Note: This is a beta release of Red Hat Bugzilla 5.0. The data contained within is a snapshot of the live data so any changes you make will not be reflected in the production Bugzilla. Also email is disabled so feel free to test any aspect of the site that you want. File any problems you find or give feedback here.

Bug 88227

Summary: Internal compiler error in splice_child_die, at dwarf2out.c
Product: [Retired] Red Hat Linux Reporter: Dan Morrill <morrildl>
Component: gcc3Assignee: Jakub Jelinek <jakub>
Status: CLOSED UPSTREAM QA Contact:
Severity: medium Docs Contact:
Priority: medium    
Version: 8.0   
Target Milestone: ---   
Target Release: ---   
Hardware: athlon   
OS: Linux   
Whiteboard:
Fixed In Version: Doc Type: Bug Fix
Doc Text:
Story Points: ---
Clone Of: Environment:
Last Closed: 2004-10-03 19:42:00 UTC Type: ---
Regression: --- Mount Type: ---
Documentation: --- CRM:
Verified Versions: Category: ---
oVirt Team: --- RHEL 7.3 requirements from Atomic Host:
Cloudforms Team: --- Target Upstream Version:
Attachments:
Description Flags
area.h - base class definition
none
framework.h - derived class declaration
none
event.h - a dependency framework.h has
none
framework.cpp - class definition none

Description Dan Morrill 2003-04-07 22:04:41 UTC
From Bugzilla Helper:
User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.3a) Gecko/20021212

Description of problem:
gcc -v:
Thread model: posix
gcc version 3.2 20020903 (Red Hat Linux 8.0 3.2-7)

I will attach input files:  framework.c, framework.h, area.h, framework.cpp.  I
do not have a more reduced test case.

Error is this:
g++  -g -Wall  -c -o framework.o framework.cpp
framework.cpp:23: Internal compiler error in splice_child_die, at dwarf2out.c:
   5093
Please submit a full bug report,
with preprocessed source if appropriate.
See <URL:http://bugzilla.redhat.com/bugzilla/> for instructions.

See attached source files.  Should be fairly self-explanatory, but let me know
if it's not.

Version-Release number of selected component (if applicable):


How reproducible:
Always

Steps to Reproduce:
See code.

Additional info:

Mere formality, but... source code is not a trivial test case, but it's
BSD-licensed, so it's okay to throw around as needed.

Comment 1 Dan Morrill 2003-04-07 22:05:52 UTC
Created attachment 90979 [details]
area.h - base class definition

Comment 2 Dan Morrill 2003-04-07 22:06:34 UTC
Created attachment 90980 [details]
framework.h - derived class declaration

Comment 3 Dan Morrill 2003-04-07 22:09:40 UTC
Created attachment 90981 [details]
event.h - a dependency framework.h has

Comment 4 Dan Morrill 2003-04-07 22:10:13 UTC
Created attachment 90982 [details]
framework.cpp - class definition

Comment 5 Dan Morrill 2003-04-07 22:11:35 UTC
I have attached the problematic files -- I think this should compile but I am
not sure since gcc dies out. :)

If I can work around it I will add that info to this entry.

Comment 6 Dan Morrill 2003-04-07 22:18:07 UTC
Apparent workaround:  due to a copy-and-paste issue from area.h (the parent
class) there are a number of methods in framework.h that are defined to be
abstract but for which bodies exist in framework.cpp.  So, this is a
syntactically incorrect C++ file.

Obviously, that wasn't a graceful failure mode, though.

I will attempt to reproduce with a simpler test case...

Comment 7 Richard Henderson 2004-10-03 19:42:00 UTC
Actually, function bodies for abstract virtual functions is just fine.
You aren't allowed to invoke them implicitly (this->foo()), but you are
allowed to invoke them explicitly (this->base::foo()).

The real problem is that Framework::getName is implemented twice.  If
you remove the -g flag, the compiler won't crash, but you'll get 

/tmp/ccJPpA05.s: Assembler messages:
/tmp/ccJPpA05.s:505: Error: symbol `_ZN9Framework7getNameEv' is
already defined

Current versions of the compiler don't crash in the debug info, but
they don't diagnose the duplicate function definition either. Pushed
upstream to http://gcc.gnu.org/bugzilla/show_bug.cgi?id=17816