- 15 Apr, 2010 1 commit
-
-
Jeremy C. Reed authored
(so is included in tarball). Also sorted a couple entries. This was not reviewed. But is minor. git-svn-id: svn://bind10.isc.org/svn/bind10/trunk@1714 e5f2f494-b856-4b98-b285-d166d9295462
-
- 14 Apr, 2010 2 commits
-
-
JINMEI Tatuya authored
applied the same fix as r1709 and r1710. This will make it possible to run tests in a separate build tree. (trac ticket #143) also applied the DRY principle to unittest_util*: remove the local copies and simply referred to the dns/tests versions of these files. git-svn-id: svn://bind10.isc.org/svn/bind10/trunk@1713 e5f2f494-b856-4b98-b285-d166d9295462
-
Jeremy C. Reed authored
I also sorted them so later it will be easier to recognize changes. (Sanity checked by jinmei.) git-svn-id: svn://bind10.isc.org/svn/bind10/trunk@1712 e5f2f494-b856-4b98-b285-d166d9295462
-
- 13 Apr, 2010 4 commits
-
-
JINMEI Tatuya authored
made test data paths dynamically configurable with the default path of @srcdir@/testdata. This will allow testdata to be referred to from a non-src build directory. We can also specify mulitiple directories, which may happen if we generate some data files from spec at build time and run tests under a non-src directory. git-svn-id: svn://bind10.isc.org/svn/bind10/trunk@1709 e5f2f494-b856-4b98-b285-d166d9295462
-
JINMEI Tatuya authored
git-svn-id: svn://bind10.isc.org/svn/bind10/trunk@1708 e5f2f494-b856-4b98-b285-d166d9295462
-
JINMEI Tatuya authored
cleanup: construct RRClasses from integer rather than strings when how to construct them doesn't matter. git-svn-id: svn://bind10.isc.org/svn/bind10/trunk@1707 e5f2f494-b856-4b98-b285-d166d9295462
-
JINMEI Tatuya authored
git-svn-id: svn://bind10.isc.org/svn/bind10/trunk@1706 e5f2f494-b856-4b98-b285-d166d9295462
-
- 12 Apr, 2010 1 commit
-
-
JINMEI Tatuya authored
cleanup: use relative path to header files from "src/lib", i.e., s/"name.h"/<dns/name.h", etc. also removed unnecessary header files to be included. git-svn-id: svn://bind10.isc.org/svn/bind10/trunk@1705 e5f2f494-b856-4b98-b285-d166d9295462
-
- 10 Apr, 2010 1 commit
-
-
JINMEI Tatuya authored
test cases and documentation were also fully provided. (this changeset includes a minor unrelated cleanup: move the pimpl structure inside the MessageRenderer class declaration to clarify the intent) git-svn-id: svn://bind10.isc.org/svn/bind10/trunk@1704 e5f2f494-b856-4b98-b285-d166d9295462
-
- 09 Apr, 2010 3 commits
-
-
JINMEI Tatuya authored
git-svn-id: svn://bind10.isc.org/svn/bind10/trunk@1701 e5f2f494-b856-4b98-b285-d166d9295462
-
JINMEI Tatuya authored
git-svn-id: svn://bind10.isc.org/svn/bind10/trunk@1700 e5f2f494-b856-4b98-b285-d166d9295462
-
JINMEI Tatuya authored
- constify as much as possible git-svn-id: svn://bind10.isc.org/svn/bind10/trunk@1699 e5f2f494-b856-4b98-b285-d166d9295462
-
- 08 Apr, 2010 2 commits
-
-
JINMEI Tatuya authored
git-svn-id: svn://bind10.isc.org/svn/bind10/trunk@1695 e5f2f494-b856-4b98-b285-d166d9295462
-
JINMEI Tatuya authored
- pass xfr request data length in network byte order, and receive it as such. receiving this as native data (unpack('H')) would be a bit naive because it may assume some padding. - avoid cast as a bonus side effect of the first fix - make cmsg space handling more portable. use CMSG_SPACE when possible; otherwise calculate the space using dummy data - overall simplied fd_share code: trying to handle multiple FDs is not correct in terms of API, and in any case we don't need that; verify cmsg_level and cmsg_type on reception; avoid unnecessary data initialization; stop naively assume the exact memory allocation of cmsg_data (assuming no padding) - use cmsg_space() so that padding, if any, will be taken intou account - catch the error case of recv_fd failure in XfroutSession.handle() explicitly, and raise an exception git-svn-id: svn://bind10.isc.org/svn/bind10/trunk@1694 e5f2f494-b856-4b98-b285-d166d9295462
-
- 07 Apr, 2010 5 commits
-
-
JINMEI Tatuya authored
git-svn-id: svn://bind10.isc.org/svn/bind10/trunk@1692 e5f2f494-b856-4b98-b285-d166d9295462
-
JINMEI Tatuya authored
- style fixes according to bind10 coding guideline - constify things - propset git-svn-id: svn://bind10.isc.org/svn/bind10/trunk@1691 e5f2f494-b856-4b98-b285-d166d9295462
-
JINMEI Tatuya authored
git-svn-id: svn://bind10.isc.org/svn/bind10/trunk@1687 e5f2f494-b856-4b98-b285-d166d9295462
-
JINMEI Tatuya authored
git-svn-id: svn://bind10.isc.org/svn/bind10/trunk@1686 e5f2f494-b856-4b98-b285-d166d9295462
-
JINMEI Tatuya authored
cleanup: eliminated the need for isc.auth.sqlite3_ds.AXFRInDB by directly calling sqlite3_ds.load() from the xfrin module. Other cleanups: - cosmetic: removed redundant blank lines and white spaces after EOL - grammar fix in comments - catch Sqlite3DSError explicitly (but I suspect the exception handling in the xfrin module is naive overall, which should be fixed) git-svn-id: svn://bind10.isc.org/svn/bind10/trunk@1684 e5f2f494-b856-4b98-b285-d166d9295462
-
- 06 Apr, 2010 3 commits
-
-
JINMEI Tatuya authored
git-svn-id: svn://bind10.isc.org/svn/bind10/trunk@1681 e5f2f494-b856-4b98-b285-d166d9295462
-
JINMEI Tatuya authored
git-svn-id: svn://bind10.isc.org/svn/bind10/trunk@1680 e5f2f494-b856-4b98-b285-d166d9295462
-
JINMEI Tatuya authored
git-svn-id: svn://bind10.isc.org/svn/bind10/trunk@1678 e5f2f494-b856-4b98-b285-d166d9295462
-
- 05 Apr, 2010 1 commit
-
-
JINMEI Tatuya authored
git-svn-id: svn://bind10.isc.org/svn/bind10/trunk@1673 e5f2f494-b856-4b98-b285-d166d9295462
-
- 01 Apr, 2010 2 commits
-
-
Jelte Jansen authored
check map tag size in create functions too, for consistency, i think we should indeed remove the direct map access and constructor, see comment in ticket 129 git-svn-id: svn://bind10.isc.org/svn/bind10/trunk@1668 e5f2f494-b856-4b98-b285-d166d9295462
-
JINMEI Tatuya authored
git-svn-id: svn://bind10.isc.org/svn/bind10/trunk@1667 e5f2f494-b856-4b98-b285-d166d9295462
-
- 31 Mar, 2010 3 commits
-
-
JINMEI Tatuya authored
it cannot exceed 255 bytes, but the "from string" constructor didn't check it. added test cases for long tags including this specific case. git-svn-id: svn://bind10.isc.org/svn/bind10/trunk@1666 e5f2f494-b856-4b98-b285-d166d9295462
-
JINMEI Tatuya authored
git-svn-id: svn://bind10.isc.org/svn/bind10/trunk@1665 e5f2f494-b856-4b98-b285-d166d9295462
-
JINMEI Tatuya authored
additional cleanups: - unify decode_blob() and decode_utf() since these two are identical. moved the XXX comment to the caller side. - costified the length argument to decode_blob(); it doesn't have to be adjusted. - removed an unnecessary temporary variable (it didn't seem to improve readability) git-svn-id: svn://bind10.isc.org/svn/bind10/trunk@1664 e5f2f494-b856-4b98-b285-d166d9295462
-
- 26 Mar, 2010 7 commits
-
-
Jeremy C. Reed authored
also specifically within "dns/". git-svn-id: svn://bind10.isc.org/svn/bind10/trunk@1657 e5f2f494-b856-4b98-b285-d166d9295462
-
Jeremy C. Reed authored
I didn't notice this before because didn't configure with gtest when testing distcheck. git-svn-id: svn://bind10.isc.org/svn/bind10/trunk@1656 e5f2f494-b856-4b98-b285-d166d9295462
-
JINMEI Tatuya authored
git-svn-id: svn://bind10.isc.org/svn/bind10/trunk@1655 e5f2f494-b856-4b98-b285-d166d9295462
-
Jeremy C. Reed authored
-I$(top_srcdir)/src/lib/dns -I$(top_builddir)/src/lib/dns Because some includes aren't prefixed with "dns/" like rdata.h, name.h, and rrttl.h in the srcdir and rrtype.h on the builddir. git-svn-id: svn://bind10.isc.org/svn/bind10/trunk@1654 e5f2f494-b856-4b98-b285-d166d9295462
-
Jeremy C. Reed authored
files are generated. I noticed that these two files were placed in the original source tree and also later in the build tree -- the rrclass.h was identical but the rrtype.h files were different. For the tests look in the top_srcdir for includes also. This should fix the problem I saw on FreeBSD when it was building tests using a tarball from source and building in a different location than the source: In file included from ../../../../../src/lib/dns/tests/unittest_util.cc:29: ../../../../../src/lib/dns/tests/unittest_util.h:77: error: expected unqualified-id before '::' token git-svn-id: svn://bind10.isc.org/svn/bind10/trunk@1653 e5f2f494-b856-4b98-b285-d166d9295462
-
Jeremy C. Reed authored
git-svn-id: svn://bind10.isc.org/svn/bind10/trunk@1650 e5f2f494-b856-4b98-b285-d166d9295462
-
Jeremy C. Reed authored
This line was removed in r1617. I assume that was just an oversight. The problem was this header was not included in the tarball. git-svn-id: svn://bind10.isc.org/svn/bind10/trunk@1649 e5f2f494-b856-4b98-b285-d166d9295462
-
- 25 Mar, 2010 2 commits
-
-
Jelte Jansen authored
git-svn-id: svn://bind10.isc.org/svn/bind10/trunk@1645 e5f2f494-b856-4b98-b285-d166d9295462
-
Likun Zhang authored
git-svn-id: svn://bind10.isc.org/svn/bind10/trunk@1644 e5f2f494-b856-4b98-b285-d166d9295462
-
- 24 Mar, 2010 1 commit
-
-
JINMEI Tatuya authored
(first aid workaround with ugly ifdef's, but worked) also fixed the problem that "from source" didn't work. git-svn-id: svn://bind10.isc.org/svn/bind10/trunk@1636 e5f2f494-b856-4b98-b285-d166d9295462
-
- 23 Mar, 2010 2 commits
-
-
Likun Zhang authored
git-svn-id: svn://bind10.isc.org/svn/bind10/trunk@1633 e5f2f494-b856-4b98-b285-d166d9295462
-
Likun Zhang authored
1. When auth server get axfr query in tcp, the query message and tcp socket will be sent to xfrout process over sendmsg/recvmsg, then auth server will close the tcp socket, or get the next tcp query on the socket. Xfrout send the query response over the tcp socket it received from auth server. 2. Since python3.1 didn't provide the wrapper of sendmsg() and recvmsg() for C socket module(Seems they plan do it in python3.2), I have to write the code in c++, and make python binding for it. See the code in file /src/lib/xfr/fd_share.cc and /src/lib/xfr/python_xfr.cc. git-svn-id: svn://bind10.isc.org/svn/bind10/trunk@1630 e5f2f494-b856-4b98-b285-d166d9295462
-