flyhorse_jtRrT has quit [Ping timeout: 265 seconds]
flyhorse has quit [Read error: Connection reset by peer]
flyhorse has joined #linux-sunxi
naobsd has quit [Quit: naobsd]
mmarker has quit [Read error: Connection reset by peer]
ninolein has joined #linux-sunxi
mmarker has joined #linux-sunxi
ninolein_ has quit [Ping timeout: 256 seconds]
khuey|away is now known as khuey
egbert has quit [Disconnected by services]
egbert has joined #linux-sunxi
naobsd has joined #linux-sunxi
flyhorse_DgmZa has joined #linux-sunxi
flyhorse has quit [Read error: Connection reset by peer]
kaspter has quit [Ping timeout: 255 seconds]
Gerwin_J has quit [Quit: Gerwin_J]
Gerwin_J has joined #linux-sunxi
Gerwin_J has quit [Quit: Gerwin_J]
kaspter has joined #linux-sunxi
Gerwin_J has joined #linux-sunxi
[7] has quit [Disconnected by services]
TheSeven has joined #linux-sunxi
Gerwin_J has quit [Quit: Gerwin_J]
flyhorse has joined #linux-sunxi
flyhorse_DgmZa has quit [Read error: Connection reset by peer]
naobsd has quit [Quit: naobsd]
khuey is now known as khuey|away
bobryan has quit [Ping timeout: 256 seconds]
bobryan has joined #linux-sunxi
gzamboni has quit [Ping timeout: 272 seconds]
flyhorse_nFCAG has joined #linux-sunxi
flyhorse has quit [Read error: Connection reset by peer]
naobsd has joined #linux-sunxi
flyhorse has joined #linux-sunxi
flyhorse_nFCAG has quit [Read error: Connection reset by peer]
<plaes>
why do we have Xunlong stuff in Community devices category?
cubeast has joined #linux-sunxi
JohnDoe_71Rus has joined #linux-sunxi
pstef has joined #linux-sunxi
domidumont has joined #linux-sunxi
domidumont has quit [Remote host closed the connection]
domidumont has joined #linux-sunxi
domidumont has quit [Client Quit]
gzamboni has joined #linux-sunxi
domidumont has joined #linux-sunxi
<rellla>
plaes: because the wiki was spammed (kind of) with the orange pages. incomplete, faulty ...
<rellla>
I didn't have the time to strip the pages yet
<plaes>
ok, I dropped some of the xunlong items from the mainline kernel category
prz has joined #linux-sunxi
<rellla>
plaes, i see you try to take over the label "mr. wiki" from libv ;)
<plaes>
nah.. ;)
<whitesn>
are there any way to install rom on cubieboard 2's NAND without using sd?
<RSpliet>
whitesn: to the best of my knowledge no, but you should be able to build such a tool quite easily, borrowing from the sunxi reverse engineering tools
<RSpliet>
in particular the one that partitions your NAND
<whitesn>
RSpliet: does this mean i can install directly using FEL mode from another machine?
<RSpliet>
oh good, there's closed source tools too... I guess you can
JohnDoe71rus has joined #linux-sunxi
JohnDoe_71Rus has quit [Read error: Connection reset by peer]
<whitesn>
hmm i will attach a sata hdd on the cubie
<whitesn>
i wonder if it's better to boot from sata or the nand
Andy-D has quit [Ping timeout: 252 seconds]
<RSpliet>
I'm not sure whether the first level bootloader (BROM) will attempt to read U-boot from SATA
awe00 has joined #linux-sunxi
Andy-D has joined #linux-sunxi
<jackdaniel>
it won't, it tries mmc, nand and emmc on a20
<jackdaniel>
hi o/
mawe242 has joined #linux-sunxi
trepidaciousMB-1 has quit [Quit: trepidaciousMB-1]
awe00 has quit [Ping timeout: 256 seconds]
<RSpliet>
jackdaniel: hey!
<RSpliet>
sorry for the lack of feedback from my side, been away for a long weekend
<jackdaniel>
me either, no problem :)
<jackdaniel>
I've reinstalled my OS, and it's semi working now ^^
<RSpliet>
right now I'm looking at using the existing mtdpart for nand partitioning, see how much adapting is required for ECC support
<RSpliet>
bbrezillon: is there a particular reason why you attempted to write your own partitioning code rather than use mtdpart? problems you ran in to?
<wens>
mripard: are you sending out pull reqs for the remaining 4.2 stuff in your tree? (maybe update sunxi-next as well? :) )
awe00 has joined #linux-sunxi
paulk-collins has joined #linux-sunxi
<mripard>
wens: yes
trepidaciousMBR2 has joined #linux-sunxi
vishnup has joined #linux-sunxi
<wens>
mripard: btw, i have some sun8i usb patches lying around, though i need to bring up my a33 board to test them
<mripard>
wens: I'll work on that too in the next few weeks, so if you prefer to spend that on something else... :)
<mripard>
I'm not trying to take anything from you though, if you want to do it, go ahead
<wens>
mripard: it was done to get otg working on my a23 tablet, with an earlier version of the musb patches
<wens>
iirc the phy and clock stuff are already done, just need to test ordinary usb hci
heffer has quit [Ping timeout: 245 seconds]
ricardocrudo has joined #linux-sunxi
<cubeast>
Hey! Which embedded boards have: a) SATA and b) Gigabit ethernet?
<wens>
plaes: u-boot on my a33... can't get it working
<wens>
oh, i used the wrong blob :|
<bbrezillon>
RSpliet: there was some limitations in the mtdpart system, the main one being that the mtdpart system does not callback the master layer to delegate part of the initialization
<bbrezillon>
and in case of per-partition suff we need to parse and store partition specific stuff (ECC, randomizer, ...)
<RSpliet>
bbrezillon: ok... I'm currently looking at mtdpart and ofpart... from what I can tell the partitions do have room for storing ECC information
<bbrezillon>
modifying the mtdpart layer to do that should be doable, but at the time I did it I wanted to test the feasibility, so I decided to go for tha nand_part approach
<RSpliet>
cool :-)
<RSpliet>
for U-boot I kind of have to
<bbrezillon>
RSpliet: no, we need to have a private pointer so that the NAND controller can store its own private info
<RSpliet>
bbrezillon: yes, hence I'm thinking of an "allocate eccinfo" hook in drivers that can allocate a container struct containing the nand_ecclayout
<RSpliet>
so you can use container_of to find your private data too
<libv>
NiteHawk: not that we would wish to promote banana pi here :)
<libv>
but then, it is not as if tom or benn profit from more future sales of cubietruck either
<NiteHawk>
i know that bpi is the black sheep around here :P
<libv>
one of several ;)
<NiteHawk>
which community device would you recommend currently? the lime2?
<libv>
yeah, OSHW :)
<libv>
plaes: i thought i removed that xulong crap
<libv>
or was that for yet another golddigger?
<NiteHawk>
the xulong copy&paste orgy left us with a lot of misleading pages / info - it's simply not straightened out all yet, i think
<lemonzest>
xulong is the orange pi people?
<NiteHawk>
yup
<lemonzest>
i have one of their Orange Pi 2 boards
<lemonzest>
Quad H3
<ganbold_>
how fast is it?
<lemonzest>
not had much chance to test it really
<lemonzest>
cores are 1.5GHz
<ganbold_>
I meant compared to RPI2 maybe
<lemonzest>
dunno, not tested them against each other yet
<ganbold_>
because orange pi mini 2 is like 25$
<lemonzest>
yeah, i got the non-mini with wifi
<lemonzest>
on the forums they are announcing a new "Orange PC" for $15
<ganbold_>
where?
<lemonzest>
lemnmie find it
<ganbold_>
I know odroidc1 which has amlogic SoC is fast
<lemonzest>
I have a C1, RPi2, Orange Pi 2, A20-OLINUXINO-LIME2 and a Banana Pro
<ganbold_>
ok, 15$ nice
<lemonzest>
hardly any support for current models tho, so dunno why they keep pumping out hardware
sehraf has joined #linux-sunxi
<libv>
i am thinking about getting a c1, as it has mali-450
<libv>
but i doubt that there is much support for it
<libv>
hardkernel also is not the most proactive player
<ganbold_>
yeah, but at least it is fast little thing
<cubeast>
lemonzest: what's the maximum ethernet speed you can get on lime2?
<lemonzest>
cubeast: dunno, not tried it on ethernet yet, just small tests on the serial line
<ganbold_>
I was thinking to get orange pi mini 2, and canceled today the purchase so I did the right thing :)
reinforce has quit [Quit: Leaving.]
reinforce has joined #linux-sunxi
<bbrezillon>
RSpliet: I'm more in favor of an mtdparse hook
<bbrezillon>
and a private field in the mtdpart
<bbrezillon>
remember that per partition scrambler/randomizer config is needed too
philippe_fouquet has joined #linux-sunxi
<RSpliet>
I guess it's both the same to me
<RSpliet>
any reason why you favour this?
adj__ has joined #linux-sunxi
sunxi_fan1 has joined #linux-sunxi
arokux1 has joined #linux-sunxi
pstef has quit [Ping timeout: 250 seconds]
adj_ has quit [Ping timeout: 250 seconds]
sunxi_fan has quit [Ping timeout: 250 seconds]
arossdotme has quit [Ping timeout: 250 seconds]
arokux has quit [Ping timeout: 250 seconds]
<philippe_fouquet>
Hi
<philippe_fouquet>
Somebody have some doc in flash ecc on SoC A20
<RSpliet>
that's good to know
arossdotme has joined #linux-sunxi
<philippe_fouquet>
I try to read back the flash that be wrote with phoenix tool before the data are good in mem but the ecc is falt
cajg has quit [Ping timeout: 276 seconds]
<philippe_fouquet>
I try with/without random, seed to 0x4a80 and some other value but alway the same result data are reading but ecc is fail
bbrezill1 has joined #linux-sunxi
dlan_ has joined #linux-sunxi
lemonzest_ has joined #linux-sunxi
bbrezillon has quit [Ping timeout: 258 seconds]
lemonzest has quit [Ping timeout: 258 seconds]
dlan has quit [Ping timeout: 258 seconds]
sehraf has quit [Ping timeout: 258 seconds]
speakman has quit [Ping timeout: 258 seconds]
iflema has joined #linux-sunxi
Gerwin_J has joined #linux-sunxi
reinforce has quit [Ping timeout: 276 seconds]
reinforce has joined #linux-sunxi
Gerwin_J has quit [Quit: Gerwin_J]
naobsd has joined #linux-sunxi
sehraf has joined #linux-sunxi
trepidaciousMBR2 has quit [Quit: trepidaciousMBR2]
Gerwin_J has joined #linux-sunxi
Gerwin_J has quit [Quit: Gerwin_J]
Gerwin_J has joined #linux-sunxi
trepidaciousMBR2 has joined #linux-sunxi
<bbrezill1>
RSpliet: because if we want to add per-partition stuff for controller specific features (and currently, the randomizer is controller specific) we need the hook to be as much generic as possible, and the NAND controller should be able to parse its own properties (I'm refering to the DT case, but I guess we could have similar mechanisms with other mtdpart parsers)
trepidaciousMBR2 has quit [Client Quit]
Gerwin_J has quit [Ping timeout: 264 seconds]
<bbrezill1>
RSpliet: IIUC, you're suggesting to fill the ecc related fields in the mtd device exposed by the mtdpart, but remember that the controller driver has to initiliaze things accordingly if the ECC constraints have changed
arossdotme has quit [Ping timeout: 252 seconds]
Black_Horseman has quit [Remote host closed the connection]
trepidaciousMBR2 has joined #linux-sunxi
arossdotme has joined #linux-sunxi
<montjoie>
hello, does someone work on A83T ?
cubeast has quit [Quit: Leaving]
<RSpliet>
bbrezil1: yes... well, one step back: you think ofpart needs to hook into nand? I think it's easier to post-process from the sunxi-nand driver
<bbrezill1>
RSpliet: but how will you retrieve the created mtdparts ?
<bbrezill1>
and this also mean parsing the DT twice
<RSpliet>
that's not a big problem, is it? it's a one-time operation
<bbrezill1>
and with your solution you'll have to iterate over the MTD partitions each time you need to retrieve the priv field
<RSpliet>
yes, that's not so nice, but that's the downside of using a list for partitions instead of a hashmap
<bbrezill1>
and that's why I'm suggesting to store priv data in the partition itself, and to pass the mtd pointer of the partition and not the master one
<RSpliet>
yes, that would be the nicest solution, but also one that breaks legacy in ways I can't predict
oliv3r has quit [Ping timeout: 272 seconds]
<bbrezill1>
RSpliet: okay
<bbrezill1>
RSpliet: BTW, did you manage to boot from an SLC chip ?
kaspter has quit [Remote host closed the connection]
<RSpliet>
I didn't
<RSpliet>
we decided to focus on MLC for now because it... well, works
kaspter has joined #linux-sunxi
<RSpliet>
bbrezil1: yes, the proper solution is going to be messy because partitions mtd struct have their size set
<RSpliet>
and the read functions don't take the offset of a partiition into account when verifying against out-of-bound reads/writes
kaspter has quit [Ping timeout: 245 seconds]
flyhorse_YwjdW has joined #linux-sunxi
flyhorse has quit [Read error: Connection reset by peer]
oliv3r has joined #linux-sunxi
<bbrezill1>
RSpliet: yes interfacing with MLC works, but I wouldn't use it on a real product without some modification to the NAND, UBI and UBIFS layers
JohnDoe71rus has quit [Quit: Miranda IM! Smaller, Faster, Easier. http://miranda-im.org]
<bbrezill1>
RSpliet: if it's all about testing stuff, then yes it should work
<RSpliet>
bbrezill1: people keep saying that, but nobody is able to explain to me why :-) but if I set the maximum erase count diff in kernel to something low and use the randomiser the way Allwinner does, what are the major drawbacks>
flyhorse has joined #linux-sunxi
iflema has quit [Quit: WeeChat 1.0.1]
flyhorse_YwjdW has quit [Ping timeout: 250 seconds]
<bbrezill1>
RSpliet: and if you did read it, but still don't understand why MLC need special care, then maybe I can explain the parts you don't understand
<RSpliet>
you haven't, thanks
trepidaciousMBR2 has quit [Ping timeout: 255 seconds]
flyhorse_yxMLL has joined #linux-sunxi
trepidaciousMBR2 has joined #linux-sunxi
Gerwin_J has joined #linux-sunxi
flyhorse has quit [Read error: Connection reset by peer]
Gerwin_J has quit [Ping timeout: 264 seconds]
Gerwin_J has joined #linux-sunxi
<RSpliet>
bbrezill1: thanks... So voltage threshold adjustment needs work and it needs page scheduling for optimal perf
<RSpliet>
(where perf includes durability etc.)
reinforce has quit [Ping timeout: 276 seconds]
flyhorse has joined #linux-sunxi
flyhorse_yxMLL has quit [Read error: Connection reset by peer]
reinforce has joined #linux-sunxi
<bbrezill1>
RSpliet: but that's not the biggest problem with MLC, corruption after power-cut is
<bbrezill1>
RSpliet: what do you mean by page scheduling ?
diego_r has joined #linux-sunxi
Gerwin_J has quit [Ping timeout: 264 seconds]
<RSpliet>
your presentation mentions paired pages. If you want to increase reliability with that property in mind you'd need to do some sort of page scheduling I reckon
<RSpliet>
well, scheduling I guess is temporal, while you want the planning to be spatial... but you catch my drift I hope
sehraf has quit [Read error: Connection reset by peer]
sehraf has joined #linux-sunxi
arossdotme has quit [Ping timeout: 256 seconds]
prz has quit [Quit: Leaving.]
Gerwin_J has joined #linux-sunxi
arossdotme has joined #linux-sunxi
<philippe_fouquet>
bbrezill1: I see your path for booting on nand It work for an A20 SoC and how you flash it
mawe242 has quit [Remote host closed the connection]
iflema has quit [Ping timeout: 258 seconds]
philippe_fouquet has quit [Remote host closed the connection]
domidumont has quit [Quit: Leaving.]
trepidaciousMBR2 has quit [Quit: trepidaciousMBR2]
trepidaciousMBR2 has joined #linux-sunxi
bonbons has joined #linux-sunxi
trepidaciousMBR2 has quit [Client Quit]
<wens>
interesting, allwinner's shares are going up 10% (capped) everyday
<SQUelcher>
Make Money Fast?
<wens>
oh, they just listed last month
<wens>
the stock exchange caps price fluctuations at 10% per day
<wens>
so it's just the market price catching up with estimated value
ganbold_ has quit [Remote host closed the connection]
ganbold_ has joined #linux-sunxi
<wens>
anyone using systemd? can't get my debian systemd to automatically create a login console on ttyS0 :(
lemonzest_ has quit [Ping timeout: 246 seconds]
awe00 has quit [Ping timeout: 264 seconds]
<jackdaniel>
wens: enable fhandle in kernel
<jackdaniel>
or use runit
<jackdaniel>
:)
<wens>
it is enabled :(
<jackdaniel>
oh
<jackdaniel>
then no idea
<wens>
systemd loads fine, just doesn't create a console on ttyS0
<jackdaniel>
yes, that's why I suspected lack of fhandle - it refused to create console after succesfull loading
<wens>
just dropping back to sysvinit :|
<wens>
rather spend my time testing patches than fixing this