lkml.org 
[lkml]   [2004]   [Dec]   [28]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
Patch in this message
/
Date
From
Subject[PATCH 8/8] s390: SCLP device driver cleanup
[PATCH 8/8] s390: sclp driver.

From: Peter Oberparleiter <peter.oberparleiter@de.ibm.com>

sclp: core driver cleanup

Details:
- moved signal shutdown (quiesce) handling into a separate file
- cleanup of SCLP core driver:
. introduced driver states instead of bits
. introduced request retry count and retry limit
. sclp_add_request now returns an error code if a request couldn't be started
. introduced separate request structure for init_mask requests to simplify
code
. request timer is now manually checked in sclp_sync_wait because timer
interrupts are disabled in this context
. removed busy timer - request timer now handles both cases
. split up sclp_start_request into __sclp_start_request and sclp_process
queue
. removed sclp_error_message (unused)
. introduced sclp_check_handler function to split up initial init mask
test from standard init mask request processing
. introduced sclp_deactivate and sclp_reactivate for simplified reboot
event handling (and potential use in suspend/resume scenario)
. added protection against multiple concurrent init mask calls
- minor changes in SCLP core driver:
. updated comments
. renamed functions to be consistent with "function name starts with __ =>
needs lock"
. renamed internal functions for consistency reasons
. introduced inlined helper functions to simplify code
. moved EXPORT_SYMBOL definitions next to function definition
- changes in sclp console driver
. removed callback recursion to prevent stack overflow
- changes to CPI module
. added check for sclp_add_request return code
. changed printks to specify a message level
- changes to generic sclp tty layer
. removed timed buffer retry after error (timers may not work in some
situations)
. introduced return code for sclp_emit_buffer
- changes to sclp tty driver
. removed callback recursion
- changes to sclp vt220 driver
. removed callback recursion
. removed timed buffer retry after error
- modified sclp_init_mask to prevent problems with some compiler versions

Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>

diffstat:
drivers/s390/char/Makefile | 2
drivers/s390/char/sclp.c | 1266 ++++++++++++++++++++-------------------
drivers/s390/char/sclp.h | 6
drivers/s390/char/sclp_con.c | 39 -
drivers/s390/char/sclp_cpi.c | 19
drivers/s390/char/sclp_quiesce.c | 114 +++
drivers/s390/char/sclp_rw.c | 62 -
drivers/s390/char/sclp_rw.h | 4
drivers/s390/char/sclp_tty.c | 40 -
drivers/s390/char/sclp_vt220.c | 94 +-
10 files changed, 896 insertions(+), 750 deletions(-)

diff -urN linux-2.6/drivers/s390/char/Makefile linux-2.6-patched/drivers/s390/char/Makefile
--- linux-2.6/drivers/s390/char/Makefile 2004-12-28 08:50:53.000000000 +0100
+++ linux-2.6-patched/drivers/s390/char/Makefile 2004-12-28 08:50:53.000000000 +0100
@@ -11,7 +11,7 @@

obj-$(CONFIG_TN3215) += con3215.o

-obj-$(CONFIG_SCLP) += sclp.o sclp_rw.o
+obj-$(CONFIG_SCLP) += sclp.o sclp_rw.o sclp_quiesce.o
obj-$(CONFIG_SCLP_TTY) += sclp_tty.o
obj-$(CONFIG_SCLP_CONSOLE) += sclp_con.o
obj-$(CONFIG_SCLP_VT220_TTY) += sclp_vt220.o
diff -urN linux-2.6/drivers/s390/char/sclp.c linux-2.6-patched/drivers/s390/char/sclp.c
--- linux-2.6/drivers/s390/char/sclp.c 2004-12-24 22:35:00.000000000 +0100
+++ linux-2.6-patched/drivers/s390/char/sclp.c 2004-12-28 08:50:53.000000000 +0100
@@ -8,82 +8,99 @@
* Martin Schwidefsky <schwidefsky@de.ibm.com>
*/

-#include <linux/config.h>
#include <linux/module.h>
-#include <linux/kmod.h>
-#include <linux/bootmem.h>
#include <linux/err.h>
-#include <linux/ptrace.h>
-#include <linux/slab.h>
#include <linux/spinlock.h>
#include <linux/interrupt.h>
#include <linux/timer.h>
-#include <linux/init.h>
-#include <linux/cpumask.h>
#include <linux/reboot.h>
+#include <linux/jiffies.h>
+#include <asm/types.h>
#include <asm/s390_ext.h>
-#include <asm/processor.h>

#include "sclp.h"

-#define SCLP_CORE_PRINT_HEADER "sclp low level driver: "
+#define SCLP_HEADER "sclp: "

/* Structure for register_early_external_interrupt. */
static ext_int_info_t ext_int_info_hwc;

-/* spinlock to protect global variables of sclp_core */
-static spinlock_t sclp_lock;
+/* Lock to protect internal data consistency. */
+static spinlock_t sclp_lock = SPIN_LOCK_UNLOCKED;

-/* Mask of valid sclp events */
+/* Mask of events that we can receive from the sclp interface. */
static sccb_mask_t sclp_receive_mask;
+
+/* Mask of events that we can send to the sclp interface. */
static sccb_mask_t sclp_send_mask;

-/* List of registered event types */
+/* List of registered event listeners and senders. */
static struct list_head sclp_reg_list;

-/* sccb queue */
+/* List of queued requests. */
static struct list_head sclp_req_queue;

-/* sccb for unconditional read */
+/* Data for read and and init requests. */
static struct sclp_req sclp_read_req;
+static struct sclp_req sclp_init_req;
static char sclp_read_sccb[PAGE_SIZE] __attribute__((__aligned__(PAGE_SIZE)));
-/* sccb for write mask sccb */
static char sclp_init_sccb[PAGE_SIZE] __attribute__((__aligned__(PAGE_SIZE)));

-/* Timer for init mask retries. */
-static struct timer_list retry_timer;
-
-/* Timer for busy retries. */
-static struct timer_list sclp_busy_timer;
+/* Timer for request retries. */
+static struct timer_list sclp_request_timer;

-static volatile unsigned long sclp_status = 0;
-/* some status flags */
-#define SCLP_INIT 0
-#define SCLP_RUNNING 1
-#define SCLP_READING 2
-#define SCLP_SHUTDOWN 3
-
-#define SCLP_INIT_POLL_INTERVAL 1
-#define SCLP_BUSY_POLL_INTERVAL 1
-
-#define SCLP_COMMAND_INITIATED 0
-#define SCLP_BUSY 2
-#define SCLP_NOT_OPERATIONAL 3
+/* Internal state: is the driver initialized? */
+static volatile enum sclp_init_state_t {
+ sclp_init_state_uninitialized,
+ sclp_init_state_initializing,
+ sclp_init_state_initialized
+} sclp_init_state = sclp_init_state_uninitialized;
+
+/* Internal state: is a request active at the sclp? */
+static volatile enum sclp_running_state_t {
+ sclp_running_state_idle,
+ sclp_running_state_running
+} sclp_running_state = sclp_running_state_idle;
+
+/* Internal state: is a read request pending? */
+static volatile enum sclp_reading_state_t {
+ sclp_reading_state_idle,
+ sclp_reading_state_reading
+} sclp_reading_state = sclp_reading_state_idle;
+
+/* Internal state: is the driver currently serving requests? */
+static volatile enum sclp_activation_state_t {
+ sclp_activation_state_active,
+ sclp_activation_state_deactivating,
+ sclp_activation_state_inactive,
+ sclp_activation_state_activating
+} sclp_activation_state = sclp_activation_state_active;
+
+/* Internal state: is an init mask request pending? */
+static volatile enum sclp_mask_state_t {
+ sclp_mask_state_idle,
+ sclp_mask_state_initializing
+} sclp_mask_state = sclp_mask_state_idle;
+
+/* Maximum retry counts */
+#define SCLP_INIT_RETRY 3
+#define SCLP_MASK_RETRY 3
+#define SCLP_REQUEST_RETRY 3
+
+/* Timeout intervals in seconds.*/
+#define SCLP_BUSY_INTERVAL 2
+#define SCLP_RETRY_INTERVAL 5
+
+static void sclp_process_queue(void);
+static int sclp_init_mask(int calculate);
+static int sclp_init(void);

-/*
- * assembler instruction for Service Call
- */
+/* Perform service call. Return 0 on success, non-zero otherwise. */
static int
-__service_call(sclp_cmdw_t command, void *sccb)
+service_call(sclp_cmdw_t command, void *sccb)
{
int cc;

- /*
- * Mnemonic: SERVC Rx, Ry [RRE]
- *
- * Rx: SCLP command word
- * Ry: address of SCCB
- */
__asm__ __volatile__(
" .insn rre,0xb2200000,%1,%2\n" /* servc %1,%2 */
" ipm %0\n"
@@ -91,61 +108,94 @@
: "=&d" (cc)
: "d" (command), "a" (__pa(sccb))
: "cc", "memory" );
- /*
- * cc == 0: Service Call succesful initiated
- * cc == 2: SCLP busy, new Service Call not initiated,
- * new SCCB unchanged
- * cc == 3: SCLP function not operational
- */
- if (cc == SCLP_NOT_OPERATIONAL)
+ if (cc == 3)
return -EIO;
- if (cc == SCLP_BUSY)
+ if (cc == 2)
return -EBUSY;
return 0;
}

+/* Request timeout handler. Restart the request queue. If DATA is non-zero,
+ * force restart of running request. */
static void
-sclp_start_request(void)
+sclp_request_timeout(unsigned long data)
+{
+ unsigned long flags;
+
+ if (data) {
+ spin_lock_irqsave(&sclp_lock, flags);
+ sclp_running_state = sclp_running_state_idle;
+ spin_unlock_irqrestore(&sclp_lock, flags);
+ }
+ sclp_process_queue();
+}
+
+/* Set up request retry timer. Called while sclp_lock is locked. */
+static inline void
+__sclp_set_request_timer(unsigned long time, void (*function)(unsigned long),
+ unsigned long data)
+{
+ del_timer(&sclp_request_timer);
+ sclp_request_timer.function = function;
+ sclp_request_timer.data = data;
+ sclp_request_timer.expires = jiffies + time;
+ add_timer(&sclp_request_timer);
+}
+
+/* Try to start a request. Return zero if the request was successfully
+ * started or if it will be started at a later time. Return non-zero otherwise.
+ * Called while sclp_lock is locked. */
+static int
+__sclp_start_request(struct sclp_req *req)
+{
+ int rc;
+
+ if (sclp_running_state != sclp_running_state_idle)
+ return 0;
+ del_timer(&sclp_request_timer);
+ if (req->start_count <= SCLP_REQUEST_RETRY) {
+ rc = service_call(req->command, req->sccb);
+ req->start_count++;
+ } else
+ rc = -EIO;
+ if (rc == 0) {
+ /* Sucessfully started request */
+ req->status = SCLP_REQ_RUNNING;
+ sclp_running_state = sclp_running_state_running;
+ __sclp_set_request_timer(SCLP_RETRY_INTERVAL * HZ,
+ sclp_request_timeout, 1);
+ return 0;
+ } else if (rc == -EBUSY) {
+ /* Try again later */
+ __sclp_set_request_timer(SCLP_BUSY_INTERVAL * HZ,
+ sclp_request_timeout, 0);
+ return 0;
+ }
+ /* Request failed */
+ req->status = SCLP_REQ_FAILED;
+ return rc;
+}
+
+/* Try to start queued requests. */
+static void
+sclp_process_queue(void)
{
struct sclp_req *req;
int rc;
unsigned long flags;

spin_lock_irqsave(&sclp_lock, flags);
- /* quick exit if sclp is already in use */
- if (test_bit(SCLP_RUNNING, &sclp_status)) {
+ if (sclp_running_state != sclp_running_state_idle) {
spin_unlock_irqrestore(&sclp_lock, flags);
return;
}
- /* Try to start requests from the request queue. */
+ del_timer(&sclp_request_timer);
while (!list_empty(&sclp_req_queue)) {
req = list_entry(sclp_req_queue.next, struct sclp_req, list);
- rc = __service_call(req->command, req->sccb);
- if (rc == 0) {
- /* Sucessfully started request. */
- req->status = SCLP_REQ_RUNNING;
- /* Request active. Set running indication. */
- set_bit(SCLP_RUNNING, &sclp_status);
+ rc = __sclp_start_request(req);
+ if (rc == 0)
break;
- }
- if (rc == -EBUSY) {
- /**
- * SCLP is busy but no request is running.
- * Try again later.
- */
- if (!timer_pending(&sclp_busy_timer) ||
- !mod_timer(&sclp_busy_timer,
- jiffies + SCLP_BUSY_POLL_INTERVAL*HZ)) {
- sclp_busy_timer.function =
- (void *) sclp_start_request;
- sclp_busy_timer.expires =
- jiffies + SCLP_BUSY_POLL_INTERVAL*HZ;
- add_timer(&sclp_busy_timer);
- }
- break;
- }
/* Request failed. */
- req->status = SCLP_REQ_FAILED;
list_del(&req->list);
if (req->callback) {
spin_unlock_irqrestore(&sclp_lock, flags);
@@ -156,265 +206,258 @@
spin_unlock_irqrestore(&sclp_lock, flags);
}

+/* Queue a new request. Return zero on success, non-zero otherwise. */
+int
+sclp_add_request(struct sclp_req *req)
+{
+ unsigned long flags;
+ int rc;
+
+ spin_lock_irqsave(&sclp_lock, flags);
+ if ((sclp_init_state != sclp_init_state_initialized ||
+ sclp_activation_state != sclp_activation_state_active) &&
+ req != &sclp_init_req) {
+ spin_unlock_irqrestore(&sclp_lock, flags);
+ return -EIO;
+ }
+ req->status = SCLP_REQ_QUEUED;
+ req->start_count = 0;
+ list_add_tail(&req->list, &sclp_req_queue);
+ rc = 0;
+ /* Start if request is first in list */
+ if (req->list.prev == &sclp_req_queue) {
+ rc = __sclp_start_request(req);
+ if (rc)
+ list_del(&req->list);
+ }
+ spin_unlock_irqrestore(&sclp_lock, flags);
+ return rc;
+}
+
+EXPORT_SYMBOL(sclp_add_request);
+
+/* Dispatch events found in request buffer to registered listeners. Return 0
+ * if all events were dispatched, non-zero otherwise. */
static int
-sclp_process_evbufs(struct sccb_header *sccb)
+sclp_dispatch_evbufs(struct sccb_header *sccb)
{
- int result;
unsigned long flags;
struct evbuf_header *evbuf;
struct list_head *l;
- struct sclp_register *t;
+ struct sclp_register *reg;
+ int offset;
+ int rc;

spin_lock_irqsave(&sclp_lock, flags);
- evbuf = (struct evbuf_header *) (sccb + 1);
- result = 0;
- while ((addr_t) evbuf < (addr_t) sccb + sccb->length) {
- /* check registered event */
- t = NULL;
+ rc = 0;
+ for (offset = sizeof(struct sccb_header); offset < sccb->length;
+ offset += evbuf->length) {
+ /* Search for event handler */
+ evbuf = (struct evbuf_header *) ((addr_t) sccb + offset);
+ reg = NULL;
list_for_each(l, &sclp_reg_list) {
- t = list_entry(l, struct sclp_register, list);
- if (t->receive_mask & (1 << (32 - evbuf->type))) {
- if (t->receiver_fn != NULL) {
- spin_unlock_irqrestore(&sclp_lock,
- flags);
- t->receiver_fn(evbuf);
- spin_lock_irqsave(&sclp_lock, flags);
- }
+ reg = list_entry(l, struct sclp_register, list);
+ if (reg->receive_mask & (1 << (32 - evbuf->type)))
break;
- }
else
- t = NULL;
+ reg = NULL;
}
- /* Check for unrequested event buffer */
- if (t == NULL)
- result = -ENOSYS;
- evbuf = (struct evbuf_header *)
- ((addr_t) evbuf + evbuf->length);
+ if (reg && reg->receiver_fn) {
+ spin_unlock_irqrestore(&sclp_lock, flags);
+ reg->receiver_fn(evbuf);
+ spin_lock_irqsave(&sclp_lock, flags);
+ } else if (reg == NULL)
+ rc = -ENOSYS;
}
spin_unlock_irqrestore(&sclp_lock, flags);
- return result;
-}
-
-char *
-sclp_error_message(u16 rc)
-{
- static struct {
- u16 code; char *msg;
- } sclp_errors[] = {
- { 0x0000, "No response code stored (machine malfunction)" },
- { 0x0020, "Normal Completion" },
- { 0x0040, "SCLP equipment check" },
- { 0x0100, "SCCB boundary violation" },
- { 0x01f0, "Invalid command" },
- { 0x0220, "Normal Completion; suppressed buffers pending" },
- { 0x0300, "Insufficient SCCB length" },
- { 0x0340, "Contained SCLP equipment check" },
- { 0x05f0, "Target resource in improper state" },
- { 0x40f0, "Invalid function code/not installed" },
- { 0x60f0, "No buffers stored" },
- { 0x62f0, "No buffers stored; suppressed buffers pending" },
- { 0x70f0, "Invalid selection mask" },
- { 0x71f0, "Event buffer exceeds available space" },
- { 0x72f0, "Inconsistent lengths" },
- { 0x73f0, "Event buffer syntax error" }
- };
- int i;
- for (i = 0; i < sizeof(sclp_errors)/sizeof(sclp_errors[0]); i++)
- if (rc == sclp_errors[i].code)
- return sclp_errors[i].msg;
- return "Invalid response code";
+ return rc;
}

-/*
- * postprocessing of unconditional read service call
- */
+/* Read event data request callback. */
static void
-sclp_unconditional_read_cb(struct sclp_req *read_req, void *data)
+sclp_read_cb(struct sclp_req *req, void *data)
{
+ unsigned long flags;
struct sccb_header *sccb;

- sccb = read_req->sccb;
- if (sccb->response_code == 0x0020 ||
- sccb->response_code == 0x0220) {
- if (sclp_process_evbufs(sccb) != 0)
- printk(KERN_WARNING SCLP_CORE_PRINT_HEADER
- "unconditional read: "
- "unrequested event buffer received.\n");
- }
-
- if (sccb->response_code != 0x0020)
- printk(KERN_WARNING SCLP_CORE_PRINT_HEADER
- "unconditional read: %s (response code=0x%x).\n",
- sclp_error_message(sccb->response_code),
- sccb->response_code);
-
- clear_bit(SCLP_READING, &sclp_status);
+ sccb = (struct sccb_header *) req->sccb;
+ if (req->status == SCLP_REQ_DONE && (sccb->response_code == 0x20 ||
+ sccb->response_code == 0x220))
+ sclp_dispatch_evbufs(sccb);
+ spin_lock_irqsave(&sclp_lock, flags);
+ sclp_reading_state = sclp_reading_state_idle;
+ spin_unlock_irqrestore(&sclp_lock, flags);
}

-/*
- * Function to queue Read Event Data/Unconditional Read
- */
-static void
-__sclp_unconditional_read(void)
+/* Prepare read event data request. Called while sclp_lock is locked. */
+static inline void
+__sclp_make_read_req(void)
{
struct sccb_header *sccb;
- struct sclp_req *read_req;

- /*
- * Don't try to initiate Unconditional Read if we are not able to
- * receive anything
- */
- if (sclp_receive_mask == 0)
- return;
- /* Don't try reading if a read is already outstanding */
- if (test_and_set_bit(SCLP_READING, &sclp_status))
- return;
- /* Initialize read sccb */
sccb = (struct sccb_header *) sclp_read_sccb;
clear_page(sccb);
+ memset(&sclp_read_req, 0, sizeof(struct sclp_req));
+ sclp_read_req.command = SCLP_CMDW_READDATA;
+ sclp_read_req.status = SCLP_REQ_QUEUED;
+ sclp_read_req.start_count = 0;
+ sclp_read_req.callback = sclp_read_cb;
+ sclp_read_req.sccb = sccb;
sccb->length = PAGE_SIZE;
- sccb->function_code = 0; /* unconditional read */
- sccb->control_mask[2] = 0x80; /* variable length response */
- /* Initialize request structure */
- read_req = &sclp_read_req;
- read_req->command = SCLP_CMDW_READDATA;
- read_req->status = SCLP_REQ_QUEUED;
- read_req->callback = sclp_unconditional_read_cb;
- read_req->sccb = sccb;
- /* Add read request to the head of queue */
- list_add(&read_req->list, &sclp_req_queue);
-}
-
-/* Bit masks to interpret external interruption parameter contents. */
-#define EXT_INT_SCCB_MASK 0xfffffff8
-#define EXT_INT_STATECHANGE_PENDING 0x00000002
-#define EXT_INT_EVBUF_PENDING 0x00000001
+ sccb->function_code = 0;
+ sccb->control_mask[2] = 0x80;
+}

-/*
- * Handler for service-signal external interruptions
- */
+/* Search request list for request with matching sccb. Return request if found,
+ * NULL otherwise. Called while sclp_lock is locked. */
+static inline struct sclp_req *
+__sclp_find_req(u32 sccb)
+{
+ struct list_head *l;
+ struct sclp_req *req;
+
+ list_for_each(l, &sclp_req_queue) {
+ req = list_entry(l, struct sclp_req, list);
+ if (sccb == (u32) (addr_t) req->sccb)
+ return req;
+ }
+ return NULL;
+}
+
+/* Handler for external interruption. Perform request post-processing.
+ * Prepare read event data request if necessary. Start processing of next
+ * request on queue. */
static void
sclp_interrupt_handler(struct pt_regs *regs, __u16 code)
{
- u32 ext_int_param, finished_sccb, evbuf_pending;
- struct list_head *l;
- struct sclp_req *req, *tmp;
+ struct sclp_req *req;
+ u32 finished_sccb;
+ u32 evbuf_pending;

spin_lock(&sclp_lock);
- /*
- * Only process interrupt if sclp is initialized.
- * This avoids strange effects for a pending request
- * from before the last re-ipl.
- */
- if (!test_bit(SCLP_INIT, &sclp_status)) {
- /* Now clear the running bit */
- clear_bit(SCLP_RUNNING, &sclp_status);
- spin_unlock(&sclp_lock);
- return;
- }
- ext_int_param = S390_lowcore.ext_params;
- finished_sccb = ext_int_param & EXT_INT_SCCB_MASK;
- evbuf_pending = ext_int_param & (EXT_INT_EVBUF_PENDING |
- EXT_INT_STATECHANGE_PENDING);
- req = NULL;
- if (finished_sccb != 0U) {
- list_for_each(l, &sclp_req_queue) {
- tmp = list_entry(l, struct sclp_req, list);
- if (finished_sccb == (u32)(addr_t) tmp->sccb) {
- list_del(&tmp->list);
- req = tmp;
- break;
+ finished_sccb = S390_lowcore.ext_params & 0xfffffff8;
+ evbuf_pending = S390_lowcore.ext_params & 0x3;
+ if (finished_sccb) {
+ req = __sclp_find_req(finished_sccb);
+ if (req) {
+ /* Request post-processing */
+ list_del(&req->list);
+ req->status = SCLP_REQ_DONE;
+ if (req->callback) {
+ spin_unlock(&sclp_lock);
+ req->callback(req, req->callback_data);
+ spin_lock(&sclp_lock);
}
}
+ sclp_running_state = sclp_running_state_idle;
}
- spin_unlock(&sclp_lock);
- /* Perform callback */
- if (req != NULL) {
- req->status = SCLP_REQ_DONE;
- if (req->callback != NULL)
- req->callback(req, req->callback_data);
+ if (evbuf_pending && sclp_receive_mask != 0 &&
+ sclp_reading_state == sclp_reading_state_idle &&
+ sclp_activation_state == sclp_activation_state_active ) {
+ sclp_reading_state = sclp_reading_state_reading;
+ __sclp_make_read_req();
+ /* Add request to head of queue */
+ list_add(&sclp_read_req.list, &sclp_req_queue);
}
- spin_lock(&sclp_lock);
- /* Head queue a read sccb if an event buffer is pending */
- if (evbuf_pending)
- __sclp_unconditional_read();
- /* Now clear the running bit if SCLP indicated a finished SCCB */
- if (finished_sccb != 0U)
- clear_bit(SCLP_RUNNING, &sclp_status);
spin_unlock(&sclp_lock);
- /* and start next request on the queue */
- sclp_start_request();
+ sclp_process_queue();
}

-/*
- * Wait synchronously for external interrupt of sclp. We may not receive
- * any other external interrupt, so we disable all other external interrupts
- * in control register 0.
- */
+/* Return current Time-Of-Day clock. */
+static inline u64
+sclp_get_clock(void)
+{
+ u64 result;
+
+ asm volatile ("STCK 0(%1)" : "=m" (result) : "a" (&(result)) : "cc");
+ return result;
+}
+
+/* Convert interval in jiffies to TOD ticks. */
+static inline u64
+sclp_tod_from_jiffies(unsigned long jiffies)
+{
+ return (u64) (jiffies / HZ) << 32;
+}
+
+/* Wait until a currently running request finished. Note: while this function
+ * is running, no timers are served on the calling CPU. */
void
sclp_sync_wait(void)
{
unsigned long psw_mask;
unsigned long cr0, cr0_sync;
+ u64 timeout;

- /* Prevent BH from executing. */
+ /* We'll be disabling timer interrupts, so we need a custom timeout
+ * mechanism */
+ timeout = 0;
+ if (timer_pending(&sclp_request_timer)) {
+ /* Get timeout TOD value */
+ timeout = sclp_get_clock() +
+ sclp_tod_from_jiffies(sclp_request_timer.expires -
+ jiffies);
+ }
+ /* Prevent bottom half from executing once we force interrupts open */
local_bh_disable();
- /*
- * save cr0
- * enable service signal external interruption (cr0.22)
- * disable cr0.20-21, cr0.25, cr0.27, cr0.30-31
- * don't touch any other bit in cr0
- */
+ /* Enable service-signal interruption, disable timer interrupts */
__ctl_store(cr0, 0, 0);
cr0_sync = cr0;
cr0_sync |= 0x00000200;
cr0_sync &= 0xFFFFF3AC;
__ctl_load(cr0_sync, 0, 0);
-
- /* enable external interruptions (PSW-mask.7) */
asm volatile ("STOSM 0(%1),0x01"
: "=m" (psw_mask) : "a" (&psw_mask) : "memory");
-
- /* wait until ISR signals receipt of interrupt */
- while (test_bit(SCLP_RUNNING, &sclp_status)) {
+ /* Loop until driver state indicates finished request */
+ while (sclp_running_state != sclp_running_state_idle) {
+ /* Check for expired request timer */
+ if (timer_pending(&sclp_request_timer) &&
+ sclp_get_clock() > timeout &&
+ del_timer(&sclp_request_timer))
+ sclp_request_timer.function(sclp_request_timer.data);
barrier();
cpu_relax();
}
-
- /* disable external interruptions */
+ /* Restore interrupt settings */
asm volatile ("SSM 0(%0)"
: : "a" (&psw_mask) : "memory");
-
- /* restore cr0 */
__ctl_load(cr0, 0, 0);
__local_bh_enable();
}

-/*
- * Queue an SCLP request. Request will immediately be processed if queue is
- * empty.
- */
-void
-sclp_add_request(struct sclp_req *req)
+EXPORT_SYMBOL(sclp_sync_wait);
+
+/* Dispatch changes in send and receive mask to registered listeners. */
+static inline void
+sclp_dispatch_state_change(void)
{
+ struct list_head *l;
+ struct sclp_register *reg;
unsigned long flags;
+ sccb_mask_t receive_mask;
+ sccb_mask_t send_mask;

- if (!test_bit(SCLP_INIT, &sclp_status)) {
- req->status = SCLP_REQ_FAILED;
- if (req->callback != NULL)
- req->callback(req, req->callback_data);
- return;
- }
- spin_lock_irqsave(&sclp_lock, flags);
- /* queue the request */
- req->status = SCLP_REQ_QUEUED;
- list_add_tail(&req->list, &sclp_req_queue);
- spin_unlock_irqrestore(&sclp_lock, flags);
- /* try to start the first request on the queue */
- sclp_start_request();
+ do {
+ spin_lock_irqsave(&sclp_lock, flags);
+ reg = NULL;
+ list_for_each(l, &sclp_reg_list) {
+ reg = list_entry(l, struct sclp_register, list);
+ receive_mask = reg->receive_mask & sclp_receive_mask;
+ send_mask = reg->send_mask & sclp_send_mask;
+ if (reg->sclp_receive_mask != receive_mask ||
+ reg->sclp_send_mask != send_mask) {
+ reg->sclp_receive_mask = receive_mask;
+ reg->sclp_send_mask = send_mask;
+ break;
+ } else
+ reg = NULL;
+ }
+ spin_unlock_irqrestore(&sclp_lock, flags);
+ if (reg && reg->state_change_fn)
+ reg->state_change_fn(reg);
+ } while (reg);
}

-/* state change notification */
struct sclp_statechangebuf {
struct evbuf_header header;
u8 validity_sclp_active_facility_mask : 1;
@@ -429,148 +472,126 @@
u32 read_data_function_mask;
} __attribute__((packed));

-static inline void
-__sclp_notify_state_change(void)
-{
- struct list_head *l;
- struct sclp_register *t;
- sccb_mask_t receive_mask, send_mask;
-
- list_for_each(l, &sclp_reg_list) {
- t = list_entry(l, struct sclp_register, list);
- receive_mask = t->receive_mask & sclp_receive_mask;
- send_mask = t->send_mask & sclp_send_mask;
- if (t->sclp_receive_mask != receive_mask ||
- t->sclp_send_mask != send_mask) {
- t->sclp_receive_mask = receive_mask;
- t->sclp_send_mask = send_mask;
- if (t->state_change_fn != NULL)
- t->state_change_fn(t);
- }
- }
-}

+/* State change event callback. Inform listeners of changes. */
static void
-sclp_state_change(struct evbuf_header *evbuf)
+sclp_state_change_cb(struct evbuf_header *evbuf)
{
unsigned long flags;
struct sclp_statechangebuf *scbuf;

- spin_lock_irqsave(&sclp_lock, flags);
scbuf = (struct sclp_statechangebuf *) evbuf;
-
- if (scbuf->validity_sclp_receive_mask) {
- if (scbuf->mask_length != sizeof(sccb_mask_t))
- printk(KERN_WARNING SCLP_CORE_PRINT_HEADER
- "state change event with mask length %i\n",
- scbuf->mask_length);
- else
- /* set new receive mask */
- sclp_receive_mask = scbuf->sclp_receive_mask;
- }
-
- if (scbuf->validity_sclp_send_mask) {
- if (scbuf->mask_length != sizeof(sccb_mask_t))
- printk(KERN_WARNING SCLP_CORE_PRINT_HEADER
- "state change event with mask length %i\n",
- scbuf->mask_length);
- else
- /* set new send mask */
- sclp_send_mask = scbuf->sclp_send_mask;
- }
-
- __sclp_notify_state_change();
+ if (scbuf->mask_length != sizeof(sccb_mask_t))
+ return;
+ spin_lock_irqsave(&sclp_lock, flags);
+ if (scbuf->validity_sclp_receive_mask)
+ sclp_receive_mask = scbuf->sclp_receive_mask;
+ if (scbuf->validity_sclp_send_mask)
+ sclp_send_mask = scbuf->sclp_send_mask;
spin_unlock_irqrestore(&sclp_lock, flags);
+ sclp_dispatch_state_change();
}

static struct sclp_register sclp_state_change_event = {
.receive_mask = EvTyp_StateChange_Mask,
- .receiver_fn = sclp_state_change
+ .receiver_fn = sclp_state_change_cb
};

-
-/*
- * SCLP quiesce event handler
- */
-#ifdef CONFIG_SMP
-static void
-do_load_quiesce_psw(void * __unused)
+/* Calculate receive and send mask of currently registered listeners.
+ * Called while sclp_lock is locked. */
+static inline void
+__sclp_get_mask(sccb_mask_t *receive_mask, sccb_mask_t *send_mask)
{
- static atomic_t cpuid = ATOMIC_INIT(-1);
- psw_t quiesce_psw;
- __u32 status;
- int i;
-
- if (atomic_compare_and_swap(-1, smp_processor_id(), &cpuid))
- signal_processor(smp_processor_id(), sigp_stop);
- /* Wait for all other cpus to enter stopped state */
- i = 1;
- while (i < NR_CPUS) {
- if (!cpu_online(i)) {
- i++;
- continue;
- }
- switch (signal_processor_ps(&status, 0, i, sigp_sense)) {
- case sigp_order_code_accepted:
- case sigp_status_stored:
- /* Check for stopped and check stop state */
- if (status & 0x50)
- i++;
- break;
- case sigp_busy:
- break;
- case sigp_not_operational:
- i++;
- break;
- }
+ struct list_head *l;
+ struct sclp_register *t;
+
+ *receive_mask = 0;
+ *send_mask = 0;
+ list_for_each(l, &sclp_reg_list) {
+ t = list_entry(l, struct sclp_register, list);
+ *receive_mask |= t->receive_mask;
+ *send_mask |= t->send_mask;
}
- /* Quiesce the last cpu with the special psw */
- quiesce_psw.mask = PSW_BASE_BITS | PSW_MASK_WAIT;
- quiesce_psw.addr = 0xfff;
- __load_psw(quiesce_psw);
}

-static void
-do_machine_quiesce(void)
+/* Register event listener. Return 0 on success, non-zero otherwise. */
+int
+sclp_register(struct sclp_register *reg)
{
- on_each_cpu(do_load_quiesce_psw, NULL, 0, 0);
+ unsigned long flags;
+ sccb_mask_t receive_mask;
+ sccb_mask_t send_mask;
+ int rc;
+
+ rc = sclp_init();
+ if (rc)
+ return rc;
+ spin_lock_irqsave(&sclp_lock, flags);
+ /* Check event mask for collisions */
+ __sclp_get_mask(&receive_mask, &send_mask);
+ if (reg->receive_mask & receive_mask || reg->send_mask & send_mask) {
+ spin_unlock_irqrestore(&sclp_lock, flags);
+ return -EBUSY;
+ }
+ /* Trigger initial state change callback */
+ reg->sclp_receive_mask = 0;
+ reg->sclp_send_mask = 0;
+ list_add(&reg->list, &sclp_reg_list);
+ spin_unlock_irqrestore(&sclp_lock, flags);
+ rc = sclp_init_mask(1);
+ if (rc) {
+ spin_lock_irqsave(&sclp_lock, flags);
+ list_del(&reg->list);
+ spin_unlock_irqrestore(&sclp_lock, flags);
+ }
+ return rc;
}
-#else
-static void
-do_machine_quiesce(void)
+
+EXPORT_SYMBOL(sclp_register);
+
+/* Unregister event listener. */
+void
+sclp_unregister(struct sclp_register *reg)
{
- psw_t quiesce_psw;
+ unsigned long flags;

- quiesce_psw.mask = PSW_BASE_BITS | PSW_MASK_WAIT;
- quiesce_psw.addr = 0xfff;
- __load_psw(quiesce_psw);
+ spin_lock_irqsave(&sclp_lock, flags);
+ list_del(&reg->list);
+ spin_unlock_irqrestore(&sclp_lock, flags);
+ sclp_init_mask(1);
}
-#endif

-extern void ctrl_alt_del(void);
+EXPORT_SYMBOL(sclp_unregister);

-static void
-sclp_quiesce(struct evbuf_header *evbuf)
+/* Remove event buffers which are marked processed. Return the number of
+ * remaining event buffers. */
+int
+sclp_remove_processed(struct sccb_header *sccb)
{
- /*
- * We got a "shutdown" request.
- * Add a call to an appropriate "shutdown" routine here. This
- * routine should set all PSWs to 'disabled-wait', 'stopped'
- * or 'check-stopped' - except 1 PSW which needs to carry a
- * special bit pattern called 'quiesce PSW'.
- */
- _machine_restart = (void *) do_machine_quiesce;
- _machine_halt = do_machine_quiesce;
- _machine_power_off = do_machine_quiesce;
- ctrl_alt_del();
-}
-
-static struct sclp_register sclp_quiesce_event = {
- .receive_mask = EvTyp_SigQuiesce_Mask,
- .receiver_fn = sclp_quiesce
-};
+ struct evbuf_header *evbuf;
+ int unprocessed;
+ u16 remaining;
+
+ evbuf = (struct evbuf_header *) (sccb + 1);
+ unprocessed = 0;
+ remaining = sccb->length - sizeof(struct sccb_header);
+ while (remaining > 0) {
+ remaining -= evbuf->length;
+ if (evbuf->flags & 0x80) {
+ sccb->length -= evbuf->length;
+ memcpy(evbuf, (void *) ((addr_t) evbuf + evbuf->length),
+ remaining);
+ } else {
+ unprocessed++;
+ evbuf = (struct evbuf_header *)
+ ((addr_t) evbuf + evbuf->length);
+ }
+ }
+ return unprocessed;
+}
+
+EXPORT_SYMBOL(sclp_remove_processed);

-/* initialisation of SCLP */
struct init_sccb {
struct sccb_header header;
u16 _reserved;
@@ -581,273 +602,314 @@
sccb_mask_t sclp_receive_mask;
} __attribute__((packed));

-static void sclp_init_mask_retry(unsigned long);
-
-static int
-sclp_init_mask(void)
+/* Prepare init mask request. Called while sclp_lock is locked. */
+static inline void
+__sclp_make_init_req(u32 receive_mask, u32 send_mask)
{
- unsigned long flags;
struct init_sccb *sccb;
- struct sclp_req *req;
- struct list_head *l;
- struct sclp_register *t;
- int rc;

sccb = (struct init_sccb *) sclp_init_sccb;
- /* stick the request structure to the end of the init sccb page */
- req = (struct sclp_req *) ((addr_t) sccb + PAGE_SIZE) - 1;
-
- /* SCLP setup concerning receiving and sending Event Buffers */
- req->command = SCLP_CMDW_WRITEMASK;
- req->status = SCLP_REQ_QUEUED;
- req->callback = NULL;
- req->sccb = sccb;
- /* setup sccb for writemask command */
- memset(sccb, 0, sizeof(struct init_sccb));
+ clear_page(sccb);
+ memset(&sclp_init_req, 0, sizeof(struct sclp_req));
+ sclp_init_req.command = SCLP_CMDW_WRITEMASK;
+ sclp_init_req.status = SCLP_REQ_FILLED;
+ sclp_init_req.start_count = 0;
+ sclp_init_req.callback = NULL;
+ sclp_init_req.callback_data = NULL;
+ sclp_init_req.sccb = sccb;
sccb->header.length = sizeof(struct init_sccb);
sccb->mask_length = sizeof(sccb_mask_t);
- /* copy in the sccb mask of the registered event types */
- spin_lock_irqsave(&sclp_lock, flags);
- if (!test_bit(SCLP_SHUTDOWN, &sclp_status)) {
- list_for_each(l, &sclp_reg_list) {
- t = list_entry(l, struct sclp_register, list);
- sccb->receive_mask |= t->receive_mask;
- sccb->send_mask |= t->send_mask;
- }
- }
+ sccb->receive_mask = receive_mask;
+ sccb->send_mask = send_mask;
sccb->sclp_receive_mask = 0;
sccb->sclp_send_mask = 0;
- if (test_bit(SCLP_INIT, &sclp_status)) {
- /* add request to sclp queue */
- list_add_tail(&req->list, &sclp_req_queue);
- spin_unlock_irqrestore(&sclp_lock, flags);
- /* and start if SCLP is idle */
- sclp_start_request();
- /* now wait for completion */
- while (req->status != SCLP_REQ_DONE &&
- req->status != SCLP_REQ_FAILED)
+}
+
+/* Start init mask request. If calculate is non-zero, calculate the mask as
+ * requested by registered listeners. Use zero mask otherwise. Return 0 on
+ * success, non-zero otherwise. */
+static int
+sclp_init_mask(int calculate)
+{
+ unsigned long flags;
+ struct init_sccb *sccb = (struct init_sccb *) sclp_init_sccb;
+ sccb_mask_t receive_mask;
+ sccb_mask_t send_mask;
+ int retry;
+ int rc;
+ unsigned long wait;
+
+ spin_lock_irqsave(&sclp_lock, flags);
+ /* Check if interface is in appropriate state */
+ if (sclp_mask_state != sclp_mask_state_idle) {
+ spin_unlock_irqrestore(&sclp_lock, flags);
+ return -EBUSY;
+ }
+ if (sclp_activation_state == sclp_activation_state_inactive) {
+ spin_unlock_irqrestore(&sclp_lock, flags);
+ return -EINVAL;
+ }
+ sclp_mask_state = sclp_mask_state_initializing;
+ /* Determine mask */
+ if (calculate)
+ __sclp_get_mask(&receive_mask, &send_mask);
+ else {
+ receive_mask = 0;
+ send_mask = 0;
+ }
+ rc = -EIO;
+ for (retry = 0; retry <= SCLP_MASK_RETRY; retry++) {
+ /* Prepare request */
+ __sclp_make_init_req(receive_mask, send_mask);
+ spin_unlock_irqrestore(&sclp_lock, flags);
+ if (sclp_add_request(&sclp_init_req)) {
+ /* Try again later */
+ wait = jiffies + SCLP_BUSY_INTERVAL * HZ;
+ while (time_before(jiffies, wait))
+ sclp_sync_wait();
+ spin_lock_irqsave(&sclp_lock, flags);
+ continue;
+ }
+ while (sclp_init_req.status != SCLP_REQ_DONE &&
+ sclp_init_req.status != SCLP_REQ_FAILED)
sclp_sync_wait();
spin_lock_irqsave(&sclp_lock, flags);
- } else {
- /*
- * Special case for the very first write mask command.
- * The interrupt handler is not removing request from
- * the request queue and doesn't call callbacks yet
- * because there might be an pending old interrupt
- * after a Re-IPL. We have to receive and ignore it.
- */
- do {
- rc = __service_call(req->command, req->sccb);
- if (rc == 0)
- set_bit(SCLP_RUNNING, &sclp_status);
+ if (sclp_init_req.status == SCLP_REQ_DONE &&
+ sccb->header.response_code == 0x20) {
+ /* Successful request */
+ if (calculate) {
+ sclp_receive_mask = sccb->sclp_receive_mask;
+ sclp_send_mask = sccb->sclp_send_mask;
+ } else {
+ sclp_receive_mask = 0;
+ sclp_send_mask = 0;
+ }
spin_unlock_irqrestore(&sclp_lock, flags);
- if (rc == -EIO)
- return -ENOSYS;
- sclp_sync_wait();
+ sclp_dispatch_state_change();
spin_lock_irqsave(&sclp_lock, flags);
- } while (rc == -EBUSY);
- }
- if (sccb->header.response_code != 0x0020) {
- /* WRITEMASK failed - we cannot rely on receiving a state
- change event, so initially, polling is the only alternative
- for us to ever become operational. */
- if (!test_bit(SCLP_SHUTDOWN, &sclp_status) &&
- (!timer_pending(&retry_timer) ||
- !mod_timer(&retry_timer,
- jiffies + SCLP_INIT_POLL_INTERVAL*HZ))) {
- retry_timer.function = sclp_init_mask_retry;
- retry_timer.data = 0;
- retry_timer.expires = jiffies +
- SCLP_INIT_POLL_INTERVAL*HZ;
- add_timer(&retry_timer);
+ rc = 0;
+ break;
}
- } else {
- sclp_receive_mask = sccb->sclp_receive_mask;
- sclp_send_mask = sccb->sclp_send_mask;
- __sclp_notify_state_change();
}
+ sclp_mask_state = sclp_mask_state_idle;
spin_unlock_irqrestore(&sclp_lock, flags);
- return 0;
-}
-
-static void
-sclp_init_mask_retry(unsigned long data)
-{
- sclp_init_mask();
+ return rc;
}

-/* Reboot event handler - reset send and receive mask to prevent pending SCLP
- * events from interfering with rebooted system. */
-static int
-sclp_reboot_event(struct notifier_block *this, unsigned long event, void *ptr)
+/* Deactivate SCLP interface. On success, new requests will be rejected,
+ * events will no longer be dispatched. Return 0 on success, non-zero
+ * otherwise. */
+int
+sclp_deactivate(void)
{
unsigned long flags;
+ int rc;

- /* Note: need spinlock to maintain atomicity when accessing global
- * variables. */
spin_lock_irqsave(&sclp_lock, flags);
- set_bit(SCLP_SHUTDOWN, &sclp_status);
+ /* Deactivate can only be called when active */
+ if (sclp_activation_state != sclp_activation_state_active) {
+ spin_unlock_irqrestore(&sclp_lock, flags);
+ return -EINVAL;
+ }
+ sclp_activation_state = sclp_activation_state_deactivating;
spin_unlock_irqrestore(&sclp_lock, flags);
- sclp_init_mask();
- return NOTIFY_DONE;
+ rc = sclp_init_mask(0);
+ spin_lock_irqsave(&sclp_lock, flags);
+ if (rc == 0)
+ sclp_activation_state = sclp_activation_state_inactive;
+ else
+ sclp_activation_state = sclp_activation_state_active;
+ spin_unlock_irqrestore(&sclp_lock, flags);
+ return rc;
}

-static struct notifier_block sclp_reboot_notifier = {
- .notifier_call = sclp_reboot_event
-};
+EXPORT_SYMBOL(sclp_deactivate);

-/*
- * sclp setup function. Called early (no kmalloc!) from sclp_console_init().
- */
-static int
-sclp_init(void)
+/* Reactivate SCLP interface after sclp_deactivate. On success, new
+ * requests will be accepted, events will be dispatched again. Return 0 on
+ * success, non-zero otherwise. */
+int
+sclp_reactivate(void)
{
+ unsigned long flags;
int rc;

- if (test_bit(SCLP_INIT, &sclp_status))
- /* Already initialized. */
- return 0;
-
- spin_lock_init(&sclp_lock);
- INIT_LIST_HEAD(&sclp_req_queue);
-
- /* init event list */
- INIT_LIST_HEAD(&sclp_reg_list);
- list_add(&sclp_state_change_event.list, &sclp_reg_list);
- list_add(&sclp_quiesce_event.list, &sclp_reg_list);
-
- rc = register_reboot_notifier(&sclp_reboot_notifier);
- if (rc)
- return rc;
+ spin_lock_irqsave(&sclp_lock, flags);
+ /* Reactivate can only be called when inactive */
+ if (sclp_activation_state != sclp_activation_state_inactive) {
+ spin_unlock_irqrestore(&sclp_lock, flags);
+ return -EINVAL;
+ }
+ sclp_activation_state = sclp_activation_state_activating;
+ spin_unlock_irqrestore(&sclp_lock, flags);
+ rc = sclp_init_mask(1);
+ spin_lock_irqsave(&sclp_lock, flags);
+ if (rc == 0)
+ sclp_activation_state = sclp_activation_state_active;
+ else
+ sclp_activation_state = sclp_activation_state_inactive;
+ spin_unlock_irqrestore(&sclp_lock, flags);
+ return rc;
+}

- /*
- * request the 0x2401 external interrupt
- * The sclp driver is initialized early (before kmalloc works). We
- * need to use register_early_external_interrupt.
- */
- if (register_early_external_interrupt(0x2401, sclp_interrupt_handler,
- &ext_int_info_hwc) != 0)
- return -EBUSY;
+EXPORT_SYMBOL(sclp_reactivate);

- /* enable service-signal external interruptions,
- * Control Register 0 bit 22 := 1
- * (besides PSW bit 7 must be set to 1 sometimes for external
- * interruptions)
- */
- ctl_set_bit(0, 9);
+/* Handler for external interruption used during initialization. Modify
+ * request state to done. */
+static void
+sclp_check_handler(struct pt_regs *regs, __u16 code)
+{
+ u32 finished_sccb;

- init_timer(&retry_timer);
- init_timer(&sclp_busy_timer);
- /* do the initial write event mask */
- rc = sclp_init_mask();
- if (rc == 0) {
- /* Ok, now everything is setup right. */
- set_bit(SCLP_INIT, &sclp_status);
- return 0;
+ finished_sccb = S390_lowcore.ext_params & 0xfffffff8;
+ /* Is this the interrupt we are waiting for? */
+ if (finished_sccb == 0)
+ return;
+ if (finished_sccb != (u32) (addr_t) sclp_init_sccb) {
+ printk(KERN_WARNING SCLP_HEADER "unsolicited interrupt "
+ "for buffer at 0x%x\n", finished_sccb);
+ return;
}
-
- /* The sclp_init_mask failed. SCLP is broken, unregister and exit. */
- ctl_clear_bit(0,9);
- unregister_early_external_interrupt(0x2401, sclp_interrupt_handler,
- &ext_int_info_hwc);
-
- return rc;
+ spin_lock(&sclp_lock);
+ if (sclp_running_state == sclp_running_state_running) {
+ sclp_init_req.status = SCLP_REQ_DONE;
+ sclp_running_state = sclp_running_state_idle;
+ }
+ spin_unlock(&sclp_lock);
}

-/*
- * Register the SCLP event listener identified by REG. Return 0 on success.
- * Some error codes and their meaning:
- *
- * -ENODEV = SCLP interface is not supported on this machine
- * -EBUSY = there is already a listener registered for the requested
- * event type
- * -EIO = SCLP interface is currently not operational
- */
-int
-sclp_register(struct sclp_register *reg)
+/* Initial init mask request timed out. Modify request state to failed. */
+static void
+sclp_check_timeout(unsigned long data)
{
unsigned long flags;
- struct list_head *l;
- struct sclp_register *t;
-
- if (!MACHINE_HAS_SCLP)
- return -ENODEV;

- if (!test_bit(SCLP_INIT, &sclp_status))
- sclp_init();
spin_lock_irqsave(&sclp_lock, flags);
- /* check already registered event masks for collisions */
- list_for_each(l, &sclp_reg_list) {
- t = list_entry(l, struct sclp_register, list);
- if (t->receive_mask & reg->receive_mask ||
- t->send_mask & reg->send_mask) {
- spin_unlock_irqrestore(&sclp_lock, flags);
- return -EBUSY;
- }
+ if (sclp_running_state == sclp_running_state_running) {
+ sclp_init_req.status = SCLP_REQ_FAILED;
+ sclp_running_state = sclp_running_state_idle;
}
- /*
- * set present mask to 0 to trigger state change
- * callback in sclp_init_mask
- */
- reg->sclp_receive_mask = 0;
- reg->sclp_send_mask = 0;
- list_add(&reg->list, &sclp_reg_list);
spin_unlock_irqrestore(&sclp_lock, flags);
- sclp_init_mask();
- return 0;
}

-/*
- * Unregister the SCLP event listener identified by REG.
- */
-void
-sclp_unregister(struct sclp_register *reg)
+/* Perform a check of the SCLP interface. Return zero if the interface is
+ * available and there are no pending requests from a previous instance.
+ * Return non-zero otherwise. */
+static int
+sclp_check_interface(void)
{
+ struct init_sccb *sccb;
unsigned long flags;
+ int retry;
+ int rc;

spin_lock_irqsave(&sclp_lock, flags);
- list_del(&reg->list);
+ /* Prepare init mask command */
+ rc = register_early_external_interrupt(0x2401, sclp_check_handler,
+ &ext_int_info_hwc);
+ if (rc) {
+ spin_unlock_irqrestore(&sclp_lock, flags);
+ return rc;
+ }
+ for (retry = 0; retry <= SCLP_INIT_RETRY; retry++) {
+ __sclp_make_init_req(0, 0);
+ sccb = (struct init_sccb *) sclp_init_req.sccb;
+ rc = service_call(sclp_init_req.command, sccb);
+ if (rc == -EIO)
+ break;
+ sclp_init_req.status = SCLP_REQ_RUNNING;
+ sclp_running_state = sclp_running_state_running;
+ __sclp_set_request_timer(SCLP_RETRY_INTERVAL * HZ,
+ sclp_check_timeout, 0);
+ spin_unlock_irqrestore(&sclp_lock, flags);
+ /* Enable service-signal interruption - needs to happen
+ * with IRQs enabled. */
+ ctl_set_bit(0, 9);
+ /* Wait for signal from interrupt or timeout */
+ sclp_sync_wait();
+ /* Disable service-signal interruption - needs to happen
+ * with IRQs enabled. */
+ ctl_clear_bit(0,9);
+ spin_lock_irqsave(&sclp_lock, flags);
+ del_timer(&sclp_request_timer);
+ if (sclp_init_req.status == SCLP_REQ_DONE &&
+ sccb->header.response_code == 0x20) {
+ rc = 0;
+ break;
+ } else
+ rc = -EBUSY;
+ }
+ unregister_early_external_interrupt(0x2401, sclp_check_handler,
+ &ext_int_info_hwc);
spin_unlock_irqrestore(&sclp_lock, flags);
- sclp_init_mask();
+ return rc;
}

-#define SCLP_EVBUF_PROCESSED 0x80
+/* Reboot event handler. Reset send and receive mask to prevent pending SCLP
+ * events from interfering with rebooted system. */
+static int
+sclp_reboot_event(struct notifier_block *this, unsigned long event, void *ptr)
+{
+ sclp_deactivate();
+ return NOTIFY_DONE;
+}

-/*
- * Traverse array of event buffers contained in SCCB and remove all buffers
- * with a set "processed" flag. Return the number of unprocessed buffers.
- */
-int
-sclp_remove_processed(struct sccb_header *sccb)
+static struct notifier_block sclp_reboot_notifier = {
+ .notifier_call = sclp_reboot_event
+};
+
+/* Initialize SCLP driver. Return zero if driver is operational, non-zero
+ * otherwise. */
+static int
+sclp_init(void)
{
- struct evbuf_header *evbuf;
- int unprocessed;
- u16 remaining;
+ unsigned long flags;
+ int rc;

- evbuf = (struct evbuf_header *) (sccb + 1);
- unprocessed = 0;
- remaining = sccb->length - sizeof(struct sccb_header);
- while (remaining > 0) {
- remaining -= evbuf->length;
- if (evbuf->flags & SCLP_EVBUF_PROCESSED) {
- sccb->length -= evbuf->length;
- memcpy((void *) evbuf,
- (void *) ((addr_t) evbuf + evbuf->length),
- remaining);
- } else {
- unprocessed++;
- evbuf = (struct evbuf_header *)
- ((addr_t) evbuf + evbuf->length);
- }
+ if (!MACHINE_HAS_SCLP)
+ return -ENODEV;
+ spin_lock_irqsave(&sclp_lock, flags);
+ /* Check for previous or running initialization */
+ if (sclp_init_state != sclp_init_state_uninitialized) {
+ spin_unlock_irqrestore(&sclp_lock, flags);
+ return 0;
}
-
- return unprocessed;
+ sclp_init_state = sclp_init_state_initializing;
+ /* Set up variables */
+ INIT_LIST_HEAD(&sclp_req_queue);
+ INIT_LIST_HEAD(&sclp_reg_list);
+ list_add(&sclp_state_change_event.list, &sclp_reg_list);
+ init_timer(&sclp_request_timer);
+ /* Check interface */
+ spin_unlock_irqrestore(&sclp_lock, flags);
+ rc = sclp_check_interface();
+ spin_lock_irqsave(&sclp_lock, flags);
+ if (rc) {
+ sclp_init_state = sclp_init_state_uninitialized;
+ spin_unlock_irqrestore(&sclp_lock, flags);
+ return rc;
+ }
+ /* Register reboot handler */
+ rc = register_reboot_notifier(&sclp_reboot_notifier);
+ if (rc) {
+ sclp_init_state = sclp_init_state_uninitialized;
+ spin_unlock_irqrestore(&sclp_lock, flags);
+ return rc;
+ }
+ /* Register interrupt handler */
+ rc = register_early_external_interrupt(0x2401, sclp_interrupt_handler,
+ &ext_int_info_hwc);
+ if (rc) {
+ unregister_reboot_notifier(&sclp_reboot_notifier);
+ sclp_init_state = sclp_init_state_uninitialized;
+ spin_unlock_irqrestore(&sclp_lock, flags);
+ return rc;
+ }
+ sclp_init_state = sclp_init_state_initialized;
+ spin_unlock_irqrestore(&sclp_lock, flags);
+ /* Enable service-signal external interruption - needs to happen with
+ * IRQs enabled. */
+ ctl_set_bit(0, 9);
+ sclp_init_mask(1);
+ return 0;
}
-
-module_init(sclp_init);
-
-EXPORT_SYMBOL(sclp_add_request);
-EXPORT_SYMBOL(sclp_sync_wait);
-EXPORT_SYMBOL(sclp_register);
-EXPORT_SYMBOL(sclp_unregister);
-EXPORT_SYMBOL(sclp_error_message);
diff -urN linux-2.6/drivers/s390/char/sclp_con.c linux-2.6-patched/drivers/s390/char/sclp_con.c
--- linux-2.6/drivers/s390/char/sclp_con.c 2004-12-24 22:34:46.000000000 +0100
+++ linux-2.6-patched/drivers/s390/char/sclp_con.c 2004-12-28 08:50:53.000000000 +0100
@@ -49,25 +49,22 @@
sclp_conbuf_callback(struct sclp_buffer *buffer, int rc)
{
unsigned long flags;
- struct sclp_buffer *next;
void *page;

- /* Ignore return code - because console-writes aren't critical,
- we do without a sophisticated error recovery mechanism. */
- page = sclp_unmake_buffer(buffer);
- spin_lock_irqsave(&sclp_con_lock, flags);
- /* Remove buffer from outqueue */
- list_del(&buffer->list);
- sclp_con_buffer_count--;
- list_add_tail((struct list_head *) page, &sclp_con_pages);
- /* Check if there is a pending buffer on the out queue. */
- next = NULL;
- if (!list_empty(&sclp_con_outqueue))
- next = list_entry(sclp_con_outqueue.next,
- struct sclp_buffer, list);
- spin_unlock_irqrestore(&sclp_con_lock, flags);
- if (next != NULL)
- sclp_emit_buffer(next, sclp_conbuf_callback);
+ do {
+ page = sclp_unmake_buffer(buffer);
+ spin_lock_irqsave(&sclp_con_lock, flags);
+ /* Remove buffer from outqueue */
+ list_del(&buffer->list);
+ sclp_con_buffer_count--;
+ list_add_tail((struct list_head *) page, &sclp_con_pages);
+ /* Check if there is a pending buffer on the out queue. */
+ buffer = NULL;
+ if (!list_empty(&sclp_con_outqueue))
+ buffer = list_entry(sclp_con_outqueue.next,
+ struct sclp_buffer, list);
+ spin_unlock_irqrestore(&sclp_con_lock, flags);
+ } while (buffer && sclp_emit_buffer(buffer, sclp_conbuf_callback));
}

static inline void
@@ -76,6 +73,7 @@
struct sclp_buffer* buffer;
unsigned long flags;
int count;
+ int rc;

spin_lock_irqsave(&sclp_con_lock, flags);
buffer = sclp_conbuf;
@@ -87,8 +85,11 @@
list_add_tail(&buffer->list, &sclp_con_outqueue);
count = sclp_con_buffer_count++;
spin_unlock_irqrestore(&sclp_con_lock, flags);
- if (count == 0)
- sclp_emit_buffer(buffer, sclp_conbuf_callback);
+ if (count)
+ return;
+ rc = sclp_emit_buffer(buffer, sclp_conbuf_callback);
+ if (rc)
+ sclp_conbuf_callback(buffer, rc);
}

/*
diff -urN linux-2.6/drivers/s390/char/sclp_cpi.c linux-2.6-patched/drivers/s390/char/sclp_cpi.c
--- linux-2.6/drivers/s390/char/sclp_cpi.c 2004-12-24 22:34:57.000000000 +0100
+++ linux-2.6-patched/drivers/s390/char/sclp_cpi.c 2004-12-28 08:50:53.000000000 +0100
@@ -196,18 +196,20 @@
rc = sclp_register(&sclp_cpi_event);
if (rc) {
/* could not register sclp event. Die. */
- printk("cpi: could not register to hardware console.\n");
+ printk(KERN_WARNING "cpi: could not register to hardware "
+ "console.\n");
return -EINVAL;
}
if (!(sclp_cpi_event.sclp_send_mask & EvTyp_CtlProgIdent_Mask)) {
- printk("cpi: no control program identification support\n");
+ printk(KERN_WARNING "cpi: no control program identification "
+ "support\n");
sclp_unregister(&sclp_cpi_event);
return -ENOTSUPP;
}

req = cpi_prepare_req();
if (IS_ERR(req)) {
- printk("cpi: couldn't allocate request\n");
+ printk(KERN_WARNING "cpi: couldn't allocate request\n");
sclp_unregister(&sclp_cpi_event);
return PTR_ERR(req);
}
@@ -216,13 +218,20 @@
sema_init(&sem, 0);
req->callback_data = &sem;
/* Add request to sclp queue */
- sclp_add_request(req);
+ rc = sclp_add_request(req);
+ if (rc) {
+ printk(KERN_WARNING "cpi: could not start request\n");
+ cpi_free_req(req);
+ sclp_unregister(&sclp_cpi_event);
+ return rc;
+ }
/* make "insmod" sleep until callback arrives */
down(&sem);

rc = ((struct cpi_sccb *) req->sccb)->header.response_code;
if (rc != 0x0020) {
- printk("cpi: failed with response code 0x%x\n", rc);
+ printk(KERN_WARNING "cpi: failed with response code 0x%x\n",
+ rc);
rc = -ECOMM;
} else
rc = 0;
diff -urN linux-2.6/drivers/s390/char/sclp.h linux-2.6-patched/drivers/s390/char/sclp.h
--- linux-2.6/drivers/s390/char/sclp.h 2004-12-24 22:34:01.000000000 +0100
+++ linux-2.6-patched/drivers/s390/char/sclp.h 2004-12-28 08:50:53.000000000 +0100
@@ -95,6 +95,7 @@
sclp_cmdw_t command; /* sclp command to execute */
void *sccb; /* pointer to the sccb to execute */
char status; /* status of this request */
+ int start_count; /* number of SVCs done for this req */
/* Callback that is called after reaching final status. */
void (*callback)(struct sclp_req *, void *data);
void *callback_data;
@@ -123,12 +124,13 @@
};

/* externals from sclp.c */
-void sclp_add_request(struct sclp_req *req);
+int sclp_add_request(struct sclp_req *req);
void sclp_sync_wait(void);
int sclp_register(struct sclp_register *reg);
void sclp_unregister(struct sclp_register *reg);
-char *sclp_error_message(u16 response_code);
int sclp_remove_processed(struct sccb_header *sccb);
+int sclp_deactivate(void);
+int sclp_reactivate(void);

/* useful inlines */

diff -urN linux-2.6/drivers/s390/char/sclp_quiesce.c linux-2.6-patched/drivers/s390/char/sclp_quiesce.c
--- linux-2.6/drivers/s390/char/sclp_quiesce.c 1970-01-01 01:00:00.000000000 +0100
+++ linux-2.6-patched/drivers/s390/char/sclp_quiesce.c 2004-12-28 08:50:53.000000000 +0100
@@ -0,0 +1,114 @@
+/*
+ * drivers/s390/char/sclp_quiesce.c
+ * signal quiesce handler
+ *
+ * (C) Copyright IBM Corp. 1999,2004
+ * Author(s): Martin Schwidefsky <schwidefsky@de.ibm.com>
+ * Peter Oberparleiter <peter.oberparleiter@de.ibm.com>
+ */
+
+#include <linux/config.h>
+#include <linux/module.h>
+#include <linux/types.h>
+#include <linux/cpumask.h>
+#include <linux/smp.h>
+#include <linux/init.h>
+#include <asm/atomic.h>
+#include <asm/ptrace.h>
+#include <asm/sigp.h>
+
+#include "sclp.h"
+
+
+#ifdef CONFIG_SMP
+/* Signal completion of shutdown process. All CPUs except the first to enter
+ * this function: go to stopped state. First CPU: wait until all other
+ * CPUs are in stopped or check stop state. Afterwards, load special PSW
+ * to indicate completion. */
+static void
+do_load_quiesce_psw(void * __unused)
+{
+ static atomic_t cpuid = ATOMIC_INIT(-1);
+ psw_t quiesce_psw;
+ __u32 status;
+ int i;
+
+ if (atomic_compare_and_swap(-1, smp_processor_id(), &cpuid))
+ signal_processor(smp_processor_id(), sigp_stop);
+ /* Wait for all other cpus to enter stopped state */
+ i = 1;
+ while (i < NR_CPUS) {
+ if (!cpu_online(i)) {
+ i++;
+ continue;
+ }
+ switch (signal_processor_ps(&status, 0, i, sigp_sense)) {
+ case sigp_order_code_accepted:
+ case sigp_status_stored:
+ /* Check for stopped and check stop state */
+ if (status & 0x50)
+ i++;
+ break;
+ case sigp_busy:
+ break;
+ case sigp_not_operational:
+ i++;
+ break;
+ }
+ }
+ /* Quiesce the last cpu with the special psw */
+ quiesce_psw.mask = PSW_BASE_BITS | PSW_MASK_WAIT;
+ quiesce_psw.addr = 0xfff;
+ __load_psw(quiesce_psw);
+}
+
+/* Shutdown handler. Perform shutdown function on all CPUs. */
+static void
+do_machine_quiesce(void)
+{
+ on_each_cpu(do_load_quiesce_psw, NULL, 0, 0);
+}
+#else
+/* Shutdown handler. Signal completion of shutdown by loading special PSW. */
+static void
+do_machine_quiesce(void)
+{
+ psw_t quiesce_psw;
+
+ quiesce_psw.mask = PSW_BASE_BITS | PSW_MASK_WAIT;
+ quiesce_psw.addr = 0xfff;
+ __load_psw(quiesce_psw);
+}
+#endif
+
+extern void ctrl_alt_del(void);
+
+/* Handler for quiesce event. Start shutdown procedure. */
+static void
+sclp_quiesce_handler(struct evbuf_header *evbuf)
+{
+ _machine_restart = (void *) do_machine_quiesce;
+ _machine_halt = do_machine_quiesce;
+ _machine_power_off = do_machine_quiesce;
+ ctrl_alt_del();
+}
+
+static struct sclp_register sclp_quiesce_event = {
+ .receive_mask = EvTyp_SigQuiesce_Mask,
+ .receiver_fn = sclp_quiesce_handler
+};
+
+/* Initialize quiesce driver. */
+static int __init
+sclp_quiesce_init(void)
+{
+ int rc;
+
+ rc = sclp_register(&sclp_quiesce_event);
+ if (rc)
+ printk(KERN_WARNING "sclp: could not register quiesce handler "
+ "(rc=%d)\n", rc);
+ return rc;
+}
+
+module_init(sclp_quiesce_init);
diff -urN linux-2.6/drivers/s390/char/sclp_rw.c linux-2.6-patched/drivers/s390/char/sclp_rw.c
--- linux-2.6/drivers/s390/char/sclp_rw.c 2004-12-24 22:34:44.000000000 +0100
+++ linux-2.6-patched/drivers/s390/char/sclp_rw.c 2004-12-28 08:50:53.000000000 +0100
@@ -54,7 +54,6 @@
buffer = ((struct sclp_buffer *) ((addr_t) sccb + PAGE_SIZE)) - 1;
buffer->sccb = sccb;
buffer->retry_count = 0;
- init_timer(&buffer->retry_timer);
buffer->mto_number = 0;
buffer->mto_char_sum = 0;
buffer->current_line = NULL;
@@ -365,17 +364,7 @@
return rc;
}

-static void
-sclp_buffer_retry(unsigned long data)
-{
- struct sclp_buffer *buffer = (struct sclp_buffer *) data;
- buffer->request.status = SCLP_REQ_FILLED;
- buffer->sccb->header.response_code = 0x0000;
- sclp_add_request(&buffer->request);
-}
-
-#define SCLP_BUFFER_MAX_RETRY 5
-#define SCLP_BUFFER_RETRY_INTERVAL 2
+#define SCLP_BUFFER_MAX_RETRY 1

/*
* second half of Write Event Data-function that has to be done after
@@ -404,7 +393,7 @@
break;

case 0x0340: /* Contained SCLP equipment check */
- if (buffer->retry_count++ > SCLP_BUFFER_MAX_RETRY) {
+ if (++buffer->retry_count > SCLP_BUFFER_MAX_RETRY) {
rc = -EIO;
break;
}
@@ -413,26 +402,26 @@
/* not all buffers were processed */
sccb->header.response_code = 0x0000;
buffer->request.status = SCLP_REQ_FILLED;
- sclp_add_request(request);
- return;
- }
- rc = 0;
+ rc = sclp_add_request(request);
+ if (rc == 0)
+ return;
+ } else
+ rc = 0;
break;

case 0x0040: /* SCLP equipment check */
case 0x05f0: /* Target resource in improper state */
- if (buffer->retry_count++ > SCLP_BUFFER_MAX_RETRY) {
+ if (++buffer->retry_count > SCLP_BUFFER_MAX_RETRY) {
rc = -EIO;
break;
}
- /* wait some time, then retry request */
- buffer->retry_timer.function = sclp_buffer_retry;
- buffer->retry_timer.data = (unsigned long) buffer;
- buffer->retry_timer.expires = jiffies +
- SCLP_BUFFER_RETRY_INTERVAL*HZ;
- add_timer(&buffer->retry_timer);
- return;
-
+ /* retry request */
+ sccb->header.response_code = 0x0000;
+ buffer->request.status = SCLP_REQ_FILLED;
+ rc = sclp_add_request(request);
+ if (rc == 0)
+ return;
+ break;
default:
if (sccb->header.response_code == 0x71f0)
rc = -ENOMEM;
@@ -446,9 +435,10 @@

/*
* Setup the request structure in the struct sclp_buffer to do SCLP Write
- * Event Data and pass the request to the core SCLP loop.
+ * Event Data and pass the request to the core SCLP loop. Return zero on
+ * success, non-zero otherwise.
*/
-void
+int
sclp_emit_buffer(struct sclp_buffer *buffer,
void (*callback)(struct sclp_buffer *, int))
{
@@ -459,11 +449,8 @@
sclp_finalize_mto(buffer);

/* Are there messages in the output buffer ? */
- if (buffer->mto_number == 0) {
- if (callback != NULL)
- callback(buffer, 0);
- return;
- }
+ if (buffer->mto_number == 0)
+ return -EIO;

sccb = buffer->sccb;
if (sclp_rw_event.sclp_send_mask & EvTyp_Msg_Mask)
@@ -472,16 +459,13 @@
else if (sclp_rw_event.sclp_send_mask & EvTyp_PMsgCmd_Mask)
/* Use write priority message */
sccb->msg_buf.header.type = EvTyp_PMsgCmd;
- else {
- if (callback != NULL)
- callback(buffer, -ENOSYS);
- return;
- }
+ else
+ return -ENOSYS;
buffer->request.command = SCLP_CMDW_WRITEDATA;
buffer->request.status = SCLP_REQ_FILLED;
buffer->request.callback = sclp_writedata_callback;
buffer->request.callback_data = buffer;
buffer->request.sccb = sccb;
buffer->callback = callback;
- sclp_add_request(&buffer->request);
+ return sclp_add_request(&buffer->request);
}
diff -urN linux-2.6/drivers/s390/char/sclp_rw.h linux-2.6-patched/drivers/s390/char/sclp_rw.h
--- linux-2.6/drivers/s390/char/sclp_rw.h 2004-12-24 22:35:50.000000000 +0100
+++ linux-2.6-patched/drivers/s390/char/sclp_rw.h 2004-12-28 08:50:53.000000000 +0100
@@ -12,7 +12,6 @@
#define __SCLP_RW_H__

#include <linux/list.h>
-#include <linux/timer.h>

struct mto {
u16 length;
@@ -74,7 +73,6 @@
char *current_line;
int current_length;
int retry_count;
- struct timer_list retry_timer;
/* output format settings */
unsigned short columns;
unsigned short htab;
@@ -90,7 +88,7 @@
void *sclp_unmake_buffer(struct sclp_buffer *);
int sclp_buffer_space(struct sclp_buffer *);
int sclp_write(struct sclp_buffer *buffer, const unsigned char *, int);
-void sclp_emit_buffer(struct sclp_buffer *,void (*)(struct sclp_buffer *,int));
+int sclp_emit_buffer(struct sclp_buffer *,void (*)(struct sclp_buffer *,int));
void sclp_set_columns(struct sclp_buffer *, unsigned short);
void sclp_set_htab(struct sclp_buffer *, unsigned short);
int sclp_chars_in_buffer(struct sclp_buffer *);
diff -urN linux-2.6/drivers/s390/char/sclp_tty.c linux-2.6-patched/drivers/s390/char/sclp_tty.c
--- linux-2.6/drivers/s390/char/sclp_tty.c 2004-12-24 22:35:50.000000000 +0100
+++ linux-2.6-patched/drivers/s390/char/sclp_tty.c 2004-12-28 08:50:53.000000000 +0100
@@ -255,25 +255,22 @@
sclp_ttybuf_callback(struct sclp_buffer *buffer, int rc)
{
unsigned long flags;
- struct sclp_buffer *next;
void *page;

- /* Ignore return code - because tty-writes aren't critical,
- we do without a sophisticated error recovery mechanism. */
- page = sclp_unmake_buffer(buffer);
- spin_lock_irqsave(&sclp_tty_lock, flags);
- /* Remove buffer from outqueue */
- list_del(&buffer->list);
- sclp_tty_buffer_count--;
- list_add_tail((struct list_head *) page, &sclp_tty_pages);
- /* Check if there is a pending buffer on the out queue. */
- next = NULL;
- if (!list_empty(&sclp_tty_outqueue))
- next = list_entry(sclp_tty_outqueue.next,
- struct sclp_buffer, list);
- spin_unlock_irqrestore(&sclp_tty_lock, flags);
- if (next != NULL)
- sclp_emit_buffer(next, sclp_ttybuf_callback);
+ do {
+ page = sclp_unmake_buffer(buffer);
+ spin_lock_irqsave(&sclp_tty_lock, flags);
+ /* Remove buffer from outqueue */
+ list_del(&buffer->list);
+ sclp_tty_buffer_count--;
+ list_add_tail((struct list_head *) page, &sclp_tty_pages);
+ /* Check if there is a pending buffer on the out queue. */
+ buffer = NULL;
+ if (!list_empty(&sclp_tty_outqueue))
+ buffer = list_entry(sclp_tty_outqueue.next,
+ struct sclp_buffer, list);
+ spin_unlock_irqrestore(&sclp_tty_lock, flags);
+ } while (buffer && sclp_emit_buffer(buffer, sclp_ttybuf_callback));
wake_up(&sclp_tty_waitq);
/* check if the tty needs a wake up call */
if (sclp_tty != NULL) {
@@ -286,14 +283,17 @@
{
unsigned long flags;
int count;
+ int rc;

spin_lock_irqsave(&sclp_tty_lock, flags);
list_add_tail(&buffer->list, &sclp_tty_outqueue);
count = sclp_tty_buffer_count++;
spin_unlock_irqrestore(&sclp_tty_lock, flags);
-
- if (count == 0)
- sclp_emit_buffer(buffer, sclp_ttybuf_callback);
+ if (count)
+ return;
+ rc = sclp_emit_buffer(buffer, sclp_ttybuf_callback);
+ if (rc)
+ sclp_ttybuf_callback(buffer, rc);
}

/*
diff -urN linux-2.6/drivers/s390/char/sclp_vt220.c linux-2.6-patched/drivers/s390/char/sclp_vt220.c
--- linux-2.6/drivers/s390/char/sclp_vt220.c 2004-12-24 22:34:30.000000000 +0100
+++ linux-2.6-patched/drivers/s390/char/sclp_vt220.c 2004-12-28 08:50:53.000000000 +0100
@@ -42,7 +42,6 @@
struct list_head list;
struct sclp_req sclp_req;
int retry_count;
- struct timer_list retry_timer;
};

/* VT220 SCCB */
@@ -96,7 +95,7 @@
static int sclp_vt220_flush_later;

static void sclp_vt220_receiver_fn(struct evbuf_header *evbuf);
-static void __sclp_vt220_emit(struct sclp_vt220_request *request);
+static int __sclp_vt220_emit(struct sclp_vt220_request *request);
static void sclp_vt220_emit_current(void);

/* Registration structure for our interest in SCLP event buffers */
@@ -116,25 +115,24 @@
sclp_vt220_process_queue(struct sclp_vt220_request *request)
{
unsigned long flags;
- struct sclp_vt220_request *next;
void *page;

- /* Put buffer back to list of empty buffers */
- page = request->sclp_req.sccb;
- spin_lock_irqsave(&sclp_vt220_lock, flags);
- /* Move request from outqueue to empty queue */
- list_del(&request->list);
- sclp_vt220_outqueue_count--;
- list_add_tail((struct list_head *) page, &sclp_vt220_empty);
- /* Check if there is a pending buffer on the out queue. */
- next = NULL;
- if (!list_empty(&sclp_vt220_outqueue))
- next = list_entry(sclp_vt220_outqueue.next,
- struct sclp_vt220_request, list);
- spin_unlock_irqrestore(&sclp_vt220_lock, flags);
- if (next != NULL)
- __sclp_vt220_emit(next);
- else if (sclp_vt220_flush_later)
+ do {
+ /* Put buffer back to list of empty buffers */
+ page = request->sclp_req.sccb;
+ spin_lock_irqsave(&sclp_vt220_lock, flags);
+ /* Move request from outqueue to empty queue */
+ list_del(&request->list);
+ sclp_vt220_outqueue_count--;
+ list_add_tail((struct list_head *) page, &sclp_vt220_empty);
+ /* Check if there is a pending buffer on the out queue. */
+ request = NULL;
+ if (!list_empty(&sclp_vt220_outqueue))
+ request = list_entry(sclp_vt220_outqueue.next,
+ struct sclp_vt220_request, list);
+ spin_unlock_irqrestore(&sclp_vt220_lock, flags);
+ } while (request && __sclp_vt220_emit(request));
+ if (request == NULL && sclp_vt220_flush_later)
sclp_vt220_emit_current();
wake_up(&sclp_vt220_waitq);
/* Check if the tty needs a wake up call */
@@ -143,25 +141,7 @@
}
}

-/*
- * Retry sclp write request after waiting some time for an sclp equipment
- * check to pass.
- */
-static void
-sclp_vt220_retry(unsigned long data)
-{
- struct sclp_vt220_request *request;
- struct sclp_vt220_sccb *sccb;
-
- request = (struct sclp_vt220_request *) data;
- request->sclp_req.status = SCLP_REQ_FILLED;
- sccb = (struct sclp_vt220_sccb *) request->sclp_req.sccb;
- sccb->header.response_code = 0x0000;
- sclp_add_request(&request->sclp_req);
-}
-
-#define SCLP_BUFFER_MAX_RETRY 5
-#define SCLP_BUFFER_RETRY_INTERVAL 2
+#define SCLP_BUFFER_MAX_RETRY 1

/*
* Callback through which the result of a write request is reported by the
@@ -189,29 +169,26 @@
break;

case 0x0340: /* Contained SCLP equipment check */
- if (vt220_request->retry_count++ > SCLP_BUFFER_MAX_RETRY)
+ if (++vt220_request->retry_count > SCLP_BUFFER_MAX_RETRY)
break;
/* Remove processed buffers and requeue rest */
if (sclp_remove_processed((struct sccb_header *) sccb) > 0) {
/* Not all buffers were processed */
sccb->header.response_code = 0x0000;
vt220_request->sclp_req.status = SCLP_REQ_FILLED;
- sclp_add_request(request);
- return;
+ if (sclp_add_request(request) == 0)
+ return;
}
break;

case 0x0040: /* SCLP equipment check */
- if (vt220_request->retry_count++ > SCLP_BUFFER_MAX_RETRY)
+ if (++vt220_request->retry_count > SCLP_BUFFER_MAX_RETRY)
break;
- /* Wait some time, then retry request */
- vt220_request->retry_timer.function = sclp_vt220_retry;
- vt220_request->retry_timer.data =
- (unsigned long) vt220_request;
- vt220_request->retry_timer.expires =
- jiffies + SCLP_BUFFER_RETRY_INTERVAL*HZ;
- add_timer(&vt220_request->retry_timer);
- return;
+ sccb->header.response_code = 0x0000;
+ vt220_request->sclp_req.status = SCLP_REQ_FILLED;
+ if (sclp_add_request(request) == 0)
+ return;
+ break;

default:
break;
@@ -220,22 +197,22 @@
}

/*
- * Emit vt220 request buffer to SCLP.
+ * Emit vt220 request buffer to SCLP. Return zero on success, non-zero
+ * otherwise.
*/
-static void
+static int
__sclp_vt220_emit(struct sclp_vt220_request *request)
{
if (!(sclp_vt220_register.sclp_send_mask & EvTyp_VT220Msg_Mask)) {
request->sclp_req.status = SCLP_REQ_FAILED;
- sclp_vt220_callback(&request->sclp_req, (void *) request);
- return;
+ return -EIO;
}
request->sclp_req.command = SCLP_CMDW_WRITEDATA;
request->sclp_req.status = SCLP_REQ_FILLED;
request->sclp_req.callback = sclp_vt220_callback;
request->sclp_req.callback_data = (void *) request;

- sclp_add_request(&request->sclp_req);
+ return sclp_add_request(&request->sclp_req);
}

/*
@@ -253,12 +230,12 @@
spin_unlock_irqrestore(&sclp_vt220_lock, flags);
/* Emit only the first buffer immediately - callback takes care of
* the rest */
- if (count == 0)
- __sclp_vt220_emit(request);
+ if (count == 0 && __sclp_vt220_emit(request))
+ sclp_vt220_process_queue(request);
}

/*
- * Queue and emit current request.
+ * Queue and emit current request. Return zero on success, non-zero otherwise.
*/
static void
sclp_vt220_emit_current(void)
@@ -300,7 +277,6 @@
/* Place request structure at end of page */
request = ((struct sclp_vt220_request *)
((addr_t) page + PAGE_SIZE)) - 1;
- init_timer(&request->retry_timer);
request->retry_count = 0;
request->sclp_req.sccb = page;
/* SCCB goes at start of page */
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
\
 
 \ /
  Last update: 2005-03-22 14:09    [W:1.493 / U:1.188 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site