Bug 70153 - New port audio/mac (Monkey's audio codec)
Summary: New port audio/mac (Monkey's audio codec)
Status: Closed FIXED
Alias: None
Product: Ports & Packages
Classification: Unclassified
Component: Individual Port(s) (show other bugs)
Version: Latest
Hardware: Any Any
: Normal Affects Only Me
Assignee: Pav Lucistnik
URL:
Keywords:
Depends on:
Blocks:
 
Reported: 2004-08-08 08:50 UTC by Martin Dieringer
Modified: 2004-08-11 08:53 UTC (History)
0 users

See Also:


Attachments
mac.sh (7.37 KB, text/plain)
2004-08-08 08:50 UTC, Martin Dieringer
no flags Details
patch-ab (812 bytes, TEXT/PLAIN; charset=US-ASCII)
2004-08-10 16:55 UTC, Martin Dieringer
no flags Details

Note You need to log in before you can comment on or make changes to this bug.
Description Martin Dieringer 2004-08-08 08:50:19 UTC

    
Comment 1 Pav Lucistnik freebsd_committer freebsd_triage 2004-08-09 22:59:19 UTC
Responsible Changed
From-To: freebsd-ports-bugs->pav

Take.
Comment 2 Pav Lucistnik freebsd_committer freebsd_triage 2004-08-09 23:00:21 UTC
c++ -DHAVE_CONFIG_H -I. -I. -I../../src/Shared -I../../src/Shared -g -O -pipe -s -O3 -Wall -pedantic -Wno-long-long -DBUILD_CROSS_PLATFORM -Wp,-MD,.deps/APESimple.pp -c APESimple.cpp  -fPIC -DPIC -o .libs/APESimple.o
APESimple.cpp: In function `int DecompressFile(const str_ansi *, const str_ansi *, int *, void (*)(int), int *)':
APESimple.cpp:34: ambiguous overload for `const str_ansi *& ? CSmartPtr<__wchar_t> & : int'
APESimple.cpp:34: candidates are: operator ?:(bool, CSmartPtr<__wchar_t>, CSmartPtr<__wchar_t>) <builtin>
APESimple.cpp:34:                 operator ?:(bool, str_utf16 *, str_utf16 *) <builtin>

Any idea? That's with gcc-2.95 on 4.10-STABLE system.

-- 
Pav Lucistnik <pav@oook.cz>
              <pav@FreeBSD.org>

Orfax's hip flask contained Old Janx Spirit, a mixture of potions of
stupidity, paralysis, naivery, ruination, lose memories, confusion,
and slime mold juice, with a twist of lemon.
    - from a post on Angband Community Forum
Comment 3 Martin Dieringer 2004-08-10 08:10:29 UTC
On Tue, 10 Aug 2004, Pav Lucistnik wrote:

> c++ -DHAVE_CONFIG_H -I. -I. -I../../src/Shared -I../../src/Shared -g -O -pipe -s -O3 -Wall -pedantic -Wno-long-long -DBUILD_CROSS_PLATFORM -Wp,-MD,.deps/APESimple.pp -c APESimple.cpp  -fPIC -DPIC -o .libs/APESimple.o
> APESimple.cpp: In function `int DecompressFile(const str_ansi *, const str_ansi *, int *, void (*)(int), int *)':
> APESimple.cpp:34: ambiguous overload for `const str_ansi *& ? CSmartPtr<__wchar_t> & : int'
> APESimple.cpp:34: candidates are: operator ?:(bool, CSmartPtr<__wchar_t>, CSmartPtr<__wchar_t>) <builtin>
> APESimple.cpp:34:                 operator ?:(bool, str_utf16 *, str_utf16 *) <builtin>
>
> Any idea? That's with gcc-2.95 on 4.10-STABLE system.


I guess it should be the first "candidate", which is defined just a line
above. But how to make gcc-2.95 use this I don't really know...

I just tried building with USE_GCC=2.95 and I get lots and lots of
errors even before this... probably one should use a newer gcc
version?
would adding USE_GCC=3 to the Makefile be sufficient even on a 4.x
system?


m.
Comment 4 Pav Lucistnik freebsd_committer freebsd_triage 2004-08-10 08:39:42 UTC
V út, 10. 08. 2004 v 09:10, Martin Dieringer pí¹e:

> > c++ -DHAVE_CONFIG_H -I. -I. -I../../src/Shared -I../../src/Shared -g -O -pipe -s -O3 -Wall -pedantic -Wno-long-long -DBUILD_CROSS_PLATFORM -Wp,-MD,.deps/APESimple.pp -c APESimple.cpp  -fPIC -DPIC -o .libs/APESimple.o
> > APESimple.cpp: In function `int DecompressFile(const str_ansi *, const str_ansi *, int *, void (*)(int), int *)':
> > APESimple.cpp:34: ambiguous overload for `const str_ansi *& ? CSmartPtr<__wchar_t> & : int'
> > APESimple.cpp:34: candidates are: operator ?:(bool, CSmartPtr<__wchar_t>, CSmartPtr<__wchar_t>) <builtin>
> > APESimple.cpp:34:                 operator ?:(bool, str_utf16 *, str_utf16 *) <builtin>
> >
> > Any idea? That's with gcc-2.95 on 4.10-STABLE system.


> I guess it should be the first "candidate", which is defined just a line
> above. But how to make gcc-2.95 use this I don't really know...


Quirk is to convert all DOS newlines to UNIX before trying with
gcc-2.95.

> I just tried building with USE_GCC=2.95 and I get lots and lots of
> errors even before this... probably one should use a newer gcc
> version?
> would adding USE_GCC=3 to the Makefile be sufficient even on a 4.x
> system?


I would prefer patching it to allow build with 2.95, unless it's really
a lot of work. First, it's a big hit to force user to install gcc-3.3 if
he only wants to build mac. Second, you would get unuable shared library
which can't be linked anywhere with 2.95, only with 3.3 again. :(

-- 
Pav Lucistnik <pav@oook.cz>
              <pav@FreeBSD.org>

What luck - a dagger!  This is your holder.  Rest there and let me die.
Comment 5 Martin Dieringer 2004-08-10 09:28:14 UTC
On Tue, 10 Aug 2004, Pav Lucistnik wrote:

> Quirk is to convert all DOS newlines to UNIX before trying with
> gcc-2.95.

How do I do this in a port, automatically?

>> would adding USE_GCC=3 to the Makefile be sufficient even on a 4.x
>> system?
>
> it's a big hit to force user to install gcc-3.3 if he only wants
> to build mac.

that's right...
but how long will there be users without ggc-3?

m.
Comment 6 Martin Dieringer 2004-08-10 16:55:01 UTC

I hope my understanding of the problem is correct.
Please have a look at the attached "patch-ab"




in the port-Makefile I used following:

USE_GCC=		2.95
GNU_CONFIGURE=		yes
#USE_LIBTOOL_VER=       15


because USE_LIBTOOL_VER=15 doesn't compile:

===>  Building for mac-3.99.4
Making all in src
gmake[1]: Entering directory /BIG3/macfbsd/port/mac/work/mac-3.99-u4/src'
Making all in Shared
gmake[2]: Entering directory /BIG3/macfbsd/port/mac/work/mac-3.99-u4/src/Shared'
/bin/sh /usr/local/bin/libtool15 --mode=compile g++295 -DHAVE_CONFIG_H -I. -I. -I.     -O -pipe -s -O3 -Wall -pedantic -Wno-long-long -DBUILD_CROSS_PLATFORM -c GlobalFunctions.cpp
libtool15: compile: unable to infer tagged configuration
libtool15: compile: specify a tag with --tag'
gmake[2]: *** [GlobalFunctions.lo] Error 1
gmake[2]: Leaving directory /BIG3/macfbsd/port/mac/work/mac-3.99-u4/src/Shared'
gmake[1]: *** [all-recursive] Error 1
gmake[1]: Leaving directory /BIG3/macfbsd/port/mac/work/mac-3.99-u4/src'
gmake: *** [all-recursive] Error 1
*** Error code 2


m.





On Tue, 10 Aug 2004, Pav Lucistnik wrote:

> V út, 10. 08. 2004 v 09:10, Martin Dieringer píae:
>
>>> c++ -DHAVE_CONFIG_H -I. -I. -I../../src/Shared -I../../src/Shared -g -O -pipe -s -O3 -Wall -pedantic -Wno-long-long -DBUILD_CROSS_PLATFORM -Wp,-MD,.deps/APESimple.pp -c APESimple.cpp  -fPIC -DPIC -o .libs/APESimple.o
>>> APESimple.cpp: In function `int DecompressFile(const str_ansi *, const str_ansi *, int *, void (*)(int), int *)':
>>> APESimple.cpp:34: ambiguous overload for `const str_ansi *& ? CSmartPtr<__wchar_t> & : int'
>>> APESimple.cpp:34: candidates are: operator ?:(bool, CSmartPtr<__wchar_t>, CSmartPtr<__wchar_t>) <builtin>
>>> APESimple.cpp:34:                 operator ?:(bool, str_utf16 *, str_utf16 *) <builtin>
>>>
>>> Any idea? That's with gcc-2.95 on 4.10-STABLE system.

>
Comment 7 Pav Lucistnik freebsd_committer freebsd_triage 2004-08-10 17:08:24 UTC
V út, 10. 08. 2004 v 17:55, Martin Dieringer pí¹e:

> I hope my understanding of the problem is correct.
> Please have a look at the attached "patch-ab"


That got us one step further, now it fails on the fact that 4.x libc
does not have wcstol. Would be replacing wcstol call with strtol
acceptable?

> in the port-Makefile I used following:
> 
> USE_GCC=		2.95
> GNU_CONFIGURE=		yes
> #USE_LIBTOOL_VER=       15
> because USE_LIBTOOL_VER=15 doesn't compile:


That's weird but it's not relevant to this particular issue.

-- 
Pav Lucistnik <pav@oook.cz>
              <pav@FreeBSD.org>

And please, please, please add COMMENTS to your code. Reading
uncommented PERL is like chewing on chunks of broken glass,
only without the tasty blood sauce to go with it.
  -- John Rowan in rec.games.roguelike.adom
Comment 8 Pav Lucistnik freebsd_committer freebsd_triage 2004-08-11 08:53:00 UTC
State Changed
From-To: open->closed

Committed after some work with submitter