ChanServ changed the topic of #armlinux to: ARM kernel talk [Upstream kernel, find your vendor forums for questions about their kernels] | https://libera.irclog.whitequark.org/armlinux
bps2 has quit [Ping timeout: 264 seconds]
apritzel has quit [Ping timeout: 264 seconds]
heat_ has quit [Remote host closed the connection]
heat_ has joined #armlinux
heat_ has quit [Remote host closed the connection]
heat has joined #armlinux
heat has quit [Remote host closed the connection]
jclsn has quit [Ping timeout: 245 seconds]
jclsn has joined #armlinux
amitk has joined #armlinux
elastic_dog has quit [Ping timeout: 245 seconds]
elastic_dog has joined #armlinux
monstr has joined #armlinux
prabhakarlad has quit [Quit: Client closed]
cbeznea has joined #armlinux
frieder has joined #armlinux
bjoto has joined #armlinux
amitk_ has joined #armlinux
sszy has joined #armlinux
heat has joined #armlinux
frieder has quit [Ping timeout: 250 seconds]
Forty-Bot has quit [Server closed connection]
Forty-Bot has joined #armlinux
monstr has quit [Remote host closed the connection]
luispm has joined #armlinux
frieder has joined #armlinux
headless has joined #armlinux
apritzel has joined #armlinux
<conchuod>
ardb: Half that "added a firmware information passing method FFI" is missing from lore. You seem to make a good argument that regardless of breaking arm64, it is a bad idea.
<ardb>
conchuod: yeah it get a bit tired of the 'efi sucks so whatever we dream up is necessarily better' attitude
<ardb>
i can forward you the thread if you want (assuming gmail let's me)
<ardb>
i guess it depends on whether RISC-V is as finicky about mismatched memory type attributes as ARM is
<conchuod>
It's just the bits from Ron that are missing.
<conchuod>
>i guess it depends on whether RISC-V is as finicky about mismatched memory type attributes as ARM is
<conchuod>
Probably a question for Alex Ghiti.
<javierm>
ardb: I never fully understood why people conflate replacing vendor provided proprietary firmware with replacing EFI
<javierm>
I mean, EFI is just a spec, one can have a 100% open source firmware that implements that API...
<conchuod>
ardb: Either way, extremely wary of something you call out as a maint. burden that you do not want on arm..
<ardb>
javierm: not sure we should be debating religion here :-)
<javierm>
ardb: ok sorry then :)
amitk_ has quit [Ping timeout: 260 seconds]
<ardb>
conchuod: i think there is nothing fundamentally wrong with doing ACPI without UEFI as long as the linux plumbing for ACPI does not depend on UEFI
<ardb>
which it does for arm64
<ardb>
but not for riscv (right now, at least)
amitk_ has joined #armlinux
<conchuod>
ardb: thanks for sending those 👍
<ardb>
my pleasure - i'll put them into lore once i figure out how to contribute them
<ardb>
note that most of what's in those is either incorrect or irrelevant
<ardb>
but it explains why the OP seems confused or at least misinformed
<javierm>
ukleinek: did a drm-misc maintainer mentioned to you that commit d13b5d2b2b45 ("drm/i2c: Switch i2c drivers back to use .probe()") won't make it to v6.5-rc1 ?
headless has quit [Quit: Konversation terminated!]
<javierm>
ukleinek: it seems you are correct, I just asked tzimmermann
<javierm>
ukleinek: sorry that I didn't land in time then :(
<ukleinek>
javierm: My initial idea was to let Wolfram take this patch via his i2c tree, but now that there are some more drivers using .probe_new, there is no urgency
<ukleinek>
javierm: no need to be sorry. Compared to how other patches go (not) in, I prefer a maintainer who cares to get my code in for the 2nd merge window from now.
<javierm>
ukleinek: hopefully in v6.6 you could finally finish that rework
frieder has quit [Ping timeout: 252 seconds]
prabhakarlad has joined #armlinux
frieder has joined #armlinux
gurevitz has joined #armlinux
amitk_ has quit [Ping timeout: 264 seconds]
bryanb has quit [Ping timeout: 246 seconds]
bryanb has joined #armlinux
headless has joined #armlinux
cbeznea has quit [Quit: Leaving.]
<gurevitz>
HI all, I was asked to support Debian on ARM Cortex-A53 in a lab environment for my company. I am looking to build a cluster of 20 machines with full IPMI so I can manage all remotely. Anyone have a suggestion for hardware?
frieder has quit [Remote host closed the connection]
geertu has quit [Ping timeout: 246 seconds]
monstr has joined #armlinux
gurevitz has quit [Quit: Client closed]
monstr has quit [Remote host closed the connection]
headless has quit [Quit: Konversation terminated!]
luispm has quit [Quit: Leaving]
punit has joined #armlinux
apritzel has quit [Ping timeout: 240 seconds]
rvalue has quit [Ping timeout: 246 seconds]
amitk_ has joined #armlinux
rvalue has joined #armlinux
apritzel has joined #armlinux
ardb has quit [Server closed connection]
ardb has joined #armlinux
mkorpershoek has quit [Server closed connection]
mkorpershoek has joined #armlinux
apritzel has quit [Ping timeout: 245 seconds]
cbeznea has joined #armlinux
bps2 has joined #armlinux
apritzel has joined #armlinux
amitk_ has quit [Ping timeout: 264 seconds]
headless has joined #armlinux
geertu has joined #armlinux
<headless>
clk API and PM related question...
<headless>
if a driver's PM resume() method calls clk_prepare_enable() and continues with reading controller regs ignoring the result, wouldn't it cause s/th like async bus abort on the reg read then?
<headless>
or s/th of that sort... rather then just returning all 0s or 1s (I understand this is platform specific)
<headless>
(I'm looking at drivers/usb/host/xhci-plat.c)
<geertu>
headless: On e.g. R-Car Gen2 that would cause an imprecise external abort.
<geertu>
On e.g. R_Car Gen3, it is usually caught in the firmware, so reads back e.g. zeroes
<headless>
geertu: aha, that's what I suspected
<headless>
(except firmware)
<headless>
geertu: but how does it know where to put 0s if that abort is imprecise?
<headless>
ah, gen2 vs gen3
apritzel has quit [Ping timeout: 252 seconds]
<headless>
geertu: TY :-)
<headless>
geertu: are you in Prague?
<geertu>
headless: Yes I am
<geertu>
headless: Are you in Moscow?
<headless>
geertu: yeah
<headless>
geertu: didn't even renew my passport last year
<headless>
heck! tried booking a flight on booking.com now
<headless>
actually, I just had a 4-week time-off and I need to visit 3 doctors this week, so decidec not to bother mgnt with a virtual option
<headless>
*mgmt
<headless>
*decided
<headless>
*decided
<geertu>
headless: IC
<headless>
booking a flight apparently doesn't work :-)
<headless>
(expectedly)
<headless>
geertu: another question: how serious is a kernel oops (that imprecise abort) to the resume process
<headless>
you prolly know that stuff better than me :-)
<geertu>
headless: crashes are always serious
<headless>
geertu: having a speech this time?
<headless>
geertu: say hi from me to the Renesas crowd :-)
<headless>
night
headless has quit [Quit: Konversation terminated!]