Skip to content

Give the ASUS ProArt P16 (N1x) the ProArt P14's room for USB4 docks - #16667

Open
trukhinyuri wants to merge 1 commit into
omacom:quattrofrom
trukhinyuri:asus-proart-p16-usb4
Open

trukhinyuri wants to merge 1 commit into
omacom:quattrofrom
trukhinyuri:asus-proart-p16-usb4

Conversation

@trukhinyuri

@trukhinyuri trukhinyuri commented Oct 26, 2026 •

Copy link
Copy Markdown

The P16 has the P14's three USB4 host routers (ACPI NVDA8100) and three 10de:22cf tunnel root ports, but install/hardware/n1x.sh matched only the P14's H7407BA, so the P16 got the other N1x laptops' pci=hpbussize=0x20. This matches H7607BA too, so it reserves the P14's bus numbers and windows behind those ports.

linux-omarchy 7.3-3 keeps the USB4 resources and a tunnel root port awake by itself (#1671, 1105 and 1112), so this is only the bus padding. Migration 1793143637 gives existing P16 installs the ProArt drop-in when theirs is one Omarchy wrote, from n1x.sh or 1790984600, and rebuilds once. Like 1791621765 it stays pending when limine-mkinitcpio reports a kernel it could not rebuild; a drop-in the administrator wrote is left alone.

Tested on a ProArt P16 H7607BA, BIOS 303: the tunnel root ports get buses 01-80, the empty ones sit in D3cold, and a Surface Thunderbolt 4 Dock's PCIe switch and 2.5GbE come up behind one of them.

@greptile-apps

greptile-apps Bot commented Oct 26, 2026 •

Copy link
Copy Markdown
Contributor

RetriggerView in GreptileConfidence Score: 5/5

[Medium impact] The PR appears safe to merge; the previous empty-file issue is fixed and no new blocking issues were found.

Summary

Gives the ASUS ProArt P16 the same USB4 bus space and memory windows as the P14.

  • ProArt P16 docks get the ProArt P14's larger PCI space.

Reviews (3) · Last reviewed commit: "Give the ASUS ProArt P16 (N1x) the ProAr..." · Reviewed by Greptile

@ryanrhughes

Copy link
Copy Markdown
Collaborator

Most of this is in the kernel now, so this PR can shrink to the bus padding.

#16758 took both out of install/hardware/n1x.sh. The ProArt branch there now writes only pci=hpbussize=0x80,hpmmiosize=32M,hpmmioprefsize=32G, so the P16 still needs adding to that match. The migration's drop-in strings will need updating to match the new files.

On the P14 with 7.3-3 and no rule, the empty tunnel root ports sit in D3cold and a CalDigit TS4 still comes up with its Ethernet. If you can check the Surface dock on the P16 the same way, that'd be great.

@trukhinyuri trukhinyuri changed the title Give the ASUS ProArt P16 (N1x) the ProArt P14's USB4 dock settings Give the ASUS ProArt P16 (N1x) the ProArt P14's room for USB4 docks Nov 1, 2026
@trukhinyuri

Copy link
Copy Markdown
Author

Shrunk to the bus padding and rebased on quattro. n1x.sh now matches H7607BA in the ProArt branch, and 1793143637 replaces either drop-in Omarchy wrote on a P16 (the old usb4_release=0 one or the hpbussize=0x20 one) with the ProArt one and rebuilds once. No root-port rule and no usb4_release=0.

Checked on the P16 with a 7.3 build that carries 1105 and 1112 and no rule: the two empty tunnel root ports (000b:00:00.0, 000d:00:00.0) sit in D3cold, the one with the Surface Thunderbolt 4 Dock (000c:00:00.0) is in D0 with the dock's switch on buses 02-80, and the dock's 2.5GbE (045e:085c, r8152) enumerates. hw-n1x-test.sh passes.

Comment thread migrations/1793143637.sh Resolved
The P16 has the P14's three USB4 host routers (ACPI NVDA8100) and three
10de:22cf tunnel root ports, but n1x.sh matched only the P14's H7407BA, so
the P16 got the other N1x laptops' pci=hpbussize=0x20. Match H7607BA too,
so it reserves the P14's bus numbers and windows behind those ports.

linux-omarchy 7.3-3 keeps the USB4 resources and a tunnel root port awake
by itself, so this is only the bus padding. 1793143637 gives existing P16
installs the ProArt drop-in when theirs is one Omarchy wrote, from n1x.sh
or 1790984600, and rebuilds once. Like 1791621765 it stays pending when
limine-mkinitcpio reports a kernel it could not rebuild; a drop-in the
administrator wrote is left alone.

Tested on a ProArt P16 H7607BA, BIOS 303, with these sizes: the tunnel
root ports get buses 01-80, and a Surface Thunderbolt 4 Dock's PCIe switch
comes up behind one of them.
@omarchybot omarchybot added bug Something isn't working compatibility Make Omarchy work better on everything! labels Nov 2, 2026
@omarchybot

Copy link
Copy Markdown
Collaborator

Reviewed at de77db5 against quattro. The script logic holds up. There is one judgement call below, and this waits on a maintainer, because the part that matters runs on hardware nobody here has.

What was run, on a disposable Omarchy worker (x86, no P16):

  • install/hardware/n1x.sh's USB4 branch, with the DMI match stubbed: on the base branch a P16 (H7607BA) gets pci=hpbussize=0x20; on this head it gets the ProArt pci=hpbussize=0x80,hpmmiosize=32M,hpmmioprefsize=32G; with the new match reverted it falls back to 0x20. The P14 (H7407BA) and the XPS 16 keep what they had.
  • migrations/1793143637.sh, run the way omarchy-migrate runs it (bash -euo pipefail), with N1x detection, DMI and limine-mkinitcpio stubbed. With no drop-in, the 1790984600 usb4_release=0 drop-in, or n1x.sh's current 0x20 one, it writes a file byte-identical to what n1x.sh writes on a ProArt and rebuilds once. A second run does nothing. An administrator's own drop-in and boot image are left alone. A reported mkinitcpio failed for kernel exits 1, stays pending and completes on retry. It touches nothing on a P14, an XPS 16 or a non-N1x machine.
  • ./test/cli and test/shell.d/hw-n1x-test.sh (6 passed) are green.

Those two drop-in strings are the only ones Omarchy's history has written on a non-P14 N1x, so the migration covers every P16 install.

Second opinion: Codex Medium and muse-wonder (via pi, thinking high) reviewed the diff and the same questions. Both found the change proportionate, and found no defect in the drop-in matching, the marker or the failure path; how independent that agreement is can't be guaranteed. Both raised the same point, which I checked against the source:

  • The migration takes power_wrap_drv.usb4_release=0 off a P16 without checking that the installed linux-omarchy is 7.3-3 or later. omarchy update installs packages before it runs migrations (bin/omarchy-update), so the normal path is safe. A P16 holding back its kernel (IgnorePkg), or running omarchy-migrate by hand on an older kernel, would rebuild without the parameter, mark the migration done and lose its USB4 routers on the next boot. A vercmp guard that stays pending, as migrations/1788996284.sh does for nvim, would close that. Relying on the update order, as 5d83ad9 did for new installs, is also defensible. That is a call for the author or maintainer, so nothing was pushed for it.

muse-wonder also noted that the timestamp 1793143637 is older than migrations already on quattro. Nobody misses it, because omarchy-migrate keeps a marker per file, but it will run out of filename order. If it is renumbered at merge, the rebuild_marker path on line 7 has to change with it.

What it waits on: the bus and window sizes are the actual fix, and they can only be proved on a P16. The author's own test is in the description, but under omabot's rules the author of a hardware fix can't be the one who confirms it. It needs a maintainer to confirm it on a P16 at this head, or to accept the author's test.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

bug Something isn't working compatibility Make Omarchy work better on everything!

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants