stikonas has quit [Remote host closed the connection]
ganbold has joined #linux-rockchip
vstehle has quit [Ping timeout: 246 seconds]
vstehle has joined #linux-rockchip
<wens>
repk: sounds like you might be using the last (2020.01) release of U-boot, or even older?
vagrantc has quit [Quit: leaving]
chewitt has quit [Quit: Zzz..]
chewitt has joined #linux-rockchip
_whitelogger has joined #linux-rockchip
_whitelogger has joined #linux-rockchip
<tuxd3v>
hello everyone,
<tuxd3v>
does anybody have a nice .config file for Rockpro64 Mainline kernel?
wens has quit [Remote host closed the connection]
ldevulder_ has joined #linux-rockchip
ldevulder has quit [Ping timeout: 256 seconds]
wens has joined #linux-rockchip
<wens>
anarsoul: I modified ATF to use the standard GPIOs (TSADC_INT / AP_PWROFF) for reset and shutdown
<repk>
wens: I compiling 2020.01 yes, But looking at the code bl2_plat_get_bl31_params() I don't think master goes to V2 bl31_params
dlezcano- has joined #linux-rockchip
dlezcano- is now known as dlezcano
<dlezcano>
Hi all,
<dlezcano>
is there a place where to fill a bug for the rock960 ?
ldevulder_ is now known as ldevulder
<dlezcano>
mmind00, mani_s : when running dhrystone on my rock960, the console is getting wild and not working properly. As soon as dhrystone stops, the console works correctly again. I'm observing this issue since some releases.
<mmind00>
dlezcano: I guess mailing the people that did submit the board
<mmind00>
dlezcano: interesting question would be if rock960 is the only one affected
<dlezcano>
yeah, unfortunately I don't another rk3399 board
<dlezcano>
s/don't/don't have/
matthias_bgg has joined #linux-rockchip
lkcl has joined #linux-rockchip
kevery has quit [Ping timeout: 240 seconds]
<wens>
mmind00: are the dts files in u-boot synced from linux?
<mmind00>
wens: they're supposed to be
<mmind00>
wens: i.e. when adding new dts files to u-boot people do seem to check wether they're present in Linux as well
<mmind00>
wens: of course mostly without the memory-controller stuff, which is somewhat u-boot specific
<wens>
the extra stuff is in -u-boot.dtsi files, so it shouldn't interfere
stikonas has joined #linux-rockchip
sb35 has quit [Ping timeout: 250 seconds]
matthias_bgg has quit [Ping timeout: 256 seconds]
<obbardc>
having issues with rock64 on linux-next and v5.5, u-boot gets to "Starting kernel..." then hangs
<obbardc>
other rk3328 board boots fine (with its DTS obviously)
<obbardc>
using the rockchip generated trust from rkbin repo
robmur01_ has joined #linux-rockchip
JohnDoe_71Rus has joined #linux-rockchip
<repk>
obbardc: Does it really hang or you just don't have console ?
<wens>
obbardc: use upstream ATF master & u-boot?
<wens>
upstream
<wens>
I just rebuilt both for my rock64, and it seems to work ok, without stable MAC address nor CPUID though
<obbardc>
ah i am not booting with u-boot.itb
<obbardc>
so no bl31
<wens>
any particular reason for that?
<robmur01_>
yeah, you kinda need a BL31 to boot - unless you've entirely removed all traces of PSCI from the DT, the kernel will start making SMC calls pretty early on, so *something* has to be resident to answer the call or it'll just get lost in EL3 and never come back
<mmind00>
and I guess in general ... take linux-next or wait for 5.7-rc1
<alyssa>
Is linux-next newer or older than git://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git
<mmind00>
as there may be more px30 groundwork still sitting in some trees waiting for 5.7 (and I generally stop following my patches once a maintainer has taken them)
<mmind00>
linux-next is newer ... aka contains the stuff that Linus will merge during the 5.7 merge window
<alyssa>
Oh, joy :-)
<alyssa>
This is all very good to know for Later(TM)
<alyssa>
which might become Sooner(TM) if we can't get this amlogic thing sorted :|
chiastre has quit [Remote host closed the connection]
<mmind00>
depending on how long you want to wait ... in 2.5 weeks everything except (1) should be in Linus tree
cristian__c has joined #linux-rockchip
<alyssa>
Makes sense
<mmind00>
[oh and the analog joystick needs separate patches to get merged first ... but it also works nicely here so far already ;-) ]
<mmind00>
but I guess you won't care about that very much in the beginning :-D
cristian_c has quit [Ping timeout: 250 seconds]
<wens>
what amlogic thing? :p
thefloweringash has quit [Ping timeout: 256 seconds]
<alyssa>
wens: Gorey details are in the #panfrost logs but
<alyssa>
The GPU fails often and totally nondeterministically
<alyssa>
on ODORID N2 (G12A soc, mali g52)
<alyssa>
And everyone is pretty certain this is an integration hw bug that is worked around with the blob/kbase... somehow
thefloweringash has joined #linux-rockchip
<maz>
amlogic is such a nightmare...
<maz>
they've fsck'd up the A55-based SoC as well in a really classy way (here's a bunch of wires, let's tie them all together)
<mps>
Esmil: hi! no luck with config options picked from your .config for chromebook, still same problem with analogix/rockchip-drm
<mps>
anyway, thanks for help
<Esmil>
mps: did you try the kernel I compiled for you?
<Esmil>
it's just odd that the same hardware with the same bootloader and almost identically built kernel behaves so differently
<mps>
well, no. if you built it with my config I think it will work same as mine
<Esmil>
try it
<mps>
did you make it as partition ready
<Esmil>
no i don't know your setup
<mps>
ahm, ok
<Esmil>
i build my kernels as distro packages, and then i have scripts to do all the vboot/uboot stuff when i install them, so i completely forgot how to do it :P
<mps>
problem is, it needs random time to show the problem, sometimes few minutes but sometimes more than 20 hours
<mps>
np, I have script to build kernel for chromebooks and create kpart images
<mps>
what gcc version you use to build it
<Esmil>
whatever is in rawhide. right now it's aarch64-linux-gnu-gcc (GCC) 9.2.1 20190827 (Red Hat Cross 9.2.1-3)
<Esmil>
mps: also don't forgot to loose all the earlycon and console stuff from your cmdline and just do console=tty1
<mps>
ok, will try in a half hour
<mps>
maybe sooner if I'm fast enough to put ssd in usb-c case :)
robmur01_ has quit [Quit: robmur01_]
robmur01_ has joined #linux-rockchip
<mps>
Esmil: you didn't changed anything in config ?
chiastre has joined #linux-rockchip
<Esmil>
mps: i just compiled the config you pasted
<mps>
aha. why do you think it will solve the problem
<Esmil>
as i said earlier it's weird that my screen works on the same hardware with the same bootloader and almost identical kernel
<Esmil>
..so i'm just trying to eliminate on variable, which is the toolchain used to build the kernel
<Esmil>
*one more
<mps>
I thought so, ok
<mps>
I doubt that toolchain can fix bug in driver but who knows, I will test and see. thanks again
<mps>
forgot to tell, it boots. lets see
<Esmil>
mps: wait. i'm confused. said earlier "built kernel with your .config (adding f2fs) and boot, but I don't have display working, it is blank", but just now you said "problem is, it needs random time to show the problem, sometimes few minutes but sometimes more than 20 hours"
<Esmil>
checking if the display comes up shouldn't take 20 hours
<mps>
Esmil: display was blank when I built kernel with your config, and added f2fs and some small changes
<Esmil>
ok, but what about with this kernel?
<mps>
kernel you sent is the same as mine
<Esmil>
so does the display work or not?
<mps>
maybe we didn't understand each other
<mps>
display works on it from the time I bought it
<mps>
but I have problem it hangs at random time
<Esmil>
right, but does this kernel show anything on the display?
<mps>
yes, it boots normally
<Esmil>
Great! Then your first problem is fixed
<mps>
well, this was not a problem
<Esmil>
so my .config *does* work then :)
<mps>
no, your .config doesn't work in my case, but that is expected
<Esmil>
you just said it was showing stuff on the display right?
<mps>
right, kernel you built with my config
<Esmil>
Cool. then this problem is fixed right: "built kernel with your .config (adding f2fs) and boot, but I don't have display working, it is blank"
<mps>
uh, we are lost :)
<mps>
kernel built with your .config doesn't boot
<Esmil>
ah so you mean that when you build the kernel it doesn't boot, but the one I sent you boots?
<mps>
I enabled f2fs to .config you posted and build kernel, this one doesn't boot
<mps>
right, on you sent boots
<Esmil>
Ok, that's even better. Then we know that it's your toolchain that doesn't w ork
<Esmil>
good thing i made you try the one I built then ;)
<mps>
no again
<mps>
it boots with kernels I build about three years for this box (or two, I forgot when they made them)
<mps>
and worked fine with chromeOS kernels which I also build and use
<mps>
but when I switched to mainline kernels (5.1 or 5.2) it started to make problems with displays (freeze sometimes) and emmc driver crashing
<Esmil>
right right, but before you can get to that you need a kernel that boots *and* shows something on the display right
<Esmil>
So at least you're that far now
<mps>
hmm, I'm using this machine about three years now, and always built kernels locally for it
<mps>
three or two years, I forgot when I bought it, but it was new thing on the market
<Esmil>
right, but you said you wanted to try my .config, and now you can try my .config with a kernel that boots and shows something on the display
<mps>
uh, looks like we don't understand each other. I want bug fixed which I reported in above URL to freedesktop.org ML
<Esmil>
i understand that, but i was just trying to solve this: "built kernel with your .config (adding f2fs) and boot, but I don't have display working, it is blank"
<Esmil>
..so you can get to try my .config, as you said you would like to
<mps>
chromebook works for about three years and about 10 months bug I reported started to irritate me
alyssa has left #linux-rockchip [#linux-rockchip]
lopsided98 has quit [Quit: Disconnected]
lopsided98 has joined #linux-rockchip
vicencb has joined #linux-rockchip
chewitt has quit [Quit: Zzz..]
matthias_bgg has quit [Ping timeout: 240 seconds]
<tuxd3v>
does anyone managed to get sound out of the Jack, or the hdmi on RockPi4 ?
matthias_bgg has joined #linux-rockchip
<tuxd3v>
also any hint on how to fix display resolution in bootargs ?