1RANDOM(4) Linux Programmer's Manual RANDOM(4)
2
6 random, urandom - kernel random number source devices
7
9 #include <linux/random.h>
10 int ioctl(fd, RNDrequest, param);
12
14 The character special files /dev/random and /dev/urandom (present since
15 Linux 1.3.30) provide an interface to the kernel's random number gener‐
16 ator. The file /dev/random has major device number 1 and minor device
17 number 8. The file /dev/urandom has major device number 1 and minor
18 device number 9.
19 The random number generator gathers environmental noise from device
21 drivers and other sources into an entropy pool. The generator also
22 keeps an estimate of the number of bits of noise in the entropy pool.
23 From this entropy pool, random numbers are created.
24 Linux 3.17 and later provides the simpler and safer getrandom(2) inter‐
26 face which requires no special files; see the getrandom(2) manual page
27 for details.
28 When read, the /dev/urandom device returns random bytes using a pseudo‐
30 random number generator seeded from the entropy pool. Reads from this
31 device do not block (i.e., the CPU is not yielded), but can incur an
32 appreciable delay when requesting large amounts of data.
33 When read during early boot time, /dev/urandom may return data prior to
35 the entropy pool being initialized. If this is of concern in your ap‐
36 plication, use getrandom(2) or /dev/random instead.
37 The /dev/random device is a legacy interface which dates back to a time
39 where the cryptographic primitives used in the implementation of
40 /dev/urandom were not widely trusted. It will return random bytes only
41 within the estimated number of bits of fresh noise in the entropy pool,
42 blocking if necessary. /dev/random is suitable for applications that
43 need high quality randomness, and can afford indeterminate delays.
44 When the entropy pool is empty, reads from /dev/random will block until
46 additional environmental noise is gathered. If open(2) is called for
47 /dev/random with the O_NONBLOCK flag, a subsequent read(2) will not
48 block if the requested number of bytes is not available. Instead, the
49 available bytes are returned. If no byte is available, read(2) will
50 return -1 and errno will be set to EAGAIN.
51 The O_NONBLOCK flag has no effect when opening /dev/urandom. When
53 calling read(2) for the device /dev/urandom, reads of up to 256 bytes
54 will return as many bytes as are requested and will not be interrupted
55 by a signal handler. Reads with a buffer over this limit may return
56 less than the requested number of bytes or fail with the error EINTR,
57 if interrupted by a signal handler.
58 Since Linux 3.16, a read(2) from /dev/urandom will return at most
60 32 MB. A read(2) from /dev/random will return at most 512 bytes (340
61 bytes on Linux kernels before version 2.6.12).
62 Writing to /dev/random or /dev/urandom will update the entropy pool
64 with the data written, but this will not result in a higher entropy
65 count. This means that it will impact the contents read from both
66 files, but it will not make reads from /dev/random faster.
67 Usage
69 The /dev/random interface is considered a legacy interface, and
70 /dev/urandom is preferred and sufficient in all use cases, with the ex‐
71 ception of applications which require randomness during early boot
72 time; for these applications, getrandom(2) must be used instead, be‐
73 cause it will block until the entropy pool is initialized.
74 If a seed file is saved across reboots as recommended below, the output
76 is cryptographically secure against attackers without local root access
77 as soon as it is reloaded in the boot sequence, and perfectly adequate
78 for network encryption session keys. (All major Linux distributions
79 have saved the seed file across reboots since 2000 at least.) Since
80 reads from /dev/random may block, users will usually want to open it in
81 nonblocking mode (or perform a read with timeout), and provide some
82 sort of user notification if the desired entropy is not immediately
83 available.
84 Configuration
86 If your system does not have /dev/random and /dev/urandom created al‐
87 ready, they can be created with the following commands:
88 mknod -m 666 /dev/random c 1 8
90 mknod -m 666 /dev/urandom c 1 9
91 chown root:root /dev/random /dev/urandom
92 When a Linux system starts up without much operator interaction, the
94 entropy pool may be in a fairly predictable state. This reduces the
95 actual amount of noise in the entropy pool below the estimate. In or‐
96 der to counteract this effect, it helps to carry entropy pool informa‐
97 tion across shut-downs and start-ups. To do this, add the lines to an
98 appropriate script which is run during the Linux system start-up se‐
99 quence:
100 echo "Initializing random number generator..."
102 random_seed=/var/run/random-seed
103 # Carry a random seed from start-up to start-up
104 # Load and then save the whole entropy pool
105 if [ -f $random_seed ]; then
106 cat $random_seed >/dev/urandom
107 else
108 touch $random_seed
109 fi
110 chmod 600 $random_seed
111 poolfile=/proc/sys/kernel/random/poolsize
112 [ -r $poolfile ] && bits=$(cat $poolfile) || bits=4096
113 bytes=$(expr $bits / 8)
114 dd if=/dev/urandom of=$random_seed count=1 bs=$bytes
115 Also, add the following lines in an appropriate script which is run
117 during the Linux system shutdown:
118 # Carry a random seed from shut-down to start-up
120 # Save the whole entropy pool
121 echo "Saving random seed..."
122 random_seed=/var/run/random-seed
123 touch $random_seed
124 chmod 600 $random_seed
125 poolfile=/proc/sys/kernel/random/poolsize
126 [ -r $poolfile ] && bits=$(cat $poolfile) || bits=4096
127 bytes=$(expr $bits / 8)
128 dd if=/dev/urandom of=$random_seed count=1 bs=$bytes
129 In the above examples, we assume Linux 2.6.0 or later, where
131 /proc/sys/kernel/random/poolsize returns the size of the entropy pool
132 in bits (see below).
133 /proc interfaces
135 The files in the directory /proc/sys/kernel/random (present since
136 2.3.16) provide additional information about the /dev/random device:
137 entropy_avail
139 This read-only file gives the available entropy, in bits. This
140 will be a number in the range 0 to 4096.
141 poolsize
143 This file gives the size of the entropy pool. The semantics of
144 this file vary across kernel versions:
145 Linux 2.4:
147 This file gives the size of the entropy pool in bytes.
148 Normally, this file will have the value 512, but it is
149 writable, and can be changed to any value for which an
150 algorithm is available. The choices are 32, 64, 128,
151 256, 512, 1024, or 2048.
152 Linux 2.6 and later:
154 This file is read-only, and gives the size of the entropy
155 pool in bits. It contains the value 4096.
156 read_wakeup_threshold
158 This file contains the number of bits of entropy required for
159 waking up processes that sleep waiting for entropy from
160 /dev/random. The default is 64.
161 write_wakeup_threshold
163 This file contains the number of bits of entropy below which we
164 wake up processes that do a select(2) or poll(2) for write ac‐
165 cess to /dev/random. These values can be changed by writing to
166 the files.
167 uuid and boot_id
169 These read-only files contain random strings like
170 6fd5a44b-35f4-4ad4-a9b9-6b9be13e1fe9. The former is generated
171 afresh for each read, the latter was generated once.
172 ioctl(2) interface
174 The following ioctl(2) requests are defined on file descriptors con‐
175 nected to either /dev/random or /dev/urandom. All requests performed
176 will interact with the input entropy pool impacting both /dev/random
177 and /dev/urandom. The CAP_SYS_ADMIN capability is required for all re‐
178 quests except RNDGETENTCNT.
179 RNDGETENTCNT
181 Retrieve the entropy count of the input pool, the contents will
182 be the same as the entropy_avail file under proc. The result
183 will be stored in the int pointed to by the argument.
184 RNDADDTOENTCNT
186 Increment or decrement the entropy count of the input pool by
187 the value pointed to by the argument.
188 RNDGETPOOL
190 Removed in Linux 2.6.9.
191 RNDADDENTROPY
193 Add some additional entropy to the input pool, incrementing the
194 entropy count. This differs from writing to /dev/random or
195 /dev/urandom, which only adds some data but does not increment
196 the entropy count. The following structure is used:
197 struct rand_pool_info {
199 int entropy_count;
200 int buf_size;
201 __u32 buf[0];
202 };
203 Here entropy_count is the value added to (or subtracted from)
205 the entropy count, and buf is the buffer of size buf_size which
206 gets added to the entropy pool.
207 RNDZAPENTCNT, RNDCLEARPOOL
209 Zero the entropy count of all pools and add some system data
210 (such as wall clock) to the pools.
211
213 /dev/random
214 /dev/urandom
215
217 For an overview and comparison of the various interfaces that can be
218 used to obtain randomness, see random(7).
219
221 During early boot time, reads from /dev/urandom may return data prior
222 to the entropy pool being initialized.
223
225 mknod(1), getrandom(2), random(7)
226 RFC 1750, "Randomness Recommendations for Security"
228
230 This page is part of release 5.13 of the Linux man-pages project. A
231 description of the project, information about reporting bugs, and the
232 latest version of this page, can be found at
233 https://www.kernel.org/doc/man-pages/.
234Linux 2021-03-22 RANDOM(4)