1 .\" Copyright (c) 2000 Jeroen Ruigrok van der Werven
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
31 .Nm BUS_TEARDOWN_INTR ,
33 .Nd create, attach and teardown an interrupt handler
39 .Fa "device_t dev" "device_t child" "struct resource *irq" "int flags"
40 .Fa "driver_filter_t *filter" "driver_intr_t *ithread" "void *arg"
45 .Fa "device_t dev" "struct resource *r" "int flags"
46 .Fa "driver_filter_t filter" "driver_intr_t ithread" "void *arg"
51 .Fa "device_t dev" "device_t child" "struct resource *irq" "void *cookiep"
54 .Fn bus_teardown_intr "device_t dev" "struct resource *r" "void *cookiep"
59 will create and attach an interrupt handler to an interrupt
60 previously allocated by the resource manager's
61 .Xr BUS_ALLOC_RESOURCE 9
67 and give the broad category of interrupt.
70 also tell the interrupt handlers about certain
71 device driver characteristics.
73 marks the handler as being
74 an exclusive handler for this interrupt.
76 tells the scheduler that the interrupt handler
77 is well behaved in a preemptive environment
80 to be protected by the ``Giant Lock'' mutex.
82 marks the interrupt as being a good source of entropy -
83 this may be used by the entropy device
86 To define a time-critical handler that will not execute any potentially
87 blocking operation, use the
92 section below for information on writing a filter.
97 will be called with the value
101 .Sx "ithread Routines"
102 section below for more information on writing an interrupt handler.
106 argument is a pointer to a
110 will write a cookie for the parent bus' use to if it is successful in
111 establishing an interrupt.
112 Driver writers may assume that this cookie will be non-zero.
113 The nexus driver will write 0 on failure to
116 The interrupt handler will be detached by
117 .Fn BUS_TEARDOWN_INTR .
118 The cookie needs to be passed to
119 .Fn BUS_TEARDOWN_INTR
120 in order to tear down the correct interrupt handler.
122 .Fn BUS_TEARDOWN_INTR
123 returns, it is guaranteed that the interrupt function is not active and
124 will no longer be called.
126 Mutexes are not allowed to be held across calls to these functions.
127 .Ss "Filter Routines"
128 A filter runs in primary interrupt context.
129 In this context, normal mutexes cannot be used.
130 Only the spin lock version of these can be used (specified by passing
134 when initializing the mutex).
136 and similar routines can be called.
137 Atomic operations from
140 Reads and writes to hardware through
143 PCI configuration registers may be read and written.
144 All other kernel interfaces cannot be used.
146 In this restricted environment, care must be taken to account for all
148 A careful analysis of races should be done as well.
149 It is generally cheaper to take an extra interrupt, for example, than
150 to protect variables with spinlocks.
151 Read, modify, write cycles of hardware registers need to be carefully
152 analyzed if other threads are accessing the same registers.
154 Generally, a filter routine will use one of two strategies.
155 The first strategy is to simply mask the interrupt in hardware and
158 routine to read the state from the hardware and then reenable
162 also acknowledges the interrupt before re-enabling the interrupt
164 Most PCI hardware can mask its interrupt source.
166 The second common approach is to use a filter with multiple
169 In this case, the filter acknowledges the interrupts and queues the
170 work to the appropriate taskqueue.
171 Where one has to multiplex different kinds of interrupt sources, like
172 a network card's transmit and receive paths, this can reduce lock
173 contention and increase performance.
177 from inside a filter.
178 You may not call anything that uses a normal mutex.
179 Witness may complain about these.
180 .Ss "ithread Routines"
181 You can do whatever you want in an ithread routine, except sleep.
182 Care must be taken not to sleep in an ithread.
183 In addition, one should minimize lock contention in an ithread routine
184 because contested locks ripple over to all other ithread routines on
187 Sleeping is voluntarily giving up control of your thread.
188 All the sleep routine found in
191 Waiting for a condition variable described in
194 Calling any function that does any of these things is sleeping.
196 Zero is returned on success,
197 otherwise an appropriate error is returned.
205 This manual page was written by
206 .An Jeroen Ruigrok van der Werven Aq Mt asmodai@FreeBSD.org
207 based on the manual pages for
212 .An Doug Rabson Aq Mt dfr@FreeBSD.org .