[alsa-devel] Hang up at .prepare( )
Hi
I've started debugging our ASoC drivers. I've faced issue when installing these drivers, system becomes hang up at async_synchronize_full_domain() of dapm_power_widgets() in .prepare().
Do you have any information about this hang up ?
I show details below.
======Details============
* Procedure install codec-driver install platform-driver install machine driver
echo ml26124 0x1A > /sys/bus/i2c/devices/i2c-1/new_device After this command, occur system hang up within a few seconds
* Call Trees
platform-driver:.open() machine-driver:.hw_params() codec-driver:.hw_params() platform-driver:.dai_hw_params() platform-driver:.hw_params() platform-driver:.prepare() platform-driver:.dai_digital_mute() platform-driver:.prepare() platform-driver:.dai_digital_mute()
platform-driver:.open() machine-driver:.hw_params() codec-driver:.hw_params() platform-driver:.dai_hw_params() platform-driver:.hw_params() platform-driver:.prepare() platform-driver:.dai_digital_mute() platform-driver:.prepare() platform-driver:.dai_digital_mute() platform-driver:.hw_free platform-driver:.dai_hw_free platform-driver:.hw_free platform-driver:.dai_hw_free platform-driver:.close()
platform-driver:.open() machine-driver:.hw_params() codec-driver:.hw_params() platform-driver:.dai_hw_params() platform-driver:.hw_params() platform-driver:.prepare() platform-driver:.dai_digital_mute() platform-driver:.prepare() platform-driver:.dai_digital_mute() platform-driver:.hw_free platform-driver:.dai_hw_free platform-driver:.hw_free platform-driver:.dai_hw_free platform-driver:.close() platform-driver:.hw_free platform-driver:.dai_hw_free platform-driver:.hw_free platform-driver:.dai_hw_free platform-driver:.close() platform-driver:.dai_digital_mute()
platform-driver:.open() machine-driver:.hw_params() codec-driver:.hw_params() platform-driver:.dai_hw_params() platform-driver:.hw_params() platform-driver:.prepare() platform-driver:.dai_digital_mute() platform-driver:.prepare() platform-driver:.dai_digital_mute()
platform-driver:.open() machine-driver:.hw_params() codec-driver:.hw_params() platform-driver:.dai_hw_params() platform-driver:.hw_params() platform-driver:.prepare() platform-driver:.dai_digital_mute() platform-driver:.prepare() platform-driver:.dai_digital_mute() platform-driver:.hw_free platform-driver:.dai_hw_free platform-driver:.hw_free platform-driver:.dai_hw_free platform-driver:.close()
platform-driver:.open() machine-driver:.hw_params() codec-driver:.hw_params() platform-driver:.dai_hw_params() platform-driver:.hw_params() platform-driver:.prepare()
==> Occur system hang up.
According to system log, the hang up seems occuring at async_synchronize_full_domain(), which is at soc_pcm_prepare -> soc_dapm_stream_event -> dapm_power_widgets -> async_synchronize_full_domain [2nd call]
* Test condition CPU=Intel Atom E6xx series IOH=LAPIS Semiconductor ML7213 Fedora15 Kernel=3.2
thanks in advance. tomoya
On Tue, Jan 24, 2012 at 06:38:45PM +0900, Tomoya MORINAGA wrote:
I've faced issue when installing these drivers, system becomes hang up at async_synchronize_full_domain() of dapm_power_widgets() in .prepare().
Do you have any information about this hang up ?
I show details below.
What efforts have you made to debug this yourself? Clearly this is not happening for other people, perhaps it is related to something you have in your drivers - have you addressed the issues raised with them?
2012/1/24 Mark Brown broonie@opensource.wolfsonmicro.com:
What efforts have you made to debug this yourself? Clearly this is not happening for other people, perhaps it is related to something you have in your drivers - have you addressed the issues raised with them?
I said previously, "our ASoC drivers" not others. Of course, I understand current upstreamed drivers don't have this issue. I just asked "If someone faced the same issue when debugging, let me know about it."
thanks,
On Tue, Jan 24, 2012 at 08:30:17PM +0900, Tomoya MORINAGA wrote:
2012/1/24 Mark Brown broonie@opensource.wolfsonmicro.com:
What efforts have you made to debug this yourself? Clearly this is not happening for other people, perhaps it is related to something you have in your drivers - have you addressed the issues raised with them?
I said previously, "our ASoC drivers" not others. Of course, I understand current upstreamed drivers don't have this issue. I just asked "If someone faced the same issue when debugging, let me know about it."
That's not answering my question - I'm asking you what steps you've taken to progress, to explain the analysis you've done so far. For example, what is the blocked thread blocked on - what is it waiting for and why has that thing happened?
If you have not addressed the review issues with your driver then take a look at those, perhaps the problem has already been identified through code review.
participants (2)
-
Mark Brown -
Tomoya MORINAGA