1 .\" Copyright (c) 2002 Luigi Rizzo
2 .\" All rights reserved.
4 .\" Redistribution and use in source and binary forms, with or without
5 .\" modification, are permitted provided that the following conditions
7 .\" 1. Redistributions of source code must retain the above copyright
8 .\" notice, this list of conditions and the following disclaimer.
9 .\" 2. Redistributions in binary form must reproduce the above copyright
10 .\" notice, this list of conditions and the following disclaimer in the
11 .\" documentation and/or other materials provided with the distribution.
13 .\" THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND
14 .\" ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
15 .\" IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
16 .\" ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE
17 .\" FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
18 .\" DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
19 .\" OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
20 .\" HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
21 .\" LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
22 .\" OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
25 .\" $FreeBSD: src/share/man/man4/polling.4,v 1.27 2007/04/06 14:25:14 brueffer Exp $
26 .\" $DragonFly: src/share/man/man4/polling.4,v 1.12 2007/10/03 09:55:25 sephe Exp $
33 .Nd device polling support
35 .Cd "options DEVICE_POLLING"
39 for brevity) refers to a technique that
40 lets the operating system periodically poll devices, instead of
41 relying on the devices to generate interrupts when they need attention.
42 This might seem inefficient and counterintuitive, but when done
45 gives more control to the operating system on
46 when and how to handle devices, with a number of advantages in terms
47 of system responsiveness and performance.
51 reduces the overhead for context
52 switches which is incurred when servicing interrupts, and
53 gives more control on the scheduling of a CPU between various
54 tasks (user processes, software interrupts, device handling)
55 which ultimately reduces the chances of livelock in the system.
56 .Ss Principles of Operation
57 In the normal, interrupt-based mode, devices generate an interrupt
58 whenever they need attention.
60 context switch and the execution of an interrupt handler
61 which performs whatever processing is needed by the device.
62 The duration of the interrupt handler is potentially unbounded
63 unless the device driver has been programmed with real-time
64 concerns in mind (which is generally not the case for
67 Furthermore, under heavy traffic load, the system might be
68 persistently processing interrupts without being able to
69 complete other work, either in the kernel or in userland.
71 Device polling disables interrupts by polling devices on clock
73 This way, the context switch overhead is removed.
75 the operating system can control accurately how much work to spend
76 in handling device events, and thus prevent livelock by reserving
77 some amount of CPU to other tasks.
81 also changes the way software network interrupts
82 are scheduled, so there is never the risk of livelock because
83 packets are not processed to completion.
85 Currently only network interface drivers support the
88 It is turned on and off with help of
92 The following tunables can be set from
94 .Bl -tag -width indent -compact
95 .It Va kern.polling.enable
101 .It Va kern.polling.cpumask
102 A bitmask that controls which CPUs support device polling.
103 Default is 0xffffffff.
108 is controlled by the following per CPU
114 .Bl -tag -width indent -compact
115 .It Va kern.polling.X.enable
121 .It Va kern.polling.X.pollhz
122 The polling frequency, whose range is 1 to 30000.
125 .It Va kern.polling.cpumask
126 A read only bitmask of the CPUs that support device polling.
128 .It Va kern.polling.defcpu
129 The default CPU used to run device polling (read only).
131 .It Va kern.polling.X.user_frac
134 is enabled, and provided that there is some work to do,
135 up to this percent of the CPU cycles is reserved to userland tasks,
136 the remaining fraction being available for
141 .It Va kern.polling.X.burst
142 Maximum number of packets grabbed from each network interface in
144 This number is dynamically adjusted by the kernel,
145 according to the programmed
146 .Va user_frac , burst_max ,
147 CPU speed, and system load.
149 .It Va kern.polling.X.each_burst
150 The burst above is split into smaller chunks of this number of
151 packets, going round-robin among all interfaces registered for
153 This prevents the case that a large burst from a single interface
154 can saturate the IP interrupt queue
155 .Pq Va net.inet.ip.intr_queue_maxlen .
158 .It Va kern.polling.X.burst_max
160 .Va kern.polling.burst .
163 is enabled, each interface can receive at most
164 .Pq Va pollhz No * Va burst_max
165 packets per second unless there are spare CPU cycles available for
168 This number should be tuned to match the expected load
169 (which can be quite high with GigE cards).
170 Default is 150 which is adequate for 100Mbit network and pollhz=1000.
172 .It Va kern.polling.X.reg_frac
173 Controls how often (every
174 .Va reg_frac No / Va pollhz
175 seconds) the status registers of the device are checked for error
176 conditions and the like.
177 Increasing this value reduces the load on the bus, but also delays
181 .It Va kern.polling.X.handlers
182 How many active devices have registered for
185 .It Va kern.polling.X.short_ticks
186 .It Va kern.polling.X.lost_polls
187 .It Va kern.polling.X.pending_polls
188 .It Va kern.polling.X.residual_burst
189 .It Va kern.polling.X.phase
190 .It Va kern.polling.X.suspect
191 .It Va kern.polling.X.stalled
194 .Sh SUPPORTED DEVICES
195 Device polling requires explicit modifications to the device drivers.
196 As of this writing, the
214 devices are supported, with others in the works.
215 The modifications are rather straightforward, consisting in
216 the extraction of the inner part of the interrupt service routine
217 and writing a callback function,
220 to probe the device for events and process them.
222 conditionally compiled sections of the devices mentioned above
225 In order to reduce the latency in processing packets,
226 it is advisable to set the
229 .Va kern.polling.X.pollhz
232 Device polling first appeared in
238 The device polling code was rewritten by
240 based on the original code by
241 .An Luigi Rizzo Aq luigi@iet.unipi.it .
243 made the polling frequency settable at runtime and added per CPU polling.