linux/drivers/s390/block/dasd_eer.c

710 lines
20 KiB
C
Raw Normal View History

/*
* Character device driver for extended error reporting.
*
* Copyright (C) 2005 IBM Corporation
* extended error reporting for DASD ECKD devices
* Author(s): Stefan Weinhuber <wein@de.ibm.com>
*/
#define KMSG_COMPONENT "dasd-eckd"
#include <linux/init.h>
#include <linux/fs.h>
#include <linux/kernel.h>
#include <linux/miscdevice.h>
#include <linux/module.h>
#include <linux/moduleparam.h>
#include <linux/device.h>
#include <linux/poll.h>
#include <linux/mutex.h>
#include <linux/err.h>
include cleanup: Update gfp.h and slab.h includes to prepare for breaking implicit slab.h inclusion from percpu.h percpu.h is included by sched.h and module.h and thus ends up being included when building most .c files. percpu.h includes slab.h which in turn includes gfp.h making everything defined by the two files universally available and complicating inclusion dependencies. percpu.h -> slab.h dependency is about to be removed. Prepare for this change by updating users of gfp and slab facilities include those headers directly instead of assuming availability. As this conversion needs to touch large number of source files, the following script is used as the basis of conversion. http://userweb.kernel.org/~tj/misc/slabh-sweep.py The script does the followings. * Scan files for gfp and slab usages and update includes such that only the necessary includes are there. ie. if only gfp is used, gfp.h, if slab is used, slab.h. * When the script inserts a new include, it looks at the include blocks and try to put the new include such that its order conforms to its surrounding. It's put in the include block which contains core kernel includes, in the same order that the rest are ordered - alphabetical, Christmas tree, rev-Xmas-tree or at the end if there doesn't seem to be any matching order. * If the script can't find a place to put a new include (mostly because the file doesn't have fitting include block), it prints out an error message indicating which .h file needs to be added to the file. The conversion was done in the following steps. 1. The initial automatic conversion of all .c files updated slightly over 4000 files, deleting around 700 includes and adding ~480 gfp.h and ~3000 slab.h inclusions. The script emitted errors for ~400 files. 2. Each error was manually checked. Some didn't need the inclusion, some needed manual addition while adding it to implementation .h or embedding .c file was more appropriate for others. This step added inclusions to around 150 files. 3. The script was run again and the output was compared to the edits from #2 to make sure no file was left behind. 4. Several build tests were done and a couple of problems were fixed. e.g. lib/decompress_*.c used malloc/free() wrappers around slab APIs requiring slab.h to be added manually. 5. The script was run on all .h files but without automatically editing them as sprinkling gfp.h and slab.h inclusions around .h files could easily lead to inclusion dependency hell. Most gfp.h inclusion directives were ignored as stuff from gfp.h was usually wildly available and often used in preprocessor macros. Each slab.h inclusion directive was examined and added manually as necessary. 6. percpu.h was updated not to include slab.h. 7. Build test were done on the following configurations and failures were fixed. CONFIG_GCOV_KERNEL was turned off for all tests (as my distributed build env didn't work with gcov compiles) and a few more options had to be turned off depending on archs to make things build (like ipr on powerpc/64 which failed due to missing writeq). * x86 and x86_64 UP and SMP allmodconfig and a custom test config. * powerpc and powerpc64 SMP allmodconfig * sparc and sparc64 SMP allmodconfig * ia64 SMP allmodconfig * s390 SMP allmodconfig * alpha SMP allmodconfig * um on x86_64 SMP allmodconfig 8. percpu.h modifications were reverted so that it could be applied as a separate patch and serve as bisection point. Given the fact that I had only a couple of failures from tests on step 6, I'm fairly confident about the coverage of this conversion patch. If there is a breakage, it's likely to be something in one of the arch headers which should be easily discoverable easily on most builds of the specific arch. Signed-off-by: Tejun Heo <tj@kernel.org> Guess-its-ok-by: Christoph Lameter <cl@linux-foundation.org> Cc: Ingo Molnar <mingo@redhat.com> Cc: Lee Schermerhorn <Lee.Schermerhorn@hp.com>
2010-03-24 08:04:11 +00:00
#include <linux/slab.h>
#include <asm/uaccess.h>
#include <linux/atomic.h>
#include <asm/ebcdic.h>
#include "dasd_int.h"
#include "dasd_eckd.h"
#ifdef PRINTK_HEADER
#undef PRINTK_HEADER
#endif /* PRINTK_HEADER */
#define PRINTK_HEADER "dasd(eer):"
/*
* SECTION: the internal buffer
*/
/*
* The internal buffer is meant to store obaque blobs of data, so it does
* not know of higher level concepts like triggers.
* It consists of a number of pages that are used as a ringbuffer. Each data
* blob is stored in a simple record that consists of an integer, which
* contains the size of the following data, and the data bytes themselfes.
*
* To allow for multiple independent readers we create one internal buffer
* each time the device is opened and destroy the buffer when the file is
* closed again. The number of pages used for this buffer is determined by
* the module parmeter eer_pages.
*
* One record can be written to a buffer by using the functions
* - dasd_eer_start_record (one time per record to write the size to the
* buffer and reserve the space for the data)
* - dasd_eer_write_buffer (one or more times per record to write the data)
* The data can be written in several steps but you will have to compute
* the total size up front for the invocation of dasd_eer_start_record.
* If the ringbuffer is full, dasd_eer_start_record will remove the required
* number of old records.
*
* A record is typically read in two steps, first read the integer that
* specifies the size of the following data, then read the data.
* Both can be done by
* - dasd_eer_read_buffer
*
* For all mentioned functions you need to get the bufferlock first and keep
* it until a complete record is written or read.
*
* All information necessary to keep track of an internal buffer is kept in
* a struct eerbuffer. The buffer specific to a file pointer is strored in
* the private_data field of that file. To be able to write data to all
* existing buffers, each buffer is also added to the bufferlist.
* If the user does not want to read a complete record in one go, we have to
* keep track of the rest of the record. residual stores the number of bytes
* that are still to deliver. If the rest of the record is invalidated between
* two reads then residual will be set to -1 so that the next read will fail.
* All entries in the eerbuffer structure are protected with the bufferlock.
* To avoid races between writing to a buffer on the one side and creating
* and destroying buffers on the other side, the bufferlock must also be used
* to protect the bufferlist.
*/
static int eer_pages = 5;
module_param(eer_pages, int, S_IRUGO|S_IWUSR);
struct eerbuffer {
struct list_head list;
char **buffer;
int buffersize;
int buffer_page_count;
int head;
int tail;
int residual;
};
static LIST_HEAD(bufferlist);
static DEFINE_SPINLOCK(bufferlock);
static DECLARE_WAIT_QUEUE_HEAD(dasd_eer_read_wait_queue);
/*
* How many free bytes are available on the buffer.
* Needs to be called with bufferlock held.
*/
static int dasd_eer_get_free_bytes(struct eerbuffer *eerb)
{
if (eerb->head < eerb->tail)
return eerb->tail - eerb->head - 1;
return eerb->buffersize - eerb->head + eerb->tail -1;
}
/*
* How many bytes of buffer space are used.
* Needs to be called with bufferlock held.
*/
static int dasd_eer_get_filled_bytes(struct eerbuffer *eerb)
{
if (eerb->head >= eerb->tail)
return eerb->head - eerb->tail;
return eerb->buffersize - eerb->tail + eerb->head;
}
/*
* The dasd_eer_write_buffer function just copies count bytes of data
* to the buffer. Make sure to call dasd_eer_start_record first, to
* make sure that enough free space is available.
* Needs to be called with bufferlock held.
*/
static void dasd_eer_write_buffer(struct eerbuffer *eerb,
char *data, int count)
{
unsigned long headindex,localhead;
unsigned long rest, len;
char *nextdata;
nextdata = data;
rest = count;
while (rest > 0) {
headindex = eerb->head / PAGE_SIZE;
localhead = eerb->head % PAGE_SIZE;
len = min(rest, PAGE_SIZE - localhead);
memcpy(eerb->buffer[headindex]+localhead, nextdata, len);
nextdata += len;
rest -= len;
eerb->head += len;
if (eerb->head == eerb->buffersize)
eerb->head = 0; /* wrap around */
BUG_ON(eerb->head > eerb->buffersize);
}
}
/*
* Needs to be called with bufferlock held.
*/
static int dasd_eer_read_buffer(struct eerbuffer *eerb, char *data, int count)
{
unsigned long tailindex,localtail;
unsigned long rest, len, finalcount;
char *nextdata;
finalcount = min(count, dasd_eer_get_filled_bytes(eerb));
nextdata = data;
rest = finalcount;
while (rest > 0) {
tailindex = eerb->tail / PAGE_SIZE;
localtail = eerb->tail % PAGE_SIZE;
len = min(rest, PAGE_SIZE - localtail);
memcpy(nextdata, eerb->buffer[tailindex] + localtail, len);
nextdata += len;
rest -= len;
eerb->tail += len;
if (eerb->tail == eerb->buffersize)
eerb->tail = 0; /* wrap around */
BUG_ON(eerb->tail > eerb->buffersize);
}
return finalcount;
}
/*
* Whenever you want to write a blob of data to the internal buffer you
* have to start by using this function first. It will write the number
* of bytes that will be written to the buffer. If necessary it will remove
* old records to make room for the new one.
* Needs to be called with bufferlock held.
*/
static int dasd_eer_start_record(struct eerbuffer *eerb, int count)
{
int tailcount;
if (count + sizeof(count) > eerb->buffersize)
return -ENOMEM;
while (dasd_eer_get_free_bytes(eerb) < count + sizeof(count)) {
if (eerb->residual > 0) {
eerb->tail += eerb->residual;
if (eerb->tail >= eerb->buffersize)
eerb->tail -= eerb->buffersize;
eerb->residual = -1;
}
dasd_eer_read_buffer(eerb, (char *) &tailcount,
sizeof(tailcount));
eerb->tail += tailcount;
if (eerb->tail >= eerb->buffersize)
eerb->tail -= eerb->buffersize;
}
dasd_eer_write_buffer(eerb, (char*) &count, sizeof(count));
return 0;
};
/*
* Release pages that are not used anymore.
*/
static void dasd_eer_free_buffer_pages(char **buf, int no_pages)
{
int i;
for (i = 0; i < no_pages; i++)
free_page((unsigned long) buf[i]);
}
/*
* Allocate a new set of memory pages.
*/
static int dasd_eer_allocate_buffer_pages(char **buf, int no_pages)
{
int i;
for (i = 0; i < no_pages; i++) {
buf[i] = (char *) get_zeroed_page(GFP_KERNEL);
if (!buf[i]) {
dasd_eer_free_buffer_pages(buf, i);
return -ENOMEM;
}
}
return 0;
}
/*
* SECTION: The extended error reporting functionality
*/
/*
* When a DASD device driver wants to report an error, it calls the
* function dasd_eer_write and gives the respective trigger ID as
* parameter. Currently there are four kinds of triggers:
*
* DASD_EER_FATALERROR: all kinds of unrecoverable I/O problems
* DASD_EER_PPRCSUSPEND: PPRC was suspended
* DASD_EER_NOPATH: There is no path to the device left.
* DASD_EER_STATECHANGE: The state of the device has changed.
*
* For the first three triggers all required information can be supplied by
* the caller. For these triggers a record is written by the function
* dasd_eer_write_standard_trigger.
*
* The DASD_EER_STATECHANGE trigger is special since a sense subsystem
* status ccw need to be executed to gather the necessary sense data first.
* The dasd_eer_snss function will queue the SNSS request and the request
* callback will then call dasd_eer_write with the DASD_EER_STATCHANGE
* trigger.
*
* To avoid memory allocations at runtime, the necessary memory is allocated
* when the extended error reporting is enabled for a device (by
* dasd_eer_probe). There is one sense subsystem status request for each
* eer enabled DASD device. The presence of the cqr in device->eer_cqr
* indicates that eer is enable for the device. The use of the snss request
* is protected by the DASD_FLAG_EER_IN_USE bit. When this flag indicates
* that the cqr is currently in use, dasd_eer_snss cannot start a second
* request but sets the DASD_FLAG_EER_SNSS flag instead. The callback of
* the SNSS request will check the bit and call dasd_eer_snss again.
*/
#define SNSS_DATA_SIZE 44
#define DASD_EER_BUSID_SIZE 10
struct dasd_eer_header {
__u32 total_size;
__u32 trigger;
__u64 tv_sec;
__u64 tv_usec;
char busid[DASD_EER_BUSID_SIZE];
} __attribute__ ((packed));
/*
* The following function can be used for those triggers that have
* all necessary data available when the function is called.
* If the parameter cqr is not NULL, the chain of requests will be searched
* for valid sense data, and all valid sense data sets will be added to
* the triggers data.
*/
static void dasd_eer_write_standard_trigger(struct dasd_device *device,
struct dasd_ccw_req *cqr,
int trigger)
{
struct dasd_ccw_req *temp_cqr;
int data_size;
struct timeval tv;
struct dasd_eer_header header;
unsigned long flags;
struct eerbuffer *eerb;
char *sense;
/* go through cqr chain and count the valid sense data sets */
data_size = 0;
for (temp_cqr = cqr; temp_cqr; temp_cqr = temp_cqr->refers)
if (dasd_get_sense(&temp_cqr->irb))
data_size += 32;
header.total_size = sizeof(header) + data_size + 4; /* "EOR" */
header.trigger = trigger;
do_gettimeofday(&tv);
header.tv_sec = tv.tv_sec;
header.tv_usec = tv.tv_usec;
strncpy(header.busid, dev_name(&device->cdev->dev),
DASD_EER_BUSID_SIZE);
spin_lock_irqsave(&bufferlock, flags);
list_for_each_entry(eerb, &bufferlist, list) {
dasd_eer_start_record(eerb, header.total_size);
dasd_eer_write_buffer(eerb, (char *) &header, sizeof(header));
for (temp_cqr = cqr; temp_cqr; temp_cqr = temp_cqr->refers) {
sense = dasd_get_sense(&temp_cqr->irb);
if (sense)
dasd_eer_write_buffer(eerb, sense, 32);
}
dasd_eer_write_buffer(eerb, "EOR", 4);
}
spin_unlock_irqrestore(&bufferlock, flags);
wake_up_interruptible(&dasd_eer_read_wait_queue);
}
/*
* This function writes a DASD_EER_STATECHANGE trigger.
*/
static void dasd_eer_write_snss_trigger(struct dasd_device *device,
struct dasd_ccw_req *cqr,
int trigger)
{
int data_size;
int snss_rc;
struct timeval tv;
struct dasd_eer_header header;
unsigned long flags;
struct eerbuffer *eerb;
[S390] dasd: add hyper PAV support to DASD device driver, part 1 Parallel access volumes (PAV) is a storage server feature, that allows to start multiple channel programs on the same DASD in parallel. It defines alias devices which can be used as alternative paths to the same disk. With the old base PAV support we only needed rudimentary functionality in the DASD device driver. As the mapping between base and alias devices was static, we just had to export an identifier (uid) and could leave the combining of devices to external layers like a device mapper multipath. Now hyper PAV removes the requirement to dedicate alias devices to specific base devices. Instead each alias devices can be combined with multiple base device on a per request basis. This requires full support by the DASD device driver as now each channel program itself has to identify the target base device. The changes to the dasd device driver and the ECKD discipline are: - Separate subchannel device representation (dasd_device) from block device representation (dasd_block). Only base devices are block devices. - Gather information about base and alias devices and possible combinations. - For each request decide which dasd_device should be used (base or alias) and build specific channel program. - Support summary unit checks, which allow the storage server to upgrade / downgrade between base and hyper PAV at runtime (support is mandatory). Signed-off-by: Stefan Weinhuber <wein@de.ibm.com> Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
2008-01-26 13:11:23 +00:00
snss_rc = (cqr->status == DASD_CQR_DONE) ? 0 : -EIO;
if (snss_rc)
data_size = 0;
else
data_size = SNSS_DATA_SIZE;
header.total_size = sizeof(header) + data_size + 4; /* "EOR" */
header.trigger = DASD_EER_STATECHANGE;
do_gettimeofday(&tv);
header.tv_sec = tv.tv_sec;
header.tv_usec = tv.tv_usec;
strncpy(header.busid, dev_name(&device->cdev->dev),
DASD_EER_BUSID_SIZE);
spin_lock_irqsave(&bufferlock, flags);
list_for_each_entry(eerb, &bufferlist, list) {
dasd_eer_start_record(eerb, header.total_size);
dasd_eer_write_buffer(eerb, (char *) &header , sizeof(header));
if (!snss_rc)
dasd_eer_write_buffer(eerb, cqr->data, SNSS_DATA_SIZE);
dasd_eer_write_buffer(eerb, "EOR", 4);
}
spin_unlock_irqrestore(&bufferlock, flags);
wake_up_interruptible(&dasd_eer_read_wait_queue);
}
/*
* This function is called for all triggers. It calls the appropriate
* function that writes the actual trigger records.
*/
void dasd_eer_write(struct dasd_device *device, struct dasd_ccw_req *cqr,
unsigned int id)
{
if (!device->eer_cqr)
return;
switch (id) {
case DASD_EER_FATALERROR:
case DASD_EER_PPRCSUSPEND:
dasd_eer_write_standard_trigger(device, cqr, id);
break;
case DASD_EER_NOPATH:
dasd_eer_write_standard_trigger(device, NULL, id);
break;
case DASD_EER_STATECHANGE:
dasd_eer_write_snss_trigger(device, cqr, id);
break;
default: /* unknown trigger, so we write it without any sense data */
dasd_eer_write_standard_trigger(device, NULL, id);
break;
}
}
EXPORT_SYMBOL(dasd_eer_write);
/*
* Start a sense subsystem status request.
* Needs to be called with the device held.
*/
void dasd_eer_snss(struct dasd_device *device)
{
struct dasd_ccw_req *cqr;
cqr = device->eer_cqr;
if (!cqr) /* Device not eer enabled. */
return;
if (test_and_set_bit(DASD_FLAG_EER_IN_USE, &device->flags)) {
/* Sense subsystem status request in use. */
set_bit(DASD_FLAG_EER_SNSS, &device->flags);
return;
}
[S390] dasd: add hyper PAV support to DASD device driver, part 1 Parallel access volumes (PAV) is a storage server feature, that allows to start multiple channel programs on the same DASD in parallel. It defines alias devices which can be used as alternative paths to the same disk. With the old base PAV support we only needed rudimentary functionality in the DASD device driver. As the mapping between base and alias devices was static, we just had to export an identifier (uid) and could leave the combining of devices to external layers like a device mapper multipath. Now hyper PAV removes the requirement to dedicate alias devices to specific base devices. Instead each alias devices can be combined with multiple base device on a per request basis. This requires full support by the DASD device driver as now each channel program itself has to identify the target base device. The changes to the dasd device driver and the ECKD discipline are: - Separate subchannel device representation (dasd_device) from block device representation (dasd_block). Only base devices are block devices. - Gather information about base and alias devices and possible combinations. - For each request decide which dasd_device should be used (base or alias) and build specific channel program. - Support summary unit checks, which allow the storage server to upgrade / downgrade between base and hyper PAV at runtime (support is mandatory). Signed-off-by: Stefan Weinhuber <wein@de.ibm.com> Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
2008-01-26 13:11:23 +00:00
/* cdev is already locked, can't use dasd_add_request_head */
clear_bit(DASD_FLAG_EER_SNSS, &device->flags);
cqr->status = DASD_CQR_QUEUED;
[S390] dasd: add hyper PAV support to DASD device driver, part 1 Parallel access volumes (PAV) is a storage server feature, that allows to start multiple channel programs on the same DASD in parallel. It defines alias devices which can be used as alternative paths to the same disk. With the old base PAV support we only needed rudimentary functionality in the DASD device driver. As the mapping between base and alias devices was static, we just had to export an identifier (uid) and could leave the combining of devices to external layers like a device mapper multipath. Now hyper PAV removes the requirement to dedicate alias devices to specific base devices. Instead each alias devices can be combined with multiple base device on a per request basis. This requires full support by the DASD device driver as now each channel program itself has to identify the target base device. The changes to the dasd device driver and the ECKD discipline are: - Separate subchannel device representation (dasd_device) from block device representation (dasd_block). Only base devices are block devices. - Gather information about base and alias devices and possible combinations. - For each request decide which dasd_device should be used (base or alias) and build specific channel program. - Support summary unit checks, which allow the storage server to upgrade / downgrade between base and hyper PAV at runtime (support is mandatory). Signed-off-by: Stefan Weinhuber <wein@de.ibm.com> Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
2008-01-26 13:11:23 +00:00
list_add(&cqr->devlist, &device->ccw_queue);
dasd_schedule_device_bh(device);
}
/*
* Callback function for use with sense subsystem status request.
*/
static void dasd_eer_snss_cb(struct dasd_ccw_req *cqr, void *data)
{
[S390] dasd: add hyper PAV support to DASD device driver, part 1 Parallel access volumes (PAV) is a storage server feature, that allows to start multiple channel programs on the same DASD in parallel. It defines alias devices which can be used as alternative paths to the same disk. With the old base PAV support we only needed rudimentary functionality in the DASD device driver. As the mapping between base and alias devices was static, we just had to export an identifier (uid) and could leave the combining of devices to external layers like a device mapper multipath. Now hyper PAV removes the requirement to dedicate alias devices to specific base devices. Instead each alias devices can be combined with multiple base device on a per request basis. This requires full support by the DASD device driver as now each channel program itself has to identify the target base device. The changes to the dasd device driver and the ECKD discipline are: - Separate subchannel device representation (dasd_device) from block device representation (dasd_block). Only base devices are block devices. - Gather information about base and alias devices and possible combinations. - For each request decide which dasd_device should be used (base or alias) and build specific channel program. - Support summary unit checks, which allow the storage server to upgrade / downgrade between base and hyper PAV at runtime (support is mandatory). Signed-off-by: Stefan Weinhuber <wein@de.ibm.com> Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
2008-01-26 13:11:23 +00:00
struct dasd_device *device = cqr->startdev;
unsigned long flags;
dasd_eer_write(device, cqr, DASD_EER_STATECHANGE);
spin_lock_irqsave(get_ccwdev_lock(device->cdev), flags);
if (device->eer_cqr == cqr) {
clear_bit(DASD_FLAG_EER_IN_USE, &device->flags);
if (test_bit(DASD_FLAG_EER_SNSS, &device->flags))
/* Another SNSS has been requested in the meantime. */
dasd_eer_snss(device);
cqr = NULL;
}
spin_unlock_irqrestore(get_ccwdev_lock(device->cdev), flags);
if (cqr)
/*
* Extended error recovery has been switched off while
* the SNSS request was running. It could even have
* been switched off and on again in which case there
* is a new ccw in device->eer_cqr. Free the "old"
* snss request now.
*/
dasd_kfree_request(cqr, device);
}
/*
* Enable error reporting on a given device.
*/
int dasd_eer_enable(struct dasd_device *device)
{
struct dasd_ccw_req *cqr;
unsigned long flags;
struct ccw1 *ccw;
if (device->eer_cqr)
return 0;
if (!device->discipline || strcmp(device->discipline->name, "ECKD"))
return -EPERM; /* FIXME: -EMEDIUMTYPE ? */
cqr = dasd_kmalloc_request(DASD_ECKD_MAGIC, 1 /* SNSS */,
SNSS_DATA_SIZE, device);
if (IS_ERR(cqr))
return -ENOMEM;
[S390] dasd: add hyper PAV support to DASD device driver, part 1 Parallel access volumes (PAV) is a storage server feature, that allows to start multiple channel programs on the same DASD in parallel. It defines alias devices which can be used as alternative paths to the same disk. With the old base PAV support we only needed rudimentary functionality in the DASD device driver. As the mapping between base and alias devices was static, we just had to export an identifier (uid) and could leave the combining of devices to external layers like a device mapper multipath. Now hyper PAV removes the requirement to dedicate alias devices to specific base devices. Instead each alias devices can be combined with multiple base device on a per request basis. This requires full support by the DASD device driver as now each channel program itself has to identify the target base device. The changes to the dasd device driver and the ECKD discipline are: - Separate subchannel device representation (dasd_device) from block device representation (dasd_block). Only base devices are block devices. - Gather information about base and alias devices and possible combinations. - For each request decide which dasd_device should be used (base or alias) and build specific channel program. - Support summary unit checks, which allow the storage server to upgrade / downgrade between base and hyper PAV at runtime (support is mandatory). Signed-off-by: Stefan Weinhuber <wein@de.ibm.com> Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
2008-01-26 13:11:23 +00:00
cqr->startdev = device;
cqr->retries = 255;
cqr->expires = 10 * HZ;
clear_bit(DASD_CQR_FLAGS_USE_ERP, &cqr->flags);
set_bit(DASD_CQR_ALLOW_SLOCK, &cqr->flags);
ccw = cqr->cpaddr;
ccw->cmd_code = DASD_ECKD_CCW_SNSS;
ccw->count = SNSS_DATA_SIZE;
ccw->flags = 0;
ccw->cda = (__u32)(addr_t) cqr->data;
cqr->buildclk = get_clock();
cqr->status = DASD_CQR_FILLED;
cqr->callback = dasd_eer_snss_cb;
spin_lock_irqsave(get_ccwdev_lock(device->cdev), flags);
if (!device->eer_cqr) {
device->eer_cqr = cqr;
cqr = NULL;
}
spin_unlock_irqrestore(get_ccwdev_lock(device->cdev), flags);
if (cqr)
dasd_kfree_request(cqr, device);
return 0;
}
/*
* Disable error reporting on a given device.
*/
void dasd_eer_disable(struct dasd_device *device)
{
struct dasd_ccw_req *cqr;
unsigned long flags;
int in_use;
if (!device->eer_cqr)
return;
spin_lock_irqsave(get_ccwdev_lock(device->cdev), flags);
cqr = device->eer_cqr;
device->eer_cqr = NULL;
clear_bit(DASD_FLAG_EER_SNSS, &device->flags);
in_use = test_and_clear_bit(DASD_FLAG_EER_IN_USE, &device->flags);
spin_unlock_irqrestore(get_ccwdev_lock(device->cdev), flags);
if (cqr && !in_use)
dasd_kfree_request(cqr, device);
}
/*
* SECTION: the device operations
*/
/*
* On the one side we need a lock to access our internal buffer, on the
* other side a copy_to_user can sleep. So we need to copy the data we have
* to transfer in a readbuffer, which is protected by the readbuffer_mutex.
*/
static char readbuffer[PAGE_SIZE];
static DEFINE_MUTEX(readbuffer_mutex);
static int dasd_eer_open(struct inode *inp, struct file *filp)
{
struct eerbuffer *eerb;
unsigned long flags;
eerb = kzalloc(sizeof(struct eerbuffer), GFP_KERNEL);
if (!eerb)
return -ENOMEM;
eerb->buffer_page_count = eer_pages;
if (eerb->buffer_page_count < 1 ||
eerb->buffer_page_count > INT_MAX / PAGE_SIZE) {
kfree(eerb);
DBF_EVENT(DBF_WARNING, "can't open device since module "
"parameter eer_pages is smaller than 1 or"
" bigger than %d", (int)(INT_MAX / PAGE_SIZE));
return -EINVAL;
}
eerb->buffersize = eerb->buffer_page_count * PAGE_SIZE;
eerb->buffer = kmalloc(eerb->buffer_page_count * sizeof(char *),
GFP_KERNEL);
if (!eerb->buffer) {
kfree(eerb);
return -ENOMEM;
}
if (dasd_eer_allocate_buffer_pages(eerb->buffer,
eerb->buffer_page_count)) {
kfree(eerb->buffer);
kfree(eerb);
return -ENOMEM;
}
filp->private_data = eerb;
spin_lock_irqsave(&bufferlock, flags);
list_add(&eerb->list, &bufferlist);
spin_unlock_irqrestore(&bufferlock, flags);
return nonseekable_open(inp,filp);
}
static int dasd_eer_close(struct inode *inp, struct file *filp)
{
struct eerbuffer *eerb;
unsigned long flags;
eerb = (struct eerbuffer *) filp->private_data;
spin_lock_irqsave(&bufferlock, flags);
list_del(&eerb->list);
spin_unlock_irqrestore(&bufferlock, flags);
dasd_eer_free_buffer_pages(eerb->buffer, eerb->buffer_page_count);
kfree(eerb->buffer);
kfree(eerb);
return 0;
}
static ssize_t dasd_eer_read(struct file *filp, char __user *buf,
size_t count, loff_t *ppos)
{
int tc,rc;
int tailcount,effective_count;
unsigned long flags;
struct eerbuffer *eerb;
eerb = (struct eerbuffer *) filp->private_data;
if (mutex_lock_interruptible(&readbuffer_mutex))
return -ERESTARTSYS;
spin_lock_irqsave(&bufferlock, flags);
if (eerb->residual < 0) { /* the remainder of this record */
/* has been deleted */
eerb->residual = 0;
spin_unlock_irqrestore(&bufferlock, flags);
mutex_unlock(&readbuffer_mutex);
return -EIO;
} else if (eerb->residual > 0) {
/* OK we still have a second half of a record to deliver */
effective_count = min(eerb->residual, (int) count);
eerb->residual -= effective_count;
} else {
tc = 0;
while (!tc) {
tc = dasd_eer_read_buffer(eerb, (char *) &tailcount,
sizeof(tailcount));
if (!tc) {
/* no data available */
spin_unlock_irqrestore(&bufferlock, flags);
mutex_unlock(&readbuffer_mutex);
if (filp->f_flags & O_NONBLOCK)
return -EAGAIN;
rc = wait_event_interruptible(
dasd_eer_read_wait_queue,
eerb->head != eerb->tail);
if (rc)
return rc;
if (mutex_lock_interruptible(&readbuffer_mutex))
return -ERESTARTSYS;
spin_lock_irqsave(&bufferlock, flags);
}
}
WARN_ON(tc != sizeof(tailcount));
effective_count = min(tailcount,(int)count);
eerb->residual = tailcount - effective_count;
}
tc = dasd_eer_read_buffer(eerb, readbuffer, effective_count);
WARN_ON(tc != effective_count);
spin_unlock_irqrestore(&bufferlock, flags);
if (copy_to_user(buf, readbuffer, effective_count)) {
mutex_unlock(&readbuffer_mutex);
return -EFAULT;
}
mutex_unlock(&readbuffer_mutex);
return effective_count;
}
static unsigned int dasd_eer_poll(struct file *filp, poll_table *ptable)
{
unsigned int mask;
unsigned long flags;
struct eerbuffer *eerb;
eerb = (struct eerbuffer *) filp->private_data;
poll_wait(filp, &dasd_eer_read_wait_queue, ptable);
spin_lock_irqsave(&bufferlock, flags);
if (eerb->head != eerb->tail)
mask = POLLIN | POLLRDNORM ;
else
mask = 0;
spin_unlock_irqrestore(&bufferlock, flags);
return mask;
}
static const struct file_operations dasd_eer_fops = {
.open = &dasd_eer_open,
.release = &dasd_eer_close,
.read = &dasd_eer_read,
.poll = &dasd_eer_poll,
.owner = THIS_MODULE,
llseek: automatically add .llseek fop All file_operations should get a .llseek operation so we can make nonseekable_open the default for future file operations without a .llseek pointer. The three cases that we can automatically detect are no_llseek, seq_lseek and default_llseek. For cases where we can we can automatically prove that the file offset is always ignored, we use noop_llseek, which maintains the current behavior of not returning an error from a seek. New drivers should normally not use noop_llseek but instead use no_llseek and call nonseekable_open at open time. Existing drivers can be converted to do the same when the maintainer knows for certain that no user code relies on calling seek on the device file. The generated code is often incorrectly indented and right now contains comments that clarify for each added line why a specific variant was chosen. In the version that gets submitted upstream, the comments will be gone and I will manually fix the indentation, because there does not seem to be a way to do that using coccinelle. Some amount of new code is currently sitting in linux-next that should get the same modifications, which I will do at the end of the merge window. Many thanks to Julia Lawall for helping me learn to write a semantic patch that does all this. ===== begin semantic patch ===== // This adds an llseek= method to all file operations, // as a preparation for making no_llseek the default. // // The rules are // - use no_llseek explicitly if we do nonseekable_open // - use seq_lseek for sequential files // - use default_llseek if we know we access f_pos // - use noop_llseek if we know we don't access f_pos, // but we still want to allow users to call lseek // @ open1 exists @ identifier nested_open; @@ nested_open(...) { <+... nonseekable_open(...) ...+> } @ open exists@ identifier open_f; identifier i, f; identifier open1.nested_open; @@ int open_f(struct inode *i, struct file *f) { <+... ( nonseekable_open(...) | nested_open(...) ) ...+> } @ read disable optional_qualifier exists @ identifier read_f; identifier f, p, s, off; type ssize_t, size_t, loff_t; expression E; identifier func; @@ ssize_t read_f(struct file *f, char *p, size_t s, loff_t *off) { <+... ( *off = E | *off += E | func(..., off, ...) | E = *off ) ...+> } @ read_no_fpos disable optional_qualifier exists @ identifier read_f; identifier f, p, s, off; type ssize_t, size_t, loff_t; @@ ssize_t read_f(struct file *f, char *p, size_t s, loff_t *off) { ... when != off } @ write @ identifier write_f; identifier f, p, s, off; type ssize_t, size_t, loff_t; expression E; identifier func; @@ ssize_t write_f(struct file *f, const char *p, size_t s, loff_t *off) { <+... ( *off = E | *off += E | func(..., off, ...) | E = *off ) ...+> } @ write_no_fpos @ identifier write_f; identifier f, p, s, off; type ssize_t, size_t, loff_t; @@ ssize_t write_f(struct file *f, const char *p, size_t s, loff_t *off) { ... when != off } @ fops0 @ identifier fops; @@ struct file_operations fops = { ... }; @ has_llseek depends on fops0 @ identifier fops0.fops; identifier llseek_f; @@ struct file_operations fops = { ... .llseek = llseek_f, ... }; @ has_read depends on fops0 @ identifier fops0.fops; identifier read_f; @@ struct file_operations fops = { ... .read = read_f, ... }; @ has_write depends on fops0 @ identifier fops0.fops; identifier write_f; @@ struct file_operations fops = { ... .write = write_f, ... }; @ has_open depends on fops0 @ identifier fops0.fops; identifier open_f; @@ struct file_operations fops = { ... .open = open_f, ... }; // use no_llseek if we call nonseekable_open //////////////////////////////////////////// @ nonseekable1 depends on !has_llseek && has_open @ identifier fops0.fops; identifier nso ~= "nonseekable_open"; @@ struct file_operations fops = { ... .open = nso, ... +.llseek = no_llseek, /* nonseekable */ }; @ nonseekable2 depends on !has_llseek @ identifier fops0.fops; identifier open.open_f; @@ struct file_operations fops = { ... .open = open_f, ... +.llseek = no_llseek, /* open uses nonseekable */ }; // use seq_lseek for sequential files ///////////////////////////////////// @ seq depends on !has_llseek @ identifier fops0.fops; identifier sr ~= "seq_read"; @@ struct file_operations fops = { ... .read = sr, ... +.llseek = seq_lseek, /* we have seq_read */ }; // use default_llseek if there is a readdir /////////////////////////////////////////// @ fops1 depends on !has_llseek && !nonseekable1 && !nonseekable2 && !seq @ identifier fops0.fops; identifier readdir_e; @@ // any other fop is used that changes pos struct file_operations fops = { ... .readdir = readdir_e, ... +.llseek = default_llseek, /* readdir is present */ }; // use default_llseek if at least one of read/write touches f_pos ///////////////////////////////////////////////////////////////// @ fops2 depends on !fops1 && !has_llseek && !nonseekable1 && !nonseekable2 && !seq @ identifier fops0.fops; identifier read.read_f; @@ // read fops use offset struct file_operations fops = { ... .read = read_f, ... +.llseek = default_llseek, /* read accesses f_pos */ }; @ fops3 depends on !fops1 && !fops2 && !has_llseek && !nonseekable1 && !nonseekable2 && !seq @ identifier fops0.fops; identifier write.write_f; @@ // write fops use offset struct file_operations fops = { ... .write = write_f, ... + .llseek = default_llseek, /* write accesses f_pos */ }; // Use noop_llseek if neither read nor write accesses f_pos /////////////////////////////////////////////////////////// @ fops4 depends on !fops1 && !fops2 && !fops3 && !has_llseek && !nonseekable1 && !nonseekable2 && !seq @ identifier fops0.fops; identifier read_no_fpos.read_f; identifier write_no_fpos.write_f; @@ // write fops use offset struct file_operations fops = { ... .write = write_f, .read = read_f, ... +.llseek = noop_llseek, /* read and write both use no f_pos */ }; @ depends on has_write && !has_read && !fops1 && !fops2 && !has_llseek && !nonseekable1 && !nonseekable2 && !seq @ identifier fops0.fops; identifier write_no_fpos.write_f; @@ struct file_operations fops = { ... .write = write_f, ... +.llseek = noop_llseek, /* write uses no f_pos */ }; @ depends on has_read && !has_write && !fops1 && !fops2 && !has_llseek && !nonseekable1 && !nonseekable2 && !seq @ identifier fops0.fops; identifier read_no_fpos.read_f; @@ struct file_operations fops = { ... .read = read_f, ... +.llseek = noop_llseek, /* read uses no f_pos */ }; @ depends on !has_read && !has_write && !fops1 && !fops2 && !has_llseek && !nonseekable1 && !nonseekable2 && !seq @ identifier fops0.fops; @@ struct file_operations fops = { ... +.llseek = noop_llseek, /* no read or write fn */ }; ===== End semantic patch ===== Signed-off-by: Arnd Bergmann <arnd@arndb.de> Cc: Julia Lawall <julia@diku.dk> Cc: Christoph Hellwig <hch@infradead.org>
2010-08-15 16:52:59 +00:00
.llseek = noop_llseek,
};
static struct miscdevice *dasd_eer_dev = NULL;
int __init dasd_eer_init(void)
{
int rc;
dasd_eer_dev = kzalloc(sizeof(*dasd_eer_dev), GFP_KERNEL);
if (!dasd_eer_dev)
return -ENOMEM;
dasd_eer_dev->minor = MISC_DYNAMIC_MINOR;
dasd_eer_dev->name = "dasd_eer";
dasd_eer_dev->fops = &dasd_eer_fops;
rc = misc_register(dasd_eer_dev);
if (rc) {
kfree(dasd_eer_dev);
dasd_eer_dev = NULL;
DBF_EVENT(DBF_ERR, "%s", "dasd_eer_init could not "
"register misc device");
return rc;
}
return 0;
}
void dasd_eer_exit(void)
{
if (dasd_eer_dev) {
misc_deregister(dasd_eer_dev);
kfree(dasd_eer_dev);
dasd_eer_dev = NULL;
}
}