ChanServ changed the topic of #linux-msm to:
BobBeck has quit [Remote host closed the connection]
marvin24 has joined #linux-msm
marvin24_ has quit [Ping timeout: 480 seconds]
<Tooniis[m]> any idea why cpuidle would make big cpus lock up on msm8996pro?
pevik has joined #linux-msm
jhovold has joined #linux-msm
pespin has joined #linux-msm
jhovold has quit [Quit: WeeChat 3.3]
jhovold has joined #linux-msm
BobBeck has joined #linux-msm
xerpi[m] has quit []
x[m] has quit []
<aka_[m]> ok im still out of idea how to get that debug info, i have changed 32 line of wcn36xx/main.c to
<aka_[m]> unsigned int wcn36xx_dbg_mask = 0xFFFFFFFF;
<aka_[m]> compiled and no idea how to progress
<aka_[m]> now it just shows "No adapter found"
<Mis012[m]> uh, isn't it supposed to be set on cmdline?
<aka_[m]> uh somehow some logs have appear in my dmesg
<aka_[m]> no idea how
<Mis012[m]> `[ 538.265697] wcn36xx: FW Cap UNKNOWN`
<Mis012[m]> fun
<aka_[m]> ok i have log
<aka_[m]> MIs how to fetch it now>_>
<Mis012[m]> hm?
<Mis012[m]> scp?
<aka_[m]> and how to use that?
<Mis012[m]> `scp user@172.16.42.1:/home/user/x.log /tmp/`?
<aka_[m]> bryanodonoghue:
<aka_[m]> bryanodonoghue:
pevik has quit [Ping timeout: 480 seconds]
xiaer1921 has quit [Remote host closed the connection]
<bryanodonoghue> aka_[m]
<bryanodonoghue> interestingly your firmware declares
<bryanodonoghue> [ 125.033964] wcn36xx: FW Cap DOT11AC
<bryanodonoghue> => wcn3680b
<bryanodonoghue> [ 125.062599] wcn36xx: Not in BMPS mode, no need to exit [ 125.062685] wcn36xx: HAL >>> 00000000: 4c 00 00 00 08 00 00 00 [ 125.063515] wcn36xx: SMD <<< 00000000: 5f 00 00 00 0c 00 00 00 00 00 00 00 [ 125.063638] wcn36xx: SMD command (req 76, rsp 95) completed in 0ms [ 125.063799] wcn36xx: Entered idle mode
<bryanodonoghue> so..
<bryanodonoghue> you've successfully loaded there
<bryanodonoghue> and entered into IMPS
<bryanodonoghue> Idle Mode Power Save
<bryanodonoghue> try
<bryanodonoghue> wpa_passwd "myssid" "mysuper-secretpassword"
<bryanodonoghue> > wpa_supplicant.conf
<bryanodonoghue> cp wpa_supplicant.conf /etc/wpa_supplicant
<bryanodonoghue> cat wu.sh
<bryanodonoghue> #!/bin/bash
<bryanodonoghue> modprobe qrtr_smd
<bryanodonoghue> modprobe qcom_wcnss_pil
<bryanodonoghue> modprobe rfcomm
<bryanodonoghue> sleep 8
<bryanodonoghue> systemctl stop NetworkManager.service
<bryanodonoghue> systemctl stop wpa_supplicant.service
<bryanodonoghue> wpa_supplicant -B -c /etc/wpa_supplicant/wpa_supplicant.conf -i wlan0 &
<bryanodonoghue> dhclient wlan0 &
<bryanodonoghue> iw phy0 wowlan show
<bryanodonoghue> iw phy0 wowlan enable any
<bryanodonoghue> that's my dirty hack script for just about any linux user-space
<bryanodonoghue> it *looks*
<bryanodonoghue> like you have successfully loaded wcn36xx
<bryanodonoghue> like you have a wcn3680b
<bryanodonoghue> and all you need to do is associate / authenticate with your AP
<aka_[m]> Rn I'm in job and it's 4pm I will be back 11pm,shoudl I change compatible to 3680?
<aka_[m]> I was using 3660b like in msm8953 as I seen this chip used on redmi note 3 Pro for which I found schematic(its 8956 which is 8976 without 2 big cores)
<aka_[m]> So I should change to 3680b Then network manager should show me networks in phosh settings?
<aka_[m]> Out of curiosity why iris reg on downstream reported 3660b?
<bryanodonoghue> I don't know if you have a wcn3660
<bryanodonoghue> The firmware you are running as 802.11ac support
<bryanodonoghue> only 3680b can do AC
<bryanodonoghue> we should really write some better detection logic
<bryanodonoghue> anyway supporting DOT11AC
<bryanodonoghue> means your firmware declares it can process the extended messages between AP and WCNSS
<bryanodonoghue> so far I've not encountered a fw that declares DOT11AC and lies
<bryanodonoghue> yep
<bryanodonoghue> my downstream image for wcn380b
<bryanodonoghue> says this
<bryanodonoghue> [ 19.267720] wcnss: IRIS Reg: 04000004
<bryanodonoghue> aka_[m]
<bryanodonoghue> from the command line try
<aka_[m]> Ah
<bryanodonoghue> iw wlan0 scan
<aka_[m]> Ok so I will try after I change compatible
<aka_[m]> Now I hope there will appear other pro in adsp space xd
<bryanodonoghue> heh srinik
<srinik> bryanodonoghue: hi Bryan..
<bryanodonoghue> you're the adsp expert afaik..
<srinik> bryanodonoghue: what was the adsp question ? sorry am unable to find it.
<bryanodonoghue> idk aka_[m] has some sort of problem I think
<bryanodonoghue> just pinging you so they know who to bug
<aka_[m]> Well it's more qdsp issue probably, once it tries to touch primary mi2s stuff it crash remoteproc
<aka_[m]> There is no schematic for this device but I assume its using pm8952 codec and tfa9890 is connected to quinary interface
<aka_[m]> It feels like once qdsp code tries to set primary mi2s cpu dai Clk it crash
<srinik> aka_[m]: sounds like clks are not enabled
<aka_[m]> srinik: I added q6afe-clocks under lpass but not sure if they are supported
<aka_[m]> Clk_summary shows some clocks for that enabled
<srinik> aka_[m]: which SoC is this?
<aka_[m]> MSM8976
<aka_[m]> I'm trying with MSM8953 qdsp compatible:
<aka_[m]> Oh damn i forgot something there, bit clk map missing quinary
<srinik> aka_[m]: once that is fixed, if sending out AFE CMD to enable clock still crashes dsp, then am suspecting some adsp clocks or power-domains are not voted from the cpu side.
<aka_[m]> Maybe this device is just funky? I see a lot of changes in downstream code for that device
<aka_[m]> It's this one
<aka_[m]> I see some mi2s_rx_clk being set to lpass_osr_clk_12p288
<aka_[m]> But that shouldn't influence primary right?
<srinik> aka_[m]: It depends on the clk mux setup inside LPASS. But this is OSR clk, even if its shared you should not see any impact, dsp will probably collect votes on this..
<aka_[m]> Piece of old log:
pespin has quit [Remote host closed the connection]
jhovold has quit [Ping timeout: 480 seconds]
pevik has joined #linux-msm
<aka_[m]> bryanodonoghue: with 3680 it was showing me networks for a bit but failed anyway... (full message at https://matrix.org/_matrix/media/r0/download/matrix.org/WKPEjpJGVHRounfKzvPRPTPs)
<aka_[m]> i wonder if there is difference if wcnss is remoteproc0 or remoteproc1
<aka_[m]> my downstream node have this property:
<aka_[m]> notsure if it even matters
<aka_[m]> qcom,pc-disable-latency = <0x50>;
<aka_[m]> optional is also after it does "platform remove" and there is
<aka_[m]> "mac start" it ends in oops with "unable to handle kernel paging request at virtual address fff-more-letters
<aka_[m]> s/optional//, s/is//
<aka_[m]> some log
<aka_[m]> after forth crash
pevik has quit [Ping timeout: 480 seconds]
cxl000 has quit [Ping timeout: 480 seconds]