mmarker2 has quit [Remote host closed the connection]
mmarker has joined #linux-sunxi
amitk has joined #linux-sunxi
kaspter has joined #linux-sunxi
lkcl has quit [Ping timeout: 245 seconds]
Lorenzo64 has joined #linux-sunxi
codekipper has joined #linux-sunxi
<codekipper>
wens: morning!
<codekipper>
did you ever do a sdk build for your merrii a31?
<codekipper>
or did you go straight into mainlining?
NegativeForty has quit [Ping timeout: 240 seconds]
NegativeForty has joined #linux-sunxi
xeros has quit [Ping timeout: 260 seconds]
lkcl has joined #linux-sunxi
NegativeForty is now known as FreezingCold
PulkoMandy has joined #linux-sunxi
<wens>
codekipper: straight to mainlining
Nyuutwo has quit [Ping timeout: 246 seconds]
andoma_ has quit [Ping timeout: 246 seconds]
<codekipper>
\m/
<codekipper>
i thought so....I've downloaded the sdk and will have a look at building the kernel
andoma has joined #linux-sunxi
<codekipper>
is it a pain with the fel switch and booting from sdcard?
Nyuutwo has joined #linux-sunxi
PaulFertser has joined #linux-sunxi
<wens>
the onboard boot1 or u-boot checks for sdcard boot signature
Black_Horseman has quit [Remote host closed the connection]
<quitte>
actially that's eGON
<PaulFertser>
Hi. I'm pondering buying an ARM board to use as my home server (mcabber/irssi/mutt/emacs with an external SATA driver). But I'm somewhat reluctant to get anything-allwinner because of their blatant GPL violations. What are my options anyway?
<PaulFertser>
s/driver/drive/
<wens>
if it finds one, it will chain load
<quitte>
oh
quitte has quit [Read error: Connection reset by peer]
<codekipper>
ok...I'll get the board properly moded and a card prepared. Did you use the allwinner-sunxi-a31 u-boot at http://git.rhombus-tech.net
<wens>
mripard_ has an sdcard image somewhere, which has boot0/boot1, and fat partition for u-boot
<codekipper>
I've seen that but haven't tried...thanks
quitte has joined #linux-sunxi
Skaag is now known as Skaag
<oliv3r>
libv: i have no objections if it works :p
<oliv3r>
libv: except to name it sunxi-meminfo :p
lkcl has quit [Ping timeout: 240 seconds]
<oliv3r>
i can't seem to find the a31 user manual on dl.linux-sunxi/A31 just the datasheet and lots of chinese documents
<oliv3r>
libv: btw, we do have sun6i (A31) u-boot memory support
<oliv3r>
just nobody worked on it/with it implemented it in our u-boot tested it etc
<oliv3r>
i think hans had a protoge that would wanna work on it
<oliv3r>
but nothing ccame of it that i know
bertrik has joined #linux-sunxi
JohnDoe_71Rus has quit [Ping timeout: 240 seconds]
JohnDoe_71Rus has joined #linux-sunxi
wingrime has joined #linux-sunxi
_massi has joined #linux-sunxi
sehraf has joined #linux-sunxi
diego_r has joined #linux-sunxi
Nyuutwo has quit [Ping timeout: 250 seconds]
<rellla>
ssvb: so if i understand correctly, one has to enable scaler mode in fex to use scaled layers at all, or can i access and use a scaler during runtime when scaler mode = 0?
<rellla>
ssvb: or asked the other way round: can these 2 scalers (on A10/A20) be requested during runtime when scaler_mode is on in fex? sorry, but i'm a bit confused...
<ssvb>
rellla: yes, the scalers can be enabled/disabled at runtime for any layer
<ssvb>
rellla: the fex settings are just configuring the initial state
<rellla>
ssvb: so the better way is to disable it, to make sure, that there is at least 1 scaler available. and ..
<wens>
a80 optimus comes with emmc flash
<rellla>
because in the image0_pixel_sequence errors the value given back seems to be a random value, may it help to define all the neccessary values fb.seq, fb.format .... before doing the SET_PARAM ioctl? maybe this random "value" is the result of a missing definition.
<ssvb>
rellla: you can't be completely sure, if all the scalers are used up, then trying to enable scaler for your layer just fails
<ssvb>
rellla: as for the image0_pixel_sequence errors, there is no other way but to debug this issue :)
bertrik has quit [Ping timeout: 240 seconds]
lkcl has joined #linux-sunxi
<mripard_>
codekipper: the sdcard image is meant for the A31 evk, use at your own peril.
<mripard_>
wens: yeah, I have to work on a v2 for these clocks (:
<wens>
mripard_: what parts are you changing mostly?
<wens>
i need the part that splits out factors clk
<wingrime>
wens: than, witch boot order a80 have?
<wens>
wingrime: not clear actually, haven't tested putting an sd card in for it to read
<mripard_>
wens: not much actually, only a few typos here and there
<mripard_>
you can base your work on the v1 I suppose
<wens>
that's pretty much what i'm doing
lkcl has quit [Ping timeout: 260 seconds]
<codekipper>
mripard_: shudder....I saw that it had a mainline kernel on it. I was going to replace the kernel with latest + add sun6i-a31-m9.dtb
<codekipper>
I won't attempt anything until I see boot0 coming out of the uart
<wens>
codekipper: it worked on the hummingbird :)
<codekipper>
I'm pretty confident that it should work
<wens>
wingrime: looks like bootrom goes straight to mmc2
<wens>
maybe it's the same as a31, some external pins choose which device to boot from
<wens>
stock u-boot checks mmc0 for valid signature
<wens>
i wonder where the fex file is hidden :(
alexvf has joined #linux-sunxi
PulkoMandy has quit [Quit: Vision[0.9.7-H-20140108]: i've been blurred!]
PulkoMandy has joined #linux-sunxi
Quarx has joined #linux-sunxi
shineworld has joined #linux-sunxi
PlkMndy has joined #linux-sunxi
PulkoMandy has quit [Read error: Connection reset by peer]
deasy has quit [Remote host closed the connection]
notmart has joined #linux-sunxi
tm512 has quit [Quit: Just relax and enjoy this pleasant adventure]
<deepe>
Hello I am using custom display, I set up everithing, but look like there is problem with Enable signal on A10-olinuxino-Lime board on LCD connector
<deepe>
it is inverted, should be 1 it is 0
ecelis has quit [Ping timeout: 240 seconds]
<deepe>
I can not find out settings to "invert " signal in fex file
<deepe>
Can anybody help me with way, where to look in source code, where are this settings
<deepe>
and I would like to add feature like this, for future use for custom displays
Master_Crumble has quit [Ping timeout: 246 seconds]
<wens>
hmm, mali 400/450 ddk team moved to shanghai
<mnemoc>
thinking about moving to .cn? :p
<wingrime>
wens:from where?
<mnemoc>
wens: btw, how do you call "mainline china"?
<mripard_>
codekipper: I'm not afraid of the kernel, but rather the PMIC?
<mripard_>
s/PMIC?/PMIC and RAM setup/
<mripard_>
the kernel has a DT for it btw
lkcl has quit [Ping timeout: 246 seconds]
jemk has joined #linux-sunxi
<deepe>
codekipper: thanks for hint, going to check it
<wens>
wingrime: no idea, just saw a tweet from a friend at mediatek
<wens>
mnemoc: the mainland, or just china
<wens>
about the pmic, the axp regulators still aren't fully merged :|
<mnemoc>
wens: so it's an outdated myth that you call the formosa/taiwan island "china"?
<wens>
mnemoc: the island is taiwan/formosa
<mnemoc>
and how do you call the country?
<wens>
the governing body is the "republic of china"
<wens>
but most just use taiwan
<mnemoc>
i see
<wens>
using "china" tends to get confused with the PRC, and attracts some hostility
<wens>
chinese tourists are infamous around the world it seems
<mnemoc>
people's vs. not-people's ... odd
<wens>
hehe
<mnemoc>
china vs. taiwan is much simpler :)
<wens>
well communist countries seem to like people's
<wens>
mnemoc: informally that is how everyone refers to the two countries
<wens>
(well maybe not the hardcore people in the PRC)
<mnemoc>
they probably insist the island is still under their control :p
<wens>
but since our constitution says ROC, that is what official documents use
<wens>
mnemoc: nah, deep inside they know it's not under their control, they just insist it's part of china, and should be reunited
<mnemoc>
thanks, everything is more clear now. i always wondered how normal-people dealt with the problem when not under the influence of political idealisms
popolon has joined #linux-sunxi
<wens>
mnemoc: normal people are sane and respect the difference i guess, or just don't talk about it
afaerber has joined #linux-sunxi
<wens>
i guess some places in europe may have gone through stuff like this?
<wens>
albeit maybe a couple hundred years ago
joost_dt1 has quit [Remote host closed the connection]
<wens>
hah, but that's way off topic
<mnemoc>
europe has moved it's internal borders far more often... and the cultural differences are stronger I think. the middle kingdom has always given the idea of a single... empire
<mnemoc>
even while most europe was hidding in trees and caves
<mnemoc>
wens: offtopic indeed :p
joost_dtn has joined #linux-sunxi
<deepe>
sdfjhsdjkljklj;kl
* mnemoc
nods
<codekipper>
mripard_: does the uboot do anything specific?, the m9 dts is pretty minimal. I'll compare fex's to see how they're set up
<wens>
codekipper: you can replace the uboot binary with uboot-sunxi
balog has joined #linux-sunxi
<skoperst>
wens, did you get uart cable with the optimus board?
<codekipper>
wens: thanks....just use the Colombus config
<skoperst>
thats interesting.. cause libnand is a piece of ****
<wens>
yeah, mines emmc: Samsung KLMAG2WEMB-B031
<wens>
guess i got the better deal
<skoperst>
jesus.. :/
<skoperst>
How much did you pay for it?
kuldeepdhaka has joined #linux-sunxi
<wens>
skoperst: around 169 usd
<wens>
after they halved the price
<bbrezill1>
quitte: I don't see anything related to UBI fastmap in your logs, don't you get the same kind of errors without fastmap ?
<quitte>
no. those are only with fastmap
<bbrezill1>
quitte: and sometime it works ?
<quitte>
yes.
<quitte>
I think the works/doesn't work is determined at ubiformat time.
<quitte>
or on initial autoconversion, more likely
shineworld has joined #linux-sunxi
shineworld has joined #linux-sunxi
<bbrezill1>
so you've passe fw_autoconvert to the kernel cmdline (or module param)
<bbrezill1>
right ?
<bbrezill1>
quitte: ^
<quitte>
line 41 in the paste
<quitte>
ubi.fm_autoconvert=1
<bbrezill1>
:-)
Quarx has joined #linux-sunxi
<bbrezill1>
quitte: you mean that when it works one time (I mean attaching the mtd part using fastmap, not creating the fastmap metadata) it keeps working on the following attach
<quitte>
that theory just went bust. i did a reboot and it worked.
joost_dtn has quit [Quit: leaving]
paulk-collins has joined #linux-sunxi
<bbrezill1>
quitte: if this is the case, then yes, I suspect the corruption comes during fastmap metadata creation
joost_dtn has joined #linux-sunxi
<bbrezill1>
quitte: still experimenting on a cubietruck ?
<quitte>
yes
<bbrezill1>
quitte: no read retry problem then
<quitte>
are there rr problems on others?
<bbrezill1>
yep
<bbrezill1>
I mean on other NAND chips
<quitte>
i did about 5 reboots just now. it worked every time. maybe ubi has moved the problem into better erase blocks?
<bbrezill1>
quitte: this patch has a drawback => it prevents UBI from detection pages where bitflips are almost exceding the ECC strength
<bbrezill1>
quitte: meaning that UBI won't be able to copy its data to another page :-(
Black_Horseman has joined #linux-sunxi
Andy-D has joined #linux-sunxi
leviathanch has joined #linux-sunxi
sehraf has joined #linux-sunxi
Tenebrous has joined #linux-sunxi
<bbrezill1>
quitte: we have to rework the read_retry to test for all available internal bit levels and choose the bit level that generate the less bitflips, instead of sticking on the first one that can be corrected by ECC
<quitte>
for every read?
<bbrezill1>
quitte: hopefully this will prevent from raising the ECC threshold on a lot of pages, and will let UBI do its "copy almost unreliable pages to another place"
<bbrezill1>
quitte: not for every read => for those who generate ECC errors (uncorrectable bitflips) or that exceed the bitflip threshold
polto has joined #linux-sunxi
PlkMndy has quit [Read error: No route to host]
PulkoMandy has joined #linux-sunxi
polto has quit [Client Quit]
<bbrezill1>
quitte: Actually, what you've seen with your fastmap problem might be related to my threshold patch
<quitte>
I can't take it out. then ubiformat fails way too often, while with the patch it hasn't
<quitte>
in as many formats as the mean erase times
<ddc>
bbrezill1: Do you have any materials on the HW randomizer implementation. I'm having some troubles understanding the reason behind it.
<quitte>
ddc: it's kind of like similar to a cd where not too many pits and lands may follow each other or they will be indistinguishable, is how i got the gist of it. of course that's technically incorrect.
<bbrezill1>
quitte: ubiformat shouldn't fail because of this patch, but you might end moving data from one block to another each time you're booting your platform
<ddc>
bbrezill1; is just a security feature ( scrambling data)
<quitte>
bbrezill1: ubiformat writing a working ubi filesystem fails without the patch
<bbrezill1>
quitte: that shouldn't happen :-(
<bbrezill1>
quitte: could you paste the errors ?
<bbrezill1>
ddc: AFAIU this has nothing to do with a security feature
<ddc>
bbrezill1: so it is used to avoid large island of identical data
<bbrezill1>
ddc: yep
<quitte>
bbrezill1: that's the reason you gave me the patch in the first place. to clarify: now, with the patch applied, ubiformat works every time.
<bbrezill1>
quitte: yes, I think I gave you this for testing purpose in the first place
<bbrezill1>
quitte: but this definitely not the appropriate fix
<quitte>
okay. I'll remove the patch and rebuild the kernel then?
kaspter has quit [Ping timeout: 255 seconds]
shineworld has quit [Quit: Leaving]
<ddc>
ibbrezill1:s it possible to avoid specifying a large seeds array in DT by using other means of generation the seeds?
<ddc>
bbrezill1: s/generation/generating
<quitte>
what is the difference between reading mtd nand and mlc-nand? I did this: http://pastebin.com/KZ8hZJVc to u-boots env tools. but no matter how i configure it, it will always fail due to CRC errors. even if I write the environment from an image from within linux
<quitte>
this env image will then work as the environment for u-boot just fine
<quitte>
bbrezill1: the above clearly is not something I expect you to look into. but maybe you can shed some light on the situation?
<quitte>
for the flag_scheme i simply guessed. i have no idea what that means
yann_s has quit [Ping timeout: 260 seconds]
<ddc>
quitte: Does yuq driver use HW randomization?
yann_s|AFK has joined #linux-sunxi
yann_s|AFK is now known as yann_s
Quarx has quit [Remote host closed the connection]
<ddc>
quitte: u-boot nand driver
<quitte>
ddc: only for the spl. I haven't looked if it can be configured for the rest. it probably could
<quitte>
after all the hw-rand and ecc code is there
Quarx has joined #linux-sunxi
<quitte>
at least for the nad read.1k and write.1k commands
<quitte>
s/nad/nand
<quitte>
ddc: if you have a cubietruck - openwrt is getting to a state where I'd love to hear feedback - before adding support for other boards
<mripard_>
longsleep: DT overlays have been merged last night.
<wens>
mripard_: muxing ahb1 to pll6 makes my a23 hang :(
<ddc>
quitte: I've ported most of bbrezillion kernel driver to u-boot . I'm wondering if you would like to host in your repos
<ddc>
quitte: host it
<mripard_>
wens: damn
<quitte>
ddc: great :) but no. I'd prefer if you hosted it. I'm not in it for the long-term or even the medium term, as i explained a couple of days ago
<mripard_>
is it at least a viable mux for ahb1?
<quitte>
ddc: I'd like to use it however.
avsm has quit [Quit: Leaving.]
<petrosagg>
bbrezill1: hey, I was wondering if you had a chance to look into my pull request for the cubieboard2 nand stuff
<quitte>
petrosagg: send one to me,too.
<quitte>
or even better a patchset, if that's not a lot of trouble
<ddc>
quitte: I want to get some feedback from bbrezillion first it may not be ready till the weekend. I will send u an email by then
<libv>
oliv3r: ah, cool
<petrosagg>
quitte: I'm still working on the u-boot part, I'll send a patchset to your email. Haven't done it before so more things to learn there :)
<libv>
ah, yes, rename might not be too bad
<quitte>
ddc: if you tell me to not put it out there that's fine. please send me what you have.
<quitte>
petrosagg: u-boot is actually better than i thought. just don't use it to write to nand.
<petrosagg>
quitte: the u-boot patch was for the sunxi-spl.bin generation for the Cubieboard2
<petrosagg>
quitte: you meant the kernel patches? I don't see a linux-sunxi repo in your github account
<quitte>
that's why i was hoping for a patchset. the kernel lives here:https://github.com/quitte/openwrt/tree/sunxi-mtd/target/linux/sunxi/patches-3.16
<quitte>
petrosagg: it's mostly nand partitioning in the dts, i assume?
wingrime has quit [Read error: Connection reset by peer]
<petrosagg>
quitte: yes, you're correct
wingrime has joined #linux-sunxi
<petrosagg>
quitte: I can create those patchsets, will send today
<quitte>
great. I'll ask you to give openwrt a try once I feel there is a chance it works on the CB2
<petrosagg>
quitte: sure, I'll be around :)
wingrime has quit [Read error: Connection reset by peer]
<quitte>
I just spent a couple of seconds trying to figure out how everybody copes with the html formatting of emailed patches. then it hit me. it has been a about a decade since i used a proper email client
<petrosagg>
quitte: done
<petrosagg>
was my email html formatted? :S I used git send-mail for the first attempt
<quitte>
petrosagg: gmail web client
apo_ has quit [Remote host closed the connection]
apo_ has joined #linux-sunxi
<quitte>
petrosagg: what about the spl one? it's just adding nand to the defconfig?
<wens>
mripard_: it is, but maybe the clock rate is too high without setting the predivider
<petrosagg>
quitte: I haven't looked into it, should be a simple change
Nyuutwo has joined #linux-sunxi
kuldeepdhaka has quit [Ping timeout: 255 seconds]
shineworld has joined #linux-sunxi
Nyuutwo has quit [Ping timeout: 246 seconds]
kaspter has joined #linux-sunxi
kaspter has quit [Remote host closed the connection]
Black_Horseman has quit [Remote host closed the connection]
<mripard_>
wens: and what happens if you remove the reparenting?
<libv>
skoperst: btw, can you ndh the pcduino?
<wens>
mripard_: system loads properly
<bbrezill1>
petrosagg: I saw your pull request, and I'll take it soon ;-)
<wens>
mripard_: does dmatest work?
<quitte>
bbrezill1: 3rd ubiformat in a row without the threshold patch and no problem whatsoever. odd. I could erase the flash and see if that breaks it again
<quitte>
well - except for fixable bitflips detected
<bbrezill1>
quitte: yes, but that's expected
<bbrezill1>
quitte: still using the fastmap option ?
<quitte>
yes
<mripard_>
wens: you tell me :)
<mripard_>
it does on an A31, last time i checked
<skoperst>
libv: sure, I'll give it a try
Nyuutwo has joined #linux-sunxi
PulkoMandy has quit [Ping timeout: 246 seconds]
avsm has joined #linux-sunxi
lkcl has quit [Ping timeout: 264 seconds]
<libv>
skoperst: thanks :)
<libv>
wens: does your optimus really read 672MHz from the registers?
<libv>
or is that halved as well?
<libv>
not that the ccmu registers have to be the same for a80, but there is a good chance that they are
JohnDoe_71Rus has quit [Quit: Miranda IM! Smaller, Faster, Easier. http://miranda-im.org]
<petrosagg>
bbrezill1: from what I understand mtd->_read() will return the number of bits corrected
<petrosagg>
bbrezillon1: but that number can't be higher than the ECC strength, since by definition this is the maximum amount of bitflips that can be corrected
<petrosagg>
bbrezill1: so by doing strength + 1 isn't essentially making all mtd_read() calls never return -EUCLEAN?
<bbrezill1>
petrosagg: this was patch intended for test
dack has joined #linux-sunxi
nove has joined #linux-sunxi
<bbrezill1>
petrosagg: as I explained to quitte earlier, it should not be part of a stable version of the sunxi NAND driver
<petrosagg>
bbrezill1: Got that, but will it cause what I described or am I missing something?
<petrosagg>
bbrezill1: can mtd->_read ever return more than strength bits corrected?
<wens>
libv: i haven't looked, put in what the spec sheet said
<libv>
wens: since when do you believe spec sheets ;p
<libv>
wens: nah, this ties in with the a23 clocks we read a few h ago
<bbrezill1>
petrosagg: you're asumptions are correct
<dack>
wens: I'm trying to get a supplier to send a flash image for my device (http://linux-sunxi.org/TXCZ_A20) but their English is barely functional
<dack>
wens: I sent one email in Chinese by my wife to try to clarify my situation, but she doesn't know anything about technical things so the message probably wasn't very good.
<dack>
wens: I think the guy is just writing responses in Chinese and passing it through Google translate (or some equivalent)
<bbrezill1>
petrosagg: s/you're/your/
<dack>
wens: eh.. but then you probably only do traditional?
lkcl has joined #linux-sunxi
<dack>
does 刷 == flash when talking about firmware?
lkcl has quit [Ping timeout: 245 seconds]
<dack>
I guess it's "刷机" for "reflash"...
* nove
's little motivation build over this last weeks, went in a poof of smoke.
<dack>
nove: you let out the magic smoke!
<nove>
still no A80 android sdk sources?
<libv>
nove: check the a80 page
<libv>
hramrach: can you check the inet 86vs?
<libv>
hramrach: then we should iirc, be able to mark that one as done
<linkmauve1>
Are those Chinese docs publics? A friend of mine would like to give a try at making a proper translation.
<skoperst>
nove: look at the a80 page, there is a device repo, its enough to compile recovery/boot images
<nove>
libv: i did, but is only kernel/uboot
<wens>
dack: right
<libv>
nove: and buildroot
<libv>
nove: what else were you expecting?
<dack>
wens: Does the following make sense: "I need to 烧录/刷机 the device. I already have the 把烧录工具 and know how to use it. I need 烧录软件（后缀为img格式） for BBA22-UI4-V2.0.4"
<nove>
skoperst: yes, i saw, but i would like to look at the openmax sources, and cedarx use/headers
<wens>
dack: my terminal doesn't do simplified
<dack>
wens: tamade!
<wens>
dack: so half of what you just posted is ???
<dack>
wens: okay... :) anyone on here speak 中文?
jinzo has joined #linux-sunxi
jinzo has quit [Changing host]
jinzo has joined #linux-sunxi
<nove>
libv: the same as previous sdks, that includes everything
<nove>
libv: only empty tarballs and 2011 old cedarxtest application with a sample video file
<phh>
nove: google has been pushing manufacturesr to have hardware vp9
<libv>
skoperst, wens: is it me, or are pcduino8 and the optimusboard very similar in layout?
<libv>
ah, right, skoperst said so at the top of the page
<wens>
libv: i said the same thing earlier
<nove>
phh: but why is avc(h264?) and mpeg2 under allwinner
<phh>
ah yes sorry
<phh>
nove: OMX.google are software codecs
<phh>
OMX.allwinner are hardware ones
<phh>
so vp9 is sw actually
<libv>
wens: ah, missed that
<nove>
phh: that is the odd part, as A80 is advertised to have h265/VP9 1080p30fps
<phh>
sw is just not ready for it ?
<phh>
well the file you gave doesn't even mention hevc actually
<nove>
phh: well that is why i am curious, if is by hardware(video engine), or possible gpu compute, or ?
<jemk>
nove: "We have partnered with MultiCoreWare and Google to create an efficient software decoder based on GPU compute processing on our PowerVR Rogue GPUs."
<dack>
hramrach: I'm going to go with wens on this one... I got the text 烧录教程 from the title of a doc explaining how to reflash using an sdcard.
<dack>
libv: lol :) that's the opposite of what I meant... I meant adding the category to http://linux-sunxi.org/Template:Remove_only_when_finished so that when you added that template into a page that the category would be added to that page as well. Or is that possible?
<dack>
hramrach: but... WHERE CAN I GET THAT DICTIONARY?? :)
<libv>
ooh, that is a great idea indeed
<hramrach>
dack: pera pera extension to firefox
MY123 has joined #linux-sunxi
<hramrach>
in combination with google translate Chinese kind of makes sense
<dack>
libv: oh at least put the category tag up with the Remove_only_when_finished in the template
<libv>
yeah
<MY123>
hramrach: Not at all times.
<libv>
yup, works like a charm
<libv>
dack: great stuff
<dack>
hramrach: yeah, I use Google translate a lot...
<dack>
libv: sweet
<MY123>
libv: Between the Github page and the linux-sunxi.org, one says offset=40 for newer images (GitHub) , and linux-sunxi says offset 32. Who is right?
<MY123>
(U-boot)
<quitte>
MY123: 40
<libv>
MY123: where does our wiki say 40?
<libv>
MY123: if it does, it clearly states "for older u-boot"
<libv>
err, say 32
<quitte>
MY123: it depends on where spl tries to load u-boot from
<MY123>
libv: On the Wiki, it says offset 32 for newer images , and 40 for the older ones
<dack>
libv: now you can remove {{Remove_only_when_finished|[[Category:NDH_TODO]]}} from the bottom of the new device example page
<libv>
dack: yeah, i was just answering MY123 first
<dack>
libv: ;)
<libv>
dack: thanks for being so eagle eyed :)
lkcl has joined #linux-sunxi
<dack>
libv: just trying to simplify things for you... you spend too much time on the wiki. Any time I can save you on there the better!
<libv>
i do spend a lot of time on there, but i find it very important
<libv>
device pages == users
avsm has quit [Quit: Leaving.]
<libv>
and for most of the questions asked here, we can today just refer to the wiki
<libv>
that wasn't always so
<dack>
libv: yes, it is very important, but you're skills are better served elsewhere.. so, the more we can automate and reduce your time on there the better!
<dack>
libv: reducing time on questions by allowing you to point to the wiki is golden too
<libv>
i know that my skill should be used elsewhere
<libv>
but only recently have i been getting more people to work the wiki
<libv>
and for many things it still is: if libv doesn't sit down and do it, it'll never happen, as most seem to be above this sort of thing (while at the same time depending on it)
xavia has quit [Ping timeout: 250 seconds]
<quitte>
hmm. despite the CRC errors fw_setenv stored a data pair just fine
<hramrach>
what are some useful debug options for kernel when you have just small screen and no other console?
<hramrach>
the last thing that kernel says is it detected mmc0 as sdhc card
<hramrach>
but does not mount root
<hramrach>
nor does it print partition map
<MY123>
hramrach: UART
<MY123>
or JTAG
<hramrach>
ENOUART
<libv>
hramrach: hrm, is this the 86vs?
<hramrach>
no, a86
<hramrach>
but very similar hardware
<hramrach>
could check the fex if there are free pins for console
<hramrach>
if I remove the card it registers that as well
<WarheadsSE>
PaulFertser: you can also have a look @ the Kirkwood lines. Sure, they are older, but are designed for NAS, and while they have no graphics, armv5te, .. they do have GbE, decent SATA performance, and are mainlined.
<hramrach>
but does not panic because of missing root as it should
wingrime has joined #linux-sunxi
<MY123>
libv: Does the Allwinner A20 boot1 manage multiboot?
<wingrime>
libv: I increase gpl violation on main page just by little
<wingrime>
*font
<PaulFertser>
WarheadsSE: thank you. I know Dockstars were quite popular some time ago, but haven't seen much about modern boards.
<libv>
wingrime: i am not going dis or encourage you to do that, but imho it's perhaps better to tell simos that the gpl violations are not conditional
<WarheadsSE>
1.6Ghz, I do need to make a DTB for it . . but I have done the work to open the kernel, the uboot, et al
<PaulFertser>
WarheadsSE: looks interesting. Where can we move to avoid offtopic here?
<MY123>
libv: Is there an API to use the OpenRISC core in the A31 as a coprocessor? Can't seem to find that.
<libv>
MY123: i am not sure whether the GPL can be enforced on a microcontroller firmware
<libv>
depends on the interface and interdependency
<libv>
and this needs very thorough investigation
<WarheadsSE>
PaulFertser: feel free to #archlinux-arm
<libv>
which i am not going to do, as i have wasted a lot of time on this stuff already
<nove>
MY123: do you see the AR100 wiki page, there is a link to github
ricardocrudo has quit [Remote host closed the connection]
ricardocrudo has joined #linux-sunxi
bertrik has joined #linux-sunxi
bonbons has joined #linux-sunxi
petr has quit [Ping timeout: 250 seconds]
lkcl has quit [Ping timeout: 244 seconds]
petr has joined #linux-sunxi
lkcl has joined #linux-sunxi
<skoperst>
Is the DTB is loaded by the kernel? or the bootloader? (I want to add DTB to device that has no DTB before)
<skoperst>
And another thing, I want to take sunxi-next and add A80 support, is it theoretically possible?
<mripard_>
skoperst: it's loaded by the bootloader
<mripard_>
and yes.
<skoperst>
mripard_: If its loaded by the bootloader so I need to get a DTB capable bootloader..
<skoperst>
am i right? because with the current Bootloader it will not go
<hramrach>
what does the bootloader need to do other than loading the DTB blob at certain address?
<skoperst>
AFAIK, it need to load the kernel, ramdisk,DTB and also give a mode to flash these parts quickly(in android its called fastboot)
notmart has quit [Quit: notmart terminated!]
ricardocrudo has quit [Remote host closed the connection]
ricardocrudo has joined #linux-sunxi
uwe__ has joined #linux-sunxi
uwe_ has quit [Ping timeout: 260 seconds]
MY123 has quit [Remote host closed the connection]
<wens>
mripard_: on my a23, if i don't mux ahb1 to pll6, dmatest fails on prep
lkcl has quit [Ping timeout: 245 seconds]
<wens>
libv: if both boards are the same, then we bring up 2 products at the same time, and probably have more people to work and test it
jemk has quit [Quit: leaving]
<libv>
wens: yeah, hopefully
<libv>
wens: but don't unify the work too soon
<wens>
don't plan to, but initial bringup is only uarts, which everyone can test
<libv>
wens: before you run off an play with upstream
<libv>
wens: can you document how to get things going with the sdk uboot and kernel?
<wens>
libv: i'll give it a run tomorrow (hopefully)
<wens>
cnx-soft's attempt should be useful
uwe__ is now known as uwe_
<wens>
which reminds me, i ran out of sd cards
<libv>
wens: you can never have enough of those
<wens>
libv: i agree, but having a lot around tends to get them mixed up
<libv>
micro-sd sucks that way
deasy has quit [Remote host closed the connection]
libcg has joined #linux-sunxi
BorgCuba has joined #linux-sunxi
enrico_ has quit [Quit: Bye]
skoperst has quit [Quit: Page closed]
libcg has quit [Quit: libcg]
_massi has quit [Remote host closed the connection]
diego_r has quit [Ping timeout: 260 seconds]
<petrosagg>
skoperst: the bootloader doesn't use the dtb file in any way
<petrosagg>
skoperst: it loads the file in memory and then the kernel reads it from there
<petrosagg>
skoperst: so in the end, it's the kernel that reads and parses that file but it is already loaded in memory by the bootloader
kuldeepdhaka has joined #linux-sunxi
<petrosagg>
skoperst: you shuldn't need any special u-boot to load your dtb
<petrosagg>
skoperst: s/shuldn't/shouldn't
<deepe>
q
<skoperst1>
petrosagg: tnx for the clearifying
konradoo77 has joined #linux-sunxi
Seppoz has quit [Ping timeout: 250 seconds]
uwe_ has quit [Read error: Connection reset by peer]
uwe_ has joined #linux-sunxi
uwe__ has joined #linux-sunxi
uwe_ has quit [Ping timeout: 244 seconds]
F1skr has joined #linux-sunxi
rz2k has quit []
<BorgCuba>
whats -m switch for disassembling thumb2 or in general the objdump syntax?
<libv>
BorgCuba: a recent enough binutils should just do it for you
balog has quit [Quit: Leaving.]
<BorgCuba>
its a binary file and it is: --disassembler-options=force-thumb
<BorgCuba>
like "arm-linux-gnueabihf-objdump -D -m arm --disassembler-options=force-thumb -b binary $INFILE > $OUTFILE"
konradoo77 has quit [Ping timeout: 240 seconds]
konradoo77 has joined #linux-sunxi
<BorgCuba>
libv, I see you are hanging out on the rockchip channel as well. I currently try to run the u-boot version from this repo "https://github.com/neo-technologies/rockchip_u-boot" from sd-card but it stays silent while the std. proprietary bootloader works. Any ideas? Do you know somebody who did this?
lkcl has joined #linux-sunxi
deasy has joined #linux-sunxi
F1skr has quit [Quit: WeeChat 1.0]
deasy has quit [Remote host closed the connection]
nedko has quit [Quit: kernel panic]
BorgCuba has quit [Quit: leaving]
nove has quit [Quit: nove]
skoperst has joined #linux-sunxi
formruga has joined #linux-sunxi
skoperst1 has quit [Ping timeout: 260 seconds]
<mripard_>
petrosagg: that's completely wrong.
<mripard_>
skoperst: in a perfect world, the bootloader should have DT support
konradoo77 has quit [Ping timeout: 260 seconds]
<mripard_>
to load it, but also to pass it to the krenel in the proper way
<mripard_>
moreover, most of the bootloaders patch the DTB just before jumping into the kernel
amitk has quit [Quit: leaving]
<mripard_>
for example to give to give its command line to the kernel, or to set the mac address
<mripard_>
in a !perfect wordl, you have a legacy option in the kernel to append the DTB right after the kernel image, so that !DT-enabled bootloaders can still boot a DT kernel
<mripard_>
obviously, all the DT mangling won't work in this case, but still...
sehraf has quit [Read error: Connection reset by peer]
konradoo77 has joined #linux-sunxi
deasy has joined #linux-sunxi
RzBo has joined #linux-sunxi
<RzBo>
Hello! I have a question about AXP209 chip :
<RzBo>
I'm trying to read its internal temperature at 5Eh, 5Fh
<RzBo>
I have written a bash script (I am using a pcDuino board)
<RzBo>
then i convert hex value to °C
<RzBo>
but values doesn't look reliable but i'm always reading three different values that changing about 1°C each time. example : 28.3°C then 29.3°C then 30.3°C
<RzBo>
when i try to heat the chip, temperature doesn't change :-(