[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: Bug#301370: marked as done (quota postrm fails due to missing dependency)



Steve,

> There appear to be a number of undocumented changes, relative to -4, which
> are not documentation fixes. :/  Would you remind removing them and
> uploading a -4sarge1 to testing-proposed-updates, or else explain why these
> changes are warranted during the freeze?

Please take my apologies. I was interrupted while working on the package
and simply forgot to remove some parts of the whole patch I applied.
Also I didn't complete the changelog entry. And you probably noticed
that stray patch file lying around in the source tree. Argh!

I fixed this and am currently uploading -6 which should be okay. It
still contains two non-doc patches:

diff -ruN tmp/quota-3.12/quotaops.c quota-3.12/quotaops.c
--- tmp/quota-3.12/quotaops.c	2005-05-11 11:58:35.000000000 +0200
+++ quota-3.12/quotaops.c	2005-05-11 11:58:16.000000000 +0200
@@ -34,7 +34,7 @@
 
 #ident "$Copyright: (c) 1980, 1990 Regents of the University of California. $"
 #ident "$Copyright: All rights reserved. $"
-#ident "$Id: quotaops.c,v 1.14 2004/05/24 19:39:15 jkar8572 Exp $"
+#ident "$Id: quotaops.c,v 1.16 2005/03/31 11:48:02 jkar8572 Exp $"
 
 #include <rpc/rpc.h>
 #include <sys/types.h>
@@ -72,13 +72,13 @@
 	time_t now;
 
 	time(&now);
-	if (q->dq_dqb.dqb_bsoftlimit && toqb(q->dq_dqb.dqb_curspace) >= q->dq_dqb.dqb_bsoftlimit) {
+	if (q->dq_dqb.dqb_bsoftlimit && toqb(q->dq_dqb.dqb_curspace) > q->dq_dqb.dqb_bsoftlimit) {
 		if (!q->dq_dqb.dqb_btime)
 			q->dq_dqb.dqb_btime = now + q->dq_h->qh_info.dqi_bgrace;
 	}
 	else
 		q->dq_dqb.dqb_btime = 0;
-	if (q->dq_dqb.dqb_isoftlimit && q->dq_dqb.dqb_curinodes >= q->dq_dqb.dqb_isoftlimit) {
+	if (q->dq_dqb.dqb_isoftlimit && q->dq_dqb.dqb_curinodes > q->dq_dqb.dqb_isoftlimit) {
 		if (!q->dq_dqb.dqb_itime)
 			q->dq_dqb.dqb_itime = now + q->dq_h->qh_info.dqi_igrace;
 	}
[...]

This one fixes the bug that grace time was set when reaching the limit
instead of starting it when exceeding the limit.

diff -ruN tmp/quota-3.12/rquota_client.c quota-3.12/rquota_client.c
--- tmp/quota-3.12/rquota_client.c	2004-02-12 10:45:14.000000000 +0100
+++ quota-3.12/rquota_client.c	2005-05-11 11:58:16.000000000 +0200
@@ -149,8 +149,10 @@
 	 * Strip off pathname on nfs mounted dir. Ignore entries of any
 	 * automounter.
 	 */
-	if ((pathname = strchr(fsname_tmp, ':')) == (char *)0 || *(pathname + 1) == '(')
+	if ((pathname = strchr(fsname_tmp, ':')) == (char *)0 || *(pathname + 1) == '(') {
+		free(fsname_tmp);
 		return -ENOENT;
+	}
 
 	*pathname++ = '\0';

This one obviously fixes a memory leak. It's not that serious but I
rather have it fixed for sarge too.

> Incidentally, I'm puzzled why quota is calling update-inetd at all from the
> postrm, since there's no code in the postinst which adds an inetd entry. 
> That shouldn't be a blocker for sarge, though.

You're right. That's why I removed the line completely.

Michael
-- 
Michael Meskes
Email: Michael at Fam-Meskes dot De, Michael at Meskes dot (De|Com|Net|Org)
ICQ: 179140304, AIM/Yahoo: michaelmeskes, Jabber: meskes@jabber.org
Go SF 49ers! Go Rhein Fire! Use Debian GNU/Linux! Use PostgreSQL!

Attachment: signature.asc
Description: Digital signature


Reply to: