blob: a24ec89ba4420ad38e8848422ca5f69baa5ffe1c [file] [log] [blame]
  1. Linux kernel release 3.x <>
  2. These are the release notes for Linux version 3. Read them carefully,
  3. as they tell you what this is all about, explain how to install the
  4. kernel, and what to do if something goes wrong.
  6. Linux is a clone of the operating system Unix, written from scratch by
  7. Linus Torvalds with assistance from a loosely-knit team of hackers across
  8. the Net. It aims towards POSIX and Single UNIX Specification compliance.
  9. It has all the features you would expect in a modern fully-fledged Unix,
  10. including true multitasking, virtual memory, shared libraries, demand
  11. loading, shared copy-on-write executables, proper memory management,
  12. and multistack networking including IPv4 and IPv6.
  13. It is distributed under the GNU General Public License - see the
  14. accompanying COPYING file for more details.
  16. Although originally developed first for 32-bit x86-based PCs (386 or higher),
  17. today Linux also runs on (at least) the Compaq Alpha AXP, Sun SPARC and
  18. UltraSPARC, Motorola 68000, PowerPC, PowerPC64, ARM, Hitachi SuperH, Cell,
  19. IBM S/390, MIPS, HP PA-RISC, Intel IA-64, DEC VAX, AMD x86-64, AXIS CRIS,
  20. Xtensa, Tilera TILE, AVR32 and Renesas M32R architectures.
  21. Linux is easily portable to most general-purpose 32- or 64-bit architectures
  22. as long as they have a paged memory management unit (PMMU) and a port of the
  23. GNU C compiler (gcc) (part of The GNU Compiler Collection, GCC). Linux has
  24. also been ported to a number of architectures without a PMMU, although
  25. functionality is then obviously somewhat limited.
  26. Linux has also been ported to itself. You can now run the kernel as a
  27. userspace application - this is called UserMode Linux (UML).
  29. - There is a lot of documentation available both in electronic form on
  30. the Internet and in books, both Linux-specific and pertaining to
  31. general UNIX questions. I'd recommend looking into the documentation
  32. subdirectories on any Linux FTP site for the LDP (Linux Documentation
  33. Project) books. This README is not meant to be documentation on the
  34. system: there are much better sources available.
  35. - There are various README files in the Documentation/ subdirectory:
  36. these typically contain kernel-specific installation notes for some
  37. drivers for example. See Documentation/00-INDEX for a list of what
  38. is contained in each file. Please read the Changes file, as it
  39. contains information about the problems, which may result by upgrading
  40. your kernel.
  41. - The Documentation/DocBook/ subdirectory contains several guides for
  42. kernel developers and users. These guides can be rendered in a
  43. number of formats: PostScript (.ps), PDF, HTML, & man-pages, among others.
  44. After installation, "make psdocs", "make pdfdocs", "make htmldocs",
  45. or "make mandocs" will render the documentation in the requested format.
  46. INSTALLING the kernel source:
  47. - If you install the full sources, put the kernel tarball in a
  48. directory where you have permissions (eg. your home directory) and
  49. unpack it:
  50. gzip -cd linux-3.X.tar.gz | tar xvf -
  51. or
  52. bzip2 -dc linux-3.X.tar.bz2 | tar xvf -
  53. Replace "X" with the version number of the latest kernel.
  54. Do NOT use the /usr/src/linux area! This area has a (usually
  55. incomplete) set of kernel headers that are used by the library header
  56. files. They should match the library, and not get messed up by
  57. whatever the kernel-du-jour happens to be.
  58. - You can also upgrade between 3.x releases by patching. Patches are
  59. distributed in the traditional gzip and the newer bzip2 format. To
  60. install by patching, get all the newer patch files, enter the
  61. top level directory of the kernel source (linux-3.X) and execute:
  62. gzip -cd ../patch-3.x.gz | patch -p1
  63. or
  64. bzip2 -dc ../patch-3.x.bz2 | patch -p1
  65. Replace "x" for all versions bigger than the version "X" of your current
  66. source tree, _in_order_, and you should be ok. You may want to remove
  67. the backup files (some-file-name~ or some-file-name.orig), and make sure
  68. that there are no failed patches (some-file-name# or some-file-name.rej).
  69. If there are, either you or I have made a mistake.
  70. Unlike patches for the 3.x kernels, patches for the 3.x.y kernels
  71. (also known as the -stable kernels) are not incremental but instead apply
  72. directly to the base 3.x kernel. For example, if your base kernel is 3.0
  73. and you want to apply the 3.0.3 patch, you must not first apply the 3.0.1
  74. and 3.0.2 patches. Similarly, if you are running kernel version 3.0.2 and
  75. want to jump to 3.0.3, you must first reverse the 3.0.2 patch (that is,
  76. patch -R) _before_ applying the 3.0.3 patch. You can read more on this in
  77. Documentation/applying-patches.txt
  78. Alternatively, the script patch-kernel can be used to automate this
  79. process. It determines the current kernel version and applies any
  80. patches found.
  81. linux/scripts/patch-kernel linux
  82. The first argument in the command above is the location of the
  83. kernel source. Patches are applied from the current directory, but
  84. an alternative directory can be specified as the second argument.
  85. - Make sure you have no stale .o files and dependencies lying around:
  86. cd linux
  87. make mrproper
  88. You should now have the sources correctly installed.
  90. Compiling and running the 3.x kernels requires up-to-date
  91. versions of various software packages. Consult
  92. Documentation/Changes for the minimum version numbers required
  93. and how to get updates for these packages. Beware that using
  94. excessively old versions of these packages can cause indirect
  95. errors that are very difficult to track down, so don't assume that
  96. you can just update packages when obvious problems arise during
  97. build or operation.
  98. BUILD directory for the kernel:
  99. When compiling the kernel, all output files will per default be
  100. stored together with the kernel source code.
  101. Using the option "make O=output/dir" allow you to specify an alternate
  102. place for the output files (including .config).
  103. Example:
  104. kernel source code: /usr/src/linux-3.X
  105. build directory: /home/name/build/kernel
  106. To configure and build the kernel, use:
  107. cd /usr/src/linux-3.X
  108. make O=/home/name/build/kernel menuconfig
  109. make O=/home/name/build/kernel
  110. sudo make O=/home/name/build/kernel modules_install install
  111. Please note: If the 'O=output/dir' option is used, then it must be
  112. used for all invocations of make.
  113. CONFIGURING the kernel:
  114. Do not skip this step even if you are only upgrading one minor
  115. version. New configuration options are added in each release, and
  116. odd problems will turn up if the configuration files are not set up
  117. as expected. If you want to carry your existing configuration to a
  118. new version with minimal work, use "make oldconfig", which will
  119. only ask you for the answers to new questions.
  120. - Alternative configuration commands are:
  121. "make config" Plain text interface.
  122. "make menuconfig" Text based color menus, radiolists & dialogs.
  123. "make nconfig" Enhanced text based color menus.
  124. "make xconfig" X windows (Qt) based configuration tool.
  125. "make gconfig" X windows (Gtk) based configuration tool.
  126. "make oldconfig" Default all questions based on the contents of
  127. your existing ./.config file and asking about
  128. new config symbols.
  129. "make silentoldconfig"
  130. Like above, but avoids cluttering the screen
  131. with questions already answered.
  132. Additionally updates the dependencies.
  133. "make olddefconfig"
  134. Like above, but sets new symbols to their default
  135. values without prompting.
  136. "make defconfig" Create a ./.config file by using the default
  137. symbol values from either arch/$ARCH/defconfig
  138. or arch/$ARCH/configs/${PLATFORM}_defconfig,
  139. depending on the architecture.
  140. "make ${PLATFORM}_defconfig"
  141. Create a ./.config file by using the default
  142. symbol values from
  143. arch/$ARCH/configs/${PLATFORM}_defconfig.
  144. Use "make help" to get a list of all available
  145. platforms of your architecture.
  146. "make allyesconfig"
  147. Create a ./.config file by setting symbol
  148. values to 'y' as much as possible.
  149. "make allmodconfig"
  150. Create a ./.config file by setting symbol
  151. values to 'm' as much as possible.
  152. "make allnoconfig" Create a ./.config file by setting symbol
  153. values to 'n' as much as possible.
  154. "make randconfig" Create a ./.config file by setting symbol
  155. values to random values.
  156. "make localmodconfig" Create a config based on current config and
  157. loaded modules (lsmod). Disables any module
  158. option that is not needed for the loaded modules.
  159. To create a localmodconfig for another machine,
  160. store the lsmod of that machine into a file
  161. and pass it in as a LSMOD parameter.
  162. target$ lsmod > /tmp/mylsmod
  163. target$ scp /tmp/mylsmod host:/tmp
  164. host$ make LSMOD=/tmp/mylsmod localmodconfig
  165. The above also works when cross compiling.
  166. "make localyesconfig" Similar to localmodconfig, except it will convert
  167. all module options to built in (=y) options.
  168. You can find more information on using the Linux kernel config tools
  169. in Documentation/kbuild/kconfig.txt.
  170. - NOTES on "make config":
  171. - Having unnecessary drivers will make the kernel bigger, and can
  172. under some circumstances lead to problems: probing for a
  173. nonexistent controller card may confuse your other controllers
  174. - Compiling the kernel with "Processor type" set higher than 386
  175. will result in a kernel that does NOT work on a 386. The
  176. kernel will detect this on bootup, and give up.
  177. - A kernel with math-emulation compiled in will still use the
  178. coprocessor if one is present: the math emulation will just
  179. never get used in that case. The kernel will be slightly larger,
  180. but will work on different machines regardless of whether they
  181. have a math coprocessor or not.
  182. - The "kernel hacking" configuration details usually result in a
  183. bigger or slower kernel (or both), and can even make the kernel
  184. less stable by configuring some routines to actively try to
  185. break bad code to find kernel problems (kmalloc()). Thus you
  186. should probably answer 'n' to the questions for "development",
  187. "experimental", or "debugging" features.
  188. COMPILING the kernel:
  189. - Make sure you have at least gcc 3.2 available.
  190. For more information, refer to Documentation/Changes.
  191. Please note that you can still run a.out user programs with this kernel.
  192. - Do a "make" to create a compressed kernel image. It is also
  193. possible to do "make install" if you have lilo installed to suit the
  194. kernel makefiles, but you may want to check your particular lilo setup first.
  195. To do the actual install, you have to be root, but none of the normal
  196. build should require that. Don't take the name of root in vain.
  197. - If you configured any of the parts of the kernel as `modules', you
  198. will also have to do "make modules_install".
  199. - Verbose kernel compile/build output:
  200. Normally, the kernel build system runs in a fairly quiet mode (but not
  201. totally silent). However, sometimes you or other kernel developers need
  202. to see compile, link, or other commands exactly as they are executed.
  203. For this, use "verbose" build mode. This is done by inserting
  204. "V=1" in the "make" command. E.g.:
  205. make V=1 all
  206. To have the build system also tell the reason for the rebuild of each
  207. target, use "V=2". The default is "V=0".
  208. - Keep a backup kernel handy in case something goes wrong. This is
  209. especially true for the development releases, since each new release
  210. contains new code which has not been debugged. Make sure you keep a
  211. backup of the modules corresponding to that kernel, as well. If you
  212. are installing a new kernel with the same version number as your
  213. working kernel, make a backup of your modules directory before you
  214. do a "make modules_install".
  215. Alternatively, before compiling, use the kernel config option
  216. "LOCALVERSION" to append a unique suffix to the regular kernel version.
  217. LOCALVERSION can be set in the "General Setup" menu.
  218. - In order to boot your new kernel, you'll need to copy the kernel
  219. image (e.g. .../linux/arch/i386/boot/bzImage after compilation)
  220. to the place where your regular bootable kernel is found.
  221. - Booting a kernel directly from a floppy without the assistance of a
  222. bootloader such as LILO, is no longer supported.
  223. If you boot Linux from the hard drive, chances are you use LILO, which
  224. uses the kernel image as specified in the file /etc/lilo.conf. The
  225. kernel image file is usually /vmlinuz, /boot/vmlinuz, /bzImage or
  226. /boot/bzImage. To use the new kernel, save a copy of the old image
  227. and copy the new image over the old one. Then, you MUST RERUN LILO
  228. to update the loading map!! If you don't, you won't be able to boot
  229. the new kernel image.
  230. Reinstalling LILO is usually a matter of running /sbin/lilo.
  231. You may wish to edit /etc/lilo.conf to specify an entry for your
  232. old kernel image (say, /vmlinux.old) in case the new one does not
  233. work. See the LILO docs for more information.
  234. After reinstalling LILO, you should be all set. Shutdown the system,
  235. reboot, and enjoy!
  236. If you ever need to change the default root device, video mode,
  237. ramdisk size, etc. in the kernel image, use the 'rdev' program (or
  238. alternatively the LILO boot options when appropriate). No need to
  239. recompile the kernel to change these parameters.
  240. - Reboot with the new kernel and enjoy.
  242. - If you have problems that seem to be due to kernel bugs, please check
  243. the file MAINTAINERS to see if there is a particular person associated
  244. with the part of the kernel that you are having trouble with. If there
  245. isn't anyone listed there, then the second best thing is to mail
  246. them to me (, and possibly to any other
  247. relevant mailing-list or to the newsgroup.
  248. - In all bug-reports, *please* tell what kernel you are talking about,
  249. how to duplicate the problem, and what your setup is (use your common
  250. sense). If the problem is new, tell me so, and if the problem is
  251. old, please try to tell me when you first noticed it.
  252. - If the bug results in a message like
  253. unable to handle kernel paging request at address C0000010
  254. Oops: 0002
  255. EIP: 0010:XXXXXXXX
  256. eax: xxxxxxxx ebx: xxxxxxxx ecx: xxxxxxxx edx: xxxxxxxx
  257. esi: xxxxxxxx edi: xxxxxxxx ebp: xxxxxxxx
  258. ds: xxxx es: xxxx fs: xxxx gs: xxxx
  259. Pid: xx, process nr: xx
  260. xx xx xx xx xx xx xx xx xx xx
  261. or similar kernel debugging information on your screen or in your
  262. system log, please duplicate it *exactly*. The dump may look
  263. incomprehensible to you, but it does contain information that may
  264. help debugging the problem. The text above the dump is also
  265. important: it tells something about why the kernel dumped code (in
  266. the above example, it's due to a bad kernel pointer). More information
  267. on making sense of the dump is in Documentation/oops-tracing.txt
  268. - If you compiled the kernel with CONFIG_KALLSYMS you can send the dump
  269. as is, otherwise you will have to use the "ksymoops" program to make
  270. sense of the dump (but compiling with CONFIG_KALLSYMS is usually preferred).
  271. This utility can be downloaded from
  272. ftp://ftp.<country> .
  273. Alternatively, you can do the dump lookup by hand:
  274. - In debugging dumps like the above, it helps enormously if you can
  275. look up what the EIP value means. The hex value as such doesn't help
  276. me or anybody else very much: it will depend on your particular
  277. kernel setup. What you should do is take the hex value from the EIP
  278. line (ignore the "0010:"), and look it up in the kernel namelist to
  279. see which kernel function contains the offending address.
  280. To find out the kernel function name, you'll need to find the system
  281. binary associated with the kernel that exhibited the symptom. This is
  282. the file 'linux/vmlinux'. To extract the namelist and match it against
  283. the EIP from the kernel crash, do:
  284. nm vmlinux | sort | less
  285. This will give you a list of kernel addresses sorted in ascending
  286. order, from which it is simple to find the function that contains the
  287. offending address. Note that the address given by the kernel
  288. debugging messages will not necessarily match exactly with the
  289. function addresses (in fact, that is very unlikely), so you can't
  290. just 'grep' the list: the list will, however, give you the starting
  291. point of each kernel function, so by looking for the function that
  292. has a starting address lower than the one you are searching for but
  293. is followed by a function with a higher address you will find the one
  294. you want. In fact, it may be a good idea to include a bit of
  295. "context" in your problem report, giving a few lines around the
  296. interesting one.
  297. If you for some reason cannot do the above (you have a pre-compiled
  298. kernel image or similar), telling me as much about your setup as
  299. possible will help. Please read the REPORTING-BUGS document for details.
  300. - Alternatively, you can use gdb on a running kernel. (read-only; i.e. you
  301. cannot change values or set break points.) To do this, first compile the
  302. kernel with -g; edit arch/i386/Makefile appropriately, then do a "make
  303. clean". You'll also need to enable CONFIG_PROC_FS (via "make config").
  304. After you've rebooted with the new kernel, do "gdb vmlinux /proc/kcore".
  305. You can now use all the usual gdb commands. The command to look up the
  306. point where your system crashed is "l *0xXXXXXXXX". (Replace the XXXes
  307. with the EIP value.)
  308. gdb'ing a non-running kernel currently fails because gdb (wrongly)
  309. disregards the starting offset for which the kernel is compiled.