#asahi-gpu · 2022-10-29 14:17
jannaubut yes, I sure they have more demand for "broken" chips than their yield produces#asahi-gpu · 2022-10-29 14:14
linajannau: Sure! More data is great ^^#asahi-gpu · 2022-10-29 14:14
linaSo then both would have to line up, which is a 1/16 chance#asahi-gpu · 2022-10-29 14:14
jannauI can do a dump of the m1 max (24 core gpu) if that helps#asahi-gpu · 2022-10-29 14:14
linaThough they get to pick both dies, and presumably they would pick two bad ones if they had enough of them#asahi-gpu · 2022-10-29 14:13
linaTrue...#asahi-gpu · 2022-10-29 14:13
jannauat this point == since they started shipping. if I have 1 broken core there is a 1/4 chance that it would still look regular#asahi-gpu · 2022-10-29 14:03
linaphire: If you get a chance, can you do the same trace? initdata with that pwr script + hv/trace_all_more.py#asahi-gpu · 2022-10-29 14:02
lina... 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#asahi-gpu · 2022-10-29 14:02
linajannau: Heh, looking at the SGX registers it reads at the beginning, yours has what looks like a mask of active cores...#asahi-gpu · 2022-10-29 12:53
jannauupdated#asahi-gpu · 2022-10-29 12:53
jannaunot done, I forgot to update the command#asahi-gpu · 2022-10-29 12:45
jannaudone#asahi-gpu · 2022-10-29 12:41
linaI just realized trace_all was untracing pmgr/error_handler which definitely have reg ranges related to fuses in them...#asahi-gpu · 2022-10-29 12:40
linajannau: Actually actually, use hv/trace_all_more.py which I just pushed...#asahi-gpu · 2022-10-29 12:34
linaIt'll be a huge log but it might have something useful...#asahi-gpu · 2022-10-29 12:34
linajannau: Actually, use trace_all (on top of trace_agx) too if you can#asahi-gpu · 2022-10-29 12:34
jannaulina: one moment, I have reboot 2 machines#asahi-gpu · 2022-10-29 12:27
linaI 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#asahi-gpu · 2022-10-29 12:27
linajannau: 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?#asahi-gpu · 2022-10-29 12:21
phireYeah, suspected as much #asahi-gpu · 2022-10-29 12:00
linaAh yeah, my weirdo old mesa for this still had the "force linear" patch in it...#asahi-gpu · 2022-10-29 10:14
phireJust needed a little unswizling: https://i.imgur.com/FTLx9qa.png#asahi-gpu · 2022-10-29 10:03
phirewell, it's close to being a triangle: https://i.imgur.com/eBs0AWQ.png#asahi-gpu · 2022-10-29 08:35
phireremoving the subprocess out to gcc finally got it working#asahi-gpu · 2022-10-29 08:34
phirefixed it. It was a bad interaction between drm_shim and python's subprocess#asahi-gpu · 2022-10-29 07:17
phiredrm_shim is implemented by hooking a bunch of libc functions (open/close/dup/mmap), and it seems to be screwing with python #asahi-gpu · 2022-10-29 07:14
phireI think this might just be cursed#asahi-gpu · 2022-10-29 06:30
phireum... the drm shim init function gets called repeatedly until init completes #asahi-gpu · 2022-10-29 06:00
linaI don't think I broke anything but the render command though?#asahi-gpu · 2022-10-29 06:00
linaOh also, the Python side does not implement the up to date UABI yet... I'm running on some old mesa branch still for that#asahi-gpu · 2022-10-29 05:38
phirepython is corrupting it?#asahi-gpu · 2022-10-29 05:09
phirethe shim_fd appears to be corrupted#asahi-gpu · 2022-10-29 05:08
phirecurrently trying to work out why it's deadlocking on the first DRM_IOCTL_ASAHI_CREATE_BO#asahi-gpu · 2022-10-29 05:04
linaMake sure to check render.py, I think I put in some dumb condition for what to blit to the FB#asahi-gpu · 2022-10-29 04:36
phireah, eglGetPlatformDisplay(EGL_PLATFORM_SURFACELESS_MESA)#asahi-gpu · 2022-10-29 04:18
phireoh, mesa doesn't consider asahi to be a valid driver for the current display?#asahi-gpu · 2022-10-29 03:35
phireyeah, just trying to get surfaceless egl working#asahi-gpu · 2022-10-29 03:22
linaThere's no mechanism for actually doing WSI right now#asahi-gpu · 2022-10-29 03:22
linaYeah, it can't do GLX, you need to use surfaceless EGL and it'll just blit some surfaces into the framebuffer...#asahi-gpu · 2022-10-29 01:31
phireoh, it probally doesn't work with glx and I'll have to use egl instead#asahi-gpu · 2022-10-29 01:30
phirenow, how do I get this m1n1 shim driver working?#asahi-gpu · 2022-10-29 00:12
phireotherwise all the initdata looks identical to me on my machine (ignoring rel_max_powers and rel_boost_powers)#asahi-gpu · 2022-10-29 00:08
phirealso max_freq_mhz is missing a divide by 1000000#asahi-gpu · 2022-10-28 23:30
phirelina: you have two pwr_filter_a_neg's#asahi-gpu · 2022-10-28 23:30
phirehmmm... it would be nice if construct warned you when you named two fields the same thing#asahi-gpu · 2022-10-28 17:55
alyssaOr nothing. Completely untested :p#asahi-gpu · 2022-10-28 17:55
alyssaIDK if this will fix glamor but it'll probably fix something!#asahi-gpu · 2022-10-28 17:55
alyssahttps://gitlab.freedesktop.org/alyssa/mesa/-/commit/5af2421d20d9f39c1e60db8ea277193267c1b91e#asahi-gpu · 2022-10-28 14:06
alyssa_jannau_: Ooh, got it, thank you