Fwd: Kernel tickets verify

classic Classic list List threaded Threaded
14 messages Options
Reply | Threaded
Open this post in threaded view
|

Fwd: Kernel tickets verify

wendy-2-3


Hi Andy,

the testing team plan to test the 2.6.29 kernel on Om2009, but would like to
ask your help to verify some of them.
List the bugs below and might have more coming! ;)

#1319 sysfs RTC class device missing wakealarm capability
[https://docs.openmoko.org/trac/ticket/1319]
Do we have this issue? how we can test and verify this one?

#1493 when press power button makes some noise during suspend time.
This one has kernel patch in trac, does this fix? Last week we use Om2009in
our device but it seems still have this problem. We'll verify it more, just
want to make sure this could be in_testing or not.

#2211 Windows instant reboot while connecting Freerunner on USB
.....we want to close this one as community, is that okay?

Thanks for your help!

Regards,
Wendy

_______________________________________________
testing mailing list
[hidden email]
https://lists.openmoko.org/mailman/listinfo/testing
Reply | Threaded
Open this post in threaded view
|

Re: Fwd: Kernel tickets verify

andy green-3
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Somebody in the thread at some point said:

| the testing team plan to test the 2.6.29 kernel on Om2009, but would
like to
| ask your help to verify some of them.

Well that will be very good to hear about the functional testing
criteria again.

| List the bugs below and might have more coming! ;)
|
| #1319 sysfs RTC class device missing wakealarm capability
| [https://docs.openmoko.org/trac/ticket/1319]
| Do we have this issue? how we can test and verify this one?

Mickey do you still need this?  We seem to be able to wake off RTC alarm?

| #1493 when press power button makes some noise during suspend time.
| This one has kernel patch in trac, does this fix? Last week we use
Om2009in
| our device but it seems still have this problem. We'll verify it more,
just
| want to make sure this could be in_testing or not.

No there's no patch aimed at this AFAIK.  I guess it's to do with AMP
enable timing or chargeup of Vref / Vmid.  There is actually a delayed
work associated with this as part of 2.6.28 changes though, you should
definitely reconfirm with recent kernel.  If it's still there it's
possible we can do something.

| #2211 Windows instant reboot while connecting Freerunner on USB
| .....we want to close this one as community, is that okay?

Well I think we can wait and see what happens now somebody kindly
supplied a log.  Actually it does need community to help solve it but we
probably need to track it.

- -Andy
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.9 (GNU/Linux)
Comment: Using GnuPG with Fedora - http://enigmail.mozdev.org

iEYEARECAAYFAkmlGXEACgkQOjLpvpq7dMrsjwCbBxwZpNDjGBL0XDE53jNWkGMq
M7sAnitF9JWCOQxhQMzutDxygSj7TvhO
=9dCW
-----END PGP SIGNATURE-----

_______________________________________________
testing mailing list
[hidden email]
https://lists.openmoko.org/mailman/listinfo/testing
Reply | Threaded
Open this post in threaded view
|

Re: Fwd: Kernel tickets verify

Timo Juhani Lindfors
Andy Green <[hidden email]> writes:
> | #1493 when press power button makes some noise during suspend time.

> No there's no patch aimed at this AFAIK.  I guess it's to do with AMP
> enable timing or chargeup of Vref / Vmid.  There is actually a delayed

amixer -q -d sset "Amp Spk" mute
sleep 4
apm -s
sleep 1
amixer -q -d sset "Amp Spk" unmute

works for me. I don't hear a thing on suspend/resume even if I hold
the phone near my ear.

(Also added as http://docs.openmoko.org/trac/ticket/1493#comment:14)

_______________________________________________
testing mailing list
[hidden email]
https://lists.openmoko.org/mailman/listinfo/testing
Reply | Threaded
Open this post in threaded view
|

Re: Fwd: Kernel tickets verify

andy green-3
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Somebody in the thread at some point said:
| Andy Green <[hidden email]> writes:
|> | #1493 when press power button makes some noise during suspend time.
|
|> No there's no patch aimed at this AFAIK.  I guess it's to do with AMP
|> enable timing or chargeup of Vref / Vmid.  There is actually a delayed
|
| amixer -q -d sset "Amp Spk" mute
| sleep 4
| apm -s
| sleep 1
| amixer -q -d sset "Amp Spk" unmute
|
| works for me. I don't hear a thing on suspend/resume even if I hold
| the phone near my ear.
|
| (Also added as http://docs.openmoko.org/trac/ticket/1493#comment:14)

Well it's quite convincing... is the sleep 4 the business end of it, or
the sleep 1?

If it's coming only on resume then the deferred register dump that
happens to get that out of the synchronous resume path would be a good
place to re-enable the amp state, it's 750ms or so away from the start
of resume IIRC already.

- -Andy

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.9 (GNU/Linux)
Comment: Using GnuPG with Fedora - http://enigmail.mozdev.org

iEYEARECAAYFAkmlIGUACgkQOjLpvpq7dMqVrgCfUpnHtyaxX93/HplFnlIUB6Wo
TAsAn0KnitnQISmfwOCAVK1GyyhtUPOP
=AD8r
-----END PGP SIGNATURE-----

_______________________________________________
testing mailing list
[hidden email]
https://lists.openmoko.org/mailman/listinfo/testing
Reply | Threaded
Open this post in threaded view
|

Re: Fwd: Kernel tickets verify

Timo Juhani Lindfors
Andy Green <[hidden email]> writes:
> Well it's quite convincing... is the sleep 4 the business end of it, or
> the sleep 1?

sleep 1 is to avoid Xglamo crash on resume. sleep 4 is to make sure
ogsmd has finished talking to GSM. :-)

With Xglamo stopped I tested this without sleeps and could not hear
anything either.

_______________________________________________
testing mailing list
[hidden email]
https://lists.openmoko.org/mailman/listinfo/testing
Reply | Threaded
Open this post in threaded view
|

Re: Fwd: Kernel tickets verify

andy green-3
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Somebody in the thread at some point said:
| Andy Green <[hidden email]> writes:
|> Well it's quite convincing... is the sleep 4 the business end of it, or
|> the sleep 1?
|
| sleep 1 is to avoid Xglamo crash on resume. sleep 4 is to make sure
| ogsmd has finished talking to GSM. :-)
|
| With Xglamo stopped I tested this without sleeps and could not hear
| anything either.

OK so it's the timing of amp mute and unmute vs everything else going in
and out of suspend then.  Thanks for the hints.

- -Andy
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.9 (GNU/Linux)
Comment: Using GnuPG with Fedora - http://enigmail.mozdev.org

iEYEARECAAYFAkmlJQYACgkQOjLpvpq7dMqIIgCginJhnaFyKWnN3Hykty1oNqTp
gKIAn0/YrTUYr/9DYigleZXTY12q59+D
=FO/v
-----END PGP SIGNATURE-----

_______________________________________________
testing mailing list
[hidden email]
https://lists.openmoko.org/mailman/listinfo/testing
Reply | Threaded
Open this post in threaded view
|

Re: Fwd: Kernel tickets verify

andy green-3
In reply to this post by Timo Juhani Lindfors
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Somebody in the thread at some point said:
| Andy Green <[hidden email]> writes:
|> | #1493 when press power button makes some noise during suspend time.
|
|> No there's no patch aimed at this AFAIK.  I guess it's to do with AMP
|> enable timing or chargeup of Vref / Vmid.  There is actually a delayed
|
| amixer -q -d sset "Amp Spk" mute

Oh I see it diverts the output to headphones there?  That is why it
impacts the click?

Does it still actually click if you have headphones in?  That could be
objectionable if it is happening right at your ear.

- -Andy
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.9 (GNU/Linux)
Comment: Using GnuPG with Fedora - http://enigmail.mozdev.org

iEYEARECAAYFAkmlJ6EACgkQOjLpvpq7dMq7eACaA/FrNAt0/CpuffAq0ceMTGYZ
VhUAn2I2Ph/05P2b7UwYrZOodOY078WA
=ob5v
-----END PGP SIGNATURE-----

_______________________________________________
testing mailing list
[hidden email]
https://lists.openmoko.org/mailman/listinfo/testing
Reply | Threaded
Open this post in threaded view
|

Re: Fwd: Kernel tickets verify

Paul Fertser


Andy Green <andy-4Bgg8jF3iZdWk0Htik3J/[hidden email]> writes:

> Somebody in the thread at some point said:
> | Andy Green <andy-4Bgg8jF3iZdWk0Htik3J/[hidden email]> writes:
> |> | #1493 when press power button makes some noise during suspend time.
> |
> |> No there's no patch aimed at this AFAIK.  I guess it's to do with AMP
> |> enable timing or chargeup of Vref / Vmid.  There is actually a delayed
> |
> | amixer -q -d sset "Amp Spk" mute
>
> Oh I see it diverts the output to headphones there?  That is why it
> impacts the click?

Sure.

> Does it still actually click if you have headphones in?  That could be
> objectionable if it is happening right at your ear.

BTW, if they read WM8753 datasheet instead of proposing to introduce
some ugly workarounds in kernel, they'd see that the codec itself
supports silent method of turning on/off. Needs further investigation
though. It can be even theoretically possible that the sound comes
from some glitch on stupid DL_GSM line. This matter needs further
investigation imho, but i can't see why it is considered to be
important and especially why it needs kernel work right now; sorry for
expressing my stupid opinion.

--
Be free, use free (http://www.gnu.org/philosophy/free-sw.html) software!
mailto:[hidden email]


_______________________________________________
testing mailing list
[hidden email]
https://lists.openmoko.org/mailman/listinfo/testing
Reply | Threaded
Open this post in threaded view
|

Re: Fwd: Kernel tickets verify

Timo Juhani Lindfors
In reply to this post by andy green-3
Andy Green <[hidden email]> writes:
> Does it still actually click if you have headphones in?  That could be
> objectionable if it is happening right at your ear.

I don't have headphones from openmoko.

_______________________________________________
testing mailing list
[hidden email]
https://lists.openmoko.org/mailman/listinfo/testing
Reply | Threaded
Open this post in threaded view
|

Re: Fwd: Kernel tickets verify

andy green-3
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Somebody in the thread at some point said:
| Andy Green <[hidden email]> writes:
|> Does it still actually click if you have headphones in?  That could be
|> objectionable if it is happening right at your ear.
|
| I don't have headphones from openmoko.

Understood, thanks again for telling about your script.

- -Andy

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.9 (GNU/Linux)
Comment: Using GnuPG with Fedora - http://enigmail.mozdev.org

iEYEARECAAYFAkmlSmEACgkQOjLpvpq7dMpr+gCfabNENcbFb0h4iy/lBcRpthnk
RZEAmwYb7o1M9GQ9iXLSTSXDMBocok01
=N5wy
-----END PGP SIGNATURE-----

_______________________________________________
testing mailing list
[hidden email]
https://lists.openmoko.org/mailman/listinfo/testing
Reply | Threaded
Open this post in threaded view
|

Re: Fwd: Kernel tickets verify

Mark Brown-3
In reply to this post by andy green-3
On Wed, Feb 25, 2009 at 06:12:02PM +0800, Andy Green wrote:

> No there's no patch aimed at this AFAIK.  I guess it's to do with AMP
> enable timing or chargeup of Vref / Vmid.  There is actually a delayed
> work associated with this as part of 2.6.28 changes though, you should
> definitely reconfirm with recent kernel.  If it's still there it's
> possible we can do something.

With current kernels the speaker amp pops *very* noticably every time it
is enabled or disabled.

_______________________________________________
testing mailing list
[hidden email]
https://lists.openmoko.org/mailman/listinfo/testing
Reply | Threaded
Open this post in threaded view
|

Re: Fwd: Kernel tickets verify

andy green-3
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Somebody in the thread at some point said:
| On Wed, Feb 25, 2009 at 06:12:02PM +0800, Andy Green wrote:
|
|> No there's no patch aimed at this AFAIK.  I guess it's to do with AMP
|> enable timing or chargeup of Vref / Vmid.  There is actually a delayed
|> work associated with this as part of 2.6.28 changes though, you should
|> definitely reconfirm with recent kernel.  If it's still there it's
|> possible we can do something.
|
| With current kernels the speaker amp pops *very* noticably every time it
| is enabled or disabled.

You mentioned you're planning to take care of this, it's much appreciated.

- -Andy
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.9 (GNU/Linux)
Comment: Using GnuPG with Fedora - http://enigmail.mozdev.org

iEYEARECAAYFAkmlcaAACgkQOjLpvpq7dMrhagCfbLF+l1pej199HEMUmsC+loDf
TPgAmwSrPBtRvTTcZDemCdfMOrF56/To
=Fmsu
-----END PGP SIGNATURE-----

_______________________________________________
testing mailing list
[hidden email]
https://lists.openmoko.org/mailman/listinfo/testing
Reply | Threaded
Open this post in threaded view
|

Re: Fwd: Kernel tickets verify

Carl Lobo
With the latest andy-tracking & qi from people.openmoko.org I get
plenty of errors from gsm0710muxd. With all services stopped when I
cat /dev/ttySAC0 I get "AT Command Interpreter ready". As soon as I
start frameworkd (and hence gsm-muxd) it starts complaining about the
modem not responding to AT commands. Sending reset commands, powering
down & up etc over ssh doesn't help.
Exactly the same image and kernel don't give me this problem if I boot
through the NOR boot.

_______________________________________________
testing mailing list
[hidden email]
https://lists.openmoko.org/mailman/listinfo/testing
Reply | Threaded
Open this post in threaded view
|

Re: Fwd: Kernel tickets verify

andy green-3
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Somebody in the thread at some point said:
| With the latest andy-tracking & qi from people.openmoko.org I get
| plenty of errors from gsm0710muxd. With all services stopped when I
| cat /dev/ttySAC0 I get "AT Command Interpreter ready". As soon as I
| start frameworkd (and hence gsm-muxd) it starts complaining about the
| modem not responding to AT commands. Sending reset commands, powering
| down & up etc over ssh doesn't help.
| Exactly the same image and kernel don't give me this problem if I boot
| through the NOR boot.

What's changed recently here is the UART to GSM defaults to GPIO mode in
Qi now, to stop us driving the powered-down GSM unit

http://git.openmoko.org/?p=qi.git;a=commitdiff;h=4aeebf9ecad292bbff8c0bce8fad64bdd7c7b269

However when your rootfs asks to power up the GSM unit, the mode on
those pins should be changed to UART mode

http://git.openmoko.org/?p=kernel.git;a=commitdiff;h=8df75a337131d835d4dd8509696d684e4fa8970e

how does this fit with what frameworkd is doing?

- -Andy
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.9 (GNU/Linux)
Comment: Using GnuPG with Fedora - http://enigmail.mozdev.org

iEYEARECAAYFAkml/jwACgkQOjLpvpq7dMqf4ACfWqm303M069haCJhqAvafZGVK
yPoAnArjJhCra+qQtFxahaKFIFdQC9qH
=lNpY
-----END PGP SIGNATURE-----

_______________________________________________
testing mailing list
[hidden email]
https://lists.openmoko.org/mailman/listinfo/testing