[<prev] [next>] [thread-next>] [day] [month] [year] [list]
Message-Id: <20200316230223.242532-1-edumazet@google.com>
Date: Mon, 16 Mar 2020 16:02:20 -0700
From: Eric Dumazet <edumazet@...gle.com>
To: "David S . Miller" <davem@...emloft.net>
Cc: netdev <netdev@...r.kernel.org>,
Willem de Bruijn <willemb@...gle.com>,
Eric Dumazet <edumazet@...gle.com>,
Eric Dumazet <eric.dumazet@...il.com>
Subject: [PATCH net-next 0/3] net_sched: allow use of hrtimer slack
Packet schedulers have used hrtimers with exact expiry times.
Some of them can afford having a slack, in order to reduce
the number of timer interrupts and feed bigger batches
to increase efficiency.
FQ for example does not care if throttled packets are
sent with an additional (small) delay.
Original observation of having maybe too many interrupts
was made by Willem de Bruijn.
Eric Dumazet (3):
net_sched: add qdisc_watchdog_schedule_range_ns()
net_sched: do not reprogram a timer about to expire
net_sched: sch_fq: enable use of hrtimer slack
include/net/pkt_sched.h | 10 +++++++++-
include/uapi/linux/pkt_sched.h | 2 ++
net/sched/sch_api.c | 21 ++++++++++++++-------
net/sched/sch_fq.c | 19 +++++++++++++++----
4 files changed, 40 insertions(+), 12 deletions(-)
--
2.25.1.481.gfbce0eb801-goog
Powered by blists - more mailing lists