Re: [AD] OSX qtmidi.m build failure fix

["Miguel A. Gavidia" <juvinious@xxxxxxxxxx>]

On Saturday 15 April 2006 20:02, Peter Hull wrote:
> On my system, both TARGET_RT_BIG_ENDIAN and TARGET_RT_LITTLE_ENDIAN
> are defined, just that one is defined as 1, the other as 0.
>
> Can you check the attached program?
>
>
> Pete
>
> On 4/15/06, Miguel A. Gavidia <juvinious@xxxxxxxxxx> wrote:
> > On Saturday 15 April 2006 16:41, Peter Hull wrote:
> > > OK, I am wrong and you are (both) right - I emailed
> > > quicktime-api@xxxxxxxxxx and got a reply from one of their
> > > engineers. The solution* is either to do the *(short*)&whatever as
> > > Ultio suggested, _or_ to put in a conditional as Juvinious suggested.
> > >
> > > However, they recommend using the constant TARGET_RT_BIG_ENDIAN rather
> > > that __i386__, so I suggest the following patch against the current
> > > svn
> > >
> > > Index: src/macosx/qtmidi.m
> > > ===================================================================
> > > --- src/macosx/qtmidi.m (revision 5778)
> > > +++ src/macosx/qtmidi.m (working copy)
> > > @@ -125,7 +125,7 @@
> > >        voice[i].vol = -1;
> > >        voice[i].pan = -1;
> > >        memset(&note_request, 0, sizeof(note_request));
> > > -      #ifndef __i386__
> > > +      #if TARGET_RT_BIG_ENDIAN
> > >         note_request.info.polyphony = 8;
> > >         note_request.info.typicalPolyphony = 0x00010000;
> > >        #else
> > >
> > > Please could someone check this?
> > >
> > > Pete
> > >
> > > * Have a look at the top of qtmusic.c in
> > > http://developer.apple.com/samplecode/qtmusic/qtmusic.html
> > >
> > > On 4/14/06, Peter Hull <peterhull90@xxxxxxxxxx> wrote:
> > > > I just don't think that's the best way to do it. Please can someone
> > > > test what I wrote in my previous email.
> > > >
> > > > EndianU16_NtoB() and friends are _designed_ to overcome differences
> > > > in processors, it seems unnecessary to be wrapping things in #ifdefs.
> > > >
> > > > Pete
> > > >
> > > > On 4/14/06, Elias Pschernig <elias@xxxxxxxxxx> wrote:
> > > > > On Fri, 2006-04-14 at 17:21 +0800, Miguel A. Gavidia wrote:
> > > > > > On Friday 14 April 2006 07:13, Jay Bernardo wrote:
> > > > > > > This is my first attempt at flagging something as broken and
> > > > > > > supplying a patch to fix it. If I improperly followed
> > > > > > > procedure, please let me know.
> > > > > > >
> > > > > > > When building Allegro for MacOS X from the 4.2 branch (revision
> > > > > > > 5777) the following errors occur:
> > > > > > >
> > > > > > > src/macosx/qtmidi.m: In function 'osx_midi_init':
> > > > > > > src/macosx/qtmidi.m:128: error: request for member
> > > > > > > 'bigEndianValue' in something not a structure or union
> > > > > > > src/macosx/qtmidi.m:129: error: request for member
> > > > > > > 'bigEndianValue' in something not a structure or union
> > > > > > >
> > > > > > > Below is a diff of a fix that remedies the problem. Testing may
> > > > > > > be required, as I'm not sure exactly what should be tested to
> > > > > > > ensure it does everything correctly. My poor test consisted of
> > > > > > > running exmidi and verifying that it properly plays a midi
> > > > > > > file.
> > > > >
> > > > > Ok, committed it, after discussion in #allegro.
> > > > >
> > > > > --
> > > > > Elias Pschernig
> > > > >
> > > > >
> > > > >
> > > > > -------------------------------------------------------
> > > > > This SF.Net email is sponsored by xPML, a groundbreaking scripting
> > > > > language that extends applications into web and mobile media.
> > > > > Attend the live webcast and join the prime developer group breaking
> > > > > into this new coding territory!
> > > > > http://sel.as-us.falkag.net/sel?cmd=lnk&kid=110944&bid=241720&dat=1
> > > > >2164 2 --
> > > > > https://lists.sourceforge.net/lists/listinfo/alleg-developers
> > >
> > > -------------------------------------------------------
> > > This SF.Net email is sponsored by xPML, a groundbreaking scripting
> > > language that extends applications into web and mobile media. Attend
> > > the live webcast and join the prime developer group breaking into this
> > > new coding territory!
> > > http://sel.as-us.falkag.net/sel?cmd=lnk&kid0944&bid$1720&dat1642
> >
> > Works fine for me on x86, built with no issues. Need to get a PPC check
> > now. Although don't change that to:
> > #if TARGET_RT_BIG_ENDIAN
> > leave it as:
> > #ifndef TARGET_RT_BIG_ENDIAN
> >
> > --
> > -juvinious
> >
> > <a href="http://chasm.hopto.org";>The Chasm</a> - My Site.
> >
> >
> > -------------------------------------------------------
> > This SF.Net email is sponsored by xPML, a groundbreaking scripting
> > language that extends applications into web and mobile media. Attend the
> > live webcast and join the prime developer group breaking into this new
> > coding territory!
> > http://sel.as-us.falkag.net/sel?cmd=lnk&kid=110944&bid=241720&dat=121642
> > --
> > https://lists.sourceforge.net/lists/listinfo/alleg-developers

Ok my output is this:
TARGET_RT_BIG_ENDIAN = 0
TARGET_RT_LITTLE_ENDIAN = 1

You know though it doesn't matter though, either doing:
#if or #ifndef TARGET_RT_BIG_ENDIAN
will work either way.
As one is defined for each, so either platform won't have both.
ie:
ppc will only have TARGET_RT_BIG_ENDIAN
x86 will only have TARGET_RT_LITTLE_ENDIAN

So I would just go ahead and apply the patch peterhull proposes.

-- 
-juvinious

<a href="http://chasm.hopto.org";>The Chasm</a> - My Site.