Bump rockchip64-edge kernel to 6.17

[paolosabatino]

Description

Lot's of adaptations for rockchip64-edge kernel to bump to version 6.17:

• all the clashing things for rk3576 have been upstreamed and thus removed from the patch archive.
• rkvdec patches floating around have been removed: rkvdec has been promoted from staging and 6.16 patches did not apply anymore, plus they were a bit of a mess. A new set of patches has been taken from LibreELEC repository since they are already appear in the kernel mailing lists and are already rebased upon 6.17. I did not test the outcome, but surely now the kernel compiles and does not complain.
• kernel .config updated to 6.17
• fix for uwe5622 driver
• fix for realtek drivers (see downstream pull requests below), code is already taking the commits from the pull requests
• all patches have been adapted, there are no *.disabled files left in the rockchip64-6.17 patch archive

GitHub issue reference:
Jira reference number AR-2761

Downstream pull requests:

• rtl8723ds: Patch for kernel 6.17 rtl8723ds#3
• rtl8852bs: Patch for kernel 6.17 wifi-rtl8852bs#3

How Has This Been Tested?

• Kernel compiles
• Tested successfully on a live rk3328 system: board boots and dmesg is clean
• Test on rk3399
• Test on rk3588
• Test on rk3576
• Test on rk356x

Anyone who has any chance to test on a board with the untested chips above is welcome. I guess the merge window for this PR should be very short as these kind of PRs rot quickly. Thanks.

Checklist:

• My code follows the style guidelines of this project
• I have performed a self-review of my own code
• My changes generate no new warnings
• Any dependent changes have been merged and published in downstream modules

[coderabbitai]

Walkthrough

Bump linux-rockchip64-edge kernel from 6.16 to 6.17 and update related kernel config and driver patching: many netfilter/iptables/netfilter (XT/NF/IP6/bridge) options were added or removed; CONFIG_RTW89_8851BU and CONFIG_RTW89_8852BU enabled; CONFIG_RTC_DRV_PCF85063 added; CONFIG_REGULATOR removed; VIDEO/TRACEFS flags adjusted. lib/functions/compilation/patch/drivers_network.sh updated Realtek driver commit references and added conditional application of an uwe5622 patch for kernel ≥ 6.17.

Estimated code review effort

🎯 3 (Moderate) | ⏱️ ~25 minutes

Possibly related PRs

• uefi-arm64: bump to 6.16 #8453 — Modifies the same family include kernel-version variable and adds kernel-version-specific wireless driver patches.
• armsom-sige5: add edge kernel support #8585 — Updates rtl8852bs driver handling and touches drivers_network.sh and rockchip edge config.
• uefi-all/x13s/meson64/rockchip64: bump edge to 6.15(final); legacy wifi bumps; HDMIRX for some rk3588 boards #8234 — Edits KERNEL_MAJOR_MINOR and adds versioned uwe5622 patch logic in drivers_network.sh.

Suggested reviewers

• joekhoobyar
• prahal
• rpardini
• amazingfate
• HeyMeco
• SuperKali
• igorpecovnik
• pyavitz

Pre-merge checks and finishing touches

❌ Failed checks (1 warning)

Check name	Status	Explanation	Resolution
Docstring Coverage	⚠️ Warning	Docstring coverage is 33.33% which is insufficient. The required threshold is 80.00%.	You can run @coderabbitai generate docstrings to improve docstring coverage.

✅ Passed checks (4 passed)

Check name	Status	Explanation
Title Check	✅ Passed	The title accurately and concisely summarizes the primary change of bumping the rockchip64-edge kernel version to 6.17, matching the core intent of the pull request.
Linked Issues Check	✅ Passed	The pull request fully implements the objective of linked issue AR-2761 by updating the kernel version string, regenerating the configuration for 6.17, adapting or removing patches, and updating driver logic to support the new kernel, thereby satisfying the issue requirement.
Out of Scope Changes Check	✅ Passed	All modifications, including config updates, driver commit bumps, patch logic, and kernel version changes, are directly related to upgrading the rockchip64-edge kernel to 6.17 and do not introduce unrelated or extraneous changes outside the scope of the linked issue.
Description Check	✅ Passed	The pull request description clearly outlines the bump to kernel version 6.17, details the removal and replacement of patches, updates to the .config, driver fixes, and summarizes testing steps, all of which directly correspond to the changes in the diff.

✨ Finishing touches

• 📝 Generate docstrings

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Post copyable unit tests in a comment

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands and usage tips.}

[github-actions]

✅ This PR has been reviewed and approved — all set for merge!

[EvilOlaf]

I can test on zero 3e (rk3566) and rock 3a (rk3568), though earliest I can do is upcoming week.

[paolosabatino]

Tested on OrangPi 4 LTS (rk3399), boots and dmesg is kinda clean. I don't see rkvdec though, got to make a check in the kernel configs...

edit:
it looks like is there at /dev/video1, it just does not announce itself in dmesg (hantro is /dev/video3):

also checked against a h.265 video and rkvdec works fine

[paolosabatino]

Well it looks to me that general state is good, I would like to proceed and merging to avoid this PR becoming rot and blocking other possible contributions. @EvilOlaf do you think it is ok for you even though rk356x has not been tested yet? Since this is edge kernel, I think we can address fixes later.

[HeyMeco]

I can test RK3566 / RK3568 this weekend but I agree that this can already be merged

[EvilOlaf]

My bet is that everything will be just fine. And it is bleeding edge anyway so yeah, let's do it.

[HeyMeco]

Just updated my NanoPi R5S (RK3568) via the beta branch to 6.17. Everything works

[paolosabatino]

Thank you all!

[EvilOlaf]

tested with zero 3e, so rk356x seems good. https://paste.armbian.com/acacogerih

[rpardini]

Hey @paolosabatino -- do you remember anything about CONFIG_TRACEFS_AUTOMOUNT_DEPRECATED here?

I can't reason where it came from (it is default y in Kconfig, introduced in 6.17-rc1).

Turns out some eBPF tooling still depends on it, I'll re-enable in the 6.18-rc3 PR if you don't mind.

[paolosabatino]

Hello @rpardini, yes I remember it appeared automatically from make oldconfig after bumping the kernel config from 6.16 to 6.17.

Absolutely I don't mind if you alter that config option, I just accepted it as the proposed default from kernel.

[rpardini]

Thanks Paolo. I'll keep an eye to see if it happens in other families (thus far it hasn't).