[PATCH v2 12/14] cxl/pci: Add error handler for CXL PCIe port RAS errors

Terry Bowman posted 14 patches 1 month ago
There is a newer version of this series
[PATCH v2 12/14] cxl/pci: Add error handler for CXL PCIe port RAS errors
Posted by Terry Bowman 1 month ago
Introduce correctable and uncorrectable CXL PCIe port handlers.

Use the PCIe port's device object to find the matching port or
downstream port in the CXL topology. The matching port or downstream
port will include the cached RAS register block.

Invoke the existing __cxl_handle_ras() with the RAS registers as a
parameter. __cxl_handle_ras() will log the RAS errors (if present)
and clear the RAS status.

Signed-off-by: Terry Bowman <terry.bowman@amd.com>
---
 drivers/cxl/core/pci.c | 59 ++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 59 insertions(+)

diff --git a/drivers/cxl/core/pci.c b/drivers/cxl/core/pci.c
index bb2fd7d04c4f..adb184d346ae 100644
--- a/drivers/cxl/core/pci.c
+++ b/drivers/cxl/core/pci.c
@@ -772,6 +772,65 @@ static void cxl_disable_rch_root_ints(struct cxl_dport *dport)
 	writel(aer_cmd, aer_base + PCI_ERR_ROOT_COMMAND);
 }
 
+static int match_uport(struct device *dev, const void *data)
+{
+	struct device *uport_dev = (struct device *)data;
+	struct cxl_port *port;
+
+	if (!is_cxl_port(dev))
+		return 0;
+
+	port = to_cxl_port(dev);
+
+	return port->uport_dev == uport_dev;
+}
+
+static void __iomem *cxl_pci_port_ras(struct pci_dev *pdev)
+{
+	struct cxl_port *port __free(put_cxl_port) = NULL;
+	void __iomem *ras_base = NULL;
+
+	if (!pdev)
+		return NULL;
+
+	if ((pci_pcie_type(pdev) == PCI_EXP_TYPE_ROOT_PORT) ||
+	    (pci_pcie_type(pdev) == PCI_EXP_TYPE_DOWNSTREAM)) {
+		struct cxl_dport *dport;
+
+		port = find_cxl_port(&pdev->dev, &dport);
+		ras_base = dport ? dport->regs.ras : NULL;
+	} else if (pci_pcie_type(pdev) == PCI_EXP_TYPE_UPSTREAM) {
+		struct device *port_dev;
+
+		port_dev = bus_find_device(&cxl_bus_type, NULL, &pdev->dev,
+					   match_uport);
+		if (!port_dev)
+			return NULL;
+
+		port = to_cxl_port(port_dev);
+		ras_base = port ? port->uport_regs.ras : NULL;
+	}
+
+	return ras_base;
+}
+
+static void cxl_port_cor_error_detected(struct pci_dev *pdev)
+{
+	void __iomem *ras_base = cxl_pci_port_ras(pdev);
+
+	__cxl_handle_cor_ras(&pdev->dev, ras_base);
+}
+
+static bool cxl_port_error_detected(struct pci_dev *pdev)
+{
+	void __iomem *ras_base = cxl_pci_port_ras(pdev);
+	bool ue;
+
+	ue = __cxl_handle_ras(&pdev->dev, ras_base);
+
+	return ue;
+}
+
 void cxl_uport_init_ras_reporting(struct cxl_port *port)
 {
 	/* uport may have more than 1 downstream EP. Check if already mapped. */
-- 
2.34.1
Re: [PATCH v2 12/14] cxl/pci: Add error handler for CXL PCIe port RAS errors
Posted by Jonathan Cameron 3 weeks, 4 days ago
On Fri, 25 Oct 2024 16:03:03 -0500
Terry Bowman <terry.bowman@amd.com> wrote:

> Introduce correctable and uncorrectable CXL PCIe port handlers.
> 
> Use the PCIe port's device object to find the matching port or
> downstream port in the CXL topology. The matching port or downstream
> port will include the cached RAS register block.
> 
> Invoke the existing __cxl_handle_ras() with the RAS registers as a
> parameter. __cxl_handle_ras() will log the RAS errors (if present)
> and clear the RAS status.
> 
> Signed-off-by: Terry Bowman <terry.bowman@amd.com>
> ---
>  drivers/cxl/core/pci.c | 59 ++++++++++++++++++++++++++++++++++++++++++
>  1 file changed, 59 insertions(+)
> 
> diff --git a/drivers/cxl/core/pci.c b/drivers/cxl/core/pci.c
> index bb2fd7d04c4f..adb184d346ae 100644
> --- a/drivers/cxl/core/pci.c
> +++ b/drivers/cxl/core/pci.c
> @@ -772,6 +772,65 @@ static void cxl_disable_rch_root_ints(struct cxl_dport *dport)
>  	writel(aer_cmd, aer_base + PCI_ERR_ROOT_COMMAND);
>  }
>  
> +static int match_uport(struct device *dev, const void *data)
> +{
> +	struct device *uport_dev = (struct device *)data;
> +	struct cxl_port *port;
> +
> +	if (!is_cxl_port(dev))
> +		return 0;
> +
> +	port = to_cxl_port(dev);
> +
> +	return port->uport_dev == uport_dev;
> +}
> +
> +static void __iomem *cxl_pci_port_ras(struct pci_dev *pdev)
> +{
> +	struct cxl_port *port __free(put_cxl_port) = NULL;
> +	void __iomem *ras_base = NULL;
> +
> +	if (!pdev)
> +		return NULL;
> +
> +	if ((pci_pcie_type(pdev) == PCI_EXP_TYPE_ROOT_PORT) ||
> +	    (pci_pcie_type(pdev) == PCI_EXP_TYPE_DOWNSTREAM)) {
> +		struct cxl_dport *dport;
> +
> +		port = find_cxl_port(&pdev->dev, &dport);

Scope of port is messy as the constructor and destructor
are not well associated. I'd drag a copy into each leg so they can
remain closer to each other.
Or don't use __free() as it's not adding much here.

> +		ras_base = dport ? dport->regs.ras : NULL;	
> +	} else if (pci_pcie_type(pdev) == PCI_EXP_TYPE_UPSTREAM) {
> +		struct device *port_dev;
> +
> +		port_dev = bus_find_device(&cxl_bus_type, NULL, &pdev->dev,
> +					   match_uport);
> +		if (!port_dev)
> +			return NULL;
> +
> +		port = to_cxl_port(port_dev);
> +		ras_base = port ? port->uport_regs.ras : NULL;
> +	}
> +
> +	return ras_base;
> +}
> +
> +static void cxl_port_cor_error_detected(struct pci_dev *pdev)
> +{
> +	void __iomem *ras_base = cxl_pci_port_ras(pdev);
> +
> +	__cxl_handle_cor_ras(&pdev->dev, ras_base);
> +}
> +
> +static bool cxl_port_error_detected(struct pci_dev *pdev)
> +{
> +	void __iomem *ras_base = cxl_pci_port_ras(pdev);
> +	bool ue;
> +
> +	ue = __cxl_handle_ras(&pdev->dev, ras_base);
> +
> +	return ue;
> +}
> +
>  void cxl_uport_init_ras_reporting(struct cxl_port *port)
>  {
>  	/* uport may have more than 1 downstream EP. Check if already mapped. */