pullup from trunk

also pull up ChangeLog:1.99-1.101

ticket: 2861
version_fixed: 1.4

git-svn-id: svn://anonsvn.mit.edu/krb5/branches/krb5-1-4@16994 dc483132-0cff-0310-8789-dd5450dbe970

diff --git a/doc/ChangeLog b/doc/ChangeLog
index 46f45851943b53c4792c809f091d1c9093ff2da0..704360603faca38079f828454cd62f42855f2189 100644 (file)
--- a/doc/ChangeLog
+++ b/doc/ChangeLog
@@ -1,3 +1,13 @@
+2005-01-03  Ken Raeburn  <raeburn@mit.edu>
+
+       * build.texinfo (The util Directory): Add the new support
+       library.
+       (Installing the Binaries): Discuss parallel builds under GNU
+       make.
+
+       * implementor.texinfo (Host Address Lookup): Document Mac OS X
+       issues.
+
 2004-11-19  Tom Yu  <tlyu@mit.edu>
 
        * build.texinfo (Solaris 9): Document Solaris patches for pty

diff --git a/doc/build.texinfo b/doc/build.texinfo
index 87833c80f5558a9066733ad57a49a7eb7105a93f..64a048fb76bd7bbeeb2babd29e423bcc3c797e84 100644 (file)
--- a/doc/build.texinfo
+++ b/doc/build.texinfo
@@ -212,7 +212,8 @@ functions which parse the Kerberos configuration files (@code{krb5.conf}
 and @code{kdc.conf}).  Also in this directory are the Kerberos error table
 library and utilities (@i{et}), the Sub-system library and utilities
 (@i{ss}), database utilities (@i{db2}), pseudo-terminal utilities
-(@i{pty}), and bug-reporting program @code{send-pr}.
+(@i{pty}), bug-reporting program @code{send-pr}, and a generic
+support library @code{support} used by several of our other libraries.
 
 @node Build Requirements, Unpacking the Sources, Organization of the Source Directory, Building Kerberos V5
 @section Build Requirements
@@ -233,6 +234,8 @@ will need a @samp{make} program which supports @samp{VPATH}, or
 you will need to use a tool such as @samp{lndir} to produce a symbolic
 link tree for your build tree.
 
+@c Library support...
+
 @node Unpacking the Sources, Doing the Build, Build Requirements, Building Kerberos V5
 @section Unpacking the Sources
 
@@ -367,6 +370,11 @@ libraries into @code{DESTDIR/PREFIX/lib}, etc.
 Note that if you want to test the build (see @ref{Testing the Build}),
 you usually do not need to do a @code{make install} first.
 
+Some implementations of @samp{make} allow multiple commands to be run in
+parallel, for faster builds.  We test our Makefiles in parallel builds with
+GNU @same{make} only; they may not be compatible with other parallel build
+implementations.
+
 @node Testing the Build, Options to Configure, Installing the Binaries, Building Kerberos V5
 @section Testing the Build
 

diff --git a/doc/implementor.texinfo b/doc/implementor.texinfo
index d9e00a66bdf4945ff207e85a0f00edd168dc59a8..6609676205a292f4595f58f9c0d4274ac0fdce4a 100644 (file)
--- a/doc/implementor.texinfo
+++ b/doc/implementor.texinfo
@@ -395,6 +395,25 @@ is, both of which are wrong.
 No known bugs here, but as of IRIX 6.5.7, the version we're using at
 MIT, these functions had not been implemented.
 
+@item Mac OS X
+Two problems have been found with @code{getaddrinfo} on Mac OS X, at
+least under version 10.3.  First, while @code{gethostbyname} data is
+cached to make multiple lookups of the same name (@i{e.g.}, by
+different parts of the code that need to know about the same server
+host), @code{getaddrinfo} results are not cached, so multiple queries
+mean multiple DNS requests, which means more delays if the DNS servers
+are not close by and fast to respond.  We've implemented a cache of
+our own to work around this, though it only applies to multiple
+lookups in a short period of time within the same application process,
+and it's only implemented for the Mac at the moment.
+
+Second, the Mac libraries will generate a DNS SRV RR query; as far as
+I [Ken] can tell this is a bug, but Apple seems to consider it a
+feature.  (Call @code{getaddrinfo("example.com", "telnet", ...)} and
+you get a SRV record query, but the spec on SRV records says you must
+not use them unless the specification for the service in question says
+to.)  Yet more network traffic for each name to look up.
+
 @item NetBSD
 As of NetBSD 1.5, this function is not thread-safe.  In 1.5X
 (intermediate code snapshot between 1.5 and 1.6 releases), the