#
#asahi-dev · 2026-08-12 09:39
chadmedim pretty sure the kernel branch i linked should Just Work with ycc surfaces too but i've not got an easy way to test that
#asahi-dev · 2026-08-12 09:37
chadmedwell at least interchange works on all supported iosurfaces so im happy with that
#asahi-dev · 2026-08-12 09:37
chadmedi just saw that too :D
#asahi-dev · 2026-08-12 09:37
jannauI suspect this might be somewhat intentionally avoided in released kwin due to multi-gpu issues. might work kwin master after https://zamundaaa.github.io/wayland/2026/07/31/fixing-multi-gpu.html
#asahi-dev · 2026-08-12 09:28
chadmediirc mutter does
#asahi-dev · 2026-08-12 09:28
chadmedkwin doesnt expose a linux-dmabuf tranche with dcp as the target
#asahi-dev · 2026-08-12 09:06
chadmedive updated mesa to do the right thing
#asahi-dev · 2026-08-12 08:47
chadmedok turns out it was neither of those things, it was allowing all scanout buffers to use interchange. i guess we have to ensure that the buffer has a valid pixel format before we allow it or something
#asahi-dev · 2026-08-12 08:37
chadmed(cc oliver)
#asahi-dev · 2026-08-12 08:37
chadmedif mutter also explodes horribly then its probably the former
#asahi-dev · 2026-08-12 08:36
chadmedi suspect theres either something wrong with doing interchange on >2 surfaces, or kwin does weird things with the framebuffers that dcp does not like
#asahi-dev · 2026-08-12 08:36
chadmedcould someone with an up to date gnome system please test https://github.com/chadmed/linux/tree/dcp/interchange along with https://gitlab.freedesktop.org/chadmed/mesa/-/tree/asahi/interchange?ref_type=heads
#asahi-dev · 2026-08-12 08:27
Rafaychaos_princess: sorry, I meant TCP.
#asahi-dev · 2026-08-12 08:26
Rafaychaos_princess: UDP
#asahi-dev · 2026-08-12 08:24
leiofwiw, the GRO message happens for end0 for 10G Aquantia AQtion AQC113 (Antigua 10G) as well, been idly wondering about that, but no speed issues there at 400 Mbps levels at least
#asahi-dev · 2026-08-12 08:24
chaos_princessare you using udp
#asahi-dev · 2026-08-12 08:23
Rafayiperf output: https://pastebin.com/mVR30GkX
#asahi-dev · 2026-08-12 08:20
Rafaysven: The above line is also in my dmesg quite late after the thunderbolt messages.
#asahi-dev · 2026-08-12 08:19
Rafay[ 2862.192205] TCP: thunderbolt0: Driver has suspect GRO implementation, TCP performance may be compromised.
#asahi-dev · 2026-08-12 08:17
RafayI have no other interface, besides loopback and disconnected ethernet.
#asahi-dev · 2026-08-12 08:11
sveneveryone has had issue getting it to actually use the thunderbolt0 interface so far
#asahi-dev · 2026-08-12 08:11
svenno idea why perf is so low, are you sure it’s going over tbt?
#asahi-dev · 2026-08-12 08:11
sventhat’s just the same thing essentially
#asahi-dev · 2026-08-12 08:04
Rafayhigh_speed_lane.c:289: Gen2/3 link error. lane=1, error=82
#asahi-dev · 2026-08-12 08:04
RafayAlso aside from error 88: [ 63.138266] thunderbolt-apple-acio 501ac0000.nhi: RTKit: syslog message: high_speed_lane.c:289: Gen2/3 link error. lane=0, error=84
#asahi-dev · 2026-08-12 08:00
RafayThe iperf3 speeds are very low and it breaks by the end of the test. I don't know why it's breaking at the end but speeds are around 10.5 Mbits/s which is quite low.
#asahi-dev · 2026-08-12 04:44
sventhe only error i see is "[ 65.440040] thunderbolt-apple-acio 501ac0000.nhi: RTKit: syslog message: high_speed_lane.c:289: Gen2/3 link error. lane=1, error=88" and i got that locally as well which made me suspect that something with lane bonding goes wrong yesterday
#asahi-dev · 2026-08-12 04:37
svenrafay: what makes you think it breaks anything? that looks good actually
#asahi-dev · 2026-08-12 03:53
chadmedall i want is an interchange buffer smdh
#asahi-dev · 2026-08-12 03:52
chadmed[ 184.342412] apple-dcp 289c00000.dcp: RTKit: syslog message: PPipeDCP_H13P.cpp:10259: IOMFB: polarity stuck! Resetting display
#asahi-dev · 2026-08-12 00:17
Rafayhttps://pastebin.com/QS9tAabE
#asahi-dev · 2026-08-12 00:15
RafayUghh, that forcing 0x413 is breaking link init/training and therefore I am getting USB 2 speeds.
#asahi-dev · 2026-08-11 23:26
RafayThe other host is x86.
#asahi-dev · 2026-08-11 23:24
Rafaysven: I forced 0x413 in apple.c and now I am able to ping and nc.
#asahi-dev · 2026-08-11 20:22
svenprobably just need to move that return further down i guess
#asahi-dev · 2026-08-11 20:22
svenanyway, will take another look tomorrow
#asahi-dev · 2026-08-11 20:22
svenugh, sorry. should've tested that first locally but i already shut down my test rig. anyway, maybe lane bonding is somehow broken because i don't see that with my m1 setup but i think yours tries to do that
#asahi-dev · 2026-08-11 20:19
chaos_princess"failed to add NVM devices: -19"
#asahi-dev · 2026-08-11 20:19
chaos_princessgets stuck in connect/disconnect loop
#asahi-dev · 2026-08-11 20:09
svenchaos_princess: can you add return -ENODEV or whatever at the beginning of tb_xdomain_lane_bonding_enable in xdomain.c?
#asahi-dev · 2026-08-11 19:57
svenso you not only have ~every mac out there, you also have ~every usb4 cable out there :D
#asahi-dev · 2026-08-11 19:57
svenhahaha, nice!
#asahi-dev · 2026-08-11 19:57
jannauincluding an optical usb4 cable
#asahi-dev · 2026-08-11 19:57
sveni think everything that doesn't use EUDO is very broken right now though. there's a bug somewhere in tipd that i'll have to figure out
#asahi-dev · 2026-08-11 19:56
svennice!
#asahi-dev · 2026-08-11 19:56
jannausven: cables are all working now. the issues before were probably caused by smc's charging handling breaking things when running HV and connecting a power source into the other port
#asahi-dev · 2026-08-11 19:56
svenhrm... okay
#asahi-dev · 2026-08-11 19:55
chaos_princessno, one is macos
#asahi-dev · 2026-08-11 19:55
svenare you running linux on both machines?
#asahi-dev · 2026-08-11 19:51
sveni just must be missing something dumb somewhere

plus ›