Problematic error checking in lmtpd.c

[Rob Siemborski]

On Wed, 1 Oct 2003, Michael Bacon wrote:

> The following is a diff against 2.0.17, just because it was what I had
> handy, but the code appears to be roughly the same in the current CVS.
> This seems to fix it, but there may be better ways to go about it that I
> haven't considered.

Yeah, this is a bug.

I'm going to change the fix though -- instead I'm going to have
open_sendmail set sm to NULL if sm_pid < 0.  That way there's still only
one error condition for the caller to check... Though probably it should
always return -1 in case of failure and we should check that instead.

Patch will be in CVS shortly.

-Rob

Index: lmtpd.c
===================================================================
RCS file: /afs/andrew.cmu.edu/system/cvs/src/cyrus/imap/lmtpd.c,v
retrieving revision 1.118
diff -u -r1.118 lmtpd.c
--- lmtpd.c     13 Aug 2003 18:39:38 -0000      1.118
+++ lmtpd.c     1 Oct 2003 18:15:53 -0000
@@ -384,7 +384,14 @@
        printf("451 lmtpd: didn't exec?!?\r\n");
        fatal("couldn't exec", EC_OSERR);
     }
-    /* i'm the parent */
+
+    if(p < 0) {
+       /* failure */
+       *sm = NULL;
+       return p;
+    }
+
+    /* parent */
     close(fds[0]);
     ret = fdopen(fds[1], "w");
     *sm = ret;

-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
Rob Siemborski * Andrew Systems Group * Cyert Hall 207 * 412-268-7456
Research Systems Programmer * /usr/contributed Gatekeeper