Re: [LIP] NPTL behaviour differences
Shridhar Daithankar <[email protected]>
| Newsgroups | gmane.user-groups.linux.india.programmers |
|---|---|
| Organization | Persistent Systems Pvt. Ltd. |
| Message-ID | <[email protected]> |
Arun Sharma wrote: > On Wed, Mar 03, 2004 at 10:27:17PM +0530, Shantanu Kumar wrote: > >>Hi, >> >>Has anybody tried out using NPTL on 2.6 or the patched 2.4 branch that >>came with RH9 onwards? I was wondering what are the behaviour >>differences compared to the "straight" kernel. And whether there's any >>change in API? I heard it breaks some apps. >> >>Most importantly, what are the caveats or the important points to keep >>in mind while coding with NPTL? Links will be appreciated. > > > - Signal handling (you can send signals to a whole process now) I believe signals can be per thread now. Earlier it was to the entire process. > - All threads in a process get the same return value for getpid() Additionally there are futexes now. It is not a behavioural difference though. Not to mention O(1) proc cleanups which is a big thing.. Shridhar ------------------------------------------------------- This SF.Net email is sponsored by: IBM Linux Tutorials Free Linux tutorial presented by Daniel Robbins, President and CEO of GenToo technologies. Learn everything from fundamentals to system administration.http://ads.osdn.com/?ad_id=1470&alloc_id=3638&op=click