17:31 <CosmicPenguin> Has anybody looked into crypto on sm8250 or newer? Downstream shares the part with trustzone so there are no clocks defined in the DT and I can't tell which clock I need or if it is even defined in gcc-sm8250.c

17:36 <bamse> CosmicPenguin: https://lore.kernel.org/linux-arm-msm/20211110105922.217895-1-bhupesh.sharma@linaro.org/ <- this crypto?

17:44 <CosmicPenguin> yes!

17:45 <bamse> :)

17:45 <CosmicPenguin> Odd, the reg address doesn't match downstream

17:45 <bamse> isn't that due to the qcrypto vs qcedev split downstream?

17:46 <CosmicPenguin> maybe. Really hard to tell from the DT. my qcedev has the same definitions as qcrypto

17:47 <bamse> i will ask Bhupesh to follow up on those patches, so we get that landed...but any feedback you have is welcome

17:47 <CosmicPenguin> Let me try that patch series really quickly and I can offer suggestions

17:47 <CosmicPenguin> The bam definition seems a little bit different to me too, but I barely understand bam anyway

18:23 <CosmicPenguin> I agree that the error messages should be more informative but I agree with Bhupesh that it was annoying that the driver didn't error on failure

18:24 <Mis012[m]> with their commitment to Windows on Arm, I guess we will see more disgusting abuses of non-HLOS software as HALs?

19:42 <CosmicPenguin> Bhupesh's stack puts qce specific ciphers in /proc/crypto, so great start

19:43 <CosmicPenguin> But both aes-xts entries have the same priority , so I'm not sure which one ends up getting used by cryptsetup

21:04 <calebccff> bamse: are there any PMICs with more than 2 USIDs?

21:06 <calebccff> finally finishing up the next rev of the rradc series and this is what I've come up with for walking the PMIC list to find the siblind USID which actually contains the PMIC data https://github.com/aospm/linux/commit/829a6ffc50b1b42ce0d15ee5ce2bc51474b9b6ed#diff-6e909837287e36266eb2d3f287332106da3c007c3e177e84c618f3fe330042c2R67