#
…

2022-10-29

00:08phirealso max_freq_mhz is missing a divide by 1000000
00:08Dcow_ has joined #asahi-gpu
00:12phireotherwise all the initdata looks identical to me on my machine (ignoring rel_max_powers and rel_boost_powers)
00:13Dcow_ has quit [Remote host closed the connection]
00:15Dcow has quit [Ping timeout: 480 seconds]
00:21millenia_ has joined #asahi-gpu
00:24millenia_ has quit [Read error: Connection reset by peer]
00:29Dcow has joined #asahi-gpu
00:37Dcow has quit [Ping timeout: 480 seconds]
00:40millenialhacker has joined #asahi-gpu
00:48millenialhacker has quit [Ping timeout: 480 seconds]
01:00DragoonAethis has quit [Quit: hej-hej!]
01:00DragoonAethis has joined #asahi-gpu
01:02Dcow has joined #asahi-gpu
01:10Dcow has quit [Ping timeout: 480 seconds]
01:20Emantor has quit [Quit: ZNC - http://znc.in]
01:20Emantor has joined #asahi-gpu
01:30phirenow, how do I get this m1n1 shim driver working?
01:31phireoh, it probally doesn't work with glx and I'll have to use egl instead
01:39Dcow has joined #asahi-gpu
01:45amarioguy has joined #asahi-gpu
01:47Dcow has quit [Ping timeout: 480 seconds]
01:48Etrien has joined #asahi-gpu
01:55Etrien__ has quit [Ping timeout: 480 seconds]
02:12Etrien__ has joined #asahi-gpu
02:15Dcow has joined #asahi-gpu
02:17Etrien has quit [Ping timeout: 480 seconds]
02:21VinDuv has quit [Ping timeout: 480 seconds]
02:23Dcow has quit [Ping timeout: 480 seconds]
02:34SSJ_GZ has quit [Ping timeout: 480 seconds]
02:34SSJ_GZ has joined #asahi-gpu
02:38Core9066 has joined #asahi-gpu
02:51Dcow has joined #asahi-gpu
03:00Dcow has quit [Ping timeout: 480 seconds]
03:22linaYeah, it can't do GLX, you need to use surfaceless EGL and it'll just blit some surfaces into the framebuffer...
03:22linaThere's no mechanism for actually doing WSI right now
03:27Dcow has joined #asahi-gpu
03:35Dcow has quit [Ping timeout: 480 seconds]
03:35phireyeah, just trying to get surfaceless egl working
03:53chadmed has joined #asahi-gpu
03:58ella-0 has joined #asahi-gpu
04:01ella-0_ has quit [Read error: Connection reset by peer]
04:18phireoh, mesa doesn't consider asahi to be a valid driver for the current display?
04:36phireah, eglGetPlatformDisplay(EGL_PLATFORM_SURFACELESS_MESA)
04:41millenia_ has joined #asahi-gpu
04:49millenia_ has quit [Ping timeout: 480 seconds]
05:04linaMake sure to check render.py, I think I put in some dumb condition for what to blit to the FB
05:08phirecurrently trying to work out why it's deadlocking on the first DRM_IOCTL_ASAHI_CREATE_BO
05:09phirethe shim_fd appears to be corrupted
05:10chadmed has quit [Quit: Konversation terminated!]
05:10Dcow has joined #asahi-gpu
05:10chadmed has joined #asahi-gpu
05:13chadmed has quit []
05:18Dcow has quit [Ping timeout: 480 seconds]
05:38phirepython is corrupting it?
05:46Dcow has joined #asahi-gpu
05:55Dcow has quit [Ping timeout: 480 seconds]
06:00linaOh also, the Python side does not implement the up to date UABI yet... I'm running on some old mesa branch still for that
06:00linaI don't think I broke anything but the render command though?
06:30phireum... the drm shim init function gets called repeatedly until init completes
07:00millenia_ has joined #asahi-gpu
07:09millenia_ has quit [Ping timeout: 480 seconds]
07:14phireI think this might just be cursed
07:17phiredrm_shim is implemented by hooking a bunch of libc functions (open/close/dup/mmap), and it seems to be screwing with python
08:01cr1901 has quit [Read error: Connection reset by peer]
08:02cr1901 has joined #asahi-gpu
08:13chadmed has joined #asahi-gpu
08:34phirefixed it. It was a bad interaction between drm_shim and python's subprocess
08:35phireremoving the subprocess out to gcc finally got it working
08:49Etrien has joined #asahi-gpu
08:52VinDuv has joined #asahi-gpu
08:55Etrien__ has quit [Ping timeout: 480 seconds]
09:00VinDuv has quit [Quit: ZNC 1.8.2+deb2+b1 - https://znc.in]
09:00VinDuv has joined #asahi-gpu
09:09Etrien__ has joined #asahi-gpu
09:13Etrien has quit [Ping timeout: 480 seconds]
09:13chadmed has quit [Quit: Konversation terminated!]
09:16chadmed has joined #asahi-gpu
09:16chadmed has quit []
09:51millenialhacker has joined #asahi-gpu
09:54millenialhacker has quit [Read error: No route to host]
09:54millenialhacker has joined #asahi-gpu
09:58millenialhacker has quit [Read error: Connection reset by peer]
09:58millenialhacker has joined #asahi-gpu
09:58millenialhacker has quit [Remote host closed the connection]
09:59millenialhacker has joined #asahi-gpu
09:59millenialhacker has quit [Remote host closed the connection]
10:00millenia_ has joined #asahi-gpu
10:01millenia_ has quit [Remote host closed the connection]
10:01millenia_ has joined #asahi-gpu
10:01millenia_ has quit [Remote host closed the connection]
10:02millenia_ has joined #asahi-gpu
10:03millenia_ has quit [Remote host closed the connection]
10:03millenia_ has joined #asahi-gpu
10:03millenia_ has quit [Remote host closed the connection]
10:03phirewell, it's close to being a triangle: https://i.imgur.com/eBs0AWQ.png
10:14phireJust needed a little unswizling: https://i.imgur.com/FTLx9qa.png
10:16millenia_ has joined #asahi-gpu
11:11chadmed has joined #asahi-gpu
11:42chadmed has quit [Quit: Konversation terminated!]
11:47uur has joined #asahi-gpu
11:57uur has quit [Remote host closed the connection]
12:00linaAh yeah, my weirdo old mesa for this still had the "force linear" patch in it...
12:02uur has joined #asahi-gpu
12:07Dementor has quit [Remote host closed the connection]
12:08Dementor has joined #asahi-gpu
12:11uur has quit [Remote host closed the connection]
12:12uur has joined #asahi-gpu
12:14chadmed has joined #asahi-gpu
12:21phireYeah, suspected as much
12:22chadmed has quit [Quit: Konversation terminated!]
12:27linajannau: Can you take another dump of the M1 Ultra with that agx trace script, but save the HV log (-l log.txt) and send me that?
12:27linaI want to figure out if any of the few SGX registers read at startup have like a variant/model ID, if I can convince my macOS that I have a lower-core model I can find out if that's really what matters or not
12:28chadmed has joined #asahi-gpu
12:32Dcow has joined #asahi-gpu
12:34jannaulina: one moment, I have reboot 2 machines
12:34uur has quit [Remote host closed the connection]
12:34linajannau: Actually, use trace_all (on top of trace_agx) too if you can
12:34linaIt'll be a huge log but it might have something useful...
12:36uur has joined #asahi-gpu
12:38chadmed has quit [Quit: Konversation terminated!]
12:40linajannau: Actually actually, use hv/trace_all_more.py which I just pushed...
12:41linaI just realized trace_all was untracing pmgr/error_handler which definitely have reg ranges related to fuses in them...
12:53jannaunot done, I forgot to update the command
13:33uur has quit [Remote host closed the connection]
14:02linajannau: Heh, looking at the SGX registers it reads at the beginning, yours has what looks like a mask of active cores...
14:02lina... and yours are completely regular, so I'm pretty sure Apple is turning cores off for marketing reasons and not yield at this point, at least for you
14:03linaphire: If you get a chance, can you do the same trace? initdata with that pwr script + hv/trace_all_more.py
14:13jannauat this point == since they started shipping. if I have 1 broken core there is a 1/4 chance that it would still look regular
14:13linaTrue...
14:14linaThough they get to pick both dies, and presumably they would pick two bad ones if they had enough of them
14:14jannauI can do a dump of the m1 max (24 core gpu) if that helps
14:14linaSo then both would have to line up, which is a 1/16 chance
14:14linajannau: Sure! More data is great ^^
14:17jannaubut yes, I sure they have more demand for "broken" chips than their yield produces
14:20wyes has joined #asahi-gpu
14:21linaOK, so at least some things are adding up...
14:22linaThe core leakage/static power data is additive. If I subtract f=0 data from f=1000 data, the total power becomes constant
14:22linaSo that means there is no per-core shenanigans for dynamic power, and the static and dynamic power are just added together
14:23linaAnd the f=0 data is proportional (to within the available significant digits) to the leakage table values.
14:30jannaulina: at the same place with the obvious s/8/4/ fix
14:31jannauphire: change line 76 to "for j in range(4):" for the m1 max
14:58uur has joined #asahi-gpu
15:02uur has quit [Remote host closed the connection]
15:08uur has joined #asahi-gpu
15:26millenia_ has quit []
15:51Retr0id8 has joined #asahi-gpu
15:51Retr0id has quit [Read error: Connection reset by peer]
15:51Retr0id8 is now known as Retr0id
16:05uur has quit [Remote host closed the connection]
16:09uur has joined #asahi-gpu
16:22uur has quit [Remote host closed the connection]
16:24wyes has quit [Quit: Leaving]
16:50uur has joined #asahi-gpu
16:52uur has quit [Remote host closed the connection]
16:56uur has joined #asahi-gpu
16:57uur has quit [Remote host closed the connection]
17:07chengsun_ has quit [Ping timeout: 480 seconds]
18:08alyssa has left #asahi-gpu [#asahi-gpu]
18:15uur has joined #asahi-gpu
18:15ChaosPrincess has quit [Quit: WeeChat 3.6]
18:16ChaosPrincess has joined #asahi-gpu
18:28chengsun has joined #asahi-gpu
18:30pthariensflame has joined #asahi-gpu
18:44uur has quit [Remote host closed the connection]
18:46chengsun has quit [Ping timeout: 480 seconds]
18:50pthariensflame has quit [Quit: Textual IRC Client: www.textualapp.com]
19:05yuyichao has quit [Quit: Konversation terminated!]
19:06yuyichao has joined #asahi-gpu
20:09c10l5 has joined #asahi-gpu
20:09c10l has quit [Ping timeout: 480 seconds]
20:31phireLina, just let me wake up
20:43pthariensflame has joined #asahi-gpu
20:43pthariensflame has quit []
21:46SSJ_GZ has quit [Read error: No route to host]
22:05phiretrace_all_more times out and panics: panic(cpu 2 caller 0xfffffe0014d15fe4): cannot find IOAESAccelerator: timeout = 120 seconds
22:08jannauin combination with trace_agx_pwr? the latter reboots when it is done
22:12phirerunning both tracers just made it panic sooner (diffrent panic?), didn't get as far as agx
22:15jannaumacos panics?
22:16phirethis time: "assertion failed /AppleInternal/Library/BuildRoots/b19089b6-9308-11ec-ba99-4e67803e3fd9/Library/Caches/com.apple.xbs/Sources/AppleConvergedIPC/Core/ACIPCPort.cpp:105" @ACIPCPort.cpp:105
22:17jannau'<@lina> jannau: mkdir initdata; python tools/run_guest.py -m hv/trace_agx_pwr.py 12.3/kernelcache.release.mac13j.macho -- "debug=0x14e serial=3 -enable-kprintf-spam wdt=-1"'
22:19Etrien has joined #asahi-gpu
22:23Etrien__ has quit [Ping timeout: 480 seconds]
22:39Etrien has quit [Read error: Connection reset by peer]
22:40Etrien has joined #asahi-gpu
22:51Etrien__ has joined #asahi-gpu
22:53Etrien has quit [Ping timeout: 480 seconds]
…