#asahi-dev · 2025-04-16 09:36
jannauADP_CTRL is 0x0412 at that point and writing 0 to it does't change anything#asahi-dev · 2025-04-16 09:32
jannauyes#asahi-dev · 2025-04-16 09:28
jannauchaos_princess: have you seen the adp fe irq firing with ADP_INT_STATUS being consistently 0?#asahi-dev · 2025-04-16 06:34
jannauI suspect a code issue which manifests only with specific compilers / optimisations#asahi-dev · 2025-04-15 19:12
jannaustill broken. I commented the whole pm_ops#asahi-dev · 2025-04-15 19:03
jannauand for the last 5-10 boots with uart and no_console_suspend it reproduced in all attempts#asahi-dev · 2025-04-15 19:02
jannauit did not reproduce reliably for me but from what I gather from the user reports it's always (or almost always)#asahi-dev · 2025-04-15 18:58
jannauI don't think so just wanted to mention it. there was an user report that still occurs with working tiny-dfr#asahi-dev · 2025-04-15 18:56
jannaustill running without fixed tiny-dfr so that might be factor as well#asahi-dev · 2025-04-15 18:55
jannauold one, I updated to upstream in 6.14#asahi-dev · 2025-04-15 18:54
jannauI've seen only complaints about asahi-6.14.2-1 with explicit mentions that 6.13.8 was ok#asahi-dev · 2025-04-15 18:49
jannauapaprently an adp interrupt storm, so macbook pro 13-inch specific#asahi-dev · 2025-04-15 18:23
jannauas if irqs to some CPU cores are not arriving (due to the soft lockup). this is m2 so irqs are distributed by aic on the cores#asahi-dev · 2025-04-15 18:21
jannausystem is half broken though. touchpad input does not work continously#asahi-dev · 2025-04-15 18:20
jannaugenerating irqs seem to help. after pressing nothing happens except for the smc button press log lines but pressing keys repeated brings the display back on#asahi-dev · 2025-04-15 18:18
jannauno compliants :(#asahi-dev · 2025-04-15 18:15
jannaulet's hope lockdep finds ssomething#asahi-dev · 2025-04-15 18:14
jannauin 6.14.2, not reliebly broken#asahi-dev · 2025-04-15 18:11
jannauthe broken resume seems to lock related. CPUs seems to soft lockup after resume#asahi-dev · 2025-04-15 16:56
jannausven: did you already try different compiler (versions / options)?#asahi-dev · 2025-04-14 11:02
jannauI think that would cause more obvious problems #asahi-dev · 2025-04-14 10:45
jannaume neither. I suspect aop's audio endpoint works slightly different which triggers the strange error mode with aop_audio. I don't think I can do anymore tracing on t602x devices without mics#asahi-dev · 2025-04-13 20:22
jannaulooks that way. does always break at starting cpu 2?#asahi-dev · 2025-04-13 20:15
jannauthat reproduces#asahi-dev · 2025-04-13 20:05
jannaudoes the execption happens before trying to mount the rootfs? it boots until that here#asahi-dev · 2025-04-13 20:00
jannauthe binary fails for on macho parsing: https://paste.debian.net/1369328/#asahi-dev · 2025-04-13 19:47
jannaucan you upload the m1n1+payload?#asahi-dev · 2025-04-13 13:18
jannauI'm using raw m1n1 payloads though#asahi-dev · 2025-04-13 13:18
jannausven: no issues on linux but I have seen proxy communication issues during exception handling#asahi-dev · 2025-04-12 18:25
jannaufl0_id: do not interact with that person. they were banned for repeated code of conduct violations#asahi-dev · 2025-04-11 13:41
jannauI'll do direct comparison with j314. this was just a quick test of the new kernel tag#asahi-dev · 2025-04-11 13:39
jannaumuffled#asahi-dev · 2025-04-11 12:28
jannauchadmed: tested mics on j493. I think quality is noticeable worse. unsure if that the mic array layout itself, beam-forming parameters or triforce with this mic layout#asahi-dev · 2025-04-10 19:28
jannauagreed#asahi-dev · 2025-04-10 19:27
jannauclocks would be a standard property but with rust we would need of course more bindings to query the rate#asahi-dev · 2025-04-10 19:13
jannauin the m2 pro/max adt I've seen that aop has clock_ids#asahi-dev · 2025-04-10 19:12
jannauthe clocks could probably modelled with fixed rate clocks like we do with other devices#asahi-dev · 2025-04-10 19:12
jannauchaos_princess: we could transfer the values from the ADT in m1n1 but the majority of the parameters looks indeed not as they would describe hardware#asahi-dev · 2025-04-10 16:53
jannauthat's at least what the pdf devicetree spec suggests#asahi-dev · 2025-04-10 16:53
jannauI suppose we should have used the top level model proerty for that instead of our strings#asahi-dev · 2025-04-10 13:42
jannauwith that we have macvdmtool functionality on linux#asahi-dev · 2025-04-10 13:42
jannaumartinr1: welcome back. Yes, we're still interested in this. The other interesting vdm command is enabling the uart (remote and locally)#asahi-dev · 2025-04-09 16:55
jannauchaos_princess: fixes pushed to asahi-wip / asahi-6.14.1-2#asahi-dev · 2025-04-09 15:15
jannauI think it was added to avoid letting dart_reset() blow iboot mappings#asahi-dev · 2025-04-09 15:14
jannauno, m1n1 locks it. iboot locks all dcp*-darts. (I guess it could differ on a per device basis)#asahi-dev · 2025-04-09 15:09
jannauthis does not work anymore since locked and non-locked darts have different iommu ops#asahi-dev · 2025-04-09 15:08
jannauproblem is known. one issue is that it's mixing the locked disp0-dart (locked by m1n1) with dispext0-dart which is not locked#asahi-dev · 2025-04-08 12:52
jannauI'll revert the blocking patch#asahi-dev · 2025-04-08 12:52
jannauchadmed: I wasn't sure that we were convinced of its fidelity and I've never managed to test on j493#asahi-dev · 2025-04-07 22:41
jannauchadmed: no