#
✕ nick:rkjnsn
#asahi-dev · 2022-09-13 04:46
rkjnsnmarcan, yeah, 44.1k. You can also pay 50¢ extra for a 24-bit download (also at 44.1k). I assume that's automatic because the creator uploaded it in 24-bits, but it's kind of absurd given that the whole track is effectively one-bit samples.
#asahi-dev · 2022-09-13 04:39
rkjnsn(reading the backlog) I take it Apple Music ended up having a sufficient copy of “I Won the Loudness War”? If not, I purchased as a FLAC from Qobuz once, and the samples were all pegged at one extreme or the other as expected.
#asahi-dev · 2022-09-05 01:46
rkjnsn#asahi-alt is probably the right channel for that.
#asahi-dev · 2022-07-29 07:44
rkjnsnIndeed.
#asahi-dev · 2022-07-29 07:39
rkjnsnThere are some SATA port multiplier enclosures that support hot plug and look pretty nice, if one is okay sharing one 6Gb/s link across multiple drives. The problem is that in my experience, Linux port multiplier support has been pretty terrible with every host controller + port multiplier combination I've ever tried it with.
#asahi-dev · 2022-07-28 22:08
rkjnsn(It's Thunderbolt only, though, not USB.)
#asahi-dev · 2022-07-28 22:05
rkjnsnIt's an OWC Envoy Express with a Seeed Studio 5-port SATA M.2 card in it, which uses a JMicron JMB58x AHCI SATA controller.
#asahi-dev · 2022-07-28 22:03
rkjnsnI realize the storage discussion was over a week ago now (just catching up on the back log), but this has been working well for me
#asahi-dev · 2022-07-28 22:03
rkjnsnuploaded an image: (396KiB) < https://matrix.org/_matrix/media/r0/download/matrix.org/fFTwoqmKIwJYoTIuapJxTuZD/thunderbolt-sata.jpeg >
#asahi-dev · 2022-06-30 21:45
rkjnsnI use it to send files to/from my Android phone.
#asahi-dev · 2022-06-30 21:45
rkjnsn“<sven> is there anything except for audio that people use Bluetooth for”
#asahi-dev · 2022-03-24 19:30
rkjnsnI wonder if the M1 Max studio uses the same AS Media chip for the front USB-C ports, or something else.
#asahi-dev · 2022-03-09 09:42
rkjnsnSo, SEP doesn't access the SSD directly, but asks the OS to do so on its behalf (and then verifies the data the OS gives it back is legit), right? Could U-Boot do something similar for EFI vars, where it gives the OS a signed blob to store for it that it can read on the next boot?
#asahi-dev · 2022-03-09 09:12
rkjnsnAh, that's right.
#asahi-dev · 2022-03-09 09:10
rkjnsnm1n1 SEP driver? 😛 (I have no idea if that would actually work.)
#asahi-dev · 2022-03-09 09:07
rkjnsnAh. Fun…
#asahi-dev · 2022-03-09 09:07
rkjnsnThat's kind of annoying, though. When I've used target disk mode in the past, it was because the system was too dead to even boot recovery..
#asahi-dev · 2022-03-09 09:06
rkjnsnGood to know. I was thinking I had to worry about target disk mode, but it looks like the Apple Silicon require entering recovery (and thus authenticating) to enable it, instead of using a key combination?
#asahi-dev · 2022-03-09 08:58
rkjnsnPersonally, I'll be happy just to have everything through Grub (and configuration) verified. Since I keep my kernel and rootfs on an encrypted partition unlocked by Grub, that should be enough to make offline attacks difficult.
#asahi-dev · 2022-02-19 08:30
rkjnsn(I also think abxoe mentioned that flushing by default would be needed for it to be accepted upstream.)
#asahi-dev · 2022-02-19 08:29
rkjnsnAs a random person on the internet, I would feel better defaulting to always flushing.
#asahi-dev · 2022-02-19 07:52
rkjnsnMaybe it's worth seeing if there's a similar speedup for F_BARRIERFSYNC vs F_BARRIERFSYNC on an HFS+ partition, which presumably can't be doing any CoW shenanigans try to enforce ordering with fewer flushes.
#asahi-dev · 2022-02-19 07:46
rkjnsnWithout some way to pass the barrier on to the ANS, how would it enforce ordering to persistent storage without requiring individual flushes?
#asahi-dev · 2022-02-19 07:41
rkjnsnRight. I mean if there was an option to translate fsyncs into drive barriers instead of full flushes.
#asahi-dev · 2022-02-19 07:08
rkjnsnmarcan, yeah, 300-600 IOPS isn't great, but it's a lot better than 46 for cases where one wants to ensure internal consistency but doesn't care about losing a second or two of writes. It's almost certainly the mode I'll end up using for fsync if we provide it as an option.
#asahi-dev · 2022-02-17 17:35
rkjnsnmarcan, would it be easy to test F_BARRIERFSYNC using your python script to see what it does and how it performs on these machines?
#asahi-dev · 2022-02-17 11:11
rkjnsnI saw a reference to a F_BARRIERFSYNC. Not sure if that's a macOS thing or just an iOS thing, but if it's available on macOS, it might be worth seeing if it works as advertised and results in less performance loss.
#asahi-dev · 2022-02-17 07:31
rkjnsnEven without support for write ordering in the kernel, it seems like (IIUC) translating fsyncs to write barriers (assuming the drive supports them) rather than flushes could at least help avoid corruption from the drive reordering things, even if it doesn't protect against writes that were supposedly committed being lost.
#asahi-dev · 2022-02-17 07:31
rkjnsnI think they were talking about barriers in Linux, as opposed to barriers in the drive, though?
#asahi-dev · 2022-02-17 07:09
rkjnsnI know it could still eat data that was supposed to be safely stored, but it'd be nice to avoid broken internal consistency.
#asahi-dev · 2022-02-17 07:09
rkjnsnOut of curiosity (since I don't know too much about this topic), if folks do want to ignore/postpone flushes on Linux, does the SSD support any other kind of ordering operation that could ensure data written after a flush doesn't hit the disk before data written prior to it?
#asahi-dev · 2022-02-17 07:06
rkjnsnAh.
#asahi-dev · 2022-02-17 07:04
rkjnsnDo we know how much of an issue it would is outside of apt and synthetic benchmarks? I'm curious if it would make sense just to make apt less aggressive with its syncs, given that in can still be a significant slowdown (albeit less of one) on other drives. Presumably most software isn't trying to sync more than 50 times a second…
#asahi-dev · 2022-02-17 06:26
rkjnsnSounds like it matches how kettenis was describing OpenBSD as working, then, which I suppose is not terribly surprising.