[3/4] usb: host: xhci-plat: revert "usb: host: xhci-plat: enable clk in resume timing"

[3/4] usb: host: xhci-plat: revert "usb: host: xhci-plat: enable clk in resume timing"

[Greg Kroah-Hartman]

On Thu, Mar 08, 2018 at 05:17:16PM +0200, Mathias Nyman wrote:
> From: Yoshihiro Shimoda <yoshihiro.shimoda.uh@renesas.com>
> 
> This patch reverts the commit 835e4241e714 ("usb: host: xhci-plat:
> enable clk in resume timing") because this driver also has runtime PM
> and the commit 560869100b99 ("clk: renesas: cpg-mssr: Restore module
> clocks during resume") will restore the clock on R-Car H3 environment.
> 
> If the xhci_plat_suspend() disables the clk, the system cannot enable
> the clk in resume like the following behavior:
> 
> < In resume >
>  - genpd_resume_noirq() runs and enable the clk (enable_count = 1)
>  - cpg_mssr_resume_noirq() restores the clk register.
>   -- Since the clk was disabled in suspend, cpg_mssr_resume_noirq()
>      will disable the clk and keep the enable_count.
>  - Even if xhci_plat_resume() calls clk_prepare_enable(), since
>    the enable_count is 1, the clk will be not enabled.
> 
> After this patch is applied, the cpg-mssr driver will save the clk
> as enable, so the clk will be enabled in resume.
> 
> Fixes: 835e4241e714 ("usb: host: xhci-plat: enable clk in resume timing")
> Signed-off-by: Yoshihiro Shimoda <yoshihiro.shimoda.uh@renesas.com>
> Signed-off-by: Mathias Nyman <mathias.nyman@linux.intel.com>

This really should go to 4.15-stable, right?  I'll go add the tag if
needed, but want your confirmation first.

thanks,

greg k-h
---
To unsubscribe from this list: send the line "unsubscribe linux-usb" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

[3/4] usb: host: xhci-plat: revert "usb: host: xhci-plat: enable clk in resume timing"

[Yoshihiro Shimoda]

Hi Greg,

> From: Greg KH, Sent: Saturday, March 10, 2018 2:08 AM
> 
> On Fri, Mar 09, 2018 at 01:48:37AM +0000, Yoshihiro Shimoda wrote:
> > Hi Greg,
> >
> > > From: Greg KH, Sent: Friday, March 9, 2018 2:06 AM
> > >
> > > On Thu, Mar 08, 2018 at 05:17:16PM +0200, Mathias Nyman wrote:
> > > > From: Yoshihiro Shimoda <yoshihiro.shimoda.uh@renesas.com>
< snip >
> > > > Fixes: 835e4241e714 ("usb: host: xhci-plat: enable clk in resume timing")
> > > > Signed-off-by: Yoshihiro Shimoda <yoshihiro.shimoda.uh@renesas.com>
> > > > Signed-off-by: Mathias Nyman <mathias.nyman@linux.intel.com>
> > >
> > > This really should go to 4.15-stable, right?  I'll go add the tag if
> > > needed, but want your confirmation first.
> >
> > Since 4.15 is not longterm release, I think Renesas will not consume 4.15.
> 
> But it fixes a bug in 4.15, so someone might care :)

:)

> > So, I don't think this really should go to 4.15-stable.
> > In such case, should I not add Fixes: tag? If not, I'll be careful in the future.
> 
> No, the fixes tag was great.  I'll just leave it as-is.

I got it. Thank you for your comment!

Best regards,
Yoshihiro Shimoda

> thanks,
> 
> greg k-h
---
To unsubscribe from this list: send the line "unsubscribe linux-usb" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

[3/4] usb: host: xhci-plat: revert "usb: host: xhci-plat: enable clk in resume timing"

[Greg Kroah-Hartman]

On Fri, Mar 09, 2018 at 01:48:37AM +0000, Yoshihiro Shimoda wrote:
> Hi Greg,
> 
> > From: Greg KH, Sent: Friday, March 9, 2018 2:06 AM
> > 
> > On Thu, Mar 08, 2018 at 05:17:16PM +0200, Mathias Nyman wrote:
> > > From: Yoshihiro Shimoda <yoshihiro.shimoda.uh@renesas.com>
> > >
> > > This patch reverts the commit 835e4241e714 ("usb: host: xhci-plat:
> > > enable clk in resume timing") because this driver also has runtime PM
> > > and the commit 560869100b99 ("clk: renesas: cpg-mssr: Restore module
> > > clocks during resume") will restore the clock on R-Car H3 environment.
> > >
> > > If the xhci_plat_suspend() disables the clk, the system cannot enable
> > > the clk in resume like the following behavior:
> > >
> > > < In resume >
> > >  - genpd_resume_noirq() runs and enable the clk (enable_count = 1)
> > >  - cpg_mssr_resume_noirq() restores the clk register.
> > >   -- Since the clk was disabled in suspend, cpg_mssr_resume_noirq()
> > >      will disable the clk and keep the enable_count.
> > >  - Even if xhci_plat_resume() calls clk_prepare_enable(), since
> > >    the enable_count is 1, the clk will be not enabled.
> > >
> > > After this patch is applied, the cpg-mssr driver will save the clk
> > > as enable, so the clk will be enabled in resume.
> > >
> > > Fixes: 835e4241e714 ("usb: host: xhci-plat: enable clk in resume timing")
> > > Signed-off-by: Yoshihiro Shimoda <yoshihiro.shimoda.uh@renesas.com>
> > > Signed-off-by: Mathias Nyman <mathias.nyman@linux.intel.com>
> > 
> > This really should go to 4.15-stable, right?  I'll go add the tag if
> > needed, but want your confirmation first.
> 
> Since 4.15 is not longterm release, I think Renesas will not consume 4.15.

But it fixes a bug in 4.15, so someone might care :)

> So, I don't think this really should go to 4.15-stable.
> In such case, should I not add Fixes: tag? If not, I'll be careful in the future.

No, the fixes tag was great.  I'll just leave it as-is.

thanks,

greg k-h
---
To unsubscribe from this list: send the line "unsubscribe linux-usb" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

[3/4] usb: host: xhci-plat: revert "usb: host: xhci-plat: enable clk in resume timing"

[Yoshihiro Shimoda]

Hi Greg,

> From: Greg KH, Sent: Friday, March 9, 2018 2:06 AM
> 
> On Thu, Mar 08, 2018 at 05:17:16PM +0200, Mathias Nyman wrote:
> > From: Yoshihiro Shimoda <yoshihiro.shimoda.uh@renesas.com>
> >
> > This patch reverts the commit 835e4241e714 ("usb: host: xhci-plat:
> > enable clk in resume timing") because this driver also has runtime PM
> > and the commit 560869100b99 ("clk: renesas: cpg-mssr: Restore module
> > clocks during resume") will restore the clock on R-Car H3 environment.
> >
> > If the xhci_plat_suspend() disables the clk, the system cannot enable
> > the clk in resume like the following behavior:
> >
> > < In resume >
> >  - genpd_resume_noirq() runs and enable the clk (enable_count = 1)
> >  - cpg_mssr_resume_noirq() restores the clk register.
> >   -- Since the clk was disabled in suspend, cpg_mssr_resume_noirq()
> >      will disable the clk and keep the enable_count.
> >  - Even if xhci_plat_resume() calls clk_prepare_enable(), since
> >    the enable_count is 1, the clk will be not enabled.
> >
> > After this patch is applied, the cpg-mssr driver will save the clk
> > as enable, so the clk will be enabled in resume.
> >
> > Fixes: 835e4241e714 ("usb: host: xhci-plat: enable clk in resume timing")
> > Signed-off-by: Yoshihiro Shimoda <yoshihiro.shimoda.uh@renesas.com>
> > Signed-off-by: Mathias Nyman <mathias.nyman@linux.intel.com>
> 
> This really should go to 4.15-stable, right?  I'll go add the tag if
> needed, but want your confirmation first.

Since 4.15 is not longterm release, I think Renesas will not consume 4.15.
So, I don't think this really should go to 4.15-stable.
In such case, should I not add Fixes: tag? If not, I'll be careful in the future.

Best regards,
Yoshihiro Shimoda

> thanks,
> 
> greg k-h
---
To unsubscribe from this list: send the line "unsubscribe linux-usb" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

[3/4] usb: host: xhci-plat: revert "usb: host: xhci-plat: enable clk in resume timing"

[Mathias Nyman]

From: Yoshihiro Shimoda <yoshihiro.shimoda.uh@renesas.com>

This patch reverts the commit 835e4241e714 ("usb: host: xhci-plat:
enable clk in resume timing") because this driver also has runtime PM
and the commit 560869100b99 ("clk: renesas: cpg-mssr: Restore module
clocks during resume") will restore the clock on R-Car H3 environment.

If the xhci_plat_suspend() disables the clk, the system cannot enable
the clk in resume like the following behavior:

< In resume >
 - genpd_resume_noirq() runs and enable the clk (enable_count = 1)
 - cpg_mssr_resume_noirq() restores the clk register.
  -- Since the clk was disabled in suspend, cpg_mssr_resume_noirq()
     will disable the clk and keep the enable_count.
 - Even if xhci_plat_resume() calls clk_prepare_enable(), since
   the enable_count is 1, the clk will be not enabled.

After this patch is applied, the cpg-mssr driver will save the clk
as enable, so the clk will be enabled in resume.

Fixes: 835e4241e714 ("usb: host: xhci-plat: enable clk in resume timing")
Signed-off-by: Yoshihiro Shimoda <yoshihiro.shimoda.uh@renesas.com>
Signed-off-by: Mathias Nyman <mathias.nyman@linux.intel.com>
---
 drivers/usb/host/xhci-plat.c | 11 +----------
 1 file changed, 1 insertion(+), 10 deletions(-)

diff --git a/drivers/usb/host/xhci-plat.c b/drivers/usb/host/xhci-plat.c
index 6f03830..6652e2d 100644
--- a/drivers/usb/host/xhci-plat.c
+++ b/drivers/usb/host/xhci-plat.c
@@ -360,7 +360,6 @@ static int __maybe_unused xhci_plat_suspend(struct device *dev)
 {
 	struct usb_hcd	*hcd = dev_get_drvdata(dev);
 	struct xhci_hcd	*xhci = hcd_to_xhci(hcd);
-	int ret;
 
 	/*
 	 * xhci_suspend() needs `do_wakeup` to know whether host is allowed
@@ -370,12 +369,7 @@ static int __maybe_unused xhci_plat_suspend(struct device *dev)
 	 * reconsider this when xhci_plat_suspend enlarges its scope, e.g.,
 	 * also applies to runtime suspend.
 	 */
-	ret = xhci_suspend(xhci, device_may_wakeup(dev));
-
-	if (!device_may_wakeup(dev) && !IS_ERR(xhci->clk))
-		clk_disable_unprepare(xhci->clk);
-
-	return ret;
+	return xhci_suspend(xhci, device_may_wakeup(dev));
 }
 
 static int __maybe_unused xhci_plat_resume(struct device *dev)
@@ -384,9 +378,6 @@ static int __maybe_unused xhci_plat_resume(struct device *dev)
 	struct xhci_hcd	*xhci = hcd_to_xhci(hcd);
 	int ret;
 
-	if (!device_may_wakeup(dev) && !IS_ERR(xhci->clk))
-		clk_prepare_enable(xhci->clk);
-
 	ret = xhci_priv_resume_quirk(hcd);
 	if (ret)
 		return ret;