Add HMCAD15XX support and Sharkbyte board support

[PopPaul2021]

PR Description

Add support for the HMCAD1511 and HMCAD1520 high-speed ADCs
with LVDS outputs, along with board support for the Sharkbyte dual-ADC data acquisition platform.

PR Type

• Bug fix (a change that fixes an issue)
• New feature (a change that adds new functionality)
• Breaking change (a change that affects other repos or cause CIs to fail)

PR Checklist

• I have conducted a self-review of my own code changes
• I have compiled my changes, including the documentation
• I have tested the changes on the relevant hardware
• I have updated the documentation outside this repo accordingly
• I have provided links for the relevant upstream lore

[nunojsa]

Given that this requires much more changes, stopping the review with that one comment

[nunojsa]

AFAICT, There's no reason to support new high speed converters using the old framework! These days we should iio backends which is also upstream and therefore makes possible to upstream this part.

[github-actions]

LLM review

This series adds the HMCAD1511/HMCAD1520 high-speed ADC driver and Zynq Sharkbyte platform support, plus an ADF5355 clock-flag fix.

run: 24855823681

---

a291d2b83a - iio: adc: add support for HMCAD15xx high-speed ADC

HMCAD1520 14-bit precision mode register bug:

The HMCAD1511 and HMCAD1520 have different register layouts for 0x31 (OPERATION_MODE). On the HMCAD1520, bit D[3] is precision_mode — it must be set to 1 to enable 14-bit operation. The driver sends HMCAD15XX_OP_MODE_QUAD_CH = 0x0204 for all 14-bit paths, which leaves precision_mode=0, putting the HMCAD1520 into 12-bit high-speed quad mode instead of 14-bit precision mode.

Per the HMCAD1520 datasheet (Table 17, p. 17; Table 57, p. 35), the correct value for 14-bit precision mode is 0x0308 (precision_mode=1, high_speed_mode=0, clk_divide=8).

The zynq-sharkbyte_14b.dts overlay is directly affected.

Fixup patch: 0003-fixup-iio-adc-add-support-for-HMCAD15xx-high-speed-A.patch

---

0fb61f6f5f - iio: frequency: adf5355: set CLK_GET_RATE_NOCACHE for clock output

Double-free of devm-allocated memory (coccicheck):

clk_priv is allocated with devm_kzalloc() (line 975) but kfree(clk_priv) is called manually on clk_register() failure (line 1006). The devm framework will attempt to free the same pointer again on device removal, causing a double-free.

Fix: remove the kfree(clk_priv) call; devm manages the lifetime unconditionally.

Fixup patch: 0001-fixup-iio-frequency-adf5355-set-CLK_GET_RATE_NOCACHE.patch

The CLK_GET_RATE_NOCACHE flag change itself is correct.

---

f399f54493 - dt-bindings: iio: adc: add adi,hmcad15xx

HMCAD1511 resolution is 8-bit, not 12-bit:

The binding description and Kconfig both claim "The HMCAD1511 is a 12-bit ADC". Per the HMCAD1511 datasheet the device is 8-bit (ENOB ~7.9 bits, SNR ~49 dBFS). The 12-bit variant in this family is the HMCAD1512. Also, product URLs pointed to ti.com instead of analog.com.

Fixup patch: 0002-fixup-dt-bindings-iio-adc-add-adi-hmcad15xx.patch

---

CI warnings

• checkpatch _idx macro argument reuse: _idx is used in .address, .channel, and .scan_index in HMCAD15XX_CHAN(). All call sites pass compile-time constants so there are no actual side-effects; this is a false positive.
• smatch warnings at adf5355.c lines 604 and 953: pre-existing issues in the base driver, not introduced by this series.
• DTS unit_address_vs_reg and unique_unit_address warnings for zynq-sharkbyte.dts: consistent with other ADI platform device-trees; lint-level only.

---

Verification data

HMCAD1511.pdf and HMCAD1520.pdf were downloaded from analog.com and register maps extracted with camelot-py. HMCAD1512.pdf was also fetched for cross-reference of the 8-bit vs 12-bit distinction. The full kernel was compiled for ARM with zynq_sharkbyte_defconfig both before and after the fixups; no errors.

Suggested patches

Apply the suggested patches with:

cd path/to/repository
export GITHUB_TOKEN=ghp_***
apply-patches --repo=analogdevicesinc/linux 24855823681

Install instructions

The following one-liner installs the script if not present already:

grep "/apply-patches.sh" ~/.bashrc ||  { curl "https://raw.githubusercontent.com/analogdevicesinc/doctools/refs/heads/main/ci/scripts/apply-patches.sh"    -o ~/.local/bin/apply-patches.sh &&  echo "source ~/.local/bin/apply-patches.sh" >> ~/.bashrc ; source ~/.bashrc ; }

More information at AI Usage.