CFS Scheduler

Hi, What was the reason for changing the CFS mechanism from fair_clock/wait_runtime to vruntime? What were the kind of workloads where the fairclock approach faired badly ? Also, when adding a blocked task to the runqueue (rbtree), why is it's vruntime set to less than min_vruntime ? This means

Re: LIBS?

On Thu, 28 Aug 2008 10:49:11 +0500 Shaz [EMAIL PROTECTED] wrote: Hi, Whats the difference between /lib and /usr/lib? Apart from this what make or gcc variables should we use to set these? This seems to cover it: http://en.wikipedia.org/wiki/Filesystem_Hierarchy_Standard Martyn --

Re: CFS Scheduler

On Thu, 2008-08-28 at 12:29 +0530, Sukanto Ghosh wrote: Hi, What was the reason for changing the CFS mechanism from fair_clock/wait_runtime to vruntime? Numerical integrety mostly. What were the kind of workloads where the fairclock approach faired badly ? They are analytically identical

Re: Sleep methods in kernel

Le Wed, 27 Aug 2008 16:58:00 +0530 (IST), [EMAIL PROTECTED] a écrit : 1. What is the difference between schedule() and wait_event_interruptible_timeout execpt for the fact that wait_event_interruptible_timeout will also respond to a timeout period? Are both not meant to put to sleep the user

Re: Preventing module-init from being freed

Recovery without reload/disappearing the device. -Asim On 8/27/08, Greg KH [EMAIL PROTECTED] wrote: On Wed, Aug 27, 2008 at 04:04:30PM -0500, Asim wrote: Hi, Linux has this feature where module-init() is freed for some drivers once insmod modprobe finishes. I want to prevent this from

Re: Preventing module-init from being freed

On Thu, Aug 28, 2008 at 11:09:26AM -0500, Asim wrote: Recovery without reload/disappearing the device. recovery from what? The module init code should have nothing to do with this if you have a properly written driver. Do you have a pointer to your code which shows this need? What type of

Re: Preventing module-init from being freed

On Thu, Aug 28, 2008 at 6:35 PM, Greg KH [EMAIL PROTECTED] wrote: On Thu, Aug 28, 2008 at 11:09:26AM -0500, Asim wrote: Recovery without reload/disappearing the device. recovery from what? The module init code should have nothing to do with this if you have a properly written driver. Do you

Re: Preventing module-init from being freed

On Thu, Aug 28, 2008 at 07:06:59PM -0500, Asim wrote: Well - I'm working on something that performs online recovery as a subsytem of my project. There has been many past work which does this or uses such approach. Some famous examples are :-

Re: Preventing module-init from being freed

Well - I wanted to re-initalized the device to its old state by clearing the .bss, copying the old .data, calling the init function followed by replaying the device history on it. I was not getting the .init function by calling the module-init() as it is deleted by sys_init_module on a successful

Re: Preventing module-init from being freed

At a glance it seems nosense, but I'd guess he would like to reuse a block of code from the initialization code [which should fall back to the straightforward solution where he might extract that block and create another function]. Yes - but then that would be very driver specific or worse

Re: Preventing module-init from being freed

Well - I'm working on something that performs online recovery as a subsytem of my project. There has been many past work which does this or uses such approach. Some famous examples are :- 1.www.cs.washington.edu/homes/levy/shadowdrivers.pdf 2.

Re: Preventing module-init from being freed

On Thu, Aug 28, 2008 at 07:17:25PM -0500, Asim wrote: Well - I wanted to re-initalized the device to its old state by clearing the .bss, copying the old .data, calling the init function followed by replaying the device history on it. I was not getting the .init function by calling the

Question on swapping behavior of new kernels

Hi, I'm using a 2.6.26 kernel on an AMD64 machine. Im running some timing experiments on the Linux swapping system. My system has 2GB of RAM but I boot linux with mem=1GB. There's is some behavior happening which I don't understand and I'm hoping someone can explain it to me. Here is what my

Re: Question on swapping behavior of new kernels

Hi Arn... On Fri, Aug 29, 2008 at 8:33 AM, Arn [EMAIL PROTECTED] wrote: Hi, I'm using a 2.6.26 kernel on an AMD64 machine. Im running some timing experiments on the Linux swapping system. My system has 2GB of RAM but I boot linux with mem=1GB. There's is some behavior happening which I

Is there any programs illustrating VESA programming in protection mode?

Hi, all I want to enter graphics mode after the kernel initialization, just like x-windows does. The kernel is a Linux-like one which is written in my lab course, as a result, there's no any driver support. So i need a example showing how to enter graphics mode. Many thanks for your reply

Re: LIBS?

On Thu, Aug 28, 2008 at 12:25 PM, Martyn Welch [EMAIL PROTECTED] wrote: On Thu, 28 Aug 2008 10:49:11 +0500 Shaz [EMAIL PROTECTED] wrote: Hi, Whats the difference between /lib and /usr/lib? Apart from this what make or gcc variables should we use to set these? This seems to cover it:

gcc 4.1, -fno-toplevel-reorder

I replaced a failing harddrive in one of my Debian machines today. Instead of reinstalling Debian I decided to put Ubuntu on there. The new Ubuntu comes with gcc 4.2.3: /usr/bin/gcc-4.2 -v Using built-in specs. Target: i486-linux-gnu Configured with: ../src/configure -v

Makefiles and make

Hi, When we have some makefile variables in a makefile then can we add our own? I mean are these defaults of some kind like DESTDIR, PREFIX, BINDIR etc? Are these standards? If we add ours then how can they affect builds if not assigned to the present variables? thanx. -- Shaz -- To