Current Status
Andrey Slepuhin
pooh@msu.ru
Thu, 23 Oct 1997 01:45:30 +0300
Hi Sai-Lai,
I've done the promised testing of omniORB2.2.0 under egcs on Linux.
Unfortunately, results are bad. Only thread examples and eg1 work.
Other examples were failed both with shared and static libraries
(ofcause, standard libraries i.e. libc (glibc-2.0.5c) and libpthread
(glibc-linuxthreads-2.0.5) are shared). I tried to debug eg2_impl but
unsuccessfully. Below there is a *strange* gdb output (client was run
on other computer, I marked actions related to client with *****).
---------------------------------------------------------
pooh@nest:~/src/omniORB_2.2.0/src/examples/echo$ gdb eg2_impl
GNU gdb 970817
Copyright 1997 Free Software Foundation, Inc.
GDB is free software, covered by the GNU General Public License, and you are
welcome to change it and/or distribute copies of it under certain conditions.
Type "show copying" to see the conditions.
There is absolutely no warranty for GDB. Type "show warranty" for details.
This GDB was configured as "i386-any-linux"...
(gdb) run
Starting program: /home/pooh/src/omniORB_2.2.0/src/examples/echo/eg2_impl
'IOR:01120d080d00000049444c3a4563686f3a312e3000120d0801000000000000002c000000
010100000f0000003139332e3132342e3134382e37320000771600000c000000344e679173c4c
90a00000001'
Program received signal SIGUSR1, User defined signal 1.
0x400ee544 in __syscall_sigsuspend () at soinit.c:27
soinit.c:27: No such file or directory.
(gdb) cont
Continuing.
*****starting remote client
*****client successfully received echo end exited
Program received signal SIGSEGV, Segmentation fault.
__pthread_sighandler (sig=-1073743192) at pthread.c:318
pthread.c:318: No such file or directory.
(gdb) bt
#0 __pthread_sighandler (sig=-1073743192) at pthread.c:318
#1 0x400140c5 in pthread_create (thread=0x8138440, attr=0x4010028e,
start_routine=0xbffff7f8, arg=0x4017a508) at restart.h:32
#2 0x1 in ?? ()
Cannot access memory at address 0x1.
(gdb)
----------------------------------------------------------------
I am curious about such gdb output, I have no starting point for
further debugging and I cannot say is this caused by egcs or omniORB
bug :-(.
Well, I'll wait for upcoming new release and then I'll try to look at it
more closely. Below I put a summary of omniORB2.2.0 test results on
platforms available to me.
-------------------------------------------------------------------------
platform\example |thread | eg1 | eg2 | eg3 |
-------------------------------------------------------------------------
AIX 4.2, native | | | | |
C Set++ 3.1.4 compiler, | ok | ok | ok | ok |
static libs | | | | |
-------------------------------------------------------------------------
AIX 4.2, native | | | | |
C Set++ 3.1.4 compiler, | ok | ok | ok | ok |
shared libs | | | | |
-------------------------------------------------------------------------
AIX 4.2, egcs-971016, | ok | ok | ok | ok |
native ld, static libs | | | | |
-------------------------------------------------------------------------
AIX 4.2, egcs-971016, | ok | ok | ok | ok |
native ld, shared libs | | | | |
-------------------------------------------------------------------------
AIX 4.2, egcs-971016, | ok | ok | ok | ok |
gnu ld, static libs | | | | |
-------------------------------------------------------------------------
AIX 4.2, egcs-971016, | ok | ok | fail | fail |
gnu ld, shared libs | | | | |
-------------------------------------------------------------------------
Linux 2.0.30, gcc-2.7.2, | ok | ok | ok | ok |
libc-5.3.12, linuxthreads-0.5, | | | | |
static libs | | | | |
-------------------------------------------------------------------------
Linux 2.0.30, egcs-971016, | | | | |
glibc-2.0.5c, glibc-linuxthreads-2.0.5 | ok | ok | fail | fail |
(linuxthreads-0.6), static libs | | | | |
-------------------------------------------------------------------------
Linux 2.0.30, egcs-971016, | | | | |
glibc-2.0.5c, glibc-linuxthreads-2.0.5 | ok | ok | fail | fail |
(linuxthreads-0.6), shared libs | | | | |
-------------------------------------------------------------------------
Best regards,
Andrey.