#asahi · 2021-11-16 08:07
marcanyeah#asahi · 2021-11-16 08:07
marcan~35kB or so, depending on the IP core version#asahi · 2021-11-16 08:07
marcanat least it's not signed though, so if someone is bored enough it could be reverse engineered / reimplemented#asahi · 2021-11-16 08:07
marcanit's in AVD, and we need to load its firmware, and it has to come from cutting up the kext :/#asahi · 2021-11-16 08:06
marcanoh yeah, did you see I found *another* M3?#asahi · 2021-11-16 08:06
marcanyeah, I guess those m3s are probably in charge of doing all the LS channel stuff for TB#asahi · 2021-11-16 08:05
marcanmakes sense#asahi · 2021-11-16 08:05
marcanI guess that's mostly just for TB?#asahi · 2021-11-16 08:05
marcansven: USB3 doesn't need the ACIO m3s right?#asahi · 2021-11-16 08:04
marcanand from when I took a look at the existing code... it wasn't the best engineered to begin with :P#asahi · 2021-11-16 08:03
marcanbut we need to add TB to that mix#asahi · 2021-11-16 08:03
marcanwe have some ARM boards doing DP/USB3 over Type C which is why at least we have all the infra for connector mode switching for that#asahi · 2021-11-16 08:03
marcanso this is all new to the kernel#asahi · 2021-11-16 08:03
marcanAIUI this is the first device in existence that does thunderbolt outside of intel controllers/firmware#asahi · 2021-11-16 08:00
marcanI get the feeling we aren't going to get away without sticking apple tunables in the DT for that one...#asahi · 2021-11-16 08:00
marcanit'll be more fun when we support USB3/USB4/TB3 anyway though#asahi · 2021-11-16 07:59
marcanah, I thought it was for some reason#asahi · 2021-11-16 07:59
marcanwasn't this all merged already anyway?#asahi · 2021-11-16 07:58
marcanbut the eUSB2 repeater has a bunch of weirdo state machines and even a config protocol, so it's possible that some kind of transaction wedges in a nasty way when it gets reset#asahi · 2021-11-16 07:57
marcanthe phy does not get reset externally#asahi · 2021-11-16 07:57
marcanagain it's not the phy, it's eUSB2 that is dropping out :)#asahi · 2021-11-16 07:57
marcansven: you know more about the hard dwc3 wedging in device mode, how bad was it?#asahi · 2021-11-16 07:56
marcanbut as I said, we tried pretty hard to do this a saner way; feel free to try :)#asahi · 2021-11-16 07:55
marcanespecially when USB is involved :p#asahi · 2021-11-16 07:55
marcanhardware often smells bad#asahi · 2021-11-16 07:54
marcanit would've been a lot saner if they'd handled the eUSB2 problem more like a phy, but instead they did some crazy stuff, typical for USB-IF :p#asahi · 2021-11-16 07:54
marcandon't confuse the phy and the eUSB2 repeater#asahi · 2021-11-16 07:54
marcanthe phy is internal#asahi · 2021-11-16 07:53
marcanbut the problem seems to be dwc3 wedges itself in a bad way, not just a phy link issue; ask sven about it#asahi · 2021-11-16 07:53
marcanthat's all internal anyway#asahi · 2021-11-16 07:52
marcanit's on the list of things we currently do better than macos :p#asahi · 2021-11-16 07:52
marcanbut even macOS gets wedged by this bug right now#asahi · 2021-11-16 07:52
marcanof course, if you think you can find one, feel absolutely free to dig around :)#asahi · 2021-11-16 07:52
marcanand yes, there will be a phy driver, but as I said resetting the phy is not enough (I'd have *loved* if it were, but we haven't found a way to bring it back to life after the big wedge without a bigger reset)#asahi · 2021-11-16 07:51
marcanso it does the reset when going through disconnected mode#asahi · 2021-11-16 07:51
marcanwe're just adding a quirk that changes the way dwc3 handles role switches#asahi · 2021-11-16 07:51
marcanthat's what we're using#asahi · 2021-11-16 07:51
marcanso it can switch roles#asahi · 2021-11-16 07:51
marcandwc3 already gets notifications from the connector stuff about what mode the port is in#asahi · 2021-11-16 07:50
marcanthere is already a standard way#asahi · 2021-11-16 07:50
marcanin particular when transitioning from host/device mode to none mode#asahi · 2021-11-16 07:50
marcanwe have a notification from the ace2 that a hotplug happened, which is how we currently know to reset the dwc3#asahi · 2021-11-16 07:49
marcanI forget if there's an irq...#asahi · 2021-11-16 07:49
marcanthe i2c is wired to the ace2 master bus#asahi · 2021-11-16 07:48
marcaninterrupt?#asahi · 2021-11-16 07:48
marcanbut then there's this phy business in front#asahi · 2021-11-16 07:48
marcanat least it's not dwc2, sure#asahi · 2021-11-16 07:47
marcan(I've been doing USB for ~18 years now too)#asahi · 2021-11-16 07:47
marcanit's worse :p#asahi · 2021-11-16 07:46
marcanwell, now we have Type C and eUSB2 to deal with