<Tartarus>
That hopefully isn't too much to add in
<apalos>
for the record, mwalle is not wrong
<apalos>
what happens right *now* is what he claims
<apalos>
OTOH, I think he has a part of it wrong
<apalos>
the spec doesn't mandate (at least for now) that the firmware *always* provides the DT
<apalos>
it'
<apalos>
it's desireable, yes, but not a strict requirement atm
<apalos>
IOW the spec doesn't say "if you load an OS DTB you loose compliance"
<apalos>
however personally I think we can have a sane way of syncing the u-boot dtb's, provide the nodes u-boot extends and have a standard way of overriding that DTB
<mwalle>
apalos: i don't say, that shouldn't be a possibilty
<mwalle>
apalos: just that the second case should also be a first class citizen
<apalos>
mwalle: and I intend to fix it
<apalos>
SR-IR implies EFI, so I'll go write the same code I did for initrd
<apalos>
so you will be able to have the bootmanager boot an OS (or grub), with an alternative DTB
astroid has quit [Quit: IRCNow!]
FredO has quit [Quit: Leaving]
<apalos>
mwalle: so I assume you volunteer to test it :)
<apalos>
?
<mwalle>
apalos: but having the bootloader ready is just one part. the distros also need support
<mwalle>
apalos: sure, feel free to CC me on patches. If I have time, I'll try em out
<apalos>
mwalle: yea one step at a time, let's fix the loader first and see what distros think
<mwalle>
apalos: although I don't know the state of the distros regarding that. I know that debian isn't providing a device tree. because that was something I've tried some time ago
<apalos>
mwalle: there's a DTRB folder on fedora last time I checked,
camus has quit [Quit: camus]
<apalos>
Although I'll admit I dont know if they include multiple versions of the same dtb, tweaking config bits
<mwalle>
apalos: but its hard to test something, when I don't have any distros supporting it ;)
<apalos>
you can load your special dtb manually though!
<mwalle>
apalos: debian has /lib/kernel-bla/, IIRC. But whats missing is that its somehow copied to the ESP and somehow communicated to the bootloader
<apalos>
there's a bigger probnlem there, since setvariable at runtime is a different kind of mess
<mwalle>
apalos: thats already possible, today, isn't it
<apalos>
nop
<apalos>
well I have patches that fix that for u-boot + rpmb
<apalos>
but the general usecase of the sd card doesn't and it probably never will
<apalos>
there's ways around it though
<mwalle>
apalos: not the variables, loading the fdt manually
astroid has joined #u-boot
astroid has quit [Changing host]
astroid has joined #u-boot
<apalos>
ah yea, that works on distro_bootcmd already
<apalos>
you need to define fdt_file or something
<apalos>
or you need to run bootefi with the new dtb address
FredO has joined #u-boot
camus has joined #u-boot
monstr has quit [Remote host closed the connection]
tre has quit [Remote host closed the connection]
redbrain has joined #u-boot
sszy has quit [Ping timeout: 268 seconds]
mckoan is now known as mckoan|away
camus has quit [Quit: camus]
joaoSchim__ has joined #u-boot
joaoSchim_ has quit [Read error: Connection reset by peer]
___nick___ has joined #u-boot
as2334 has joined #u-boot
<as2334>
question about rockchip SoCs : what is the "ID block" header that the rom looks for? Is it just the string "RK31" for rk31xx chips? And at what address(es) of a nand chip is it supposed to be?
tambarus has quit [Quit: Connection closed for inactivity]
mranostaj has quit [Remote host closed the connection]