<oliv3r>
this is how pinmuxing is done for regular gpio, which is fine
<oliv3r>
but how does this deal with multiple address spaces?
<oliv3r>
it doesn't and can't :S
<mripard>
ah, I see
rm has quit [Ping timeout: 264 seconds]
<oliv3r>
but i'm not that smart :)
<oliv3r>
so either it needs to be somewhat rewritten, to take care of double addresses properly
<oliv3r>
or, not rely on register -> struct mapping :)
<oliv3r>
i'll show you my hack in an update mail later today
<oliv3r>
it was the least invasive way
<mripard>
ok
rm has joined #linux-sunxi
<oliv3r>
until it's rewritten ;)
<oliv3r>
mripard: do you have any way to test if pinmux is setup properly from u-boot-spl?
paulk-desktop has joined #linux-sunxi
BJfreeman has quit [Quit: had a good time]
<mripard>
oliv3r: JTAG/printf? :)d
bsdfox_ has joined #linux-sunxi
bsdfox_ has quit [Changing host]
bsdfox_ has joined #linux-sunxi
bsdfox has quit [Ping timeout: 256 seconds]
stekern has quit [Ping timeout: 246 seconds]
stekern has joined #linux-sunxi
hglm has joined #linux-sunxi
<hglm>
I was surprised to find Debian's armhf userland is already using Thumb2 mode, which results in 20-30% smaller code size than ARM mode and performs better (less cache misses etc).
<rm>
for everything?
<hglm>
Only thing missing is a kernel compiled in Thumb2 mode, when trying it compiles OK (with gcc 4.8) but doesn't seem to boot.
<rm>
doesn't this have some downsides too?
<rm>
I thought thumb2 implied smaller, but not at all necessarily faster code
<hglm>
rm: I think almost everything is Thumb2 compiled.
<hglm>
rm: In my experiments Thumb2 doesn't seem to have much downsides -- it's a little faster. Of course it also depends on the compiler used because Thumb2 has some extra instructions and a few instructions are different.
<oliv3r>
mripard: i hope your jtag works thenh :)O
<mripard>
I don't have one :)
<mripard>
oh, speaking of which
<mripard>
hno: Tsvetan: I'll probably get one JTAG interface, do you have any recommendation (ideally working with OpenOCD)?
<oliv3r>
bus blaster!
<oliv3r>
i have bus blaster and bus pirate :)
<oliv3r>
they both work with openOCD afaik
<hglm>
I have been looking for references of a working Thumb2 *kernel* for sunxi but haven't found one -- it should work in principle but I think there must be a non-compliant piece of code in the sunxi-specific part of the kernel that fails on Thumb2.
<mripard>
oliv3r: wow, nice.
<oliv3r>
mripard: i have never used my bus blaster
<oliv3r>
it was an impulsive buy
<oliv3r>
i even have one of those rigol osciloscopes
<oliv3r>
i have used that one :p
<oliv3r>
but i admit, i don't use it as much as I should, concidering the cost
<oliv3r>
but it's still very nice to have and i have examined tons of stuff with it :)
<oliv3r>
i'm not an EE
<mripard>
yeah, me neither
<oliv3r>
i know how to solder, i know how to make very basic PCB's in geda
<oliv3r>
and that's where it ends :p
<mripard>
:)
dragonn has joined #linux-sunxi
<atiti>
hm, can someone shed some light on how to use the meminfo dump + script.bin to make my device work? :p
ykchavan has quit [Quit: Leaving]
<atiti>
i have an mk802+ with A10s and I guess a different memory chip then the ones hans made his image for
<geecko>
i've been scratching my head all morning on this
<oliv3r>
the wiki is resourcefull
<geecko>
since it will be used as an headless server, i can disable all this stuff without a problem?
<oliv3r>
yep
<geecko>
anyway, i will try ;)
<oliv3r>
i suppose even the framebuffer can go
<geecko>
i was searching for "allwinner a10" in google without much results
<oliv3r>
so you could have the full ~512mb
<geecko>
\o/
<oliv3r>
linux-sunxi :)
<geecko>
ok :D
tinti has quit [Ping timeout: 256 seconds]
<hglm>
Is there any way to get early kernel messages from a kernel that doesn't boot (no video output) on a tablet? Do you need to use a USB-to-serial converter?
Tartarus_ has quit [Quit: Ex-Chat]
Black_Horseman has joined #linux-sunxi
tinti has joined #linux-sunxi
<atiti>
hansg: are you here?
<hansg>
yes, but busy, try again in an hour
wingrime has quit [Ping timeout: 248 seconds]
<atiti>
alrighty :)
wingrime has joined #linux-sunxi
[hawk] has joined #linux-sunxi
vincenzo has quit [Ping timeout: 256 seconds]
rz2k has joined #linux-sunxi
<oliv3r>
wow, just got a 27" to replace my 24" at $work
<oliv3r>
4 terminals in a row :D sexeh
<andoma>
sweet
<oliv3r>
hglm: usb to serial or uSD to serial
<oliv3r>
andoma: hell yeah; wish i could take it home and duplicate it a few times :)
[hawk] has quit [Quit: Konversation terminated!]
[hawk] has joined #linux-sunxi
[hawk] has quit [Remote host closed the connection]
[hawk] has joined #linux-sunxi
<hansg>
atiti, I'm available now
<atiti>
hi
<hansg>
hi :)
<atiti>
any chance you could take a very quick glance on my meminfo dump and see if the small difference between the one you have for the mk802+ and mine could be causing some issues?
<atiti>
i have no idea how to check it :(
<hansg>
atiti, sure, I had that on my todo list for today anyways, let me check it right away
<atiti>
sweet, the only difference I saw was the io_width being 16 instead of 8
<oliv3r>
henrik commented on that on the Mailing list
rellla2 has joined #linux-sunxi
hansg has quit [Read error: Operation timed out]
shalem has joined #linux-sunxi
shalem is now known as hansg
<hansg>
atiti, my pc thought this was a good moment to crash, looking at your memdump now
<atiti>
hehe ok :)
<hansg>
atiti, so you're right the only difference is 8 versus 16. And the 8 is correct, even for your board. with a bus width of 32 you either have 2 16 bit ram chips, or 4 8 bit ram chips. And you've 4 8 bit ram chips
<atiti>
hm alright
<atiti>
do you have any good tips on how to figure out why it doesn't boot then? :p
<oliv3r>
the setting is used for drive strength if i recall correctly
<oliv3r>
do you boot from SD?
<hansg>
atiti, what was the url for the PCB pictures again (I want to google the dram chips on your pcb)
<hansg>
Hmm, those are 1.5v 2Gbit (8x256M) ddr3-1600 ic-s, which matches the config we're using
<hansg>
atiti, does the hdmi out work when you boot the android ?
<atiti>
it goes completely black
<hansg>
fex file also has nothing interesting
<atiti>
theres some signal probably because the monitor doesn't complain about not having any signal
<hansg>
atiti, with android it goes completely black, and with the Fedora 18 sdcard ?
<atiti>
no, with android it displays the android boot screen and boots up
<atiti>
with fedora 18 it goes black
<atiti>
oh and the other thing, the two leds (red and blue) light up almost immediately on power-on
<atiti>
and after a couple of seconds the red light flashes once
<hansg>
ok, can you try to edid uEnv.txt on the first partition of the sdcard, and remove the EDID: in front of 1280x720p60
<atiti>
kind of like the device reboots or smtg
<hansg>
Also why you've the sdcard in a pc, can you check /var/log/messages on the 3th partition. Maybe even upload it to your website ?
tinti has quit [Ping timeout: 248 seconds]
<atiti>
messages is empty
<atiti>
wooo, removing EDID did make it output stuff :D
rellla2 has quit [Remote host closed the connection]
<atiti>
fails to mount the rootfs
<atiti>
damn cant see all the messages cos of this monitor overscan
<atiti>
the last two lines i see complains that it couldnt mount mmcblk0p3 as ext3 or 2 due to feature incompatibilities
<hglm>
atiti: I get the same on normal boots but below it is says it succeeds mounting as ext4, so the root fs may be mounted correctly.
<hansg>
atiti, you can also try removing the other EDID: in uEnv.txt. I don't expect much from that, but you can try :)
<hansg>
atiti, I assume you are using the script.bin from the Fedora image, right? Or did you copy over the one from android ?
\\Mr_C\\ has quit []
<atiti>
hansg: ye, the one from the fedora image
<atiti>
hm now i get to the fedora firstboot
<atiti>
:D
<atiti>
just the input devices don't work
<oliv3r>
usb keyboard should work
<hansg>
atiti, you need to use a hub in the "real" usb port and then plug both your keyboard and mouse into the hub. It does not need be a powered hub.
<oliv3r>
touchscreen is highly likly to not work
<oliv3r>
does the mk802+ even have a real usb port? my tablet only has otg port and works quite well with a keyboard (or a mouse, not together, but that's known)
<hansg>
oliv3r, yes the mk802+ has a real port. Which for some reason does not work if you plug a device in directly, only works with a hub in between ...
tinti has joined #linux-sunxi
<atiti>
sweet, that did it
<atiti>
woohoo, this is great! Thanks guys :D
andoma has quit [Ping timeout: 252 seconds]
<hansg>
atiti, can you please collect the output of "dmesg" and mail it to me, also try to boot with the EDID: back, then wait say 10 minutes, then power off, and see if there is anything (new) in /var/log/messages ?
<hansg>
atiti, I would like to get your tv / monitor to work with EDID.
<oliv3r>
hansg: ah, strange bug :S mk802+ not one of those HDMI sticks?
<oliv3r>
hansg: my monitor refuses to work with fedora
<oliv3r>
+edid*
<oliv3r>
it's a HP which tries to scan before displaying or something
<atiti>
hansg: sure, doing it now
andoma has joined #linux-sunxi
<atiti>
it's an LG L245WP-BNQ
Soru has joined #linux-sunxi
<hansg>
oliv3r, the mk802_a10s which atiti has is a hdmi stick
<hansg>
oliv3r, about your not working with edid, I would not mind getting a dmesg output of both a successful (so no edid) as an unsuccessful boot, then I can see if anything stands out (and I can fix it)
andoma has quit [Ping timeout: 276 seconds]
<wingrime>
oliv3r: I get hdmi using fex but sceen have unseeable borders
<geecko>
oliv3r, how can i change u-boot's environment variables
<geecko>
are these stored on the sd-card?
<geecko>
i tried to extract it using dd but i failed
<geecko>
now i'm trying to use u-boot's shell
<geecko>
is it possible through the usb otg connector?
<atiti>
hansg: it seems it never boots once I add back the EDID again
<atiti>
at least after waiting 10 mins, nothing new ever shows up in /var/log/messages
andoma has joined #linux-sunxi
wingrime has quit [Ping timeout: 246 seconds]
_BJFreeman has joined #linux-sunxi
<hansg>
atiti, and we don't have a serial console to see the kernel crash, bummer. Well maybe something useful will be in the dmesg of a working boot, send that my way please, and I'll see what can do.
_BJFreeman is now known as BJfreeman
<atiti>
sent to your rh address, hope you can use it for smtg :)
<geecko>
okay, i managed to mount the boot partition
<geecko>
i need to create boot.scr with my options
<geecko>
how ;_;
wingrime has joined #linux-sunxi
<oliv3r>
hansg: will try, i have a rc2 SD image ready to go with fd kernel and fd script.bin etc
<oliv3r>
hansg: actually, it was on the cubie1, not my tablet, tablet worked flawlessly, but it's LCD obivously
eebrah is now known as eebrah|away
<oliv3r>
geecko: you can change the uEnv.txt on the fat partition; editing the uEnv partition is trickier, not sure what happens if you erease it, and only rely on uEnv.txt; hansg should know that
<geecko>
which fat partition?
<hansg>
oliv3r, there is no uEnv.txt partition, the actual uEnv is a part of the sdcard before the first partition. If you erase that u-boot will detect a bad checksum there and fallback to its build-in defaults.
<geecko>
hansg, i'm wondering if the default environment will be loaded if i put a partial boot.scr here?
<geecko>
i only want to put 'sunxi_ve_mem_reserve=0' for example
<hansg>
geecko, it is best for these things to simply edit the uEnv.txt text file
<geecko>
hansg, ok, where can i find it?
<oliv3r>
true, it's not really a partition, we just write it at offset 54something
<hansg>
It is part of the uboot partition of the Fedora-18 images I maintain.
<oliv3r>
but uENv.txt is ON the fat partition :)
<hansg>
geecko, if you're rolling your own images you need to figure this out yourself
<geecko>
i dunno, i'm on a new project here
<geecko>
i have one sd card with debian
<geecko>
two partitions
<geecko>
1 with 'script_a10s_linux_HDMIscript_a10s_linux_LCD_800x480 script.bin uImage'
<geecko>
the other with the / filesystem
<geecko>
there's also an onboard NAND
<geecko>
i'm not even sure how to access it
\\Mr_C\\ has joined #linux-sunxi
<geecko>
hmmm
<geecko>
oh yay!
<geecko>
the first partition is using FAT
<geecko>
hansg, so i only have to put one file here and it's done? let's see ;)
<hglm>
Using uEnv.txt is not explained anywhere on the wiki, it would be a nice addition. The only info is about boot.scr/boot.cmd.
<hansg>
hglm, IIRC for uEnv,txt to work, some extra stuff is needed in boot.cmd, it was a while ago I put this all together for the Fedora images..
<hansg>
It is time for me to go, see you all tomorrow.
hansg has quit [Quit: Leaving]
<hglm>
geecko: Your Debian image may not support uEnv.txt yet, but editing boot.cmd and converting it to boot.scr will work.
<geecko>
hglm, can i put a boot.cmd file containing:
<geecko>
sunxi_ve_mem_reserve=0
<geecko>
sunxi_g2d_mem_reserve=0
<geecko>
sunxi_fb_mem_reserve=16
<geecko>
sunxi_no_mali_mem_reserve
<geecko>
?
<hglm>
You have have to add them to the setenv bootargs line (the kernel arguments)
<geecko>
hglm, but i don't know the default arguments...
<hglm>
geecko: Possible, the latest stable 3.0 may also support those options. But generally I think 3.4 is preferable.
<ssvb>
geecko: if you are not interested in android, then 3.4 is probably a better choice nowadays
<geecko>
ssvb, i'm using debian as an headless server
atiti has quit [Ping timeout: 240 seconds]
<ssvb>
then go for 3.4, afaik the only reason 3.0 is still not dropped is that android has some android specific drivers not properly forward ported to 3.4
<hglm>
ssvb: Do you have experience with kernels compiled with Thumb2? It doesn't seem to work on sunxi.
<geecko>
ssvb, hglm, will the inux-sunxi 3.4 kernel work without any modification on a oLinuXino board?
<ssvb>
hglm: no, imho it is not worth the effort
<hglm>
ssvb: OK, but from limited testing I get the impression that it could bring some benefits in real-world performance (old thumb gives it a bad reputation). Benchmarks that fit in the L1 cache won't show much difference, but for large memory footprints (such as real-world usage) using Thumb2 might be faster.
<ssvb>
hglm: the kernel itself is not that large, and I prefer it to be as much stable as possible because debugging it is a pain
<ssvb>
hglm: but even in the userland, thumb2 heavily depends on the compiler code generation quality (sometimes it is good, sometimes it is not so much)
<hglm>
ssvb: point taken, and I agree on the compiler quality affecting Thumb2 code (Thumb2 has some clever new instructions/instruction combinations that can be used to speed things up). Maybe gcc 4.8+ improves on this.
<geecko>
ssvb, hglm, can i use sun4i_defconfig without any problem?
<ssvb>
hglm: but at least thumb2 is good for getting measurably higher power consumption in cpuburn :)
<ssvb>
hglm: this seems to be cortex-a8 specific, because cortex-a9 and cortex-a7 are rather neutral to arm/thumb2 mode selection in this respect
<ssvb>
hglm: maybe the thumb2 instruction decoder in cortex-a8 was not properly optimized for lower power consumption
<geecko>
oh nevermind i found a defconfig <3
<ssvb>
hglm: have you tried to run some benchmarks for arm vs. thumb2?
<hglm>
ssvb: thumb2 is supposed to help with power consumption, maybe it's specific to cpuburn.
<hglm>
ssvb: Yes, I am running debian which has thumb2 userland, and optimizing my experimental memcpy function shows a slight benefit to thumb2 in optimized asm. But I haven't really tested larger benchmarks that are affected by C compiler quality.
<ssvb>
hglm: that's what they are writing in their advertisement materials (they can't write that their super-duper new feature is crap, do they?)
<ssvb>
hglm: and the reduction of I-cache misses may be indeed very good for lower consumption
<hglm>
thumb2 does seem to be the way forward with ARM to the point of some ARM processors only supporting Thumb2 and not ARM mode.
<ssvb>
hglm: 64-bit ARM ISA does not have any thumb counterpart yet (any sort of variable length instructions encoding for code size reduction)
<ssvb>
maybe it turned out to be not so great as they initially anticipated
<hglm>
ssvb: I guess 64-bit would need a whole new thumb instruction set.
^hawk^ has quit [Remote host closed the connection]
^hawk^ has joined #linux-sunxi
wingrime has quit [Ping timeout: 264 seconds]
atiti has quit [Ping timeout: 248 seconds]
<oliv3r>
samsung buys boxee
atiti has joined #linux-sunxi
<Turl>
oliv3r: heh
geecko has joined #linux-sunxi
nove has joined #linux-sunxi
Superpelican has joined #linux-sunxi
Soru_ has joined #linux-sunxi
Soru has quit [Ping timeout: 276 seconds]
<geecko>
hard float kernel on a soft float debian image => would it run?
<oliv3r>
why not
<geecko>
oliv3r, i dunno, that's why i'm asking :D
<hno>
geecko, kernel is always hard float on a cpu with hard float support.
<hno>
the kernel as such do not use float, but needs to be built for the CPU it's running on.
<geecko>
hno, i am asking because, i see in a tutorial that gnueabi is used. this is a mess to install for me (i'm using Arch Linux). on the other hand, i successfully compiled the kernel with gnueabihf
<hno>
geecko, you'll be fine.
<geecko>
i was thinking that it would not work on the stock debian image
<hno>
kernel is ABI neutral.
<geecko>
great :)
<geecko>
i now have to turn zImage into a uImage
<hno>
geecko, sunxi kernels have an uImage build target, just "make ... uImage"
<geecko>
hno, strange, i used it but i only have a zImage in the output folder
<hno>
the uImage is in arch/arm/boot/ I think. It says where during the build.
<geecko>
hno, yep, it isn't there :(
eebrah|away has quit [Ping timeout: 264 seconds]
dwilkins has quit [Ping timeout: 264 seconds]
dwilkins has joined #linux-sunxi
hglm has quit [Quit: leaving]
eebrah|away has joined #linux-sunxi
eebrah|away is now known as eebrah
eebrah is now known as Guest53935
Guest53935 is now known as eebrah_
<oliv3r>
hno: Hey, hows your vacation going :)
<geecko>
hno, oliv3r, what happens if i don't copy the compiled modules with the kernel?
<geecko>
3.0.56 vs 3.0.7x
<geecko>
will it boot?
<Turl>
geecko: it should, but without the modules you might lose functionality
<geecko>
Turl, ok
<Turl>
hno: as far as I know, the uImage target is everywhere, not just on sunxi
\\Mr_C\\ has joined #linux-sunxi
<geecko>
oh lol
<geecko>
"mkimage" command not found - U-Boot images will not be built
<geecko>
here's why
Soru has joined #linux-sunxi
Soru_ has quit [Ping timeout: 260 seconds]
<ssvb>
geecko: you need u-boot tools package installed :)
<geecko>
Image arch/arm/boot/uImage is ready
<geecko>
yay <3
<Dreadlish>
now only modules and can boot up ;p
<geecko>
drachensun, done
<geecko>
now... wait for tomorrow morning :3
Superpelican has quit [Quit: Konversation terminated!]
n01_ has joined #linux-sunxi
nove has quit [Quit: nove]
bsdfox_ has quit [Read error: Connection reset by peer]
^hawk^ has quit [Remote host closed the connection]
^hawk^ has joined #linux-sunxi
<Turl>
rellla: spammers attack still? :(
eebrah_ is now known as eebrah|away
^hawk^ has quit [Read error: Connection reset by peer]
^hawk^ has joined #linux-sunxi
<hno>
Turl, spammers always look for attack points,
^hawk^ has quit [Remote host closed the connection]
^hawk^ has joined #linux-sunxi
cajg has quit [Quit: WeeChat 0.4.0]
cajg has joined #linux-sunxi
<Turl>
hno: yes, but I even enabled DNSBL blocking of them :(
^hawk^ has quit [Remote host closed the connection]
^hawk^ has joined #linux-sunxi
_BJFreeman has joined #linux-sunxi
^hawk^ has quit [Remote host closed the connection]
_BJFreeman is now known as BJfreeman
^hawk^ has joined #linux-sunxi
tinti has quit [Quit: Leaving]
<rellla>
Turl: yeah it's like playing cat and mouse
^vincenzo^ has joined #linux-sunxi
geecko has quit [Quit: Quitte]
^hawk^ has quit [Ping timeout: 246 seconds]
hramrach__ has quit [Ping timeout: 240 seconds]
^vincenzo^ has quit [Remote host closed the connection]
^vincenzo^ has joined #linux-sunxi
<drachensun>
geecko: ???
<drachensun>
and he is gone
andoma has quit [*.net *.split]
andoma has joined #linux-sunxi
winocm has quit [Ping timeout: 260 seconds]
notmart has quit [Quit: notmart terminated!]
winocm has joined #linux-sunxi
^vincenzo^ has quit [Remote host closed the connection]
^vincenzo^ has joined #linux-sunxi
rellla has quit [Remote host closed the connection]
rz2k has quit []
ykchavan has quit [Quit: Leaving]
^vincenzo^ has quit [Remote host closed the connection]
^vincenzo^ has joined #linux-sunxi
hramrach__ has joined #linux-sunxi
n01_ has quit [Ping timeout: 276 seconds]
paulk-desktop has quit [Quit: Ex-Chat]
_BJFreeman has joined #linux-sunxi
BJfreeman is now known as Guest16491
_BJFreeman is now known as BJfreeman
Guest16491 has quit [Ping timeout: 252 seconds]
\\Mr_C\\ has quit []
^vincenzo^ has quit [Remote host closed the connection]
ZaEarl_ has quit [Ping timeout: 248 seconds]
BJfreeman has quit [Quit: had a good time]
^vincenzo^ has joined #linux-sunxi
ZaEarl__ has joined #linux-sunxi
^vincenzo^ has quit [Remote host closed the connection]
^vincenzo^ has joined #linux-sunxi
tinti has joined #linux-sunxi
specing has quit [Ping timeout: 264 seconds]
specing has joined #linux-sunxi
servili007 has joined #linux-sunxi
<servili007>
libv: how mature is lima at this point? I'm trying to wrap my head around how usable it is since I'm seeing projects like picuntu claiming to use it