|
|
|
Use this tool to learn about websites, specifically the one you just entered.
If you find some aspect of it inappropriate it is not our fault.
If you are the owner of this website: yes we are a real search engine, we do have a real web crawler called FyberSpider and you can block it if you feel the urge.
We are in the process of updating this tool. Until we are done just use our search results to check the inclusion status of your site.
Submit your site to major search engines within 48 hours.
Find out if your site has been cataloged by top search engines for only $8.99.
Below you will see site info taken directly from the URL you entered in real time. This is also known as our URL Breakdown tool and can be used independently of our site info tool.
New GNU thread library
This is just a sample of the content found on this website. Please visit the website to read the entire page.
"Update: this document is obsolete. Most of what is said here
fortunately did not have to come true. The premise, that scheduler
activation are needed for scalable threads, is not true. A much
simpler and far superior solution has been implemented in the form
of NPTL. If we would have known the method to scale the number of threads
on IA32 which we use now, we never would have devised this complicated
mechanism. Hindsight is 20-20. But there are still people who think
scheduler activations are a viable method. I pity the fools.
Design of the New GNU Thread Library
Ulrich Drepper
2002-4-13
On March 26th and 27th a meeting was held at Compaq's
facilities in Nashua, NH, with the goal to design a new thread
library for the GNU C library which would be used as the
replacement of the current LinuxThreads library. The participants
were:
Bill Abt, IBM
Dave Butenhof, Compaq
Ed Sharpe, HP (only on the 27th)
Finnbarr Murphy, Compaq
Jerry Harrow, Compaq
John Piacente, Compaq
Karl Runge, Sun
Kaz Kylheku
Larry Dwyer, HP (only on the 27th)
Peter Portante, Compaq
Ulrich Drepper, Red Hat
Summary
Provide Linux kernel with improved dynamic control of system
load distribution, while allowing threaded applications to
dynamically adapt to the kernel's decisions. This will result in
better performing applications with fewer kernel resources.
The following is the result of the discussions. We tried to
provide a complete overview of the problem at hand and suggest a
solution. We could not do more than that since collaboration from
other parties is necessary to make a final decision. These
parties are the developers of the OS kernels, Linux and Hurd. In
a first step this document is meant to help communicating the
group's opinion to the kernel developers. If an agreement on the
"
....
read entire page
|
Links to Pages on the Same Domain Name
|
|
