Hi, do you have any experience with the plugdet interrupt?
We have set up the Pyra hardware that the plugdet pin is pulled low if no headset is connected and released to "high" if it is.
This seems to make the kernel hang immediately, sometimes with strange interrupt or nmi warnings.
We probably could live without this feature since we have a ts3a227e in parallel which can also do plug/unplug detection and issue an interrupt (but that one is also not working).
Is this a known bug or untested area of the twl6040 (driver)?
BR and thanks, Nikolaus
Hi,
On 10/19/2016 11:00 PM, H. Nikolaus Schaller wrote:
Hi, do you have any experience with the plugdet interrupt?
We have set up the Pyra hardware that the plugdet pin is pulled low if no headset is connected and released to "high" if it is.
This seems to make the kernel hang immediately, sometimes with strange interrupt or nmi warnings.
We probably could live without this feature since we have a ts3a227e in parallel which can also do plug/unplug detection and issue an interrupt (but that one is also not working).
Is this a known bug or untested area of the twl6040 (driver)?
As I recall it was working on omap5-uevm and on omap4-SDP/Blaze.
I will retest it again on both.
BR and thanks, Nikolaus
Hi Peter,
Am 24.10.2016 um 16:02 schrieb Peter Ujfalusi peter.ujfalusi@ti.com:
Hi,
On 10/19/2016 11:00 PM, H. Nikolaus Schaller wrote:
Hi, do you have any experience with the plugdet interrupt?
We have set up the Pyra hardware that the plugdet pin is pulled low if no headset is connected and released to "high" if it is.
This seems to make the kernel hang immediately, sometimes with strange interrupt or nmi warnings.
We probably could live without this feature since we have a ts3a227e in parallel which can also do plug/unplug detection and issue an interrupt (but that one is also not working).
Is this a known bug or untested area of the twl6040 (driver)?
As I recall it was working on omap5-uevm and on omap4-SDP/Blaze.
I will retest it again on both.
it looks as if ti,jack_detection is missing for the omap5-uevm.
I have a tiny patch/fix in our production kernel now:
http://git.goldelico.com/?p=gta04-kernel.git;a=patch;h=7771a9fb3b9133fc361c8...
but I haven't tested deeply. Especially not on the EVM.
Could it be, that an activated plugdet pin without "ti,jack_detection" being defined, makes problems when switched?
BR and thanks, Nikolaus
On 10/24/16 18:30, H. Nikolaus Schaller wrote:
Hi Peter,
Am 24.10.2016 um 16:02 schrieb Peter Ujfalusi peter.ujfalusi@ti.com:
Hi,
On 10/19/2016 11:00 PM, H. Nikolaus Schaller wrote:
Hi, do you have any experience with the plugdet interrupt?
We have set up the Pyra hardware that the plugdet pin is pulled low if no headset is connected and released to "high" if it is.
This seems to make the kernel hang immediately, sometimes with strange interrupt or nmi warnings.
We probably could live without this feature since we have a ts3a227e in parallel which can also do plug/unplug detection and issue an interrupt (but that one is also not working).
Is this a known bug or untested area of the twl6040 (driver)?
As I recall it was working on omap5-uevm and on omap4-SDP/Blaze.
I will retest it again on both.
it looks as if ti,jack_detection is missing for the omap5-uevm.
I have a tiny patch/fix in our production kernel now:
http://git.goldelico.com/?p=gta04-kernel.git;a=patch;h=7771a9fb3b9133fc361c8...
but I haven't tested deeply. Especially not on the EVM.
Could it be, that an activated plugdet pin without "ti,jack_detection" being defined, makes problems when switched?
on omap5-uevm I have tested first w/o the 'ti,jack-detection' in DTS. Plugging/un-plugging the headset does not cause any issue and the twl6040_irq_plug is incrementing.
After adding the 'ti,jack-detection' Plugging/un-plugging works also and I see that the widget states are also changing.
But I see no kernel hang on the board in either case.
Hi Peter,
Am 25.10.2016 um 15:50 schrieb Peter Ujfalusi peter.ujfalusi@ti.com:
On 10/24/16 18:30, H. Nikolaus Schaller wrote:
Hi Peter,
Am 24.10.2016 um 16:02 schrieb Peter Ujfalusi peter.ujfalusi@ti.com:
Hi,
On 10/19/2016 11:00 PM, H. Nikolaus Schaller wrote:
Hi, do you have any experience with the plugdet interrupt?
We have set up the Pyra hardware that the plugdet pin is pulled low if no headset is connected and released to "high" if it is.
This seems to make the kernel hang immediately, sometimes with strange interrupt or nmi warnings.
We probably could live without this feature since we have a ts3a227e in parallel which can also do plug/unplug detection and issue an interrupt (but that one is also not working).
Is this a known bug or untested area of the twl6040 (driver)?
As I recall it was working on omap5-uevm and on omap4-SDP/Blaze.
I will retest it again on both.
it looks as if ti,jack_detection is missing for the omap5-uevm.
I have a tiny patch/fix in our production kernel now:
http://git.goldelico.com/?p=gta04-kernel.git;a=patch;h=7771a9fb3b9133fc361c8...
but I haven't tested deeply. Especially not on the EVM.
Could it be, that an activated plugdet pin without "ti,jack_detection" being defined, makes problems when switched?
on omap5-uevm I have tested first w/o the 'ti,jack-detection' in DTS. Plugging/un-plugging the headset does not cause any issue and the twl6040_irq_plug is incrementing.
After adding the 'ti,jack-detection' Plugging/un-plugging works also and I see that the widget states are also changing.
But I see no kernel hang on the board in either case.
Ok, then it could be some coincidence with some other problem. I will keep an eye on it.
So is there something speaking against that I officially submit the patch?
Many thanks for testing, Nikolaus
On 10/25/2016 06:52 PM, H. Nikolaus Schaller wrote:
Ok, then it could be some coincidence with some other problem. I will keep an eye on it.
So is there something speaking against that I officially submit the patch?
I don't see any reason why you should not ;)
Many thanks for testing, Nikolaus
-
H. Nikolaus Schaller -
Peter Ujfalusi