summaryrefslogtreecommitdiffstats
path: root/src
diff options
context:
space:
mode:
authorTheodore Tso <tytso@mit.edu>1998-12-01 20:41:18 +0000
committerTheodore Tso <tytso@mit.edu>1998-12-01 20:41:18 +0000
commit63ba75c5ce159af40b2548aa1e9f48926792998d (patch)
treed6e1e30d3e6da781fd9df50481410a401fa8cf72 /src
parente89537ec6a010b9f8a018be0a8086621a57c99f1 (diff)
Update the Changelog with a better documentation of the whole SUBDIRS,
LOCAL_SUBDIRS, and MY_SUBDIRS mess. git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@11049 dc483132-0cff-0310-8789-dd5450dbe970
Diffstat (limited to 'src')
-rw-r--r--src/config/ChangeLog23
1 files changed, 22 insertions, 1 deletions
diff --git a/src/config/ChangeLog b/src/config/ChangeLog
index 3101ba08b..0fd5fa6d1 100644
--- a/src/config/ChangeLog
+++ b/src/config/ChangeLog
@@ -5,7 +5,28 @@
over. (This is needed so we can fold a directory's
configure.in scripts into the parent's configure.in
without having to move all of its siblings as well into
- the parent directory's configure.in.)
+ the parent directory's configure.in. This is because if
+ the parent's configure.in file as a AC_CONFIG_DIRS macro,
+ it gets included into all the $(SUBDIRS) macro of *all*
+ of its generated makefiles, and this is wrong for all
+ but the Makefile.in of the parent directory. For example,
+ if the configure.in in src/appl contains
+ AC_CONFIG_DIRS(bsd telnet gssftp), and that configure
+ script also generates the Makefile.in for src/appl/sample,
+ then SUBDIRS will contain "bsd telnet gssftp" plus
+ whatever LOCAL_SUBDIRS is defined to contain. This
+ doesn't work. So instead of defining LOCAL_SUBDIRS,
+ the Makefile.in in src/appl/sample defines MY_SUBDIRS,
+ which completely overrides the SUBDIRS macro.
+ Eventually, we should replace all of the LOCAL_SUBDIRS
+ definitions in the Makefile.in's with MY_SUBDIRS, and
+ add MY_SUBDIRS definitions into all Makefile.in's that
+ don't currently have then, and depend on
+ AC_CONFIG_SUBDIRS to insert the right magic value into
+ SUBDIRS. At that point, we can get rid of the SUBDIR
+ definition in pre.in, and remove this test from
+ post.in which tests of MY_SUBDIRS and uses it in
+ preference to SUBDIRS, and simply always use MY_SUBDIRS.)
1998-05-27 Theodore Ts'o <tytso@rsts-11.mit.edu>