<libv>
it's just going to be a few people who work on mainlining, probably to pad resumes, and to work on the coolest project of them all. who cares if there are actual users for it.
<libv>
it'll be years before the code gets backed out, if at all, and by then, nobody remembers anything anymore anyway
<wens>
looking at the a80 datasheet, some of the new stuff is nice
<wens>
a80 headless bring-up should be easy to do, assuming aw is following past conventions
<wens>
troublesome part would be smp and big.LITTLE stuff
<Turl>
and bootloader :\
<wens>
yeah, bootloader is a pain :(
<wens>
i suppose i should go get an optimus board
<Turl>
for like 300$? :p
<wens>
it's down to $165-ish
ZetaNeta has joined #linux-sunxi
<ZetaNeta>
Hi
<ZetaNeta>
I am now banned from #gentoo-chat....
<ZetaNeta>
And i will have to return back to a less, but still very familiar channel
<wens>
and assuming they didn't change much, we already have some code for the a80 (in the a23 sdk)
<ZetaNeta>
Yay! Some news about A80!
<ZetaNeta>
Do we have here a bot which can be feed with cookies, and a mooing lemur which i can hug every 1.5 hour?
<wens>
Turl: once you're done with audio, i think i can port it over to a23
<wens>
or if you have the hardware, you could do it yourself :)
<libv>
ZetaNeta: here you have a wiki which you can feed ndh.
<Turl>
wens: does it share the ip with the older SoCs?
<ZetaNeta>
libv, And what about lemur?
<libv>
ZetaNeta: i will happily moo for you each time you ndh one of your devices.
<wens>
Turl: i don't see separate drivers, so it should :)
<wens>
which means i can easily do some bring-up code, even if i don't have the hardware lol
<wens>
pinctrl drivers look right
<Turl>
wens: the A31 audio is different though, so I'd have expected it to be like that instead
<wens>
Turl: nah, the A23 looks more like A20 :)
<wens>
except for the analog bits, which have been moved into the prcm
leviathanch2 has quit [Ping timeout: 246 seconds]
<wens>
probably need "mfd: syscon: Decouple syscon interface from syscon devices" to make a syscon node for that
<libv>
ooh
<libv>
allwinner uses vp62 code
<libv>
vp6 was once used by macromedia flash for video playback
<libv>
someone release libvp62 as free software
<libv>
but the creator of vp6 (on2) got it removed
<libv>
and here allwinner is, using vp62 code :)
<libv>
how hot do we want allwinners water?
hipboi has quit [Ping timeout: 255 seconds]
<libv>
hrm, anti compiled from java
<wens>
doesn't ffmpeg have vp6 support?
<libv>
it does
<libv>
but this vp62 code appears around the ff_huff_build_tree code
<libv>
which is an ffmpeg symbol
<libv>
so my guess is that vp62 used ffmpeg code
<libv>
and allwinner then grabbed the forbidden vp62 code
hipboi has joined #linux-sunxi
<Turl>
didn't google buy on2's stuff later on?
<libv>
apparently
<libv>
but it is still very amazing that this code pops up now
<libv>
so parts of the vp62 code and a bit of ffmpeg are used here...
<libv>
there's all sorts of nasty in cedar code
<Turl>
libv: how do you know this is vp62 btw? :)
<libv>
google still has these files
<libv>
just the actual files though
<libv>
no tarballs or anything
<libv>
on pretty hairy sites as well
<libv>
but it's still there
<libv>
which is why getting pages removed from the internet is bad.
<libv>
here we have a clear case where the fact that some sources are still available gives us the ability to verify that this is indeed very very bad code.
<libv>
if it had vanished completely, only very few people would've been able to know this
<mripard_>
libv: so I guess what you meant by "you have to keep the pressure constant for anything to happen" is that we should keep reminding you to work on the KMS driver instead of doing pointless and/or backward changes to the wiki, right?
ZetaNeta has quit [Remote host closed the connection]
nabblet has quit [Ping timeout: 272 seconds]
hipboi has quit [Ping timeout: 264 seconds]
<mripard_>
libv: no wonder why we never had any formal contact with them before...
<mripard_>
if as soon as someone reach out to our community, they're being insulted
<mripard_>
I know you're doing your best. really.
<mripard_>
you like to brag about how important you are in this community, but really, I don't see what you bring to the table beside insults and whinings.
hipboi has joined #linux-sunxi
eagles0513875 has quit [Max SendQ exceeded]
PulkoMandy has joined #linux-sunxi
eagles0513875 has joined #linux-sunxi
lkcl has quit [Ping timeout: 272 seconds]
hipboi has quit [Ping timeout: 245 seconds]
eagles0513875 has quit [Changing host]
eagles0513875 has joined #linux-sunxi
eagles0513875 has quit [Changing host]
eagles0513875 has joined #linux-sunxi
hipboi has joined #linux-sunxi
lkcl has joined #linux-sunxi
Quarx has joined #linux-sunxi
yann_s|AFK has joined #linux-sunxi
yann_s|AFK is now known as yann_s
paulk-aldrin has joined #linux-sunxi
nabblet has joined #linux-sunxi
arokux has quit [Ping timeout: 245 seconds]
kuldeepdhaka has joined #linux-sunxi
nabblet has quit [Ping timeout: 260 seconds]
kuldeepdhaka has quit [Max SendQ exceeded]
arokux has joined #linux-sunxi
bengal has joined #linux-sunxi
kuldeepdhaka has joined #linux-sunxi
kuldeepdhaka_ has joined #linux-sunxi
nabblet has joined #linux-sunxi
kuldeepdhaka has quit [Ping timeout: 255 seconds]
hipboi has quit [Ping timeout: 272 seconds]
lkcl has quit [Ping timeout: 245 seconds]
kuldeepdhaka has joined #linux-sunxi
kuldeepdhaka_ has quit [Ping timeout: 250 seconds]
Nyuutwo has quit [Read error: No route to host]
Nyuutwo has joined #linux-sunxi
nabblet has quit [Quit: leaving]
akaizen_ has joined #linux-sunxi
akaizen has quit [Ping timeout: 245 seconds]
hipboi has joined #linux-sunxi
bonbons has joined #linux-sunxi
kuldeepdhaka has quit [Ping timeout: 250 seconds]
hipboi has quit [Quit: Leaving]
kuldeepdhaka has joined #linux-sunxi
quitte has quit [Read error: Connection reset by peer]
<libv>
wens: does that sound like a sensible way to go about things today? just dump the a23/a31/a33 dram and ccmu register range
<Turl>
oliv3r: pong
<oliv3r>
Turl: nvm :)
<oliv3r>
just backread a little here :)
<libv>
oliv3r: you should be able to test such code on another allwinner device, and verify the results there
<oliv3r>
i'll dd the partitions to my sd card first
<libv>
i doubt that the base addresses will have changed
<ssvb>
libv: IMHO Allwinner started taking the importance of mainline kernel support more seriously (and hence considering to provide better assistance to the people who are actually doing the work) because of the recent moves from Rockchip and Mediatec
<ssvb>
libv: and not because somebody accused them of GPL violations
<libv>
ssvb: i'd say wait and see, but that's not going to prove things either way
<libv>
they will keep on dragging their feet anyway, and it remains to be seen how useful these allwinner engineers end up being
konradoo87 has joined #linux-sunxi
pwhalen has quit [Ping timeout: 245 seconds]
<ssvb>
libv: I mean, joining Linaro was good for PR purposes, but now Allwinner people have probably noticed that Rockchip developers have started to contribute actual code :)
naobsd has joined #linux-sunxi
konradoo77 has quit [Ping timeout: 246 seconds]
<libv>
ssvb: in any case, these blatant and continuing gpl violations cannot be ignored just because they gave us two email addresses and a "promise"
<libv>
the latter i know is worthless, i have been there too often before
<libv>
via was full of such idiocy, and some parties still call that a chinese company
<libv>
you'd think that .tw companies would have been far ahead a decade ago, in dealing with western culture and in solving problems, compared to chinese companies today
<ssvb>
libv: at any rate, talking about gpl violations with the allwinner *engineers* assigned to support linux-sunxi is probably not very productive
<libv>
ssvb: but acting as if all is good now, and as if there are no issues there is plain wrong.
<ssvb>
libv: you probably need to reach their lawyers or managers
<Turl>
I wonder if they have lawyers
<libv>
ssvb: we've said this before: from the time they joined linaro, until now, there was no talking with eva
<libv>
before that time, eva seemed to have been directly reachable
<libv>
ben el-baz, i am not sure what he really means
<libv>
he wrote "olinuxino" right after cubietech, and not olimex
<libv>
i am not sure whether he is more than just marketing
<libv>
if he had dealt with olimex directly, or had even just spoken to them, he would've gotten that name right
<wens>
libv: what were you replying to?
<ssvb>
libv: also having the sources of their proprietary code is less important than getting accurate and detailed hardware documentation
<libv>
ssvb: the latter never happens
<libv>
you get "some" documentation, always
<libv>
and you know it
<ssvb>
libv: yes, nothing is ever 100% perfect
FreezingCold has quit [Ping timeout: 240 seconds]
pwhalen has joined #linux-sunxi
<libv>
ssvb: gpl is a legal requirement, docs aren't
<libv>
ssvb: allwinner is required to give us that code, but it only does docs on a goodwill basis
<ssvb>
libv: these things are completely orthogonal
<libv>
the latter is very ethereal
<libv>
yes, that is true
<libv>
but that doesn't mean that the first should be swept under the rug
<libv>
the first has to happen, the second, depends on the mood of people inside allwinner, which could be different next week due to no fault of our own
<libv>
code can always be rescued.
lkcl has joined #linux-sunxi
<ssvb>
libv: not always :) sometimes the code may be irreversibly lost for old EOL products
<libv>
ssvb: that's an excuse these days
<libv>
ssvb: i heard that excuse from ati all the time
<libv>
"the people have gone, the documentation was never written"
<libv>
that was just one of the many lies ATI fed us on the radeonhd project
<ssvb>
libv: it's not an excuse, sometimes this happens for real
<libv>
ssvb: very very very rarely
<libv>
ssvb: i got told about a situation like that from amd, after they threw out their german developers 1-2 years ago
<libv>
turns out that this was a lie too
<libv>
the code was there, they just didn't want to reveal it
MackBoy_ has quit [Remote host closed the connection]
<ssvb>
libv: more than a decade ago I was applying for a job in a company, which had lost the original sources of their old application
<libv>
ssvb: and it's not as if allwinner has been seriously downsizing in the last few years
<libv>
they have the code
<libv>
ssvb: how old? 90s?
MackBoy has joined #linux-sunxi
<libv>
it was common then, but it isn't today
libcg has quit [Remote host closed the connection]
<ssvb>
libv: yes, the application was from mid-nineties, and around 2000 they decided that they wanted to add some new features to it and release a new version :)
<libv>
ssvb: none of the code we are talking about is older than 2007, max.
<libv>
allwinner has the code for libnand, libisp and cedarx
<libv>
there are no excuses there.
FreezingCold has joined #linux-sunxi
<libv>
it happened a lot 20ys ago, it might very very rarely happen today, but if anyone says that today, especially in a situation like this, you have 99% certainty that it is a lie
<libv>
especially if it is a reason for not releasing any code, if some ancient version cannot be retrieved then so be it
<libv>
but none of that applies in this situation
<mripard_>
libv: feel free to think that you're important. Of course, that has nothing to do with Simos, the interview, or anything, it's only thanks to you...
<mripard_>
and I'm sorry, but saying that Jon and I are sucking up is definitely an insult.
pwhalen has quit [Ping timeout: 260 seconds]
<libv>
mripard_: heh.
lkcl has quit [Ping timeout: 255 seconds]
<libv>
mripard_: remind me, how did you get into allwinner soc development?
<libv>
mripard_: did you find linux-sunxi.org and did you go from there?
<mripard_>
bought a board, started hacking.
<mripard_>
no
<mripard_>
it was the other way around
<mripard_>
I started working on it and then was pointed here
<libv>
heh, ok.
<libv>
understood.
<libv>
continue ignoring the gpl violations as well then, and continue to hope for good things to come from allwinner just because they want good for the world.
<mripard_>
I'm not saying that either
<mripard_>
but I don't believe in being rude with every newcomer.
<mripard_>
be it allwinner or anyone else
<libv>
mripard_: these are not newcomers, these are people (supposedly) tasked to solve the issues
<mripard_>
honestly, I would have been welcome here like you're used to these day, I wouldn't have even bother.
<libv>
...
<ssvb>
libv: they are tasked to solve some very specific issues, and may be simply not authorized to talk about legal stuff and licenses
<libv>
ssvb: so when will alwinner be telling their managers and legal department to go talk to us?
<libv>
ssvb: or do we tell these guys, and these guys have to feed that up the foodchain in allwinner?
<wens>
it won't work that way # feed up the foodchain
<ssvb>
libv: imho you are just putting these guys in an awkward position
atsampson has quit [Quit: I'm going to move my PC 40cm to the left]
<libv>
they are in an awkward position.
<libv>
their company blatantly violates the gpl, and now hopes that this will make that go away?
<petrosagg>
hi guys, if I have an u-boot-spl.bin file how do I write it on the nand boot partition? I'm using a kernel that has the new /dev/mtd0 interface instead of the /dev/nanda etc of the sunxi-3.4 kernel
<oliv3r>
petrosagg: you can't :)
<libv>
petrosagg: why?
<oliv3r>
petrosagg: u-boot-spl is for microsd only
<petrosagg>
libv: I'm trying to make my board boot from NAND
<libv>
petrosagg: what are you trying to achieve?
<oliv3r>
petrosagg: unless your hacking on u-boot + mtd stuff, and then you'll probably know more then us :)
<libv>
petrosagg: did you try reading our wiki?
<ssvb>
libv: you have worked for a big company yourself and should know that some developers may have no idea what the other departments are doing
<oliv3r>
petrosagg: you are hacking on mtd then; so you should probbaly tell us how to write to the bootloader area ;)
<petrosagg>
libv: cubieboard2
<libv>
petrosagg: do you have the thing booting from sd card already?
<petrosagg>
libv: yes
<oliv3r>
petrosagg: mtd stuff is Highly highly highly experimental ...
<libv>
petrosagg: why don't you have access to nanda from the sdcard?
<oliv3r>
petrosagg: so as I said, if you are playing with that, you probably should be very 'into' it
<petrosagg>
oliv3r: apparently you can replace boot0 with the u-boot spl
<oliv3r>
only with yuq's patches applied
<oliv3r>
and even then, I don't think you can 'just write' it yet
<petrosagg>
libv: the nanda interface is when your kernel is complied with the old allwinner's drivers
pwhalen has joined #linux-sunxi
<libv>
petrosagg: so you wish to use mtd?
<petrosagg>
libv: I was working on this board a few days ago with the help of bbrezillon and now I can have proper mtd device support for my nand
<libv>
ah, ok
<petrosagg>
libv: yes
<libv>
petrosagg: ask bbrezillon and the others who are working on this to document how to get uboot running from the nand.
<libv>
or find out yourself, and document then
<oliv3r>
i belive yuq had a special code path in his kernel driver to enable that
<oliv3r>
but i don't think we can write the bootloader to mtd yet
<oliv3r>
since it requires a 'special mode'
<petrosagg>
alright. I've done some research already. I guess I'll keep looking
<petrosagg>
oliv3r: I think the special mode you're referring to is a specific hw randomiser setting
<libv>
ssvb: these guys were put in the firing line by their company, they have to hear about the issues that cannot be ignored.
<oliv3r>
i think there was some talk about having a sperate driver to unlock the bootloader area and use that to write the bootloader
<oliv3r>
petrosagg: could be close to yeah
<petrosagg>
oliv3r: which I've already have set up
<oliv3r>
petrosagg: i'm not sure this code exists yet
<petrosagg>
oliv3r: I can surely write there from linux as I've destroyed the u-boot that was there before. It's just that when I try to write things it fails for some reason
<oliv3r>
to write the bootloader
<petrosagg>
oliv3r: I used nandwrite and nand_erase to do that
<oliv3r>
petrosagg: but your pionering here :) no made manuals and howto's up to you to write it :D
<petrosagg>
oliv3r: haha, ok. My question was a more general of "how to write a .bin file to and mtd partition"
<petrosagg>
oliv3r: I can't just write to it as it isn't a block device and nandwrite complains often
<petrosagg>
oliv3r: also if I write something with nandwrite and then nanddump the partition I get different data
<petrosagg>
oliv3r: anyway, I'll continue trying and report back
<oliv3r>
petrosagg: good luck :)
<petrosagg>
oliv3r: thanks :)
<libv>
paulk-aldrin: can you create a close-up picture of the uart?
<paulk-aldrin>
well, now it's soldered
<libv>
paulk-aldrin: nah, by editing the big board picture
<paulk-aldrin>
ah ok, so just zooming in
<paulk-aldrin>
sure
<libv>
paulk-aldrin: you just need to have part of the touch controller connection, nand chip, and soc in that closeup picture
<libv>
but cool, the usual 2 pads near the SoC :)
<paulk-aldrin>
it's the camer a
<paulk-aldrin>
camera*
<libv>
paulk-aldrin: what test is under that tape btw
<libv>
paulk-aldrin: ah, ok
popolon has quit [Quit: Quitte]
<libv>
...6Q-V1.1
<paulk-aldrin>
libv, well, plenty of them, but they're not labelled
<paulk-aldrin>
ah ok
konradoo87 has quit [Ping timeout: 260 seconds]
<paulk-aldrin>
I'll write it down fully in identification
<paulk-aldrin>
good point
<libv>
underneath the ram
<libv>
ok :)
<paulk-aldrin>
yeah
<libv>
oooh, i just learnt that 94V-0 is a generic marking denoting flammability
<oliv3r>
paulk-aldrin: i dint' root it, but adb ran as root
<ssvb>
nove: don't worry, g2d is an even better security hole without the 256M restriction :)
<libv>
hehe :)
<ssvb>
and, I guess, the disp driver too (if the writeback pipeline can be utilized)
<Turl>
there's /dev/mem too
<oliv3r>
paulk-aldrin: i think on our devices, the 'scratchpad' is nandb; but everything not sun5i has some bytes available in the rtc
<nove>
and as allwinner joined linaro home group, suppose to get it soc in settopbox, is funny to see what software will run in it
<libv>
Turl: you'd need to be root to access that, right?
<ssvb>
nove: we are all grateful to allwinner for their marvelous security model, with non-crippled ARM processors booting in secure mode (and have full privileges), no signed bootloaders and also using the SD card as the primary boot device :)
<petrosagg>
I'm really confused. Has anyone used mtd_debug read/write or nanddump/nandwrite?
<libv>
ssvb: true :)
<nove>
ssvb: yes, that is good side
<oliv3r>
nove: how do you mean millions of allwinner devices can be trashed
<Turl>
libv: yeah, same with g2d, disp and cedar on default config right?
lkcl has joined #linux-sunxi
<nove>
oliv3r: "trashed" as in, is easy to write trash to memory locations
<ssvb>
Turl: unlike /dev/mem, there might be a legitimate reason for non-root user applications to access g2d, disp and cedar
<ssvb>
Turl: the users from the 'video' group, or something like this
<oliv3r>
nove: ah ok, 'hackable'
<ssvb>
nove: if allwinner eventually gets the security implemented right, then we might end up with non-hackable devices unable to run proper linux without the cat and mouse jailbreaking fun enjoyed by the others
konradoo77 has quit [Ping timeout: 264 seconds]
Quarx has quit [Ping timeout: 250 seconds]
<ssvb>
nove: except for the development boards
lkcl has quit [Ping timeout: 240 seconds]
<petrosagg>
bbrezillon: I think something is wrong with the boot and boot-rescue partitions on the NAND. Do they work for you on the cubietruck?
<petrosagg>
bbrezillon: specifically, can you run nandtest on them?
konradoo77 has joined #linux-sunxi
<oliv3r>
mnemoc: E: script.bin:1: data must follow a section.
<oliv3r>
mnemoc: this on an a10s script.bin
<libv>
mnemoc: i think i might have seen this before
<libv>
errm oliv3r
<libv>
not sure what i did though
<libv>
oliv3r: what does strings say? is it actually a script.bin?
<libv>
oliv3r: are you using the latest fexc?
<oliv3r>
libv: yeah it's a script.bin, less shows the comon para's
<oliv3r>
i did obtain it via adb pull, but that shouldn't cause corruption though?
<libv>
indeed
<libv>
fix fexc to stop throwing in the towel :)
<libv>
oliv3r: it's amazing what one runs into when ndhing hw, no?
<oliv3r>
i AM on the ndh
<libv>
yeah, i know, but you run into all sorts of random things
<oliv3r>
strange fexc works, bin2fex not
<oliv3r>
strange
<libv>
which many normal users have a hard time dealing with
<libv>
eh?
<oliv3r>
oh nvm, i'm being an idiot, fex2bin instead of bin2fex
<libv>
shouldn't that be a link?
<libv>
ah
<oliv3r>
machine = "A10s-h25-V1.2"
<oliv3r>
they actually setup the machine properly
<oliv3r>
h25 is printed on the PCB too
Gerwin_J has joined #linux-sunxi
ricardocrudo has joined #linux-sunxi
<libv>
nice
<libv>
how come you are already retrieving device info while there is no wiki page :p
<libv>
oliv3r: the fact that those jesurun letters are a bit too much to the right, and that they are attached, that points to the fact that this is just stuck on
<libv>
look at the line between the J and the e
<libv>
Xplus was printed on, by machine
<libv>
or at least it's centered
<libv>
jesurun was stuck on by hand
<oliv3r>
btw, the A33 seems to run at 1200 stock
<oliv3r>
in your pic, i'm sure
<libv>
oliv3r: no, your pic shows what i just explained
<oliv3r>
i'm just saying, the device i have in my hand here, has the jesurun 'logo' come 'out' in 3D
<libv>
oliv3r: why does that discount it being stuck on?
<libv>
the url i posted also seems to have these letters raised
<oliv3r>
well you can't just remove it i mean
<libv>
otherwise the layer of plastic on top wouldn't have been that visible
<libv>
good glue?
<oliv3r>
lol no
<oliv3r>
it's plastic
<libv>
it's just not part of the case mold
<oliv3r>
it looks like it is
<oliv3r>
or it was melted ontop
<libv>
again, good glue
<oliv3r>
i can't peel it off in any case :p
<oliv3r>
i'll try a knife
<libv>
why?
<libv>
why bother?
<oliv3r>
curious!
<libv>
it's not straight, it's too much to the right
<oliv3r>
what is the 'remove|' bit for in the infobox
<libv>
the letters are attached to eachother artificially
<libv>
oliv3r: to make it stand out that it needs to be edited
<libv>
oliv3r: check other device pages
<oliv3r>
so the remove| bit needs to be actually removed
<libv>
yes
<libv>
this page should not have any red left afterwards
<libv>
i should actually make 2 more of those
<libv>
Edit and Remove_when_fully_finished
lkcl has quit [Ping timeout: 260 seconds]
imcsk8 has quit [Quit: Reconnecting]
imcsk8 has joined #linux-sunxi
<oliv3r>
WiFi 802.11 b/g/n probably should read a/b/g/n/ac :p
<libv>
how often do you see a and ac?
<libv>
today still?
<libv>
iirc, there was talk of the fcc outlawing a
<oliv3r>
oh really?
<oliv3r>
all new ac devices will do a
<oliv3r>
and a is used on all 5ghz 'n' devices
<oliv3r>
since there's 2 'n's 'smartly'
<oliv3r>
n @ 2.4ghz and n @ 5ghz
<libv>
oh, i see
<libv>
then what wifi band was being outlawed
<paulk-aldrin>
I finally got around running meminfo
<paulk-aldrin>
it doesn't match script.bin at all
<paulk-aldrin>
and it also happens that the A13_MID config works
<libv>
paulk-aldrin: try to stick with meminfo
<oliv3r>
paulk-aldrin: yeah script.bin's meminfo is filled either by livesuit or manually by engineers
<libv>
and the values from that
<paulk-aldrin>
ok
<libv>
loads of things are compatible with eachother though, like for instance a lower frequency setting
<oliv3r>
meminfo reads the actual registers
deasy has joined #linux-sunxi
<paulk-aldrin>
what's CONS_INDEX=2 in boards.cfg btw?
<libv>
selects a different uart for console
<paulk-aldrin>
okay
ninolein has quit [Ping timeout: 260 seconds]
ninolein has joined #linux-sunxi
<oliv3r>
can we find the 'strings from within android' in one of the .prop files?
<oliv3r>
i'm runnning it only via otg and don't have it connected to a hdmi connector