The question is how to execute aout-format binary (I mean old format which for example used on FreeBSD before it has migrated to ELF) on Linux system. Is there a possibility to do so without extra coding (is there some existing solution)? Probably it should be in form of kernel module or patch for the Linux kernel. Another solution could be user-space launcher (may be even run-time linker). I have searched for something similar but was unable to found something. I have not yet checked difference in system calls interfaces, if you have some comments about that, you are welcome to provide them.
P.S. I know that writing user-space launcher for aout static binary is quite trivial but the question is about some existing solution.
Check for CONFIG_BINFMT_AOUT in your kernel config.
If your kernel has /proc/config.gz:
zgrep CONFIG_BINFMT_AOUT /proc/config.gz
On Ubuntu and the like:
grep CONFIG_BINFMT_AOUT /boot/config-$(uname -r)
Kernel option was CONFIG_BINFMT_AOUT, not sure if it's still around or necessary.
Related
I am looking if there is any API through which we can get OS distribution name and version from a Linux kernel module?
For example, I am using SLES 12 service pack 4. This information is present in /etc/os-release file. I want to know if there is any way to get this information from kernel code.
linux:/ # cat /etc/os-release
NAME="SLES"
VERSION="12-SP4"
VERSION_ID="12.4"
PRETTY_NAME="SUSE Linux Enterprise Server 12 SP4"
ID="sles"
ANSI_COLOR="0;32"
CPE_NAME="cpe:/o:suse:sles:12:sp4"
linux:/ #
There's no kernel API for detecting the current OS distribution, simply because it's not really needed. The Linux kernel itself is distribution-agnostic, and it couldn't care less which distribution is being run on top of it (having the kernel depend on what's being run on top of it wouldn't make much sense).
If you really want, you can open, read and parse the file yourself from kernel space. See more in this other post for an example, and in particular this answer for modern kernels. In any case, remember that filesystem interaction from kernel space is generally discouraged, and could easily lead to bugs and compromise the security of the kernel if done wrong, so be careful.
If you are developing a kernel module, I would suggest you to parse the /etc/os-release file from userspace when compiling/installing the module and use a set of #defines, or even module parameters. In any case, you should ask yourself why you need this information in kernel code in the first place, as you really shouldn't.
I'm new to linux.
I want to edit "This kernel requires an x86-64 CPU, but only detected an i686 CPU.Unable to boot - Please use a kernel appropriate for your CPU" from vmlinuz file(RHEL6.0)
If editing this message is possible Can anyone suggest a way to modify message.
The message is in /arch/.../boot/cpu.c in the linux kernel source. You can modify it, recompile the kernel and install the new kernel.
you can grep the kernel sources for this string, and replace it with whatever you want, and then recompile your kernel. Usually replacing a string directly inside a binary (though it is technically possible), especially for the kernel is a very bad idea
I was wondering if anyone knows a way to prevent building unneeded device drivers when building the 2.6.32 kernel in Ubuntu 10.4 on VB? The reason I ask is we have to do a project for my operating systems class that involves adding some system calls to the kernel and the instructions say that after you add your call you need to rebuild the kernel(which takes like 3 freakin hours) and I know its because Ubuntu doesn't know which device drivers I need on so it builds them all so I'm wondering if there is a way to have it only build the ones I need? and if so how to go about that? or if anyone knows a way of being able to test added system calls without rebuilding the whole kernel(as this is really the issue)?
Thanks in advance
You can manually change kernel configuration with with rather friendly menus. Just type make nconfig (or menuconfig, or xconfig for gui). And remove drivers that you don't need.
Here are some links that may help you:
http://www.linuxtopia.org/online_books/linux_kernel/kernel_configuration/ch05.html
http://www.gentoo.org/doc/en/handbook/handbook-x86.xml?part=1&chap=7
http://kernel.xc.net/
Also, do you have a multicore processor? If so do you use advantages of it like here?
UPDATE: I've remembered a faster way. You can wrap a new syscall in module, thus avoiding recompilation of the whole kernel. Look here and there.
You can easily find everything here with the help of Google, though.
I would like to expose the settings and statistics of my program in a 'everything is a file' manner - sort of how /proc/ and /sys/ works.
As an example, imagine for a moment that apache2 had this type of interface. You would then be able to do something like this (hypothetical):
cd /apache2/virtual_hosts
mkdir 172.20.30.50
cd 172.20.30.50
echo '/www/example1' > DocumentRoot
echo 'www.example1.com' > ServerName
echo 1 > control/enabled
cat control/status
enabled true
uptime 4080
hits 0
Now, are there any tutorials or similar on how to do this? I'm mainly looking for techniques for 'pretending to be a file or dir'. I'm on linux, POSIX or other more portable method would be preferable, but not required.
On Linux, have a look at Fuse: implement a fully functional filesystem in a userspace program.
Simple library API
Simple installation (no need to patch or recompile the kernel)
Secure implementation
Userspace - kernel interface is very efficient
Usable by non privileged users
Runs on Linux kernels 2.4.X and 2.6.X
Has proven very stable over time
Look at compatible platforms here. In terms of tutorial, one of good ones I've came across is here.
In addition to FUSE, another solution is to export a 9p filesystem. wmii does this, for example.
Perhaps the way to do this is just use "real" files and use a change-notification library (inotify preferred) to detect when they change and update your behaviour accordingly.
The /proc and /sys are for kernel-user communication and not really intended for userspace programs' IPC - you're expected to use named pipes, sockets, shared memory etc for that.
(ab)using FUSE is not really a nice idea in this case, I think.
When I use top the iowait on the host is really high.
iostat tells me which disk is utilized more but I want to find out which process is the culprit?
I am trying to find this out on a red hat linux host. Any suggestions.
EDIT: My linux flavor does not either have atop or ntop and since building kernel is not an option for me don't ask me why :) (since this is not my personal box). are there any other alternatives
I usually use atop. There's a really good article at Debian Package A Day about it. It does require kernel patching (although Ubuntu already has the patch applied, I'm not sure about any other distributions.)
Use iotop.
Or you can get it standalone, it's a simple python script which requires a recent kernel (can't remember, but at least > 2.6.20)