ChanServ changed the topic of #linux-rockchip to: Rockchip development discussion | Wiki at http://linux-rockchip.info | Logs at http://irclog.whitequark.org/linux-rockchip | ML at http://groups.google.com/group/linux-rockchip
<bbelos> might someone have any insight on these DVFS ERRs at line 152? http://pastebin.com/gix0ZCTc
<naobsd> ganbold_: there might be code that "check lock or unlock" around boot code in u-boot
<naobsd> ganbold_: I'll check it
<naobsd> ganbold_: but I think you can send u-boot command via fastboot
<naobsd> ganbold_: I cannot remember syntax now, I'll check it too
<jmcneill> naobsd, I see you added chipver=, is there a list of possible values?
<naobsd> jmcneill: only in GPL code... I'll paste RK3o66/RK3188 soon
<naobsd> rk3066 0x33303041 0x32303131 0x31313131 0x56313031
<naobsd> 300A 2011 1111 V101
<jmcneill> it is passed as a string?
<naobsd> rk3188 0x33313042 0x32303132 0x31313330 0x56313030
<naobsd> 310B 2012 1130 V100
<naobsd> jmcneill: in my implementation, yes
<naobsd> but as 16B stream. it's actually values of four 32bit width register
<naobsd> so
<naobsd> rk3066 A00311021111101V
<naobsd> rk3188 B01321020311001V
<naobsd> is passed as chipver=
<naobsd> first 4B is (rough) SoC id
<naobsd> next 4B is probably year, next 4B is probably month/day
<naobsd> last 4B is version
<naobsd> I think I have some RK3188 which shows another ymd and/or version
<jmcneill> should it be read 300A or A003 ?
<naobsd> if you don't convert 16B to 4x uint32, "A003" is correct
<jmcneill> I get that part, just wondering how to interpret the fields. They don't seem arbitrary.
<jmcneill> I see 2012 in there, that is year right?
<jmcneill> Might make the chipver= param easier to read if you passed them in that order.
<naobsd> 2nd 4B is probably year
<naobsd> I see
<naobsd> I'll convert it on u-boot
<jmcneill> Thanks. So to detect RK3188, should I only look for "310B" or are there other chips that use the same prefix?
<naobsd> it seems there is rk3188plus uses same 1st 4B
<jmcneill> ok
<naobsd> I'm sure that 16B is not perfect to identify all RK SoCs
<naobsd> but I think it's ok for now
<jmcneill> The other thing you could do is pass chipver=rk3188 :)
<jmcneill> and let me not have to worry about it!
<naobsd> ;)
<naobsd> hmm
<naobsd> no, I want to keep it as 16B string
<jmcneill> chipver=B01331021310101V
<naobsd> as like as debug printf
<jmcneill> that's what I get
<jmcneill> for RK3188
<jmcneill> 2013 edition
<naobsd> oh
<jmcneill> fb.addr=0x63c08000 <-- does that work?
<naobsd> ah, it's plus
<naobsd> I'm not sure which is common, plus or non-plus
<naobsd> there are some if (plus) in ddr config code and
<naobsd> power management code (cru setting) and
<naobsd> rk3188_delayline_init (what's this) and
<jmcneill> interesting
<jmcneill> i had no idea, marketed as a 3188
<naobsd> plus_pll_clk_set_rate etc
<naobsd> currently I'm not sure we're (mostly) using which one
<naobsd> anyway
<naobsd> I want to pass all 16B for now ;)
<jmcneill> well now I have no idea if my cru code is correct
<jmcneill> argrg
<naobsd> and I'm not sure mainline clk-rk3188.c is for which ;)
<naobsd> I guess all we have plus, then it's no problem ;)
<naobsd> then, fb,
<naobsd> u-boot for rk30/31 has code to init fb/lcdc
<naobsd> I'm not sure it's really working
<naobsd> because hdmi out needs external transmitter such as it66121
<naobsd> jmcneill: your x7 has vga out?
<jmcneill> hdmi
<naobsd> naobsd: hm. what Radxa Rock did you bought?
<naobsd> oops
<jmcneill> lite, not sure when that will arrive
<naobsd> I see
<jmcneill> actually i think i have at tracking number, sec
field^Mop has joined #linux-rockchip
<naobsd> it has RK616 for HDMI/LVDS out and RK1000 for CVBS out, both needs configration via i2c
<jmcneill> which post starts with RF
<jmcneill> singapore post
<naobsd> (and maybe power control via act8846 via i2c)
<jmcneill> status: Despatched to overseas (Country code: CA)
<jmcneill> on the 27th
<jmcneill> Canada Post doesn't know anything about it yet
<naobsd> jmcneill: if you have parallel RGB LCD unit, you may use it on RR
<jmcneill> i do not
<naobsd> then you need to init external video signal converter unit via i2c to get picture...
<jmcneill> sounds like a job for naobsd
<naobsd> hehe
<jmcneill> skrll@ finally received his a80 board, hoping for a gic fix soon
<jmcneill> and hey, the cubie4 code i wrote works on optimusboard
<naobsd> I have pararell RGB to VGA converter module which doesn't require i2c control ;)
<naobsd> sounds good :)
<jmcneill> awin a20/a31 has built in hdmi tx
<jmcneill> not sure about a80 though
<jmcneill> cubie4 uboot sets it up enough to use it at 1280x720 at least
<naobsd> oh RK3066/PX2/RK3288 has internal HDMI unit... do you want to hear rest of my words? ;)
<jmcneill> you are going to send me an rk3288 and trm
<naobsd> s/trm//
<naobsd> s/^/px2/
<jmcneill> is px2 really so bad that you can't even give it away
<naobsd> I never thought px2 is bad... I like it too ;)
<naobsd> ah, I don't send _my_ board
<naobsd> I need it :)
levd has joined #linux-rockchip
<jmcneill> well the cpu freq scaling is definitely working
<naobsd> jmcneill: btw if you need RR _pro_, you can ask hipboi
<jmcneill> here is 600MHz (openssl speed md5):
<naobsd> jmcneill: at run time?
<jmcneill> md5 1470.27k 5416.79k 17489.92k 40214.96k 66266.94k
<jmcneill> and 1602MHz:
<jmcneill> md5 4075.09k 15118.14k 48740.52k 112278.87k 182635.18k
<naobsd> jmcneill: very clear :)
<jmcneill> openssl speed my go-to for checking cpu freq scaling stuff
<naobsd> openssl speed -multi N is good for baking CPU
<naobsd> then, I forgot what I need to do... ah, chipver= string
cosm has quit [Ping timeout: 265 seconds]
<naobsd> ganbold_: can you post bootlog around u-boot propmt? I cannot put 2 board on my desk same time :(
<jmcneill> RR pro looks like the same specs as x7
<naobsd> RR lite <-> pro is: RAM 1GB <-> 2GB, +8GB NAND, different USB-WLAN/BT unit
<naobsd> actually I should ask before you bought lite
<naobsd> too late, sorry ;)
zombu2 has quit [Ping timeout: 255 seconds]
zombu2 has joined #linux-rockchip
Avagetto has quit [Quit: Leaving.]
levd1 has joined #linux-rockchip
Avagetto has joined #linux-rockchip
Avagetto has quit [Client Quit]
<naobsd> ah
<naobsd> well
<naobsd> u-boot think 310B V101 is RK3188B...
levd has quit [Ping timeout: 272 seconds]
<naobsd> jmcneill: u-boot says "CPU: RK3188B"?
<jmcneill> ok
<naobsd> ah sorry, that print is defined at compile time
<jmcneill> that would explain why the cpufreq stuff works i think
<naobsd> I'm confusing a little
<naobsd> wait a moment...
levd has joined #linux-rockchip
<jmcneill> when you are ready..
<jmcneill> I have it set so you can set cpu.frequency=1608 (or some other supported speed) in parameters file to select freq at boot
<jmcneill> just need those IDs from you
levd1 has quit [Ping timeout: 264 seconds]
FreezingCold has joined #linux-rockchip
<naobsd> jmcneill: my RR pro is 310B20121130V100 :(
<naobsd> well
<naobsd> there is a cpu id code in u-boot rk
<naobsd> I can pass result code instead of chipver string
<naobsd> but u-boot rk30/31 is obsoleted, not maintained by rk long time ago
<jmcneill> chipver=rk3188 seems a much more stable interface :)
<naobsd> u-boot rk30/31 code is not so reliable
<naobsd> if u-boot passes id result, I should make it reliable
<naobsd> I think we drop this u-boot in future, upstreaming should be done...
<jmcneill> are you volunteering? :)
<naobsd> I'm lazy, if I do something for u-boot rk30/31 now, it just passes RK3066 or RK3188 or RK3188B, others will be ignored ;)
<naobsd> mm, RK3188B or RK3188plus, which is correct ;)
<naobsd> rk uses a/b suffix for some SoC
<naobsd> but strangely T or plus is used for 3188 in kernel source
<naobsd> there is a code which uses "rk3188+" for plus/B
<ganbold_> naobsd: http://pastebin.ca/2894538
Avagetto has joined #linux-rockchip
<naobsd> hmm.......
<naobsd> jmcneill: how about socid=rk3188+ (or rk3188 or rk3066)
<jmcneill> works for me!
<naobsd> I still prefer chipver= with raw info...
<naobsd> but ok
<jmcneill> i don't really have a preference
<jmcneill> for chipver can you guarantee that all past and future socs will have that 16-byte ID?
<naobsd> what I'm thinking is I have to edit u-boot in future when new SoC is supported to NetBSD
<naobsd> jmcneill: anything I can guarantee _for now_ is only for rk3066 and rk3188(+)
<jmcneill> heh
Avagetto1 has joined #linux-rockchip
<jmcneill> that's the only reason why i lean toward socid solution
<jmcneill> i see what you're saying though
Avagetto has quit [Ping timeout: 240 seconds]
<jmcneill> lets go with chipver, and you can fake an ID for boards without a real one
<jmcneill> chipver=BOOTLEG or whatever
<naobsd> at least reading chipver must be always done in u-boot rk30/31
<naobsd> it should work on SoCs that current u-boot rk30/31 support
<jmcneill> yeah I think that's fine
<jmcneill> rk3288 will need a different kernel anyway i think
<naobsd> yes
<jmcneill> so let me know when you are ready, and give me the IDs to use :)
<naobsd> btw u-boot rk32 is based different u-boot ver, many things are different ;)
<naobsd> both u-boot should be upstreamed...
<naobsd> well
<naobsd> chipid=(raw 16B str) is ok, right?
<naobsd> only I have to do is swapping
<jmcneill> yep, just swap the byte order so it makes sense
<naobsd> I see
<jmcneill> 310B20121130V100 style
cnxsoft has joined #linux-rockchip
<jmcneill> there is one other option we didn't consider
<jmcneill> you could pass the equivalent of the "compatible" string from linux dts
<jmcneill> rockchip,rk3188
<jmcneill> same problem as socid i suppose, if you don't have dts
cnxsoft1 has joined #linux-rockchip
cnxsoft has quit [Ping timeout: 244 seconds]
<naobsd> sorry, need to take (late) breakfast....
<naobsd> ganbold_: where is boot countdown?
Avagetto has joined #linux-rockchip
Avagetto1 has quit [Ping timeout: 264 seconds]
<ganbold_> naobsd: no countdown
<ganbold_> naobsd: that is disappeared after I did oem unlock via fastboot I think
Avagetto1 has joined #linux-rockchip
Avagetto has quit [Ping timeout: 245 seconds]
Avagetto has joined #linux-rockchip
Avagetto1 has quit [Ping timeout: 244 seconds]
Avagetto1 has joined #linux-rockchip
Avagetto has quit [Ping timeout: 244 seconds]
Avagetto has joined #linux-rockchip
field^Mop has quit [Ping timeout: 244 seconds]
Avagetto1 has quit [Ping timeout: 245 seconds]
levd1 has joined #linux-rockchip
levd has quit [Ping timeout: 244 seconds]
Avagetto has quit [Ping timeout: 240 seconds]
ojn_ has quit [Ping timeout: 258 seconds]
ojn has joined #linux-rockchip
<naobsd> grr
<naobsd> I somewhat make u-boot unbootable ;)
<naobsd> where is needle...
<jmcneill> are you planning on keeping the name chipver or chipid?
levd has joined #linux-rockchip
<jmcneill> zzz
levd1 has quit [Ping timeout: 255 seconds]
levd1 has joined #linux-rockchip
<naobsd> chipver
<naobsd> mmm
levd has quit [Ping timeout: 265 seconds]
<naobsd> how can I keep content of global var while booting u-boot ;)
<naobsd> strange...
Astralix1 has joined #linux-rockchip
Astralix has quit [Ping timeout: 250 seconds]
Avagetto has joined #linux-rockchip
<naobsd> ok, my C coding skill is really terrible ;)
<naobsd> u-boot binary is updated
<naobsd> new u-boot says like ChipVer: 310B20121130V100
<naobsd> on boot
<naobsd> jmcneill: you should get CPU: RK3188+ now, please check
bengal has quit [Ping timeout: 255 seconds]
<naobsd> PX2(RK3066) 300A20111111V101
<naobsd> RK3068(RK3066) 300A20111111V101
<naobsd> hm, same
<naobsd> my RR lite 2013 310B20121130V100
<naobsd> same as RR pro
<naobsd> where is "different one"... I think I have it
<ganbold_> naobsd: do you know if rk3288 uart is different than rk3288?
<ganbold_> I mean 3188
<naobsd> ganbold_: yes
<naobsd> address is different
hipboi has joined #linux-rockchip
<ganbold_> I mean other than that
<naobsd> ganbold_: I think no
hipboi_ has quit [Ping timeout: 244 seconds]
<naobsd> oops, that was RR full
<naobsd> RR lite is 310B20121130V100 same
<naobsd> RR ES 310B20121130V100 same
<naobsd> oops, I remembered my RK3188 stick (HTC-T010-V2) doesn't boot with u-boot
bengal has joined #linux-rockchip
levd has joined #linux-rockchip
levd1 has quit [Ping timeout: 244 seconds]
<naobsd> jmcneill: CRU_PLL_CON0_CLKR same
<naobsd> jmcneill: CRU_PLL_CON0_CLKOD __BITS(5,0) -> __BITS(3,0) for rk3188+ and rk3066
<naobsd> jmcneill: CRU_PLL_CON1_CLKF __BITS(15,0) -> __BITS(12,0) for 3188+ and 3066
<naobsd> jmcneill: CRU_PLL_CON3_POWER_DOWN__BIT(1) -> 3188+ uses BIT(5) (reset)
<naobsd> jmcneill: in addition to change CON[01], 3188+ also change CON2
<naobsd> it seems delay just before waiting lock is different...
levd has quit [Remote host closed the connection]
levd has joined #linux-rockchip
<naobsd> ok, set CON2 and use CON3[5] is same for rk3066
levd1 has joined #linux-rockchip
<naobsd> and some delay before waiting lock state
<naobsd> very strange, why jmcneill's 3188+ works fine with current code for 3188 ;)
<naobsd> and "do same as 3066" doesn't sound like "plus/B"
<naobsd> ;)
levd has quit [Ping timeout: 240 seconds]
<naobsd> btw, CON3[5] reset bit seems to be "rest" bit by comment in rk u-boot/kernel
<naobsd> /* enter rest */
hipboi has quit [Read error: Connection timed out]
hipboi has joined #linux-rockchip
levd has joined #linux-rockchip
levd1 has quit [Ping timeout: 256 seconds]
Avagetto1 has joined #linux-rockchip
Avagetto has quit [Ping timeout: 265 seconds]
<naobsd> I have to learn about clksel more...
Avagetto1 has quit [Ping timeout: 240 seconds]
Avagetto has joined #linux-rockchip
levd has quit [Remote host closed the connection]
levd has joined #linux-rockchip
levd1 has joined #linux-rockchip
levd has quit [Ping timeout: 272 seconds]
Avagetto has quit [Ping timeout: 265 seconds]
Avagetto has joined #linux-rockchip
levd has joined #linux-rockchip
Avagetto1 has joined #linux-rockchip
Avagetto has quit [Ping timeout: 245 seconds]
levd1 has quit [Ping timeout: 244 seconds]
Avagetto has joined #linux-rockchip
Avagetto2 has joined #linux-rockchip
Avagetto1 has quit [Ping timeout: 255 seconds]
Avagetto has quit [Ping timeout: 250 seconds]
Avagetto has joined #linux-rockchip
Avagetto2 has quit [Ping timeout: 255 seconds]
GriefNorth has joined #linux-rockchip
Avagetto1 has joined #linux-rockchip
Avagetto has quit [Ping timeout: 258 seconds]
Avagetto has joined #linux-rockchip
Avagetto1 has quit [Ping timeout: 272 seconds]
Avagetto1 has joined #linux-rockchip
Avagetto has quit [Ping timeout: 245 seconds]
Avagetto has joined #linux-rockchip
Avagetto1 has quit [Ping timeout: 244 seconds]
Avagetto1 has joined #linux-rockchip
jas-hacks has joined #linux-rockchip
Avagetto has quit [Ping timeout: 256 seconds]
levd1 has joined #linux-rockchip
Avagetto has joined #linux-rockchip
levd has quit [Ping timeout: 255 seconds]
Avagetto1 has quit [Ping timeout: 244 seconds]
Avagetto1 has joined #linux-rockchip
Avagetto has quit [Ping timeout: 245 seconds]
<jas-hacks> naobsd: hi, with 3.18.0-next-20141210 on firefly problem with SD/MMC driver. http://pastebin.com/5jZxTmqj
Avagetto has joined #linux-rockchip
Avagetto2 has joined #linux-rockchip
Avagetto1 has quit [Ping timeout: 244 seconds]
mrcan__ is now known as mrcan
<naobsd> jas-hacks: I see, thank you
Avagetto has quit [Ping timeout: 250 seconds]
Avagetto has joined #linux-rockchip
bengal_ has joined #linux-rockchip
bengal has quit [Ping timeout: 244 seconds]
Avagetto2 has quit [Ping timeout: 255 seconds]
Avagetto1 has joined #linux-rockchip
Avagetto has quit [Ping timeout: 258 seconds]
Avagetto has joined #linux-rockchip
Avagetto1 has quit [Ping timeout: 245 seconds]
Avagetto1 has joined #linux-rockchip
Avagetto has quit [Ping timeout: 244 seconds]
Avagetto has joined #linux-rockchip
Avagetto1 has quit [Ping timeout: 244 seconds]
Avagetto1 has joined #linux-rockchip
Avagetto has quit [Ping timeout: 258 seconds]
Avagetto has joined #linux-rockchip
Avagetto1 has quit [Ping timeout: 250 seconds]
Avagetto1 has joined #linux-rockchip
Avagetto has quit [Ping timeout: 244 seconds]
levd has joined #linux-rockchip
levd1 has quit [Ping timeout: 255 seconds]
Avagetto has joined #linux-rockchip
Avagetto1 has quit [Ping timeout: 240 seconds]
Avagetto1 has joined #linux-rockchip
Avagetto has quit [Ping timeout: 240 seconds]
Avagetto has joined #linux-rockchip
Avagetto1 has quit [Ping timeout: 244 seconds]
levd1 has joined #linux-rockchip
Avagetto1 has joined #linux-rockchip
levd has quit [Ping timeout: 258 seconds]
Avagetto has quit [Ping timeout: 256 seconds]
Avagetto has joined #linux-rockchip
Avagetto1 has quit [Ping timeout: 258 seconds]
Avagetto1 has joined #linux-rockchip
bengal_ has quit [Ping timeout: 245 seconds]
Avagetto has quit [Ping timeout: 264 seconds]
Avagetto has joined #linux-rockchip
Avagetto1 has quit [Ping timeout: 265 seconds]
Avagetto1 has joined #linux-rockchip
Avagetto has quit [Ping timeout: 264 seconds]
Avagetto has joined #linux-rockchip
Avagetto1 has quit [Ping timeout: 255 seconds]
Avagetto1 has joined #linux-rockchip
Avagetto has quit [Ping timeout: 255 seconds]
Avagetto has joined #linux-rockchip
bengal has joined #linux-rockchip
Avagetto1 has quit [Ping timeout: 240 seconds]
Avagetto1 has joined #linux-rockchip
Avagetto has quit [Ping timeout: 255 seconds]
Avagetto has joined #linux-rockchip
Avagetto1 has quit [Ping timeout: 244 seconds]
Avagetto1 has joined #linux-rockchip
Avagetto has quit [Ping timeout: 264 seconds]
Avagetto1 has quit [Ping timeout: 264 seconds]
Avagetto has joined #linux-rockchip
Avagetto1 has joined #linux-rockchip
Avagetto has quit [Ping timeout: 244 seconds]
Avagetto has joined #linux-rockchip
Avagetto1 has quit [Ping timeout: 245 seconds]
Avagetto1 has joined #linux-rockchip
Avagetto has quit [Ping timeout: 244 seconds]
Avagetto has joined #linux-rockchip
Avagetto1 has quit [Ping timeout: 272 seconds]
Avagetto1 has joined #linux-rockchip
Avagetto has quit [Ping timeout: 244 seconds]
cosm has joined #linux-rockchip
Avagetto has joined #linux-rockchip
Avagetto2 has joined #linux-rockchip
Avagetto1 has quit [Ping timeout: 244 seconds]
Avagetto has quit [Ping timeout: 244 seconds]
<mmind00> jas-hacks: try https://git.kernel.org/cgit/linux/kernel/git/mmind/linux-rockchip.git/commit/?h=v3.19-armsoc/dts-fixes&id=6618e478954d2866754798c3b07805dc9dbb2fc4
Avagetto has joined #linux-rockchip
Avagetto2 has quit [Ping timeout: 264 seconds]
<jas-hacks> mmind00: I tried that already it makes the card detectable by the driver. however I don't think the card is setup correctly because of
Avagetto has quit [Ping timeout: 256 seconds]
<mmind00> jas-hacks: you mean the "mount: mounting /dev/mmcblk0p1 on /rfs failed: Invalid argument"? Does the kernel you use has the necessary filesystem compiled in?
Avagetto has joined #linux-rockchip
<jas-hacks> mmind00: i checked that one
Avagetto1 has joined #linux-rockchip
Avagetto has quit [Ping timeout: 244 seconds]
Avagetto1 has quit [Ping timeout: 245 seconds]
cosm has quit [Ping timeout: 244 seconds]
Avagetto has joined #linux-rockchip
Avagetto1 has joined #linux-rockchip
Avagetto has quit [Ping timeout: 240 seconds]
levd1 has quit [Remote host closed the connection]
levd has joined #linux-rockchip
<naobsd> jmcneill: dwc mmc card detect bit is 0: present 1: not present
<naobsd> jmcneill: and grf/gpio should be configured properly
Avagetto has joined #linux-rockchip
Avagetto1 has quit [Ping timeout: 240 seconds]
cosm has joined #linux-rockchip
Avagetto1 has joined #linux-rockchip
Avagetto has quit [Ping timeout: 250 seconds]
Avagetto2 has joined #linux-rockchip
Avagetto1 has quit [Ping timeout: 245 seconds]
Avagetto has joined #linux-rockchip
Avagetto2 has quit [Ping timeout: 272 seconds]
McBitter has joined #linux-rockchip
Avagetto1 has joined #linux-rockchip
Avagetto has quit [Ping timeout: 258 seconds]
<jas-hacks> The dw_mmc driver (rk_sdmmc.c) is heavily patched up in Android, I'm guessing the extension driver dw_mmc-rockchip.c does have the same work arounds.
cosm has quit [Ping timeout: 265 seconds]
Avagetto1 has quit [Ping timeout: 240 seconds]
Avagetto has joined #linux-rockchip
Avagetto1 has joined #linux-rockchip
Avagetto has quit [Ping timeout: 244 seconds]
Avagetto1 has quit [Ping timeout: 258 seconds]
<naobsd> jas-hacks: do you know how much work was done in mainline for rk3288?
<naobsd> jmcneill: sdmmc works on px2 board too
levd has quit [Ping timeout: 245 seconds]
<jas-hacks> naobsd: sdmmc works on px2, what class card did you test?
<jmcneill> naobsd, it does? thats great!
<naobsd> jas-hacks: do you understand I'm talking about NetBSD?
<naobsd> jmcneill: thank you for your great work :)
<jmcneill> always good to hear my code works for someone other than me :p
<naobsd> :)
<naobsd> I need to take dinner, later
<naobsd> jmcneill: return !!(cdetect & DWC_MMC_CDETECT_CARD_DETECT_N); <- remove one "!"
<jmcneill> did i misread datasheet?
dlezcano has joined #linux-rockchip
<naobsd> jmcneill: I checked linux mainline, 0 is present
<naobsd> jmcneill: you may need
<naobsd> jmcneill: obio_iomux(RK3188_GRF_GPIO3A_IOMUX_OFFSET, 0x55555554);
bengal_ has joined #linux-rockchip
<naobsd> jmcneill: obio_iomux(RK3188_GRF_GPIO3B_IOMUX_OFFSET, 0x00050001);
<jmcneill> email patches please
<naobsd> yes, later
bengal has quit [Ping timeout: 245 seconds]
<ganbold_> naobsd: rk3188 uart has reg-io-width = <1>; but rk3288 has reg-io-width = <4>;
<ganbold_> maybe little bit different
_bengal_ has joined #linux-rockchip
bengal_ has quit [Ping timeout: 245 seconds]
<naobsd> jmcneill: do you know which gpio pins are used for otg/host vbus on X7?
<McBitter> naobsd: anyway - do you have file structure that maskrom expects before starts to execute?
<jmcneill> no idea, can i get that from linux dmesg?
<jmcneill> or android i guess it came with
<naobsd> jmcneill: I think I can get from sdk for your board, but it's ok, I'll check later
<naobsd> McBitter: I don't know for now
<naobsd> McBitter: why do you want to modify DDR init blob?
<naobsd> hmm, I cannot see uhub2 on px2 board...
<McBitter> naobsd: to get uboot init dram on it's own?
<McBitter> rather than done by that procedure
<naobsd> McBitter: I see.
<McBitter> anyway - it seems that you put rkcrc16 value at the end in big endian form
GriefNorth has quit [Quit: WeeChat 1.0.1]
<jmcneill> naobsd: doubt there's an sdk for it, but usb does already work
<jmcneill> not the type of product that comes with an sdk :p
<naobsd> jmcneill: I think I wrote uri here...
<jmcneill> ah yes you did
<jmcneill> anyway, usb works already :)
<naobsd> maybe: until grf/gpio code is cleaned
<naobsd> mmc and usb otg/host power pin is one of function of pin mux
<naobsd> on linux these pins are configured as gpio and it's controlled as gpio
<jmcneill> how to detect specific boards?
<naobsd> I'm not sure but dwc controller unit can controll these pins if it's configured as "power control function pin" not gpio
<jmcneill> the way it is handled for awin ports works well, but the board type is generally compiled in
<naobsd> no good idea to specify board type
<naobsd> bootargs or dts
<jmcneill> downloading now, thx
<jmcneill> slow though
<jmcneill> i think i started before and realized it was going to take an hour and a half and gave up
<karlp> is that kernel source for th x7?!
<karlp> I've asked them before for the kernel source for the x5mini and just got answers of "why do you need it?"
<jmcneill> change x7.tar.gz to x5.tar.gz :)
<jmcneill> (there's something there)
<karlp> that's not the same as x5 mini however
<karlp> but yes, downloading now :)
<karlp> 3gig?1 yuk.
<naobsd> I'll downloading, I'll put it somewhere
<naobsd> jmcneill: btw, if my reading is correct, rk3188+ need to do same as rk3066 for set pll rate
<naobsd> naobsd: but your current code is for 3188
<naobsd> ah
<naobsd> jmcneill: but your current code is for 3188
<jmcneill> yeah not sure how it works :p
<jmcneill> but openssl speed doesn't lie
<jmcneill> does the current code work on your rk3188?
<naobsd> I don't have userland yet...
<naobsd> jmcneill: 3188/plus major difference is reg bit width, offset is same, I think current value is not so large
<naobsd> or maybe plus is same as 3188, and linux code is wrong ;)
<jmcneill> ah
<naobsd> I found px2 trm error... there are 2 mmc write-protect pin, and no mmc power pin ;)
<jmcneill> btw x7 has ite hdmi tx
<naobsd> jmcneill: as far as I know, there is no ite66121 doc
<naobsd> only ultimately dirty code in RK linux kernel
<jmcneill> maybe i will look at vmac next then
<naobsd> it may need regulator driver
<jmcneill> maybe i will look at awin a80 next then :)
<naobsd> and phy power or reset pin
<naobsd> hehe :)
<naobsd> what's pmic on x7? act8846?
<jmcneill> can't tell
<jmcneill> lets see what android says
<jmcneill> i can't believe how long it takes to boot
<jmcneill> arg that's not right
<jmcneill> start at line 122
<naobsd> act8846
<jmcneill> next you'll tell me there's no docs for that either
<jmcneill> ty
<naobsd> sorry!
<jmcneill> i guess i'll do i2c next then
<naobsd> i2c is same on both rk3066/rk3188 :)
<jmcneill> lucky you
<naobsd> no problem, pmic is different!
hipboi has quit [Read error: Connection timed out]
<naobsd> act8846 is common on rk3188/rk3288
<jmcneill> oh i should try your new uboot
<naobsd> for rk3066 tps(forgot number) or axp2xx
<naobsd> axp2xx on rk3066 is very minor I think
<naobsd> jmcneill: it should say RK3188+ ;)
<naobsd> (I'm away for a while)
<jmcneill> ChipVer: 310B20130131V101
<jmcneill> CPU: RK3188+
<jmcneill> looks good
<jmcneill> and my cpu came up at 1.6GHz so netbsd was happy with it
<jmcneill> hm, sdmmc doesn't work anymore
BorgCuba has joined #linux-rockchip
<BorgCuba> jmcneill, for me sdmmc works
<BorgCuba> with 3.19-rc1
<jmcneill> ah that is because i applied your cdetect change
<naobsd> jmcneill: did you apply grf lines too?
<jmcneill> did you send me a patch?
<naobsd> yes
<jmcneill> ah you should have sent to the mail at the top of the file
<jmcneill> rarely check the other one:)
<naobsd> ah, I see
<naobsd> anyway, gpio3b0, sdmmc0 detect pin is mis-configured as gpio
<BorgCuba> btw, has anybody checked the dwc2 otg gadget patches (29) or has a working configuration for it? I think rperier also had problems with it.
hipboi has joined #linux-rockchip
<jmcneill> I wonder if that's why PWREN logic is inverted
<jmcneill> with your full patch sdmmc works again
<naobsd> I'm sure mmc power pin on RR is active low, but I'm not sure it's dwc mmc requirement or just board configuration
<jmcneill> uboot sources have a comment to use 0 for power on mmc0 and 1 for power on emmc
<jmcneill> i couldn't find the same logic in the linux driver
<jmcneill> looked to me like it was 1 all the time there
<naobsd> linux mainline _mmc_ driver doesn't controll power pin, it's configured as gpio and described in board specific dts
<naobsd> mmc power pin is one of pinmux "function", but it's possible to use another pin. later case, it must be handled as gpio controlled power pin
<naobsd> but generally another pin shouldn't be used, "boot from SD" is impossible
<jmcneill> I applied your patch, thanks!
<naobsd> I'm trying pll things for rk3066...
<naobsd> (I'm away for a while again...)
McBitter has quit [Quit: Leaving]
<karlp> naobsd: where did you get that minix link? I can't find it linked in the forums at all, I was going to refer to it to ask again for x5mini source
<jmcneill> i searched "minix neo x7 sdk" on google and it showed up in the second result
<karlp> hrm, the facebook post mentions it's for x7/x7mini
<karlp> maybe the x5 is also for the mini too
<karlp> unfortunately the link went to the forum that they shut down and deleted the posts from :|
<naobsd> yeah, google search told me that ;)
<naobsd> I downloaded both x5 and x7 tarball
<naobsd> if you need I can put it somewhere
<karlp> well, x5 is nearly here,
<karlp> at 2.9/3.1gig
<naobsd> I see
<karlp> would be nice to get a self built android kernel to build and boot
<karlp> I hadn't found any existing source that booted on this x5mini,
<karlp> (of the 3.0+ android sources)
<naobsd> I remembered why uhub2 is not attached on px2
<naobsd> it needs DC plugged...
<naobsd> I forgot it every time...
<naobsd> and I rememred
<naobsd> remembered that
<naobsd> rockchip specific dwctwo parameters should be applied only for otg, for host, generic one should be used...
foo_ has joined #linux-rockchip
foo_ has quit [Ping timeout: 246 seconds]
<karlp> huh, that download is actually a git repo
<karlp> hrm, I guess I need to clone it again locally to make it "unbare"
<jmcneill> naobsd: can you find me i2c clock info?
<naobsd> jmcneill: it should be same as px2, pclk_cpu for i2c[01], pclk_periph for i2c[234]
<jmcneill> datasheet says "I2C controller uses the APB clock as the system clock"
libv has quit [Ping timeout: 250 seconds]
<naobsd> apll -> aclk_cpu -> pclk_cpu
<naobsd> sorry, pll/clk thing is still difficult for me :(
Avagetto has joined #linux-rockchip
<naobsd> gpll/cpll -> aclk_periph -> pclk_periph
<naobsd> I think i2c clock is same as px2 trm
<naobsd> mmm, I forgot what I was trying...
<naobsd> hmm, I want "reboot bootloader"...
<naobsd> let hardcode it ;)
<jmcneill> i2c driver done except for the src clk part
nighty^ has joined #linux-rockchip
<naobsd> jmcneill: where it comes from? CRU_CLKSEL_CON11_MMC0_PLL_SEL __BIT(6)
hipboi has quit [Read error: Connection timed out]
hipboi has joined #linux-rockchip
libv has joined #linux-rockchip
<jmcneill> I forget where I saw that. I think RK3188 specific.
cnxsoft1 has quit [Remote host closed the connection]
Avagetto has quit [Quit: Leaving.]
Avagetto has joined #linux-rockchip
Luke-Jr has quit [Quit: Konversation terminated!]
Luke-Jr has joined #linux-rockchip
<naobsd> hmmm...
<naobsd> no idea why emmc cannot be detected...
Avagetto has quit [Quit: Leaving.]
plumser has joined #linux-rockchip
plumser has quit [Client Quit]
<ganbold_> I guess enough for today, was able to use early printf in initarm()
nighty^ has quit [Ping timeout: 244 seconds]
<jmcneill> naobsd: different clock, inverted PWREN logic?
nighty^ has joined #linux-rockchip
<jmcneill> ganbold, that's great!
FreezingCold has quit [Quit: Out]
ganbold_ has quit [Remote host closed the connection]
need_a_ko has joined #linux-rockchip
<need_a_ko> hi
<need_a_ko> Is there a tun.ko for rk3026? or at least any kernel source?
ganbold_ has joined #linux-rockchip
FreezingCold has joined #linux-rockchip
need_a_ko has quit [Quit: Page closed]
jas-hacks1 has joined #linux-rockchip
jas-hacks has quit [Ping timeout: 244 seconds]
jas-hacks1 has left #linux-rockchip [#linux-rockchip]
ganbold_ has quit [Remote host closed the connection]
ganbold_ has joined #linux-rockchip
<naobsd> sleepy
nighty^ has quit [Ping timeout: 244 seconds]
Luke-Jr has quit [Excess Flood]
Luke-Jr has joined #linux-rockchip
luke-jr_ has joined #linux-rockchip
Luke-Jr has quit [Read error: Connection reset by peer]
<naobsd> aaaaah
<naobsd> I broke px2 board...
<naobsd> I scratched board hardly while removing HDD...
luke-jr_ is now known as Luke-Jr
<jmcneill> now you can focus on rk3188 :)
<naobsd> it still boot...
<naobsd> small 3pin chip is floating ...
<naobsd> 8050 J3Y SOT-23 ?
<naobsd> probably some power line is dead :(
<jmcneill> how to tell if this board has emmc or nand?
<naobsd> configure grf & probe for both emmc and nand?
<jmcneill> heh
<jmcneill> well linux dmesg looks like nand, bummer
<naobsd> ah, rk3066/rk3188 with eMMC is minor
<jmcneill> rockchip_mmc_set_div(self->dv_unit, 1);
<jmcneill> few comments
<jmcneill> first, should never access members of device_t directly
<naobsd> eMMC is more expensive/fast than NAND, cheap product uses NAND
<jmcneill> device_unit(self) is preferred mechanism
<jmcneill> second, i added a "port" locator that you should use for that (using it with i2c)
<jmcneill> dwcmmc0 at obio0 ... port 0
<jmcneill> dwcmmc1 at obio0 ... port 1
<jmcneill> etc
<naobsd> I see
<jmcneill> i'll probably move it over to config similar to awin_io.c at some point
<jmcneill> but lazy
<jmcneill> probably spend the next couple days trying to figure out why i2c driver doesn't work
<naobsd> I see
<jmcneill> or at least, i can't talk to the hdmi tx where i think it is
<jmcneill> rk3188 cru trm sure would be helpful
<naobsd> sleepy
<naobsd> good night!
<jmcneill> cya
<Astralix1> Good evening
dlezcano has quit [Ping timeout: 272 seconds]
FreezingCold has quit [Ping timeout: 256 seconds]
<bbelos> hello Astralix1
<Astralix1> Hey, bbelos
<bbelos> my kernel seems to be in pretty decent shape if i can get the display working
<Astralix1> That is one important step forward.
<Astralix1> My kernel does set up the LCD correctly too, but you can hear that the DC/DC for some core functionality isn't working right
<Astralix1> It issues a high frequency beep that changes frequency under changing system load
<bbelos> that's pretty weird
<Astralix1> And depending on the test setting, SOC and DRAM is getting pretty warm
<Astralix1> But as I told, the sources and the kernel do not match. But the german vendor tried several times to get the correct sources from the chinese OEM... no chance
<Astralix1> Got two different packages, both not matching
<bbelos> comparing my dmesg to stock shows a difference in dclk on lcdc1. trying to figure that one out and then have some issues with gpu clk
<bbelos> also need to find an axp_power driver
<Astralix1> I am currently replicating the whole build process of firefly android
field^Mop has joined #linux-rockchip
<Astralix1> hmm... my rkflashtools do not get parameters from rk3288 booting via uboot...
<Astralix1> How can I backup parameters from uboot started rk3288?
_bengal_ has quit [Quit: Leaving]
zombu2 has quit [Ping timeout: 244 seconds]
zombu2 has joined #linux-rockchip
zombu2 has quit [Ping timeout: 258 seconds]
zombu2 has joined #linux-rockchip
<jmcneill> naobsd: smp fixed but it doesnt get along well with cpu freq changing