[PATCH v4 02/16] staging: unisys: Convert pending_msg_hdr to a pointer
Benjamin Romer
benjamin.romer at unisys.com
Mon Jun 1 17:00:27 UTC 2015
From: Don Zickus <dzickus at redhat.com>
In order for bus/dev_info structs to become public structs, one
element, pending_msg_hdr, needs to become opaque. This is to keep
all the internals of the controlvm struct private to the bus layer.
So a simple conversion of embedding the pending_msg_hdr struct into
a pointer is done. The rest of the patch is the fallout.
The rules are modified slightly. Instead of relying on the 'id' to be
CONTROLVM_INVALID to indicate invalid, just use the pointer set to NULL.
In addition, because bus/dev_info can be NULL and we still need to send
a
response, pass pending_msg_hdr to all 'responders' instead of bus/
That change causes some fallout in the success case. Instead of setting
state
bits and clearing info in the responders, do all that magic in the
responder
wrappers.
Signed-off-by: Don Zickus <dzickus at redhat.com>
Signed-off-by: Benjamin Romer <benjamin.romer at unisys.com>
---
drivers/staging/unisys/visorbus/visorbus_private.h | 4 +-
drivers/staging/unisys/visorbus/visorchipset.c | 185 +++++++++++++--------
2 files changed, 118 insertions(+), 71 deletions(-)
diff --git a/drivers/staging/unisys/visorbus/visorbus_private.h b/drivers/staging/unisys/visorbus/visorbus_private.h
index 912b6cd..f371f9d 100644
--- a/drivers/staging/unisys/visorbus/visorbus_private.h
+++ b/drivers/staging/unisys/visorbus/visorbus_private.h
@@ -57,7 +57,7 @@ struct visorchipset_device_info {
u64 reserved2;
u32 switch_no; /* when devState.attached==1 */
u32 internal_port_no; /* when devState.attached==1 */
- struct controlvm_message_header pending_msg_hdr;/* CONTROLVM_MESSAGE */
+ struct controlvm_message_header *pending_msg_hdr;/* CONTROLVM_MESSAGE */
/** For private use by the bus driver */
void *bus_driver_context;
};
@@ -84,7 +84,7 @@ struct visorchipset_bus_info {
/* Add new fields above. */
/* Remaining bits in this 32-bit word are unused. */
} flags;
- struct controlvm_message_header pending_msg_hdr;/* CONTROLVM MsgHdr */
+ struct controlvm_message_header *pending_msg_hdr;/* CONTROLVM MsgHdr */
/** For private use by the bus driver */
void *bus_driver_context;
};
diff --git a/drivers/staging/unisys/visorbus/visorchipset.c b/drivers/staging/unisys/visorbus/visorchipset.c
index 3d08661..2478889 100644
--- a/drivers/staging/unisys/visorbus/visorchipset.c
+++ b/drivers/staging/unisys/visorbus/visorchipset.c
@@ -962,37 +962,17 @@ enum crash_obj_type {
};
static void
-bus_responder(enum controlvm_id cmd_id, struct visorchipset_bus_info *p,
+bus_responder(enum controlvm_id cmd_id,
+ struct controlvm_message_header *pending_msg_hdr,
int response)
{
- bool need_clear = false;
- u32 bus_no = p->bus_no;
+ if (pending_msg_hdr == NULL)
+ return; /* no controlvm response needed */
- if (!p)
+ if (pending_msg_hdr->id != (u32)cmd_id)
return;
- if (response < 0) {
- if ((cmd_id == CONTROLVM_BUS_CREATE) &&
- (response != (-CONTROLVM_RESP_ERROR_ALREADY_DONE)))
- /* undo the row we just created... */
- busdevices_del(&dev_info_list, bus_no);
- } else {
- if (cmd_id == CONTROLVM_BUS_CREATE)
- p->state.created = 1;
- if (cmd_id == CONTROLVM_BUS_DESTROY)
- need_clear = true;
- }
-
- if (p->pending_msg_hdr.id == CONTROLVM_INVALID)
- return; /* no controlvm response needed */
- if (p->pending_msg_hdr.id != (u32)cmd_id)
- return;
- controlvm_respond(&p->pending_msg_hdr, response);
- p->pending_msg_hdr.id = CONTROLVM_INVALID;
- if (need_clear) {
- bus_info_clear(p);
- busdevices_del(&dev_info_list, bus_no);
- }
+ controlvm_respond(pending_msg_hdr, response);
}
static void
@@ -1004,14 +984,12 @@ device_changestate_responder(enum controlvm_id cmd_id,
u32 bus_no = p->bus_no;
u32 dev_no = p->dev_no;
- if (!p)
- return;
- if (p->pending_msg_hdr.id == CONTROLVM_INVALID)
+ if (p->pending_msg_hdr == NULL)
return; /* no controlvm response needed */
- if (p->pending_msg_hdr.id != cmd_id)
+ if (p->pending_msg_hdr->id != cmd_id)
return;
- controlvm_init_response(&outmsg, &p->pending_msg_hdr, response);
+ controlvm_init_response(&outmsg, p->pending_msg_hdr, response);
outmsg.cmd.device_change_state.bus_no = bus_no;
outmsg.cmd.device_change_state.dev_no = dev_no;
@@ -1020,35 +998,20 @@ device_changestate_responder(enum controlvm_id cmd_id,
if (!visorchannel_signalinsert(controlvm_channel,
CONTROLVM_QUEUE_REQUEST, &outmsg))
return;
-
- p->pending_msg_hdr.id = CONTROLVM_INVALID;
}
static void
-device_responder(enum controlvm_id cmd_id, struct visorchipset_device_info *p,
+device_responder(enum controlvm_id cmd_id,
+ struct controlvm_message_header *pending_msg_hdr,
int response)
{
- bool need_clear = false;
-
- if (!p)
- return;
- if (response >= 0) {
- if (cmd_id == CONTROLVM_DEVICE_CREATE)
- p->state.created = 1;
- if (cmd_id == CONTROLVM_DEVICE_DESTROY)
- need_clear = true;
- }
-
- if (p->pending_msg_hdr.id == CONTROLVM_INVALID)
+ if (pending_msg_hdr == NULL)
return; /* no controlvm response needed */
- if (p->pending_msg_hdr.id != (u32)cmd_id)
+ if (pending_msg_hdr->id != (u32)cmd_id)
return;
- controlvm_respond(&p->pending_msg_hdr, response);
- p->pending_msg_hdr.id = CONTROLVM_INVALID;
- if (need_clear)
- dev_info_clear(p);
+ controlvm_respond(pending_msg_hdr, response);
}
static void
@@ -1057,15 +1020,32 @@ bus_epilog(struct visorchipset_bus_info *bus_info,
int response, bool need_response)
{
bool notified = false;
+ struct controlvm_message_header *pmsg_hdr = NULL;
- if (!bus_info)
- return;
+ if (!bus_info) {
+ /* relying on a valid passed in response code */
+ /* be lazy and re-use msg_hdr for this failure, is this ok?? */
+ pmsg_hdr = msg_hdr;
+ goto away;
+ }
+
+ if (bus_info->pending_msg_hdr) {
+ /* only non-NULL if dev is still waiting on a response */
+ response = -CONTROLVM_RESP_ERROR_MESSAGE_ID_INVALID_FOR_CLIENT;
+ pmsg_hdr = bus_info->pending_msg_hdr;
+ goto away;
+ }
if (need_response) {
- memcpy(&bus_info->pending_msg_hdr, msg_hdr,
+ pmsg_hdr = kzalloc(sizeof(*pmsg_hdr), GFP_KERNEL);
+ if (!pmsg_hdr) {
+ response = -CONTROLVM_RESP_ERROR_KMALLOC_FAILED;
+ goto away;
+ }
+
+ memcpy(pmsg_hdr, msg_hdr,
sizeof(struct controlvm_message_header));
- } else {
- bus_info->pending_msg_hdr.id = CONTROLVM_INVALID;
+ bus_info->pending_msg_hdr = pmsg_hdr;
}
down(¬ifier_lock);
@@ -1085,6 +1065,7 @@ bus_epilog(struct visorchipset_bus_info *bus_info,
break;
}
}
+away:
if (notified)
/* The callback function just called above is responsible
* for calling the appropriate visorchipset_busdev_responders
@@ -1092,7 +1073,12 @@ bus_epilog(struct visorchipset_bus_info *bus_info,
*/
;
else
- bus_responder(cmd, bus_info, response);
+ /*
+ * Do not kfree(pmsg_hdr) as this is the failure path.
+ * The success path ('notified') will call the responder
+ * directly and kfree() there.
+ */
+ bus_responder(cmd, pmsg_hdr, response);
up(¬ifier_lock);
}
@@ -1106,22 +1092,39 @@ device_epilog(struct visorchipset_device_info *dev_info,
bool notified = false;
u32 bus_no = dev_info->bus_no;
u32 dev_no = dev_info->dev_no;
+ struct controlvm_message_header *pmsg_hdr = NULL;
char *envp[] = {
"SPARSP_DIAGPOOL_PAUSED_STATE = 1",
NULL
};
- if (!dev_info)
- return;
-
notifiers = &busdev_notifiers;
+ if (!dev_info) {
+ /* relying on a valid passed in response code */
+ /* be lazy and re-use msg_hdr for this failure, is this ok?? */
+ pmsg_hdr = msg_hdr;
+ goto away;
+ }
+
+ if (dev_info->pending_msg_hdr) {
+ /* only non-NULL if dev is still waiting on a response */
+ response = -CONTROLVM_RESP_ERROR_MESSAGE_ID_INVALID_FOR_CLIENT;
+ pmsg_hdr = dev_info->pending_msg_hdr;
+ goto away;
+ }
+
if (need_response) {
- memcpy(&dev_info->pending_msg_hdr, msg_hdr,
+ pmsg_hdr = kzalloc(sizeof(*pmsg_hdr), GFP_KERNEL);
+ if (!pmsg_hdr) {
+ response = -CONTROLVM_RESP_ERROR_KMALLOC_FAILED;
+ goto away;
+ }
+
+ memcpy(pmsg_hdr, msg_hdr,
sizeof(struct controlvm_message_header));
- } else {
- dev_info->pending_msg_hdr.id = CONTROLVM_INVALID;
+ dev_info->pending_msg_hdr = pmsg_hdr;
}
down(¬ifier_lock);
@@ -1179,6 +1182,7 @@ device_epilog(struct visorchipset_device_info *dev_info,
break;
}
}
+away:
if (notified)
/* The callback function just called above is responsible
* for calling the appropriate visorchipset_busdev_responders
@@ -1186,7 +1190,12 @@ device_epilog(struct visorchipset_device_info *dev_info,
*/
;
else
- device_responder(cmd, dev_info, response);
+ /*
+ * Do not kfree(pmsg_hdr) as this is the failure path.
+ * The success path ('notified') will call the responder
+ * directly and kfree() there.
+ */
+ device_responder(cmd, pmsg_hdr, response);
up(¬ifier_lock);
}
@@ -1285,7 +1294,7 @@ bus_configure(struct controlvm_message *inmsg,
POSTCODE_LINUX_3(BUS_CONFIGURE_FAILURE_PC, bus_no,
POSTCODE_SEVERITY_ERR);
rc = -CONTROLVM_RESP_ERROR_BUS_INVALID;
- } else if (bus_info->pending_msg_hdr.id != CONTROLVM_INVALID) {
+ } else if (bus_info->pending_msg_hdr != NULL) {
POSTCODE_LINUX_3(BUS_CONFIGURE_FAILURE_PC, bus_no,
POSTCODE_SEVERITY_ERR);
rc = -CONTROLVM_RESP_ERROR_MESSAGE_ID_INVALID_FOR_CLIENT;
@@ -2127,25 +2136,57 @@ cleanup:
static void
bus_create_response(struct visorchipset_bus_info *bus_info, int response)
{
- bus_responder(CONTROLVM_BUS_CREATE, bus_info, response);
+ if (response >= 0) {
+ bus_info->state.created = 1;
+ } else {
+ if (response != -CONTROLVM_RESP_ERROR_ALREADY_DONE)
+ /* undo the row we just created... */
+ busdevices_del(&dev_info_list, bus_info->bus_no);
+ }
+
+ bus_responder(CONTROLVM_BUS_CREATE, bus_info->pending_msg_hdr,
+ response);
+
+ kfree(bus_info->pending_msg_hdr);
+ bus_info->pending_msg_hdr = NULL;
}
static void
bus_destroy_response(struct visorchipset_bus_info *bus_info, int response)
{
- bus_responder(CONTROLVM_BUS_DESTROY, bus_info, response);
+ bus_responder(CONTROLVM_BUS_DESTROY, bus_info->pending_msg_hdr,
+ response);
+
+ kfree(bus_info->pending_msg_hdr);
+ bus_info->pending_msg_hdr = NULL;
+
+ bus_info_clear(bus_info);
+ busdevices_del(&dev_info_list, bus_info->bus_no);
}
static void
device_create_response(struct visorchipset_device_info *dev_info, int response)
{
- device_responder(CONTROLVM_DEVICE_CREATE, dev_info, response);
+ if (response >= 0)
+ dev_info->state.created = 1;
+
+ device_responder(CONTROLVM_DEVICE_CREATE, dev_info->pending_msg_hdr,
+ response);
+
+ kfree(dev_info->pending_msg_hdr);
+ dev_info->pending_msg_hdr = NULL;
}
static void
device_destroy_response(struct visorchipset_device_info *dev_info, int response)
{
- device_responder(CONTROLVM_DEVICE_DESTROY, dev_info, response);
+ device_responder(CONTROLVM_DEVICE_DESTROY, dev_info->pending_msg_hdr,
+ response);
+
+ kfree(dev_info->pending_msg_hdr);
+ dev_info->pending_msg_hdr = NULL;
+
+ dev_info_clear(dev_info);
}
static void
@@ -2155,6 +2196,9 @@ visorchipset_device_pause_response(struct visorchipset_device_info *dev_info,
device_changestate_responder(CONTROLVM_DEVICE_CHANGESTATE,
dev_info, response,
segment_state_standby);
+
+ kfree(dev_info->pending_msg_hdr);
+ dev_info->pending_msg_hdr = NULL;
}
static void
@@ -2163,6 +2207,9 @@ device_resume_response(struct visorchipset_device_info *dev_info, int response)
device_changestate_responder(CONTROLVM_DEVICE_CHANGESTATE,
dev_info, response,
segment_state_running);
+
+ kfree(dev_info->pending_msg_hdr);
+ dev_info->pending_msg_hdr = NULL;
}
bool
--
2.1.4
More information about the devel
mailing list