netdev - [PATCH V2] sch_red: fix red_calc_qavg_from_idle

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <1322691053.2602.24.camel@edumazet-laptop>
Date:	Wed, 30 Nov 2011 23:10:53 +0100
From:	Eric Dumazet <eric.dumazet@...il.com>
To:	David Miller <davem@...emloft.net>
Cc:	netdev <netdev@...r.kernel.org>,
	Stephen Hemminger <shemminger@...tta.com>,
	David Täht <dave.taht@...il.com>
Subject: [PATCH V2] sch_red: fix red_calc_qavg_from_idle_time

Since commit a4a710c4a7490587 (pkt_sched: Change PSCHED_SHIFT from 10 to
6) it seems RED/GRED are broken.

red_calc_qavg_from_idle_time() computes a delay in us units, but this
delay is now 16 times bigger than real delay, so the final qavg result
smaller than expected.

Use standard kernel time services since there is no need to obfuscate
them.

Signed-off-by: Eric Dumazet <eric.dumazet@...il.com>
---
 include/net/red.h |   15 ++++++---------
 1 file changed, 6 insertions(+), 9 deletions(-)

diff --git a/include/net/red.h b/include/net/red.h
index 3319f16..b72a3b8 100644
--- a/include/net/red.h
+++ b/include/net/red.h
@@ -116,7 +116,7 @@ struct red_parms {
 	u32		qR;		/* Cached random number */
 
 	unsigned long	qavg;		/* Average queue length: A scaled */
-	psched_time_t	qidlestart;	/* Start of current idle period */
+	ktime_t		qidlestart;	/* Start of current idle period */
 };
 
 static inline u32 red_rmask(u8 Plog)
@@ -148,17 +148,17 @@ static inline void red_set_parms(struct red_parms *p,
 
 static inline int red_is_idling(struct red_parms *p)
 {
-	return p->qidlestart != PSCHED_PASTPERFECT;
+	return p->qidlestart.tv64 != 0;
 }
 
 static inline void red_start_of_idle_period(struct red_parms *p)
 {
-	p->qidlestart = psched_get_time();
+	p->qidlestart = ktime_get();
 }
 
 static inline void red_end_of_idle_period(struct red_parms *p)
 {
-	p->qidlestart = PSCHED_PASTPERFECT;
+	p->qidlestart.tv64 = 0;
 }
 
 static inline void red_restart(struct red_parms *p)
@@ -170,13 +170,10 @@ static inline void red_restart(struct red_parms *p)
 
 static inline unsigned long red_calc_qavg_from_idle_time(struct red_parms *p)
 {
-	psched_time_t now;
-	long us_idle;
+	s64 delta = ktime_us_delta(ktime_get(), p->qidlestart);
+	long us_idle = min_t(s64, delta, p->Scell_max);
 	int  shift;
 
-	now = psched_get_time();
-	us_idle = psched_tdiff_bounded(now, p->qidlestart, p->Scell_max);
-
 	/*
 	 * The problem: ideally, average length queue recalcultion should
 	 * be done over constant clock intervals. This is too expensive, so


--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html