1 .. _kernel_hacking_hack: 1 .. _kernel_hacking_hack:
2 2
3 ============================================ 3 ============================================
4 Unreliable Guide To Hacking The Linux Kernel 4 Unreliable Guide To Hacking The Linux Kernel
5 ============================================ 5 ============================================
6 6
7 :Author: Rusty Russell 7 :Author: Rusty Russell
8 8
9 Introduction 9 Introduction
10 ============ 10 ============
11 11
12 Welcome, gentle reader, to Rusty's Remarkably 12 Welcome, gentle reader, to Rusty's Remarkably Unreliable Guide to Linux
13 Kernel Hacking. This document describes the co 13 Kernel Hacking. This document describes the common routines and general
14 requirements for kernel code: its goal is to s 14 requirements for kernel code: its goal is to serve as a primer for Linux
15 kernel development for experienced C programme 15 kernel development for experienced C programmers. I avoid implementation
16 details: that's what the code is for, and I ig 16 details: that's what the code is for, and I ignore whole tracts of
17 useful routines. 17 useful routines.
18 18
19 Before you read this, please understand that I 19 Before you read this, please understand that I never wanted to write
20 this document, being grossly under-qualified, 20 this document, being grossly under-qualified, but I always wanted to
21 read it, and this was the only way. I hope it 21 read it, and this was the only way. I hope it will grow into a
22 compendium of best practice, common starting p 22 compendium of best practice, common starting points and random
23 information. 23 information.
24 24
25 The Players 25 The Players
26 =========== 26 ===========
27 27
28 At any time each of the CPUs in a system can b 28 At any time each of the CPUs in a system can be:
29 29
30 - not associated with any process, serving a 30 - not associated with any process, serving a hardware interrupt;
31 31
32 - not associated with any process, serving a 32 - not associated with any process, serving a softirq or tasklet;
33 33
34 - running in kernel space, associated with a 34 - running in kernel space, associated with a process (user context);
35 35
36 - running a process in user space. 36 - running a process in user space.
37 37
38 There is an ordering between these. The bottom 38 There is an ordering between these. The bottom two can preempt each
39 other, but above that is a strict hierarchy: e 39 other, but above that is a strict hierarchy: each can only be preempted
40 by the ones above it. For example, while a sof 40 by the ones above it. For example, while a softirq is running on a CPU,
41 no other softirq will preempt it, but a hardwa 41 no other softirq will preempt it, but a hardware interrupt can. However,
42 any other CPUs in the system execute independe 42 any other CPUs in the system execute independently.
43 43
44 We'll see a number of ways that the user conte 44 We'll see a number of ways that the user context can block interrupts,
45 to become truly non-preemptable. 45 to become truly non-preemptable.
46 46
47 User Context 47 User Context
48 ------------ 48 ------------
49 49
50 User context is when you are coming in from a 50 User context is when you are coming in from a system call or other trap:
51 like userspace, you can be preempted by more i 51 like userspace, you can be preempted by more important tasks and by
52 interrupts. You can sleep, by calling :c:func: 52 interrupts. You can sleep, by calling :c:func:`schedule()`.
53 53
54 .. note:: 54 .. note::
55 55
56 You are always in user context on module l 56 You are always in user context on module load and unload, and on
57 operations on the block device layer. 57 operations on the block device layer.
58 58
59 In user context, the ``current`` pointer (indi 59 In user context, the ``current`` pointer (indicating the task we are
60 currently executing) is valid, and :c:func:`in 60 currently executing) is valid, and :c:func:`in_interrupt()`
61 (``include/linux/preempt.h``) is false. 61 (``include/linux/preempt.h``) is false.
62 62
63 .. warning:: 63 .. warning::
64 64
65 Beware that if you have preemption or soft 65 Beware that if you have preemption or softirqs disabled (see below),
66 :c:func:`in_interrupt()` will return a fal 66 :c:func:`in_interrupt()` will return a false positive.
67 67
68 Hardware Interrupts (Hard IRQs) 68 Hardware Interrupts (Hard IRQs)
69 ------------------------------- 69 -------------------------------
70 70
71 Timer ticks, network cards and keyboard are ex 71 Timer ticks, network cards and keyboard are examples of real hardware
72 which produce interrupts at any time. The kern 72 which produce interrupts at any time. The kernel runs interrupt
73 handlers, which services the hardware. The ker 73 handlers, which services the hardware. The kernel guarantees that this
74 handler is never re-entered: if the same inter 74 handler is never re-entered: if the same interrupt arrives, it is queued
75 (or dropped). Because it disables interrupts, 75 (or dropped). Because it disables interrupts, this handler has to be
76 fast: frequently it simply acknowledges the in 76 fast: frequently it simply acknowledges the interrupt, marks a 'software
77 interrupt' for execution and exits. 77 interrupt' for execution and exits.
78 78
79 You can tell you are in a hardware interrupt, 79 You can tell you are in a hardware interrupt, because in_hardirq() returns
80 true. 80 true.
81 81
82 .. warning:: 82 .. warning::
83 83
84 Beware that this will return a false posit 84 Beware that this will return a false positive if interrupts are
85 disabled (see below). 85 disabled (see below).
86 86
87 Software Interrupt Context: Softirqs and Taskl 87 Software Interrupt Context: Softirqs and Tasklets
88 ---------------------------------------------- 88 -------------------------------------------------
89 89
90 Whenever a system call is about to return to u 90 Whenever a system call is about to return to userspace, or a hardware
91 interrupt handler exits, any 'software interru 91 interrupt handler exits, any 'software interrupts' which are marked
92 pending (usually by hardware interrupts) are r 92 pending (usually by hardware interrupts) are run (``kernel/softirq.c``).
93 93
94 Much of the real interrupt handling work is do 94 Much of the real interrupt handling work is done here. Early in the
95 transition to SMP, there were only 'bottom hal 95 transition to SMP, there were only 'bottom halves' (BHs), which didn't
96 take advantage of multiple CPUs. Shortly after 96 take advantage of multiple CPUs. Shortly after we switched from wind-up
97 computers made of match-sticks and snot, we ab 97 computers made of match-sticks and snot, we abandoned this limitation
98 and switched to 'softirqs'. 98 and switched to 'softirqs'.
99 99
100 ``include/linux/interrupt.h`` lists the differ 100 ``include/linux/interrupt.h`` lists the different softirqs. A very
101 important softirq is the timer softirq (``incl 101 important softirq is the timer softirq (``include/linux/timer.h``): you
102 can register to have it call functions for you 102 can register to have it call functions for you in a given length of
103 time. 103 time.
104 104
105 Softirqs are often a pain to deal with, since 105 Softirqs are often a pain to deal with, since the same softirq will run
106 simultaneously on more than one CPU. For this 106 simultaneously on more than one CPU. For this reason, tasklets
107 (``include/linux/interrupt.h``) are more often 107 (``include/linux/interrupt.h``) are more often used: they are
108 dynamically-registrable (meaning you can have 108 dynamically-registrable (meaning you can have as many as you want), and
109 they also guarantee that any tasklet will only 109 they also guarantee that any tasklet will only run on one CPU at any
110 time, although different tasklets can run simu 110 time, although different tasklets can run simultaneously.
111 111
112 .. warning:: 112 .. warning::
113 113
114 The name 'tasklet' is misleading: they hav 114 The name 'tasklet' is misleading: they have nothing to do with
115 'tasks'. 115 'tasks'.
116 116
117 You can tell you are in a softirq (or tasklet) 117 You can tell you are in a softirq (or tasklet) using the
118 :c:func:`in_softirq()` macro (``include/linux/ 118 :c:func:`in_softirq()` macro (``include/linux/preempt.h``).
119 119
120 .. warning:: 120 .. warning::
121 121
122 Beware that this will return a false posit 122 Beware that this will return a false positive if a
123 :ref:`bottom half lock <local_bh_disable>` 123 :ref:`bottom half lock <local_bh_disable>` is held.
124 124
125 Some Basic Rules 125 Some Basic Rules
126 ================ 126 ================
127 127
128 No memory protection 128 No memory protection
129 If you corrupt memory, whether in user con 129 If you corrupt memory, whether in user context or interrupt context,
130 the whole machine will crash. Are you sure 130 the whole machine will crash. Are you sure you can't do what you
131 want in userspace? 131 want in userspace?
132 132
133 No floating point or MMX 133 No floating point or MMX
134 The FPU context is not saved; even in user 134 The FPU context is not saved; even in user context the FPU state
135 probably won't correspond with the current 135 probably won't correspond with the current process: you would mess
136 with some user process' FPU state. If you 136 with some user process' FPU state. If you really want to do this,
137 you would have to explicitly save/restore 137 you would have to explicitly save/restore the full FPU state (and
138 avoid context switches). It is generally a 138 avoid context switches). It is generally a bad idea; use fixed point
139 arithmetic first. 139 arithmetic first.
140 140
141 A rigid stack limit 141 A rigid stack limit
142 Depending on configuration options the ker 142 Depending on configuration options the kernel stack is about 3K to
143 6K for most 32-bit architectures: it's abo 143 6K for most 32-bit architectures: it's about 14K on most 64-bit
144 archs, and often shared with interrupts so 144 archs, and often shared with interrupts so you can't use it all.
145 Avoid deep recursion and huge local arrays 145 Avoid deep recursion and huge local arrays on the stack (allocate
146 them dynamically instead). 146 them dynamically instead).
147 147
148 The Linux kernel is portable 148 The Linux kernel is portable
149 Let's keep it that way. Your code should b 149 Let's keep it that way. Your code should be 64-bit clean, and
150 endian-independent. You should also minimi 150 endian-independent. You should also minimize CPU specific stuff,
151 e.g. inline assembly should be cleanly enc 151 e.g. inline assembly should be cleanly encapsulated and minimized to
152 ease porting. Generally it should be restr 152 ease porting. Generally it should be restricted to the
153 architecture-dependent part of the kernel 153 architecture-dependent part of the kernel tree.
154 154
155 ioctls: Not writing a new system call 155 ioctls: Not writing a new system call
156 ===================================== 156 =====================================
157 157
158 A system call generally looks like this:: 158 A system call generally looks like this::
159 159
160 asmlinkage long sys_mycall(int arg) 160 asmlinkage long sys_mycall(int arg)
161 { 161 {
162 return 0; 162 return 0;
163 } 163 }
164 164
165 165
166 First, in most cases you don't want to create 166 First, in most cases you don't want to create a new system call. You
167 create a character device and implement an app 167 create a character device and implement an appropriate ioctl for it.
168 This is much more flexible than system calls, 168 This is much more flexible than system calls, doesn't have to be entered
169 in every architecture's ``include/asm/unistd.h 169 in every architecture's ``include/asm/unistd.h`` and
170 ``arch/kernel/entry.S`` file, and is much more 170 ``arch/kernel/entry.S`` file, and is much more likely to be accepted by
171 Linus. 171 Linus.
172 172
173 If all your routine does is read or write some 173 If all your routine does is read or write some parameter, consider
174 implementing a :c:func:`sysfs()` interface ins 174 implementing a :c:func:`sysfs()` interface instead.
175 175
176 Inside the ioctl you're in user context to a p 176 Inside the ioctl you're in user context to a process. When a error
177 occurs you return a negated errno (see 177 occurs you return a negated errno (see
178 ``include/uapi/asm-generic/errno-base.h``, 178 ``include/uapi/asm-generic/errno-base.h``,
179 ``include/uapi/asm-generic/errno.h`` and ``inc 179 ``include/uapi/asm-generic/errno.h`` and ``include/linux/errno.h``),
180 otherwise you return 0. 180 otherwise you return 0.
181 181
182 After you slept you should check if a signal o 182 After you slept you should check if a signal occurred: the Unix/Linux
183 way of handling signals is to temporarily exit 183 way of handling signals is to temporarily exit the system call with the
184 ``-ERESTARTSYS`` error. The system call entry 184 ``-ERESTARTSYS`` error. The system call entry code will switch back to
185 user context, process the signal handler and t 185 user context, process the signal handler and then your system call will
186 be restarted (unless the user disabled that). 186 be restarted (unless the user disabled that). So you should be prepared
187 to process the restart, e.g. if you're in the 187 to process the restart, e.g. if you're in the middle of manipulating
188 some data structure. 188 some data structure.
189 189
190 :: 190 ::
191 191
192 if (signal_pending(current)) 192 if (signal_pending(current))
193 return -ERESTARTSYS; 193 return -ERESTARTSYS;
194 194
195 195
196 If you're doing longer computations: first thi 196 If you're doing longer computations: first think userspace. If you
197 **really** want to do it in kernel you should 197 **really** want to do it in kernel you should regularly check if you need
198 to give up the CPU (remember there is cooperat 198 to give up the CPU (remember there is cooperative multitasking per CPU).
199 Idiom:: 199 Idiom::
200 200
201 cond_resched(); /* Will sleep */ 201 cond_resched(); /* Will sleep */
202 202
203 203
204 A short note on interface design: the UNIX sys 204 A short note on interface design: the UNIX system call motto is "Provide
205 mechanism not policy". 205 mechanism not policy".
206 206
207 Recipes for Deadlock 207 Recipes for Deadlock
208 ==================== 208 ====================
209 209
210 You cannot call any routines which may sleep, 210 You cannot call any routines which may sleep, unless:
211 211
212 - You are in user context. 212 - You are in user context.
213 213
214 - You do not own any spinlocks. 214 - You do not own any spinlocks.
215 215
216 - You have interrupts enabled (actually, Andi 216 - You have interrupts enabled (actually, Andi Kleen says that the
217 scheduling code will enable them for you, b 217 scheduling code will enable them for you, but that's probably not
218 what you wanted). 218 what you wanted).
219 219
220 Note that some functions may sleep implicitly: 220 Note that some functions may sleep implicitly: common ones are the user
221 space access functions (\*_user) and memory al 221 space access functions (\*_user) and memory allocation functions
222 without ``GFP_ATOMIC``. 222 without ``GFP_ATOMIC``.
223 223
224 You should always compile your kernel ``CONFIG 224 You should always compile your kernel ``CONFIG_DEBUG_ATOMIC_SLEEP`` on,
225 and it will warn you if you break these rules. 225 and it will warn you if you break these rules. If you **do** break the
226 rules, you will eventually lock up your box. 226 rules, you will eventually lock up your box.
227 227
228 Really. 228 Really.
229 229
230 Common Routines 230 Common Routines
231 =============== 231 ===============
232 232
233 :c:func:`printk()` 233 :c:func:`printk()`
234 ------------------ 234 ------------------
235 235
236 Defined in ``include/linux/printk.h`` 236 Defined in ``include/linux/printk.h``
237 237
238 :c:func:`printk()` feeds kernel messages to th 238 :c:func:`printk()` feeds kernel messages to the console, dmesg, and
239 the syslog daemon. It is useful for debugging 239 the syslog daemon. It is useful for debugging and reporting errors, and
240 can be used inside interrupt context, but use 240 can be used inside interrupt context, but use with caution: a machine
241 which has its console flooded with printk mess 241 which has its console flooded with printk messages is unusable. It uses
242 a format string mostly compatible with ANSI C 242 a format string mostly compatible with ANSI C printf, and C string
243 concatenation to give it a first "priority" ar 243 concatenation to give it a first "priority" argument::
244 244
245 printk(KERN_INFO "i = %u\n", i); 245 printk(KERN_INFO "i = %u\n", i);
246 246
247 247
248 See ``include/linux/kern_levels.h``; for other 248 See ``include/linux/kern_levels.h``; for other ``KERN_`` values; these are
249 interpreted by syslog as the level. Special ca 249 interpreted by syslog as the level. Special case: for printing an IP
250 address use:: 250 address use::
251 251
252 __be32 ipaddress; 252 __be32 ipaddress;
253 printk(KERN_INFO "my ip: %pI4\n", &ipaddre 253 printk(KERN_INFO "my ip: %pI4\n", &ipaddress);
254 254
255 255
256 :c:func:`printk()` internally uses a 1K buffer 256 :c:func:`printk()` internally uses a 1K buffer and does not catch
257 overruns. Make sure that will be enough. 257 overruns. Make sure that will be enough.
258 258
259 .. note:: 259 .. note::
260 260
261 You will know when you are a real kernel h 261 You will know when you are a real kernel hacker when you start
262 typoing printf as printk in your user prog 262 typoing printf as printk in your user programs :)
263 263
264 .. note:: 264 .. note::
265 265
266 Another sidenote: the original Unix Versio 266 Another sidenote: the original Unix Version 6 sources had a comment
267 on top of its printf function: "Printf sho 267 on top of its printf function: "Printf should not be used for
268 chit-chat". You should follow that advice. 268 chit-chat". You should follow that advice.
269 269
270 :c:func:`copy_to_user()` / :c:func:`copy_from_ 270 :c:func:`copy_to_user()` / :c:func:`copy_from_user()` / :c:func:`get_user()` / :c:func:`put_user()`
271 ---------------------------------------------- 271 ---------------------------------------------------------------------------------------------------
272 272
273 Defined in ``include/linux/uaccess.h`` / ``asm 273 Defined in ``include/linux/uaccess.h`` / ``asm/uaccess.h``
274 274
275 **[SLEEPS]** 275 **[SLEEPS]**
276 276
277 :c:func:`put_user()` and :c:func:`get_user()` 277 :c:func:`put_user()` and :c:func:`get_user()` are used to get
278 and put single values (such as an int, char, o 278 and put single values (such as an int, char, or long) from and to
279 userspace. A pointer into userspace should nev 279 userspace. A pointer into userspace should never be simply dereferenced:
280 data should be copied using these routines. Bo 280 data should be copied using these routines. Both return ``-EFAULT`` or
281 0. 281 0.
282 282
283 :c:func:`copy_to_user()` and :c:func:`copy_fro 283 :c:func:`copy_to_user()` and :c:func:`copy_from_user()` are
284 more general: they copy an arbitrary amount of 284 more general: they copy an arbitrary amount of data to and from
285 userspace. 285 userspace.
286 286
287 .. warning:: 287 .. warning::
288 288
289 Unlike :c:func:`put_user()` and :c:func:`g 289 Unlike :c:func:`put_user()` and :c:func:`get_user()`, they
290 return the amount of uncopied data (ie. 0 290 return the amount of uncopied data (ie. 0 still means success).
291 291
292 [Yes, this objectionable interface makes me cr 292 [Yes, this objectionable interface makes me cringe. The flamewar comes
293 up every year or so. --RR.] 293 up every year or so. --RR.]
294 294
295 The functions may sleep implicitly. This shoul 295 The functions may sleep implicitly. This should never be called outside
296 user context (it makes no sense), with interru 296 user context (it makes no sense), with interrupts disabled, or a
297 spinlock held. 297 spinlock held.
298 298
299 :c:func:`kmalloc()`/:c:func:`kfree()` 299 :c:func:`kmalloc()`/:c:func:`kfree()`
300 ------------------------------------- 300 -------------------------------------
301 301
302 Defined in ``include/linux/slab.h`` 302 Defined in ``include/linux/slab.h``
303 303
304 **[MAY SLEEP: SEE BELOW]** 304 **[MAY SLEEP: SEE BELOW]**
305 305
306 These routines are used to dynamically request 306 These routines are used to dynamically request pointer-aligned chunks of
307 memory, like malloc and free do in userspace, 307 memory, like malloc and free do in userspace, but
308 :c:func:`kmalloc()` takes an extra flag word. 308 :c:func:`kmalloc()` takes an extra flag word. Important values:
309 309
310 ``GFP_KERNEL`` 310 ``GFP_KERNEL``
311 May sleep and swap to free memory. Only al 311 May sleep and swap to free memory. Only allowed in user context, but
312 is the most reliable way to allocate memor 312 is the most reliable way to allocate memory.
313 313
314 ``GFP_ATOMIC`` 314 ``GFP_ATOMIC``
315 Don't sleep. Less reliable than ``GFP_KERN 315 Don't sleep. Less reliable than ``GFP_KERNEL``, but may be called
316 from interrupt context. You should **reall 316 from interrupt context. You should **really** have a good
317 out-of-memory error-handling strategy. 317 out-of-memory error-handling strategy.
318 318
319 ``GFP_DMA`` 319 ``GFP_DMA``
320 Allocate ISA DMA lower than 16MB. If you d 320 Allocate ISA DMA lower than 16MB. If you don't know what that is you
321 don't need it. Very unreliable. 321 don't need it. Very unreliable.
322 322
323 If you see a sleeping function called from inv 323 If you see a sleeping function called from invalid context warning
324 message, then maybe you called a sleeping allo 324 message, then maybe you called a sleeping allocation function from
325 interrupt context without ``GFP_ATOMIC``. You 325 interrupt context without ``GFP_ATOMIC``. You should really fix that.
326 Run, don't walk. 326 Run, don't walk.
327 327
328 If you are allocating at least ``PAGE_SIZE`` ( 328 If you are allocating at least ``PAGE_SIZE`` (``asm/page.h`` or
329 ``asm/page_types.h``) bytes, consider using :c 329 ``asm/page_types.h``) bytes, consider using :c:func:`__get_free_pages()`
330 (``include/linux/gfp.h``). It takes an order a 330 (``include/linux/gfp.h``). It takes an order argument (0 for page sized,
331 1 for double page, 2 for four pages etc.) and 331 1 for double page, 2 for four pages etc.) and the same memory priority
332 flag word as above. 332 flag word as above.
333 333
334 If you are allocating more than a page worth o 334 If you are allocating more than a page worth of bytes you can use
335 :c:func:`vmalloc()`. It'll allocate virtual me 335 :c:func:`vmalloc()`. It'll allocate virtual memory in the kernel
336 map. This block is not contiguous in physical 336 map. This block is not contiguous in physical memory, but the MMU makes
337 it look like it is for you (so it'll only look 337 it look like it is for you (so it'll only look contiguous to the CPUs,
338 not to external device drivers). If you really 338 not to external device drivers). If you really need large physically
339 contiguous memory for some weird device, you h 339 contiguous memory for some weird device, you have a problem: it is
340 poorly supported in Linux because after some t 340 poorly supported in Linux because after some time memory fragmentation
341 in a running kernel makes it hard. The best wa 341 in a running kernel makes it hard. The best way is to allocate the block
342 early in the boot process via the :c:func:`all 342 early in the boot process via the :c:func:`alloc_bootmem()`
343 routine. 343 routine.
344 344
345 Before inventing your own cache of often-used 345 Before inventing your own cache of often-used objects consider using a
346 slab cache in ``include/linux/slab.h`` 346 slab cache in ``include/linux/slab.h``
347 347
348 :c:macro:`current` 348 :c:macro:`current`
349 ------------------ 349 ------------------
350 350
351 Defined in ``include/asm/current.h`` 351 Defined in ``include/asm/current.h``
352 352
353 This global variable (really a macro) contains 353 This global variable (really a macro) contains a pointer to the current
354 task structure, so is only valid in user conte 354 task structure, so is only valid in user context. For example, when a
355 process makes a system call, this will point t 355 process makes a system call, this will point to the task structure of
356 the calling process. It is **not NULL** in int 356 the calling process. It is **not NULL** in interrupt context.
357 357
358 :c:func:`mdelay()`/:c:func:`udelay()` 358 :c:func:`mdelay()`/:c:func:`udelay()`
359 ------------------------------------- 359 -------------------------------------
360 360
361 Defined in ``include/asm/delay.h`` / ``include 361 Defined in ``include/asm/delay.h`` / ``include/linux/delay.h``
362 362
363 The :c:func:`udelay()` and :c:func:`ndelay()` 363 The :c:func:`udelay()` and :c:func:`ndelay()` functions can be
364 used for small pauses. Do not use large values 364 used for small pauses. Do not use large values with them as you risk
365 overflow - the helper function :c:func:`mdelay 365 overflow - the helper function :c:func:`mdelay()` is useful here, or
366 consider :c:func:`msleep()`. 366 consider :c:func:`msleep()`.
367 367
368 :c:func:`cpu_to_be32()`/:c:func:`be32_to_cpu() 368 :c:func:`cpu_to_be32()`/:c:func:`be32_to_cpu()`/:c:func:`cpu_to_le32()`/:c:func:`le32_to_cpu()`
369 ---------------------------------------------- 369 -----------------------------------------------------------------------------------------------
370 370
371 Defined in ``include/asm/byteorder.h`` 371 Defined in ``include/asm/byteorder.h``
372 372
373 The :c:func:`cpu_to_be32()` family (where the 373 The :c:func:`cpu_to_be32()` family (where the "32" can be replaced
374 by 64 or 16, and the "be" can be replaced by " 374 by 64 or 16, and the "be" can be replaced by "le") are the general way
375 to do endian conversions in the kernel: they r 375 to do endian conversions in the kernel: they return the converted value.
376 All variations supply the reverse as well: 376 All variations supply the reverse as well:
377 :c:func:`be32_to_cpu()`, etc. 377 :c:func:`be32_to_cpu()`, etc.
378 378
379 There are two major variations of these functi 379 There are two major variations of these functions: the pointer
380 variation, such as :c:func:`cpu_to_be32p()`, w 380 variation, such as :c:func:`cpu_to_be32p()`, which take a pointer
381 to the given type, and return the converted va 381 to the given type, and return the converted value. The other variation
382 is the "in-situ" family, such as :c:func:`cpu_ 382 is the "in-situ" family, such as :c:func:`cpu_to_be32s()`, which
383 convert value referred to by the pointer, and 383 convert value referred to by the pointer, and return void.
384 384
385 :c:func:`local_irq_save()`/:c:func:`local_irq_ 385 :c:func:`local_irq_save()`/:c:func:`local_irq_restore()`
386 ---------------------------------------------- 386 --------------------------------------------------------
387 387
388 Defined in ``include/linux/irqflags.h`` 388 Defined in ``include/linux/irqflags.h``
389 389
390 These routines disable hard interrupts on the 390 These routines disable hard interrupts on the local CPU, and restore
391 them. They are reentrant; saving the previous 391 them. They are reentrant; saving the previous state in their one
392 ``unsigned long flags`` argument. If you know 392 ``unsigned long flags`` argument. If you know that interrupts are
393 enabled, you can simply use :c:func:`local_irq 393 enabled, you can simply use :c:func:`local_irq_disable()` and
394 :c:func:`local_irq_enable()`. 394 :c:func:`local_irq_enable()`.
395 395
396 .. _local_bh_disable: 396 .. _local_bh_disable:
397 397
398 :c:func:`local_bh_disable()`/:c:func:`local_bh 398 :c:func:`local_bh_disable()`/:c:func:`local_bh_enable()`
399 ---------------------------------------------- 399 --------------------------------------------------------
400 400
401 Defined in ``include/linux/bottom_half.h`` 401 Defined in ``include/linux/bottom_half.h``
402 402
403 403
404 These routines disable soft interrupts on the 404 These routines disable soft interrupts on the local CPU, and restore
405 them. They are reentrant; if soft interrupts w 405 them. They are reentrant; if soft interrupts were disabled before, they
406 will still be disabled after this pair of func 406 will still be disabled after this pair of functions has been called.
407 They prevent softirqs and tasklets from runnin 407 They prevent softirqs and tasklets from running on the current CPU.
408 408
409 :c:func:`smp_processor_id()` 409 :c:func:`smp_processor_id()`
410 ---------------------------- 410 ----------------------------
411 411
412 Defined in ``include/linux/smp.h`` 412 Defined in ``include/linux/smp.h``
413 413
414 :c:func:`get_cpu()` disables preemption (so yo 414 :c:func:`get_cpu()` disables preemption (so you won't suddenly get
415 moved to another CPU) and returns the current 415 moved to another CPU) and returns the current processor number, between
416 0 and ``NR_CPUS``. Note that the CPU numbers a 416 0 and ``NR_CPUS``. Note that the CPU numbers are not necessarily
417 continuous. You return it again with :c:func:` 417 continuous. You return it again with :c:func:`put_cpu()` when you
418 are done. 418 are done.
419 419
420 If you know you cannot be preempted by another 420 If you know you cannot be preempted by another task (ie. you are in
421 interrupt context, or have preemption disabled 421 interrupt context, or have preemption disabled) you can use
422 smp_processor_id(). 422 smp_processor_id().
423 423
424 ``__init``/``__exit``/``__initdata`` 424 ``__init``/``__exit``/``__initdata``
425 ------------------------------------ 425 ------------------------------------
426 426
427 Defined in ``include/linux/init.h`` 427 Defined in ``include/linux/init.h``
428 428
429 After boot, the kernel frees up a special sect 429 After boot, the kernel frees up a special section; functions marked with
430 ``__init`` and data structures marked with ``_ 430 ``__init`` and data structures marked with ``__initdata`` are dropped
431 after boot is complete: similarly modules disc 431 after boot is complete: similarly modules discard this memory after
432 initialization. ``__exit`` is used to declare 432 initialization. ``__exit`` is used to declare a function which is only
433 required on exit: the function will be dropped 433 required on exit: the function will be dropped if this file is not
434 compiled as a module. See the header file for 434 compiled as a module. See the header file for use. Note that it makes no
435 sense for a function marked with ``__init`` to 435 sense for a function marked with ``__init`` to be exported to modules
436 with :c:func:`EXPORT_SYMBOL()` or :c:func:`EXP 436 with :c:func:`EXPORT_SYMBOL()` or :c:func:`EXPORT_SYMBOL_GPL()`- this
437 will break. 437 will break.
438 438
439 :c:func:`__initcall()`/:c:func:`module_init()` 439 :c:func:`__initcall()`/:c:func:`module_init()`
440 ---------------------------------------------- 440 ----------------------------------------------
441 441
442 Defined in ``include/linux/init.h`` / ``inclu 442 Defined in ``include/linux/init.h`` / ``include/linux/module.h``
443 443
444 Many parts of the kernel are well served as a 444 Many parts of the kernel are well served as a module
445 (dynamically-loadable parts of the kernel). Us 445 (dynamically-loadable parts of the kernel). Using the
446 :c:func:`module_init()` and :c:func:`module_ex 446 :c:func:`module_init()` and :c:func:`module_exit()` macros it
447 is easy to write code without #ifdefs which ca 447 is easy to write code without #ifdefs which can operate both as a module
448 or built into the kernel. 448 or built into the kernel.
449 449
450 The :c:func:`module_init()` macro defines whic 450 The :c:func:`module_init()` macro defines which function is to be
451 called at module insertion time (if the file i 451 called at module insertion time (if the file is compiled as a module),
452 or at boot time: if the file is not compiled a 452 or at boot time: if the file is not compiled as a module the
453 :c:func:`module_init()` macro becomes equivale 453 :c:func:`module_init()` macro becomes equivalent to
454 :c:func:`__initcall()`, which through linker m 454 :c:func:`__initcall()`, which through linker magic ensures that
455 the function is called on boot. 455 the function is called on boot.
456 456
457 The function can return a negative error numbe 457 The function can return a negative error number to cause module loading
458 to fail (unfortunately, this has no effect if 458 to fail (unfortunately, this has no effect if the module is compiled
459 into the kernel). This function is called in u 459 into the kernel). This function is called in user context with
460 interrupts enabled, so it can sleep. 460 interrupts enabled, so it can sleep.
461 461
462 :c:func:`module_exit()` 462 :c:func:`module_exit()`
463 ----------------------- 463 -----------------------
464 464
465 465
466 Defined in ``include/linux/module.h`` 466 Defined in ``include/linux/module.h``
467 467
468 This macro defines the function to be called a 468 This macro defines the function to be called at module removal time (or
469 never, in the case of the file compiled into t 469 never, in the case of the file compiled into the kernel). It will only
470 be called if the module usage count has reache 470 be called if the module usage count has reached zero. This function can
471 also sleep, but cannot fail: everything must b 471 also sleep, but cannot fail: everything must be cleaned up by the time
472 it returns. 472 it returns.
473 473
474 Note that this macro is optional: if it is not 474 Note that this macro is optional: if it is not present, your module will
475 not be removable (except for 'rmmod -f'). 475 not be removable (except for 'rmmod -f').
476 476
477 :c:func:`try_module_get()`/:c:func:`module_put 477 :c:func:`try_module_get()`/:c:func:`module_put()`
478 ---------------------------------------------- 478 -------------------------------------------------
479 479
480 Defined in ``include/linux/module.h`` 480 Defined in ``include/linux/module.h``
481 481
482 These manipulate the module usage count, to pr 482 These manipulate the module usage count, to protect against removal (a
483 module also can't be removed if another module 483 module also can't be removed if another module uses one of its exported
484 symbols: see below). Before calling into modul 484 symbols: see below). Before calling into module code, you should call
485 :c:func:`try_module_get()` on that module: if 485 :c:func:`try_module_get()` on that module: if it fails, then the
486 module is being removed and you should act as 486 module is being removed and you should act as if it wasn't there.
487 Otherwise, you can safely enter the module, an 487 Otherwise, you can safely enter the module, and call
488 :c:func:`module_put()` when you're finished. 488 :c:func:`module_put()` when you're finished.
489 489
490 Most registerable structures have an owner fie 490 Most registerable structures have an owner field, such as in the
491 :c:type:`struct file_operations <file_operatio 491 :c:type:`struct file_operations <file_operations>` structure.
492 Set this field to the macro ``THIS_MODULE``. 492 Set this field to the macro ``THIS_MODULE``.
493 493
494 Wait Queues ``include/linux/wait.h`` 494 Wait Queues ``include/linux/wait.h``
495 ==================================== 495 ====================================
496 496
497 **[SLEEPS]** 497 **[SLEEPS]**
498 498
499 A wait queue is used to wait for someone to wa 499 A wait queue is used to wait for someone to wake you up when a certain
500 condition is true. They must be used carefully 500 condition is true. They must be used carefully to ensure there is no
501 race condition. You declare a :c:type:`wait_qu 501 race condition. You declare a :c:type:`wait_queue_head_t`, and then processes
502 which want to wait for that condition declare 502 which want to wait for that condition declare a :c:type:`wait_queue_entry_t`
503 referring to themselves, and place that in the 503 referring to themselves, and place that in the queue.
504 504
505 Declaring 505 Declaring
506 --------- 506 ---------
507 507
508 You declare a ``wait_queue_head_t`` using the 508 You declare a ``wait_queue_head_t`` using the
509 :c:func:`DECLARE_WAIT_QUEUE_HEAD()` macro, or 509 :c:func:`DECLARE_WAIT_QUEUE_HEAD()` macro, or using the
510 :c:func:`init_waitqueue_head()` routine in you 510 :c:func:`init_waitqueue_head()` routine in your initialization
511 code. 511 code.
512 512
513 Queuing 513 Queuing
514 ------- 514 -------
515 515
516 Placing yourself in the waitqueue is fairly co 516 Placing yourself in the waitqueue is fairly complex, because you must
517 put yourself in the queue before checking the 517 put yourself in the queue before checking the condition. There is a
518 macro to do this: :c:func:`wait_event_interrup 518 macro to do this: :c:func:`wait_event_interruptible()`
519 (``include/linux/wait.h``) The first argument 519 (``include/linux/wait.h``) The first argument is the wait queue head, and
520 the second is an expression which is evaluated 520 the second is an expression which is evaluated; the macro returns 0 when
521 this expression is true, or ``-ERESTARTSYS`` i 521 this expression is true, or ``-ERESTARTSYS`` if a signal is received. The
522 :c:func:`wait_event()` version ignores signals 522 :c:func:`wait_event()` version ignores signals.
523 523
524 Waking Up Queued Tasks 524 Waking Up Queued Tasks
525 ---------------------- 525 ----------------------
526 526
527 Call :c:func:`wake_up()` (``include/linux/wait 527 Call :c:func:`wake_up()` (``include/linux/wait.h``), which will wake
528 up every process in the queue. The exception i 528 up every process in the queue. The exception is if one has
529 ``TASK_EXCLUSIVE`` set, in which case the rema 529 ``TASK_EXCLUSIVE`` set, in which case the remainder of the queue will
530 not be woken. There are other variants of this 530 not be woken. There are other variants of this basic function available
531 in the same header. 531 in the same header.
532 532
533 Atomic Operations 533 Atomic Operations
534 ================= 534 =================
535 535
536 Certain operations are guaranteed atomic on al 536 Certain operations are guaranteed atomic on all platforms. The first
537 class of operations work on :c:type:`atomic_t` 537 class of operations work on :c:type:`atomic_t` (``include/asm/atomic.h``);
538 this contains a signed integer (at least 32 bi 538 this contains a signed integer (at least 32 bits long), and you must use
539 these functions to manipulate or read :c:type: 539 these functions to manipulate or read :c:type:`atomic_t` variables.
540 :c:func:`atomic_read()` and :c:func:`atomic_se 540 :c:func:`atomic_read()` and :c:func:`atomic_set()` get and set
541 the counter, :c:func:`atomic_add()`, :c:func:` 541 the counter, :c:func:`atomic_add()`, :c:func:`atomic_sub()`,
542 :c:func:`atomic_inc()`, :c:func:`atomic_dec()` 542 :c:func:`atomic_inc()`, :c:func:`atomic_dec()`, and
543 :c:func:`atomic_dec_and_test()` (returns true 543 :c:func:`atomic_dec_and_test()` (returns true if it was
544 decremented to zero). 544 decremented to zero).
545 545
546 Yes. It returns true (i.e. != 0) if the atomic 546 Yes. It returns true (i.e. != 0) if the atomic variable is zero.
547 547
548 Note that these functions are slower than norm 548 Note that these functions are slower than normal arithmetic, and so
549 should not be used unnecessarily. 549 should not be used unnecessarily.
550 550
551 The second class of atomic operations is atomi 551 The second class of atomic operations is atomic bit operations on an
552 ``unsigned long``, defined in ``include/linux/ 552 ``unsigned long``, defined in ``include/linux/bitops.h``. These
553 operations generally take a pointer to the bit 553 operations generally take a pointer to the bit pattern, and a bit
554 number: 0 is the least significant bit. :c:fun 554 number: 0 is the least significant bit. :c:func:`set_bit()`,
555 :c:func:`clear_bit()` and :c:func:`change_bit( 555 :c:func:`clear_bit()` and :c:func:`change_bit()` set, clear,
556 and flip the given bit. :c:func:`test_and_set_ 556 and flip the given bit. :c:func:`test_and_set_bit()`,
557 :c:func:`test_and_clear_bit()` and 557 :c:func:`test_and_clear_bit()` and
558 :c:func:`test_and_change_bit()` do the same th 558 :c:func:`test_and_change_bit()` do the same thing, except return
559 true if the bit was previously set; these are 559 true if the bit was previously set; these are particularly useful for
560 atomically setting flags. 560 atomically setting flags.
561 561
562 It is possible to call these operations with b 562 It is possible to call these operations with bit indices greater than
563 ``BITS_PER_LONG``. The resulting behavior is s 563 ``BITS_PER_LONG``. The resulting behavior is strange on big-endian
564 platforms though so it is a good idea not to d 564 platforms though so it is a good idea not to do this.
565 565
566 Symbols 566 Symbols
567 ======= 567 =======
568 568
569 Within the kernel proper, the normal linking r 569 Within the kernel proper, the normal linking rules apply (ie. unless a
570 symbol is declared to be file scope with the ` 570 symbol is declared to be file scope with the ``static`` keyword, it can
571 be used anywhere in the kernel). However, for 571 be used anywhere in the kernel). However, for modules, a special
572 exported symbol table is kept which limits the 572 exported symbol table is kept which limits the entry points to the
573 kernel proper. Modules can also export symbols 573 kernel proper. Modules can also export symbols.
574 574
575 :c:func:`EXPORT_SYMBOL()` 575 :c:func:`EXPORT_SYMBOL()`
576 ------------------------- 576 -------------------------
577 577
578 Defined in ``include/linux/export.h`` 578 Defined in ``include/linux/export.h``
579 579
580 This is the classic method of exporting a symb 580 This is the classic method of exporting a symbol: dynamically loaded
581 modules will be able to use the symbol as norm 581 modules will be able to use the symbol as normal.
582 582
583 :c:func:`EXPORT_SYMBOL_GPL()` 583 :c:func:`EXPORT_SYMBOL_GPL()`
584 ----------------------------- 584 -----------------------------
585 585
586 Defined in ``include/linux/export.h`` 586 Defined in ``include/linux/export.h``
587 587
588 Similar to :c:func:`EXPORT_SYMBOL()` except th 588 Similar to :c:func:`EXPORT_SYMBOL()` except that the symbols
589 exported by :c:func:`EXPORT_SYMBOL_GPL()` can 589 exported by :c:func:`EXPORT_SYMBOL_GPL()` can only be seen by
590 modules with a :c:func:`MODULE_LICENSE()` that 590 modules with a :c:func:`MODULE_LICENSE()` that specifies a GPL
591 compatible license. It implies that the functi 591 compatible license. It implies that the function is considered an
592 internal implementation issue, and not really 592 internal implementation issue, and not really an interface. Some
593 maintainers and developers may however require 593 maintainers and developers may however require EXPORT_SYMBOL_GPL()
594 when adding any new APIs or functionality. 594 when adding any new APIs or functionality.
595 595
596 :c:func:`EXPORT_SYMBOL_NS()` 596 :c:func:`EXPORT_SYMBOL_NS()`
597 ---------------------------- 597 ----------------------------
598 598
599 Defined in ``include/linux/export.h`` 599 Defined in ``include/linux/export.h``
600 600
601 This is the variant of `EXPORT_SYMBOL()` that 601 This is the variant of `EXPORT_SYMBOL()` that allows specifying a symbol
602 namespace. Symbol Namespaces are documented in 602 namespace. Symbol Namespaces are documented in
603 Documentation/core-api/symbol-namespaces.rst 603 Documentation/core-api/symbol-namespaces.rst
604 604
605 :c:func:`EXPORT_SYMBOL_NS_GPL()` 605 :c:func:`EXPORT_SYMBOL_NS_GPL()`
606 -------------------------------- 606 --------------------------------
607 607
608 Defined in ``include/linux/export.h`` 608 Defined in ``include/linux/export.h``
609 609
610 This is the variant of `EXPORT_SYMBOL_GPL()` t 610 This is the variant of `EXPORT_SYMBOL_GPL()` that allows specifying a symbol
611 namespace. Symbol Namespaces are documented in 611 namespace. Symbol Namespaces are documented in
612 Documentation/core-api/symbol-namespaces.rst 612 Documentation/core-api/symbol-namespaces.rst
613 613
614 Routines and Conventions 614 Routines and Conventions
615 ======================== 615 ========================
616 616
617 Double-linked lists ``include/linux/list.h`` 617 Double-linked lists ``include/linux/list.h``
618 -------------------------------------------- 618 --------------------------------------------
619 619
620 There used to be three sets of linked-list rou 620 There used to be three sets of linked-list routines in the kernel
621 headers, but this one is the winner. If you do 621 headers, but this one is the winner. If you don't have some particular
622 pressing need for a single list, it's a good c 622 pressing need for a single list, it's a good choice.
623 623
624 In particular, :c:func:`list_for_each_entry()` 624 In particular, :c:func:`list_for_each_entry()` is useful.
625 625
626 Return Conventions 626 Return Conventions
627 ------------------ 627 ------------------
628 628
629 For code called in user context, it's very com 629 For code called in user context, it's very common to defy C convention,
630 and return 0 for success, and a negative error 630 and return 0 for success, and a negative error number (eg. ``-EFAULT``) for
631 failure. This can be unintuitive at first, but 631 failure. This can be unintuitive at first, but it's fairly widespread in
632 the kernel. 632 the kernel.
633 633
634 Using :c:func:`ERR_PTR()` (``include/linux/err 634 Using :c:func:`ERR_PTR()` (``include/linux/err.h``) to encode a
635 negative error number into a pointer, and :c:f 635 negative error number into a pointer, and :c:func:`IS_ERR()` and
636 :c:func:`PTR_ERR()` to get it back out again: 636 :c:func:`PTR_ERR()` to get it back out again: avoids a separate
637 pointer parameter for the error number. Icky, 637 pointer parameter for the error number. Icky, but in a good way.
638 638
639 Breaking Compilation 639 Breaking Compilation
640 -------------------- 640 --------------------
641 641
642 Linus and the other developers sometimes chang 642 Linus and the other developers sometimes change function or structure
643 names in development kernels; this is not done 643 names in development kernels; this is not done just to keep everyone on
644 their toes: it reflects a fundamental change ( 644 their toes: it reflects a fundamental change (eg. can no longer be
645 called with interrupts on, or does extra check 645 called with interrupts on, or does extra checks, or doesn't do checks
646 which were caught before). Usually this is acc 646 which were caught before). Usually this is accompanied by a fairly
647 complete note to the appropriate kernel develo 647 complete note to the appropriate kernel development mailing list; search
648 the archives. Simply doing a global replace on 648 the archives. Simply doing a global replace on the file usually makes
649 things **worse**. 649 things **worse**.
650 650
651 Initializing structure members 651 Initializing structure members
652 ------------------------------ 652 ------------------------------
653 653
654 The preferred method of initializing structure 654 The preferred method of initializing structures is to use designated
655 initialisers, as defined by ISO C99, eg:: 655 initialisers, as defined by ISO C99, eg::
656 656
657 static struct block_device_operations opt_ 657 static struct block_device_operations opt_fops = {
658 .open = opt_open, 658 .open = opt_open,
659 .release = opt_release, 659 .release = opt_release,
660 .ioctl = opt_ioctl, 660 .ioctl = opt_ioctl,
661 .check_media_change = opt_media_ch 661 .check_media_change = opt_media_change,
662 }; 662 };
663 663
664 664
665 This makes it easy to grep for, and makes it c 665 This makes it easy to grep for, and makes it clear which structure
666 fields are set. You should do this because it 666 fields are set. You should do this because it looks cool.
667 667
668 GNU Extensions 668 GNU Extensions
669 -------------- 669 --------------
670 670
671 GNU Extensions are explicitly allowed in the L 671 GNU Extensions are explicitly allowed in the Linux kernel. Note that
672 some of the more complex ones are not very wel 672 some of the more complex ones are not very well supported, due to lack
673 of general use, but the following are consider 673 of general use, but the following are considered standard (see the GCC
674 info page section "C Extensions" for more deta 674 info page section "C Extensions" for more details - Yes, really the info
675 page, the man page is only a short summary of 675 page, the man page is only a short summary of the stuff in info).
676 676
677 - Inline functions 677 - Inline functions
678 678
679 - Statement expressions (ie. the ({ and }) co 679 - Statement expressions (ie. the ({ and }) constructs).
680 680
681 - Declaring attributes of a function / variab 681 - Declaring attributes of a function / variable / type
682 (__attribute__) 682 (__attribute__)
683 683
684 - typeof 684 - typeof
685 685
686 - Zero length arrays 686 - Zero length arrays
687 687
688 - Macro varargs 688 - Macro varargs
689 689
690 - Arithmetic on void pointers 690 - Arithmetic on void pointers
691 691
692 - Non-Constant initializers 692 - Non-Constant initializers
693 693
694 - Assembler Instructions (not outside arch/ a 694 - Assembler Instructions (not outside arch/ and include/asm/)
695 695
696 - Function names as strings (__func__). 696 - Function names as strings (__func__).
697 697
698 - __builtin_constant_p() 698 - __builtin_constant_p()
699 699
700 Be wary when using long long in the kernel, th 700 Be wary when using long long in the kernel, the code gcc generates for
701 it is horrible and worse: division and multipl 701 it is horrible and worse: division and multiplication does not work on
702 i386 because the GCC runtime functions for it 702 i386 because the GCC runtime functions for it are missing from the
703 kernel environment. 703 kernel environment.
704 704
705 C++ 705 C++
706 --- 706 ---
707 707
708 Using C++ in the kernel is usually a bad idea, 708 Using C++ in the kernel is usually a bad idea, because the kernel does
709 not provide the necessary runtime environment 709 not provide the necessary runtime environment and the include files are
710 not tested for it. It is still possible, but n 710 not tested for it. It is still possible, but not recommended. If you
711 really want to do this, forget about exception 711 really want to do this, forget about exceptions at least.
712 712
713 #if 713 #if
714 --- 714 ---
715 715
716 It is generally considered cleaner to use macr 716 It is generally considered cleaner to use macros in header files (or at
717 the top of .c files) to abstract away function 717 the top of .c files) to abstract away functions rather than using \`#if'
718 pre-processor statements throughout the source 718 pre-processor statements throughout the source code.
719 719
720 Putting Your Stuff in the Kernel 720 Putting Your Stuff in the Kernel
721 ================================ 721 ================================
722 722
723 In order to get your stuff into shape for offi 723 In order to get your stuff into shape for official inclusion, or even to
724 make a neat patch, there's administrative work 724 make a neat patch, there's administrative work to be done:
725 725
726 - Figure out who are the owners of the code y 726 - Figure out who are the owners of the code you've been modifying. Look
727 at the top of the source files, inside the 727 at the top of the source files, inside the ``MAINTAINERS`` file, and
728 last of all in the ``CREDITS`` file. You sh 728 last of all in the ``CREDITS`` file. You should coordinate with these
729 people to make sure you're not duplicating 729 people to make sure you're not duplicating effort, or trying something
730 that's already been rejected. 730 that's already been rejected.
731 731
732 Make sure you put your name and email addre 732 Make sure you put your name and email address at the top of any files
733 you create or modify significantly. This is 733 you create or modify significantly. This is the first place people
734 will look when they find a bug, or when **t 734 will look when they find a bug, or when **they** want to make a change.
735 735
736 - Usually you want a configuration option for 736 - Usually you want a configuration option for your kernel hack. Edit
737 ``Kconfig`` in the appropriate directory. T 737 ``Kconfig`` in the appropriate directory. The Config language is
738 simple to use by cut and paste, and there's 738 simple to use by cut and paste, and there's complete documentation in
739 ``Documentation/kbuild/kconfig-language.rst 739 ``Documentation/kbuild/kconfig-language.rst``.
740 740
741 In your description of the option, make sur 741 In your description of the option, make sure you address both the
742 expert user and the user who knows nothing 742 expert user and the user who knows nothing about your feature.
743 Mention incompatibilities and issues here. 743 Mention incompatibilities and issues here. **Definitely** end your
744 description with “if in doubt, say N” ( 744 description with “if in doubt, say N” (or, occasionally, \`Y'); this
745 is for people who have no idea what you are 745 is for people who have no idea what you are talking about.
746 746
747 - Edit the ``Makefile``: the CONFIG variables 747 - Edit the ``Makefile``: the CONFIG variables are exported here so you
748 can usually just add a "obj-$(CONFIG_xxx) + 748 can usually just add a "obj-$(CONFIG_xxx) += xxx.o" line. The syntax
749 is documented in ``Documentation/kbuild/mak 749 is documented in ``Documentation/kbuild/makefiles.rst``.
750 750
751 - Put yourself in ``CREDITS`` if you consider 751 - Put yourself in ``CREDITS`` if you consider what you've done
752 noteworthy, usually beyond a single file (y 752 noteworthy, usually beyond a single file (your name should be at the
753 top of the source files anyway). ``MAINTAIN 753 top of the source files anyway). ``MAINTAINERS`` means you want to be
754 consulted when changes are made to a subsys 754 consulted when changes are made to a subsystem, and hear about bugs;
755 it implies a more-than-passing commitment t 755 it implies a more-than-passing commitment to some part of the code.
756 756
757 - Finally, don't forget to read 757 - Finally, don't forget to read
758 ``Documentation/process/submitting-patches. 758 ``Documentation/process/submitting-patches.rst``
759 759
760 Kernel Cantrips 760 Kernel Cantrips
761 =============== 761 ===============
762 762
763 Some favorites from browsing the source. Feel 763 Some favorites from browsing the source. Feel free to add to this list.
764 764
765 ``arch/x86/include/asm/delay.h``:: 765 ``arch/x86/include/asm/delay.h``::
766 766
767 #define ndelay(n) (__builtin_constant_p(n) 767 #define ndelay(n) (__builtin_constant_p(n) ? \
768 ((n) > 20000 ? __bad_ndelay() : __ 768 ((n) > 20000 ? __bad_ndelay() : __const_udelay((n) * 5ul)) : \
769 __ndelay(n)) 769 __ndelay(n))
770 770
771 771
772 ``include/linux/fs.h``:: 772 ``include/linux/fs.h``::
773 773
774 /* 774 /*
775 * Kernel pointers have redundant informat 775 * Kernel pointers have redundant information, so we can use a
776 * scheme where we can return either an er 776 * scheme where we can return either an error code or a dentry
777 * pointer with the same return value. 777 * pointer with the same return value.
778 * 778 *
779 * This should be a per-architecture thing 779 * This should be a per-architecture thing, to allow different
780 * error and pointer decisions. 780 * error and pointer decisions.
781 */ 781 */
782 #define ERR_PTR(err) ((void *)((long)( 782 #define ERR_PTR(err) ((void *)((long)(err)))
783 #define PTR_ERR(ptr) ((long)(ptr)) 783 #define PTR_ERR(ptr) ((long)(ptr))
784 #define IS_ERR(ptr) ((unsigned long)( 784 #define IS_ERR(ptr) ((unsigned long)(ptr) > (unsigned long)(-1000))
785 785
786 ``arch/x86/include/asm/uaccess_32.h:``:: 786 ``arch/x86/include/asm/uaccess_32.h:``::
787 787
788 #define copy_to_user(to,from,n) 788 #define copy_to_user(to,from,n) \
789 (__builtin_constant_p(n) ? 789 (__builtin_constant_p(n) ? \
790 __constant_copy_to_user((to),(fro 790 __constant_copy_to_user((to),(from),(n)) : \
791 __generic_copy_to_user((to),(from 791 __generic_copy_to_user((to),(from),(n)))
792 792
793 793
794 ``arch/sparc/kernel/head.S:``:: 794 ``arch/sparc/kernel/head.S:``::
795 795
796 /* 796 /*
797 * Sun people can't spell worth damn. "com 797 * Sun people can't spell worth damn. "compatability" indeed.
798 * At least we *know* we can't spell, and 798 * At least we *know* we can't spell, and use a spell-checker.
799 */ 799 */
800 800
801 /* Uh, actually Linus it is I who cannot s 801 /* Uh, actually Linus it is I who cannot spell. Too much murky
802 * Sparc assembly will do this to ya. 802 * Sparc assembly will do this to ya.
803 */ 803 */
804 C_LABEL(cputypvar): 804 C_LABEL(cputypvar):
805 .asciz "compatibility" 805 .asciz "compatibility"
806 806
807 /* Tested on SS-5, SS-10. Probably someone 807 /* Tested on SS-5, SS-10. Probably someone at Sun applied a spell-checker. */
808 .align 4 808 .align 4
809 C_LABEL(cputypvar_sun4m): 809 C_LABEL(cputypvar_sun4m):
810 .asciz "compatible" 810 .asciz "compatible"
811 811
812 812
813 ``arch/sparc/lib/checksum.S:``:: 813 ``arch/sparc/lib/checksum.S:``::
814 814
815 /* Sun, you just can't beat me, yo 815 /* Sun, you just can't beat me, you just can't. Stop trying,
816 * give up. I'm serious, I am goi 816 * give up. I'm serious, I am going to kick the living shit
817 * out of you, game over, lights o 817 * out of you, game over, lights out.
818 */ 818 */
819 819
820 820
821 Thanks 821 Thanks
822 ====== 822 ======
823 823
824 Thanks to Andi Kleen for the idea, answering m 824 Thanks to Andi Kleen for the idea, answering my questions, fixing my
825 mistakes, filling content, etc. Philipp Rumpf 825 mistakes, filling content, etc. Philipp Rumpf for more spelling and
826 clarity fixes, and some excellent non-obvious 826 clarity fixes, and some excellent non-obvious points. Werner Almesberger
827 for giving me a great summary of :c:func:`disa 827 for giving me a great summary of :c:func:`disable_irq()`, and Jes
828 Sorensen and Andrea Arcangeli added caveats. M 828 Sorensen and Andrea Arcangeli added caveats. Michael Elizabeth Chastain
829 for checking and adding to the Configure secti 829 for checking and adding to the Configure section. Telsa Gwynne for
830 teaching me DocBook. 830 teaching me DocBook.
Linux® is a registered trademark of Linus Torvalds in the United States and other countries.
TOMOYO® is a registered trademark of NTT DATA CORPORATION.