On Thu, Jun 09, 2022 at 06:20:16AM +0000, Hongxing Zhu wrote: > > -----Original Message----- > > From: Bjorn Helgaas <helgaas@xxxxxxxxxx> > > Sent: 2022年6月9日 2:58 > > To: Hongxing Zhu <hongxing.zhu@xxxxxxx> > > Cc: l.stach@xxxxxxxxxxxxxx; bhelgaas@xxxxxxxxxx; robh+dt@xxxxxxxxxx; > > broonie@xxxxxxxxxx; lorenzo.pieralisi@xxxxxxx; jingoohan1@xxxxxxxxx; > > festevam@xxxxxxxxx; francesco.dolcini@xxxxxxxxxxx; > > linux-pci@xxxxxxxxxxxxxxx; linux-arm-kernel@xxxxxxxxxxxxxxxxxxx; > > linux-kernel@xxxxxxxxxxxxxxx; kernel@xxxxxxxxxxxxxx; dl-linux-imx > > <linux-imx@xxxxxxx> > > Subject: Re: [PATCH v9 7/8] PCI: imx6: Move the phy driver callbacks to the > > proper places > > > > On Fri, May 06, 2022 at 09:47:08AM +0800, Richard Zhu wrote: > > > To make it more reasonable, move the phy_power_on/phy_init callbacks > > > to the proper places. > > > - move the phy_power_on() out of imx6_pcie_clk_enable(). > > > - move the phy_init() out of imx6_pcie_deassert_core_reset(). > > > > I'm not sure what "make it more reasonable" is telling me. In > > subject line and commit log, please say something more specific > > than "the proper places." > > > > It's probably more important to say where they are moving *to* > > than where they're moving *out of*. > Thanks for your comments. > In another review loop listed below, Lucas used said that it's not > good to hide PHY init in imx6_pcie_assert_core_reset() So, I make a > try to move the phy_init() out of imx6_pcie_assert_core_reset(). > And move phy_power_on() out of imx6_pcie_clk_enable() accordingly. > https://patchwork.kernel.org/project/linux-pci/patch/1646289275-17813-1-git-send-email-hongxing.zhu@xxxxxxx/ > Okay, I would specific that they are moving *to* later. > > > > In order to save power consumption, turn off the clocks and > > > regulators when the imx6_pcie_host_init() return error. > > > > Is the power savings the *reason* for this change? I can't tell > > from the commit log. > > The error handling of the imx6_pcie_host_init() is not mentioned in > the subject. Should I split these changes into two patches? If they can be split, they probably should be split.