tom 2009/10/16 09:40:53 BST
Modified files:
exim-src/src dbstuff.h
Log:
Bugzilla #871
Revision Changes Path
1.8 +12 -1 exim/exim-src/src/dbstuff.h
Index: dbstuff.h
===================================================================
RCS file: /home/cvs/exim/exim-src/src/dbstuff.h,v
retrieving revision 1.7
retrieving revision 1.8
diff -u -r1.7 -r1.8
--- dbstuff.h 29 Aug 2007 14:02:22 -0000 1.7
+++ dbstuff.h 16 Oct 2009 08:40:53 -0000 1.8
@@ -1,4 +1,4 @@
-/* $Cambridge: exim/exim-src/src/dbstuff.h,v 1.7 2007/08/29 14:02:22 ph10 Exp $ */
+/* $Cambridge: exim/exim-src/src/dbstuff.h,v 1.8 2009/10/16 08:40:53 tom Exp $ */
/*************************************************
* Exim - an Internet mail transport agent *
@@ -300,6 +300,17 @@
/* Some text for messages */
#define EXIM_DBTYPE "db (v1)"
+/* When scanning, for the non-first case we historically just passed 0
+as the flags field and it worked. On FreeBSD 8 it no longer works and
+instead leads to memory exhaustion. The man-page on FreeBSD says to use
+R_NEXT, but this 1.x is a historical fallback and I've no idea how portable
+the use of that flag is; so the solution is to define R_NEXT here if it's not
+already defined, with a default value of 0 because that's what we've always
+before been able to pass successfully. */
+#ifndef R_NEXT
+#define R_NEXT 0
+#endif
+
/* Access functions */
/* EXIM_DBOPEN - sets *dbpp to point to an EXIM_DB, NULL if failed */
@@ -331,7 +342,7 @@
/* EXIM_DBSCAN - returns TRUE if data is returned, FALSE at end */
#define EXIM_DBSCAN(db, key, data, first, cursor) \
- ((db)->seq(db, &key, &data, (first? R_FIRST : 0)) == 0)
+ ((db)->seq(db, &key, &data, (first? R_FIRST : R_NEXT)) == 0)
/* EXIM_DBDELETE_CURSOR - terminate scanning operation (null). Make it
refer to cursor, to keep picky compilers happy. */