00:03 <deathmist> diederik: I did, they figured out that I could use ControlPortOverNL80211=false, from the discussion: "EAPoL handshake is performed but we keep receiving Packet 3/4 again and again on a 1 second interval; the Replay Key Counter is never being increased, so iwd just drops the packet"

00:03 <deathmist> "so after a while the AP gives up. either the firmware on the local card is doing something wrong, or the AP is"

00:03 <deathmist> "or maybe the driver isn't support control port over nl80211 properly; the AP should really be incrementing the Key Replay Counter if it is performing retransmits"

00:03 <diederik> ok, was worth a shot

00:09 <deathmist> the crash could be a side-effect of being disconnected though; on 8998 every time the Wi-Fi connection drops for any reason the firmware crashes always

00:11 <deathmist> https://gitlab.com/msm8998-mainline/linux/-/issues/27 fwiw here's the issue I wrote up

00:21 <Mis012[m]> randomly rebooting really is qcom's favorite use for error handlers...

22:07 <Marijn[m]> lumag_: Worked my way through handling your review, you can find the results here if you wish to test it: https://github.com/SoMainline/linux/commits/marijn/dsi-parent-clk-cleanup. No pressure to review it yet, feel free to leave that to the official channels when I send it in as a v2. Will wait with that until collecting reviews from more maintainers, ie. sboyd for the clk framework/header changes. Thanks!

22:09 <Marijn[m]> lumag_: Also I've been wondering about additionally renaming everything that currently uses `clk_hw *hw` (the ones that are only used for assignment to `provided_clocks`) to the respective clk name just like I did for all the clks that are explicitly referenced as parent and needed a stable name.

22:09 <Marijn[m]> lumag_: Final question: should we replace the `32` in `snprintf` with `sizeof(clk_name)`?

23:12 <lumag> Marijn[m]: yes, replace 32 please

23:12 <lumag> No strict opinion regarding hw renaming.