From mboxrd@z Thu Jan  1 00:00:00 1970
Received: from mailman by lists.gnu.org with tmda-scanned (Exim 4.43)
	id 1DmAFy-0007x3-2S
	for qemu-devel@nongnu.org; Sat, 25 Jun 2005 08:59:34 -0400
Received: from exim by lists.gnu.org with spam-scanned (Exim 4.43)
	id 1DmAFo-0007rb-Sy
	for qemu-devel@nongnu.org; Sat, 25 Jun 2005 08:59:25 -0400
Received: from [199.232.76.173] (helo=monty-python.gnu.org)
	by lists.gnu.org with esmtp (Exim 4.43) id 1DmAFm-0007q9-24
	for qemu-devel@nongnu.org; Sat, 25 Jun 2005 08:59:22 -0400
Received: from [195.130.132.56] (helo=adicia.telenet-ops.be)
	by monty-python.gnu.org with esmtp (Exim 4.34) id 1DmAB5-0001ZZ-EY
	for qemu-devel@nongnu.org; Sat, 25 Jun 2005 08:54:31 -0400
Received: from localhost (localhost.localdomain [127.0.0.1])
	by adicia.telenet-ops.be (Postfix) with SMTP id 85B3144229
	for <qemu-devel@nongnu.org>; Sat, 25 Jun 2005 14:50:37 +0200 (MEST)
Received: from [192.168.1.2] (d51A51CF1.access.telenet.be [81.165.28.241])
	by adicia.telenet-ops.be (Postfix) with ESMTP id 6630544225
	for <qemu-devel@nongnu.org>; Sat, 25 Jun 2005 14:50:37 +0200 (MEST)
From: Bob Deblier <bob.deblier@telenet.be>
Content-Type: text/plain
Date: Sat, 25 Jun 2005 14:50:45 +0200
Message-Id: <1119703845.2902.10.camel@orion>
Mime-Version: 1.0
Content-Transfer-Encoding: 7bit
Subject: [Qemu-devel] Regression compiling sparc-user on x86_64 host
Reply-To: qemu-devel@nongnu.org
List-Id: qemu-devel.nongnu.org
List-Unsubscribe: <http://lists.nongnu.org/mailman/listinfo/qemu-devel>,
	<mailto:qemu-devel-request@nongnu.org?subject=unsubscribe>
List-Archive: <http://lists.gnu.org/pipermail/qemu-devel>
List-Post: <mailto:qemu-devel@nongnu.org>
List-Help: <mailto:qemu-devel-request@nongnu.org?subject=help>
List-Subscribe: <http://lists.nongnu.org/mailman/listinfo/qemu-devel>,
	<mailto:qemu-devel-request@nongnu.org?subject=subscribe>
To: qemu-devel@nongnu.org

In October last year I reported a problem compiling qemu-sparc on FC3
x86_64, for which a fix was then found. Trying to recompile the latest
CVS version on FC4 x86_64 (with compat-gcc-32), I find that the code in
dyngen.c has been overhauled, but that compilation fails at:

make[1]: Entering directory `/home/bob/cvs/qemu/sparc-user'
gcc32 -Wall -O2 -g -fno-strict-aliasing -I.
-I/home/bob/cvs/qemu/target-sparc -I/home/bob/cvs/qemu
-I/home/bob/cvs/qemu/linux-user -I/home/bob/cvs/qemu/linux-user/sparc
-D_GNU_SOURCE -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
-I/home/bob/cvs/qemu/fpu -I/home/bob/cvs/qemu/slirp -c -o
translate-op.o /home/bob/cvs/qemu/translate-op.c
In file included from /home/bob/cvs/qemu/translate-op.c:36:
op.h: In function `dyngen_code':
op.h:4631: syntax error before '.' token
op.h:4633: syntax error before '.' token
make[1]: *** [translate-op.o] Error 1

The code from op.h:

case INDEX_op_fnegs: {
    extern void op_fnegs();
extern char .LC96;
    memcpy(gen_code_ptr, (void *)((char *)&op_fnegs+0), 31);
    *(uint32_t *)(gen_code_ptr + 10) = (long)(&.LC96) -
(long)(gen_code_ptr + 10) + -4;
    gen_code_ptr += 31;
}
break;

Unfortunately this time, looking at the relevant section in dyngen.c, I
find at line 1593:

#if defined(HOST_SPARC)
            if (sym_name[0] == '.') {
            fprintf(outfile,
                "extern char __dot_%s __asm__(\"%s\");\n",
                sym_name+1, sym_name);
            continue;
            }
#endif

Extending this to #if defined(HOST_SPARC) || defined(HOST_X86_64)
produces another error, and none of my attempts to fix this seems to
work.

It looks like a working symbol generation in dyngen.c was replaced by
code which regresses to incorrect behavior on x86_64.

Can anyone with more experience please suggest a fix?

Sincerely,

Bob Deblier