boycottg00gle has quit [Remote host closed the connection]
tuliom has quit [Quit: Konversation terminated!]
cnxsoft has joined #arm-netbook
rz2k has quit []
Quarx has joined #arm-netbook
Kebianizao|work has joined #arm-netbook
ssvb has quit [Remote host closed the connection]
panda84kde has joined #arm-netbook
_massi_ has joined #arm-netbook
cnxsoft has quit [Ping timeout: 252 seconds]
cnxsoft has joined #arm-netbook
popolon has joined #arm-netbook
popolon has quit [Changing host]
popolon has joined #arm-netbook
theborger has quit [Ping timeout: 240 seconds]
hramrach has quit [Ping timeout: 240 seconds]
theborger has joined #arm-netbook
uwe_ has quit [Ping timeout: 272 seconds]
atiti has joined #arm-netbook
uwe_mobile has joined #arm-netbook
uwe_ has joined #arm-netbook
cnxsoft has quit [Ping timeout: 252 seconds]
torindel has joined #arm-netbook
diego has joined #arm-netbook
diego is now known as Guest88545
torindel_ has joined #arm-netbook
popolon_ has joined #arm-netbook
Gujs has joined #arm-netbook
nashpa_ has joined #arm-netbook
Mazon_ has joined #arm-netbook
Gujs_ has quit [Ping timeout: 272 seconds]
Mazon has quit [Ping timeout: 272 seconds]
torindel has quit [Ping timeout: 272 seconds]
popolon has quit [Ping timeout: 272 seconds]
panda84kde has quit [Ping timeout: 272 seconds]
nashpa has quit [Ping timeout: 272 seconds]
Mazon_ is now known as Mazon
nashpa_ is now known as nashpa
torindel_ is now known as torindel
hero100 has joined #arm-netbook
<oliv3r>
lkcl: a few days have passed; so i'll try to ask the question again :)
<oliv3r>
lkcl: in u-boot currently we have MMC=3 set in boards.cfg; but as far as I know, this is not valid, the BROM cannot boot from MMC3, only MMC2 and thus this statement won't work
<oliv3r>
it WILL work, IF you manage to store the SPL elsewhere (nand, SPI, mmc0) which loads whatever is on mmc3, but that's really not what you'd want
<oliv3r>
I think MMC=2 should be in boards.cfg which yes, means you can only board from mmc2 with those pins actually connected
hero100 has left #arm-netbook ["Leaving"]
popolon_ has quit [Quit: #freenode]
popolon has joined #arm-netbook
popolon has joined #arm-netbook
popolon has quit [Changing host]
popolon has quit [Read error: Connection reset by peer]
popolon has joined #arm-netbook
acassis has joined #arm-netbook
Guest88545 is now known as panda84kde
acassis has quit [Ping timeout: 272 seconds]
acassis has joined #arm-netbook
Kebianizao|work has quit [Remote host closed the connection]
rz2k has joined #arm-netbook
slapin has quit [*.net *.split]
slapin has joined #arm-netbook
fredy has quit [Excess Flood]
fredy has joined #arm-netbook
cnxsoft has joined #arm-netbook
popolon has quit [Ping timeout: 264 seconds]
traeak has joined #arm-netbook
traeak has left #arm-netbook [#arm-netbook]
popolon has joined #arm-netbook
popolon has joined #arm-netbook
popolon has quit [Changing host]
slapin has quit [Ping timeout: 272 seconds]
cnxsoft has quit [Quit: cnxsoft]
slapin has joined #arm-netbook
Mr_Electronico has joined #arm-netbook
fredy has quit [Excess Flood]
fredy has joined #arm-netbook
Mr_Electronico has quit [Quit: Leaving]
MMlosh has joined #arm-netbook
<lkcl>
oliv3r: yep if you're asking "should that be changed" please go ahead.
<lkcl>
oliv3r: there will only be one user boot method: NAND.
<lkcl>
oliv3r: so now that i think about it, it really doesn't matter. the SPL will only be on NAND. when doing factory-install that is a completely different matter, and booting will only be done from "stock" anyway.
<lkcl>
probably USB-FEX
<lkcl>
but maybe mm32.
<lkcl>
s/mm32/mmc2/
<infobot>
lkcl meant: but maybe mmc2.
<lkcl>
however, once loaded from NAND, yes booting should be possible from SDcard, which is MMC *3*.
<lkcl>
which is the micro-sd card slot on the front.
<lkcl>
so if you are asking "should u-boot try to load the linux kernel from the 68-pin header mmc2" the answer is ABSOLUTELY NOT.
<lkcl>
if you're asking "should u-boot try to load the linux kernel from mmc3 by default" the answer is YES
<lkcl>
if you're asking "should u-boot/spl try to load from the 68-pin header mmc2" the answer is ABSOLUTELY NOT.
<lkcl>
if you're asking "should u-boot/spl try to load script.fex and u-boot from the front micro-sd card slot (mmc3)" the answer is YES.
<lkcl>
in short, MMC=3 is correct.
<lkcl>
that's my understanding, anyway.
ssvb has joined #arm-netbook
panda84kde has quit [Ping timeout: 260 seconds]
Quarx has quit [Read error: Connection reset by peer]
fredy has quit [Excess Flood]
fredy has joined #arm-netbook
greyfire has joined #arm-netbook
greyfire has quit [Remote host closed the connection]
_massi_ has quit [Remote host closed the connection]
acassis has quit [Quit: Leaving.]
replexer has joined #arm-netbook
replexer has quit [Remote host closed the connection]
dfletcher has quit [Read error: Connection reset by peer]
ssvb has quit [Ping timeout: 252 seconds]
nashpa has left #arm-netbook [#arm-netbook]
dfletcher has joined #arm-netbook
<oliv3r>
lkcl: i think the mmc= option applies to the SPL only though
<oliv3r>
so shall the SPL try to load u-boot from MMC=
<oliv3r>
since u-boot can boot from whatever I thought
<oliv3r>
anyway, just gotta make sure that we dot the i's and cross the t's here
ssvb has joined #arm-netbook
eFfeM has joined #arm-netbook
<lkcl>
oliv3r: SPL should be attempting to load from MMC=3 because that would mean that anyone putting a boot-compatible MicroSD card into the front slot stands a chance of it working
<lkcl>
if SPL has been loaded onto the NAND
<lkcl>
... or uploaded to DDR3 RAM using felboot
<lkcl>
that's the most useful thing.
<lkcl>
for "factory" use it would... i know what: if you'd be happy to add an extra line to the u-boot boards.cfg ... say.... i dunno... EOMA68-A20-FACTORYSPL or something, _that_ should have MMC=2
<lkcl>
because only at the factory would we make a special engineering board to boot off of the 2nd MMC.
<lkcl>
but, to be honest, that's not needed
<lkcl>
becauuuuse, if you boot off of MMC2 (from the 68 pins) we would make it a boot0/boot1 system *anyway*.
<lkcl>
so yeah don't worry about it.
<lkcl>
MMC=3 is absolutely fine.
techn_ has quit [Ping timeout: 272 seconds]
<lkcl>
btw does anyone know how to put SD/MMC on the A20 into "SPI compatibility" mode? or is it automatic?
techn_ has joined #arm-netbook
<oliv3r>
lkcl: you can't upload to ddr via fel; only to sram :)
<oliv3r>
lkcl: it really really sucks they messed up the mmc connector on the board :(
techn__ has joined #arm-netbook
techn_ has quit [Ping timeout: 265 seconds]
<lkcl>
oliv3r: ok ok sram, that's fiiine :)
<lkcl>
oliv3r: *sigh* it's to do with the orientation of the CPU. wits-tech turned it through 180 degrees without consulting me, which is why the ethernet PHY is in a truly arse-about-face corner.
<lkcl>
they did a really _really_ beautiful and i truly mean that job on the DDR3 RAM wiring
<lkcl>
_but_ because they also only did 3 signal layers, there simply wasn't enough room to cross over MMC0 to the other side of the board
<lkcl>
so... _again_ without consulting me they dropped MMC0, put MMC3 through to the MicroSD slot and put MMC2 onto the EOMA68 header
<oliv3r>
really shit
<lkcl>
and then proceeded to fuck everything up by adding in MMC pull-up resistors onto the EOMA68 GPIO lines
<oliv3r>
will it ever get 'fixed' in a v2 so to say?
<oliv3r>
(new soc)
<lkcl>
etc. etc.
<lkcl>
we'll really need to do a new revision, use those lower-cost x16 DDR3 RAM ICs that tom uses on the cubieboards
<lkcl>
lower-cost _and_ smaller, give us a bit more space
<oliv3r>
very cool
<lkcl>
and upgrade to 2gbyte
<lkcl>
and use a GbE PHY
<lkcl>
but the SD/MMC? weeelll, to be honest, after i got over freaking out and jumping up and down, i think it really works.
<oliv3r>
yeah sounds sweet :)
<lkcl>
but we have to sell these cards first!
<oliv3r>
well if they re-do the board
<oliv3r>
make sure they do it proper ;)
<lkcl>
*grins* yeahhh.... it'll be another $12k though. we need to sell boards first.
<lkcl>
either that or i need to learn how to do DDR3 RAM routing.
<oliv3r>
heh
<oliv3r>
ddr ram routing is a bitch :(
* lkcl
shudders
<lkcl>
it's fiiiine!
<lkcl>
btw when i spoke to allwinner's tech director several months back i mentioned several things, including that we _really_ want to see a 28nm quad-core version of the A20.
<lkcl>
he mentioned that he was concerned about MALI 400 overloading the main CPU: ARM's MALI driver really is pretty shit.
<lkcl>
i reassured him about libv's work, that there are performance increases possible and that it's really a software issue
<oliv3r>
a23 si pretty cool actually
<lkcl>
and also suggested they look at both vivante and tegra. he was _really_ intrigued to hear that tegra are licensing their GPU stuff
<lkcl>
oliv3r: what's an A23? :)
<lkcl>
*sigh* just looked it up...
<oliv3r>
552 mhz ram
<oliv3r>
in a OEM tablet
<oliv3r>
1.4ghz cpu
<oliv3r>
and there might be a quad version
<lkcl>
gooood.
<oliv3r>
but you may know more then we
<oliv3r>
we only got a23 SDK
<oliv3r>
and a tablet 'in the wild'
<oliv3r>
but no sata, no hdmi etc
<oliv3r>
tablet soc :(
<lkcl>
my associate will be calling eva in the morning.
<lkcl>
if they've done the same thing that they did with the A10/A10s, they *might* have just used the same die but not brought out some of the pins.
<lkcl>
i gotta make a phone call .... :)
<oliv3r>
i invited eva to fosdem
<oliv3r>
haven't heard anything
<lkcl>
the call's in the morning
<lkcl>
oh right!
<lkcl>
hey that would be cool.
<oliv3r>
so double invite her :)
<lkcl>
oliv3r: i have to make a phone call so i can catch my associate for his call tomorrow morning ok?
<lkcl>
gotta go
rz2k has quit [Read error: Connection reset by peer]
<lkcl>
oliv3r: ok, thank you (a lot!) - my associate's going to ask about the A23, quad-core versions etc. tomorrow
acassis has joined #arm-netbook
<oliv3r>
very cool
<oliv3r>
and invite eva to fosdem :)
<oliv3r>
but no sata/hdmi is sad :(
lerc has quit [Read error: Connection reset by peer]
ganbold_ has quit [Ping timeout: 245 seconds]
lerc has joined #arm-netbook
eFfeM has quit [Remote host closed the connection]