[PATCH] rcu: add rcu_sync torture type to rcutorture
[deliverable/linux.git] / Documentation / RCU / torture.txt
CommitLineData
a241ec65
PM
1RCU Torture Test Operation
2
3
4CONFIG_RCU_TORTURE_TEST
5
6The CONFIG_RCU_TORTURE_TEST config option is available for all RCU
7implementations. It creates an rcutorture kernel module that can
8be loaded to run a torture test. The test periodically outputs
9status messages via printk(), which can be examined via the dmesg
72e9bb54 10command (perhaps grepping for "torture"). The test is started
a241ec65
PM
11when the module is loaded, and stops when the module is unloaded.
12
13However, actually setting this config option to "y" results in the system
14running the test immediately upon boot, and ending only when the system
15is taken down. Normally, one will instead want to build the system
16with CONFIG_RCU_TORTURE_TEST=m and to use modprobe and rmmod to control
17the test, perhaps using a script similar to the one shown at the end of
18this document. Note that you will need CONFIG_MODULE_UNLOAD in order
19to be able to end the test.
20
21
22MODULE PARAMETERS
23
24This module has the following parameters:
25
26nreaders This is the number of RCU reading threads supported.
27 The default is twice the number of CPUs. Why twice?
28 To properly exercise RCU implementations with preemptible
29 read-side critical sections.
30
b772e1dd
JT
31nfakewriters This is the number of RCU fake writer threads to run. Fake
32 writer threads repeatedly use the synchronous "wait for
33 current readers" function of the interface selected by
34 torture_type, with a delay between calls to allow for various
35 different numbers of writers running in parallel.
36 nfakewriters defaults to 4, which provides enough parallelism
37 to trigger special cases caused by multiple writers, such as
38 the synchronize_srcu() early return optimization.
39
a241ec65
PM
40stat_interval The number of seconds between output of torture
41 statistics (via printk()). Regardless of the interval,
42 statistics are printed when the module is unloaded.
43 Setting the interval to zero causes the statistics to
44 be printed -only- when the module is unloaded, and this
45 is the default.
46
29766f1e
PM
47shuffle_interval
48 The number of seconds to keep the test threads affinitied
49 to a particular subset of the CPUs. Used in conjunction
50 with test_no_idle_hz.
51
52test_no_idle_hz Whether or not to test the ability of RCU to operate in
53 a kernel that disables the scheduling-clock interrupt to
54 idle CPUs. Boolean parameter, "1" to test, "0" otherwise.
55
20d2e428
JT
56torture_type The type of RCU to test: "rcu" for the rcu_read_lock() API,
57 "rcu_sync" for rcu_read_lock() with synchronous reclamation,
58 "rcu_bh" for the rcu_read_lock_bh() API, and "srcu" for the
59 "srcu_read_lock()" API.
72e9bb54 60
a241ec65
PM
61verbose Enable debug printk()s. Default is disabled.
62
63
64OUTPUT
65
66The statistics output is as follows:
67
72e9bb54
PM
68 rcu-torture: --- Start of test: nreaders=16 stat_interval=0 verbose=0
69 rcu-torture: rtc: 0000000000000000 ver: 1916 tfle: 0 rta: 1916 rtaf: 0 rtf: 1915
70 rcu-torture: Reader Pipe: 1466408 9747 0 0 0 0 0 0 0 0 0
71 rcu-torture: Reader Batch: 1464477 11678 0 0 0 0 0 0 0 0
72 rcu-torture: Free-Block Circulation: 1915 1915 1915 1915 1915 1915 1915 1915 1915 1915 0
73 rcu-torture: --- End of test
a241ec65 74
72e9bb54 75The command "dmesg | grep torture:" will extract this information on
a241ec65
PM
76most systems. On more esoteric configurations, it may be necessary to
77use other commands to access the output of the printk()s used by
78the RCU torture test. The printk()s use KERN_ALERT, so they should
79be evident. ;-)
80
81The entries are as follows:
82
83o "ggp": The number of counter flips (or batches) since boot.
84
85o "rtc": The hexadecimal address of the structure currently visible
86 to readers.
87
88o "ver": The number of times since boot that the rcutw writer task
89 has changed the structure visible to readers.
90
91o "tfle": If non-zero, indicates that the "torture freelist"
92 containing structure to be placed into the "rtc" area is empty.
93 This condition is important, since it can fool you into thinking
94 that RCU is working when it is not. :-/
95
96o "rta": Number of structures allocated from the torture freelist.
97
98o "rtaf": Number of allocations from the torture freelist that have
99 failed due to the list being empty.
100
101o "rtf": Number of frees into the torture freelist.
102
103o "Reader Pipe": Histogram of "ages" of structures seen by readers.
104 If any entries past the first two are non-zero, RCU is broken.
105 And rcutorture prints the error flag string "!!!" to make sure
106 you notice. The age of a newly allocated structure is zero,
107 it becomes one when removed from reader visibility, and is
108 incremented once per grace period subsequently -- and is freed
109 after passing through (RCU_TORTURE_PIPE_LEN-2) grace periods.
110
111 The output displayed above was taken from a correctly working
112 RCU. If you want to see what it looks like when broken, break
113 it yourself. ;-)
114
115o "Reader Batch": Another histogram of "ages" of structures seen
116 by readers, but in terms of counter flips (or batches) rather
117 than in terms of grace periods. The legal number of non-zero
118 entries is again two. The reason for this separate view is
119 that it is easier to get the third entry to show up in the
120 "Reader Batch" list than in the "Reader Pipe" list.
121
122o "Free-Block Circulation": Shows the number of torture structures
123 that have reached a given point in the pipeline. The first element
124 should closely correspond to the number of structures allocated,
125 the second to the number that have been removed from reader view,
126 and all but the last remaining to the corresponding number of
127 passes through a grace period. The last entry should be zero,
128 as it is only incremented if a torture structure's counter
129 somehow gets incremented farther than it should.
130
b2896d2e
PM
131Different implementations of RCU can provide implementation-specific
132additional information. For example, SRCU provides the following:
133
134 srcu-torture: rtc: f8cf46a8 ver: 355 tfle: 0 rta: 356 rtaf: 0 rtf: 346 rtmbe: 0
135 srcu-torture: Reader Pipe: 559738 939 0 0 0 0 0 0 0 0 0
136 srcu-torture: Reader Batch: 560434 243 0 0 0 0 0 0 0 0
137 srcu-torture: Free-Block Circulation: 355 354 353 352 351 350 349 348 347 346 0
138 srcu-torture: per-CPU(idx=1): 0(0,1) 1(0,1) 2(0,0) 3(0,1)
139
140The first four lines are similar to those for RCU. The last line shows
141the per-CPU counter state. The numbers in parentheses are the values
142of the "old" and "current" counters for the corresponding CPU. The
143"idx" value maps the "old" and "current" values to the underlying array,
144and is useful for debugging.
145
a241ec65
PM
146
147USAGE
148
149The following script may be used to torture RCU:
150
151 #!/bin/sh
152
153 modprobe rcutorture
154 sleep 100
155 rmmod rcutorture
72e9bb54 156 dmesg | grep torture:
a241ec65
PM
157
158The output can be manually inspected for the error flag of "!!!".
159One could of course create a more elaborate script that automatically
29766f1e
PM
160checked for such errors. The "rmmod" command forces a "SUCCESS" or
161"FAILURE" indication to be printk()ed.
This page took 0.125544 seconds and 5 git commands to generate.