#
…

2022-02-18

00:07axboe has quit [Ping timeout: 480 seconds]
00:30axboe has joined #asahi-dev
01:06yuyichao has quit [Ping timeout: 480 seconds]
02:16yuyichao has joined #asahi-dev
02:17refi64 has quit [Remote host closed the connection]
02:18refi64 has joined #asahi-dev
02:20mmarchini has joined #asahi-dev
02:21refi64 has quit []
02:23refi64 has joined #asahi-dev
03:27phiologe has joined #asahi-dev
03:28kov has quit [Quit: Coyote finally caught me]
03:31PhilippvK has quit [Ping timeout: 480 seconds]
04:11axboe has quit [Quit: leaving]
05:19the_lanetly_052___ has joined #asahi-dev
05:40marcanrkjnsn: it does 300-600 IOPS, which is still slow
05:40marcandepending on the machine, apparently?
05:40marcanI got 350 on an M1 Max lol
05:41marcanaxboe: nice :)
07:10marcanaxboe: wait, shouldn't that be F_FULL*F*SYNC on the patch? did you compile test it? :-)
07:10marcanoh, jannau already mentioned it
07:27_jannau_and already fixed and tested in fio master
08:07yuyichao_ has joined #asahi-dev
08:09Major_Biscuit has joined #asahi-dev
08:11yuyichao has quit [Ping timeout: 480 seconds]
08:12MajorBiscuit has joined #asahi-dev
08:18Major_Biscuit has quit [Ping timeout: 480 seconds]
08:47the_lanetly_052___ has quit [Ping timeout: 480 seconds]
10:07kov has joined #asahi-dev
10:56bps has joined #asahi-dev
11:27kloenk has quit [Remote host closed the connection]
12:53roxfan2 has joined #asahi-dev
13:00roxfan has quit [Ping timeout: 480 seconds]
13:04axboe has joined #asahi-dev
13:19axboesorry, jumpy touchpad...
13:42landscape15[m]In order to test out latest power management features, do I need to use smc/work branch?
13:55_jannau_spmi/work
13:59_jannau_sven: is it expected that dart_map_pages() just maps to the end of a l2 table? the imac vram/framebuffer is huge
14:01svenno, that sounds like a bug
14:01sveniirc io-pgtbl writes the number of mapped pages to one of its arguments
14:01svenI probably don't check that
14:01_jannau_it writes the mapped size
14:02svenI'm not sure what io-pgtbl is supposed to do but having that as an output sounds like it's not guaranteed to always map everything
14:02_jannau_so dart_map_pages would be responsible for looping
14:02svenpossibly. i'm just guessing from the API here :D
14:03svencould also be a bug in there. I'd ask robin and the iommu list
14:04_jannau_it has basicly the same api as io-pgtbl. I'll check the decumention tonight.
14:04svenlooks like it's intended though, check e.g. __iommu_map
14:05svenit calls __iommu_map_pages in a loop until it has mapped everything
14:05_jannau_handling that for the resrved regions myself is not a problem but I had the cma alloc of the framebuffer fail as well
14:05_jannau_ok, the cma alloc might be a different problem then
14:06svencma sounds wrong when we have an iommu
14:06svenoh.. wait.. is that thing inside drm that is just called cma but actually just uses the coherent dma api?
14:07_jannau_probably, I haven't looked yet, it's called cma in drm's debug output
14:08svenyeah, i think it's probably cma-in-name-only then
14:09landscape15[m]jannau: thanks. Is there a plan to merge it in smc/work then? Or maybe they will remain separated
14:09svencma is only required when you really need physically contiguous memory. once you have an iommu you can just stich discontiguous pages together for the both the CPU and the device
14:10svenbut for what I believe are historic reasons there's a thing inside drm that calls itself cma_something_something but actually just uses the coherent dma api which makes everything work fine with iommus
14:18alyssaoh i actually know this one
14:19alyssa_jannau_: sven is correct
14:20alyssathe contiguous memory allocator in DRM was designed for physically contiguous allocations, for devices without an IOMMU
14:20alyssahowever it uses the DMA API for everything
14:20alyssaand DMA uses the underlying IOMMU API if the device has an IOMMU
14:20alyssawhich means DRM's CMA works fine if you have an IOMMU ...
14:20alyssa... but instead of allocating physically contiguous buffers, it allocates _device virtual address_ contiguous buffers
14:21alyssaof course, that is exactly what you want for IOMMU-ful display hardware where you specify a single framebuffer VA (as opposed to a list of pages)
14:22alyssawe all agreed that CMA should probably be renamed in DRM, but that it's too much effort and we'd rather bite the noobies because we're mean kernel hackers
14:23svenohhh... does that thing use map_sg while assuming that it'll only get a single output?
14:24svenif you're using a 4K kernel with our lovely 16K IOMMU that's not guaranteed to be true anymore
14:29kovlandscape15[m], afaik spmi/work includes smc/work already (you can search for the top commit messages from smc/work on spmi/work to verify)
14:40landscape15[m]kov: yeah sorry, I found them
14:40yuyichao_ has quit [Ping timeout: 480 seconds]
14:47axboe Glanzmann: I'm already using that, copied from you :)
14:47GlanzmannI see, I I copied it from mps. :-)
14:47axboeit seems the "disable touchpad on typing" is probably what bit me, doesn't work so well
14:47GlanzmannYep, it doesn't.
14:48axboeapart from that, with those settings, I like how the touchpad behaves
14:50mpsaxboe: Glanzmann: `libinput Disable While Typing Enabled (268): 1` is set in my case and it works fine
14:51mpsthough I built latest beta of libinput and rebuilt xf86-input-libinput with it
14:52mpsi.e. libinput-1.19.901
14:52axboemps: promising
14:53mpsbut still didn't got it to detect thumb
14:59yuyichao_ has joined #asahi-dev
15:00jbowen_ has joined #asahi-dev
15:02jbowen_ has quit []
15:05axboe has quit [Ping timeout: 480 seconds]
15:07mpsGlanzmann: `libinput quirks list /dev/input/event1` will show you features of input device
15:09mpsand `libinput quirks list /dev/input/event1` will show quirks from libinput database
15:30kendfinger has quit [Remote host closed the connection]
15:30robher has quit [Read error: No route to host]
15:30kendfinger has joined #asahi-dev
15:31robher has joined #asahi-dev
15:34the_lanetly_052___ has joined #asahi-dev
15:38axboe has joined #asahi-dev
16:17kendfinger has quit [Read error: No route to host]
16:18tardyp has quit [Remote host closed the connection]
16:18mmarchini has quit [Remote host closed the connection]
16:18robher has quit [Remote host closed the connection]
16:18kendfinger has joined #asahi-dev
16:18mmarchini has joined #asahi-dev
16:18robher has joined #asahi-dev
16:18tardyp has joined #asahi-dev
16:28the_lanetly_052___ has quit [Remote host closed the connection]
16:30mmarchini has quit [Read error: No route to host]
16:30mmarchini has joined #asahi-dev
16:30tardyp has quit [Read error: No route to host]
16:31robher has quit [Read error: No route to host]
16:31kendfinger has quit [Read error: No route to host]
16:31kendfinger has joined #asahi-dev
16:31tardyp has joined #asahi-dev
16:32robher has joined #asahi-dev
16:32the_lanetly_052 has joined #asahi-dev
17:01the_lanetly_052__ has joined #asahi-dev
17:07the_lanetly_052 has quit [Ping timeout: 480 seconds]
17:09robher has quit [Read error: No route to host]
17:09tardyp has quit [Read error: No route to host]
17:09robher has joined #asahi-dev
17:10tardyp has joined #asahi-dev
18:15bisko has joined #asahi-dev
18:23MajorBiscuit has quit [Ping timeout: 480 seconds]
18:25axboe has quit [Quit: leaving]
18:49axboe has joined #asahi-dev
19:11jannaufailed framebuffer allocation solved. I somehow managed to allocate the whole display processor's iova space
19:23Gaspare has joined #asahi-dev
19:32the_lanetly_052__ has quit [Remote host closed the connection]
19:41jannaustill not sure how
19:55axboe has quit [Ping timeout: 480 seconds]
20:00jannauadding a size=0 region with iommu_dma_get_resv_regions() seems to do the trick
20:06bisko has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
20:13axboe has joined #asahi-dev
20:14bisko has joined #asahi-dev
20:29tardyp has quit [Read error: No route to host]
20:29tardyp has joined #asahi-dev
20:46sebastinez has joined #asahi-dev
21:00sebastinez has quit [Quit: Textual IRC Client: www.textualapp.com]
21:03axboe_ has joined #asahi-dev
21:05axboe has quit [Ping timeout: 480 seconds]
21:50Gaspare has quit [Quit: Gaspare]
23:44jeffmiwdoes the spmi/work branch supposed to compile ? it seems dts files are broken on my side: t8103.dtsi referring to non existent node
23:45j`eyjeffmiw: markan was only testing on t6000, I guess he'll merge t8301 in the next few days
23:51jeffmiwj'ey; thanks !
23:58j`eyjeffmiw: de rien
…