Problem on Pro*C compilation on HP-UX

Hi,
This is my first post to this forum. I have been facing a strange compilation error message. When I try to make a pro*C file on HP-UX ( uname -a shows HP-UX aopc7449 B.11.11 U 9000/800 2416083493 unlimited-user license),
::: Linking...
/usr/ccs/bin/ld: profls.o: Not a valid object file (invalid system id)
*** Error exit code 1

Stop.
The same file was built perfectly OK with HP-UX (HP-UX awhq6392 B.11.00 U 9000/800 1714987616 unlimited-user license). HP-UX (from 11.00 to 11.11)and Oracle (from 9 to 10.2.0.4) were upgraded recently and now I am getting this error. Can you please suggest how to resolve the error completely ?
Thanks in advance,
Asutosh

Further to my previous query, I would like to add that the compiler is cc, the proC flag are as below.
PROCFLAGS=dbms=native sqlcheck=semantics userid=${USERID} include=${INCDIR}
.pc.c:
@$(PROC) $(PROCFLAGS) iname=$
.pc

Sorry for missong in my oroginal thread.
Thnx
Asutosh

This is supposed to be an object file: profls.o You need to rebuild it, or restore it from tape. It is corrupted.

You are using a makefile. It may "know" how to make profls.o if the source exists for it.

Thank you for your response. Yes, it is an object file. The file is getting created by the makefile only from its source file. I am doing a clean make and the .o file is created, but the error is shown during final stage of linking. This is happening consistently and repeatedly. I can not make out why the .o file is getting generated corrupt everytime. But there was no issue before OS upgrade.

Thnx
Asutosh

which cc?

 $ ll /usr/bin/cc
lr-xr-xr-t   1 root       sys             17 Aug 22  2001 /usr/bin/cc -> /opt/ansic/bin/cc

Usually oracle require also HP aC++ Compiler (product: B3913DB )from 11.11

You could try this evaluation:
Hewlett-Packard Co.

Our Oracle makefile requires the ansi/cc version at a minimum. You would be getting hundreds of syntax errors if you were running against the old K & R compiler - unless your code is really old-fashioned. vbe is right

Can you get the good .o files from tape somewhere? AND. Are you positive the .o files are not also parked in a directory that gets scanned by the -I direcretive? Meaning the .o files could be ancient. And the linker uses them before ld's include path finds the new ones?