The jtreg test is:
[j2se workspace]/test/java/io/PipedInputStream/CloseAndAvailableRC.java
Seems to spin on a CPU (100% busy) while trying to reach a safepoint. Several pstacks show the only active thread is t@6:
dbx on a fastdebug libjvm.so shows t@6 doing this:
current thread: t@6
[1] _libc_poll(0x0, 0x0, 0x1), at 0xdfad9599
[2] poll(0x1, 0x0, 0x0, 0x817a400), at 0xdfb98194
[3] os::sleep(0x817a400, 0x1, 0x0, 0x0, 0xddf97131), at 0xddf96f2a
=>[4] os::yield_all(0x6474), at 0xddf9722e
[5] SafepointSynchronize::begin(0x808cc38, 0x817a400, 0xdeb3395c, 0xdd7f0bf8, 0xdd7f0bc0, 0xddf9792e), at 0xde05b661
[6] VMThread::loop(0x817a400), at 0xde1a8d9f
[7] VMThread::run(0x817a400), at 0xde1a85bb
[8] _start(0x817a400, 0xddf935c8), at 0xddf9
See attached files:
16232.pstack.out
16232.pstack.out.2
18165.jstack.out
18165.pstack.out
18165.pstack.out.2
18165.pstack.out.3
18165.pstack.out.4
core.16232
core.18165
doit.ksh
###@###.### 2005-07-21 00:48:09 GMT
[j2se workspace]/test/java/io/PipedInputStream/CloseAndAvailableRC.java
Seems to spin on a CPU (100% busy) while trying to reach a safepoint. Several pstacks show the only active thread is t@6:
dbx on a fastdebug libjvm.so shows t@6 doing this:
current thread: t@6
[1] _libc_poll(0x0, 0x0, 0x1), at 0xdfad9599
[2] poll(0x1, 0x0, 0x0, 0x817a400), at 0xdfb98194
[3] os::sleep(0x817a400, 0x1, 0x0, 0x0, 0xddf97131), at 0xddf96f2a
=>[4] os::yield_all(0x6474), at 0xddf9722e
[5] SafepointSynchronize::begin(0x808cc38, 0x817a400, 0xdeb3395c, 0xdd7f0bf8, 0xdd7f0bc0, 0xddf9792e), at 0xde05b661
[6] VMThread::loop(0x817a400), at 0xde1a8d9f
[7] VMThread::run(0x817a400), at 0xde1a85bb
[8] _start(0x817a400, 0xddf935c8), at 0xddf9
See attached files:
16232.pstack.out
16232.pstack.out.2
18165.jstack.out
18165.pstack.out
18165.pstack.out.2
18165.pstack.out.3
18165.pstack.out.4
core.16232
core.18165
doit.ksh
###@###.### 2005-07-21 00:48:09 GMT
- duplicates
-
JDK-6295451 CPU hogging with latest bias locking changes.
-
- Resolved
-