Re: PATCH: server/trace.c cleanup (RETRY)

[Alexandre Julliard]

Gerald Pfeifer  writes:
> [ 2nd attempt. I really don't see any viable alternative to these
>   casts; keeping the status quo certainly seems bad. ]

That code is generated, there's no point in fixing it. You need to
change tools/make_requests.

PATCH: server/trace.c cleanup (RETRY)

[Gerald Pfeifer]

On Tue, 30 Sep 2003, Alexandre Julliard wrote:
> That code is generated, there's no point in fixing it. You need to
> change tools/make_requests.

Thanks for the hint; I'll give it a try.

However, server/registry.c doesn't seem to be generated, so I believe
the patch below is fine as is?

Gerald

-------- forwarded message --------
From: Gerald Pfeifer 
To: [e-mail]
Date: Sun, 21 Sep 2003 22:36:05 +0200 (CEST)
Subject: PATCH: server/registry.c

Similar to the previous one.

Gerald

ChangeLog:
Properly cast parameters to fprintf(...,"%ld",...) to long.
Index: registry.c
===================================================================
RCS file: /home/wine/wine/server/registry.c,v
retrieving revision 1.53
diff -u -3 -p -r1.53 registry.c
--- registry.c	7 Sep 2003 05:08:14 -0000	1.53
+++ registry.c	21 Sep 2003 20:35:08 -0000
@@ -265,7 +265,7 @@ static void save_subkeys( const struct k
     {
         fprintf( f, "\n[" );
         if (key != base) dump_path( key, base, f );
-        fprintf( f, "] %ld\n", key->modif );
+        fprintf( f, "] %ld\n", (long)key->modif );
         for (i = 0; i <= key->last_value; i++) dump_value( &key->values[i], f );
     }
     for (i = 0; i <= key->last_subkey; i++) save_subkeys( key->subkeys[i], base, f );

PATCH: tools/make_requests (was: PATCH: server/trace.c cleanup)

[Gerald Pfeifer]

On Tue, 30 Sep 2003, Alexandre Julliard wrote:
> That code is generated, there's no point in fixing it. You need to
> change tools/make_requests.

Here we go. ;-)

(I never had a look at this script before, and it's not documented too
much, but it seems I found the right place to patch nevertheless, and
the script definitely seems nice.)

Gerald

ChangeLog:
Properly handle time_t in DO_DUMP_FUNC.

Index: make_requests
===================================================================
RCS file: /home/wine/wine/tools/make_requests,v
retrieving revision 1.36
diff -u -3 -p -r1.36 make_requests
--- make_requests	2 Apr 2003 22:48:59 -0000	1.36
+++ make_requests	30 Sep 2003 17:57:53 -0000
@@ -269,6 +269,9 @@ sub DO_DUMP_FUNC
                 push @trace_lines, "    fprintf( stderr, " $var=$formats{$type}";
                 push @trace_lines, "," if ($#_ > 0);
                 push @trace_lines, "", ";
+                # In the case of time_t we need to cast the parameter to
+                # long to match the associated "%ld" printf modifier.
+                push @trace_lines, "(long)" if( $type eq "time_t" );
                 push @trace_lines, "req->$var );\n";
             }
    }