ports/87853: [fix] multimedia/mplayer: no bsdbt848 driver compiled in

[Danny Pansters]

On Thursday 3 November 2005 22:30, Simun Mikecin wrote:
> The following reply was made to PR ports/87853; it has been noted by GNATS.
>
> From: Simun Mikecin <numisemis at yahoo.com>
> To: bug-followup at FreeBSD.org
> Cc:
> Subject: Re: ports/87853: [fix] multimedia/mplayer: no bsdbt848 driver
> compiled in Date: Thu, 3 Nov 2005 14:27:14 -0800 (PST)
>
>  --- "Thomas E. Zander" <riggs at rrr.de> wrote:
>  > Hi,
>  >
>  > I'm a bit confused....referring to your PR you were
>  > running a 6.0-RC
>  > when this problem occurred.
>  >
>  > On a RELENG_6 box as of today (yes, I have cleaned
>  > up /usr/include
>  > entirely) there are both
>  >
>  > /usr/include/machine/ioctl_bt848.h *and*
>  > /usr/include/dev/bktr/ioctl_bt848.h
>  >
>  > in place.
>
>  I did a bit more research on this, and actually we are
>  both right.
>  I have two RELENG_6 boxes. One is i386 that is
>  actually an upgrade from RELENG_5. The other one is
>  amd64 that is a clean RELENG_6 install.
>
>  On an i386 box I have:
>
>  /usr/include/dev/bktr/ioctl_bt848.h (9767 bytes)
>  /usr/include/machine/ioctl_bt848.h (1675 bytes)
>
>  And on an amd64 box I only have:
>
>  /usr/include/dev/bktr/ioctl_bt848.h (9767 bytes)
>
>  If you try to find ioctl_bt848.h in the /usr/src/
>  tree, you will find this:
>
>  /usr/src/sys/alpha/include/ioctl_bt848.h (1675 bytes)
>  /usr/src/sys/dev/bktr/ioctl_bt848.h (9767 bytes)
>  /usr/src/sys/i386/include/ioctl_bt848.h (1675 bytes)
>  /usr/src/sys/pc98/include/ioctl_bt848.h (165 bytes)
>
>  So, my conclusion is: on the architectures that have
>  their "machine" version of ioctl_bt848.h (and those
>  are alpha, i386 and pc98), those "machine" version
>  files get installed as
>  /usr/include/machine/ioctl_bt848.h.
>  But *ALL* architectures get
>  /usr/include/dev/bktr/ioctl_bt848.h installed.
>
>  So on an i386 you have both, on the amd64 you only get
>  /usr/include/dev/bktr/ioctl_bt848.h.
>
>  > However the first header doesn't to much but adding
>  > the second one.
>  > Thus, for me it detects bt848, compiles fine and the
>  > driver is
>  > alright. Are your includes really okay?
>  >
>  > The xmms patch and the cdrom/dvdrom device names are
>  > reinplaced in the
>  > post-patch stage during port build.
>  >
>  > I'm sorry but I don't think this pr should be
>  > committed.
>
>  I probably made an error in the patch file. But by now
>  you understand that only thing that is needed is to
>  replace /usr/include/machine/ioctl_bsdbt848.h with
>  /usr/include/dev/bktr/ioctl_bsdbt848.h in the
>  configure script to make it work on *all*
>  architectures.
>  Your port in the current state is broken concerning
>  the usage of the bsdbt848 driver on amd64 platform.
>  Hope this changed your mind...


I ran into this recently when people were testing my kbtv that uses mplayer to 
display tv. (please see thread on multimedia@) and I think indeed that now 
<dev/bktr/ioctl_bsdbt848.h> is fine but older versions (pre 5.2something) 
still want <machine/ioctl_bsdbt848.h>. That's how I understand it at this 
point anyway (I think amd64 wasnt even there at 5.2 time). I'll be following 
this thread because it also posed/s a problem for me. I'm on i386 myself so 
I'm having to work with what other people report to me. Currently, in the 
tool (seperate from mplayer) that's used to change frequency and stuff I 
follow the same fallthrough as mplayer's bsdbt848 has:

#include <sys/param.h>
#ifdef __NetBSD__
#include <dev/ic/bt8xx.h>          /* OpenBSD also? */
#elif __FreeBSD_version >= 502100
#include <dev/bktr/ioctl_meteor.h>
#include <dev/bktr/ioctl_bt848.h>
#else
#include <machine/ioctl_meteor.h>  /* Dfly also? */
#include <machine/ioctl_bt848.h>
#endif

If you think any response is relevant to me also, please CC me. Thanks!