Tag Info

New answers tagged

1

I know it's a bit late but this may be the source of your problem. : [Mon May 12 18:33:00.589 2014] spl: error reading image system.dtb, err - -1 A wrong device tree can lead to this kind of errors.


2

The boot fails for the same reason as in the mentioned question - just booting a kernel without anything else doesn't do much good. You must provide a disk. Or an initrd image. But just enabling initrd doesn't give you an initrd image magically. You need to prepare one and provide it to qemu like so: qemu-system-i386 -kernel <your kernel> -initrd ...


1

There is a system.conf configuration option, DefaultControllers, that controls which cgroup hierarchies are attached to. By default it's cpu. I set it to null and /proc/$$$/cgroup no longer lists the getty process under cpuacct,cpu, and the test program works. Why the same configuration file -- I was using the default which is in use on both systems -- ...


1

If you have space, please back up the disk as a whole (e.g. dd if=/dev/sdb of=disk.img bs=1M), before running random programs like fsck on things that you don't think are valid partitions :p. I'm not saying you've damaged it, but there's a very good chance of doing so while experimenting. The partition table shown by parted & the kernel looks ...


1

It's done! I forgot to replace old ipv6.ko module with the new one.


1

You can't. There used to be a bdi_add_flusher_task() that got dropped around 2009 iirc. Also it was originally intended for filesystems but I don't think you're writing a filesystem :-) But.. I can't see why you would have 100% cpu, that doesn't make sense - unless you're doing 10+ GB/s writes to some fantastic $100,000 array. RAM is so much faster than ...


0

As goldilocks pointed out in comments, there's no "reliable way" to do this. There are many ways the driver could be broken that even if you manage to import it, it may present the same or other problems. The best way to go is to report a bug, explaining the problem and what kernel you didn't found the problem, so they can review the relevant code and revert ...


1

When patch fails, it saves the bit that fails in a .rej file. Since you have already run it, you will have those files already. You can find them with find . -name \*.rej. To get an actual log, you'd have to pipe the output and errors from patch through a pager, such as less: gzip -dc patch.gz | patch -p1 2>&1 | less. patch does not maintain any log ...


1

If you are using Buildroot, select a uImage kernel output file (BR2_LINUX_KERNEL_UIMAGE=y). Compile normally and Buildroot will also create a vmlinux file in an intermediate step. While the vmlinux file is not copied to the /output/images directory it can be found in ./output/build/linux-custom/(or by running find . -name "vmlinux").


0

You have to make sure the toolchain is aware of the kernel source tree you're working with. You have to set the environment variable KERNEL_TREE to the directory where your sources are. See http://unix.stackexchange.com/a/147700/37538 for reference.


3

You could use make oldconfig. After you copy the 2.6 .config file, this make option will prompt to you for options in the current kernel source that are not found in the file. However, you will have to deal with choosing options out of the context, being difficult to give the right answer Further reading: What does “make oldconfig” do exactly - Linux ...


1

Officially that is not recommended but you can do it and that's what I do. You have to go through the build menus very carefully. Things change (especially from 2.6 to 3.2), and without review you might end up with a lot of features and drivers you don't need. I don't know of a migration guide but for kernel tweaking you may want to look at Linux Kernel ...


1

Despite what file says, it turns out to be debugging symbols after all. A thread about this on the LKML led me to try: make INSTALL_MOD_STRIP=1 modules_install And low and behold, a comparison from within the /lib/modules/x.x.x directory; before: > ls -hs kernel/crypto/anubis.ko 112K kernel/crypto/anubis.ko And after: > ls -hs ...


0

After your make oldconfig, do a make vmlinuz. I think you'll find that the pre-compiled kernel is a "executable bzImage", which means it's compressed on-disk. If you watch boot messages closely, you'll see it uncompressing the kernel very early on in the process.


0

As derobert pointed out, hda -> sda is an expected change since a long time. The raid array name change was weird, but it resolved itself in the end. I tried to boot from a live cd distribution, mount the raid array, mount the boot partition, then apt-get install the new kernel. This procedure produced an error because I didn't know that I should have ...


1

kernel.org has some excellent documentation on Applying Patches to the Linux Kernel. Essentially, you use the patch command. Once you have acquired the patch (here called patchfile), place it in your build directory and then issue the command: patch -p <num> < patchfile where <num> is the number of leading slashes to remove from the ...


1

the new ATA driver in the kernel use /dev/sda, the old drivers are still supported but you will have to edit your kernel by chrooting into you system using a livecd. Device drivers ---> <*> ATA/ATAPI/MFM/TLL support (deprecated) <*> Serial ATA and Parallel ATA drivers ---> For chrooting I always use the gentoo minimal installation cd ...


0

In reality, nearly each Linux distribution maintains its own kernel The Kernels maintained by distro's are "flavoured" versions of the vanilla kernel (kernel.org). This means the programmers working on the distro added their own code and fixes for modules, but it really is just the same kernel underneath. For distributions, it is dangerous to ...


1

Compression algorithms are declared in lib/decompress.c. Gzip is defined in lib/decompress_inflate.c and doesn't get any special status; it'll only be there if CONFIG_DECOMPRESS_GZIP is y when the kernel is compiled. The list of available compression algorithms is the compressed_formats structure. Since it's defined as static, it isn't available in other ...


0

Is is as simple as comparing /proc/filesystems with lsmod? No: $ comm -31 <(lsmod | awk 'NR!=1 {print $1}' |sort) \ <(</proc/filesystems awk '{print $NF}' |sort) | fmt anon_inodefs autofs bdev cgroup cpuset debugfs devpts devtmpfs ext2 ext3 fuseblk fusectl hugetlbfs mqueue nfs4 pipefs proc pstore ramfs rootfs rpc_pipefs securityfs ...


1

The documentation for the pam_loginuid PAM module gives a pretty good hint: The pam_loginuid module sets the loginuid process attribute for the process that was authenticated. This is necessary for applications to be correctly audited. This PAM module should only be used for entry point applications like: login, sshd, gdm, vsftpd, crond and atd. ...


2

You might see some information from here. Generally, a file descriptor is an index for an entry in a kernel-resident data structure containing the details of all open files. In POSIX, this data structure is called a file descriptor table, and each process has its own file descriptor table. The user application passes the abstract key to the kernel through a ...


3

A FILE structure in C is typically called the file handle and is a bit of abstraction around a file descriptor: The data type FILE is a structure that contains information about a file or specified data stream. It includes such information as a file descriptor, current position, status flags, and more. It is most often used as a pointer to a ...


1

Ok, so here's the way the boot process works: firmware > bootloader maybe > kernel ${parameters} > initramfs > userspace maybe On a redhat installation disk their dracut system of scripts is what builds and constitutes initramfs and their anaconda installation system constitutes the final userspace. It is udev that handles the device setup - as in, it ...


2

go into the BIOS of the host and rearrange the order of the hard drives and removable drives. This will adjust the order as it appears to the Linux kernel.



Top 50 recent answers are included