marvin24 has joined #linux-msm
marvin24_ has quit [Ping timeout: 480 seconds]
jhovold has joined #linux-msm
<
aka_[m]>
have anyone tested recent next?
<
aka_[m]>
no idea if i should rebase over it or just go with torvalds/master
svarbanov has joined #linux-msm
Danct12 has quit [Quit: A-lined: User has been AVIVA lined]
Danct12 has joined #linux-msm
davidinux has joined #linux-msm
davidinux has quit [Ping timeout: 480 seconds]
<
aka_[m]>
appears to work with few patches from @kanuj:matrix.org
jhovold has quit [Ping timeout: 480 seconds]
davidinux has joined #linux-msm
<
aka_[m]>
bryanodonoghue: are you going to add XO to apcs-msm8916.c?
<
aka_[m]>
apcs1_mbox has "ref" passed in clocks
<
aka_[m]>
how does flow of booting HFPLL work?
davidinux has quit [Ping timeout: 480 seconds]
svarbanov has quit [charon.oftc.net liquid.oftc.net]
marvin24 has quit [charon.oftc.net liquid.oftc.net]
aat596_2 has quit [charon.oftc.net liquid.oftc.net]
telent has quit [charon.oftc.net liquid.oftc.net]
kholk has quit [charon.oftc.net liquid.oftc.net]
rmsilva has quit [charon.oftc.net liquid.oftc.net]
nergzd723 has quit [charon.oftc.net liquid.oftc.net]
Marijn[m] has quit [charon.oftc.net liquid.oftc.net]
konradybcio has quit [charon.oftc.net liquid.oftc.net]
Tooniis[m] has quit [charon.oftc.net liquid.oftc.net]
krzk has quit [charon.oftc.net liquid.oftc.net]
_lore_ has quit [charon.oftc.net liquid.oftc.net]
rawoul has quit [charon.oftc.net liquid.oftc.net]
suihkulokki has quit [charon.oftc.net liquid.oftc.net]
telent has joined #linux-msm
aat596_2 has joined #linux-msm
svarbanov has joined #linux-msm
marvin24 has joined #linux-msm
krzk has joined #linux-msm
nergzd723 has joined #linux-msm
Marijn[m] has joined #linux-msm
_lore_ has joined #linux-msm
Tooniis[m] has joined #linux-msm
rmsilva has joined #linux-msm
konradybcio has joined #linux-msm
kholk has joined #linux-msm
suihkulokki has joined #linux-msm
rawoul has joined #linux-msm
<
aka_[m]>
ok without pre-programming l_val i believe there is situation where PLL is inited without rate set
<
aka_[m]>
lumag: is there a chance for that?
<
aka_[m]>
after adding l_val write inside init_once it appears to boot well
<
lumag>
aka_[m], I think so. L is PLL frequency multiplier
<
aka_[m]>
yes i know
<
lumag>
If it was not set up by bootloader, you have to do it
<
aka_[m]>
so its set on 8974?
<
lumag>
Maybe I missed the previous question.
<
aka_[m]>
8974 have few hfplls in it
<
lumag>
Haven't tried that board for a while, but I don't remember any issue with it.
<
aka_[m]>
i decided to go for dirtierapproach which would be faster on 8976
<
lumag>
Maybe I missed it.
<
lumag>
I haven't yet looked at the cpufreq on 8974
<
aka_[m]>
but without writing L in init_once it does reboot
<
lumag>
Sounds sensible
<
aka_[m]>
i tried assigned-clocks/parents/rates
<
aka_[m]>
and it appears to first boot pll then maybe later do set_rate
<
lumag>
Yes, as far as I understand.
<
lumag>
You can do the initial set_rate manually by setting .l to a known value
<
lumag>
But that's it
<
aka_[m]>
gonna guard behind if obv
<
aka_[m]>
maybe one day i would unfold user_reg but this is quite bothersome
<
aka_[m]>
are these okish?
<
aka_[m]>
quite dirty bringup but i experience platform hang on latest torvalds/master
<
aka_[m]>
some clocks might be shut down
<
aka_[m]>
after a while it reboots
<
bryanodonoghue>
aka_[m] I don't have many plans for 8916 save verification of recent camss patch
<
lumag>
aka_[m], I'd notice that other platforms don't instantiate hfplls directly.
<
lumag>
You most likely need some code around it to prepare core clocks, etc.
<
lumag>
So, I'd say, qcs404 road might be not the best one
<
lumag>
on 8960/8064 the hfpll's are handled by the gcc driver, because they are a part of the same region.
<
lumag>
Ok, maybe your approach is correct