> -----Original Message----- > From: Bjorn Helgaas <helgaas@xxxxxxxxxx> > Sent: 2022年6月16日 23:56 > To: Hongxing Zhu <hongxing.zhu@xxxxxxx> > Cc: Lucas Stach <l.stach@xxxxxxxxxxxxxx>; Rob Herring <robh+dt@xxxxxxxxxx>; > Mark Brown <broonie@xxxxxxxxxx>; Lorenzo Pieralisi > <lorenzo.pieralisi@xxxxxxx>; Fabio Estevam <festevam@xxxxxxxxx>; > Francesco Dolcini <francesco.dolcini@xxxxxxxxxxx>; linux-pci@xxxxxxxxxxxxxxx; > linux-arm-kernel@xxxxxxxxxxxxxxxxxxx; linux-kernel@xxxxxxxxxxxxxxx; > kernel@xxxxxxxxxxxxxx; dl-linux-imx <linux-imx@xxxxxxx>; Bjorn Helgaas > <bhelgaas@xxxxxxxxxx> > Subject: Re: [PATCH v12 09/13] PCI: imx6: Turn off regulator when system is in > suspend mode > > On Thu, Jun 16, 2022 at 02:23:29AM +0000, Hongxing Zhu wrote: > > > -----Original Message----- > > > From: Bjorn Helgaas <helgaas@xxxxxxxxxx> > > > Sent: 2022年6月16日 7:31 > > > To: Hongxing Zhu <hongxing.zhu@xxxxxxx>; Lucas Stach > > > <l.stach@xxxxxxxxxxxxxx>; Rob Herring <robh+dt@xxxxxxxxxx>; Mark > > > Brown <broonie@xxxxxxxxxx>; Lorenzo Pieralisi > > > <lorenzo.pieralisi@xxxxxxx>; Fabio Estevam <festevam@xxxxxxxxx>; > > > Francesco Dolcini <francesco.dolcini@xxxxxxxxxxx> > > > Cc: linux-pci@xxxxxxxxxxxxxxx; linux-arm-kernel@xxxxxxxxxxxxxxxxxxx; > > > linux-kernel@xxxxxxxxxxxxxxx; kernel@xxxxxxxxxxxxxx; dl-linux-imx > > > <linux-imx@xxxxxxx>; Bjorn Helgaas <bhelgaas@xxxxxxxxxx> > > > Subject: Re: [PATCH v12 09/13] PCI: imx6: Turn off regulator when > > > system is in suspend mode > > > > > > On Wed, Jun 15, 2022 at 06:15:47PM -0500, Bjorn Helgaas wrote: > > > > From: Richard Zhu <hongxing.zhu@xxxxxxx> > > > > > > > > The driver should undo any enables it did itself. The regulator > > > > disable shouldn't be basing decisions on regulator_is_enabled(). > > > > > > > > Move the regulator_disable to the suspend function, turn off > > > > regulator when the system is in suspend mode. > > > > > > > > To keep the balance of the regulator usage counter, disable the > > > > regulator in shutdown. > > > > > > > > Link: > > > > https://eur01.safelinks.protection.outlook.com/?url=https%3A%2F%2F > > > > lore > > > > .kernel.org%2Fr%2F1655189942-12678-6-git-send-email-hongxing.z&am > p > > > > ;d > > > at > > > > > > > > a=05%7C01%7Chongxing.zhu%40nxp.com%7C8dea46d908d34cb4825808da4 > > > f270eaf% > > > > > > > > 7C686ea1d3bc2b4c6fa92cd99c5c301635%7C0%7C0%7C6379093264156415 > > > 35%7CUnkn > > > > > > > > own%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1 > > > haWwi > > > > > > > > LCJXVCI6Mn0%3D%7C3000%7C%7C%7C&sdata=tqzlcCmtSkUNxNzJ0Fr5o > > > CR3X2CTH > > > > 4dwUJS4NXt5KRc%3D&reserved=0 > > > > hu@xxxxxxx > > > > Signed-off-by: Richard Zhu <hongxing.zhu@xxxxxxx> > > > > Signed-off-by: Bjorn Helgaas <bhelgaas@xxxxxxxxxx> > > > > --- > > > > drivers/pci/controller/dwc/pci-imx6.c | 19 +++++++------------ > > > > 1 file changed, 7 insertions(+), 12 deletions(-) > > > > > > > > diff --git a/drivers/pci/controller/dwc/pci-imx6.c > > > > b/drivers/pci/controller/dwc/pci-imx6.c > > > > index 6eddd0b5f628..537b8a2e0e3b 100644 > > > > --- a/drivers/pci/controller/dwc/pci-imx6.c > > > > +++ b/drivers/pci/controller/dwc/pci-imx6.c > > > > @@ -670,8 +670,6 @@ static void imx6_pcie_clk_disable(struct > > > > imx6_pcie > > > > *imx6_pcie) > > > > > > > > static void imx6_pcie_assert_core_reset(struct imx6_pcie > > > > *imx6_pcie) { > > > > - struct device *dev = imx6_pcie->pci->dev; > > > > - > > > > switch (imx6_pcie->drvdata->variant) { > > > > case IMX7D: > > > > case IMX8MQ: > > > > @@ -702,14 +700,6 @@ static void > > > > imx6_pcie_assert_core_reset(struct > > > imx6_pcie *imx6_pcie) > > > > break; > > > > } > > > > > > > > - if (imx6_pcie->vpcie && regulator_is_enabled(imx6_pcie->vpcie) > 0) > { > > > > - int ret = regulator_disable(imx6_pcie->vpcie); > > > > - > > > > - if (ret) > > > > - dev_err(dev, "failed to disable vpcie regulator: %d\n", > > > > - ret); > > > > - } > > > > - > > > > /* Some boards don't have PCIe reset GPIO. */ > > > > if (gpio_is_valid(imx6_pcie->reset_gpio)) > > > > gpio_set_value_cansleep(imx6_pcie->reset_gpio, > > > > @@ -722,7 +712,7 @@ static int > > > > imx6_pcie_deassert_core_reset(struct > > > imx6_pcie *imx6_pcie) > > > > struct device *dev = pci->dev; > > > > int ret, err; > > > > > > > > - if (imx6_pcie->vpcie && !regulator_is_enabled(imx6_pcie->vpcie)) { > > > > + if (imx6_pcie->vpcie) { > > > > ret = regulator_enable(imx6_pcie->vpcie); > > > > if (ret) { > > > > dev_err(dev, "failed to enable vpcie regulator: %d\n", @@ > > > -796,7 > > > > +786,7 @@ static int imx6_pcie_deassert_core_reset(struct > > > > +imx6_pcie > > > *imx6_pcie) > > > > return 0; > > > > > > > > err_clks: > > > > - if (imx6_pcie->vpcie && regulator_is_enabled(imx6_pcie->vpcie) > 0) > { > > > > + if (imx6_pcie->vpcie) { > > > > ret = regulator_disable(imx6_pcie->vpcie); > > > > if (ret) > > > > dev_err(dev, "failed to disable vpcie regulator: %d\n", @@ > > > -1023,6 > > > > +1013,9 @@ static int imx6_pcie_suspend_noirq(struct device *dev) > > > > break; > > > > } > > > > > > > > + if (imx6_pcie->vpcie) > > > > + regulator_disable(imx6_pcie->vpcie); > > > > > > This is a little bit ugly because imx6_pcie_suspend_noirq() and > > > imx6_pcie_resume_noirq() are not symmetric. > > > > > > We call regulator_disable() directly here in > > > imx6_pcie_suspend_noirq(), but the corresponding regulator_enable() is > buried in imx6_pcie_deassert_core_reset(). > > > > > > It would be nicer if the suspend and resume paths looked more > > > similar, as > > > rockchip_pcie_suspend_noirq() and rockchip_pcie_resume_noirq() do, > > > for example. > > Yes, it is. The regulator_disable()/regulator_enable() are not invoked > > symmetrically. > > > > In the original codes, the regulator_disable() is contained in > > imx6_pcie_assert_core_reset() and with _eanbled() check. > > And regulator_enable() is embedded in imx6_pcie_deassert_core_reset(). > > Both the regulator disable and enable are invoked in resume before. > > I move the regulator_disable to suspend mode refer to Lucas' comments. > > > > Refer to the current situation, how about to move the > > regulator_disable() out of imx6_pcie_deassert_core_reset() too? > > I think it's important to tweak this closer to the pattern of other drivers like > exynos and rockchip. > > exynos_pcie_resume_noirq() calls exynos_pcie_host_init(). > imx6_pcie_resume_noirq() is VERY similar to imx6_pcie_host_init(), and I think > it should call imx6_pcie_host_init() instead of duplicating that code. The only > real difference is that imx6_pcie_host_init() calls imx6_setup_phy_mpll() while > imx6_pcie_resume_noirq() does not. > I suspect that's a bug. > > I don't think "regulator_enable(imx6_pcie->vpcie)" really has anything to do > with imx6_pcie_deassert_core_reset() and it should be moved out. > The typical pattern is to enable regulators in *_probe() and > *_resume_noirq() and disable them in *_remove() and *_suspend_noirq(). > I think you should do the same unless there's some reason you can't. > > Can you give this a try, starting with this v12 posting? Thanks a lot for your inspiration. Exactly, these codes are almost same. The host_init can be invoked directly in resume. I would re-change the #12/13 of v12, and send the v13 later. Best Regards Richard Zhu > > Bjorn
