z8k sim breakage

z8k sim breakage

[Andrew Cagney]

Alan,

I suspect the change:

2002-12-13  Alan Modra  <amodra@bigpond.net.au>

	* h8500-opc.h (h8500_table): Add missing initializers to quiet
	warnings.
	* pj-dis.c (print_insn_pj): Adjust for pj_opc_info_t change.
	* pj-opc.c (pj_opc_info): Add braces around union initializer.
	* z8kgen.c: Include "libiberty.h".
	(opt, args, toks): Fix initializer warnings.
	(chewname): Make "name" a char **.  Return mnemonic trimmed of
	operands.
	(gas): Improve emitted "DO NOT EDIT" warning.  Format emitted
	opcode_entry_type, and make "nicename" and "name" const.  Make
	z8k_table const too.  Formatting.  Generate idx as gas needs it.
	* z8k-opc.h: Regenerate.

broke the building of sim/z8k vis:

$ make
gcc -c -DHAVE_CONFIG_H     -DPROFILE=1 -DWITH_PROFILE=-1 
-DDEFAULT_INLINE=0 
-I/notnfs/cagney/GDB/src/sim/z8k/../../newlib/libc/sys/z8k   -I. 
-I/notnfs/cagney/GDB/src/sim/z8k -I../common 
-I/notnfs/cagney/GDB/src/sim/z8k/../common -I../../include 
-I/notnfs/cagney/GDB/src/sim/z8k/../../include -I../../bfd 
-I/notnfs/cagney/GDB/src/sim/z8k/../../bfd -I../../opcodes 
-I/notnfs/cagney/GDB/src/sim/z8k/../../opcodes -I../../intl 
-I/notnfs/cagney/GDB/src/sim/z8k/../../intl -g -O2 
/notnfs/cagney/GDB/src/sim/z8k/comped3.c
In file included from /notnfs/cagney/GDB/src/sim/z8k/comped3.c:26:
tc-gen3.h:68: redefinition of `sfop_2'
tc-gen3.h:46: `sfop_2' previously defined here
tc-gen3.h:91: redefinition of `sfop_2'
tc-gen3.h:68: `sfop_2' previously defined here

That sim's writecode program drags in the opcodes table and then uses it 
to generate part of the simulator (hence my suspicion).  I'm still 
trying to figure out how that `wonderful' code works.

Anyway, off hand, what exactly did you change here? (perhaphs that will 
explain what went wrong)
Also, was it because of a reported z8k bug or just a cleanup? (i'm also 
looking for evidence of z8k life)

Andrew

Re: z8k sim breakage

[Alan Modra]

On Mon, Jan 06, 2003 at 05:27:18PM -0500, Andrew Cagney wrote:
> Anyway, off hand, what exactly did you change here? (perhaphs that will 
> explain what went wrong)

I was fixing a z8k gas segfault caused by opcodes 2002-11-18 Klee Dienes
constification of z8k_table.  I'll take a look at the sim problem today.

-- 
Alan Modra
IBM OzLabs - Linux Technology Centre