ChanServ changed the topic of #linux-rockchip to: Rockchip development discussion
stikonas has quit [Read error: Connection reset by peer]
lurchi_ is now known as lurchi__
vagrantc has quit [Quit: leaving]
SallyAhaj has quit [Remote host closed the connection]
Tenkawa has quit [Quit: Was I really ever here?]
SallyAhaj has joined #linux-rockchip
SallyAhaj has quit [Remote host closed the connection]
lurchi__ has quit [Ping timeout: 240 seconds]
lurchi_ has joined #linux-rockchip
lurchi_ is now known as lurchi__
robmur01 has quit [*.net *.split]
mps has quit [*.net *.split]
urja has quit [*.net *.split]
maz has quit [*.net *.split]
Esmil has quit [*.net *.split]
mps has joined #linux-rockchip
maz has joined #linux-rockchip
robmur01 has joined #linux-rockchip
urja has joined #linux-rockchip
Werner has quit [*.net *.split]
marvs has quit [*.net *.split]
marvs has joined #linux-rockchip
Werner has joined #linux-rockchip
Werner has quit [Changing host]
Werner has joined #linux-rockchip
Esmil has joined #linux-rockchip
SallyAhaj has joined #linux-rockchip
matthias_bgg has joined #linux-rockchip
SallyAhaj has quit [Remote host closed the connection]
shoragan has joined #linux-rockchip
stikonas has joined #linux-rockchip
maz has quit [Quit: ZNC 1.8.2+deb2+b1 - https://znc.in]
maz has joined #linux-rockchip
lkcl has quit [Ping timeout: 240 seconds]
lkcl has joined #linux-rockchip
Jmabsd has joined #linux-rockchip
<Jmabsd> Hi all! Just curious, the RK3588 support for Linux how mature is it, has anyone of you guys tried it? To your awareness will a standard Linux distribution work quite well with it???
<Jmabsd> (crosspost with #linux-sunxi , sorry for any spam)
lurchi__ is now known as lurchi_
SallyAhaj has joined #linux-rockchip
lurchi_ is now known as lurchi__
Tenkawa has joined #linux-rockchip
SallyAhaj has quit [Remote host closed the connection]
lurchi__ is now known as lurchi_
matthias_bgg has quit [Ping timeout: 246 seconds]
<wens> Jmabsd: probably not very mature, given that boards aren't quite available yet?
<Jmabsd> wens: I think one board is available
<Tenkawa> which board (didn't get the prior discussion however I have 2 new rk boards coming so I'm curious)
matthias_bgg has joined #linux-rockchip
SallyAhaj has joined #linux-rockchip
<Jmabsd> Tenkawa: which RK boards do you have coming?
<Jmabsd> Tenkawa: So, I saw on Firefly's web site that the RK3588 SBC is buyable.
<wens> IIRC that happened very recently?
<Tenkawa> I have the Rock5 "ordered" and in theory the Odroid-M1 should be shipping anytime
<Tenkawa> they have no eta on Rock5 release yet
<Tenkawa> M1 suppose to be shipping this week
<Tenkawa> so its the RK3588 and RK3568B2
SallyAhaj has quit [Remote host closed the connection]
SallyAhaj has joined #linux-rockchip
SallyAhaj has quit [Remote host closed the connection]
SallyAhaj has joined #linux-rockchip
<Jmabsd> RK3568B2 is not at all as wholesome as RK3588 is it
<Jmabsd> RK3588 has PCIe and 32GB RAM cap, and is fast
<Jmabsd> Tenkawa: do you have any clue about when they will ship the Rock5
<daniels> it's mainly that RK3588 has 4x A78 + 4x A55, whereas RK3568 is just 4x A55
<daniels> and those cores are ... not fast
<Jmabsd> Oh right, you are right, the RK3568 indeed has PCIe. Cool.
<Tenkawa> Jmabsd: no idea... I'm anxious though for mine..
<Jmabsd> daniels: you mean the A55 cores. Really. Ok.
<Tenkawa> daniels: no.. the key with the M1 is its IO
<Tenkawa> its native NVME
<Tenkawa> its being marketted as a RPI CM4 competitor
<Jmabsd> RockChip RK3588 versus Apple M1: https://browser.geekbench.com/v5/cpu/compare/14087462?baseline=11836244 RK3588 is a quite humble performer compared to the M1. But, its ecosystem is more open, there will be more hardware variants available, also probably better Linux etc. support because Apple is not sharing any hardware documentation.
<Tenkawa> (which is what I plan on repurposing)
<Tenkawa> Jmabsd: completely different type of M1
<Jmabsd> sure i know Apple M1 and Odroid M1 are totally different things.
<daniels> Tenkawa: yeah, there are some parts of it that are nice, but you also top out I/O surprisingly quickly on 3568
<daniels> it looks like they've ganged together some peripherals in such a way that you don't get close to the theoretical bandwidth you should achieve
<daniels> luckily slow I/O isn't really a problem because your CPU's never ready to do anything anyway
<Jmabsd> RK3588 is ~~50% faster than RK3399, interesting it's not more considering the litography upgrade.
<Jmabsd> daniel: The RK3588 in contrast is "how fast"?
<Jmabsd> looks like being the fastest ARM64 desktop SoC today apart from Apple M1??
<Tenkawa> daniels: I'll be running plenty of tests when I get it.
<daniels> Tenkawa: enjoy :)
<Tenkawa> Jmabsd: a lot of it is going to be the tuning though
<Tenkawa> I use "all" of my socs as servers anyway so I have very specific tuning parameters
<robmur01> I'd hesitate to call RK3588 "desktop" ;)
<Tenkawa> my rockpro64 RK3399 is nice... only annoyance with it is its hwrng "issue"
<Tenkawa> robmur01: in the right hamds it would make a nice one though
<robmur01> I mean, yeah, I run a GNOME desktop on my RK3399 board, mostly out of stubbornness, but nothing Rockchip makes is likely to worry my 32-core 3GHz eMAG workstation, much less those newer Altra boxes
<Tenkawa> robmur01: thats extreme overkill for a desktop in that same statement
<Tenkawa> the rk3399 could do just fine... but that emag or even a Honeycomb LX2 for example.... way more than what they call it "workstation"
<Tenkawa> 16 cpus?
<Tenkawa> (cores rather)
<Jmabsd> robmur01: what is not "desktop" about it, you think it's too slow still?
<Jmabsd> robmur01: oh, Altra, eMag, woo! can you buy normal mini-ITX/mini-ATX/ATX motherboard of those high-end ARM64?
<Jmabsd> robmur01: the RK3588 is a breakthrough ARM64 in that it brings >>4GB RAM cap into a small form factor and relatively low cost device segment
<daniels> ^
<daniels> robmur01: traditionally desktops are smaller than desks
<Jmabsd> Tenkawa,robmur01: Some activities like software development, are really tight with 4GB RAM. it's not a pleasant environment. the Raspberry Pi 4 recently brought 8GB into the small&low cost segment, that was great done. (though its PCIe is slow.)
<Jmabsd> Tenkawa: Which of Altra, eMag, Honeycomb LX2 exist in a small form factor?
<Tenkawa> the pi4 8gb is actually still only usb
<Tenkawa> the only one that uses an actual active pcie interface is the cm4
<robmur01> daniels: admittedly it does actually sit *under* the desk :D
<Tenkawa> and even its very unwieldly on the default io board (and single lane)
SallyAhaj has quit [Read error: Connection reset by peer]
<Jmabsd> Tenkawa: The Pi 4 Compute Module is PCIe.
<Jmabsd> Tenkawa: Yeah the default IO board for CM4 is useless. You want the ... what's the name..
<Tenkawa> much faster though... if only the module came in a 8gb version wjen I bought it
SallyAhaj has joined #linux-rockchip
<Tenkawa> Jmabsd: I just said the cm4 was
SallyAhaj has quit [Remote host closed the connection]
<Tenkawa> the "PI4" is not
SallyAhaj has joined #linux-rockchip
<Jmabsd> Tenkawa: www.oratek.com , that is the good one
<Tenkawa> waveshare also makes a good one
<Jmabsd> Oratek Tofu. Has an M.2 slot for NVMe SSD.
<Jmabsd> URL to the Wavehare?
<Tenkawa> ahh yes the tofu
<Jmabsd> AFAIK they sell a 3D printed chassis too.
<Jmabsd> Tenkawa,robmur01: Can you tell me if there are mATX/ATX/ITX/mITX motherboards with those high end ARM64:s you just talked about
<Tenkawa> the Rock5 isn't out yet
<Tenkawa> so no on that one
<Tenkawa> if you are in the states Ameridroiud is a good RockChip vendor
<Tenkawa> er Ameridroid
<robmur01> Jmabsd: the LX2 carrier board claims to be mini-ITX; that AVA thing looks to be eATX
<Jmabsd> Tenkawa: Looking at the Waveshare page I think the Oratek Tofu is vastly superior as desktop type base board+chassi. the Waveshare looks like embedded industrial use
<Jmabsd> robmur01: could you send me URL to both?
<robmur01> realistically, the beefier SoCs are going to be constrained by the board space for DIMM sockets and PCIe slots to make them useful, so aren't likely to ever be available in tiny form factors
<Jmabsd> the eMAG or was it Altra I saw, was one big chassi with a special form factor motherboard, for like 6,000 USD, it was not a perfect fit
<Jmabsd> robmur01: yea.
<Tenkawa> Jmabsd: the whole idea for waveshare is efficiency in space.. why make it big and bulky?
<Tenkawa> the arm architecture doesn't suffer from the same problems that x86 does mostly in that regard.
<jakllsch> why do you say that?
<Tenkawa> jakllsch: none of my arm devices (11 including my M1 Mac) need active cooling for heavy usage all day. My 3 x86-64 overheats if you look at them wrong
<Tenkawa> the chassis on these arm boxes are tiny in comparison as well
<lurchi_> The M1 requires a massive amount of silicon combined with the very latest fabrication node to reach it. And you have to pay a hefty price for it
<lurchi_> The other ARM devices are just in a complete different league computing wise ...
<jakllsch> yeah.
<Jmabsd> robmur01,Tenkawa: any URL much appreicated
SallyAhaj has quit [Read error: Connection reset by peer]
SallyAhaj has joined #linux-rockchip
SallyAhaj has quit [Remote host closed the connection]
SallyAhaj has joined #linux-rockchip
chewitt has joined #linux-rockchip
lurchi_ is now known as lurchi__
<macromorgan> mmind00: You're listed as the author of rockchip-saradc.c. Can I ask you a question about it?
<macromorgan> should the index for values in the data struct that gets pushed via iio_push_to_buffers_with_timestamp() match the scan_index? https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/tree/drivers/iio/adc/rockchip_saradc.c?h=v5.17#n267
<macromorgan> if yes, that driver needs to be corrected. If no, there is a bug in this driver here I need to reach out and resolve: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/tree/drivers/input/joystick/adc-joystick.c#n42
zoiahorn_ has joined #linux-rockchip
zoiahorn has quit [Ping timeout: 272 seconds]
SallyAhaj has quit [Remote host closed the connection]
SallyAhaj has joined #linux-rockchip
SallyAhaj has quit [Remote host closed the connection]
SallyAhaj has joined #linux-rockchip
zoiahorn_ has quit [Ping timeout: 268 seconds]
SallyAhaj has quit [Remote host closed the connection]
SallyAhaj has joined #linux-rockchip
warpme___ has joined #linux-rockchip
lurchi__ is now known as lurchi_
SallyAhaj has quit [Remote host closed the connection]
Tenkawa has quit [Quit: ... ... ... ... ...]
Tenkawa has joined #linux-rockchip
Tenkawa has quit [Client Quit]
Jmabsd has quit [Quit: Leaving]
warpme___ has quit [Quit: Connection closed for inactivity]