Add board: Luckfox Nova (RK3308B)#9985
Merged
igorpecovnik merged 9 commits intoJun 19, 2026
Merged
Conversation
added 4 commits
June 16, 2026 14:49
Compact RK3308B board: 512MB DDR3, 8GB eMMC, 100M Ethernet (RTL8201F), USB-C OTG peripheral, USB-A host, microSD. Debug console is UART4 at 1500000 baud. Fully mainline: U-Boot 2025.04 (binman, rkbin DDR/BL31) and mainline kernel for both current (6.18) and edge (7.0). Board quirks handled: - the microSD slot only has DAT0 usable - 4-bit transfers fail with a Start Bit Error in U-Boot and I/O errors in Linux (verified with two cards), so both the U-Boot and kernel DTS pin the slot to bus-width=1 (SD High Speed at 50MHz works fine). The U-Boot side also needs dwmci_setup_cfg() to honor a 1-bit DT bus-width plus two dw_mmc fixes backported from upstream (dwmci_control_clken CDONE ack, setup_bus same-frequency skip). - the evb_rk3308 download-key handler hardcodes the pre-DM "saradc" device name; fixed to the actual device name so the recovery key works, and the key now enters RockUSB directly (usable with rkdeveloptool/upgrade_tool without entering maskrom). - ethernet PHY is strapped at MDIO address 0; the gmac node lets stmmac scan the bus instead of assuming an address. Tested on hardware: boots from eMMC and SD (SD preferred when inserted), ethernet, USB host/OTG, PWM, SARADC, recovery key.
The mainline kernels (current 6.18, edge 7.0) had no audio enabled. Add a simple-audio-card binding i2s_8ch_2 to the RK3308 internal codec and enable both nodes, following the Rock S0 pattern. This brings up the on-board analog mic (and the lineout) on card 0. Signed-off-by: crackerjacques <jack@supremeoverlordjabs.co>
rk3308.dtsi carries the PDM pin groups but no controller node, so the PDM interface on the P1 header is unusable. Define pdm@ff380000 (values from the Rockchip BSP) on the M2 pin mux plus a dmic-codec simple-card. Disabled by default - the pins are shared with i2s_8ch_0 and GPIO. Enable CONFIG_SND_SOC_DMIC (current + edge) for the dmic-codec. Signed-off-by: crackerjacques <jack@supremeoverlordjabs.co>
Enable the dedicated SDIO controller (mmc@ff4a0000) for the on-board AIC8800DC WiFi module with an mmc-pwrseq-simple driving WL_REG_ON (GPIO0_A2). Harmless on a plain Nova (empty SDIO scan). The AIC8800DC fullmac driver is out-of-tree; build it via DKMS on the running system (LYU4662/aic8800-sdio-linux + the 6.16+ API compat fixups). current(6.18) and edge(7.0). Signed-off-by: crackerjacques <jack@supremeoverlordjabs.co>
crackerjacques
requested review from
ColorfulRhino,
HeyMeco,
JohnTheCoolingFan,
NicoD-SBC,
SuperKali,
TheSnowfield,
Tonymac32,
adeepn,
ahoneybun,
amazingfate,
brentr,
catalinii,
chainsx,
efectn,
iav,
igorpecovnik,
joekhoobyar,
krachlatte,
lanefu,
paolosabatino,
prahal,
pyavitz,
rpardini and
schwar3kat
as code owners
Contributor
|
No actionable comments were generated in the recent review. 🎉 ℹ️ Recent review info⚙️ Run configurationConfiguration used: Path: .coderabbit.yaml Review profile: CHILL Plan: Pro Run ID: 📒 Files selected for processing (2)
🚧 Files skipped from review as they are similar to previous changes (2)
📝 WalkthroughWalkthroughAdds full Armbian board support for the Luckfox Nova (RK3308B/rockchip64): Armbian board config, UART4 U-Boot boot script, U-Boot DTS/defconfig and four dw_mmc/download-key driver patches, kernel device trees for rockchip64-6.18 and 7.1, DMIC kernel module enablement, and a BSP CLI package with ALSA configs plus GPIO/PWM/audio/overlay management scripts. ChangesLuckfox Nova RK3308B board support
Sequence Diagram(s)sequenceDiagram
rect rgba(70, 130, 180, 0.5)
note over UBoot,BOOTROM: U-Boot recovery key detection
participant UBoot as U-Boot
participant SARADC as saradc@ff1e0000
participant RockUSB as rockusb gadget
participant BOOTROM as BOOTROM
end
UBoot->>SARADC: adc_channel_single_shot("saradc@ff1e0000")
SARADC-->>UBoot: ADC key value
alt download key pressed
UBoot->>RockUSB: rockusb 0 mmc 0
RockUSB-->>RockUSB: serve USB flashing commands
RockUSB-->>UBoot: command returns
UBoot->>BOOTROM: set_back_to_bootrom_dnl_flag + do_reset()
else normal boot
UBoot->>UBoot: continue to boot script
end
sequenceDiagram
rect rgba(60, 179, 113, 0.5)
note over User,BootConfig: novaconfig overlay management
participant User
participant novaconfig
participant dtc as dtc compiler
participant BootConfig as armbianEnv.txt / extlinux.conf
participant DTB as live device tree
end
User->>novaconfig: run (root required)
novaconfig->>BootConfig: detect Armbian vs extlinux mode
novaconfig->>User: show interface menu with enabled state
User->>novaconfig: toggle interface (e.g. pdm)
novaconfig->>novaconfig: check pin-group conflicts
novaconfig->>dtc: compile nova-pdm.dts → nova-pdm.dtbo
dtc-->>novaconfig: dtbo written to overlays dir
novaconfig->>BootConfig: write_list updates user_overlays / fdtoverlays
novaconfig->>DTB: live_status reads /proc/device-tree
DTB-->>novaconfig: current node status
novaconfig->>User: show updated menu
Estimated code review effort🎯 5 (Critical) | ⏱️ ~120 minutes Suggested labels
Suggested reviewers
Poem
🚥 Pre-merge checks | ✅ 5✅ Passed checks (5 passed)
✏️ Tip: You can configure your own custom pre-merge checks in the settings. ✨ Finishing Touches🧪 Generate unit tests (beta)
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. Comment |
github-actions
Bot
added
08
Contributor
🚫 Missing required board assetsThis PR adds new board configuration(s). Required assets must already exist in github/armbian/armbian.github.io.
Missing items
Once the missing files are added (or a PR is opened in armbian/armbian.github.io), re-run this check. |
![coderabbitai[bot]](https://avatars.githubusercontent.com/in/347564?s=60&v=4){kind=small}
Contributor
There was a problem hiding this comment.
Actionable comments posted: 7
🤖 Prompt for all review comments with AI agents
Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.
Inline comments:
In
`@config/optional/boards/luckfox-rk3308b-nova/_packages/bsp-cli/usr/sbin/gpiocheck`:
- Around line 148-152: The gpiocheck function currently accepts any non-empty
value for the third parameter and does not handle failures from the sysfs
writes. Add validation to ensure the third parameter is restricted to only "0"
or "1" values, and modify the echo commands that write to the direction and
value files to fail fatally on error using the die function (similar to the
pattern used elsewhere in the script) so that the success message on the final
line is only printed when all sysfs operations have actually succeeded.
In
`@config/optional/boards/luckfox-rk3308b-nova/_packages/bsp-cli/usr/sbin/mictest`:
- Around line 30-37: Both mictest and pdmtest scripts use predictable hardcoded
/tmp paths (mictest8.wav, mictest.wav, and pdmtest.wav) that are vulnerable to
symlink-based file clobber attacks. In
config/optional/boards/luckfox-rk3308b-nova/_packages/bsp-cli/usr/sbin/mictest
at lines 30-37, replace the fixed OUT=/tmp/mictest8.wav and OUT=/tmp/mictest.wav
assignments with secure temporary file creation using mktemp before invoking
arecord. Similarly, in
config/optional/boards/luckfox-rk3308b-nova/_packages/bsp-cli/usr/sbin/pdmtest
at lines 22-25, replace the fixed OUT=/tmp/pdmtest.wav with a securely created
unique temp file using mktemp before invoking arecord. Ensure both fixes follow
the same secure pattern for consistency.
In
`@config/optional/boards/luckfox-rk3308b-nova/_packages/bsp-cli/usr/sbin/novaconfig`:
- Around line 141-150: The enabled_list function correctly filters to show only
nova-* overlays, but the code that rewrites the overlay configuration (at lines
161-169) is dropping non-Nova entries from the boot config. When rewriting
overlays after any toggle operation, the code must preserve existing non-Nova
overlays that were already present in the boot configuration. Modify the code
that reconstructs the overlay line to extract and preserve any non-Nova overlays
from the original configuration (both for armbian mode using user_overlays and
extlinux mode using fdtoverlays), then combine those with the modified Nova
overlays when writing back the complete overlay line to the configuration file.
- Around line 64-67: Replace the predictable temporary file path assignment for
the tmp variable with a secure alternative using mktemp. Instead of hardcoding
/tmp/nova-$1.dts, use mktemp with an appropriate template (such as mktemp
/tmp/nova-XXXXXX.dts or similar) to generate a random temporary filename. This
eliminates the symlink clobber attack vulnerability by ensuring the temporary
file path cannot be predicted by an attacker, even when the script runs as root.
In
`@config/optional/boards/luckfox-rk3308b-nova/_packages/bsp-cli/usr/sbin/pwmtest`:
- Around line 54-56: The kernel-use detection logic at line 54 uses a glob
pattern `"$c"/../*/pwm-*` that checks sibling directories rather than limiting
the scope to the current pwmchip represented by variable `c`. Replace this path
pattern with one that only checks for pwm devices within the current pwmchip
directory (change `"$c"/../*/pwm-*` to `"$c"/pwm-*`) so that the "in use
(kernel)" state is correctly set only when channels under the current pwmchip
are in use, not when any sibling pwmchip has in-use channels.
- Around line 81-96: Add a guard after computing period at line 81 to check if
period equals zero and die with an appropriate error message (e.g., "frequency
too high, period would be 0"). Then add error checking to the PWM configuration
writes at lines 93-95 by changing the redirects from "2>/dev/null" to "|| die"
to fail fast if any write fails. This ensures the success message printed at
line 96 only executes when all configurations have been successfully applied.
In
`@patch/u-boot/v2025.04/board_luckfox-rk3308b-nova/mmc-dw-mmc-fix-clken-cmd-completion.patch`:
- Around line 45-54: The `dwmci_control_clken()` function returns success (0) on
Line 54 regardless of whether the CDONE bit was actually observed before the
timeout expired. Add proper timeout failure handling after the for loop: if the
loop exhausts all `timeout` iterations without finding the DWMCI_INTMSK_CDONE
bit set (i.e., the break statement was never executed), the function should
return an error code instead of 0. Only return 0 if the CDONE bit was
successfully detected and the loop exited via break.
🪄 Autofix (Beta)
Fix all unresolved CodeRabbit comments on this PR:
- Push a commit to this branch (recommended)
- Create a new PR with the fixes
ℹ️ Review info
⚙️ Run configuration
Configuration used: Path: .coderabbit.yaml
Review profile: CHILL
Plan: Pro
Run ID: 42a48b07-3265-4a30-ac40-bd9f264a5f9b
📒 Files selected for processing (18)
config/boards/luckfox-rk3308b-nova.cscconfig/bootscripts/boot-rockchip64-ttyS4.cmdconfig/kernel/linux-rockchip64-current.configconfig/kernel/linux-rockchip64-edge.configconfig/optional/boards/luckfox-rk3308b-nova/_packages/bsp-cli/etc/asound.confconfig/optional/boards/luckfox-rk3308b-nova/_packages/bsp-cli/usr/sbin/gpiocheckconfig/optional/boards/luckfox-rk3308b-nova/_packages/bsp-cli/usr/sbin/mictestconfig/optional/boards/luckfox-rk3308b-nova/_packages/bsp-cli/usr/sbin/novaconfigconfig/optional/boards/luckfox-rk3308b-nova/_packages/bsp-cli/usr/sbin/pdmtestconfig/optional/boards/luckfox-rk3308b-nova/_packages/bsp-cli/usr/sbin/pwmtestpatch/kernel/archive/rockchip64-6.18/dt/rk3308-luckfox-nova.dtspatch/kernel/archive/rockchip64-7.0/dt/rk3308-luckfox-nova.dtspatch/u-boot/v2025.04/board_luckfox-rk3308b-nova/add-board-luckfox-rk3308b-nova.patchpatch/u-boot/v2025.04/board_luckfox-rk3308b-nova/boot-mode-enter-rockusb-on-download-key.patchpatch/u-boot/v2025.04/board_luckfox-rk3308b-nova/fix-evb-rk3308-dnl-key-saradc-name.patchpatch/u-boot/v2025.04/board_luckfox-rk3308b-nova/mmc-dw-mmc-fix-clken-cmd-completion.patchpatch/u-boot/v2025.04/board_luckfox-rk3308b-nova/mmc-dw-mmc-honor-1bit-bus-width.patchpatch/u-boot/v2025.04/board_luckfox-rk3308b-nova/mmc-dw-mmc-no-skip-setup-bus-same-freq.patch
Contributor
Author
|
The Verify assets for newly added boards check is red only because the board image lives in the companion website PR armbian/armbian.github.io#332, which is still open. That PR adds Once #332 is merged, the asset check can be re-run and should go green. Same pattern as the recent Anbernic RG DS board (armbian.github.io#326). Thanks! |
crackerjacques
force-pushed
the
board-luckfox-rk3308b-nova
branch
from
408aedb to
d6327f7
Compare
Member
|
edge was moved to 7.1, the 7.0 patchset is no longer used and will be remove at some point in the future |
Member
Some corrections are needed. I updated automation to tell which. |
added 4 commits
June 17, 2026 12:13
The Nova is headless and has a few board-specific bits that need poking
from userspace: enabling the PDM mic / PWM / SPI / I2C (mainline has no
runtime DT-overlay interface, so these are toggled via boot-time overlays),
capturing the on-board codec mic and the PDM mics, and inspecting GPIO/PWM
while bringing the board up. Ship the helpers in the board bsp-cli package
so they land in the image:
- novaconfig: enable/disable interfaces by generating + compiling a DT
overlay on the device and wiring it into armbianEnv.txt user_overlays
- mictest: record from the RK3308 internal-codec on-board mic
- pdmtest: record from the PDM digital mics on the P1 header
- gpiocheck / pwmtest: GPIO and PWM inspectors
- /etc/asound.conf: 'mic'/'mics' pcms for the channel-8 on-board mic
novaconfig compiles overlays on-device and mictest/pdmtest use arecord, so
pull in device-tree-compiler + alsa-utils via PACKAGE_LIST_BOARD.
The eMMC node only declared cap-mmc-highspeed, so mode negotiation capped at eMMC High Speed (~52 MHz) even though max-frequency is 150 MHz - about a third of the eMMC's real throughput. The vendor DTS and our U-Boot already run it at HS200; only the kernel DT was missing the cap. Add mmc-hs200-1_8v to &emmc (current 6.18 and edge 7.0). The eMMC VCCQ is a fixed 1.8 V rail on this board and the dw_mmc controller has the tuning clocks, so HS200 negotiates; on tuning failure the kernel falls back to High Speed, so this is low risk. Verified on hardware: "mmc0: new HS200 MMC card", phase tuned, bus clock 52 -> 147 MHz. The microSD slot (&sdmmc) is untouched and stays bus-width=1 (DAT1-3 are not usable on this board); this only affects the 8-bit eMMC.
The P1 header brings out i2s_8ch_0 (the RK3308 I2S/TDM controller at 0xff300000), but rk3308.dtsi only defines its pin groups, not the controller node, so mainline has no way to drive it. Add the controller node (clocks/dmas/resets from the vendor BSP, modelled on the in-tree i2s_8ch_2 binding), disabled by default since its pins are shared with the PDM interface (mutually exclusive). novaconfig grows an i2s_tx toggle that enables it for raw master/TX use (SCLK=GPIO2_A5 LRCK=GPIO2_A7 SDO=GPIO2_B1 SDI=GPIO2_B5). Verified on hardware: with the overlay applied the controller binds (rockchip-i2s-tdm), its clocks come up, and the pins are muxed to the i2s_8ch_0 function (sclktx/lrcktx/sdo0/sdi0).
Building on the raw i2s_8ch_0 controller, add a capture path so an
external I2S master can be recorded on the Nova. A simple-audio-card
("i2s-rx-test") binds i2s_8ch_0 to a dmic-codec; both nodes are disabled
by default. novaconfig grows an i2s_rx toggle whose overlay enables the
card, switches i2s_8ch_0 to its RX clock pins (sclkrx=GPIO2_A6,
lrckrx=GPIO2_B0, sdi0=GPIO2_B5) and frees those pins from the two header
PWMs (pwm@ff160030/ff170030) that otherwise squat on them. Mutually
exclusive with pdm and i2s_tx (shared GPIO2 audio pins). The overlay uses
labelled phandles, so emit_overlay now compiles with dtc -@.
Because that overlay references base-DTB labels, novaconfig refuses to
enable i2s_rx when the running DTB has no /__symbols__ node (a kernel not
built with dtc -@), where the bootloader could not apply the overlay and
the boot would fail. The other overlays use target-path and are unaffected.
Verified on hardware: a Raspberry Pi Pico driving I2S (1 kHz sine, 48 kHz,
16-bit stereo) into GPIO2_A6/B0/B5 is captured via arecord -D hw:i2srxtest.
Contributor
Author
|
okay, I will fix those. |
crackerjacques
force-pushed
the
board-luckfox-rk3308b-nova
branch
from
ddbc5f3 to
d313b09
Compare
Member
|
take your time, no need to hurry anything |
Contributor
There was a problem hiding this comment.
Actionable comments posted: 2
♻️ Duplicate comments (1)
patch/u-boot/v2025.04/board_luckfox-rk3308b-nova/mmc-dw-mmc-fix-clken-cmd-completion.patch (1)
45-54:⚠️ Potential issue | 🔴 Critical | ⚡ Quick winRestore timeout failure handling in
dwmci_control_clken().The function now returns success even when the CDONE interrupt bit is never observed. If the loop exhausts all
timeoutiterations without the CDONE bit being set, Line 54 still returns0, masking a failed clock-update command. This can cause MMC instability and data corruption.🐛 Proposed fix
for (i = 0; i < timeout; i++) { mask = dwmci_readl(host, DWMCI_RINTSTS); if (mask & DWMCI_INTMSK_CDONE) { dwmci_writel(host, DWMCI_RINTSTS, DWMCI_INTMSK_CDONE); break; } } + if (i == timeout) { + debug("%s: Timeout!\n", __func__); + return -ETIMEDOUT; + } return 0; }🤖 Prompt for AI Agents
Verify each finding against current code. Fix only still-valid issues, skip the rest with a brief reason, keep changes minimal, and validate. In `@patch/u-boot/v2025.04/board_luckfox-rk3308b-nova/mmc-dw-mmc-fix-clken-cmd-completion.patch` around lines 45 - 54, The dwmci_control_clken() function currently returns 0 (success) unconditionally after the timeout loop completes, even if the CDONE interrupt bit was never detected. To fix this, add a check after the loop to determine if the CDONE bit was found during the loop iterations. If the loop completed without detecting the DWMCI_INTMSK_CDONE bit in the mask variable (meaning timeout was exhausted), return a non-zero error code to indicate failure. Only return 0 if the CDONE bit was actually detected and the loop was exited via the break statement.
🤖 Prompt for all review comments with AI agents
Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.
Inline comments:
In `@patch/kernel/archive/rockchip64-7.1/dt/rk3308-luckfox-nova.dts`:
- Around line 124-136: The vbus_host regulator block is using the deprecated
`gpio` property instead of the modern `gpios` property. In the
regulator-5v0-vbus-host node, replace the singular `gpio` property with `gpios`
to align with current kernel device tree conventions and maintain consistency
with other regulators like vcc_sd in the same file.
- Around line 207-208: In the i2s_8ch_0 device tree node, swap the dmas
assignment and the dma-names assignment to match the RK3308 standard binding.
Change the dmas property from <&dmac1 1>, <&dmac1 0> to <&dmac1 0>, <&dmac1 1>,
and change the dma-names property from "rx", "tx" to "tx", "rx". Both properties
must be updated together to maintain correct DMA channel mapping for the I2S
controller.
---
Duplicate comments:
In
`@patch/u-boot/v2025.04/board_luckfox-rk3308b-nova/mmc-dw-mmc-fix-clken-cmd-completion.patch`:
- Around line 45-54: The dwmci_control_clken() function currently returns 0
(success) unconditionally after the timeout loop completes, even if the CDONE
interrupt bit was never detected. To fix this, add a check after the loop to
determine if the CDONE bit was found during the loop iterations. If the loop
completed without detecting the DWMCI_INTMSK_CDONE bit in the mask variable
(meaning timeout was exhausted), return a non-zero error code to indicate
failure. Only return 0 if the CDONE bit was actually detected and the loop was
exited via the break statement.
🪄 Autofix (Beta)
Fix all unresolved CodeRabbit comments on this PR:
- Push a commit to this branch (recommended)
- Create a new PR with the fixes
ℹ️ Review info
⚙️ Run configuration
Configuration used: Path: .coderabbit.yaml
Review profile: CHILL
Plan: Pro
Run ID: 43cc9d26-cf1d-4044-bbd3-0f7db3e53770
📒 Files selected for processing (18)
config/boards/luckfox-rk3308b-nova.cscconfig/bootscripts/boot-rockchip64-ttyS4.cmdconfig/kernel/linux-rockchip64-current.configconfig/kernel/linux-rockchip64-edge.configconfig/optional/boards/luckfox-rk3308b-nova/_packages/bsp-cli/etc/asound.confconfig/optional/boards/luckfox-rk3308b-nova/_packages/bsp-cli/usr/sbin/gpiocheckconfig/optional/boards/luckfox-rk3308b-nova/_packages/bsp-cli/usr/sbin/mictestconfig/optional/boards/luckfox-rk3308b-nova/_packages/bsp-cli/usr/sbin/novaconfigconfig/optional/boards/luckfox-rk3308b-nova/_packages/bsp-cli/usr/sbin/pdmtestconfig/optional/boards/luckfox-rk3308b-nova/_packages/bsp-cli/usr/sbin/pwmtestpatch/kernel/archive/rockchip64-6.18/dt/rk3308-luckfox-nova.dtspatch/kernel/archive/rockchip64-7.1/dt/rk3308-luckfox-nova.dtspatch/u-boot/v2025.04/board_luckfox-rk3308b-nova/add-board-luckfox-rk3308b-nova.patchpatch/u-boot/v2025.04/board_luckfox-rk3308b-nova/boot-mode-enter-rockusb-on-download-key.patchpatch/u-boot/v2025.04/board_luckfox-rk3308b-nova/fix-evb-rk3308-dnl-key-saradc-name.patchpatch/u-boot/v2025.04/board_luckfox-rk3308b-nova/mmc-dw-mmc-fix-clken-cmd-completion.patchpatch/u-boot/v2025.04/board_luckfox-rk3308b-nova/mmc-dw-mmc-honor-1bit-bus-width.patchpatch/u-boot/v2025.04/board_luckfox-rk3308b-nova/mmc-dw-mmc-no-skip-setup-bus-same-freq.patch
✅ Files skipped from review due to trivial changes (1)
- config/kernel/linux-rockchip64-edge.config
🚧 Files skipped from review as they are similar to previous changes (12)
- config/kernel/linux-rockchip64-current.config
- config/optional/boards/luckfox-rk3308b-nova/_packages/bsp-cli/usr/sbin/pdmtest
- config/optional/boards/luckfox-rk3308b-nova/_packages/bsp-cli/usr/sbin/mictest
- patch/u-boot/v2025.04/board_luckfox-rk3308b-nova/mmc-dw-mmc-no-skip-setup-bus-same-freq.patch
- patch/u-boot/v2025.04/board_luckfox-rk3308b-nova/fix-evb-rk3308-dnl-key-saradc-name.patch
- config/optional/boards/luckfox-rk3308b-nova/_packages/bsp-cli/etc/asound.conf
- config/boards/luckfox-rk3308b-nova.csc
- patch/u-boot/v2025.04/board_luckfox-rk3308b-nova/add-board-luckfox-rk3308b-nova.patch
- patch/u-boot/v2025.04/board_luckfox-rk3308b-nova/boot-mode-enter-rockusb-on-download-key.patch
- config/optional/boards/luckfox-rk3308b-nova/_packages/bsp-cli/usr/sbin/novaconfig
- config/optional/boards/luckfox-rk3308b-nova/_packages/bsp-cli/usr/sbin/gpiocheck
- patch/kernel/archive/rockchip64-6.18/dt/rk3308-luckfox-nova.dts
Contributor
Author
|
Thanks! edge is now on rockchip64-7.1 (current = 6.18); the 7.0 patchset isn't used. Also pushed a small DT tidy ( |
EvilOlaf
approved these changes
Jun 19, 2026
Contributor
|
✅ This PR has been reviewed and approved — all set for merge! |
Member
|
Nice. I wonder why u-boot v2025.04 and not v2026.07? |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
4 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Luckfox Nova (RK3308B)
Adds support for the Luckfox Nova: Rockchip RK3308B (quad Cortex-A35), 512 MB DDR3, 8 GB eMMC, 100M Ethernet (RTL8201F), USB-C OTG + USB-A host, microSD, no video out. Debug console is UART4 (ttyS4) @ 1500000n8.
The board boots on mainline-style native U-Boot v2025.04 (binman + rkbin rk3308 DDR/BL31/miniloader) from eMMC — no vendor U-Boot and no Android boot image. SPL → U-Boot 2025.04 → standard
boot.scr→booti→ Armbian. Verified end to end on hardware.KERNEL_TARGET="current,edge"(current = 6.18, edge = 7.0).What's in the PR (all verified on hardware)
dmic-codeccapture cardPDM, I2S TX and I2S RX share the GPIO2 audio pins and are mutually exclusive;
novaconfigenforces this. The mainline kernel has no runtime DT-overlay interface, so these are toggled via boot-time overlays thatnovaconfiggenerates and compiles on the device.Verification highlights
mmc0: new HS200 MMC card, bus clock 52 → 147 MHz.card: pdmmics [pdm-mics], captured from a PDM mic.rockchip-i2s-tdm), clocks come up, pins are muxed to thei2s_8ch_0function; the RX path captures an external I2S master (RP2040/Pico, 1 kHz sine, 48 kHz/16-bit) viaarecord -D hw:i2srxtest.Flashing
The image is written to the on-board eMMC over USB (Rockchip rockusb/maskrom), not
dd'd to an SD card. Enter download mode (hold BOOT while plugging in USB-C, orsudo reboot loaderfrom a running system), then:Windows users can use RKDevTool. Luckfox wiki: https://wiki.luckfox.com/Luckfox-Nova/Flash-image/
Notes
Summary by CodeRabbit
New Features
Bug Fixes