Re: Calling sys_sysinfo from sched.c

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



On Wed, 20 Apr 2005, A.M. Fradley wrote:

> From the looks of things, that's how the functions in thrash.c work but 
> I have no idea how I might create my own verion of these or modify them 
> for my assignment.  Do you know if there's any good documentation of 
> where these functions are called from?  I'm kind of worried now as I 
> think this is the exact type of thing I think I should have tried to 
> develop myself but since the experts have already done this, there's 
> very little left I can do to improve it.

The "expert" (that's me) spent 2 hours reading a conference
paper, and then a day and a half writing proof of concept
code.

That proof of concept code ended up helping with some heavy
VM workloads, so it got merged into the kernel quite quickly.
In short, the mechanism is there, but the policy has received
only about 3 hours of my time ;)

This means there are a lot of things left to do:
1) the swap token mechanism is bad for performance in very
   light VM loads, and switched off by default in the upstream
   kernel - it would be good if the swap token mechanism could
   detect the VM load and switch itself on and off on demand
2) the policy for moving the swap token between processes is
   pretty braindead - having a more intelligent policy could
   probably get big performance gains!
3) the paper linked from mm/thrash.c has some hints on what a
   better policy could look like - I just never got around to
   implementing it ;(

I have a feeling there is a LOT of interesting research left to
do for the swap token mechanism.  I'm saying "interesting research"
because the actual code needed for most of these policy changes
should be relatively small, but the amount of thinking about the
problem may need to be big.

Of course, I'm always willing to discuss things with you, as long
as it's on one of the public mailing lists, so I won't have to have
the same discussion with 10 people ;)

-- 
"Debugging is twice as hard as writing the code in the first place.
Therefore, if you write the code as cleverly as possible, you are,
by definition, not smart enough to debug it." - Brian W. Kernighan

--
Kernelnewbies: Help each other learn about the Linux kernel.
Archive:       http://mail.nl.linux.org/kernelnewbies/
FAQ:           http://kernelnewbies.org/faq/


[Index of Archives]     [Newbies FAQ]     [Linux Kernel Mentors]     [Linux Kernel Development]     [IETF Annouce]     [Git]     [Networking]     [Security]     [Bugtraq]     [Yosemite]     [MIPS Linux]     [ARM Linux]     [Linux RAID]     [Linux SCSI]     [Linux ACPI]
  Powered by Linux