Message ID | 1653387228-28110-1-git-send-email-quic_kriskura@quicinc.com |
---|---|
Headers | show |
Series | USB DWC3 host wake up support from system suspend | expand |
On Tue, May 24, 2022 at 03:43:48PM +0530, Krishna Kurapati wrote: > From: Sandeep Maheswaram <quic_c_sanm@quicinc.com> > > If dwc3 is wakeup capable, keep the power domain always ON so that > wakeup from system suspend can be supported. Otherwise, keep the > power domain ON only during runtime suspend to support wakeup from > runtime suspend. > > Signed-off-by: Sandeep Maheswaram <quic_c_sanm@quicinc.com> > Signed-off-by: Krishna Kurapati <quic_kriskura@quicinc.com> > --- > drivers/usb/dwc3/dwc3-qcom.c | 24 +++++++++++++++++------- > 1 file changed, 17 insertions(+), 7 deletions(-) > > diff --git a/drivers/usb/dwc3/dwc3-qcom.c b/drivers/usb/dwc3/dwc3-qcom.c > index 5d5db62..7bffb31 100644 > --- a/drivers/usb/dwc3/dwc3-qcom.c > +++ b/drivers/usb/dwc3/dwc3-qcom.c > @@ -17,6 +17,7 @@ > #include <linux/of_platform.h> > #include <linux/platform_device.h> > #include <linux/phy/phy.h> > +#include <linux/pm_domain.h> > #include <linux/usb/of.h> > #include <linux/reset.h> > #include <linux/iopoll.h> > @@ -760,12 +761,13 @@ dwc3_qcom_create_urs_usb_platdev(struct device *dev) > > static int dwc3_qcom_probe(struct platform_device *pdev) > { > - struct device_node *np = pdev->dev.of_node; > - struct device *dev = &pdev->dev; > - struct dwc3_qcom *qcom; > - struct resource *res, *parent_res = NULL; > - int ret, i; > - bool ignore_pipe_clk; > + struct device_node *np = pdev->dev.of_node; > + struct device *dev = &pdev->dev; > + struct dwc3_qcom *qcom; > + struct resource *res, *parent_res = NULL; > + int ret, i; > + bool ignore_pipe_clk; > + struct generic_pm_domain *genpd; > > qcom = devm_kzalloc(&pdev->dev, sizeof(*qcom), GFP_KERNEL); > if (!qcom) > @@ -774,6 +776,8 @@ static int dwc3_qcom_probe(struct platform_device *pdev) > platform_set_drvdata(pdev, qcom); > qcom->dev = &pdev->dev; > > + genpd = pd_to_genpd(qcom->dev->pm_domain); > + > if (has_acpi_companion(dev)) { > qcom->acpi_pdata = acpi_device_get_match_data(dev); > if (!qcom->acpi_pdata) { > @@ -881,7 +885,13 @@ static int dwc3_qcom_probe(struct platform_device *pdev) > if (ret) > goto interconnect_exit; > > - device_init_wakeup(&pdev->dev, 1); > + if (device_can_wakeup(&qcom->dwc3->dev)) { > + genpd->flags |= GENPD_FLAG_ALWAYS_ON; > + device_init_wakeup(&pdev->dev, true); > + } else { > + genpd->flags |= GENPD_FLAG_RPM_ALWAYS_ON; > + } > + Can you please write a comment here explaining why we do this? commit description has some details already. Thanks, Pavan
Hi Krishna, On Tue, May 24, 2022 at 03:43:47PM +0530, Krishna Kurapati wrote: > From: Sandeep Maheswaram <quic_c_sanm@quicinc.com> > > Configure DP/DM line interrupts based on the USB2 device attached to > the root hub port. When HS/FS device is connected, configure the DP line > as falling edge to detect both disconnect and remote wakeup scenarios. When > LS device is connected, configure DM line as falling edge to detect both > disconnect and remote wakeup. When no device is connected, configure both > DP and DM lines as rising edge to detect HS/HS/LS device connect scenario. > > Signed-off-by: Sandeep Maheswaram <quic_c_sanm@quicinc.com> > Reviewed-by: Matthias Kaehlcke <mka@chromium.org> > Signed-off-by: Krishna Kurapati <quic_kriskura@quicinc.com> > --- > drivers/usb/dwc3/dwc3-qcom.c | 74 ++++++++++++++++++++++++++++++++++++++------ > 1 file changed, 65 insertions(+), 9 deletions(-) > > diff --git a/drivers/usb/dwc3/dwc3-qcom.c b/drivers/usb/dwc3/dwc3-qcom.c > index 7352124..5d5db62 100644 > --- a/drivers/usb/dwc3/dwc3-qcom.c > +++ b/drivers/usb/dwc3/dwc3-qcom.c > @@ -20,7 +20,8 @@ > #include <linux/usb/of.h> > #include <linux/reset.h> > #include <linux/iopoll.h> > - > +#include <linux/usb/hcd.h> > +#include <linux/usb.h> > #include "core.h" > > /* USB QSCRATCH Hardware registers */ > @@ -296,11 +297,37 @@ static void dwc3_qcom_interconnect_exit(struct dwc3_qcom *qcom) > icc_put(qcom->icc_path_apps); > } > > -static void dwc3_qcom_enable_wakeup_irq(int irq) > +enum usb_device_speed dwc3_qcom_update_usb2_speed(struct dwc3_qcom *qcom) > +{ > + struct dwc3 *dwc = platform_get_drvdata(qcom->dwc3); > + struct usb_hcd *hcd = platform_get_drvdata(dwc->xhci); > + struct usb_device *udev; > + enum usb_device_speed usb2_speed; > + > + /* > + * It is possible to query the speed of all children of > + * USB2.0 root hub via usb_hub_for_each_child(). DWC3 code > + * currently supports only 1 port per controller. So > + * this is sufficient. > + */ > + udev = usb_hub_find_child(hcd->self.root_hub, 1); > + > + if (udev) > + usb2_speed = udev->speed; > + else > + usb2_speed = USB_SPEED_UNKNOWN; > + > + return usb2_speed; > +} > + Can you rename this function since we are not updating anything here. %s/dwc3_qcom_update_usb2_speed/dwc3_qcom_get_usb2_speed > +static void dwc3_qcom_enable_wakeup_irq(int irq, unsigned int polarity) > { > if (!irq) > return; > > + if (polarity) > + irq_set_irq_type(irq, polarity); > + > enable_irq(irq); > enable_irq_wake(irq); > } > @@ -316,24 +343,53 @@ static void dwc3_qcom_disable_wakeup_irq(int irq) > > static void dwc3_qcom_disable_interrupts(struct dwc3_qcom *qcom) > { > - dwc3_qcom_disable_wakeup_irq(qcom->hs_phy_irq); > + enum usb_device_speed usb2_speed = dwc3_qcom_update_usb2_speed(qcom); I am sorry for misleading you on this. Looks like caching usb2_speed in suspend routine dwc3_qcom struct has a benefit. Can we please change to your previous patch-set style? Thanks, Pavan
On Tue, May 24, 2022 at 03:43:46PM +0530, Krishna Kurapati wrote: > From: Sandeep Maheswaram <quic_c_sanm@quicinc.com> > > Adding helper functions to enable,disable wake irqs to make > the code simple and readable. > > Signed-off-by: Sandeep Maheswaram <quic_c_sanm@quicinc.com> > Reviewed-by: Matthias Kaehlcke <mka@chromium.org> > Signed-off-by: Krishna Kurapati <quic_kriskura@quicinc.com> > --- > drivers/usb/dwc3/dwc3-qcom.c | 58 ++++++++++++++++++++------------------------ > 1 file changed, 26 insertions(+), 32 deletions(-) > > diff --git a/drivers/usb/dwc3/dwc3-qcom.c b/drivers/usb/dwc3/dwc3-qcom.c > index 6cba990..7352124 100644 > --- a/drivers/usb/dwc3/dwc3-qcom.c > +++ b/drivers/usb/dwc3/dwc3-qcom.c > @@ -296,50 +296,44 @@ static void dwc3_qcom_interconnect_exit(struct dwc3_qcom *qcom) > icc_put(qcom->icc_path_apps); > } > > +static void dwc3_qcom_enable_wakeup_irq(int irq) > +{ > + if (!irq) > + return; > + > + enable_irq(irq); > + enable_irq_wake(irq); > +} > + > +static void dwc3_qcom_disable_wakeup_irq(int irq) > +{ > + if (!irq) > + return; > + > + disable_irq_wake(irq); > + disable_irq_nosync(irq); > +} > + > static void dwc3_qcom_disable_interrupts(struct dwc3_qcom *qcom) > { > - if (qcom->hs_phy_irq) { > - disable_irq_wake(qcom->hs_phy_irq); > - disable_irq_nosync(qcom->hs_phy_irq); > - } > + dwc3_qcom_disable_wakeup_irq(qcom->hs_phy_irq); > > - if (qcom->dp_hs_phy_irq) { > - disable_irq_wake(qcom->dp_hs_phy_irq); > - disable_irq_nosync(qcom->dp_hs_phy_irq); > - } > + dwc3_qcom_disable_wakeup_irq(qcom->dp_hs_phy_irq); > > - if (qcom->dm_hs_phy_irq) { > - disable_irq_wake(qcom->dm_hs_phy_irq); > - disable_irq_nosync(qcom->dm_hs_phy_irq); > - } > + dwc3_qcom_disable_wakeup_irq(qcom->dm_hs_phy_irq); > > - if (qcom->ss_phy_irq) { > - disable_irq_wake(qcom->ss_phy_irq); > - disable_irq_nosync(qcom->ss_phy_irq); > - } > + dwc3_qcom_disable_wakeup_irq(qcom->ss_phy_irq); > } > > static void dwc3_qcom_enable_interrupts(struct dwc3_qcom *qcom) > { > - if (qcom->hs_phy_irq) { > - enable_irq(qcom->hs_phy_irq); > - enable_irq_wake(qcom->hs_phy_irq); > - } > + dwc3_qcom_enable_wakeup_irq(qcom->hs_phy_irq); > > - if (qcom->dp_hs_phy_irq) { > - enable_irq(qcom->dp_hs_phy_irq); > - enable_irq_wake(qcom->dp_hs_phy_irq); > - } > + dwc3_qcom_enable_wakeup_irq(qcom->dp_hs_phy_irq); > > - if (qcom->dm_hs_phy_irq) { > - enable_irq(qcom->dm_hs_phy_irq); > - enable_irq_wake(qcom->dm_hs_phy_irq); > - } > + dwc3_qcom_enable_wakeup_irq(qcom->dm_hs_phy_irq); > > - if (qcom->ss_phy_irq) { > - enable_irq(qcom->ss_phy_irq); > - enable_irq_wake(qcom->ss_phy_irq); > - } > + dwc3_qcom_enable_wakeup_irq(qcom->ss_phy_irq); > } > > static int dwc3_qcom_suspend(struct dwc3_qcom *qcom) > -- > 2.7.4 > Looks good to me. Feel free to add Reviewed-by: Pavankumar Kondeti <quic_pkondeti@quicinc.com> Thanks, Pavan
On Tue, May 24, 2022 at 03:43:45PM +0530, Krishna Kurapati wrote: > From: Sandeep Maheswaram <quic_c_sanm@quicinc.com> > > Check wakeup-source property for dwc3 core node to set the > wakeup capability. Drop the device_init_wakeup call from > runtime suspend and resume. > > If the dwc3 is wakeup capable, don't power down the USB PHY(s). > The glue drivers are expected to take care of configuring the > additional wakeup settings if needed based on the dwc3 wakeup > capability status. In some SOC designs, powering off the PHY is > resulting in higher leakage, so this patch save power on such boards. > > Signed-off-by: Sandeep Maheswaram <quic_c_sanm@quicinc.com> > Signed-off-by: Krishna Kurapati <quic_kriskura@quicinc.com> > --- > drivers/usb/dwc3/core.c | 26 ++++++++++++-------------- > 1 file changed, 12 insertions(+), 14 deletions(-) > > diff --git a/drivers/usb/dwc3/core.c b/drivers/usb/dwc3/core.c > index 5734219..4a7c82e 100644 > --- a/drivers/usb/dwc3/core.c > +++ b/drivers/usb/dwc3/core.c > @@ -1782,6 +1782,7 @@ static int dwc3_probe(struct platform_device *pdev) > > platform_set_drvdata(pdev, dwc); > dwc3_cache_hwparams(dwc); > + device_init_wakeup(&pdev->dev, of_property_read_bool(dev->of_node, "wakeup-source")); > > spin_lock_init(&dwc->lock); > mutex_init(&dwc->mutex); > @@ -1943,11 +1944,6 @@ static int dwc3_suspend_common(struct dwc3 *dwc, pm_message_t msg) > dwc3_core_exit(dwc); > break; > case DWC3_GCTL_PRTCAP_HOST: > - if (!PMSG_IS_AUTO(msg)) { > - dwc3_core_exit(dwc); > - break; > - } > - > /* Let controller to suspend HSPHY before PHY driver suspends */ > if (dwc->dis_u2_susphy_quirk || > dwc->dis_enblslpm_quirk) { > @@ -1962,6 +1958,11 @@ static int dwc3_suspend_common(struct dwc3 *dwc, pm_message_t msg) > > phy_pm_runtime_put_sync(dwc->usb2_generic_phy); > phy_pm_runtime_put_sync(dwc->usb3_generic_phy); > + > + if (!PMSG_IS_AUTO(msg)) { > + if (!device_can_wakeup(dwc->dev)) > + dwc3_core_exit(dwc); > + } > break; > case DWC3_GCTL_PRTCAP_OTG: > /* do nothing during runtime_suspend */ > @@ -2005,11 +2006,12 @@ static int dwc3_resume_common(struct dwc3 *dwc, pm_message_t msg) > break; > case DWC3_GCTL_PRTCAP_HOST: > if (!PMSG_IS_AUTO(msg)) { > - ret = dwc3_core_init_for_resume(dwc); > - if (ret) > - return ret; > - dwc3_set_prtcap(dwc, DWC3_GCTL_PRTCAP_HOST); > - break; > + if (!device_can_wakeup(dwc->dev)) { > + ret = dwc3_core_init_for_resume(dwc); > + if (ret) > + return ret; > + dwc3_set_prtcap(dwc, DWC3_GCTL_PRTCAP_HOST); > + } > } > /* Restore GUSB2PHYCFG bits that were modified in suspend */ > reg = dwc3_readl(dwc->regs, DWC3_GUSB2PHYCFG(0)); > @@ -2081,8 +2083,6 @@ static int dwc3_runtime_suspend(struct device *dev) > if (ret) > return ret; > > - device_init_wakeup(dev, true); > - > return 0; > } > > @@ -2091,8 +2091,6 @@ static int dwc3_runtime_resume(struct device *dev) > struct dwc3 *dwc = dev_get_drvdata(dev); > int ret; > > - device_init_wakeup(dev, false); > - > ret = dwc3_resume_common(dwc, PMSG_AUTO_RESUME); > if (ret) > return ret; > -- > 2.7.4 > Looks good to me. Feel free to add Reviewed-by: Pavankumar Kondeti <quic_pkondeti@quicinc.com> Thanks, Pavan
On Tue, May 24, 2022 at 03:43:47PM +0530, Krishna Kurapati wrote: > From: Sandeep Maheswaram <quic_c_sanm@quicinc.com> > > Configure DP/DM line interrupts based on the USB2 device attached to > the root hub port. When HS/FS device is connected, configure the DP line > as falling edge to detect both disconnect and remote wakeup scenarios. When > LS device is connected, configure DM line as falling edge to detect both > disconnect and remote wakeup. When no device is connected, configure both > DP and DM lines as rising edge to detect HS/HS/LS device connect scenario. > > Signed-off-by: Sandeep Maheswaram <quic_c_sanm@quicinc.com> > Reviewed-by: Matthias Kaehlcke <mka@chromium.org> With the addition of dwc3_qcom_update_usb2_speed() the patch has changed a bit since my review, I'm generally on board with the direction this is going, but please remove my R-B tag for now. > Signed-off-by: Krishna Kurapati <quic_kriskura@quicinc.com> > --- > drivers/usb/dwc3/dwc3-qcom.c | 74 ++++++++++++++++++++++++++++++++++++++------ > 1 file changed, 65 insertions(+), 9 deletions(-) > > diff --git a/drivers/usb/dwc3/dwc3-qcom.c b/drivers/usb/dwc3/dwc3-qcom.c > index 7352124..5d5db62 100644 > --- a/drivers/usb/dwc3/dwc3-qcom.c > +++ b/drivers/usb/dwc3/dwc3-qcom.c > @@ -20,7 +20,8 @@ > #include <linux/usb/of.h> > #include <linux/reset.h> > #include <linux/iopoll.h> > - > +#include <linux/usb/hcd.h> > +#include <linux/usb.h> > #include "core.h" > > /* USB QSCRATCH Hardware registers */ > @@ -296,11 +297,37 @@ static void dwc3_qcom_interconnect_exit(struct dwc3_qcom *qcom) > icc_put(qcom->icc_path_apps); > } > > -static void dwc3_qcom_enable_wakeup_irq(int irq) > +enum usb_device_speed dwc3_qcom_update_usb2_speed(struct dwc3_qcom *qcom) > +{ > + struct dwc3 *dwc = platform_get_drvdata(qcom->dwc3); > + struct usb_hcd *hcd = platform_get_drvdata(dwc->xhci); > + struct usb_device *udev; > + enum usb_device_speed usb2_speed; nit: dwc and hcd are indented to align, udev and usb2_speed not. Either align the indentation or not, but don't be inconsistent. My vote goes for removing the alignment. > + > + /* > + * It is possible to query the speed of all children of > + * USB2.0 root hub via usb_hub_for_each_child(). DWC3 code > + * currently supports only 1 port per controller. So > + * this is sufficient. > + */ > + udev = usb_hub_find_child(hcd->self.root_hub, 1); > + > + if (udev) > + usb2_speed = udev->speed; > + else > + usb2_speed = USB_SPEED_UNKNOWN; > + > + return usb2_speed; nit: you could get away without 'usb2_speed': if (!udev) return USB_SPEED_UNKNOWN; return udev->speed;
On Tue, May 24, 2022 at 03:43:45PM +0530, Krishna Kurapati wrote: > From: Sandeep Maheswaram <quic_c_sanm@quicinc.com> > > Check wakeup-source property for dwc3 core node to set the > wakeup capability. Drop the device_init_wakeup call from > runtime suspend and resume. > > If the dwc3 is wakeup capable, don't power down the USB PHY(s). > The glue drivers are expected to take care of configuring the > additional wakeup settings if needed based on the dwc3 wakeup > capability status. In some SOC designs, powering off the PHY is > resulting in higher leakage, so this patch save power on such boards. > > Signed-off-by: Sandeep Maheswaram <quic_c_sanm@quicinc.com> > Signed-off-by: Krishna Kurapati <quic_kriskura@quicinc.com> Reviewed-by: Matthias Kaehlcke <mka@chromium.org>