<palmer>
bjdooks: was it ever anything else? all I really care is to figure out if these new B-based string routines are faster on it, I can't even tell if it properly has any B-related stuff
wbn has quit [Read error: Connection reset by peer]
wbn has joined #riscv
Sofia has quit [Remote host closed the connection]
Starfoxxes has quit [Ping timeout: 260 seconds]
Starfoxxes has joined #riscv
jjido has quit [Quit: My laptop has gone to sleep. ZZZzzz…]
prabhakarlad has quit [Quit: Client closed]
Sofia has joined #riscv
craigo has quit [Quit: Leaving]
vagrantc has quit [Quit: leaving]
Trifton has joined #riscv
rurtty has quit [Quit: Leaving]
XYZ has joined #riscv
XYZ has quit [Remote host closed the connection]
XYZ has joined #riscv
XYZ89 has joined #riscv
XYZ has quit [Remote host closed the connection]
XYZ89 has quit [Remote host closed the connection]
paulk has quit [Read error: Connection reset by peer]
paulk-bis has quit [Quit: WeeChat 3.0]
paulk has joined #riscv
paulk has quit [Client Quit]
wbn has quit [Ping timeout: 256 seconds]
paulk has joined #riscv
lockna has quit [Quit: lockna]
wingsorc__ has quit [Quit: Leaving]
wingsorc__ has joined #riscv
wingsorc has quit [Remote host closed the connection]
davidlt has quit [Ping timeout: 252 seconds]
wingsorc__ has quit [Quit: Leaving]
wbn has joined #riscv
Andre_H has joined #riscv
davidlt has joined #riscv
wingsorc has joined #riscv
Stat_headcrabed has joined #riscv
lockna has joined #riscv
Stat_headcrabed has quit [Remote host closed the connection]
Stat_headcrabed has joined #riscv
davidlt_ has joined #riscv
davidlt has quit [Remote host closed the connection]
davidlt_ has quit [Remote host closed the connection]
davidlt has joined #riscv
<davidlt>
conchuod, got VF2 and booted, updated firmware, etc. The initial impression it's good enough :) Months of upstreaming and users reporting issues/testings should make it good.
Gravis has quit [Ping timeout: 272 seconds]
Gravis has joined #riscv
Stat_headcrabed has quit [Ping timeout: 260 seconds]
tusko has quit [Ping timeout: 255 seconds]
tusko has joined #riscv
micttyl has joined #riscv
Stat_headcrabed has joined #riscv
Andre_H has quit [Quit: Leaving.]
ldevulder has quit [Remote host closed the connection]
Stat_headcrabed has quit [Ping timeout: 256 seconds]
craigo has joined #riscv
<conchuod>
davidlt: At the reported perf level it should be pretty good as an SBC. Did yours come with an OS installed?
<davidlt>
conchuod, now. I just took a release (builroot) from GitHub which was released several hours ago.
<davidlt>
conchuod, note, that wget is built without HTTPS support, dropbear (SSH) doesn't have sftp support and netcat (nc) is not included :)
<davidlt>
So if you want to flash new firmware use USB drive :)
<davidlt>
conchuod, do you plan to start testing all those patchsets from StarFive?
<conchuod>
All my boards do tftpboot with nfs
<conchuod>
I've tested clock I suppose and pinctrl to a minimal extent
<davidlt>
I used to do PXE boot with Unleashed (FU540) a long time ago with U-Boot.
<conchuod>
Unfortunately ethernet is kinda broken but I reported that
<conchuod>
I'll try the other stuff as it comes in now that I'm bootable
<davidlt>
Cool. This will be extremely popular board until those TH1520 show up.
<conchuod>
I'll be happy once I'm set up to natively build stuff :)
<conchuod>
Which I think means nvme operational with mainline on it
<davidlt>
oh yea, but PCIe drive will probably come later.
<davidlt>
It took me weeks to recompile Fedora/RISCV 37 natively.
jacklsw has joined #riscv
elastic_dog has quit [Read error: Connection reset by peer]
elastic_dog has joined #riscv
<drmpeg>
Slow builds show you how much bloat is added to each release of Linux.
<davidlt>
drmpeg, you can always make Linux slim and boot fast. You can fit it within a few MBs.
<davidlt>
So far majority RISCV SBCs are lowest-end smartphone level stuff.
<drmpeg>
With the same config on Unmatched, the build time had gone from 6 hours for 5.12 to 7.5 hours for 6.1
<drmpeg>
What's really needed is about 20X the performance of the FU740.
<drmpeg>
For native builds.
<conchuod>
Yah
<conchuod>
davidlt: I just want a honeycomb lx2 type thing :)
<conchuod>
But perhaps not at $1k
tusko has quit [Ping timeout: 255 seconds]
<davidlt>
conchuod, that cannot be cheap :)
Stat_headcrabed has joined #riscv
<conchuod>
Yeah, I know... I just need to convince someone else that I need one and have them pay :)
<davidlt>
conchuod, there is 64-core RISCV server SoC, but not yet available in a product folks could buy.
Stat_headcrabed has quit [Client Quit]
<conchuod>
My employer for instance
<conchuod>
Yah, I was at one of those riscv open hours and wei fu was presenting about it
<davidlt>
conchuod, aren't you #1 person these days reviewing all the patches? :)
<conchuod>
I fear for that server being rather non standard though!
<davidlt>
Standards a still away :) We just got psABI ratified.
<davidlt>
Profiles comes next probably.
tusko has joined #riscv
<conchuod>
Yah, which is why I figure that one will be rather "special"
<davidlt>
Even Ventana stuff is not compliant with standards, because they don't exist yet :)
<davidlt>
OS-A SEE is not there yet too. Platforms same.
<davidlt>
I wouldn't be surprised if OS-A server stuff requires RVA23 profile sooner than later.
<conchuod>
Maybe im biased because I see ventana people and not that company around, but I at least figure that ventana will endeavour to meet the proposals
<davidlt>
We don't have standard IOMMU too IIRC (yet).
<davidlt>
There is definitely a difference between Ventana and that 64-core SoC vendor :)
<davidlt>
That 64-core SoC doubtfully will reach that level of compliance. It's more probably more like a large SBC, which might improve a bit later as FW/SW improves.
<davidlt>
But remember that in early ARM64 days servers were built as embedded Linux system too.
<davidlt>
That's a phase we will have too.
<conchuod>
> 2:47 PM <davidlt> conchuod, aren't you #1 person these days reviewing all the patches? :(
<conchuod>
I'm not sure about the #1, but that would be my excuse!
<davidlt>
I think, I haven't seen patch for riscv without your reviewed tag :)
<conchuod>
Oh ye, I figure it'll be that way for us too. Just if I'm gonna purchase one, I may as well get something (approaching) standard
<davidlt>
True standards systems are a bit away.
<davidlt>
Listening to Ventana CEO I get impression they want to be sooner and iterate faster.
<davidlt>
I remember HPE early ARM64 servers shipping with U-Boot and later via shim (?) booting UEFI firmware and into the system.
<davidlt>
All the funny things will begin again.
ldevulder has joined #riscv
<davidlt>
I feel the final ISA level specs will be ratified in 2023 (or that's the hope). RVA23 will ratified later this year as a major variant, which is the true base for whatever standard will be.
<davidlt>
Some non-ISA specs like IOMMU are also important.
ldevulder has quit [Client Quit]
<conchuod>
tbh the upcoming non ISA stuff seems more interesting
<davidlt>
Agreed, non-ISA specs are probably more important things right now.
ldevulder has joined #riscv
pecastro has quit [Ping timeout: 256 seconds]
lockna has quit [Quit: lockna]
lockna has joined #riscv
Stat_headcrabed has joined #riscv
Stat_headcrabed has quit [Client Quit]
<conchuod>
> 2:53 PM <davidlt> I think, I haven't seen patch for riscv without your reviewed tag :)
<conchuod>
Heh, that's a stretch. There's plenty of stuff I don't feel I can review, but I try to help where I can.
prabhakarlad has quit [Quit: Client closed]
billchenchina has joined #riscv
gdd has quit [Ping timeout: 252 seconds]
craigo has quit [Read error: Connection reset by peer]
craigo has joined #riscv
EchelonX has joined #riscv
EchelonX has quit [Client Quit]
EchelonX has joined #riscv
pecastro has joined #riscv
davidlt has quit [Ping timeout: 268 seconds]
vagrantc has joined #riscv
junaid_ has joined #riscv
wingsorc has quit [Quit: Leaving]
billchenchina- has joined #riscv
wingsorc has joined #riscv
lockna has quit [Quit: lockna]
billchenchina has quit [Ping timeout: 256 seconds]
Andre_H has joined #riscv
Andre_H has quit [Client Quit]
Andre_H1 has joined #riscv
elastic_dog has quit [Ping timeout: 246 seconds]
Starfoxxes has quit [Ping timeout: 272 seconds]
elastic_dog has joined #riscv
eroux has quit [Ping timeout: 260 seconds]
Starfoxxes has joined #riscv
Andre_H1 has quit [Quit: Leaving.]
eroux has joined #riscv
jacklsw has quit [Read error: Connection reset by peer]
<conchuod>
davidlt: did you try using any of the Zbb instructions on your board?
<conchuod>
I can't get my ethernet to do anything more than wget google.ie before it goes kaput (if it even works in the first place)
Trifton has quit [Ping timeout: 252 seconds]
Leopold_ has quit [Ping timeout: 255 seconds]
lockna has joined #riscv
<bjdooks>
davidlt: te moonshot? we got piles of complaints over having one in our datacentre
Leopold has joined #riscv
pedja has joined #riscv
Andre_H has quit [Quit: Leaving.]
JanC has quit [Remote host closed the connection]
JanC has joined #riscv
epony has quit [Quit: QUIT]
JanC has quit [Remote host closed the connection]
JanC has joined #riscv
lockna has quit [Quit: lockna]
aburgess has joined #riscv
lockna has joined #riscv
<palmer>
I bumped riscv-systemts-ci to a GCC-13 prerelease, it's still got some warnings on rc1 but patches have started to show up so I'll stash stuff on not-for-next
Leopold has quit [Ping timeout: 255 seconds]
<palmer>
I've also got this futex_time64 issue with libstdc++, not sure changing the kernel headers is the right way to go but it's kind of in line with the off64 stuff so maybe it is?
Leopold has joined #riscv
<conchuod>
Arnd's been shunting out a lot of stuff for gcc-13 issues
pabs3 has quit [Read error: Connection reset by peer]
pabs3 has joined #riscv
iooi has quit [Read error: Connection reset by peer]
aburgess has quit [Ping timeout: 246 seconds]
prabhakarlad has quit [Quit: Client closed]
<conchuod>
So uh, anyone got anything handy for poking at Zbb stuff?
iooi has joined #riscv
BootLayer has quit [Quit: Leaving]
iooi has quit [Read error: Connection reset by peer]
iooi has joined #riscv
micttyl has quit [Quit: leaving]
<palmer>
conchuod: yep, I'm just turning off Werror for now
lockna has quit [Quit: lockna]
<conchuod>
palmer: looks like it does in fact have (some sort of) Zbb
<conchuod>
Tried the couple instructions that are used in Appendix A of the bitmanip spec. They all seem to be present...
<conchuod>
I hope there's some testcase thing I can copy-paste cos I am not typing out all of these extensions.
<conchuod>
instructions*
JanC has quit [Remote host closed the connection]
JanC has joined #riscv
cousteau has joined #riscv
JanC_ has joined #riscv
JanC is now known as Guest3686
JanC_ is now known as JanC
Guest3686 has quit [Read error: Connection reset by peer]
davidlt has quit [Ping timeout: 264 seconds]
Trifton has joined #riscv
aburgess has joined #riscv
___nick___ has quit [Ping timeout: 252 seconds]
lockna has joined #riscv
JanC has quit [Ping timeout: 265 seconds]
JanC has joined #riscv
bjoto has quit [Ping timeout: 272 seconds]
bjoto has joined #riscv
lockna has quit [Quit: lockna]
fabs has quit [Read error: Connection reset by peer]
<conchuod>
davidlt: 0x4210427 looks to be impid?
junaid_ has quit [Ping timeout: 256 seconds]
EchelonX has quit [Quit: Leaving]
<jrtc27>
21G1.02.00 / llama.02.00-general
<jrtc27>
whatever that means
<jrtc27>
of course, SiFive's U74 page *still* claims it's RV64GBC
<jrtc27>
which is hard to achieve when B was never ratified and got canned as an alias
<conchuod>
In the core complex manual it claims RV64GC_Zba_Zbb_Sscofpmf
<jrtc27>
yeah though that's for 21G3
<conchuod>
Which is certainly more reasonable, but I dunno how much of that applies to 0x4210427
<jrtc27>
B.2 gives you all the mimpid <-> version mappings, but that's unhelpful when there's nothing telling you how those versions differ
<conchuod>
There's one online for 21G2.01.00
<conchuod>
Says Zbb .94
Trifton has quit [Ping timeout: 268 seconds]
crabbedhaloablut has quit [Remote host closed the connection]
<conchuod>
21G2.01.00 for S21 and E76 both say 0.94 too. Think it could be a safe enough guess that it has 0.94