<palmer>
geertu: conor say you need an ack on the RZ/Five patches? I'm at my home computer and email is screwed up, I'll try and remember when I get to the office
<pabs3>
hmm, interesting project
<drewfustini>
palmer: did the call just end?
<bjoto>
drewfustini: yeah, it was a quick one today
<palmer>
ya, sorry
<palmer>
I didn't really get any LInux stuff done over the last few days because of the profile stuff, so it's kind of just in the same state as last week
<drewfustini>
Thanks, i couldn't tell if the app crashed or it was actually the end
<bjoto>
:D
<geertu>
palmer: OK. I'll send a ping next week when it hasn't happened (I need to send my PR to soc before rc6)
<palmer>
thanks
<palmer>
I'll try and get email working today, I have a bit of WFH time this morning...
<palmer>
never mind...
qrqrtty[m] has quit [Quit: You have been kicked for being idle]
jellydonut has joined #riscv
jacklsw has quit [Read error: Connection reset by peer]
vagrantc has joined #riscv
___nick___ has joined #riscv
pecastro has quit [Ping timeout: 248 seconds]
<drewfustini>
I finally got a HiFive Unmatched board and that tiny CPU fan is rather noisy. Anyone with the board replace it with something else?
<q66>
yeah I just replaced it with a bigger heatsink and no fan
<q66>
or you can use a bigger fan if you replace the heatsink
epony has quit [Remote host closed the connection]
aerkiaga has joined #riscv
<drewfustini>
Thanks
<drewfustini>
Did you just read an on-die sensor to know if it was cooling well enough?
pecastro has joined #riscv
<rneese>
I want a jifive unmatched but the cost to high
<rneese>
want to get it supporting efi uefi booting and make shure our images work on it
elastic_dog has quit [Ping timeout: 246 seconds]
elastic_dog has joined #riscv
<q66>
drewfustini: touch the heatsink? :P
<q66>
(sensors work as well)
frkzoid has joined #riscv
<drewfustini>
:)
frkazoid333 has quit [Ping timeout: 260 seconds]
prabhakarlad has quit [Quit: Client closed]
<mps>
I played with SPI fan control on visionfivev1 and temperature measuring. It works but not so perfect
<mps>
also for simpler solution used 3.3V pin from GPIO header to power fan
<mps>
with 3.3V it is less noisy but still not quiet
<muurkha>
it's so funny people use SPI instead of I²C for fans\
<muurkha>
less hassle with addresses I guess
<mps>
muurkha: yes, less hassle :)
<muurkha>
not that 400kbps is too slow for the fan speed data ;)
BootLayer has quit [Quit: Leaving]
wingsorc has quit [Quit: Leaving]
<palmer>
drewfustini: IDK if the Unmatched is better, the Unleashed didn't have any thermal sensors or protection so you needed to be pretty careful with custom cooling
prabhakarlad has joined #riscv
<drewfustini>
I replaced the tiny fan with a 120mm fan that is USB powered. I see this for the temps which seems okay:
<muurkha>
there must be some browser extension for this
<geertu>
muurkha: Oh, you're not doing fan PWM using SPI ;-)
davidlt has quit [Ping timeout: 260 seconds]
<geertu>
conchuod: yeah, else Amazon thinks drewfustini is everywhere in the world at once ;-)
<muurkha>
geertu: hmm, the time constant of a fan is on the order of a second, isn't it? so you probably *could* do fan PWM using SPI, but you probably shouldn't
<muurkha>
hmm, that makes me think that fans are probably the cheapest servos available, too bad you can only servo speed and not position using their normal interface
<bjdooks>
iirc there are some 3d printed verisons of the unmatched fan holder for bigger less whiny fans
<muurkha>
it's funny how the whiniest fans are rarely the biggest fans
wingsorc__ has joined #riscv
<geertu>
muurkha: Exactly. On the recent R-Car gen4 devboards, the fan is controlled through a temperature sensor and PWM. After power-up, but when still cold, the fan makes strange noises, but doesn't spin yet.
<conchuod>
It's usually somewhat inversely proportional muurkha
<muurkha>
conchuod: yeah, though sometimes the people who donate to your Patreon do feel more entitled as a result
<geertu>
When warmed up and spinning, the sound changes. Although the fan is not really loud, I do find the changing sound annoying, so I'm happy I could move it to the farm in another room ;-)
___nick___ has joined #riscv
<geertu>
s/changes/changes all the time/
___nick___ has quit [Client Quit]
___nick___ has joined #riscv
<conchuod>
muurkha: I have no idea what that means but okay.
<conchuod>
Yeah, it's mostly not the sound itself but variations in it.
<conchuod>
muurkha: Oh I get you now...
prabhakarlad has quit [Quit: Client closed]
<rneese>
what we working on
prabhakarlad has joined #riscv
<conchuod>
palmer: esmil: I sent the maintainers patches - lmk if it doesn't match what you expected...
rneese has quit []
<palmer>
thanks, I'll have to go look a bit more -- I'm never sure quite how those entries work, and some of these don't have lists/trees so IDK where get_maintainers will end up putting stuff
<palmer>
also I guess SiFive should Ack their driver change, but they've kind of just abandoned it
<conchuod>
You're in the SiFive entry so technically your ack qualifies ;)
<palmer>
ya
<palmer>
it's just kind of odd to have a vendor completely out of the loop on their drivers? maybe that's not so bad, though
<palmer>
it's not like they're really doing the work, so whatever
<conchuod>
If they don't support it, then yea. Plenty of drivers don't get upstreamed by the vendor /shrug
<muurkha>
who is really doing the work?
<conchuod>
I'm a customer for the only driver covered by drivers/soc/sifive so I've got a vested interest anyway.
<conchuod>
Depends muurkha, they do contribute new stuff but (apart from Ben) don't look at patches.
<palmer>
ya, it's kind of just all over the place
<palmer>
I've been merging the code, but mostly just because otherwise nobody does it
<conchuod>
Paul's on the CC for it all so they do at least (in theory) see it.
<bjdooks>
paulw? he's really busy
<conchuod>
Hence the in-theory ;)
<conchuod>
I'd ask for your Ack Ben, but I assume it's not really your call.
* bjdooks
is sitting at 21:30 trying to get to the bottom of a plic issue
<conchuod>
>some of these don't have lists/trees so IDK where get_maintainers will end up putting stuff
<conchuod>
palmer: the parent directory's entries should add the lists where I have omited it from the entries that I added.
<palmer>
OK, cool
<palmer>
I just never remember how any of that stuff works so I go test it ;)
<palmer>
as long as the patches end up _somewhere_
<conchuod>
Things should only end up in more places than they did before :)
talos has joined #riscv
jmdaemon has joined #riscv
<bjdooks>
urgh, so boot on hart 0, hart0's irqs work from plic... boot on hart 3, plic enables irqs on hart 3 and irqs never get triggered
crabbedhaloablut has quit [Ping timeout: 255 seconds]
crabbedhaloablut has joined #riscv
elastic_dog has quit [Ping timeout: 248 seconds]
elastic_dog has joined #riscv
mpido has quit []
aerkiaga has quit [Remote host closed the connection]
pedja has quit [Quit: Leaving]
somlo has quit [Remote host closed the connection]