#
…

2021-11-21

00:33nsklaus has quit [Quit: WeeChat 3.3]
00:55nsklaus has joined #asahi-dev
01:07kenzie has quit [Quit: The Lounge - https://thelounge.chat]
02:10kode54what even is the point of disabling linear-sq?
02:10kode54faster random access?
03:06yuyichao_ has joined #asahi-dev
03:08aquijoule_ has joined #asahi-dev
03:14richbridger has quit [Ping timeout: 480 seconds]
03:14psykose has quit [Remote host closed the connection]
03:15psykose has joined #asahi-dev
03:16PhilippvK has joined #asahi-dev
03:20phiologe has quit [Ping timeout: 480 seconds]
03:38kov has quit [Quit: Coyote finally caught me]
04:02palmer_ has joined #asahi-dev
04:02palmer_ has quit [Remote host closed the connection]
04:35phire_ has joined #asahi-dev
04:35phire is now known as Guest6421
04:35phire_ is now known as phire
04:37Guest6421 has quit [Ping timeout: 480 seconds]
06:02Dcow has joined #asahi-dev
06:10Dcow has quit [Ping timeout: 480 seconds]
07:23rkt has joined #asahi-dev
07:47aquijoule_ is now known as richbridger
07:49richbridger has quit []
07:49richbridger has joined #asahi-dev
07:50richbridger has quit [Remote host closed the connection]
07:50richbridger has joined #asahi-dev
07:57Dcow has joined #asahi-dev
08:30richbridger has quit [Remote host closed the connection]
08:45rkt has left #asahi-dev [#asahi-dev]
10:15kov has joined #asahi-dev
10:21aleasto has joined #asahi-dev
10:53gabuscus has quit [Quit: https://quassel-irc.org - Chat comfortably. Anywhere.]
10:53gabuscus has joined #asahi-dev
11:15the_lanetly_052___ has joined #asahi-dev
11:29svenI guess it’s for the t2 where the whole nvmmu doesn’t make much sense
11:30svenAnd there nvme is attached over pcie so you “just” need a few quirks to the normal driver to make it work in that mode
11:43gladiac is now known as Guest6447
11:43gladiac has joined #asahi-dev
11:47gladiac has quit []
11:48gladiac has joined #asahi-dev
11:49Guest6447 has quit [Ping timeout: 480 seconds]
12:17gladiac is now known as Guest6448
12:17gladiac has joined #asahi-dev
12:23Guest6448 has quit [Ping timeout: 480 seconds]
12:47the_lanetly_052___ has quit [Ping timeout: 480 seconds]
12:53aleasto has quit [Remote host closed the connection]
13:20gladiac is now known as Guest6452
13:20gladiac has joined #asahi-dev
13:26Guest6452 has quit [Ping timeout: 480 seconds]
14:19nsklaus has quit [Quit: WeeChat 3.3]
14:23klaus has joined #asahi-dev
15:04marcanmaz: so in apple-pcie, we don't actually do anything with the max link speed
15:05marcanalso there's some extra stuff that has to happen with the designware registers to make it train at the right speed on the first go (ASPM ends up making it retrain anyway, but that's after linux complains about a speed bottleneck)
15:05marcanremind me again why we aren't reusing any code from drivers/pci/controller/dwc? It's the same registers for the parts that are shared...
15:06marcan(that codebase does the speed dance properly that apple-pcie doesn't)
15:12kettenisthe existing dwc code has the 1 contoller 1 port assumption throughly built in
15:13ketteniss/throughly/thoroughly/
15:21Dcow has quit [Quit: My Mac Mini has gone to sleep. ZZZzzz…]
15:57marcanI kind of hate that. It's poking config space registers, but since this is before the PCI bus gets initialized, we can't use the usual accessors...
15:59marcanAlso, the pmgr stuff in t6000/bringup-work is what I'm going to send as v3, modulo I need to do a final check to make sure I didn't do anything silly :)
16:00marcanmain changes are I enable auto-PM now (unconditionally, seems to be fine) and also force on "always-on" things that aren't, but complain. There's also a bit of cleanup/fixes, and also the t6000 compatible in the bindings.
16:00marcanalso I'm outright adding the entire t8103 pmgr tree instead of just a subset
16:03kettenismarcan: some of the link speed related setup is actually done by the tunables applied by m1n1
16:03marcannot this one though
16:04marcanyou get this: [ 0.266469] pci 0000:02:00.0: 2.000 Gb/s available PCIe bandwidth, limited by 2.5 GT/s PCIe x1 link at 0000:00:01.0 (capable of 4.000 Gb/s with 5.0 GT/s PCIe x1 link)
16:04marcanbecause without PORT_LOGIC_SPEED_CHANGE preset (can be done before initial link training), it just comes up at Gen1
16:05marcanthe max speed thing is tangential, it defaults to Gen4 but still comes up at Gen1 without that bit
17:00jkkm has quit [Remote host closed the connection]
17:00cptcobalt has quit [Remote host closed the connection]
17:00nkaretnikov has quit [Remote host closed the connection]
17:00nathanchance has quit [Remote host closed the connection]
17:07jabashque has quit [Remote host closed the connection]
17:11weems_ has quit [Remote host closed the connection]
17:11brinly has quit [Remote host closed the connection]
17:23austriancoder has quit [Ping timeout: 480 seconds]
17:24esden has quit [Ping timeout: 480 seconds]
17:24philpax has quit [Ping timeout: 480 seconds]
17:25tom-w has quit [Ping timeout: 480 seconds]
17:25robher has quit [Ping timeout: 480 seconds]
17:25kendfinger has quit [Ping timeout: 480 seconds]
17:25Chainsaw has quit [Ping timeout: 480 seconds]
17:25Ariadne has quit [Ping timeout: 480 seconds]
17:26sjg1 has quit [Ping timeout: 480 seconds]
17:26arnd_ has quit [Ping timeout: 480 seconds]
17:27daniels has quit [Ping timeout: 480 seconds]
17:35kettenismarcan: since you rebased the PMGR stuff on top of a tree that has more stuff than what has been submitted upstream so far, the bits that actually add the nodes to the device tree won't apply
17:36kettenisnot really a problem since you're probably going to end up applying those yourself
17:36Dcow has joined #asahi-dev
17:46kettenisotherwise it looks good
18:50weems_ has joined #asahi-dev
18:55tardyp has joined #asahi-dev
18:55brinly has joined #asahi-dev
18:57NekomimiScience has joined #asahi-dev
19:01jabashque has joined #asahi-dev
19:04arnd_ has joined #asahi-dev
19:04kendfinger has joined #asahi-dev
19:04esden has joined #asahi-dev
19:04robher has joined #asahi-dev
19:05sjg1 has joined #asahi-dev
19:05daniels has joined #asahi-dev
19:08cptcobalt has joined #asahi-dev
19:09jkkm has joined #asahi-dev
19:09nkaretnikov has joined #asahi-dev
19:10Chainsaw has joined #asahi-dev
19:10nathanchance has joined #asahi-dev
19:10philpax has joined #asahi-dev
19:10Ariadne has joined #asahi-dev
19:11austriancoder has joined #asahi-dev
19:14tom-w has joined #asahi-dev
19:41maor26 has joined #asahi-dev
19:50thunfisch has quit [Remote host closed the connection]
19:50thunfisch has joined #asahi-dev
19:56thunfisch has quit [Remote host closed the connection]
19:57thunfisch has joined #asahi-dev
20:05thunfisch is now known as Guest6468
20:05thunfisch has joined #asahi-dev
20:05Guest6468 has quit [Ping timeout: 480 seconds]
20:16thunfisch has quit [Remote host closed the connection]
20:16thunfisch has joined #asahi-dev
20:26gladiac is now known as Guest6469
20:26gladiac has joined #asahi-dev
20:31Guest6469 has quit [Ping timeout: 480 seconds]
20:42kenzie has joined #asahi-dev
20:50Dcow_ has joined #asahi-dev
20:55Dcow has quit [Ping timeout: 480 seconds]
20:57kenzie has quit [Quit: The Lounge - https://thelounge.chat]
20:58kenzie has joined #asahi-dev
21:08X-Scale has joined #asahi-dev
21:13X-Scale` has quit [Ping timeout: 480 seconds]
21:43kenzie has quit [Quit: The Lounge - https://thelounge.chat]
21:45kenzie has joined #asahi-dev
21:45X-Scale` has joined #asahi-dev
21:50X-Scale has quit [Ping timeout: 480 seconds]
21:52marcankettenis: yeah, I know, I'll rebase that anyway; I'm just using this branch as my kitchen sink to exercise the patches
21:54kenzie has quit [Quit: The Lounge - https://thelounge.chat]
21:58jannaumarcan: does the branch work for you on the mac mini? pmgr seems to cause here a SError in apple_dart_hw_reset of the PCIe dart
21:59marcanit does; that sounds like a power domain not set properly?
21:59marcanactually I'm testing in the hypervisor, haven't tested bare metal
22:00jannautesting baremetal here
22:03jannauseems to be the same problem in the hv
22:03kenzie has joined #asahi-dev
22:04jannaucan you test without a display?
22:04kenzie has quit []
22:05kenzie has joined #asahi-dev
22:07marcanstill no SError without a display
22:07marcancan you paste it somewhere?
22:07marcandmesg too
22:07marcanalso did you compile in the PMGR driver? :)
22:10marcanunrelated (probably) but that dispdfr_be thing is interesting
22:10marcanthat's saying there's a missing parent relationship or so
22:11marcanI don't get that though :/
22:11marcanoh wait, I do lol
22:14marcanhold on, what
22:14marcan[cpu1][0xffffc0001049cf50] PMGR: W.4 0x23b7003a8 (DISPDFR_FE.pstate + 0x00) = 0x300 -> 0xf
22:14marcan[cpu1][0xffffc0001049d034] PMGR: R.4 0x23b7003a8 (DISPDFR_FE.pstate + 0x00) = 0x3ff
22:14marcan[cpu1][0xffffc0001049cf50] PMGR: W.4 0x23b7003a8 (DISPDFR_FE.pstate + 0x00) = 0x3ff -> 0x10000300
22:14marcan.... that doesn't make any sense
22:14marcanwhy is it shutting it down again
22:14marcan... because I'm an idiot
22:19marcanjannau: pushed a fixup, though I don't know if it'll help your pcie problem
22:22marcanjannau: can you run it with `-m hv/trace_pmgr.py` (which I just pushed) and paste the HV side log including the SError dump?
22:22jannausure. it does not fix the dart issue
22:24kettenisdoes your DT have the necessary power-domain properties on the dart nodes?
22:26jannauyes, the dts in the branch has the power-domain properties
22:27marcanjannau: read32(0x23b7003e8)
22:28jannau0x1f0020ff
22:31marcanread32(0x23b7001b8)
22:31jannau0x1f0000ff
22:32j_ey(m1n1 supports irc proxy now, cool!)
22:33marcanI assume read32(0x682008060) SErrors?
22:34marcan^\ out of the hypervisor (without rebooting) and run tools/dump_pmgr.py
22:34marcan(this is a bit of a limbo state inside the HV but not on the host side, but it should work)
22:35marcanonly thing I can think of is we're using different iBoot versions and there's some difference
22:36jannaushould be iboot 12.0.1 on my side
22:38jannauanother potential difference is that iboot doesn't detect my display but we should be good if you tested without display connected. also unclear how it would be related
22:40RealityVoid has quit [Remote host closed the connection]
22:40RealityVoid has joined #asahi-dev
22:41marcanI did test without display
22:41marcanI need to get some sleep, but I'll try updating to 12.0.1 tomorrow
22:41marcannot seeing anything obvious in the pmgr dump :/
22:41marcanyou could try removing pmgr nodes and seeing if it works at some point
22:42marcanthough I think you aren't even at the point where it starts turning things off
22:42marcanso I really don'tknow
22:42j_eymarcan: I thought you woke up early.. lol
22:42marcanyeah... this cold has also screwed up my sleep schedule
23:00yrlf has quit [Quit: The Lounge - https://thelounge.chat]
23:07jannauit boots with all not always on pmgr nodes removed and the power domains commented out
23:12jannauit looks like it's the gpio power domain
23:12yrlf has joined #asahi-dev
23:23maor26 has quit [Ping timeout: 480 seconds]
23:31palmer_ has joined #asahi-dev
23:32palmer_ has quit [Remote host closed the connection]
23:40jannauseems like a dependency problem. with the power-domain pinctrl_ap doesn't get probed
23:49jannauthe pinctrl devices are probed before the pmgr ones
…