<smaeul>
I think he applied it and then dropped it for some reason
<Esmil>
Ah, cool
coldfeet has quit [Quit: Lost terminal]
cousteau_ has joined #riscv
luca_ has joined #riscv
luca_ is now known as OwlWizard
cousteau has quit [Ping timeout: 255 seconds]
<Esmil>
smaeul: hmm.. that patch by itself does not fix booting for me. the errors above are gone, but the plic is still not initialized
cousteau_ is now known as cousteau
alperak has quit [Quit: Connection closed for inactivity]
OwlWizard has quit [Quit: OwlWizard]
<smaeul>
Esmil: hmm, I see this on 6.9.0-rc4: [ 0.130202] riscv-plic 10000000.interrupt-controller: mapped 175 interrupts with 1 handlers for 2 contexts.
<smaeul>
oh, I think the issue is that since it gets probed later, the SBI timer gets used during part of the boot process, which doesn't work in your setup because the CLINT isn't in OpenSBI's DT
<smaeul>
arguably seeing this at all is a kernel bug: [ T0] Timer extension is not available in SBI v2.0
<smaeul>
because Linux should be smart enough to not register the clockevent when the SBI timer extension is unavailable
<Esmil>
ah. yeah, as you see after the wdt reset the SoC it's running vaniall OpenSBI v1.4 and your d1-wip u-boot 2e89b706f5c9 ("sunxi: riscv: Add defconfigs for several boards")
naoki has joined #riscv
esv has quit [Ping timeout: 240 seconds]
<Esmil>
smaeul: enabling CONFIG_RISCV_SBI_V01=y gets rid of those messages, but it still doesn't initialize the plic: http://sprunge.us/mvV0Sw
<conchuod>
courmisch: sorry, didn't notice your message til now. Missing anything other than a uart!
jfsimon1981 has quit [Remote host closed the connection]
lagash has quit [Remote host closed the connection]
<smaeul>
Esmil: right, the issue isn't with the PLIC driver. it's that the system hangs before any platform driver probes because timer interrupts aren't firing
cousteau has quit [Quit: ♫ I can't forget the day I shot that network down ♫]