Users complain (not only in this RHBZ ticket) about very slow performance and this issue: btrfs should use crc32c-intel instead of crc32c-generic on a SSE4.2 processor.

@atim, as mentioned in the BZ this is likely a red herring. Can you link to the other places where this is being discussed / complained about so we can weigh in? Thanks!

@dcavalca here and here. It is non-EN Fedora user group in telegram messenger. User said that UX like on 5400rpm HDD on his Intel(R) Core(TM) i7-1065G7 and Intel Corporation SSD 660P Series (rev 03), all defaults, Fedora 33 Workstation. I'll ask both of them provide more details and file a bug properly, but no luck, sorry. I cannot do much there. I've triggered personally only because i found that on my system i have crc32c-generic as well but Josef already explained that this is not a problem.

There many talks and people keep talking that btrfs is terribly slow and such, but since i've asked many times provide details and numbers, so i consider this false alert and useless noise.

👀 Some users complain and saying they noticed huge regression in new 5.10 kernel https://www.reddit.com/r/linux/comments/kieqyu/warning_linux_510_has_a_500_to_2000_btrfs/

This appears to have been reported to the Btrfs mailing list: https://lore.kernel.org/linux-btrfs/cc99c0a9-6b2f-559f-c867-d2064ab46e09@exactcode.com/T/

@josef Have you seen this yet?

My kingdom for a standardized performance suite. Everything we'd been running was showing massive performance gains, and then something simple like this slips through. I'll dig in tomorrow morning.