sorear changed the topic of #riscv to: RISC-V instruction set architecture | https://riscv.org | Logs: https://libera.irclog.whitequark.org/riscv | Matrix: #riscv:catircservices.org
aredridel9 has joined #riscv
aredridel has quit [Ping timeout: 246 seconds]
aredridel9 is now known as aredridel
EchelonX has quit [Quit: Leaving]
aredridel has quit [Ping timeout: 245 seconds]
vagrantc has quit [Quit: leaving]
Stat_headcrabed has joined #riscv
<Stat_headcrabed> mps: what configuration would work for nvme booting?
<Stat_headcrabed> Downstream u-boot could boot but upstream can't?
<Stat_headcrabed> Also, what kernel you are using now?
<Stat_headcrabed> 🤔
Stat_headcrabed has quit [Remote host closed the connection]
Tenkawa has quit [Quit: Was I really ever here?]
Tenkawa has joined #riscv
aburgess has quit [Remote host closed the connection]
aburgess has joined #riscv
handsome_feng has joined #riscv
Tenkawa has quit [Quit: Was I really ever here?]
simpl_e has quit [Ping timeout: 246 seconds]
stolen has joined #riscv
davidlt has joined #riscv
powderhorn has joined #riscv
GenTooMan has quit [Remote host closed the connection]
GenTooMan has joined #riscv
Stat_headcrabed has joined #riscv
Stat_headcrabed1 has joined #riscv
The_Decryptor0 has joined #riscv
stolen has quit [Ping timeout: 246 seconds]
Stat_headcrabed1 has quit [Client Quit]
Stat_headcrabed has quit [Ping timeout: 246 seconds]
The_Decryptor has quit [Ping timeout: 246 seconds]
The_Decryptor0 is now known as The_Decryptor
stolen has joined #riscv
Stat_headcrabed has joined #riscv
Stat_headcrabed has quit [Quit: Stat_headcrabed]
davidlt has quit [Ping timeout: 245 seconds]
stolen has quit [Quit: Connection closed for inactivity]
ntwk has quit [Quit: ntwk]
BootLayer has joined #riscv
davidlt has joined #riscv
Valeria22 has quit [Quit: Konversation terminated!]
jacklsw has joined #riscv
crabbedhaloablut has joined #riscv
Bluefoxicy has quit [Ping timeout: 245 seconds]
zBeeble42 has joined #riscv
zBeeble has quit [Ping timeout: 246 seconds]
sakman_ has joined #riscv
clandmeter6 has joined #riscv
SpaceCoaster_ has joined #riscv
ximet437 has joined #riscv
Perflosopher has joined #riscv
sympt3 has joined #riscv
cwittlut has joined #riscv
ahs3_ has joined #riscv
TMM__ has joined #riscv
cousteau has joined #riscv
gurki_ has joined #riscv
wmat_ has joined #riscv
agraf_ has joined #riscv
wbx_ has joined #riscv
mmind00_ has joined #riscv
kaol_ has joined #riscv
carefish1 has joined #riscv
Leopold_ has joined #riscv
TMM_ has quit [*.net *.split]
pabs3 has quit [*.net *.split]
sakman has quit [*.net *.split]
Leopold has quit [*.net *.split]
ahs3 has quit [*.net *.split]
Perflosopher5 has quit [*.net *.split]
ximet43 has quit [*.net *.split]
SpaceCoaster has quit [*.net *.split]
kgz has quit [*.net *.split]
sympt has quit [*.net *.split]
wmat has quit [*.net *.split]
cwittlut_ has quit [*.net *.split]
wbx has quit [*.net *.split]
Esmil has quit [*.net *.split]
clandmeter has quit [*.net *.split]
gurki has quit [*.net *.split]
agraf has quit [*.net *.split]
carefish has quit [*.net *.split]
kaol has quit [*.net *.split]
mmind00 has quit [*.net *.split]
SpaceCoaster_ is now known as SpaceCoaster
sympt3 is now known as sympt
clandmeter6 is now known as clandmeter
agraf_ is now known as agraf
frkzoid has joined #riscv
Esmil has joined #riscv
freakazoid332 has quit [Ping timeout: 248 seconds]
kgz has joined #riscv
pabs3 has joined #riscv
Leopold_ has quit [Ping timeout: 245 seconds]
MaxGanzII_ has joined #riscv
wbx_ is now known as wbx
ldevulder has quit [Quit: Leaving]
ldevulder has joined #riscv
danilogondolfo has joined #riscv
Bluefoxicy has joined #riscv
ezulian has joined #riscv
shachaf has joined #riscv
powderhorn has quit [Ping timeout: 246 seconds]
Wickram has joined #riscv
gurki_ is now known as gurki
mmind00_ is now known as mmind00
mmind00 has quit [Quit: mmind00]
mmind00 has joined #riscv
joev1 has quit [Ping timeout: 246 seconds]
joev1 has joined #riscv
jacklsw has quit [Ping timeout: 244 seconds]
joev1 has quit [Ping timeout: 252 seconds]
joev1 has joined #riscv
kaol_ is now known as kaol
cousteau has quit [Quit: ♫ I can't forget the day I shot that network down ♫]
valdemaras has joined #riscv
TMM__ has quit [Quit: https://quassel-irc.org - Chat comfortably. Anywhere.]
joev1 has quit [Ping timeout: 252 seconds]
TMM_ has joined #riscv
joev1 has joined #riscv
joev1 has quit [Ping timeout: 252 seconds]
joev1 has joined #riscv
Stat_headcrabed has joined #riscv
joev1 has quit [Ping timeout: 245 seconds]
joev1 has joined #riscv
joev1 has quit [Ping timeout: 252 seconds]
joev1 has joined #riscv
jmdaemon has quit [Ping timeout: 245 seconds]
valdemaras has quit [Quit: valdemaras]
Leopold has joined #riscv
Tenkawa has joined #riscv
joev1 has quit [Ping timeout: 245 seconds]
joev1 has joined #riscv
BootLayer has quit [Quit: Leaving]
Stat_headcrabed has quit [Quit: Stat_headcrabed]
joev1 has quit [Ping timeout: 260 seconds]
joev1 has joined #riscv
valdemaras has joined #riscv
andyc has joined #riscv
handsome_feng has quit [Quit: Connection closed for inactivity]
ntwk has joined #riscv
BootLayer has joined #riscv
Wickram has quit [Quit: WeeChat 4.0.2]
valdemaras has quit [Ping timeout: 244 seconds]
valdemaras has joined #riscv
<bjoto> palmer: andyc: For the V ptrace stuff. *Only* patch 1 and 2, *not* 3, right? Or the whole series? Are the libc folks (Maciej) happy with patch 3?
jacklsw has joined #riscv
<palmer> just 1+2
<bjoto> FWIW, here's the diff of prabhakar's v11 and v12: http://ix.io/4Ejl/diff Looking in the IRC logs, it seems to be what arnd asked for. arnd?
phoooo has joined #riscv
MaxGanzII__ has joined #riscv
<palmer> ya, I think so too. I'm OK taking this for the merge window, if Arnd is?
<Tenkawa> its getting very close
<Tenkawa> cut off the ?u forward
MaxGanzII_ has quit [Ping timeout: 246 seconds]
crabbedhaloablut has quit [Ping timeout: 258 seconds]
crabbedhaloablut has joined #riscv
<Tenkawa> looks like they might be getting fairly close to fixing the mainline u-boot soon
iooi has quit [Quit: iooi]
iooi has joined #riscv
<andyc> Ok, I will try define a struct for ptrace to make them happy. Do you expect to have it on 6.5 release or next
iooi has quit [Client Quit]
<andyc> So the struct proposal for signal is too late, right? Even though it will look the same on binary level.
powderhorn has joined #riscv
<arnd> bjoto, palmer: yes, please merge the dma series
<bjoto> That'll make Esmil happy! :D
<mps> Tenkawa: I didn't but I see now
<mps> Tenkawa: I have built mainline u-boot and it works if I remove nvme from the board
iooi has joined #riscv
<Tenkawa> Yes.. you obviously didn't read the post
<mps> Tenkawa: but with nvme it simply hangs probing nvme
<Tenkawa> please apply or boot would fail. Also you could check this series if you want to try to use usb/nvme in u-boot.
<Tenkawa> there is more that needs to be done for nvme .. did you verify those?
<mps> I applied these patches, at the end I found this issue by bisecting u-boot and posted report upstream
<mps> on first try I simply reverted this problematic commit
<Tenkawa> mps: you applied both sets of patches?
<mps> yes
<Tenkawa> Its definitely something with your setup then
iooi has quit [Ping timeout: 252 seconds]
<mps> could be 'disagreement' of u-boot and kernel DTS
MaxGanzII__ has quit [Remote host closed the connection]
<Tenkawa> I would respond back to them and see if you can get a precompiled... and yes that is entirely possible too
MaxGanzII__ has joined #riscv
<Tenkawa> I ran into that here at first too
junaid_ has joined #riscv
<mps> and I applied also config patch posted by Stat_headcrabed but it also doesn't solve problem
<Tenkawa> How far into the boot does it get before it crashes?
<Tenkawa> (or stops)
Leopold has quit [Remote host closed the connection]
Leopold has joined #riscv
<mps> kernel hangs probing nvme
<Tenkawa> That's what I was running into before I added boot settings
<mps> if I let it for 20-30 minutes it actually do some tasks
<Tenkawa> (and some of my drives still don't play nice)
<Tenkawa> these are the settings I'm using currently
<Tenkawa> nvme.io_queue_depth=2048 nvme.max_host_mem_size_mb=512 nvme.write_queues=4 nvme.poll_queues=4 nvme_core.default_ps_max_latency_us=0 nvme_core.apst_secondary_timeout_ms=100
<mps> Tenkawa: is this your repo https://gitlab.com/pnru/VF2_boot
<Tenkawa> no
<Tenkawa> I didn't touch u-boot or opensbi.. only the kernel
<mps> hm, and you are not paul63 on forum?
<Tenkawa> but the kernel was not playing nice with 6.x until I started tuning
<Tenkawa> no
<Tenkawa> I'm arakeen9
<Tenkawa> or Tenkawa/Orac on all forums
<mps> and what kernel you use
phoooo has quit [Quit: Client closed]
Valeria22 has joined #riscv
ntwk has quit [Read error: Connection reset by peer]
<Tenkawa> uname -a
<Tenkawa> Linux vf2 6.4.11-vf2+ #33 SMP Sat Aug 19 14:57:11 EDT 2023 riscv64 GNU/Linux
<mps> if something sudden doesn't happen I plan on weekend to go through all these patches again and try to find problem
<mps> I tried with 6.4.11 and 6.5-rc7 with aurel32 patches, result is same
<aurel32> not that in my case the NVME only works fine, after power-on, it hangs during kernel boot after a reboot
<mps> aurel32: with mainline u-boot or current starfive/debian?
<aurel32> starfive/debian
<mps> in my case kernel works fine with starfive/debian u-boot. problem happens only with mainline u-boot
<mps> and it also works with reboots fine
<mps> btw, I create small patch for DTS to fix e/mmc order in kernel, not sure where to post it
<mps> I prefer to have fixed devices order on every boot and not randomly
jack_lsw has joined #riscv
stolen has joined #riscv
knolle has joined #riscv
valdemaras has quit [Ping timeout: 260 seconds]
jacklsw has quit [Quit: Back to the real life]
iooi has joined #riscv
jack_lsw has quit [Ping timeout: 250 seconds]
ezulian has quit [Ping timeout: 252 seconds]
<ardb> conchuod: jrtc27 ^^^
junaid_ has quit [Remote host closed the connection]
<conchuod> Yeah, I saw your reply to that this afternoon. It's on my to-do list for tomorrow
<jrtc27> maybe I'm missing something, but if you don't even have BS, why do you need to care about UEFI using memory?
<jrtc27> just pave over it?..
<ardb> it's for inferring the memory types
<ardb> e.g., tables in RAM reserved for RAS may require non-cacheable mappings
<ardb> if AML code accesses those regions, it needs to use matching memory attributes (on ARM) or it may see garbage
<ardb> these guys basically rip out the boot picker part of UEFI and replace it with linux
vagrantc has joined #riscv
<ardb> but all the firmware pieces that underly it still need to interact as before
<jrtc27> I want linuxboot to go die in a fire
<jrtc27> but *shrug* do what you want for linux I guess
<ardb> yeah it's awful
<jrtc27> (well, either go die in a fire or come up with an actual specified standard that isn't just "what linux does")
<jrtc27> imagine the uproar if platforms only did windowsboot
<jrtc27> bunch of hypocrites...
<jrtc27> anyway /rant
<ardb> no disagreement here
<ardb> if only folks were a bit more pragmatic when it comes to boot support across OSes and architectures
<ardb> past experiences on x86 are wholly irrelevant here
<muurkha> probably you could put together a simple boot speciication linuxboot could implement which would enable it to boot other OSes?
<jrtc27> and suddenly you end up finding that the UEFI parts in EBBR are there for a reason
<jrtc27> because you've just implemented something that's entirely equivalent
<jrtc27> save for naming and not using GUIDs all over the place
<muurkha> it sounds like you're saying UEFI is approximately the simplest workable solution for providing the necessary information to boot, rather than the astonishingly overcomplex design I see it as?
<muurkha> if that's what you say, well, you know a lot more about UEFI and booting than I do, so maybe I'm just wrong
iooi has quit [Quit: iooi]
<jrtc27> the EBBR subset thereof
EchelonX has joined #riscv
<jrtc27> other than I guess runtime services which you could do without
<jrtc27> but those are pretty boring themselves
<ardb> and mostly useless
<ardb> in linux, at least - setvariable is typically used once to set the boot path at installation time
<jrtc27> on arm64 the standard pl*mumble* rtc isn't exposed via acpi, only uefi
<ardb> ah right
<ardb> and there's ResetSystem too
<jrtc27> and if you don't have acpi you might want reset
<jrtc27> yeah
<jrtc27> but that's about it other than efibootmgr
<ardb> yeah and we made them all optional now
<ardb> interestingly, the architects that were involved from the start are quite happy to rip stuff out
<ardb> it is mostly the BIOS vendors that see UEFI as a way to add value and stick to their 90s business model
<ardb> linux itself only uses a tiny subset of UEFI in practice
<ardb> (but the bootloader will use the block and network device protocols etc)
<Tenkawa> uefi in general is a major overengineer for most practical purposes
<Tenkawa> In this hardware especially using Linux
Andre_Z has joined #riscv
junaid_ has joined #riscv
valdemaras has joined #riscv
ezulian has joined #riscv
vagrantc has quit [Quit: leaving]
heat has joined #riscv
andyc has quit [Quit: Connection closed for inactivity]
heat has quit [Read error: Connection reset by peer]
heat has joined #riscv
heat has quit [Ping timeout: 252 seconds]
heat has joined #riscv
Trifton has joined #riscv
davidlt has quit [Ping timeout: 250 seconds]
Trifton_ has quit [Ping timeout: 260 seconds]
heat has quit [Read error: Connection reset by peer]
ezulian has quit [Ping timeout: 260 seconds]
prabhakarlad has quit [Quit: Client closed]
cousteau has joined #riscv
ntwk has joined #riscv
wmat_ is now known as wmat
crabbedhaloablut has quit []
theruran has quit [Quit: Connection closed for inactivity]
prabhakarlad has joined #riscv
junaid_ has quit [Remote host closed the connection]
iooi has joined #riscv
<mps> Tenkawa: source of alpine files used to build kernel I use is here https://gitlab.alpinelinux.org/nmeum/alpine-visionfive/-/tree/main/starfive/linux-starfive
<Tenkawa> Yeah let me take a look
<Tenkawa> the config will tell me a lot
<mps> diff to mainline u-boot repo is here https://tpaste.us/ogdq
BootLayer has quit [Quit: Leaving]
<Tenkawa> I think its more the kernel than u-boot (I'm running vendor u-boot so it wouldn't help my comparison anyway
<mps> complete boot log with u-boot mainline is here https://dev.alpinelinux.org/~mps/riscv64/boot-fail
<mps> this kernel boots fine with vendor u-boot
cousteau has quit [Read error: Connection reset by peer]
cousteau_ has joined #riscv
<mps> Stat_headcrabed: ^
stolen has quit [Quit: Connection closed for inactivity]
<Tenkawa> Default ASPM policy (BIOS default) ---> that not set to performance is going to be problematic
<Tenkawa> PCI Express hierarchy optimization setting (Performance) ---> PCI-E isnt set at all on your config
<mps> it is not needed
<Tenkawa> actually it will impact your NVMe heavily
<Tenkawa> since thats what it uses
<mps> I have impression that nvme is faster without it
<Tenkawa> no.. its not
<Tenkawa> I ran hundreds of tests
<mps> `hdparm -t --direct /dev/nvme0n1` => Timing O_DIRECT disk reads: 994 MB in 3.00 seconds = 331.00 MB/sec
<mps> I can test with kernel which have it set
<Tenkawa> hdparm is "not" an accurate assesment of the viability of the drive
<mps> give me a minute
<Tenkawa> and run it without direct and note how different your numbers are
<mps> I know hdparm is very good but with it fast results can be seen
<Tenkawa> the controller will give you much differen numbers
<mps> Timing buffered disk reads: 652 MB in 3.00 seconds = 217.03 MB/sec withot -t direct
<Tenkawa> I also gave you a set of boot params to try earlier
<mps> `hdparm -t --direct /dev/nvme0n1` => Timing O_DIRECT disk reads: 998 MB in 3.00 seconds = 332.26 MB/sec
<mps> this is with kernel which PCI-E and performance enabled
<Tenkawa> That was 4/5 of the IO contention errors I was having
<Tenkawa> mps: if its that config you just sent. it is "not" enabled
<mps> no, last one with config where it is enabled
<Tenkawa> what one?
<mps> it is in not yet published repo
<Tenkawa> ok
<mps> I have few different kernels with different configs for testing
aerkiaga has joined #riscv
jmdaemon has joined #riscv
aerkiaga has quit [Remote host closed the connection]
valdemaras has quit [Read error: Connection reset by peer]
EchelonX has quit [Quit: Leaving]
vagrantc has joined #riscv
EchelonX has joined #riscv
mps has quit [Ping timeout: 245 seconds]
mps has joined #riscv
TMM_ has quit [Quit: https://quassel-irc.org - Chat comfortably. Anywhere.]
TMM_ has joined #riscv
Andre_Z has quit [Quit: Leaving.]
Trifton has quit [Remote host closed the connection]
Trifton has joined #riscv
danilogondolfo has quit [Remote host closed the connection]
vagrantc has quit [Quit: leaving]
KombuchaKip has quit [Quit: Leaving.]
MaxGanzII__ has quit [Ping timeout: 246 seconds]
aburgess has quit [Remote host closed the connection]
aburgess has joined #riscv