PCI: dwc: Fix resume failure if no EP is connected at some platforms

[PATCH v2] PCI: dwc: Fix resume failure if no EP is connected at some platforms

Posted by Richard Zhu 1 year, 4 months ago

The dw_pcie_suspend_noirq() function currently returns success directly
if no endpoint (EP) device is connected. However, on some platforms, power
loss occurs during suspend, causing dw_resume() to do nothing in this case.
This results in a system halt because the DWC controller is not initialized
after power-on during resume.

Change call to deinit() in suspend and init() at resume regardless of
whether there are EP device connections or not. It is not harmful to
perform deinit() and init() again for the no power-off case, and it keeps
the code simple and consistent in logic.

Fixes: 4774faf854f5 ("PCI: dwc: Implement generic suspend/resume functionality")
Signed-off-by: Richard Zhu <hongxing.zhu@nxp.com>
Reviewed-by: Frank Li <Frank.Li@nxp.com>
---
 .../pci/controller/dwc/pcie-designware-host.c | 30 +++++++++----------
 1 file changed, 15 insertions(+), 15 deletions(-)

diff --git a/drivers/pci/controller/dwc/pcie-designware-host.c b/drivers/pci/controller/dwc/pcie-designware-host.c
index a0822d5371bc5..cb8c3c2bcc790 100644
--- a/drivers/pci/controller/dwc/pcie-designware-host.c
+++ b/drivers/pci/controller/dwc/pcie-designware-host.c
@@ -933,23 +933,23 @@ int dw_pcie_suspend_noirq(struct dw_pcie *pci)
 	if (dw_pcie_readw_dbi(pci, offset + PCI_EXP_LNKCTL) & PCI_EXP_LNKCTL_ASPM_L1)
 		return 0;
 
-	if (dw_pcie_get_ltssm(pci) <= DW_PCIE_LTSSM_DETECT_ACT)
-		return 0;
-
-	if (pci->pp.ops->pme_turn_off)
-		pci->pp.ops->pme_turn_off(&pci->pp);
-	else
-		ret = dw_pcie_pme_turn_off(pci);
+	if (dw_pcie_get_ltssm(pci) > DW_PCIE_LTSSM_DETECT_ACT) {
+		/* Only send out PME_TURN_OFF when PCIE link is up */
+		if (pci->pp.ops->pme_turn_off)
+			pci->pp.ops->pme_turn_off(&pci->pp);
+		else
+			ret = dw_pcie_pme_turn_off(pci);
 
-	if (ret)
-		return ret;
+		if (ret)
+			return ret;
 
-	ret = read_poll_timeout(dw_pcie_get_ltssm, val, val == DW_PCIE_LTSSM_L2_IDLE,
-				PCIE_PME_TO_L2_TIMEOUT_US/10,
-				PCIE_PME_TO_L2_TIMEOUT_US, false, pci);
-	if (ret) {
-		dev_err(pci->dev, "Timeout waiting for L2 entry! LTSSM: 0x%x\n", val);
-		return ret;
+		ret = read_poll_timeout(dw_pcie_get_ltssm, val, val == DW_PCIE_LTSSM_L2_IDLE,
+					PCIE_PME_TO_L2_TIMEOUT_US/10,
+					PCIE_PME_TO_L2_TIMEOUT_US, false, pci);
+		if (ret) {
+			dev_err(pci->dev, "Timeout waiting for L2 entry! LTSSM: 0x%x\n", val);
+			return ret;
+		}
 	}
 
 	if (pci->pp.ops->deinit)
-- 
2.37.1

Re: [PATCH v2] PCI: dwc: Fix resume failure if no EP is connected at some platforms

Posted by Bjorn Helgaas 1 year, 1 month ago

On Mon, Jul 22, 2024 at 02:15:13PM +0800, Richard Zhu wrote:
> The dw_pcie_suspend_noirq() function currently returns success directly
> if no endpoint (EP) device is connected. However, on some platforms, power
> loss occurs during suspend, causing dw_resume() to do nothing in this case.
> This results in a system halt because the DWC controller is not initialized
> after power-on during resume.
> 
> Change call to deinit() in suspend and init() at resume regardless of
> whether there are EP device connections or not. It is not harmful to
> perform deinit() and init() again for the no power-off case, and it keeps
> the code simple and consistent in logic.
> ...

> -	ret = read_poll_timeout(dw_pcie_get_ltssm, val, val == DW_PCIE_LTSSM_L2_IDLE,
> -				PCIE_PME_TO_L2_TIMEOUT_US/10,
> -				PCIE_PME_TO_L2_TIMEOUT_US, false, pci);
> -	if (ret) {
> -		dev_err(pci->dev, "Timeout waiting for L2 entry! LTSSM: 0x%x\n", val);
> -		return ret;
> +		ret = read_poll_timeout(dw_pcie_get_ltssm, val, val == DW_PCIE_LTSSM_L2_IDLE,
> +					PCIE_PME_TO_L2_TIMEOUT_US/10,
> +					PCIE_PME_TO_L2_TIMEOUT_US, false, pci);
> +		if (ret) {
> +			dev_err(pci->dev, "Timeout waiting for L2 entry! LTSSM: 0x%x\n", val);
> +			return ret;
> +		}

Not related to this patch, but what's the reason for waiting for the
link to enter L2?  There are a few other drivers that do this, but
most don't.  Is there something else the driver needs to do after the
link is in L2?

>  	}
>  
>  	if (pci->pp.ops->deinit)
> -- 
> 2.37.1
>

Re: [PATCH v2] PCI: dwc: Fix resume failure if no EP is connected at some platforms

Posted by Bjorn Helgaas 1 year, 1 month ago

On Mon, Jul 22, 2024 at 02:15:13PM +0800, Richard Zhu wrote:
> The dw_pcie_suspend_noirq() function currently returns success directly
> if no endpoint (EP) device is connected. However, on some platforms, power
> loss occurs during suspend, causing dw_resume() to do nothing in this case.
> This results in a system halt because the DWC controller is not initialized
> after power-on during resume.

dw_resume() doesn't exist.  What function did you mean?

System halt?  In dw_pcie_resume_noirq()?  What causes the halt?  A
NULL pointer dereference?  A CPU hang because a read of some
controller register never completes?  Feels a little hand-wavy.

Another comment below.

> Change call to deinit() in suspend and init() at resume regardless of
> whether there are EP device connections or not. It is not harmful to
> perform deinit() and init() again for the no power-off case, and it keeps
> the code simple and consistent in logic.
> 
> Fixes: 4774faf854f5 ("PCI: dwc: Implement generic suspend/resume functionality")
> Signed-off-by: Richard Zhu <hongxing.zhu@nxp.com>
> Reviewed-by: Frank Li <Frank.Li@nxp.com>
> ---
>  .../pci/controller/dwc/pcie-designware-host.c | 30 +++++++++----------
>  1 file changed, 15 insertions(+), 15 deletions(-)
> 
> diff --git a/drivers/pci/controller/dwc/pcie-designware-host.c b/drivers/pci/controller/dwc/pcie-designware-host.c
> index a0822d5371bc5..cb8c3c2bcc790 100644
> --- a/drivers/pci/controller/dwc/pcie-designware-host.c
> +++ b/drivers/pci/controller/dwc/pcie-designware-host.c
> @@ -933,23 +933,23 @@ int dw_pcie_suspend_noirq(struct dw_pcie *pci)
>  	if (dw_pcie_readw_dbi(pci, offset + PCI_EXP_LNKCTL) & PCI_EXP_LNKCTL_ASPM_L1)
>  		return 0;
>  
> -	if (dw_pcie_get_ltssm(pci) <= DW_PCIE_LTSSM_DETECT_ACT)
> -		return 0;
> -
> -	if (pci->pp.ops->pme_turn_off)
> -		pci->pp.ops->pme_turn_off(&pci->pp);
> -	else
> -		ret = dw_pcie_pme_turn_off(pci);
> +	if (dw_pcie_get_ltssm(pci) > DW_PCIE_LTSSM_DETECT_ACT) {
> +		/* Only send out PME_TURN_OFF when PCIE link is up */
> +		if (pci->pp.ops->pme_turn_off)
> +			pci->pp.ops->pme_turn_off(&pci->pp);
> +		else
> +			ret = dw_pcie_pme_turn_off(pci);

This looks possibly racy since the link can go down at any point.

> -	if (ret)
> -		return ret;
> +		if (ret)
> +			return ret;
>  
> -	ret = read_poll_timeout(dw_pcie_get_ltssm, val, val == DW_PCIE_LTSSM_L2_IDLE,
> -				PCIE_PME_TO_L2_TIMEOUT_US/10,
> -				PCIE_PME_TO_L2_TIMEOUT_US, false, pci);
> -	if (ret) {
> -		dev_err(pci->dev, "Timeout waiting for L2 entry! LTSSM: 0x%x\n", val);
> -		return ret;
> +		ret = read_poll_timeout(dw_pcie_get_ltssm, val, val == DW_PCIE_LTSSM_L2_IDLE,
> +					PCIE_PME_TO_L2_TIMEOUT_US/10,
> +					PCIE_PME_TO_L2_TIMEOUT_US, false, pci);
> +		if (ret) {
> +			dev_err(pci->dev, "Timeout waiting for L2 entry! LTSSM: 0x%x\n", val);
> +			return ret;
> +		}
>  	}
>  
>  	if (pci->pp.ops->deinit)
> -- 
> 2.37.1
>

Re: [PATCH v2] PCI: dwc: Fix resume failure if no EP is connected at some platforms

Posted by Krzysztof Wilczyński 1 year, 1 month ago

Hello,

> The dw_pcie_suspend_noirq() function currently returns success directly
> if no endpoint (EP) device is connected. However, on some platforms, power
> loss occurs during suspend, causing dw_resume() to do nothing in this case.
> This results in a system halt because the DWC controller is not initialized
> after power-on during resume.
> 
> Change call to deinit() in suspend and init() at resume regardless of
> whether there are EP device connections or not. It is not harmful to
> perform deinit() and init() again for the no power-off case, and it keeps
> the code simple and consistent in logic.

Applied to controller/dwc, thank you!

[01/01] PCI: dwc: Fix resume failure if no EP is connected at some platforms
        https://git.kernel.org/pci/pci/c/ec008c493c25

	Krzysztof