1        Linux kernel release 2.6.xx <>
   3These are the release notes for Linux version 2.6.  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,
  26  IBM S/390, MIPS, HP PA-RISC, Intel IA-64, DEC VAX, AMD x86-64, AXIS CRIS,
  27  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.
  37 - There is a lot of documentation available both in electronic form on
  38   the Internet and in books, both Linux-specific and pertaining to
  39   general UNIX questions.  I'd recommend looking into the documentation
  40   subdirectories on any Linux FTP site for the LDP (Linux Documentation
  41   Project) books.  This README is not meant to be documentation on the
  42   system: there are much better sources available.
  44 - There are various README files in the Documentation/ subdirectory:
  45   these typically contain kernel-specific installation notes for some 
  46   drivers for example. See Documentation/00-INDEX for a list of what
  47   is contained in each file.  Please read the Changes file, as it
  48   contains information about the problems, which may result by upgrading
  49   your kernel.
  51 - The Documentation/DocBook/ subdirectory contains several guides for
  52   kernel developers and users.  These guides can be rendered in a
  53   number of formats:  PostScript (.ps), PDF, and HTML, among others.
  54   After installation, "make psdocs", "make pdfdocs", or "make htmldocs"
  55   will render the documentation in the requested format.
  57INSTALLING the kernel:
  59 - If you install the full sources, put the kernel tarball in a
  60   directory where you have permissions (eg. your home directory) and
  61   unpack it:
  63                gzip -cd linux-2.6.XX.tar.gz | tar xvf -
  65   or
  66                bzip2 -dc linux-2.6.XX.tar.bz2 | tar xvf -
  69   Replace "XX" with the version number of the latest kernel.
  71   Do NOT use the /usr/src/linux area! This area has a (usually
  72   incomplete) set of kernel headers that are used by the library header
  73   files.  They should match the library, and not get messed up by
  74   whatever the kernel-du-jour happens to be.
  76 - You can also upgrade between 2.6.xx releases by patching.  Patches are
  77   distributed in the traditional gzip and the newer bzip2 format.  To
  78   install by patching, get all the newer patch files, enter the
  79   top level directory of the kernel source (linux-2.6.xx) and execute:
  81                gzip -cd ../patch-2.6.xx.gz | patch -p1
  83   or
  84                bzip2 -dc ../patch-2.6.xx.bz2 | patch -p1
  86   (repeat xx for all versions bigger than the version of your current
  87   source tree, _in_order_) and you should be ok.  You may want to remove
  88   the backup files (xxx~ or xxx.orig), and make sure that there are no
  89   failed patches (xxx# or xxx.rej). If there are, either you or me has
  90   made a mistake.
  92   Unlike patches for the 2.6.x kernels, patches for the 2.6.x.y kernels
  93   (also known as the -stable kernels) are not incremental but instead apply
  94   directly to the base 2.6.x kernel.  Please read
  95   Documentation/applying-patches.txt for more information.
  97   Alternatively, the script patch-kernel can be used to automate this
  98   process.  It determines the current kernel version and applies any
  99   patches found.
 101                linux/scripts/patch-kernel linux
 103   The first argument in the command above is the location of the
 104   kernel source.  Patches are applied from the current directory, but
 105   an alternative directory can be specified as the second argument.
 107 - If you are upgrading between releases using the stable series patches
 108   (for example, patch-2.6.xx.y), note that these "dot-releases" are
 109   not incremental and must be applied to the 2.6.xx base tree. For
 110   example, if your base kernel is 2.6.12 and you want to apply the
 111 patch, you do not and indeed must not first apply the
 112 and patches. Similarly, if you are running kernel
 113   version and want to jump to, you must first
 114   reverse the patch (that is, patch -R) _before_ applying
 115   the patch.
 117 - Make sure you have no stale .o files and dependencies lying around:
 119                cd linux
 120                make mrproper
 122   You should now have the sources correctly installed.
 126   Compiling and running the 2.6.xx kernels requires up-to-date
 127   versions of various software packages.  Consult
 128   Documentation/Changes for the minimum version numbers required
 129   and how to get updates for these packages.  Beware that using
 130   excessively old versions of these packages can cause indirect
 131   errors that are very difficult to track down, so don't assume that
 132   you can just update packages when obvious problems arise during
 133   build or operation.
 135BUILD directory for the kernel:
 137   When compiling the kernel all output files will per default be
 138   stored together with the kernel source code.
 139   Using the option "make O=output/dir" allow you to specify an alternate
 140   place for the output files (including .config).
 141   Example:
 142     kernel source code:        /usr/src/linux-2.6.N
 143     build directory:           /home/name/build/kernel
 145   To configure and build the kernel use:
 146   cd /usr/src/linux-2.6.N
 147   make O=/home/name/build/kernel menuconfig
 148   make O=/home/name/build/kernel
 149   sudo make O=/home/name/build/kernel modules_install install
 151   Please note: If the 'O=output/dir' option is used then it must be
 152   used for all invocations of make.
 154CONFIGURING the kernel:
 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.
 163 - Alternate configuration commands are:
 164        "make menuconfig"  Text based color menus, radiolists & dialogs.
 165        "make xconfig"     X windows (Qt) based configuration tool.
 166        "make gconfig"     X windows (Gtk) based configuration tool.
 167        "make oldconfig"   Default all questions based on the contents of
 168                           your existing ./.config file and asking about
 169                           new config symbols.
 170        "make silentoldconfig"
 171                           Like above, but avoids cluttering the screen
 172                           with questions already answered.
 173        "make defconfig"   Create a ./.config file by using the default
 174                           symbol values from arch/$ARCH/defconfig.
 175        "make allyesconfig"
 176                           Create a ./.config file by setting symbol
 177                           values to 'y' as much as possible.
 178        "make allmodconfig"
 179                           Create a ./.config file by setting symbol
 180                           values to 'm' as much as possible.
 181        "make allnoconfig" Create a ./.config file by setting symbol
 182                           values to 'n' as much as possible.
 183        "make randconfig"  Create a ./.config file by setting symbol
 184                           values to random values.
 186   The allyesconfig/allmodconfig/allnoconfig/randconfig variants can
 187   also use the environment variable KCONFIG_ALLCONFIG to specify a
 188   filename that contains config options that the user requires to be
 189   set to a specific value.  If KCONFIG_ALLCONFIG=filename is not used,
 190   "make *config" checks for a file named "all{yes/mod/no/random}.config"
 191   for symbol values that are to be forced.  If this file is not found,
 192   it checks for a file named "all.config" to contain forced values.
 194        NOTES on "make config":
 195        - having unnecessary drivers will make the kernel bigger, and can
 196          under some circumstances lead to problems: probing for a
 197          nonexistent controller card may confuse your other controllers
 198        - compiling the kernel with "Processor type" set higher than 386
 199          will result in a kernel that does NOT work on a 386.  The
 200          kernel will detect this on bootup, and give up.
 201        - A kernel with math-emulation compiled in will still use the
 202          coprocessor if one is present: the math emulation will just
 203          never get used in that case.  The kernel will be slightly larger,
 204          but will work on different machines regardless of whether they
 205          have a math coprocessor or not. 
 206        - the "kernel hacking" configuration details usually result in a
 207          bigger or slower kernel (or both), and can even make the kernel
 208          less stable by configuring some routines to actively try to
 209          break bad code to find kernel problems (kmalloc()).  Thus you
 210          should probably answer 'n' to the questions for
 211          "development", "experimental", or "debugging" features.
 213COMPILING the kernel:
 215 - Make sure you have at least gcc 3.2 available.
 216   For more information, refer to Documentation/Changes.
 218   Please note that you can still run a.out user programs with this kernel.
 220 - Do a "make" to create a compressed kernel image. It is also
 221   possible to do "make install" if you have lilo installed to suit the
 222   kernel makefiles, but you may want to check your particular lilo setup first.
 224   To do the actual install you have to be root, but none of the normal
 225   build should require that. Don't take the name of root in vain.
 227 - If you configured any of the parts of the kernel as `modules', you
 228   will also have to do "make modules_install".
 230 - Keep a backup kernel handy in case something goes wrong.  This is 
 231   especially true for the development releases, since each new release
 232   contains new code which has not been debugged.  Make sure you keep a
 233   backup of the modules corresponding to that kernel, as well.  If you
 234   are installing a new kernel with the same version number as your
 235   working kernel, make a backup of your modules directory before you
 236   do a "make modules_install".
 237   Alternatively, before compiling, use the kernel config option
 238   "LOCALVERSION" to append a unique suffix to the regular kernel version.
 239   LOCALVERSION can be set in the "General Setup" menu.
 241 - In order to boot your new kernel, you'll need to copy the kernel
 242   image (e.g. .../linux/arch/i386/boot/bzImage after compilation)
 243   to the place where your regular bootable kernel is found. 
 245 - Booting a kernel directly from a floppy without the assistance of a
 246   bootloader such as LILO, is no longer supported.
 248   If you boot Linux from the hard drive, chances are you use LILO which
 249   uses the kernel image as specified in the file /etc/lilo.conf.  The
 250   kernel image file is usually /vmlinuz, /boot/vmlinuz, /bzImage or
 251   /boot/bzImage.  To use the new kernel, save a copy of the old image
 252   and copy the new image over the old one.  Then, you MUST RERUN LILO
 253   to update the loading map!! If you don't, you won't be able to boot
 254   the new kernel image.
 256   Reinstalling LILO is usually a matter of running /sbin/lilo. 
 257   You may wish to edit /etc/lilo.conf to specify an entry for your
 258   old kernel image (say, /vmlinux.old) in case the new one does not
 259   work.  See the LILO docs for more information. 
 261   After reinstalling LILO, you should be all set.  Shutdown the system,
 262   reboot, and enjoy!
 264   If you ever need to change the default root device, video mode,
 265   ramdisk size, etc.  in the kernel image, use the 'rdev' program (or
 266   alternatively the LILO boot options when appropriate).  No need to
 267   recompile the kernel to change these parameters. 
 269 - Reboot with the new kernel and enjoy. 
 273 - If you have problems that seem to be due to kernel bugs, please check
 274   the file MAINTAINERS to see if there is a particular person associated
 275   with the part of the kernel that you are having trouble with. If there
 276   isn't anyone listed there, then the second best thing is to mail
 277   them to me (, and possibly to any other relevant
 278   mailing-list or to the newsgroup.
 280 - In all bug-reports, *please* tell what kernel you are talking about,
 281   how to duplicate the problem, and what your setup is (use your common
 282   sense).  If the problem is new, tell me so, and if the problem is
 283   old, please try to tell me when you first noticed it.
 285 - If the bug results in a message like
 287        unable to handle kernel paging request at address C0000010
 288        Oops: 0002
 289        EIP:   0010:XXXXXXXX
 290        eax: xxxxxxxx   ebx: xxxxxxxx   ecx: xxxxxxxx   edx: xxxxxxxx
 291        esi: xxxxxxxx   edi: xxxxxxxx   ebp: xxxxxxxx
 292        ds: xxxx  es: xxxx  fs: xxxx  gs: xxxx
 293        Pid: xx, process nr: xx
 294        xx xx xx xx xx xx xx xx xx xx
 296   or similar kernel debugging information on your screen or in your
 297   system log, please duplicate it *exactly*.  The dump may look
 298   incomprehensible to you, but it does contain information that may
 299   help debugging the problem.  The text above the dump is also
 300   important: it tells something about why the kernel dumped code (in
 301   the above example it's due to a bad kernel pointer). More information
 302   on making sense of the dump is in Documentation/oops-tracing.txt
 304 - If you compiled the kernel with CONFIG_KALLSYMS you can send the dump
 305   as is, otherwise you will have to use the "ksymoops" program to make
 306   sense of the dump.  This utility can be downloaded from
 307   ftp://ftp.<country>
 308   Alternately you can do the dump lookup by hand:
 310 - In debugging dumps like the above, it helps enormously if you can
 311   look up what the EIP value means.  The hex value as such doesn't help
 312   me or anybody else very much: it will depend on your particular
 313   kernel setup.  What you should do is take the hex value from the EIP
 314   line (ignore the "0010:"), and look it up in the kernel namelist to
 315   see which kernel function contains the offending address.
 317   To find out the kernel function name, you'll need to find the system
 318   binary associated with the kernel that exhibited the symptom.  This is
 319   the file 'linux/vmlinux'.  To extract the namelist and match it against
 320   the EIP from the kernel crash, do:
 322                nm vmlinux | sort | less
 324   This will give you a list of kernel addresses sorted in ascending
 325   order, from which it is simple to find the function that contains the
 326   offending address.  Note that the address given by the kernel
 327   debugging messages will not necessarily match exactly with the
 328   function addresses (in fact, that is very unlikely), so you can't
 329   just 'grep' the list: the list will, however, give you the starting
 330   point of each kernel function, so by looking for the function that
 331   has a starting address lower than the one you are searching for but
 332   is followed by a function with a higher address you will find the one
 333   you want.  In fact, it may be a good idea to include a bit of
 334   "context" in your problem report, giving a few lines around the
 335   interesting one. 
 337   If you for some reason cannot do the above (you have a pre-compiled
 338   kernel image or similar), telling me as much about your setup as
 339   possible will help. 
 341 - Alternately, you can use gdb on a running kernel. (read-only; i.e. you
 342   cannot change values or set break points.) To do this, first compile the
 343   kernel with -g; edit arch/i386/Makefile appropriately, then do a "make
 344   clean". You'll also need to enable CONFIG_PROC_FS (via "make config").
 346   After you've rebooted with the new kernel, do "gdb vmlinux /proc/kcore".
 347   You can now use all the usual gdb commands. The command to look up the
 348   point where your system crashed is "l *0xXXXXXXXX". (Replace the XXXes
 349   with the EIP value.)
 351   gdb'ing a non-running kernel currently fails because gdb (wrongly)
 352   disregards the starting offset for which the kernel is compiled.