From: Cliff Young <cyoung@eecs.harvard.edu>
Date: Thu, 4 Jan 1996 11:40:46 -0500
Subject: Re:  baseparsuif-1.0.0 on HPUX
Message-Id: <199601041640.LAA06526@betamax.eecs.harvard.edu>

Yes, I've seen this code->data bug before with gcc/g++ 2.6.3
on HPs. Some of the time, it goes away if you turn optimization off
on the library that should be supplying the (code) symbol.  
I haven't found a general way to fix this.

--Cliff

From: jeroen@leonardo.et.tudelft.nl (Jeroen Hordijk)
Date: Thu, 4 Jan 1996 15:14:43 +0100 (MET)
Subject: baseparsuif-1.0.0 on HPUX
Message-Id: <199601041414.AA012294884@leonardo.et.tudelft.nl>

Hi,


I'm trying to install SUIF on a HP machine using gcc/g++ 2.6.3.
Both basesuif and suifbuilder were no problem (after fixing the
problem reported by William Waite). But baseparsuif results in
the following errors:
******************************************************************************
/opt/suif/gcc-2.6.3/bin/g++ -o pgen -g -Wall -Wno-unused   -I/opt/suif/suif-1.1.
0/basesuif-1.1.0/include  -DDEAL_WITH_GCC_BRAIN_DAMAGE sched.o parent.o task.o p
ar_region.o pgen.o main.o suif_init.o \
                 -L/opt/suif/suif-1.1.0/basesuif-1.1.0/hppa1.1-hp-hpux9.03/lib  
-ltransform -ldependence -lsuifmath -lbuilder -lipmath -list -luseful -lsuif -lm
 prog_ver.cc
/bin/ld: Unsatisfied symbols:
   append__7tn_listP9tn_list_e (data)
   append__19tree_node_list_baseP16tree_node_list_e (data)
   append__15immed_list_baseP12immed_list_e (data)
   push__15immed_list_baseP12immed_list_e (data)
   set_null__7operand (data)
   init__16name_table_entry (data)
   clear__8any_oper (data)
   set_instr__7operandP11instruction (data)
collect2: ld returned 1 exit status
gmake[3]: *** [pgen] Error 1
gmake[2]: *** [pgen] Error 2
gmake[1]: *** [real-reinstall-keep-src] Error 1
gmake: *** [reinstall] Error 2
******************************************************************************

This has probably to do with bugs in gcc/g++. Do these faults sounds
familiar to anybody ?


Jeroen.