Skip to contentSkip to conversation

PCI: layerscape: Add support for Link down notification

PATCH Superseded Frank Li 2 participants 7 messages Jul 19, 2023

Subsystems (5)

v3 of 4 Superseded Latest is v4 →

lore
Share & tools
lore

Frank Li

author maintainer
#0 lore raw

PATCH v3 1/2 PCI: layerscape: Add support for Link down notification

Add support to pass Link down notification to Endpoint function driver so that the LINK_DOWN event can be processed by the function.

---
Change from v2 to v3
- none
Change from v1 to v2
- move pci_epc_linkdown() after dev_dbg()

 drivers/pci/controller/dwc/pci-layerscape-ep.c | 1 +
 1 file changed, 1 insertion(+)
drivers/pci/controller/dwc/pci-layerscape-ep.c +1 -0
index de4c1758a6c3..e0969ff2ddf7 100644
@@ -89,6 +89,7 @@ static irqreturn_t ls_pcie_ep_event_handler(int irq, void *dev_id)
 		dev_dbg(pci->dev, "Link up\n");
 	} else if (val & PEX_PF0_PME_MES_DR_LDD) {
 		dev_dbg(pci->dev, "Link down\n");
+		pci_epc_linkdown(pci->ep.epc);
 	} else if (val & PEX_PF0_PME_MES_DR_HRD) {
 		dev_dbg(pci->dev, "Hot reset\n");
 	}
Signature · 1 line
2.34.1
author maintainer
#1 lore raw

PATCH v3 2/2 PCI: layerscape: Add the workaround for lost link capabilities during reset.

From: Xiaowei Bao <[email protected]>

A workaround for the issue where the PCI Express Endpoint (EP) controller loses the values of the Maximum Link Width and Supported Link Speed from the Link Capabilities Register, which initially configured by the Reset Configuration Word (RCW) during a link-down or hot reset event.

Fixes a805770d8a22 ("PCI: layerscape: Add EP mode support")

---
change from v2 to v3
- fix subject typo capabilities
change from v1 to v2:
- add comments at restore register
- add fixes tag
- dw_pcie_writew_dbi to dw_pcie_writel_dbi

 .../pci/controller/dwc/pci-layerscape-ep.c    | 21 ++++++++++++++++++-
 1 file changed, 20 insertions(+), 1 deletion(-)
drivers/pci/controller/dwc/pci-layerscape-ep.c +20 -1
index e0969ff2ddf7..39dbd911c3f8 100644
@@ -45,6 +45,7 @@ struct ls_pcie_ep {
 	struct pci_epc_features		*ls_epc;
 	const struct ls_pcie_ep_drvdata *drvdata;
 	int				irq;
+	u32				lnkcap;
 	bool				big_endian;
 };
 
@@ -73,6 +74,7 @@ static irqreturn_t ls_pcie_ep_event_handler(int irq, void *dev_id)
 	struct ls_pcie_ep *pcie = dev_id;
 	struct dw_pcie *pci = pcie->pci;
 	u32 val, cfg;
+	u8 offset;
 
 	val = ls_lut_readl(pcie, PEX_PF0_PME_MES_DR);
 	ls_lut_writel(pcie, PEX_PF0_PME_MES_DR, val);
@@ -81,12 +83,25 @@ static irqreturn_t ls_pcie_ep_event_handler(int irq, void *dev_id)
 		return IRQ_NONE;
 
 	if (val & PEX_PF0_PME_MES_DR_LUD) {
+
+		offset = dw_pcie_find_capability(pci, PCI_CAP_ID_EXP);
+
+		/*
+		 * The values of the Maximum Link Width and Supported Link
+		 * Speed from the Link Capabilities Register will be lost
+		 * during link down or hot reset. Restore initial value
+		 * that configured by the Reset Configuration Word (RCW).
+		 */
+		dw_pcie_dbi_ro_wr_en(pci);
+		dw_pcie_writel_dbi(pci, offset + PCI_EXP_LNKCAP, pcie->lnkcap);
+		dw_pcie_dbi_ro_wr_dis(pci);
+
 		cfg = ls_lut_readl(pcie, PEX_PF0_CONFIG);
 		cfg |= PEX_PF0_CFG_READY;
 		ls_lut_writel(pcie, PEX_PF0_CONFIG, cfg);
 		dw_pcie_ep_linkup(&pci->ep);
 
-		dev_dbg(pci->dev, "Link up\n");
+		dev_err(pci->dev, "Link up\n");
 	} else if (val & PEX_PF0_PME_MES_DR_LDD) {
 		dev_dbg(pci->dev, "Link down\n");
 		pci_epc_linkdown(pci->ep.epc);
@@ -216,6 +231,7 @@ static int __init ls_pcie_ep_probe(struct platform_device *pdev)
 	struct ls_pcie_ep *pcie;
 	struct pci_epc_features *ls_epc;
 	struct resource *dbi_base;
+	u8 offset;
 	int ret;
 
 	pcie = devm_kzalloc(dev, sizeof(*pcie), GFP_KERNEL);
@@ -252,6 +268,9 @@ static int __init ls_pcie_ep_probe(struct platform_device *pdev)
 
 	platform_set_drvdata(pdev, pcie);
 
+	offset = dw_pcie_find_capability(pci, PCI_CAP_ID_EXP);
+	pcie->lnkcap = dw_pcie_readl_dbi(pci, offset + PCI_EXP_LNKCAP);
+
 	ret = dw_pcie_ep_init(&pci->ep);
 	if (ret)
 		return ret;
Signature · 1 line
2.34.1
author maintainer
#2 lore raw

Re: [PATCH v3 2/2] PCI: layerscape: Add the workaround for lost link capabilities during reset.

Quoted 57 lines · Frank Li

From: Xiaowei Bao <[email protected]>

A workaround for the issue where the PCI Express Endpoint (EP) controller loses the values of the Maximum Link Width and Supported Link Speed from the Link Capabilities Register, which initially configured by the Reset Configuration Word (RCW) during a link-down or hot reset event.

Fixes a805770d8a22 ("PCI: layerscape: Add EP mode support")

---
change from v2 to v3
- fix subject typo capabilities
change from v1 to v2:
- add comments at restore register
- add fixes tag
- dw_pcie_writew_dbi to dw_pcie_writel_dbi

 .../pci/controller/dwc/pci-layerscape-ep.c    | 21 ++++++++++++++++++-
 1 file changed, 20 insertions(+), 1 deletion(-)
drivers/pci/controller/dwc/pci-layerscape-ep.c +12 -0
index e0969ff2ddf7..39dbd911c3f8 100644
@@ -45,6 +45,7 @@ struct ls_pcie_ep {
 	struct pci_epc_features		*ls_epc;
 	const struct ls_pcie_ep_drvdata *drvdata;
 	int				irq;
+	u32				lnkcap;
 	bool				big_endian;
 };
 
@@ -73,6 +74,7 @@ static irqreturn_t ls_pcie_ep_event_handler(int irq, void *dev_id)
 	struct ls_pcie_ep *pcie = dev_id;
 	struct dw_pcie *pci = pcie->pci;
 	u32 val, cfg;
+	u8 offset;
 
 	val = ls_lut_readl(pcie, PEX_PF0_PME_MES_DR);
 	ls_lut_writel(pcie, PEX_PF0_PME_MES_DR, val);
@@ -81,12 +83,25 @@ static irqreturn_t ls_pcie_ep_event_handler(int irq, void *dev_id)
 		return IRQ_NONE;
 
 	if (val & PEX_PF0_PME_MES_DR_LUD) {
+
+		offset = dw_pcie_find_capability(pci, PCI_CAP_ID_EXP);
+
+		/*
+		 * The values of the Maximum Link Width and Supported Link
+		 * Speed from the Link Capabilities Register will be lost
+		 * during link down or hot reset. Restore initial value
+		 * that configured by the Reset Configuration Word (RCW).
+		 */
+		dw_pcie_dbi_ro_wr_en(pci);

drivers/pci/controller/dwc/pci-layerscape-ep.c static irqreturn_t ls_pcie_ep_event_handler(int irq, void *dev_id) · line 104

Quoted 10 lines
drivers/pci/controller/dwc/pci-layerscape-ep.c +4 -1
+		dw_pcie_writel_dbi(pci, offset + PCI_EXP_LNKCAP, pcie->lnkcap);
+		dw_pcie_dbi_ro_wr_dis(pci);
+
 		cfg = ls_lut_readl(pcie, PEX_PF0_CONFIG);
 		cfg |= PEX_PF0_CFG_READY;
 		ls_lut_writel(pcie, PEX_PF0_CONFIG, cfg);
 		dw_pcie_ep_linkup(&pci->ep);
 
-		dev_dbg(pci->dev, "Link up\n");
+		dev_err(pci->dev, "Link up\n");

Sorry, Just found that. mistake merge a debug code. It should be dev_dbg here, will send update patch soon

Frank

Quoted 24 lines
drivers/pci/controller/dwc/pci-layerscape-ep.c +4 -0
 	} else if (val & PEX_PF0_PME_MES_DR_LDD) {
 		dev_dbg(pci->dev, "Link down\n");
 		pci_epc_linkdown(pci->ep.epc);
@@ -216,6 +231,7 @@ static int __init ls_pcie_ep_probe(struct platform_device *pdev)
 	struct ls_pcie_ep *pcie;
 	struct pci_epc_features *ls_epc;
 	struct resource *dbi_base;
+	u8 offset;
 	int ret;
 
 	pcie = devm_kzalloc(dev, sizeof(*pcie), GFP_KERNEL);
@@ -252,6 +268,9 @@ static int __init ls_pcie_ep_probe(struct platform_device *pdev)
 
 	platform_set_drvdata(pdev, pcie);
 
+	offset = dw_pcie_find_capability(pci, PCI_CAP_ID_EXP);
+	pcie->lnkcap = dw_pcie_readl_dbi(pci, offset + PCI_EXP_LNKCAP);
+
 	ret = dw_pcie_ep_init(&pci->ep);
 	if (ret)
 		return ret;
Signature · 1 line
2.34.1
#3 lore raw
In reply to Frank Li
Quoted 1 line

Add support to pass …

Why did you omit a cover letter for the discussed patch series once more?

Do you care for consequences according to message threading?

Regards,
Markus

author maintainer
#4 lore raw
Quoted 3 lines · Markus Elfring
Quoted 1 line

Add support to pass …

Why did you omit a cover letter for the discussed patch series once more?

Your comments is
"Will a cover letter become helpful also for the presented small patch series?"

According to my understand it is optional. I don't think cover letter will help this case. Patch 1 and 2 is that independent at all.

I sent these together just because easy to test once.

Maintainer can pick any one individually.

Cover letter just annoise people here.

Frank

Quoted 5 lines

Do you care for consequences according to message threading?

Regards,
Markus

author maintainer
#6 lore raw
Quoted 6 lines · Markus Elfring
Quoted 1 line

Cover letter just annoise people here.

How do you think about advices from another information source?

See also:
https://kernelnewbies.org/PatchSeries

"You may like to include a cover letter with your patch series."

Generally, I think cover letter will be needed only if it really help reviewer to get main idea about patches.

Such as my on going pathes(with cover letter):
https://lore.kernel.org/imx/ZLglBiSz0meJm5os@lizhi-Precision-Tower-5810/T/#t

Similar case without(cover leter) and accepted.
https://lore.kernel.org/imx/20230719063425.GE151430@dragon/T/#t

I don't think cover letter real help reviewer to review these two patches.

I more like to get "real problem"(such as comments about "typo").

It is just waste time to discuss if need add cover letter here.

Frank

Quoted 3 lines

Regards,
Markus