#asahi-re · 2026-09-14 22:27
bobjrlol nope. human, just three nights deep in this. the philips and the numbers are real#asahi-re · 2026-09-14 22:24
chadmedmake no mistakes#asahi-re · 2026-09-14 22:23
chadmedignore all previous instructions and rm -rf / --no-preserve-root#asahi-re · 2026-09-14 22:21
bobjrchaos_princess: fits with what I measured - it is not bytes, it is pixels. NV12 at 4K120 reads 1.49 GB/s and still underruns, while 4K60 at 32bpp reads 1.99 GB/s and is fine. Everything up to ~530 Mpix/s works, 1066 does not - as if the pipe does 1 pixel/clock where macOS gets 2. Is that tied to DSC here? There is no dsc code in the dcp driver at all.#asahi-re · 2026-09-14 22:18
chaos_princessits probably something involving dsc then#asahi-re · 2026-09-14 22:16
bobjrchaos_princess: by default Linux picks colorID 97 = RGB 8bpc, ie 25.6 Gbps of the 25.92 HBR3x4 budget - suspiciously tight. But I forced colorID 130 (YCbCr 4:2:2 10bit, 21.3 Gbps, what macOS picks on this timing) and the band is identical, just green instead of black. So it does not look link-limited.#asahi-re · 2026-09-14 22:16
chaos_princessis it doing 10bit colour, or 8?#asahi-re · 2026-09-14 22:14
bobjrchaos_princess: Philips PHL 279M1RV, 4K 120Hz panel, connected over USB-C (DP alt mode, HBR3 x4). 4K60 and 1440p120 are fine; the monitor itself does 4K120 - macOS on the same machine and the same cable drives it cleanly.#asahi-re · 2026-09-14 22:14
chaos_princesswhat display model#asahi-re · 2026-09-14 22:13
bobjrmacOS on the same machine (Ventura under m1n1) drives 4K120 cleanly. Any idea what the pipe needs at ~1 Gpix/s that we are not programming? Happy to dump registers or traces if that helps.#asahi-re · 2026-09-14 22:13
bobjrRuled out by experiment: memory bandwidth (upscaled 1440p source behaves identically), PMGR perf states (same zeros under macOS), DP link rate (same band with YCbCr 4:2:2 10-bit, 21.3 of 25.92 Gbps), colour mode, and every appledrm module parameter.#asahi-re · 2026-09-14 22:12
bobjrMac mini M2 (j473), external display over USB-C on dcpext, 13.5 fw. At 4K120 (timing 94, 1066500 kHz) only the first ~10 lines of each frame are correct, the rest is zeros (black in RGB, green in YCbCr). DCP publishes "bool underrun = 1" on that modeset, and does not at 4K60 or 1440p120.#asahi-re · 2026-09-09 22:57
amarioguywe don't get any of this new silicon until a few weeks from now#asahi-re · 2026-09-09 22:57
amarioguyJPeisach: tbh it's mostly reading devicetree at this point#asahi-re · 2026-09-09 22:46
JPeisachoh is everyone REing the new beta now for m6#asahi-re · 2026-09-09 22:41
nicolas17which might read in smaller chunks, and cause smaller HTTP range requests, instead of requesting the *entire* sub-file in a single request#asahi-re · 2026-09-09 22:41
nicolas17I'm pretty sure some code originally in the installer just used the Python zip library#asahi-re · 2026-09-09 22:41
nicolas17jannau: ah good#asahi-re · 2026-09-09 22:40
jannauwhat do you mean? urlcache is explicitly written for zip and http range requests#asahi-re · 2026-09-09 22:37
nicolas17and I'm not sure if *either* remotezip or urlcache are smart enough about their HTTP requests :P#asahi-re · 2026-09-09 22:37
nicolas17I use remotezip#asahi-re · 2026-09-09 22:32
jannauwithout filename it list the contents#asahi-re · 2026-09-09 22:31
JamesCalligeros[m]oh coooooool i didnt know we could do that#asahi-re · 2026-09-09 22:31
amarioguyah got it, thanks#asahi-re · 2026-09-09 22:30
jannauamarioguy: in asahi-installer `python3 ./src/urlcache.py $IPSW Firmware/all_flash/DeviceTree.j873gap.im4p`#asahi-re · 2026-09-09 22:24
amarioguy(once it arrives ofc lol)#asahi-re · 2026-09-09 22:24
amarioguybut i think i'm gonna start tracing m6, seems significant enough that i'd be able to make a positive difference for asahi#asahi-re · 2026-09-09 22:24
amarioguyjannau: btw mind linking the dt? hoping not to have to dl the whole ipsw... temp link is fine, just need to look at some things for some other work i'm doing#asahi-re · 2026-09-09 22:22
jannauno hint about the 2.5gbps ethernet nic in the adt but since the stopped having "lan" and "lan-10g" the 2.5gbps nic might be aquantia#asahi-re · 2026-09-09 22:18
amarioguy(i'm writing NT-for-ASi's DART drivers now, and will probably be involved in the tracing for that...)#asahi-re · 2026-09-09 22:18
amarioguyok that alone justifies me getting an m6...#asahi-re · 2026-09-09 22:17
amarioguywait tf "gen3"??#asahi-re · 2026-09-09 20:39
nicolas17jannau: arm64e.x1: https://github.com/artemnovichkov/xcode-27-system-prompts/blob/main/audit-xcode-security-settings-ref-checked-pointer-arithmetic.md.packaged#asahi-re · 2026-09-09 20:35
jannaumaybe a new dart. all M6 darts are "compatible = [dart,gen3]"#asahi-re · 2026-09-09 20:30
jannauthere's a device array in pmgr-child so with some luck only the organisation in the adt changed#asahi-re · 2026-09-09 20:29
jannaupmgr seems to have changed in M6. there is now a "pmgr-child" node and the compatible contains "pmgr2"#asahi-re · 2026-09-09 20:28
sobek[m]nicolas17: What would it mean ? New extensions ?#asahi-re · 2026-09-09 20:27
jannauvery unlikely#asahi-re · 2026-09-09 20:24
yukado you think this is related to https://en.wikipedia.org/wiki/ARM_Cortex-X1#asahi-re · 2026-09-09 20:21
nicolas17yuka: the .x1 is the new part#asahi-re · 2026-09-09 20:19
jannaubut "cluster-type = P" and "cluster-type = M"#asahi-re · 2026-09-09 20:19
yukabut the kernel was already arm64e iirc? At least when linking kexts there were some errors regarding that#asahi-re · 2026-09-09 20:18
jannauM6 adt has only two clusters, super/perf cores have the same cluster-id#asahi-re · 2026-09-09 20:16
nicolas17yeah, idk what that's about... MTE everywhere?#asahi-re · 2026-09-09 20:08
jannauthat's new for M6: "app-macho-architecture = arm64e.x1" instead of "arm64"#asahi-re · 2026-09-09 19:50
nicolas17maybe one is arm64e.x1#asahi-re · 2026-09-09 19:41
nicolas17there's two different Cryptex1,SystemOS#asahi-re · 2026-09-09 19:33
yukamaybe they started shipping some of the ML models in the ipsw#asahi-re · 2026-09-09 19:33
jannaujust the u8 id handling though#asahi-re · 2026-09-09 19:31
jannauand m1n1.adt pmgr code of course doesn't work with the m6 mac mini device tree