Rob Norris
Rob Norris
Huh, thanks. I need to fix my test rig some then, because it didn't throw those up (I mean yeah, the CI did, but I hadn't got into that yet)....
@amotin thanks for doing that! I asked on socials and a couple of people pointed out that they have various old-style tunables set in various boot configs and such, so...
Strong opening move! I've had a little play with it, here's what I got. Before the final thing, you should run `make checkstyle`, and fix the errors it throws out....
Oh the other thing I forgot to add, after loading the module it says `fastest`, and fails as above. So it is selecting `avx2` as fastest, which is good! (we...
With #17089, looks pretty great: ``` $ ./tests/zfs-tests/cmd/crypto_test -c tests/zfs-tests/tests/functional/crypto/aes_gcm_test.txt AES-GCM [generic+generic]: tests=316: passed=316 failed=0 AES-GCM [x86_64+generic]: tests=316: passed=316 failed=0 AES-GCM [aesni+generic]: tests=316: passed=316 failed=0 AES-GCM [generic+pclmulqdq]: tests=316: passed=316 failed=0...
@lowjoel here's two commits you may or may not wish to include here: - robn/zfs@a1cdddc63bb92d33c355a32b1a01e22c895e42c6: actually test `avx2` in `crypto_test` - robn/zfs@cedd824e2ba30d5167abdfdd5ed08d1763e03507: include `vaes` and `vpclmulqdq` in `/proc/spl/kstat/zfs/simd`
@AttilaFueloep is merging this PR going to mess up your work? I am all in favour of better abstractions inside the ICP, but that's going to take time to finish,...
@AttilaFueloep sounds like merging now suits everyone: you get some breathing room, @lowjoel gets his gold star, and fancy folk with AVX2 machines get a nice bump (maybe in time...
@manschwetus could you get a screenshot of the panic? It'd be _really_ useful! What compiler & toolchain versions are being used here. Are they different on the Zen5 systems? `CONFIG_X86_NATIVE_CPU`...
Not an oops, just an internal warning inside the kernel. Was there any other effect? Did things just appear to keep going? Anything else happeing at the time; errors reported...