Re: [PATCH net-next v8 08/11] r8152: split rtl8156_enable/up/down into per-generation functions

[sashiko-bot@kernel.org]

> rtl8156_enable(), rtl8156_up() and rtl8156_down() had accumulated
> tp->version range checks to cover RTL8156, RTL8156B, and
> RTL8157/RTL8159, including an early bail-out in rtl8156_enable()
> that skipped most of the function body for RTL8157/RTL8159. The
> result hides which settings belong to which chip, so chip-specific
> register differences cannot be reviewed, and cannot be added without
> touching the other families.
> 
> Give each generation its own enable callback: rtl8156_enable() for
> RTL8156 (VER_10/11), rtl8156b_enable() for RTL8156B (VER_12/13/15),
> and rtl8157_enable() for RTL8157/RTL8159 (VER_16/17). RTL8157/RTL8159
> also get their own rtl8157_up()/rtl8157_down(); RTL8156 and RTL8156B
> still share rtl8156_up()/rtl8156_down().
> 
> The four chips share most of the up/down sequence, so move it into
> r8156_exit_oob() and r8156_enter_oob(), as RTL8152B/8153 do.
> Likewise, rtl8156_enable() and rtl8156b_enable() share the idle
> spin-down and L1 NYET steps, so move those into small helpers. The
> FC patch restart is also done by rtl8153_enable(), so its helper,
> rtl_fc_patch_restart(), is shared with it.
> 
> Each extracted function reproduces what its chips already did, with
> the guards that were true for them inlined and the ones that were
> false dropped as unreachable.
> 
> No functional change intended.
> 
> Signed-off-by: Chih Kai Hsu <hsu.chih.kai@realtek.com>

Sashiko has reviewed this patch and found no issues.

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/20261110032416.16968-1-nic_swsd@realtek.com?part=8