Segmentation fault when delivering music

[LinusU]

[../deps/mpg123/src/output/coreaudio.c:81] warning: Didn't have any audio data in callback (buffer underflow)

Program received signal EXC_BAD_ACCESS, Could not access memory.
Reason: KERN_INVALID_ADDRESS at address: 0x0000000000000000
0x000008c4719245b9 in ?? ()
(gdb) bt
#0  0x000008c4719245b9 in ?? ()
#1  0x000008c4719118b7 in ?? ()
#2  0x000000010018a28c in v8::internal::Invoke ()
#3  0x0000000100146364 in v8::Function::Call ()
#4  0x00000001021055b4 in read_delivered_music (handle=<value temporarily unavailable, due to optimizations>, status=<value temporarily unavailable, due to optimizations>) at ../src/player.cc:132
#5  0x000000010012f60c in uv__run_timers ()
#6  0x0000000100125f7e in uv_run ()
#7  0x000000010000c6e3 in node::Start ()
#8  0x00000001000039f4 in start ()
(gdb) 

[LinusU]

Got one more, it seems that there is always at least one buffer underflow before this happens. Last time there was one, this time four.

[../deps/mpg123/src/output/coreaudio.c:81] warning: Didn't have any audio data in callback (buffer underflow)
[../deps/mpg123/src/output/coreaudio.c:81] warning: Didn't have any audio data in callback (buffer underflow)
[../deps/mpg123/src/output/coreaudio.c:81] warning: Didn't have any audio data in callback (buffer underflow)
[../deps/mpg123/src/output/coreaudio.c:81] warning: Didn't have any audio data in callback (buffer underflow)

Program received signal EXC_BAD_ACCESS, Could not access memory.
Reason: KERN_INVALID_ADDRESS at address: 0x0000000000000000
0x0000360b14f245b9 in ?? ()
(gdb) bt   
#0  0x0000360b14f245b9 in ?? ()
#1  0x0000360b14f118b7 in ?? ()
#2  0x000000010018a28c in v8::internal::Invoke ()
#3  0x0000000100146364 in v8::Function::Call ()
#4  0x00000001021055b4 in read_delivered_music (handle=<value temporarily unavailable, due to optimizations>, status=<value temporarily unavailable, due to optimizations>) at ../src/player.cc:132
#5  0x000000010012f60c in uv__run_timers ()
#6  0x0000000100125f7e in uv_run ()
#7  0x000000010000c6e3 in node::Start ()
#8  0x00000001000039f4 in start ()
(gdb) 

[Floby]

Are you sure this isn't a node core audio error?
Le 22 oct. 2013 02:11, "Linus Unnebäck" notifications@github.com a écrit :

Got one more, it seems that there is always at least one buffer underflow
before this happens. Last time there was one, this time four.

[../deps/mpg123/src/output/coreaudio.c:81] warning: Didn't have any audio data in callback (buffer underflow)
[../deps/mpg123/src/output/coreaudio.c:81] warning: Didn't have any audio data in callback (buffer underflow)
[../deps/mpg123/src/output/coreaudio.c:81] warning: Didn't have any audio data in callback (buffer underflow)
[../deps/mpg123/src/output/coreaudio.c:81] warning: Didn't have any audio data in callback (buffer underflow)

Program received signal EXC_BAD_ACCESS, Could not access memory.
Reason: KERN_INVALID_ADDRESS at address: 0x0000000000000000
0x0000360b14f245b9 in ?? ()
(gdb) bt
#0 0x0000360b14f245b9 in ?? ()
#1 0x0000360b14f118b7 in ?? ()
#2 0x000000010018a28c in v8::internal::Invoke ()
#3 0x0000000100146364 in v8::Function::Call ()
#4 0x00000001021055b4 in read_delivered_music (handle=<value temporarily unavailable, due to optimizations>, status=<value temporarily unavailable, due to optimizations>) at ../src/player.cc:132
#5 0x000000010012f60c in uv__run_timers ()
#6 0x0000000100125f7e in uv_run ()
#7 0x000000010000c6e3 in node::Start ()
#8 0x00000001000039f4 in start ()
(gdb)

—
Reply to this email directly or view it on GitHubhttps://github.com//issues/41#issuecomment-26767967
.

[LinusU]

coreaudio.c is from node-speaker, however, I don't think that the problem lies there. This is just what I think and I might be totally wrong:

#3 0x0000000100146364 in v8::Function::Call () Is player.cc line 132, where it calls into this._session._sp_session.music_delivery which I think is #1 which just forwards the call to Node.js Readable stream which actually uses the buffer (#0). When the buffer is read it tries to read it from address 0 (NULL) which of course results in EXC_BAD_ACCESS.

So I think that somehow we create the buffer with a NULL pointer...

[Floby]

That's unfortunate... ^^

On 22 October 2013 15:38, Linus Unnebäck notifications@github.com wrote:

coreaudio.c is from node-speaker, however, I don't think that the problem
lies there. This is just what I think and I might be totally wrong:

#3 0x0000000100146364 in v8::Function::Call () Is player.cc line 132,
where it calls into this._session._sp_session.music_delivery which I
think is #1 which just forwards the call to Node.js Readable stream which
actually uses the buffer (#0). When the buffer is read it tries to read
it from address 0 (NULL) which of course results in EXC_BAD_ACCESS.

So I think that somehow we create the buffer with a NULL pointer...

—
Reply to this email directly or view it on GitHubhttps://github.com//issues/41#issuecomment-26803169
.

[Floby]

As I've said, I should be able to get more time to work on this next week.

On 22 October 2013 17:41, Florent JABY florent.jaby@gmail.com wrote:

That's unfortunate... ^^

On 22 October 2013 15:38, Linus Unnebäck notifications@github.com wrote:

coreaudio.c is from node-speaker, however, I don't think that the
problem lies there. This is just what I think and I might be totally wrong:

#3 0x0000000100146364 in v8::Function::Call () Is player.cc line 132,
where it calls into this._session._sp_session.music_delivery which I
think is #1 which just forwards the call to Node.js Readable stream
which actually uses the buffer (#0). When the buffer is read it tries to
read it from address 0 (NULL) which of course results in EXC_BAD_ACCESS.

So I think that somehow we create the buffer with a NULL pointer...

—
Reply to this email directly or view it on GitHubhttps://github.com//issues/41#issuecomment-26803169
.

[LinusU]

Yeah, no worries. I'll look at aswell if I get the time.

[IainCole]

I have made a change in this commit that should hopefully address this issue, I'm still testing it for a while but it seems to do the trick so far. I'll create a PR once I'm happy that it has actually fixed it. IainCole@85a2cdd

[IainCole]

Looks like this didn't work :(

[IainCole]

This one IainCole@ec2fbc5 might work better though

[tim-gp]

Did it Iain?

[IainCole]

I still get a segfault in my particular app but I don't get this error any more and it's much more stable, so I think it worked for fixing this issue. I haven't been able to get to the root cause of my other issue though so there might still be something wrong.