#asahi-dev · 2023-05-11 06:05
marcanhonestly it being a bit toastly doesn't bother me that much, it *is* a metal case and they probably use it as a big heatsink#asahi-dev · 2023-05-11 06:05
marcanI guess we need to get that hwmon stuff in...#asahi-dev · 2023-05-11 06:05
marcanI assume the fans are clearly running when it gets hot?#asahi-dev · 2023-05-11 06:03
marcanthere's a GPU in there too#asahi-dev · 2023-05-11 06:03
marcanbut it shouldn't be overheating with just CPU workload, that's nuts#asahi-dev · 2023-05-11 06:03
marcanyeah, pretty much#asahi-dev · 2023-05-11 06:02
marcanbut if it causes a plain old USB connection to drop that's not good#asahi-dev · 2023-05-11 06:01
marcanI wonder if this is something SMC does on purpose when things get too hot...#asahi-dev · 2023-05-11 06:01
marcanoh, so the PD link drops? huh#asahi-dev · 2023-05-11 04:58
marcan(I don't have a lockdep build lying around, probably easier for you :p)#asahi-dev · 2023-05-11 04:58
marcanmaz: can you check if adding that flag to the cpuidle driver fixes it? (on the idle states array, for both)#asahi-dev · 2023-05-11 04:56
marcanmaz: huh. the PD controller, as in device mode? or host mode?#asahi-dev · 2023-05-11 04:56
marcanarnd: did you ever get to talk to Janni?#asahi-dev · 2023-05-10 01:58
marcanprobably just needs flags = CPUIDLE_FLAG_RCU_IDLE in the idle states#asahi-dev · 2023-05-10 01:57
marcanprobably caused by that#asahi-dev · 2023-05-10 01:56
marcanmaz: https://www.spinics.net/lists/kernel/msg4649758.html#asahi-dev · 2023-05-10 01:46
marcanthe only supported configuration is pipewire, the daemon will work with anything else but it's up to the user to set it up to do what pipewire does in some way, and if it can't do it that's their problem#asahi-dev · 2023-05-10 01:46
marcanand no I don't particularly care about people without sound servers :p#asahi-dev · 2023-05-10 01:43
marcanand if you're running without the safety daemon you have a broken system#asahi-dev · 2023-05-10 01:43
marcanso it doesn't matter that the kernel takes over them when it's not there#asahi-dev · 2023-05-10 01:42
marcanthe safety daemon needs to own the controls to do what it does#asahi-dev · 2023-05-10 01:42
marcansoftware volumes thanks#asahi-dev · 2023-05-10 01:42
marcanso nobody should be expecting to use them for anything#asahi-dev · 2023-05-10 01:42
marcanplan is to outright lock them#asahi-dev · 2023-05-10 01:41
marcanchadmed_: the DVC volumes will be owned by the safety daemon anyway#asahi-dev · 2023-05-09 15:31
marcanbecause of course it's on NVMe :p#asahi-dev · 2023-05-09 15:31
marcanthe funny thing is once it's all set up and the default boot partition, you can unplug the disk and it still boots into m1n1#asahi-dev · 2023-05-09 15:31
marcanthis is already PoCed in expert mode, you can do a USB install of just m1n1#asahi-dev · 2023-05-09 15:30
marcanso you end up with source files in Preboot (on USB) and the actual boot/ folder on iSCPreboot#asahi-dev · 2023-05-09 15:30
marcanit basically does the actual bootability provisioning in iSCPreboot instead of the Preboot partition itself#asahi-dev · 2023-05-09 15:30
marcanyes#asahi-dev · 2023-05-09 15:30
marcani.e. it works like macOS on USB#asahi-dev · 2023-05-09 15:29
marcanso no repartitioning of nvme at all#asahi-dev · 2023-05-09 15:29
marcanwhen your stub partition is on USB#asahi-dev · 2023-05-09 15:29
marcanj`ey: yes, which is done automatically by bputil/kmutil#asahi-dev · 2023-05-09 15:27
marcanxHCI has lovely things like "the buffer is full message goes in the buffer"#asahi-dev · 2023-05-09 15:27
marcanright...#asahi-dev · 2023-05-09 15:26
marcanfair#asahi-dev · 2023-05-09 15:25
marcanI dread the day I get back to figuring out why dwc3 has broken port sleep...#asahi-dev · 2023-05-09 15:25
marcanyeah but xHCI takes the standard level of USB insanity and wraps it in a mixed register/ring buffer/DMA interface ;)#asahi-dev · 2023-05-09 15:24
marcansven: this is worse, this is USB core stuff. you're lucky you just have to sequence PHYs :p#asahi-dev · 2023-05-09 14:57
marcanyeah, they're orthogonal#asahi-dev · 2023-05-09 14:46
marcanalso it would be mildly simpler to have a single global cap instead of per-speaker caps#asahi-dev · 2023-05-09 14:46
marcanbut I'd defer to chadmed's opinion on that since he knows how that'd sound#asahi-dev · 2023-05-09 14:46
marcanmeans it just gets quieter, not weird#asahi-dev · 2023-05-09 14:42
marcanI guess this is why people pay me ;)#asahi-dev · 2023-05-09 14:41
marcanI'll fix this before the next update#asahi-dev · 2023-05-09 14:41
marcanglobal pros: the speakers remain in balance. global cons: quieter#asahi-dev · 2023-05-09 14:41
marcanchadmed: practical issue: do you think we should have the kernel safe fallback volume cap global, or per-speaker?#asahi-dev · 2023-05-09 14:39
marcankettenis: ack