<ikeeki>
Any help will be highly appreciated. Bye.
ikeeki has quit [Quit: Saliendo]
gianMOD has joined #linux-sunxi
gianMOD has quit [Read error: Connection reset by peer]
gianMOD has joined #linux-sunxi
a1d3s is now known as a1d3s|away
gianMOD has quit [Remote host closed the connection]
gianMOD_ has joined #linux-sunxi
gianMOD_ has quit [Ping timeout: 264 seconds]
<el3>
Hi I finally managed to build an image with yocto. In meta-sunxi I can not choose pcduino, so I made a cubieboard2 image. Is it possible to make it run on pcDuino3?
jinzo has quit [Ping timeout: 250 seconds]
egbert has quit [Disconnected by services]
egbert has joined #linux-sunxi
cnxsoft has joined #linux-sunxi
ktemkin has joined #linux-sunxi
gianMOD has joined #linux-sunxi
gianMOD has quit [Ping timeout: 264 seconds]
eagles0513875_ has quit [Ping timeout: 245 seconds]
HeavyMetal has quit [Ping timeout: 245 seconds]
<ktemkin>
ssvb: I just saw that we were editing the Primo 73 wiki page at the same time, and thought I'd say hello. (Just a short while ago, I added board pictures and instructions for adding a UART that doesn't take up the microSD slot.)
HeavyMetal has joined #linux-sunxi
<ssvb>
ktemkin: hi, thanks a lot
<ktemkin>
No problem-- I need some excuse to play with cool lab equipment. ;)
<ssvb>
was it one of that dirt cheap tablets from newegg?
<ktemkin>
Yeah, I grabbed it once the price came down to ~$50.
<ssvb>
pink color is a bit unusual choice, unless this was a bargain :)
<ktemkin>
I haven't seen any non-pink ones left-- and the color doesn't matter as much as the stuff inside.
<ktemkin>
I needed to test out a series of tablets with easily replaceable bootloaders, so I could ensure that I could remain in non-secure hypervisor mode after I left the bootloader.
eagles0513875_ has joined #linux-sunxi
faisal has joined #linux-sunxi
akaizen_ has quit [Remote host closed the connection]
akaizen has joined #linux-sunxi
akaizen has quit [Ping timeout: 250 seconds]
gianMOD has joined #linux-sunxi
gianMOD has quit [Ping timeout: 246 seconds]
p1u3sch1 has joined #linux-sunxi
Andy-D has joined #linux-sunxi
p1u3sch1_ has quit [Ping timeout: 255 seconds]
marcin has quit [Ping timeout: 264 seconds]
HeavyMetal has quit [Ping timeout: 245 seconds]
xenoxaos has quit [Ping timeout: 245 seconds]
<el3>
I am about to compile kernel http://linux-sunxi.org/Mainline_Kernel_Howto. But the compiling stops saying -> #error Your compiler is too buggy; it is known to miscompile kernels
HeavyMetal has joined #linux-sunxi
<el3>
How do I install another. I installed this one apt-get install gcc-arm-linux-gnueabihf
<el3>
?
<ktemkin>
el3: Is your (cross)-gcc version between 4.8.0 and 4.8.2?
<el3>
ktemkin: yes
<ktemkin>
You'll need to go older or newer-- I'm assuming you're on Ubuntu, and not using a "platform" like Yocto/Angstrom/OE/etc?
<el3>
ktemkin: when I try to follow the guide here http://linux-sunxi.org/Toolchain to get another toolchain, it says there should be a bin folder. But I cannot see a bin folder
<ktemkin>
You grabbed one of the gcc-linaro tarballs and extracted it?
<el3>
ktemkin: yes
<el3>
ktemkin: tried 4.7 and 4.9
<el3>
no bin folders
<ktemkin>
Let me grab one of the tarballs listed and take a look. (Those look like they may be toolchain source tarballs, and not the binary ones you're looking for).
<el3>
ktemkin: thank you.
arossdotme has quit [Remote host closed the connection]
<ktemkin>
I've been using gcc 4.9.1 without issue-- but YMMV.
TheSeven has joined #linux-sunxi
<el3>
ktemkin: I will try. Have been struggliong with this forever now :D
<ktemkin>
A little bit of struggle is good for you: that's how you learn things. ;)
<el3>
ktemkin: yea lets hope so. Now I moved the bin directory to my build enviroment. And then I shall run this ARCH=arm CROSS_COMPILE=<toolchain-prefix> LOADADDR=0x40008000 make uImage dtbs . Do you know what should I put in <toolchain-prefix>
<el3>
?
marcin has quit [Remote host closed the connection]
<ktemkin>
In that bin directory, you should see a file that looks like arm-linux-gnueabihf-gcc. It'll be whatever the name is before gcc, including the final "-"
<el3>
ok
<ktemkin>
so, for "arm-linux-gnueabihf-gcc", you'd set CROSS_COMPILE=arm-linux-gnueabihf-
marcin has joined #linux-sunxi
<el3>
ktemkin: next error :D arm-linux-gnueabihf-gcc: error while loading shared libraries: libstdc++.so.6: cannot open shared object file: No such file or directory
rah has quit [Ping timeout: 245 seconds]
rah has joined #linux-sunxi
jero has quit [Ping timeout: 245 seconds]
<ktemkin>
I don't recall if there's a Ubuntu package that provides that-- try sudo apt-get install libstdc++6 ?
jero has joined #linux-sunxi
<el3>
ktemkin: I allready got the newest it says
<el3>
ktemkin: I Might have done something wrong on this step. export PATH="$PATH":/home/user/folder/gcc-linaro-arm-linux-gnueabihf-*_linux/bin/
<ktemkin>
Doesn't seem likely. More likely those binaries are 32-bit and you have a 64-bit libstdc++6.
<el3>
ktemkin: I will try the original bin. Not the one I copied to the enviroment
<ktemkin>
Can you run "file $(which arm-linux-gnueabihf-gcc)" , and report the output?
<el3>
2sec
marcin has quit [Read error: Connection reset by peer]
<el3>
ktemkin: /home/el3/linuxSunxi/linux/bin//arm-linux-gnueabihf-gcc: symbolic link to `arm-linux-gnueabihf-gcc-4.9.2'
<ktemkin>
Then "file $(which arm-linux-gnueabihf-gcc-4.9.2)", since that was a symlink.
<el3>
/home/el3/linuxSunxi/linux/bin//arm-linux-gnueabihf-gcc-4.9.2: ELF 32-bit LSB executable, Intel 80386, version 1 (SYSV), dynamically linked (uses shared libs), for GNU/Linux 2.6.15, stripped
<el3>
ktemkin: I see 2 // between bin//arm is that normal?
gianMOD has joined #linux-sunxi
arossdotme has joined #linux-sunxi
gianMOD has quit [Ping timeout: 255 seconds]
<el3>
ktemkin: Finally its compiling. Had to use the full path like ->ARCH=arm CROSS_COMPILE=/home/el3/Downloads/gcc-linaro-arm-linux-gnueabihf-4.9-2014.09_linux LOADADDR=0x40008000 make uImage dtbs
<el3>
*/bin/gcc-lina...gcc
marcin has joined #linux-sunxi
<ktemkin>
Ah, yeah-- you were appending it to your path, but there was another result first.
<ktemkin>
It uses the first matching result on your path-- so you can fix that by running this line instead:
<el3>
Oh yes it looks very messy when I do echo $PATH
<el3>
How can I clean the $PATH
<ktemkin>
close and re-open your terminal; your exported modifications only last for that shell session
<ktemkin>
For now, you can stick the correct path at the front of your PATH:
<ktemkin>
(To answer your earlier question: the // wasn't hurting anything-- it was there because there aren't really supposed to be /'s at the end of the entries in PATH. Fortunately, they don't do anything. / winds up the same as /.)
<el3>
ktemkin: Okey thank you
<ktemkin>
No problem.
<el3>
I did build a debian image before. Not the mainline. It was very buggy, crashed all the time. So I was adviced to try yocto. Made an image for cubieboard2, but pcduino3 would not boot on that.
<el3>
Then I went back to try the mainline. Failed last time. The problem we fixed right now. :)
<ktemkin>
There's no Yocto BSP for the pcduino3, eh?
<el3>
So I hope this runs more smoothly. I only need a headless system with g_mass_storage. And gpio
<el3>
No there is not
<el3>
But someone adviced me to look into that
<ktemkin>
That could have worked, but you would have had to tweak the configuration-- provided your own kernel (and possibly u-boot) defconfig and overridden the configuration variables in the BSP section of meta-sunxi.
<el3>
Yea that seems a bit difficoult. I need to know this stuff better
marcin has quit [Read error: Connection reset by peer]
<ktemkin>
If you're using an initramfs, then you'll have to enter the address you loaded it to in memory there; otherwise (IIRC) a dash.
<el3>
Okey, before compiling the kernel, I saw initramfs enabled. But I have no Idea what the address is
<ktemkin>
If you are, you should have an option that loads the initramfs into memory-- like the "fatload" lines you have. The address to which that is loaded is the same address you want to pass to bootm.
marcin has quit [Ping timeout: 264 seconds]
<el3>
ktemkin: Can I choose something there like initrd=0x47000000
<ktemkin>
If you want to use one, yes. The initramfs isn't necessary in many cases-- unless, for example, you have a module you need to load in order to attach the main disk (or you want the recovery busybox it provides).
<el3>
ktemkin: Okey so the image will work if I omit it, even if in the kernel config I enabled it?
nicksydney has quit [Ping timeout: 244 seconds]
<ktemkin>
Enabling initrd support doesn't imply that you'll need to generate one-- that just means the kernel has support for using them.
<ktemkin>
[There's a second option that will allow you to build an initramfs directly into the kernel, but since you haven't generated one of those yet, I doubt you've set it. ;)]
<el3>
ok
<ktemkin>
If it makes you feel better, there's a kernel command line argument you can pass that will explicitly tell the kernel to ignore any initrd/initramfs provided. (IIRC, it's noinitrd).
<el3>
I will just try. I suppose I can easily change the boot partition if it does not work
<ktemkin>
That's assuming you're not trying to do anything fancy like boot off SATA while SATA support was configured as a loadable module rather than built in to your kernel.
zombu2 has quit [Read error: Connection reset by peer]
Wizzup has quit [Quit: reboot]
zombu2 has joined #linux-sunxi
Wizzup has joined #linux-sunxi
Netlynx_ is now known as Netlynx
Netlynx has quit [Changing host]
Netlynx has joined #linux-sunxi
ganbold_ has quit [Ping timeout: 255 seconds]
konradoo77 has quit [Ping timeout: 265 seconds]
interrobangd has quit [Read error: Connection reset by peer]
konradoo77 has joined #linux-sunxi
Renard has joined #linux-sunxi
faisal has quit [Quit: Leaving]
f15h has quit [Quit: Leaving]
ganbold_ has joined #linux-sunxi
popolon has joined #linux-sunxi
<ssvb>
plaes: the -CG Nanya chips are DDR3-1333 9-9-9 (or DDR3-1333H) according to the table from page 6
<ssvb>
plaes: regarding down binning to DDR3-1066F support, the table for the DDR3-1333 9-9-9 (-CG) variant at the page 107 in the datasheet says that tAA/tRCD/tRP delay is 13.125 ns, which is normal for DDR3-1066F speed bin
<ssvb>
plaes: and also CL=7 / CWL = 6 for settings for tCK 1.875 are listed in the table too (albeit marked as " Optional")
<ssvb>
plaes: so it seems like the DDR3-1066F timings are supported by NT5CB256M8DN-CG, though they are kind of beating around the bush instead of saying this explicitly
konradoo77 has quit [Ping timeout: 252 seconds]
<plaes>
how do you calculate the MHz?
<plaes>
also, density is 2G for this chip?
<ssvb>
DDR3 is double data rate, so DDR3-1333 means clock speeds up to 666.5MHz and DDR3-1066 means clock speeds up to 533MHz
<zombu2>
1mhz + 1mhz = 11mhz
<zombu2>
sounds about right .... right?
konradoo77 has joined #linux-sunxi
dubito has joined #linux-sunxi
<ssvb>
also 1000 / 533MHz = ~1.876 (which corresponds to the tCK value 1.875ns)
<plaes>
ok.. current speed is 432MHz
<ssvb>
yes, the density is 2 Gb (Gbit) for this chip
<ssvb>
plaes: it is clocked at 432MHz, but the DDR3 chip itself supports clock speeds up to 667MHz
<ssvb>
plaes: and the 400MHz - 533MHz is the clock speed range, which is covered by DDR3-1066 timings
<plaes>
and max tRFC is 107ns?
<ssvb>
DDR3-1066 is either the faster 'F' speed bin or slower 'G' speed bin variant
<plaes>
page 105
<plaes>
or do I have to multiply by 1.5?
<ssvb>
yes, it is 107 cycles or ~160ns
<ssvb>
more detailed information about the standard speed bins is in the JEDEC DDR3 spec
ricardocrudo has quit [Ping timeout: 265 seconds]
<ssvb>
plaes: the standard JEDEC tRFC values, sepending on chip density - http://git.denx.de/?p=u-boot.git;a=blob;f=arch/arm/cpu/armv7/sunxi/dram_sun4i.c;h=c736fa3b474f5bbb3ed704fea566985c6c9b1011;hb=HEAD#l424
<plaes>
ok, so the line in the wiki should be following>
<plaes>
DDR3-1600K (supports down binning to DDR3-1066F), density: 2G, page size: 1K, tRFC: 160.5ns
<ssvb>
plaes: 160ns is pretty normal for 2Gb
<ssvb>
plaes: the NT5CB256M8DN-CG chip is DDR3-1333H
<ssvb>
plaes: page 6
<plaes>
yup
<plaes>
I read the table wrong
jinzo_ has quit [Ping timeout: 252 seconds]
<plaes>
DDR3-1333H (supports down binning to DDR3-1066F), density: 2G, page size: 1K, tRFC: 160.5ns
<ssvb>
looks right
atsampson has quit [Quit: new kernel]
sehraf has joined #linux-sunxi
<plaes>
current u-boot supports the +S:CONFIG_DRAM_TIMINGS_DDR3_1066F_1333H=y ?
<plaes>
u-boot master
<plaes>
hmm.. nope
<plaes>
duh.. I was wrong
<ssvb>
u-boot from git master supports this - it should get to http://git.denx.de/?p=u-boot.git;a=blob;f=board/sunxi/dram_timings_sun4i.h;h=29b934da639c86b1830d1e26b7a751b6b227a21a;hb=HEAD#l28
<plaes>
yeah.. I was grepping in configs directory
<ssvb>
the cas/tpr0/tpr1/tpr2/emr2 settings for 432MHz through the battery of ifdefs :)
<plaes>
because it was identical to the script.bin from my tablet
gianMOD has joined #linux-sunxi
<ssvb>
ok, just the vendor dram settings are not exactly great in general, but something has to be used as the initial approximation of a proper config
<ssvb>
plaes: have you run lima-memtester on this tablet?
<plaes>
it didn't work with mainline kernel :S
<ssvb>
just use sunxi-3.4 :)
<plaes>
:(
<ssvb>
does sunxi-3.4 not work correctly there or what?
<plaes>
it works
<ssvb>
so just boot it with the mainline u-boot and maybe run lima-memtester?
<plaes>
mhm
* plaes
tries setting the DRAM clk to 528MHz
premoboss has quit [Remote host closed the connection]
<plaes>
oh wait.. I can actually go up to 648?
<ssvb>
no, you can't
<ssvb>
unless you really find good zq settings and increase the vdd-dll voltage first
<ssvb>
the bottleneck is the soc, not the ddr3 chip
<plaes>
ok.. 528MHz then :P
<ssvb>
528MHz could be also unstable or even fail to boot
<ssvb>
you may want to ensure that 432MHz is really working properly first
<ssvb>
and then gradually increase it
<ssvb>
I do have the DDR3 clock speed set to 648MHz on my Cubietruck though :)
gianMOD has quit [Remote host closed the connection]
FlibberTGibbet has joined #linux-sunxi
<ssvb>
so it's not something really fantastic, but needs parameters tuning instead of just bumping up the dram clock speed and hoping for the best :)
<plaes>
every cell is a single test?
<plaes>
or do you extrapolate?
<ssvb>
yes every cell is a single test, but they are run automatically by a script
<ssvb>
it does not mean that you have to run each of them manually
gianMOD has joined #linux-sunxi
<ssvb>
each cell has a different 'tpr3' value
ikeeki has joined #linux-sunxi
konradoo77 has quit [Ping timeout: 250 seconds]
<ikeeki>
Hi, I have openbox+lxde+3.4.106+binary drivers (not working)+fbturbo+vdpau+X11. 1080p60 runs smoothly (even with >20GB movies) but mpv player shows movies ina bit dark way. Could anybody please tell me how to increase brightness and decrease gamma in mpv? (standard mpv controls do not recognize gamma and brightness control is recognized but does not change anything). Thanks
<diego71_>
plaes: i've already see that error, with latest ubuntu. With debian I've no problems.
cnxsoft has joined #linux-sunxi
<plaes>
ugh.. seems to be broken toolchain (gcc-4.9.2 b0rked, 4.8.4 seems to be working)
FlibberTGibbet has quit [Quit: Leaving]
domidumont has joined #linux-sunxi
gianMOD has quit [Remote host closed the connection]
faisal has joined #linux-sunxi
<diego71_>
plaes: in teory you can disable ssp in kernel configuration, through menuconfig, but someone reporting the same problem, told me that was useless
<diego71_>
so changing the toolchain is better
dubito has quit [Ping timeout: 246 seconds]
domidumont has quit [Remote host closed the connection]
domidumont has joined #linux-sunxi
dubito has joined #linux-sunxi
<ssvb>
plaes: be sure to keep nand support disabled in the sunxi-3.4 kernel
<plaes>
ssvb: it automatically breaks existing partitions?
<ssvb>
I'm not sure what is the current status of the android nand filesystem corruption story due to the changes in the nand filesystem format in later versions of libnand
<ssvb>
I alway keep nand support disabled myself :)
<plaes>
ok, thanks
dubito has quit [Client Quit]
MY123 has joined #linux-sunxi
<MY123>
Is there a tutorial about bare-metal on Allwinner A20 with the machineID information using a DTB?
ricardocrudo has quit [Ping timeout: 245 seconds]
<MY123>
( I did one RTOS with the same binary working on a BBB, RPi1, and RPi2, I would like porting it to then occupy about a HDMI/LCD driver, before attempting to do it with KMS)
cnxsoft has quit [Remote host closed the connection]
<el3>
Hi I built mainline kernel for pcduino3. http://linux-sunxi.org/Manual_build_howto and set up the boot partition. But it wont boot. It says that I should copy the kernel modules like this cp -r <PATH_TO_KERNEL_TREE>/output/lib /mnt/ . But I cannot find the output/lib folder.
<MY123>
el3: serial dump?
<el3>
MY123: when I do screen /dev/ttyUSB0 115200
<el3>
MY123: It says terminated
<MY123>
el3: Get your serial adapter working and pastebin the result.
<MY123>
el3: I never used a serial adapter( only a Bus Pirate)
<el3>
MY123: Okey I try another serial adapter
konradoo77 has quit [Ping timeout: 255 seconds]
konradoo77 has joined #linux-sunxi
<el3>
MY123: Okey works now. Turned out I needed sudo screen. It keeps booting on nand where ubuntu is installed. Can I stop it from doing that. I cannot scroll in the serial output.
<MY123>
el3: Try reinstalling the SPL and uboot
gianMOD has joined #linux-sunxi
gianMOD has quit [Remote host closed the connection]
<el3>
MY123: Yes think you are right. I got the outpout now ttp://pastebin.com/dLmqshkD board seems to be cubieboard2 and something wrong with spl to
<MY123>
el3: INSTALL THE SPL FOR YOUR BOARD
<MY123>
el3: Or you will get very nice surprises !
<el3>
MY123: got it
<MY123>
el3: Because of wrong SPL settings , I lost a full year trying to run linux-sunxi
<MY123>
(on my tablet)
<el3>
MY123: lost a full year?
<MY123>
el3: a year of debugging with JTAG
<libv>
...
<MY123>
( before discovering that something was really wrong and RAM corrupted randomly)
<el3>
MY123: damn, I would give up
<MY123>
el3: at least, you have a SPL config ready for your board. Compile that
<plaes>
el3: you need to install the modules first
<el3>
MY123: or actually maybe not. I have been struggling getting this to run stabile for 3 months
<plaes>
el3: make ARCH=arm CROSS_COMPILE=... INSTALL_MOD_PATH=... modules_install
<libv>
MY123: i helped you with the kurio 7s last august.
<libv>
on the 24th to be exact
<MY123>
libv: I got it on December 2013
<libv>
i have no tangible data on you before that.
<el3>
plaes: Yes it says that in the guide. But I dont find the output folder. And I also need to fix this SPL
<MY123>
libv: and tried since this day before discovering this channel and linux-sunxi.org
<el3>
I got it working buildiing with help from the upstream guide. But the mainline is a bit different
<MY123>
el3: Is lima-memtester stable?
<el3>
MY123: on the pcduino? yes tried that
<el3>
MY123: all ok
afaerber_ is now known as afaerber
<MY123>
el3: mainline or u-boot-sunxi?
<MY123>
( I have the original PCdruino)
<el3>
MY123: mainline
<el3>
MY123: should try u-boot-sunxi with mainline kernel
<el3>
MY123: it says on the side, mainline is better
gianMOD has joined #linux-sunxi
<plaes>
ssvb: companion cube is visible :P
gianMOD_ has joined #linux-sunxi
gianMOD has quit [Read error: Connection reset by peer]
<ssvb>
plaes: good :) just let it run for a while, a few minutes should be enough for a short test
<ssvb>
then you can try increasing the dram clock speed until it starts failing
<plaes>
mhm :)
gianMOD_ has quit [Remote host closed the connection]
<ssvb>
just to see whether 432mhz is far from the real limit or not
<ssvb>
and also to see whether you have any safety headroom or not
<ssvb>
for example, passing the test at 432mhz but failing it at 456mhz would mean that 432mhz might be not very safe
<fest>
libv: is that a recent change, or was it incorrect in wiki all the time?
<ssvb>
plaes: just power off or reset the device :) or press Ctrl-C in console
<fest>
I was wondering why I still get dynamic mac address even with that setting but didn't bother looking into
pmattern has joined #linux-sunxi
cubear has quit [Quit: Leaving]
<libv>
very recent change
<libv>
not sure what to make of it
<plaes>
could it be different in 3.4 and mainline?
<libv>
i think this editor has been editing the module options on his host machine
<libv>
nope
<plaes>
both are available
<libv>
yeah, but which is correct in this context?
gianMOD has joined #linux-sunxi
gianMOD has quit [Remote host closed the connection]
<ssvb>
I think one is the mac address that is visible on the device itself (run ifconfig on the device) and another is the mac address that is visible to the other machines (run ifconfig on your desktop pc)
<el3>
I cannot see that on that site. What should I do?
Renard has quit [Remote host closed the connection]
* plaes
adds extra conf...
<plaes>
el3: I'll update wiki page
<el3>
plaes: Great thank you vbery much. msg when done
<plaes>
el3: refresh
<el3>
plaes: you are god
<plaes>
just let me know if I was wrong
Renard has joined #linux-sunxi
<el3>
yes I will try now
<plaes>
ssvb: 508 seems to be good
<plaes>
err.. 504
<plaes>
how does it meminfo determine density?
<plaes>
it reports currently 4096, although I have 2G chips
<ssvb>
plaes: it's best to run the test longer (overnight), and still leave some safety headroom even if it can't detect problems at 504
<el3>
plaes: You were right :) great thank you
<ssvb>
plaes: the other people may have different dram overclocking limits even if they have exactly the same tablet model, so unless we collect more statistics, it's safer to keep 432 in u-boot :)
<plaes>
so far I seem to be only one :P
<plaes>
although Turl has similar tablet (zatab)
<ssvb>
plaes: regarding the chip density, what is the reported bus and io width?
ricardocrudo has quit [Ping timeout: 252 seconds]
gianMOD has joined #linux-sunxi
<plaes>
io_width = 16, bus_width=32
<ssvb>
plaes: you have four x8 dram chips, but u-boot detects them as two x16 chips with twice higher density
endlezzz has quit [Ping timeout: 256 seconds]
<plaes>
might be
<ssvb>
this is safe, just the refresh timings are a bit worse because of assuming higher density
<ssvb>
plaes: then 504 is not really reliable, step down to 480 and don't forget to do the final overnight run after you find something that does not fail quickly :)
<plaes>
but in kernel config?
<el3>
I think I did
<el3>
I will try to rebuild. Is it make mrproper, before I build again?
<plaes>
dunno
afaerber_ has joined #linux-sunxi
<ssvb>
plaes: btw, assuming 4Gb density instead of 2Gb just means that refresh takes 300ns instead of 160ns - http://git.denx.de/?p=u-boot.git;a=blob;f=arch/arm/cpu/armv7/sunxi/dram_sun4i.c;h=c736fa3b474f5bbb3ed704fea566985c6c9b1011;hb=HEAD#l424
<ssvb>
plaes: your dram runs a bit slower
<plaes>
what about this calculation from datasheet?
<ssvb>
tRFC is different timing parameter, it is not part of tpr0/tpr2/tpr2
afaerber has quit [Ping timeout: 245 seconds]
<ssvb>
and u-boot currently uses the default tRFC values from the JEDEC spec, based on chip density
<ssvb>
if we take the values from datasheets, we may get better tRFC timing (like 260ns instead of 300ns for some of the chips)
<ssvb>
but this is just a minor performance tweak, safety first :)
<ssvb>
autodetecting x8 ddr3 chips as a twice lower number of x16 chips results in a minor performance loss, but without sacrificing reliability
konradoo87 has quit [Ping timeout: 246 seconds]
endlezzz has joined #linux-sunxi
konradoo77 has joined #linux-sunxi
<plaes>
cool.. I actually managed to find stock firmware
<el3>
okey never mind. I will try to build the kernel again. I only need a minimal system that can use g_mass_storage and wifi. Any suggestions are wellcome. Images on linksprite dont suppor g_mass_storage. And the last upstream I built with debian, worked, but crashed after ½ hour running each time
konradoo77 has joined #linux-sunxi
<el3>
plaes: Is it enough to build kernel again and move the modules to the lib folder. Or do I need to delete anything on the rootfs?
<el3>
ofc the uImage too
<plaes>
yes
<plaes>
I have uImage and uImage-3.4 on my device and /lib/modules wihch contains modules for different kernel
<el3>
okey
<ktemkin>
<late>libv: thanks</late>
gianMOD has quit [Remote host closed the connection]
gianMOD has joined #linux-sunxi
gianMOD has quit [Remote host closed the connection]
konradoo77 has quit [Ping timeout: 250 seconds]
<el3>
plaes: ssvb: Finally it worked. Thanks for your patience
konradoo77 has joined #linux-sunxi
Zboonet has quit [Ping timeout: 252 seconds]
Zboonet has joined #linux-sunxi
Zboonet has quit [Ping timeout: 256 seconds]
bengal has joined #linux-sunxi
Netlynx has quit [Quit: Leaving]
gianMOD has joined #linux-sunxi
gianMOD has quit [Ping timeout: 264 seconds]
exslestonec| has quit [Ping timeout: 246 seconds]
forest has joined #linux-sunxi
<el3>
Does any of you know where in kernel config I can enable nand, so I can access nand-flash while booted on sd-card
<el3>
?
MY123 has quit [Ping timeout: 246 seconds]
<el3>
found it
<plaes>
what was wrong?
<el3>
plaes: the option - "Initramdisk something" in general options should not be included
<el3>
dont know if that is allways the case
<plaes>
oh
tomcheng86 has quit [Read error: Connection reset by peer]
endlezzz has quit [Read error: Connection reset by peer]