1        Linux kernel release 3.x <>
   3These are the release notes for Linux version 3.  Read them carefully,
   4as they tell you what this is all about, explain how to install the
   5kernel, and what to do if something goes wrong. 
   9  Linux is a clone of the operating system Unix, written from scratch by
  10  Linus Torvalds with assistance from a loosely-knit team of hackers across
  11  the Net. It aims towards POSIX and Single UNIX Specification compliance.
  13  It has all the features you would expect in a modern fully-fledged Unix,
  14  including true multitasking, virtual memory, shared libraries, demand
  15  loading, shared copy-on-write executables, proper memory management,
  16  and multistack networking including IPv4 and IPv6.
  18  It is distributed under the GNU General Public License - see the
  19  accompanying COPYING file for more details. 
  23  Although originally developed first for 32-bit x86-based PCs (386 or higher),
  24  today Linux also runs on (at least) the Compaq Alpha AXP, Sun SPARC and
  25  UltraSPARC, Motorola 68000, PowerPC, PowerPC64, ARM, Hitachi SuperH, Cell,
  26  IBM S/390, MIPS, HP PA-RISC, Intel IA-64, DEC VAX, AMD x86-64, AXIS CRIS,
  27  Xtensa, Tilera TILE, AVR32 and Renesas M32R architectures.
  29  Linux is easily portable to most general-purpose 32- or 64-bit architectures
  30  as long as they have a paged memory management unit (PMMU) and a port of the
  31  GNU C compiler (gcc) (part of The GNU Compiler Collection, GCC). Linux has
  32  also been ported to a number of architectures without a PMMU, although
  33  functionality is then obviously somewhat limited.
  34  Linux has also been ported to itself. You can now run the kernel as a
  35  userspace application - this is called UserMode Linux (UML).
  39 - There is a lot of documentation available both in electronic form on
  40   the Internet and in books, both Linux-specific and pertaining to
  41   general UNIX questions.  I'd recommend looking into the documentation
  42   subdirectories on any Linux FTP site for the LDP (Linux Documentation
  43   Project) books.  This README is not meant to be documentation on the
  44   system: there are much better sources available.
  46 - There are various README files in the Documentation/ subdirectory:
  47   these typically contain kernel-specific installation notes for some 
  48   drivers for example. See Documentation/00-INDEX for a list of what
  49   is contained in each file.  Please read the Changes file, as it
  50   contains information about the problems, which may result by upgrading
  51   your kernel.
  53 - The Documentation/DocBook/ subdirectory contains several guides for
  54   kernel developers and users.  These guides can be rendered in a
  55   number of formats:  PostScript (.ps), PDF, HTML, & man-pages, among others.
  56   After installation, "make psdocs", "make pdfdocs", "make htmldocs",
  57   or "make mandocs" will render the documentation in the requested format.
  59INSTALLING the kernel source:
  61 - If you install the full sources, put the kernel tarball in a
  62   directory where you have permissions (eg. your home directory) and
  63   unpack it:
  65     gzip -cd linux-3.X.tar.gz | tar xvf -
  67   or
  69     bzip2 -dc linux-3.X.tar.bz2 | tar xvf -
  71   Replace "X" with the version number of the latest kernel.
  73   Do NOT use the /usr/src/linux area! This area has a (usually
  74   incomplete) set of kernel headers that are used by the library header
  75   files.  They should match the library, and not get messed up by
  76   whatever the kernel-du-jour happens to be.
  78 - You can also upgrade between 3.x releases by patching.  Patches are
  79   distributed in the traditional gzip and the newer bzip2 format.  To
  80   install by patching, get all the newer patch files, enter the
  81   top level directory of the kernel source (linux-3.X) and execute:
  83     gzip -cd ../patch-3.x.gz | patch -p1
  85   or
  87     bzip2 -dc ../patch-3.x.bz2 | patch -p1
  89   Replace "x" for all versions bigger than the version "X" of your current
  90   source tree, _in_order_, and you should be ok.  You may want to remove
  91   the backup files (some-file-name~ or some-file-name.orig), and make sure
  92   that there are no failed patches (some-file-name# or some-file-name.rej).
  93   If there are, either you or I have made a mistake.
  95   Unlike patches for the 3.x kernels, patches for the 3.x.y kernels
  96   (also known as the -stable kernels) are not incremental but instead apply
  97   directly to the base 3.x kernel.  For example, if your base kernel is 3.0
  98   and you want to apply the 3.0.3 patch, you must not first apply the 3.0.1
  99   and 3.0.2 patches. Similarly, if you are running kernel version 3.0.2 and
 100   want to jump to 3.0.3, you must first reverse the 3.0.2 patch (that is,
 101   patch -R) _before_ applying the 3.0.3 patch. You can read more on this in
 102   Documentation/applying-patches.txt
 104   Alternatively, the script patch-kernel can be used to automate this
 105   process.  It determines the current kernel version and applies any
 106   patches found.
 108     linux/scripts/patch-kernel linux
 110   The first argument in the command above is the location of the
 111   kernel source.  Patches are applied from the current directory, but
 112   an alternative directory can be specified as the second argument.
 114 - Make sure you have no stale .o files and dependencies lying around:
 116     cd linux
 117     make mrproper
 119   You should now have the sources correctly installed.
 123   Compiling and running the 3.x kernels requires up-to-date
 124   versions of various software packages.  Consult
 125   Documentation/Changes for the minimum version numbers required
 126   and how to get updates for these packages.  Beware that using
 127   excessively old versions of these packages can cause indirect
 128   errors that are very difficult to track down, so don't assume that
 129   you can just update packages when obvious problems arise during
 130   build or operation.
 132BUILD directory for the kernel:
 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" allow you to specify an alternate
 137   place for the output files (including .config).
 138   Example:
 140     kernel source code: /usr/src/linux-3.X
 141     build directory:    /home/name/build/kernel
 143   To configure and build the kernel, use:
 145     cd /usr/src/linux-3.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
 150   Please note: If the 'O=output/dir' option is used, then it must be
 151   used for all invocations of make.
 153CONFIGURING the kernel:
 155   Do not skip this step even if you are only upgrading one minor
 156   version.  New configuration options are added in each release, and
 157   odd problems will turn up if the configuration files are not set up
 158   as expected.  If you want to carry your existing configuration to a
 159   new version with minimal work, use "make oldconfig", which will
 160   only ask you for the answers to new questions.
 162 - Alternative configuration commands are:
 164     "make config"      Plain text interface.
 166     "make menuconfig"  Text based color menus, radiolists & dialogs.
 168     "make nconfig"     Enhanced text based color menus.
 170     "make xconfig"     X windows (Qt) based configuration tool.
 172     "make gconfig"     X windows (Gtk) based configuration tool.
 174     "make oldconfig"   Default all questions based on the contents of
 175                        your existing ./.config file and asking about
 176                        new config symbols.
 178     "make silentoldconfig"
 179                        Like above, but avoids cluttering the screen
 180                        with questions already answered.
 181                        Additionally updates the dependencies.
 183     "make defconfig"   Create a ./.config file by using the default
 184                        symbol values from either arch/$ARCH/defconfig
 185                        or arch/$ARCH/configs/${PLATFORM}_defconfig,
 186                        depending on the architecture.
 188     "make ${PLATFORM}_defconfig"
 189                        Create a ./.config file by using the default
 190                        symbol values from
 191                        arch/$ARCH/configs/${PLATFORM}_defconfig.
 192                        Use "make help" to get a list of all available
 193                        platforms of your architecture.
 195     "make allyesconfig"
 196                        Create a ./.config file by setting symbol
 197                        values to 'y' as much as possible.
 199     "make allmodconfig"
 200                        Create a ./.config file by setting symbol
 201                        values to 'm' as much as possible.
 203     "make allnoconfig" Create a ./.config file by setting symbol
 204                        values to 'n' as much as possible.
 206     "make randconfig"  Create a ./.config file by setting symbol
 207                        values to random values.
 209     "make localmodconfig" Create a config based on current config and
 210                           loaded modules (lsmod). Disables any module
 211                           option that is not needed for the loaded modules.
 213                           To create a localmodconfig for another machine,
 214                           store the lsmod of that machine into a file
 215                           and pass it in as a LSMOD parameter.
 217                   target$ lsmod > /tmp/mylsmod
 218                   target$ scp /tmp/mylsmod host:/tmp
 220                   host$ make LSMOD=/tmp/mylsmod localmodconfig
 222                           The above also works when cross compiling.
 224     "make localyesconfig" Similar to localmodconfig, except it will convert
 225                           all module options to built in (=y) options.
 227   You can find more information on using the Linux kernel config tools
 228   in Documentation/kbuild/kconfig.txt.
 230 - NOTES on "make config":
 232    - Having unnecessary drivers will make the kernel bigger, and can
 233      under some circumstances lead to problems: probing for a
 234      nonexistent controller card may confuse your other controllers
 236    - Compiling the kernel with "Processor type" set higher than 386
 237      will result in a kernel that does NOT work on a 386.  The
 238      kernel will detect this on bootup, and give up.
 240    - A kernel with math-emulation compiled in will still use the
 241      coprocessor if one is present: the math emulation will just
 242      never get used in that case.  The kernel will be slightly larger,
 243      but will work on different machines regardless of whether they
 244      have a math coprocessor or not.
 246    - The "kernel hacking" configuration details usually result in a
 247      bigger or slower kernel (or both), and can even make the kernel
 248      less stable by configuring some routines to actively try to
 249      break bad code to find kernel problems (kmalloc()).  Thus you
 250      should probably answer 'n' to the questions for "development",
 251      "experimental", or "debugging" features.
 253COMPILING the kernel:
 255 - Make sure you have at least gcc 3.2 available.
 256   For more information, refer to Documentation/Changes.
 258   Please note that you can still run a.out user programs with this kernel.
 260 - Do a "make" to create a compressed kernel image. It is also
 261   possible to do "make install" if you have lilo installed to suit the
 262   kernel makefiles, but you may want to check your particular lilo setup first.
 264   To do the actual install, you have to be root, but none of the normal
 265   build should require that. Don't take the name of root in vain.
 267 - If you configured any of the parts of the kernel as `modules', you
 268   will also have to do "make modules_install".
 270 - Verbose kernel compile/build output:
 272   Normally, the kernel build system runs in a fairly quiet mode (but not
 273   totally silent).  However, sometimes you or other kernel developers need
 274   to see compile, link, or other commands exactly as they are executed.
 275   For this, use "verbose" build mode.  This is done by inserting
 276   "V=1" in the "make" command.  E.g.:
 278     make V=1 all
 280   To have the build system also tell the reason for the rebuild of each
 281   target, use "V=2".  The default is "V=0".
 283 - Keep a backup kernel handy in case something goes wrong.  This is 
 284   especially true for the development releases, since each new release
 285   contains new code which has not been debugged.  Make sure you keep a
 286   backup of the modules corresponding to that kernel, as well.  If you
 287   are installing a new kernel with the same version number as your
 288   working kernel, make a backup of your modules directory before you
 289   do a "make modules_install".
 291   Alternatively, before compiling, use the kernel config option
 292   "LOCALVERSION" to append a unique suffix to the regular kernel version.
 293   LOCALVERSION can be set in the "General Setup" menu.
 295 - In order to boot your new kernel, you'll need to copy the kernel
 296   image (e.g. .../linux/arch/i386/boot/bzImage after compilation)
 297   to the place where your regular bootable kernel is found. 
 299 - Booting a kernel directly from a floppy without the assistance of a
 300   bootloader such as LILO, is no longer supported.
 302   If you boot Linux from the hard drive, chances are you use LILO, which
 303   uses the kernel image as specified in the file /etc/lilo.conf.  The
 304   kernel image file is usually /vmlinuz, /boot/vmlinuz, /bzImage or
 305   /boot/bzImage.  To use the new kernel, save a copy of the old image
 306   and copy the new image over the old one.  Then, you MUST RERUN LILO
 307   to update the loading map!! If you don't, you won't be able to boot
 308   the new kernel image.
 310   Reinstalling LILO is usually a matter of running /sbin/lilo. 
 311   You may wish to edit /etc/lilo.conf to specify an entry for your
 312   old kernel image (say, /vmlinux.old) in case the new one does not
 313   work.  See the LILO docs for more information. 
 315   After reinstalling LILO, you should be all set.  Shutdown the system,
 316   reboot, and enjoy!
 318   If you ever need to change the default root device, video mode,
 319   ramdisk size, etc.  in the kernel image, use the 'rdev' program (or
 320   alternatively the LILO boot options when appropriate).  No need to
 321   recompile the kernel to change these parameters. 
 323 - Reboot with the new kernel and enjoy. 
 327 - If you have problems that seem to be due to kernel bugs, please check
 328   the file MAINTAINERS to see if there is a particular person associated
 329   with the part of the kernel that you are having trouble with. If there
 330   isn't anyone listed there, then the second best thing is to mail
 331   them to me (, and possibly to any other
 332   relevant mailing-list or to the newsgroup.
 334 - In all bug-reports, *please* tell what kernel you are talking about,
 335   how to duplicate the problem, and what your setup is (use your common
 336   sense).  If the problem is new, tell me so, and if the problem is
 337   old, please try to tell me when you first noticed it.
 339 - If the bug results in a message like
 341     unable to handle kernel paging request at address C0000010
 342     Oops: 0002
 343     EIP:   0010:XXXXXXXX
 344     eax: xxxxxxxx   ebx: xxxxxxxx   ecx: xxxxxxxx   edx: xxxxxxxx
 345     esi: xxxxxxxx   edi: xxxxxxxx   ebp: xxxxxxxx
 346     ds: xxxx  es: xxxx  fs: xxxx  gs: xxxx
 347     Pid: xx, process nr: xx
 348     xx xx xx xx xx xx xx xx xx xx
 350   or similar kernel debugging information on your screen or in your
 351   system log, please duplicate it *exactly*.  The dump may look
 352   incomprehensible to you, but it does contain information that may
 353   help debugging the problem.  The text above the dump is also
 354   important: it tells something about why the kernel dumped code (in
 355   the above example, it's due to a bad kernel pointer). More information
 356   on making sense of the dump is in Documentation/oops-tracing.txt
 358 - If you compiled the kernel with CONFIG_KALLSYMS you can send the dump
 359   as is, otherwise you will have to use the "ksymoops" program to make
 360   sense of the dump (but compiling with CONFIG_KALLSYMS is usually preferred).
 361   This utility can be downloaded from
 362   ftp://ftp.<country> .
 363   Alternatively, you can do the dump lookup by hand:
 365 - In debugging dumps like the above, it helps enormously if you can
 366   look up what the EIP value means.  The hex value as such doesn't help
 367   me or anybody else very much: it will depend on your particular
 368   kernel setup.  What you should do is take the hex value from the EIP
 369   line (ignore the "0010:"), and look it up in the kernel namelist to
 370   see which kernel function contains the offending address.
 372   To find out the kernel function name, you'll need to find the system
 373   binary associated with the kernel that exhibited the symptom.  This is
 374   the file 'linux/vmlinux'.  To extract the namelist and match it against
 375   the EIP from the kernel crash, do:
 377     nm vmlinux | sort | less
 379   This will give you a list of kernel addresses sorted in ascending
 380   order, from which it is simple to find the function that contains the
 381   offending address.  Note that the address given by the kernel
 382   debugging messages will not necessarily match exactly with the
 383   function addresses (in fact, that is very unlikely), so you can't
 384   just 'grep' the list: the list will, however, give you the starting
 385   point of each kernel function, so by looking for the function that
 386   has a starting address lower than the one you are searching for but
 387   is followed by a function with a higher address you will find the one
 388   you want.  In fact, it may be a good idea to include a bit of
 389   "context" in your problem report, giving a few lines around the
 390   interesting one. 
 392   If you for some reason cannot do the above (you have a pre-compiled
 393   kernel image or similar), telling me as much about your setup as
 394   possible will help.  Please read the REPORTING-BUGS document for details.
 396 - Alternatively, you can use gdb on a running kernel. (read-only; i.e. you
 397   cannot change values or set break points.) To do this, first compile the
 398   kernel with -g; edit arch/i386/Makefile appropriately, then do a "make
 399   clean". You'll also need to enable CONFIG_PROC_FS (via "make config").
 401   After you've rebooted with the new kernel, do "gdb vmlinux /proc/kcore".
 402   You can now use all the usual gdb commands. The command to look up the
 403   point where your system crashed is "l *0xXXXXXXXX". (Replace the XXXes
 404   with the EIP value.)
 406   gdb'ing a non-running kernel currently fails because gdb (wrongly)
 407   disregards the starting offset for which the kernel is compiled.