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
bkeys has joined #fedora-riscv
bkeys has quit [Quit: With every step we take, danger will follow closely]
bkeys has joined #fedora-riscv
bkeys has quit [Ping timeout: 256 seconds]
davidlt has joined #fedora-riscv
<davidlt[m]> So the last build I launched before going to sleep also failed :(
<davidlt[m]> I guess more defconfig tweaking today.
davidlt has quit [Ping timeout: 248 seconds]
davidlt has joined #fedora-riscv
<davidlt[m]> During the morning walk I remembered that I forgot to enable LITEX_SOC stuff too.
<rwmjones> morning
<davidlt[m]> by the time I manage to get it right there might be v5.18.5 :)
<davidlt[m]> and new attempt is submitted (incl. LITEX stuff, I just went with =y for now)
<davidlt[m]> but MMC and Ethernet might become =m in the future
<davidlt[m]> somlo: ^^
somlo_ has joined #fedora-riscv
somlo has quit [Read error: Connection reset by peer]
<davidlt[m]> somlo: I am settings CONFIG_SERIAL_LITEUART_MAX_PORTS=1
<davidlt[m]> and a new attempt is pushed :)
<davidlt[m]> hmm...
<davidlt[m]> rwmjones: I get impression your board (nujive) is somehow slowish
<davidlt[m]> buildArch (kernel-5.18.4-201.0.riscv64.fc33.src.rpm, riscv64)
<davidlt[m]> buildArch (kernel-5.18.4-201.0.riscv64.fc33.src.rpm, noarch)
<davidlt[m]> The 1st one is on djdelorie board, the noach is on nujive. Non scientific, but based on logs it seems it took a bit longer than I expected.
<davidlt[m]> It was processing defconfings while djdelorie already was building a kernel image.
<davidlt[m]> Just something unexpected, but the timing will tell everything.
<rwmjones> davidlt[m]: checking ..
<davidlt[m]> Other than that, the kernel is finally building :)
<rwmjones> davidlt[m]: nufive dead AGAIN!
<rwmjones> nujive is building some stuff using sphinx (documentation IIRC)
<rwmjones> but seems to be ok
<davidlt[m]> as long as nujive is fine, it's all good.
<davidlt[m]> Yeah, kernel noarch.
<rwmjones> seems ok, I won't touch it
<davidlt[m]> But again just felt slowish. To be seen in 6-12 hours or so
<rwmjones> will reboot nufive in 30 mins when I get to the office
<davidlt[m]> if this v5.18.4 boots, I will need another build to backport some stuff
<rwmjones> actually I'm going to take the case off nufive and have a look at the airflow & fans
<davidlt[m]> this build is mostly for v5.14 -> v5.18 defconfig adjustment
<davidlt[m]> touch the heatsink for SoC and PCIe switch
<davidlt[m]> they should be slightly warm to the touch, but nothing too hot
<davidlt[m]> damn, I need to clean the dusts from mine :/
<davidlt[m]> I assume we will get the final kernel by with v5.18.5 or so
<rwmjones> ok will d
<rwmjones> do
<davidlt[m]> According to Koji records it should take 9+ hours to build
<davidlt[m]> I basically get two shots a day at this :)
<davidlt[m]> Way faster to crosscompile, but not enough pain :D
pierosimonet has joined #fedora-riscv
pierosimonet has quit [Remote host closed the connection]
Guest55 has joined #fedora-riscv
Guest55 has quit [Client Quit]
<rwmjones> I just rebooted nufive with the case off
<rwmjones> because the power was shut off, I can't test for hot components
<rwmjones> but I'll go back and have a look in an hour
<rwmjones> the case fan seems to be running fine
jakfrost has joined #fedora-riscv
acharles has joined #fedora-riscv
masami has joined #fedora-riscv
warren_ is now known as warren
jakfrost has quit [Quit: Leaving]
bkeys has joined #fedora-riscv
<somlo_> davidlt: thanks! (=y should be perfectly OK, as well as CONFIG_SERIAL_LITEUART_MAX_PORTS=1)
<davidlt[m]> rwmjones: all OK, noarch package is built, and it was faster
<rwmjones> davidlt[m]: oh the sphinx/python package that was building?
<rwmjones> both boards are still
<rwmjones> up
<rwmjones> nufive still has the case off, but I didn't notice anything getting hot
<rwmjones> nujive has been up for a few days now I think
<davidlt> ha, matrix bridge is lagging again :/
<rwmjones> i see
somlo_ is now known as somlo
cyberpear has quit [Quit: Connection closed for inactivity]
<davidlt[m]> kernel close to finishing (an hour or more)
<davidlt[m]> I will need to rebuild some other kernel-* packages
<davidlt[m]> rwmjones: do you plan to just install it on nujive directly?
<rwmjones> davidlt[m]: I could try it if you like
<rwmjones> in other news, nufive crashed again
<rwmjones> with the lid off
<rwmjones> nothing hot inside
<rwmjones> seems to be hardware-ish probably
<rwmjones> will wait to see if it still happens with the new kernel though
<davidlt> doubt it :)
<davidlt> I don't trust the kernel first :)
<rwmjones> well, yeah, I'm not certain either, but the fact that one machine is crashing often and the other not very often is a bit suspicious
<davidlt> true, but could be SW related
<rwmjones> we'll see anyway with the new kernel
<rwmjones> nujive is idle now I think
<davidlt> you could also try to boot without NVMe and see if that solves the problem
<davidlt> disconnect everything and go with basics to see if anything else is causing issues
<davidlt> you could also try running FUSDK image, like 2022.03 and see if that crashes
<rwmjones> nujive:
<rwmjones> 18:24:13 up 4 days, 23:35, 2 users, load average: 0.00, 0.00, 0.00
<davidlt> there is a kept v5.13.X kernels for so long for FUSDK
<davidlt> anything above always had some issues
<davidlt> fingers crossed that v5.18.4 is OK :)
<davidlt> but djdelorie board so far didn't hand and it had a high load (full kernel build)
<davidlt> now running xargs -P4 -r xz, which is CPU intensive
<rwmjones> maybe dj's got air conditioning :-?
<djdelorie> it's in a beige box in my basement rack
<davidlt> I don't have one too ;)
<davidlt> but I definitely know that my silicon is lucky as it has quite low temps
<davidlt> mine is on the desk on the open air bench table, full of dusts on the fans
<davidlt> I have no idea how dust manages to stick to the fan blades :/
<djdelorie> so that python script that took days to run, wasn't a hang?
<davidlt[m]> djdelorie: something was wrong, I stopped it at some point
cyberpear has joined #fedora-riscv
<davidlt> we got a new kernel!
<davidlt> launching kernel-tools and kernel-headers
davidlt has quit [Ping timeout: 258 seconds]
<rwmjones> nufive crashed again!
masami has quit [Quit: Leaving]
<rwmjones> & again ..
<rwmjones> definitely hardware
bkeys has quit [Ping timeout: 276 seconds]