gdb/1552: debugging on PowerPC - Thread problem - Linux

[ccavanna at ca dot ibm dot com]

>Number:         1552
>Category:       gdb
>Synopsis:       debugging on PowerPC - Thread problem - Linux
>Confidential:   no
>Severity:       serious
>Priority:       medium
>Responsible:    unassigned
>State:          open
>Class:          sw-bug
>Submitter-Id:   net
>Arrival-Date:   Thu Feb 12 21:28:01 UTC 2004
>Closed-Date:
>Last-Modified:
>Originator:     Carlos Cavanna
>Release:        5.1.X 6.0 and latest snapshot
>Organization:
>Environment:
OS: Power4+, 4-way.
Linux (SuSE SLES 8) 2.4.19-u11-ppc64-SMP.

GCC:
Reading specs from /usr/lib/gcc-lib/powerpc-suse-linux/3.2/specs
Configured with: ../configure --enable-threads=posix --prefix=/usr --with-local-prefix=/usr/local --infodir=/usr/share/info --mandir
=/usr/share/man --libdir=/usr/lib --enable-languages=c,c++,f77,objc,java,ada --enable-libgcj --with-gxx-include-dir=/usr/include/g++
 --with-slibdir=/lib --with-system-zlib --enable-shared --enable-__cxa_atexit powerpc-suse-linux
Thread model: posix
gcc version 3.2

GDB:
This GDB was configured as "powerpc64-unknown-linux-gnu".
and also
This GDB was configured as "powerpc-unknown-linux-gnu"

>Description:
The application starts with 5 threads (and more threads are created along the run). gdb is attached in the middle of the
multithreaded application run. Then, the run is continued inside gdb. gdb then hungs at some point during the execution.

The hang is intermittent, but usually happens after we receive the [New Thread # (LWP #)] message.

Also, we sometimes get the following:

[New Thread 3076 (LWP 30055)]
[New Thread 4101 (LWP 30058)]
[New Thread 5126 (LWP 30059)]
[New Thread 6151 (LWP 30060)]
lin-lwp.c:1166: gdb-internal-error: lin_lwp_wait: Assertion `lp->resumed' failed.
An internal GDB error was detected.  This may make further
debugging unreliable.  Quit this debugging session? (y or n) 

In case it is relevant, the application we use consists of a main binary and a set of additional libraries.

The application was compiled as 32-bit, and gdb as 64 binary for the above tests. Though it is possible this detail doesn't matter, it might still be relevant.

So we also tried using gdb as a 32-bit binary.
And we got a different message.

gdb) cont
Continuing.
[New Thread 3076 (LWP 21526)]
[New Thread 4101 (LWP 21527)]
Cannot get thread event message: generic error

If you need further details (that I may have ommited), please let me know.
>How-To-Repeat:
Unfortunately, we are still unable to reproduce it using a simple testcase. But it happens regularly with our application. We cannot disclose any internal details there, sorry.

I hope this is a know problem, though I haven't found any clear references to it in the bug database so far.
>Fix:

>Release-Note:
>Audit-Trail:
>Unformatted: