02:07
jakogut_ has quit [Ping timeout: 256 seconds]
02:41
sputnik_ has quit [*.net *.split]
02:42
mayorga has quit [Ping timeout: 277 seconds]
02:42
sputnik_ has joined #linux-amlogic
02:44
libv has quit [*.net *.split]
02:44
dlan has quit [*.net *.split]
02:44
wens has quit [*.net *.split]
02:44
nashpa has quit [*.net *.split]
02:44
douardda has quit [*.net *.split]
02:45
dlan has joined #linux-amlogic
02:45
libv has joined #linux-amlogic
02:45
wens has joined #linux-amlogic
02:45
nashpa has joined #linux-amlogic
02:45
douardda has joined #linux-amlogic
02:46
cyrozap has quit [*.net *.split]
02:46
gnufan has quit [*.net *.split]
02:46
phh has quit [*.net *.split]
02:46
jakogut has quit [*.net *.split]
02:46
naobsd has quit [*.net *.split]
02:47
cyrozap has joined #linux-amlogic
02:47
naobsd has joined #linux-amlogic
02:47
jakogut has joined #linux-amlogic
02:47
phh has joined #linux-amlogic
02:47
gnufan has joined #linux-amlogic
02:50
ChanServ has quit [*.net *.split]
02:50
mag has quit [*.net *.split]
02:50
xdarklight has quit [*.net *.split]
02:50
shadeslayer has quit [*.net *.split]
02:50
Ivanovic has quit [*.net *.split]
02:50
brujah has quit [*.net *.split]
02:50
edcragg has quit [*.net *.split]
02:50
jelly has quit [*.net *.split]
02:50
Simonious_ has quit [*.net *.split]
02:50
mastertheknife has quit [*.net *.split]
02:50
gab has quit [*.net *.split]
02:50
gnufan has quit [*.net *.split]
02:50
phh has quit [*.net *.split]
02:50
jakogut has quit [*.net *.split]
02:50
naobsd has quit [*.net *.split]
02:50
cyrozap has quit [*.net *.split]
02:50
libv has quit [*.net *.split]
02:50
dlan has quit [*.net *.split]
02:50
wens has quit [*.net *.split]
02:50
nashpa has quit [*.net *.split]
02:50
douardda has quit [*.net *.split]
02:50
sputnik_ has quit [*.net *.split]
02:50
nemunaire has quit [*.net *.split]
02:50
lvrp16 has quit [*.net *.split]
02:50
saintdev has quit [*.net *.split]
02:50
Guest15584 has quit [*.net *.split]
02:50
LoysO has quit [*.net *.split]
02:50
cosm has quit [*.net *.split]
02:50
vagrantc has quit [*.net *.split]
02:50
[TheBug] has quit [*.net *.split]
02:50
aballier has quit [*.net *.split]
02:50
droman0 has quit [*.net *.split]
02:50
scc23456 has quit [*.net *.split]
02:50
miceopede has quit [*.net *.split]
02:50
Pix has quit [*.net *.split]
02:50
Elpaulo has quit [*.net *.split]
02:50
webczat has quit [*.net *.split]
02:50
return0e_ has quit [*.net *.split]
02:51
naobsd has joined #linux-amlogic
02:51
dlan has joined #linux-amlogic
02:51
[TheBug] has joined #linux-amlogic
02:51
commavir has joined #linux-amlogic
02:51
Pix has joined #linux-amlogic
02:59
_whitelogger has joined #linux-amlogic
03:01
vagrantc has quit [Quit: leaving]
03:13
mayorga has joined #linux-amlogic
04:32
sputnik_ has quit [Read error: Connection reset by peer]
04:32
sputnik_ has joined #linux-amlogic
04:56
lissyx has joined #linux-amlogic
05:12
Barada has joined #linux-amlogic
05:47
gnufan has quit [Ping timeout: 264 seconds]
05:52
jakogut has quit [Remote host closed the connection]
05:59
lissyx has quit [Ping timeout: 248 seconds]
06:14
fedux has joined #linux-amlogic
06:22
gnufan has joined #linux-amlogic
06:40
return0e_ has quit []
07:05
lissyx has joined #linux-amlogic
07:32
a5m has joined #linux-amlogic
07:38
cthugha has joined #linux-amlogic
07:42
cthugha is now known as ldevulder
10:51
afaerber has joined #linux-amlogic
11:01
Guest97734 has joined #linux-amlogic
11:29
Guest97734 has quit [Quit: Page closed]
11:46
chewitt has joined #linux-amlogic
12:20
<
chewitt >
jbrunet: ping!
12:22
<
jbrunet >
chewitt: pong!
12:23
<
chewitt >
since I bumped from 4.16.anything to 4.17-rc's I am seeing a ton of this in dmesg
12:23
<
chewitt >
[ 187.783493] meson-i2s-dai meson-i2s-dai.0.auto: Unable set to the i2s clock rates
12:23
<
chewitt >
[ 187.783501] meson-i2s-dai meson-i2s-dai.0.auto: ASoC: can't set meson-i2s-dai.0.auto hw params: -22
12:23
<
chewitt >
narmstrong thinks it may be clock changes
12:24
<
chewitt >
this is the kernel patch set i'm using
12:25
<
chewitt >
b-* are current "waiting to go upstream" patches
12:26
<
chewitt >
d-* are the initial HDMI audio patches that one of the LE team created; adding multi-channel output
12:26
<
chewitt >
it was working okay (some bugs to hunt, but working) on 4.16.x
12:27
<
chewitt >
any ideas?
12:28
<
jbrunet >
Unlikely to be clock related
12:29
<
jbrunet >
at least not directly
12:30
<
jbrunet >
look at the code : "Unable set to the i2s clock rates " is reported when __bclks_set_rate() fails
12:31
<
jbrunet >
IIRC -22 is -EINVAL which means fs is not a multiple of 64
12:32
<
jbrunet >
It worth pointing out that this code (and this part in particular) is far from finalized
12:33
<
jbrunet >
Did you specifiy an mclk factor in your sound card DT ?
12:34
<
chewitt >
mclk-fs = <0x00000080>;
12:35
<
chewitt >
^ from the dtb
12:36
<
chewitt >
mclk-fs = <128>;
12:36
<
chewitt >
^ from dts
12:39
<
jbrunet >
when the error happen, you could dump the clock, see want you got
12:41
<
jbrunet >
We indeed changed a few things in the clock tree in last cycle. xdarklight already found a few mistakes
12:41
<
chewitt >
can you explain how I dump the clock?
12:41
<
jbrunet >
Maybe there are other things we missed, or I could be this audio driver, which is still fragile as it is
12:42
<
jbrunet >
cat /sys/kernel/debug/clk/clk_summary (which debugfs mounted there , of course)
12:43
<
jbrunet >
s/which/with
12:44
<
jbrunet >
what sample rate are you trying to play exactly ?
12:44
<
chewitt >
this is on cold boot
12:45
<
jbrunet >
This trace happen asoc is preparing a playback, and setting the clock for a particular ... so yes you are
12:46
<
jbrunet >
This trace happens when asoc is preparing a playback, and is setting the clock for a particular sample rate ... so yes you are
12:46
<
chewitt >
Kodi starts and enumerates alsa for devices
12:46
<
jbrunet >
Does more than t
12:50
<
jbrunet >
I can't look at the driver right now, I'll try later on. In the meantime you try to check meson_i2s_dai_set_sysclk() and see what clock rate is requested and what you actually get (add a few printk and clk_get_rate())
12:52
<
jbrunet >
it should be the the sample rate * mclk-fs ... which you reported being 128 but I mainly see 256 in your patch list
12:52
<
jbrunet >
It is expected that the frequency actually set is a bit off what is requested ... but it should be fairly close
12:52
nashpa has quit [Quit: Going away]
12:53
<
jbrunet >
In your trace "cts_amclk 0 0 0 5709664"
12:53
<
jbrunet >
5709664 / 128 = 44600 ... odd
12:54
<
jbrunet >
I would not expect 44100 to be that off
12:54
<
chewitt >
^ that's the dts for the box
12:56
<
chewitt >
I see the same on wetek hub/play2 which have dts in the mainline tree
12:56
<
chewitt >
current box is gxl/s905d .. hub/play2 are gxbb based
13:01
<
chewitt >
I forget exactly what the code is doing, but opening the device and reading properties
13:02
<
chewitt >
it's very well proven code, so I'm confident there's no issue there
13:06
<
jbrunet >
I'm telling you there is an issue, I'm telling that ,as a far as the audio driver is concerned, you are preparing a playback
13:06
<
jbrunet >
I'm
*NOT* ;)
13:07
<
jbrunet >
knowing the sample rate used would help
13:26
Elpaulo has quit [Read error: Connection reset by peer]
13:27
Elpaulo has joined #linux-amlogic
13:57
<
chewitt >
an educated hunch would be 44100
14:03
Barada has quit [Quit: Barada]
14:16
<
chewitt >
but I'm asking Kodi devs for input
14:30
<
chewitt >
jbrunet: comment from a colleague who saw something similar with rockchip hardware; this was needed
14:34
<
jbrunet >
I probably end up removing this driver completely
14:35
<
jbrunet >
Its purpose was to force the PLL to select higher rate (with significant post divider)
14:36
<
jbrunet >
but it is actually better to select some "magic rate" and stop the rate propagation
14:40
<
chewitt >
am rebuilding the kernel
14:52
<
chewitt >
that seems to work
14:58
<
chewitt >
the remaining audio problem I had when testing on 4.16 was that audio just drops out after a random time period
14:59
<
chewitt >
forcing an audio hardware change in Kodi, e.g. switching from HDMI (alsa) to pulse output and back to HDMI again, reawakens audio
14:59
<
chewitt >
nothing logged in dmesg
14:59
<
chewitt >
nothing logged on the Kodi side
15:06
<
chewitt >
completely independent of the audio format being played.. can be 2-channel mp3 or mult-channel dts
15:21
a5m has quit [Remote host closed the connection]
15:55
lissyx has quit [Ping timeout: 260 seconds]
16:33
vagrantc has joined #linux-amlogic
16:36
lissyx has joined #linux-amlogic
17:33
lissyx has quit [Ping timeout: 252 seconds]
18:02
afaerber has quit [Quit: Leaving]
18:07
jakogut has joined #linux-amlogic
18:07
jakogut has quit [Remote host closed the connection]
18:09
jakogut has joined #linux-amlogic
18:44
Ivanovic_ has joined #linux-amlogic
18:44
sputnik__ has joined #linux-amlogic
18:47
lissyx has joined #linux-amlogic
18:48
shadeslayer_ has joined #linux-amlogic
18:51
sputnik__ has quit [Ping timeout: 260 seconds]
18:51
sputnik__ has joined #linux-amlogic
18:52
xdarklight_ has joined #linux-amlogic
18:52
mag- has joined #linux-amlogic
18:53
sputnik_ has quit [*.net *.split]
18:53
mag has quit [*.net *.split]
18:53
xdarklight has quit [*.net *.split]
18:53
shadeslayer has quit [*.net *.split]
18:53
Ivanovic has quit [*.net *.split]
18:53
brujah has quit [*.net *.split]
18:54
Ivanovic_ is now known as Ivanovic
19:00
brujah has joined #linux-amlogic
19:11
ldevulder has quit [Ping timeout: 260 seconds]
19:14
fedux has joined #linux-amlogic
19:19
trem has joined #linux-amlogic
20:00
lissyx has quit [Ping timeout: 240 seconds]
20:04
fedux has joined #linux-amlogic
20:54
dsd has joined #linux-amlogic
21:31
trem has quit [Quit: Leaving]
21:36
ldevulder has joined #linux-amlogic
22:10
chewitt has quit [Read error: Connection reset by peer]
22:11
chewitt has joined #linux-amlogic
22:12
return0e has joined #linux-amlogic
22:18
return0e has quit [Read error: Connection reset by peer]
22:18
return0e_ has joined #linux-amlogic
22:22
dsd has quit [Quit: Lost terminal]
22:53
return0e has joined #linux-amlogic
22:56
return0e_ has quit [Ping timeout: 256 seconds]
23:57
return0e has quit []