[RFC] media: assertion to check that transport exists

[RFC] media: assertion to check that transport exists

[Mikel Astiz]

From: Mikel Astiz <mikel.astiz@bmw-carit.de>

>From my understanding, a transport should exist for any non-disconnected
gateway.

These assertions sometimes fail though. So I would like to clarify if
that's a consistent state in BlueZ or there is some bug.
---
 audio/media.c |    3 +++
 1 files changed, 3 insertions(+), 0 deletions(-)

diff --git a/audio/media.c b/audio/media.c
index a2ef437..c5fe3d9 100644
--- a/audio/media.c
+++ b/audio/media.c
@@ -28,6 +28,7 @@
 #endif
 
 #include <errno.h>
+#include <assert.h>
 
 #include <glib.h>
 #include <gdbus.h>
@@ -620,8 +621,10 @@ static void gateway_state_changed(struct audio_device *dev,
 					gateway_setconf_cb, dev, NULL);
 		break;
 	case GATEWAY_STATE_CONNECTED:
+		assert(endpoint->transport != NULL);
 		break;
 	case GATEWAY_STATE_PLAYING:
+		assert(endpoint->transport != NULL);
 		break;
 	}
 }
-- 
1.7.6.4

Re: [RFC] media: assertion to check that transport exists

[Luiz Augusto von Dentz]

Hi Mikel,

On Thu, Dec 15, 2011 at 2:33 PM, Mikel Astiz <mikel.astiz.oss@gmail.com> wrote:
> From: Mikel Astiz <mikel.astiz@bmw-carit.de>
>
> From my understanding, a transport should exist for any non-disconnected
> gateway.
>
> These assertions sometimes fail though. So I would like to clarify if
> that's a consistent state in BlueZ or there is some bug.
> ---
>  audio/media.c |    3 +++
>  1 files changed, 3 insertions(+), 0 deletions(-)
>
> diff --git a/audio/media.c b/audio/media.c
> index a2ef437..c5fe3d9 100644
> --- a/audio/media.c
> +++ b/audio/media.c
> @@ -28,6 +28,7 @@
>  #endif
>
>  #include <errno.h>
> +#include <assert.h>
>
>  #include <glib.h>
>  #include <gdbus.h>
> @@ -620,8 +621,10 @@ static void gateway_state_changed(struct audio_device *dev,
>                                        gateway_setconf_cb, dev, NULL);
>                break;
>        case GATEWAY_STATE_CONNECTED:
> +               assert(endpoint->transport != NULL);
>                break;
>        case GATEWAY_STATE_PLAYING:
> +               assert(endpoint->transport != NULL);
>                break;
>        }
>  }
> --
> 1.7.6.4

IMO assert on daemon are not that great, it may help while developing
but why not run with valgrind and let it crash?

-- 
Luiz Augusto von Dentz

Re: [RFC] media: assertion to check that transport exists

[Mikel Astiz]

Hi Luiz,

On 12/16/2011 09:43 AM, Luiz Augusto von Dentz wrote:
> IMO assert on daemon are not that great, it may help while developing 
> but why not run with valgrind and let it crash? 

I was not that much proposing to add the assertion, but to discuss if 
that assertion should hold or not.

Basically my question is: is it guaranteed that a transport will exist 
while a gateway is connected or playing?

If yes, I think there is a bug somewhere, because those assertions do 
fail sometimes.

I was specifically asking this regarding the internal state of the 
daemon, but the same question could be formulated for the state exposed 
in D-Bus.

Regards,
Mikel

Re: [RFC] media: assertion to check that transport exists

[Luiz Augusto von Dentz]

Hi Mikel,

On Fri, Dec 16, 2011 at 2:06 PM, Mikel Astiz <mikel.astiz@bmw-carit.de> wrote:
> Hi Luiz,
>
>
> On 12/16/2011 09:43 AM, Luiz Augusto von Dentz wrote:
>>
>> IMO assert on daemon are not that great, it may help while developing but
>> why not run with valgrind and let it crash?
>
>
> I was not that much proposing to add the assertion, but to discuss if that
> assertion should hold or not.
>
> Basically my question is: is it guaranteed that a transport will exist while
> a gateway is connected or playing?

Yes it should exist if media API is being used, while we are
connecting we attempt to create the transport.

> If yes, I think there is a bug somewhere, because those assertions do fail
> sometimes.

Probably, but it should be easier if you debug the state transitions.

> I was specifically asking this regarding the internal state of the daemon,
> but the same question could be formulated for the state exposed in D-Bus.

It should be consistent, I would start by analyzing the logs, there
seems to be at least one place where it call clear_configuration
without being disconnected see media.c:endpoint_reply, that is when
the endpoint stop responding.

-- 
Luiz Augusto von Dentz

Re: [RFC] media: assertion to check that transport exists

[Vinicius Costa Gomes]

Hi Luiz,

On 10:43 Fri 16 Dec, Luiz Augusto von Dentz wrote:
> Hi Mikel,
> 
> On Thu, Dec 15, 2011 at 2:33 PM, Mikel Astiz <mikel.astiz.oss@gmail.com> wrote:
> > From: Mikel Astiz <mikel.astiz@bmw-carit.de>
> >
> > From my understanding, a transport should exist for any non-disconnected
> > gateway.
> >
> > These assertions sometimes fail though. So I would like to clarify if
> > that's a consistent state in BlueZ or there is some bug.
> > ---
> >  audio/media.c |    3 +++
> >  1 files changed, 3 insertions(+), 0 deletions(-)
> >
> > diff --git a/audio/media.c b/audio/media.c
> > index a2ef437..c5fe3d9 100644
> > --- a/audio/media.c
> > +++ b/audio/media.c
> > @@ -28,6 +28,7 @@
> >  #endif
> >
> >  #include <errno.h>
> > +#include <assert.h>
> >
> >  #include <glib.h>
> >  #include <gdbus.h>
> > @@ -620,8 +621,10 @@ static void gateway_state_changed(struct audio_device *dev,
> >                                        gateway_setconf_cb, dev, NULL);
> >                break;
> >        case GATEWAY_STATE_CONNECTED:
> > +               assert(endpoint->transport != NULL);
> >                break;
> >        case GATEWAY_STATE_PLAYING:
> > +               assert(endpoint->transport != NULL);
> >                break;
> >        }
> >  }
> > --
> > 1.7.6.4
> 
> IMO assert on daemon are not that great, it may help while developing
> but why not run with valgrind and let it crash?

This information may help with your worries, from the assert(3) man page:
"If the macro NDEBUG was defined at the moment <assert.h> was last included,
the macro assert() generates no code, and hence does nothing at all."

So we can have assert() do something only when it is compiled in developer mode.

> 
> -- 
> Luiz Augusto von Dentz
> --
> To unsubscribe from this list: send the line "unsubscribe linux-bluetooth" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html


Cheers,
-- 
Vinicius