From: john on 14 Dec 2009 04:05 in my project, sometimes some java thread can't be killed immediately, is there any good ways to slove the problem? in java API Thread.java, there isn't stop method, whether it means in java, App writer can't kill a thread forcely or there is some around methods?
From: Peter Duniho on 14 Dec 2009 04:17 john wrote: > in my project, sometimes some java thread can't be killed immediately, > is there any good ways to slove the problem? > in java API Thread.java, there isn't stop method, whether it means in > java, App writer can't kill a thread forcely or there is some around > methods? Don't write a thread that you need to kill forcefully. Always provide some mechanism by which your code in the thread can exit gracefully and in a predictable way. Corallary: don't use third-party code that also does not provide this feature in its code. For related information, see: http://java.sun.com/javase/6/docs/technotes/guides/concurrency/threadPrimitiveDeprecation.html Pete
From: Eric Sosman on 14 Dec 2009 08:16 On 12/14/2009 4:05 AM, john wrote: > in my project, sometimes some java thread can't be killed immediately, > is there any good ways to slove the problem? > in java API Thread.java, there isn't stop method, whether it means in > java, App writer can't kill a thread forcely or there is some around > methods? If the victim thread isn't cooperating (by checking a "stop now" flag occasionally or some such), the only way I can think of to stop it safely is System.exit(). The problem with hurling a hand grenade into a thread's vitals at some arbitrary moment is that you don't know what the thread was doing at the moment you blew it up. If it was in the middle of updating a data structure somewhere, the data structure may now be half-updated, useless or even poisonous to the rest of the program. You've got to make sure the victim is in a "safe" state at the moment of termination, and you can't do that without the victim's cooperation. -- Eric Sosman esosman(a)ieee-dot-org.invalid
From: Lothar Kimmeringer on 14 Dec 2009 12:22 Eric Sosman wrote: > On 12/14/2009 4:05 AM, john wrote: >> in my project, sometimes some java thread can't be killed immediately, >> is there any good ways to slove the problem? >> in java API Thread.java, there isn't stop method, whether it means in >> java, App writer can't kill a thread forcely or there is some around >> methods? > > If the victim thread isn't cooperating (by checking a > "stop now" flag occasionally or some such), the only way I > can think of to stop it safely is System.exit(). System.exit waits for all non-daemon-threads to stop. So if the Thread you want to kill is not a daemon-thread and it ignores all attempts to stop, you're screwed. Regards, Lothar -- Lothar Kimmeringer E-Mail: spamfang(a)kimmeringer.de PGP-encrypted mails preferred (Key-ID: 0x8BC3CD81) Always remember: The answer is forty-two, there can only be wrong questions!
From: Daniel Pitts on 14 Dec 2009 12:51 Lothar Kimmeringer wrote: > Eric Sosman wrote: > >> On 12/14/2009 4:05 AM, john wrote: >>> in my project, sometimes some java thread can't be killed immediately, >>> is there any good ways to slove the problem? >>> in java API Thread.java, there isn't stop method, whether it means in >>> java, App writer can't kill a thread forcely or there is some around >>> methods? >> If the victim thread isn't cooperating (by checking a >> "stop now" flag occasionally or some such), the only way I >> can think of to stop it safely is System.exit(). > > System.exit waits for all non-daemon-threads to stop. > So if the Thread you want to kill is not a daemon-thread > and it ignores all attempts to stop, you're screwed. Actually, barring a security exception, System.exit forcefully exits the JVM. Without a call to system.exit, non-daemon threads will keep the JVM alive. -- Daniel Pitts' Tech Blog: <http://virtualinfinity.net/wordpress/>
|
Next
|
Last
Pages: 1 2 Prev: Why can nsmc, local classes or anonymous classes have staticmembers ? Next: Checked Exceptions |