summaryrefslogtreecommitdiffstats
path: root/drivers/s390
diff options
context:
space:
mode:
Diffstat (limited to 'drivers/s390')
-rw-r--r--drivers/s390/cio/cmf.c623
-rw-r--r--drivers/s390/cio/device.h4
-rw-r--r--drivers/s390/cio/device_fsm.c18
3 files changed, 468 insertions, 177 deletions
diff --git a/drivers/s390/cio/cmf.c b/drivers/s390/cio/cmf.c
index 07ef3f6..1c3e8e9 100644
--- a/drivers/s390/cio/cmf.c
+++ b/drivers/s390/cio/cmf.c
@@ -3,9 +3,10 @@
*
* Linux on zSeries Channel Measurement Facility support
*
- * Copyright 2000,2003 IBM Corporation
+ * Copyright 2000,2006 IBM Corporation
*
- * Author: Arnd Bergmann <arndb@de.ibm.com>
+ * Authors: Arnd Bergmann <arndb@de.ibm.com>
+ * Cornelia Huck <cornelia.huck@de.ibm.com>
*
* original idea from Natarajan Krishnaswami <nkrishna@us.ibm.com>
*
@@ -96,9 +97,9 @@ module_param(format, bool, 0444);
/**
* struct cmb_operations - functions to use depending on cmb_format
*
- * all these functions operate on a struct cmf_device. There is only
- * one instance of struct cmb_operations because all cmf_device
- * objects are guaranteed to be of the same type.
+ * Most of these functions operate on a struct ccw_device. There is only
+ * one instance of struct cmb_operations because the format of the measurement
+ * data is guaranteed to be the same for every ccw_device.
*
* @alloc: allocate memory for a channel measurement block,
* either with the help of a special pool or with kmalloc
@@ -107,6 +108,7 @@ module_param(format, bool, 0444);
* @readall: read a measurement block in a common format
* @reset: clear the data in the associated measurement block and
* reset its time stamp
+ * @align: align an allocated block so that the hardware can use it
*/
struct cmb_operations {
int (*alloc) (struct ccw_device*);
@@ -115,11 +117,19 @@ struct cmb_operations {
u64 (*read) (struct ccw_device*, int);
int (*readall)(struct ccw_device*, struct cmbdata *);
void (*reset) (struct ccw_device*);
+ void * (*align) (void *);
struct attribute_group *attr_group;
};
static struct cmb_operations *cmbops;
+struct cmb_data {
+ void *hw_block; /* Pointer to block updated by hardware */
+ void *last_block; /* Last changed block copied from hardware block */
+ int size; /* Size of hw_block and last_block */
+ unsigned long long last_update; /* when last_block was updated */
+};
+
/* our user interface is designed in terms of nanoseconds,
* while the hardware measures total times in its own
* unit.*/
@@ -226,63 +236,229 @@ struct set_schib_struct {
unsigned long address;
wait_queue_head_t wait;
int ret;
+ struct kref kref;
};
+static void cmf_set_schib_release(struct kref *kref)
+{
+ struct set_schib_struct *set_data;
+
+ set_data = container_of(kref, struct set_schib_struct, kref);
+ kfree(set_data);
+}
+
+#define CMF_PENDING 1
+
static int set_schib_wait(struct ccw_device *cdev, u32 mme,
int mbfc, unsigned long address)
{
- struct set_schib_struct s = {
- .mme = mme,
- .mbfc = mbfc,
- .address = address,
- .wait = __WAIT_QUEUE_HEAD_INITIALIZER(s.wait),
- };
+ struct set_schib_struct *set_data;
+ int ret;
spin_lock_irq(cdev->ccwlock);
- s.ret = set_schib(cdev, mme, mbfc, address);
- if (s.ret != -EBUSY) {
- goto out_nowait;
+ if (!cdev->private->cmb) {
+ ret = -ENODEV;
+ goto out;
}
+ set_data = kzalloc(sizeof(struct set_schib_struct), GFP_ATOMIC);
+ if (!set_data) {
+ ret = -ENOMEM;
+ goto out;
+ }
+ init_waitqueue_head(&set_data->wait);
+ kref_init(&set_data->kref);
+ set_data->mme = mme;
+ set_data->mbfc = mbfc;
+ set_data->address = address;
+
+ ret = set_schib(cdev, mme, mbfc, address);
+ if (ret != -EBUSY)
+ goto out_put;
if (cdev->private->state != DEV_STATE_ONLINE) {
- s.ret = -EBUSY;
/* if the device is not online, don't even try again */
- goto out_nowait;
+ ret = -EBUSY;
+ goto out_put;
}
+
cdev->private->state = DEV_STATE_CMFCHANGE;
- cdev->private->cmb_wait = &s;
- s.ret = 1;
+ set_data->ret = CMF_PENDING;
+ cdev->private->cmb_wait = set_data;
spin_unlock_irq(cdev->ccwlock);
- if (wait_event_interruptible(s.wait, s.ret != 1)) {
+ if (wait_event_interruptible(set_data->wait,
+ set_data->ret != CMF_PENDING)) {
spin_lock_irq(cdev->ccwlock);
- if (s.ret == 1) {
- s.ret = -ERESTARTSYS;
- cdev->private->cmb_wait = 0;
+ if (set_data->ret == CMF_PENDING) {
+ set_data->ret = -ERESTARTSYS;
if (cdev->private->state == DEV_STATE_CMFCHANGE)
cdev->private->state = DEV_STATE_ONLINE;
}
spin_unlock_irq(cdev->ccwlock);
}
- return s.ret;
-
-out_nowait:
+ spin_lock_irq(cdev->ccwlock);
+ cdev->private->cmb_wait = NULL;
+ ret = set_data->ret;
+out_put:
+ kref_put(&set_data->kref, cmf_set_schib_release);
+out:
spin_unlock_irq(cdev->ccwlock);
- return s.ret;
+ return ret;
}
void retry_set_schib(struct ccw_device *cdev)
{
- struct set_schib_struct *s;
+ struct set_schib_struct *set_data;
+
+ set_data = cdev->private->cmb_wait;
+ if (!set_data) {
+ WARN_ON(1);
+ return;
+ }
+ kref_get(&set_data->kref);
+ set_data->ret = set_schib(cdev, set_data->mme, set_data->mbfc,
+ set_data->address);
+ wake_up(&set_data->wait);
+ kref_put(&set_data->kref, cmf_set_schib_release);
+}
+
+static int cmf_copy_block(struct ccw_device *cdev)
+{
+ struct subchannel *sch;
+ void *reference_buf;
+ void *hw_block;
+ struct cmb_data *cmb_data;
+
+ sch = to_subchannel(cdev->dev.parent);
+
+ if (stsch(sch->schid, &sch->schib))
+ return -ENODEV;
+
+ if (sch->schib.scsw.fctl & SCSW_FCTL_START_FUNC) {
+ /* Don't copy if a start function is in progress. */
+ if ((!sch->schib.scsw.actl & SCSW_ACTL_SUSPENDED) &&
+ (sch->schib.scsw.actl &
+ (SCSW_ACTL_DEVACT | SCSW_ACTL_SCHACT)) &&
+ (!sch->schib.scsw.stctl & SCSW_STCTL_SEC_STATUS))
+ return -EBUSY;
+ }
+ cmb_data = cdev->private->cmb;
+ hw_block = cmbops->align(cmb_data->hw_block);
+ if (!memcmp(cmb_data->last_block, hw_block, cmb_data->size))
+ /* No need to copy. */
+ return 0;
+ reference_buf = kzalloc(cmb_data->size, GFP_ATOMIC);
+ if (!reference_buf)
+ return -ENOMEM;
+ /* Ensure consistency of block copied from hardware. */
+ do {
+ memcpy(cmb_data->last_block, hw_block, cmb_data->size);
+ memcpy(reference_buf, hw_block, cmb_data->size);
+ } while (memcmp(cmb_data->last_block, reference_buf, cmb_data->size));
+ cmb_data->last_update = get_clock();
+ kfree(reference_buf);
+ return 0;
+}
+
+struct copy_block_struct {
+ wait_queue_head_t wait;
+ int ret;
+ struct kref kref;
+};
+
+static void cmf_copy_block_release(struct kref *kref)
+{
+ struct copy_block_struct *copy_block;
+
+ copy_block = container_of(kref, struct copy_block_struct, kref);
+ kfree(copy_block);
+}
+
+static int cmf_cmb_copy_wait(struct ccw_device *cdev)
+{
+ struct copy_block_struct *copy_block;
+ int ret;
+ unsigned long flags;
+
+ spin_lock_irqsave(cdev->ccwlock, flags);
+ if (!cdev->private->cmb) {
+ ret = -ENODEV;
+ goto out;
+ }
+ copy_block = kzalloc(sizeof(struct copy_block_struct), GFP_ATOMIC);
+ if (!copy_block) {
+ ret = -ENOMEM;
+ goto out;
+ }
+ init_waitqueue_head(&copy_block->wait);
+ kref_init(&copy_block->kref);
+
+ ret = cmf_copy_block(cdev);
+ if (ret != -EBUSY)
+ goto out_put;
+
+ if (cdev->private->state != DEV_STATE_ONLINE) {
+ ret = -EBUSY;
+ goto out_put;
+ }
+
+ cdev->private->state = DEV_STATE_CMFUPDATE;
+ copy_block->ret = CMF_PENDING;
+ cdev->private->cmb_wait = copy_block;
+
+ spin_unlock_irqrestore(cdev->ccwlock, flags);
+ if (wait_event_interruptible(copy_block->wait,
+ copy_block->ret != CMF_PENDING)) {
+ spin_lock_irqsave(cdev->ccwlock, flags);
+ if (copy_block->ret == CMF_PENDING) {
+ copy_block->ret = -ERESTARTSYS;
+ if (cdev->private->state == DEV_STATE_CMFUPDATE)
+ cdev->private->state = DEV_STATE_ONLINE;
+ }
+ spin_unlock_irqrestore(cdev->ccwlock, flags);
+ }
+ spin_lock_irqsave(cdev->ccwlock, flags);
+ cdev->private->cmb_wait = NULL;
+ ret = copy_block->ret;
+out_put:
+ kref_put(&copy_block->kref, cmf_copy_block_release);
+out:
+ spin_unlock_irqrestore(cdev->ccwlock, flags);
+ return ret;
+}
+
+void cmf_retry_copy_block(struct ccw_device *cdev)
+{
+ struct copy_block_struct *copy_block;
- s = cdev->private->cmb_wait;
- cdev->private->cmb_wait = 0;
- if (!s) {
+ copy_block = cdev->private->cmb_wait;
+ if (!copy_block) {
WARN_ON(1);
return;
}
- s->ret = set_schib(cdev, s->mme, s->mbfc, s->address);
- wake_up(&s->wait);
+ kref_get(&copy_block->kref);
+ copy_block->ret = cmf_copy_block(cdev);
+ wake_up(&copy_block->wait);
+ kref_put(&copy_block->kref, cmf_copy_block_release);
+}
+
+static void cmf_generic_reset(struct ccw_device *cdev)
+{
+ struct cmb_data *cmb_data;
+
+ spin_lock_irq(cdev->ccwlock);
+ cmb_data = cdev->private->cmb;
+ if (cmb_data) {
+ memset(cmb_data->last_block, 0, cmb_data->size);
+ /*
+ * Need to reset hw block as well to make the hardware start
+ * from 0 again.
+ */
+ memset(cmbops->align(cmb_data->hw_block), 0, cmb_data->size);
+ cmb_data->last_update = 0;
+ }
+ cdev->private->cmb_start_time = get_clock();
+ spin_unlock_irq(cdev->ccwlock);
}
/**
@@ -343,8 +519,8 @@ struct cmb {
/* insert a single device into the cmb_area list
* called with cmb_area.lock held from alloc_cmb
*/
-static inline int
-alloc_cmb_single (struct ccw_device *cdev)
+static inline int alloc_cmb_single (struct ccw_device *cdev,
+ struct cmb_data *cmb_data)
{
struct cmb *cmb;
struct ccw_device_private *node;
@@ -358,10 +534,12 @@ alloc_cmb_single (struct ccw_device *cdev)
/* find first unused cmb in cmb_area.mem.
* this is a little tricky: cmb_area.list
- * remains sorted by ->cmb pointers */
+ * remains sorted by ->cmb->hw_data pointers */
cmb = cmb_area.mem;
list_for_each_entry(node, &cmb_area.list, cmb_list) {
- if ((struct cmb*)node->cmb > cmb)
+ struct cmb_data *data;
+ data = node->cmb;
+ if ((struct cmb*)data->hw_block > cmb)
break;
cmb++;
}
@@ -372,7 +550,8 @@ alloc_cmb_single (struct ccw_device *cdev)
/* insert new cmb */
list_add_tail(&cdev->private->cmb_list, &node->cmb_list);
- cdev->private->cmb = cmb;
+ cmb_data->hw_block = cmb;
+ cdev->private->cmb = cmb_data;
ret = 0;
out:
spin_unlock_irq(cdev->ccwlock);
@@ -385,7 +564,19 @@ alloc_cmb (struct ccw_device *cdev)
int ret;
struct cmb *mem;
ssize_t size;
+ struct cmb_data *cmb_data;
+
+ /* Allocate private cmb_data. */
+ cmb_data = kzalloc(sizeof(struct cmb_data), GFP_KERNEL);
+ if (!cmb_data)
+ return -ENOMEM;
+ cmb_data->last_block = kzalloc(sizeof(struct cmb), GFP_KERNEL);
+ if (!cmb_data->last_block) {
+ kfree(cmb_data);
+ return -ENOMEM;
+ }
+ cmb_data->size = sizeof(struct cmb);
spin_lock(&cmb_area.lock);
if (!cmb_area.mem) {
@@ -414,29 +605,36 @@ alloc_cmb (struct ccw_device *cdev)
}
/* do the actual allocation */
- ret = alloc_cmb_single(cdev);
+ ret = alloc_cmb_single(cdev, cmb_data);
out:
spin_unlock(&cmb_area.lock);
-
+ if (ret) {
+ kfree(cmb_data->last_block);
+ kfree(cmb_data);
+ }
return ret;
}
-static void
-free_cmb(struct ccw_device *cdev)
+static void free_cmb(struct ccw_device *cdev)
{
struct ccw_device_private *priv;
-
- priv = cdev->private;
+ struct cmb_data *cmb_data;
spin_lock(&cmb_area.lock);
spin_lock_irq(cdev->ccwlock);
+ priv = cdev->private;
+
if (list_empty(&priv->cmb_list)) {
/* already freed */
goto out;
}
+ cmb_data = priv->cmb;
priv->cmb = NULL;
+ if (cmb_data)
+ kfree(cmb_data->last_block);
+ kfree(cmb_data);
list_del_init(&priv->cmb_list);
if (list_empty(&cmb_area.list)) {
@@ -451,83 +649,97 @@ out:
spin_unlock(&cmb_area.lock);
}
-static int
-set_cmb(struct ccw_device *cdev, u32 mme)
+static int set_cmb(struct ccw_device *cdev, u32 mme)
{
u16 offset;
+ struct cmb_data *cmb_data;
+ unsigned long flags;
- if (!cdev->private->cmb)
+ spin_lock_irqsave(cdev->ccwlock, flags);
+ if (!cdev->private->cmb) {
+ spin_unlock_irqrestore(cdev->ccwlock, flags);
return -EINVAL;
-
- offset = mme ? (struct cmb *)cdev->private->cmb - cmb_area.mem : 0;
+ }
+ cmb_data = cdev->private->cmb;
+ offset = mme ? (struct cmb *)cmb_data->hw_block - cmb_area.mem : 0;
+ spin_unlock_irqrestore(cdev->ccwlock, flags);
return set_schib_wait(cdev, mme, 0, offset);
}
-static u64
-read_cmb (struct ccw_device *cdev, int index)
+static u64 read_cmb (struct ccw_device *cdev, int index)
{
- /* yes, we have to put it on the stack
- * because the cmb must only be accessed
- * atomically, e.g. with mvc */
- struct cmb cmb;
- unsigned long flags;
+ struct cmb *cmb;
u32 val;
+ int ret;
+ unsigned long flags;
+
+ ret = cmf_cmb_copy_wait(cdev);
+ if (ret < 0)
+ return 0;
spin_lock_irqsave(cdev->ccwlock, flags);
if (!cdev->private->cmb) {
- spin_unlock_irqrestore(cdev->ccwlock, flags);
- return 0;
+ ret = 0;
+ goto out;
}
-
- cmb = *(struct cmb*)cdev->private->cmb;
- spin_unlock_irqrestore(cdev->ccwlock, flags);
+ cmb = ((struct cmb_data *)cdev->private->cmb)->last_block;
switch (index) {
case cmb_ssch_rsch_count:
- return cmb.ssch_rsch_count;
+ ret = cmb->ssch_rsch_count;
+ goto out;
case cmb_sample_count:
- return cmb.sample_count;
+ ret = cmb->sample_count;
+ goto out;
case cmb_device_connect_time:
- val = cmb.device_connect_time;
+ val = cmb->device_connect_time;
break;
case cmb_function_pending_time:
- val = cmb.function_pending_time;
+ val = cmb->function_pending_time;
break;
case cmb_device_disconnect_time:
- val = cmb.device_disconnect_time;
+ val = cmb->device_disconnect_time;
break;
case cmb_control_unit_queuing_time:
- val = cmb.control_unit_queuing_time;
+ val = cmb->control_unit_queuing_time;
break;
case cmb_device_active_only_time:
- val = cmb.device_active_only_time;
+ val = cmb->device_active_only_time;
break;
default:
- return 0;
+ ret = 0;
+ goto out;
}
- return time_to_avg_nsec(val, cmb.sample_count);
+ ret = time_to_avg_nsec(val, cmb->sample_count);
+out:
+ spin_unlock_irqrestore(cdev->ccwlock, flags);
+ return ret;
}
-static int
-readall_cmb (struct ccw_device *cdev, struct cmbdata *data)
+static int readall_cmb (struct ccw_device *cdev, struct cmbdata *data)
{
- /* yes, we have to put it on the stack
- * because the cmb must only be accessed
- * atomically, e.g. with mvc */
- struct cmb cmb;
- unsigned long flags;
+ struct cmb *cmb;
+ struct cmb_data *cmb_data;
u64 time;
+ unsigned long flags;
+ int ret;
+ ret = cmf_cmb_copy_wait(cdev);
+ if (ret < 0)
+ return ret;
spin_lock_irqsave(cdev->ccwlock, flags);
- if (!cdev->private->cmb) {
- spin_unlock_irqrestore(cdev->ccwlock, flags);
- return -ENODEV;
+ cmb_data = cdev->private->cmb;
+ if (!cmb_data) {
+ ret = -ENODEV;
+ goto out;
}
-
- cmb = *(struct cmb*)cdev->private->cmb;
- time = get_clock() - cdev->private->cmb_start_time;
- spin_unlock_irqrestore(cdev->ccwlock, flags);
+ if (cmb_data->last_update == 0) {
+ ret = -EAGAIN;
+ goto out;
+ }
+ cmb = cmb_data->last_block;
+ time = cmb_data->last_update - cdev->private->cmb_start_time;
memset(data, 0, sizeof(struct cmbdata));
@@ -538,31 +750,32 @@ readall_cmb (struct ccw_device *cdev, struct cmbdata *data)
data->elapsed_time = (time * 1000) >> 12;
/* copy data to new structure */
- data->ssch_rsch_count = cmb.ssch_rsch_count;
- data->sample_count = cmb.sample_count;
+ data->ssch_rsch_count = cmb->ssch_rsch_count;
+ data->sample_count = cmb->sample_count;
/* time fields are converted to nanoseconds while copying */
- data->device_connect_time = time_to_nsec(cmb.device_connect_time);
- data->function_pending_time = time_to_nsec(cmb.function_pending_time);
- data->device_disconnect_time = time_to_nsec(cmb.device_disconnect_time);
+ data->device_connect_time = time_to_nsec(cmb->device_connect_time);
+ data->function_pending_time = time_to_nsec(cmb->function_pending_time);
+ data->device_disconnect_time =
+ time_to_nsec(cmb->device_disconnect_time);
data->control_unit_queuing_time
- = time_to_nsec(cmb.control_unit_queuing_time);
+ = time_to_nsec(cmb->control_unit_queuing_time);
data->device_active_only_time
- = time_to_nsec(cmb.device_active_only_time);
+ = time_to_nsec(cmb->device_active_only_time);
+ ret = 0;
+out:
+ spin_unlock_irqrestore(cdev->ccwlock, flags);
+ return ret;
+}
- return 0;
+static void reset_cmb(struct ccw_device *cdev)
+{
+ cmf_generic_reset(cdev);
}
-static void
-reset_cmb(struct ccw_device *cdev)
+static void * align_cmb(void *area)
{
- struct cmb *cmb;
- spin_lock_irq(cdev->ccwlock);
- cmb = cdev->private->cmb;
- if (cmb)
- memset (cmb, 0, sizeof (*cmb));
- cdev->private->cmb_start_time = get_clock();
- spin_unlock_irq(cdev->ccwlock);
+ return area;
}
static struct attribute_group cmf_attr_group;
@@ -574,6 +787,7 @@ static struct cmb_operations cmbops_basic = {
.read = read_cmb,
.readall = readall_cmb,
.reset = reset_cmb,
+ .align = align_cmb,
.attr_group = &cmf_attr_group,
};
@@ -610,22 +824,34 @@ static inline struct cmbe* cmbe_align(struct cmbe *c)
return (struct cmbe*)addr;
}
-static int
-alloc_cmbe (struct ccw_device *cdev)
+static int alloc_cmbe (struct ccw_device *cdev)
{
struct cmbe *cmbe;
- cmbe = kmalloc (sizeof (*cmbe) * 2, GFP_KERNEL);
+ struct cmb_data *cmb_data;
+ int ret;
+
+ cmbe = kzalloc (sizeof (*cmbe) * 2, GFP_KERNEL);
if (!cmbe)
return -ENOMEM;
-
+ cmb_data = kzalloc(sizeof(struct cmb_data), GFP_KERNEL);
+ if (!cmb_data) {
+ ret = -ENOMEM;
+ goto out_free;
+ }
+ cmb_data->last_block = kzalloc(sizeof(struct cmbe), GFP_KERNEL);
+ if (!cmb_data->last_block) {
+ ret = -ENOMEM;
+ goto out_free;
+ }
+ cmb_data->size = sizeof(struct cmbe);
spin_lock_irq(cdev->ccwlock);
if (cdev->private->cmb) {
- kfree(cmbe);
spin_unlock_irq(cdev->ccwlock);
- return -EBUSY;
+ ret = -EBUSY;
+ goto out_free;
}
-
- cdev->private->cmb = cmbe;
+ cmb_data->hw_block = cmbe;
+ cdev->private->cmb = cmb_data;
spin_unlock_irq(cdev->ccwlock);
/* activate global measurement if this is the first channel */
@@ -636,14 +862,24 @@ alloc_cmbe (struct ccw_device *cdev)
spin_unlock(&cmb_area.lock);
return 0;
+out_free:
+ if (cmb_data)
+ kfree(cmb_data->last_block);
+ kfree(cmb_data);
+ kfree(cmbe);
+ return ret;
}
-static void
-free_cmbe (struct ccw_device *cdev)
+static void free_cmbe (struct ccw_device *cdev)
{
+ struct cmb_data *cmb_data;
+
spin_lock_irq(cdev->ccwlock);
- kfree(cdev->private->cmb);
+ cmb_data = cdev->private->cmb;
cdev->private->cmb = NULL;
+ if (cmb_data)
+ kfree(cmb_data->last_block);
+ kfree(cmb_data);
spin_unlock_irq(cdev->ccwlock);
/* deactivate global measurement if this is the last channel */
@@ -654,89 +890,105 @@ free_cmbe (struct ccw_device *cdev)
spin_unlock(&cmb_area.lock);
}
-static int
-set_cmbe(struct ccw_device *cdev, u32 mme)
+static int set_cmbe(struct ccw_device *cdev, u32 mme)
{
unsigned long mba;
+ struct cmb_data *cmb_data;
+ unsigned long flags;
- if (!cdev->private->cmb)
+ spin_lock_irqsave(cdev->ccwlock, flags);
+ if (!cdev->private->cmb) {
+ spin_unlock_irqrestore(cdev->ccwlock, flags);
return -EINVAL;
- mba = mme ? (unsigned long) cmbe_align(cdev->private->cmb) : 0;
+ }
+ cmb_data = cdev->private->cmb;
+ mba = mme ? (unsigned long) cmbe_align(cmb_data->hw_block) : 0;
+ spin_unlock_irqrestore(cdev->ccwlock, flags);
return set_schib_wait(cdev, mme, 1, mba);
}
-u64
-read_cmbe (struct ccw_device *cdev, int index)
+static u64 read_cmbe (struct ccw_device *cdev, int index)
{
- /* yes, we have to put it on the stack
- * because the cmb must only be accessed
- * atomically, e.g. with mvc */
- struct cmbe cmb;
- unsigned long flags;
+ struct cmbe *cmb;
+ struct cmb_data *cmb_data;
u32 val;
+ int ret;
+ unsigned long flags;
- spin_lock_irqsave(cdev->ccwlock, flags);
- if (!cdev->private->cmb) {
- spin_unlock_irqrestore(cdev->ccwlock, flags);
+ ret = cmf_cmb_copy_wait(cdev);
+ if (ret < 0)
return 0;
- }
- cmb = *cmbe_align(cdev->private->cmb);
- spin_unlock_irqrestore(cdev->ccwlock, flags);
+ spin_lock_irqsave(cdev->ccwlock, flags);
+ cmb_data = cdev->private->cmb;
+ if (!cmb_data) {
+ ret = 0;
+ goto out;
+ }
+ cmb = cmb_data->last_block;
switch (index) {
case cmb_ssch_rsch_count:
- return cmb.ssch_rsch_count;
+ ret = cmb->ssch_rsch_count;
+ goto out;
case cmb_sample_count:
- return cmb.sample_count;
+ ret = cmb->sample_count;
+ goto out;
case cmb_device_connect_time:
- val = cmb.device_connect_time;
+ val = cmb->device_connect_time;
break;
case cmb_function_pending_time:
- val = cmb.function_pending_time;
+ val = cmb->function_pending_time;
break;
case cmb_device_disconnect_time:
- val = cmb.device_disconnect_time;
+ val = cmb->device_disconnect_time;
break;
case cmb_control_unit_queuing_time:
- val = cmb.control_unit_queuing_time;
+ val = cmb->control_unit_queuing_time;
break;
case cmb_device_active_only_time:
- val = cmb.device_active_only_time;
+ val = cmb->device_active_only_time;
break;
case cmb_device_busy_time:
- val = cmb.device_busy_time;
+ val = cmb->device_busy_time;
break;
case cmb_initial_command_response_time:
- val = cmb.initial_command_response_time;
+ val = cmb->initial_command_response_time;
break;
default:
- return 0;
+ ret = 0;
+ goto out;
}
- return time_to_avg_nsec(val, cmb.sample_count);
+ ret = time_to_avg_nsec(val, cmb->sample_count);
+out:
+ spin_unlock_irqrestore(cdev->ccwlock, flags);
+ return ret;
}
-static int
-readall_cmbe (struct ccw_device *cdev, struct cmbdata *data)
+static int readall_cmbe (struct ccw_device *cdev, struct cmbdata *data)
{
- /* yes, we have to put it on the stack
- * because the cmb must only be accessed
- * atomically, e.g. with mvc */
- struct cmbe cmb;
- unsigned long flags;
+ struct cmbe *cmb;
+ struct cmb_data *cmb_data;
u64 time;
+ unsigned long flags;
+ int ret;
+ ret = cmf_cmb_copy_wait(cdev);
+ if (ret < 0)
+ return ret;
spin_lock_irqsave(cdev->ccwlock, flags);
- if (!cdev->private->cmb) {
- spin_unlock_irqrestore(cdev->ccwlock, flags);
- return -ENODEV;
+ cmb_data = cdev->private->cmb;
+ if (!cmb_data) {
+ ret = -ENODEV;
+ goto out;
}
-
- cmb = *cmbe_align(cdev->private->cmb);
- time = get_clock() - cdev->private->cmb_start_time;
- spin_unlock_irqrestore(cdev->ccwlock, flags);
+ if (cmb_data->last_update == 0) {
+ ret = -EAGAIN;
+ goto out;
+ }
+ time = cmb_data->last_update - cdev->private->cmb_start_time;
memset (data, 0, sizeof(struct cmbdata));
@@ -746,35 +998,38 @@ readall_cmbe (struct ccw_device *cdev, struct cmbdata *data)
/* conver to nanoseconds */
data->elapsed_time = (time * 1000) >> 12;
+ cmb = cmb_data->last_block;
/* copy data to new structure */
- data->ssch_rsch_count = cmb.ssch_rsch_count;
- data->sample_count = cmb.sample_count;
+ data->ssch_rsch_count = cmb->ssch_rsch_count;
+ data->sample_count = cmb->sample_count;
/* time fields are converted to nanoseconds while copying */
- data->device_connect_time = time_to_nsec(cmb.device_connect_time);
- data->function_pending_time = time_to_nsec(cmb.function_pending_time);
- data->device_disconnect_time = time_to_nsec(cmb.device_disconnect_time);
+ data->device_connect_time = time_to_nsec(cmb->device_connect_time);
+ data->function_pending_time = time_to_nsec(cmb->function_pending_time);
+ data->device_disconnect_time =
+ time_to_nsec(cmb->device_disconnect_time);
data->control_unit_queuing_time
- = time_to_nsec(cmb.control_unit_queuing_time);
+ = time_to_nsec(cmb->control_unit_queuing_time);
data->device_active_only_time
- = time_to_nsec(cmb.device_active_only_time);
- data->device_busy_time = time_to_nsec(cmb.device_busy_time);
+ = time_to_nsec(cmb->device_active_only_time);
+ data->device_busy_time = time_to_nsec(cmb->device_busy_time);
data->initial_command_response_time
- = time_to_nsec(cmb.initial_command_response_time);
+ = time_to_nsec(cmb->initial_command_response_time);
- return 0;
+ ret = 0;
+out:
+ spin_unlock_irqrestore(cdev->ccwlock, flags);
+ return ret;
}
-static void
-reset_cmbe(struct ccw_device *cdev)
+static void reset_cmbe(struct ccw_device *cdev)
{
- struct cmbe *cmb;
- spin_lock_irq(cdev->ccwlock);
- cmb = cmbe_align(cdev->private->cmb);
- if (cmb)
- memset (cmb, 0, sizeof (*cmb));
- cdev->private->cmb_start_time = get_clock();
- spin_unlock_irq(cdev->ccwlock);
+ cmf_generic_reset(cdev);
+}
+
+static void * align_cmbe(void *area)
+{
+ return cmbe_align(area);
}
static struct attribute_group cmf_attr_group_ext;
@@ -786,6 +1041,7 @@ static struct cmb_operations cmbops_extended = {
.read = read_cmbe,
.readall = readall_cmbe,
.reset = reset_cmbe,
+ .align = align_cmbe,
.attr_group = &cmf_attr_group_ext,
};
@@ -803,14 +1059,19 @@ cmb_show_avg_sample_interval(struct device *dev, struct device_attribute *attr,
struct ccw_device *cdev;
long interval;
unsigned long count;
+ struct cmb_data *cmb_data;
cdev = to_ccwdev(dev);
- interval = get_clock() - cdev->private->cmb_start_time;
count = cmf_read(cdev, cmb_sample_count);
- if (count)
+ spin_lock_irq(cdev->ccwlock);
+ cmb_data = cdev->private->cmb;
+ if (count) {
+ interval = cmb_data->last_update -
+ cdev->private->cmb_start_time;
interval /= count;
- else
+ } else
interval = -1;
+ spin_unlock_irq(cdev->ccwlock);
return sprintf(buf, "%ld\n", interval);
}
@@ -823,7 +1084,10 @@ cmb_show_avg_utilization(struct device *dev, struct device_attribute *attr, char
int ret;
ret = cmf_readall(to_ccwdev(dev), &data);
- if (ret)
+ if (ret == -EAGAIN || ret == -ENODEV)
+ /* No data (yet/currently) available to use for calculation. */
+ return sprintf(buf, "n/a\n");
+ else if (ret)
return ret;
utilization = data.device_connect_time +
@@ -982,6 +1246,13 @@ cmf_readall(struct ccw_device *cdev, struct cmbdata *data)
return cmbops->readall(cdev, data);
}
+/* Reenable cmf when a disconnected device becomes available again. */
+int cmf_reenable(struct ccw_device *cdev)
+{
+ cmbops->reset(cdev);
+ return cmbops->set(cdev, 2);
+}
+
static int __init
init_cmf(void)
{
diff --git a/drivers/s390/cio/device.h b/drivers/s390/cio/device.h
index 8e0d1db..00be9a5 100644
--- a/drivers/s390/cio/device.h
+++ b/drivers/s390/cio/device.h
@@ -27,6 +27,7 @@ enum dev_state {
DEV_STATE_DISCONNECTED,
DEV_STATE_DISCONNECTED_SENSE_ID,
DEV_STATE_CMFCHANGE,
+ DEV_STATE_CMFUPDATE,
/* last element! */
NR_DEV_STATES
};
@@ -118,5 +119,8 @@ int ccw_device_stlck(struct ccw_device *);
void ccw_device_set_timeout(struct ccw_device *, int);
extern struct subchannel_id ccw_device_get_subchannel_id(struct ccw_device *);
+/* Channel measurement facility related */
void retry_set_schib(struct ccw_device *cdev);
+void cmf_retry_copy_block(struct ccw_device *);
+int cmf_reenable(struct ccw_device *);
#endif
diff --git a/drivers/s390/cio/device_fsm.c b/drivers/s390/cio/device_fsm.c
index 5ec8ef1..7d0dd72 100644
--- a/drivers/s390/cio/device_fsm.c
+++ b/drivers/s390/cio/device_fsm.c
@@ -336,8 +336,11 @@ ccw_device_oper_notify(void *data)
if (!ret)
/* Driver doesn't want device back. */
ccw_device_do_unreg_rereg((void *)cdev);
- else
+ else {
+ /* Reenable channel measurements, if needed. */
+ cmf_reenable(cdev);
wake_up(&cdev->private->wait_q);
+ }
}
/*
@@ -1095,6 +1098,13 @@ ccw_device_change_cmfstate(struct ccw_device *cdev, enum dev_event dev_event)
dev_fsm_event(cdev, dev_event);
}
+static void ccw_device_update_cmfblock(struct ccw_device *cdev,
+ enum dev_event dev_event)
+{
+ cmf_retry_copy_block(cdev);
+ cdev->private->state = DEV_STATE_ONLINE;
+ dev_fsm_event(cdev, dev_event);
+}
static void
ccw_device_quiesce_done(struct ccw_device *cdev, enum dev_event dev_event)
@@ -1249,6 +1259,12 @@ fsm_func_t *dev_jumptable[NR_DEV_STATES][NR_DEV_EVENTS] = {
[DEV_EVENT_TIMEOUT] = ccw_device_change_cmfstate,
[DEV_EVENT_VERIFY] = ccw_device_change_cmfstate,
},
+ [DEV_STATE_CMFUPDATE] = {
+ [DEV_EVENT_NOTOPER] = ccw_device_update_cmfblock,
+ [DEV_EVENT_INTERRUPT] = ccw_device_update_cmfblock,
+ [DEV_EVENT_TIMEOUT] = ccw_device_update_cmfblock,
+ [DEV_EVENT_VERIFY] = ccw_device_update_cmfblock,
+ },
};
/*
OpenPOWER on IntegriCloud