<suihkulokki>
What's the best alternative to earlyprintk on chromebook (snow) ?
<Wizzup>
Can't you pass earlyprintk?
<suihkulokki>
I don't have serialport
<suihkulokki>
I'm trying to boot 4.2, and I get nothing after "starting kernel.." and "stashed 20 records"
<Wizzup>
hm... I'm still on 4.1
<Wizzup>
Guess I should try 4.2.
<Wizzup>
I was going to wait for the grsec. patch
<javier__>
suihkulokki: what bootloader are you using?
<javier__>
suihkulokki: 4.2 works for sure
<suihkulokki>
javier__: u-boot v2015.10-rc3
<javier__>
suihkulokki: and how are you building your kernel and booting it?
<javier__>
there is certainly a problem on your setup since v4.2 with mainline u-boot should work
<suihkulokki>
javier__: exynos_defconfig, then mkimage to uImage, and copy the uImage and dtb the sd card where the mainline u-boot script picks them up
<javier__>
suihkulokki: can you share your mkimage command and also the mainline u-boot bootcmd and bootargs?
<javier__>
I don't remember what is the mainline u-boot default boot args since I've my own...
<javier__>
s/boot args/env vars
<suihkulokki>
moment
<javier__>
sure, no rush :)
<zombah>
4.2 exynos_defconfig works fine for me with 2015.10-rc2 chainloaded from sdcard
<suihkulokki>
so any example of bootargs that get a framebuffer console up?
<javier__>
suihkulokki: sorry, I was busy and didn't pay attention to this channel
* javier__
reads the backlog
<javier__>
suihkulokki: your bootargs looks correct
<javier__>
suihkulokki: you mentioned that you are using mainline u-boot right? that doesn't have simplefb but is not needed since Exynos DRM should be working
<suihkulokki>
javier__: yeah, mainline u-boot
<suihkulokki>
javier__: with those, I get nothing but a blank screen after u-boot
<javier__>
suihkulokki: strange, did you share your complete process to build the kernel?
<javier__>
suihkulokki: your load and entry address that you are using for u-boot is wrong
<javier__>
should be 0x40008000
<javier__>
s/u-boot/the uImage
<liquidAcid>
somebody an idea how to properly invalidate way on an l2c-310? i can't make it working properly in a smp setting and i'm getting the impression that the way ops are only useful when brining the system up or down
<liquidAcid>
raw spin_lock_irqsave doesn't help since it only disabled irqs on the local cpu, but apparantly for the way ops to work properly no "background activity" (whatever that means) is allowed -- in particular probably activity on the non-local cpus
zombah has quit [Quit: Leaving]
<suihkulokki>
javier__: didn't change anything, will have to look another day
<javier__>
suihkulokki: that was certainly wrong, maybe there is another issue