Re: [PATCH net-next v6 01/10] r8152: split r8156_init per chip and factor out wait_autoload_done

[sashiko-bot@kernel.org]

> RTL8156, RTL8156B, RTL8157 and RTL8159 shared a single r8156_init(),
> gated by a dense set of tp->version range checks. That made the init
> sequence hard to follow.
> 
> Split r8156_init() into one function per chip family:
> - r8156_init  (VER_10/11)
> - r8156b_init (VER_12/13/15)
> - r8157_init  (VER_16)
> - r8159_init  (VER_17)
> 
> Also factor the repeated "poll PLA_BOOT_CTRL for AUTOLOAD_DONE" loop
> into wait_autoload_done(), reusing read_poll_timeout() instead of the
> 500 * msleep(20) busy loops. r8152b_init(), r8153_init(),
> r8153b_init(), r8153c_init(), r8153b_ups_en() and r8153c_ups_en() are
> converted to use it too.
> 
> wait_autoload_done() only reports failure (nonzero) when
> RTL8152_INACCESSIBLE is set; a plain AUTOLOAD_DONE timeout is logged
> but otherwise treated as success, same as every one of the loops it
> replaces, which only bailed out early on RTL8152_INACCESSIBLE and
> otherwise fell through to the rest of the init sequence. The two
> failure modes are not equivalent: once RTL8152_INACCESSIBLE is set,
> generic_ocp_read()/generic_ocp_write() short-circuit to -ENODEV
> without touching the device, so returning early costs nothing. A plain
> timeout means the device is still reachable, so skipping the rest of
> the bring-up would leave real hardware state half-configured.
> 
> No functional change intended.
> 
> Signed-off-by: Chih Kai Hsu <hsu.chih.kai@realtek.com>

Sashiko has reviewed this patch and found no issues. It looks great!

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/20261008081556.47901-1-nic_swsd@realtek.com?part=1