#
✕ nick:lina
#asahi-gpu · 2022-12-03 17:53
linaIf you ever notice weird breakage again (especially clustering-related), try that flag. It should quickly find if a buffer is undersized somewhere.
#asahi-gpu · 2022-12-03 17:41
linajannau: Should be fixed, but I'm really not happy with the magic *2 factor...
#asahi-gpu · 2022-12-03 16:50
linaOK, looks like we have double the tilemaps as usual?
#asahi-gpu · 2022-12-03 16:19
linaActually I think in this particular case you don't even need to set it at boot time, that only matters for the kernel allocators anyway
#asahi-gpu · 2022-12-03 16:17
linaOr maybe I still got the stride somewhere wrong...
#asahi-gpu · 2022-12-03 16:16
linaSo why are those undersized...
#asahi-gpu · 2022-12-03 16:16
linaIt's the cluster tilemaps!
#asahi-gpu · 2022-12-03 16:15
linaYay, my allocator debug stuff is useful! ^^
#asahi-gpu · 2022-12-03 16:15
linaasahi.debug_flags=0xf00002000 at boot time (important) makes it work, and also spam dmesg with buffer overflow errors
#asahi-gpu · 2022-12-03 16:14
linaOkay, now I know what it is
#asahi-gpu · 2022-12-03 16:14
linaBut I still have no idea what it is...
#asahi-gpu · 2022-12-03 16:14
linajannau: I can reproduce it, by the way ^^
#asahi-gpu · 2022-12-03 16:01
linaIt seems to make things 5-10% faster (at least on M1 Ultra), but I have no idea what it does or whether always enabling that has a downside
#asahi-gpu · 2022-12-03 16:00
linaalyssa: If you get a chance, try echo 0x01000000000000 > /sys/module/asahi/parameters/debug_flags
#asahi-gpu · 2022-12-03 15:10
linaIt sounds like this is probably broken ever since I enabled clustering (so 44113d0e62)
#asahi-gpu · 2022-12-03 15:10
linaI don't think the previous commit could've caused it?
#asahi-gpu · 2022-12-03 13:35
lina(I basically just moved that store)
#asahi-gpu · 2022-12-03 13:35
linaYou can also try removing the matching store in auto_grow() which also runs unconditionally
#asahi-gpu · 2022-12-03 13:34
linaIf that's not it then I don't understand
#asahi-gpu · 2022-12-03 13:34
linaCan you try reverting (re-adding) those and seeing if it does anything?
#asahi-gpu · 2022-12-03 13:33
linaThe only part of bedf1a6cc218669 that changed that doesn't only run with TVB resizes is the 4 lines of code around @@ -360,10 +386,6 @@ impl Buffer::ver {
#asahi-gpu · 2022-12-03 13:32
linaBut larger sizes always broke
#asahi-gpu · 2022-12-03 13:32
linaThis stuff sometimes isn't reproducible... yesterday's clustering stuff broke glmark2-refract at 1080p initially but not later
#asahi-gpu · 2022-12-03 13:27
linajannau: That sounds like a null pointer, but then it can't bisect to that commit because that commit's code is never running...
#asahi-gpu · 2022-12-03 11:21
linajannau: The grep cut off the fault info, can you get me that? ^^
#asahi-gpu · 2022-12-03 11:20
linaOh wait, those are other apps aren't they
#asahi-gpu · 2022-12-03 11:16
linaUnless it creates multiple GL contexts by opening the device more than once somehow?
#asahi-gpu · 2022-12-03 11:16
linajannau: Okay, that's weird. I don't think supertuxkart is more than one process? So that should be one TVB, so it not growing linearly is wrong...
#asahi-gpu · 2022-12-03 11:14
linajannau: If the growth is what breaks then setting the initial TVB size higher would avoid that
#asahi-gpu · 2022-12-03 10:39
linaI know it's missing tile size but I think that mesa branch doesn't have varying tile size yet I think?
#asahi-gpu · 2022-12-03 10:38
linaThis might just be that the TVB is too small to work properly and the logic for the initial size is wrong, or missing some factors
#asahi-gpu · 2022-12-03 10:37
linaYou can also add debug_flags=0x2000 to get messages when TVB growth happens, which might be useful
#asahi-gpu · 2022-12-03 10:37
linajannau: setting initial_tvb_size large enough should also avoid the problem, right?
#asahi-gpu · 2022-12-03 10:36
linaI think I'll have to go back to tracing macOS VM accesses around that part of the code, there's probably something I'm doing wrong...
#asahi-gpu · 2022-12-03 10:36
linaI did that late last night so it didn't get much testing...
#asahi-gpu · 2022-12-03 10:36
linaInteresting!
#asahi-gpu · 2022-12-03 07:40
linasteven: I replied on #asahi-stream but we should probably talk about texture issues here since alyssa is here ^^
#asahi-gpu · 2022-12-03 07:40
linajannau: Is this with yesterday's clustering fixes, or only the incomplete ones from wednesday?
#asahi-gpu · 2022-12-02 06:00
linaalyssa: Thank you!!! ^^
#asahi-gpu · 2022-11-29 17:31
linaThank you!! ^^
#asahi-gpu · 2022-11-29 16:51
linaIt's finally up!! ✨✨ https://asahilinux.org/2022/11/tales-of-the-m1-gpu/
#asahi-gpu · 2022-11-27 16:06
linaAlong with some cleanup patches and another thing which is probably unrelated but sounded like a good idea?
#asahi-gpu · 2022-11-27 16:06
linaPushed to my usual branch ^^
#asahi-gpu · 2022-11-27 16:05
linaThat took a while to work out... I didn't understand why it was ignoring my requests...
#asahi-gpu · 2022-11-27 16:05
linaSo we not only need to force a set on poweron, but also turn it off on poweroff, otherwise it just ignores the set as a no-op
#asahi-gpu · 2022-11-27 16:05
linaI think it's a firmware bug... it turns out it desyncs its idea of backlight brightness on poweroff/on
#asahi-gpu · 2022-11-27 16:04
linajannau: I figured out the backlight restore problem on M2 Air (and probably others) ^^
#asahi-gpu · 2022-11-27 15:52
linamarcan: Thank you ^^
#asahi-gpu · 2022-11-27 14:39
linaNow I see why alyssa keeps her WIP branches private... she sends me a branch to implement occlusion queries just so I can add the UAPI and random people start running it T_T
#asahi-gpu · 2022-11-27 14:38
linaReporting issues in random branches is never useful...

plus ›