Posts by axboe@fosstodon.org
(DIR) Post #ATmKa6cJ9Al4UNAzD6 by axboe@fosstodon.org
2023-03-19T16:10:26Z
0 likes, 0 repeats
@penguin42 Thanks, that may work. Too bad they don't have debian, as everything I need to reproduce is readily available there. Don't want to waste too much time on setup, and all the enterprise'y stuff is so outdated.Can I boot custom kernels on these?
(DIR) Post #ATmKtrS3F1QZFInanQ by axboe@fosstodon.org
2023-03-19T16:13:58Z
0 likes, 0 repeats
@penguin42 That's what I assumed too. Pretty wild configuring these, I end up anywhere betwwn $740/month to $40k/month!
(DIR) Post #ATmLB3iurqJgEIt756 by axboe@fosstodon.org
2023-03-19T16:17:06Z
0 likes, 0 repeats
@penguin42 No kidding... This is why I've never used any cloud services for any testing and have my own lab for this kind of stuff. No openpower in in, though, unfortunately.
(DIR) Post #AUmdz15DCs0LsY6iSO by axboe@fosstodon.org
2023-04-18T14:22:21Z
0 likes, 1 repeats
libuv finally gets io_uring support:https://github.com/libuv/libuv/commit/d2c31f429b87b476a7f1344d145dad4752a406d4
(DIR) Post #AXZnzpU3YxEwmjQfmy by axboe@fosstodon.org
2023-07-11T01:46:55Z
0 likes, 0 repeats
Added support for waitid(2) in io_uring. Fully async, no thread offloading required. Very lightly tested... You could even imagine a multi-shot variant of this - get notified every time a process changes state.https://git.kernel.dk/cgit/linux/log/?h=io_uring-waitid
(DIR) Post #AXZnzqPU7RB3eqIXIm by axboe@fosstodon.org
2023-07-11T02:03:00Z
0 likes, 1 repeats
BTW, don't ever try to explain process management to your kids. "So you see, you fork to create a child. If someone kills that child or it exits, the parent can get notified about it. Then the parent has to reap it, nobody likes zombies."Made for a fun dinner conversation.
(DIR) Post #AXcCQ4Gn34dVh5vm3U by axboe@fosstodon.org
2023-07-12T00:48:47Z
0 likes, 0 repeats
Busy day, three patchsets posted:xfs/iomap async dio improvements:https://lore.kernel.org/io-uring/20230711203325.208957-1-axboe@kernel.dk/async waitid support for io_uring:https://lore.kernel.org/io-uring/20230711204352.214086-1-axboe@kernel.dk/futex support (wait/wake/waitv) for io_uring:https://lore.kernel.org/io-uring/20230712004705.316157-1-axboe@kernel.dk/
(DIR) Post #AXcQjKGpBKdwBB1YPY by axboe@fosstodon.org
2023-07-12T13:00:35Z
0 likes, 0 repeats
@penguin42 don’t think anyone has tried it yet, but agree that would be a useful case.
(DIR) Post #AXcZ5g8BNtZFz09RoW by axboe@fosstodon.org
2023-07-12T14:34:16Z
0 likes, 0 repeats
@penguin42 Yep, it's impossible to do that efficiently. And it's a general problem, two of the three patchsets above is literally catering to problems like that 😉
(DIR) Post #AbzSLLiw5mpiJMocQi by axboe@fosstodon.org
2023-11-19T13:04:38Z
1 likes, 2 repeats
If you enjoy the hairiest of bug hunts with a thrilling conclusion, this one is for you. The hunt and hair pulling:https://lore.kernel.org/regressions/480932026.45576726.1699374859845.JavaMail.zimbra@raptorengineeringinc.com/and the conclusion:https://lore.kernel.org/regressions/1105090647.48374193.1700351103830.JavaMail.zimbra@raptorengineeringinc.com/Hats off to Timothy for seeing this one through to completion!
(DIR) Post #AbzT0nyNy2ExnVoFJg by axboe@fosstodon.org
2023-02-15T00:09:33Z
1 likes, 0 repeats
@ljs @marcan > I guess Jens point is that dancing around who it is adds to the 'fuck (other :) kernel maintainers' vibes.Yes indeed, it hurts somewhat that a few (maybe more than a few, but definitely the minority) rotten apples ends up being a generalization against the kernel as a whole. And if I or others can help make that better, I'm surely all for it.
(DIR) Post #AmptbFpBUmxKwEzQDw by axboe@fosstodon.org
2024-10-09T15:40:31Z
0 likes, 0 repeats
David recently posted the first non-rfc version of the io_uring zerocopy receive work, you can find it here:https://lore.kernel.org/io-uring/ZwW7_cRr_UpbEC-X@LQ3V64L9R2/T/I noticed Broadcom has public fw that supports this now, so really very few (and minor) dependencies left for this work.I ran some testing. Details in the reply below:https://lore.kernel.org/io-uring/2e475d9f-8d39-43f4-adc5-501897c951a8@kernel.dk/tldr - line rate, 59% faster.
(DIR) Post #AmpueJcZw5nxXK7JdQ by axboe@fosstodon.org
2024-10-09T16:58:28Z
0 likes, 0 repeats
@pro I have no horse in that race.
(DIR) Post #AmpxaL58rZ0ClrGyPI by axboe@fosstodon.org
2024-10-09T17:28:41Z
0 likes, 0 repeats
@pro I mean, any kind of bypass solution is, in my opinion, the wrong direction. So if it renders AF_XDP less useful, then I won't shed any tears.
(DIR) Post #AoH82sNYWJGTvqwwhU by axboe@fosstodon.org
2024-11-21T13:42:32Z
1 likes, 1 repeats
In today’s news: man with zero self reflection goes on lengthy one sided rant highlighting just that.
(DIR) Post #AqHSgvTqNGwIsreKzQ by axboe@fosstodon.org
2025-01-20T00:26:33Z
0 likes, 0 repeats
@osandov man those comments are a whole other level of insane. I never read phoronix comments, this confirms that is a good decision!
(DIR) Post #AqsoYktYRbUMCc58Ay by axboe@fosstodon.org
2025-02-07T14:17:29Z
0 likes, 1 repeats
The networking side of the io_uring receive zero-copy support was finally merged, it's queued for the 6.15 kernel. See this merge for details:https://git.kernel.org/pub/scm/linux/kernel/git/netdev/net-next.git/commit/?id=71f0dd5a3293d75d26d405ffbaedfdda4836af32Now we just need to stage the io_uring side on top of that! Exciting to finally see this happening.
(DIR) Post #Ar5Riv2TZhIqQygibg by axboe@fosstodon.org
2025-02-13T19:06:56Z
0 likes, 1 repeats
marcan on leaving the Asahi project: "No matter how much we did, how many impossible feats we pulled off, people always wanted more. And more."Yep this is basically ANY open source project. Doesn't matter what you give people, they will always complain about something.
(DIR) Post #Ar5vTAcqqy0elc6oam by axboe@fosstodon.org
2025-02-13T19:36:31Z
0 likes, 0 repeats
@jani I don't have any experience with donations/patreon/etc, so can't say if that makes it worse or not. But I do know that it's already bad enough when you just give stuff away for free and have never solicited users for any donations. It's just a thankless never ending job with little to no community rewards. That can very quickly suck the fun out of it and remove that initial motivation you had for working on it, which is that it was fun.
(DIR) Post #Ar5vTCGakhONr9i4Aq by axboe@fosstodon.org
2025-02-13T19:47:17Z
1 likes, 0 repeats
@jani It's akin to the "don't make your hobby your job" advice. If you end up not having tons of fun, now you need both a new hobby AND job 🙂