wingsorc has quit [Remote host closed the connection]
jacklsw has joined #riscv
handsome_feng has joined #riscv
wingsorc has joined #riscv
seninha has quit [Remote host closed the connection]
ivii has quit [Remote host closed the connection]
pabs3 has joined #riscv
mahmutov has joined #riscv
littlebobeep has quit [Ping timeout: 240 seconds]
jimwilson has quit [Quit: Leaving]
jimwilson has joined #riscv
seninha has joined #riscv
jmd_ has joined #riscv
jmdaemon has quit [Ping timeout: 250 seconds]
elastic_1 has joined #riscv
BOKALDO has joined #riscv
elastic_dog has quit [Ping timeout: 260 seconds]
BOKALDO has quit [Read error: Connection reset by peer]
BOKALDO has joined #riscv
elastic_1 is now known as elastic_dog
vagrantc has quit [Quit: leaving]
mahmutov has quit [Ping timeout: 272 seconds]
davidlt has joined #riscv
littlebobeep has joined #riscv
seninha has quit [Ping timeout: 240 seconds]
jmd_ has quit [Ping timeout: 256 seconds]
jmdaemon has joined #riscv
davidlt has quit [Ping timeout: 246 seconds]
jmd_ has joined #riscv
jmdaemon has quit [Ping timeout: 256 seconds]
rektide has quit [Quit: Lost terminal]
rektide has joined #riscv
jmd_ has quit [Ping timeout: 248 seconds]
jmdaemon has joined #riscv
raym has quit [Remote host closed the connection]
raym has joined #riscv
riff-IRC has quit [Remote host closed the connection]
jmd_ has joined #riscv
jmdaemon has quit [Ping timeout: 248 seconds]
riff-IRC has joined #riscv
raym has quit [Quit: new kernel, rebooting...]
raym has joined #riscv
handsome_feng has quit [Quit: Connection closed for inactivity]
Sofia has quit [Remote host closed the connection]
Sofia has joined #riscv
kaph_ has quit [Remote host closed the connection]
kaph_ has joined #riscv
BOKALDO has quit [Quit: Leaving]
m5zs7k has quit [Ping timeout: 248 seconds]
m5zs7k has joined #riscv
conordooley has joined #riscv
duckworld has quit [*.net *.split]
solrize has quit [Ping timeout: 240 seconds]
<conordooley>
palmer - what is your plan for device drivers' config in defconfigs? do you want me to create an icicle defconfig or do you want things added to the regular defconfig (as modules?)?
davidlt has joined #riscv
motherfsck has quit [Quit: quit]
conordooley has quit [Quit: Client closed]
conordooley has joined #riscv
kaph_ has quit [Remote host closed the connection]
kaph_ has joined #riscv
___nick___ has joined #riscv
aerkiaga has joined #riscv
ZipCPU_ has joined #riscv
ZipCPU has quit [Ping timeout: 248 seconds]
ZipCPU_ is now known as ZipCPU
adjtm has quit [Read error: Connection reset by peer]
adjtm_ has joined #riscv
ivii has joined #riscv
toulene8 has joined #riscv
rvalles_ has joined #riscv
toulene has quit [Quit: Ping timeout (120 seconds)]
rvalles has quit [Ping timeout: 256 seconds]
toulene8 is now known as toulene
jmd_ has quit [Ping timeout: 248 seconds]
BOKALDO has joined #riscv
Andre_H has joined #riscv
aerkiaga has quit [Remote host closed the connection]
kaph_ has quit [Ping timeout: 240 seconds]
Andre_H has quit [Quit: Leaving.]
jacklsw has quit [Quit: Back to the real world]
jljusten has quit [Quit: WeeChat 3.4]
jljusten has joined #riscv
<la_mettrie>
hmm... any rookie-level tasks in linux-riscv kernel development or should i attempt to get freebsd running on visionfive
conordooley has quit [Quit: Client closed]
wingsorc has quit [Quit: Leaving]
<pabs3>
la_mettrie: upstreaming support for devices that only have vendor trees?
<la_mettrie>
thanks, have to think about it
<pabs3>
or general testing of all the Linux kernel features on the devices you have
<la_mettrie>
that might be fine as well
Sofia has quit [Ping timeout: 240 seconds]
conordooley has joined #riscv
Andre_H has joined #riscv
Andre_H has quit [Client Quit]
Sofia has joined #riscv
Sofia has quit [Remote host closed the connection]
Sofia has joined #riscv
littlebobeep has quit [Remote host closed the connection]
littlebobeep has joined #riscv
seninha has joined #riscv
cwebber has quit [Remote host closed the connection]
elastic_dog has quit [Ping timeout: 248 seconds]
bauruine has joined #riscv
elastic_dog has joined #riscv
Andre_H has joined #riscv
Andre_H has quit [Client Quit]
aerkiaga has joined #riscv
Sofia has quit [Ping timeout: 240 seconds]
Sofia has joined #riscv
raiaq has left #riscv [#riscv]
kaph has joined #riscv
EchelonX has joined #riscv
kaph_ has joined #riscv
kaph has quit [Ping timeout: 248 seconds]
<courmisch>
the VisionFive has a lot of missing stuff even on the dedicated tree
<courmisch>
e.g. HDMI audio and CEC are missing entirely
<courmisch>
not to speak of the VP6 drivers but that's probably hard to clean up
PyroPeter has joined #riscv
vagrantc has joined #riscv
motherfsck has joined #riscv
<nmeum>
hi, I have a question regarding regarding the 20191213 spec where csr support is split from the i extension. if I want to cross-compile embedded software which uses newlib for rv32imac_zicsr do I specfically need to compile the newlib files for rv32imac_zicsr as well or should binutils regard rv32imac and rv32imac_zicsr as binary-compatible and pick up the former automatically when
<nmeum>
-march=rv32imac_zicsr is requested?
kaph_ has quit [Ping timeout: 248 seconds]
Andre_H has joined #riscv
bauruine has quit [Remote host closed the connection]
<pjw>
conordooley: probably best just to add the drivers to the existing common defconfig
<palmer>
conordooley pjw: ya, that's what we usually do, with an entry in Kconfig.socs as well (which itself is a bit broken)
<la_mettrie>
courmisch: "missing stuff" means that a driver exists for them but it's not linked to the system?
<conordooley>
@palmer, pjw - is there any real reason to put it in the defconfig if kconfig.socs is selecting it automatically?
<palmer>
you can just put your SOC in defconfig
<palmer>
but IIUC savedefconfig will still add the rest of them, so it'll happen either way (that's how the selects work, it's not like I thought they did when we did the Kconfig.socs)
<conordooley>
Aye, soc is already there :)
JanC has quit [Remote host closed the connection]
JanC has joined #riscv
<conordooley>
idk, I thought it didn't pick them up? clock driver is not in the defconfig but the kconfig.socs selects it
<conordooley>
but maybe thats a bad example since it was added to kconfig.socs before the symbol actually meant anything
jacklsw has quit [Read error: Connection reset by peer]
JanC has quit [Remote host closed the connection]
JanC has joined #riscv
Sofia has quit [Remote host closed the connection]
Sofia has joined #riscv
<palmer>
conordooley: ah, sorry, for some reason I thought it wasn't. If you've got all your drivers selected in Kconfig.socs, your SOC selected in the defconfig, and you're getting your drivers then you should be good to go
jamtorus has joined #riscv
<palmer>
whatever we do to fix Kconfig.socs will avoid breaking what's there already
<conordooley>
Aight, I'll send you something later in the cycle so, when I know what's made it in this time or not
<palmer>
SG
<palmer>
you can also just include the defconfig/Kconfig.socs updates on those patch sets, I'll Ack them
<conordooley>
One last question then, what about things that aren't the soc but are on boards? module those in the defconfig?
<conordooley>
s/aren't the/aren't in the
foton_x has joined #riscv
octav1a_ has joined #riscv
geertu_ has joined #riscv
dlan_ has joined #riscv
<courmisch>
la_mettrie: yes
<conordooley>
also ye, I will add the kconfig.socs stuff if I sent more versions :)
toulene has quit [*.net *.split]
adjtm_ has quit [*.net *.split]
wolfshappen has quit [*.net *.split]
dlan has quit [*.net *.split]
geertu has quit [*.net *.split]
jrjsmrtn has quit [*.net *.split]
jellydonut has quit [*.net *.split]
foton has quit [*.net *.split]
sirn has quit [*.net *.split]
octav1a has quit [*.net *.split]
adjtm_ has joined #riscv
wolfshappen has joined #riscv
sirn has joined #riscv
jrjsmrtn has joined #riscv
wolfshappen has quit [Max SendQ exceeded]
wolfshappen has joined #riscv
kaph has joined #riscv
KombuchaKip has quit [Quit: Leaving.]
kaph_ has joined #riscv
kaph has quit [Read error: Connection reset by peer]
kaph_ has quit [Read error: Connection reset by peer]
kaph has joined #riscv
<palmer>
conordooley: IMO you can just add a board option under your SOC, that all needs to be refactored anyway
<mps>
nmeum: what binutils version you use
seninha_ has quit [Quit: Leaving]
<palmer>
the goal is to have defconfig boot everywhere, at least until it gets too crazy to do that
<palmer>
having the necessary bits in Kconfig.socs just lets us more easily know why they're enabled, if it's in defconfig then we'll never know how many targets use it
kaph has quit [Ping timeout: 250 seconds]
<conordooley>
palmer: when you say "board option", i assume you mean a kconfig symbol, say along the lines of CONFIG_MACH_MPFS_ICICLE?
<palmer>
something like that, I'd do a sub-menu under the SOC
<palmer>
(shouldn't be that hard to tell which is which, though)
<palmer>
historically we haven't had that many SOCs and there's been one board per SOC, but things are starting to grow so we'll need to sort something scalable out at some point
<conordooley>
I have 4 different boards with the same soc on my desk ;)
<palmer>
I guess that's the problem with building production parts, people use them ;)
<palmer>
the D1 stuff is going to be even crazier, so we'll have to do something better than this mess in Kconfig.socs
<palmer>
but for now, I think it's OK to just stick more stuff in there
<palmer>
as long as it's clear what's board stuff and what's SOC stuff, so we can sort it out later
Andre_H has quit [Quit: Leaving.]
PyroPeter has left #riscv [WeeChat 3.1]
<geertu_>
palmer: conordooley: Do you really want to add per-board Kconfig options?
geertu_ is now known as geertu
<palmer>
TBH, I'm not sure how we manage this
<geertu>
(I know older architectures that do not support multi-platform (e.g. SH) have them)
<palmer>
i'd really like the RISC-V folks to have some sort of common base, so we can at least boot and get a console on all these systems, but that's not going to happen
<geertu>
palmer: Sure, but that's a defconfig issue
<palmer>
ya, but Kconfig.socs (at least as it was originally desgined) is really just there to make managing the configs easier
<geertu>
On arm64, the single defconfig is supposed to work everywhere (but it's huge, so I never use it)
<palmer>
it was originally meant to just collect all the drivers necessary for an SOC, then turn it on
<palmer>
ya, I'm definately not thinking our defconfig is going to be what everyone runs in production
<geertu>
That's something you can derive from DTS ;-)
<conordooley>
I just want to add the driver for the phy thats on the board (for now at least, im sure there will be more in the future).
<palmer>
that's a good point
<palmer>
so maybe we just abandon Kconfig.socs entirely, and do everything via defconfig?
pecastro has joined #riscv
<palmer>
conordooley: just send a patch that adds it to the SOC, with a comment saying it's for that board
<palmer>
that's good enough for now, we can sort the rest out later
<palmer>
at least folks will be able to boot ;)
<conordooley>
Well they can boot as is, that PM stuff breaks ethernet regardless of PHY driver.
<palmer>
so you're saying it's broken with or without the phy driver?
<geertu>
conordooley: When I started using the Icicle, Ethernet didn't work with my own .config (derived from starlight), but it did work with a .config derived from the MPFS BSP. I never got into bisecting the configs, but I knew CONFIG_PM was one of the differences...
<conordooley>
Well know you know why haha
<geertu>
conordooley: You don't need a separate driver for resets. As resets and clocks are deeply entangled, you can just add reset controller functionality to the clock driver.
<geertu>
It's similar on many other SoCs (e.g. most Renesas SoCs)
<conordooley>
Oh nice, that makes handling the dt-bindings a lot easier since the clock and reset registers are intermingled.
<conordooley>
Thanks :)
<geertu>
conordooley: You're welcome. Just add #reset-cells = <...> to the clock controller node.
jmdaemon has joined #riscv
<geertu>
conordooley: If you need any help, please ping me
<conordooley>
Willdo :)
peeps[zen] has quit [Ping timeout: 248 seconds]
BOKALDO has quit [Quit: Leaving]
<geertu>
conordooley: BTW, does the clock/reset controller control power to on-chip devices, too? If yes, you want to add power-controller functionality (#power-domain-cells = <...>), and start using Runtime PM in device drivers.
<conordooley>
To be honest geertu, I don't know. I'll have to check.
<palmer>
IIRC the Microchip clock/reset is entirely different than the SiFive block?
<conordooley>
I'm pretty sure that the reset register that it is currently using does not disable power, just asserts the reset
<conordooley>
but there may be another register that could disable the power. If there is, I'd assume it is in the DRI - and I am not sure if that regmap is public or not.
<conordooley>
I think it is *supposed* to be public, but not sure if it is in the regmap on the site
<conordooley>
palmer: I don't know what the sifive block is, but I kinda doubt it is the same. There's a bunch of muxes & other clocks related to the FPGA mixed in.
<palmer>
ya, it's different then
sobkas has joined #riscv
kaph has joined #riscv
<conordooley>
A lot of that is just hidden from Linux & handled by the bootloader based on the FPGA bitstream.
davidlt has quit [Ping timeout: 256 seconds]
___nick___ has quit [Ping timeout: 240 seconds]
seninha has joined #riscv
jamtorus has quit [Quit: jamtorus]
jellydonut has joined #riscv
seninha has quit [Remote host closed the connection]
kaph has quit [Ping timeout: 256 seconds]
kaph has joined #riscv
zjason` has joined #riscv
zjason has quit [Ping timeout: 272 seconds]
wingsorc has joined #riscv
Gravis has quit [Read error: Connection reset by peer]
Gravis_ has joined #riscv
peepsalot has joined #riscv
nun has quit [Read error: Connection reset by peer]