Tony_ has quit [Remote host closed the connection]
<naobsd>
Tony_: are you using same kernel for android and linux?
markm_ has quit [Ping timeout: 255 seconds]
levd has quit [Ping timeout: 272 seconds]
dlezcano_ has quit [Ping timeout: 244 seconds]
dlezcano_ has joined #linux-rockchip
<cnxsoft1>
naobsd
BorgCuba has joined #linux-rockchip
<naobsd>
cnxsoft1: it seems "sign-in with twitter" on chipspark.com is working now
<naobsd>
cnxsoft1: and chipspark said that documents/source code for PopMetal will be provided soon
<BorgCuba>
hi, I am looking into the usb-phy stuff for the rk3066 (maybe this applies to the other socs as well) and it looks like its a synopsys "nano-phy".
<BorgCuba>
is there already some driver available for this ip in the kernel?
<cnxsoft1>
Good.
sats861 has joined #linux-rockchip
<sats861>
hi!
<sats861>
who have kernek 3288.git
<sats861>
?
antoinemaillard has quit [Quit: antoinemaillard]
dlezcano_ has quit [Ping timeout: 250 seconds]
dlezcano_ has joined #linux-rockchip
markm_ has joined #linux-rockchip
<AstralixNB1>
BorgCuba, I started to do the same, but I ignored a little cold that I had and it cam back and brought some friends... so the last two days wheren't so happy to me.
dlezcano_ has quit [Ping timeout: 272 seconds]
dlezcano_ has joined #linux-rockchip
naobsd has quit [Ping timeout: 246 seconds]
<BorgCuba>
AstralixNB1, so you were looking into adding missing sub phy support? thats great! I am only starting and havent done something like this before. But I am pretty sure that its the designware/synopsys nanophy
cnxsoft1 has quit [Ping timeout: 265 seconds]
<AstralixNB1>
There is a driver for 3288 proposed to mainline. And I guess the best is to use that one and extend it to 31/30
<AstralixNB1>
I already dfownloaded these sources, but then got interrupted ...#
<BorgCuba>
the block diagram for the phy from the rk30xx trm is the same like the one in this doc http://wenku.baidu.com/view/7f1aad6a58fafab069dc020d.html and in the description for GRF_UOC0_CON0:12 the term "USB 2.0 nanoPHY" is mentioned
<AstralixNB1>
I do have all needed information for all the chip. But thanks
<BorgCuba>
AstralixNB1, I installed that patch
<BorgCuba>
although I haven't yet tested it
<AstralixNB1>
Yes, RK reuses very much of their IPs. Why not, they have paid the license fees for usage and if they work...
<AstralixNB1>
I am a bit locked at xmas preparations for the family, so I will not do much more nbefore tonight.
<BorgCuba>
btw, maybe the exynos 4 also uses this phy (they are using dwc2 at least)
<BorgCuba>
do your preparations and get rid of your cold! :)
<BorgCuba>
I am already half prepared
<AstralixNB1>
That is possible. but the last times I checked the drivers, the Exynos varied veery mauch from the RK part.
<naobsd>
ganbold_: -kernel repo should be kernel subtree of full sdk tree
<naobsd>
ganbold_: firefly-rk3288.dts is the right one
<BorgCuba>
I am especially interested in "reg = <0x320>;" for rk3288
<AstralixNB1>
BorgCuba, the comparision of RK30xxx and 32xx is very complicated as the manuals have grown... The RK30xx are some first trials to document something, while the 31xx is the first thing that you might call a TRM. However I'll try to line these addresses and bits up for all three of them as good as I can
<AstralixNB1>
0x320 is an offset?
<ganbold_>
naobsd: yeah, still includes rk3288.dtsi
<naobsd>
ganbold_: ah, yes
<naobsd>
ganbold_: is your firefly stable now?
<AstralixNB1>
at least, both, 30xx and 32xx have basic USB and USB PHY bits well spread around GRF registers
dlezcano_ has quit [Ping timeout: 245 seconds]
dlezcano_ has joined #linux-rockchip
<ganbold_>
naobsd: it should be ok, with original power adapter boot didn't loop
<AstralixNB1>
BorgCuba, in RK3288 this offset is GRF but in RK30xx it is not available. So there will be some remapping tables needed
<AstralixNB1>
But I have to leave now. See you later
<BorgCuba>
AstralixNB1, I think for rk30xx its 0x178
<BorgCuba>
(bit 13 - siddq)
<BorgCuba>
AstralixNB1, for the rk30xx we have GRF_UOC0_CON0..2 and GRF_UOC1_CON0..3 (0x17c .. 0x194)
<BorgCuba>
well, and its not 0x178 but 0x17c (the grf reg offset)
dlezcano_ has quit [Ping timeout: 256 seconds]
dlezcano_ has joined #linux-rockchip
<rperier>
you need to implement a devicetree variant for that, reuse the phy-usb driver for rockchip, refactorize it and then add a devicetree variant to support rk3066/r3188
jmcneill has joined #linux-rockchip
<AstralixNB1>
hi reperier
<rperier>
hi AstralixNB1
<AstralixNB1>
rperier, do you have a already accepted example for this refactoring. So we can avoid style issues when implementing that.
dlezcano_ has quit [*.net *.split]
c0d3z3r0 has quit [*.net *.split]
eebrah has quit [*.net *.split]
akaizen_ has quit [*.net *.split]
mmind00 has quit [*.net *.split]
zombu2 has quit [*.net *.split]
ssvb has quit [*.net *.split]
leirbag has quit [*.net *.split]
else- has quit [*.net *.split]
honx has quit [*.net *.split]
lerc has quit [*.net *.split]
mrueg has quit [*.net *.split]
Luke-Jr has quit [*.net *.split]
Danukeru has quit [*.net *.split]
ChanServ has quit [*.net *.split]
RaYmAn has quit [*.net *.split]
mrcan has quit [*.net *.split]
phh has quit [*.net *.split]
FreezingCold has quit [*.net *.split]
hipboi has quit [*.net *.split]
FergusL has quit [*.net *.split]
ganbold_ has quit [*.net *.split]
nashpa has quit [*.net *.split]
dianders has quit [*.net *.split]
rperier has quit [*.net *.split]
killbit has quit [*.net *.split]
architekt has quit [*.net *.split]
khilman has quit [*.net *.split]
VargaD has quit [*.net *.split]
npcomp has quit [*.net *.split]
karlp has quit [*.net *.split]
maz has quit [*.net *.split]
naobsd has quit [*.net *.split]
sats861 has quit [*.net *.split]
BorgCuba has quit [*.net *.split]
AstralixNB1 has quit [*.net *.split]
libv has quit [*.net *.split]
strange has quit [*.net *.split]
ynezz has quit [*.net *.split]
amstan has quit [*.net *.split]
johnsu01 has quit [*.net *.split]
zoobab_ has quit [*.net *.split]
khilman has joined #linux-rockchip
RaYmAn has joined #linux-rockchip
ynezz has joined #linux-rockchip
architekt has joined #linux-rockchip
ChanServ has joined #linux-rockchip
dianders has joined #linux-rockchip
maz has joined #linux-rockchip
phh has joined #linux-rockchip
mrcan has joined #linux-rockchip
karlp has joined #linux-rockchip
strange has joined #linux-rockchip
rperier has joined #linux-rockchip
zoobab_ has joined #linux-rockchip
libv has joined #linux-rockchip
npcomp has joined #linux-rockchip
hipboi has joined #linux-rockchip
johnsu01 has joined #linux-rockchip
FergusL has joined #linux-rockchip
killbit has joined #linux-rockchip
nashpa has joined #linux-rockchip
Luke-Jr has joined #linux-rockchip
VargaD has joined #linux-rockchip
amstan has joined #linux-rockchip
ganbold_ has joined #linux-rockchip
FreezingCold has joined #linux-rockchip
AstralixNB1 has joined #linux-rockchip
Danukeru has joined #linux-rockchip
BorgCuba has joined #linux-rockchip
sats861 has joined #linux-rockchip
naobsd has joined #linux-rockchip
dlezcano_ has joined #linux-rockchip
c0d3z3r0 has joined #linux-rockchip
eebrah has joined #linux-rockchip
leirbag has joined #linux-rockchip
<BorgCuba>
yes
cnxsoft has joined #linux-rockchip
lerc has joined #linux-rockchip
mrueg has joined #linux-rockchip
<BorgCuba>
rperier, any idea?
<rperier>
when dwc2 try to use/enable your phy usb... something goes wrong apparently... did you look at /home/jan/devel/rk30xx/linux-next-3.19.0-rc1-next-20141221/drivers/usb/dwc2/gadget.c:202 ?
<rperier>
also try to find the point from where the message "insufficient fifo memory" is emitted, it might help
AstralixNB1 has quit [Remote host closed the connection]
<BorgCuba>
ohh, the reg values are 0x17c for phy0 and 0x188 for phy1 (otg)
<BorgCuba>
for rk3066
dlezcano_ has quit [Ping timeout: 250 seconds]
dlezcano_ has joined #linux-rockchip
dlezcano_ has quit [Ping timeout: 252 seconds]
dlezcano_ has joined #linux-rockchip
sats861 has quit [Ping timeout: 246 seconds]
<mmind00>
BorgCuba: I see the pll locking issue now too on my marsboard ... the pll-init code itself is most likely innocent, but exposes a problem of the pll code
<mmind00>
BorgCuba: I'll try to hunt this down
dlezcano_ has quit [Ping timeout: 264 seconds]
dlezcano_ has joined #linux-rockchip
_andrew_ has joined #linux-rockchip
mrcan has quit [Ping timeout: 244 seconds]
field^Mop has joined #linux-rockchip
mrcan has joined #linux-rockchip
mrcan has quit [Changing host]
mrcan has joined #linux-rockchip
field^Mop has quit [Ping timeout: 256 seconds]
<BorgCuba>
mmind00, thank you for taking a look at the pll issue
<BorgCuba>
mmind00, the dwc2 gadget seems to have some problems. But as you said this is probably wip and I already saw someone on the mailing list reported something similar (http://www.kernelhub.org/?msg=652138&p=2)
<BorgCuba>
mmind00, if you are interested in enabling the usb2 phy driver for your marsboard I could prepare something for you?
paowz has joined #linux-rockchip
<mmind00>
BorgCuba: as rperier already said, you should extend the proposed rk3288 phy driver with support for the older variants ... the dw-nano-phy doesn't count here, as the access to the phys on all socs is done via the GRF registers
<BorgCuba>
I think it is working?
<BorgCuba>
I mean I can see on the pc that it tries to enumerate and something inside the gadget driver seems to fail
<BorgCuba>
I build the dwc2 with all debugging message options activated
JohnDoe_71Rus has joined #linux-rockchip
<BorgCuba>
mmind00, basically the driver is very simple, it only turns on the phy (bit 13 of GRF_UOC0_CON0). So I didnt even have to change anything in the driver, I only had to plug in the correct register offset in the dts file (0x17c and 0x188 for rk30xx)
<BorgCuba>
what makes you guys think that the issue I am facing is related to the phy driver?
<BorgCuba>
I mean its possible, but I dont see this at the moment.
PaowZ_ has joined #linux-rockchip
paowz has quit [Ping timeout: 252 seconds]
<mmind00>
BorgCuba: I haven't had to much contact with the dwc2 itself except following patches for it a bit ... so I've also only skimmed through the stuff you talked with rperier(?) and Astraix(?) about
<mmind00>
BorgCuba: so I'm probably not the best person to talk to about dwc2 issues
<BorgCuba>
btw, I think Ive never seen a rk3xxx kernel with working usb gadget driver. I only have some uboot code for rk30xxx that I have tested and that works
PaowZ__ has joined #linux-rockchip
PaowZ_ has quit [Ping timeout: 244 seconds]
paowz has joined #linux-rockchip
antoinemaillard has quit [Quit: antoinemaillard]
PaowZ_ has joined #linux-rockchip
PaowZ__ has quit [Ping timeout: 255 seconds]
antoinemaillard has joined #linux-rockchip
antoinemaillard has quit [Client Quit]
paowz has quit [Ping timeout: 264 seconds]
dlezcano_ has quit [Ping timeout: 264 seconds]
<rperier>
dwc2 is buggy with recent changes on some stuffs, I can confirm that
dlezcano_ has joined #linux-rockchip
antoinemaillard has joined #linux-rockchip
paowz has joined #linux-rockchip
<BorgCuba>
rperier, have you already tested the phy???
dlezcano_ has quit [Ping timeout: 240 seconds]
dlezcano_ has joined #linux-rockchip
<BorgCuba>
I think it works, but its always good to have someone else to confirm/deny
PaowZ_ has quit [Ping timeout: 240 seconds]
antoinemaillard has quit [Ping timeout: 245 seconds]
dlezcano_ has quit [Ping timeout: 245 seconds]
dlezcano_ has joined #linux-rockchip
<rperier>
BorgCuba: no, however I already tested and fixed some things dwc2, and I found that it was buggy since 3.18 (I use 3.19-rc1 here, same issues with linux-next)
<rperier>
mhhh... interesting... with the usb ttl converter from radxa (the green one) my uart debug is very stable
<BorgCuba>
too bad, I am using 3.19.0-rc1-next-20141221 (only the gadget part)
<rperier>
so it has really to do with the usb ttl converter...
<rperier>
(sorry, it's another issue I had early)
<rperier>
:)
<BorgCuba>
rperier, so you also have uart problems?
<rperier>
I had uart issues on my radxa rock pro
<rperier>
which I did not have with my radxa rock first edition
<rperier>
but with the good usb ttl converter, it seems to be solved
<rperier>
so good news
<BorgCuba>
I have problems with my rk3188 and rk3188t boxes
<rperier>
with uart2 ?
<BorgCuba>
with the default debug uart
<BorgCuba>
i.e. the one also used by the rock-loader
<BorgCuba>
and the ddr init
<BorgCuba>
I measured the voltage level on one board and it was like approx 3V while my ft232 also pulls the rx line up to 3v3
<BorgCuba>
I think next time I will build a little opamp buffer maybe a use a fast opto (but I'll have to look at the ds before - how many mA the rk-tx pin can drive)