<pfdm>
ssvb: when i use ump_mapped_pointer_get to write something is it locking automatically ? or i need to do it manually ?
<ssvb>
pfdm: only ARM knows for sure how this all interacts with the blob
Soru has quit [Read error: Connection reset by peer]
<pfdm>
ssvb: ok so I m guessing this is the right call function to write directly something in the address. Oh i see i'm correct if not you have to use ump_write which is less convenient. ;)
Soru has joined #linux-sunxi
torindel has quit [Ping timeout: 272 seconds]
torindel has joined #linux-sunxi
Black_Horseman has quit [Quit: Αποχώρησε]
geecko has quit [Quit: Quitte]
Black_Horseman has joined #linux-sunxi
torindel has quit [Ping timeout: 245 seconds]
torindel has joined #linux-sunxi
<pfdm>
ssvb: Isn't that weird that EGL_NATIVE_BUFFER_ANDROID is not supported in the proprietary mali ?
<pfdm>
oh sorry, i guess they include only the linux version
hramrach has quit [Ping timeout: 240 seconds]
eebrah_ has joined #linux-sunxi
torindel has quit [Ping timeout: 272 seconds]
torindel has joined #linux-sunxi
wingrime has quit [Ping timeout: 245 seconds]
Sonicadvance1 has joined #linux-sunxi
deasy has quit [Quit: Nom d'un quark, c'est Edmonton !]
Black_Horseman has quit [Remote host closed the connection]
benn has joined #linux-sunxi
eebrah_ has quit [Read error: Connection reset by peer]
egbert has quit [Disconnected by services]
egbert has joined #linux-sunxi
torindel has quit [Ping timeout: 240 seconds]
torindel has joined #linux-sunxi
benn_ has joined #linux-sunxi
benn has quit [Ping timeout: 240 seconds]
benn_ has quit [Ping timeout: 272 seconds]
torindel has quit [Ping timeout: 245 seconds]
torindel has joined #linux-sunxi
benn_ has joined #linux-sunxi
eebrah_ has joined #linux-sunxi
eebrah_ has quit [Read error: Connection reset by peer]
torindel has quit [Ping timeout: 245 seconds]
torindel has joined #linux-sunxi
benn____ has joined #linux-sunxi
hipboi has joined #linux-sunxi
torindel has quit [Ping timeout: 245 seconds]
torindel has joined #linux-sunxi
eebrah_ has joined #linux-sunxi
benn_ has quit [Ping timeout: 272 seconds]
torindel has quit [Ping timeout: 240 seconds]
torindel has joined #linux-sunxi
benn____ has quit [Quit: Ex-Chat]
torindel has quit [Remote host closed the connection]
torindel has joined #linux-sunxi
eebrah_ has quit [Ping timeout: 272 seconds]
eebrah_ has joined #linux-sunxi
eebrah_ has quit [Ping timeout: 246 seconds]
eebrah_ has joined #linux-sunxi
deasy has joined #linux-sunxi
torindel has quit [Ping timeout: 248 seconds]
torindel has joined #linux-sunxi
eebrah_ has quit [Ping timeout: 240 seconds]
TheSeven has quit [Disconnected by services]
[7] has joined #linux-sunxi
torindel has quit [Ping timeout: 240 seconds]
torindel has joined #linux-sunxi
torindel has quit [Ping timeout: 245 seconds]
torindel has joined #linux-sunxi
wingrime has joined #linux-sunxi
torindel has quit [Ping timeout: 272 seconds]
torindel has joined #linux-sunxi
torindel has quit [Remote host closed the connection]
torindel has joined #linux-sunxi
<libv>
kenny: what would you like us to spend our limited free time on?
<libv>
kenny: getting things to actually work, or, chasing the latest upstream version and (further) breaking things all the time?
naobsd has quit [Ping timeout: 250 seconds]
<wingrime>
Checking for ARM NEON ... no
<wingrime>
Checking for ARM THUMB ... no
<wingrime>
mplayer so funny
<libv>
they are too keen on assembler "optimizations", and are so bad that they cannot even manage to do so in runtime (as that smells like autoconf output)
<libv>
slapin: you have verified that you have copied the right bits to those places on your sdcard, right?
* wingrime
trys run 4k video with cedarx
<slapin>
libv: I've just expected for u-boot prompt to cum up :)
<slapin>
libv: I just setup open embedded stuff to build everything, so to have more friendly environment and now a bit surprised meeting this one
<slapin>
libv: I know how to resolve this a hacky way, I just want to know what was really intended to get this way
<libv>
slapin: still, verify that all bits and pieces are installed appropriately, as if you had done it the hacky way
<slapin>
libv: I want just u-boot prompt :) I don't want to put kernel on this SD card just to get u-boot prompt
<libv>
if the oe stuff is repeatable, set up the basics the hacky way on top, and see if that fixes it
<slapin>
just waiting for hno to answer
<libv>
ok
n01 has joined #linux-sunxi
<slapin>
libv: I already did that, so i can happyly wait for hno to answer, thanks. I just want to know what way was really intended. I've just set up falconless old boot of cubieboard.
pokus has quit [Ping timeout: 272 seconds]
pokus has joined #linux-sunxi
rellla2 has joined #linux-sunxi
<wingrime>
ssvb: how much VE ram will be enought for 4k video?
pokus has quit [Ping timeout: 248 seconds]
pokus has joined #linux-sunxi
cubear has joined #linux-sunxi
andoma_ is now known as andoma
pokus has quit [Ping timeout: 272 seconds]
<wingrime>
ssvb: currently I see that memory maybe currupts at even 100Mb allocated
<hno>
slapin, what?
<wingrime>
ssvb: also, seeing this, I suggest, that when you run mem-test same time with mali, It was just memory corrupton not realy releated with DDR3 workflow
<hno>
slapin, see sunxi u-boot wiki.
<hno>
SPL changed somewhat upstream with effect that installation procedure differs in a tiny bt critical bt (which image to write to the card)
<wingrime>
hno: magic related with uboot , not realy ended
panda84kde has joined #linux-sunxi
<wingrime>
hno: kernel boots normaly if I turn of power
<wingrime>
hno: and plug it agian
<wingrime>
hno: but if I do "reboot" using console
<wingrime>
hno: uboot enter endless loop with "Reset..."
BluesBoy has joined #linux-sunxi
<wingrime>
hno: ideas?
<wingrime>
hno: you are realy shure that cpu1 is killed?
<wingrime>
*sure
pokus has joined #linux-sunxi
<oliv3r>
goood morning
<hno>
wingrime, I am not sure, but I would suppose it is. reboot is done using the watchdog which triggers a reset.
<oliv3r>
kenny: not likly
<hno>
wingrime, might be worth trying the SMP change that brings both CPUs online in u-boot.
<wingrime>
hno: maybe we should do more actions to consider clear boot
<wingrime>
hno: also, dmesg saying , that IRQs was enabled before it do by kernel
<wingrime>
hno: and maybe sure in SPL, that watchdog realy disabled
<wingrime>
hno: I just guessing, whats wrong,
<wingrime>
hno: but, I can calim, that reset using watchdog not do proper cleanup actions
<wingrime>
[ 0.000000] start_kernel(): bug: interrupts were enabled early
<steev>
indeed they are
naobsd has joined #linux-sunxi
pokus has quit [Ping timeout: 240 seconds]
pokus has joined #linux-sunxi
pokus has quit [Ping timeout: 272 seconds]
pokus has joined #linux-sunxi
notmart has joined #linux-sunxi
torindel has quit [Ping timeout: 240 seconds]
<ssvb>
slapin: try to "dd if=u-boot-sunxi-with-spl.bin of=/dev/sdX bs=1024 seek=8"
<ssvb>
slapin: this has changed some time ago, and maybe you have picked some stale outdated instructions for building u-boot and writing it to the SD card
torindel has joined #linux-sunxi
<oliv3r>
that is most def. the easiest way
<oliv3r>
mripard: the a31 SMP stuff, is it based on the A20 patches we saw some time ago? or does that SoC work completly different
FR^2 has joined #linux-sunxi
<mripard>
oliv3r: as far as SMP is concerned, it's a bit different, so you can't exactly use the SMP ops of the A20
<mripard>
there might be some opportunity to factorise some code, but since no SMP has been merged yet...
<oliv3r>
mripard: just curious because cinifr hasn't properly fixed everything has he?
<mripard>
nope
<mripard>
and he keeps making the same mistakes and ignoring the comments and help he could receive
<slapin>
hno: do you understand that you can't ever het into u-boot prompt on cubieboard with such configuration?
<oliv3r>
mripard: frustrating :(
<slapin>
hno: at least if you write kernel?
<mripard>
oliv3r: yep.
<oliv3r>
mripard: how are things with you?
<slapin>
hno: I mean into SD card
<oliv3r>
slapin: why can't you get into the uboot prompt? if you invalidate your env partition, it should get forced into the u-boot propmt
<oliv3r>
slapin: what exactly is your problem right now?
<slapin>
oliv3r: this is not inapropriate for production
fluo75 has joined #linux-sunxi
<slapin>
oliv3r: you have to be able to get into u-boot prompt with standard controls
<slapin>
oliv3r: and no, you have to invalidate your kernel, actually
<mripard>
oliv3r: how are things with me?
<slapin>
oliv3r: actually I'm fairly able to reconfigure, but I need to warn about bad defaults
<oliv3r>
slapin: 'normally' on our u-boot, there is a 3 second delay
<oliv3r>
slapin: you can override this obviously, from the env partition/scr file
fluo75|2 has quit [Read error: Connection reset by peer]
<oliv3r>
mripard: yeah how are things, how's life, etc etc :)
pokus has quit [Ping timeout: 260 seconds]
<slapin>
oliv3r: 'normally' you just get SPL getting directly to kernel and no way to do anything if your kernel is faulty
pokus has joined #linux-sunxi
<oliv3r>
mripard: haven't seen /talked to you a lot recently :)
<oliv3r>
slapin: not with our u-boot
<mripard>
oliv3r: fine, I'm done with all the travelling, now I can sleep a bit :)
<slapin>
oliv3r: I use master of github:linux-sunix/u-boot
vicenteH has quit [Ping timeout: 272 seconds]
<oliv3r>
slapin: that one has a 3 second delay by default; which you can intterupt with a keypress
<slapin>
oliv3r: for me I use my fork and I can swith this off, too.
<slapin>
oliv3r: seen the link? it just doesn't have ANY chance to run
<slapin>
oliv3r: to provide the delay
<slapin>
oliv3r: as helping SPL does damn nice job running kernel directly
<slapin>
oliv3r: without any chance of interrupting
<ssvb>
slapin: sorry, have you tried the non-modified recent linux-sunxi u-boot and that "dd" command already?
<slapin>
ssvb: this is what I've tried
<oliv3r>
mripard: sleep, i need lots of it! but also time, such is life
<slapin>
ssvb: as I try to make some production images I want to make the picture on how the official stuff is, and that is not nice one, at least for cubieboard.
<arokux1>
mripard: have you seen my reply to your e-mail? I hope you understand I wasn't trolling. the OP wanted to have things mainlined but was missing official mainlining status, so I have given it to him
<arokux1>
hno: please take a look at (fresh) u-boot usb patches (on ML)
<oliv3r>
slapin: i'm not sure what you want/need or what's going wrong then
torindel has quit [Ping timeout: 272 seconds]
<mripard>
arokux1: it's in my inbox, haven't read it yet
torindel has joined #linux-sunxi
<slapin>
oliv3r: if you have normally made sdcard you can't get u-boot prompt on cubieboard
<oliv3r>
arokux1: i saw them; will try to look at them
<slapin>
oliv3r: until you clear out kernel
<slapin>
oliv3r: shich is error
<slapin>
which
<arokux1>
oliv3r: thanks, take a look at ehci-sunxi.c first, I've got some ideas how to improve the rest, will do tonight.
<slapin>
oliv3r: as cubieboard doesn't have any means to make SPL run u-boot instead of kernel directly
<slapin>
oliv3r: so it is better to disable falconboot for cubie by default
<ssvb>
slapin: and yes, the changed u-boot installation instructions have bitten some people :(
<oliv3r>
slapin: i'm looking at your openstack paste, and that looks like a generic boot problem, most the SPL can't load the u-boot binary and simply fails
<oliv3r>
it smells like something is old, My first advise would be, checkout a clean u-boot somewhere, do a fresh clean build and write it as ssvb linked above
<slapin>
ssvb: and where is mentioned they can have u-boot prompt?
<slapin>
ssvb: the board boots, that is all
<oliv3r>
u-boot promt is from u-boot, you are creashing before that, at the SPL
<oliv3r>
the SPL can't load u-boot so it can't do anything anymore, and the SPL is too limited to support a prompt
<ssvb>
slapin: with a properly installed u-boot and kernel, you can either boot your system, or interrupt u-boot from the serial console if you wish
<slapin>
ssvb: I need a way to get into u-boot prompt without damaging anything, cubieboard doean't have any buttons for that, so why bother with falcon on it at all?
<ssvb>
slapin: if your u-boot is fubared, then you can do neither
<oliv3r>
i think falcon mode is currently not supported
<oliv3r>
we don't have enough space in the SPL
<slapin>
ssvb: you can't interrupt on serial console, that is a problem. The fubared one on link is my attempt to have wtf answered, which I've got reading git log, so to know that now we need u-boot.img (uimage) instead of bin
<oliv3r>
i still think you are using something old or an old config; as falcon mode was cmopletly disabled a few commits ago
<slapin>
oliv3r: is is supported for sun4i as SPL is build with thumb
<slapin>
oliv3r: was this stuff upstreamed to Denx I wonder
<oliv3r>
slapin: no, sunxi isn't usptreamed yet; it needs some work (cleanup mostly)
<oliv3r>
naobsd: ah, yeah the commit subject seems wrong
* slapin
thinks this all image generation stuff will have most problems with upstreaming
<naobsd>
slapin: then why you use falcon mode? it should not enabled by default, as you said.
<oliv3r>
if you use a clean u-boot tree, it should work without much trouble
<oliv3r>
you can still flash spl + u-boot speratly, or just use the full binary that is much easier for users
<naobsd>
(I'm talking about configuration, not about source tree/compile time thing)
<ssvb>
wingrime: sorry, what kind of problem are you having now? I have some difficulties understanding your last messages (are they cedar or mali related?)
<slapin>
naobsd: it should not, but it is.
<naobsd>
slapin: how did you erase configuration for using falcon mode?
<slapin>
notmart: It just tries to run kernel first.
torindel has quit [Ping timeout: 272 seconds]
<naobsd>
well...
torindel has joined #linux-sunxi
<naobsd>
slapin: did you erase your SD card before flashing new image?
<mnemoc>
arokux1: at least I can't see anything sunxi related there, so it's at least off-topic
vicenteH has joined #linux-sunxi
<mnemoc>
arokux1: can you sync your "own" sunxi-next-usb with the other? to start using developer trees in the nightlies
<arokux1>
mnemoc: why? I'd delete my own. shouldn't I?
pokus has quit [Ping timeout: 272 seconds]
pokus has joined #linux-sunxi
<mnemoc>
arokux1: as oliver said before in this case it belongs to your personal tree (been a wip) more than the official
<mnemoc>
arokux1: so I would like to start building your usb linux and u-boot branches directly from your developer tree
<arokux1>
ah..
<mnemoc>
wip branches should not be in the official repo
<mnemoc>
repos*
<mnemoc>
and sunxi-nightlies is almost ready to support that :)
<mnemoc>
but need some developer branches to test ;-)
<mnemoc>
like, yours :)
<arokux1>
mnemoc: as side note, there are still lots of wip/ branches, will you prune them too? :p
<arokux1>
mnemoc: feel free to prune my branch, I'll sync tonight with my personal tree.
<mnemoc>
I planned to rename the dir, not to wipe it
<mnemoc>
for the nightlies I would prefer long-run wip branches, like this usb driver thing
<mnemoc>
than small wips
<arokux1>
mnemoc: ok, then wait till tonight, I'll ping you once I've synced to my personal tree.
pokus has quit [Ping timeout: 260 seconds]
<mnemoc>
arokux1: thanks! please don't forget your u-boot branch too
<arokux1>
mnemoc: yes, will do. thank YOU for the hourly build service!
<mnemoc>
thank you for giving code to build/test ;-)
<mnemoc>
a building service without new code is kind of useless :p
<arokux1>
mnemoc: btw, if building of the u-boot, can you specify which boards to build for? for the usb staff it makes sense to build for 3-5 boards only. (now)
<mnemoc>
currently I just build all sun.i boards, might add a pattern filter later
torindel has quit [Remote host closed the connection]
torindel has joined #linux-sunxi
pokus has joined #linux-sunxi
fredy has quit [Excess Flood]
fredy has joined #linux-sunxi
codingrobot has joined #linux-sunxi
pokus has quit [Ping timeout: 245 seconds]
pokus has joined #linux-sunxi
<codingrobot>
Anyone tried running mainline 3.12 on a cubieboard2? the sunxi-devel branch has a defconfig which I'm about to try.
enrico_ has joined #linux-sunxi
pokus has quit [Ping timeout: 240 seconds]
pokus has joined #linux-sunxi
Soru has quit [Read error: No route to host]
torindel has quit [Ping timeout: 245 seconds]
torindel has joined #linux-sunxi
Soru has joined #linux-sunxi
pokus has quit [Ping timeout: 245 seconds]
pokus has joined #linux-sunxi
<mnemoc>
codingrobot: the key is to use the right .dtb and setup u-boot accordingly
<mnemoc>
codingrobot: linux-sunxi has a wiki page about how to do it
<wingrime>
ssvb: I tring get 4k video with cedar
<wingrime>
ssvb: but, I see glitches too
<wingrime>
ssvb: I just thinking that glitches are resulut of absence region checks in mali allocator could produce "mem-test" error, you saw when run mali and mem-test same time
pokus has quit [Ping timeout: 245 seconds]
pokus has joined #linux-sunxi
torindel has quit [Ping timeout: 245 seconds]
torindel has joined #linux-sunxi
pokus has quit [Ping timeout: 240 seconds]
pitillo has quit [Changing host]
pitillo has joined #linux-sunxi
torindel has quit [Ping timeout: 245 seconds]
torindel has joined #linux-sunxi
pokus has joined #linux-sunxi
torindel has quit [Ping timeout: 245 seconds]
pokus has quit [Ping timeout: 240 seconds]
torindel has joined #linux-sunxi
<ssvb>
wingrime: if you are decoding video with cedar and using vlc or vdpau, then mali is not involved in this activity
<wingrime>
ssvb: yes, but fb, I just guessing that your mem glitches related to software nor hw
Black_Horseman has joined #linux-sunxi
Black_Horseman has joined #linux-sunxi
<ssvb>
wingrime: the memory failures that I observed are very clearly related to overclocking dram too much
torindel has quit [Ping timeout: 248 seconds]
<ssvb>
wingrime: if one removes dram overclocking, the problems disappear
<wingrime>
heh
torindel has joined #linux-sunxi
pokus has joined #linux-sunxi
rz2k has joined #linux-sunxi
torindel has quit [Ping timeout: 246 seconds]
torindel has joined #linux-sunxi
<arokux1>
wingrime: have you tried to add disp to u-boot? :p
<wingrime>
arokux1: I tryed get test-bars to HDMI
<wingrime>
but not successed at all
<arokux1>
wingrime: what about VGA?
shineworld has joined #linux-sunxi
<wingrime>
arokux1: I just get ct, not tryed
<arokux1>
wingrime: ok
<wingrime>
arokux1: now I playing with 4k video and cedar
<ssvb>
wingrime: with the highest cedar hw clock frequency?
<mnemoc>
libv: my "graphic" skills are crap
<wingrime>
320 Mhz
<wingrime>
not realy HI-est
fnchen has joined #linux-sunxi
<wingrime>
ssvb: video looks strange, looks "I" frames are loaded not fully
<wingrime>
ssvb: but moving vectors works
<ssvb>
wingrime: does changing the cedar clock frequency affect FPS? and if yes, then how much (is it scaling linearly)?
<wingrime>
ssvb: defetly affects
<wingrime>
ssvb: but I can't measure it by eye
<ssvb>
wingrime: the clock frequency limit should be around 370MHz
deasy has quit [Quit: Nom d'un quark, c'est Edmonton !]
<ssvb>
wingrime: try to implement/improve the benchmarking tools
<wingrime>
ssvb: I don't care mali, so possible use pll
<wingrime>
ssvb: cpu not realy hot
arokux1 has quit [Remote host closed the connection]
arokux has joined #linux-sunxi
pokus has quit [Ping timeout: 246 seconds]
fnchen has quit [Remote host closed the connection]
derethor has joined #linux-sunxi
pokus has joined #linux-sunxi
mouchon has quit [Ping timeout: 245 seconds]
mouchon has joined #linux-sunxi
Soru has quit [Read error: Connection reset by peer]
pokus has quit [Ping timeout: 246 seconds]
pokus has joined #linux-sunxi
soul has joined #linux-sunxi
soul is now known as Guest29599
\\Mr_C\\ has quit [Quit: .]
torindel has quit [Ping timeout: 245 seconds]
torindel has joined #linux-sunxi
pokus has quit [Ping timeout: 246 seconds]
torindel has quit [Ping timeout: 240 seconds]
torindel has joined #linux-sunxi
codingrobot has quit [Ping timeout: 264 seconds]
Black_Horseman has quit [Quit: Αποχώρησε]
popolon has quit [Read error: Connection reset by peer]
pokus has joined #linux-sunxi
Sonicadvance1 has quit [Remote host closed the connection]
heffer has quit [Read error: Connection reset by peer]
slapin has quit [Ping timeout: 245 seconds]
heffer has joined #linux-sunxi
heffer has joined #linux-sunxi
heffer has quit [Changing host]
slapin has joined #linux-sunxi
pokus has quit [Ping timeout: 248 seconds]
rz2k has quit []
\\Mr_C\\ has joined #linux-sunxi
shineworld is now known as shinealien
pokus has joined #linux-sunxi
shinealien is now known as alienworld
Guest29599 has quit [Ping timeout: 272 seconds]
popolon has joined #linux-sunxi
<kenny>
libv: good point. The reason I asked was because it appears 3.12 is the first mainline to have native CB2 support. It seemed a good candidate for a new stable.
cubear has quit [Remote host closed the connection]
<wingrime>
mripard: ping
vicenteH has quit [Ping timeout: 272 seconds]
Sonicadvance1 has joined #linux-sunxi
<mripard>
pong
<Sonicadvance1>
plong
Quarx has joined #linux-sunxi
<wingrime>
mripard: you can test VDPAU on a31
<mripard>
nope, the Android image is completely broken on it, I can't even boot it anymore :(
hipboi has quit [Ping timeout: 272 seconds]
vicenteH has joined #linux-sunxi
moofree_ has quit [Ping timeout: 245 seconds]
<arokux>
from http://linux-sunxi.org/AL-A13-RT713 "Over time the host controller driver has gotten broken and only a small portion of the devices that actually work with the stock android kernel currently work right with the linux-sunxi kernel."
deasy has quit [Quit: Nom d'un quark, c'est Edmonton !]
wingrime has quit [Quit: leaving]
<Nyuutwo>
Is there tool for dumping registers from kernel (i know sunxi-dbgreg but it prints out to kernel log)
nove has joined #linux-sunxi
pfdm has quit [Ping timeout: 250 seconds]
enrico_ has quit [Quit: Bye]
xeloweb has joined #linux-sunxi
<xeloweb>
hi all
FR^2 has quit [Quit: Connection reset by peer]
<xeloweb>
any here can etll me on german whats current status for A20 h264 decoding with xbmc ß
<Nyuutwo>
now I dumped TCON0 section and it is configured properly, probably in android is something encoded in wakeup sequence
<wigyori>
evening
Quarx has quit []
pfdm has joined #linux-sunxi
moofree has joined #linux-sunxi
kz1 has joined #linux-sunxi
n01 has joined #linux-sunxi
fredy has quit [Excess Flood]
fredy has joined #linux-sunxi
torindel has quit [Ping timeout: 240 seconds]
panda84kde has quit [Ping timeout: 245 seconds]
torindel has joined #linux-sunxi
<kz1>
is there a linux method to replace the pheonix suite?
torindel has quit [Ping timeout: 245 seconds]
torindel has joined #linux-sunxi
deasy has joined #linux-sunxi
<mnemoc>
hno: is it possible to make MyBoard_config not clean the world if the resulting setup would be identical?
<mnemoc>
hno: and so only rebuild the needed parts
FR^2 has joined #linux-sunxi
<Nyuutwo>
what does Rocky LCD_POWER_EN vb_en=1 from drv_de.drv does?
<wigyori>
arokux: what could be the reason if olinuxino-a13 brings up the usb hub, but finds no devices on it? to test it, i've added in usbstorage (for a pendrive) and rtl8192cu (for the onboard wifi), but doesn't seem to find them
rellla has joined #linux-sunxi
vinifr has quit [Ping timeout: 272 seconds]
notmart has quit [Quit: notmart terminated!]
<Turl>
wigyori: can you see them on lsusb?
<Turl>
(or dmesg)
moofree has quit [Ping timeout: 272 seconds]
Fusing2 has joined #linux-sunxi
moofree has joined #linux-sunxi
Fusing has quit [Ping timeout: 240 seconds]
<hramrach>
mnemoc: it's not cleared. not everything is rebuilt
<hramrach>
build takes much longer when you do clear
<mnemoc>
weird... in my nightlies, using a separated build dir for each board everything seems to be rebuilt
<mnemoc>
maybe just a side effect of touching the config file
<hramrach>
there are some libraries that are kept but the board specific parts are rebuilt - something like that
<hramrach>
it does not save the board choice anywhere so it does not know it's the same
<hramrach>
And there's probably some make deficiency which makes this less evil design than making a targer for each object as configures with different option combination
zoobab has quit [Read error: Connection reset by peer]
<mnemoc>
what I usually do in _config like Make rules is to generate a temporal file, and only replace if they don't cmp
<mnemoc>
so avoiding touching it when not needed
torindel has quit [Ping timeout: 245 seconds]
<mnemoc>
and I really hate mali's driver ALWAYS getting recompiled
torindel has joined #linux-sunxi
jemk has joined #linux-sunxi
<pfdm>
jemk: can i ask you why "ve_virt2phys(os->vs->data) + 0x40000000; " adding 0x40000000 ?
<wigyori>
Turl: no emac on this one so can't get it to nfsroot - the plan is to have usbethernet, but without usb detecting devices plugged in, it won't detect a usbethernet either
<wigyori>
it finds the ehci, the usb hub, but that's it:
<wigyori>
[ 0.655961] hub 1-0:1.0: USB hub found
<wigyori>
[ 0.659732] hub 1-0:1.0: 1 port detected
<jemk>
pfdm: for cedar the ram starts at address 0x00000000, for all other modules like cpu and disp it starts at 0x40000000
<jemk>
pfdm: and the ve_ functions are for cedar and thus return cedar addresses
<mnemoc>
that might deserve a help macro
<mnemoc>
like #define cpu_virt2phys(N) (ve_virt2phys(N) + 0x40000000)
<pfdm>
jemk: ok. so the adress you keep in the surface output need to be transform back in the same way before any use ?
rz2k has quit []
<jemk>
pfdm: where do you mean? all address from ve_virt2phys are for cedar, if you want to give them to disp you have to add 0x4..
<pfdm>
jemk: i want to use the decoded picture to give them to another render than disp in xbmc for example
<pfdm>
jemk: Does this 0x4 is related to that flag CEDARV_FLAG_DECODE_NO_DELAY ? as it appears no where in avheap.c or in the code of libcedarv. I've seen this address in the sunxi_allocator too.
<jemk>
pfdm: sry, 0x40000000, not 0x4, that's simply a address offset, no flag
<jemk>
pfdm: if you need a virtual address (for memcpy for example) you can use the stored address, if you need a physical you can use ve_virt2phys + 0x40000000
geecko has joined #linux-sunxi
<pfdm>
jemk: result of yesterday discussion is to try to access the physical memory by UMP and give that adress to mali for gles rendering
torindel has quit [Ping timeout: 240 seconds]
torindel has joined #linux-sunxi
<jemk>
pfdm: ok, don't know how ump works, but i think it needs physical address, but as far as i know in kernel space.
<pfdm>
jemk: but actually i patched xbmc to use your libvdpau, if i don't use your presentation_queue_display the yvbuffer in xbmc display a green picture. that's why i thought the adress in the surface_output needs to be transform to virtual before.
<pfdm>
jemk: yes sure, it's just a way to avoid copy the texture before using it. I'll try and let you know. I just wanted to be sure of the nature of the surface_output data pointer you use in libvdpau.
<jemk>
pfdm: ah, be careful, surface_output doesn't really exist yet, it only has pointers to surface_video. the whole vdpau thing is only some proof of concept
arokux2 has joined #linux-sunxi
<arokux2>
wigyori: there?
<arokux2>
wigyori: I've just read your messages, have you resolved the issue?
popolon has joined #linux-sunxi
<pfdm>
jemk: ah i see. btw I've posted an issue on the github. Is there any tutorial on how to RE other codec ?
ZetaNeta has joined #linux-sunxi
tinti has joined #linux-sunxi
<wigyori>
arokux2: here, not yet
<arokux2>
wigyori: it worked with my branch, right?
<mnemoc>
arokux2: if your u-boot branch only cares about a few boards, can you just remove the others from boards.cfg in a temporal commit?
<mnemoc>
it will avoid confusions
<jemk>
pfdm: are you sure about this issue? if i'm not totaly wrong this is valid C, but this function is stub anyway, it needs to be finished some time.
<arokux2>
mnemoc: I've figured a better way. I'll add USB_EHCI_SUNXI to the boards where it is known to work instead of adding it to the include/configs/sunxi-common.h
<jemk>
pfdm: tutorial for RE... not that i know of, i believe this is learning by doing
<wigyori>
arokux2: with your branch, it only goes up to the same level, f.e. detecting the usb hub and 1 port, but does not find any dev on the usb chain
<arokux2>
mnemoc: then in include/configs/sunxi-common.h I'll do #ifdef and add USB stuff if needed.
<mnemoc>
ok
<hno>
mnemoc, there is a common header file updated when you make .._config, so no. makefile deps thinks it all need to be rebuilt.
<hno>
arokux2, ok. did a quick scan and looks reasoable.
<mnemoc>
hno: can I touch that part of the code? :p
<arokux2>
hno: thanks
<arokux2>
wigyori: it works for me
<hno>
mnemoc, sure, but it's not a sunxi change. Post to main mailintlist please.
<mnemoc>
ok
<nove>
pfdm, there is no RE tutorial, but there is this page http://linux-sunxi.org/CedarX/RE_Toolkit, is empty, but i was trying to fill out in how to start with the h264 encoder
<jemk>
pfdm: there are many things missing/wrong in vdpau, you are welcome to help, i don't have much time at the moment and would really like to use this little time for RE work instead of "simple" software work
<wigyori>
arokux2: can you pastebin a dmesg for me pls?
<arokux2>
wigyori: of course, sec
<pfdm>
jemk: yes , even how to start would be helpfull, also i'd love to see xbmc +hw decode on my cb2 , as soon as my test with gles and ump will be sucessfull , i could help improve libvdpau.
torindel has quit [Ping timeout: 265 seconds]
torindel has joined #linux-sunxi
n01 has quit [Read error: Connection reset by peer]
<arokux2>
mnemoc: sorry lots of things to do and this hackberry is driving me crazy
<mnemoc>
arokux2: I just mean that it might be a nice idea to improve the page we have about the hackberry while you fight it
<arokux2>
mnemoc: ok
<mnemoc>
and why the f* my nighlty script insist in rebuilding u-boot now?!!!!
<mnemoc>
it's the sae freaking hash every time
jeremb_ has quit [Remote host closed the connection]
<arokux2>
mnemoc: u-boot is shit... on an noop make lots of the things get rebuilt, additionally you can be fucked up without distclean and.... there is NOISE during the build process
<mnemoc>
arokux2: fine, but my script is supposed to not call the building script at all if the branch has not been touched
<arokux2>
mnemoc: yeah, I've just this very moment to complain about u-boot
FR^2 has quit [Quit: und weg...]
<mnemoc>
later I'll fix u-boot to not touch the common header on _config when not needed
<arokux2>
mnemoc: ehci-sunxi.c:59:18: error: 'CONFIG_USB_MAX_CONTROLLER_COUNT' undeclared here (not in a function)
<mnemoc>
uhm
<mnemoc>
that means Make is "accepting" the error
<mnemoc>
in the case of linux I have to `elif grep -q -e '\[sub-make\]' "$log.out"; then` to detect errors :|
<mnemoc>
I suppose I'll have to do something similar with u-boot
<arokux2>
mnemoc: even kernel behaves that stupid?!
<mnemoc>
$ make; echo $?
<mnemoc>
make: *** No targets specified and no makefile found. Stop.
<mnemoc>
2
<mnemoc>
$ make 2>&1 | tee /dev/null; echo $?
<mnemoc>
0
<mnemoc>
make: *** No targets specified and no makefile found. Stop.
<mnemoc>
f* sh
<arokux2>
mnemoc: u-boot wants CROSS_COMPILE=arm-linux-gnueabihf- even for distclean
<wigyori>
arokux2: okay, no luck with your sunxi-usb-next tree, the config you've provided and the sun5i-a13-olinuxino.dtb - http://pastebin.com/fsu9SaBn
<mnemoc>
arokux2: as I use build dirs, I don't distclean, I just rm -rf :p
<arokux2>
wigyori: o_O
<arokux2>
wigyori: ah...
<arokux2>
wigyori: the kernel is unable to find root, so it bails out and does not continue to boot