hno changed the topic of #linux-sunxi to: Allwinner/sunxi development discussion - Don't ask to ask. Just ask! - See | | Logs at
hglm has quit [Quit: leaving]
dlan has quit [Ping timeout: 248 seconds]
techn_ has quit [Read error: Connection reset by peer]
forcev has quit [Read error: Connection reset by peer]
FunkyPenguin has joined #linux-sunxi
dlan has joined #linux-sunxi
e-ndy has quit [Read error: Operation timed out]
e-ndy has joined #linux-sunxi
utente has joined #linux-sunxi
egbert has quit [Disconnected by services]
egbert has joined #linux-sunxi
_BJFreeman has joined #linux-sunxi
_BJFreeman is now known as BJfreeman
zeRez_ has joined #linux-sunxi
zeRez has quit [Ping timeout: 256 seconds]
TheSeven has quit [Disconnected by services]
[7] has joined #linux-sunxi
guest0815 has joined #linux-sunxi
zeRez_ has quit [Ping timeout: 246 seconds]
guest0815 has quit [Quit: Leaving]
BJfreeman has quit [Quit: had a good time]
gzamboni has quit [Ping timeout: 246 seconds]
<Tsvetan> Turl I guess this is same for every electonics
atiti has quit [Ping timeout: 246 seconds]
<hramrach__> hello :)
yo5phz has joined #linux-sunxi
yo5phz has quit [Client Quit]
atiti has joined #linux-sunxi
<oliv3r> gooooood morning everybody
rellla has joined #linux-sunxi
<oliv3r> hi rel
<rellla> cubierock ;)
jemk has joined #linux-sunxi
tkoskine_ is now known as tkoskine
<mripard_> oliv3r: I won't be mad at you because you ask something (but I might be if you do it behind my back saying that you don't want to talk to me)
Tsvetan2 has joined #linux-sunxi
<jemk> morning, i've heared there would be some openmax wrapper for cedar in a20-sdk, anyone knows more?
Tsvetan has quit [Ping timeout: 240 seconds]
Tsvetan2 is now known as Tsvetan
wingrime has joined #linux-sunxi
FR^2 has joined #linux-sunxi
<rellla> jemk: seems there is something in android4.2/hardware/aw
Black_Horseman has quit [Remote host closed the connection]
dragonn has joined #linux-sunxi
<hramrach__> jemk: I haerd that for a10 too. like 2 months ago
<oliv3r> mripard_: i don't do it behind your back! We do it right here in public! but I don't wanna bother with the same stuff too much because I don't understand stuff :)
notmart has joined #linux-sunxi
notmart has quit [Changing host]
notmart has joined #linux-sunxi
damask has joined #linux-sunxi
lunatik has joined #linux-sunxi
<damask> Hi. At this moment what is the right git kernel repo to use for A20 ?
<oliv3r> hansg's wip repo
<oliv3r> but it's highly experimental
<damask> I'm searching for hangs
<damask> uh... 31 hansg users on git :)
<utente> i dont understand why there are difficults to support A20: it is just a dual core fersion of A10, so - it is a suppose -it shyould be 100% complatible with all done alreadi for A10: worsta case should be that only one core is recognized and used. instead it seems it is so difficult to use A20: why?
<damask> A20 is almost 100% pin compatible with A10 but seem is a different CPU and device (A7 vs A8, different GPU, etc)
<FR^2> "almost" ;)
<utente> both ARMv7, so ASM is the same.... maybe there are some other parameters to take in count?
<damask> nah ... Cortext-A8 (A10) vs Cortex-A7(A20)
<utente> both cortex-a8 and cortex-a7 are ARMv7, afaik
<damask> Ok for that ... I was not able yet to find A20 repo of a user called hansg ... could be more precise ? Thank you for help
<utente> i dont understnd what are the difference between cortex-a7 and cortex-a8 to require a different repository to build a installation.
<damask> actually I'm looking differences in sunxi-3.4-a20-wip
focus has joined #linux-sunxi
damask has quit [Quit: Page closed]
vicenteH has joined #linux-sunxi
ganbold__ has quit [Ping timeout: 246 seconds]
ganbold__ has joined #linux-sunxi
<buZz> jwrdegoede = hansg
notmart has quit [Quit: notmart terminated!]
jukivili has quit [Remote host closed the connection]
jukivili has joined #linux-sunxi
drachensun has quit [Read error: Operation timed out]
Tartarus has quit [Ping timeout: 276 seconds]
Tartarus has joined #linux-sunxi
<oliv3r> a20 uses different IRQ controller
<oliv3r> a20's code base is like what a10 was when we stared, so 'back in time by tons of work'
<oliv3r> so yeah, getting a20 support up is quite a daunting task
<utente> oliv3r, so, another amazing coding challenge to make A20 usable :)
<oliv3r> it's almost like starting over
<oliv3r> also sun7i is like sun4i + sun5i
<utente> i guess, who is already slylle in A10 *should* have less difficulties to work on A20
<oliv3r> never that easy
<oliv3r> but feel free to send patches :)
<utente> i dont understand meaning of: "also sun7i is like sun4i + sun5i"
derethor has quit [Read error: Connection reset by peer]
<oliv3r> ok sun7i == a20
<oliv3r> sun4i = a10, sun5i = a10s/a13
<oliv3r> sun7i is NOT 2x sun4i
<oliv3r> the cpu is as said slighhtly different, but ok sure, that's identical
<oliv3r> there's a different IRQ controller, an upgraded mmc controller
<oliv3r> i guess it's the mmc controller from sun5i actually
<oliv3r> they did actually change around quite a few things, so it really isn't sun4i x2; it's more like sun4i + sun5i and then some
<utente> now is more clear, thanks. but wher edo you find all these tecdhnical ino? it isi interensting, i wisho to know more.
<oliv3r> we have 3.3 kernel code drop for a20
<oliv3r> so we can compare and see how its done there
<utente> ah ok
<utente> do you know nm of A20?
<oliv3r> same as a10
<utente> 55nm?
<oliv3r> i think so
<utente> so at 1.2Ghz, A20 will be melted by hotness.
<ijc> The a20 IRQ controller is a GICv2 which is a fairly standard ARM part, for which there is an upstream driver already. You just need to know the base address and the irq routing for the SoC...
<ijc> I just asked on #cubieboard but perhaps here is better:
<ijc> Has anyone run a recent (e.g. 3.9+) kernel on a cubieboard2? Or perhaps someone has started writing an a20 dts file? Last resort doesn't anyone have access to a datasheet which enumerates the peripheral MMIO addresses and the interrupt mappings etc?
<utente> i blame A20. 55nm are to much. i hoped in at least 40nm, but....
<oliv3r> ijc: it's not so sure if it goes that way in 3.4 ;) but for mainline, sure
<oliv3r> ijc: we have zero documentation, only the 3.3 code drop
<ijc> 3.4 is the android/allwinner kernel, right?
<oliv3r> ijc: work on a20 for mainline is slowly starting, but we're short on time :)
<ijc> documentation> that's what I thought :-/
<oliv3r> ijc: 3.4 is 'our' kernel yeah, 3.3 is pure allwinner kenrel
<ijc> My ultimate aim is to run Xen on the cubieboard, which requires 3.7+. But getting any newish mainline kernel booting (without Xen) would be a good start...
<oliv3r> ijc: right now, now way :)
<ijc> I've got a very thin dts file, but I'm guessing things like the uart irq mapping, and it doesn't bootvery far. Gets to calibrating delay -- which suggests Ive either not guessed the arch timer IRQ correctly or I've not turned on the appropriate clocks. Pretty hard toguess which without docs...
gzamboni has joined #linux-sunxi
<oliv3r> ijc: are you hosting your patches anywhere?
<ijc> oliv3r: not yet, I've just been hacking in the spare moments at linaroconnect this week and hadn't really "succeeded". I can throw them up somewhere though..
lunatik has quit [Quit: Page closed]
jemk has quit [Ping timeout: 246 seconds]
focus has quit [Ping timeout: 260 seconds]
<ijc> oliv3r: -- be aware that 90% of the interupt numbers are a pure fiction!
<mripard_> ijc: I was planning on starting it like this week end or next week
<ijc> mripard_: Sweet ;-) I should just be patient I think...
<oliv3r> mripard_: pff, you siad you where starting it 3 weeks ago! ;)
<mripard_> ijc: but if you want, you can start it right away if you have some time
<oliv3r> on that note, i should continue on my a31 stuff, p2wi should be complete now
<mripard_> oliv3r: the nightmare of procrastination, every week has a next week :)
<mripard_> oliv3r: but I wanted first to get that A31 boot up properly
<mripard_> I think I have a clue on what's going on now
<mripard_> so A20 will be the next step if it's right
<mripard_> (and the problem is likely to arise on A20)
paulk-desktop has joined #linux-sunxi
<oliv3r> mripard_: so 1 fix for 2, that's good
<zumbi> do you guys have run into license checks when building MALI400?
<oliv3r> i wanted to do some more a31 uboot work today, so lets see how taht goes
<oliv3r> i'm unfamiliar with mali building, haven't really used it zumbi
<zumbi> MALI400 proprietary module seems to be incompatible with PM_RUNTIME, PROFILING and DMA_SHARED_BUFFER
<mripard_> oliv3r: yeah, that's why I didn't start working on the A20 yet :)
<oliv3r> mripard_: so what's causing the issue?
<mripard_> the SMP bit is enforced in the kernel only when you have CONFIG_SMP
<mripard_> even if you use only a single CPU
<mripard_> so there's kind of a weird logic here
<oliv3r> and you where setting up a single a31 cpu only?
<mripard_> since it won't be set at all if you build the kernel with !CONFIG_SMP
<oliv3r> figuring, the other cpu's come later?
<mripard_> for now, I didn't care at all about SMP.
<oliv3r> i can agree and imagine
<mripard_> I was only using the CPU u-boot brought up
<mripard_> and that's it
<oliv3r> so the SMP bit IS properly set in the BROM?
<mripard_> I haven't check that yet
<oliv3r> ah ok
<mripard_> but it's quite easy to see if the SMP bit is the root cause of this
<mripard_> I'll just enable CONFIG_SMP and see how it goes.
<oliv3r> living on the edge!
<oliv3r> mripard_: are you planning to go to fosdem 2014?
atiti has quit [Ping timeout: 256 seconds]
<mripard_> oliv3r: using CONFIG_SMP doesn't mean you actually use several CPUs
<mripard_> just that you *can* do it :)
<mripard_> and I haven't thought about it yet
<mripard_> it's in a loooong time :)
<oliv3r> < 6 months, it'll be done and gone before you know it! ;0
orsic has joined #linux-sunxi
Black_Horseman has joined #linux-sunxi
<ssvb> zumbi: how is mali400 module proprietary when it is GPL licensed?
leowt has joined #linux-sunxi
<zumbi> ssvb: it seems I am doing out of tree build and it thinks it's proprietary stuff
<zumbi> ssvb: also mali400 seems to be dual license, proprietary and gpl
<zumbi> # For customer releases the Linux Device Drivers will be provided as ARM proprietary and GPL releases:
<zumbi> # The ARM proprietary product will only include the license/proprietary directory
<zumbi> # The GPL product will only include the license/gpl directory
focus has joined #linux-sunxi
techn_ has joined #linux-sunxi
Black_Horseman has quit [Quit: Zwi se logou mou!!!]
Tartarus_ has joined #linux-sunxi
Tartarus_ has quit [Remote host closed the connection]
wingrime has quit [Ping timeout: 246 seconds]
leowt has quit [Quit: leowt]
atiti has joined #linux-sunxi
wingrime has joined #linux-sunxi
H6r6tiqu6 has quit [Quit: leaving]
dragonn has quit [Ping timeout: 256 seconds]
zeRez has joined #linux-sunxi
<zeRez> hi
vicenteH has quit [Ping timeout: 240 seconds]
<oliv3r> lo
<zeRez> :)
<zeRez> sup?
<oliv3r> stuck with things :)
<zeRez> :)
Undertasker has joined #linux-sunxi
<hramrach__> hi
<oliv3r> mripard_: what was the axp on the a31 again? axp229 or 221?
<hramrach__> is ther a source for the r3p2 mali blob that would be licensed for use with a20?
<hramrach__> looking in the 3.3 source drop there is only kernel side module but no bloblib nor is there one on hte factory Ubuntu image
<hramrach__> so I am wonderin if we are getting any 3d on a20 at all
<zeRez> hramrach__, hi :)
<mripard_> oliv3r: 221 iirc
<zeRez> hramrach__, so it will be a bad idea for me to purchase an a20 board? :o
<Turl> oliv3r: 221 "monster chip" as seen on olimex I think :P
<mripard_> Turl: see, you're getting famous
<utente> im recompiling kernel for cubieboad. I wish OTG usb port can be also a "normal" usb port per mouse, keyboard, etc. how to congigure it in menuconfig? it is ok to select "otg support"?
<oliv3r> famous turl!
<oliv3r> how are you famous?
<oliv3r> ok i have the following define now: #define XPOWERS_AXP221_DCDC_VOL_0.7V_+_25mV_X(n) ((n) & 1f)
<oliv3r> but it's easier then having a table
<oliv3r> how would you make a define, that does 0.7V - 2.25V in 0.025V steps/
tinti_ has joined #linux-sunxi
tinti_ has joined #linux-sunxi
tinti_ has quit [Changing host]
<mripard_> Mike Turquette is the clock framework maintainer
<utente> im following, and im recompiling kernel now. the "problem" is this kernel is 3.0.x, i wish to move to 3.4.x. where can i found it?
<oliv3r> maxi? :D
<oliv3r> Kudos maxime, pff where's emilio's kudos
<oliv3r> bad mike!
<Turl> mripard_: :)
<oliv3r> but deffo famous!
<Turl> oliv3r: the post is by maxime's coworker :P
<oliv3r> ahh :)
<oliv3r> maxi :)
<wingrime> Turl: have you uart cable nearbly?
<Turl> mripard_: are they streaming these talks?
<Turl> wingrime: yeah, on top of my desk, connected to an A10S :)
<hramrach__> utente: just check oput the 3.4 branch
<wingrime> Turl: I noticed that on my a13 kernel not boots without AXP options in config turned on
<hramrach__> that's probably a bug
<Turl> wingrime: mainline?
<utente> hramrach__, i duno where to gind 3.4: can you tell me how to dowloand using git?
wooy has quit [Read error: Connection reset by peer]
<wingrime> Turl: linux sunxi 3.0
<Turl> wingrime: ah, I dunno, I've only booted mainline on A10S
<hramrach__> utente: if you clenet the linux-sunxi repo the default branch is 3.0
<hramrach__> to get 3.4 'git checkout sunxi-3.4'
<utente> hramrach__, thannsks, i do immediately
<hramrach__> to get branch list 'git branch -a'
<wingrime> Turl: I borrow my cable to someone, so not have uart now, also need solder
<utente> im a newbie in git :-(
<mripard_> Turl: I guess it's on youtube
<hramrach__> there are git tutorials :p
<utente> hramrach__, wat do you meand with "clenet"?
<mripard_> oliv3r: Look at the comments :)
<hramrach__> cloned
<utente> ah ok
<hramrach__> 2 typos in 1 word :/
<utente> hramrach__, i can do worse :S
<mripard_> this was in the "Implementing power management for a new SoC" talk iirc
<Turl> mripard_: thanks :)
rellla has quit [Quit: Nettalk6 -]
<utente> now i must wait compiling task ends, then i go for linux 3.4 :)
\\Mr_C\\ has quit []
<oliv3r> mripard_: i saw; you made emilio more famous ;)
<oliv3r> mripard_: youtube link us!
<oliv3r> oh there is is allready
<oliv3r> i'm a slow reader
<oliv3r> mripard_: btw, for u-boot i've only done a single byte p2wi, even though it can burst 8 bytes every time
<mripard_> that sounds sensible
<Turl> mripard_: looking at the davinci suspend, I might implement something for sunxi :)
<Turl> mripard_: do you have hw to make power usage measurements?
<mripard_> Turl: none :(
<Turl> oliv3r: what about you? :P
<oliv3r> we only use p2wi to setup ddr voltages n stuff, not much is done so writing only the single bytes isn't a huge performance loss, I guess the kernel will want/need the higher speed bus
<oliv3r> Turl: depends on what you need
<oliv3r> Turl: i have a scope
<oliv3r> bus pirate
<oliv3r> a really fucked up 220V power measuring thing
<oliv3r> that's so wrong, it's not funny (it needs to be calibrated)
<Turl> oliv3r: :P
<Turl> oliv3r: well, an amperimeter would be good
<oliv3r> i have a multimeter :p
<oliv3r> but getting that in line with the circuit will be hard
<oliv3r> but possible
<oliv3r> i don't think i can measure amperage with my scope :p
<mripard_> oliv3r: yes, I definitely agree for you p2wi
<oliv3r> i'd have to make a circuit for that
<oliv3r> i'll mail an status update later tonight
<oliv3r> just so you know what i'm doing :)
wingrime has quit [Ping timeout: 276 seconds]
wingrime has joined #linux-sunxi
\\Mr_C\\ has joined #linux-sunxi
dragonn has joined #linux-sunxi
<oliv3r> yay, setup of ddr voltage shoudl work now
<oliv3r> in theory anyway
<Turl> mripard_: is there a "modern way" to use SRAM?
<mripard_> Turl: I've never used sram, but I came accross a drivers/misc/sram.c driver
<mripard_> that seem to be what we'd want
<mripard_> and basically just mapps the sram address space and calls a function to set it in a pool
<mripard_> so I guess there's all needed to use it
<mripard_> 2s
<Turl> mripard_: great, looks like exactly what we need
<Turl> and other platforms are duplicating on mach-platform/sram.c :)
<Turl> I can see a lot of potential cleanup patches there :p
<mripard_> yeah, genalloc.h seems to provide everything we need.
<Turl> mripard_: hm, how are you supposed to allocate sram when you use this driver?
<mripard_> gen_pool_(first|best)_fit?
<Turl> of_get_named_gen_pool :)
eebrah has joined #linux-sunxi
<Turl> apparently you add a sram = <&sram> thingy on dt, and then use of_get_named_gen_pool(np, "sram", 0)
<Turl> looking at media/platform/coda.c
<mripard_> yes, I was just looking at this as well
_BJFreeman has joined #linux-sunxi
<Turl> mripard_: it's the only user I could find of this
_BJFreeman is now known as BJfreeman
<mripard_> Turl: yep, me too :)
zeRez has quit [Quit: Leaving]
\\Mr_C\\ has quit []
zeRez has joined #linux-sunxi
\\Mr_C\\ has joined #linux-sunxi
Undertasker has quit [Read error: Connection reset by peer]
Undertasker has joined #linux-sunxi
hipboi has joined #linux-sunxi
<zeRez> is there any howto for cb2 images on qemu?
<zeRez> never worked with before
pacopad has joined #linux-sunxi
<oliv3r> qemu ON the cb2?
<oliv3r> hipboi: hey tom
<oliv3r> mripard_: Turl i actually spotted a generic SRAM controller the other day
<hipboi> oliv3r, hi
<pacopad> Hi Guys, little question about the 3.3.0 kernel on A20, is it a way to activate the cedarX ?
<zeRez> oliv3r, no, cb2 over qemu on x64^^
<oliv3r> ./drivers/misc/sram.c
<oliv3r> zeRez: you want to emulate ARMv7 on x86 via qemu; many howtoś for that, totall yunrelated to cb2
<mripard_> oliv3r: yep, it's what we were talking about :)
<mripard_> oliv3r: and qemu still has to emulate properly the board to boot something :)
<zeRez> oliv3r, read one about rpi on qemu, but cant boot cb2 image...right arm was choosen by me
<zeRez> ^^
<zeRez> do i have to compile a kernel for it anyway?
<oliv3r> mripard_: incidentally, i looked at the generic sram controller too; figuring we can use atleast the trustzone cache for it
hglm has joined #linux-sunxi
<zeRez> hglm, hi :)
<hglm> zeRez: I think it is non trivial to run qemu on a sunxi kernel -- qemu needs to know how to deal with all the sunxi-specific hardware details I think.
<hglm> zeRez: Hi, tried assembler yet?
<zeRez> zeRez, not at all...get stucked cause of reallife :P...
<zeRez> but wanna run an arm image on qemu to try it
<zeRez> ^^
<zeRez> but wait...isnt it crossassemble?
<hglm> zeRez: I think it is easier to try asm on the native device.
rz2k has joined #linux-sunxi
<zeRez> hglm, thats why i wanna run qemu with an arm image
<zeRez> ^^
<hglm> zeRez: Maybe you can setup qemu not use a kernel image but just "interpret" assembler files (sort of debugging mode).
<zeRez> okay
<zeRez> thats nice to know anyway!
<zeRez> found an howto
<zeRez> its for rpi
<zeRez> but will work with cubie images and slightly customizing too i think!
<zeRez> ?
<zeRez> ^^
<hglm> zeRez: Not sure, sounds difficult to me, qemu will need to know about the sunxi hardware. Maybe it requires a custom (minimal) kernel.
<zeRez> hm okay
<zeRez> loading image
<zeRez> will try it when done
<zeRez> than check errors
<zeRez> ^^
<zeRez> But i think it will run anyhow...i want that its running...have no device yet but wanna learn ^^...
<zeRez> hglm, but to learn basics of arm(asm), a simple rpi qemu vm should work also for me, isnt it?
<zeRez> ^^
drachensun has joined #linux-sunxi
FR^2 has quit [Quit: Connection reset by peer]
BJfreeman has quit [Ping timeout: 252 seconds]
lunatik has joined #linux-sunxi
_BJFreeman has joined #linux-sunxi
Tsvetan has quit [Read error: Connection reset by peer]
_BJFreeman is now known as BJfreeman
lunatik has quit [Client Quit]
<hglm> zeRez: Yes, you should be able to use a rpi vm. You can also cross-compile ARM assembler source files to ARM object files or executables with an ARM cross-compile toolchain.
<Turl> mripard_: any idea what might this be?
<Turl> anything reading the process tree hangs too
<Turl> It's the second time I see it I think
<Turl> it's 3.4 btw, so it might be something fixed on mainline
naobsd has joined #linux-sunxi
<mripard_> Turl: I have no idea
<zeRez> hgml sounds like a good idea to crosscompile
<Turl> mripard_: that file is not even there anymore on mainline :p
<Turl> $ uptime 13:11:43 up 14 days, 18:08, 3 users, load average: 15,16, 14,71, 13,46
<Turl> interesting, top works but ps and htop hang
<hglm> zeRez: You could run something like "armhf-...-as assembler.S -o assembler.o and then run armhf-...-objdump --disassemble assembler.o
eebrah is now known as eebrah|away
<zeRez> i should write an asm file, assemble it, and then dump it back to assembly? ^^
<zeRez> To understand what the assembler is doing with my code or why should i?
<hglm> zeRez: To verify that the assembler works. But you may be able to test a compiled object file or binary in qemu.
<zeRez> ahh okay :)
<zeRez> btw which assembler is common to use on arm?
<hglm> zeRez: GNU as, like everywhere Linux.
<hglm> zeRez: Try unified syntax (.syntax unified).
<zeRez> what is it?
<zeRez> the unified syntax?
<zeRez> ^^
dragonn has quit [Remote host closed the connection]
<hglm> zeRez: It's a notation for ARM assembler instructions that is compatible with both old and new ARM architectures. It looks almost the same as old syntax but is more flexible.
<zeRez> ah okay, thanks for hint! will be useful i think :)
pacopad has quit [Quit: pacopad]
<zeRez> rpi vm is running :)
hipboi has quit [Quit: Leaving]
jukivil1 has joined #linux-sunxi
BJfreeman has quit [Read error: Connection reset by peer]
jukivili has quit [*.net *.split]
steev has quit [*.net *.split]
_BJFreeman has joined #linux-sunxi
_BJFreeman is now known as BJfreeman
Tartarus_ has joined #linux-sunxi
hglm_ has joined #linux-sunxi
Undertasker1 has joined #linux-sunxi
Undertasker has quit [Ping timeout: 246 seconds]
hglm has quit [Ping timeout: 246 seconds]
Tartarus has quit [Ping timeout: 246 seconds]
vicenteH has joined #linux-sunxi
<hramrach__> Turl: never seen that crash
<hramrach__> maybe different VM options?
techn__ has joined #linux-sunxi
<hramrach__> and what device?
techn_ has quit [Ping timeout: 248 seconds]
<hramrach__> Turl: /* Leave this BUG_ON till prio_tree patch stabilizes */
<hramrach__> it has not stabilized in this version, obviously :p
zeRez has quit [Quit: Verlassend]
dragonn has joined #linux-sunxi
<Turl> hramrach__: :p
<Turl> mripard_: where are mappings defined on the DT?
wingrime has quit [Ping timeout: 264 seconds]
<Turl> mripard_: virtual to physical more specifically
<Turl> (or the other way round, rather)
<Turl> looks like we're using 1:1 mapping hm :)
n01 has joined #linux-sunxi
steev has joined #linux-sunxi
eebrah|away is now known as eebrah
<utente> i recompiled linux 3.4.43, but my wireless zd1211 doesnt work. in kernel source there is option only related to zd1201 but zd1211 doesnt exist, the module zd1211rw is not created. Same problem with RTL8188SU, not recognized by cubie... some suggestions?
<utente> the wireless usb is recognized by "lsusb" but "iwconfig" dont report the new interface :-(
Tartarus_ has quit [Quit: ZNC -]
Tartarus has joined #linux-sunxi
<oliv3r> utente: iwconfig wlan0?
<oliv3r> manually load the module?
<utente> oliv3r, wlan0 no such device
<utente> lsusb report preset RTL8188SU
<utente> in /lib/modules/3.4.43/kernel/drivers/net/wireless/ there are directory: rtl8188es, rtl8189es,rtl8192cu, rtl8723as.
eebrah is now known as eebrah|away
<utente> on x86 both wireless usb are recognized, but not on cubie. x86 has linux 3.2, cubie 3.4.43. it seems not all wireless drivers are avaiable for arm
<utente> rd1211rw is present since many years in linux source tree, but not in 3.4.43 i get from linux-sunxi...
n01 has quit [Ping timeout: 240 seconds]
<oliv3r> we have very limited wireless drivers
<oliv3r> especially in the older kernels
<hramrach__> the drivers are there
<hramrach__> just not enabled
<hramrach__> utente: did you enable the wireless drivers in your config?
<hramrach__> ls /lib/modules/3.4.43+/kernel/drivers/net/wireless/
<hramrach__> ath brcm80211 libertas rtl8188eu rtl8723as
<hramrach__> bcm4330 hostap mwifiex rtl8189es rtxx7x
<hramrach__> bcmdhd iwmc3200wifi rndis_wlan.ko rtl8192cu zd1201.ko
<orsic> hi there
<orsic> i try to get a picture out of a 1680x1050 monitor using a hackberry
<orsic> the device itself works, i have verified it with an other monitor (same resolution)
<orsic> but my monitor wont display anything
<orsic> is there any "easy" solution? or does this mean i have to tinker with the disp driver?
<hramrach__> if it works with another then the easy solution does not apply
<orsic> damn...
<hramrach__> try getting the dmesg output with the non-working monitor attached
<orsic> those are the relvant bits
<orsic> [ 1.530000] ParseEDID
<orsic> [ 1.550000] EDID version: 1.3
<orsic> [ 1.550000] PCLK=119000000 X 1680 1728 1760 1840 Y 1050 1053 1059 1080 fr 59 PN
<orsic> [ 1.550000] Using above mode as preferred EDID mode
<hramrach__> pastebin please
<orsic> [ 1.550000] PCLK=146250000 X 1680 1784 1960 2240 Y 1050 1053 1059 1089 fr 59 NP
<orsic> [ 1.550000] disp_clk: Could not find a matching pll-freq for 78800000 pclk
<orsic> [ 1.550000] disp_clk: Could not find a matching pll-freq for 53450000 pclk
<orsic> [ 1.560000] disp_clk: Could not find a matching pll-freq for 26150000 pclk
<orsic> [ 1.570000] disp_clk: Could not find a matching pll-freq for 146250000 pclk
<orsic> [ 1.570000] waited 0 ms for EDID info
<orsic> [ 1.570000] disp clks: lcd 119000000 pre_scale 1 hdmi 119000000 pll 357000000 2x 0
<orsic> (i think :) )
<orsic> kk
<hramrach__> also the monitor model
<hramrach__> which might be in the log too
<orsic> It's a LG Flatron W2042T
<orsic> i've also put the EDID data in the pastebin
<orsic> i did read it with my pc
<orsic> but it looks somewhat bogous to me
<orsic> on the other hand i have no idea what i should expect there...
<hramrach__> it failed to read
<hramrach__> the same as on the sunxi board
<orsic> so is the EDID data somehow corrupted?
<orsic> but then how does xrandr detect the right modes?
<hramrach__> the one you received on PC is blank
<hramrach__> the one you have on sunxi is probably multi-block
<hramrach__> I don't get that many messaged for single block
<hramrach__> what is the kernel version?
<orsic> 3.4.43.sun4i+ fedora 18
<orsic> hansg's build
<hramrach__> try to wait till hansg is around I guess
<orsic> all right
<hramrach__> or build your own kernel with some debug prints
<orsic> which one would be useful?
<hramrach__> like how many blocks you got and from which block the mode is picked
<hramrach__> In particular it is suspicious that you have 2 1680x1050 modes
<orsic> and getting this data from userspace is not possible?
eebrah|away is now known as eebrah
<hramrach__> it is to some extent
<orsic> well i'm also willing to build a kernel if this is better in this case
<hramrach__> cat /sys/devices/platform/disp/graphics/fb0/modes
<hramrach__> you can try settign a different one in /sys/devices/platform/disp/graphics/fb0/mode
<hramrach__> but if the edid data is interpreted wrongly you have only wrong modes
ChanServ has quit [shutting down]
<utente> hramrach__, another wireless dongle that i have it works, so it leans wireless stack is enabled. simply, it seems there are no driver for the other 2 usb donfles wireless .-(
<orsic> cat /sys/devices/platform/disp/graphics/fb0/mode doesn't return anything, does this mean there is no mode set?
<hramrach__> it is set but not displayed
<atiti> orsic: hansg has some EDID fixes
<hramrach__> the reader is not implemented I guess
<orsic> ah okay
<utente> root@debian:/lib/modules/3.4.43/kernel/drivers/net/wireless# ls
<utente> bcm4330 hostap libertas rtl8188eu rtl8192cu rtxx7x
<utente> ath bcmdhd iwmc3200wifi rndis_wlan.ko rtl8189es rtl8723as zd1201.ko
<orsic> i will play with fb0/modes a little
<orsic> thanks for the help so far
<hramrach__> yes, seems zd1211 is not enabled
<hramrach__> possibly kconfig error
<utente> hramrach__, do you know how to solve?
ChanServ has joined #linux-sunxi
<hglm_> orsic: Maybe the a10disp utility can be useful. It is a utility to change video mode in userspace. You can try "a10disp info" to see what HDMI modes are available.
<orsic> thanks, i will have a look at it
<hramrach__> utente: you need to enable MAC80211
<utente> mmm... mac80211... o check immediately
<utente> hramrach__, in /proc/config.gz it is reported as NOT SET. i go to recompile kernel
<utente> hramrach__, yes, now in source tree there are more wireless driver avaiable! also what im looking for! thaks!
<hramrach__> np
<utente> hramrach__, u are skilled, more than me
<hramrach__> you need to master the grep utility ;-)
<utente> ehehe :)
<utente> hramrach__, kernel recompiled and istalled. all 3 usb wireless dongle works. thanks again.
eebrah is now known as eebrah|away
Black_Horseman has joined #linux-sunxi
atiti has quit [Remote host closed the connection]
hramrach__ has quit [Ping timeout: 240 seconds]
hglm_ has quit [Quit: leaving]
hramrach__ has joined #linux-sunxi
<orsic> so i tired a10disp
<orsic> and also to set modes diretly like this:
<orsic> echo S:1280x1024p-60 > /sys/devices/platform/disp/graphics/fb0/mode
<orsic> no change
<orsic> atiti: is this the repository with the fixes?
<Turl> mripard_: ping
orsic has left #linux-sunxi [#linux-sunxi]
orsic has joined #linux-sunxi
orsic has quit [Client Quit]
zeRez has joined #linux-sunxi
<oliv3r> pong!
<oliv3r> :p
atiti has joined #linux-sunxi
<Turl> oliv3r: do you happen to know how ioremap works? :p
<ssvb> hramrach__: iirc you had some problems with similar stuff
<oliv3r> Turl: a little bit, i've had to use it
<oliv3r> i think
<oliv3r> Turl: nvm i used:
<oliv3r> -------res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
<oliv3r> >-------sid_reg_base = devm_ioremap_resource(&pdev->dev, res);
<oliv3r> unless you actually ment devm-ioremap_resources :)
<hramrach__> oh, now they fix it
<hramrach__> it's against their own spex :p
<hramrach__> spec
<hramrach__> it will land in Debian like 3 yrs from now if the patch is not dropped xD
<Turl> oliv3r: well, internally that uses ioremap :p
<Turl> oliv3r: it's just that when I use it it gives me an address somewhere else and I can't understand why
* hramrach__ does a voodoo ritual sicking pins into Xlib manual
paulk-desktop has quit [Quit: Ex-Chat]
<oliv3r> Turl: why would you want to use ioremap directly instead of the devm/platform ones?
<oliv3r> i thought the higher up the better, that's what the abstraction layers where for
<Turl> oliv3r: yeah but they use it internally
<Turl> oliv3r: SRAM is 0x0 PA right? or am I reading the manual wrong?
tinti_ has quit [Quit: Leaving]
<oliv3r> pa 0x0 what?
Black_Horseman has quit [Remote host closed the connection]
Black_Horseman has joined #linux-sunxi
Black_Horseman has joined #linux-sunxi
<Turl> oliv3r: physical address 0x0
dragonn has quit [Ping timeout: 246 seconds]
<oliv3r> Turl: you mean ALL of the sram?
<oliv3r> i'm not sure if we use A1 at all; uboot obviously uses it, i don't know if the kernel does
<oliv3r> if not, A; B are free for use
<oliv3r> C is off limits, as they are mapped to the hardware
<oliv3r> D (2kb only) is free
<oliv3r> if you don't have nand, you can use 2KB from 'nand' too i think, but i don't know if that's the sram address, or some shadow address
hramrach__ has quit [Ping timeout: 240 seconds]
slapin_nb has quit [Ping timeout: 264 seconds]
slapin_nb has joined #linux-sunxi
<Turl> oliv3r: yeah, say SRAM A, it has physical address 0x0 right?
<oliv3r> 112 kiB should be usable
<oliv3r> yes
<oliv3r> best test with B; that's unused 100% for sure
ykchavan has joined #linux-sunxi
utente has quit [Ping timeout: 248 seconds]
hramrach__ has joined #linux-sunxi
gzamboni has quit [Read error: No route to host]
_BJFreeman has joined #linux-sunxi
BJfreeman is now known as Guest40713
_BJFreeman is now known as BJfreeman
Guest40713 has quit [Ping timeout: 252 seconds]
Undertasker1 has quit [Quit: Leaving.]