#asahi-dev · 2026-08-04 17:16
chaos_princessoh, that's annoying, on t603x, the power states basically just worked(tm)#asahi-dev · 2026-08-04 17:13
jannauthe other messages hanging in the same way. I guess I'll be monitoring pmgr + 0x4000. those power states are copied from t8112. the t8122 adt has none of them#asahi-dev · 2026-08-04 16:27
chaos_princesswhat if you skip sending print_enable?#asahi-dev · 2026-08-04 16:20
jannauthe IO message is not acked. all other IO and DEBUG slots have ISP_IPC_FLAG_ACK set but are otherwise empty#asahi-dev · 2026-08-04 15:43
chaos_princessright, we do not set that#asahi-dev · 2026-08-04 15:39
jannauerr, from offset 0x38 but the hint was helpful as that explains the LOG malloc request missing in the macos trace as it sets logbuf*#asahi-dev · 2026-08-04 15:30
chaos_princesshttps://github.com/AsahiLinux/linux/blob/bits/240-isp/drivers/media/platform/apple/isp/isp-fw.c#L122C2-L122C5#asahi-dev · 2026-08-04 15:29
chaos_princessyes, it should be non-zero#asahi-dev · 2026-08-04 15:26
jannaudo you see non-zero values in the ISP bootargs from offset 0x30 - 0x50?#asahi-dev · 2026-08-04 15:25
chaos_princessline 812. the device sent something#asahi-dev · 2026-08-04 15:24
jannauchaos_princess: what let's you think there is an answer to CISP_CMD_PRINT_ENABLE? as far as I can tell there is none and the memory is untouched#asahi-dev · 2026-08-03 22:08
fl0_idyou can switch to #asahi if you don't want to clutter here#asahi-dev · 2026-08-03 21:56
Tachyonray[m]fl0_id: last time I did it slightly differently. I chainloaded m1n1 first, then ran run_guest_kernel.sh. It’s probably me doing something wrong though. (hope not to clutter this channel with my noob explorations)#asahi-dev · 2026-08-03 21:53
chaos_princessmaybe print that?#asahi-dev · 2026-08-03 21:53
chaos_princessif i am reading the linux trace correctly, it does reply with something.#asahi-dev · 2026-08-03 21:48
fl0_idit's relatively recent, but still#asahi-dev · 2026-08-03 21:47
fl0_idalso to be sure, maybe also chainload m1n1, to make sure it is current/matcing#asahi-dev · 2026-08-03 21:46
fl0_idsometimes serial error can also be an issue if you didn't reboot after some failed attempt#asahi-dev · 2026-08-03 21:45
fl0_iddid you also use connection thorugh dockchannel uart before? (this is relatively new, it prob should work but I haven't used it much)#asahi-dev · 2026-08-03 21:44
Tachyonray[m]when I tried before the rebase to 7.1.5 with this setup I didn’t get an error about serial connection#asahi-dev · 2026-08-03 21:43
Tachyonray[m]I have a usb3 dock and usba-> usbc cable to the M3#asahi-dev · 2026-08-03 21:41
fl0_idTachyonray[m] how are you connected? maybe try without dockchannel uart or just chainload a m1n1 with a payload directly for testing#asahi-dev · 2026-08-03 21:34
jannaumacos trace https://paste.centos.org/view/8cf7cfb8#asahi-dev · 2026-08-03 21:34
jannaulinux trace https://paste.centos.org/view/1ab3fb67#asahi-dev · 2026-08-03 21:31
jannaulooks similar enough. With linux I see '<== [SHAREDMALLOC](0): [Malloc REQ: size: 0x18000, name: LOG' after the channel table which the macos does not have#asahi-dev · 2026-08-03 21:27
Tachyonray[m]I gave it a go, but don’t understand it very well. built the asahi-wip branch: https://paste.gentoo.zip/4DirfbME#asahi-dev · 2026-08-03 21:21
chaos_princessdoes it do the gpio part correctly?#asahi-dev · 2026-08-03 21:21
jannauwith linux. I havent tried the m1n1 client either#asahi-dev · 2026-08-03 21:21
chaos_princess(i have never tried the m1n1 client with new isp, it probably won't work)#asahi-dev · 2026-08-03 21:20
chaos_princesswith linux too?#asahi-dev · 2026-08-03 21:16
jannaufails to send the first command#asahi-dev · 2026-08-03 21:16
jannauno success so far with https://github.com/jannau/m1n1/commits/isp-t8122/ and https://github.com/jannau/linux/commit/abbb89250863cf05578de993c25e81328ee15741#asahi-dev · 2026-08-03 19:04
jannauthat seems to work#asahi-dev · 2026-08-03 19:00
chaos_princessok, try the 0x24c4454 as ISP_IRQ_DOORBELL and the other as ISP_IRQ_ACK_#asahi-dev · 2026-08-03 19:00
jannau0x24c4454 or 0x24c4460#asahi-dev · 2026-08-03 18:58
chaos_princessare there unknown writes to the general 0x24c44?? area#asahi-dev · 2026-08-03 18:56
jannauthe dorbell is missing#asahi-dev · 2026-08-03 18:55
chaos_princessare there accesses of w_ISP_IRQ_DOORBELL_T6031 and r_ISP_IRQ_INTERRUPT_T8112#asahi-dev · 2026-08-03 18:55
jannaualso for ISP_GPIO_7_T6031#asahi-dev · 2026-08-03 18:55
jannauyes, MMIO: R.4 ISP_GPIO_0_T6031 = 0x8042006 ()#asahi-dev · 2026-08-03 18:53
chaos_princessis there a read of 0x8042006 in ISP_GPIO_0_T6031#asahi-dev · 2026-08-03 18:52
jannau0xf1001#asahi-dev · 2026-08-03 18:50
chaos_princesswhat does ISP_REVISION contain?#asahi-dev · 2026-08-03 18:49
jannauchaos_princess: I'm confused because I only see BUF_T2H, IO_T2H and SHAREDMALLOC messages in the trace#asahi-dev · 2026-08-03 18:19
chaos_princessyes, that should be sufficient#asahi-dev · 2026-08-03 18:19
jannaustart the camera in macos? I did#asahi-dev · 2026-08-03 18:17
chaos_princessbut you need to actually activate it#asahi-dev · 2026-08-03 18:17
chaos_princessjust isp#asahi-dev · 2026-08-03 18:17
jannauchaos_princess: is everything I need in the isp trace or do I need to trace pmgr as well?#asahi-dev · 2026-08-03 18:16
jannaustill running into pmgr _syncDeviceStatusChange timeouts. patched macos increase the factors by ~16