, a handle lookup requires _no_ locks at all. If you don't
believe me, read up Windows Internals (4th Edition). The algorithms to
allocate subtables are implemented in such a way that it is impossible
to page fault on a lookup. However, it is not allowed to do a lookup in
a table unless attached to the process that owns it (except it is a
global handle table).

Best Regards,
Thomas