---------------------------------------------------------------- BEFORE YOU POST, search the faq at <http://java.apache.org/faq/> WHEN YOU POST, include all relevant version numbers, log files, and configuration files. Don't make us guess your problem!!! ---------------------------------------------------------------- Hello all..... Quick fix..... see the excerpt that follows the "##### Message from Jeff Crowder..." below, for a more on the problem, keep reading....:) First, many thanks to punkytse@geocites for suggesting the IBM Newsgroup (news://news.software.ibm.com/ibm.software.java.linux), and way thanks to Jeff at One.Net (see below)! The following 2 part thread seems to be the answer to the MDK6.1/IBMJDK118 integration problem. I spoke with a couple of the best in UNIX/Java Engineering at my company, and they felt that the thread lib could very well be the culprit of Jserv not being able to bind to port 8007. The reasoning is, that when the JVM initializes the Servlet engine, a separate thread is started to bind to a TCP port that the Servlet engine will listen on (as specified in the directives for ApacheJserv). I have not had time to implement/test this fix, but will do so today, and report my findings. Even if this does not fix the Servlet engine initialization problem, it does seem to indicate that the current set of 2.1.1 libs for Linux (MDK, at least.....will need to check out RH6.1) have backwards compatibility with 2.1 libs of the 6.0 releases, but only as far as Linux applications go. Seems that somewhere in the upgrade process, depreciation of certain c-functions/system calls were made that are not supported in the older IBMJDK118 java binary. After doing a strings and file on the actual java binary, and armed with the insight about the libpthread mentioned in the IBM newsgroup excerpt below, it would seem that the glibc thing is where the answer lies (could this be a result of the Java1.2(2.0) . Maybe need to install the lib packages from RH Dist Archives in a separate directory and do a lib-by-lib comparison!!!! Ugh. Stay Tuned...., JN BTY: Here is an output of the command strings and file on the "/<path to actual java binary in native-threads>/java" binary: >file java ELF 32-bit LSB executable, Intel 80386, version 1, dynamically linked (uses shared libs), not stripped >strings java /lib/ld-linux.so.2 __gmon_start__ libm.so.6 libpthread.so.0 libc.so.6 __finitef __deregister_frame_info __finite _IO_stdin_used __libc_start_main __register_frame_info __finitel libdl.so.2 libjava.so _DYNAMIC _GLOBAL_OFFSET_TABLE_ java_main _etext _edata __bss_start _end GLIBC_2.1 GLIBC_2.0 PTRh QVhX ##### Message from Jeff Crowder of One.Net and reply thread from Brian Watt of IBM ####################### ##### The folowing excerpt was copied from the ibm.software.java.linux newsgroup ####################### To all, I have been using the 10/13 build of JDK118 on both RedHat 6.0 (SMP) systems and typical Mandrake 6.1 systems with greatly varying results. As should be expected, the tested platform (RedHat 6.0 in this case), has performed well beyond my expectations, while the Mandrake 6.1 system has nothing but problems running the JDK. Typically, applications dealing with more than a couple executing threads (especially Swing) have caused SIGSEGVs or runaway processes. The attached class is as simple an example of this problem as can be illustrated. To make a long story short, I grabbed libpthread.so.8 from the RedHat 6.0 box and put it in place in Mandrake 6.1 and all previous crash scenarios immediately dissappeared within the JDK. Use at your own risk. Regards, -Jeff ---------------------------------------------------------------------------- ---- import java.util.*; public class TestCase implements Runnable { public static void main(String[] args) { TestCase t1 = new TestCase("test1"); TestCase t2 = new TestCase("test2"); // TestCase t3 = new TestCase("test3"); // TestCase t4 = new TestCase("test4"); // TestCase t5 = new TestCase("test5"); (new Thread(t1)).start(); (new Thread(t2)).start(); // (new Thread(t3)).start(); // (new Thread(t4)).start(); // (new Thread(t5)).start(); } protected String name; public TestCase(String name) { this.name = name; System.out.println("Created " + name); } public void run() { Vector list = new Vector(); System.out.println("Beginning loop in " + name); for (int i = 0; i < 100000; i++) { list.addElement(new Integer(i)); } System.out.println("Loop complete in " + name); } } Brian Watt <[EMAIL PROTECTED]> wrote in message news:<[EMAIL PROTECTED]>... > Jeff, > > Thanks for the insight. I'm wondering what Mandrake 6.0 has done to > libpthread-0.8.so. The RedHat 6.0 lib is 235,419 bytes and the Mandrake 6.1 > lib is only 79,218. Whoa! I've tried it here, and it has helped greatly reduce > the SIGSEGVs (w/o JIT) and hangs (w/ JIT) on Mandrake 6.1. However, I agree > "Use at your own risk." > > Sincerely, > Brian Watt ########################## Message and reply thread from IBM newsgroup above ########################### -- -------------------------------------------------------------- Please read the FAQ! <http://java.apache.org/faq/> To subscribe: [EMAIL PROTECTED] To unsubscribe: [EMAIL PROTECTED] Archives and Other: <http://java.apache.org/main/mail.html> Problems?: [EMAIL PROTECTED]
