jazzplusplus-devel

Re: [jazzplusplus-devel] configure.ac update -please test-

[Kevin C. <ke...@co...>]

On 10 May 2008 at 12:31, Pete Stieber <pst...@ve...> wrote:

> Donald and Kevin, could I get you to try the Linux build and make sure 
> the jack/flsynth/qsynth is working.  I don't have access to a Linux 
> setup at the moment and that would really help me out.

I'm a bit busy at the moment, but I'll work on that ASAP.

> BTW I always seem to have a heck of a time with Mandriva.  Jazz++ can 
> hang it, it doesn't want to recognize my keyboard...

Hmmm.  I wonder what would happen with a Linus kernel on Mandriva?

Thanks.....
 
--
Kevin

Re: [jazzplusplus-devel] configure.ac update -please test-

[D.B. M. <db...@ho...>]

> Date: Mon, 12 May 2008 09:40:48 -0700
> Subject: Re: [jazzplusplus-devel] configure.ac update -please test-
>
>
> On 10 May 2008 at 12:31, Pete Stieber wrote:
>
>> Donald and Kevin, could I get you to try the Linux build and make sure
>> the jack/flsynth/qsynth is working.  I don't have access to a Linux
>> setup at the moment and that would really help me out.

..for some reason I didn't receive this original posting - thankfully Kevin
responded to it and I was able to pick up the thread....

>
>
>> BTW I always seem to have a heck of a time with Mandriva.  Jazz++ can
>> hang it, it doesn't want to recognize my keyboard...

Curious...I hadn't noticed that here myself with my initial builds, however
subsequent to this call for testing....yes..there's something strange going on
with input devices on linux. Here's the detail....

My svn source version here is current at rev. 518

Distcleaned my local svn, rebuilt the source and installed (completes successfully)
The build-log is attached just in case there's something to see in it.

Started up jack/fluidsynth/qsynth as previously configured and tested okay.

Launching jazz almost immediately caused a segfault....

jazz[3277]: segfault at 1 ip 424c82 sp 7fffe79bf840 error 4 in jazz[400000+573000]
jazz[3287]: segfault at 1 ip 424c82 sp 7fff0dd7b040 error 4 in jazz[400000+573000]
jazz[3288]: segfault at 1 ip 424c82 sp 7fff0af091d0 error 4 in jazz[400000+573000]
jazz[3299]: segfault at 1 ip 424c82 sp 7fff5cf1c1f0 error 4 in jazz[400000+573000]
jazz[3300]: segfault at 1 ip 424c82 sp 7fff12a95d60 error 4 in jazz[400000+573000]
jazz[3301]: segfault at 1 ip 424c82 sp 7fff8092bbf0 error 4 in jazz[400000+573000]

This is me (re)starting jazz half-a-dozen times looking for clues. On the 7th attempt ;

jazz[3310]: segfault at 1 ip 426cea sp 7fff8e2aa1e0 error 4 in jazz[400000+5ad000]


That last one 'stole' both my mouse&keyboard completely, and left me with no X input.
The system was still running though, so I moved to one of my other boxes, and
ssh'd into this system to see if there were any clues...unfortunately and AFAICT there
were no clues or errors to seen...in fact, everything looked fine...just X had lost input.


One of the last things I saw in console that launched jazz leading up to this was ;

"Invalid input device"


I believe I might have recovered the situation with an simple X restart, however I
decided to reboot the system to be sure of a 'clean slate'.

Taking Pete's obtrusion about mouse&keyboard weirdness into account, and on
a hunch thinking 'invalid input device' might be somehow related, I decided on a
different approach...

I downloaded Takashi Iwai's 'vkeybd' source, compiled & installed. I then started
this virtual midi keyboard and check it registered in alsa midi devices -- it does,
along with jack/fluidsynth/qsynth.

I then restarted jazz to discover things were actually sane again. It prompted me
for input device (I of course chose the vkeybd client), and output device (selected
qsynth)....hey presto, it's all working again as currently expected..well...caveat ;

subscribe: Operation not permitted

I see this in dmesg when jazz finishes configuration dialogs and starts.

Occasionally;

jazz[3458]: segfault at 0 ip 7f606dd1dd39 sp 7fff77552410 error 4 in libasound.so.2.0.0[7f606dc87000+d2000]

is seen in dmesg -- I'm still trying to accurately recreate this one to find the cause,
but it's being elusive...


I hope this helps some -- anything else/specific you want to know?


>Kevin C wote;
>
> Hmmm.  I wonder what would happen with a Linus kernel on Mandriva?
>

Just for the record, this account of proceedings here was using linux 2.6.25.3

I think my next move here is to double-check my alsalib installation - it's
been a while since I checked what's current.

Regards,

Donald B


_________________________________________________________________
It's simple! Sell your car for just $30 at CarPoint.com.au
http://a.ninemsn.com.au/b.aspx?URL=http%3A%2F%2Fsecure%2Dau%2Eimrworldwide%2Ecom%2Fcgi%2Dbin%2Fa%2Fci%5F450304%2Fet%5F2%2Fcg%5F801459%2Fpi%5F1004813%2Fai%5F859641&_t=762955845&_r=tig_OCT07&_m=EXT

Re: [jazzplusplus-devel] configure.ac update -please test-

[D.B. M. <db...@ho...>]

>> Date: Mon, 12 May 2008 09:40:48 -0700
>> Subject: Re: [jazzplusplus-devel] configure.ac update -please test-
>>
>>
>> On 10 May 2008 at 12:31, Pete Stieber wrote:
>>

Just confirming that this type of error;

jazz[3247]: segfault at 0 ip 7fa62e0e1b49 sp 7fff379157d0 error 4 in libasound.so.2.0.0[7fa62e04b000+d1000]


is actually emitted when jazz cannot seemingly find a valid input device,
ie; I didn't have vkeybd running. Previous (elusive) reports of this I 
posted, were apparently related to me changing things in the qsynth
connection panel 'on the fly'....sorry if that created any confusion.

I did update my alsalib btw, but such made no difference here.


Regards,

Donald B

_________________________________________________________________
Search for local singles online @ Lavalife - Click here
http://a.ninemsn.com.au/b.aspx?URL=http%3A%2F%2Flavalife9%2Eninemsn%2Ecom%2Eau%2Fclickthru%2Fclickthru%2Eact%3Fid%3Dninemsn%26context%3Dan99%26locale%3Den%5FAU%26a%3D30290&_t=764581033&_r=email_taglines_Search_OCT07&_m=EXT

Re: [jazzplusplus-devel] configure.ac update -please test-

[Pete S. <pst...@gm...>]

DM = D.B. Moore
DM> Just confirming that this type of error;
DM>
DM> jazz[3247]: segfault at 0 ip 7fa62e0e1b49 sp 7fff379157d0 error 4 in 
libasound.so.2.0.0[7fa62e04b000+d1000]
DM>
DM> is actually emitted when jazz cannot seemingly
DM> find a valid input device, ie; I didn't have
DM> vkeybd running. Previous (elusive) reports of
DM> this I posted, were apparently related to me
DM> changing things in the qsynth connection panel
DM> 'on the fly'....sorry if that created any confusion.
DM>
DM> I did update my alsalib btw, but such made no
DM> difference here.

Donald,

When you are debugging add --enable-debug to your configure line.

After you finish the build, run the code using gdb...

gdb jazz

Then type "run" when you get a prompt.  When the code crashes, generate 
a back trace using the "bt" command.  Full gdb documentation can be 
found at http://sourceware.org/gdb/current/onlinedocs/gdb_toc.html

Reply back with the result of the back trace.

It will be a while before I get to it.  Maybe not until this weekend.

Pete

Re: [jazzplusplus-devel] configure.ac update -please test-

[D.B. M. <db...@ho...>]

----------------------------------------
> Date: Mon, 12 May 2008 21:48:25 -0700

 
> DM = D.B. Moore
> DM> Just confirming that this type of error;
> DM>
> DM> jazz[3247]: segfault at 0 ip 7fa62e0e1b49 sp 7fff379157d0 error 4 in 
> libasound.so.2.0.0[7fa62e04b000+d1000]
> DM>
> DM> is actually emitted when jazz cannot seemingly
> DM> find a valid input device, ie; I didn't have
> DM> vkeybd running. Previous (elusive) reports of
> DM> this I posted, were apparently related to me
> DM> changing things in the qsynth connection panel
> DM> 'on the fly'....sorry if that created any confusion.
> DM>
> DM> I did update my alsalib btw, but such made no
> DM> difference here.
> 
> Donald,
> 
> When you are debugging add --enable-debug to your configure line.
> 
> After you finish the build, run the code using gdb...
> 
> gdb jazz
> 
> Then type "run" when you get a prompt.  When the code crashes, generate 
> a back trace using the "bt" command.  Full gdb documentation can be 
> found at http://sourceware.org/gdb/current/onlinedocs/gdb_toc.html
> 
> Reply back with the result of the back trace.
> 
> It will be a while before I get to it.  Maybe not until this weekend.
> 
> Pete
> 

Yep, no problem.... the back trace without vkeybd running but with
jack/fluidsynth/qsynth running and ready to go (complete session);

(gdb) run
Starting program: /root/Jazz++/TestInstall/bin/jazz 
Failed to read a valid object file image from memory.
[Thread debugging using libthread_db enabled]
[New Thread 140090596853472 (LWP 3341)]
JZProject::ReadConfiguration() ConfFileNameAndPath:
  "/root/Jazz++/TestInstall/share/Jazz++/jazz.cfg"
JZConfiguration::LoadConfig:
  "/root/Jazz++/TestInstall/share/Jazz++/jazz.cfg"
Include synthesizer configuration file "gs.jzi"
FindFile: Immediate hit on file "gs.jzi"
Include file "gsdrmset.jzi"
FindFile: Immediate hit on file "gsdrmset.jzi"
Include file "gmdrmnam.jzi"
FindFile: Immediate hit on file "gmdrmnam.jzi"
Include file "gsvoices.jzi"
FindFile: Immediate hit on file "gsvoices.jzi"
Include file "ctrlnam.jzi"
FindFile: Immediate hit on file "ctrlnam.jzi"
created client:port = 131:0
invalid output device, so selecting one
Program received signal SIGSEGV, Segmentation fault.
[Switching to Thread 140090596853472 (LWP 3341)]
snd_seq_port_subscribe_set_sender (info=0x7fff6a45d310, addr=0x0) at seq.c:2432
2432            memcpy(&info->sender, addr, sizeof(*addr));
Current language:  auto; currently c

(gdb) bt
#0  snd_seq_port_subscribe_set_sender (info=0x7fff6a45d310, addr=0x0)
    at seq.c:2432
#1  0x0000000000426a59 in tAlsaPlayer::subscribe_inp (this=0xb68020, inp=0)
    at ../../jazz/src/AlsaPlayer.cpp:219
#2  0x0000000000427793 in tAlsaPlayer (this=0xb68020, song=0x7f69623ea010)
    at ../../jazz/src/AlsaPlayer.cpp:122
#3  0x0000000000423638 in tAlsaAudioPlayer (this=0xb68020, 
    pSong=0x7f69623ea010) at ../../jazz/src/AlsaDriver.cpp:127
#4  0x00000000004b9ed2 in JZProject (this=0x7f69623ea010)
    at ../../jazz/src/Project.cpp:194
#5  0x0000000000498948 in JZJazzPlusPlusApplication::OnInit (this=0xafe690)
    at ../../jazz/src/JazzPlusPlusApplication.cpp:127
#6  0x0000000000498d96 in wxAppConsole::CallOnInit (this=0xafe690)
    at /usr/local/wx287/include/wx-2.8/wx/app.h:76
#7  0x000000000079a70a in wxEntry ()
#8  0x0000000000498d10 in main (argc=1, argv=0x7fff6a45dc68)
    at ../../jazz/src/JazzPlusPlusApplication.cpp:63



Hope this helps put some light on the issue.


Regards,

Donald B
_________________________________________________________________
Search for local singles online @ Lavalife - Click here
http://a.ninemsn.com.au/b.aspx?URL=http%3A%2F%2Flavalife9%2Eninemsn%2Ecom%2Eau%2Fclickthru%2Fclickthru%2Eact%3Fid%3Dninemsn%26context%3Dan99%26locale%3Den%5FAU%26a%3D30290&_t=764581033&_r=email_taglines_Search_OCT07&_m=EXT

Re: [jazzplusplus-devel] configure.ac update -please test-

[Pete S. <pst...@gm...>]

DM = D.B. Moore
DM> Yep, no problem.... the back trace without vkeybd
DM> running but with jack/fluidsynth/qsynth running
DM> and ready to go (complete session);
<Snip>
> Program received signal SIGSEGV, Segmentation fault.
> [Switching to Thread 140090596853472 (LWP 3341)]
> snd_seq_port_subscribe_set_sender (info=0x7fff6a45d310, addr=0x0) at seq.c:2432
> 2432            memcpy(&info->sender, addr, sizeof(*addr));
> Current language:  auto; currently c
> 
> (gdb) bt
> #0  snd_seq_port_subscribe_set_sender (info=0x7fff6a45d310, addr=0x0)
>     at seq.c:2432
> #1  0x0000000000426a59 in tAlsaPlayer::subscribe_inp (this=0xb68020, inp=0)
>     at ../../jazz/src/AlsaPlayer.cpp:219

My current guess is an invalid input or output device index.  Could you 
update you source tree, rebuild, run again and let me know how it goes?

Thanks,
Pete

Re: [jazzplusplus-devel] configure.ac update -please test-

[D.B. M. <db...@ho...>]

Greetings,

----------------------------------------
> Date: Tue, 13 May 2008 21:38:17 -0700
> Subject: Re: [jazzplusplus-devel] configure.ac update -please test-
> 
> DM = D.B. Moore
> DM> Yep, no problem.... the back trace without vkeybd
> DM> running but with jack/fluidsynth/qsynth running
> DM> and ready to go (complete session);
> 
>> Program received signal SIGSEGV, Segmentation fault.
>> [Switching to Thread 140090596853472 (LWP 3341)]
>> snd_seq_port_subscribe_set_sender (info=0x7fff6a45d310, addr=0x0) at seq.c:2432
>> 2432            memcpy(&info->sender, addr, sizeof(*addr));
>> Current language:  auto; currently c
>> 
>> (gdb) bt
>> #0  snd_seq_port_subscribe_set_sender (info=0x7fff6a45d310, addr=0x0)
>>     at seq.c:2432
>> #1  0x0000000000426a59 in tAlsaPlayer::subscribe_inp (this=0xb68020, inp=0)
>>     at ../../jazz/src/AlsaPlayer.cpp:219
> 
> My current guess is an invalid input or output device index.  Could you 
> update you source tree, rebuild, run again and let me know how it goes?
> 


Same env as before..ie; vkeybd not running but with jack/fluidsynth/qsynth running
and ready to go.

Test build @ revision 527  (complete session);

(gdb) run
Starting program: /root/Jazz++/TestInstall/bin/jazz 
warning: no loadable sections found in added symbol-file system-supplied DSO at 0x7fff30bfe000
[Thread debugging using libthread_db enabled]
[New Thread 140725286655712 (LWP 11863)]
JZProject::ReadConfiguration() ConfFileNameAndPath:
  "/root/Jazz++/TestInstall/share/Jazz++/jazz.cfg"
JZConfiguration::LoadConfig:
  "/root/Jazz++/TestInstall/share/Jazz++/jazz.cfg"
Include synthesizer configuration file "gs.jzi"
FindFile: Immediate hit on file "gs.jzi"
Include file "gsdrmset.jzi"
FindFile: Immediate hit on file "gsdrmset.jzi"
Include file "gmdrmnam.jzi"
FindFile: Immediate hit on file "gmdrmnam.jzi"
Include file "gsvoices.jzi"
FindFile: Immediate hit on file "gsvoices.jzi"
Include file "ctrlnam.jzi"
FindFile: Immediate hit on file "ctrlnam.jzi"
created client:port = 131:0
Input device count: 0
INFO: input device is -1, so selecting one.
no device found!
Input device is: -1
INFO: output device is -1, so selecting one.

//pops 'select output device' dialog window
// candidates -- client 128 qjackctl | fluid synth (11837)  Synth input port (11837.0) 
//selecting fluid synth entry, click on  OK


WARNING: The output device index is out of range!

Program received signal SIGSEGV, Segmentation fault.
[Switching to Thread 140569115023072 (LWP 11874)]
0x0000000000426cea in tAlsaPlayer::SetSoftThru (this=0xb68010, on=1, idev=0, odev=0) at ../../jazz/src/AlsaPlayer.cpp:224
224         thru->SetSource(iaddr[ithru].client, iaddr[ithru].port);

(gdb) bt
#0  0x0000000000426cea in tAlsaPlayer::SetSoftThru (this=0xb68010, on=1, idev=0, odev=0) at ../../jazz/src/AlsaPlayer.cpp:224
#1  0x0000000000427ce0 in tAlsaPlayer (this=0xb68010, song=0x7fd8cc267010) at ../../jazz/src/AlsaPlayer.cpp:182
#2  0x0000000000423638 in tAlsaAudioPlayer (this=0xb68010, pSong=0x7fd8cc267010) at ../../jazz/src/AlsaDriver.cpp:127
#3  0x00000000004ba196 in JZProject (this=0x7fd8cc267010) at ../../jazz/src/Project.cpp:194
#4  0x0000000000498db0 in JZJazzPlusPlusApplication::OnInit (this=0xafe690) at ../../jazz/src/JazzPlusPlusApplication.cpp:127
#5  0x00000000004991fe in wxAppConsole::CallOnInit (this=0xafe690) at /usr/local/wx287/include/wx-2.8/wx/app.h:76
#6  0x000000000079a9ca in wxEntry ()
#7  0x0000000000499178 in main (argc=1, argv=0x7fffd42daae8) at ../../jazz/src/JazzPlusPlusApplication.cpp:63



Hope this helps, regards

Donald B
_________________________________________________________________
Are you paid what you're worth? Find out: SEEK Salary Centre
http://a.ninemsn.com.au/b.aspx?URL=http%3A%2F%2Fninemsn%2Eseek%2Ecom%2Eau%2Fcareer%2Dresources%2Fsalary%2Dcentre%2F%3Ftracking%3Dsk%3Ahet%3Asc%3Anine%3A0%3Ahot%3Atext&_t=764565661&_r=OCT07_endtext_salary&_m=EXT

Re: [jazzplusplus-devel] configure.ac update -please test-

[D.B. M. <db...@ho...>]

----------------------------------------
> Date: Thu, 15 May 2008 02:14:28 +0000
> Subject: Re: [jazzplusplus-devel] configure.ac update -please test-
> 
> 
> Greetings,
> 
> ----------------------------------------
>> Date: Tue, 13 May 2008 21:38:17 -0700
>> Subject: Re: [jazzplusplus-devel] configure.ac update -please test-
>> 
>> DM = D.B. Moore
>> DM> Yep, no problem.... the back trace without vkeybd
>> DM> running but with jack/fluidsynth/qsynth running
>> DM> and ready to go (complete session);
>> 
>>> Program received signal SIGSEGV, Segmentation fault.
>>> [Switching to Thread 140090596853472 (LWP 3341)]
>>> snd_seq_port_subscribe_set_sender (info=0x7fff6a45d310, addr=0x0) at seq.c:2432
>>> 2432            memcpy(&info->sender, addr, sizeof(*addr));
>>> Current language:  auto; currently c
>>> 
>>> (gdb) bt
>>> #0  snd_seq_port_subscribe_set_sender (info=0x7fff6a45d310, addr=0x0)
>>>     at seq.c:2432
>>> #1  0x0000000000426a59 in tAlsaPlayer::subscribe_inp (this=0xb68020, inp=0)
>>>     at ../../jazz/src/AlsaPlayer.cpp:219
>> 
>> My current guess is an invalid input or output device index.  Could you 
>> update you source tree, rebuild, run again and let me know how it goes?
>> 
> 

I had a closer look at this problem just now....here's what I've found ;

~/Jazz++/TestInstall/bin# aconnect -l -o
client 130: 'AlsaModularSynth' [type=user]
    0 'ams             '
        Connected From: 129:0
client 131: 'FLUID Synth (qsynth)' [type=user]
    0 'Synth input port (qsynth:0)'
        Connected From: 129:0

~/Jazz++/TestInstall/bin# aconnect -l -i
client 0: 'System' [type=kernel]
    0 'Timer           '
    1 'Announce        '
        Connecting To: 128:0, 129:0
client 130: 'AlsaModularSynth' [type=user]
    1 'ams             '
        Connecting To: 129:0[real:0]
    2 'ams             '
        Connecting To: 129:0[real:0]


Qsynth is not creating/registering any alsa_input_ports,
whereas AMS (and vkeybd) do.

If I start jazz and  if finds no alsa_input_ports, it will
*not* pop the 'select input' dialog window, and instead
continue bravely forth and pop the 'select output' dialog
window instead.

In this condition, it doesn't matter which output port
is selected, clicking on 'OK' to confirm the output selection
results in the crash observed.

Provided jazz finds at least one (or more) valid alsa_input_ports
to try and connect to at init, the 'select input' dialog box is
displayed, followed by the 'select output' dialog box, and jazz
starts and runs as currently expected.

Regards,

Donald B   
_________________________________________________________________
Search for local singles online @ Lavalife - Click here
http://a.ninemsn.com.au/b.aspx?URL=http%3A%2F%2Flavalife9%2Eninemsn%2Ecom%2Eau%2Fclickthru%2Fclickthru%2Eact%3Fid%3Dninemsn%26context%3Dan99%26locale%3Den%5FAU%26a%3D30290&_t=764581033&_r=email_taglines_Search_OCT07&_m=EXT

Re: [jazzplusplus-devel] configure.ac update -please test-

[Pete S. <pst...@gm...>]

DM> Same env as before..ie; vkeybd not running but
DM> with jack/fluidsynth/qsynth running and ready to go.
DM>
DM> Test build @ revision 527  (complete session);
DM>
DM> (gdb) run
DM> Starting program: /root/Jazz++/TestInstall/bin/jazz
DM> warning: no loadable sections found in added symbol-file
DM> system-supplied DSO at 0x7fff30bfe000
DM> [Thread debugging using libthread_db enabled]
DM> [New Thread 140725286655712 (LWP 11863)]
DM> JZProject::ReadConfiguration() ConfFileNameAndPath:
DM>   "/root/Jazz++/TestInstall/share/Jazz++/jazz.cfg"
DM> JZConfiguration::LoadConfig:
DM>   "/root/Jazz++/TestInstall/share/Jazz++/jazz.cfg"
DM> Include synthesizer configuration file "gs.jzi"
DM> FindFile: Immediate hit on file "gs.jzi"
DM> Include file "gsdrmset.jzi"
DM> FindFile: Immediate hit on file "gsdrmset.jzi"
DM> Include file "gmdrmnam.jzi"
DM> FindFile: Immediate hit on file "gmdrmnam.jzi"
DM> Include file "gsvoices.jzi"
DM> FindFile: Immediate hit on file "gsvoices.jzi"
DM> Include file "ctrlnam.jzi"
DM> FindFile: Immediate hit on file "ctrlnam.jzi"
DM> created client:port = 131:0
DM> Input device count: 0
DM> INFO: input device is -1, so selecting one.
DM> no device found!
DM> Input device is: -1
DM> INFO: output device is -1, so selecting one.
DM>
DM> //pops 'select output device' dialog window
DM> // candidates --
DM> client 128 qjackctl
DM> fluid synth (11837)  Synth input port (11837.0)
DM> //selecting fluid synth entry, click on  OK
DM>
DM>
DM> WARNING: The output device index is out of range!
DM>
DM> Program received signal SIGSEGV, Segmentation fault.
DM> [Switching to Thread 140569115023072 (LWP 11874)]
DM> 0x0000000000426cea in tAlsaPlayer::SetSoftThru
DM> (this=0xb68010, on=1, idev=0, odev=0) at
DM> ../../jazz/src/AlsaPlayer.cpp:224
DM> 224  thru->SetSource(iaddr[ithru].client, iaddr[ithru].port);
DM>
DM> Hope this helps, regards

It does.

After you last crash report I guessed that there were no value inputs on 
your machine, and I believe I fixed the cause of that initial crash. 
The debug output that reads...

Input device count: 0
INFO: input device is -1, so selecting one.
no device found!
Input device is: -1

Shows me we are now getting past that problem.

Now you have discovered another place where the code attempts to set the 
  ALSA thru for the input device, but there are no input devices 
available.  I'll bet this is the same problem Kevin C. is having. 
Unfortunately the Mandriva box I attempt debugging on hangs completely 
when I run Jazz++.

I'm hoping Kevin will reply to my Mandriva post indicating if the 
"Spring version" of Mandriva may help me out.

Otherwise, I'm applying blind patches (i.e. patches that haven't been 
tested on anything).

I'll see what I can do.
Pete

Re: [jazzplusplus-devel] configure.ac update -please test-

[D.B. M. <db...@ho...>]

Greetings,

----------------------------------------
> Date: Sat, 17 May 2008 15:27:25 -0700
> Subject: Re: [jazzplusplus-devel] configure.ac update -please test-
> 
> DM> Same env as before..ie; vkeybd not running but
> DM> with jack/fluidsynth/qsynth running and ready to go.
> DM>
> DM> Test build @ revision 527  (complete session);
> DM>

> DM> WARNING: The output device index is out of range!
> DM>
> DM> Program received signal SIGSEGV, Segmentation fault.
> DM> [Switching to Thread 140569115023072 (LWP 11874)]
> DM> 0x0000000000426cea in tAlsaPlayer::SetSoftThru
> DM> (this=0xb68010, on=1, idev=0, odev=0) at
> DM> ../../jazz/src/AlsaPlayer.cpp:224
> DM> 224  thru->SetSource(iaddr[ithru].client, iaddr[ithru].port);
> DM>
> DM> Hope this helps, regards
> 
> It does.
> 
> After you last crash report I guessed that there were no value inputs on 
> your machine, and I believe I fixed the cause of that initial crash. 
> The debug output that reads...
> 
> Input device count: 0
> INFO: input device is -1, so selecting one.
> no device found!
> Input device is: -1
> 
> Shows me we are now getting past that problem.
> 
> Now you have discovered another place where the code attempts to set the 
>   ALSA thru for the input device, but there are no input devices 
> available.  I'll bet this is the same problem Kevin C. is having. 
> Unfortunately the Mandriva box I attempt debugging on hangs completely 
> when I run Jazz++.

..that's curious..

> 
> I'm hoping Kevin will reply to my Mandriva post indicating if the 
> "Spring version" of Mandriva may help me out.
> 
> Otherwise, I'm applying blind patches (i.e. patches that haven't been 
> tested on anything).
> 
> I'll see what I can do.


I wasn't actually aware your mandriva was that broken by jazz.
Or is that the other way around?... ;-)

I'm pretty prompt testing things however...so patch on...


Regards,

Donald B
_________________________________________________________________
Are you paid what you're worth? Find out: SEEK Salary Centre
http://a.ninemsn.com.au/b.aspx?URL=http%3A%2F%2Fninemsn%2Eseek%2Ecom%2Eau%2Fcareer%2Dresources%2Fsalary%2Dcentre%2F%3Ftracking%3Dsk%3Ahet%3Asc%3Anine%3A0%3Ahot%3Atext&_t=764565661&_r=OCT07_endtext_salary&_m=EXT

Re: [jazzplusplus-devel] configure.ac update -please test-

[Pete S. <pst...@gm...>]

DM = D.B. Moore
DM>>> Same env as before..ie; vkeybd not running but
DM>>> with jack/fluidsynth/qsynth running and ready to go.
DM>>>
DM>>> Test build @ revision 527  (complete session);
DM>>>
DM>>> WARNING: The output device index is out of range!
DM>>>
DM>>> Program received signal SIGSEGV, Segmentation fault.
DM>>> [Switching to Thread 140569115023072 (LWP 11874)]
DM>>> 0x0000000000426cea in tAlsaPlayer::SetSoftThru
DM>>> (this=0xb68010, on=1, idev=0, odev=0) at
DM>>> ../../jazz/src/AlsaPlayer.cpp:224
DM>>> 224  thru->SetSource(iaddr[ithru].client, iaddr[ithru].port);
DM>>>
DM>>> Hope this helps, regards

PS = Pete Stieber
PS>> It does.
PS>>
PS>> After you last crash report I guessed that there
PS>> were no value inputs on your machine, and I
PS>> believe I fixed the cause of that initial crash.
PS>> The debug output that reads...
PS>>
PS>> Input device count: 0
PS>> INFO: input device is -1, so selecting one.
PS>> no device found!
PS>> Input device is: -1
PS>>
PS>> Shows me we are now getting past that problem.
PS>>
PS>> Now you have discovered another place where the
PS>> code attempts to set the ALSA thru for the input
PS>> device, but there are no input devices available.
PS>> I'll bet this is the same problem Kevin C. is
PS>> having.  Unfortunately the Mandriva box I attempt
PS>> debugging on hangs completely when I run Jazz++.

DM> ..that's curious..

PS>> I'm hoping Kevin will reply to my Mandriva post
PS>> indicating if the "Spring version" of Mandriva
PS>> may help me out.
PS>>
PS>> Otherwise, I'm applying blind patches (i.e. patches
PS>> that haven't been tested on anything).
PS>>
PS>> I'll see what I can do.

DM> I wasn't actually aware your mandriva was
DM> that broken by jazz.  Or is that the other
DM> way around?... ;-)
DM>
DM> I'm pretty prompt testing things however...so
DM> patch on...

OK...  I just made some changed to AlsaPlayer.cpp that should write more 
debug output and check for invalid device IDs.  Jazz++ isn't hanging my 
Mandriva install any more (maybe that was keyboard/mouse related), but 
it isn't playing either.  This isn't a Jazz++ issue, but I am having 
problems starting the jack daemon in Mandriva.  See the "Starting jackd 
on Mandriva" thread.


Donald & Kevin,

Could you update to the latest svn and give it a try for me.

Thanks,
Pete

Re: [jazzplusplus-devel] configure.ac update -please test-

[D.B. M. <db...@ho...>]

Greetings,

----------------------------------------
> Date: Sun, 18 May 2008 16:49:25 -0700
> Subject: Re: [jazzplusplus-devel] configure.ac update -please test-
> 
> DM = D.B. Moore
> DM>>> Same env as before..ie; vkeybd not running but
> DM>>> with jack/fluidsynth/qsynth running and ready to go.
> DM>>>

...this above bit remains the same....

> DM>>> Test build @ revision 544  

...that's where I am now....


Good work Pete!  It seems you've got this problem nailed down;


 (gdb) run
Starting program: /root/Jazz++/TestInstall/bin/jazz 
warning: no loadable sections found in added symbol-file system-supplied DSO at 
0x7fff3a5fe000
[Thread debugging using libthread_db enabled]
[New Thread 140051137894112 (LWP 13419)]
JZProject::ReadConfiguration() ConfFileNameAndPath:
  "/root/Jazz++/TestInstall/share/Jazz++/jazz.cfg"
JZConfiguration::LoadConfig:
  "/root/Jazz++/TestInstall/share/Jazz++/jazz.cfg"
Include synthesizer configuration file "gs.jzi"
FindFile: Immediate hit on file "gs.jzi"
Include file "gsdrmset.jzi"
FindFile: Immediate hit on file "gsdrmset.jzi"
Include file "gmdrmnam.jzi"
FindFile: Immediate hit on file "gmdrmnam.jzi"
Include file "gsvoices.jzi"
FindFile: Immediate hit on file "gsvoices.jzi"
Include file "ctrlnam.jzi"
FindFile: Immediate hit on file "ctrlnam.jzi"
INFO: Created client:port = 131:0
INFO: Input device count: 0
INFO: Output device count: 2
Output Devices
Client-128 qjackctl = 128:0
FLUID Synth (qsynth) Synth input port (qsynth:0) = 130:0
INFO: input device is -1, so selecting one.
INFO: No device found!
INFO: Input device is: -1
INFO: Output device is -1, so selecting one.
WARNING: The input MIDI thru device index (0) is out of range!
Setting the value to -1
subscribe: Operation not permitted

///here I muse, muse, muse, poke at buttons etc etc///
///and then I quit the program//


Program exited normally.



Wonderful!....just as currently expected.

Mind you, just *how* I got it to work a week or so
ago eludes my rational mind....but not to worry,
we seem to be over it now....


Regards,

Donald B

_________________________________________________________________
Are you paid what you're worth? Find out: SEEK Salary Centre
http://a.ninemsn.com.au/b.aspx?URL=http%3A%2F%2Fninemsn%2Eseek%2Ecom%2Eau%2Fcareer%2Dresources%2Fsalary%2Dcentre%2F%3Ftracking%3Dsk%3Ahet%3Asc%3Anine%3A0%3Ahot%3Atext&_t=764565661&_r=OCT07_endtext_salary&_m=EXT

Re: [jazzplusplus-devel] configure.ac update -please test-

[Pete S. <pst...@gm...>]

DM = D.B. Moore
DM>> Same env as before..ie; vkeybd not running but
DM>> with jack/fluidsynth/qsynth running and ready to go.

DM> ...this above bit remains the same....

DM> Test build @ revision 544
DM> ...that's where I am now....
DM>
DM> Good work Pete!  It seems you've got this
DM> problem nailed down;
DM>
DM> (gdb) run
DM> Starting program: /root/Jazz++/TestInstall/bin/jazz
DM> warning: no loadable sections found in added symbol-file
DM> system-supplied DSO at 0x7fff3a5fe000
DM> [Thread debugging using libthread_db enabled]
DM> [New Thread 140051137894112 (LWP 13419)]
DM> JZProject::ReadConfiguration() ConfFileNameAndPath:
DM>   "/root/Jazz++/TestInstall/share/Jazz++/jazz.cfg"
DM> JZConfiguration::LoadConfig:
DM>   "/root/Jazz++/TestInstall/share/Jazz++/jazz.cfg"
DM> Include synthesizer configuration file "gs.jzi"
DM> FindFile: Immediate hit on file "gs.jzi"
DM> Include file "gsdrmset.jzi"
DM> FindFile: Immediate hit on file "gsdrmset.jzi"
DM> Include file "gmdrmnam.jzi"
DM> FindFile: Immediate hit on file "gmdrmnam.jzi"
DM> Include file "gsvoices.jzi"
DM> FindFile: Immediate hit on file "gsvoices.jzi"
DM> Include file "ctrlnam.jzi"
DM> FindFile: Immediate hit on file "ctrlnam.jzi"
DM> INFO: Created client:port = 131:0
DM> INFO: Input device count: 0
DM> INFO: Output device count: 2
DM> Output Devices
DM> Client-128 qjackctl = 128:0
DM> FLUID Synth (qsynth) Synth input port (qsynth:0) = 130:0
DM> INFO: input device is -1, so selecting one.
DM> INFO: No device found!
DM> INFO: Input device is: -1
DM> INFO: Output device is -1, so selecting one.
DM> WARNING: The input MIDI thru device index (0) is out of range!
DM> Setting the value to -1
DM> subscribe: Operation not permitted
DM>
DM> ///here I muse, muse, muse, poke at buttons etc etc///
DM> ///and then I quit the program//
DM>
DM> Program exited normally.
DM>
DM> Wonderful!....just as currently expected.

Great to hear!  How about you Kevin... any better?

Pete

Re: [jazzplusplus-devel] configure.ac update -please test-

[D.B. M. <db...@ho...>]

----------------------------------------
> Date: Tue, 13 May 2008 05:41:08 +0000
> Subject: Re: [jazzplusplus-devel] configure.ac update -please test-
> 
>>> 
>> When you are debugging add --enable-debug to your configure line.
>> 
>> After you finish the build, run the code using gdb...
>> 
>> gdb jazz
>> 
>> Then type "run" when you get a prompt.  When the code crashes, generate 
>> a back trace using the "bt" command.  Full gdb documentation can be 
>> found at http://sourceware.org/gdb/current/onlinedocs/gdb_toc.html
>> 
>> Reply back with the result of the back trace.
>> 

I actually know about gdb, backtraces and so forth. I doubt one could
build a Free software based systems for as long as I have without needing this
'software scalpel' at some stage or another. However, I tend to 'follow the book'
very closely, as to expose 'userland', because sometimes that's a useful strategy.
So we see here...the configuration switch '--enable-debug' (which is part of
the current svn build instructions anyhow), and the quick gdb primer of how
to obtain useful debugging information to help developers out here.

I think somehow that for as long as a debug build is the default build, these
brief gdb instructions would be quite pertinent for users to read somewhere
on the website.  Test builders are a valuable resource, and so I believe it might
useful to include such documentation on the site so users who do build jazz++
can quickly and easily supply (and know how to create) useful information 
if they find their jazz++ build crashes. If nothing else, it'll save having to spell
that backtrace passage out, over and over again. Remember, everyone building
from svn is effectively a test builder at present.

Let me know what we think here - if it sounds like a good idea and we agree,
I'll edit the htdocs in some appropriate spot to include such instructions.  


>> It will be a while before I get to it.  Maybe not until this weekend.
>> 
>> Pete
>> 
> 
> Yep, no problem.... the back trace without vkeybd running but with
> jack/fluidsynth/qsynth running and ready to go (complete session);
> 
> (gdb) run
> Starting program: /root/Jazz++/TestInstall/bin/jazz 
> Failed to read a valid object file image from memory.
> [Thread debugging using libthread_db enabled]
> [New Thread 140090596853472 (LWP 3341)]
> JZProject::ReadConfiguration() ConfFileNameAndPath:
>   "/root/Jazz++/TestInstall/share/Jazz++/jazz.cfg"
> JZConfiguration::LoadConfig:
>   "/root/Jazz++/TestInstall/share/Jazz++/jazz.cfg"
> Include synthesizer configuration file "gs.jzi"
> FindFile: Immediate hit on file "gs.jzi"
> Include file "gsdrmset.jzi"
> FindFile: Immediate hit on file "gsdrmset.jzi"
> Include file "gmdrmnam.jzi"
> FindFile: Immediate hit on file "gmdrmnam.jzi"
> Include file "gsvoices.jzi"
> FindFile: Immediate hit on file "gsvoices.jzi"
> Include file "ctrlnam.jzi"
> FindFile: Immediate hit on file "ctrlnam.jzi"
> created client:port = 131:0
> invalid output device, so selecting one
> Program received signal SIGSEGV, Segmentation fault.
> [Switching to Thread 140090596853472 (LWP 3341)]
> snd_seq_port_subscribe_set_sender (info=0x7fff6a45d310, addr=0x0) at seq.c:2432
> 2432            memcpy(&info->sender, addr, sizeof(*addr));
> Current language:  auto; currently c
> 
> (gdb) bt
> #0  snd_seq_port_subscribe_set_sender (info=0x7fff6a45d310, addr=0x0)
>     at seq.c:2432
> #1  0x0000000000426a59 in tAlsaPlayer::subscribe_inp (this=0xb68020, inp=0)
>     at ../../jazz/src/AlsaPlayer.cpp:219
> #2  0x0000000000427793 in tAlsaPlayer (this=0xb68020, song=0x7f69623ea010)
>     at ../../jazz/src/AlsaPlayer.cpp:122
> #3  0x0000000000423638 in tAlsaAudioPlayer (this=0xb68020, 
>     pSong=0x7f69623ea010) at ../../jazz/src/AlsaDriver.cpp:127
> #4  0x00000000004b9ed2 in JZProject (this=0x7f69623ea010)
>     at ../../jazz/src/Project.cpp:194
> #5  0x0000000000498948 in JZJazzPlusPlusApplication::OnInit (this=0xafe690)
>     at ../../jazz/src/JazzPlusPlusApplication.cpp:127
> #6  0x0000000000498d96 in wxAppConsole::CallOnInit (this=0xafe690)
>     at /usr/local/wx287/include/wx-2.8/wx/app.h:76
> #7  0x000000000079a70a in wxEntry ()
> #8  0x0000000000498d10 in main (argc=1, argv=0x7fff6a45dc68)
>     at ../../jazz/src/JazzPlusPlusApplication.cpp:63
> 

Just a further note on this -- as I mentioned in a previous posting, 
I'm keen to test out and document jazz++'s interoperability and setup
with other midi based applications on linux. Last night, I got around
to trying the AlsaModularSynth (AMS) application with jazz++.

To use AMS without a hardware midi/synth soundcard, relies on a
very similar setup to jack/fluidsynth/qsynth. One starts AMS with 
the --jack cmdline option, and when jazz++ is started and prompts
for an output device, the user selects the AMS output device entry.

Yes...it works, and curiously enough it never once spawned the above
crash in a two hour session of me mucking around with things, stopping
and restarting things...I could not seemingly reproduce this error above
using jack/AMS and jazz++ at all. Nor did it ever steal my mouse and
keyboard I hasten to add -- everything behaved itself.

I'm no coder, but I thought I'd mention this here, just in case there's
more to the above error than immediately meets the eye.

Regards,

Donald B 

   



> 
> 
> Hope this helps put some light on the issue.
> 
> 
> Regards,
> 
> Donald B
> _________________________________________________________________
> Search for local singles online @ Lavalife - Click here
> http://a.ninemsn.com.au/b.aspx?URL=http%3A%2F%2Flavalife9%2Eninemsn%2Ecom%2Eau%2Fclickthru%2Fclickthru%2Eact%3Fid%3Dninemsn%26context%3Dan99%26locale%3Den%5FAU%26a%3D30290&_t=764581033&_r=email_taglines_Search_OCT07&_m=EXT
> -------------------------------------------------------------------------
> This SF.net email is sponsored by: Microsoft 
> Defy all challenges. Microsoft(R) Visual Studio 2008. 
> http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
> _______________________________________________
> jazzplusplus-devel mailing list
> jaz...@li...
> https://lists.sourceforge.net/lists/listinfo/jazzplusplus-devel

_________________________________________________________________
Are you paid what you're worth? Find out: SEEK Salary Centre
http://a.ninemsn.com.au/b.aspx?URL=http%3A%2F%2Fninemsn%2Eseek%2Ecom%2Eau%2Fcareer%2Dresources%2Fsalary%2Dcentre%2F%3Ftracking%3Dsk%3Ahet%3Asc%3Anine%3A0%3Ahot%3Atext&_t=764565661&_r=OCT07_endtext_salary&_m=EXT

Re: [jazzplusplus-devel] configure.ac update -please test-

[Pete S. <pst...@gm...>]

DM = D.B. Moore
DM> I actually know about gdb, backtraces and so forth.
DM> I doubt one could build a Free software based systems
DM> for as long as I have without needing this 'software
DM> scalpel' at some stage or another. However, I tend to
DM> 'follow the book' very closely, as to expose
DM> 'userland', because sometimes that's a useful strategy.
DM> So we see here...the configuration switch
DM> '--enable-debug' (which is part of the current svn
DM> build instructions anyhow), and the quick gdb primer of
DM> how to obtain useful debugging information to help
DM> developers out here.
DM>
DM> I think somehow that for as long as a debug build is the
DM> default build, these brief gdb instructions would be
DM> quite pertinent for users to read somewhere on the
DM> website.  Test builders are a valuable resource, and so
DM> I believe it might useful to include such documentation
DM> on the site so users who do build jazz++ can quickly and
DM> easily supply (and know how to create) useful information
DM> if they find their jazz++ build crashes. If nothing else,
DM> it'll save having to spell that backtrace passage out,
DM> over and over again. Remember, everyone building from svn
DM> is effectively a test builder at present.
DM>
DM> Let me know what we think here - if it sounds like a good
DM> idea and we agree, I'll edit the htdocs in some appropriate
DM> spot to include such instructions.

I added a web directory (bugreport) and put an index.php in it for you 
to edit.  Have at it.

Pete