[PATCH 0/6] ibmvscsi: code cleanup

Fixed up a couple spots that were out of line with the PAPR in regards
to its defined VSCSI protocol. Did away with some magic numbers directly
in the code. Fixed a minor endian issue.

Tyrel Datwyler (6):
  ibmvscsi: Correct values for several viosrp_crq_format enums
  ibmvscsi: Add and use enums for valid CRQ header values
  ibmvscsi: Replace magic values in set_adpater_info() with defines
  ibmvscsi: Use of_root to access OF device tree root node
  ibmvscsi: Remove unsupported host config MAD and sysfs interface
  ibmvscsi: Add endian conversions to sysfs attribute show functions

 drivers/scsi/ibmvscsi/ibmvscsi.c | 117 +++++++--------------------------------
 drivers/scsi/ibmvscsi/viosrp.h   |  20 ++++---
 2 files changed, 30 insertions(+), 107 deletions(-)

-- 
2.5.0

The values returned by the show functions for the host os_type,
mad_version, and partition_number attributes get their values
directly from the madapter_info struct whose associated fields are
__be32 typed. Added endian conversion to ensure these values are
sane on LE platforms.

Signed-off-by: Tyrel Datwyler <redacted>
---
 drivers/scsi/ibmvscsi/ibmvscsi.c | 7 ++++---
 1 file changed, 4 insertions(+), 3 deletions(-)

diff --git a/drivers/scsi/ibmvscsi/ibmvscsi.c b/drivers/scsi/ibmvscsi/ibmvscsi.c
index a2b8db1..3621ac5 100644
--- a/drivers/scsi/ibmvscsi/ibmvscsi.c
+++ b/drivers/scsi/ibmvscsi/ibmvscsi.c

@@ -1983,7 +1983,7 @@ static ssize_t show_host_partition_number(struct device *dev,
 	int len;
 
 	len = snprintf(buf, PAGE_SIZE, "%d\n",
-		       hostdata->madapter_info.partition_number);
+		       be32_to_cpu(hostdata->madapter_info.partition_number));
 	return len;
 }
 

@@ -2003,7 +2003,7 @@ static ssize_t show_host_mad_version(struct device *dev,
 	int len;
 
 	len = snprintf(buf, PAGE_SIZE, "%d\n",
-		       hostdata->madapter_info.mad_version);
+		       be32_to_cpu(hostdata->madapter_info.mad_version));
 	return len;
 }
 

@@ -2022,7 +2022,8 @@ static ssize_t show_host_os_type(struct device *dev,
 	struct ibmvscsi_host_data *hostdata = shost_priv(shost);
 	int len;
 
-	len = snprintf(buf, PAGE_SIZE, "%d\n", hostdata->madapter_info.os_type);
+	len = snprintf(buf, PAGE_SIZE, "%d\n",
+		       be32_to_cpu(hostdata->madapter_info.os_type));
 	return len;
 }
 

-- 
2.5.0

On Wed, Feb 03, 2016 at 05:28:34PM -0600, Tyrel Datwyler wrote:

Reviewed-by: Johannes Thumshirn <redacted>
-- 
Johannes Thumshirn                                          Storage
jthumshirn@suse.de                                +49 911 74053 689
SUSE LINUX GmbH, Maxfeldstr. 5, 90409 Nürnberg
GF: Felix Imendörffer, Jane Smithard, Graham Norton
HRB 21284 (AG Nürnberg)
Key fingerprint = EC38 9CAB C2C4 F25D 8600 D0D0 0393 969D 2D76 0850

Add defines for mad version and mad os_type, and replace the magic
numbers in set_adapter_info() accordingly.

Signed-off-by: Tyrel Datwyler <redacted>
---
 drivers/scsi/ibmvscsi/ibmvscsi.c | 4 ++--
 drivers/scsi/ibmvscsi/viosrp.h   | 2 ++
 2 files changed, 4 insertions(+), 2 deletions(-)

diff --git a/drivers/scsi/ibmvscsi/ibmvscsi.c b/drivers/scsi/ibmvscsi/ibmvscsi.c
index 176260d..7e51615 100644
--- a/drivers/scsi/ibmvscsi/ibmvscsi.c
+++ b/drivers/scsi/ibmvscsi/ibmvscsi.c

@@ -283,8 +283,8 @@ static void set_adapter_info(struct ibmvscsi_host_data *hostdata)
 	hostdata->madapter_info.partition_number =
 					cpu_to_be32(partition_number);
 
-	hostdata->madapter_info.mad_version = cpu_to_be32(1);
-	hostdata->madapter_info.os_type = cpu_to_be32(2);
+	hostdata->madapter_info.mad_version = cpu_to_be32(SRP_MAD_VERSION_1);
+	hostdata->madapter_info.os_type = cpu_to_be32(SRP_MAD_OS_LINUX);
 }
 
 /**

diff --git a/drivers/scsi/ibmvscsi/viosrp.h b/drivers/scsi/ibmvscsi/viosrp.h
index 17f2de0..da9bb29 100644
--- a/drivers/scsi/ibmvscsi/viosrp.h
+++ b/drivers/scsi/ibmvscsi/viosrp.h

@@ -38,6 +38,8 @@
 #define SRP_VERSION "16.a"
 #define SRP_MAX_IU_LEN	256
 #define SRP_MAX_LOC_LEN 32
+#define SRP_MAD_VERSION_1 1
+#define SRP_MAD_OS_LINUX 2
 
 union srp_iu {
 	struct srp_login_req login_req;

-- 
2.5.0

On Wed, Feb 03, 2016 at 05:28:31PM -0600, Tyrel Datwyler wrote:

Reviewed-by: Johannes Thumshirn <redacted>
-- 
Johannes Thumshirn                                          Storage
jthumshirn@suse.de                                +49 911 74053 689
SUSE LINUX GmbH, Maxfeldstr. 5, 90409 Nürnberg
GF: Felix Imendörffer, Jane Smithard, Graham Norton
HRB 21284 (AG Nürnberg)
Key fingerprint = EC38 9CAB C2C4 F25D 8600 D0D0 0393 969D 2D76 0850

A VIOSRP_HOST_CONFIG_TYPE management datagram (MAD) has existed in
the code for some time. From what information I've gathered from
Brian King this was likely implemented on the host side in a SLES 9
based VIOS, which is no longer supported anywhere. Further, it is
not defined in PAPR or supported by any AIX based VIOS.

Treating as bit rot and removing the sysfs interface and associated
host config code accordingly.

Signed-off-by: Tyrel Datwyler <redacted>
---
 drivers/scsi/ibmvscsi/ibmvscsi.c | 78 ----------------------------------------
 drivers/scsi/ibmvscsi/viosrp.h   |  7 ----
 2 files changed, 85 deletions(-)

diff --git a/drivers/scsi/ibmvscsi/ibmvscsi.c b/drivers/scsi/ibmvscsi/ibmvscsi.c
index 47cfe33..a2b8db1 100644
--- a/drivers/scsi/ibmvscsi/ibmvscsi.c
+++ b/drivers/scsi/ibmvscsi/ibmvscsi.c

@@ -1853,62 +1853,6 @@ static void ibmvscsi_handle_crq(struct viosrp_crq *crq,
 }
 
 /**
- * ibmvscsi_get_host_config: Send the command to the server to get host
- * configuration data.  The data is opaque to us.
- */
-static int ibmvscsi_do_host_config(struct ibmvscsi_host_data *hostdata,
-				   unsigned char *buffer, int length)
-{
-	struct viosrp_host_config *host_config;
-	struct srp_event_struct *evt_struct;
-	unsigned long flags;
-	dma_addr_t addr;
-	int rc;
-
-	evt_struct = get_event_struct(&hostdata->pool);
-	if (!evt_struct) {
-		dev_err(hostdata->dev, "couldn't allocate event for HOST_CONFIG!\n");
-		return -1;
-	}
-
-	init_event_struct(evt_struct,
-			  sync_completion,
-			  VIOSRP_MAD_FORMAT,
-			  info_timeout);
-
-	host_config = &evt_struct->iu.mad.host_config;
-
-	/* The transport length field is only 16-bit */
-	length = min(0xffff, length);
-
-	/* Set up a lun reset SRP command */
-	memset(host_config, 0x00, sizeof(*host_config));
-	host_config->common.type = cpu_to_be32(VIOSRP_HOST_CONFIG_TYPE);
-	host_config->common.length = cpu_to_be16(length);
-	addr = dma_map_single(hostdata->dev, buffer, length, DMA_BIDIRECTIONAL);
-
-	if (dma_mapping_error(hostdata->dev, addr)) {
-		if (!firmware_has_feature(FW_FEATURE_CMO))
-			dev_err(hostdata->dev,
-			        "dma_mapping error getting host config\n");
-		free_event_struct(&hostdata->pool, evt_struct);
-		return -1;
-	}
-
-	host_config->buffer = cpu_to_be64(addr);
-
-	init_completion(&evt_struct->comp);
-	spin_lock_irqsave(hostdata->host->host_lock, flags);
-	rc = ibmvscsi_send_srp_event(evt_struct, hostdata, info_timeout * 2);
-	spin_unlock_irqrestore(hostdata->host->host_lock, flags);
-	if (rc == 0)
-		wait_for_completion(&evt_struct->comp);
-	dma_unmap_single(hostdata->dev, addr, length, DMA_BIDIRECTIONAL);
-
-	return rc;
-}
-
-/**
  * ibmvscsi_slave_configure: Set the "allow_restart" flag for each disk.
  * @sdev:	struct scsi_device device to configure
  *

@@ -2090,27 +2034,6 @@ static struct device_attribute ibmvscsi_host_os_type = {
 	.show = show_host_os_type,
 };
 
-static ssize_t show_host_config(struct device *dev,
-				struct device_attribute *attr, char *buf)
-{
-	struct Scsi_Host *shost = class_to_shost(dev);
-	struct ibmvscsi_host_data *hostdata = shost_priv(shost);
-
-	/* returns null-terminated host config data */
-	if (ibmvscsi_do_host_config(hostdata, buf, PAGE_SIZE) == 0)
-		return strlen(buf);
-	else
-		return 0;
-}
-
-static struct device_attribute ibmvscsi_host_config = {
-	.attr = {
-		 .name = "config",
-		 .mode = S_IRUGO,
-		 },
-	.show = show_host_config,
-};
-
 static struct device_attribute *ibmvscsi_attrs[] = {
 	&ibmvscsi_host_vhost_loc,
 	&ibmvscsi_host_vhost_name,

@@ -2119,7 +2042,6 @@ static struct device_attribute *ibmvscsi_attrs[] = {
 	&ibmvscsi_host_partition_number,
 	&ibmvscsi_host_mad_version,
 	&ibmvscsi_host_os_type,
-	&ibmvscsi_host_config,
 	NULL
 };
 

diff --git a/drivers/scsi/ibmvscsi/viosrp.h b/drivers/scsi/ibmvscsi/viosrp.h
index da9bb29..04bae63 100644
--- a/drivers/scsi/ibmvscsi/viosrp.h
+++ b/drivers/scsi/ibmvscsi/viosrp.h

@@ -96,7 +96,6 @@ enum viosrp_mad_types {
 	VIOSRP_EMPTY_IU_TYPE = 0x01,
 	VIOSRP_ERROR_LOG_TYPE = 0x02,
 	VIOSRP_ADAPTER_INFO_TYPE = 0x03,
-	VIOSRP_HOST_CONFIG_TYPE = 0x04,
 	VIOSRP_CAPABILITIES_TYPE = 0x05,
 	VIOSRP_ENABLE_FAST_FAIL = 0x08,
 };

@@ -162,11 +161,6 @@ struct viosrp_adapter_info {
 	__be64 buffer;
 };
 
-struct viosrp_host_config {
-	struct mad_common common;
-	__be64 buffer;
-};
-
 struct viosrp_fast_fail {
 	struct mad_common common;
 };

@@ -204,7 +198,6 @@ union mad_iu {
 	struct viosrp_empty_iu empty_iu;
 	struct viosrp_error_log error_log;
 	struct viosrp_adapter_info adapter_info;
-	struct viosrp_host_config host_config;
 	struct viosrp_fast_fail fast_fail;
 	struct viosrp_capabilities capabilities;
 };

-- 
2.5.0

On Wed, Feb 03, 2016 at 05:28:33PM -0600, Tyrel Datwyler wrote:

A VIOSRP_HOST_CONFIG_TYPE management datagram (MAD) has existed in
the code for some time. From what information I've gathered from
Brian King this was likely implemented on the host side in a SLES 9
based VIOS, which is no longer supported anywhere. Further, it is
not defined in PAPR or supported by any AIX based VIOS.

Treating as bit rot and removing the sysfs interface and associated
host config code accordingly.

Doesn't removing a sysfs interface potentially break userspace code?

-- 
Johannes Thumshirn                                          Storage
jthumshirn@suse.de                                +49 911 74053 689
SUSE LINUX GmbH, Maxfeldstr. 5, 90409 Nürnberg
GF: Felix Imendörffer, Jane Smithard, Graham Norton
HRB 21284 (AG Nürnberg)
Key fingerprint = EC38 9CAB C2C4 F25D 8600 D0D0 0393 969D 2D76 0850

On 02/04/2016 12:03 AM, Johannes Thumshirn wrote:

On Wed, Feb 03, 2016 at 05:28:33PM -0600, Tyrel Datwyler wrote:

quoted

A VIOSRP_HOST_CONFIG_TYPE management datagram (MAD) has existed in
the code for some time. From what information I've gathered from
Brian King this was likely implemented on the host side in a SLES 9
based VIOS, which is no longer supported anywhere. Further, it is
not defined in PAPR or supported by any AIX based VIOS.

Treating as bit rot and removing the sysfs interface and associated
host config code accordingly.

Doesn't removing a sysfs interface potentially break userspace code?

In the general case yes, but I feel in this case no. First, Reading from
this config attribute of a vscsi host adapter always returns nothing.
Second, any userspace code using this attribute better be checking for
the existence of config. Just a quick look for
/sys/class/scsi_host/host*/config under other host adapters on my system
I find that attribute doesn't exist for any of them.

If there is truly enough concern that somebody may actually be accessing
this useless attribute from userspace then we can still strip out the
unsupported code, but leave the attribute and return nothing directly
from the show function.

-Tyrel

On Thu, Feb 04, 2016 at 09:48:23AM -0800, Tyrel Datwyler wrote:

On 02/04/2016 12:03 AM, Johannes Thumshirn wrote:

quoted

On Wed, Feb 03, 2016 at 05:28:33PM -0600, Tyrel Datwyler wrote:

quoted

A VIOSRP_HOST_CONFIG_TYPE management datagram (MAD) has existed in
the code for some time. From what information I've gathered from
Brian King this was likely implemented on the host side in a SLES 9
based VIOS, which is no longer supported anywhere. Further, it is
not defined in PAPR or supported by any AIX based VIOS.

Treating as bit rot and removing the sysfs interface and associated
host config code accordingly.

Doesn't removing a sysfs interface potentially break userspace code?

In the general case yes, but I feel in this case no. First, Reading from
this config attribute of a vscsi host adapter always returns nothing.
Second, any userspace code using this attribute better be checking for
the existence of config. Just a quick look for
/sys/class/scsi_host/host*/config under other host adapters on my system
I find that attribute doesn't exist for any of them.

If there is truly enough concern that somebody may actually be accessing
this useless attribute from userspace then we can still strip out the
unsupported code, but leave the attribute and return nothing directly
from the show function.

Which is what I kinda prefer. Slowly deprecate and phase out, but don't break
userspace. You never know who is writing some obscure piece of code relaying on
some sysfs attribute.

Thanks,
	Johannes

-- 
Johannes Thumshirn                                          Storage
jthumshirn@suse.de                                +49 911 74053 689
SUSE LINUX GmbH, Maxfeldstr. 5, 90409 Nürnberg
GF: Felix Imendörffer, Jane Smithard, Graham Norton
HRB 21284 (AG Nürnberg)
Key fingerprint = EC38 9CAB C2C4 F25D 8600 D0D0 0393 969D 2D76 0850

The root node of the OF device tree is exported as of_root. No need
to look up the root by path name. Instead just get a reference
directly via of_root.

Signed-off-by: Tyrel Datwyler <redacted>
---
 drivers/scsi/ibmvscsi/ibmvscsi.c | 14 ++++++--------
 1 file changed, 6 insertions(+), 8 deletions(-)

diff --git a/drivers/scsi/ibmvscsi/ibmvscsi.c b/drivers/scsi/ibmvscsi/ibmvscsi.c
index 7e51615..47cfe33 100644
--- a/drivers/scsi/ibmvscsi/ibmvscsi.c
+++ b/drivers/scsi/ibmvscsi/ibmvscsi.c

@@ -248,25 +248,23 @@ static void ibmvscsi_task(void *data)
 
 static void gather_partition_info(void)
 {
-	struct device_node *rootdn;
-
 	const char *ppartition_name;
 	const __be32 *p_number_ptr;
 
 	/* Retrieve information about this partition */
-	rootdn = of_find_node_by_path("/");
-	if (!rootdn) {
+	if (!of_root)
 		return;
-	}
 
-	ppartition_name = of_get_property(rootdn, "ibm,partition-name", NULL);
+	of_node_get(of_root);
+
+	ppartition_name = of_get_property(of_root, "ibm,partition-name", NULL);
 	if (ppartition_name)
 		strncpy(partition_name, ppartition_name,
 				sizeof(partition_name));
-	p_number_ptr = of_get_property(rootdn, "ibm,partition-no", NULL);
+	p_number_ptr = of_get_property(of_root, "ibm,partition-no", NULL);
 	if (p_number_ptr)
 		partition_number = of_read_number(p_number_ptr, 1);
-	of_node_put(rootdn);
+	of_node_put(of_root);
 }
 
 static void set_adapter_info(struct ibmvscsi_host_data *hostdata)

-- 
2.5.0

On Wed, Feb 03, 2016 at 05:28:32PM -0600, Tyrel Datwyler wrote:

Reviewed-by: Johannes Thumshirn <redacted>
-- 
Johannes Thumshirn                                          Storage
jthumshirn@suse.de                                +49 911 74053 689
SUSE LINUX GmbH, Maxfeldstr. 5, 90409 Nürnberg
GF: Felix Imendörffer, Jane Smithard, Graham Norton
HRB 21284 (AG Nürnberg)
Key fingerprint = EC38 9CAB C2C4 F25D 8600 D0D0 0393 969D 2D76 0850

The PAPR defines four valid header values for the first byte of a
CRQ message. Namely, an unused/empty message (0x00), a valid
command/response entry (0x80), a valid initialization entry (0xC0),
and a transport event (0xFF). Define these values as enums and use
them in the code in place of their magic number equivalents.

Signed-off-by: Tyrel Datwyler <redacted>
---
 drivers/scsi/ibmvscsi/ibmvscsi.c | 14 +++++++-------
 drivers/scsi/ibmvscsi/viosrp.h   |  7 +++++++
 2 files changed, 14 insertions(+), 7 deletions(-)

diff --git a/drivers/scsi/ibmvscsi/ibmvscsi.c b/drivers/scsi/ibmvscsi/ibmvscsi.c
index adfef9d..176260d 100644
--- a/drivers/scsi/ibmvscsi/ibmvscsi.c
+++ b/drivers/scsi/ibmvscsi/ibmvscsi.c

@@ -182,7 +182,7 @@ static struct viosrp_crq *crq_queue_next_crq(struct crq_queue *queue)
 
 	spin_lock_irqsave(&queue->lock, flags);
 	crq = &queue->msgs[queue->cur];
-	if (crq->valid & 0x80) {
+	if (crq->valid & VIOSRP_CRQ_VALID) {
 		if (++queue->cur == queue->size)
 			queue->cur = 0;
 

@@ -231,7 +231,7 @@ static void ibmvscsi_task(void *data)
 		/* Pull all the valid messages off the CRQ */
 		while ((crq = crq_queue_next_crq(&hostdata->queue)) != NULL) {
 			ibmvscsi_handle_crq(crq, hostdata);
-			crq->valid = 0x00;
+			crq->valid = VIOSRP_CRQ_FREE;
 		}
 
 		vio_enable_interrupts(vdev);

@@ -239,7 +239,7 @@ static void ibmvscsi_task(void *data)
 		if (crq != NULL) {
 			vio_disable_interrupts(vdev);
 			ibmvscsi_handle_crq(crq, hostdata);
-			crq->valid = 0x00;
+			crq->valid = VIOSRP_CRQ_FREE;
 		} else {
 			done = 1;
 		}

@@ -474,7 +474,7 @@ static int initialize_event_pool(struct event_pool *pool,
 		struct srp_event_struct *evt = &pool->events[i];
 		memset(&evt->crq, 0x00, sizeof(evt->crq));
 		atomic_set(&evt->free, 1);
-		evt->crq.valid = 0x80;
+		evt->crq.valid = VIOSRP_CRQ_VALID;
 		evt->crq.IU_length = cpu_to_be16(sizeof(*evt->xfer_iu));
 		evt->crq.IU_data_ptr = cpu_to_be64(pool->iu_token +
 			sizeof(*evt->xfer_iu) * i);

@@ -1767,7 +1767,7 @@ static void ibmvscsi_handle_crq(struct viosrp_crq *crq,
 	struct srp_event_struct *evt_struct =
 			(__force struct srp_event_struct *)crq->IU_data_ptr;
 	switch (crq->valid) {
-	case 0xC0:		/* initialization */
+	case VIOSRP_CRQ_INIT:		/* initialization */
 		switch (crq->format) {
 		case 0x01:	/* Initialization message */
 			dev_info(hostdata->dev, "partner initialized\n");

@@ -1791,7 +1791,7 @@ static void ibmvscsi_handle_crq(struct viosrp_crq *crq,
 			dev_err(hostdata->dev, "unknown crq message type: %d\n", crq->format);
 		}
 		return;
-	case 0xFF:	/* Hypervisor telling us the connection is closed */
+	case VIOSRP_CRQ_TRANSPORT:	/* Hypervisor telling us the connection is closed */
 		scsi_block_requests(hostdata->host);
 		atomic_set(&hostdata->request_limit, 0);
 		if (crq->format == 0x06) {

@@ -1807,7 +1807,7 @@ static void ibmvscsi_handle_crq(struct viosrp_crq *crq,
 			ibmvscsi_reset_host(hostdata);
 		}
 		return;
-	case 0x80:		/* real payload */
+	case VIOSRP_CRQ_VALID:		/* real payload */
 		break;
 	default:
 		dev_err(hostdata->dev, "got an invalid message type 0x%02x\n",

diff --git a/drivers/scsi/ibmvscsi/viosrp.h b/drivers/scsi/ibmvscsi/viosrp.h
index d1044e9..17f2de0 100644
--- a/drivers/scsi/ibmvscsi/viosrp.h
+++ b/drivers/scsi/ibmvscsi/viosrp.h

@@ -51,6 +51,13 @@ union srp_iu {
 	u8 reserved[SRP_MAX_IU_LEN];
 };
 
+enum viosrp_crq_headers {
+	VIOSRP_CRQ_FREE = 0x00,
+	VIOSRP_CRQ_VALID = 0x80,
+	VIOSRP_CRQ_INIT = 0xC0,
+	VIOSRP_CRQ_TRANSPORT = 0xFF
+};
+
 enum viosrp_crq_formats {
 	VIOSRP_SRP_FORMAT = 0x01,
 	VIOSRP_MAD_FORMAT = 0x02,

-- 
2.5.0

On Wed, Feb 03, 2016 at 05:28:30PM -0600, Tyrel Datwyler wrote:

Reviewed-by: Johannes Thumshirn <redacted>
-- 
Johannes Thumshirn                                          Storage
jthumshirn@suse.de                                +49 911 74053 689
SUSE LINUX GmbH, Maxfeldstr. 5, 90409 Nürnberg
GF: Felix Imendörffer, Jane Smithard, Graham Norton
HRB 21284 (AG Nürnberg)
Key fingerprint = EC38 9CAB C2C4 F25D 8600 D0D0 0393 969D 2D76 0850

On 2/3/2016 5:28 PM, Tyrel Datwyler wrote:

After the switch to enums, bitwise operators are a bit misleading.
Especially in this case since multiple values would satisfy this
condition: VIOSRP_CRQ_VALID, VIOSRP_CRQ_INIT and
VIOSRP_CRQ_TRANSPORT.

If 'valid' will only have one of these four enums defined, would
this be better written as:

	if (crq->valid != VIOSRP_CRQ_FREE)

On 02/04/2016 11:55 AM, Manoj Kumar wrote:

On 2/3/2016 5:28 PM, Tyrel Datwyler wrote:

quoted

The PAPR defines four valid header values for the first byte of a
CRQ message. Namely, an unused/empty message (0x00), a valid
command/response entry (0x80), a valid initialization entry (0xC0),
and a transport event (0xFF). Define these values as enums and use
them in the code in place of their magic number equivalents.

Signed-off-by: Tyrel Datwyler <redacted>
---
  drivers/scsi/ibmvscsi/ibmvscsi.c | 14 +++++++-------
  drivers/scsi/ibmvscsi/viosrp.h   |  7 +++++++
  2 files changed, 14 insertions(+), 7 deletions(-)

diff --git a/drivers/scsi/ibmvscsi/ibmvscsi.c b/drivers/scsi/ibmvscsi/ibmvscsi.c
index adfef9d..176260d 100644
--- a/drivers/scsi/ibmvscsi/ibmvscsi.c
+++ b/drivers/scsi/ibmvscsi/ibmvscsi.c

@@ -182,7 +182,7 @@ static struct viosrp_crq *crq_queue_next_crq(struct crq_queue *queue)

  	spin_lock_irqsave(&queue->lock, flags);
  	crq = &queue->msgs[queue->cur];
-	if (crq->valid & 0x80) {
+	if (crq->valid & VIOSRP_CRQ_VALID) {

After the switch to enums, bitwise operators are a bit misleading.
Especially in this case since multiple values would satisfy this
condition: VIOSRP_CRQ_VALID, VIOSRP_CRQ_INIT and
VIOSRP_CRQ_TRANSPORT.

Yeah, I can see how that is confusing. Since, all three possible valid
crq message types have the first bit set I think this was originally a
cute hack to grab anything that was likely valid. Then in
ibmvscsi_handle_crq() we explicitly match the full header value in a
switch statement logging anything that turned out actually invalid.

If 'valid' will only have one of these four enums defined, would
this be better written as:

	if (crq->valid != VIOSRP_CRQ_FREE)

This definitely would make the logic easier to read and follow. Also,
this would make sure any crq with an invalid header that doesn't have
its first bit set will also be logged by the ibmvscsi_handle_crq()
switch statement default block and not silently ignored.

-Tyrel

quoted

  		if (++queue->cur == queue->size)
  			queue->cur = 0;

@@ -231,7 +231,7 @@ static void ibmvscsi_task(void *data)
  		/* Pull all the valid messages off the CRQ */
  		while ((crq = crq_queue_next_crq(&hostdata->queue)) != NULL) {
  			ibmvscsi_handle_crq(crq, hostdata);
-			crq->valid = 0x00;
+			crq->valid = VIOSRP_CRQ_FREE;
  		}

  		vio_enable_interrupts(vdev);

@@ -239,7 +239,7 @@ static void ibmvscsi_task(void *data)
  		if (crq != NULL) {
  			vio_disable_interrupts(vdev);
  			ibmvscsi_handle_crq(crq, hostdata);
-			crq->valid = 0x00;
+			crq->valid = VIOSRP_CRQ_FREE;
  		} else {
  			done = 1;
  		}

@@ -474,7 +474,7 @@ static int initialize_event_pool(struct event_pool *pool,
  		struct srp_event_struct *evt = &pool->events[i];
  		memset(&evt->crq, 0x00, sizeof(evt->crq));
  		atomic_set(&evt->free, 1);
-		evt->crq.valid = 0x80;
+		evt->crq.valid = VIOSRP_CRQ_VALID;
  		evt->crq.IU_length = cpu_to_be16(sizeof(*evt->xfer_iu));
  		evt->crq.IU_data_ptr = cpu_to_be64(pool->iu_token +
  			sizeof(*evt->xfer_iu) * i);

@@ -1767,7 +1767,7 @@ static void ibmvscsi_handle_crq(struct viosrp_crq *crq,
  	struct srp_event_struct *evt_struct =
  			(__force struct srp_event_struct *)crq->IU_data_ptr;
  	switch (crq->valid) {
-	case 0xC0:		/* initialization */
+	case VIOSRP_CRQ_INIT:		/* initialization */
  		switch (crq->format) {
  		case 0x01:	/* Initialization message */
  			dev_info(hostdata->dev, "partner initialized\n");

@@ -1791,7 +1791,7 @@ static void ibmvscsi_handle_crq(struct viosrp_crq *crq,
  			dev_err(hostdata->dev, "unknown crq message type: %d\n", crq->format);
  		}
  		return;
-	case 0xFF:	/* Hypervisor telling us the connection is closed */
+	case VIOSRP_CRQ_TRANSPORT:	/* Hypervisor telling us the connection is closed */
  		scsi_block_requests(hostdata->host);
  		atomic_set(&hostdata->request_limit, 0);
  		if (crq->format == 0x06) {

@@ -1807,7 +1807,7 @@ static void ibmvscsi_handle_crq(struct viosrp_crq *crq,
  			ibmvscsi_reset_host(hostdata);
  		}
  		return;
-	case 0x80:		/* real payload */
+	case VIOSRP_CRQ_VALID:		/* real payload */
  		break;
  	default:
  		dev_err(hostdata->dev, "got an invalid message type 0x%02x\n",

diff --git a/drivers/scsi/ibmvscsi/viosrp.h b/drivers/scsi/ibmvscsi/viosrp.h
index d1044e9..17f2de0 100644
--- a/drivers/scsi/ibmvscsi/viosrp.h
+++ b/drivers/scsi/ibmvscsi/viosrp.h

@@ -51,6 +51,13 @@ union srp_iu {
  	u8 reserved[SRP_MAX_IU_LEN];
  };

+enum viosrp_crq_headers {
+	VIOSRP_CRQ_FREE = 0x00,
+	VIOSRP_CRQ_VALID = 0x80,
+	VIOSRP_CRQ_INIT = 0xC0,
+	VIOSRP_CRQ_TRANSPORT = 0xFF
+};
+
  enum viosrp_crq_formats {
  	VIOSRP_SRP_FORMAT = 0x01,
  	VIOSRP_MAD_FORMAT = 0x02,

_______________________________________________
Linuxppc-dev mailing list
Linuxppc-dev@lists.ozlabs.org
https://lists.ozlabs.org/listinfo/linuxppc-dev

Yeah, I can see how that is confusing. Since, all three possible valid
crq message types have the first bit set I think this was originally a
cute hack to grab anything that was likely valid. Then in
ibmvscsi_handle_crq() we explicitly match the full header value in a
switch statement logging anything that turned out actually invalid.

quoted

If 'valid' will only have one of these four enums defined, would
this be better written as:

	if (crq->valid != VIOSRP_CRQ_FREE)

This definitely would make the logic easier to read and follow. Also,
this would make sure any crq with an invalid header that doesn't have
its first bit set will also be logged by the ibmvscsi_handle_crq()
switch statement default block and not silently ignored.

-Tyrel

Sounds good, Tyrel. Does this mean I should expect a v2 of this patch
series?

- Manoj N. Kumar

On 02/09/2016 09:41 AM, Manoj Kumar wrote:

quoted

Yeah, I can see how that is confusing. Since, all three possible valid
crq message types have the first bit set I think this was originally a
cute hack to grab anything that was likely valid. Then in
ibmvscsi_handle_crq() we explicitly match the full header value in a
switch statement logging anything that turned out actually invalid.

quoted

If 'valid' will only have one of these four enums defined, would
this be better written as:

	if (crq->valid != VIOSRP_CRQ_FREE)

This definitely would make the logic easier to read and follow. Also,
this would make sure any crq with an invalid header that doesn't have
its first bit set will also be logged by the ibmvscsi_handle_crq()
switch statement default block and not silently ignored.

-Tyrel

Sounds good, Tyrel. Does this mean I should expect a v2 of this patch
series?

- Manoj N. Kumar

Haven't had a chance to clean up and resubmit, but yes there will be a
v2 coming along soon.

-Tyrel

_______________________________________________
Linuxppc-dev mailing list
Linuxppc-dev@lists.ozlabs.org
https://lists.ozlabs.org/listinfo/linuxppc-dev

The enum values for VIOSRP_LINUX_FORMAT and VIOSRP_INLINE_FORMAT are
off by one. They are currently defined as 0x06 and 0x07 respetively.
These values are defined in PAPR correctly as 0x05 and 0x06. This
inconsistency has gone unnoticed as neither enum is currently used.
The possible future support of PING messages between the VIOS and
client adapter relies on VIOSRP_INLINE_FORMAT crq messages.
Corrected these enum values to match PAPR definitions.

Signed-off-by: Tyrel Datwyler <redacted>
---
 drivers/scsi/ibmvscsi/viosrp.h | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/drivers/scsi/ibmvscsi/viosrp.h b/drivers/scsi/ibmvscsi/viosrp.h
index 1162430..d1044e9 100644
--- a/drivers/scsi/ibmvscsi/viosrp.h
+++ b/drivers/scsi/ibmvscsi/viosrp.h

@@ -56,8 +56,8 @@ enum viosrp_crq_formats {
 	VIOSRP_MAD_FORMAT = 0x02,
 	VIOSRP_OS400_FORMAT = 0x03,
 	VIOSRP_AIX_FORMAT = 0x04,
-	VIOSRP_LINUX_FORMAT = 0x06,
-	VIOSRP_INLINE_FORMAT = 0x07
+	VIOSRP_LINUX_FORMAT = 0x05,
+	VIOSRP_INLINE_FORMAT = 0x06
 };
 
 enum viosrp_crq_status {

-- 
2.5.0

On Wed, Feb 03, 2016 at 05:28:29PM -0600, Tyrel Datwyler wrote:

Reviewed-by: Johannes Thumshirn <redacted>
-- 
Johannes Thumshirn                                          Storage
jthumshirn@suse.de                                +49 911 74053 689
SUSE LINUX GmbH, Maxfeldstr. 5, 90409 Nürnberg
GF: Felix Imendörffer, Jane Smithard, Graham Norton
HRB 21284 (AG Nürnberg)
Key fingerprint = EC38 9CAB C2C4 F25D 8600 D0D0 0393 969D 2D76 0850