Commit 67a670bd authored by Linus Torvalds's avatar Linus Torvalds

Merge http://linux-isdn.bkbits.net/linux-2.5.make-next

into home.transmeta.com:/home/torvalds/v2.5/linux
parents 8876c643 66ebd50b
......@@ -436,14 +436,15 @@ static int alloc_array_sb(mddev_t * mddev)
static int alloc_disk_sb(mdk_rdev_t * rdev)
{
if (rdev->sb)
if (rdev->sb_page)
MD_BUG();
rdev->sb = (mdp_super_t *) __get_free_page(GFP_KERNEL);
if (!rdev->sb) {
rdev->sb_page = alloc_page(GFP_KERNEL);
if (!rdev->sb_page) {
printk(OUT_OF_MEM);
return -EINVAL;
}
rdev->sb = (mdp_super_t *) page_address(rdev->sb_page);
clear_page(rdev->sb);
return 0;
......@@ -451,9 +452,10 @@ static int alloc_disk_sb(mdk_rdev_t * rdev)
static void free_disk_sb(mdk_rdev_t * rdev)
{
if (rdev->sb) {
free_page((unsigned long) rdev->sb);
if (rdev->sb_page) {
page_cache_release(rdev->sb_page);
rdev->sb = NULL;
rdev->sb_page = NULL;
rdev->sb_offset = 0;
rdev->size = 0;
} else {
......@@ -462,13 +464,42 @@ static void free_disk_sb(mdk_rdev_t * rdev)
}
}
static void bi_complete(struct bio *bio)
{
complete((struct completion*)bio->bi_private);
}
static int sync_page_io(struct block_device *bdev, sector_t sector, int size,
struct page *page, int rw)
{
struct bio bio;
struct bio_vec vec;
struct completion event;
bio_init(&bio);
bio.bi_io_vec = &vec;
vec.bv_page = page;
vec.bv_len = size;
vec.bv_offset = 0;
bio.bi_vcnt = 1;
bio.bi_idx = 0;
bio.bi_size = size;
bio.bi_bdev = bdev;
bio.bi_sector = sector;
init_completion(&event);
bio.bi_private = &event;
bio.bi_end_io = bi_complete;
submit_bio(rw, &bio);
run_task_queue(&tq_disk);
wait_for_completion(&event);
return test_bit(BIO_UPTODATE, &bio.bi_flags);
}
static int read_disk_sb(mdk_rdev_t * rdev)
{
struct address_space *mapping = rdev->bdev->bd_inode->i_mapping;
struct page *page;
char *p;
unsigned long sb_offset;
int n = PAGE_CACHE_SIZE / BLOCK_SIZE;
if (!rdev->sb) {
MD_BUG();
......@@ -483,24 +514,14 @@ static int read_disk_sb(mdk_rdev_t * rdev)
*/
sb_offset = calc_dev_sboffset(rdev->dev, rdev->mddev, 1);
rdev->sb_offset = sb_offset;
page = read_cache_page(mapping, sb_offset/n,
(filler_t *)mapping->a_ops->readpage, NULL);
if (IS_ERR(page))
goto out;
wait_on_page_locked(page);
if (!PageUptodate(page))
goto fail;
if (PageError(page))
if (!sync_page_io(rdev->bdev, sb_offset<<1, MD_SB_BYTES, rdev->sb_page, READ))
goto fail;
p = (char *)page_address(page) + BLOCK_SIZE * (sb_offset % n);
memcpy((char*)rdev->sb, p, MD_SB_BYTES);
page_cache_release(page);
printk(KERN_INFO " [events: %08lx]\n", (unsigned long)rdev->sb->events_lo);
return 0;
fail:
page_cache_release(page);
out:
printk(NO_SB,partition_name(rdev->dev));
return -EINVAL;
}
......@@ -893,11 +914,6 @@ static mdk_rdev_t * find_rdev_all(kdev_t dev)
static int write_disk_sb(mdk_rdev_t * rdev)
{
struct block_device *bdev = rdev->bdev;
struct address_space *mapping = bdev->bd_inode->i_mapping;
struct page *page;
unsigned offs;
int error;
kdev_t dev = rdev->dev;
unsigned long sb_offset, size;
......@@ -933,29 +949,11 @@ static int write_disk_sb(mdk_rdev_t * rdev)
}
printk(KERN_INFO "(write) %s's sb offset: %ld\n", partition_name(dev), sb_offset);
fsync_bdev(bdev);
page = grab_cache_page(mapping, sb_offset/(PAGE_CACHE_SIZE/BLOCK_SIZE));
offs = sb_offset % (PAGE_CACHE_SIZE/BLOCK_SIZE);
if (!page)
if (!sync_page_io(rdev->bdev, sb_offset<<1, MD_SB_BYTES, rdev->sb_page, WRITE))
goto fail;
error = mapping->a_ops->prepare_write(NULL, page, offs,
offs + MD_SB_BYTES);
if (error)
goto unlock;
memcpy((char *)page_address(page) + offs, rdev->sb, MD_SB_BYTES);
error = mapping->a_ops->commit_write(NULL, page, offs,
offs + MD_SB_BYTES);
if (error)
goto unlock;
unlock_page(page);
wait_on_page_locked(page);
page_cache_release(page);
fsync_bdev(bdev);
skip:
return 0;
unlock:
unlock_page(page);
page_cache_release(page);
fail:
printk("md: write_disk_sb failed for device %s\n", partition_name(dev));
return 1;
......
......@@ -24,19 +24,19 @@
#include <asm/bitops.h>
#include <asm/atomic.h>
static mdk_personality_t raid5_personality;
/*
* Stripe cache
*/
#define NR_STRIPES 256
#define STRIPE_SIZE PAGE_SIZE
#define STRIPE_SECTORS (STRIPE_SIZE>>9)
#define IO_THRESHOLD 1
#define HASH_PAGES 1
#define HASH_PAGES_ORDER 0
#define NR_HASH (HASH_PAGES * PAGE_SIZE / sizeof(struct stripe_head *))
#define HASH_MASK (NR_HASH - 1)
#define stripe_hash(conf, sect) ((conf)->stripe_hashtbl[((sect) / ((conf)->buffer_size >> 9)) & HASH_MASK])
#define stripe_hash(conf, sect) ((conf)->stripe_hashtbl[((sect) / STRIPE_SECTORS) & HASH_MASK])
/*
* The following can be used to debug the driver
......@@ -142,47 +142,36 @@ static struct stripe_head *get_free_stripe(raid5_conf_t *conf)
static void shrink_buffers(struct stripe_head *sh, int num)
{
struct buffer_head *bh;
struct page *p;
int i;
for (i=0; i<num ; i++) {
bh = sh->bh_cache[i];
if (!bh)
return;
sh->bh_cache[i] = NULL;
free_page((unsigned long) bh->b_data);
kfree(bh);
p = sh->dev[i].page;
if (!p)
continue;
sh->dev[i].page = NULL;
page_cache_release(p);
}
}
static int grow_buffers(struct stripe_head *sh, int num, int b_size, int priority)
static int grow_buffers(struct stripe_head *sh, int num)
{
struct buffer_head *bh;
int i;
for (i=0; i<num; i++) {
struct page *page;
bh = kmalloc(sizeof(struct buffer_head), priority);
if (!bh)
return 1;
memset(bh, 0, sizeof (struct buffer_head));
if ((page = alloc_page(priority)))
bh->b_data = page_address(page);
else {
kfree(bh);
if (!(page = alloc_page(GFP_KERNEL))) {
return 1;
}
atomic_set(&bh->b_count, 0);
bh->b_page = page;
sh->bh_cache[i] = bh;
sh->dev[i].page = page;
}
return 0;
}
static struct buffer_head *raid5_build_block (struct stripe_head *sh, int i);
static void raid5_build_block (struct stripe_head *sh, int i);
static inline void init_stripe(struct stripe_head *sh, unsigned long sector)
static inline void init_stripe(struct stripe_head *sh, unsigned long sector, int pd_idx)
{
raid5_conf_t *conf = sh->raid_conf;
int disks = conf->raid_disks, i;
......@@ -198,40 +187,26 @@ static inline void init_stripe(struct stripe_head *sh, unsigned long sector)
remove_hash(sh);
sh->sector = sector;
sh->size = conf->buffer_size;
sh->pd_idx = pd_idx;
sh->state = 0;
for (i=disks; i--; ) {
if (sh->bh_read[i] || sh->bh_write[i] || sh->bh_written[i] ||
buffer_locked(sh->bh_cache[i])) {
struct r5dev *dev = &sh->dev[i];
if (dev->toread || dev->towrite || dev->written ||
test_bit(R5_LOCKED, &dev->flags)) {
printk("sector=%lx i=%d %p %p %p %d\n",
sh->sector, i, sh->bh_read[i],
sh->bh_write[i], sh->bh_written[i],
buffer_locked(sh->bh_cache[i]));
sh->sector, i, dev->toread,
dev->towrite, dev->written,
test_bit(R5_LOCKED, &dev->flags));
BUG();
}
clear_buffer_uptodate(sh->bh_cache[i]);
dev->flags = 0;
raid5_build_block(sh, i);
}
insert_hash(conf, sh);
}
/* the buffer size has changed, so unhash all stripes
* as active stripes complete, they will go onto inactive list
*/
static void shrink_stripe_cache(raid5_conf_t *conf)
{
int i;
CHECK_DEVLOCK();
if (atomic_read(&conf->active_stripes))
BUG();
for (i=0; i < NR_HASH; i++) {
struct stripe_head *sh;
while ((sh = conf->stripe_hashtbl[i]))
remove_hash(sh);
}
}
static struct stripe_head *__find_stripe(raid5_conf_t *conf, unsigned long sector)
{
struct stripe_head *sh;
......@@ -245,53 +220,16 @@ static struct stripe_head *__find_stripe(raid5_conf_t *conf, unsigned long secto
return NULL;
}
static struct stripe_head *get_active_stripe(raid5_conf_t *conf, unsigned long sector, int size, int noblock)
static struct stripe_head *get_active_stripe(raid5_conf_t *conf, unsigned long sector,
int pd_idx, int noblock)
{
struct stripe_head *sh;
PRINTK("get_stripe, sector %lu\n", sector);
md_spin_lock_irq(&conf->device_lock);
spin_lock_irq(&conf->device_lock);
do {
if (conf->buffer_size == 0 ||
(size && size != conf->buffer_size)) {
/* either the size is being changed (buffer_size==0) or
* we need to change it.
* If size==0, we can proceed as soon as buffer_size gets set.
* If size>0, we can proceed when active_stripes reaches 0, or
* when someone else sets the buffer_size to size.
* If someone sets the buffer size to something else, we will need to
* assert that we want to change it again
*/
int oldsize = conf->buffer_size;
PRINTK("get_stripe %ld/%d buffer_size is %d, %d active\n", sector, size, conf->buffer_size, atomic_read(&conf->active_stripes));
if (size==0)
wait_event_lock_irq(conf->wait_for_stripe,
conf->buffer_size,
conf->device_lock);
else {
while (conf->buffer_size != size && atomic_read(&conf->active_stripes)) {
conf->buffer_size = 0;
wait_event_lock_irq(conf->wait_for_stripe,
atomic_read(&conf->active_stripes)==0 || conf->buffer_size,
conf->device_lock);
PRINTK("waited and now %ld/%d buffer_size is %d - %d active\n", sector, size,
conf->buffer_size, atomic_read(&conf->active_stripes));
}
if (conf->buffer_size != size) {
printk("raid5: switching cache buffer size, %d --> %d\n", oldsize, size);
shrink_stripe_cache(conf);
if (size==0) BUG();
conf->buffer_size = size;
PRINTK("size now %d\n", conf->buffer_size);
}
}
}
if (size == 0)
sector -= sector & ((conf->buffer_size>>9)-1);
sh = __find_stripe(conf, sector);
if (!sh) {
if (!conf->inactive_blocked)
......@@ -307,7 +245,7 @@ static struct stripe_head *get_active_stripe(raid5_conf_t *conf, unsigned long s
conf->device_lock);
conf->inactive_blocked = 0;
} else
init_stripe(sh, sector);
init_stripe(sh, sector, pd_idx);
} else {
if (atomic_read(&sh->count)) {
if (!list_empty(&sh->lru))
......@@ -325,25 +263,35 @@ static struct stripe_head *get_active_stripe(raid5_conf_t *conf, unsigned long s
if (sh)
atomic_inc(&sh->count);
md_spin_unlock_irq(&conf->device_lock);
spin_unlock_irq(&conf->device_lock);
return sh;
}
static int grow_stripes(raid5_conf_t *conf, int num, int priority)
static int grow_stripes(raid5_conf_t *conf, int num)
{
struct stripe_head *sh;
kmem_cache_t *sc;
int devs = conf->raid_disks;
sprintf(conf->cache_name, "md/raid5-%d", conf->mddev->__minor);
sc = kmem_cache_create(conf->cache_name,
sizeof(struct stripe_head)+(devs-1)*sizeof(struct r5dev),
0, 0, NULL, NULL);
if (!sc)
return 1;
conf->slab_cache = sc;
while (num--) {
sh = kmalloc(sizeof(struct stripe_head), priority);
sh = kmem_cache_alloc(sc, GFP_KERNEL);
if (!sh)
return 1;
memset(sh, 0, sizeof(*sh));
memset(sh, 0, sizeof(*sh) + (devs-1)*sizeof(struct r5dev));
sh->raid_conf = conf;
sh->lock = SPIN_LOCK_UNLOCKED;
if (grow_buffers(sh, conf->raid_disks, PAGE_SIZE, priority)) {
if (grow_buffers(sh, conf->raid_disks)) {
shrink_buffers(sh, conf->raid_disks);
kfree(sh);
kmem_cache_free(sc, sh);
return 1;
}
/* we just created an active stripe so... */
......@@ -355,11 +303,11 @@ static int grow_stripes(raid5_conf_t *conf, int num, int priority)
return 0;
}
static void shrink_stripes(raid5_conf_t *conf, int num)
static void shrink_stripes(raid5_conf_t *conf)
{
struct stripe_head *sh;
while (num--) {
while (1) {
spin_lock_irq(&conf->device_lock);
sh = get_free_stripe(conf);
spin_unlock_irq(&conf->device_lock);
......@@ -368,21 +316,22 @@ static void shrink_stripes(raid5_conf_t *conf, int num)
if (atomic_read(&sh->count))
BUG();
shrink_buffers(sh, conf->raid_disks);
kfree(sh);
kmem_cache_free(conf->slab_cache, sh);
atomic_dec(&conf->active_stripes);
}
kmem_cache_destroy(conf->slab_cache);
conf->slab_cache = NULL;
}
static void raid5_end_read_request (struct buffer_head * bh, int uptodate)
static void raid5_end_read_request (struct bio * bi)
{
struct stripe_head *sh = bh->b_private;
struct stripe_head *sh = bi->bi_private;
raid5_conf_t *conf = sh->raid_conf;
int disks = conf->raid_disks, i;
unsigned long flags;
int uptodate = test_bit(BIO_UPTODATE, &bi->bi_flags);
for (i=0 ; i<disks; i++)
if (bh == sh->bh_cache[i])
if (bi == &sh->dev[i].req)
break;
PRINTK("end_read_request %lu/%d, count: %d, uptodate %d.\n", sh->sector, i, atomic_read(&sh->count), uptodate);
......@@ -392,7 +341,9 @@ static void raid5_end_read_request (struct buffer_head * bh, int uptodate)
}
if (uptodate) {
struct buffer_head *buffer;
#if 0
struct bio *bio;
unsigned long flags;
spin_lock_irqsave(&conf->device_lock, flags);
/* we can return a buffer if we bypassed the cache or
* if the top buffer is not in highmem. If there are
......@@ -409,38 +360,43 @@ static void raid5_end_read_request (struct buffer_head * bh, int uptodate)
} else
buffer = NULL;
spin_unlock_irqrestore(&conf->device_lock, flags);
if (sh->bh_page[i]==NULL)
if (sh->bh_page[i]==bh->b_page)
set_buffer_uptodate(bh);
if (buffer) {
if (buffer->b_page != bh->b_page)
memcpy(buffer->b_data, bh->b_data, bh->b_size);
buffer->b_end_io(buffer, 1);
}
#else
set_bit(R5_UPTODATE, &sh->dev[i].flags);
#endif
} else {
md_error(conf->mddev, bh->b_bdev);
clear_buffer_uptodate(bh);
md_error(conf->mddev, bi->bi_bdev);
clear_bit(R5_UPTODATE, &sh->dev[i].flags);
}
#if 0
/* must restore b_page before unlocking buffer... */
if (sh->bh_page[i]) {
if (sh->bh_page[i] != bh->b_page) {
bh->b_page = sh->bh_page[i];
bh->b_data = page_address(bh->b_page);
sh->bh_page[i] = NULL;
clear_buffer_uptodate(bh);
}
clear_buffer_locked(bh);
#endif
clear_bit(R5_LOCKED, &sh->dev[i].flags);
set_bit(STRIPE_HANDLE, &sh->state);
release_stripe(sh);
}
static void raid5_end_write_request (struct buffer_head *bh, int uptodate)
static void raid5_end_write_request (struct bio *bi)
{
struct stripe_head *sh = bh->b_private;
struct stripe_head *sh = bi->bi_private;
raid5_conf_t *conf = sh->raid_conf;
int disks = conf->raid_disks, i;
unsigned long flags;
int uptodate = test_bit(BIO_UPTODATE, &bi->bi_flags);
for (i=0 ; i<disks; i++)
if (bh == sh->bh_cache[i])
if (bi == &sh->dev[i].req)
break;
PRINTK("end_write_request %lu/%d, count %d, uptodate: %d.\n", sh->sector, i, atomic_read(&sh->count), uptodate);
......@@ -449,41 +405,48 @@ static void raid5_end_write_request (struct buffer_head *bh, int uptodate)
return;
}
md_spin_lock_irqsave(&conf->device_lock, flags);
spin_lock_irqsave(&conf->device_lock, flags);
if (!uptodate)
md_error(conf->mddev, bh->b_bdev);
clear_buffer_locked(bh);
md_error(conf->mddev, bi->bi_bdev);
clear_bit(R5_LOCKED, &sh->dev[i].flags);
set_bit(STRIPE_HANDLE, &sh->state);
__release_stripe(conf, sh);
md_spin_unlock_irqrestore(&conf->device_lock, flags);
spin_unlock_irqrestore(&conf->device_lock, flags);
}
static struct buffer_head *raid5_build_block (struct stripe_head *sh, int i)
static unsigned long compute_blocknr(struct stripe_head *sh, int i);
static void raid5_build_block (struct stripe_head *sh, int i)
{
raid5_conf_t *conf = sh->raid_conf;
struct buffer_head *bh = sh->bh_cache[i];
unsigned long block = sh->sector / (sh->size >> 9);
init_buffer(bh, raid5_end_read_request, sh);
bh->b_dev = conf->disks[i].dev;
/* FIXME - later we will need bdev here */
bh->b_blocknr = block;
bh->b_state = (1 << BH_Req) | (1 << BH_Mapped);
bh->b_size = sh->size;
return bh;
struct r5dev *dev = &sh->dev[i];
bio_init(&dev->req);
dev->req.bi_io_vec = &dev->vec;
dev->req.bi_vcnt++;
dev->vec.bv_page = dev->page;
dev->vec.bv_len = STRIPE_SIZE;
dev->vec.bv_offset = 0;
dev->req.bi_bdev = conf->disks[i].bdev;
dev->req.bi_sector = sh->sector;
dev->req.bi_private = sh;
dev->flags = 0;
if (i != sh->pd_idx)
dev->sector = compute_blocknr(sh, i);
}
static int raid5_error (mddev_t *mddev, kdev_t dev)
static int error (mddev_t *mddev, kdev_t dev)
{
raid5_conf_t *conf = (raid5_conf_t *) mddev->private;
mdp_super_t *sb = mddev->sb;
struct disk_info *disk;
int i;
PRINTK("raid5_error called\n");
PRINTK("raid5: error called\n");
for (i = 0, disk = conf->disks; i < conf->raid_disks; i++, disk++) {
if (kdev_same(disk->dev, dev)) {
......@@ -544,14 +507,14 @@ static int raid5_error (mddev_t *mddev, kdev_t dev)
* Input: a 'big' sector number,
* Output: index of the data and parity disk, and the sector # in them.
*/
static unsigned long raid5_compute_sector(unsigned long r_sector, unsigned int raid_disks,
static unsigned long raid5_compute_sector(sector_t r_sector, unsigned int raid_disks,
unsigned int data_disks, unsigned int * dd_idx,
unsigned int * pd_idx, raid5_conf_t *conf)
{
unsigned long stripe;
sector_t stripe;
unsigned long chunk_number;
unsigned int chunk_offset;
unsigned long new_sector;
sector_t new_sector;
int sectors_per_chunk = conf->chunk_size >> 9;
/* First compute the information on this sector */
......@@ -607,17 +570,17 @@ static unsigned long raid5_compute_sector(unsigned long r_sector, unsigned int r
return new_sector;
}
#if 0
static unsigned long compute_blocknr(struct stripe_head *sh, int i)
static sector_t compute_blocknr(struct stripe_head *sh, int i)
{
raid5_conf_t *conf = sh->raid_conf;
int raid_disks = conf->raid_disks, data_disks = raid_disks - 1;
unsigned long new_sector = sh->sector, check;
sector_t new_sector = sh->sector, check;
int sectors_per_chunk = conf->chunk_size >> 9;
unsigned long stripe = new_sector / sectors_per_chunk;
sector_t stripe = new_sector / sectors_per_chunk;
int chunk_offset = new_sector % sectors_per_chunk;
int chunk_number, dummy1, dummy2, dd_idx = i;
unsigned long r_sector, blocknr;
sector_t r_sector;
switch (conf->algorithm) {
case ALGORITHM_LEFT_ASYMMETRIC:
......@@ -637,22 +600,72 @@ static unsigned long compute_blocknr(struct stripe_head *sh, int i)
chunk_number = stripe * data_disks + i;
r_sector = chunk_number * sectors_per_chunk + chunk_offset;
blocknr = r_sector / (sh->size >> 9);
check = raid5_compute_sector (r_sector, raid_disks, data_disks, &dummy1, &dummy2, conf);
if (check != sh->sector || dummy1 != dd_idx || dummy2 != sh->pd_idx) {
printk("compute_blocknr: map not correct\n");
return 0;
}
return blocknr;
return r_sector;
}
#endif
#define check_xor() do { \
if (count == MAX_XOR_BLOCKS) { \
xor_block(count, bh_ptr); \
count = 1; \
} \
/*
* Copy data between a page in the stripe cache, and one or more bion
* The page could align with the middle of the bio, or there could be
* several bion, each with several bio_vecs, which cover part of the page
* Multiple bion are linked together on bi_next. There may be extras
* at the end of this list. We ignore them.
*/
static void copy_data(int frombio, struct bio *bio,
struct page *page,
sector_t sector)
{
char *pa = page_address(page);
struct bio_vec *bvl;
int i;
for (;bio && bio->bi_sector < sector+STRIPE_SECTORS;
bio = bio->bi_next) {
int page_offset;
if (bio->bi_sector >= sector)
page_offset = (signed)(bio->bi_sector - sector) * 512;
else
page_offset = (signed)(sector - bio->bi_sector) * -512;
bio_for_each_segment(bvl, bio, i) {
char *ba = __bio_kmap(bio, i);
int len = bio_iovec_idx(bio,i)->bv_len;
int clen;
int b_offset = 0;
if (page_offset < 0) {
b_offset = -page_offset;
page_offset += b_offset;
len -= b_offset;
}
if (len > 0 && page_offset + len > STRIPE_SIZE)
clen = STRIPE_SIZE - page_offset;
else clen = len;
if (len > 0) {
if (frombio)
memcpy(pa+page_offset, ba+b_offset, clen);
else
memcpy(ba+b_offset, pa+page_offset, clen);
}
__bio_kunmap(bio, i);
page_offset += len;
}
}
}
#define check_xor() do { \
if (count == MAX_XOR_BLOCKS) { \
xor_block(count, STRIPE_SIZE, ptr); \
count = 1; \
} \
} while(0)
......@@ -660,88 +673,84 @@ static void compute_block(struct stripe_head *sh, int dd_idx)
{
raid5_conf_t *conf = sh->raid_conf;
int i, count, disks = conf->raid_disks;
struct buffer_head *bh_ptr[MAX_XOR_BLOCKS], *bh;
void *ptr[MAX_XOR_BLOCKS], *p;
PRINTK("compute_block, stripe %lu, idx %d\n", sh->sector, dd_idx);
memset(sh->bh_cache[dd_idx]->b_data, 0, sh->size);
bh_ptr[0] = sh->bh_cache[dd_idx];
ptr[0] = page_address(sh->dev[dd_idx].page);
memset(ptr[0], 0, STRIPE_SIZE);
count = 1;
for (i = disks ; i--; ) {
if (i == dd_idx)
continue;
bh = sh->bh_cache[i];
if (buffer_uptodate(bh))
bh_ptr[count++] = bh;
p = page_address(sh->dev[i].page);
if (test_bit(R5_UPTODATE, &sh->dev[i].flags))
ptr[count++] = p;
else
printk("compute_block() %d, stripe %lu, %d not present\n", dd_idx, sh->sector, i);
check_xor();
}
if (count != 1)
xor_block(count, bh_ptr);
set_buffer_uptodate(sh->bh_cache[dd_idx]);
xor_block(count, STRIPE_SIZE, ptr);
set_bit(R5_UPTODATE, &sh->dev[i].flags);
}
static void compute_parity(struct stripe_head *sh, int method)
{
raid5_conf_t *conf = sh->raid_conf;
int i, pd_idx = sh->pd_idx, disks = conf->raid_disks, count;
struct buffer_head *bh_ptr[MAX_XOR_BLOCKS];
struct buffer_head *chosen[MD_SB_DISKS];
void *ptr[MAX_XOR_BLOCKS];
struct bio *chosen[MD_SB_DISKS];
PRINTK("compute_parity, stripe %lu, method %d\n", sh->sector, method);
memset(chosen, 0, sizeof(chosen));
count = 1;
bh_ptr[0] = sh->bh_cache[pd_idx];
ptr[0] = page_address(sh->dev[pd_idx].page);
switch(method) {
case READ_MODIFY_WRITE:
if (!buffer_uptodate(sh->bh_cache[pd_idx]))
if (!test_bit(R5_UPTODATE, &sh->dev[pd_idx].flags))
BUG();
for (i=disks ; i-- ;) {
if (i==pd_idx)
continue;
if (sh->bh_write[i] &&
buffer_uptodate(sh->bh_cache[i])) {
bh_ptr[count++] = sh->bh_cache[i];
chosen[i] = sh->bh_write[i];
sh->bh_write[i] = sh->bh_write[i]->b_reqnext;
chosen[i]->b_reqnext = sh->bh_written[i];
sh->bh_written[i] = chosen[i];
if (sh->dev[i].towrite &&
test_bit(R5_UPTODATE, &sh->dev[i].flags)) {
ptr[count++] = page_address(sh->dev[i].page);
chosen[i] = sh->dev[i].towrite;
sh->dev[i].towrite = NULL;
if (sh->dev[i].written) BUG();
sh->dev[i].written = chosen[i];
check_xor();
}
}
break;
case RECONSTRUCT_WRITE:
memset(sh->bh_cache[pd_idx]->b_data, 0, sh->size);
memset(ptr[0], 0, STRIPE_SIZE);
for (i= disks; i-- ;)
if (i!=pd_idx && sh->bh_write[i]) {
chosen[i] = sh->bh_write[i];
sh->bh_write[i] = sh->bh_write[i]->b_reqnext;
chosen[i]->b_reqnext = sh->bh_written[i];
sh->bh_written[i] = chosen[i];
if (i!=pd_idx && sh->dev[i].towrite) {
chosen[i] = sh->dev[i].towrite;
sh->dev[i].towrite = NULL;
if (sh->dev[i].written) BUG();
sh->dev[i].written = chosen[i];
}
break;
case CHECK_PARITY:
break;
}
if (count>1) {
xor_block(count, bh_ptr);
xor_block(count, STRIPE_SIZE, ptr);
count = 1;
}
for (i = disks; i--;)
if (chosen[i]) {
struct buffer_head *bh = sh->bh_cache[i];
char *bdata;
bdata = bh_kmap(chosen[i]);
memcpy(bh->b_data,
bdata,sh->size);
bh_kunmap(chosen[i]);
set_buffer_locked(bh);
set_buffer_uptodate(bh);
sector_t sector = sh->dev[i].sector;
copy_data(1, chosen[i], sh->dev[i].page, sector);
set_bit(R5_LOCKED, &sh->dev[i].flags);
set_bit(R5_UPTODATE, &sh->dev[i].flags);
}
switch(method) {
......@@ -749,55 +758,74 @@ static void compute_parity(struct stripe_head *sh, int method)
case CHECK_PARITY:
for (i=disks; i--;)
if (i != pd_idx) {
bh_ptr[count++] = sh->bh_cache[i];
ptr[count++] = page_address(sh->dev[i].page);
check_xor();
}
break;
case READ_MODIFY_WRITE:
for (i = disks; i--;)
if (chosen[i]) {
bh_ptr[count++] = sh->bh_cache[i];
ptr[count++] = page_address(sh->dev[i].page);
check_xor();
}
}
if (count != 1)
xor_block(count, bh_ptr);
xor_block(count, STRIPE_SIZE, ptr);
if (method != CHECK_PARITY) {
set_buffer_uptodate(sh->bh_cache[pd_idx]);
set_buffer_locked(sh->bh_cache[pd_idx]);
set_bit(R5_UPTODATE, &sh->dev[pd_idx].flags);
set_bit(R5_LOCKED, &sh->dev[pd_idx].flags);
} else
clear_buffer_uptodate(sh->bh_cache[pd_idx]);
clear_bit(R5_UPTODATE, &sh->dev[pd_idx].flags);
}
static void add_stripe_bh (struct stripe_head *sh, struct buffer_head *bh, int dd_idx, int rw)
/*
* Each stripe/dev can have one or more bion attached.
* toread/towrite point to the first in a chain.
* The bi_next chain must be in order.
*/
static void add_stripe_bio (struct stripe_head *sh, struct bio *bi, int dd_idx, int forwrite)
{
struct buffer_head **bhp;
struct bio **bip;
raid5_conf_t *conf = sh->raid_conf;
PRINTK("adding bh b#%lu to stripe s#%lu\n", bh->b_blocknr, sh->sector);
PRINTK("adding bh b#%lu to stripe s#%lu\n", bi->bi_sector, sh->sector);
spin_lock(&sh->lock);
spin_lock_irq(&conf->device_lock);
bh->b_reqnext = NULL;
if (rw == READ)
bhp = &sh->bh_read[dd_idx];
if (forwrite)
bip = &sh->dev[dd_idx].towrite;
else
bhp = &sh->bh_write[dd_idx];
while (*bhp) {
printk(KERN_NOTICE "raid5: multiple %d requests for sector %ld\n", rw, sh->sector);
bhp = & (*bhp)->b_reqnext;
}
*bhp = bh;
bip = &sh->dev[dd_idx].toread;
while (*bip && (*bip)->bi_sector < bi->bi_sector)
bip = & (*bip)->bi_next;
/* FIXME do I need to worry about overlapping bion */
if (*bip && bi->bi_next && (*bip) != bi->bi_next)
BUG();
if (*bip)
bi->bi_next = *bip;
*bip = bi;
bi->bi_phys_segments ++;
spin_unlock_irq(&conf->device_lock);
spin_unlock(&sh->lock);
PRINTK("added bh b#%lu to stripe s#%lu, disk %d.\n", bh->b_blocknr, sh->sector, dd_idx);
}
if (forwrite) {
/* check if page is coverred */
sector_t sector = sh->dev[dd_idx].sector;
for (bi=sh->dev[dd_idx].towrite;
sector < sh->dev[dd_idx].sector + STRIPE_SECTORS &&
bi && bi->bi_sector <= sector;
bi = bi->bi_next) {
if (bi->bi_sector + (bi->bi_size>>9) >= sector)
sector = bi->bi_sector + (bi->bi_size>>9);
}
if (sector >= sh->dev[dd_idx].sector + STRIPE_SECTORS)
set_bit(R5_OVERWRITE, &sh->dev[dd_idx].flags);
}
PRINTK("added bi b#%lu to stripe s#%lu, disk %d.\n", bi->bi_sector, sh->sector, dd_idx);
}
/*
......@@ -822,13 +850,14 @@ static void handle_stripe(struct stripe_head *sh)
{
raid5_conf_t *conf = sh->raid_conf;
int disks = conf->raid_disks;
struct buffer_head *return_ok= NULL, *return_fail = NULL;
struct bio *return_bi= NULL;
struct bio *bi;
int action[MD_SB_DISKS];
int i;
int syncing;
int locked=0, uptodate=0, to_read=0, to_write=0, failed=0, written=0;
int failed_num=0;
struct buffer_head *bh;
struct r5dev *dev;
PRINTK("handling stripe %ld, cnt=%d, pd_idx=%d\n", sh->sector, atomic_read(&sh->count), sh->pd_idx);
memset(action, 0, sizeof(action));
......@@ -841,36 +870,38 @@ static void handle_stripe(struct stripe_head *sh)
/* Now to look around and see what can be done */
for (i=disks; i--; ) {
bh = sh->bh_cache[i];
PRINTK("check %d: state 0x%lx read %p write %p written %p\n", i, bh->b_state, sh->bh_read[i], sh->bh_write[i], sh->bh_written[i]);
dev = &sh->dev[i];
PRINTK("check %d: state 0x%lx read %p write %p written %p\n", i,
dev->flags, dev->toread, dev->towrite, dev->written);
/* maybe we can reply to a read */
if (buffer_uptodate(bh) && sh->bh_read[i]) {
struct buffer_head *rbh, *rbh2;
if (test_bit(R5_UPTODATE, &dev->flags) && dev->toread) {
struct bio *rbi, *rbi2;
PRINTK("Return read for disc %d\n", i);
spin_lock_irq(&conf->device_lock);
rbh = sh->bh_read[i];
sh->bh_read[i] = NULL;
rbi = dev->toread;
dev->toread = NULL;
spin_unlock_irq(&conf->device_lock);
while (rbh) {
char *bdata;
bdata = bh_kmap(rbh);
memcpy(bdata, bh->b_data, bh->b_size);
bh_kunmap(rbh);
rbh2 = rbh->b_reqnext;
rbh->b_reqnext = return_ok;
return_ok = rbh;
rbh = rbh2;
while (rbi && rbi->bi_sector < dev->sector + STRIPE_SECTORS) {
copy_data(0, rbi, dev->page, dev->sector);
rbi2 = rbi->bi_next;
spin_lock_irq(&conf->device_lock);
if (--rbi->bi_phys_segments == 0) {
rbi->bi_next = return_bi;
return_bi = rbi;
}
spin_unlock_irq(&conf->device_lock);
rbi = rbi2;
}
}
/* now count some things */
if (buffer_locked(bh)) locked++;
if (buffer_uptodate(bh)) uptodate++;
if (test_bit(R5_LOCKED, &dev->flags)) locked++;
if (test_bit(R5_UPTODATE, &dev->flags)) uptodate++;
if (sh->bh_read[i]) to_read++;
if (sh->bh_write[i]) to_write++;
if (sh->bh_written[i]) written++;
if (dev->toread) to_read++;
if (dev->towrite) to_write++;
if (dev->written) written++;
if (!conf->disks[i].operational) {
failed++;
failed_num = i;
......@@ -882,29 +913,42 @@ static void handle_stripe(struct stripe_head *sh)
* need to be failed
*/
if (failed > 1 && to_read+to_write) {
spin_lock_irq(&conf->device_lock);
for (i=disks; i--; ) {
/* fail all writes first */
if (sh->bh_write[i]) to_write--;
while ((bh = sh->bh_write[i])) {
sh->bh_write[i] = bh->b_reqnext;
bh->b_reqnext = return_fail;
return_fail = bh;
bi = sh->dev[i].towrite;
sh->dev[i].towrite = NULL;
if (bi) to_write--;
while (bi && bi->bi_sector < sh->dev[i].sector + STRIPE_SECTORS){
struct bio *nextbi = bi->bi_next;
clear_bit(BIO_UPTODATE, &bi->bi_flags);
if (--bi->bi_phys_segments == 0) {
bi->bi_next = return_bi;
return_bi = bi;
}
bi = nextbi;
}
/* fail any reads if this device is non-operational */
if (!conf->disks[i].operational) {
spin_lock_irq(&conf->device_lock);
if (sh->bh_read[i]) to_read--;
while ((bh = sh->bh_read[i])) {
sh->bh_read[i] = bh->b_reqnext;
bh->b_reqnext = return_fail;
return_fail = bh;
bi = sh->dev[i].toread;
sh->dev[i].toread = NULL;
if (bi) to_read--;
while (bi && bi->bi_sector < sh->dev[i].sector + STRIPE_SECTORS){
struct bio *nextbi = bi->bi_next;
clear_bit(BIO_UPTODATE, &bi->bi_flags);
if (--bi->bi_phys_segments == 0) {
bi->bi_next = return_bi;
return_bi = bi;
}
bi = nextbi;
}
spin_unlock_irq(&conf->device_lock);
}
}
spin_unlock_irq(&conf->device_lock);
}
if (failed > 1 && syncing) {
md_done_sync(conf->mddev, (sh->size>>9) - sh->sync_redone,0);
md_done_sync(conf->mddev, STRIPE_SECTORS,0);
clear_bit(STRIPE_SYNCING, &sh->state);
syncing = 0;
}
......@@ -912,40 +956,43 @@ static void handle_stripe(struct stripe_head *sh)
/* might be able to return some write requests if the parity block
* is safe, or on a failed drive
*/
bh = sh->bh_cache[sh->pd_idx];
dev = &sh->dev[sh->pd_idx];
if ( written &&
( (conf->disks[sh->pd_idx].operational && !buffer_locked(bh) && buffer_uptodate(bh))
( (conf->disks[sh->pd_idx].operational && !test_bit(R5_LOCKED, &dev->flags) &&
test_bit(R5_UPTODATE, &dev->flags))
|| (failed == 1 && failed_num == sh->pd_idx))
) {
/* any written block on a uptodate or failed drive can be returned */
/* any written block on an uptodate or failed drive can be returned */
for (i=disks; i--; )
if (sh->bh_written[i]) {
bh = sh->bh_cache[i];
if (sh->dev[i].written) {
dev = &sh->dev[i];
if (!conf->disks[sh->pd_idx].operational ||
(!buffer_locked(bh) && buffer_uptodate(bh)) ) {
(!test_bit(R5_LOCKED, &dev->flags) && test_bit(R5_UPTODATE, &dev->flags)) ) {
/* maybe we can return some write requests */
struct buffer_head *wbh, *wbh2;
PRINTK("Return write for disc %d\n", i);
wbh = sh->bh_written[i];
sh->bh_written[i] = NULL;
while (wbh) {
wbh2 = wbh->b_reqnext;
wbh->b_reqnext = return_ok;
return_ok = wbh;
wbh = wbh2;
}
struct bio *wbi, *wbi2;
PRINTK("Return write for disc %d\n", i);
wbi = dev->written;
dev->written = NULL;
while (wbi && wbi->bi_sector < dev->sector + STRIPE_SECTORS) {
wbi2 = wbi->bi_next;
if (--wbi->bi_phys_segments == 0) {
wbi->bi_next = return_bi;
return_bi = wbi;
}
wbi = wbi2;
}
}
}
}
/* Now we might consider reading some blocks, either to check/generate
* parity, or to satisfy requests
*/
if (to_read || (syncing && (uptodate+failed < disks))) {
for (i=disks; i--;) {
bh = sh->bh_cache[i];
if (!buffer_locked(bh) && !buffer_uptodate(bh) &&
(sh->bh_read[i] || syncing || (failed && sh->bh_read[failed_num]))) {
dev = &sh->dev[i];
if (!test_bit(R5_LOCKED, &dev->flags) && !test_bit(R5_UPTODATE, &dev->flags) &&
(dev->toread || syncing || (failed && sh->dev[failed_num].toread))) {
/* we would like to get this block, possibly
* by computing it, but we might not be able to
*/
......@@ -954,21 +1001,21 @@ static void handle_stripe(struct stripe_head *sh)
compute_block(sh, i);
uptodate++;
} else if (conf->disks[i].operational) {
set_buffer_locked(bh);
set_bit(R5_LOCKED, &dev->flags);
action[i] = READ+1;
#if 0
/* if I am just reading this block and we don't have
a failed drive, or any pending writes then sidestep the cache */
if (sh->bh_page[i]) BUG();
if (sh->bh_read[i] && !sh->bh_read[i]->b_reqnext &&
! syncing && !failed && !to_write) {
sh->bh_page[i] = sh->bh_cache[i]->b_page;
sh->bh_cache[i]->b_page = sh->bh_read[i]->b_page;
sh->bh_cache[i]->b_data = sh->bh_read[i]->b_data;
}
#endif
locked++;
PRINTK("Reading block %d (sync=%d)\n", i, syncing);
if (syncing)
md_sync_acct(conf->disks[i].dev, bh->b_size>>9);
md_sync_acct(conf->disks[i].dev, STRIPE_SECTORS);
}
}
}
......@@ -980,10 +1027,14 @@ static void handle_stripe(struct stripe_head *sh)
int rmw=0, rcw=0;
for (i=disks ; i--;) {
/* would I have to read this buffer for read_modify_write */
bh = sh->bh_cache[i];
if ((sh->bh_write[i] || i == sh->pd_idx) &&
(!buffer_locked(bh) || sh->bh_page[i]) &&
!buffer_uptodate(bh)) {
dev = &sh->dev[i];
if ((dev->towrite || i == sh->pd_idx) &&
(!test_bit(R5_LOCKED, &dev->flags)
#if 0
|| sh->bh_page[i]!=bh->b_page
#endif
) &&
!test_bit(R5_UPTODATE, &dev->flags)) {
if (conf->disks[i].operational
/* && !(conf->resync_parity && i == sh->pd_idx) */
)
......@@ -991,9 +1042,13 @@ static void handle_stripe(struct stripe_head *sh)
else rmw += 2*disks; /* cannot read it */
}
/* Would I have to read this buffer for reconstruct_write */
if (!sh->bh_write[i] && i != sh->pd_idx &&
(!buffer_locked(bh) || sh->bh_page[i]) &&
!buffer_uptodate(bh)) {
if (!test_bit(R5_OVERWRITE, &dev->flags) && i != sh->pd_idx &&
(!test_bit(R5_LOCKED, &dev->flags)
#if 0
|| sh->bh_page[i] != bh->b_page
#endif
) &&
!test_bit(R5_UPTODATE, &dev->flags)) {
if (conf->disks[i].operational) rcw++;
else rcw += 2*disks;
}
......@@ -1003,14 +1058,14 @@ static void handle_stripe(struct stripe_head *sh)
if (rmw < rcw && rmw > 0)
/* prefer read-modify-write, but need to get some data */
for (i=disks; i--;) {
bh = sh->bh_cache[i];
if ((sh->bh_write[i] || i == sh->pd_idx) &&
!buffer_locked(bh) && !buffer_uptodate(bh) &&
dev = &sh->dev[i];
if ((dev->towrite || i == sh->pd_idx) &&
!test_bit(R5_LOCKED, &dev->flags) && !test_bit(R5_UPTODATE, &dev->flags) &&
conf->disks[i].operational) {
if (test_bit(STRIPE_PREREAD_ACTIVE, &sh->state))
{
PRINTK("Read_old block %d for r-m-w\n", i);
set_buffer_locked(bh);
set_bit(R5_LOCKED, &dev->flags);
action[i] = READ+1;
locked++;
} else {
......@@ -1022,14 +1077,14 @@ static void handle_stripe(struct stripe_head *sh)
if (rcw <= rmw && rcw > 0)
/* want reconstruct write, but need to get some data */
for (i=disks; i--;) {
bh = sh->bh_cache[i];
if (!sh->bh_write[i] && i != sh->pd_idx &&
!buffer_locked(bh) && !buffer_uptodate(bh) &&
dev = &sh->dev[i];
if (!test_bit(R5_OVERWRITE, &dev->flags) && i != sh->pd_idx &&
!test_bit(R5_LOCKED, &dev->flags) && !test_bit(R5_UPTODATE, &dev->flags) &&
conf->disks[i].operational) {
if (test_bit(STRIPE_PREREAD_ACTIVE, &sh->state))
{
PRINTK("Read_old block %d for Reconstruct\n", i);
set_buffer_locked(bh);
set_bit(R5_LOCKED, &dev->flags);
action[i] = READ+1;
locked++;
} else {
......@@ -1044,7 +1099,7 @@ static void handle_stripe(struct stripe_head *sh)
compute_parity(sh, rcw==0 ? RECONSTRUCT_WRITE : READ_MODIFY_WRITE);
/* now every locked buffer is ready to be written */
for (i=disks; i--;)
if (buffer_locked(sh->bh_cache[i])) {
if (test_bit(R5_LOCKED, &sh->dev[i].flags)) {
PRINTK("Writing block %d\n", i);
locked++;
action[i] = WRITE+1;
......@@ -1068,13 +1123,14 @@ static void handle_stripe(struct stripe_head *sh)
!test_bit(STRIPE_INSYNC, &sh->state) && failed <= 1) {
set_bit(STRIPE_HANDLE, &sh->state);
if (failed == 0) {
char *pagea;
if (uptodate != disks)
BUG();
compute_parity(sh, CHECK_PARITY);
uptodate--;
bh = sh->bh_cache[sh->pd_idx];
if ((*(u32*)bh->b_data) == 0 &&
!memcmp(bh->b_data, bh->b_data+4, bh->b_size-4)) {
pagea = page_address(sh->dev[sh->pd_idx].page);
if ((*(u32*)pagea) == 0 &&
!memcmp(pagea, pagea+4, STRIPE_SIZE-4)) {
/* parity is correct (on disc, not in buffer any more) */
set_bit(STRIPE_INSYNC, &sh->state);
}
......@@ -1084,7 +1140,7 @@ static void handle_stripe(struct stripe_head *sh)
if (failed==0)
failed_num = sh->pd_idx;
/* should be able to compute the missing block and write it to spare */
if (!buffer_uptodate(sh->bh_cache[failed_num])) {
if (!test_bit(R5_UPTODATE, &sh->dev[failed_num].flags)) {
if (uptodate+1 != disks)
BUG();
compute_block(sh, failed_num);
......@@ -1092,60 +1148,62 @@ static void handle_stripe(struct stripe_head *sh)
}
if (uptodate != disks)
BUG();
bh = sh->bh_cache[failed_num];
set_buffer_locked(bh);
dev = &sh->dev[failed_num];
set_bit(R5_LOCKED, &dev->flags);
action[failed_num] = WRITE+1;
locked++;
set_bit(STRIPE_INSYNC, &sh->state);
if (conf->disks[failed_num].operational)
md_sync_acct(conf->disks[failed_num].dev, bh->b_size>>9);
md_sync_acct(conf->disks[failed_num].dev, STRIPE_SECTORS);
else if ((spare=conf->spare))
md_sync_acct(spare->dev, bh->b_size>>9);
md_sync_acct(spare->dev, STRIPE_SECTORS);
}
}
if (syncing && locked == 0 && test_bit(STRIPE_INSYNC, &sh->state)) {
md_done_sync(conf->mddev, (sh->size>>9) - sh->sync_redone,1);
md_done_sync(conf->mddev, STRIPE_SECTORS,1);
clear_bit(STRIPE_SYNCING, &sh->state);
}
spin_unlock(&sh->lock);
while ((bh=return_ok)) {
return_ok = bh->b_reqnext;
bh->b_reqnext = NULL;
bh->b_end_io(bh, 1);
}
while ((bh=return_fail)) {
return_fail = bh->b_reqnext;
bh->b_reqnext = NULL;
bh->b_end_io(bh, 0);
while ((bi=return_bi)) {
return_bi = bi->bi_next;
bi->bi_next = NULL;
bi->bi_end_io(bi);
}
for (i=disks; i-- ;)
if (action[i]) {
struct buffer_head *bh = sh->bh_cache[i];
struct bio *bi = &sh->dev[i].req;
struct disk_info *spare = conf->spare;
int skip = 0;
if (action[i] == READ+1)
bh->b_end_io = raid5_end_read_request;
bi->bi_end_io = raid5_end_read_request;
else
bh->b_end_io = raid5_end_write_request;
bi->bi_end_io = raid5_end_write_request;
if (conf->disks[i].operational)
bh->b_dev = conf->disks[i].dev;
bi->bi_bdev = conf->disks[i].bdev;
else if (spare && action[i] == WRITE+1)
bh->b_dev = spare->dev;
bi->bi_bdev = spare->bdev;
else skip=1;
/* FIXME - later we will need bdev here */
if (!skip) {
PRINTK("for %ld schedule op %d on disc %d\n", sh->sector, action[i]-1, i);
atomic_inc(&sh->count);
bh->b_rdev = bh->b_dev;
bh->b_rsector = bh->b_blocknr * (bh->b_size>>9);
generic_make_request(action[i]-1, bh);
bi->bi_sector = sh->sector;
if (action[i] == READ+1)
bi->bi_rw = 0;
else
bi->bi_rw = 1;
bi->bi_flags = 0;
bi->bi_vcnt = 1;
bi->bi_idx = 0;
bi->bi_io_vec = &sh->dev[i].vec;
bi->bi_size = STRIPE_SIZE;
bi->bi_next = NULL;
generic_make_request(bi);
} else {
PRINTK("skip op %d on disc %d for sector %ld\n", action[i]-1, i, sh->sector);
clear_buffer_locked(bh);
clear_bit(R5_LOCKED, &dev->flags);
set_bit(STRIPE_HANDLE, &sh->state);
}
}
......@@ -1192,13 +1250,14 @@ static inline void raid5_plug_device(raid5_conf_t *conf)
spin_unlock_irq(&conf->device_lock);
}
static int raid5_make_request (mddev_t *mddev, int rw, struct buffer_head * bh)
static int make_request (mddev_t *mddev, int rw, struct bio * bi)
{
raid5_conf_t *conf = (raid5_conf_t *) mddev->private;
const unsigned int raid_disks = conf->raid_disks;
const unsigned int data_disks = raid_disks - 1;
unsigned int dd_idx, pd_idx;
unsigned long new_sector;
sector_t new_sector;
sector_t logical_sector, last_sector;
int read_ahead = 0;
struct stripe_head *sh;
......@@ -1208,25 +1267,39 @@ static int raid5_make_request (mddev_t *mddev, int rw, struct buffer_head * bh)
read_ahead=1;
}
new_sector = raid5_compute_sector(bh->b_rsector,
raid_disks, data_disks, &dd_idx, &pd_idx, conf);
logical_sector = bi->bi_sector & ~(STRIPE_SECTORS-1);
last_sector = bi->bi_sector + (bi->bi_size>>9);
PRINTK("raid5_make_request, sector %lu\n", new_sector);
sh = get_active_stripe(conf, new_sector, bh->b_size, read_ahead);
if (sh) {
sh->pd_idx = pd_idx;
bi->bi_next = NULL;
set_bit(BIO_UPTODATE, &bi->bi_flags); /* will be cleared if error detected */
bi->bi_phys_segments = 1; /* over-loaded to count active stripes */
for (;logical_sector < last_sector; logical_sector += STRIPE_SECTORS) {
new_sector = raid5_compute_sector(logical_sector,
raid_disks, data_disks, &dd_idx, &pd_idx, conf);
add_stripe_bh(sh, bh, dd_idx, rw);
PRINTK("raid5: make_request, sector %ul logical %ul\n",
new_sector, logical_sector);
raid5_plug_device(conf);
handle_stripe(sh);
release_stripe(sh);
} else
bh->b_end_io(bh, buffer_uptodate(bh));
sh = get_active_stripe(conf, new_sector, pd_idx, read_ahead);
if (sh) {
add_stripe_bio(sh, bi, dd_idx, rw);
raid5_plug_device(conf);
handle_stripe(sh);
release_stripe(sh);
}
}
spin_lock_irq(&conf->device_lock);
if (--bi->bi_phys_segments == 0)
bi->bi_end_io(bi);
spin_unlock_irq(&conf->device_lock);
return 0;
}
static int raid5_sync_request (mddev_t *mddev, unsigned long sector_nr)
/* FIXME go_faster isn't used */
static int sync_request (mddev_t *mddev, sector_t sector_nr, int go_faster)
{
raid5_conf_t *conf = (raid5_conf_t *) mddev->private;
struct stripe_head *sh;
......@@ -1237,25 +1310,19 @@ static int raid5_sync_request (mddev_t *mddev, unsigned long sector_nr)
unsigned long first_sector;
int raid_disks = conf->raid_disks;
int data_disks = raid_disks-1;
int redone = 0;
int bufsize;
sh = get_active_stripe(conf, sector_nr, 0, 0);
bufsize = sh->size;
redone = sector_nr - sh->sector;
first_sector = raid5_compute_sector(stripe*data_disks*sectors_per_chunk
+ chunk_offset, raid_disks, data_disks, &dd_idx, &pd_idx, conf);
sh->pd_idx = pd_idx;
sh = get_active_stripe(conf, sector_nr, pd_idx, 0);
spin_lock(&sh->lock);
set_bit(STRIPE_SYNCING, &sh->state);
clear_bit(STRIPE_INSYNC, &sh->state);
sh->sync_redone = redone;
spin_unlock(&sh->lock);
handle_stripe(sh);
release_stripe(sh);
return (bufsize>>9)-redone;
return STRIPE_SECTORS;
}
/*
......@@ -1280,7 +1347,7 @@ static void raid5d (void *data)
mddev->sb_dirty = 0;
md_update_sb(mddev);
}
md_spin_lock_irq(&conf->device_lock);
spin_lock_irq(&conf->device_lock);
while (1) {
struct list_head *first;
......@@ -1300,17 +1367,17 @@ static void raid5d (void *data)
atomic_inc(&sh->count);
if (atomic_read(&sh->count)!= 1)
BUG();
md_spin_unlock_irq(&conf->device_lock);
spin_unlock_irq(&conf->device_lock);
handled++;
handle_stripe(sh);
release_stripe(sh);
md_spin_lock_irq(&conf->device_lock);
spin_lock_irq(&conf->device_lock);
}
PRINTK("%d stripes handled\n", handled);
md_spin_unlock_irq(&conf->device_lock);
spin_unlock_irq(&conf->device_lock);
PRINTK("--- raid5d inactive\n");
}
......@@ -1340,7 +1407,7 @@ static void raid5syncd (void *data)
printk("raid5: resync finished.\n");
}
static int raid5_run (mddev_t *mddev)
static int run (mddev_t *mddev)
{
raid5_conf_t *conf;
int i, j, raid_disk, memory;
......@@ -1348,7 +1415,7 @@ static int raid5_run (mddev_t *mddev)
mdp_disk_t *desc;
mdk_rdev_t *rdev;
struct disk_info *disk;
struct md_list_head *tmp;
struct list_head *tmp;
int start_recovery = 0;
MOD_INC_USE_COUNT;
......@@ -1365,25 +1432,24 @@ static int raid5_run (mddev_t *mddev)
memset (conf, 0, sizeof (*conf));
conf->mddev = mddev;
if ((conf->stripe_hashtbl = (struct stripe_head **) md__get_free_pages(GFP_ATOMIC, HASH_PAGES_ORDER)) == NULL)
if ((conf->stripe_hashtbl = (struct stripe_head **) __get_free_pages(GFP_ATOMIC, HASH_PAGES_ORDER)) == NULL)
goto abort;
memset(conf->stripe_hashtbl, 0, HASH_PAGES * PAGE_SIZE);
conf->device_lock = MD_SPIN_LOCK_UNLOCKED;
md_init_waitqueue_head(&conf->wait_for_stripe);
conf->device_lock = SPIN_LOCK_UNLOCKED;
init_waitqueue_head(&conf->wait_for_stripe);
INIT_LIST_HEAD(&conf->handle_list);
INIT_LIST_HEAD(&conf->delayed_list);
INIT_LIST_HEAD(&conf->inactive_list);
atomic_set(&conf->active_stripes, 0);
atomic_set(&conf->preread_active_stripes, 0);
conf->buffer_size = PAGE_SIZE; /* good default for rebuild */
conf->plugged = 0;
conf->plug_tq.sync = 0;
conf->plug_tq.routine = &raid5_unplug_device;
conf->plug_tq.data = conf;
PRINTK("raid5_run(md%d) called.\n", mdidx(mddev));
PRINTK("raid5: run(md%d) called.\n", mdidx(mddev));
ITERATE_RDEV(mddev,rdev,tmp) {
/*
......@@ -1404,6 +1470,7 @@ static int raid5_run (mddev_t *mddev)
disk->number = desc->number;
disk->raid_disk = raid_disk;
disk->dev = rdev->dev;
disk->bdev = rdev->bdev;
disk->operational = 0;
disk->write_only = 0;
......@@ -1430,6 +1497,7 @@ static int raid5_run (mddev_t *mddev)
disk->number = desc->number;
disk->raid_disk = raid_disk;
disk->dev = rdev->dev;
disk->bdev = rdev->bdev;
disk->operational = 1;
disk->used_slot = 1;
......@@ -1442,6 +1510,7 @@ static int raid5_run (mddev_t *mddev)
disk->number = desc->number;
disk->raid_disk = raid_disk;
disk->dev = rdev->dev;
disk->bdev = rdev->bdev;
disk->operational = 0;
disk->write_only = 0;
......@@ -1461,6 +1530,7 @@ static int raid5_run (mddev_t *mddev)
disk->number = desc->number;
disk->raid_disk = raid_disk;
disk->dev = NODEV;
disk->bdev = NULL;
disk->operational = 0;
disk->write_only = 0;
......@@ -1518,9 +1588,9 @@ static int raid5_run (mddev_t *mddev)
memory = conf->max_nr_stripes * (sizeof(struct stripe_head) +
conf->raid_disks * ((sizeof(struct buffer_head) + PAGE_SIZE))) / 1024;
if (grow_stripes(conf, conf->max_nr_stripes, GFP_KERNEL)) {
if (grow_stripes(conf, conf->max_nr_stripes)) {
printk(KERN_ERR "raid5: couldn't allocate %dkB for buffers\n", memory);
shrink_stripes(conf, conf->max_nr_stripes);
shrink_stripes(conf);
goto abort;
} else
printk(KERN_INFO "raid5: allocated %dkB for md%d\n", memory, mdidx(mddev));
......@@ -1580,7 +1650,7 @@ static int raid5_run (mddev_t *mddev)
return -EIO;
}
static int raid5_stop_resync (mddev_t *mddev)
static int stop_resync (mddev_t *mddev)
{
raid5_conf_t *conf = mddev_to_conf(mddev);
mdk_thread_t *thread = conf->resync_thread;
......@@ -1597,7 +1667,7 @@ static int raid5_stop_resync (mddev_t *mddev)
return 0;
}
static int raid5_restart_resync (mddev_t *mddev)
static int restart_resync (mddev_t *mddev)
{
raid5_conf_t *conf = mddev_to_conf(mddev);
......@@ -1616,14 +1686,14 @@ static int raid5_restart_resync (mddev_t *mddev)
}
static int raid5_stop (mddev_t *mddev)
static int stop (mddev_t *mddev)
{
raid5_conf_t *conf = (raid5_conf_t *) mddev->private;
if (conf->resync_thread)
md_unregister_thread(conf->resync_thread);
md_unregister_thread(conf->thread);
shrink_stripes(conf, conf->max_nr_stripes);
shrink_stripes(conf);
free_pages((unsigned long) conf->stripe_hashtbl, HASH_PAGES_ORDER);
kfree(conf);
mddev->private = NULL;
......@@ -1636,12 +1706,11 @@ static void print_sh (struct stripe_head *sh)
{
int i;
printk("sh %lu, size %d, pd_idx %d, state %ld.\n", sh->sector, sh->size, sh->pd_idx, sh->state);
printk("sh %lu, pd_idx %d, state %ld.\n", sh->sector, sh->pd_idx, sh->state);
printk("sh %lu, count %d.\n", sh->sector, atomic_read(&sh->count));
printk("sh %lu, ", sh->sector);
for (i = 0; i < MD_SB_DISKS; i++) {
if (sh->bh_cache[i])
printk("(cache%d: %p %ld) ", i, sh->bh_cache[i], sh->bh_cache[i]->b_state);
for (i = 0; i < sh->raid_conf->raid_disks; i++) {
printk("(cache%d: %p %ld) ", i, sh->dev[i].page, sh->dev[i].flags);
}
printk("\n");
}
......@@ -1651,7 +1720,7 @@ static void printall (raid5_conf_t *conf)
struct stripe_head *sh;
int i;
md_spin_lock_irq(&conf->device_lock);
spin_lock_irq(&conf->device_lock);
for (i = 0; i < NR_HASH; i++) {
sh = conf->stripe_hashtbl[i];
for (; sh; sh = sh->hash_next) {
......@@ -1660,13 +1729,13 @@ static void printall (raid5_conf_t *conf)
print_sh(sh);
}
}
md_spin_unlock_irq(&conf->device_lock);
spin_unlock_irq(&conf->device_lock);
PRINTK("--- raid5d inactive\n");
}
#endif
static int raid5_status (char *page, mddev_t *mddev)
static int status (char *page, mddev_t *mddev)
{
raid5_conf_t *conf = (raid5_conf_t *) mddev->private;
mdp_super_t *sb = mddev->sb;
......@@ -1711,7 +1780,7 @@ static void print_raid5_conf (raid5_conf_t *conf)
}
}
static int raid5_diskop(mddev_t *mddev, mdp_disk_t **d, int state)
static int diskop(mddev_t *mddev, mdp_disk_t **d, int state)
{
int err = 0;
int i, failed_disk=-1, spare_disk=-1, removed_disk=-1, added_disk=-1;
......@@ -1722,7 +1791,7 @@ static int raid5_diskop(mddev_t *mddev, mdp_disk_t **d, int state)
mdk_rdev_t *spare_rdev, *failed_rdev;
print_raid5_conf(conf);
md_spin_lock_irq(&conf->device_lock);
spin_lock_irq(&conf->device_lock);
/*
* find the disk ...
*/
......@@ -1948,6 +2017,7 @@ static int raid5_diskop(mddev_t *mddev, mdp_disk_t **d, int state)
goto abort;
}
rdisk->dev = NODEV;
rdisk->bdev = NULL;
rdisk->used_slot = 0;
break;
......@@ -1965,6 +2035,8 @@ static int raid5_diskop(mddev_t *mddev, mdp_disk_t **d, int state)
adisk->number = added_desc->number;
adisk->raid_disk = added_desc->raid_disk;
adisk->dev = mk_kdev(added_desc->major,added_desc->minor);
/* it will be held open by rdev */
adisk->bdev = bdget(kdev_t_to_nr(adisk->dev));
adisk->operational = 0;
adisk->write_only = 0;
......@@ -1980,7 +2052,7 @@ static int raid5_diskop(mddev_t *mddev, mdp_disk_t **d, int state)
goto abort;
}
abort:
md_spin_unlock_irq(&conf->device_lock);
spin_unlock_irq(&conf->device_lock);
print_raid5_conf(conf);
return err;
}
......@@ -1988,18 +2060,18 @@ static int raid5_diskop(mddev_t *mddev, mdp_disk_t **d, int state)
static mdk_personality_t raid5_personality=
{
name: "raid5",
make_request: raid5_make_request,
run: raid5_run,
stop: raid5_stop,
status: raid5_status,
error_handler: raid5_error,
diskop: raid5_diskop,
stop_resync: raid5_stop_resync,
restart_resync: raid5_restart_resync,
sync_request: raid5_sync_request
make_request: make_request,
run: run,
stop: stop,
status: status,
error_handler: error,
diskop: diskop,
stop_resync: stop_resync,
restart_resync: restart_resync,
sync_request: sync_request
};
static int md__init raid5_init (void)
static int __init raid5_init (void)
{
return register_md_personality (RAID5, &raid5_personality);
}
......
......@@ -19,7 +19,6 @@
#define BH_TRACE 0
#include <linux/module.h>
#include <linux/raid/md.h>
#include <linux/raid/md_compatible.h>
#include <linux/raid/xor.h>
#include <asm/xor.h>
......@@ -27,31 +26,30 @@
static struct xor_block_template *active_template;
void
xor_block(unsigned int count, struct buffer_head **bh_ptr)
xor_block(unsigned int count, unsigned int bytes, void **ptr)
{
unsigned long *p0, *p1, *p2, *p3, *p4;
unsigned long bytes = bh_ptr[0]->b_size;
p0 = (unsigned long *) bh_ptr[0]->b_data;
p1 = (unsigned long *) bh_ptr[1]->b_data;
p0 = (unsigned long *) ptr[0];
p1 = (unsigned long *) ptr[1];
if (count == 2) {
active_template->do_2(bytes, p0, p1);
return;
}
p2 = (unsigned long *) bh_ptr[2]->b_data;
p2 = (unsigned long *) ptr[2];
if (count == 3) {
active_template->do_3(bytes, p0, p1, p2);
return;
}
p3 = (unsigned long *) bh_ptr[3]->b_data;
p3 = (unsigned long *) ptr[3];
if (count == 4) {
active_template->do_4(bytes, p0, p1, p2, p3);
return;
}
p4 = (unsigned long *) bh_ptr[4]->b_data;
p4 = (unsigned long *) ptr[4];
active_template->do_5(bytes, p0, p1, p2, p3, p4);
}
......@@ -103,7 +101,7 @@ calibrate_xor_block(void)
void *b1, *b2;
struct xor_block_template *f, *fastest;
b1 = (void *) md__get_free_pages(GFP_KERNEL, 2);
b1 = (void *) __get_free_pages(GFP_KERNEL, 2);
if (! b1) {
printk("raid5: Yikes! No memory available.\n");
return -ENOMEM;
......@@ -137,7 +135,7 @@ calibrate_xor_block(void)
return 0;
}
MD_EXPORT_SYMBOL(xor_block);
EXPORT_SYMBOL(xor_block);
MODULE_LICENSE("GPL");
module_init(calibrate_xor_block);
......@@ -105,9 +105,8 @@ int presto_set_ext_attr(struct inode *inode,
printk("InterMezzo: out of memory!!!\n");
return -ENOMEM;
}
error = copy_from_user(buf, buffer, buffer_len);
if (error)
return error;
if (copy_from_user(buf, buffer, buffer_len))
return -EFAULT;
} else
buf = buffer;
} else
......
......@@ -31,10 +31,9 @@ int begin_kml_reint (struct file *file, unsigned long arg)
ENTRY;
/* allocate buffer & copy it to kernel space */
error = copy_from_user(&input, (char *)arg, sizeof(input));
if ( error ) {
if (copy_from_user(&input, (char *)arg, sizeof(input))) {
EXIT;
return error;
return -EFAULT;
}
if (input.reclen > kml_fsdata->kml_maxsize)
......@@ -45,11 +44,10 @@ int begin_kml_reint (struct file *file, unsigned long arg)
EXIT;
return -ENOMEM;
}
error = copy_from_user(path, input.volname, input.namelen);
if ( error ) {
if (copy_from_user(path, input.volname, input.namelen)) {
PRESTO_FREE(path, input.namelen + 1);
EXIT;
return error;
return -EFAULT;
}
path[input.namelen] = '\0';
fset = kml_getfset (path);
......@@ -57,10 +55,9 @@ int begin_kml_reint (struct file *file, unsigned long arg)
kml_fsdata = FSET_GET_KMLDATA(fset);
/* read the buf from user memory here */
error = copy_from_user(kml_fsdata->kml_buf, input.recbuf, input.reclen);
if ( error ) {
if (copy_from_user(kml_fsdata->kml_buf, input.recbuf, input.reclen)) {
EXIT;
return error;
return -EFAULT;
}
kml_fsdata->kml_len = input.reclen;
......@@ -94,21 +91,19 @@ int do_kml_reint (struct file *file, unsigned long arg)
struct presto_file_set *fset;
ENTRY;
error = copy_from_user(&input, (char *)arg, sizeof(input));
if ( error ) {
if (copy_from_user(&input, (char *)arg, sizeof(input))) {
EXIT;
return error;
return -EFAULT;
}
PRESTO_ALLOC(path, char *, input.namelen + 1);
if ( !path ) {
EXIT;
return -ENOMEM;
}
error = copy_from_user(path, input.volname, input.namelen);
if ( error ) {
if (copy_from_user(path, input.volname, input.namelen)) {
PRESTO_FREE(path, input.namelen + 1);
EXIT;
return error;
return -EFAULT;
}
path[input.namelen] = '\0';
fset = kml_getfset (path);
......@@ -138,7 +133,8 @@ int do_kml_reint (struct file *file, unsigned long arg)
strlen (close->path) + 1, input.pathlen);
error = -ENOMEM;
}
copy_to_user((char *)arg, &input, sizeof (input));
if (copy_to_user((char *)arg, &input, sizeof (input)))
return -EFAULT;
}
return error;
}
......@@ -161,10 +157,9 @@ int end_kml_reint (struct file *file, unsigned long arg)
char *path;
ENTRY;
error = copy_from_user(&input, (char *)arg, sizeof(input));
if ( error ) {
if (copy_from_user(&input, (char *)arg, sizeof(input))) {
EXIT;
return error;
return -EFAULT;
}
PRESTO_ALLOC(path, char *, input.namelen + 1);
......@@ -172,11 +167,11 @@ int end_kml_reint (struct file *file, unsigned long arg)
EXIT;
return -ENOMEM;
}
error = copy_from_user(path, input.volname, input.namelen);
if (copy_from_user(path, input.volname, input.namelen)) {
if ( error ) {
PRESTO_FREE(path, input.namelen + 1);
EXIT;
return error;
return -EFAULT;
}
path[input.namelen] = '\0';
fset = kml_getfset (path);
......@@ -193,7 +188,8 @@ int end_kml_reint (struct file *file, unsigned long arg)
#if 0
input.newpos = kml_upc->newpos;
input.count = kml_upc->count;
copy_to_user((char *)arg, &input, sizeof (input));
if (copy_to_user((char *)arg, &input, sizeof (input)))
return -EFAULT;
#endif
return error;
}
......@@ -149,9 +149,8 @@ static ssize_t presto_psdev_write(struct file *file, const char *buf,
return -EINVAL;
}
error = copy_from_user(&hdr, buf, sizeof(hdr));
if ( error )
return error;
if (copy_from_user(&hdr, buf, sizeof(hdr)))
return -EFAULT;
CDEBUG(D_PSDEV, "(process,opc,uniq)=(%d,%d,%d)\n",
current->pid, hdr.opcode, hdr.unique);
......@@ -183,9 +182,8 @@ static ssize_t presto_psdev_write(struct file *file, const char *buf,
req->rq_bufsize, count, hdr.opcode, hdr.unique);
count = req->rq_bufsize; /* don't have more space! */
}
error = copy_from_user(req->rq_data, buf, count);
if ( error )
return error;
if (copy_from_user(req->rq_data, buf, count))
return -EFAULT;
/* adjust outsize: good upcalls can be aware of this */
req->rq_rep_size = count;
......@@ -280,14 +278,12 @@ static int presto_psdev_ioctl(struct inode *inode, struct file *file,
char * tmp;
int error;
error = copy_from_user(&readmount, (void *)arg,
sizeof(readmount));
if ( error ) {
if (copy_from_user(&readmount, (void *)arg, sizeof(readmount)))
printk("psdev: can't copy %Zd bytes from %p to %p\n",
sizeof(readmount), (struct readmount *) arg,
&readmount);
EXIT;
return error;
return -EFAULT;
}
len = readmount.io_len;
......@@ -307,15 +303,16 @@ static int presto_psdev_ioctl(struct inode *inode, struct file *file,
* I mean, let's let the compiler do a little work ...
* gcc suggested the extra ()
*/
error = copy_to_user(readmount.io_string, tmp, outlen);
if ( error ) {
if (copy_to_user(readmount.io_string, tmp, outlen)) {
CDEBUG(D_PSDEV, "Copy_to_user string 0x%p failed\n",
readmount.io_string);
error = -EFAULT;
}
if ((!error) && (error = copy_to_user(&(user_readmount->io_len),
&outlen, sizeof(int))) ) {
if (!error && copy_to_user(&(user_readmount->io_len),
&outlen, sizeof(int))) {
CDEBUG(D_PSDEV, "Copy_to_user len @0x%p failed\n",
&(user_readmount->io_len));
error = -EFAULT;
}
PRESTO_FREE(tmp, len);
......@@ -360,10 +357,9 @@ static int presto_psdev_ioctl(struct inode *inode, struct file *file,
int path_len;
} input;
error = copy_from_user(&input, (char *)arg, sizeof(input));
if ( error ) {
if (copy_from_user(&input, (char *)arg, sizeof(input))) {
EXIT;
return error;
return -EFAULT;
}
PRESTO_ALLOC(path, char *, input.path_len + 1);
......@@ -371,11 +367,10 @@ static int presto_psdev_ioctl(struct inode *inode, struct file *file,
EXIT;
return -ENOMEM;
}
error = copy_from_user(path, input.path, input.path_len);
if ( error ) {
if (copy_from_user(path, input.path, input.path_len)) {
PRESTO_FREE(path, input.path_len + 1);
EXIT;
return error;
return -EFAULT;
}
path[input.path_len] = '\0';
CDEBUG(D_PSDEV, "clear_fsetroot: path %s\n", path);
......@@ -401,10 +396,9 @@ static int presto_psdev_ioctl(struct inode *inode, struct file *file,
int path_len;
} input;
error = copy_from_user(&input, (char *)arg, sizeof(input));
if ( error ) {
if (copy_from_user(&input, (char *)arg, sizeof(input))) {
EXIT;
return error;
return -EFAULT;
}
PRESTO_ALLOC(path, char *, input.path_len + 1);
......@@ -412,11 +406,10 @@ static int presto_psdev_ioctl(struct inode *inode, struct file *file,
EXIT;
return -ENOMEM;
}
error = copy_from_user(path, input.path, input.path_len);
if ( error ) {
if (copy_from_user(path, input.path, input.path_len)) {
PRESTO_FREE(path, input.path_len + 1);
EXIT;
return error;
return -EFAULT;
}
path[input.path_len] = '\0';
CDEBUG(D_PSDEV, "clear_all_fsetroot: path %s\n", path);
......@@ -440,10 +433,9 @@ static int presto_psdev_ioctl(struct inode *inode, struct file *file,
int path_len;
} input;
error = copy_from_user(&input, (char *)arg, sizeof(input));
if ( error ) {
if (copy_from_user(&input, (char *)arg, sizeof(input))) {
EXIT;
return error;
return -EFAULT;
}
PRESTO_ALLOC(path, char *, input.path_len + 1);
......@@ -451,11 +443,10 @@ static int presto_psdev_ioctl(struct inode *inode, struct file *file,
EXIT;
return -ENOMEM;
}
error = copy_from_user(path, input.path, input.path_len);
if ( error ) {
if (copy_from_user(path, input.path, input.path_len)) {
PRESTO_FREE(path, input.path_len + 1);
EXIT;
return error;
return -EFAULT;
}
path[input.path_len] = '\0';
CDEBUG(D_PSDEV, "get_kmlsize: len %d path %s\n",
......@@ -474,7 +465,9 @@ static int presto_psdev_ioctl(struct inode *inode, struct file *file,
CDEBUG(D_PSDEV, "get_kmlsize: size = %Zd\n", size);
EXIT;
return copy_to_user((char *)arg, &input, sizeof(input));
if (copy_to_user((char *)arg, &input, sizeof(input)))
return -EFAULT;
return 0;
}
case PRESTO_GET_RECNO: {
......@@ -488,10 +481,9 @@ static int presto_psdev_ioctl(struct inode *inode, struct file *file,
int path_len;
} input;
error = copy_from_user(&input, (char *)arg, sizeof(input));
if ( error ) {
if (copy_from_user(&input, (char *)arg, sizeof(input))) {
EXIT;
return error;
return -EFAULT;
}
PRESTO_ALLOC(path, char *, input.path_len + 1);
......@@ -499,11 +491,10 @@ static int presto_psdev_ioctl(struct inode *inode, struct file *file,
EXIT;
return -ENOMEM;
}
error = copy_from_user(path, input.path, input.path_len);
if ( error ) {
if (copy_from_user(path, input.path, input.path_len)) {
PRESTO_FREE(path, input.path_len + 1);
EXIT;
return error;
return -EFAULT;
}
path[input.path_len] = '\0';
CDEBUG(D_PSDEV, "get_recno: len %d path %s\n",
......@@ -522,7 +513,9 @@ static int presto_psdev_ioctl(struct inode *inode, struct file *file,
CDEBUG(D_PSDEV, "get_recno: recno = %d\n", (int) recno);
EXIT;
return copy_to_user((char *)arg, &input, sizeof(input));
if (copy_to_user((char *)arg, &input, sizeof(input)))
return -EFAULT;
return 0;
}
case PRESTO_SET_FSETROOT: {
......@@ -543,10 +536,9 @@ static int presto_psdev_ioctl(struct inode *inode, struct file *file,
int flags;
} input;
error = copy_from_user(&input, (char *)arg, sizeof(input));
if ( error ) {
if (copy_from_user(&input, (char *)arg, sizeof(input))) {
EXIT;
return error;
return -EFAULT;
}
PRESTO_ALLOC(path, char *, input.path_len + 1);
......@@ -554,9 +546,9 @@ static int presto_psdev_ioctl(struct inode *inode, struct file *file,
EXIT;
return -ENOMEM;
}
error = copy_from_user(path, input.path, input.path_len);
if ( error ) {
if (copy_from_user(path, input.path, input.path_len)) {
EXIT;
error -EFAULT;
goto exit_free_path;
}
path[input.path_len] = '\0';
......@@ -567,9 +559,9 @@ static int presto_psdev_ioctl(struct inode *inode, struct file *file,
EXIT;
goto exit_free_path;
}
error = copy_from_user(fsetname, input.name, input.name_len);
if ( error ) {
if (copy_from_user(fsetname, input.name, input.name_len)) {
EXIT;
error = -EFAULT;
goto exit_free_fsetname;
}
fsetname[input.name_len] = '\0';
......@@ -621,12 +613,11 @@ static int presto_psdev_ioctl(struct inode *inode, struct file *file,
struct psdev_opt *user_opt = (struct psdev_opt *) arg;
int error;
error = copy_from_user(&kopt, (void *)arg, sizeof(kopt));
if ( error ) {
if (copy_from_user(&kopt, (void *)arg, sizeof(kopt))) {
printk("psdev: can't copyin %Zd bytes from %p to %p\n",
sizeof(kopt), (struct kopt *) arg, &kopt);
EXIT;
return error;
return -EFAULT;
}
minor = minor(dev);
if (cmd == PRESTO_SETOPT)
......@@ -650,12 +641,11 @@ static int presto_psdev_ioctl(struct inode *inode, struct file *file,
return error;
}
error = copy_to_user(user_opt, &kopt, sizeof(kopt));
if ( error ) {
if (copy_to_user(user_opt, &kopt, sizeof(kopt))) {
CDEBUG(D_PSDEV, "Copy_to_user opt 0x%p failed\n",
user_opt);
EXIT;
return error;
return -EFAULT;
}
CDEBUG(D_PSDEV, "dosetopt minor %d, opt %d, val %d return %d\n",
minor, kopt.optname, kopt.optval, error);
......@@ -668,10 +658,9 @@ static int presto_psdev_ioctl(struct inode *inode, struct file *file,
struct lento_input_attr input;
struct iattr iattr;
error = copy_from_user(&input, (char *)arg, sizeof(input));
if ( error ) {
if (copy_from_user(&input, (char *)arg, sizeof(input))) {
EXIT;
return error;
return -EFAULT;
}
iattr.ia_valid = input.valid;
iattr.ia_mode = (umode_t)input.mode;
......@@ -692,10 +681,9 @@ static int presto_psdev_ioctl(struct inode *inode, struct file *file,
int error;
struct lento_input_mode input;
error = copy_from_user(&input, (char *)arg, sizeof(input));
if ( error ) {
if (copy_from_user(&input, (char *)arg, sizeof(input))) {
EXIT;
return error;
return -EFAULT;
}
error = lento_create(input.name, input.mode, &input.info);
......@@ -707,10 +695,9 @@ static int presto_psdev_ioctl(struct inode *inode, struct file *file,
int error;
struct lento_input_old_new input;
error = copy_from_user(&input, (char *)arg, sizeof(input));
if ( error ) {
if (copy_from_user(&input, (char *)arg, sizeof(input))) {
EXIT;
return error;
return -EFAULT;
}
error = lento_link(input.oldname, input.newname, &input.info);
......@@ -722,10 +709,9 @@ static int presto_psdev_ioctl(struct inode *inode, struct file *file,
int error;
struct lento_input input;
error = copy_from_user(&input, (char *)arg, sizeof(input));
if ( error ) {
if (copy_from_user(&input, (char *)arg, sizeof(input))) {
EXIT;
return error;
return -EFAULT;
}
error = lento_unlink(input.name, &input.info);
......@@ -737,10 +723,9 @@ static int presto_psdev_ioctl(struct inode *inode, struct file *file,
int error;
struct lento_input_old_new input;
error = copy_from_user(&input, (char *)arg, sizeof(input));
if ( error ) {
if (copy_from_user(&input, (char *)arg, sizeof(input))) {
EXIT;
return error;
return -EFAULT;
}
error = lento_symlink(input.oldname, input.newname,&input.info);
......@@ -752,10 +737,9 @@ static int presto_psdev_ioctl(struct inode *inode, struct file *file,
int error;
struct lento_input_mode input;
error = copy_from_user(&input, (char *)arg, sizeof(input));
if ( error ) {
if (copy_from_user(&input, (char *)arg, sizeof(input))) {
EXIT;
return error;
return -EFAULT;
}
error = lento_mkdir(input.name, input.mode, &input.info);
......@@ -767,10 +751,9 @@ static int presto_psdev_ioctl(struct inode *inode, struct file *file,
int error;
struct lento_input input;
error = copy_from_user(&input, (char *)arg, sizeof(input));
if ( error ) {
if (copy_from_user(&input, (char *)arg, sizeof(input))) {
EXIT;
return error;
return -EFAULT;
}
error = lento_rmdir(input.name, &input.info);
......@@ -782,10 +765,9 @@ static int presto_psdev_ioctl(struct inode *inode, struct file *file,
int error;
struct lento_input_dev input;
error = copy_from_user(&input, (char *)arg, sizeof(input));
if ( error ) {
if (copy_from_user(&input, (char *)arg, sizeof(input))) {
EXIT;
return error;
return -EFAULT;
}
error = lento_mknod(input.name, input.mode,
......@@ -798,10 +780,9 @@ static int presto_psdev_ioctl(struct inode *inode, struct file *file,
int error;
struct lento_input_old_new input;
error = copy_from_user(&input, (char *)arg, sizeof(input));
if ( error ) {
if (copy_from_user(&input, (char *)arg, sizeof(input))) {
EXIT;
return error;
return -EFAULT;
}
error = lento_rename(input.oldname, input.newname, &input.info);
......@@ -817,30 +798,27 @@ static int presto_psdev_ioctl(struct inode *inode, struct file *file,
char *name;
char *buffer;
error = copy_from_user(&input, (char *)arg, sizeof(input));
if ( error ) {
if (copy_from_user(&input, (char *)arg, sizeof(input))) {
EXIT;
return error;
return -EFAULT;
}
/* Now setup the input parameters */
PRESTO_ALLOC(name, char *, input.name_len+1);
/* We need null terminated strings for attr names */
name[input.name_len] = '\0';
error=copy_from_user(name, input.name, input.name_len);
if ( error ) {
if (copy_from_user(name, input.name, input.name_len)) {
EXIT;
PRESTO_FREE(name,input.name_len+1);
return error;
return -EFAULT;
}
PRESTO_ALLOC(buffer, char *, input.buffer_len+1);
error=copy_from_user(buffer, input.buffer, input.buffer_len);
if ( error ) {
if (copy_from_user(buffer, input.buffer, input.buffer_len)) {
EXIT;
PRESTO_FREE(name,input.name_len+1);
PRESTO_FREE(buffer,input.buffer_len+1);
return error;
return -EFAULT;
}
/* Make null terminated for easy printing */
buffer[input.buffer_len]='\0';
......@@ -869,21 +847,19 @@ static int presto_psdev_ioctl(struct inode *inode, struct file *file,
struct lento_input_ext_attr input;
char *name;
error = copy_from_user(&input, (char *)arg, sizeof(input));
if ( error ) {
if (copy_from_user(&input, (char *)arg, sizeof(input))) {
EXIT;
return error;
return -EFAULT;
}
/* Now setup the input parameters */
PRESTO_ALLOC(name, char *, input.name_len+1);
/* We need null terminated strings for attr names */
name[input.name_len] = '\0';
error=copy_from_user(name, input.name, input.name_len);
if ( error ) {
if (copy_from_user(name, input.name, input.name_len)) {
EXIT;
PRESTO_FREE(name,input.name_len+1);
return error;
return -EFAULT;
}
CDEBUG(D_PSDEV," delextattr params: name %s,"
......@@ -907,10 +883,9 @@ static int presto_psdev_ioctl(struct inode *inode, struct file *file,
struct lento_input_iopen input;
int error;
error = copy_from_user(&input, (char *)arg, sizeof(input));
if ( error ) {
if (copy_from_user(&input, (char *)arg, sizeof(input))) {
EXIT;
return error;
return -EFAULT;
}
input.fd = lento_iopen(input.name, (ino_t)input.ino,
......@@ -921,17 +896,18 @@ static int presto_psdev_ioctl(struct inode *inode, struct file *file,
return input.fd;
}
EXIT;
return copy_to_user((char *)arg, &input, sizeof(input));
if (copy_to_user((char *)arg, &input, sizeof(input)))
return -EFAULT;
return 0;
}
case PRESTO_VFS_CLOSE: {
int error;
struct lento_input_close input;
error = copy_from_user(&input, (char *)arg, sizeof(input));
if ( error ) {
if (copy_from_user(&input, (char *)arg, sizeof(input))) {
EXIT;
return error;
return -EFAULT;
}
CDEBUG(D_PIOCTL, "lento_close file descriptor: %d\n", input.fd);
......@@ -952,10 +928,9 @@ static int presto_psdev_ioctl(struct inode *inode, struct file *file,
struct presto_version remote_file_version;
} input;
error = copy_from_user(&input, (char *)arg, sizeof(input));
if ( error ) {
if (copy_from_user(&input, (char *)arg, sizeof(input))) {
EXIT;
return error;
return -EFAULT;
}
user_path = input.path;
......@@ -964,11 +939,10 @@ static int presto_psdev_ioctl(struct inode *inode, struct file *file,
EXIT;
return -ENOMEM;
}
error = copy_from_user(input.path, user_path, input.path_len);
if ( error ) {
if (copy_from_user(input.path, user_path, input.path_len)) {
EXIT;
PRESTO_FREE(input.path, input.path_len + 1);
return error;
return -EFAULT;
}
input.path[input.path_len] = '\0';
......@@ -996,10 +970,9 @@ static int presto_psdev_ioctl(struct inode *inode, struct file *file,
struct lento_vfs_context info;
} input;
error = copy_from_user(&input, (char *)arg, sizeof(input));
if ( error ) {
if (copy_from_user(&input, (char *)arg, sizeof(input))) {
EXIT;
return error;
return -EFAULT;
}
user_path = input.path;
......@@ -1008,11 +981,10 @@ static int presto_psdev_ioctl(struct inode *inode, struct file *file,
EXIT;
return -ENOMEM;
}
error = copy_from_user(input.path, user_path, input.path_len);
if ( error ) {
if (copy_from_user(input.path, user_path, input.path_len)) {
EXIT;
PRESTO_FREE(input.path, input.path_len + 1);
return error;
return -EFAULT;
}
input.path[input.path_len] = '\0';
......@@ -1035,10 +1007,9 @@ static int presto_psdev_ioctl(struct inode *inode, struct file *file,
__u32 path_len;
} input;
error = copy_from_user(&input, (char *)arg, sizeof(input));
if ( error ) {
if (copy_from_user(&input, (char *)arg, sizeof(input))) {
EXIT;
return error;
return -EFAULT;
}
user_path = input.path;
......@@ -1047,11 +1018,10 @@ static int presto_psdev_ioctl(struct inode *inode, struct file *file,
EXIT;
return -ENOMEM;
}
error = copy_from_user(input.path, user_path, input.path_len);
if ( error ) {
if (copy_from_user(input.path, user_path, input.path_len)) {
EXIT;
PRESTO_FREE(input.path, input.path_len + 1);
return error;
return -EFAULT;
}
input.path[input.path_len] = '\0';
......@@ -1072,10 +1042,9 @@ static int presto_psdev_ioctl(struct inode *inode, struct file *file,
__u32 recno;
} input;
error = copy_from_user(&input, (char *)arg, sizeof(input));
if ( error ) {
if (copy_from_user(&input, (char *)arg, sizeof(input))) {
EXIT;
return error;
return -EFAULT;
}
user_path = input.path;
......@@ -1084,11 +1053,10 @@ static int presto_psdev_ioctl(struct inode *inode, struct file *file,
EXIT;
return -ENOMEM;
}
error = copy_from_user(input.path, user_path, input.path_len);
if ( error ) {
if (copy_from_user(input.path, user_path, input.path_len)) {
EXIT;
PRESTO_FREE(input.path, input.path_len + 1);
return error;
return -EFAULT;
}
input.path[input.path_len] = '\0';
......@@ -1111,10 +1079,9 @@ static int presto_psdev_ioctl(struct inode *inode, struct file *file,
char *path;
} input;
error = copy_from_user(&input, (char *)arg, sizeof(input));
if ( error ) {
if (copy_from_user(&input, (char *)arg, sizeof(input))) {
EXIT;
return error;
return -EFAULT;
}
user_path = input.path;
......@@ -1123,11 +1090,10 @@ static int presto_psdev_ioctl(struct inode *inode, struct file *file,
EXIT;
return -ENOMEM;
}
error = copy_from_user(input.path, user_path, input.path_len);
if ( error ) {
if (copy_from_user(input.path, user_path, input.path_len)) {
EXIT;
PRESTO_FREE(input.path, input.path_len + 1);
return error;
return -EFAULT;
}
input.path[input.path_len] = '\0';
......@@ -1190,7 +1156,9 @@ static int presto_psdev_ioctl(struct inode *inode, struct file *file,
}
/* return the correct cookie to wait for */
input.mark_what = res;
return copy_to_user((char *)arg, &input, sizeof(input));
if (copy_to_user((char *)arg, &input, sizeof(input)))
return -EFAULT;
return 0;
}
#ifdef CONFIG_KREINT
......@@ -1211,11 +1179,10 @@ static int presto_psdev_ioctl(struct inode *inode, struct file *file,
char *path;
} permit;
error = copy_from_user(&permit, (char *)arg, sizeof(permit));
if ( error ) {
if (copy_from_user(&permit, (char *)arg, sizeof(permit))) {
EXIT;
return error;
}
return -EFAULT;
}
user_path = permit.path;
PRESTO_ALLOC(permit.path, char *, permit.path_len + 1);
......@@ -1223,11 +1190,10 @@ static int presto_psdev_ioctl(struct inode *inode, struct file *file,
EXIT;
return -ENOMEM;
}
error = copy_from_user(permit.path, user_path, permit.path_len);
if ( error ) {
if (copy_from_user(permit.path, user_path, permit.path_len)) {
EXIT;
PRESTO_FREE(permit.path, permit.path_len + 1);
return error;
return -EFAULT;
}
permit.path[permit.path_len] = '\0';
......@@ -1241,7 +1207,9 @@ static int presto_psdev_ioctl(struct inode *inode, struct file *file,
return error;
}
/* return the correct cookie to wait for */
return copy_to_user((char *)arg, &permit, sizeof(permit));
if (copy_to_user((char *)arg, &permit, sizeof(permit)))
return -EFAULT;
return 0;
}
default:
......
......@@ -315,17 +315,19 @@ exp_export(struct nfsctl_export *nxp)
* 2: We must be able to find an inode from a filehandle.
* This means that s_export_op must be set.
*/
if (((inode->i_sb->s_type->fs_flags & FS_REQUIRES_DEV)
|| (nxp->ex_flags & NFSEXP_FSID))
&&
inode->i_sb->s_export_op)
/* Ok, we can export it */;
else {
if (!(inode->i_sb->s_type->fs_flags & FS_REQUIRES_DEV)) {
if (!(nxp->ex_flags & NFSEXP_FSID)) {
dprintk("exp_export: export of non-dev fs without fsid");
goto finish;
}
}
if (!inode->i_sb->s_export_op) {
dprintk("exp_export: export of invalid fs type.\n");
goto finish;
}
if (inode->i_sb->s_export_op &&
!inode->i_sb->s_export_op->find_exported_dentry)
/* Ok, we can export it */;
if (!inode->i_sb->s_export_op->find_exported_dentry)
inode->i_sb->s_export_op->find_exported_dentry =
find_exported_dentry;
......
......@@ -414,11 +414,13 @@ nfsd_open(struct svc_rqst *rqstp, struct svc_fh *fhp, int type,
{
struct dentry *dentry;
struct inode *inode;
int err;
int flags = O_RDONLY|O_LARGEFILE, mode = FMODE_READ, err;
/* If we get here, then the client has already done an "open", and (hopefully)
* checked permission - so allow OWNER_OVERRIDE in case a chmod has now revoked
* permission */
/*
* If we get here, then the client has already done an "open",
* and (hopefully) checked permission - so allow OWNER_OVERRIDE
* in case a chmod has now revoked permission.
*/
err = fh_verify(rqstp, fhp, type, access | MAY_OWNER_OVERRIDE);
if (err)
goto out;
......@@ -443,37 +445,24 @@ nfsd_open(struct svc_rqst *rqstp, struct svc_fh *fhp, int type,
if (err)
goto out_nfserr;
if ((access & MAY_WRITE) && (err = get_write_access(inode)) != 0)
goto out_nfserr;
memset(filp, 0, sizeof(*filp));
filp->f_op = fops_get(inode->i_fop);
atomic_set(&filp->f_count, 1);
filp->f_dentry = dentry;
filp->f_vfsmnt = fhp->fh_export->ex_mnt;
if (access & MAY_WRITE) {
filp->f_flags = O_WRONLY|O_LARGEFILE;
filp->f_mode = FMODE_WRITE;
err = get_write_access(inode);
if (err)
goto out_nfserr;
flags = O_WRONLY|O_LARGEFILE;
mode = FMODE_WRITE;
DQUOT_INIT(inode);
} else {
filp->f_flags = O_RDONLY|O_LARGEFILE;
filp->f_mode = FMODE_READ;
}
err = 0;
if (filp->f_op && filp->f_op->open) {
err = filp->f_op->open(inode, filp);
if (err) {
fops_put(filp->f_op);
if (access & MAY_WRITE)
put_write_access(inode);
/* I nearly added put_filp() call here, but this filp
* is really on callers stack frame. -DaveM
*/
atomic_dec(&filp->f_count);
}
}
err = init_private_file(filp, dentry, mode);
if (!err) {
filp->f_flags = flags;
filp->f_vfsmnt = fhp->fh_export->ex_mnt;
} else if (access & MAY_WRITE)
put_write_access(inode);
out_nfserr:
if (err)
err = nfserrno(err);
......@@ -490,9 +479,8 @@ nfsd_close(struct file *filp)
struct dentry *dentry = filp->f_dentry;
struct inode *inode = dentry->d_inode;
if (filp->f_op && filp->f_op->release)
if (filp->f_op->release)
filp->f_op->release(inode, filp);
fops_put(filp->f_op);
if (filp->f_mode & FMODE_WRITE)
put_write_access(inode);
}
......
......@@ -162,57 +162,107 @@ asmlinkage long sys_llseek(unsigned int fd, unsigned long offset_high,
}
#endif
asmlinkage ssize_t sys_read(unsigned int fd, char * buf, size_t count)
ssize_t vfs_read(struct file *file, char *buf, size_t count, loff_t *pos)
{
struct inode *inode = file->f_dentry->d_inode;
ssize_t ret;
struct file * file;
ret = -EBADF;
file = fget(fd);
if (file) {
if (file->f_mode & FMODE_READ) {
ret = locks_verify_area(FLOCK_VERIFY_READ, file->f_dentry->d_inode,
file, file->f_pos, count);
if (!ret) {
ssize_t (*read)(struct file *, char *, size_t, loff_t *);
ret = -EINVAL;
if (file->f_op && (read = file->f_op->read) != NULL)
ret = read(file, buf, count, &file->f_pos);
}
}
if (!(file->f_mode & FMODE_READ))
return -EBADF;
if (!file->f_op || !file->f_op->read)
return -EINVAL;
if (pos < 0)
return -EINVAL;
ret = locks_verify_area(FLOCK_VERIFY_READ, inode, file, *pos, count);
if (!ret) {
ret = file->f_op->read(file, buf, count, pos);
if (ret > 0)
dnotify_parent(file->f_dentry, DN_ACCESS);
}
return ret;
}
ssize_t vfs_write(struct file *file, const char *buf, size_t count, loff_t *pos)
{
struct inode *inode = file->f_dentry->d_inode;
ssize_t ret;
if (!(file->f_mode & FMODE_WRITE))
return -EBADF;
if (!file->f_op || !file->f_op->write)
return -EINVAL;
if (pos < 0)
return -EINVAL;
ret = locks_verify_area(FLOCK_VERIFY_WRITE, inode, file, *pos, count);
if (!ret) {
ret = file->f_op->write(file, buf, count, pos);
if (ret > 0)
dnotify_parent(file->f_dentry, DN_MODIFY);
}
return ret;
}
asmlinkage ssize_t sys_read(unsigned int fd, char * buf, size_t count)
{
struct file *file;
ssize_t ret = -EBADF;
file = fget(fd);
if (file) {
ret = vfs_read(file, buf, count, &file->f_pos);
fput(file);
}
return ret;
}
asmlinkage ssize_t sys_write(unsigned int fd, const char * buf, size_t count)
{
ssize_t ret;
struct file * file;
struct file *file;
ssize_t ret = -EBADF;
ret = -EBADF;
file = fget(fd);
if (file) {
if (file->f_mode & FMODE_WRITE) {
struct inode *inode = file->f_dentry->d_inode;
ret = locks_verify_area(FLOCK_VERIFY_WRITE, inode, file,
file->f_pos, count);
if (!ret) {
ssize_t (*write)(struct file *, const char *, size_t, loff_t *);
ret = -EINVAL;
if (file->f_op && (write = file->f_op->write) != NULL)
ret = write(file, buf, count, &file->f_pos);
}
}
if (ret > 0)
dnotify_parent(file->f_dentry, DN_MODIFY);
ret = vfs_write(file, buf, count, &file->f_pos);
fput(file);
}
return ret;
}
asmlinkage ssize_t sys_pread(unsigned int fd, char *buf,
size_t count, loff_t pos)
{
struct file *file;
ssize_t ret = -EBADF;
file = fget(fd);
if (file) {
ret = vfs_read(file, buf, count, &pos);
fput(file);
}
return ret;
}
asmlinkage ssize_t sys_pwrite(unsigned int fd, const char *buf,
size_t count, loff_t pos)
{
struct file *file;
ssize_t ret = -EBADF;
file = fget(fd);
if (file) {
ret = vfs_write(file, buf, count, &pos);
fput(file);
}
return ret;
}
static ssize_t do_readv_writev(int type, struct file *file,
const struct iovec * vector,
......@@ -355,70 +405,3 @@ asmlinkage ssize_t sys_writev(unsigned long fd, const struct iovec * vector,
bad_file:
return ret;
}
/* From the Single Unix Spec: pread & pwrite act like lseek to pos + op +
lseek back to original location. They fail just like lseek does on
non-seekable files. */
asmlinkage ssize_t sys_pread(unsigned int fd, char * buf,
size_t count, loff_t pos)
{
ssize_t ret;
struct file * file;
ssize_t (*read)(struct file *, char *, size_t, loff_t *);
ret = -EBADF;
file = fget(fd);
if (!file)
goto bad_file;
if (!(file->f_mode & FMODE_READ))
goto out;
ret = locks_verify_area(FLOCK_VERIFY_READ, file->f_dentry->d_inode,
file, pos, count);
if (ret)
goto out;
ret = -EINVAL;
if (!file->f_op || !(read = file->f_op->read))
goto out;
if (pos < 0)
goto out;
ret = read(file, buf, count, &pos);
if (ret > 0)
dnotify_parent(file->f_dentry, DN_ACCESS);
out:
fput(file);
bad_file:
return ret;
}
asmlinkage ssize_t sys_pwrite(unsigned int fd, const char * buf,
size_t count, loff_t pos)
{
ssize_t ret;
struct file * file;
ssize_t (*write)(struct file *, const char *, size_t, loff_t *);
ret = -EBADF;
file = fget(fd);
if (!file)
goto bad_file;
if (!(file->f_mode & FMODE_WRITE))
goto out;
ret = locks_verify_area(FLOCK_VERIFY_WRITE, file->f_dentry->d_inode,
file, pos, count);
if (ret)
goto out;
ret = -EINVAL;
if (!file->f_op || !(write = file->f_op->write))
goto out;
if (pos < 0)
goto out;
ret = write(file, buf, count, &pos);
if (ret > 0)
dnotify_parent(file->f_dentry, DN_MODIFY);
out:
fput(file);
bad_file:
return ret;
}
......@@ -1484,13 +1484,19 @@ static int reiserfs_new_symlink (struct reiserfs_transaction_handle *th,
/* inserts the stat data into the tree, and then calls
reiserfs_new_directory (to insert ".", ".." item if new object is
directory) or reiserfs_new_symlink (to insert symlink body if new
object is symlink) or nothing (if new object is regular file) */
struct inode * reiserfs_new_inode (struct reiserfs_transaction_handle *th,
struct inode * dir, int mode,
const char * symname,
int i_size, /* 0 for regular, EMTRY_DIR_SIZE for dirs,
strlen (symname) for symlinks)*/
struct dentry *dentry, struct inode *inode, int * err)
object is symlink) or nothing (if new object is regular file)
NOTE! uid and gid must already be set in the inode. If we return
non-zero due to an error, we have to drop the quota previously allocated
for the fresh inode. This can only be done outside a transaction, so
if we return non-zero, we also end the transaction. */
int reiserfs_new_inode (struct reiserfs_transaction_handle *th,
struct inode * dir, int mode,
const char * symname,
/* 0 for regular, EMTRY_DIR_SIZE for dirs,
strlen (symname) for symlinks)*/
loff_t i_size, struct dentry *dentry,
struct inode *inode)
{
struct super_block * sb;
INITIALIZE_PATH (path_to_key);
......@@ -1498,72 +1504,40 @@ struct inode * reiserfs_new_inode (struct reiserfs_transaction_handle *th,
struct item_head ih;
struct stat_data sd;
int retval;
int err;
if (!dir || !dir->i_nlink) {
*err = -EPERM;
iput(inode) ;
return NULL;
err = -EPERM;
goto out_bad_inode;
}
sb = dir->i_sb;
inode->i_flags = 0;//inode->i_sb->s_flags;
/* item head of new item */
ih.ih_key.k_dir_id = INODE_PKEY (dir)->k_objectid;
ih.ih_key.k_objectid = cpu_to_le32 (reiserfs_get_unused_objectid (th));
if (!ih.ih_key.k_objectid) {
iput(inode) ;
*err = -ENOMEM;
return NULL;
err = -ENOMEM;
goto out_bad_inode ;
}
if (old_format_only (sb))
/* not a perfect generation count, as object ids can be reused, but this
** is as good as reiserfs can do right now.
** note that the private part of inode isn't filled in yet, we have
** to use the directory.
*/
inode->i_generation = le32_to_cpu (INODE_PKEY (dir)->k_objectid);
/* not a perfect generation count, as object ids can be reused, but
** this is as good as reiserfs can do right now.
** note that the private part of inode isn't filled in yet, we have
** to use the directory.
*/
inode->i_generation = le32_to_cpu (INODE_PKEY (dir)->k_objectid);
else
#if defined( USE_INODE_GENERATION_COUNTER )
inode->i_generation =
le32_to_cpu( REISERFS_SB(sb) -> s_rs -> s_inode_generation );
inode->i_generation = le32_to_cpu(REISERFS_SB(sb)->s_rs->s_inode_generation);
#else
inode->i_generation = ++event;
inode->i_generation = ++event;
#endif
if (old_format_only (sb))
make_le_item_head (&ih, 0, KEY_FORMAT_3_5, SD_OFFSET, TYPE_STAT_DATA, SD_V1_SIZE, MAX_US_INT);
else
make_le_item_head (&ih, 0, KEY_FORMAT_3_6, SD_OFFSET, TYPE_STAT_DATA, SD_SIZE, MAX_US_INT);
/* key to search for correct place for new stat data */
_make_cpu_key (&key, KEY_FORMAT_3_6, le32_to_cpu (ih.ih_key.k_dir_id),
le32_to_cpu (ih.ih_key.k_objectid), SD_OFFSET, TYPE_STAT_DATA, 3/*key length*/);
/* find proper place for inserting of stat data */
retval = search_item (sb, &key, &path_to_key);
if (retval == IO_ERROR) {
iput (inode);
*err = -EIO;
return NULL;
}
if (retval == ITEM_FOUND) {
pathrelse (&path_to_key);
iput (inode);
*err = -EEXIST;
return NULL;
}
/* fill stat data */
inode->i_mode = mode;
inode->i_nlink = (S_ISDIR (mode) ? 2 : 1);
inode->i_uid = current->fsuid;
if (dir->i_mode & S_ISGID) {
inode->i_gid = dir->i_gid;
if (S_ISDIR(mode))
inode->i_mode |= S_ISGID;
} else
inode->i_gid = current->fsgid;
/* uid and gid must already be set by the caller for quota init */
inode->i_mtime = inode->i_atime = inode->i_ctime = CURRENT_TIME;
inode->i_size = i_size;
......@@ -1578,18 +1552,38 @@ struct inode * reiserfs_new_inode (struct reiserfs_transaction_handle *th,
REISERFS_I(inode)->i_trans_id = 0;
REISERFS_I(inode)->i_trans_index = 0;
if (old_format_only (sb))
make_le_item_head (&ih, 0, KEY_FORMAT_3_5, SD_OFFSET, TYPE_STAT_DATA, SD_V1_SIZE, MAX_US_INT);
else
make_le_item_head (&ih, 0, KEY_FORMAT_3_6, SD_OFFSET, TYPE_STAT_DATA, SD_SIZE, MAX_US_INT);
/* key to search for correct place for new stat data */
_make_cpu_key (&key, KEY_FORMAT_3_6, le32_to_cpu (ih.ih_key.k_dir_id),
le32_to_cpu (ih.ih_key.k_objectid), SD_OFFSET, TYPE_STAT_DATA, 3/*key length*/);
/* find proper place for inserting of stat data */
retval = search_item (sb, &key, &path_to_key);
if (retval == IO_ERROR) {
err = -EIO;
goto out_bad_inode;
}
if (retval == ITEM_FOUND) {
pathrelse (&path_to_key);
err = -EEXIST;
goto out_bad_inode;
}
if (old_format_only (sb)) {
if (inode->i_uid & ~0xffff || inode->i_gid & ~0xffff) {
pathrelse (&path_to_key);
/* i_uid or i_gid is too big to be stored in stat data v3.5 */
iput (inode);
*err = -EINVAL;
return NULL;
err = -EINVAL;
goto out_bad_inode;
}
inode2sd_v1 (&sd, inode);
} else
} else {
inode2sd (&sd, inode);
}
// these do not go to on-disk stat data
inode->i_ino = le32_to_cpu (ih.ih_key.k_objectid);
inode->i_blksize = PAGE_SIZE;
......@@ -1610,10 +1604,9 @@ struct inode * reiserfs_new_inode (struct reiserfs_transaction_handle *th,
/* insert the stat data into the tree */
retval = reiserfs_insert_item (th, &path_to_key, &key, &ih, (char *)(&sd));
if (retval) {
iput (inode);
*err = retval;
err = retval;
reiserfs_check_path(&path_to_key) ;
return NULL;
goto out_bad_inode;
}
if (S_ISDIR(mode)) {
......@@ -1628,19 +1621,35 @@ struct inode * reiserfs_new_inode (struct reiserfs_transaction_handle *th,
retval = reiserfs_new_symlink (th, &ih, &path_to_key, symname, i_size);
}
if (retval) {
inode->i_nlink = 0;
iput (inode);
*err = retval;
err = retval;
reiserfs_check_path(&path_to_key) ;
return NULL;
journal_end(th, th->t_super, th->t_blocks_allocated);
goto out_inserted_sd;
}
insert_inode_hash (inode);
// we do not mark inode dirty: on disk content matches to the
// in-core one
reiserfs_update_sd(th, inode);
reiserfs_check_path(&path_to_key) ;
return inode;
return 0;
/* it looks like you can easily compress these two goto targets into
* one. Keeping it like this doesn't actually hurt anything, and they
* are place holders for what the quota code actually needs.
*/
out_bad_inode:
/* Invalidate the object, nothing was inserted yet */
INODE_PKEY(inode)->k_objectid = 0;
/* dquot_drop must be done outside a transaction */
journal_end(th, th->t_super, th->t_blocks_allocated) ;
make_bad_inode(inode);
out_inserted_sd:
inode->i_nlink = 0;
th->t_trans_id = 0; /* so the caller can't use this handle later */
iput(inode);
return err;
}
/*
......
......@@ -204,7 +204,7 @@ static int set_bit_in_list_bitmap(struct super_block *p_s_sb, int block,
if (!jb->bitmaps[bmap_nr]) {
jb->bitmaps[bmap_nr] = get_bitmap_node(p_s_sb) ;
}
set_bit(bit_nr, jb->bitmaps[bmap_nr]->data) ;
set_bit(bit_nr, (unsigned long *)jb->bitmaps[bmap_nr]->data) ;
return 0 ;
}
......@@ -550,7 +550,7 @@ int reiserfs_in_journal(struct super_block *p_s_sb,
PROC_INFO_INC( p_s_sb, journal.in_journal_bitmap );
jb = SB_JOURNAL(p_s_sb)->j_list_bitmap + i ;
if (jb->journal_list && jb->bitmaps[bmap_nr] &&
test_bit(bit_nr, jb->bitmaps[bmap_nr]->data)) {
test_bit(bit_nr, (unsigned long *)jb->bitmaps[bmap_nr]->data)) {
tmp_bit = find_next_zero_bit((unsigned long *)
(jb->bitmaps[bmap_nr]->data),
p_s_sb->s_blocksize << 3, bit_nr+1) ;
......
......@@ -248,7 +248,7 @@ static int linear_search_in_dir_item (struct cpu_key * key, struct reiserfs_dir_
/* mark, that this generation number is used */
if (de->de_gen_number_bit_string)
set_bit (GET_GENERATION_NUMBER (deh_offset (deh)), de->de_gen_number_bit_string);
set_bit (GET_GENERATION_NUMBER (deh_offset (deh)), (unsigned long *)de->de_gen_number_bit_string);
// calculate pointer to name and namelen
de->de_entry_num = i;
......@@ -504,7 +504,7 @@ static int reiserfs_add_entry (struct reiserfs_transaction_handle *th, struct in
return -EEXIST;
}
gen_number = find_first_zero_bit (bit_string, MAX_GENERATION_NUMBER + 1);
gen_number = find_first_zero_bit ((unsigned long *)bit_string, MAX_GENERATION_NUMBER + 1);
if (gen_number > MAX_GENERATION_NUMBER) {
/* there is no free generation number */
reiserfs_warning ("reiserfs_add_entry: Congratulations! we have got hash function screwed up\n");
......@@ -552,6 +552,40 @@ static int reiserfs_add_entry (struct reiserfs_transaction_handle *th, struct in
return 0;
}
/* quota utility function, call if you've had to abort after calling
** new_inode_init, and have not called reiserfs_new_inode yet.
** This should only be called on inodes that do not hav stat data
** inserted into the tree yet.
*/
static int drop_new_inode(struct inode *inode) {
make_bad_inode(inode) ;
iput(inode) ;
return 0 ;
}
/* utility function that does setup for reiserfs_new_inode.
** DQUOT_ALLOC_INODE cannot be called inside a transaction, so we had
** to pull some bits of reiserfs_new_inode out into this func.
** Yes, the actual quota calls are missing, they are part of the quota
** patch.
*/
static int new_inode_init(struct inode *inode, struct inode *dir, int mode) {
/* the quota init calls have to know who to charge the quota to, so
** we have to set uid and gid here
*/
inode->i_uid = current->fsuid;
inode->i_mode = mode;
if (dir->i_mode & S_ISGID) {
inode->i_gid = dir->i_gid;
if (S_ISDIR(mode))
inode->i_mode |= S_ISGID;
} else {
inode->i_gid = current->fsgid;
}
return 0 ;
}
//
// a portion of this function, particularly the VFS interface portion,
......@@ -564,7 +598,6 @@ static int reiserfs_create (struct inode * dir, struct dentry *dentry, int mode)
{
int retval;
struct inode * inode;
int windex ;
int jbegin_count = JOURNAL_PER_BALANCE_CNT * 2 ;
struct reiserfs_transaction_handle th ;
......@@ -572,16 +605,16 @@ static int reiserfs_create (struct inode * dir, struct dentry *dentry, int mode)
if (!inode) {
return -ENOMEM ;
}
retval = new_inode_init(inode, dir, mode);
if (retval)
return retval;
lock_kernel();
journal_begin(&th, dir->i_sb, jbegin_count) ;
th.t_caller = "create" ;
windex = push_journal_writer("reiserfs_create") ;
inode = reiserfs_new_inode (&th, dir, mode, 0, 0/*i_size*/, dentry, inode, &retval);
if (!inode) {
pop_journal_writer(windex) ;
journal_end(&th, dir->i_sb, jbegin_count) ;
unlock_kernel();
return retval;
retval = reiserfs_new_inode (&th, dir, mode, 0, 0/*i_size*/, dentry, inode);
if (retval) {
goto out_failed;
}
inode->i_op = &reiserfs_file_inode_operations;
......@@ -593,22 +626,19 @@ static int reiserfs_create (struct inode * dir, struct dentry *dentry, int mode)
if (retval) {
inode->i_nlink--;
reiserfs_update_sd (&th, inode);
pop_journal_writer(windex) ;
// FIXME: should we put iput here and have stat data deleted
// in the same transactioin
journal_end(&th, dir->i_sb, jbegin_count) ;
iput (inode);
unlock_kernel();
return retval;
goto out_failed;
}
reiserfs_update_inode_transaction(inode) ;
reiserfs_update_inode_transaction(dir) ;
d_instantiate(dentry, inode);
pop_journal_writer(windex) ;
journal_end(&th, dir->i_sb, jbegin_count) ;
out_failed:
unlock_kernel();
return 0;
return retval;
}
......@@ -623,7 +653,6 @@ static int reiserfs_mknod (struct inode * dir, struct dentry *dentry, int mode,
{
int retval;
struct inode * inode;
int windex ;
struct reiserfs_transaction_handle th ;
int jbegin_count = JOURNAL_PER_BALANCE_CNT * 3;
......@@ -631,16 +660,16 @@ static int reiserfs_mknod (struct inode * dir, struct dentry *dentry, int mode,
if (!inode) {
return -ENOMEM ;
}
retval = new_inode_init(inode, dir, mode);
if (retval)
return retval;
lock_kernel();
journal_begin(&th, dir->i_sb, jbegin_count) ;
windex = push_journal_writer("reiserfs_mknod") ;
inode = reiserfs_new_inode (&th, dir, mode, 0, 0/*i_size*/, dentry, inode, &retval);
if (!inode) {
pop_journal_writer(windex) ;
journal_end(&th, dir->i_sb, jbegin_count) ;
unlock_kernel();
return retval;
retval = reiserfs_new_inode (&th, dir, mode, 0, 0/*i_size*/, dentry, inode);
if (retval) {
goto out_failed;
}
init_special_inode(inode, mode, rdev) ;
......@@ -656,18 +685,17 @@ static int reiserfs_mknod (struct inode * dir, struct dentry *dentry, int mode,
if (retval) {
inode->i_nlink--;
reiserfs_update_sd (&th, inode);
pop_journal_writer(windex) ;
journal_end(&th, dir->i_sb, jbegin_count) ;
iput (inode);
unlock_kernel();
return retval;
goto out_failed;
}
d_instantiate(dentry, inode);
pop_journal_writer(windex) ;
journal_end(&th, dir->i_sb, jbegin_count) ;
out_failed:
unlock_kernel();
return 0;
return retval;
}
......@@ -682,33 +710,33 @@ static int reiserfs_mkdir (struct inode * dir, struct dentry *dentry, int mode)
{
int retval;
struct inode * inode;
int windex ;
struct reiserfs_transaction_handle th ;
int jbegin_count = JOURNAL_PER_BALANCE_CNT * 3;
mode = S_IFDIR | mode;
inode = new_inode(dir->i_sb) ;
if (!inode) {
return -ENOMEM ;
}
retval = new_inode_init(inode, dir, mode);
if (retval)
return retval;
lock_kernel();
journal_begin(&th, dir->i_sb, jbegin_count) ;
windex = push_journal_writer("reiserfs_mkdir") ;
/* inc the link count now, so another writer doesn't overflow it while
** we sleep later on.
*/
INC_DIR_INODE_NLINK(dir)
mode = S_IFDIR | mode;
inode = reiserfs_new_inode (&th, dir, mode, 0/*symlink*/,
old_format_only (dir->i_sb) ? EMPTY_DIR_SIZE_V1 : EMPTY_DIR_SIZE,
dentry, inode, &retval);
if (!inode) {
pop_journal_writer(windex) ;
retval = reiserfs_new_inode (&th, dir, mode, 0/*symlink*/,
old_format_only (dir->i_sb) ?
EMPTY_DIR_SIZE_V1 : EMPTY_DIR_SIZE,
dentry, inode);
if (retval) {
dir->i_nlink-- ;
journal_end(&th, dir->i_sb, jbegin_count) ;
unlock_kernel();
return retval;
goto out_failed;
}
reiserfs_update_inode_transaction(inode) ;
reiserfs_update_inode_transaction(dir) ;
......@@ -723,21 +751,19 @@ static int reiserfs_mkdir (struct inode * dir, struct dentry *dentry, int mode)
inode->i_nlink = 0;
DEC_DIR_INODE_NLINK(dir);
reiserfs_update_sd (&th, inode);
pop_journal_writer(windex) ;
journal_end(&th, dir->i_sb, jbegin_count) ;
iput (inode);
unlock_kernel();
return retval;
goto out_failed;
}
// the above add_entry did not update dir's stat data
reiserfs_update_sd (&th, dir);
d_instantiate(dentry, inode);
pop_journal_writer(windex) ;
journal_end(&th, dir->i_sb, jbegin_count) ;
out_failed:
unlock_kernel();
return 0;
return retval;
}
static inline int reiserfs_empty_dir(struct inode *inode) {
......@@ -942,43 +968,43 @@ static int reiserfs_symlink (struct inode * dir, struct dentry * dentry, const c
struct inode * inode;
char * name;
int item_len;
int windex ;
struct reiserfs_transaction_handle th ;
int mode = S_IFLNK | S_IRWXUGO;
int jbegin_count = JOURNAL_PER_BALANCE_CNT * 3;
inode = new_inode(dir->i_sb) ;
if (!inode) {
return -ENOMEM ;
}
retval = new_inode_init(inode, dir, mode);
if (retval) {
return retval;
}
lock_kernel();
item_len = ROUND_UP (strlen (symname));
if (item_len > MAX_DIRECT_ITEM_LEN (dir->i_sb->s_blocksize)) {
iput(inode) ;
return -ENAMETOOLONG;
retval = -ENAMETOOLONG;
drop_new_inode(inode);
goto out_failed;
}
lock_kernel();
name = reiserfs_kmalloc (item_len, GFP_NOFS, dir->i_sb);
if (!name) {
iput(inode) ;
unlock_kernel();
return -ENOMEM;
drop_new_inode(inode);
retval = -ENOMEM;
goto out_failed;
}
memcpy (name, symname, strlen (symname));
padd_item (name, item_len, strlen (symname));
journal_begin(&th, dir->i_sb, jbegin_count) ;
windex = push_journal_writer("reiserfs_symlink") ;
inode = reiserfs_new_inode (&th, dir, S_IFLNK | S_IRWXUGO, name, strlen (symname), dentry,
inode, &retval);
retval = reiserfs_new_inode (&th, dir, mode, name, strlen (symname),
dentry, inode);
reiserfs_kfree (name, item_len, dir->i_sb);
if (inode == 0) { /* reiserfs_new_inode iputs for us */
pop_journal_writer(windex) ;
journal_end(&th, dir->i_sb, jbegin_count) ;
unlock_kernel();
return retval;
if (retval) { /* reiserfs_new_inode iputs for us */
goto out_failed;
}
reiserfs_update_inode_transaction(inode) ;
......@@ -996,18 +1022,16 @@ static int reiserfs_symlink (struct inode * dir, struct dentry * dentry, const c
if (retval) {
inode->i_nlink--;
reiserfs_update_sd (&th, inode);
pop_journal_writer(windex) ;
journal_end(&th, dir->i_sb, jbegin_count) ;
iput (inode);
unlock_kernel();
return retval;
goto out_failed;
}
d_instantiate(dentry, inode);
pop_journal_writer(windex) ;
journal_end(&th, dir->i_sb, jbegin_count) ;
out_failed:
unlock_kernel();
return 0;
return retval;
}
......
......@@ -839,6 +839,8 @@ static struct xor_block_template xor_block_pIII_sse = {
/* Also try the generic routines. */
#include <asm-generic/xor.h>
#define cpu_has_mmx (test_bit(X86_FEATURE_MMX, boot_cpu_data.x86_capability))
#undef XOR_TRY_TEMPLATES
#define XOR_TRY_TEMPLATES \
do { \
......@@ -846,7 +848,7 @@ static struct xor_block_template xor_block_pIII_sse = {
xor_speed(&xor_block_32regs); \
if (cpu_has_xmm) \
xor_speed(&xor_block_pIII_sse); \
if (md_cpu_has_mmx()) { \
if (cpu_has_mmx) { \
xor_speed(&xor_block_pII_mmx); \
xor_speed(&xor_block_p5_mmx); \
} \
......
......@@ -759,6 +759,9 @@ struct inode_operations {
struct seq_file;
extern ssize_t vfs_read(struct file *, char *, size_t, loff_t *);
extern ssize_t vfs_write(struct file *, const char *, size_t, loff_t *);
/*
* NOTE: write_inode, delete_inode, clear_inode, put_inode can be called
* without the big kernel lock held in all filesystems.
......
......@@ -169,8 +169,9 @@ struct mdk_rdev_s
struct block_device *bdev; /* block device handle */
mdp_super_t *sb;
unsigned long sb_offset;
struct page *sb_page;
mdp_super_t *sb;
unsigned long sb_offset;
int alias_device; /* device alias to the same disk */
int faulty; /* if faulty do not issue IO requests */
......
......@@ -7,21 +7,21 @@
/*
*
* Each stripe contains one buffer per disc. Each buffer can be in
* one of a number of states determined by bh_state. Changes between
* one of a number of states stored in "flags". Changes between
* these states happen *almost* exclusively under a per-stripe
* spinlock. Some very specific changes can happen in b_end_io, and
* spinlock. Some very specific changes can happen in bi_end_io, and
* these are not protected by the spin lock.
*
* The bh_state bits that are used to represent these states are:
* BH_Uptodate, BH_Lock
* The flag bits that are used to represent these states are:
* R5_UPTODATE and R5_LOCKED
*
* State Empty == !Uptodate, !Lock
* State Empty == !UPTODATE, !LOCK
* We have no data, and there is no active request
* State Want == !Uptodate, Lock
* State Want == !UPTODATE, LOCK
* A read request is being submitted for this block
* State Dirty == Uptodate, Lock
* State Dirty == UPTODATE, LOCK
* Some new data is in this buffer, and it is being written out
* State Clean == Uptodate, !Lock
* State Clean == UPTODATE, !LOCK
* We have valid data which is the same as on disc
*
* The possible state transitions are:
......@@ -124,24 +124,29 @@
* plus raid5d if it is handling it, plus one for each active request
* on a cached buffer.
*/
struct stripe_head {
struct stripe_head *hash_next, **hash_pprev; /* hash pointers */
struct list_head lru; /* inactive_list or handle_list */
struct raid5_private_data *raid_conf;
struct buffer_head *bh_cache[MD_SB_DISKS]; /* buffered copy */
struct buffer_head *bh_read[MD_SB_DISKS]; /* read request buffers of the MD device */
struct buffer_head *bh_write[MD_SB_DISKS]; /* write request buffers of the MD device */
struct buffer_head *bh_written[MD_SB_DISKS]; /* write request buffers of the MD device that have been scheduled for write */
struct page *bh_page[MD_SB_DISKS]; /* saved bh_cache[n]->b_page when reading around the cache */
unsigned long sector; /* sector of this row */
int size; /* buffers size */
sector_t sector; /* sector of this row */
int pd_idx; /* parity disk index */
unsigned long state; /* state flags */
atomic_t count; /* nr of active thread/requests */
spinlock_t lock;
int sync_redone;
struct r5dev {
struct bio req;
struct bio_vec vec;
struct page *page;
struct bio *toread, *towrite, *written;
sector_t sector; /* sector of this page */
unsigned long flags;
} dev[1]; /* allocated with extra space depending of RAID geometry */
};
/* Flags */
#define R5_UPTODATE 0 /* page contains current data */
#define R5_LOCKED 1 /* IO has been submitted on "req" */
#define R5_OVERWRITE 2 /* towrite covers whole page */
/*
* Write method
......@@ -187,6 +192,7 @@ struct stripe_head {
struct disk_info {
kdev_t dev;
struct block_device *bdev;
int operational;
int number;
int raid_disk;
......@@ -201,7 +207,6 @@ struct raid5_private_data {
mdk_thread_t *thread, *resync_thread;
struct disk_info disks[MD_SB_DISKS];
struct disk_info *spare;
int buffer_size;
int chunk_size, level, algorithm;
int raid_disks, working_disks, failed_disks;
int resync_parity;
......@@ -210,16 +215,19 @@ struct raid5_private_data {
struct list_head handle_list; /* stripes needing handling */
struct list_head delayed_list; /* stripes that have plugged requests */
atomic_t preread_active_stripes; /* stripes with scheduled io */
char cache_name[20];
kmem_cache_t *slab_cache; /* for allocating stripes */
/*
* Free stripes pool
*/
atomic_t active_stripes;
struct list_head inactive_list;
md_wait_queue_head_t wait_for_stripe;
wait_queue_head_t wait_for_stripe;
int inactive_blocked; /* release of inactive stripes blocked,
* waiting for 25% to be free
*/
md_spinlock_t device_lock;
spinlock_t device_lock;
int plugged;
struct tq_struct plug_tq;
......
......@@ -5,7 +5,7 @@
#define MAX_XOR_BLOCKS 5
extern void xor_block(unsigned int count, struct buffer_head **bh_ptr);
extern void xor_block(unsigned int count, unsigned int bytes, void **ptr);
struct xor_block_template {
struct xor_block_template *next;
......
......@@ -1841,10 +1841,10 @@ struct inode * reiserfs_iget (struct super_block * s,
const struct cpu_key * key);
struct inode * reiserfs_new_inode (struct reiserfs_transaction_handle *th,
int reiserfs_new_inode (struct reiserfs_transaction_handle *th,
struct inode * dir, int mode,
const char * symname, int item_len,
struct dentry *dentry, struct inode *inode, int * err);
const char * symname, loff_t i_size,
struct dentry *dentry, struct inode *inode);
int reiserfs_sync_inode (struct reiserfs_transaction_handle *th, struct inode * inode);
void reiserfs_update_sd (struct reiserfs_transaction_handle *th, struct inode * inode);
......
......@@ -231,7 +231,7 @@ void daemonize(void)
/*
* When we die, we re-parent all our children.
* Try to give them to another thread in our process
* Try to give them to another thread in our thread
* group, and if no such member exists, give it to
* the global child reaper process (ie "init")
*/
......@@ -241,8 +241,14 @@ static inline void forget_original_parent(struct task_struct * father)
read_lock(&tasklist_lock);
/* Next in our thread group */
reaper = next_thread(father);
/* Next in our thread group, if they're not already exiting */
reaper = father;
do {
reaper = next_thread(reaper);
if (!(reaper->flags & PF_EXITING))
break;
} while (reaper != father);
if (reaper == father)
reaper = child_reaper;
......
......@@ -243,6 +243,8 @@ EXPORT_SYMBOL(shrink_dcache_anon);
EXPORT_SYMBOL(find_inode_number);
EXPORT_SYMBOL(is_subdir);
EXPORT_SYMBOL(get_unused_fd);
EXPORT_SYMBOL(vfs_read);
EXPORT_SYMBOL(vfs_write);
EXPORT_SYMBOL(vfs_create);
EXPORT_SYMBOL(vfs_mkdir);
EXPORT_SYMBOL(vfs_mknod);
......
......@@ -949,12 +949,10 @@ svc_sock_update_bufs(struct svc_serv *serv)
if (sock->type == SOCK_DGRAM) {
/* udp sockets need large rcvbuf as all pending
* requests are still in that buffer.
* As outgoing requests do not wait for an
* ACK, only a moderate sndbuf is needed
*/
svc_sock_setbufsize(sock,
5 * serv->sv_bufsz,
(serv->sv_nrthreads+2)* serv->sv_bufsz);
(serv->sv_nrthreads+3) * serv->sv_bufsz,
(serv->sv_nrthreads+3) * serv->sv_bufsz);
} else if (svsk->sk_sk->state != TCP_LISTEN) {
printk(KERN_ERR "RPC update_bufs: permanent sock neither UDP or TCP_LISTEN\n");
}
......
Markdown is supported
0%
or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment