Jonas Dreßler

I'm a bit confused because you have a Surface Pro 3, but the commit that probably caused this applied to the Surface 3. Can you send the output of `grep...

Nevermind, it seems that Surface Pro 3 was never part of the DMI match table we have, so that's the issue here. So yeah we'll definitely need the output of...

ping @blueblue4real if you're still interested in getting this to work then I need that information as mentioned in the last comment.

cc @benjaminfaul2 @Newbytee @RussH @dplrs @vaab (you're all on the SP3 testers list). Also I wonder how wifi works in general on the SP3? If all those workarounds are not...

> Sometimes WiFi dies completely and requires a reboot (not even re-probing the module seems to help), but that only happens when resuming from suspend in my experience. Yup, that's...

Alright, would anyone like to try this patch (on top of linux-surface kernel) to enable the firmware quirk workarounds on SP3: ```patch diff --git a/drivers/net/wireless/marvell/mwifiex/pcie_quirks.c b/drivers/net/wireless/marvell/mwifiex/pcie_quirks.c index f46b06f8d643..b183c404eb8a 100644 ---...

It works pretty well in practice already you're right. The cases where a manual connect from pipewire is done right now is when headsets appear (so for example on tty...

> isn't there a risk that other subsystem start reconnecting as well? Hmm, not exactly sure what you mean by that, conflicting audio servers trying to handle the device? >...

So I finally found some time to get back to this one and did a day of hacking on the policy plugin! What I have now is an auto-connect functionality...

> in case of Central devices it is reasonable to expect that the peripheral start the ACL otherwise we may end up having to do rounds of reconnection logic just...