GitLab

if not set, we use dummy locks instead of the boost locks

instead of AbstractRRset it will now take a MessagePtr in its success()
callback. needed to change a lot of test code for it to run again

briefly talked about it on jabber, no one opposed to it, Jeremy explicitly
agreed.  changes are trivial themselves, so no explicit review.

- removed all #include <config.h> from nsas .h files (and moved them to .cc files)
- added the (hopefully temporary) workaround for boost/sunstudio (a check in configure.ac, and the redefinition of __SUNPRO_CC in the header files that include boost/thread.hpp)
- changed AC_COMPILE_IFELSE to AC_LINK_IFELSE to the compiler flags check functi
on (some flags will not fail until it gets to ld)
- changed a wrong ) in zone_entry_unittest
- changed nsas_entry_unittests so the hashkey string value would not fall out of scope
(diff checked by jreed)


git-svn-id: svn://bind10.isc.org/svn/bind10/trunk@3887 e5f2f494-b856-4b98-b285-d166d9295462

git-svn-id: svn://bind10.isc.org/svn/bind10/branches/trac397focused@3885 e5f2f494-b856-4b98-b285-d166d9295462

applying two fixed for solaris; remove 'using namespace boost' where we use uint32_t, and adds jeremy's check for the multithreading flag
(also added the latter to jinmei's ac_try_link call)
should not compile yet, we are still looking for a good solution for the -Wno-missing-initializers problem


git-svn-id: svn://bind10.isc.org/svn/bind10/trunk@3872 e5f2f494-b856-4b98-b285-d166d9295462

NameserverEntry is too low. Who would choose the nameserver to ask?

git-svn-id: svn://bind10.isc.org/svn/bind10/branches/trac408@3806 e5f2f494-b856-4b98-b285-d166d9295462

And the NameserverAddress is made safe with regard to changing indices
inside the NameserverEntry.

git-svn-id: svn://bind10.isc.org/svn/bind10/branches/trac408@3804 e5f2f494-b856-4b98-b285-d166d9295462

git-svn-id: svn://bind10.isc.org/svn/bind10/branches/trac408@3803 e5f2f494-b856-4b98-b285-d166d9295462

git-svn-id: svn://bind10.isc.org/svn/bind10/branches/trac408@3800 e5f2f494-b856-4b98-b285-d166d9295462

short -> AddressFamily
uint32_t for indices to size_t

The address_, v4_addresses_ and v6_addresses_ arrays were unified

git-svn-id: svn://bind10.isc.org/svn/bind10/branches/trac408@3794 e5f2f494-b856-4b98-b285-d166d9295462

git-svn-id: svn://bind10.isc.org/svn/bind10/branches/trac356@3760 e5f2f494-b856-4b98-b285-d166d9295462

The RTT values of previous addresses are kept and reused, if the same
addresses arrive.

git-svn-id: svn://bind10.isc.org/svn/bind10/branches/trac408@3751 e5f2f494-b856-4b98-b285-d166d9295462

git-svn-id: svn://bind10.isc.org/svn/bind10/branches/trac356@3740 e5f2f494-b856-4b98-b285-d166d9295462

git-svn-id: svn://bind10.isc.org/svn/bind10/branches/trac408@3734 e5f2f494-b856-4b98-b285-d166d9295462

git-svn-id: svn://bind10.isc.org/svn/bind10/branches/trac408@3732 e5f2f494-b856-4b98-b285-d166d9295462

git-svn-id: svn://bind10.isc.org/svn/bind10/branches/trac408@3727 e5f2f494-b856-4b98-b285-d166d9295462

git-svn-id: svn://bind10.isc.org/svn/bind10/branches/trac356@3725 e5f2f494-b856-4b98-b285-d166d9295462

git-svn-id: svn://bind10.isc.org/svn/bind10/branches/trac408@3719 e5f2f494-b856-4b98-b285-d166d9295462

git-svn-id: svn://bind10.isc.org/svn/bind10/branches/trac408@3716 e5f2f494-b856-4b98-b285-d166d9295462

With the slight possible problem of skipping to next second between
calls to time(). That needs to be fixed somehow.

git-svn-id: svn://bind10.isc.org/svn/bind10/branches/trac408@3693 e5f2f494-b856-4b98-b285-d166d9295462

git-svn-id: svn://bind10.isc.org/svn/bind10/branches/trac408@3692 e5f2f494-b856-4b98-b285-d166d9295462

* Uses RRClass instead of uint16_t
* NameserverEntry does not have constructor with rrsets, it will request
  them from resolver

It compiles but tests still do not pass

git-svn-id: svn://bind10.isc.org/svn/bind10/branches/trac408@3678 e5f2f494-b856-4b98-b285-d166d9295462

It does not return the whole message, just the RRset. This way we can
extract the data easier and it can be provided directly from cache.

git-svn-id: svn://bind10.isc.org/svn/bind10/branches/trac408@3673 e5f2f494-b856-4b98-b285-d166d9295462

git-svn-id: svn://bind10.isc.org/svn/bind10/branches/trac356@3671 e5f2f494-b856-4b98-b285-d166d9295462

git-svn-id: svn://bind10.isc.org/svn/bind10/branches/trac356@3639 e5f2f494-b856-4b98-b285-d166d9295462

Does not link, more tests needed.

git-svn-id: svn://bind10.isc.org/svn/bind10/branches/trac408@3634 e5f2f494-b856-4b98-b285-d166d9295462

It now handles many things by itself and does its own locking.

Still one test does not pass currently, as it does not handle TTL.
There's a FIXME in the code, it allows wrong behaviour, but test for
that kind of behaviour isn't written yet.

git-svn-id: svn://bind10.isc.org/svn/bind10/branches/trac408@3630 e5f2f494-b856-4b98-b285-d166d9295462

git-svn-id: svn://bind10.isc.org/svn/bind10/branches/trac408@3539 e5f2f494-b856-4b98-b285-d166d9295462

git-svn-id: svn://bind10.isc.org/svn/bind10/branches/trac408@3531 e5f2f494-b856-4b98-b285-d166d9295462

Include with " " is relative to current file - library headers need < >
and in tests .. is needed (gcc does not reject invalid relative paths
and looks into -I as well for them).

git-svn-id: svn://bind10.isc.org/svn/bind10/branches/trac408@3480 e5f2f494-b856-4b98-b285-d166d9295462

associated with the address


git-svn-id: svn://bind10.isc.org/svn/bind10/branches/trac356@3465 e5f2f494-b856-4b98-b285-d166d9295462

been created for this.
b) Move the IOAddress object into the correct namespace.


git-svn-id: svn://bind10.isc.org/svn/bind10/branches/trac356@3389 e5f2f494-b856-4b98-b285-d166d9295462

git-svn-id: svn://bind10.isc.org/svn/bind10/branches/trac356@3372 e5f2f494-b856-4b98-b285-d166d9295462

git-svn-id: svn://bind10.isc.org/svn/bind10/branches/trac356@3272 e5f2f494-b856-4b98-b285-d166d9295462