Tag Info

New answers tagged

0

Basically this logging is triggered if the CPU scheduler hasn't switched to the process in the given amount of time, with a couple of exceptions. I don't clearly understand the exceptions; FWIW the comments on the cases in the code are: Also, skip vfork and any other user process that freezer should skip. Also, when a freshly created task is scheduled ...


0

In the drivers Makefile, I added these That's not the way to go about this. See here -- you can probably skip to step #3. then tried to compile, but I get this error From your cut and paste, it looks like you ran make before make modules and nothing happened. That's no good. make should actually build the kernel, which may or may not be ...


2

In order to read /proc/[pid]/mem, a process must now PTRACE_ATTACH to it. A commonly available utility that does this is gdb Pick a running process (in my case I just opened cat in another window), then attach gdb to that process: [root@qemu ~]# gdb --pid 423 #MORE OUTPUT 0xb771dbac in __kernel_vsyscall () As part of its output while loading symbols, gdb ...


0

dmesg -w > yourlogfile & The above command will follow the dmesg log and print it to a file. Let it run until the kernel panics and check the log after booting with tail yourlogfile It may not give you the smoking gun, but it is a place to start.


1

See Apple's promo page for info about Thunderbolt. It also describes how you can chain devices together instead of needing a hub or similar. My impression is that hot-plugging such daisy-chained devices are what are not supported by the linux commit; only single devices, whatever type they are.


4

As it so happens, there is another significant interface with the kernel: the /proc and /sys virtual filesystems. While they do not hold regular files, their contents are direct gateways to the kernel: to act on them is to act directly on kernel-allocated memory. For instance, if you want to drop all memory caches, you may use... echo 3 > ...


1

It depends. What do you mean by 'verify'? If you just want to monitor what syscalls some process is triggering then it's possible.. usually.... But if you feel like digging deeper then you're in trouble... I have not heard of any tools that could do that. You can use strace to see what syscalls some particular process is firing. Of course you'll have to ...


4

I will try to answer questions as briefly as possible. The questions you are asking are usually addressed in introductory operating systems courses at universities but I will assume you have not taken such a course. Memory isolation for userspace processes is very desirable - not only to protect the kernel from malicious userspace programs, but also to ...


0

I got VGA passthrough working with an NVIDIA GTX 760 using KVM as a hypervisor with vfio-vga; I have never tried it with Virtualbox. It was a pain, but works well after getting the configuration right. KVM is just as convenient as Virtualbox for quick VMs from your desktop and you might consider it as another option. This thread has tons of information on ...


1

On systems with udev (that's more or less all for several years) every recognized device has an entry in /dev. These entries can be deleted, though. But (I guess) without a device file it is not possible for a userland process to access a device. Of course, for a device file to work it just needs the correct major and minor number. It doesn't matter what ...


2

Regarding Village's Bounty, all I can find is an archived mailing list post from August, 2011 stating that the preferred default before kernel commit r225076 was 32 GB, and up to 32 Slices for a Total of 1 TB of Ram (32^2 = 1024) After r225076 the limit per slice was removed, but the thread goes on to say that the practical limit should still be considered ...


0

I would suggest to do things in the "Linux way", which means for me, as less automatic configuration as possible. I would download the required toolchain from the distributor’s site, and set a .CC file in your project's working dir. Your CC file should look something like this: export PATH=< PATH-TO-TOOLCHAIN-BIN-FOLDER>:${PATH} export ...


0

Yes, it does support hardware entropy generators out of the box. This is necessary for high load SSL servers with lots of simultaneous connections initiated per second (Gmail, Facebook, Microsoft, etc). It is really not necessary for home servers or small organization servers. Keep in mind hardware entropy generators usually use PCI interfaces, nothing fancy ...


0

Debian now includes cross-toolchains officially, though they won't be part of the next Stable release (8.0). Virtual packages are provided.


1

By default, you should have user mode networking as explained in the docs: By default QEMU will create a SLiRP user network backend and an appropriate virtual network device for the guest (eg an E1000 PCI card for most x86 PC guests) If you want something better, consider tap networking: Add something like the following to your qemu command: -netdev ...


0

The source tarball is available from: HERE


2

You can use either dm-snapshot or NBD in copy-on-write mode. The dm-snapshot solution is provided here (sorry for not repeating it): https://raid.wiki.kernel.org/index.php/Recovering_a_failed_software_RAID#Making_the_harddisks_read-only_using_an_overlay_file As for NBD, you can install nbd-server and nbd-client, and then use it like this: mount ...


0

Here's evidence of one that we just ran into on a live system: $ sudo ps axf -O wchan 31103 - R ? 00:00:00 /bin/bash /usr/local/bin/monitorcron taskA 31104 exit X ? 00:00:00 \_ [su]


2

You can use the modinfo command to give you all kind of information about a given module. For example: $ modinfo bluetooth filename: /lib/modules/3.17.4-1-ARCH/kernel/net/bluetooth/bluetooth.ko.gz alias: net-pf-31 license: GPL version: 2.19 description: Bluetooth Core ver 2.19 author: Marcel Holtmann ...


1

Turns out to be a false warning. To quote from the (now closed) kernel bug report: Alfred Krohmer: If I unterstand your patch correctly it just removes the warning, but it won't actually fix the driver crash this bug report was submitted for. So why mark it as resolved? Emmanuel Grumbach: There is no real bug. The commit message explains this. ...


0

Have a look at quilt (http://en.wikipedia.org/wiki/Quilt_%28software%29) - that's the tool for managing and forward porting patches. Happily used by many distribution package maintainers, including maintainers of distribution kernels. However there may be significant differences between 3.0.35 and 3.10.17 - forward porting may not be straightforward. The ...


0

Quoting Linux Device Drivers, 3rd Edition. I didn't use the Quote button, as I wanted to bold the options Except where specified otherwise, all of these options are found under the "kernel hacking" menu in whatever kernel configuration tool you prefer. Note that some of these options are not supported by all architectures. CONFIG_DEBUG_KERNEL This option ...


0

If you really need an rt-kernel you will have to wait for a patch to become available for the version of kernel you seek. I would advise however to avoid the hassle of preparing an rt-kernel for your professional audio needs. You can easily get extremely low (<10msec) latencies even with the preemptive kernel. This should be more than enough.


0

I'm not sure if this is the same problem I had or not, but I recently upgraded and old internal DNS server system from OpenBSD 3.8 to 5.6, and I lost the ability to resolve hosts with ping, but the host command was working. Turned out I had to add 127.0.0.1/8 to the match-clients directive in named in addition to the 192.168.0.0/16 which I already had there, ...


1

In both cases those are not primarily warnings that break the build. For the kernel compilation: error: 'struct run_vap' has no member named 'beacon_mbuf' the compiler tells you, that the code is trying to access something that isn't there. This may have many reasons, but generally it suggests that the code is broken. Maybe you are trying to compile a ...


2

Well, from the apt-cache search linux-headers the cause should be obvious: Simply stated, the package isn't available in Ubuntu 14.10. You should instead install linux-headers-3.16.0-24, linux-headers-3.16.0-25 or save yourself from all problems and use linux-headers-generic which is a meta-package that depends on the latest available kernels (going by your ...


0

Yes, a Linux compiler can be considered as a "super-set" of a bare-metal compiler: it can compile both Linux userspace applications, and bare-metal applications (the Linux kernel, bootloaders, etc.).


0

When building a custom kernel, most people DO NOT BUILD the source inside the source tarball either. Please consider using a soft link: As you can see the /usr/src/linux link points to /usr/src/linux-3.12.21-gentoo-r1 directory Thus, when issuing the make command, the resulting output is put in the top level directory of the build tree, which doesn't ...


0

To build a kernel module, you need more than an unpacked kernel source. You need some companion programs and header files generated during the kernel build. The makefile is looking for modpost in the right place but it isn't there yet. Distributions typically ship this in a package called linux-headers-VERSION or kernel-headers-VERSION or something similar. ...



Top 50 recent answers are included