mpg123-devel — Mailing list for mpg123 developers

[mpg123-devel] Porting to WIndows/ARM64

[Kevin J. <inv...@ya...>]

Hi All,
I am currently porting libmpg123 to Windows/ARM64. From the libmpg123 web site, I gathered that Patrick and JonY are responsible for creating the Windows binaries. How can I reach either of them to include the modifications to create the Windows/ARM64 binary?

Best regards,
Kevin Jones

[mpg123-devel] mpg123 1.25.12 released

[Thomas O. <tho...@or...>]

Hello there,

we got another security bugfix release of mpg123. Still working on the
proper feature update 1.26.0. Until then you got the fun to update to
version

1.25.12
-------

More credit to OSS-Fuzz. The ID3v2 parser code is not yet as hardened
as the actual MPEG decoder. The paranoid can disable it at build-time.
If you do not need it, this is a good idea, anyway: Code that is not
there, cannot be exploited. Speaking about exploits: The recent crop
of bugs trigger a denial of service (crash) worst-case, some invalid
ID3 data normally. Code injection maybe not totally ruled out (that one
write of a zero byte?), but does not seem easy. Update to be sure that
you are only suceptible to as of yet hidden bugs.

- libmpg123
-- Fix an out-of-bounds read of maximal two bytes for truncated RVA2 frames
   (oss-fuzz-bug 15975). The earlier fix around the same location needed
   one thought more. Actually, another though was needed, oss-fuzz-bug 16009
   documents the incomplete fix.
-- Fix an invalid write of one zero byte for empty ID3v2 frames that demand
   de-unsyncing (oss-fuzz-bug 16050).
-- Correct preprocessor syntax in mangle.h, no #error in a #define line.
   (bug 273, thanks to nmlgc).
- Fix dynamic build with gcc -fsanitize=address (check for all dl functions
  before deciding that separate -ldl is not needed).

Get it from the usual places …


Alrighty then,

Thomas

Re: [mpg123-devel] Allowing 3rd party tools to correctly link to libraries libmpg123 depends on

[Thomas O. <tho...@or...>]

Am Thu, 18 Jul 2019 12:13:02 +0200
schrieb Reino Wijnsma <rwi...@xs...>: 

> I see that libmpg123.pc.in
> <http://mpg123.org/cgi-bin/scm/mpg123/trunk/libmpg123.pc.in?view=markup>
> hasn't been updated in over 11 years and that you haven't considered
> my proposal anymore. MayI ask why not?

Thanks for asking. I simply forgot. I am working a lot, since over a
year, on the mpg123 1.26 feature release, among RL pressure.

The recent release of 1.25.11 was prompted by the security bugs found
by OSS-Fuzz.

So, let's revisit your proposal to include something.

You suggested placing @LIBS@ there. That is the list of libraries used
to link mpg123 against. This is a supperset of what is needed for
libmpg213. On Linux, this is just -lm. On Windows, we got -lshlwapi in
addition (-lws2_32 is again only for the mpg123 application itself).

We need to separate the libraries in the configure script to only
include what is necessary in LIBS.private. For libout123.pc, there is
also -ldl, usually.

I see that @LIBS@ is actually used for linking libmpg123. That is
wrong, but without consequence with linkers that drop unneeded
references. We need to separate things in the build system to put the
correct set of libs into the .pc files.

I'll work on not forgetting this for the 1.26 release.


Alrighty then,

Thomas

Re: [mpg123-devel] Allowing 3rd party tools to correctly link to libraries libmpg123 depends on

[Reino W. <rwi...@xs...>]

On 5-3-2018 1:17, Thomas Orgis <tho...@or...> wrote:
> Am Sun, 4 Mar 2018 14:32:17 +0100
> schrieb Reino Wijnsma <rwi...@xs...>:
>> --- libmpg123.pc.in.bak    2017-11-30 08:55:41.000000000 +0100
>> +++ libmpg123.pc.in    2018-03-04 01:46:59.203125000 +0100
>> @@ -8,4 +8,5 @@
>>  Requires:
>>  Version: @PACKAGE_VERSION@
>>  Libs: -L${libdir} -lmpg123
>> +Libs.private: @LIBS@
>>  Cflags: -I${includedir}
> I will consider this and also check other .pc files. Thanks.

On 18-7-2019 7:52, Thomas Orgis <tho...@or...> wrote:
> mpg123 1.25.11

I see that libmpg123.pc.in <http://mpg123.org/cgi-bin/scm/mpg123/trunk/libmpg123.pc.in?view=markup> hasn't been updated in over 11 years and that you haven't considered my proposal anymore.
MayI ask why not?

-- Reino

[mpg123-devel] mpg123 1.25.11 released

[Thomas O. <tho...@or...>]

Hi folks,

the OSS-Fuzz project integrated mpg123 and thus came a bunch of bugs
they found. Please update to the current mpg123 1.25.11 to get rid of
them.

1.25.11
-------

So, here is a number of bugs found by OSS-Fuzz. Credit to OSS-Fuzz
for the bunch, then.

- libmpg123:
-- Fix out-of-bounds reads in ID3 parser for unsynced frames.
   (oss-fuzz-bug 15852) 
-- Fix out-of-bounds read for RVA2 frames with non-delimited identifier.
   (oss-fuzz-bug 15852)
-- Fix implementation-defined parsing of RVA2 values. (oss-fuzz-bug 15862)
-- Fix undefined parsing of APE header for skipping. Also prevent
   endless loop on premature end of supposed APE header. (oss-fuzz-bug 15864)
-- Fix some syntax to make pedantic compiler happy.


Alrighty then,

Thomas


PS: I got no CVEs for these. If downstream distros would like to have
some, they are free to allocate numbers. Please tell me in that case,
so that I can at least post them to the mpg123 website.

Re: [mpg123-devel] CMake support

[Thomas O. <tho...@or...>]

Yeah, getting rid of these manual ports that pollute the repositories
would really be welcomed. CMake is clearly better than curating those.

The repository is public for read access. That's no invention of github;-)
Just use

 svn://scm.orgis.org/mpg123/trunk

. Happy hacking!


Alrighty then,

Thomas
unterwegs ohne Signatur

Re: [mpg123-devel] CMake support

[<kro...@gm...>]

Well, CMake can definitely simplify things for Windows and Mac. Not sure about sony-psp<div><br /></div><div>Can I get access to the SVN repo for reading? That would make coding and patch exchanging a bit easier</div><div>
        12:40, 14 апреля 2019 г., Thomas Orgis &lt;tho...@or...&gt;:<br /><blockquote><p>Hi,<br /><br />am Sun, 14 Apr 2019 11:58:38 +0300<br />schrieb Виталий Кирсанов &lt;<a href="mailto:kro...@gm...">kro...@gm...</a>&gt;: <br /><br /></p><blockquote> I see mpg123 doesn't have CMake support and I could help with this and<br /> write a suite of CMake files.<br /><br /> Would you welcome this and find it needful?<br /></blockquote><p><br />Tricky question. Good that you ask beforehand;-) My first reaction is<br />that mpg123 does not have CMake as build system because I do not prefer<br />it, and I still have that opinion when it comes to UNIX builds. But: We<br />have that manually-curated (not since quite some time) ports directory.<br /><br />        <a href="http://scm.orgis.org/view/mpg123/trunk/ports/">http://scm.orgis.org/view/mpg123/trunk/ports/</a><br /><br />That usually is broken after some mpg123 development and needs major<br />manual intervention, involving the respective IDEs. Can CMake be a<br />better replacement for these? As I see it, a major reason vor CMake is<br />that you can generate native build files for differing platforms,<br />namely Windows stuff.<br /><br />There is a big but to this but: I'd see it as addition to the current<br />build system and the current build system will stay (until I decide to<br />replace it with something simpler, geared to UNIX-like platforms). Not<br />that I am a huge fan of autotools. Build systems suck. Heck, we had<br />plain Makefiles alongside autotools for quite some time until I gave in<br />to get rid of them. But I am way less a fan of CMake on UNIX platforms,<br />where I live. I do lots of software builds with different packaging<br />toolsets, and SCons and CMake are all-too-ready to be a source of<br />trouble/additional work there. Some of this is inherent in the<br />different platform focus of the build systems.<br /><br />If we can replace the custom ports (mainly MSVC) with CMake, and it's<br />trivial enough to keep the CMake stuff up to date when new files are<br />added/renamed, I'd welcome a patch. I won't promise to test it often,<br />but you could set up something automated that does that, anyway.<br />Of course you can use CMake for your Linux builds as well, but I do not<br />see it becoming the primary build system there.<br /><br />Oh, the ports are usually only for building libmpg123, as the console<br />application is not that interesting for most people outside UNIX-like<br />environments. You have to decide if you will keep it at that. If<br />libmpg123 is all you need, the job is a lot easier. Only complication<br />is the assembly code for different CPUs, where a request is pending to<br />convert them from AT&amp;T to Intel syntax (will do for next release). But<br />you save yourself a lot of system specifics if the application is not<br />included. So far, it can be built using Mingw/MSYS or Cygwin.<br /><br />The mpg123 project is not very busy. You can send the patch via this<br />list, or add it to the tracker. You could even send it to<br /><a href="mailto:mai...@mp...">mai...@mp...</a>. IF fact, you just reminded me  that the -devel<br />lists still exists separately. I planned on merging it with the -user<br />list, due to the very low volume. It's a mature project, there is not<br />that much to discuss. But anyhow, we can keep the discussion here.<br /><br /><br />Alrighty then,<br /><br />Thomas<br /><br /><br />_______________________________________________<br />mpg123-devel mailing list<br /><a href="mailto:mpg...@li...">mpg...@li...</a><br /><a href="https://lists.sourceforge.net/lists/listinfo/mpg123-devel">https://lists.sourceforge.net/lists/listinfo/mpg123-devel</a><br /></p></blockquote><br /><br />-- <br />Отправлено из мобильного приложения Яндекс.Почты
    </div>

Re: [mpg123-devel] CMake support

[Thomas O. <tho...@or...>]

Hi,

am Sun, 14 Apr 2019 11:58:38 +0300
schrieb Виталий Кирсанов <kro...@gm...>: 

> I see mpg123 doesn't have CMake support and I could help with this and
> write a suite of CMake files.
> 
> Would you welcome this and find it needful?

Tricky question. Good that you ask beforehand;-) My first reaction is
that mpg123 does not have CMake as build system because I do not prefer
it, and I still have that opinion when it comes to UNIX builds. But: We
have that manually-curated (not since quite some time) ports directory.

	http://scm.orgis.org/view/mpg123/trunk/ports/

That usually is broken after some mpg123 development and needs major
manual intervention, involving the respective IDEs. Can CMake be a
better replacement for these? As I see it, a major reason vor CMake is
that you can generate native build files for differing platforms,
namely Windows stuff.

There is a big but to this but: I'd see it as addition to the current
build system and the current build system will stay (until I decide to
replace it with something simpler, geared to UNIX-like platforms). Not
that I am a huge fan of autotools. Build systems suck. Heck, we had
plain Makefiles alongside autotools for quite some time until I gave in
to get rid of them. But I am way less a fan of CMake on UNIX platforms,
where I live. I do lots of software builds with different packaging
toolsets, and SCons and CMake are all-too-ready to be a source of
trouble/additional work there. Some of this is inherent in the
different platform focus of the build systems.

If we can replace the custom ports (mainly MSVC) with CMake, and it's
trivial enough to keep the CMake stuff up to date when new files are
added/renamed, I'd welcome a patch. I won't promise to test it often,
but you could set up something automated that does that, anyway.
Of course you can use CMake for your Linux builds as well, but I do not
see it becoming the primary build system there.

Oh, the ports are usually only for building libmpg123, as the console
application is not that interesting for most people outside UNIX-like
environments. You have to decide if you will keep it at that. If
libmpg123 is all you need, the job is a lot easier. Only complication
is the assembly code for different CPUs, where a request is pending to
convert them from AT&T to Intel syntax (will do for next release). But
you save yourself a lot of system specifics if the application is not
included. So far, it can be built using Mingw/MSYS or Cygwin.

The mpg123 project is not very busy. You can send the patch via this
list, or add it to the tracker. You could even send it to
mai...@mp.... IF fact, you just reminded me  that the -devel
lists still exists separately. I planned on merging it with the -user
list, due to the very low volume. It's a mature project, there is not
that much to discuss. But anyhow, we can keep the discussion here.


Alrighty then,

Thomas

[mpg123-devel] CMake support

[Виталий К. <kro...@gm...>]

Hi everyone and nice to meet you.

I see mpg123 doesn't have CMake support and I could help with this and
write a suite of CMake files.

Would you welcome this and find it needful? If yes what is the proper way
to contribute? I mean sending you a patch then passing through a review and
so on.

BR, Vitaly Kirsanov
skype: vkirsan

Re: [mpg123-devel] Code example of mpg123_index and mpg123_set_index

[Sven B. <s.b...@gm...>]

I got it myself, the destination handle was null, which is no good ;).
Thanks anyway, now it works also on C# side :).

Thank you and have a nice weekend :)!

Mit freundlichen Grüßen
Sven Baus
0175/49 50 309


-----Ursprüngliche Nachricht-----
Von: Sven Baus [mailto:s.b...@gm...] 
Gesendet: Samstag, 13. Oktober 2018 16:22
An: 'Thomas Orgis'
Cc: mpg...@li...
Betreff: Re: [mpg123-devel] Code example of mpg123_index and
mpg123_set_index

So, thanks for your help. I got one more step near:

[DllImport(Mpg123Dll)]public unsafe static extern int mpg123_index(IntPtr
mh, long **offsets, long *step, ulong *fill);
[DllImport(Mpg123Dll)]public unsafe static extern int
mpg123_set_index(IntPtr mh, long* offsets, long step, ulong fill);

public static Boolean CopyIndex(IntPtr _sourceHandle,IntPtr
_destinationHandle)
{
	Boolean copiedIndex = false;
	unsafe
	{
		long* offsets = null; 
		long step = 0;
		ulong fill = 0;
		int result = MPGImport.mpg123_index(_sourceHandle, &offsets,
&step, &fill);
		if (result == (int)MPGImport.mpg123_errors.MPG123_OK)
		{
			result =
MPGImport.mpg123_set_index(_destinationHandle, offsets, step, fill);
<---- result is always -1

			//MPG123_ERR = -1,		/**< Generic Error
*/

			if (result == (int)mpg123_errors.MPG123_OK)
			{
				copiedIndex = true;
			}
		}
	}
	return copiedIndex;
}

Do you have an idea, why this could fail?


Mit freundlichen Grüßen
Sven Baus
0175/49 50 309

-----Ursprüngliche Nachricht-----
Von: Thomas Orgis [mailto:tho...@or...] 
Gesendet: Freitag, 12. Oktober 2018 11:37
An: s.baus86
Cc: mpg...@li...
Betreff: Re: [mpg123-devel] Code example of mpg123_index and
mpg123_set_index

Am Fri, 12 Oct 2018 11:18:18 +0200
schrieb "s.baus86" <s.b...@gm...>: 

> Ok, I see, pIndex has a different initialisation. What I googled, it
> is a pointer and you hand the adress of that pointer to mpg123_index,
> correct? 

Yes. You want the address of the internal storage of the index copied
to your pointer variable.


Alrighty then,

Thomas



_______________________________________________
mpg123-devel mailing list
mpg...@li...
https://lists.sourceforge.net/lists/listinfo/mpg123-devel

Re: [mpg123-devel] Code example of mpg123_index and mpg123_set_index

[Sven B. <s.b...@gm...>]

So, thanks for your help. I got one more step near:

[DllImport(Mpg123Dll)]public unsafe static extern int mpg123_index(IntPtr
mh, long **offsets, long *step, ulong *fill);
[DllImport(Mpg123Dll)]public unsafe static extern int
mpg123_set_index(IntPtr mh, long* offsets, long step, ulong fill);

public static Boolean CopyIndex(IntPtr _sourceHandle,IntPtr
_destinationHandle)
{
	Boolean copiedIndex = false;
	unsafe
	{
		long* offsets = null; 
		long step = 0;
		ulong fill = 0;
		int result = MPGImport.mpg123_index(_sourceHandle, &offsets,
&step, &fill);
		if (result == (int)MPGImport.mpg123_errors.MPG123_OK)
		{
			result =
MPGImport.mpg123_set_index(_destinationHandle, offsets, step, fill);
<---- result is always -1

			//MPG123_ERR = -1,		/**< Generic Error
*/

			if (result == (int)mpg123_errors.MPG123_OK)
			{
				copiedIndex = true;
			}
		}
	}
	return copiedIndex;
}

Do you have an idea, why this could fail?


Mit freundlichen Grüßen
Sven Baus
0175/49 50 309

-----Ursprüngliche Nachricht-----
Von: Thomas Orgis [mailto:tho...@or...] 
Gesendet: Freitag, 12. Oktober 2018 11:37
An: s.baus86
Cc: mpg...@li...
Betreff: Re: [mpg123-devel] Code example of mpg123_index and
mpg123_set_index

Am Fri, 12 Oct 2018 11:18:18 +0200
schrieb "s.baus86" <s.b...@gm...>: 

> Ok, I see, pIndex has a different initialisation. What I googled, it
> is a pointer and you hand the adress of that pointer to mpg123_index,
> correct? 

Yes. You want the address of the internal storage of the index copied
to your pointer variable.


Alrighty then,

Thomas

Re: [mpg123-devel] Code example of mpg123_index and mpg123_set_index

[s.baus86 <s.b...@gm...>]

Ok, thanks for your mail. I'm currently handing in an empty pointer, but keep getting system.accessviolation exceptions. Any idea, why?


Mit freundlichen GrüßenSven Baus0175/4950309
-------- Ursprüngliche Nachricht --------Von: Thomas Orgis <tho...@or...> Datum: 12.10.18  11:37  (GMT+01:00) An: "s.baus86" <s.b...@gm...> Cc: mpg...@li... Betreff: Re: [mpg123-devel] Code example of mpg123_index and  
  mpg123_set_index 
Am Fri, 12 Oct 2018 11:18:18 +0200
schrieb "s.baus86" <s.b...@gm...>: 

> Ok, I see, pIndex has a different initialisation. What I googled, it
> is a pointer and you hand the adress of that pointer to mpg123_index,
> correct? 

Yes. You want the address of the internal storage of the index copied
to your pointer variable.


Alrighty then,

Thomas

Re: [mpg123-devel] Code example of mpg123_index and mpg123_set_index

[Thomas O. <tho...@or...>]

Am Fri, 12 Oct 2018 11:18:18 +0200
schrieb "s.baus86" <s.b...@gm...>: 

> Ok, I see, pIndex has a different initialisation. What I googled, it
> is a pointer and you hand the adress of that pointer to mpg123_index,
> correct? 

Yes. You want the address of the internal storage of the index copied
to your pointer variable.


Alrighty then,

Thomas

Re: [mpg123-devel] Code example of mpg123_index and mpg123_set_index

[s.baus86 <s.b...@gm...>]

Ok, I see, pIndex has a different initialisation. What I googled, it is a pointer and you hand the adress of that pointer to mpg123_index, correct? Sorry if I'm asking to much, but currently its a bit mysterious to me, how this works.


Mit freundlichen GrüßenSven Baus0175/4950309
-------- Ursprüngliche Nachricht --------Von: Thomas Orgis <tho...@or...> Datum: 29.09.18  21:58  (GMT+01:00) An: Sven Baus <s.b...@gm...> Cc: mpg...@li... Betreff: Re: [mpg123-devel] Code example of mpg123_index and
  mpg123_set_index 
Am Sat, 29 Sep 2018 16:45:28 +0200
schrieb "Sven Baus" <s.b...@gm...>: 

> So I think the following code should work, correct?
> 
> off_t*	pIndex;
> off_t*	pStep;
> size_t*	pFill;
> 
> mpg123_index(pHandle,pIndex,pStep,pFill);

Not quite. More like this:

off_t*	pIndex;
off_t	pStep;
size_t	pFill;

mpg123_index(pHandle,&pIndex,&pStep,&pFill);

The difference is crucial and explains access violations.

You see use of this in src/term.c, in print_index() (and yes, the
printing formats there should not just cast things to long, other parts
of mpg123 do better).


Alrighty then,

Thomas

Re: [mpg123-devel] Code example of mpg123_index and mpg123_set_index

[Thomas O. <tho...@or...>]

Am Sat, 29 Sep 2018 16:45:28 +0200
schrieb "Sven Baus" <s.b...@gm...>: 

> So I think the following code should work, correct?
> 
> off_t*	pIndex;
> off_t*	pStep;
> size_t*	pFill;
> 
> mpg123_index(pHandle,pIndex,pStep,pFill);

Not quite. More like this:

off_t*	pIndex;
off_t	pStep;
size_t	pFill;

mpg123_index(pHandle,&pIndex,&pStep,&pFill);

The difference is crucial and explains access violations.

You see use of this in src/term.c, in print_index() (and yes, the
printing formats there should not just cast things to long, other parts
of mpg123 do better).


Alrighty then,

Thomas

Re: [mpg123-devel] Code example of mpg123_index and mpg123_set_index

[Sven B. <s.b...@gm...>]

So I think the following code should work, correct?

off_t*	pIndex;
off_t*	pStep;
size_t*	pFill;

mpg123_index(pHandle,pIndex,pStep,pFill);

I'm currently transforming the C code to C# (I use a wrapper directly to
libmpg123-0.dll) and I'm trying to understand on how to use the function,
since I'm currently just getting access violation exceptions.

The current code looks like this:
IntPtr pOffset = Marshal.AllocHGlobal(sizeof(int));
IntPtr pStep = IntPtr.Zero;
IntPtr pFill = IntPtr.Zero;
int result = MPGImport.mpg123_index(mp3ScanHandle,pOffset,pStep,pFill);

Sincerely
Sven

Mit freundlichen Grüßen
Sven Baus
0175/49 50 309


-----Ursprüngliche Nachricht-----
Von: Thomas Orgis [mailto:tho...@or...] 
Gesendet: Donnerstag, 27. September 2018 23:14
An: Sven Baus
Cc: mpg...@li...
Betreff: Re: [mpg123-devel] Code example of mpg123_index and
mpg123_set_index

Am Thu, 27 Sep 2018 19:20:37 +0200
schrieb "Sven Baus" <s.b...@gm...>: 

> What do you mean by re-reading the index? 

Reading back into a data structure When you have stored it externally.
Generally, you are supposed to copy the index from mpg123_index() to
someplace before continuing work with the handle. You just get a
pointer to internal memory!

> I thought
> mpg123_index() wants a pointer to an array

It wants a pointer to a pointer, which can be an array in C. You need
to read up and understand more about the relationship between pointers
and arrays in C. It is not that tricky, but you must be familiar with
the meanings.

> code example would be nice, of how to retreave the index and hand it to
> another handle ;).
> 
When the original handle is not modified in between, you can just give
the value of the pointer from mpg123_index() and the fill and step
values directly to mpg123_set_index(). It will make an internal copy.
You do not need to allocate any arrays.


Alrighty then,

Thomas

Re: [mpg123-devel] Code example of mpg123_index and mpg123_set_index

[Thomas O. <tho...@or...>]

Am Thu, 27 Sep 2018 19:20:37 +0200
schrieb "Sven Baus" <s.b...@gm...>: 

> What do you mean by re-reading the index? 

Reading back into a data structure When you have stored it externally.
Generally, you are supposed to copy the index from mpg123_index() to
someplace before continuing work with the handle. You just get a
pointer to internal memory!

> I thought
> mpg123_index() wants a pointer to an array

It wants a pointer to a pointer, which can be an array in C. You need
to read up and understand more about the relationship between pointers
and arrays in C. It is not that tricky, but you must be familiar with
the meanings.

> code example would be nice, of how to retreave the index and hand it to
> another handle ;).
> 
When the original handle is not modified in between, you can just give
the value of the pointer from mpg123_index() and the fill and step
values directly to mpg123_set_index(). It will make an internal copy.
You do not need to allocate any arrays.


Alrighty then,

Thomas

Re: [mpg123-devel] Code example of mpg123_index and mpg123_set_index

[Sven B. <s.b...@gm...>]

Hello,

so just a more question: You said " Of course,
when re-reading the index for handing it to mpg123_set_index(), you
need to allocate fill elements of the correct type and then store the
values there." What do you mean by re-reading the index? I thought
mpg123_index() wants a pointer to an array and tells after the call, how
much the array has been filled, or am I missunderstanding something? Maybe a
code example would be nice, of how to retreave the index and hand it to
another handle ;).

Thanks for your answers
Sven

Mit freundlichen Grüßen
Sven Baus
0175/49 50 309

-----Ursprüngliche Nachricht-----
Von: Thomas Orgis [mailto:tho...@or...] 
Gesendet: Dienstag, 25. September 2018 12:57
An: mpg...@li...
Cc: s.baus86
Betreff: Re: [mpg123-devel] Code example of mpg123_index and
mpg123_set_index

Am Tue, 25 Sep 2018 11:30:18 +0200
schrieb "s.baus86" <s.b...@gm...>: 

> Thanks for your mail. How should the array be initialised?

The argument to mpg123_index() is just a pointer without associated
memory. You get the size of the associated memory via fill. Of course,
when re-reading the index for handing it to mpg123_set_index(), you
need to allocate fill elements of the correct type and then store the
values there.

> Sorry for asking, but I'm not that familiar with C.

You need some understanding of C with pointers and data types. If in
doubt, try to read a bit more, the net is ripe with tutorials and
examples.

> int[] indexArray = new int[100]?! Or is it that simple?

If 100 is the number of elements, not some arbitrary fixed value, this
sounds about right for C++ (I'd write int* instead of int[]) except for
the fact that int is not the correct type. It is off_t. That an be
identical to int, but mostly will not be.


Alrighty then,

Thomas

Re: [mpg123-devel] Code example of mpg123_index and mpg123_set_index

[Thomas O. <tho...@or...>]

Am Tue, 25 Sep 2018 11:30:18 +0200
schrieb "s.baus86" <s.b...@gm...>: 

> Thanks for your mail. How should the array be initialised?

The argument to mpg123_index() is just a pointer without associated
memory. You get the size of the associated memory via fill. Of course,
when re-reading the index for handing it to mpg123_set_index(), you
need to allocate fill elements of the correct type and then store the
values there.

> Sorry for asking, but I'm not that familiar with C.

You need some understanding of C with pointers and data types. If in
doubt, try to read a bit more, the net is ripe with tutorials and
examples.

> int[] indexArray = new int[100]?! Or is it that simple?

If 100 is the number of elements, not some arbitrary fixed value, this
sounds about right for C++ (I'd write int* instead of int[]) except for
the fact that int is not the correct type. It is off_t. That an be
identical to int, but mostly will not be.


Alrighty then,

Thomas

Re: [mpg123-devel] Code example of mpg123_index and mpg123_set_index

[s.baus86 <s.b...@gm...>]

Hello,
Thanks for your mail. How should the array be initialised? Sorry for asking, but I'm not that familiar with C. What kind of array should it be? I don't think its a int[] indexArray = new int[100]?! Or is it that simple?
SincerelySven


Mit freundlichen GrüßenSven Baus0175/4950309
-------- Ursprüngliche Nachricht --------Von: Thomas Orgis <tho...@or...> Datum: 25.09.18  10:38  (GMT+01:00) An: mpg...@li... Betreff: Re: [mpg123-devel] Code example of mpg123_index and mpg123_set_index 
Am Mon, 24 Sep 2018 19:49:08 +0200
schrieb "Sven Baus" <s.b...@gm...>: 

> https://www.mpg123.de/api/group__mpg123__seek.shtml the definitions, but
> what about the input parameters? How should the be initialised, etc?!

Well, the parameters to mpg123_index() are just addresses to store the
named values, including the pointer to the internal index array. You
make a copy of that, then hand a pointer to the copy and the other
values to mpg123_set_index(), which will make an internal copy again.


Alrighty then,

Thomas


_______________________________________________
mpg123-devel mailing list
mpg...@li...
https://lists.sourceforge.net/lists/listinfo/mpg123-devel

Re: [mpg123-devel] Code example of mpg123_index and mpg123_set_index

[Thomas O. <tho...@or...>]

Am Mon, 24 Sep 2018 19:49:08 +0200
schrieb "Sven Baus" <s.b...@gm...>: 

> https://www.mpg123.de/api/group__mpg123__seek.shtml the definitions, but
> what about the input parameters? How should the be initialised, etc?!

Well, the parameters to mpg123_index() are just addresses to store the
named values, including the pointer to the internal index array. You
make a copy of that, then hand a pointer to the copy and the other
values to mpg123_set_index(), which will make an internal copy again.


Alrighty then,

Thomas

[mpg123-devel] Code example of mpg123_index and mpg123_set_index

[Sven B. <s.b...@gm...>]

Hello everybody,

I'm trying to use mpg123_index and mpg123_set_index to scan a mp3 file and
afterwards apply the scanned index to a playback handle. Are there any code
examples, of how to use the functions? I can see here
https://www.mpg123.de/api/group__mpg123__seek.shtml the definitions, but
what about the input parameters? How should the be initialised, etc?!

Thanks for any help!
Sincerely 
Sven

Mit freundlichen Grüßen
Sven Baus
0175/49 50 309

Re: [mpg123-devel] mpg123_length not correct

[Thomas O. <tho...@or...>]

Hi,

I'm sorry that I lack time and fluency in C#. I have a basic issue with
such:

>                         int errorScanHandle = …
>                         mp3ScanHandle = MPGImport.mpg123_new(null,
> (IntPtr)errorScanHandle);

When you do (IntPtr)errorScanHandle, does that take the address of the
integer errorScanHandle or does it cast its value to an address?

>                                     //this.mp3Handle = mp3ScanHandle;
> <-- if uncommented, the application crashes
>                                     //TODO: Delete old handle
>  
> MPGImport.mpg123_seek(this.mp3Handle,samples,(int) SeekOrigin.Begin);

If these were just C pointers, I wouldn't see why it should crash,
unless your locking is not safe. Is the decoder thread also blocking on
this.mutex?


Alrighty then,

Thomas

Re: [mpg123-devel] mpg123_length not correct

[Sven B. <s.b...@gm...>]

Hello again,

sorry for another mail, but I got rid of the bad handle error. A process
outside the code caused the handle to be reset to IntPtr.Zero which caused
the bad handle error. Now I almost got it, but I'm stuck a bit:

This code runs only, if I don't switch the handle:

tScan = new Thread(delegate()
                    {
                        int errorScanHandle =
(int)MPGImport.mpg123_errors.MPG123_OK;
                        IntPtr mp3ScanHandle = IntPtr.Zero;
                        mp3ScanHandle = MPGImport.mpg123_new(null,
(IntPtr)errorScanHandle);
                        if (mp3ScanHandle != IntPtr.Zero)
                        {
                            if (MPGImport.mpg123_open(mp3ScanHandle,
_filename) != (int)MPGImport.mpg123_errors.MPG123_OK)
                            {
                                log.error("Error opening file: " +
MPGImport.mpg123_strerror(mp3ScanHandle));
                            }
                            log.debug("Starting mpg123_scan ...");
                            int scanResult =
MPGImport.mpg123_scan(mp3ScanHandle);
                            log.debug("mpg123_scan finished");
                            if (scanResult !=
(int)MPGImport.mpg123_errors.MPG123_OK)
                            {
                                log.error("Error scanning file: " +
MPGImport.mpg123_strerror(mp3ScanHandle));
                            }
                            else 
                            {
                                //Swap the mp3 handle
                                this.mutex.WaitOne();
                                if (this.mp3Handle != IntPtr.Zero) 
                                {
                                    int samples =
MPGImport.mpg123_tell(this.mp3Handle);
                                    log.debug("Swapping mp3 handle!");
                                    //this.mp3Handle = mp3ScanHandle;
<-- if uncommented, the application crashes
                                    //TODO: Delete old handle
 
MPGImport.mpg123_seek(this.mp3Handle,samples,(int) SeekOrigin.Begin);
                                }
                                else
                                {
                                    MPGImport.mpg123_delete(mp3ScanHandle);
                                    mp3ScanHandle = IntPtr.Zero;
                                }
                                this.mutex.ReleaseMutex();
                            }
                        }
                        else 
                        {
                            log.error("Failed to instance scan handle!");
                        }
                        tScan = null;
                    });
                    tScan.IsBackground = true;
                    tScan.Start();

Do you have an idea, why the swap could fail?


Mit freundlichen Grüßen
Sven Baus
0175/49 50 309


-----Ursprüngliche Nachricht-----
Von: Sven Baus [mailto:s.b...@gm...] 
Gesendet: Sonntag, 27. Mai 2018 10:58
An: 'Thomas Orgis'
Cc: 'mpg...@li...'
Betreff: AW: [mpg123-devel] mpg123_length not correct

Hi Thomas,

so, I tried to get in contact with the developer, but there is no response
on my mails. I also tried to get help on stackoverflow, but there is also
nobody able to help me. 

Your idea with switching the handle doesn't sound to bad, but I'm currently
facing the issue, that I always get a bad handle error.

Some code:
tScan = new Thread(delegate()
                        {
                            int errorScanHandle =
(int)MPGImport.mpg123_errors.MPG123_OK;
                            IntPtr mp3ScanHandle = IntPtr.Zero;
                            mp3ScanHandle = MPGImport.mpg123_new(null,
(IntPtr)errorScanHandle);
                            if (mp3ScanHandle != IntPtr.Zero)
                            {
                                if (MPGImport.mpg123_open(mp3ScanHandle,
_filename) != (int)MPGImport.mpg123_errors.MPG123_OK)
                                {
                                    log.error("Error opening file: " +
MPGImport.mpg123_strerror(mp3ScanHandle));
                                }
                            }
                            if (mp3ScanHandle != IntPtr.Zero)
                            {
                                log.debug("Starting mpg123_scan ...");
                                if (MPGImport.mpg123_scan(mp3ScanHandle) !=
(int)MPGImport.mpg123_errors.MPG123_OK)
                                {
                                    log.error("Error scanning file: " +
MPGImport.mpg123_strerror(mp3ScanHandle));
                                }
                                log.debug("mpg123_scan finished");
                                kvpScannedHandle = new KeyValuePair<string,
IntPtr>(_filename, mp3ScanHandle);
                                //Swap the mp3 handle
                                this.mutex.WaitOne();
                                int samples =
MPGImport.mpg123_tell(this.mp3Handle);
                                if (samples !=
(int)MPGImport.mpg123_errors.MPG123_OK)
                                {
                                    int errorCode =
MPGImport.mpg123_errcode(this.mp3Handle);
                                    if (errorCode != (int)
MPGImport.mpg123_errors.MPG123_BAD_HANDLE) 
                                    {
                                        log.error("Failed to get the sample
position from the handle! " + MPGImport.mpg123_plain_strerror(errorCode));
                                    }
                                    else 
                                    {
                                        log.error("Failed to get the sample
position from the handle! Bad Handle");
                                    }
                                }
                                else 
                                {
                                    log.debug("Swapping mp3 handle!");
                                    MPGImport.mpg123_delete(this.mp3Handle);
                                    this.mp3Handle = IntPtr.Zero;
                                    this.mp3Handle = mp3ScanHandle;
 
MPGImport.mpg123_seek(this.mp3Handle,samples,(int) SeekOrigin.Begin);
                                }
                                this.mutex.ReleaseMutex();
                            }
                            tScan = null;
                        });
                        tScan.IsBackground = true;
                        tScan.Start();

The code runs into the error " Failed to get the sample position from the
handle! Bad Handle", but I have no idea, why :(. Do you know, why this error
is triggered?

Thanks for your response.

Mit freundlichen Grüßen
Sven Baus
0175/49 50 309

-----Ursprüngliche Nachricht-----
Von: Thomas Orgis [mailto:tho...@or...] 
Gesendet: Sonntag, 25. Februar 2018 17:32
An: Sven Baus
Cc: mpg...@li...
Betreff: Re: [mpg123-devel] mpg123_length not correct

Am Sun, 25 Feb 2018 15:46:08 +0100
schrieb "Sven Baus" <s.b...@gm...>: 

> You ship a CLR ready dll with mpg123? 

Erm, well, we have the binding code / MSVC project in our repo from
an external contributor. It for shure would need some work since its last
update. But we do not provide binaries built from that. There is a DLL
in the windows binary downloads, that is built using MinGW.

I cannot contribute any personal experience with the CLR stuff.

> Concerning the error string bug, it gets really strange now:

> I have no idea, why, but it seems to me, that whenever the string
> would be a correct message, the function crashes.

Really weird stuff happening usually means that there is memory
corruption. An invalid / too large error code results in a
specific conditional being triggered. Proper values get looked up in a
char* array. Is there weird index math going on? Or is there a
fundamental difference between

	char* array[] = { "text", "more text" };
	return array[0];

and

	return "text";

?
 
> I wait also for a reply of the author of MPGIMPORT

Let's wait for that. If this works out and the behaviour can be fixed,
we might drop our 2008clr port for this project. The mpg123 project
doesn't really have the people to support this ourselves.

We have a very simple test case for what is going on. Heck, you might
even ask Microsoft;-)


Alrighty then,

Thomas

Re: [mpg123-devel] mpg123_length not correct

[Sven B. <s.b...@gm...>]

Hi Thomas,

so, I tried to get in contact with the developer, but there is no response
on my mails. I also tried to get help on stackoverflow, but there is also
nobody able to help me. 

Your idea with switching the handle doesn't sound to bad, but I'm currently
facing the issue, that I always get a bad handle error.

Some code:
tScan = new Thread(delegate()
                        {
                            int errorScanHandle =
(int)MPGImport.mpg123_errors.MPG123_OK;
                            IntPtr mp3ScanHandle = IntPtr.Zero;
                            mp3ScanHandle = MPGImport.mpg123_new(null,
(IntPtr)errorScanHandle);
                            if (mp3ScanHandle != IntPtr.Zero)
                            {
                                if (MPGImport.mpg123_open(mp3ScanHandle,
_filename) != (int)MPGImport.mpg123_errors.MPG123_OK)
                                {
                                    log.error("Error opening file: " +
MPGImport.mpg123_strerror(mp3ScanHandle));
                                }
                            }
                            if (mp3ScanHandle != IntPtr.Zero)
                            {
                                log.debug("Starting mpg123_scan ...");
                                if (MPGImport.mpg123_scan(mp3ScanHandle) !=
(int)MPGImport.mpg123_errors.MPG123_OK)
                                {
                                    log.error("Error scanning file: " +
MPGImport.mpg123_strerror(mp3ScanHandle));
                                }
                                log.debug("mpg123_scan finished");
                                kvpScannedHandle = new KeyValuePair<string,
IntPtr>(_filename, mp3ScanHandle);
                                //Swap the mp3 handle
                                this.mutex.WaitOne();
                                int samples =
MPGImport.mpg123_tell(this.mp3Handle);
                                if (samples !=
(int)MPGImport.mpg123_errors.MPG123_OK)
                                {
                                    int errorCode =
MPGImport.mpg123_errcode(this.mp3Handle);
                                    if (errorCode != (int)
MPGImport.mpg123_errors.MPG123_BAD_HANDLE) 
                                    {
                                        log.error("Failed to get the sample
position from the handle! " + MPGImport.mpg123_plain_strerror(errorCode));
                                    }
                                    else 
                                    {
                                        log.error("Failed to get the sample
position from the handle! Bad Handle");
                                    }
                                }
                                else 
                                {
                                    log.debug("Swapping mp3 handle!");
                                    MPGImport.mpg123_delete(this.mp3Handle);
                                    this.mp3Handle = IntPtr.Zero;
                                    this.mp3Handle = mp3ScanHandle;
 
MPGImport.mpg123_seek(this.mp3Handle,samples,(int) SeekOrigin.Begin);
                                }
                                this.mutex.ReleaseMutex();
                            }
                            tScan = null;
                        });
                        tScan.IsBackground = true;
                        tScan.Start();

The code runs into the error " Failed to get the sample position from the
handle! Bad Handle", but I have no idea, why :(. Do you know, why this error
is triggered?

Thanks for your response.

Mit freundlichen Grüßen
Sven Baus
0175/49 50 309

-----Ursprüngliche Nachricht-----
Von: Thomas Orgis [mailto:tho...@or...] 
Gesendet: Sonntag, 25. Februar 2018 17:32
An: Sven Baus
Cc: mpg...@li...
Betreff: Re: [mpg123-devel] mpg123_length not correct

Am Sun, 25 Feb 2018 15:46:08 +0100
schrieb "Sven Baus" <s.b...@gm...>: 

> You ship a CLR ready dll with mpg123? 

Erm, well, we have the binding code / MSVC project in our repo from
an external contributor. It for shure would need some work since its last
update. But we do not provide binaries built from that. There is a DLL
in the windows binary downloads, that is built using MinGW.

I cannot contribute any personal experience with the CLR stuff.

> Concerning the error string bug, it gets really strange now:

> I have no idea, why, but it seems to me, that whenever the string
> would be a correct message, the function crashes.

Really weird stuff happening usually means that there is memory
corruption. An invalid / too large error code results in a
specific conditional being triggered. Proper values get looked up in a
char* array. Is there weird index math going on? Or is there a
fundamental difference between

	char* array[] = { "text", "more text" };
	return array[0];

and

	return "text";

?
 
> I wait also for a reply of the author of MPGIMPORT

Let's wait for that. If this works out and the behaviour can be fixed,
we might drop our 2008clr port for this project. The mpg123 project
doesn't really have the people to support this ourselves.

We have a very simple test case for what is going on. Heck, you might
even ask Microsoft;-)


Alrighty then,

Thomas