xref: /linux/Documentation/admin-guide/README.rst (revision e58e871becec2d3b04ed91c0c16fe8deac9c9dfa)
1Linux kernel release 4.x <http://kernel.org/>
2=============================================
3
4These are the release notes for Linux version 4.  Read them carefully,
5as they tell you what this is all about, explain how to install the
6kernel, and what to do if something goes wrong.
7
8What is Linux?
9--------------
10
11  Linux is a clone of the operating system Unix, written from scratch by
12  Linus Torvalds with assistance from a loosely-knit team of hackers across
13  the Net. It aims towards POSIX and Single UNIX Specification compliance.
14
15  It has all the features you would expect in a modern fully-fledged Unix,
16  including true multitasking, virtual memory, shared libraries, demand
17  loading, shared copy-on-write executables, proper memory management,
18  and multistack networking including IPv4 and IPv6.
19
20  It is distributed under the GNU General Public License v2 - see the
21  accompanying COPYING file for more details.
22
23On what hardware does it run?
24-----------------------------
25
26  Although originally developed first for 32-bit x86-based PCs (386 or higher),
27  today Linux also runs on (at least) the Compaq Alpha AXP, Sun SPARC and
28  UltraSPARC, Motorola 68000, PowerPC, PowerPC64, ARM, Hitachi SuperH, Cell,
29  IBM S/390, MIPS, HP PA-RISC, Intel IA-64, DEC VAX, AMD x86-64, AXIS CRIS,
30  Xtensa, Tilera TILE, ARC and Renesas M32R architectures.
31
32  Linux is easily portable to most general-purpose 32- or 64-bit architectures
33  as long as they have a paged memory management unit (PMMU) and a port of the
34  GNU C compiler (gcc) (part of The GNU Compiler Collection, GCC). Linux has
35  also been ported to a number of architectures without a PMMU, although
36  functionality is then obviously somewhat limited.
37  Linux has also been ported to itself. You can now run the kernel as a
38  userspace application - this is called UserMode Linux (UML).
39
40Documentation
41-------------
42
43 - There is a lot of documentation available both in electronic form on
44   the Internet and in books, both Linux-specific and pertaining to
45   general UNIX questions.  I'd recommend looking into the documentation
46   subdirectories on any Linux FTP site for the LDP (Linux Documentation
47   Project) books.  This README is not meant to be documentation on the
48   system: there are much better sources available.
49
50 - There are various README files in the Documentation/ subdirectory:
51   these typically contain kernel-specific installation notes for some
52   drivers for example. See Documentation/00-INDEX for a list of what
53   is contained in each file.  Please read the
54   :ref:`Documentation/process/changes.rst <changes>` file, as it
55   contains information about the problems, which may result by upgrading
56   your kernel.
57
58 - The Documentation/DocBook/ subdirectory contains several guides for
59   kernel developers and users.  These guides can be rendered in a
60   number of formats:  PostScript (.ps), PDF, HTML, & man-pages, among others.
61   After installation, ``make psdocs``, ``make pdfdocs``, ``make htmldocs``,
62   or ``make mandocs`` will render the documentation in the requested format.
63
64Installing the kernel source
65----------------------------
66
67 - If you install the full sources, put the kernel tarball in a
68   directory where you have permissions (e.g. your home directory) and
69   unpack it::
70
71     xz -cd linux-4.X.tar.xz | tar xvf -
72
73   Replace "X" with the version number of the latest kernel.
74
75   Do NOT use the /usr/src/linux area! This area has a (usually
76   incomplete) set of kernel headers that are used by the library header
77   files.  They should match the library, and not get messed up by
78   whatever the kernel-du-jour happens to be.
79
80 - You can also upgrade between 4.x releases by patching.  Patches are
81   distributed in the xz format.  To install by patching, get all the
82   newer patch files, enter the top level directory of the kernel source
83   (linux-4.X) and execute::
84
85     xz -cd ../patch-4.x.xz | patch -p1
86
87   Replace "x" for all versions bigger than the version "X" of your current
88   source tree, **in_order**, and you should be ok.  You may want to remove
89   the backup files (some-file-name~ or some-file-name.orig), and make sure
90   that there are no failed patches (some-file-name# or some-file-name.rej).
91   If there are, either you or I have made a mistake.
92
93   Unlike patches for the 4.x kernels, patches for the 4.x.y kernels
94   (also known as the -stable kernels) are not incremental but instead apply
95   directly to the base 4.x kernel.  For example, if your base kernel is 4.0
96   and you want to apply the 4.0.3 patch, you must not first apply the 4.0.1
97   and 4.0.2 patches. Similarly, if you are running kernel version 4.0.2 and
98   want to jump to 4.0.3, you must first reverse the 4.0.2 patch (that is,
99   patch -R) **before** applying the 4.0.3 patch. You can read more on this in
100   :ref:`Documentation/process/applying-patches.rst <applying_patches>`.
101
102   Alternatively, the script patch-kernel can be used to automate this
103   process.  It determines the current kernel version and applies any
104   patches found::
105
106     linux/scripts/patch-kernel linux
107
108   The first argument in the command above is the location of the
109   kernel source.  Patches are applied from the current directory, but
110   an alternative directory can be specified as the second argument.
111
112 - Make sure you have no stale .o files and dependencies lying around::
113
114     cd linux
115     make mrproper
116
117   You should now have the sources correctly installed.
118
119Software requirements
120---------------------
121
122   Compiling and running the 4.x kernels requires up-to-date
123   versions of various software packages.  Consult
124   :ref:`Documentation/process/changes.rst <changes>` for the minimum version numbers
125   required and how to get updates for these packages.  Beware that using
126   excessively old versions of these packages can cause indirect
127   errors that are very difficult to track down, so don't assume that
128   you can just update packages when obvious problems arise during
129   build or operation.
130
131Build directory for the kernel
132------------------------------
133
134   When compiling the kernel, all output files will per default be
135   stored together with the kernel source code.
136   Using the option ``make O=output/dir`` allows you to specify an alternate
137   place for the output files (including .config).
138   Example::
139
140     kernel source code: /usr/src/linux-4.X
141     build directory:    /home/name/build/kernel
142
143   To configure and build the kernel, use::
144
145     cd /usr/src/linux-4.X
146     make O=/home/name/build/kernel menuconfig
147     make O=/home/name/build/kernel
148     sudo make O=/home/name/build/kernel modules_install install
149
150   Please note: If the ``O=output/dir`` option is used, then it must be
151   used for all invocations of make.
152
153Configuring the kernel
154----------------------
155
156   Do not skip this step even if you are only upgrading one minor
157   version.  New configuration options are added in each release, and
158   odd problems will turn up if the configuration files are not set up
159   as expected.  If you want to carry your existing configuration to a
160   new version with minimal work, use ``make oldconfig``, which will
161   only ask you for the answers to new questions.
162
163 - Alternative configuration commands are::
164
165     "make config"      Plain text interface.
166
167     "make menuconfig"  Text based color menus, radiolists & dialogs.
168
169     "make nconfig"     Enhanced text based color menus.
170
171     "make xconfig"     Qt based configuration tool.
172
173     "make gconfig"     GTK+ based configuration tool.
174
175     "make oldconfig"   Default all questions based on the contents of
176                        your existing ./.config file and asking about
177                        new config symbols.
178
179     "make silentoldconfig"
180                        Like above, but avoids cluttering the screen
181                        with questions already answered.
182                        Additionally updates the dependencies.
183
184     "make olddefconfig"
185                        Like above, but sets new symbols to their default
186                        values without prompting.
187
188     "make defconfig"   Create a ./.config file by using the default
189                        symbol values from either arch/$ARCH/defconfig
190                        or arch/$ARCH/configs/${PLATFORM}_defconfig,
191                        depending on the architecture.
192
193     "make ${PLATFORM}_defconfig"
194                        Create a ./.config file by using the default
195                        symbol values from
196                        arch/$ARCH/configs/${PLATFORM}_defconfig.
197                        Use "make help" to get a list of all available
198                        platforms of your architecture.
199
200     "make allyesconfig"
201                        Create a ./.config file by setting symbol
202                        values to 'y' as much as possible.
203
204     "make allmodconfig"
205                        Create a ./.config file by setting symbol
206                        values to 'm' as much as possible.
207
208     "make allnoconfig" Create a ./.config file by setting symbol
209                        values to 'n' as much as possible.
210
211     "make randconfig"  Create a ./.config file by setting symbol
212                        values to random values.
213
214     "make localmodconfig" Create a config based on current config and
215                           loaded modules (lsmod). Disables any module
216                           option that is not needed for the loaded modules.
217
218                           To create a localmodconfig for another machine,
219                           store the lsmod of that machine into a file
220                           and pass it in as a LSMOD parameter.
221
222                   target$ lsmod > /tmp/mylsmod
223                   target$ scp /tmp/mylsmod host:/tmp
224
225                   host$ make LSMOD=/tmp/mylsmod localmodconfig
226
227                           The above also works when cross compiling.
228
229     "make localyesconfig" Similar to localmodconfig, except it will convert
230                           all module options to built in (=y) options.
231
232   You can find more information on using the Linux kernel config tools
233   in Documentation/kbuild/kconfig.txt.
234
235 - NOTES on ``make config``:
236
237    - Having unnecessary drivers will make the kernel bigger, and can
238      under some circumstances lead to problems: probing for a
239      nonexistent controller card may confuse your other controllers.
240
241    - A kernel with math-emulation compiled in will still use the
242      coprocessor if one is present: the math emulation will just
243      never get used in that case.  The kernel will be slightly larger,
244      but will work on different machines regardless of whether they
245      have a math coprocessor or not.
246
247    - The "kernel hacking" configuration details usually result in a
248      bigger or slower kernel (or both), and can even make the kernel
249      less stable by configuring some routines to actively try to
250      break bad code to find kernel problems (kmalloc()).  Thus you
251      should probably answer 'n' to the questions for "development",
252      "experimental", or "debugging" features.
253
254Compiling the kernel
255--------------------
256
257 - Make sure you have at least gcc 3.2 available.
258   For more information, refer to :ref:`Documentation/process/changes.rst <changes>`.
259
260   Please note that you can still run a.out user programs with this kernel.
261
262 - Do a ``make`` to create a compressed kernel image. It is also
263   possible to do ``make install`` if you have lilo installed to suit the
264   kernel makefiles, but you may want to check your particular lilo setup first.
265
266   To do the actual install, you have to be root, but none of the normal
267   build should require that. Don't take the name of root in vain.
268
269 - If you configured any of the parts of the kernel as ``modules``, you
270   will also have to do ``make modules_install``.
271
272 - Verbose kernel compile/build output:
273
274   Normally, the kernel build system runs in a fairly quiet mode (but not
275   totally silent).  However, sometimes you or other kernel developers need
276   to see compile, link, or other commands exactly as they are executed.
277   For this, use "verbose" build mode.  This is done by passing
278   ``V=1`` to the ``make`` command, e.g.::
279
280     make V=1 all
281
282   To have the build system also tell the reason for the rebuild of each
283   target, use ``V=2``.  The default is ``V=0``.
284
285 - Keep a backup kernel handy in case something goes wrong.  This is
286   especially true for the development releases, since each new release
287   contains new code which has not been debugged.  Make sure you keep a
288   backup of the modules corresponding to that kernel, as well.  If you
289   are installing a new kernel with the same version number as your
290   working kernel, make a backup of your modules directory before you
291   do a ``make modules_install``.
292
293   Alternatively, before compiling, use the kernel config option
294   "LOCALVERSION" to append a unique suffix to the regular kernel version.
295   LOCALVERSION can be set in the "General Setup" menu.
296
297 - In order to boot your new kernel, you'll need to copy the kernel
298   image (e.g. .../linux/arch/x86/boot/bzImage after compilation)
299   to the place where your regular bootable kernel is found.
300
301 - Booting a kernel directly from a floppy without the assistance of a
302   bootloader such as LILO, is no longer supported.
303
304   If you boot Linux from the hard drive, chances are you use LILO, which
305   uses the kernel image as specified in the file /etc/lilo.conf.  The
306   kernel image file is usually /vmlinuz, /boot/vmlinuz, /bzImage or
307   /boot/bzImage.  To use the new kernel, save a copy of the old image
308   and copy the new image over the old one.  Then, you MUST RERUN LILO
309   to update the loading map! If you don't, you won't be able to boot
310   the new kernel image.
311
312   Reinstalling LILO is usually a matter of running /sbin/lilo.
313   You may wish to edit /etc/lilo.conf to specify an entry for your
314   old kernel image (say, /vmlinux.old) in case the new one does not
315   work.  See the LILO docs for more information.
316
317   After reinstalling LILO, you should be all set.  Shutdown the system,
318   reboot, and enjoy!
319
320   If you ever need to change the default root device, video mode,
321   ramdisk size, etc.  in the kernel image, use the ``rdev`` program (or
322   alternatively the LILO boot options when appropriate).  No need to
323   recompile the kernel to change these parameters.
324
325 - Reboot with the new kernel and enjoy.
326
327If something goes wrong
328-----------------------
329
330 - If you have problems that seem to be due to kernel bugs, please check
331   the file MAINTAINERS to see if there is a particular person associated
332   with the part of the kernel that you are having trouble with. If there
333   isn't anyone listed there, then the second best thing is to mail
334   them to me (torvalds@linux-foundation.org), and possibly to any other
335   relevant mailing-list or to the newsgroup.
336
337 - In all bug-reports, *please* tell what kernel you are talking about,
338   how to duplicate the problem, and what your setup is (use your common
339   sense).  If the problem is new, tell me so, and if the problem is
340   old, please try to tell me when you first noticed it.
341
342 - If the bug results in a message like::
343
344     unable to handle kernel paging request at address C0000010
345     Oops: 0002
346     EIP:   0010:XXXXXXXX
347     eax: xxxxxxxx   ebx: xxxxxxxx   ecx: xxxxxxxx   edx: xxxxxxxx
348     esi: xxxxxxxx   edi: xxxxxxxx   ebp: xxxxxxxx
349     ds: xxxx  es: xxxx  fs: xxxx  gs: xxxx
350     Pid: xx, process nr: xx
351     xx xx xx xx xx xx xx xx xx xx
352
353   or similar kernel debugging information on your screen or in your
354   system log, please duplicate it *exactly*.  The dump may look
355   incomprehensible to you, but it does contain information that may
356   help debugging the problem.  The text above the dump is also
357   important: it tells something about why the kernel dumped code (in
358   the above example, it's due to a bad kernel pointer). More information
359   on making sense of the dump is in Documentation/admin-guide/oops-tracing.rst
360
361 - If you compiled the kernel with CONFIG_KALLSYMS you can send the dump
362   as is, otherwise you will have to use the ``ksymoops`` program to make
363   sense of the dump (but compiling with CONFIG_KALLSYMS is usually preferred).
364   This utility can be downloaded from
365   https://www.kernel.org/pub/linux/utils/kernel/ksymoops/ .
366   Alternatively, you can do the dump lookup by hand:
367
368 - In debugging dumps like the above, it helps enormously if you can
369   look up what the EIP value means.  The hex value as such doesn't help
370   me or anybody else very much: it will depend on your particular
371   kernel setup.  What you should do is take the hex value from the EIP
372   line (ignore the ``0010:``), and look it up in the kernel namelist to
373   see which kernel function contains the offending address.
374
375   To find out the kernel function name, you'll need to find the system
376   binary associated with the kernel that exhibited the symptom.  This is
377   the file 'linux/vmlinux'.  To extract the namelist and match it against
378   the EIP from the kernel crash, do::
379
380     nm vmlinux | sort | less
381
382   This will give you a list of kernel addresses sorted in ascending
383   order, from which it is simple to find the function that contains the
384   offending address.  Note that the address given by the kernel
385   debugging messages will not necessarily match exactly with the
386   function addresses (in fact, that is very unlikely), so you can't
387   just 'grep' the list: the list will, however, give you the starting
388   point of each kernel function, so by looking for the function that
389   has a starting address lower than the one you are searching for but
390   is followed by a function with a higher address you will find the one
391   you want.  In fact, it may be a good idea to include a bit of
392   "context" in your problem report, giving a few lines around the
393   interesting one.
394
395   If you for some reason cannot do the above (you have a pre-compiled
396   kernel image or similar), telling me as much about your setup as
397   possible will help.  Please read the :ref:`admin-guide/reporting-bugs.rst <reportingbugs>`
398   document for details.
399
400 - Alternatively, you can use gdb on a running kernel. (read-only; i.e. you
401   cannot change values or set break points.) To do this, first compile the
402   kernel with -g; edit arch/x86/Makefile appropriately, then do a ``make
403   clean``. You'll also need to enable CONFIG_PROC_FS (via ``make config``).
404
405   After you've rebooted with the new kernel, do ``gdb vmlinux /proc/kcore``.
406   You can now use all the usual gdb commands. The command to look up the
407   point where your system crashed is ``l *0xXXXXXXXX``. (Replace the XXXes
408   with the EIP value.)
409
410   gdb'ing a non-running kernel currently fails because ``gdb`` (wrongly)
411   disregards the starting offset for which the kernel is compiled.
412