For zPCI devices, we have platform specific error information. The platform
firmware provides this error information to the operating system in an
architecture specific mechanism. To enable recovery from userspace for
these devices, we want to expose this error information to userspace. Add a
new device feature to expose this information.

Signed-off-by: Farhan Ali <alifm@linux.ibm.com>
---
 drivers/vfio/pci/vfio_pci_core.c |  2 ++
 drivers/vfio/pci/vfio_pci_priv.h |  9 ++++++++
 drivers/vfio/pci/vfio_pci_zdev.c | 35 ++++++++++++++++++++++++++++++++
 include/uapi/linux/vfio.h        | 16 +++++++++++++++
 4 files changed, 62 insertions(+)

diff --git a/drivers/vfio/pci/vfio_pci_core.c b/drivers/vfio/pci/vfio_pci_core.c
index 3a11e6f450f7..f677705921e6 100644
--- a/drivers/vfio/pci/vfio_pci_core.c
+++ b/drivers/vfio/pci/vfio_pci_core.c
@@ -1526,6 +1526,8 @@ int vfio_pci_core_ioctl_feature(struct vfio_device *device, u32 flags,
 		return vfio_pci_core_feature_token(vdev, flags, arg, argsz);
 	case VFIO_DEVICE_FEATURE_DMA_BUF:
 		return vfio_pci_core_feature_dma_buf(vdev, flags, arg, argsz);
+	case VFIO_DEVICE_FEATURE_ZPCI_ERROR:
+		return vfio_pci_zdev_feature_err(device, flags, arg, argsz);
 	default:
 		return -ENOTTY;
 	}
diff --git a/drivers/vfio/pci/vfio_pci_priv.h b/drivers/vfio/pci/vfio_pci_priv.h
index 27ac280f00b9..eed69926d8a1 100644
--- a/drivers/vfio/pci/vfio_pci_priv.h
+++ b/drivers/vfio/pci/vfio_pci_priv.h
@@ -89,6 +89,8 @@ int vfio_pci_info_zdev_add_caps(struct vfio_pci_core_device *vdev,
 				struct vfio_info_cap *caps);
 int vfio_pci_zdev_open_device(struct vfio_pci_core_device *vdev);
 void vfio_pci_zdev_close_device(struct vfio_pci_core_device *vdev);
+int vfio_pci_zdev_feature_err(struct vfio_device *device, u32 flags,
+			      void __user *arg, size_t argsz);
 #else
 static inline int vfio_pci_info_zdev_add_caps(struct vfio_pci_core_device *vdev,
 					      struct vfio_info_cap *caps)
@@ -103,6 +105,13 @@ static inline int vfio_pci_zdev_open_device(struct vfio_pci_core_device *vdev)
 
 static inline void vfio_pci_zdev_close_device(struct vfio_pci_core_device *vdev)
 {}
+
+static inline int vfio_pci_zdev_feature_err(struct vfio_device *device,
+					    u32 flags, void __user *arg,
+					    size_t argsz)
+{
+	return -ENODEV;
+}
 #endif
 
 static inline bool vfio_pci_is_vga(struct pci_dev *pdev)
diff --git a/drivers/vfio/pci/vfio_pci_zdev.c b/drivers/vfio/pci/vfio_pci_zdev.c
index 2be37eab9279..b9150782bafa 100644
--- a/drivers/vfio/pci/vfio_pci_zdev.c
+++ b/drivers/vfio/pci/vfio_pci_zdev.c
@@ -141,6 +141,41 @@ int vfio_pci_info_zdev_add_caps(struct vfio_pci_core_device *vdev,
 	return ret;
 }
 
+int vfio_pci_zdev_feature_err(struct vfio_device *device, u32 flags,
+			      void __user *arg, size_t argsz)
+{
+	struct vfio_device_feature_zpci_err err;
+	struct vfio_pci_core_device *vdev;
+	struct zpci_dev *zdev;
+	int head = 0;
+	int ret;
+
+	vdev = container_of(device, struct vfio_pci_core_device, vdev);
+	zdev = to_zpci(vdev->pdev);
+	if (!zdev)
+		return -ENODEV;
+
+	ret = vfio_check_feature(flags, argsz, VFIO_DEVICE_FEATURE_GET,
+				 sizeof(err));
+	if (ret != 1)
+		return ret;
+
+	mutex_lock(&zdev->pending_errs_lock);
+	if (zdev->pending_errs.count) {
+		head = zdev->pending_errs.head % ZPCI_ERR_PENDING_MAX;
+		err.pec = zdev->pending_errs.err[head].pec;
+		zdev->pending_errs.head++;
+		zdev->pending_errs.count--;
+		err.pending_errors = zdev->pending_errs.count;
+	}
+	mutex_unlock(&zdev->pending_errs_lock);
+
+	if (copy_to_user(arg, &err, sizeof(err)))
+		return -EFAULT;
+
+	return 0;
+}
+
 int vfio_pci_zdev_open_device(struct vfio_pci_core_device *vdev)
 {
 	struct zpci_dev *zdev = to_zpci(vdev->pdev);
diff --git a/include/uapi/linux/vfio.h b/include/uapi/linux/vfio.h
index ac2329f24141..89b44762ef02 100644
--- a/include/uapi/linux/vfio.h
+++ b/include/uapi/linux/vfio.h
@@ -1506,6 +1506,22 @@ struct vfio_device_feature_dma_buf {
 	struct vfio_region_dma_range dma_ranges[] __counted_by(nr_ranges);
 };
 
+/**
+ * VFIO_DEVICE_FEATURE_ZPCI_ERROR feature provides PCI error information to
+ * userspace for vfio-pci devices on s390x. On s390x PCI error recovery involves
+ * platform firmware and notification to operating system is done by
+ * architecture specific mechanism.  Exposing this information to userspace
+ * allows userspace to take appropriate actions to handle an error on the
+ * device.
+ */
+
+struct vfio_device_feature_zpci_err {
+	__u16 pec;
+	__u8 pending_errors;
+	__u8 pad;
+};
+#define VFIO_DEVICE_FEATURE_ZPCI_ERROR 12
+
 /* -------- API for Type1 VFIO IOMMU -------- */
 
 /**
-- 
2.43.0

On Thu, 2026-01-22 at 11:44 -0800, Farhan Ali wrote:
> For zPCI devices, we have platform specific error information. The platform
> firmware provides this error information to the operating system in an
> architecture specific mechanism. To enable recovery from userspace for
> these devices, we want to expose this error information to userspace. Add a
> new device feature to expose this information.
> 
> Signed-off-by: Farhan Ali <alifm@linux.ibm.com>
> ---
>  drivers/vfio/pci/vfio_pci_core.c |  2 ++
>  drivers/vfio/pci/vfio_pci_priv.h |  9 ++++++++
>  drivers/vfio/pci/vfio_pci_zdev.c | 35 ++++++++++++++++++++++++++++++++
>  include/uapi/linux/vfio.h        | 16 +++++++++++++++
>  4 files changed, 62 insertions(+)
> 
--- snip ---
>  
> +int vfio_pci_zdev_feature_err(struct vfio_device *device, u32 flags,
> +			      void __user *arg, size_t argsz)
> +{
> +	struct vfio_device_feature_zpci_err err;
> +	struct vfio_pci_core_device *vdev;
> +	struct zpci_dev *zdev;
> +	int head = 0;
> +	int ret;
> +
> +	vdev = container_of(device, struct vfio_pci_core_device, vdev);
> +	zdev = to_zpci(vdev->pdev);
> +	if (!zdev)
> +		return -ENODEV;
> +
> +	ret = vfio_check_feature(flags, argsz, VFIO_DEVICE_FEATURE_GET,
> +				 sizeof(err));
> +	if (ret != 1)
> +		return ret;
> +
> +	mutex_lock(&zdev->pending_errs_lock);
> +	if (zdev->pending_errs.count) {
> +		head = zdev->pending_errs.head % ZPCI_ERR_PENDING_MAX;
> +		err.pec = zdev->pending_errs.err[head].pec;

In the previous patch you saved the entire struct zpci_ccdf_err now you
only copy out and expose the PCI event code, though? If you do want to
only expose that the commit message should state this and the reason
for this restriction. Additionally I think the struct
vfio_device_feature_zpci_err should include a mechanism (version +
size?) to allow upgrading it to the full error information in the
future. 

Then again why not just expose the entire CCDF? It's an architected
data structure without and if you add it at the end of struct
vfio_device_feature_zpci_err and add a size you should even be able to
handle if it ever needs to grow. Of course you'd have to create a copy
of the struct to use the the uAPI types so I'd probably also add a
BUILD_BUG_ON() check on matching size. Or am I missing a reason to keep
just the PEC?

> +		zdev->pending_errs.head++;
> +		zdev->pending_errs.count--;
> +		err.pending_errors = zdev->pending_errs.count;
> +	}
> +	mutex_unlock(&zdev->pending_errs_lock);
> +
> +	if (copy_to_user(arg, &err, sizeof(err)))
> +		return -EFAULT;
> +
> +	return 0;
> +}
> +

On 1/27/2026 2:53 AM, Niklas Schnelle wrote:
> On Thu, 2026-01-22 at 11:44 -0800, Farhan Ali wrote:
>> For zPCI devices, we have platform specific error information. The platform
>> firmware provides this error information to the operating system in an
>> architecture specific mechanism. To enable recovery from userspace for
>> these devices, we want to expose this error information to userspace. Add a
>> new device feature to expose this information.
>>
>> Signed-off-by: Farhan Ali <alifm@linux.ibm.com>
>> ---
>>   drivers/vfio/pci/vfio_pci_core.c |  2 ++
>>   drivers/vfio/pci/vfio_pci_priv.h |  9 ++++++++
>>   drivers/vfio/pci/vfio_pci_zdev.c | 35 ++++++++++++++++++++++++++++++++
>>   include/uapi/linux/vfio.h        | 16 +++++++++++++++
>>   4 files changed, 62 insertions(+)
>>
> --- snip ---
>>   
>> +int vfio_pci_zdev_feature_err(struct vfio_device *device, u32 flags,
>> +			      void __user *arg, size_t argsz)
>> +{
>> +	struct vfio_device_feature_zpci_err err;
>> +	struct vfio_pci_core_device *vdev;
>> +	struct zpci_dev *zdev;
>> +	int head = 0;
>> +	int ret;
>> +
>> +	vdev = container_of(device, struct vfio_pci_core_device, vdev);
>> +	zdev = to_zpci(vdev->pdev);
>> +	if (!zdev)
>> +		return -ENODEV;
>> +
>> +	ret = vfio_check_feature(flags, argsz, VFIO_DEVICE_FEATURE_GET,
>> +				 sizeof(err));
>> +	if (ret != 1)
>> +		return ret;
>> +
>> +	mutex_lock(&zdev->pending_errs_lock);
>> +	if (zdev->pending_errs.count) {
>> +		head = zdev->pending_errs.head % ZPCI_ERR_PENDING_MAX;
>> +		err.pec = zdev->pending_errs.err[head].pec;
> In the previous patch you saved the entire struct zpci_ccdf_err now you
> only copy out and expose the PCI event code, though? If you do want to
> only expose that the commit message should state this and the reason
> for this restriction. Additionally I think the struct
> vfio_device_feature_zpci_err should include a mechanism (version +
> size?) to allow upgrading it to the full error information in the
> future.

I think having explicit version variable in the struct should be 
sufficient (the __pad can be replaced with version). I don't think we 
need explicit size variable? I have looked at some of the capability 
structures in vfio_zdev.h, as examples and so we could use a similar 
approach here when we need to extend the vfio_device_feature_zpci_err?

Though I don't see any other vfio device feature structure being 
explicitly versioned. I am open to any guidance/suggestions on the best 
practices on how to we version VFIO device feature structs.


>
> Then again why not just expose the entire CCDF? It's an architected
> data structure without and if you add it at the end of struct
> vfio_device_feature_zpci_err and add a size you should even be able to
> handle if it ever needs to grow. Of course you'd have to create a copy
> of the struct to use the the uAPI types so I'd probably also add a
> BUILD_BUG_ON() check on matching size. Or am I missing a reason to keep
> just the PEC?

I wanted to keep the information exposed to userspace minimal. The CCDF 
exposes far more information and may not be needed by userspace/VM. 
Today the PEC is sufficient for user space(QEMU) to take bubble up to a 
VM. I also wanted to avoid having a copy of the struct in 2 places.

Thanks

Farhan

>> +		zdev->pending_errs.head++;
>> +		zdev->pending_errs.count--;
>> +		err.pending_errors = zdev->pending_errs.count;
>> +	}
>> +	mutex_unlock(&zdev->pending_errs_lock);
>> +
>> +	if (copy_to_user(arg, &err, sizeof(err)))
>> +		return -EFAULT;
>> +
>> +	return 0;
>> +}
>> +