Hi Russell King,

Thanks for your suggestions. we’ve done some trials and investigations,
but we’d like to clarify a few points:

Moving eic7700_clks_config() into plat_dat->init and plat_dat->exit does
allow us to drop eic7700_dwmac_remove() and use devm_stmmac_pltfm_probe()
to simplify the code.

However, we don’t want clks_config() to be invoked again during
stmmac_pltfm_resume() and stmmac_pltfm_suspend(). Following your
suggestion, this means we would need to provide empty plat_dat->suspend()
and plat_dat->resume() methods.

Could you confirm whether you’re planning to add the suspend and resume
hooks into the plat_stmmacenet_data structure?
Also, regarding the cleanups you mentioned for other stmmac glue drivers,
do you have some links or reference commits so we can review the approach
you took?

Thanks!

Best regards,
Li Zhi

-----原始邮件-----
发件人: "Russell King (Oracle)" [off-list ref]
发送时间:2025-09-23 17:09:06 (星期二)
收件人: 韦尚娟 [off-list ref]
抄送: devicetree@vger.kernel.org, andrew+netdev@lunn.ch, davem@davemloft.net, edumazet@google.com, kuba@kernel.org, robh@kernel.org, krzk+dt@kernel.org, conor+dt@kernel.org, netdev@vger.kernel.org, pabeni@redhat.com, mcoquelin.stm32@gmail.com, alexandre.torgue@foss.st.com, vladimir.oltean@nxp.com, yong.liang.choong@linux.intel.com, anthony.l.nguyen@intel.com, prabhakar.mahadev-lad.rj@bp.renesas.com, jan.petrous@oss.nxp.com, jszhang@kernel.org, inochiama@gmail.com, 0x1207@gmail.com, boon.khai.ng@altera.com, linux-kernel@vger.kernel.org, linux-stm32@st-md-mailman.stormreply.com, linux-arm-kernel@lists.infradead.org, ningyu@eswincomputing.com, linmin@eswincomputing.com, lizhi2@eswincomputing.com, pinkesh.vaghela@einfochips.com
主题: Re: Re: [PATCH v7 2/2] ethernet: eswin: Add eic7700 ethernet driver

On Tue, Sep 23, 2025 at 11:06:08AM +0800, 韦尚娟 wrote:

quoted

In the current eic7700_dwmac glue driver, the regmap_read()/write()
operations(for phy_ctrl1, axi_lp_ctrl1, and the RX/TX delay registers))are 
performed directly in the probe() function. Would it be cleaner to move these
register configurations into the init() callback instead, so that they are
also reapplied during resume()?

This is a question I can't answer definitively as I don't know what
happens during a suspend on your hardware, and thus which registers
are lost / reset by the time the system resumes. So I can only give
the obvious guidance.

If the settings in the delay registers are lost over a suspend/resume
then they need to be re-initialised after resume.

-- 
RMK's Patch system: https://www.armlinux.org.uk/developer/patches/
FTTP is here! 80Mbps down 10Mbps up. Decent connectivity at last!