Posts
507
Following
144
Followers
178
The cat is not mine :(

I like cycling, powerlifting, bad video games and metal.
Otherwise, I occupy my time with various bits in RISC-V land.

~useless, placeholder, website: https://www.conchuod.ie/
@pdp7 Nah, wee bit of road rash from the crash (very lucky) and nothing from the freehub breaking.
0
0
1
@pdp7 I think there's a chrome flag, maybe force-enable-dark, that controls this. I had to fiddle with this at one point too, where I had my windows box with the dark mode & my linux one without..
1
0
0
Crashed my bike (lowside @ about 25 kph) 2 weeks ago, thought my bike had escaped pretty much unscathed. Last week or so, bike started doing some weird shit, like the chain was slipping. Today, rear wheel/axle decided to become a freewheel in both directions...
1
0
0
@palmer At the moment, we ship a fixed version of which we only compile a fraction of the files due to code size restrictions.
0
0
0
@monsieuricon Unfortunately, we are still stuck with either expensive devboards or SBCs that you need to be either a) happy with running random vendor everything on, or b) okay with pulling together something yourself from patches. At least for the VisionFive 2 @esmil has a tree that's mainline-aligned, containing all the peripheral stuff that is almost-ready-but-waiting-on-dependencies, but that's not accessible to anyone other than other kernel hackers...
0
0
1
@monsieuricon None of the RPi-likes are particularly good in terms of upstream kernel support. The VisionFive 2 is nearly there & you can run some distros on it - Ubuntu is probably the best choice. It's a bit lacking performance wise compared to the recent RPi stuff though, but nothing else SBC-wise right now runs on anything close to an upstream kernel and is compatible with standard userland.
1
0
0
@palmer Perhaps I'll just chuck it into my daily CI builds because it'll probably save me hassle in the long run, but really the project needs something official.
1
0
1
Trying to bisect OpenSBI & the thing doesn't build over a bunch of the commit range that I am interested in from the default target.

Lack of automation I suppose is the problem, but OpenSBI is an official RISC-V project, so you'd like to think something basic like github actions could be set up for it...
1
0
6
It is hard to escape the feeling that I "have" to spend my evening bisecting a QEMU boot issue I found.
0
0
3
@krzk The biggest insult to your time is not doing things that you asked for on the previous versions so that you have to give the same review again.
1
0
2
@kernellogger @chrissicool I think what might have happened is that I sent the patch based on rc1 before I saw his resort (mid may) because I have definitely been more conscious of it since, and applied patchsets on top of rc2 to avoid messing things up.
0
0
0
@kernellogger guilty as charged :/ I *could* claim that c sorts before C, but the context in the diff shows I didn't even get that right...
1
0
1
@hyeyoo We straightened out the couple of pins on my brothers & it's been working fine since.
1
0
2
@hyeyoo My wee brother had that happen to him recently w/ an AMD CPU, it's a fairly common mistake. Unfortunately you've got to twist off the cooler on pre-AM5 stuff.
1
0
2
I wonder how many different people will have to come up with different ways to say no to

https://fosstodon.org/@kernellogger/110666717786176617
0
1
3
@palmer @monsieuricon Yeah, mostly my policy too, but I took the bait today.
0
0
1
@palmer @monsieuricon I also think you need to be a member of RVI or w/e to access them, so sourcing them on lore might be helpful to my workflow - but would likely get us in hot water!
0
0
0
@monsieuricon I would describe it as "a crock of shit". It doesn't help when you're already frustrated at what people are saying & then it does something to piss you off :)
1
0
1
I really hate lists.riscv.org - the UI looks like it is designed for a mobile device & generally browsing the online interface is a miserable experience.

It also appears to be completely unable to handle me wanting to have more than one email address?
2
0
5
@kernellogger @hrw at least for me it is self deprecating, and I don't have to worry about pissing someone off ;)
0
0
2
Show older