Sonarpulse has quit [Ping timeout: 256 seconds]
<gchristensen> ok sphalerite I'm rebooting that system
<gchristensen> sphalerite: this thing boots soooooo slowly
<gchristensen> it doees like 5+min of self-testing
<gchristensen> hundreds of lines like: d759405a-b968-4966-8575-14f905f82672
<gchristensen> oops
<gchristensen> [SetAtuMemRW:717] - Base=A00A091C value=0
<gchristensen> [SetAtuConfig0RW:657] - Base=A00A0900 value=1
<samueldr> I'm curious now. googling those result in only one identifiable device
Sonarpulse has joined #nixos-aarch64
<gchristensen> quite possible
<gchristensen> wht did you find?
<gchristensen> bed time, we can try theguessing game tomorrow :P
cstrahan_ has quit [Ping timeout: 256 seconds]
cstrahan_ has joined #nixos-aarch64
orivej has quit [Ping timeout: 265 seconds]
orivej has joined #nixos-aarch64
orivej has quit [Ping timeout: 240 seconds]
disasm has quit [Ping timeout: 260 seconds]
<samueldr> while it fins stuff about the "HiSilicon D05 Development Board", I have a hunch it's instead the packet.net c1.xlarge.arm formerly 2a2, which would have the same CPUs; HiSilicon Hi1616
<gchristensen> you're a good searcher
<gchristensen> not sure how you found it is the c1.xlarge.arm though
<gchristensen> I see it as c1.large.arm.xda
<gchristensen> see anything unique about that fella?
<samueldr> the c1.xlarge.arm was from https://help.packet.net/faq/pre-sales/compute-classes, every other mention of the hi1616 at packet were with the 2a2 name
<gchristensen> nice
<samueldr> and since the documentation on packet's website doesn't seem to have the 2a2 anymore, I can't really tell
<samueldr> though, the hi1616 has that mellanox thing for 10gbps AFAICT
<samueldr> and it apparently coule be specced 2×32 cores, with 128GB of ram https://twitter.com/i/web/status/946759012313333760
<samueldr> heh
<samueldr> damn, I should start re-reading myself before pressing enter -.-
<gchristensen> but that isn't the special thing
<samueldr> if it's about CPU features, I couldn't seem to find an overview of the hi1616 :/
<gchristensen> aye
<samueldr> e.g., would it support armv7 ;)
<gchristensen> indeed
<gchristensen> it does
<samueldr> *that* was a complete guess, without supporting evidence other that who you pinged :)
<gchristensen> :D
<gchristensen> pretty much all obscure arm hw is a nightmare
<gchristensen> this one especially so
<gchristensen> it required custom, unreleased kernel patches for a long while
<samueldr> I would guess so, looking at how few things there is about it online
<gchristensen> it won't netboot my image anymore, either, for some reason
<samueldr> ubuntu apparently has hi1616-based d05 in its supported hardware :/
<samueldr> couldn't find much about the d05 though
<gchristensen> if you press up-arrow and down-arrow too many times in the boot process it'll decide something went terribly wrong and stop the boot
<samueldr> that's... fun
<gchristensen> for {set x 0} {$x<5} {incr x} {
<gchristensen> after 100
<gchristensen> send -i $console -- \x1B\[A;
<gchristensen> }
<gchristensen> indeed :)
<samueldr> this is the most I could get about the d05 http://open-estuary.org/d05/
<samueldr> I'm curious about the downloads, and if the sources could have helped beforehand
<gchristensen> if you want to try your hand at getting something to boot on it, I will try to boot it :)
<Dezgeg> at least these guys seem to boot daily linux-next builds: http://120.31.149.194:800/scheduler/device_type/hip07-d05
<gchristensen> neat
<gchristensen> who are those guys?
<Dezgeg> maybe it's an accidentaly-public instance :P but sounded related to http://open-estuary.org/d05/
<gchristensen> cool
<gchristensen> this is the console output
<Dezgeg> the LAVA log has EFI stub: Using DTB from configuration table but you get EFI stub: Generating empty DTB
<gchristensen> hrm
<gchristensen> how do I change that :%
<Dezgeg> no real idea on that, I do wonder how the patches make it work then
<gchristensen> maybe it doeesn't require patches anymore
<gchristensen> it did for a long while though
<gchristensen> hmmmm
<gchristensen> pxe maybe
<gchristensen> > U-boot will fetch the kernel image and the DTB files via TFTP.
<gchristensen> maybe I can reverse engineer this by peeking in to the iPXE files Packet uses
<Dezgeg> so which one of those 'EFI stub:' lines it prints when booting their distro?
<gchristensen> that, and actually downloading the netboot files and maybe they provide a dtb somehow
<gchristensen> you're not supposed to be able to maintain a connection to the console while it boots their internal tooling, but I've found a way :)
<Dezgeg> the 'EFI stub: Using DTB from configuration table' one not 'EFI stub: Generating empty DTB' ?
<gchristensen> (and I don't think they really mind)
<gchristensen> right, I don't know
<gchristensen> samueldr, Dezgeg: nice hunting around for this LAVA thing :D
<gchristensen> hmm
<gchristensen> they generate an empty DTB
<gchristensen> (Packet)
<gchristensen> localhost:/# grep -ri dtb .
<Dezgeg> it's a ful ACPI boot then I guess
<gchristensen> `find / -iname '*dtb*'` didn't turn anything up, but the grep is showing a update-kernel script which might be good
<Dezgeg> wonder why it's different to the LAVA setup
<gchristensen> everything we learn I'm dumping in to notes to the ARM people at Packet, so maybe they can shed some light :P
<Dezgeg> maybe this earlycon=sbsa-uart,mmio,0x602b0000 from http://120.31.149.194:800/scheduler/job/31078/log_file could help debugging
<gchristensen> I should have told the grep to skip /dev
<samueldr> :o
<sphalerite> gchristensen: aaah the joys of server hardware…
<samueldr> hardware is hard, it's in the name
<sphalerite> (that was specifically re 5 mins of self-testing)
lopsided98 has quit [Ping timeout: 260 seconds]
lopsided98 has joined #nixos-aarch64
orivej has joined #nixos-aarch64
Sonarpulse has quit [Ping timeout: 276 seconds]
Sonarpulse has joined #nixos-aarch64
orivej has quit [Ping timeout: 245 seconds]
{^_^} has quit [Remote host closed the connection]
{^_^} has joined #nixos-aarch64
{^_^} has joined #nixos-aarch64
{^_^} has quit [Remote host closed the connection]
{^_^} has joined #nixos-aarch64
{^_^} has joined #nixos-aarch64
Acou_Bass has quit [Ping timeout: 264 seconds]
Acou_Bass has joined #nixos-aarch64
Acou_Bass has quit [Quit: byeeeeeeeeeeeeeee]
Acou_Bass has joined #nixos-aarch64