compiler / linker options for compiling on IA64

[shashank k]

Dear friends,

I am facing an issue while trying to port my code to HP-UX IA64 machine.

Main problem is "The run time polymorphism is not working in the desired way. I am getting core dump in my application."

Pls provide your suggestions

[Manix]

On which compiler you are trying to port your code.Look at this one

https://h20392.www2.hp.com/portal/swdepot/displayProductInfo.do?productNumber=B9007AAEVAL

Thanks
Manix

[Zygmunt Krawczyk]

Which HP-UX version (11.23 or 11.31)?
Which compiler and version?

[Dennis Handly]

Should I ask why you quoted: "The run time polymorphism is not working ... ?

You need to mention the version of your compiler. The latest compiler is A.06.25.02.

Please provide a stack trace of your core dump.

[shashank k]

The application code is running successfully on HP-UX B.11.11 , but while i am trying to run on HP - UX 11.23 (IA64) machine it is giving core dump. The core dump stack trace is pasted below :-

Core was generated by `p2.3driver'.
Program terminated with signal 11, Segmentation fault.
SEGV_ACCERR - Invalid Permissions for object
#0 0x4100f70:0 in TSDriver::doBegin (this=0x0, fileName=@0x7fffd0d0,
fullOutput=true) at src/TSDriver.cc:198

warning: Source file is more recent than executable p2.3driver.

198 //_response = new XML_Response::ResponseGenerator(fileName);
(gdb) bt
#0 0x4100f70:0 in TSDriver::doBegin (this=0x0, fileName=@0x7fffd0d0,
fullOutput=true) at src/TSDriver.cc:198
warning: Inline error: (Possibly due to split inline instances).
Turning off inline debugging.

#1 0x41477e0:0 in executeTestCases (argc=2, argv=0x7fffdf04,
cases=0x7fffd740, start=0, howMany=1, numTests=1, outputDir=@0x7fffd560,
fullOutput=true, semid=-1) at src/main.cc:457
#2 0x41514e0:0 in main (argc=2, argv=0x7fffdf04, env=0x7fffdf10)
at src/main.cc:908



It looks like in the new environment "run time polymorphism is not working in the same way as before"

Version of compiler on new machine is
HP C/aC++ B3910B A.06.24 [Dec 04, 2009]

Version of compiler on old machine is
aCC: HP ANSI C++ B3910B A.03.33

[Dennis Handly]

>#0 0x4100f70:0 in TSDriver::doBegin (this=0x0, fileName=@0x7fffd0d0,
fullOutput=true)

This says that the "this" pointer is NULL.
How are you calling doBegin?
What is the declaration of doBegin, virtual?

>warning: Inline error: (Possibly due to split inline instances).

You may want to compile with +d to see if that helps you debug these inlines.

#1 0x41477e0:0 in executeTestCases at src/main.cc:457

What's on this line of code?

It looks like in the new environment "run time polymorphism is not working in the same way as before"

>Version of compiler on new machine is A.06.24

The latest version is A.06.25.02.

>Version of compiler on old machine is A.03.33

That obsolete compiler is nearly a decade old.

[shashank k]

#0 0x4100f70:0 in TSDriver::doBegin (this=0x0, fileName=@0x7fffd0d0,
fullOutput=true)

This says that the "this" pointer is NULL.
How are you calling doBegin?

What is the declaration of doBegin, virtual? YES

>warning: Inline error: (Possibly due to split inline instances).

You may want to compile with +d to see if that helps you debug these inlines.

#1 0x41477e0:0 in executeTestCases at src/main.cc:457

What's on this line of code?
Line 457 : driver->doBegin(resultName, fullOutput);
where driver is declared as TSDriver* driver;
TSDriver is an abstract class.

It looks like in the new environment "run time polymorphism is not working in the same way as before"

>Version of compiler on new machine is A.06.24

The latest version is A.06.25.02.

>Version of compiler on old machine is A.03.33

That obsolete compiler is nearly a decade old.

[shashank k]

Can anyone please suggest on this?

[Dennis Handly]

>where driver is declared as TSDriver* driver;

Please print the value of the pointer "driver". Is it 0?
You can't virtual call through a NULL pointer.