ia64/linux-2.6.18-xen.hg

diff README @ 0:831230e53067

Import 2.6.18 from kernel.org tarball.
author Ian Campbell <ian.campbell@xensource.com>
date Wed Apr 11 14:15:44 2007 +0100 (2007-04-11)
parents
children
line diff
     1.1 --- /dev/null	Thu Jan 01 00:00:00 1970 +0000
     1.2 +++ b/README	Wed Apr 11 14:15:44 2007 +0100
     1.3 @@ -0,0 +1,353 @@
     1.4 +	Linux kernel release 2.6.xx <http://kernel.org>
     1.5 +
     1.6 +These are the release notes for Linux version 2.6.  Read them carefully,
     1.7 +as they tell you what this is all about, explain how to install the
     1.8 +kernel, and what to do if something goes wrong. 
     1.9 +
    1.10 +WHAT IS LINUX?
    1.11 +
    1.12 +  Linux is a clone of the operating system Unix, written from scratch by
    1.13 +  Linus Torvalds with assistance from a loosely-knit team of hackers across
    1.14 +  the Net. It aims towards POSIX and Single UNIX Specification compliance.
    1.15 +
    1.16 +  It has all the features you would expect in a modern fully-fledged Unix,
    1.17 +  including true multitasking, virtual memory, shared libraries, demand
    1.18 +  loading, shared copy-on-write executables, proper memory management,
    1.19 +  and multistack networking including IPv4 and IPv6.
    1.20 +
    1.21 +  It is distributed under the GNU General Public License - see the
    1.22 +  accompanying COPYING file for more details. 
    1.23 +
    1.24 +ON WHAT HARDWARE DOES IT RUN?
    1.25 +
    1.26 +  Although originally developed first for 32-bit x86-based PCs (386 or higher),
    1.27 +  today Linux also runs on (at least) the Compaq Alpha AXP, Sun SPARC and
    1.28 +  UltraSPARC, Motorola 68000, PowerPC, PowerPC64, ARM, Hitachi SuperH,
    1.29 +  IBM S/390, MIPS, HP PA-RISC, Intel IA-64, DEC VAX, AMD x86-64, AXIS CRIS,
    1.30 +  and Renesas M32R architectures.
    1.31 +
    1.32 +  Linux is easily portable to most general-purpose 32- or 64-bit architectures
    1.33 +  as long as they have a paged memory management unit (PMMU) and a port of the
    1.34 +  GNU C compiler (gcc) (part of The GNU Compiler Collection, GCC). Linux has
    1.35 +  also been ported to a number of architectures without a PMMU, although
    1.36 +  functionality is then obviously somewhat limited.
    1.37 +
    1.38 +DOCUMENTATION:
    1.39 +
    1.40 + - There is a lot of documentation available both in electronic form on
    1.41 +   the Internet and in books, both Linux-specific and pertaining to
    1.42 +   general UNIX questions.  I'd recommend looking into the documentation
    1.43 +   subdirectories on any Linux FTP site for the LDP (Linux Documentation
    1.44 +   Project) books.  This README is not meant to be documentation on the
    1.45 +   system: there are much better sources available.
    1.46 +
    1.47 + - There are various README files in the Documentation/ subdirectory:
    1.48 +   these typically contain kernel-specific installation notes for some 
    1.49 +   drivers for example. See Documentation/00-INDEX for a list of what
    1.50 +   is contained in each file.  Please read the Changes file, as it
    1.51 +   contains information about the problems, which may result by upgrading
    1.52 +   your kernel.
    1.53 +
    1.54 + - The Documentation/DocBook/ subdirectory contains several guides for
    1.55 +   kernel developers and users.  These guides can be rendered in a
    1.56 +   number of formats:  PostScript (.ps), PDF, and HTML, among others.
    1.57 +   After installation, "make psdocs", "make pdfdocs", or "make htmldocs"
    1.58 +   will render the documentation in the requested format.
    1.59 +
    1.60 +INSTALLING the kernel:
    1.61 +
    1.62 + - If you install the full sources, put the kernel tarball in a
    1.63 +   directory where you have permissions (eg. your home directory) and
    1.64 +   unpack it:
    1.65 +
    1.66 +		gzip -cd linux-2.6.XX.tar.gz | tar xvf -
    1.67 +
    1.68 +   or
    1.69 +		bzip2 -dc linux-2.6.XX.tar.bz2 | tar xvf -
    1.70 +
    1.71 +
    1.72 +   Replace "XX" with the version number of the latest kernel.
    1.73 +
    1.74 +   Do NOT use the /usr/src/linux area! This area has a (usually
    1.75 +   incomplete) set of kernel headers that are used by the library header
    1.76 +   files.  They should match the library, and not get messed up by
    1.77 +   whatever the kernel-du-jour happens to be.
    1.78 +
    1.79 + - You can also upgrade between 2.6.xx releases by patching.  Patches are
    1.80 +   distributed in the traditional gzip and the newer bzip2 format.  To
    1.81 +   install by patching, get all the newer patch files, enter the
    1.82 +   top level directory of the kernel source (linux-2.6.xx) and execute:
    1.83 +
    1.84 +		gzip -cd ../patch-2.6.xx.gz | patch -p1
    1.85 +
    1.86 +   or
    1.87 +		bzip2 -dc ../patch-2.6.xx.bz2 | patch -p1
    1.88 +
    1.89 +   (repeat xx for all versions bigger than the version of your current
    1.90 +   source tree, _in_order_) and you should be ok.  You may want to remove
    1.91 +   the backup files (xxx~ or xxx.orig), and make sure that there are no
    1.92 +   failed patches (xxx# or xxx.rej). If there are, either you or me has
    1.93 +   made a mistake.
    1.94 +
    1.95 +   Unlike patches for the 2.6.x kernels, patches for the 2.6.x.y kernels
    1.96 +   (also known as the -stable kernels) are not incremental but instead apply
    1.97 +   directly to the base 2.6.x kernel.  Please read
    1.98 +   Documentation/applying-patches.txt for more information.
    1.99 +
   1.100 +   Alternatively, the script patch-kernel can be used to automate this
   1.101 +   process.  It determines the current kernel version and applies any
   1.102 +   patches found.
   1.103 +
   1.104 +		linux/scripts/patch-kernel linux
   1.105 +
   1.106 +   The first argument in the command above is the location of the
   1.107 +   kernel source.  Patches are applied from the current directory, but
   1.108 +   an alternative directory can be specified as the second argument.
   1.109 +
   1.110 + - If you are upgrading between releases using the stable series patches
   1.111 +   (for example, patch-2.6.xx.y), note that these "dot-releases" are
   1.112 +   not incremental and must be applied to the 2.6.xx base tree. For
   1.113 +   example, if your base kernel is 2.6.12 and you want to apply the
   1.114 +   2.6.12.3 patch, you do not and indeed must not first apply the
   1.115 +   2.6.12.1 and 2.6.12.2 patches. Similarly, if you are running kernel
   1.116 +   version 2.6.12.2 and want to jump to 2.6.12.3, you must first
   1.117 +   reverse the 2.6.12.2 patch (that is, patch -R) _before_ applying
   1.118 +   the 2.6.12.3 patch.
   1.119 +
   1.120 + - Make sure you have no stale .o files and dependencies lying around:
   1.121 +
   1.122 +		cd linux
   1.123 +		make mrproper
   1.124 +
   1.125 +   You should now have the sources correctly installed.
   1.126 +
   1.127 +SOFTWARE REQUIREMENTS
   1.128 +
   1.129 +   Compiling and running the 2.6.xx kernels requires up-to-date
   1.130 +   versions of various software packages.  Consult
   1.131 +   Documentation/Changes for the minimum version numbers required
   1.132 +   and how to get updates for these packages.  Beware that using
   1.133 +   excessively old versions of these packages can cause indirect
   1.134 +   errors that are very difficult to track down, so don't assume that
   1.135 +   you can just update packages when obvious problems arise during
   1.136 +   build or operation.
   1.137 +
   1.138 +BUILD directory for the kernel:
   1.139 +
   1.140 +   When compiling the kernel all output files will per default be
   1.141 +   stored together with the kernel source code.
   1.142 +   Using the option "make O=output/dir" allow you to specify an alternate
   1.143 +   place for the output files (including .config).
   1.144 +   Example:
   1.145 +     kernel source code:	/usr/src/linux-2.6.N
   1.146 +     build directory:		/home/name/build/kernel
   1.147 +
   1.148 +   To configure and build the kernel use:
   1.149 +   cd /usr/src/linux-2.6.N
   1.150 +   make O=/home/name/build/kernel menuconfig
   1.151 +   make O=/home/name/build/kernel
   1.152 +   sudo make O=/home/name/build/kernel modules_install install
   1.153 +
   1.154 +   Please note: If the 'O=output/dir' option is used then it must be
   1.155 +   used for all invocations of make.
   1.156 +
   1.157 +CONFIGURING the kernel:
   1.158 +
   1.159 +   Do not skip this step even if you are only upgrading one minor
   1.160 +   version.  New configuration options are added in each release, and
   1.161 +   odd problems will turn up if the configuration files are not set up
   1.162 +   as expected.  If you want to carry your existing configuration to a
   1.163 +   new version with minimal work, use "make oldconfig", which will
   1.164 +   only ask you for the answers to new questions.
   1.165 +
   1.166 + - Alternate configuration commands are:
   1.167 +	"make menuconfig"  Text based color menus, radiolists & dialogs.
   1.168 +	"make xconfig"     X windows (Qt) based configuration tool.
   1.169 +	"make gconfig"     X windows (Gtk) based configuration tool.
   1.170 +	"make oldconfig"   Default all questions based on the contents of
   1.171 +			   your existing ./.config file and asking about
   1.172 +			   new config symbols.
   1.173 +	"make silentoldconfig"
   1.174 +			   Like above, but avoids cluttering the screen
   1.175 +			   with questions already answered.
   1.176 +	"make defconfig"   Create a ./.config file by using the default
   1.177 +			   symbol values from arch/$ARCH/defconfig.
   1.178 +	"make allyesconfig"
   1.179 +			   Create a ./.config file by setting symbol
   1.180 +			   values to 'y' as much as possible.
   1.181 +	"make allmodconfig"
   1.182 +			   Create a ./.config file by setting symbol
   1.183 +			   values to 'm' as much as possible.
   1.184 +	"make allnoconfig" Create a ./.config file by setting symbol
   1.185 +			   values to 'n' as much as possible.
   1.186 +	"make randconfig"  Create a ./.config file by setting symbol
   1.187 +			   values to random values.
   1.188 +
   1.189 +   The allyesconfig/allmodconfig/allnoconfig/randconfig variants can
   1.190 +   also use the environment variable KCONFIG_ALLCONFIG to specify a
   1.191 +   filename that contains config options that the user requires to be
   1.192 +   set to a specific value.  If KCONFIG_ALLCONFIG=filename is not used,
   1.193 +   "make *config" checks for a file named "all{yes/mod/no/random}.config"
   1.194 +   for symbol values that are to be forced.  If this file is not found,
   1.195 +   it checks for a file named "all.config" to contain forced values.
   1.196 +   
   1.197 +	NOTES on "make config":
   1.198 +	- having unnecessary drivers will make the kernel bigger, and can
   1.199 +	  under some circumstances lead to problems: probing for a
   1.200 +	  nonexistent controller card may confuse your other controllers
   1.201 +	- compiling the kernel with "Processor type" set higher than 386
   1.202 +	  will result in a kernel that does NOT work on a 386.  The
   1.203 +	  kernel will detect this on bootup, and give up.
   1.204 +	- A kernel with math-emulation compiled in will still use the
   1.205 +	  coprocessor if one is present: the math emulation will just
   1.206 +	  never get used in that case.  The kernel will be slightly larger,
   1.207 +	  but will work on different machines regardless of whether they
   1.208 +	  have a math coprocessor or not. 
   1.209 +	- the "kernel hacking" configuration details usually result in a
   1.210 +	  bigger or slower kernel (or both), and can even make the kernel
   1.211 +	  less stable by configuring some routines to actively try to
   1.212 +	  break bad code to find kernel problems (kmalloc()).  Thus you
   1.213 +	  should probably answer 'n' to the questions for
   1.214 +          "development", "experimental", or "debugging" features.
   1.215 +
   1.216 +COMPILING the kernel:
   1.217 +
   1.218 + - Make sure you have at least gcc 3.2 available.
   1.219 +   For more information, refer to Documentation/Changes.
   1.220 +
   1.221 +   Please note that you can still run a.out user programs with this kernel.
   1.222 +
   1.223 + - Do a "make" to create a compressed kernel image. It is also
   1.224 +   possible to do "make install" if you have lilo installed to suit the
   1.225 +   kernel makefiles, but you may want to check your particular lilo setup first.
   1.226 +
   1.227 +   To do the actual install you have to be root, but none of the normal
   1.228 +   build should require that. Don't take the name of root in vain.
   1.229 +
   1.230 + - If you configured any of the parts of the kernel as `modules', you
   1.231 +   will also have to do "make modules_install".
   1.232 +
   1.233 + - Keep a backup kernel handy in case something goes wrong.  This is 
   1.234 +   especially true for the development releases, since each new release
   1.235 +   contains new code which has not been debugged.  Make sure you keep a
   1.236 +   backup of the modules corresponding to that kernel, as well.  If you
   1.237 +   are installing a new kernel with the same version number as your
   1.238 +   working kernel, make a backup of your modules directory before you
   1.239 +   do a "make modules_install".
   1.240 +   Alternatively, before compiling, use the kernel config option
   1.241 +   "LOCALVERSION" to append a unique suffix to the regular kernel version.
   1.242 +   LOCALVERSION can be set in the "General Setup" menu.
   1.243 +
   1.244 + - In order to boot your new kernel, you'll need to copy the kernel
   1.245 +   image (e.g. .../linux/arch/i386/boot/bzImage after compilation)
   1.246 +   to the place where your regular bootable kernel is found. 
   1.247 +
   1.248 + - Booting a kernel directly from a floppy without the assistance of a
   1.249 +   bootloader such as LILO, is no longer supported.
   1.250 +
   1.251 +   If you boot Linux from the hard drive, chances are you use LILO which
   1.252 +   uses the kernel image as specified in the file /etc/lilo.conf.  The
   1.253 +   kernel image file is usually /vmlinuz, /boot/vmlinuz, /bzImage or
   1.254 +   /boot/bzImage.  To use the new kernel, save a copy of the old image
   1.255 +   and copy the new image over the old one.  Then, you MUST RERUN LILO
   1.256 +   to update the loading map!! If you don't, you won't be able to boot
   1.257 +   the new kernel image.
   1.258 +
   1.259 +   Reinstalling LILO is usually a matter of running /sbin/lilo. 
   1.260 +   You may wish to edit /etc/lilo.conf to specify an entry for your
   1.261 +   old kernel image (say, /vmlinux.old) in case the new one does not
   1.262 +   work.  See the LILO docs for more information. 
   1.263 +
   1.264 +   After reinstalling LILO, you should be all set.  Shutdown the system,
   1.265 +   reboot, and enjoy!
   1.266 +
   1.267 +   If you ever need to change the default root device, video mode,
   1.268 +   ramdisk size, etc.  in the kernel image, use the 'rdev' program (or
   1.269 +   alternatively the LILO boot options when appropriate).  No need to
   1.270 +   recompile the kernel to change these parameters. 
   1.271 +
   1.272 + - Reboot with the new kernel and enjoy. 
   1.273 +
   1.274 +IF SOMETHING GOES WRONG:
   1.275 +
   1.276 + - If you have problems that seem to be due to kernel bugs, please check
   1.277 +   the file MAINTAINERS to see if there is a particular person associated
   1.278 +   with the part of the kernel that you are having trouble with. If there
   1.279 +   isn't anyone listed there, then the second best thing is to mail
   1.280 +   them to me (torvalds@osdl.org), and possibly to any other relevant
   1.281 +   mailing-list or to the newsgroup.
   1.282 +
   1.283 + - In all bug-reports, *please* tell what kernel you are talking about,
   1.284 +   how to duplicate the problem, and what your setup is (use your common
   1.285 +   sense).  If the problem is new, tell me so, and if the problem is
   1.286 +   old, please try to tell me when you first noticed it.
   1.287 +
   1.288 + - If the bug results in a message like
   1.289 +
   1.290 +	unable to handle kernel paging request at address C0000010
   1.291 +	Oops: 0002
   1.292 +	EIP:   0010:XXXXXXXX
   1.293 +	eax: xxxxxxxx   ebx: xxxxxxxx   ecx: xxxxxxxx   edx: xxxxxxxx
   1.294 +	esi: xxxxxxxx   edi: xxxxxxxx   ebp: xxxxxxxx
   1.295 +	ds: xxxx  es: xxxx  fs: xxxx  gs: xxxx
   1.296 +	Pid: xx, process nr: xx
   1.297 +	xx xx xx xx xx xx xx xx xx xx
   1.298 +
   1.299 +   or similar kernel debugging information on your screen or in your
   1.300 +   system log, please duplicate it *exactly*.  The dump may look
   1.301 +   incomprehensible to you, but it does contain information that may
   1.302 +   help debugging the problem.  The text above the dump is also
   1.303 +   important: it tells something about why the kernel dumped code (in
   1.304 +   the above example it's due to a bad kernel pointer). More information
   1.305 +   on making sense of the dump is in Documentation/oops-tracing.txt
   1.306 +
   1.307 + - If you compiled the kernel with CONFIG_KALLSYMS you can send the dump
   1.308 +   as is, otherwise you will have to use the "ksymoops" program to make
   1.309 +   sense of the dump.  This utility can be downloaded from
   1.310 +   ftp://ftp.<country>.kernel.org/pub/linux/utils/kernel/ksymoops.
   1.311 +   Alternately you can do the dump lookup by hand:
   1.312 +
   1.313 + - In debugging dumps like the above, it helps enormously if you can
   1.314 +   look up what the EIP value means.  The hex value as such doesn't help
   1.315 +   me or anybody else very much: it will depend on your particular
   1.316 +   kernel setup.  What you should do is take the hex value from the EIP
   1.317 +   line (ignore the "0010:"), and look it up in the kernel namelist to
   1.318 +   see which kernel function contains the offending address.
   1.319 +
   1.320 +   To find out the kernel function name, you'll need to find the system
   1.321 +   binary associated with the kernel that exhibited the symptom.  This is
   1.322 +   the file 'linux/vmlinux'.  To extract the namelist and match it against
   1.323 +   the EIP from the kernel crash, do:
   1.324 +
   1.325 +		nm vmlinux | sort | less
   1.326 +
   1.327 +   This will give you a list of kernel addresses sorted in ascending
   1.328 +   order, from which it is simple to find the function that contains the
   1.329 +   offending address.  Note that the address given by the kernel
   1.330 +   debugging messages will not necessarily match exactly with the
   1.331 +   function addresses (in fact, that is very unlikely), so you can't
   1.332 +   just 'grep' the list: the list will, however, give you the starting
   1.333 +   point of each kernel function, so by looking for the function that
   1.334 +   has a starting address lower than the one you are searching for but
   1.335 +   is followed by a function with a higher address you will find the one
   1.336 +   you want.  In fact, it may be a good idea to include a bit of
   1.337 +   "context" in your problem report, giving a few lines around the
   1.338 +   interesting one. 
   1.339 +
   1.340 +   If you for some reason cannot do the above (you have a pre-compiled
   1.341 +   kernel image or similar), telling me as much about your setup as
   1.342 +   possible will help. 
   1.343 +
   1.344 + - Alternately, you can use gdb on a running kernel. (read-only; i.e. you
   1.345 +   cannot change values or set break points.) To do this, first compile the
   1.346 +   kernel with -g; edit arch/i386/Makefile appropriately, then do a "make
   1.347 +   clean". You'll also need to enable CONFIG_PROC_FS (via "make config").
   1.348 +
   1.349 +   After you've rebooted with the new kernel, do "gdb vmlinux /proc/kcore".
   1.350 +   You can now use all the usual gdb commands. The command to look up the
   1.351 +   point where your system crashed is "l *0xXXXXXXXX". (Replace the XXXes
   1.352 +   with the EIP value.)
   1.353 +
   1.354 +   gdb'ing a non-running kernel currently fails because gdb (wrongly)
   1.355 +   disregards the starting offset for which the kernel is compiled.
   1.356 +