hi philip,
> I'm glad it's Not My Problem (tm)
heh ... i think i may, then, owe you some royalties ;-)
altho, i think it still may be ... at least a bit. pls read on.
> I suppose the next logical steps are
...
> and (2) to see if strnlen exists on your box ("man strnlen").
well 1st the easy one:
% man strnlen
No manual entry for strnlen
> (1) to find a source for libsrs and scan it for strnlen,
per libsrs_alt-1.0rc1's 'ChangeLog':
v1.0 released - 31 Mar 2005
---------------------------
30 Mar 2005
...
- Exim 4.50 patch added (will be submitted for inclusion)
...
and, per libsrs_alt-1.0rc1's 'MTAs/README.EXIM':
For later versions than 4.50, please check the src/srs.c
file as the patch will be submitted for inclusion in Exim.
comparing the relevant 'MTAs/exim-4.50-srs_update-1.0.patch' with Exim's
deliver.c, it looks as if appropriate changes/inclusions HAVE been made ...
as well, i note in Exim's 'src/srs.c':
/* SRS - Sender rewriting scheme support
(C)2004 Miles Wilton <miles@???>
SRS Support Version: 1.0a
License: GPL */
BUT, that same said patch refers to:
diff -Naur exim-4.50/src/srs.c exim-4.50-patched/src/srs.c
--- exim-4.50/src/srs.c 2005-02-17 14:49:11.000000000 +0000
+++ exim-4.50-patched/src/srs.c 2005-04-01 22:33:39.000000000 +0100
@@ -1,11 +1,12 @@
-/* $Cambridge: exim/exim-src/src/srs.c,v 1.4 2005/02/17 11:58:26 ph10 Exp $ */
-
/*************************************************
* Exim - an Internet mail transport agent *
*************************************************/
/* SRS - Sender rewriting scheme support
- )2004 Miles Wilton <miles@???>
+ (C)2004 Miles Wilton <miles@???>
+
+ SRS Support Version: 1.0
+
License: GPL */
#include "exim.h"
@@ -24,72 +25,88 @@
...
-
if((srs = srs_open(secret, strnlen(secret, SRS_MAX_SECRET_LENGTH),
- maxage, hashlen, hashlen)) == NULL)
+ maxage, hashlen, srs_hashmin)) == NULL)
...
- if((sbufp = string_nextinlist(&list, &co, sbuf, sizeof(sbuf))) != NULL)
- srs_set_option(srs, SRS_OPTION_USEHASH, atoi(sbuf));
+ /* Extra secrets? */
+ while((secret = string_nextinlist(&list, &co, secret_buf,
SRS_MAX_SECRET_LENGTH)) != NULL)
+ srs_add_secret(srs, secret, strnlen(secret, SRS_MAX_SECRET_LENGTH));
...
NOTE that it's BOTH referring to EXISTING, and adding NEW, instances of strnlen
to Exim's code ...
granted, NOT the defn, but i'll worry abt that later ...
as we exchanged in last post, there's no trace of 'strnlen' in Exim's code ...
which leads me 1st to ask --
*WAS* the patch included/integrated into the Exim tree? and if so, to what --
as there is that supposedly pre-existing strnlen instance ...
richard
