On 12/03/2015 13:52, Christopher Schultz wrote: > Mark, > > On 3/11/15 5:06 PM, ma...@apache.org wrote: >> Author: markt >> Date: Wed Mar 11 21:06:53 2015 >> New Revision: 1666013 >> >> URL: http://svn.apache.org/r1666013 >> Log: >> Add a placeholder for the descriptor merging that is going to be necessary >> for OSX and BSD. >> >> Modified: >> tomcat/trunk/java/org/apache/tomcat/util/net/AprEndpoint.java >> >> Modified: tomcat/trunk/java/org/apache/tomcat/util/net/AprEndpoint.java >> URL: >> http://svn.apache.org/viewvc/tomcat/trunk/java/org/apache/tomcat/util/net/AprEndpoint.java?rev=1666013&r1=1666012&r2=1666013&view=diff >> ============================================================================== >> --- tomcat/trunk/java/org/apache/tomcat/util/net/AprEndpoint.java (original) >> +++ tomcat/trunk/java/org/apache/tomcat/util/net/AprEndpoint.java Wed Mar 11 >> 21:06:53 2015 >> @@ -1759,6 +1759,7 @@ public class AprEndpoint extends Abstrac >> nextPollerTime += pollerTime; >> } >> if (rv > 0) { >> + rv = mergeDescriptors(desc, rv); >> pollerSpace[i] += rv; >> connectionCount.addAndGet(-rv); >> for (int n = 0; n < rv; n++) { >> @@ -1923,6 +1924,21 @@ public class AprEndpoint extends Abstrac >> this.notifyAll(); >> } >> } >> + >> + >> + private int mergeDescriptors(long[] desc, int startCount) { >> + if (OS.IS_BSD || OS.IS_MACOSX) { >> + // TODO Need to actually implement merging of the >> descriptors here. >> + // I'm currently thinking quicksort followed by running >> + // through the sorted list to merge the events. >> + return startCount; > > How many events are you expecting to have to handle, how many do you > expect to merge, and how often do you expect to do any actual work?
No idea to all of the above. That is part of the problem. > QuickSort is great if your data are completely unordered and you intend > to actually order them. In this case, though, the order of the data is > only a convenience to find matches and not the end goal. > > There may be better ways to find matches than sorting and then > performing a linear search for matches. Quote possibly. In the end I went for easy to follow code that should be good enough. If someone has a use case where this is not the case, we can deal with it then. Mark --------------------------------------------------------------------- To unsubscribe, e-mail: dev-unsubscr...@tomcat.apache.org For additional commands, e-mail: dev-h...@tomcat.apache.org
