dgilmore changed the topic of #fedora-riscv to: Fedora on RISC-V https://fedoraproject.org/wiki/Architectures/RISC-V || Logs: https://libera.irclog.whitequark.org/fedora-riscv || Alt Arch discussions are welcome in #fedora-alt-arches
peko[m] has joined #fedora-riscv
tg has quit [Quit: tg]
<davidlt[m]> Nice, but Matrix is not able to show the picture.
<davidlt[m]> rwmjones: nujive is lost?
<davidlt[m]> The picture loaded. Nice.
<davidlt[m]> Ha, so it booted at default 1.85GHz instead of 2.0GHz :)
<davidlt[m]> Anyways, I need at least 16GB variant. I will wait until the final PCB shows up.
<davidlt[m]> Interesting that their Aliexpress shop lists PCB and PCBA.
<davidlt[m]> The current view shows PCBA, and PCB is slightly smaller.
<davidlt[m]> If PCB size changes that a bigger change than some traces or/and components.
jcajka has joined #fedora-riscv
<davidlt[m]> New GCC 13 builds landed, another 2 in the pipeline.
<davidlt[m]> I am about to submit yet another GCC 13 build that will incl. libatomic inlining fixes and remove -latomic from -pthread.
<davidlt[m]> Finally (hopefully) the end of random package build failures, and hacking around the issues.
<dtometzki> davidlt: nice is you could just plug in and go. An image was already pre-installed. I installed a new image this morning with the fastboot utility and that was no problem.
<dtometzki> Compared to VF2 easy going
<davidlt[m]> Well, VF2 has upstream support, this doesn't yet :)
<davidlt[m]> And vendor is not involved, which is sad.
<dtometzki> i hope it will come
<davidlt[m]> Yeah, there are 3 patchsets posted. There are 3 potential maintainers that want to be involved, but not the vendor.
<davidlt[m]> Don't get me wrong, HW wise something TH1520 and JH8100-series is what we need at minimal.
<davidlt[m]> Ah yes, some JH8100-series details got public yesterday.
<dtometzki> do have a link with the information ?
<davidlt[m]> So it seems to use 12nm. Most likely from the same fab as TH1520.
<davidlt[m]> There is AV1 and VP9 support on this one.
<davidlt[m]> Their own cache coherent interconnect (those are difficult to get right).
<davidlt[m]> USB 3.0/2.0, 3.2 Gen 2 (Type-C support), PCIe 3.0 x8 (that's a nice bump0, some AI engines, and big.little.
<cwt[m]> 2020?
<cwt[m]> Oh the first like 3 days ago
<cwt[m]> s/like/link/
<davidlt[m]> That's probably wrong.
<davidlt[m]> They announced their own core IP last year IIRC.
<davidlt[m]> So the date is wrong.
<davidlt[m]> In 2020 were was no RVV 1.0 too.
<davidlt[m]> s/were/there/
<rwmjones> davidlt[m]: checking ...
<rwmjones> davidlt[m]: nujive is running dnf at 100% ...
<rwmjones> it seems to be working away, I can restart kojid if you like but I'll leave it alone unless you say
<davidlt[m]> It's dead
<davidlt[m]> It hasn't delivered anything for half a day or so
<rwmjones> davidlt[m]: ok I'll have a look
<rwmjones> davidlt[m]: I just killed that rogue dnf process, now mock and kojid are doing something again
<davidlt[m]> check dmesg
<davidlt[m]> It's talking to Koji now, but I typically restart once something like this happens.
<rwmjones> looks like there's a new job running:
<rwmjones> nothing really, a segv two days ago
<rwmjones> I think it's doing stuff now, let's just leave it alone for a bit and see if things recover
<rwmjones> it seems as if there was a dnf install which simply started to spin for some reason
* rwmjones wonders what lv2-ll-plugins is ..
<rwmjones> if nothing happens in an hour I'll reboot it
Kevinsadminaccou has quit [Remote host closed the connection]
<davidlt[m]> Nice!
<rwmjones> ok nujive seems like it's running again now
<thefossguy> wtf, they even have zsbl open sourced
<davidlt[m]> Yeah, it's on GH for some months now.
<davidlt[m]> SiFive also has their ZSBL open.
<davidlt[m]> At least for Unleashed.
<davidlt[m]> Not that you can do anything with it :)
<thefossguy> I just noticed that after taking a closer look at their crowdsupply page
<thefossguy> davidlt[m]: That makes the open sourcing even better imho
<thefossguy> "We know you can't do much with this but hey, you asked for src, so umm.. here you go"
<thefossguy> The StarFive guys are moving FAST
<thefossguy> Their upstream repo is already on 6.4-rc2
<thefossguy> Nice :)
<davidlt[m]> They should be, staying on a single kernel version makes life harder the longer you stay on it.
<thefossguy> I just wish their upstream page followed sane date formatting :(
<thefossguy> MM/DD/YY is downright confusing unless one field is greater than 12
<cwt[m]> Pratham Patel: I already tried that 6.4, but it is still not working yet.
<thefossguy> Yeah, some things aren't configured correctly. Mainly the MMC_DW_STARFIVE config isn't working as intended and stops at rootfs.
<thefossguy> David was right, 6.5 will be the one
<cwt[m]> I make all config options with STARFIVE in their name set to =y, but the kernel is still not working.
<davidlt[m]> Did you report the issue?
<davidlt[m]> There was one identified very recently as the bug in one of the drivers.
<thefossguy> Haven't reported it as I consider the StarFive guys to be on top of this because they are already testing this
<cwt[m]> Not yet, I can't attach the serial console on the board now as I am still recovering from covid.
<thefossguy> Get well soon cwt
<thefossguy> I will wake you when mainline boots 😆
<cwt[m]> Thanks, so weak after covid 😷
<thefossguy> Understandable... My dad got it in 2020, it was bad. But he recovered. Just rest as much as you can.
masami has joined #fedora-riscv
zsun has joined #fedora-riscv
tg has joined #fedora-riscv
zsun has quit [Quit: Leaving.]
zsun has joined #fedora-riscv
masami has quit [Quit: Leaving]
zsun has quit [Quit: Leaving.]
jcajka has quit [Quit: Leaving]
cyberpear has joined #fedora-riscv
peko[m] has quit [Excess Flood]
cyberpear has quit [Quit: Connection closed for inactivity]