[PATCH v4 17/25] nvdimm/ocxl: Add controller dump IOCTLs

["Alastair D'Silva" <alastair@d-silva.org>]

This patch adds IOCTLs to allow userspace to request & fetch dumps
of the internal controller state.

This is useful during debugging or when a fatal error on the controller
has occurred.

The expected flow of operations are:
1. IOCTL_OCXL_PMEM_CONTROLLER_DUMP to request the controller to take
   a dump. This IOCTL will complete after the dump is available for
   collection.
2. IOCTL_OCXL_PMEM_CONTROLLER_DUMP_DATA called repeatedly to fetch
   chunks from the buffer
3. IOCTL_OCXL_PMEM_CONTROLLER_DUMP_COMPLETE to notify the controller
   that it can free any internal resources used for the dump

Signed-off-by: Alastair D'Silva <alastair@d-silva.org>
---
 drivers/nvdimm/ocxl/main.c     | 161 +++++++++++++++++++++++++++++++++
 include/uapi/nvdimm/ocxlpmem.h |  16 ++++
 2 files changed, 177 insertions(+)

diff --git a/drivers/nvdimm/ocxl/main.c b/drivers/nvdimm/ocxl/main.c
index e6be0029f658..d0db358ded43 100644
--- a/drivers/nvdimm/ocxl/main.c
+++ b/drivers/nvdimm/ocxl/main.c
@@ -566,6 +566,153 @@ static int ioctl_error_log(struct ocxlpmem *ocxlpmem,
 	return 0;
 }
 
+/**
+ * controller_dump_header_parse() - Parse the first 64 bits of the controller dump command response
+ * @ocxlpmem: the device metadata
+ * @length: out, returns the number of bytes in the response (excluding the 64 bit header)
+ */
+static int controller_dump_header_parse(struct ocxlpmem *ocxlpmem, u16 *length)
+{
+	int rc;
+	u64 val;
+	u16 data_identifier;
+	u32 data_length;
+
+	rc = ocxl_global_mmio_read64(ocxlpmem->ocxl_afu,
+				     ocxlpmem->admin_command.data_offset,
+				     OCXL_LITTLE_ENDIAN, &val);
+	if (rc)
+		return rc;
+
+	data_identifier = val >> 48;
+	data_length = val & 0xFFFF;
+
+	if (data_identifier != 0x4344) { // 'CD'
+		dev_err(&ocxlpmem->dev,
+			"Bad data identifier for error log data, expected 'CD', got '%2s' (%#x), data_length=%u\n",
+			(char *)&data_identifier,
+			(unsigned int)data_identifier, data_length);
+		return -EINVAL;
+	}
+
+	*length = data_length;
+	return 0;
+}
+
+static int ioctl_controller_dump_data(struct ocxlpmem *ocxlpmem,
+				      struct ioctl_ocxlpmem_controller_dump_data __user *uarg)
+{
+	struct ioctl_ocxlpmem_controller_dump_data args;
+	u64 __user *buf;
+	u16 i, buf_size;
+	u64 val;
+	int rc;
+
+	if (copy_from_user(&args, uarg, sizeof(args)))
+		return -EFAULT;
+
+	if (args.buf_size % sizeof(u64))
+		return -EINVAL;
+
+	if (args.buf_size > ocxlpmem->admin_command.data_size)
+		return -EINVAL;
+
+	buf = (u64 *)args.buf_ptr;
+
+	mutex_lock(&ocxlpmem->admin_command.lock);
+
+	val = ((u64)args.offset) << 32;
+	val |= args.buf_size;
+	rc = ocxl_global_mmio_write64(ocxlpmem->ocxl_afu,
+				      ocxlpmem->admin_command.request_offset + 0x08,
+				      OCXL_LITTLE_ENDIAN, val);
+	if (rc)
+		goto out;
+
+	rc = admin_command_execute(ocxlpmem, ADMIN_COMMAND_CONTROLLER_DUMP);
+	if (rc)
+		goto out;
+	if (rc != STATUS_SUCCESS) {
+		warn_status(ocxlpmem,
+			    "Unexpected status from controller dump",
+			    rc);
+		goto out;
+	}
+
+	rc = controller_dump_header_parse(ocxlpmem, &buf_size);
+	if (rc)
+		goto out;
+
+	buf_size = min((u16)(buf_size + sizeof(u64)), args.buf_size);
+
+	for (i = 0; i < buf_size / sizeof(u64); i++) {
+		rc = ocxl_global_mmio_read64(ocxlpmem->ocxl_afu,
+					     ocxlpmem->admin_command.data_offset +
+							i * sizeof(u64),
+					     OCXL_HOST_ENDIAN, &val);
+		if (rc)
+			goto out;
+
+		if (copy_to_user(&buf[i], &val, sizeof(u64))) {
+			rc = -EFAULT;
+			goto out;
+		}
+	}
+
+	args.buf_size = buf_size;
+
+	if (copy_to_user(uarg, &args, sizeof(args))) {
+		rc = -EFAULT;
+		goto out;
+	}
+
+	rc = admin_response_handled(ocxlpmem);
+	if (rc)
+		goto out;
+
+out:
+	mutex_unlock(&ocxlpmem->admin_command.lock);
+	return rc;
+}
+
+int request_controller_dump(struct ocxlpmem *ocxlpmem)
+{
+	int rc;
+	u64 busy = 1;
+
+	rc = ocxl_global_mmio_set64(ocxlpmem->ocxl_afu, GLOBAL_MMIO_CHIC,
+				    OCXL_LITTLE_ENDIAN,
+				    GLOBAL_MMIO_CHI_CDA);
+	if (rc)
+		return rc;
+
+	rc = ocxl_global_mmio_set64(ocxlpmem->ocxl_afu, GLOBAL_MMIO_HCI,
+				    OCXL_LITTLE_ENDIAN,
+				    GLOBAL_MMIO_HCI_CONTROLLER_DUMP);
+	if (rc)
+		return rc;
+
+	while (busy) {
+		rc = ocxl_global_mmio_read64(ocxlpmem->ocxl_afu,
+					     GLOBAL_MMIO_HCI,
+					     OCXL_LITTLE_ENDIAN, &busy);
+		if (rc)
+			return rc;
+
+		busy &= GLOBAL_MMIO_HCI_CONTROLLER_DUMP;
+		cond_resched();
+	}
+
+	return 0;
+}
+
+static int ioctl_controller_dump_complete(struct ocxlpmem *ocxlpmem)
+{
+	return ocxl_global_mmio_set64(ocxlpmem->ocxl_afu, GLOBAL_MMIO_HCI,
+				    OCXL_LITTLE_ENDIAN,
+				    GLOBAL_MMIO_HCI_CONTROLLER_DUMP_COLLECTED);
+}
+
 static long file_ioctl(struct file *file, unsigned int cmd, unsigned long args)
 {
 	struct ocxlpmem *ocxlpmem = file->private_data;
@@ -576,7 +723,21 @@ static long file_ioctl(struct file *file, unsigned int cmd, unsigned long args)
 		rc = ioctl_error_log(ocxlpmem,
 				     (struct ioctl_ocxlpmem_error_log __user *)args);
 		break;
+
+	case IOCTL_OCXLPMEM_CONTROLLER_DUMP:
+		rc = request_controller_dump(ocxlpmem);
+		break;
+
+	case IOCTL_OCXLPMEM_CONTROLLER_DUMP_DATA:
+		rc = ioctl_controller_dump_data(ocxlpmem,
+						(struct ioctl_ocxlpmem_controller_dump_data __user *)args);
+		break;
+
+	case IOCTL_OCXLPMEM_CONTROLLER_DUMP_COMPLETE:
+		rc = ioctl_controller_dump_complete(ocxlpmem);
+		break;
 	}
+
 	return rc;
 }
 
diff --git a/include/uapi/nvdimm/ocxlpmem.h b/include/uapi/nvdimm/ocxlpmem.h
index 5d3a03ea1e08..05e2b3f7b27c 100644
--- a/include/uapi/nvdimm/ocxlpmem.h
+++ b/include/uapi/nvdimm/ocxlpmem.h
@@ -38,9 +38,25 @@ struct ioctl_ocxlpmem_error_log {
 	__u64 reserved2[2];
 };
 
+struct ioctl_ocxlpmem_controller_dump_data {
+	__u64 buf_ptr; /* coerced pointer to output buffer */
+	__u16 buf_size; /* in/out, buffer size provided/required.
+			 * If required is greater than provided, the buffer
+			 * will be truncated to the amount provided. If its
+			 * less, then only the required bytes will be populated.
+			 * If it is 0, then there is no more dump data available.
+			 */
+	__u16 reserved0;
+	__u32 offset; /* in, Offset within the dump */
+	__u64 reserved[8];
+};
+
 /* ioctl numbers */
 #define OCXLPMEM_MAGIC 0xCA
 /* OpenCAPI Persistent memory devices */
 #define IOCTL_OCXLPMEM_ERROR_LOG			_IOWR(OCXLPMEM_MAGIC, 0x30, struct ioctl_ocxlpmem_error_log)
+#define IOCTL_OCXLPMEM_CONTROLLER_DUMP			_IO(OCXLPMEM_MAGIC, 0x31)
+#define IOCTL_OCXLPMEM_CONTROLLER_DUMP_DATA		_IOWR(OCXLPMEM_MAGIC, 0x32, struct ioctl_ocxlpmem_controller_dump_data)
+#define IOCTL_OCXLPMEM_CONTROLLER_DUMP_COMPLETE		_IO(OCXLPMEM_MAGIC, 0x33)
 
 #endif /* _UAPI_OCXL_SCM_H */
-- 
2.24.1
_______________________________________________
Linux-nvdimm mailing list -- linux-nvdimm@lists.01.org
To unsubscribe send an email to linux-nvdimm-leave@lists.01.org