From: "Daniel Weismüller" <daniel.weismueller@ipfire.org>
To: development@lists.ipfire.org
Subject: [PATCH 02/10] QoS: Use Intermediate Functional Block
Date: Mon, 21 Oct 2019 17:48:58 +0200 [thread overview]
Message-ID: <20191021154906.24418-2-daniel.weismueller@ipfire.org> (raw)
In-Reply-To: <20191021154906.24418-1-daniel.weismueller@ipfire.org>
[-- Attachment #1: Type: text/plain, Size: 4061 bytes --]
From: Michael Tremer <michael.tremer(a)ipfire.org>
This is an alternative implementation to the Intermediate Queuing
Device (IMQ) which is an out-of-tree kernel patch and has been
criticised for being slow, especially with mutliple processors.
IFB is part of the mainline kernel and a lot less code.
Signed-off-by: Michael Tremer <michael.tremer(a)ipfire.org>
Signed-off-by: Daniel Weismüller <daniel.weismueller(a)ipfire.org>
---
config/qos/makeqosscripts.pl | 20 +++++++++++++-------
1 file changed, 13 insertions(+), 7 deletions(-)
diff --git a/config/qos/makeqosscripts.pl b/config/qos/makeqosscripts.pl
index 053b32161..b25aad93a 100644
--- a/config/qos/makeqosscripts.pl
+++ b/config/qos/makeqosscripts.pl
@@ -406,10 +406,20 @@ print <<END
### $qossettings{'IMQ_DEV'}
###
+ tc qdisc del dev $qossettings{'RED_DEV'} root
+ tc qdisc del dev $qossettings{'RED_DEV'} ingress
+ tc qdisc add dev $qossettings{'RED_DEV'} handle ffff: ingress
+
### BRING UP $qossettings{'IMQ_DEV'}
- modprobe imq numdevs=1 numqueues=\$(grep -c "^processor" /proc/cpuinfo || echo 1)
+ if [ ! -d "/sys/class/net/$qossettings{'IMQ_DEV'}" ]; then
+ ip link add name $qossettings{'IMQ_DEV'} type ifb
+ fi
+
ip link set $qossettings{'IMQ_DEV'} up
+ tc filter add dev $qossettings{'RED_DEV'} parent ffff: protocol all u32 match u32 0 0 \\
+ action mirred egress redirect dev $qossettings{'IMQ_DEV'}
+
### ADD HTB QDISC FOR $qossettings{'IMQ_DEV'}
tc qdisc del dev $qossettings{'IMQ_DEV'} root >/dev/null 2>&1
tc qdisc add dev $qossettings{'IMQ_DEV'} root handle 2: htb default $qossettings{'DEFCLASS_INC'}
@@ -510,7 +520,6 @@ print <<END
iptables -t mangle -A POSTROUTING -i $qossettings{'RED_DEV'} -p ah -j RETURN
iptables -t mangle -A POSTROUTING -i $qossettings{'RED_DEV'} -p esp -j RETURN
iptables -t mangle -A POSTROUTING -i $qossettings{'RED_DEV'} -p ip -j RETURN
- iptables -t mangle -A POSTROUTING -m mark ! --mark 0 ! -o $qossettings{'RED_DEV'} -j IMQ --todev 0
iptables -t mangle -I FORWARD -i $qossettings{'RED_DEV'} -j QOS-INC
iptables -t mangle -A FORWARD -i $qossettings{'RED_DEV'} -j QOS-TOS
@@ -527,7 +536,6 @@ print <<END
iptables -t mangle -A PREROUTING -i $qossettings{'RED_DEV'} -p ah -j RETURN
iptables -t mangle -A PREROUTING -i $qossettings{'RED_DEV'} -p esp -j RETURN
iptables -t mangle -A PREROUTING -i $qossettings{'RED_DEV'} -p ip -j RETURN
- iptables -t mangle -A PREROUTING -i $qossettings{'RED_DEV'} -j IMQ --todev 0
iptables -t mangle -I PREROUTING -i $qossettings{'RED_DEV'} -j QOS-INC
iptables -t mangle -A PREROUTING -i $qossettings{'RED_DEV'} -j QOS-TOS
@@ -688,16 +696,14 @@ print <<END
tc qdisc add root dev $qossettings{'IMQ_DEV'} fq_codel >/dev/null 2>&1
# STOP IMQ-DEVICE
ip link set $qossettings{'IMQ_DEV'} down >/dev/null 2>&1
+
+ # REMOVE & FLUSH CHAINS
iptables -t mangle --delete POSTROUTING -i $qossettings{'RED_DEV'} -p ah -j RETURN >/dev/null 2>&1
iptables -t mangle --delete POSTROUTING -i $qossettings{'RED_DEV'} -p esp -j RETURN >/dev/null 2>&1
iptables -t mangle --delete POSTROUTING -i $qossettings{'RED_DEV'} -p ip -j RETURN >/dev/null 2>&1
iptables -t mangle --delete PREROUTING -i $qossettings{'RED_DEV'} -p ah -j RETURN >/dev/null 2>&1
iptables -t mangle --delete PREROUTING -i $qossettings{'RED_DEV'} -p esp -j RETURN >/dev/null 2>&1
iptables -t mangle --delete PREROUTING -i $qossettings{'RED_DEV'} -p ip -j RETURN >/dev/null 2>&1
- iptables -t mangle --delete POSTROUTING -m mark ! --mark 0 ! -o $qossettings{'RED_DEV'} -j IMQ --todev 0 >/dev/null 2>&1
- iptables -t mangle --delete PREROUTING -i $qossettings{'RED_DEV'} -j IMQ --todev 0 >/dev/null 2>&1
- # rmmod imq # this crash on 2.6.25.xx
- # REMOVE & FLUSH CHAINS
iptables -t mangle --delete POSTROUTING -o $qossettings{'RED_DEV'} -j QOS-OUT >/dev/null 2>&1
iptables -t mangle --delete POSTROUTING -o $qossettings{'RED_DEV'} -j QOS-TOS >/dev/null 2>&1
iptables -t mangle --flush QOS-OUT >/dev/null 2>&1
--
2.12.2
next prev parent reply other threads:[~2019-10-21 15:48 UTC|newest]
Thread overview: 10+ messages / expand[flat|nested] mbox.gz Atom feed top
2019-10-21 15:48 [PATCH 01/10] QoS: Do not manually load iptables modules Daniel Weismüller
2019-10-21 15:48 ` Daniel Weismüller [this message]
2019-10-21 15:48 ` [PATCH 03/10] QoS: Tidy up qdiscs after QoS is being stopped Daniel Weismüller
2019-10-21 15:49 ` [PATCH 04/10] QoS: Start qosd immediately Daniel Weismüller
2019-10-21 15:49 ` [PATCH 05/10] QoS: Do not delete egress qdisc after classes have been created Daniel Weismüller
2019-10-21 15:49 ` [PATCH 06/10] linux+iptables: Drop support for IMQ Daniel Weismüller
2019-10-21 15:49 ` [PATCH 07/10] QoS: Suppress an error message when cleaning up from previous runs Daniel Weismüller
2019-10-21 15:49 ` [PATCH 08/10] QoS: Drop support for subclasses Daniel Weismüller
2019-10-21 15:49 ` [PATCH 09/10] Revert "Make IMQ Switchable between PREROUTING and POSTROUTING" Daniel Weismüller
2019-10-21 15:49 ` [PATCH 10/10] QoS: Use CONNMARK to mark connections in connection tracking Daniel Weismüller
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20191021154906.24418-2-daniel.weismueller@ipfire.org \
--to=daniel.weismueller@ipfire.org \
--cc=development@lists.ipfire.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox