#
…

2021-08-17

00:42skipwich has joined #asahi
01:49riker77_ has joined #asahi
01:54riker77 has quit [Ping timeout: 480 seconds]
01:54riker77_ is now known as riker77
02:35phiologe has joined #asahi
02:39PhilippvK has quit [Ping timeout: 480 seconds]
02:57yuyichao has quit [Ping timeout: 480 seconds]
03:02skipwich has quit [Quit: DISCONNECT]
03:02skipwich has joined #asahi
03:56marvin24 has joined #asahi
03:59marvin24_ has quit [Ping timeout: 480 seconds]
04:01yuyichao has joined #asahi
05:52tasild[m] has joined #asahi
08:16Andalu30 has joined #asahi
08:20bps has joined #asahi
08:29jeffmiw has joined #asahi
08:33balrog has quit [Quit: Bye]
08:34jkkm has quit [Read error: Connection reset by peer]
08:34balrog has joined #asahi
08:34jkkm has joined #asahi
09:04rcombs has quit [Quit: ZNC - http://znc.in/ZNC]
09:04hspak6 has joined #asahi
09:04TheJollyRoger has quit [Remote host closed the connection]
09:05austriancoder_ has joined #asahi
09:06hspak has quit [Read error: Connection reset by peer]
09:06hspak6 is now known as hspak
09:06rcombs has joined #asahi
09:09austriancoder has quit [Ping timeout: 480 seconds]
09:11austriancoder_ has quit []
09:11austriancoder_ has joined #asahi
09:11austriancoder_ is now known as austriancoder
09:12austriancoder has quit []
09:12austriancoder has joined #asahi
09:31j_eyjannau: do you understand the sizes of the gpio/pinctrl controllers? they seem way bigger than anything that looks like it is accessed
09:33sventhat's pretty common in the ADT fwiw
09:33sventhere's also one DART that has a size of 0x20000 or so but is really just the same 16k repeated over and over
09:34j_eyas in accessing reg + 0 is the same as reg + 16k?
09:36j_eythis has size 0x100000, that's 1mb, waste of ram!!
09:39svenwhat probably happens internally is that the address decoding/muxing logic uses the upper $whatever bits and some of the lower address lines either just aren't connected or are ignored by the gpio/pinctrl hw block
09:39j_eysven: is there any reason to use that size though?
09:39j_eyor can we just map 1 page for it
09:40svenyou can just map a single page
09:40j_ey(also I assume if I put 0x1000 there, it's rounded up to PAGE_SIZ)
09:40svenit'll certainly map the size rounded up to the next PAGE_SIZE, but accessing anything past 0x1000 is probably undefined then :)
09:45pipcet[m]so our theory is the ADT contains the entire decoded range, not the actually accessible range? That makes sense from a HW POV, I guess...
10:05Andalu30 has quit [Quit: Konversation terminated!]
10:40marcanit's not consistent
10:40marcanthe ADT is written by humans
10:40marcanwe should try to be more consistent than them and declare only the real ranges
10:40marcanno mirrors
10:40j_eymarcan: is this wrong: python3.9 proxyclient/tools/run_guest.py -m proxyclient/m1n1/trace/gpio.py m1n1-payload.macho ?
10:41j_eyI'm getting 'ImportError: attempted relative import with no known parent package'
10:41marcansounds like it should work
10:42j_ey File "proxyclient/m1n1/trace/gpio.py", line 3, in <module>
10:42j_ey from ..hv import TraceMode
10:42marcanhv/trace_gpio.py, not m1n1/trace/gpio.py
10:42marcanthe hv scripts are all in hv/
10:44j_eythanks, that works!
11:02cgv[m]has anyone played with perf counters in M1? I'm trying to use m1n1 to run microbenchmarks (setting up the counters and env from python), but I'm having some issues. e.g. evt=0x1 (retired uops) only seem to work in PMC1, I'm not able to set it up in PMC2, but evt=0x2 (cycle count) works well with any PMC
11:14dougallheh, yeah, i have
11:16cgv[m]dougall: i saw your repo :) did you experience anything like the above? just to be sure whether it's expected or if i'm doing something stupid
11:17cgv[m]probably the later, since i don't seem to be able to observe many events :(
11:19dougallso, I'm a little confused as I didn't think PMC1 was configurable? but yes - there are a number of events that only work on a subset of PMCs, e.g. retired uops (1) I can only use on PMC7
11:20cgv[m]no, it's not, but i was trying to count the same event with another counter; ok, that make sense, i will check what-works-where, thx
11:25dougallyeah - I've been working on that a bit - the difficulty is finding ways to trigger the events so I can see which PMCs support them, but I had some luck. Let me know if there's anything specific you're wondering about, or I'll try to put together some updated notes soon. For the officially documented ones /usr/share/kpep/a14.plist has the info
11:27dougall(no idea if you're a mac user, but i use `plutil -convert json -o - /usr/share/kpep/a14.plist` to get it as json, and the "counters_mask" has bit (1<<n) set for PMCn iirc)
11:29cgv[m]thanks, I'll start from that and see if I manage to count more events
11:56cgv[m]event and PMCs where I observe changes (i'm not yet exercising all of them)
12:12dougallnice! can i ask what you're working on/towards? just exploring? (there is a fun edge-case to watch out for with inst_int_store (0x96) and INST_LDST (0x9b) counting fused AMX instructions differently on PMC7, which counts instructions, vs PMC5/6, which count fused ops)
12:58cgv[m]just exploring, i want to get a nice low-noise environment to quickly run snippets in different contexts, after that i guess some idea will popup
13:00cgv[m]at some point i was tempted to port nanobench to arm, maybe m1n1 could be a good env for that?
13:01pipcet[m]cgv: rr support for M1 would be really nice, or at least evaluating whether it's possible (https://github.com/rr-debugger/rr)
13:13dougallsounds gret :) yeah - last i heard m1n1 doesn't do frequency scaling, so it'd be convenient in that respect - I haven't got around to setting it up myself, but I've been meaning to do that and try to reproduce something like Sushi Roll (https://gamozolabs.github.io/metrology/2019/08/19/sushi_roll.html) at some point
13:18cgv[m]that would be super cool
13:19milek7rr doesn't work on arm
13:21pipcet[m]there's some aarch64 code in the tree now, only recalled that...
13:26milek7ah, I didn't know it
14:40yuyichao has quit [Ping timeout: 480 seconds]
15:35skipwich has quit [Ping timeout: 480 seconds]
16:19bps has quit [Ping timeout: 480 seconds]
16:48bps has joined #asahi
17:04Andalu30 has joined #asahi
18:28jbowen has quit [Ping timeout: 480 seconds]
18:35Andalu30 has quit [Remote host closed the connection]
18:44thunfisch has quit [Quit: frrrp!]
18:45jbowen has joined #asahi
18:45thunfisch has joined #asahi
19:01artemist has quit [Quit: artemist]
19:03artemist has joined #asahi
19:17jbowen has quit [Ping timeout: 480 seconds]
19:21jbowen has joined #asahi
19:24Ziemas has quit [Server closed connection]
19:24Ziemas has joined #asahi
19:29segher has quit [Server closed connection]
19:36svenuh... i just realized something weird about the nvme iommu. so the normal io queue is created using an admin io queue command with the dma specified in one of the fields. that dma address is whitelisted using the nvmmu. so far so good
19:37svenbut then the nvmmu entry is deleted but ANS can still access the io queue dma buffer just fine. so where exactly is it still whitelisted?!
19:37alyssadougall: there's a cpufreq driving floating around, dunno if it's functional
19:38alyssasven: 🤔
20:03svenand weird, the shutdown process looks like a normal nvme shutdown process. so why doesn't it complain about an uncleanReboot only when macOS does it? :.
20:08alyssasven: Any side channel interactions (e.g. with SMC)?
20:10vup has quit [Server closed connection]
20:11svennothing on the ANS mailbox, but i should also trace the SMC mailbox.
20:12roxfandifferential debugging
20:13sveni'm still confuse about the io queue dma buffer :/
20:14sveni'd claim that what we call NVMMU actually isn't an IOMMU-like block, but it's handled inside PPL which is usually reserve for stuff like iommus
20:14roxfanaren't we all
20:14sven*reserved
20:18nyuhu has quit [Server closed connection]
20:18nyuhu has joined #asahi
20:20alyssasven: where does the name NVMMU come from?
20:22roxfanfrom nvme?
20:22kettenis_ has quit [Ping timeout: 480 seconds]
20:25kettenis has joined #asahi
20:34Andalu30 has joined #asahi
21:11vup has joined #asahi
21:25Shiz has quit [Server closed connection]
21:26Shiz has joined #asahi
21:55Andalu30 has quit [Remote host closed the connection]
21:57citizen1[m] has joined #asahi
23:20eta has quit [Server closed connection]
…