static int lsm_lmm_verify_v3(struct lov_mds_md *lmmv1, int lmm_bytes,
__u16 *stripe_count)
{
struct lov_mds_md_v3 *lmm;
lmm = (struct lov_mds_md_v3 *)lmmv1;
if (lmm_bytes < sizeof(*lmm)) {
CERROR("lov_mds_md_v3 too small: %d, need at least %d\n",
lmm_bytes, (int)sizeof(*lmm));
return -EINVAL;
}
*stripe_count = le16_to_cpu(lmm->lmm_stripe_count);
if (lmm_bytes < lov_mds_md_size(*stripe_count, LOV_MAGIC_V3)) {
CERROR("LOV EA V3 too small: %d, need %d\n",
lmm_bytes, lov_mds_md_size(*stripe_count, LOV_MAGIC_V3));
lov_dump_lmm_common(D_WARNING, lmm);
return -EINVAL;
}
return lsm_lmm_verify_common((struct lov_mds_md_v1 *)lmm, lmm_bytes,
*stripe_count);
}
static int lov_object_layout_get(const struct lu_env *env,
struct cl_object *obj,
struct cl_layout *cl)
{
struct lov_object *lov = cl2lov(obj);
struct lov_stripe_md *lsm = lov_lsm_addref(lov);
struct lu_buf *buf = &cl->cl_buf;
ssize_t rc;
ENTRY;
if (lsm == NULL) {
cl->cl_size = 0;
cl->cl_layout_gen = CL_LAYOUT_GEN_EMPTY;
RETURN(0);
}
cl->cl_size = lov_mds_md_size(lsm->lsm_stripe_count, lsm->lsm_magic);
cl->cl_layout_gen = lsm->lsm_layout_gen;
rc = lov_lsm_pack(lsm, buf->lb_buf, buf->lb_len);
lov_lsm_put(lsm);
RETURN(rc < 0 ? rc : 0);
}
/**
* Pack LOV striping metadata for disk storage format (in little
* endian byte order).
*
* This follows the getxattr() conventions. If \a buf_size is zero
* then return the size needed. If \a buf_size is too small then
* return -ERANGE. Otherwise return the size of the result.
*/
ssize_t lov_lsm_pack_v1v3(const struct lov_stripe_md *lsm, void *buf,
size_t buf_size)
{
struct lov_mds_md_v1 *lmmv1 = buf;
struct lov_mds_md_v3 *lmmv3 = buf;
struct lov_ost_data_v1 *lmm_objects;
size_t lmm_size;
unsigned int i;
ENTRY;
lmm_size = lov_mds_md_size(lsm->lsm_entries[0]->lsme_stripe_count,
lsm->lsm_magic);
if (buf_size == 0)
RETURN(lmm_size);
if (buf_size < lmm_size)
RETURN(-ERANGE);
/* lmmv1 and lmmv3 point to the same struct and have the
* same first fields
*/
lmmv1->lmm_magic = cpu_to_le32(lsm->lsm_magic);
lmm_oi_cpu_to_le(&lmmv1->lmm_oi, &lsm->lsm_oi);
lmmv1->lmm_stripe_size = cpu_to_le32(
lsm->lsm_entries[0]->lsme_stripe_size);
lmmv1->lmm_stripe_count = cpu_to_le16(
lsm->lsm_entries[0]->lsme_stripe_count);
lmmv1->lmm_pattern = cpu_to_le32(lsm->lsm_entries[0]->lsme_pattern);
lmmv1->lmm_layout_gen = cpu_to_le16(lsm->lsm_layout_gen);
if (lsm->lsm_magic == LOV_MAGIC_V3) {
CLASSERT(sizeof(lsm->lsm_entries[0]->lsme_pool_name) ==
sizeof(lmmv3->lmm_pool_name));
strlcpy(lmmv3->lmm_pool_name,
lsm->lsm_entries[0]->lsme_pool_name,
sizeof(lmmv3->lmm_pool_name));
lmm_objects = lmmv3->lmm_objects;
} else {
lmm_objects = lmmv1->lmm_objects;
}
if (lsm->lsm_is_released)
RETURN(lmm_size);
for (i = 0; i < lsm->lsm_entries[0]->lsme_stripe_count; i++) {
struct lov_oinfo *loi = lsm->lsm_entries[0]->lsme_oinfo[i];
ostid_cpu_to_le(&loi->loi_oi, &lmm_objects[i].l_ost_oi);
lmm_objects[i].l_ost_gen = cpu_to_le32(loi->loi_ost_gen);
lmm_objects[i].l_ost_idx = cpu_to_le32(loi->loi_ost_idx);
}
RETURN(lmm_size);
}
static int lsm_lmm_verify_v1(struct lov_mds_md_v1 *lmm, int lmm_bytes,
__u16 *stripe_count)
{
if (lmm_bytes < sizeof(*lmm)) {
CERROR("lov_mds_md_v1 too small: %d, need at least %d\n",
lmm_bytes, (int)sizeof(*lmm));
return -EINVAL;
}
*stripe_count = le16_to_cpu(lmm->lmm_stripe_count);
if (le32_to_cpu(lmm->lmm_pattern) & LOV_PATTERN_F_RELEASED)
*stripe_count = 0;
if (lmm_bytes < lov_mds_md_size(*stripe_count, LOV_MAGIC_V1)) {
CERROR("LOV EA V1 too small: %d, need %d\n",
lmm_bytes, lov_mds_md_size(*stripe_count, LOV_MAGIC_V1));
lov_dump_lmm_common(D_WARNING, lmm);
return -EINVAL;
}
return lsm_lmm_verify_common(lmm, lmm_bytes, *stripe_count);
}
/**
* Pack LOV striping metadata for disk storage format (in little
* endian byte order).
*
* This follows the getxattr() conventions. If \a buf_size is zero
* then return the size needed. If \a buf_size is too small then
* return -ERANGE. Otherwise return the size of the result.
*/
ssize_t lov_lsm_pack(const struct lov_stripe_md *lsm, void *buf,
size_t buf_size)
{
struct lov_ost_data_v1 *lmm_objects;
struct lov_mds_md_v1 *lmmv1 = buf;
struct lov_mds_md_v3 *lmmv3 = buf;
size_t lmm_size;
unsigned int i;
lmm_size = lov_mds_md_size(lsm->lsm_stripe_count, lsm->lsm_magic);
if (!buf_size)
return lmm_size;
if (buf_size < lmm_size)
return -ERANGE;
/*
* lmmv1 and lmmv3 point to the same struct and have the
* same first fields
*/
lmmv1->lmm_magic = cpu_to_le32(lsm->lsm_magic);
lmm_oi_cpu_to_le(&lmmv1->lmm_oi, &lsm->lsm_oi);
lmmv1->lmm_stripe_size = cpu_to_le32(lsm->lsm_stripe_size);
lmmv1->lmm_stripe_count = cpu_to_le16(lsm->lsm_stripe_count);
lmmv1->lmm_pattern = cpu_to_le32(lsm->lsm_pattern);
lmmv1->lmm_layout_gen = cpu_to_le16(lsm->lsm_layout_gen);
if (lsm->lsm_magic == LOV_MAGIC_V3) {
BUILD_BUG_ON(sizeof(lsm->lsm_pool_name) !=
sizeof(lmmv3->lmm_pool_name));
strlcpy(lmmv3->lmm_pool_name, lsm->lsm_pool_name,
sizeof(lmmv3->lmm_pool_name));
lmm_objects = lmmv3->lmm_objects;
} else {
lmm_objects = lmmv1->lmm_objects;
}
for (i = 0; i < lsm->lsm_stripe_count; i++) {
struct lov_oinfo *loi = lsm->lsm_oinfo[i];
ostid_cpu_to_le(&loi->loi_oi, &lmm_objects[i].l_ost_oi);
lmm_objects[i].l_ost_gen = cpu_to_le32(loi->loi_ost_gen);
lmm_objects[i].l_ost_idx = cpu_to_le32(loi->loi_ost_idx);
}
return lmm_size;
}
ssize_t lov_lsm_pack(const struct lov_stripe_md *lsm, void *buf,
size_t buf_size)
{
struct lov_comp_md_v1 *lcmv1 = buf;
struct lov_comp_md_entry_v1 *lcme;
struct lov_ost_data_v1 *lmm_objects;
size_t lmm_size;
unsigned int entry;
unsigned int offset;
unsigned int size;
unsigned int i;
ENTRY;
if (lsm->lsm_magic == LOV_MAGIC_V1 || lsm->lsm_magic == LOV_MAGIC_V3)
return lov_lsm_pack_v1v3(lsm, buf, buf_size);
lmm_size = lov_comp_md_size(lsm);
if (buf_size == 0)
RETURN(lmm_size);
if (buf_size < lmm_size)
RETURN(-ERANGE);
lcmv1->lcm_magic = cpu_to_le32(lsm->lsm_magic);
lcmv1->lcm_size = cpu_to_le32(lmm_size);
lcmv1->lcm_layout_gen = cpu_to_le32(lsm->lsm_layout_gen);
lcmv1->lcm_flags = cpu_to_le16(lsm->lsm_flags);
lcmv1->lcm_mirror_count = cpu_to_le16(lsm->lsm_mirror_count);
lcmv1->lcm_entry_count = cpu_to_le16(lsm->lsm_entry_count);
offset = sizeof(*lcmv1) + sizeof(*lcme) * lsm->lsm_entry_count;
for (entry = 0; entry < lsm->lsm_entry_count; entry++) {
struct lov_stripe_md_entry *lsme;
struct lov_mds_md *lmm;
__u16 stripe_count;
lsme = lsm->lsm_entries[entry];
lcme = &lcmv1->lcm_entries[entry];
lcme->lcme_id = cpu_to_le32(lsme->lsme_id);
lcme->lcme_flags = cpu_to_le32(lsme->lsme_flags);
lcme->lcme_extent.e_start =
cpu_to_le64(lsme->lsme_extent.e_start);
lcme->lcme_extent.e_end =
cpu_to_le64(lsme->lsme_extent.e_end);
lcme->lcme_offset = cpu_to_le32(offset);
lmm = (struct lov_mds_md *)((char *)lcmv1 + offset);
lmm->lmm_magic = cpu_to_le32(lsme->lsme_magic);
/* lmm->lmm_oi not set */
lmm->lmm_pattern = cpu_to_le32(lsme->lsme_pattern);
lmm->lmm_stripe_size = cpu_to_le32(lsme->lsme_stripe_size);
lmm->lmm_stripe_count = cpu_to_le16(lsme->lsme_stripe_count);
lmm->lmm_layout_gen = cpu_to_le16(lsme->lsme_layout_gen);
if (lsme->lsme_magic == LOV_MAGIC_V3) {
struct lov_mds_md_v3 *lmmv3 =
(struct lov_mds_md_v3 *)lmm;
strlcpy(lmmv3->lmm_pool_name, lsme->lsme_pool_name,
sizeof(lmmv3->lmm_pool_name));
lmm_objects = lmmv3->lmm_objects;
} else {
lmm_objects =
((struct lov_mds_md_v1 *)lmm)->lmm_objects;
}
if (lsme_inited(lsme) &&
!(lsme->lsme_pattern & LOV_PATTERN_F_RELEASED))
stripe_count = lsme->lsme_stripe_count;
else
stripe_count = 0;
for (i = 0; i < stripe_count; i++) {
struct lov_oinfo *loi = lsme->lsme_oinfo[i];
ostid_cpu_to_le(&loi->loi_oi, &lmm_objects[i].l_ost_oi);
lmm_objects[i].l_ost_gen =
cpu_to_le32(loi->loi_ost_gen);
lmm_objects[i].l_ost_idx =
cpu_to_le32(loi->loi_ost_idx);
}
size = lov_mds_md_size(stripe_count, lsme->lsme_magic);
lcme->lcme_size = cpu_to_le32(size);
offset += size;
} /* for each layout component */
RETURN(lmm_size);
}
static int __lov_setstripe(struct obd_export *exp, int max_lmm_size,
struct lov_stripe_md **lsmp,
struct lov_user_md *lump)
{
struct obd_device *obd = class_exp2obd(exp);
struct lov_obd *lov = &obd->u.lov;
char buffer[sizeof(struct lov_user_md_v3)];
struct lov_user_md_v3 *lumv3 = (struct lov_user_md_v3 *)&buffer[0];
struct lov_user_md_v1 *lumv1 = (struct lov_user_md_v1 *)&buffer[0];
int lmm_magic;
__u16 stripe_count;
int rc;
int cplen = 0;
rc = lov_lum_swab_if_needed(lumv3, &lmm_magic, lump);
if (rc)
return rc;
/* in the rest of the tests, as *lumv1 and lumv3 have the same
* fields, we use lumv1 to avoid code duplication */
if (lumv1->lmm_pattern == 0) {
lumv1->lmm_pattern = lov->desc.ld_pattern ?
lov->desc.ld_pattern : LOV_PATTERN_RAID0;
}
if (lov_pattern(lumv1->lmm_pattern) != LOV_PATTERN_RAID0) {
CDEBUG(D_IOCTL, "bad userland stripe pattern: %#x\n",
lumv1->lmm_pattern);
return -EINVAL;
}
/* 64kB is the largest common page size we see (ia64), and matches the
* check in lfs */
if (lumv1->lmm_stripe_size & (LOV_MIN_STRIPE_SIZE - 1)) {
CDEBUG(D_IOCTL, "stripe size %u not multiple of %u, fixing\n",
lumv1->lmm_stripe_size, LOV_MIN_STRIPE_SIZE);
lumv1->lmm_stripe_size = LOV_MIN_STRIPE_SIZE;
}
if ((lumv1->lmm_stripe_offset >= lov->desc.ld_tgt_count) &&
(lumv1->lmm_stripe_offset !=
(typeof(lumv1->lmm_stripe_offset))(-1))) {
CDEBUG(D_IOCTL, "stripe offset %u > number of OSTs %u\n",
lumv1->lmm_stripe_offset, lov->desc.ld_tgt_count);
return -EINVAL;
}
stripe_count = lov_get_stripecnt(lov, lmm_magic,
lumv1->lmm_stripe_count);
if (max_lmm_size) {
int max_stripes = (max_lmm_size -
lov_mds_md_size(0, lmm_magic)) /
sizeof(struct lov_ost_data_v1);
if (unlikely(max_stripes < stripe_count)) {
CDEBUG(D_IOCTL, "stripe count reset from %d to %d\n",
stripe_count, max_stripes);
stripe_count = max_stripes;
}
}
if (lmm_magic == LOV_USER_MAGIC_V3) {
struct pool_desc *pool;
/* In the function below, .hs_keycmp resolves to
* pool_hashkey_keycmp() */
/* coverity[overrun-buffer-val] */
pool = lov_find_pool(lov, lumv3->lmm_pool_name);
if (pool != NULL) {
if (lumv3->lmm_stripe_offset !=
(typeof(lumv3->lmm_stripe_offset))(-1)) {
rc = lov_check_index_in_pool(
lumv3->lmm_stripe_offset, pool);
if (rc < 0) {
lov_pool_putref(pool);
return -EINVAL;
}
}
if (stripe_count > pool_tgt_count(pool))
stripe_count = pool_tgt_count(pool);
lov_pool_putref(pool);
}
}
if (lumv1->lmm_pattern & LOV_PATTERN_F_RELEASED)
stripe_count = 0;
rc = lov_alloc_memmd(lsmp, stripe_count, lumv1->lmm_pattern, lmm_magic);
if (rc >= 0) {
(*lsmp)->lsm_oinfo[0]->loi_ost_idx = lumv1->lmm_stripe_offset;
(*lsmp)->lsm_stripe_size = lumv1->lmm_stripe_size;
if (lmm_magic == LOV_USER_MAGIC_V3) {
cplen = strlcpy((*lsmp)->lsm_pool_name,
lumv3->lmm_pool_name,
sizeof((*lsmp)->lsm_pool_name));
if (cplen >= sizeof((*lsmp)->lsm_pool_name))
rc = -E2BIG;
}
rc = 0;
}
return rc;
}
/* Pack LOV object metadata for disk storage. It is packed in LE byte
* order and is opaque to the networking layer.
*
* XXX In the future, this will be enhanced to get the EA size from the
* underlying OSC device(s) to get their EA sizes so we can stack
* LOVs properly. For now lov_mds_md_size() just assumes one obd_id
* per stripe.
*/
int lov_packmd(struct obd_export *exp, struct lov_mds_md **lmmp,
struct lov_stripe_md *lsm)
{
struct obd_device *obd = class_exp2obd(exp);
struct lov_obd *lov = &obd->u.lov;
struct lov_mds_md_v1 *lmmv1;
struct lov_mds_md_v3 *lmmv3;
__u16 stripe_count;
struct lov_ost_data_v1 *lmm_objects;
int lmm_size, lmm_magic;
int i;
int cplen = 0;
if (lsm) {
lmm_magic = lsm->lsm_magic;
} else {
if (lmmp && *lmmp)
lmm_magic = le32_to_cpu((*lmmp)->lmm_magic);
else
/* lsm == NULL and lmmp == NULL */
lmm_magic = LOV_MAGIC;
}
if ((lmm_magic != LOV_MAGIC_V1) &&
(lmm_magic != LOV_MAGIC_V3)) {
CERROR("bad mem LOV MAGIC: 0x%08X != 0x%08X nor 0x%08X\n",
lmm_magic, LOV_MAGIC_V1, LOV_MAGIC_V3);
return -EINVAL;
}
if (lsm) {
/* If we are just sizing the EA, limit the stripe count
* to the actual number of OSTs in this filesystem. */
if (!lmmp) {
stripe_count = lov_get_stripecnt(lov, lmm_magic,
lsm->lsm_stripe_count);
lsm->lsm_stripe_count = stripe_count;
} else if (!lsm_is_released(lsm)) {
stripe_count = lsm->lsm_stripe_count;
} else {
stripe_count = 0;
}
} else {
/* No need to allocate more than maximum supported stripes.
* Anyway, this is pretty inaccurate since ld_tgt_count now
* represents max index and we should rely on the actual number
* of OSTs instead */
stripe_count = lov_mds_md_max_stripe_count(
lov->lov_ocd.ocd_max_easize, lmm_magic);
if (stripe_count > lov->desc.ld_tgt_count)
stripe_count = lov->desc.ld_tgt_count;
}
/* XXX LOV STACKING call into osc for sizes */
lmm_size = lov_mds_md_size(stripe_count, lmm_magic);
if (!lmmp)
return lmm_size;
if (*lmmp && !lsm) {
stripe_count = le16_to_cpu((*lmmp)->lmm_stripe_count);
lmm_size = lov_mds_md_size(stripe_count, lmm_magic);
OBD_FREE_LARGE(*lmmp, lmm_size);
*lmmp = NULL;
return 0;
}
if (!*lmmp) {
OBD_ALLOC_LARGE(*lmmp, lmm_size);
if (!*lmmp)
return -ENOMEM;
}
CDEBUG(D_INFO, "lov_packmd: LOV_MAGIC 0x%08X, lmm_size = %d \n",
lmm_magic, lmm_size);
lmmv1 = *lmmp;
lmmv3 = (struct lov_mds_md_v3 *)*lmmp;
if (lmm_magic == LOV_MAGIC_V3)
lmmv3->lmm_magic = cpu_to_le32(LOV_MAGIC_V3);
else
lmmv1->lmm_magic = cpu_to_le32(LOV_MAGIC_V1);
if (!lsm)
return lmm_size;
/* lmmv1 and lmmv3 point to the same struct and have the
* same first fields
*/
lmm_oi_cpu_to_le(&lmmv1->lmm_oi, &lsm->lsm_oi);
lmmv1->lmm_stripe_size = cpu_to_le32(lsm->lsm_stripe_size);
lmmv1->lmm_stripe_count = cpu_to_le16(stripe_count);
lmmv1->lmm_pattern = cpu_to_le32(lsm->lsm_pattern);
lmmv1->lmm_layout_gen = cpu_to_le16(lsm->lsm_layout_gen);
if (lsm->lsm_magic == LOV_MAGIC_V3) {
cplen = strlcpy(lmmv3->lmm_pool_name, lsm->lsm_pool_name,
sizeof(lmmv3->lmm_pool_name));
if (cplen >= sizeof(lmmv3->lmm_pool_name))
return -E2BIG;
lmm_objects = lmmv3->lmm_objects;
} else {
lmm_objects = lmmv1->lmm_objects;
}
for (i = 0; i < stripe_count; i++) {
struct lov_oinfo *loi = lsm->lsm_oinfo[i];
/* XXX LOV STACKING call down to osc_packmd() to do packing */
LASSERTF(ostid_id(&loi->loi_oi) != 0, "lmm_oi "DOSTID
" stripe %u/%u idx %u\n", POSTID(&lmmv1->lmm_oi),
i, stripe_count, loi->loi_ost_idx);
ostid_cpu_to_le(&loi->loi_oi, &lmm_objects[i].l_ost_oi);
lmm_objects[i].l_ost_gen = cpu_to_le32(loi->loi_ost_gen);
lmm_objects[i].l_ost_idx = cpu_to_le32(loi->loi_ost_idx);
}
return lmm_size;
}
static int __lov_setstripe(struct obd_export *exp, int max_lmm_size,
struct lov_stripe_md **lsmp,
struct lov_user_md *lump)
{
struct obd_device *obd = class_exp2obd(exp);
struct lov_obd *lov = &obd->u.lov;
char buffer[sizeof(struct lov_user_md_v3)];
struct lov_user_md_v3 *lumv3 = (struct lov_user_md_v3 *)&buffer[0];
struct lov_user_md_v1 *lumv1 = (struct lov_user_md_v1 *)&buffer[0];
int lmm_magic;
int stripe_count;
int rc;
ENTRY;
if (cfs_copy_from_user(lumv3, lump, sizeof(struct lov_user_md_v1)))
RETURN(-EFAULT);
lmm_magic = lumv1->lmm_magic;
if (lmm_magic == __swab32(LOV_USER_MAGIC_V1)) {
lustre_swab_lov_user_md_v1(lumv1);
lmm_magic = LOV_USER_MAGIC_V1;
} else if (lmm_magic == LOV_USER_MAGIC_V3) {
if (cfs_copy_from_user(lumv3, lump, sizeof(*lumv3)))
RETURN(-EFAULT);
} else if (lmm_magic == __swab32(LOV_USER_MAGIC_V3)) {
if (cfs_copy_from_user(lumv3, lump, sizeof(*lumv3)))
RETURN(-EFAULT);
lustre_swab_lov_user_md_v3(lumv3);
lmm_magic = LOV_USER_MAGIC_V3;
} else if (lmm_magic != LOV_USER_MAGIC_V1) {
CDEBUG(D_IOCTL,
"bad userland LOV MAGIC: %#08x != %#08x nor %#08x\n",
lmm_magic, LOV_USER_MAGIC_V1, LOV_USER_MAGIC_V3);
RETURN(-EINVAL);
}
/* in the rest of the tests, as *lumv1 and lumv3 have the same
* fields, we use lumv1 to avoid code duplication */
if (lumv1->lmm_pattern == 0) {
lumv1->lmm_pattern = lov->desc.ld_pattern ?
lov->desc.ld_pattern : LOV_PATTERN_RAID0;
}
if (lumv1->lmm_pattern != LOV_PATTERN_RAID0) {
CDEBUG(D_IOCTL, "bad userland stripe pattern: %#x\n",
lumv1->lmm_pattern);
RETURN(-EINVAL);
}
/* 64kB is the largest common page size we see (ia64), and matches the
* check in lfs */
if (lumv1->lmm_stripe_size & (LOV_MIN_STRIPE_SIZE - 1)) {
CDEBUG(D_IOCTL, "stripe size %u not multiple of %u, fixing\n",
lumv1->lmm_stripe_size, LOV_MIN_STRIPE_SIZE);
lumv1->lmm_stripe_size = LOV_MIN_STRIPE_SIZE;
}
if ((lumv1->lmm_stripe_offset >= lov->desc.ld_tgt_count) &&
(lumv1->lmm_stripe_offset !=
(typeof(lumv1->lmm_stripe_offset))(-1))) {
CDEBUG(D_IOCTL, "stripe offset %u > number of OSTs %u\n",
lumv1->lmm_stripe_offset, lov->desc.ld_tgt_count);
RETURN(-EINVAL);
}
stripe_count = lov_get_stripecnt(lov, lumv1->lmm_stripe_count);
if (max_lmm_size) {
int max_stripes = (max_lmm_size -
lov_mds_md_size(0, lmm_magic)) /
sizeof(struct lov_ost_data_v1);
if (unlikely(max_stripes < stripe_count)) {
CDEBUG(D_IOCTL, "stripe count reset from %d to %d\n",
stripe_count, max_stripes);
stripe_count = max_stripes;
}
}
if (lmm_magic == LOV_USER_MAGIC_V3) {
struct pool_desc *pool;
pool = lov_find_pool(lov, lumv3->lmm_pool_name);
if (pool != NULL) {
if (lumv3->lmm_stripe_offset !=
(typeof(lumv3->lmm_stripe_offset))(-1)) {
rc = lov_check_index_in_pool(
lumv3->lmm_stripe_offset, pool);
if (rc < 0) {
lov_pool_putref(pool);
RETURN(-EINVAL);
}
}
if (stripe_count > pool_tgt_count(pool))
stripe_count = pool_tgt_count(pool);
lov_pool_putref(pool);
}
}
rc = lov_alloc_memmd(lsmp, stripe_count, lumv1->lmm_pattern, lmm_magic);
if (rc >= 0) {
(*lsmp)->lsm_oinfo[0]->loi_ost_idx = lumv1->lmm_stripe_offset;
(*lsmp)->lsm_stripe_size = lumv1->lmm_stripe_size;
if (lmm_magic == LOV_USER_MAGIC_V3)
strncpy((*lsmp)->lsm_pool_name, lumv3->lmm_pool_name,
LOV_MAXPOOLNAME);
rc = 0;
}
RETURN(rc);
}
/* Pack LOV object metadata for disk storage. It is packed in LE byte
* order and is opaque to the networking layer.
*
* XXX In the future, this will be enhanced to get the EA size from the
* underlying OSC device(s) to get their EA sizes so we can stack
* LOVs properly. For now lov_mds_md_size() just assumes one obd_id
* per stripe.
*/
int lov_packmd(struct obd_export *exp, struct lov_mds_md **lmmp,
struct lov_stripe_md *lsm)
{
struct obd_device *obd = class_exp2obd(exp);
struct lov_obd *lov = &obd->u.lov;
struct lov_mds_md_v1 *lmmv1;
struct lov_mds_md_v3 *lmmv3;
int stripe_count;
struct lov_ost_data_v1 *lmm_objects;
int lmm_size, lmm_magic;
int i;
ENTRY;
if (lsm) {
lmm_magic = lsm->lsm_magic;
/* If we are just sizing the EA, limit the stripe count
* to the actual number of OSTs in this filesystem. */
if (!lmmp) {
stripe_count = lov_get_stripecnt(lov,
lsm->lsm_stripe_count);
lsm->lsm_stripe_count = stripe_count;
} else {
stripe_count = lsm->lsm_stripe_count;
}
} else {
/* No needs to allocated more than LOV_MAX_STRIPE_COUNT.
* Anyway, this is pretty inaccurate since ld_tgt_count now
* represents max index and we should rely on the actual number
* of OSTs instead */
stripe_count = min((__u32)LOV_MAX_STRIPE_COUNT,
lov->desc.ld_tgt_count);
if (lmmp && *lmmp)
lmm_magic = le32_to_cpu((*lmmp)->lmm_magic);
else
/* lsm == NULL and lmmp == NULL */
lmm_magic = LOV_MAGIC;
}
if ((lmm_magic != LOV_MAGIC_V1) &&
(lmm_magic != LOV_MAGIC_V3)) {
CERROR("bad mem LOV MAGIC: 0x%08X != 0x%08X nor 0x%08X\n",
lmm_magic, LOV_MAGIC_V1, LOV_MAGIC_V3);
RETURN(-EINVAL);
}
/* XXX LOV STACKING call into osc for sizes */
lmm_size = lov_mds_md_size(stripe_count, lmm_magic);
if (!lmmp)
RETURN(lmm_size);
if (*lmmp && !lsm) {
stripe_count = le32_to_cpu((*lmmp)->lmm_stripe_count);
lmm_size = lov_mds_md_size(stripe_count, lmm_magic);
OBD_FREE(*lmmp, lmm_size);
*lmmp = NULL;
RETURN(0);
}
if (!*lmmp) {
OBD_ALLOC(*lmmp, lmm_size);
if (!*lmmp)
RETURN(-ENOMEM);
}
CDEBUG(D_INFO, "lov_packmd: LOV_MAGIC 0x%08X, lmm_size = %d \n",
lmm_magic, lmm_size);
lmmv1 = *lmmp;
lmmv3 = (struct lov_mds_md_v3 *)*lmmp;
if (lmm_magic == LOV_MAGIC_V3)
lmmv3->lmm_magic = cpu_to_le32(LOV_MAGIC_V3);
else
lmmv1->lmm_magic = cpu_to_le32(LOV_MAGIC_V1);
if (!lsm)
RETURN(lmm_size);
/* lmmv1 and lmmv3 point to the same struct and have the
* same first fields
*/
lmmv1->lmm_object_id = cpu_to_le64(lsm->lsm_object_id);
lmmv1->lmm_object_seq = cpu_to_le64(lsm->lsm_object_seq);
lmmv1->lmm_stripe_size = cpu_to_le32(lsm->lsm_stripe_size);
lmmv1->lmm_stripe_count = cpu_to_le32(stripe_count);
lmmv1->lmm_pattern = cpu_to_le32(lsm->lsm_pattern);
if (lsm->lsm_magic == LOV_MAGIC_V3) {
strncpy(lmmv3->lmm_pool_name, lsm->lsm_pool_name,
LOV_MAXPOOLNAME);
lmm_objects = lmmv3->lmm_objects;
} else {
lmm_objects = lmmv1->lmm_objects;
}
for (i = 0; i < stripe_count; i++) {
struct lov_oinfo *loi = lsm->lsm_oinfo[i];
/* XXX LOV STACKING call down to osc_packmd() to do packing */
LASSERTF(loi->loi_id, "lmm_oid "LPU64" stripe %u/%u idx %u\n",
lmmv1->lmm_object_id, i, stripe_count, loi->loi_ost_idx);
lmm_objects[i].l_object_id = cpu_to_le64(loi->loi_id);
lmm_objects[i].l_object_seq = cpu_to_le64(loi->loi_seq);
lmm_objects[i].l_ost_gen = cpu_to_le32(loi->loi_ost_gen);
lmm_objects[i].l_ost_idx = cpu_to_le32(loi->loi_ost_idx);
}
RETURN(lmm_size);
}
/* Retrieve object striping information.
*
* @lump is a pointer to an in-core struct with lmm_ost_count indicating
* the maximum number of OST indices which will fit in the user buffer.
* lmm_magic must be LOV_USER_MAGIC.
*/
int lov_getstripe(struct lov_object *obj, struct lov_stripe_md *lsm,
struct lov_user_md __user *lump)
{
/* we use lov_user_md_v3 because it is larger than lov_user_md_v1 */
struct lov_user_md_v3 lum;
struct lov_mds_md *lmmk;
u32 stripe_count;
ssize_t lmm_size;
size_t lmmk_size;
size_t lum_size;
int rc;
if (!lsm)
return -ENODATA;
if (lsm->lsm_magic != LOV_MAGIC_V1 && lsm->lsm_magic != LOV_MAGIC_V3) {
CERROR("bad LSM MAGIC: 0x%08X != 0x%08X nor 0x%08X\n",
lsm->lsm_magic, LOV_MAGIC_V1, LOV_MAGIC_V3);
rc = -EIO;
goto out;
}
if (!lsm_is_released(lsm))
stripe_count = lsm->lsm_stripe_count;
else
stripe_count = 0;
/* we only need the header part from user space to get lmm_magic and
* lmm_stripe_count, (the header part is common to v1 and v3)
*/
lum_size = sizeof(struct lov_user_md_v1);
if (copy_from_user(&lum, lump, lum_size)) {
rc = -EFAULT;
goto out;
}
if (lum.lmm_magic != LOV_USER_MAGIC_V1 &&
lum.lmm_magic != LOV_USER_MAGIC_V3 &&
lum.lmm_magic != LOV_USER_MAGIC_SPECIFIC) {
rc = -EINVAL;
goto out;
}
if (lum.lmm_stripe_count &&
(lum.lmm_stripe_count < lsm->lsm_stripe_count)) {
/* Return right size of stripe to user */
lum.lmm_stripe_count = stripe_count;
rc = copy_to_user(lump, &lum, lum_size);
rc = -EOVERFLOW;
goto out;
}
lmmk_size = lov_mds_md_size(stripe_count, lsm->lsm_magic);
lmmk = kvzalloc(lmmk_size, GFP_NOFS);
if (!lmmk) {
rc = -ENOMEM;
goto out;
}
lmm_size = lov_lsm_pack(lsm, lmmk, lmmk_size);
if (lmm_size < 0) {
rc = lmm_size;
goto out_free;
}
/* FIXME: Bug 1185 - copy fields properly when structs change */
/* struct lov_user_md_v3 and struct lov_mds_md_v3 must be the same */
BUILD_BUG_ON(sizeof(lum) != sizeof(struct lov_mds_md_v3));
BUILD_BUG_ON(sizeof(lum.lmm_objects[0]) != sizeof(lmmk->lmm_objects[0]));
if (cpu_to_le32(LOV_MAGIC) != LOV_MAGIC &&
(lmmk->lmm_magic == cpu_to_le32(LOV_MAGIC_V1) ||
lmmk->lmm_magic == cpu_to_le32(LOV_MAGIC_V3))) {
lustre_swab_lov_mds_md(lmmk);
lustre_swab_lov_user_md_objects(
(struct lov_user_ost_data *)lmmk->lmm_objects,
lmmk->lmm_stripe_count);
}
if (lum.lmm_magic == LOV_USER_MAGIC) {
/* User request for v1, we need skip lmm_pool_name */
if (lmmk->lmm_magic == LOV_MAGIC_V3) {
memmove(((struct lov_mds_md_v1 *)lmmk)->lmm_objects,
((struct lov_mds_md_v3 *)lmmk)->lmm_objects,
lmmk->lmm_stripe_count *
sizeof(struct lov_ost_data_v1));
lmm_size -= LOV_MAXPOOLNAME;
}
} else {
/* if v3 we just have to update the lum_size */
lum_size = sizeof(struct lov_user_md_v3);
}
/* User wasn't expecting this many OST entries */
if (lum.lmm_stripe_count == 0) {
lmm_size = lum_size;
} else if (lum.lmm_stripe_count < lmmk->lmm_stripe_count) {
rc = -EOVERFLOW;
goto out_free;
}
/*
* Have a difference between lov_mds_md & lov_user_md.
* So we have to re-order the data before copy to user.
*/
lum.lmm_stripe_count = lmmk->lmm_stripe_count;
lum.lmm_layout_gen = lmmk->lmm_layout_gen;
((struct lov_user_md *)lmmk)->lmm_layout_gen = lum.lmm_layout_gen;
((struct lov_user_md *)lmmk)->lmm_stripe_count = lum.lmm_stripe_count;
if (copy_to_user(lump, lmmk, lmm_size))
rc = -EFAULT;
else
rc = 0;
out_free:
kvfree(lmmk);
out:
return rc;
}
/* Pack LOV object metadata for disk storage. It is packed in LE byte
* order and is opaque to the networking layer.
*
* XXX In the future, this will be enhanced to get the EA size from the
* underlying OSC device(s) to get their EA sizes so we can stack
* LOVs properly. For now lov_mds_md_size() just assumes one u64
* per stripe.
*/
int lov_obd_packmd(struct lov_obd *lov, struct lov_mds_md **lmmp,
struct lov_stripe_md *lsm)
{
struct lov_mds_md_v1 *lmmv1;
struct lov_mds_md_v3 *lmmv3;
__u16 stripe_count;
struct lov_ost_data_v1 *lmm_objects;
int lmm_size, lmm_magic;
int i;
int cplen = 0;
if (lsm) {
lmm_magic = lsm->lsm_magic;
} else {
if (lmmp && *lmmp)
lmm_magic = le32_to_cpu((*lmmp)->lmm_magic);
else
/* lsm == NULL and lmmp == NULL */
lmm_magic = LOV_MAGIC;
}
if ((lmm_magic != LOV_MAGIC_V1) &&
(lmm_magic != LOV_MAGIC_V3)) {
CERROR("bad mem LOV MAGIC: 0x%08X != 0x%08X nor 0x%08X\n",
lmm_magic, LOV_MAGIC_V1, LOV_MAGIC_V3);
return -EINVAL;
}
if (lsm) {
/* If we are just sizing the EA, limit the stripe count
* to the actual number of OSTs in this filesystem.
*/
if (!lmmp) {
stripe_count = lov_get_stripecnt(lov, lmm_magic,
lsm->lsm_stripe_count);
lsm->lsm_stripe_count = stripe_count;
} else if (!lsm_is_released(lsm)) {
stripe_count = lsm->lsm_stripe_count;
} else {
stripe_count = 0;
}
} else {
/*
* To calculate maximum easize by active targets at present,
* which is exactly the maximum easize to be seen by LOV
*/
stripe_count = lov->desc.ld_active_tgt_count;
}
/* XXX LOV STACKING call into osc for sizes */
lmm_size = lov_mds_md_size(stripe_count, lmm_magic);
if (!lmmp)
return lmm_size;
if (*lmmp && !lsm) {
stripe_count = le16_to_cpu((*lmmp)->lmm_stripe_count);
lmm_size = lov_mds_md_size(stripe_count, lmm_magic);
kvfree(*lmmp);
*lmmp = NULL;
return 0;
}
if (!*lmmp) {
*lmmp = libcfs_kvzalloc(lmm_size, GFP_NOFS);
if (!*lmmp)
return -ENOMEM;
}
CDEBUG(D_INFO, "lov_packmd: LOV_MAGIC 0x%08X, lmm_size = %d\n",
lmm_magic, lmm_size);
lmmv1 = *lmmp;
lmmv3 = (struct lov_mds_md_v3 *)*lmmp;
if (lmm_magic == LOV_MAGIC_V3)
lmmv3->lmm_magic = cpu_to_le32(LOV_MAGIC_V3);
else
lmmv1->lmm_magic = cpu_to_le32(LOV_MAGIC_V1);
if (!lsm)
return lmm_size;
/* lmmv1 and lmmv3 point to the same struct and have the
* same first fields
*/
lmm_oi_cpu_to_le(&lmmv1->lmm_oi, &lsm->lsm_oi);
lmmv1->lmm_stripe_size = cpu_to_le32(lsm->lsm_stripe_size);
lmmv1->lmm_stripe_count = cpu_to_le16(stripe_count);
lmmv1->lmm_pattern = cpu_to_le32(lsm->lsm_pattern);
lmmv1->lmm_layout_gen = cpu_to_le16(lsm->lsm_layout_gen);
if (lsm->lsm_magic == LOV_MAGIC_V3) {
cplen = strlcpy(lmmv3->lmm_pool_name, lsm->lsm_pool_name,
sizeof(lmmv3->lmm_pool_name));
if (cplen >= sizeof(lmmv3->lmm_pool_name))
return -E2BIG;
lmm_objects = lmmv3->lmm_objects;
} else {
lmm_objects = lmmv1->lmm_objects;
}
for (i = 0; i < stripe_count; i++) {
struct lov_oinfo *loi = lsm->lsm_oinfo[i];
/* XXX LOV STACKING call down to osc_packmd() to do packing */
LASSERTF(ostid_id(&loi->loi_oi) != 0, "lmm_oi "DOSTID
" stripe %u/%u idx %u\n", POSTID(&lmmv1->lmm_oi),
i, stripe_count, loi->loi_ost_idx);
ostid_cpu_to_le(&loi->loi_oi, &lmm_objects[i].l_ost_oi);
lmm_objects[i].l_ost_gen = cpu_to_le32(loi->loi_ost_gen);
lmm_objects[i].l_ost_idx = cpu_to_le32(loi->loi_ost_idx);
}
return lmm_size;
}
int main(int argc, char **argv)
{
DIR * dir;
struct lov_user_md *lum_dir, *lum_file1 = NULL, *lum_file2 = NULL;
int rc;
int lum_size;
if (argc < 3) {
llapi_err(LLAPI_MSG_ERROR,
"Usage: %s <dirname> <filename1> [filename2]\n",
argv[0]);
return 1;
}
dir = opendir(argv[1]);
if (dir == NULL) {
rc = errno;
llapi_err(LLAPI_MSG_ERROR,
"error: %s opendir failed\n", argv[1]);
return rc;
}
lum_size = lov_mds_md_size(MAX_LOV_UUID_COUNT, LOV_MAGIC);
if ((lum_dir = (struct lov_user_md *)malloc(lum_size)) == NULL) {
rc = ENOMEM;
llapi_err(LLAPI_MSG_ERROR, "error: can't allocate %d bytes "
"for dir EA", lum_size);
goto cleanup;
}
rc = llapi_file_get_stripe(argv[1], lum_dir);
if (rc) {
if (errno == ENODATA) {
free(lum_dir);
lum_dir = NULL;
} else {
rc = errno;
llapi_err(LLAPI_MSG_ERROR,
"error: can't get EA for %s\n", argv[1]);
goto cleanup;
}
}
/* XXX should be llapi_lov_getname() */
rc = llapi_file_get_lov_uuid(argv[1], &lov_uuid);
if (rc) {
rc = errno;
llapi_err(LLAPI_MSG_ERROR, "error: can't get lov name for %s\n",
argv[1]);
return rc;
}
if ((lum_file1 = (struct lov_user_md *)malloc(lum_size)) == NULL) {
rc = ENOMEM;
llapi_err(LLAPI_MSG_ERROR,
"error: can't allocate %d bytes for EA\n", lum_size);
goto cleanup;
}
rc = llapi_file_get_stripe(argv[2], lum_file1);
if (rc) {
rc = errno;
llapi_err(LLAPI_MSG_ERROR,
"error: unable to get EA for %s\n", argv[2]);
goto cleanup;
}
if (argc == 4) {
lum_file2 = (struct lov_user_md *)malloc(lum_size);
if (lum_file2 == NULL) {
rc = ENOMEM;
llapi_err(LLAPI_MSG_ERROR, "error: can't allocate %d "
"bytes for file2 EA\n", lum_size);
goto cleanup;
}
rc = llapi_file_get_stripe(argv[3], lum_file2);
if (rc) {
rc = errno;
llapi_err(LLAPI_MSG_ERROR,
"error: can't get EA for %s\n", argv[3]);
goto cleanup;
}
}
rc = compare(lum_dir, lum_file1, lum_file2);
cleanup:
closedir(dir);
if (lum_dir != NULL)
free(lum_dir);
if (lum_file1 != NULL)
free(lum_file1);
if (lum_file2 != NULL)
free(lum_file2);
return rc;
}