|
On Saturday 07 July 2012 23:45:02 Thomas Klausner wrote: > The following reply was made to PR pkg/46662; it has been noted by GNATS. > > From: Thomas Klausner <tk%giga.or.at@localhost> > To: NetBSD bugtracking <gnats-bugs%NetBSD.org@localhost> > Cc: > Subject: Re: pkg/46662: [PATCH] upgrade timidity port to 2.14.0 > Date: Sat, 7 Jul 2012 22:43:01 +0200 > > --nFreZHaLTZJo0R7j > Content-Type: text/plain; charset=us-ascii > Content-Disposition: inline > > Hi! > > Thanks for the timidity update. > > I've tried it on NetBSD-6.99.8/amd64. > > I noticed that you removed all CVS ($NetBSD$) tags from the files -- > please don't do that. At least as long as NetBSD's using cvs(1).
Oops.
> > I've readded patch-ac because declaring errno as int will break on > some platforms. I'm not sure about patch-ae, so I've left it removed. > > I've also applied the changes you sent separately. > > The build failed for me with: <snipped> > > Can you please take a look and recommend a fix? >
That's the server interface, which can either use some defines from OSS soundcard.h or include its own server_defs.h. HAVE_SYS_SOUNDCARD_H is defined (somehow the configure finds libossaudio's soundcard.h) but libossaudio's soundcard.h doesn't include these MIDI definitions. I suggest patching server_c.c to always use its own definitions (which IMHO, sounds like the best idea in general. An interface which may mutate based on the system's soundcard.h and compilation mode is silly. Fortunately, soundcard.h is stable in this regard). An alternative fix would be to remove --enable-server, but it can be ocassionally useful (e.g. scummvm can output midi via it).
--- server_c.c 2012-07-08 02:44:39.707040910 +0300 +++ server_c.c 2012-07-08 02:44:50.672466303 +0300 @@ -58,11 +58,7 @@ #endif #include <signal.h>
-#ifdef HAVE_SYS_SOUNDCARD_H -#include <sys/soundcard.h> -#else #include "server_defs.h" -#endif /* HAVE_SYS_SOUNDCARD_H */
#include "timidity.h" #include "common.h"
|