| blob | f5f0d10c0605fd4cd0008f82a912078a0c72b697 |
2 Linux kernel release 1.3.xx
4 These are the release notes for linux version 1.3. Read them carefully,
5 as they tell you what this is all about, explain how to install the
6 kernel, and what to do if something goes wrong.
8 Linux version 1.3 is a DEVELOPMENT kernel, and not intended for general
9 public use. Different releases may have various and sometimes severe
10 bugs. It is *strongly* recommended that you back up the previous kernel
11 before installing any new 1.3.xx release.
13 If you need to use a proven and stable Linux kernel, please
14 use 1.0.9 or 1.2.13. All features found in the 1.3.xx releases
15 will be contained in 2.0 when the code base has stabilized again.
17 If you decide to use 1.3, it might be a good idea to follow the kernel
18 channel, available as newsgroup (under linux.dev.kernel) and as mailing
19 list. To subscribe to the latter, e-mail majordomo@vger.rutgers.edu,
20 and put in the body of the message "subscribe linux-kernel" or
21 "subscribe linux-kernel-digest" for a daily digest of the mailing list
22 (it is a high-traffic list). Note that getting subscribed or unsubscribed
23 sometimes takes days or even weeks.
25 WHAT IS LINUX?
27 Linux is a Unix clone for 386/486-based PCs written from scratch by
28 Linus Torvalds with assistance from a loosely-knit team of hackers
29 across the Net. It aims towards POSIX compliance.
31 It has all the features you would expect in a modern fully-fledged
32 Unix, including true multitasking, virtual memory, shared libraries,
33 demand loading, shared copy-on-write executables, proper memory
34 management and TCP/IP networking.
36 It is distributed under the GNU General Public License - see the
37 accompanying COPYING file for more details.
39 DOCUMENTATION:
41 - there is a lot of documentation available both in electronic form on
42 the internet and in books, both Linux-specific and pertaining to
43 general UNIX questions. I'd recommend looking into the documentation
44 subdirectories on any Linux ftp site for the LDP (Linux Documentation
45 Project) books. This README is not meant to be documentation on the
46 system: there are much better sources available.
48 - There are various readme's in the kernel Documentation/ subdirectory:
49 these are mainly used for kernel developers and some very kernel-specific
50 installation notes for some drivers for example.
52 INSTALLING the kernel:
54 - If you install the full sources, do a
56 cd /usr/src
57 gzip -cd linux-1.3.XX.tar.gz | tar xfv -
59 to get it all put in place. Replace "XX" with the version number of the
60 latest kernel.
62 - You can also upgrade between 1.3.xx releases by patching. For a large
63 set of revisions, it is not worth the effort since the full set of
64 patches is bigger than a new kernel distribution. Instead, get the
65 latest full source archive and install as above. Then, get all newer
66 patch files, and do
68 cd /usr/src
69 gzip -cd patchXX.gz | patch -p0
71 (repeat xx for all versions bigger than the version of your current
72 source tree, _in_order_) and you should be ok. You may want to remove
73 the backup files (xxx~ or xxx.orig), and make sure that there are no
74 failed patches (xxx# or xxx.rej). If there are, either you or me has
75 made a mistake.
77 Alternatively, the script patch-kernel can be used to automate this
78 process. It determines the current kernel version and applies any
79 patches found.
81 cd /usr/src
82 linux/scripts/patch-kernel
84 The default directory for the kernel source is /usr/src/linux, but
85 can be specified as the first argument. Patches are applied from
86 the current directory, but an alternative directory can be specified
87 as the second argument.
89 - make sure your /usr/include/linux and /usr/include/asm directories
90 are just symlinks to the kernel sources:
92 cd /usr/include
93 rm -rf linux
94 rm -rf asm
95 ln -s /usr/src/linux/include/linux linux
96 ln -s /usr/src/linux/include/asm-i386 asm
98 - make sure you have no stale .o files and dependencies lying around:
100 cd /usr/src/linux
101 make mrproper
103 You should now have the sources correctly installed.
105 CONFIGURING the kernel:
107 - do a "make config" to configure the basic kernel. "make config"
108 needs bash to work: it will search for bash in $BASH, /bin/bash and
109 /bin/sh (in that order), so hopefully one of those is correct.
111 - Alternate configuration commands are:
112 "make menuconfig" Text based color menus, radiolists & dialogs.
113 "make xconfig" X windows based configuration tool.
116 NOTES on "make config":
117 - having unnecessary drivers will make the kernel bigger, and can
118 under some circumstances lead to problems: probing for a
119 nonexistent controller card may confuse your other controllers
120 - compiling the kernel with "-m486" for a number of 486-specific
121 will result in a kernel that does NOT work on a 386. The
122 kernel will detect this on bootup, and give up.
123 - A kernel with math-emulation compiled in will still use the
124 coprocessor if one is present: the math emulation will just
125 never get used in that case. The kernel will be slightly larger,
126 but will work on different machines regardless of whether they
127 have a math coprocessor or not.
128 - the "kernel hacking" configuration details usually result in a
129 bigger or slower kernel (or both), and can even make the kernel
130 less stable by configuring some routines to actively try to
131 break bad code to find kernel problems (kmalloc()). Thus you
132 should probably answer 'n' to the questions for a "production"
133 kernel.
135 - Check the top Makefile for further site-dependent configuration
136 (default SVGA mode etc).
138 - Finally, do a "make dep" to set up all the dependencies correctly.
140 COMPILING the kernel:
142 - make sure you have gcc-2.5.8 or newer available. It seems older gcc
143 versions can have problems compiling newer versions of linux. If you
144 upgrade your compiler, remember to get the new binutils package too
145 (for as/ld/nm and company). Do not use gcc-2.6.0; it has a few serious
146 bugs. Some problems have been reported occasionally with 2.6.3 as well,
147 so use that version at your own risk. Specifying "-fno-strength-reduce"
148 in the CFLAGS of the main Makefile may help to work around a known
149 bug in gcc.
151 - do a "make zImage" to create a compressed kernel image. If you want
152 to make a bootdisk (without root filesystem or lilo), insert a floppy
153 in your A: drive, and do a "make zdisk". It is also possible to do
154 "make zlilo" if you have lilo installed to suit the kernel makefiles,
155 but you may want to check your particular lilo setup first.
157 - keep a backup kernel handy in case something goes wrong. This is
158 especially true for the development releases, since each new release
159 contains new code which has not been debugged.
161 - In order to boot your new kernel, you'll need to copy the kernel
162 image (found in /usr/src/linux/arch/i386/boot/zImage after compilation)
163 to the place where your regular bootable kernel is found.
165 For some, this is on a floppy disk, in which case you can "cp
166 /usr/src/linux/arch/i386/boot/zImage /dev/fd0" to make a bootable
167 floppy.
169 If you boot Linux from the hard drive, chances are you use LILO which
170 uses the kernel image as specified in the file /etc/lilo.conf. The
171 kernel image file is usually /vmlinuz, or /zImage, or /etc/zImage.
172 To use the new kernel, copy the new image over the old one (save a
173 backup of the original!). Then, you MUST RERUN LILO to update the
174 loading map!! If you don't, you won't be able to boot the new kernel
175 image.
177 Reinstalling LILO is usually a matter of running /sbin/lilo.
178 You may wish to edit /etc/lilo.conf to specify an entry for your
179 old kernel image (say, /vmlinux.old) in case the new one does not
180 work. See the LILO docs for more information.
182 After reinstalling LILO, you should be all set. Shutdown the system,
183 reboot, and enjoy!
185 If you ever need to change the default root device, video mode,
186 ramdisk size, etc. in the kernel image, use the 'rdev' program (or
187 alternatively the LILO boot options when appropriate). No need to
188 recompile the kernel to change these parameters.
190 - reboot with the new kernel and enjoy.
192 IF SOMETHING GOES WRONG:
194 - if you have problems that seem to be due to kernel bugs, please mail
195 them to me (Linus.Torvalds@Helsinki.FI), and possibly to any other
196 relevant mailing-list or to the newsgroup. The mailing-lists are
197 useful especially for SCSI and NETworking problems, as I can't test
198 either of those personally anyway.
200 - In all bug-reports, *please* tell what kernel you are talking about,
201 how to duplicate the problem, and what your setup is (use your common
202 sense). If the problem is new, tell me so, and if the problem is
203 old, please try to tell me when you first noticed it.
205 - if the bug results in a message like
207 unable to handle kernel paging request at address C0000010
208 Oops: 0002
209 EIP: 0010:XXXXXXXX
210 eax: xxxxxxxx ebx: xxxxxxxx ecx: xxxxxxxx edx: xxxxxxxx
211 esi: xxxxxxxx edi: xxxxxxxx ebp: xxxxxxxx
212 ds: xxxx es: xxxx fs: xxxx gs: xxxx
213 Pid: xx, process nr: xx
214 xx xx xx xx xx xx xx xx xx xx
216 or similar kernel debugging information on your screen or in your
217 system log, please duplicate it *exactly*. The dump may look
218 incomprehensible to you, but it does contain information that may
219 help debugging the problem. The text above the dump is also
220 important: it tells something about why the kernel dumped code (in
221 the above example it's due to a bad kernel pointer)
223 - You can use the "ksymoops" program to make sense of the dump. Find
224 the C++ sources under the scripts/ directory to avoid having to do
225 the dump lookup by hand:
227 - in debugging dumps like the above, it helps enormously if you can
228 look up what the EIP value means. The hex value as such doesn't help
229 me or anybody else very much: it will depend on your particular
230 kernel setup. What you should do is take the hex value from the EIP
231 line (ignore the "0010:"), and look it up in the kernel namelist to
232 see which kernel function contains the offending address.
234 To find out the kernel function name, you'll need to find the system
235 binary associated with the kernel that exhibited the symptom. This is
236 the file 'linux/vmlinux'. To extract the namelist and match it against
237 the EIP from the kernel crash, do:
239 nm vmlinux | sort | less
241 This will give you a list of kernel addresses sorted in ascending
242 order, from which it is simple to find the function that contains the
243 offending address. Note that the address given by the kernel
244 debugging messages will not necessarily match exactly with the
245 function addresses (in fact, that is very unlikely), so you can't
246 just 'grep' the list: the list will, however, give you the starting
247 point of each kernel function, so by looking for the function that
248 has a starting address lower than the one you are searching for but
249 is followed by a function with a higher address you will find the one
250 you want. In fact, it may be a good idea to include a bit of
251 "context" in your problem report, giving a few lines around the
252 interesting one.
254 If you for some reason cannot do the above (you have a pre-compiled
255 kernel image or similar), telling me as much about your setup as
256 possible will help.
258 - alternately, you can use gdb on a running kernel. (read-only; i.e. you
259 cannot change values or set break points.) To do this, first compile the
260 kernel with -g; edit arch/i386/Makefile appropriately, then do a "make
261 clean". You'll also need to enable CONFIG_PROC_FS (via "make config").
263 After you've rebooted with the new kernel, do "gdb vmlinux /proc/kcore".
264 You can now use all the usual gdb commands. The command to look up the
265 point where your system crashed is "l *0xXXXXXXXX". (Replace the XXXes
266 with the EIP value.)
268 gdb'ing a non-running kernel currently fails because gdb (wrongly)
269 disregards the starting offset for which the kernel is compiled.