#
✕ nick:kettenis
#asahi-dev · 2023-03-09 13:51
kettenisah, I missed that happening
#asahi-dev · 2023-03-09 13:50
kettenisor is this just a forward port?
#asahi-dev · 2023-03-09 13:49
kettenisjannau: did you move your development from something based on 6.1 to 6.2?
#asahi-dev · 2023-03-07 10:14
kettenismarcan: I can probably give that a shot later this week
#asahi-dev · 2023-03-04 15:23
kettenisand gpu/drm/bridge/synopsys/dw-mipi-dsi.c
#asahi-dev · 2023-03-04 13:33
kettenisbut I'd say that is graphics processing
#asahi-dev · 2023-03-04 13:32
ketteniswell, media encode/decode is blurring the lines a bit ;)
#asahi-dev · 2023-03-04 13:26
kettenis99% of the transistors on the modern AMD or nVIDIA chips are probably dedicated to the GPU functionality ;)
#asahi-dev · 2023-03-04 13:21
kettenisand SoC vendors often combine a "standard" GPU (such as ARM's mali GPU) with their own display controller block
#asahi-dev · 2023-03-04 13:20
kettenisbut display controler and gpu are (largely) independent blocks
#asahi-dev · 2023-03-04 13:19
kettenistraditionally graphics hardware has combined the display controller and GPU in a single chip
#asahi-dev · 2023-03-04 13:15
kettenisbecause they're separate drivers
#asahi-dev · 2023-03-04 13:08
kettenisyes, I plan to reimplement the GPU driver in C
#asahi-dev · 2023-03-04 13:06
kettenisplease don't
#asahi-dev · 2023-03-03 18:28
kettenisjannau: try adding .flags = DM_FLAG_OS_PREPARE to the U_BOOT_DRIVER struct
#asahi-dev · 2023-03-03 15:29
kettenisi wonder if we need to disable the LTSSM somehow in u-boot before handover
#asahi-dev · 2023-03-03 12:37
kettenislinux
#asahi-dev · 2023-03-03 12:26
kettenisjannau: so the link is reported as up? can you access config space registers?
#asahi-dev · 2023-03-02 14:56
ketteniss/supports/claims to support/
#asahi-dev · 2023-03-02 14:55
kettenisGL9755 supports SD 4.0 UHS-II
#asahi-dev · 2023-03-02 14:52
kettenisin theory at least
#asahi-dev · 2023-03-02 12:30
kettenisah, I didn't think about looking at what m1n1 did
#asahi-dev · 2023-03-02 10:48
kettenisbut maybe accessing the chip without the proper firmware crashes it?
#asahi-dev · 2023-03-02 10:47
kettenisthen the link must be up at that stage
#asahi-dev · 2023-03-02 10:46
kettenisdoas the asmedia chip show up in u-boot at all?
#asahi-dev · 2023-03-02 09:48
kettenismaybe u-boot should bring the link down again; we'd have to figure out how to do that
#asahi-dev · 2023-03-02 09:47
kettenisthe OpenBSD driver short-circuits if it detects that the link is already up
#asahi-dev · 2023-03-02 09:38
ketteniscorrect
#asahi-dev · 2023-03-02 09:35
kettenistunables even
#asahi-dev · 2023-03-02 09:35
kettenistunable survive a PERST# cycle
#asahi-dev · 2023-03-02 09:21
kettenisdoes linux provide some clues why it is unhappy?
#asahi-dev · 2023-03-01 21:35
kettenisjannau: there is no thunderbolt logo on the port and I don't think there is a PCIe root complex available for them
#asahi-dev · 2023-03-01 11:19
kettenisthat is a super-scary patcheset though
#asahi-dev · 2023-03-01 11:13
kettenisif they didn't, you wouldn't be able to boot from a drive connected to thunderbolt
#asahi-dev · 2023-03-01 11:13
kettenissven: most BIOS implementations do that indeed
#asahi-dev · 2023-02-25 10:57
ketteniskevans91: good, was going to say that there is nothing special with these machines, but that they're probably more out-of-order than anything you ran on before
#asahi-dev · 2023-02-24 18:33
kettenisthe lmb ifdefs are a complicated maze that even confused the core u-boot developers recently
#asahi-dev · 2023-02-24 18:33
kettenispovik: and what about LMB_RESERVED_REGIONS?
#asahi-dev · 2023-02-24 14:56
kettenisjannau: isn't the bandwidth of some extra audio channels negligable compared to 4K video? ;)
#asahi-dev · 2023-02-24 11:52
kettenisbut having a directory name in there is a mistake if you ask me
#asahi-dev · 2023-02-24 11:51
kettenisI'm not necessarily against having the firmware name in the binding (and there is plenty of prior art)
#asahi-dev · 2023-02-23 21:07
kettenispovik: I don't think you're supposed to reserve memory ranges that are not included in the ranges advertised by the /memory node
#asahi-dev · 2023-02-23 17:02
ketteniseven that should be fairly straightforward; some bikeshedding over the compatible name and the spien-gpios property perhaps
#asahi-dev · 2023-02-23 13:25
kettenisdo the thunderbolt ports support more than 32 MSIs?
#asahi-dev · 2023-02-23 13:10
kettenisI guess that means 128 vectors instead of 32, but I suspect everything above 32 isn't actually wired up and wouldn't work
#asahi-dev · 2023-02-23 13:07
kettenissven: not sure what the effect of that would be, but probably a good idea to fix it
#asahi-dev · 2023-02-12 16:10
kettenisor are you talking about the order in the yaml schema files?
#asahi-dev · 2023-02-12 16:08
kettenissven: order needs to be from most-specific to least-specific
#asahi-dev · 2023-02-12 13:38
kettenismy current hypothesis is that they moved the per-port registers out of that region
#asahi-dev · 2023-02-12 13:37
kettenisthe number of register regions in the ADT changed again and what we called the "rc" region in the DT is smaller now and not big enough

plus ›