viccuad has quit [Read error: Connection reset by peer]
FreezingAlt has joined #linux-sunxi
FreezingCold has quit [Ping timeout: 255 seconds]
<wens>
arokux: it has a few phys i think, but one common interface to access them :/
<wens>
hno: there's no point in supporting gmac on sun8i anyway, the pins aren't routed
<wens>
arokux: and that's what hans' usb phy driver does
<wens>
for musb, the glue layer hardware is accessed from 0x400 in the usbotg address range
<wens>
while the usb phys are at 0x404 or 0x408 (can't remember)
<wens>
i've documented what i can from allwinner's code
wenbin has joined #linux-sunxi
Renard has quit [Remote host closed the connection]
<wens>
i suggest either writing a usb-phy driver to use with musb, or export the registers using syscon
<wens>
and that is about as far as i got
<wens>
interaction between musb and phy is tricky, not to mention musb is poorly documented
ricardocrudo has quit [Ping timeout: 272 seconds]
Gerwin_J has quit [Quit: Gerwin_J]
TheSeven has quit [Ping timeout: 260 seconds]
TheSeven has joined #linux-sunxi
FreezingAlt is now known as FreezingCold
maksimlin has joined #linux-sunxi
tomcheng76 has quit [Quit: leaving]
tomcheng76 has joined #linux-sunxi
JohnDoe_71Rus has joined #linux-sunxi
wingrime has joined #linux-sunxi
wingrime1 has joined #linux-sunxi
wingrime has quit [Read error: Connection reset by peer]
FreezingCold has quit [Ping timeout: 240 seconds]
FreezingCold has joined #linux-sunxi
maksimlin has quit [Ping timeout: 258 seconds]
Quarx has joined #linux-sunxi
ganbold_ has quit [Ping timeout: 255 seconds]
philippe_fouquet has joined #linux-sunxi
Tintu has joined #linux-sunxi
<Tintu>
hello everyone...
<focus>
anyone with idea where best to get custom color LCD made? the ones i tried had moq of 100k. need someone who can settle for 10k. size 75x57mm 320x240 dots or better - to be driven by a cubieboard
<Tintu>
i want to know how fex file selects the driver file for a particular device... Pls hep me..
<Tintu>
help me..
<focus>
fex file tells the firmware built into the A10/A20 to configure the hardware
<afaerber>
Tintu, afaiu it doesn't select drivers. the drivers read from the .fex file to initialize themselves
<focus>
the driver is enabled when Linux is configured and compiled
wingrime has joined #linux-sunxi
wingrime1 has quit [Read error: Connection reset by peer]
<focus>
Tintu: the A10/A20 has firmware built into the chip at the time it ships - that firmware reads the fex file to initialise the internal hardware such as LCDs, i/o ports and so on
<wens>
i don't really use 3.4, so can't help you much
<Tintu>
yes i compiled...
_massi has joined #linux-sunxi
<Tintu>
added it manually...
<Tintu>
but an error pop-ups..
<Tintu>
system UI has stopped..
<Tintu>
also touch isn't working..
<wens>
can't help you :|
<afaerber>
hipboi, cool! any info on who manufactures that board? Merrii again?
<jelly-home>
hipboi: and 5GHz wifi, nice
<Tintu>
afaerber: yup Merrii..
<afaerber>
Tintu, hipboi: so availability of board or of chip is 2015 Q4? don't see the board mentioned alongside Optimus neither on the English nor on the Chinese page on merrii.com yet
<afaerber>
not to mention AliExpress ;)
<Tintu>
sry...i dont understand what u are telling.. if u are asking about board.. based on A20 processor .. linux kernel 3.4
<gzamboni>
wow, 64 bits ARM from allwinner
<wens>
Tintu: afaerber is refering to the link hipboi pasted
<hno>
hipboi, wtf... first 64-bit dev board and they equip the board with only 2GB RAM? The whole point of going 64-bit is to easily support 4+GB RAM. For smaller configurations it's quite pointless or even counterproductive to go 64-bit.
sehraf has joined #linux-sunxi
<Tintu>
wens: ok..
<hipboi>
i think the first 64 bit from Chinese soc maker targets at low end
<hipboi>
Actions's 64bit soc is also not high end
Tintu has quit [Ping timeout: 246 seconds]
<wens>
doesn't say which core it uses
<jelly-home>
hno: apple's gone 64bit and claims speed and power efficiency even if iPhone 6 doesn't have 4GB RAM, either
<jelly-home>
tho their soc engineers are probably a tiny bit better
<jelly-home>
oh, 5s also has a 64bit cpu
<libv>
jelly-home: did you miss that?
<jelly-home>
apparently :-)
<libv>
jelly-home: that and the fact that it was the first to ship the rogue (img pvr6) was the only thing worth talking about with the iphone 5s
Tintu has joined #linux-sunxi
<libv>
jelly-home: _everyone_ was mightilly surprised
<libv>
even ARM
<libv>
and especially samsung
leviathanch2 has joined #linux-sunxi
Tintu has quit [Client Quit]
<libv>
who, one week later, made a statement "we too will do 64bit... soon"
<libv>
samsung had been the launch partner for anything arm for several years
<libv>
and there apple went "oh, and by the way, we are 64bit now."
<libv>
i am sure that some pretty enraged telcos took place in the week after that
<libv>
so it was rather big news
leviathanch2 has quit [Read error: Connection reset by peer]
leio has joined #linux-sunxi
leviathanch2 has joined #linux-sunxi
<jelly-home>
so it wasn't quite pointless
<libv>
it was quite the technological leap forward, but it gained the consumers nothing
leviathanch2 has quit [Ping timeout: 246 seconds]
rafaelMOD has quit [Quit: Leaving]
arokux2 has joined #linux-sunxi
<arokux2>
wens: what do you mean poorly documented, is there anything except of the source code? :)
<wens>
arokux2: there is only source code, but even the source code is bad :(
<arokux2>
wens: right.
<arokux2>
wens: if there is Hans' phy driver why do we need to write another one?
<wens>
that phy driver only does usb phy initialization
<wens>
you need glue code for otg to handle stuff like id pin detection and pull-ups (some signals to the musb core are not routed out, but controlled by the glue layer)
<wens>
anyway it's a mess
leviathanch2 has joined #linux-sunxi
<arokux2>
wens: so why not extending the Hans' driver by adding needed features?
<wens>
arokux2: extending or not extending doesn't matter
<wens>
it's code that does not exist
<arokux2>
wens: sorry, I'm just trying to understand whether there is a new usb phy driver needed or the existing one can be extended...
<wens>
probably... i'm not sure whether the init sequence for the regular phys applys to otg
<wens>
applies
imcsk8 has quit [Quit: Reconnecting]
imcsk8 has joined #linux-sunxi
cubear has joined #linux-sunxi
FR^2 has joined #linux-sunxi
<arokux2>
wens: I see, thx
popolon has joined #linux-sunxi
FDCX_ has joined #linux-sunxi
FDCX__ has quit [Ping timeout: 246 seconds]
marcin_ has joined #linux-sunxi
FDCX has joined #linux-sunxi
FDCX_ has quit [Ping timeout: 272 seconds]
alexvf has joined #linux-sunxi
FDCX has quit [Ping timeout: 244 seconds]
leviathanch2 has quit [Ping timeout: 255 seconds]
FDCX has joined #linux-sunxi
leviathanch2 has joined #linux-sunxi
<mripard>
afaerber: you're at ELCE?
<afaerber>
mripard, will be at KVM Forum tomorrow, also in Düsseldorf
afaerber has quit [Quit: Verlassend]
leviathanch2 has quit [Ping timeout: 255 seconds]
megal0maniac is now known as Guest52927
Guest52927 has quit [Killed (weber.freenode.net (Nickname regained by services))]
megal0maniac has joined #linux-sunxi
indy has quit [Excess Flood]
indy has joined #linux-sunxi
megal0maniac has quit [Killed (sinisalo.freenode.net (Nickname regained by services))]
wingrime1 has quit [Read error: Connection reset by peer]
megal0maniac has joined #linux-sunxi
wingrime has joined #linux-sunxi
wingrime has quit [Read error: Connection reset by peer]
wingrime has joined #linux-sunxi
megal0maniac is now known as Guest3808
Guest3808 has quit [Killed (card.freenode.net (Nickname regained by services))]
megal0maniac has joined #linux-sunxi
wingrime1 has joined #linux-sunxi
wingrime has quit [Read error: Connection reset by peer]
megal0maniac has quit [Ping timeout: 244 seconds]
megal0maniac has joined #linux-sunxi
deasy has joined #linux-sunxi
megal0maniac is now known as Guest75919
Guest75919 has quit [Killed (morgan.freenode.net (Nickname regained by services))]
megal0maniac has joined #linux-sunxi
wenbin has quit [Read error: Connection reset by peer]
viccuad has joined #linux-sunxi
JohnDoe_71Rus has quit [Quit: Miranda IM! Smaller, Faster, Easier. http://miranda-im.org]
FDCX has quit [Ping timeout: 260 seconds]
viccuad has quit [Read error: Connection reset by peer]
FDCX has joined #linux-sunxi
leviathanch2 has joined #linux-sunxi
Renard has joined #linux-sunxi
Tintu has joined #linux-sunxi
<Tintu>
Hello... I'm trying to attach a rtp with A20 hummingbird... I configured the fexfile as described in fex guide.. But still it is not working... what will be the reason? can u help me pls...
Andy-D has joined #linux-sunxi
philippe_fouquet has quit [Remote host closed the connection]
F1skr has joined #linux-sunxi
wingrime1 has quit [Read error: Connection reset by peer]
wingrime has joined #linux-sunxi
leviathanch2 has quit [Ping timeout: 272 seconds]
wingrime1 has joined #linux-sunxi
wingrime has quit [Read error: Connection reset by peer]
wingrime has joined #linux-sunxi
wingrime1 has quit [Read error: Connection reset by peer]
Black_Horseman has quit [Remote host closed the connection]
Tintu has quit [Ping timeout: 246 seconds]
megal0maniac has quit [Ping timeout: 258 seconds]
megal0maniac has joined #linux-sunxi
megal0maniac has quit [Killed (sinisalo.freenode.net (Nickname regained by services))]
megal0maniac has joined #linux-sunxi
leviathanch2 has joined #linux-sunxi
Andy-D has quit [Read error: Connection reset by peer]
Uninstall has quit [Remote host closed the connection]
megal0maniac is now known as Guest70809
megal0maniac has joined #linux-sunxi
Guest70809 has quit [Ping timeout: 260 seconds]
Uninstall has joined #linux-sunxi
leio has quit [Quit: No Ping reply in 180 seconds.]
Andy-D has joined #linux-sunxi
Nyuutwo has quit [Quit: No Ping reply in 180 seconds.]
GeertJohan has quit [Remote host closed the connection]
RaYmAn has quit [Ping timeout: 245 seconds]
ssvb has quit [Ping timeout: 245 seconds]
slapin has quit [Ping timeout: 240 seconds]
leio has joined #linux-sunxi
RaYmAn has joined #linux-sunxi
Swabbles has quit [Ping timeout: 245 seconds]
ricardocrudo has joined #linux-sunxi
slapin has joined #linux-sunxi
Nyuutwo has joined #linux-sunxi
wigyori has quit [Ping timeout: 240 seconds]
Swabbles has joined #linux-sunxi
wigyori has joined #linux-sunxi
ssvb has joined #linux-sunxi
GeertJohan has joined #linux-sunxi
arokux2 has quit [Remote host closed the connection]
viccuad has joined #linux-sunxi
skaag has quit [Ping timeout: 255 seconds]
<rellla>
oh puneet ...
skoperst1_ has joined #linux-sunxi
hipboi has quit [Ping timeout: 260 seconds]
deasy has quit [Quit: Nom d'un quark, c'est Edmonton !]
cubear has quit [Remote host closed the connection]
<hno>
jelly-home, those speed & efficiency improvements are not from going 64-bit, but from going a leap forward on the CPU architecture in general, at which point it only makes sense to also go 64-bit to prepare for next year.
<linkmauve1>
From what I’ve heard about AArch64, the ISA is much nicer than the one used until ARMv7 (and still mandatory on ARMv8) for about everything.
leviathanch2 has quit [Ping timeout: 272 seconds]
<hno>
libv, yes it's interesting when an architecture licensor is faster to launch an implementation of the new architeccture than anyone using the reference implementation from the architecture vendor..
hipboi has joined #linux-sunxi
rz2k_ has joined #linux-sunxi
<hno>
linkmauve1, there is a whole lot more register space available, and less confusing instruction variants. Code gets a bit bigger from less special case instructions, but much more straight-forward to compile to.
<linkmauve1>
That’s what I heard too, but I haven’t touched it myself yet.
<libv>
hno: whether it actually is fully compatible with the full architecture, no-one knows
<libv>
but they did catch everyone out, and they left both arm and samsung reeling
<hno>
libv, which is good.
skaag has joined #linux-sunxi
<libv>
is it?
<libv>
the apple thing was just a meaningless trump, as it had no effect on sales
<libv>
and you can be very certain that ARM will now enforce a lot stricter set of policies with its partners and internally
<libv>
arm probably lost a lot of samsung "exclusive launchpartner" money
<libv>
and seriously hurt its relationship with samsung
<libv>
the effects of that will never really be known to us
<libv>
but it will not have made things easier inside arm
<libv>
or samsung
<libv>
and it probably just delays things even more
<libv>
and what did we gain from the apple announcement: nothing.
<libv>
so i fail to see anything good about this, the effects will be neutral or slightly negative for us.
megal0maniac has quit [Ping timeout: 260 seconds]
megal0maniac has joined #linux-sunxi
ganbold__ has joined #linux-sunxi
<hno>
libv, Apple licenses the architecture under a quite special deal, not ARMs implementation of it. Samsung exclusive launch partner is only on Cortex implementation.
<libv>
hno: there are several architecture licensees
<libv>
and samsung also seems to have been the launch partner for many of the mali releases
esperegu has joined #linux-sunxi
<libv>
in any case, when 2 big companies get trumped like this, measures are taken
megal0maniac has quit [Ping timeout: 250 seconds]
<libv>
processes will have become more involved
<libv>
and i see no advantage to outsiders
ganbold_ has quit [Ping timeout: 255 seconds]
deasy has joined #linux-sunxi
megal0maniac has joined #linux-sunxi
sehraf has quit [Read error: Connection reset by peer]
sehraf has joined #linux-sunxi
<hno>
libv, sure, but the Apple license is a bit different than the other, and so are the size of their CPU design group. And it has been known since the release of A6 that the A7 will be ARMv8 with expected release within 2014.
megal0maniac has quit [Killed (sinisalo.freenode.net (Nickname regained by services))]
megal0maniac has joined #linux-sunxi
<libv>
hno: you can google the surprise responses from both arm and samsung
<libv>
they genuinely had no idea that this was going to happen
<libv>
perhaps only the arm accountmanager(s) for apple knew
<libv>
and you can bet that he/they got a serious asskicking now
leviathanch2 has joined #linux-sunxi
<libv>
and it's been a whole year
<libv>
and noone else has actually shipped aa64 (nothing fully working anyway)
<libv>
hah
<libv>
galaxy note 4 is supposedly aarch64
<libv>
released about a month ago and actually for sale
<libv>
but it only runs 32bit code today
megal0maniac has quit [Ping timeout: 260 seconds]
<libv>
ah, will be for sale in .de in about a weeks time
<hno>
libv, my googlefu must be bad today. Only find a good one on Qualcomm panic, not much about Samsung. A lot about Samsung panicing about Iphone 6 in general, but not the A7..
<libv>
hno: i would have to try to dig out my linkedin posts from a year ago
<libv>
i am sure i could dig out the arm one if i waste enough time
<libv>
but what is the point?
wingrime has quit [Read error: Connection reset by peer]
<libv>
hno: do you really believe that apple using this marketing statement did anything good for anyone
wingrime has joined #linux-sunxi
<libv>
do you feel the need to disprove me for stating that this will not have benefitted any of us, at best, it made communication and cooperation more difficult between arm and its partners?
<libv>
5433 is not really shipping yet, and will not run aarch64 software
<libv>
one year later
<arokux>
libv: who is designing the 64-bit cores for Apple?
<libv>
arokux: i am sure that wikipedia can inform you about that
<arokux>
:p
philippe_fouquet has joined #linux-sunxi
<libv>
hno: if this sort of disruption ever led to any good, and not to the usual corporate rebound, don't you think i would be either intels, amds, or arms open source graphics driver architect today?
wickwire has joined #linux-sunxi
megal0maniac is now known as Guest30543
Guest30543 has quit [Killed (cameron.freenode.net (Nickname regained by services))]
megal0maniac has joined #linux-sunxi
skaag has quit [Ping timeout: 246 seconds]
philippe_fouquet has quit [Remote host closed the connection]
FreezingCold has quit [Ping timeout: 272 seconds]
leviathanch2 has joined #linux-sunxi
<arokux>
libv: so I've found out that it was manufactured by samsung. I've kind of assumed that. now: why is samsung surprised about this then? :)
<libv>
ah, you mean, as in fabbed.
<libv>
arokux: do you really think anyone from the chip designers knows what is being pushed through the fabs at any time?
enrico_ has joined #linux-sunxi
<libv>
plus, this is korea, i doubt that anyone circumvents the command chain and drops anyone a hint
<arokux>
libv: I have no idea, so I'm asking... so the samsung factories in principle manufacture blackboxes for the competition?
<libv>
and by the time the chip was being turned into actual silicon, samsungs chip designers really would have no more time to react
<libv>
samsung seems to be very happy to make silicon for others
<libv>
so why not?
<libv>
these are huge separate business units in an even bigger company
<arokux>
I see
FreezingCold has joined #linux-sunxi
<merbanan>
if samsungs isn't making the chips someone else will, so why not take their money
<wens>
the newest chips are made by TSMC
nove has joined #linux-sunxi
leviathanch2 has quit [Ping timeout: 272 seconds]
xavia has joined #linux-sunxi
wingrime has quit [Read error: Connection reset by peer]
<hno>
arokux, Apple designs their own ARM core. And libv I very much doubt Samsung did not expect AArch64 support in the A7. The big suprise is that they made IOS AArch64.
<hno>
and yes, it's mostly a marketing hype until you start producing devices with 4+ GB of ram.
<WarheadsSE>
hno: Is it actually, or are they just running a 32bit userland?
<hno>
and yes, ofcourse Apple bragging about it hurts competition.
<libv>
hno: i just don't see anything "good" about apple trumping arm, samsung and qualcomm.
leviathanch2 has joined #linux-sunxi
pwhalen has quit [Ping timeout: 258 seconds]
<hno>
WarheadsSE, it looks like it does since 7.0.
skaag has joined #linux-sunxi
<hno>
My guess is that Apple when designing A7 put all their optimization efforts on AArch64, making A32 more of a compat layer and no AArch32 support at all, which actually makes a great sense in their position.
<hno>
AArch32 is about as useful as X32, i.e. a special optimization what will only be useful for a blink in time.
uwe_ has quit [Ping timeout: 260 seconds]
focus has quit [Remote host closed the connection]
focus has joined #linux-sunxi
leviathanch2 has quit [Ping timeout: 260 seconds]
pwhalen has joined #linux-sunxi
FreezingCold has quit [Read error: Connection reset by peer]
<linkmauve1>
hno, isn’t AArch32 the new name of the ARMv7 ISA?
FreezingCold has joined #linux-sunxi
uwe_ has joined #linux-sunxi
<linkmauve1>
And x32 *is* useful, I would use it anytime over x86_64 if its support was existant.
uwe_ has quit [Client Quit]
megal0maniac has quit [Killed (orwell.freenode.net (Nickname regained by services))]
uwe_ has joined #linux-sunxi
megal0maniac has joined #linux-sunxi
paulk-collins has joined #linux-sunxi
tomboy64 has quit [Ping timeout: 264 seconds]
<hno>
linkmauve1, AArch32 includes ARMv7 user-space compatibility yes.
<hno>
but also supports running a 32-bit OS.
<hno>
but differently from ARMv7.
megal0maniac is now known as Guest85381
Guest85381 has quit [Killed (wolfe.freenode.net (Nickname regained by services))]
megal0maniac has joined #linux-sunxi
<linkmauve1>
Oh, how different is that?
Renard has quit [Remote host closed the connection]
tomboy64 has joined #linux-sunxi
wickwire_ has joined #linux-sunxi
wickwire has quit [Ping timeout: 255 seconds]
megal0maniac has quit [Ping timeout: 272 seconds]
bonbons has joined #linux-sunxi
FreezingCold has quit [Read error: Connection reset by peer]
<hno>
linkmauve1, hmm.. reading again and AArch32 include all the oddness of ARMv7, so actually not that different if you have full AArch32 support.
<arokux>
why don't we hear anything from Tsvetan from Olimex?
<linkmauve1>
In my understanding, ARMv7’s ISA is called AArch32 on ARMv8, but it’s the exact same ISA.
<hno>
But if you only support ARMv7 userspace and not full AArch32 then you can drop a lot of quite complex things in interrupt handling and security levels.
<hno>
arokux, should we?
<arokux>
he used to complain about bugs now and then
<libv>
arokux: we still lack a .fex for the lime2
<arokux>
you see :)
<arokux>
hm.. but how can it be true? they should have their fex anyway
<hno>
arokux, just no one have bothered to follow the ndh for it and submit one.
<libv>
the ndh is almost complete
<libv>
we just need an olimex image for this device
<libv>
the header is there of course, but does not contain the meminfo
<hno>
Our u-boot SPL do not use the same header. Only what's needed for BROM to accept it. Have many times thought about adding the DRAM parameters there but...
Lorenzo64 has joined #linux-sunxi
<libv>
no worries, i'll grab the android one
<libv>
anyway, cool, we should soon have this board ndhed
<libv>
heh... "soon"
<hno>
there is dram parameters in sunxi u-boot already.
<libv>
oh, is there?
<hno>
yes, at least in the revision they built from which is in our git.
<hno>
But can't find in current version. Strange.
<libv>
i think i might have removed them, as nothing was known back then
<hno>
Yes you did.
<hno>
Killing a perfectly fine pull request from the vendor.
esperegu has quit [Ping timeout: 255 seconds]
esperegu has joined #linux-sunxi
<libv>
well, if this torrent is at all alive, we'll soon see what the hw truly ships with
<libv>
hno: plus, no fex file and no ndh page
<hno>
The debian torrent ships with those settings.
<hno>
was a clean build.
<libv>
ok, feel free to revert my commit and push it
<hno>
revision 20a24c3
<hno>
already doing..
<hno>
and it's nowhere close to what is in the fex..
netlynx has quit [Remote host closed the connection]
<libv>
*sigh*
<libv>
hno: hold off for a bit still, i'll try the other android img over bt and look into the actual img there
<libv>
oh, lime2 schematics have been uploaded it seems
blsd1 has joined #linux-sunxi
<libv>
no peers on either android image :(
<libv>
ah, now there's one
<hno>
libv, nah. It's pushed. But feel free to fix things up if you find a difference that makes sense.
<hno>
for me the Debian version is more authorative than the Android one..
<libv>
i am more worried about the difference between fex and the version that once was in our u-boot
<hno>
The u-boot settings makes a whole lot more sense for the board than the fex ones.
<libv>
ok
<libv>
but it is not usual for the olimex folks to leave something like that out
<libv>
anyway, feel free to remove the ndh status on it
hipboi has quit [Read error: Connection timed out]
hipboi has joined #linux-sunxi
<hno>
The device page still needs a bunch of edits.
FreezingCold has quit [Ping timeout: 244 seconds]
<libv>
yes, but all can be quickly resolved now
<hno>
libv, but imho it should have been taken out if that status long ago. Support was submitted mainline ages ago, and a device with mainline support is sufficient to pass.
<libv>
hno: not everyone wants to use mainline for everything today
<libv>
hno: and extracting the fex and adding the u-boot meminfo is not the hard part of ndhing
<libv>
imho those steps should still come first
<libv>
this situation was however very weird in that there was no fex until very very recently
<libv>
i think that that one user patch was just a week old
<libv>
where he tried to hack up his own fex from the original a20 lime
konradoo77 has quit [Ping timeout: 250 seconds]
andoma has quit [Ping timeout: 272 seconds]
<arokux2>
libv: ndhing?
<libv>
arokux2: which rock have you been under ;p
<libv>
new device howto
konradoo77 has joined #linux-sunxi
<arokux2>
libv: I've missed you awesome wiki work
andoma has joined #linux-sunxi
<arokux2>
your*
<hno>
libv, debian image looks like a livesuit image, so you need a board to get anything menaingful from it.
<libv>
but that will only work with an allwinner bootloader
<arokux2>
it is a long time since I've built everything. it is quite an endevour
<libv>
the proper nand android image will here shortly
<libv>
+be
FR^2 has joined #linux-sunxi
<hno>
libv, except the little detail that livesuit updates the dram_para setting in the header while flashing...
popolon has quit [Ping timeout: 260 seconds]
<ssvb>
hno: is the dram_para struct a part of some standard header read by BROM?
<hno>
it's read by boot0 & boot1. Each have a copy in their header.
<hno>
s/read/used/.
<libv>
hno: ok?
<ssvb>
we kinda started extending 'dram_para' in u-boot, so now this stuff may diverge from boot0/boot1
<hno>
ssvb, that's fine.
<hno>
as long as we know how to go from AW dram_para to what you need then we are fine.
<hno>
libv, ok what?
FreezingCold has joined #linux-sunxi
<ssvb>
hno: so far we are backwards compatible with AW dram_para and only have added extra optional tuning parameters
<hno>
ssvb, even better
popolon has joined #linux-sunxi
blsd1 has quit [Quit: Leaving.]
blsd1 has joined #linux-sunxi
<libv>
hno: ok, questionmark, as in, this is the first i hear of that, feel free to go more in depth
<hno>
it uses dram_para from the image as base setting, but validates them a bit as part of the same code that autodetects when things are not specified.
<hno>
and when satisfied writes the new settings to the boot headers.
<libv>
i have successfully extracted valid settings before
<ssvb>
diego71: looks like it has a good potential to run at very high dram clock speeds, just like Cubietruck
<libv>
hno: do you have any idea where livesuite gathers this information from?
<hno>
not sure, but the dram_para part is sent to the board during early USB bootstrap. The flashing procedure that runs on the board then copies this to the flashed image while programming the storage media.
<diego71>
ssvb: thx, meanwhile i'm testing the other olinuxino a20 micro...
blsd1 has quit [Remote host closed the connection]
<diego71>
ssvb: and is worse than :(
<diego71>
*than the other one
<libv>
ok, so our bsp does not implement this
<ssvb>
diego71: heh, I wonder if there is something wrong with DRAM routing in OLIMEX boards in general, this would make you a bit unlucky to have them
<hno>
libv, implement what?
<diego71>
ssvb: and this one have different dram chips
<hno>
the same magics that AW applies while livesuit flashing a board?
<ssvb>
diego71: well, DRAM chips should be generally good enough, they tend to be all at least DDR3-1333
<ssvb>
diego71: maybe there could be some other important DRAM tuning parameters, which are critical for OLIMEX boards, but not so much for the other boards?
<hno>
ssvb, likely the olimex boards have somewhat different DRAM timings than most other boards. Olimex have done their own DRAM routing while most other boards ar using DRAM routing layout provided by AW.
<diego71>
ssvb: for now, I'm retesting with default dram param, at least should be all green. And start from this configuration to increse the speed
<hno>
libv, to implement that you need to be running on the board.
<ssvb>
hno: I would not be very surprised if OLIMEX happens to have a bit less experience/skills and also worse tools for DRAM routing than AW
<hno>
And less information on peculiar quirks needed for AW chips..
<ssvb>
right
gzamboni has quit [Ping timeout: 258 seconds]
<diego71>
ssvb: how is the tablet working?
gzamboni has joined #linux-sunxi
<ssvb>
diego71: well, it only has usb otg to connect with the outside world, and the driver is less than perfect
<ssvb>
diego71: right now I'm using NFS root over g_ether
<arokux2>
our wiki is great
<diego71>
with the android kernel?
<ssvb>
diego71: sunxi-3.4 kernel
<ssvb>
diego71: as for using it as an android tablet, the screen quality and screen resolution are a bit worse than what I find comfortable :)
<ssvb>
arokux2: just have a look at the patches from that my branch, except for the last two (if you are not interested in high dram clock speed for the cubietruck)
<arokux2>
first 8 KB are skipped, then u-boot (268KB), the filesystem should come after that
<hno>
arokux2, which command is broken?
<arokux2>
hno: 1,16,c
<arokux2>
it means starting after first block or?
steev has joined #linux-sunxi
<hno>
Unit is MByte. -uM option to sfdisk.
<hno>
so it's starting at 1MB and is 16MB large.
<hno>
-uS 2048 is the same. 1MByte.
<hno>
arokux2 ^
<arokux2>
hno: well, if I flash u-boot after writing fs tables then it scrambles the fs table :$