const unsigned char *buf,
unsigned len);
+#define DDF_NOTFOUND (~0U)
+#define DDF_CONTAINER (DDF_NOTFOUND-1)
+
/* The DDF metadata handling.
* DDF metadata lives at the end of the device.
* The last 512 byte block provides an 'anchor' which is used to locate
struct phys_disk *phys;
struct virtual_disk *virt;
int pdsize, vdsize;
- int max_part, mppe, conf_rec_len;
+ unsigned int max_part, mppe, conf_rec_len;
int currentdev;
int updates_pending;
struct vcl {
struct vcl *next;
__u64 *lba_offset; /* location in 'conf' of
* the lba table */
- int vcnum; /* index into ->virt */
+ unsigned int vcnum; /* index into ->virt */
+ struct vd_config **other_bvds;
__u64 *block_sizes; /* NULL if all the same */
};
};
char *devname;
int fd;
unsigned long long size; /* sectors */
+ unsigned long long primary_lba; /* sectors */
+ unsigned long long secondary_lba; /* sectors */
+ unsigned long long workspace_lba; /* sectors */
int pdnum; /* index in ->phys */
struct spare_assign *spare;
void *mdupdate; /* hold metadata update */
#define offsetof(t,f) ((size_t)&(((t*)0)->f))
#endif
+#if DEBUG
+static int all_ff(char *guid);
+static void pr_state(struct ddf_super *ddf, const char *msg)
+{
+ unsigned int i;
+ dprintf("%s/%s: ", __func__, msg);
+ for (i = 0; i < __be16_to_cpu(ddf->active->max_vd_entries); i++) {
+ if (all_ff(ddf->virt->entries[i].guid))
+ continue;
+ dprintf("%u(s=%02x i=%02x) ", i,
+ ddf->virt->entries[i].state,
+ ddf->virt->entries[i].init_state);
+ }
+ dprintf("\n");
+}
+#else
+static void pr_state(const struct ddf_super *ddf, const char *msg) {}
+#endif
+
+#define ddf_set_updates_pending(x) \
+ do { (x)->updates_pending = 1; pr_state(x, __func__); } while (0)
-static int calc_crc(void *buf, int len)
+static unsigned int calc_crc(void *buf, int len)
{
/* crcs are always at the same place as in the ddf_header */
struct ddf_header *ddf = buf;
else
offset += __be64_to_cpu(super->active->secondary_lba);
- if (lseek64(fd, offset<<9, 0) != (offset<<9)) {
+ if ((unsigned long long)lseek64(fd, offset<<9, 0) != (offset<<9)) {
if (dofree)
free(buf);
return NULL;
}
- if (read(fd, buf, len<<9) != (len<<9)) {
+ if ((unsigned long long)read(fd, buf, len<<9) != (len<<9)) {
if (dofree)
free(buf);
return NULL;
if (lseek64(fd, dsize-512, 0) < 0) {
if (devname)
- fprintf(stderr,
- Name": Cannot seek to anchor block on %s: %s\n",
- devname, strerror(errno));
+ pr_err("Cannot seek to anchor block on %s: %s\n",
+ devname, strerror(errno));
return 1;
}
if (read(fd, &super->anchor, 512) != 512) {
if (devname)
- fprintf(stderr,
- Name ": Cannot read anchor block on %s: %s\n",
- devname, strerror(errno));
+ pr_err("Cannot read anchor block on %s: %s\n",
+ devname, strerror(errno));
return 1;
}
if (super->anchor.magic != DDF_HEADER_MAGIC) {
if (devname)
- fprintf(stderr, Name ": no DDF anchor found on %s\n",
+ pr_err("no DDF anchor found on %s\n",
devname);
return 2;
}
if (calc_crc(&super->anchor, 512) != super->anchor.crc) {
if (devname)
- fprintf(stderr, Name ": bad CRC on anchor on %s\n",
+ pr_err("bad CRC on anchor on %s\n",
devname);
return 2;
}
if (memcmp(super->anchor.revision, DDF_REVISION_0, 8) != 0 &&
memcmp(super->anchor.revision, DDF_REVISION_2, 8) != 0) {
if (devname)
- fprintf(stderr, Name ": can only support super revision"
+ pr_err("can only support super revision"
" %.8s and earlier, not %.8s on %s\n",
DDF_REVISION_2, super->anchor.revision,devname);
return 2;
dsize >> 9, 1,
&super->primary, &super->anchor) == 0) {
if (devname)
- fprintf(stderr,
- Name ": Failed to load primary DDF header "
- "on %s\n", devname);
+ pr_err("Failed to load primary DDF header "
+ "on %s\n", devname);
return 2;
}
super->active = &super->primary;
return 0;
}
+static void add_other_bvd(struct vcl *vcl, struct vd_config *vd,
+ unsigned int len)
+{
+ int i;
+ for (i = 0; i < vcl->conf.sec_elmnt_count-1; i++)
+ if (vcl->other_bvds[i] != NULL &&
+ vcl->other_bvds[i]->sec_elmnt_seq == vd->sec_elmnt_seq)
+ break;
+
+ if (i < vcl->conf.sec_elmnt_count-1) {
+ if (vd->seqnum <= vcl->other_bvds[i]->seqnum)
+ return;
+ } else {
+ for (i = 0; i < vcl->conf.sec_elmnt_count-1; i++)
+ if (vcl->other_bvds[i] == NULL)
+ break;
+ if (i == vcl->conf.sec_elmnt_count-1) {
+ pr_err("no space for sec level config %u, count is %u\n",
+ vd->sec_elmnt_seq, vcl->conf.sec_elmnt_count);
+ return;
+ }
+ if (posix_memalign((void **)&vcl->other_bvds[i], 512, len)
+ != 0) {
+ pr_err("%s could not allocate vd buf\n", __func__);
+ return;
+ }
+ }
+ memcpy(vcl->other_bvds[i], vd, len);
+}
+
static int load_ddf_local(int fd, struct ddf_super *super,
char *devname, int keep)
{
struct dl *dl;
struct stat stb;
char *conf;
- int i;
- int confsec;
+ unsigned int i;
+ unsigned int confsec;
int vnum;
- int max_virt_disks = __be16_to_cpu(super->active->max_vd_entries);
+ unsigned int max_virt_disks = __be16_to_cpu(super->active->max_vd_entries);
unsigned long long dsize;
/* First the local disk info */
if (posix_memalign((void**)&dl, 512,
sizeof(*dl) +
(super->max_part) * sizeof(dl->vlist[0])) != 0) {
- fprintf(stderr, Name ": %s could not allocate disk info buffer\n",
+ pr_err("%s could not allocate disk info buffer\n",
__func__);
return 1;
}
super->active->data_section_offset,
super->active->data_section_length,
0);
- dl->devname = devname ? strdup(devname) : NULL;
+ dl->devname = devname ? xstrdup(devname) : NULL;
fstat(fd, &stb);
dl->major = major(stb.st_rdev);
dl->size = 0;
if (get_dev_size(fd, devname, &dsize))
dl->size = dsize >> 9;
+ /* If the disks have different sizes, the LBAs will differ
+ * between phys disks.
+ * At this point here, the values in super->active must be valid
+ * for this phys disk. */
+ dl->primary_lba = super->active->primary_lba;
+ dl->secondary_lba = super->active->secondary_lba;
+ dl->workspace_lba = super->active->workspace_lba;
dl->spare = NULL;
- for (i=0 ; i < super->max_part ; i++)
+ for (i = 0 ; i < super->max_part ; i++)
dl->vlist[i] = NULL;
super->dlist = dl;
dl->pdnum = -1;
- for (i=0; i < __be16_to_cpu(super->active->max_pd_entries); i++)
+ for (i = 0; i < __be16_to_cpu(super->active->max_pd_entries); i++)
if (memcmp(super->phys->entries[i].guid,
dl->disk.guid, DDF_GUID_LEN) == 0)
dl->pdnum = i;
continue;
if (posix_memalign((void**)&dl->spare, 512,
super->conf_rec_len*512) != 0) {
- fprintf(stderr, Name
- ": %s could not allocate spare info buf\n",
- __func__);
+ pr_err("%s could not allocate spare info buf\n",
+ __func__);
return 1;
}
-
+
memcpy(dl->spare, vd, super->conf_rec_len*512);
continue;
}
if (vcl) {
dl->vlist[vnum++] = vcl;
+ if (vcl->other_bvds != NULL &&
+ vcl->conf.sec_elmnt_seq != vd->sec_elmnt_seq) {
+ add_other_bvd(vcl, vd, super->conf_rec_len*512);
+ continue;
+ }
if (__be32_to_cpu(vd->seqnum) <=
__be32_to_cpu(vcl->conf.seqnum))
continue;
if (posix_memalign((void**)&vcl, 512,
(super->conf_rec_len*512 +
offsetof(struct vcl, conf))) != 0) {
- fprintf(stderr, Name
- ": %s could not allocate vcl buf\n",
- __func__);
+ pr_err("%s could not allocate vcl buf\n",
+ __func__);
return 1;
}
vcl->next = super->conflist;
vcl->block_sizes = NULL; /* FIXME not for CONCAT */
+ if (vd->sec_elmnt_count > 1)
+ vcl->other_bvds =
+ xcalloc(vd->sec_elmnt_count - 1,
+ sizeof(struct vd_config *));
+ else
+ vcl->other_bvds = NULL;
super->conflist = vcl;
dl->vlist[vnum++] = vcl;
}
#ifndef MDASSEMBLE
static int load_super_ddf_all(struct supertype *st, int fd,
- void **sbp, char *devname, int keep_fd);
+ void **sbp, char *devname);
#endif
static void free_super_ddf(struct supertype *st);
struct ddf_super *super;
int rv;
-#ifndef MDASSEMBLE
- /* if 'fd' is a container, load metadata from all the devices */
- if (load_super_ddf_all(st, fd, &st->sb, devname, 1) == 0)
- return 0;
-#endif
- if (st->subarray[0])
- return 1; /* FIXME Is this correct */
-
if (get_dev_size(fd, devname, &dsize) == 0)
return 1;
/* 32M is a lower bound */
if (dsize <= 32*1024*1024) {
if (devname)
- fprintf(stderr,
- Name ": %s is too small for ddf: "
- "size is %llu sectors.\n",
- devname, dsize>>9);
+ pr_err("%s is too small for ddf: "
+ "size is %llu sectors.\n",
+ devname, dsize>>9);
return 1;
}
if (dsize & 511) {
if (devname)
- fprintf(stderr,
- Name ": %s is an odd size for ddf: "
- "size is %llu bytes.\n",
- devname, dsize);
+ pr_err("%s is an odd size for ddf: "
+ "size is %llu bytes.\n",
+ devname, dsize);
return 1;
}
free_super_ddf(st);
if (posix_memalign((void**)&super, 512, sizeof(*super))!= 0) {
- fprintf(stderr, Name ": malloc of %zu failed.\n",
+ pr_err("malloc of %zu failed.\n",
sizeof(*super));
return 1;
}
if (rv) {
if (devname)
- fprintf(stderr,
- Name ": Failed to load all information "
- "sections on %s\n", devname);
+ pr_err("Failed to load all information "
+ "sections on %s\n", devname);
free(super);
return rv;
}
if (rv) {
if (devname)
- fprintf(stderr,
- Name ": Failed to load all information "
- "sections on %s\n", devname);
+ pr_err("Failed to load all information "
+ "sections on %s\n", devname);
free(super);
return rv;
}
- if (st->subarray[0]) {
- unsigned long val;
- struct vcl *v;
- char *ep;
-
- val = strtoul(st->subarray, &ep, 10);
- if (*ep != '\0') {
- free(super);
- return 1;
- }
-
- for (v = super->conflist; v; v = v->next)
- if (v->vcnum == val)
- super->currentconf = v;
- if (!super->currentconf) {
- free(super);
- return 1;
- }
- }
-
/* Should possibly check the sections .... */
st->sb = super;
st->minor_version = 0;
st->max_devs = 512;
}
- st->loaded_container = 0;
return 0;
}
ddf->conflist = v->next;
if (v->block_sizes)
free(v->block_sizes);
+ if (v->other_bvds) {
+ int i;
+ for (i = 0; i < v->conf.sec_elmnt_count-1; i++)
+ if (v->other_bvds[i] != NULL)
+ free(v->other_bvds[i]);
+ free(v->other_bvds);
+ }
free(v);
}
while (ddf->dlist) {
free(d->spare);
free(d);
}
+ while (ddf->add_list) {
+ struct dl *d = ddf->add_list;
+ ddf->add_list = d->next;
+ if (d->fd >= 0)
+ close(d->fd);
+ if (d->spare)
+ free(d->spare);
+ free(d);
+ }
free(ddf);
st->sb = NULL;
}
)
return NULL;
- st = malloc(sizeof(*st));
- memset(st, 0, sizeof(*st));
+ st = xcalloc(1, sizeof(*st));
st->ss = &super_ddf;
st->max_devs = 512;
st->minor_version = 0;
return st;
}
-
#ifndef MDASSEMBLE
static mapping_t ddf_state[] = {
struct vcl *vcl;
for (vcl = sb->conflist ; vcl ; vcl = vcl->next) {
- int i;
+ unsigned int i;
struct vd_config *vc = &vcl->conf;
if (calc_crc(vc, crl*512) != vc->crc)
/* Ok, we know about this VD, let's give more details */
printf(" Raid Devices[%d] : %d (", n,
__be16_to_cpu(vc->prim_elmnt_count));
- for (i=0; i<__be16_to_cpu(vc->prim_elmnt_count); i++) {
+ for (i = 0; i < __be16_to_cpu(vc->prim_elmnt_count); i++) {
int j;
int cnt = __be16_to_cpu(sb->phys->used_pdes);
for (j=0; j<cnt; j++)
}
printf(")\n");
if (vc->chunk_shift != 255)
- printf(" Chunk Size[%d] : %d sectors\n", n,
- 1 << vc->chunk_shift);
+ printf(" Chunk Size[%d] : %d sectors\n", n,
+ 1 << vc->chunk_shift);
printf(" Raid Level[%d] : %s\n", n,
map_num(ddf_level, vc->prl)?:"-unknown-");
if (vc->sec_elmnt_count != 1) {
//printf("\n");
printf(" %3d %08x ", i,
__be32_to_cpu(pd->refnum));
- printf("%8lluK ",
+ printf("%8lluK ",
(unsigned long long)__be64_to_cpu(pd->config_size)>>1);
for (dl = sb->dlist; dl ; dl = dl->next) {
if (dl->disk.refnum == pd->refnum) {
(type&8) ? "spare" : "",
(type&16)? ", foreign" : "",
(type&32)? "pass-through" : "");
+ if (state & DDF_Failed)
+ /* This over-rides these three */
+ state &= ~(DDF_Online|DDF_Rebuilding|DDF_Transition);
printf("/%s%s%s%s%s%s%s",
(state&1)? "Online": "Offline",
(state&2)? ", Failed": "",
examine_pds(sb);
}
-static void getinfo_super_ddf(struct supertype *st, struct mdinfo *info);
+static void getinfo_super_ddf(struct supertype *st, struct mdinfo *info, char *map);
+static void uuid_from_ddf_guid(const char *guid, int uuid[4]);
static void uuid_from_super_ddf(struct supertype *st, int uuid[4]);
+static unsigned int get_vd_num_of_subarray(struct supertype *st)
+{
+ /*
+ * Figure out the VD number for this supertype.
+ * Returns DDF_CONTAINER for the container itself,
+ * and DDF_NOTFOUND on error.
+ */
+ struct ddf_super *ddf = st->sb;
+ struct mdinfo *sra;
+ char *sub, *end;
+ unsigned int vcnum;
+
+ if (*st->container_devnm == '\0')
+ return DDF_CONTAINER;
+
+ sra = sysfs_read(-1, st->devnm, GET_VERSION);
+ if (!sra || sra->array.major_version != -1 ||
+ sra->array.minor_version != -2 ||
+ !is_subarray(sra->text_version))
+ return DDF_NOTFOUND;
+
+ sub = strchr(sra->text_version + 1, '/');
+ if (sub != NULL)
+ vcnum = strtoul(sub + 1, &end, 10);
+ if (sub == NULL || *sub == '\0' || *end != '\0' ||
+ vcnum >= __be16_to_cpu(ddf->active->max_vd_entries))
+ return DDF_NOTFOUND;
+
+ return vcnum;
+}
+
static void brief_examine_super_ddf(struct supertype *st, int verbose)
{
/* We just write a generic DDF ARRAY entry
*/
struct mdinfo info;
char nbuf[64];
- getinfo_super_ddf(st, &info);
+ getinfo_super_ddf(st, &info, NULL);
fname_from_uuid(st, &info, nbuf, ':');
printf("ARRAY metadata=ddf UUID=%s\n", nbuf + 5);
*/
struct ddf_super *ddf = st->sb;
struct mdinfo info;
- int i;
+ unsigned int i;
char nbuf[64];
- getinfo_super_ddf(st, &info);
+ getinfo_super_ddf(st, &info, NULL);
fname_from_uuid(st, &info, nbuf, ':');
- for (i=0; i<__be16_to_cpu(ddf->virt->max_vdes); i++) {
+ for (i = 0; i < __be16_to_cpu(ddf->virt->max_vdes); i++) {
struct virtual_entry *ve = &ddf->virt->entries[i];
struct vcl vcl;
char nbuf1[64];
{
struct mdinfo info;
char nbuf[64];
- getinfo_super_ddf(st, &info);
+ getinfo_super_ddf(st, &info, NULL);
fname_from_uuid(st, &info, nbuf, ':');
printf("MD_METADATA=ddf\n");
printf("MD_LEVEL=container\n");
printf("MD_UUID=%s\n", nbuf+5);
}
-
static void detail_super_ddf(struct supertype *st, char *homehost)
{
static void brief_detail_super_ddf(struct supertype *st)
{
- /* FIXME I really need to know which array we are detailing.
- * Can that be stored in ddf_super??
- */
-// struct ddf_super *ddf = st->sb;
struct mdinfo info;
char nbuf[64];
- getinfo_super_ddf(st, &info);
+ struct ddf_super *ddf = st->sb;
+ unsigned int vcnum = get_vd_num_of_subarray(st);
+ if (vcnum == DDF_CONTAINER)
+ uuid_from_super_ddf(st, info.uuid);
+ else if (vcnum == DDF_NOTFOUND)
+ return;
+ else
+ uuid_from_ddf_guid(ddf->virt->entries[vcnum].guid, info.uuid);
fname_from_uuid(st, &info, nbuf,':');
printf(" UUID=%s", nbuf + 5);
}
* the hostname
*/
struct ddf_super *ddf = st->sb;
- int len;
+ unsigned int len;
if (!homehost)
return 0;
}
#ifndef MDASSEMBLE
-static struct vd_config *find_vdcr(struct ddf_super *ddf, int inst)
+static struct vd_config *find_vdcr(struct ddf_super *ddf, unsigned int inst)
{
struct vcl *v;
/* Find the entry in phys_disk which has the given refnum
* and return it's index
*/
- int i;
- for (i=0; i < __be16_to_cpu(ddf->phys->max_pdes); i++)
+ unsigned int i;
+ for (i = 0; i < __be16_to_cpu(ddf->phys->max_pdes); i++)
if (ddf->phys->entries[i].refnum == phys_refnum)
return i;
return -1;
}
+static void uuid_from_ddf_guid(const char *guid, int uuid[4])
+{
+ char buf[20];
+ struct sha1_ctx ctx;
+ sha1_init_ctx(&ctx);
+ sha1_process_bytes(guid, DDF_GUID_LEN, &ctx);
+ sha1_finish_ctx(&ctx, buf);
+ memcpy(uuid, buf, 4*4);
+}
+
static void uuid_from_super_ddf(struct supertype *st, int uuid[4])
{
/* The uuid returned here is used for:
* not the device-set.
* uuid to recognise same set when adding a missing device back
* to an array. This is a uuid for the device-set.
- *
+ *
* For each of these we can make do with a truncated
* or hashed uuid rather than the original, as long as
* everyone agrees.
struct ddf_super *ddf = st->sb;
struct vcl *vcl = ddf->currentconf;
char *guid;
- char buf[20];
- struct sha1_ctx ctx;
if (vcl)
guid = vcl->conf.guid;
else
guid = ddf->anchor.guid;
-
- sha1_init_ctx(&ctx);
- sha1_process_bytes(guid, DDF_GUID_LEN, &ctx);
- sha1_finish_ctx(&ctx, buf);
- memcpy(uuid, buf, 4*4);
+ uuid_from_ddf_guid(guid, uuid);
}
-static void getinfo_super_ddf_bvd(struct supertype *st, struct mdinfo *info);
+static void getinfo_super_ddf_bvd(struct supertype *st, struct mdinfo *info, char *map);
-static void getinfo_super_ddf(struct supertype *st, struct mdinfo *info)
+static void getinfo_super_ddf(struct supertype *st, struct mdinfo *info, char *map)
{
struct ddf_super *ddf = st->sb;
+ int map_disks = info->array.raid_disks;
+ __u32 *cptr;
if (ddf->currentconf) {
- getinfo_super_ddf_bvd(st, info);
+ getinfo_super_ddf_bvd(st, info, map);
return;
}
+ memset(info, 0, sizeof(*info));
info->array.raid_disks = __be16_to_cpu(ddf->phys->used_pdes);
info->array.level = LEVEL_CONTAINER;
info->array.layout = 0;
info->array.md_minor = -1;
- info->array.ctime = DECADE + __be32_to_cpu(*(__u32*)
- (ddf->anchor.guid+16));
+ cptr = (__u32 *)(ddf->anchor.guid + 16);
+ info->array.ctime = DECADE + __be32_to_cpu(*cptr);
+
info->array.utime = 0;
info->array.chunk_size = 0;
-
+ info->container_enough = 1;
info->disk.major = 0;
info->disk.minor = 0;
info->disk.raid_disk = find_phys(ddf, ddf->dlist->disk.refnum);
info->data_offset = __be64_to_cpu(ddf->phys->
- entries[info->disk.raid_disk].
- config_size);
+ entries[info->disk.raid_disk].
+ config_size);
info->component_size = ddf->dlist->size - info->data_offset;
} else {
info->disk.number = -1;
}
info->disk.state = (1 << MD_DISK_SYNC) | (1 << MD_DISK_ACTIVE);
-
info->recovery_start = MaxSector;
info->reshape_active = 0;
+ info->recovery_blocked = 0;
info->name[0] = 0;
info->array.major_version = -1;
uuid_from_super_ddf(st, info->uuid);
+ if (map) {
+ int i;
+ for (i = 0 ; i < map_disks; i++) {
+ if (i < info->array.raid_disks &&
+ (__be16_to_cpu(ddf->phys->entries[i].state) & DDF_Online) &&
+ !(__be16_to_cpu(ddf->phys->entries[i].state) & DDF_Failed))
+ map[i] = 1;
+ else
+ map[i] = 0;
+ }
+ }
}
static int rlq_to_layout(int rlq, int prl, int raiddisks);
-static void getinfo_super_ddf_bvd(struct supertype *st, struct mdinfo *info)
+static void getinfo_super_ddf_bvd(struct supertype *st, struct mdinfo *info, char *map)
{
struct ddf_super *ddf = st->sb;
struct vcl *vc = ddf->currentconf;
int cd = ddf->currentdev;
int j;
struct dl *dl;
+ int map_disks = info->array.raid_disks;
+ __u32 *cptr;
+ memset(info, 0, sizeof(*info));
/* FIXME this returns BVD info - what if we want SVD ?? */
info->array.raid_disks = __be16_to_cpu(vc->conf.prim_elmnt_count);
info->array.layout = rlq_to_layout(vc->conf.rlq, vc->conf.prl,
info->array.raid_disks);
info->array.md_minor = -1;
- info->array.ctime = DECADE +
- __be32_to_cpu(*(__u32*)(vc->conf.guid+16));
+ cptr = (__u32 *)(vc->conf.guid + 16);
+ info->array.ctime = DECADE + __be32_to_cpu(*cptr);
info->array.utime = DECADE + __be32_to_cpu(vc->conf.timestamp);
info->array.chunk_size = 512 << vc->conf.chunk_shift;
info->custom_array_size = 0;
- if (cd >= 0 && cd < ddf->mppe) {
+ if (cd >= 0 && (unsigned)cd < ddf->mppe) {
info->data_offset = __be64_to_cpu(vc->lba_offset[cd]);
if (vc->block_sizes)
info->component_size = vc->block_sizes[cd];
}
for (dl = ddf->dlist; dl ; dl = dl->next)
- if (dl->raiddisk == info->disk.raid_disk)
+ if (dl->raiddisk == ddf->currentdev)
break;
+
info->disk.major = 0;
info->disk.minor = 0;
+ info->disk.state = 0;
if (dl) {
info->disk.major = dl->major;
info->disk.minor = dl->minor;
+ info->disk.raid_disk = dl->raiddisk;
+ info->disk.number = dl->pdnum;
+ info->disk.state = (1<<MD_DISK_SYNC)|(1<<MD_DISK_ACTIVE);
}
-// info->disk.number = __be32_to_cpu(ddf->disk.refnum);
-// info->disk.raid_disk = find refnum in the table and use index;
-// info->disk.state = ???;
info->container_member = ddf->currentconf->vcnum;
info->recovery_start = MaxSector;
info->resync_start = 0;
info->reshape_active = 0;
+ info->recovery_blocked = 0;
if (!(ddf->virt->entries[info->container_member].state
& DDF_state_inconsistent) &&
(ddf->virt->entries[info->container_member].init_state
uuid_from_super_ddf(st, info->uuid);
- info->container_member = atoi(st->subarray);
info->array.major_version = -1;
info->array.minor_version = -2;
- sprintf(info->text_version, "/%s/%s",
- devnum2devname(st->container_dev),
- st->subarray);
+ sprintf(info->text_version, "/%s/%d",
+ st->container_devnm,
+ info->container_member);
info->safe_mode_delay = 200;
memcpy(info->name, ddf->virt->entries[info->container_member].name, 16);
for(j=0; j<16; j++)
if (info->name[j] == ' ')
info->name[j] = 0;
-}
+ if (map)
+ for (j = 0; j < map_disks; j++) {
+ map[j] = 0;
+ if (j < info->array.raid_disks) {
+ int i = find_phys(ddf, vc->conf.phys_refnum[j]);
+ if (i >= 0 &&
+ (__be16_to_cpu(ddf->phys->entries[i].state) & DDF_Online) &&
+ !(__be16_to_cpu(ddf->phys->entries[i].state) & DDF_Failed))
+ map[i] = 1;
+ }
+ }
+}
static int update_super_ddf(struct supertype *st, struct mdinfo *info,
char *update,
if (strcmp(update, "grow") == 0) {
/* FIXME */
- }
- if (strcmp(update, "resync") == 0) {
+ } else if (strcmp(update, "resync") == 0) {
// info->resync_checkpoint = 0;
- }
- /* We ignore UUID updates as they make even less sense
- * with DDF
- */
- if (strcmp(update, "homehost") == 0) {
+ } else if (strcmp(update, "homehost") == 0) {
/* homehost is stored in controller->vendor_data,
* or it is when we are the vendor
*/
// if (info->vendor_is_local)
// strcpy(ddf->controller.vendor_data, homehost);
- }
- if (strcmp(update, "name") == 0) {
+ rv = -1;
+ } else if (strcmp(update, "name") == 0) {
/* name is stored in virtual_entry->name */
// memset(ve->name, ' ', 16);
// strncpy(ve->name, info->name, 16);
- }
- if (strcmp(update, "_reshape_progress") == 0) {
+ rv = -1;
+ } else if (strcmp(update, "_reshape_progress") == 0) {
/* We don't support reshape yet */
- }
+ } else if (strcmp(update, "assemble") == 0 ) {
+ /* Do nothing, just succeed */
+ rv = 0;
+ } else
+ rv = -1;
// update_all_csum(ddf);
mdu_array_info_t *info,
unsigned long long size,
char *name, char *homehost,
- int *uuid);
+ int *uuid, unsigned long long data_offset);
static int init_super_ddf(struct supertype *st,
mdu_array_info_t *info,
unsigned long long size, char *name, char *homehost,
- int *uuid)
+ int *uuid, unsigned long long data_offset)
{
/* This is primarily called by Create when creating a new array.
* We will then get add_to_super called for each component, and then
struct phys_disk *pd;
struct virtual_disk *vd;
+ if (data_offset != INVALID_SECTORS) {
+ fprintf(stderr, Name ": data-offset not supported by DDF\n");
+ return 0;
+ }
+
if (st->sb)
- return init_super_ddf_bvd(st, info, size, name, homehost, uuid);
+ return init_super_ddf_bvd(st, info, size, name, homehost, uuid,
+ data_offset);
if (posix_memalign((void**)&ddf, 512, sizeof(*ddf)) != 0) {
- fprintf(stderr, Name ": %s could not allocate superblock\n", __func__);
+ pr_err("%s could not allocate superblock\n", __func__);
return 0;
}
memset(ddf, 0, sizeof(*ddf));
strcpy((char*)ddf->controller.vendor_data, homehost);
if (posix_memalign((void**)&pd, 512, pdsize) != 0) {
- fprintf(stderr, Name ": %s could not allocate pd\n", __func__);
+ pr_err("%s could not allocate pd\n", __func__);
return 0;
}
ddf->phys = pd;
memset(pd->pad, 0xff, 52);
if (posix_memalign((void**)&vd, 512, vdsize) != 0) {
- fprintf(stderr, Name ": %s could not allocate vd\n", __func__);
+ pr_err("%s could not allocate vd\n", __func__);
return 0;
}
ddf->virt = vd;
memset(&vd->entries[i], 0xff, sizeof(struct virtual_entry));
st->sb = ddf;
- ddf->updates_pending = 1;
+ ddf_set_updates_pending(ddf);
return 1;
}
default: return -1;
}
}
+
static int layout_to_rlq(int level, int layout, int raiddisks)
{
switch(level) {
* (dnum) of the given ddf.
* Return a malloced array of 'struct extent'
-FIXME ignore DDF_Legacy devices?
+ * FIXME ignore DDF_Legacy devices?
*/
struct extent *rv;
int n = 0;
- int i, j;
+ unsigned int i, j;
- rv = malloc(sizeof(struct extent) * (ddf->max_part + 2));
- if (!rv)
- return NULL;
+ rv = xmalloc(sizeof(struct extent) * (ddf->max_part + 2));
for (i = 0; i < ddf->max_part; i++) {
struct vcl *v = dl->vlist[i];
if (v == NULL)
continue;
- for (j=0; j < v->conf.prim_elmnt_count; j++)
+ for (j = 0; j < v->conf.prim_elmnt_count; j++)
if (v->conf.phys_refnum[j] == dl->disk.refnum) {
/* This device plays role 'j' in 'v'. */
rv[n].start = __be64_to_cpu(v->lba_offset[j]);
mdu_array_info_t *info,
unsigned long long size,
char *name, char *homehost,
- int *uuid)
+ int *uuid, unsigned long long data_offset)
{
/* We are creating a BVD inside a pre-existing container.
* so st->sb is already set.
* We need to create a new vd_config and a new virtual_entry
*/
struct ddf_super *ddf = st->sb;
- int venum;
+ unsigned int venum;
struct virtual_entry *ve;
struct vcl *vcl;
struct vd_config *vc;
if (__be16_to_cpu(ddf->virt->populated_vdes)
>= __be16_to_cpu(ddf->virt->max_vdes)) {
- fprintf(stderr, Name": This ddf already has the "
- "maximum of %d virtual devices\n",
- __be16_to_cpu(ddf->virt->max_vdes));
+ pr_err("This ddf already has the "
+ "maximum of %d virtual devices\n",
+ __be16_to_cpu(ddf->virt->max_vdes));
return 0;
}
+ if (name)
+ for (venum = 0; venum < __be16_to_cpu(ddf->virt->max_vdes); venum++)
+ if (!all_ff(ddf->virt->entries[venum].guid)) {
+ char *n = ddf->virt->entries[venum].name;
+
+ if (strncmp(name, n, 16) == 0) {
+ pr_err("This ddf already"
+ " has an array called %s\n",
+ name);
+ return 0;
+ }
+ }
+
for (venum = 0; venum < __be16_to_cpu(ddf->virt->max_vdes); venum++)
if (all_ff(ddf->virt->entries[venum].guid))
break;
if (venum == __be16_to_cpu(ddf->virt->max_vdes)) {
- fprintf(stderr, Name ": Cannot find spare slot for "
- "virtual disk - DDF is corrupt\n");
+ pr_err("Cannot find spare slot for "
+ "virtual disk - DDF is corrupt\n");
return 0;
}
ve = &ddf->virt->entries[venum];
/* Now create a new vd_config */
if (posix_memalign((void**)&vcl, 512,
(offsetof(struct vcl, conf) + ddf->conf_rec_len * 512)) != 0) {
- fprintf(stderr, Name ": %s could not allocate vd_config\n", __func__);
+ pr_err("%s could not allocate vd_config\n", __func__);
return 0;
}
vcl->lba_offset = (__u64*) &vcl->conf.phys_refnum[ddf->mppe];
vcl->vcnum = venum;
- sprintf(st->subarray, "%d", venum);
vcl->block_sizes = NULL; /* FIXME not for CONCAT */
+ vcl->other_bvds = NULL;
vc = &vcl->conf;
vcl->next = ddf->conflist;
ddf->conflist = vcl;
ddf->currentconf = vcl;
- ddf->updates_pending = 1;
+ ddf_set_updates_pending(ddf);
return 1;
}
struct ddf_super *ddf = st->sb;
struct vd_config *vc;
__u64 *lba_offset;
- int working;
- int i;
+ unsigned int working;
+ unsigned int i;
unsigned long long blocks, pos, esize;
struct extent *ex;
vc->phys_refnum[dk->raid_disk] = dl->disk.refnum;
lba_offset[dk->raid_disk] = __cpu_to_be64(pos);
- for (i=0; i < ddf->max_part ; i++)
+ for (i = 0; i < ddf->max_part ; i++)
if (dl->vlist[i] == NULL)
break;
if (i == ddf->max_part)
*/
working = 0;
- for (i=0; i < __be16_to_cpu(vc->prim_elmnt_count); i++)
+ for (i = 0; i < __be16_to_cpu(vc->prim_elmnt_count); i++)
if (vc->phys_refnum[i] != 0xffffffff)
working++;
ddf->phys->entries[dl->pdnum].type &= ~__cpu_to_be16(DDF_Global_Spare);
ddf->phys->entries[dl->pdnum].type |= __cpu_to_be16(DDF_Active_in_VD);
- ddf->updates_pending = 1;
+ ddf_set_updates_pending(ddf);
}
/* add a device to a container, either while creating it or while
* expanding a pre-existing container
*/
static int add_to_super_ddf(struct supertype *st,
- mdu_disk_info_t *dk, int fd, char *devname)
+ mdu_disk_info_t *dk, int fd, char *devname,
+ unsigned long long data_offset)
{
struct ddf_super *ddf = st->sb;
struct dl *dd;
struct tm *tm;
unsigned long long size;
struct phys_disk_entry *pde;
- int n, i;
+ unsigned int n, i;
struct stat stb;
+ __u32 *tptr;
if (ddf->currentconf) {
add_to_super_ddf_bvd(st, dk, fd, devname);
fstat(fd, &stb);
if (posix_memalign((void**)&dd, 512,
sizeof(*dd) + sizeof(dd->vlist[0]) * ddf->max_part) != 0) {
- fprintf(stderr, Name
- ": %s could allocate buffer for new disk, aborting\n",
- __func__);
+ pr_err("%s could allocate buffer for new disk, aborting\n",
+ __func__);
return 1;
}
dd->major = major(stb.st_rdev);
tm = localtime(&now);
sprintf(dd->disk.guid, "%8s%04d%02d%02d",
T10, tm->tm_year+1900, tm->tm_mon+1, tm->tm_mday);
- *(__u32*)(dd->disk.guid + 16) = random32();
- *(__u32*)(dd->disk.guid + 20) = random32();
+ tptr = (__u32 *)(dd->disk.guid + 16);
+ *tptr++ = random32();
+ *tptr = random32();
do {
/* Cannot be bothered finding a CRC of some irrelevant details*/
dd->disk.refnum = random32();
- for (i = __be16_to_cpu(ddf->active->max_pd_entries) - 1;
- i >= 0; i--)
- if (ddf->phys->entries[i].refnum == dd->disk.refnum)
+ for (i = __be16_to_cpu(ddf->active->max_pd_entries);
+ i > 0; i--)
+ if (ddf->phys->entries[i-1].refnum == dd->disk.refnum)
break;
- } while (i >= 0);
+ } while (i > 0);
dd->disk.forced_ref = 1;
dd->disk.forced_guid = 1;
sizeof(struct phys_disk_entry));
struct phys_disk *pd;
- pd = malloc(len);
+ pd = xmalloc(len);
pd->magic = DDF_PHYS_RECORDS_MAGIC;
pd->used_pdes = __cpu_to_be16(n);
pde = &pd->entries[0];
} else {
dd->next = ddf->dlist;
ddf->dlist = dd;
- ddf->updates_pending = 1;
+ ddf_set_updates_pending(ddf);
}
return 0;
}
+static int remove_from_super_ddf(struct supertype *st, mdu_disk_info_t *dk)
+{
+ struct ddf_super *ddf = st->sb;
+ struct dl *dl;
+
+ /* mdmon has noticed that this disk (dk->major/dk->minor) has
+ * disappeared from the container.
+ * We need to arrange that it disappears from the metadata and
+ * internal data structures too.
+ * Most of the work is done by ddf_process_update which edits
+ * the metadata and closes the file handle and attaches the memory
+ * where free_updates will free it.
+ */
+ for (dl = ddf->dlist; dl ; dl = dl->next)
+ if (dl->major == dk->major &&
+ dl->minor == dk->minor)
+ break;
+ if (!dl)
+ return -1;
+
+ if (st->update_tail) {
+ int len = (sizeof(struct phys_disk) +
+ sizeof(struct phys_disk_entry));
+ struct phys_disk *pd;
+
+ pd = xmalloc(len);
+ pd->magic = DDF_PHYS_RECORDS_MAGIC;
+ pd->used_pdes = __cpu_to_be16(dl->pdnum);
+ pd->entries[0].state = __cpu_to_be16(DDF_Missing);
+ append_metadata_update(st, pd, len);
+ }
+ return 0;
+}
+
/*
* This is the write_init_super method for a ddf container. It is
* called when creating a container or adding another device to a
* container.
*/
+#define NULL_CONF_SZ 4096
-static unsigned char null_conf[4096+512];
+static unsigned int get_pd_index_from_refnum(const struct vcl *vc,
+ __u32 refnum, unsigned int nmax,
+ const struct vd_config **bvd,
+ unsigned int *idx);
-static int __write_init_super_ddf(struct supertype *st, int do_close)
+static int __write_ddf_structure(struct dl *d, struct ddf_super *ddf, __u8 type,
+ char *null_aligned)
{
+ unsigned long long sector;
+ struct ddf_header *header;
+ int fd, i, n_config, conf_size;
+
+ fd = d->fd;
+ switch (type) {
+ case DDF_HEADER_PRIMARY:
+ header = &ddf->primary;
+ sector = __be64_to_cpu(header->primary_lba);
+ break;
+ case DDF_HEADER_SECONDARY:
+ header = &ddf->secondary;
+ sector = __be64_to_cpu(header->secondary_lba);
+ break;
+ default:
+ return 0;
+ }
+
+ header->type = type;
+ header->openflag = 0;
+ header->crc = calc_crc(header, 512);
+
+ lseek64(fd, sector<<9, 0);
+ if (write(fd, header, 512) < 0)
+ return 0;
+
+ ddf->controller.crc = calc_crc(&ddf->controller, 512);
+ if (write(fd, &ddf->controller, 512) < 0)
+ return 0;
+
+ ddf->phys->crc = calc_crc(ddf->phys, ddf->pdsize);
+ if (write(fd, ddf->phys, ddf->pdsize) < 0)
+ return 0;
+ ddf->virt->crc = calc_crc(ddf->virt, ddf->vdsize);
+ if (write(fd, ddf->virt, ddf->vdsize) < 0)
+ return 0;
+
+ /* Now write lots of config records. */
+ n_config = ddf->max_part;
+ conf_size = ddf->conf_rec_len * 512;
+ for (i = 0 ; i <= n_config ; i++) {
+ struct vcl *c;
+ struct vd_config *vdc = NULL;
+ if (i == n_config) {
+ c = (struct vcl *)d->spare;
+ if (c)
+ vdc = &c->conf;
+ } else {
+ unsigned int dummy;
+ c = d->vlist[i];
+ if (c)
+ get_pd_index_from_refnum(
+ c, d->disk.refnum,
+ ddf->mppe,
+ (const struct vd_config **)&vdc,
+ &dummy);
+ }
+ if (c) {
+ vdc->seqnum = header->seq;
+ vdc->crc = calc_crc(vdc, conf_size);
+ if (write(fd, vdc, conf_size) < 0)
+ break;
+ } else {
+ unsigned int togo = conf_size;
+ while (togo > NULL_CONF_SZ) {
+ if (write(fd, null_aligned, NULL_CONF_SZ) < 0)
+ break;
+ togo -= NULL_CONF_SZ;
+ }
+ if (write(fd, null_aligned, togo) < 0)
+ break;
+ }
+ }
+ if (i <= n_config)
+ return 0;
+
+ d->disk.crc = calc_crc(&d->disk, 512);
+ if (write(fd, &d->disk, 512) < 0)
+ return 0;
+
+ return 1;
+}
+
+static int __write_init_super_ddf(struct supertype *st)
+{
struct ddf_super *ddf = st->sb;
- int i;
struct dl *d;
- int n_config;
- int conf_size;
int attempts = 0;
int successes = 0;
- unsigned long long size, sector;
+ unsigned long long size;
+ char *null_aligned;
+ __u32 seq;
+
+ pr_state(ddf, __func__);
+ if (posix_memalign((void**)&null_aligned, 4096, NULL_CONF_SZ) != 0) {
+ return -ENOMEM;
+ }
+ memset(null_aligned, 0xff, NULL_CONF_SZ);
+
+ seq = ddf->active->seq + 1;
/* try to write updated metadata,
* if we catch a failure move on to the next disk
*/
get_dev_size(fd, NULL, &size);
size /= 512;
- ddf->anchor.workspace_lba = __cpu_to_be64(size - 32*1024*2);
- ddf->anchor.primary_lba = __cpu_to_be64(size - 16*1024*2);
- ddf->anchor.seq = __cpu_to_be32(1);
+ if (d->workspace_lba != 0)
+ ddf->anchor.workspace_lba = d->workspace_lba;
+ else
+ ddf->anchor.workspace_lba =
+ __cpu_to_be64(size - 32*1024*2);
+ if (d->primary_lba != 0)
+ ddf->anchor.primary_lba = d->primary_lba;
+ else
+ ddf->anchor.primary_lba =
+ __cpu_to_be64(size - 16*1024*2);
+ if (d->secondary_lba != 0)
+ ddf->anchor.secondary_lba = d->secondary_lba;
+ else
+ ddf->anchor.secondary_lba =
+ __cpu_to_be64(size - 32*1024*2);
+ ddf->anchor.seq = seq;
memcpy(&ddf->primary, &ddf->anchor, 512);
memcpy(&ddf->secondary, &ddf->anchor, 512);
ddf->anchor.seq = 0xFFFFFFFF; /* no sequencing in anchor */
ddf->anchor.crc = calc_crc(&ddf->anchor, 512);
- ddf->primary.openflag = 0;
- ddf->primary.type = DDF_HEADER_PRIMARY;
-
- ddf->secondary.openflag = 0;
- ddf->secondary.type = DDF_HEADER_SECONDARY;
-
- ddf->primary.crc = calc_crc(&ddf->primary, 512);
- ddf->secondary.crc = calc_crc(&ddf->secondary, 512);
-
- sector = size - 16*1024*2;
- lseek64(fd, sector<<9, 0);
- if (write(fd, &ddf->primary, 512) < 0)
+ if (!__write_ddf_structure(d, ddf, DDF_HEADER_PRIMARY,
+ null_aligned))
continue;
- ddf->controller.crc = calc_crc(&ddf->controller, 512);
- if (write(fd, &ddf->controller, 512) < 0)
+ if (!__write_ddf_structure(d, ddf, DDF_HEADER_SECONDARY,
+ null_aligned))
continue;
- ddf->phys->crc = calc_crc(ddf->phys, ddf->pdsize);
-
- if (write(fd, ddf->phys, ddf->pdsize) < 0)
- continue;
-
- ddf->virt->crc = calc_crc(ddf->virt, ddf->vdsize);
- if (write(fd, ddf->virt, ddf->vdsize) < 0)
- continue;
-
- /* Now write lots of config records. */
- n_config = ddf->max_part;
- conf_size = ddf->conf_rec_len * 512;
- for (i = 0 ; i <= n_config ; i++) {
- struct vcl *c = d->vlist[i];
- if (i == n_config)
- c = (struct vcl*)d->spare;
-
- if (c) {
- c->conf.crc = calc_crc(&c->conf, conf_size);
- if (write(fd, &c->conf, conf_size) < 0)
- break;
- } else {
- char *null_aligned = (char*)((((unsigned long)null_conf)+511)&~511UL);
- if (null_conf[0] != 0xff)
- memset(null_conf, 0xff, sizeof(null_conf));
- int togo = conf_size;
- while (togo > sizeof(null_conf)-512) {
- if (write(fd, null_aligned, sizeof(null_conf)-512) < 0)
- break;
- togo -= sizeof(null_conf)-512;
- }
- if (write(fd, null_aligned, togo) < 0)
- break;
- }
- }
- if (i <= n_config)
- continue;
- d->disk.crc = calc_crc(&d->disk, 512);
- if (write(fd, &d->disk, 512) < 0)
- continue;
-
- /* Maybe do the same for secondary */
-
lseek64(fd, (size-1)*512, SEEK_SET);
if (write(fd, &ddf->anchor, 512) < 0)
continue;
successes++;
}
-
- if (do_close)
- for (d = ddf->dlist; d; d=d->next) {
- close(d->fd);
- d->fd = -1;
- }
+ free(null_aligned);
return attempts != successes;
}
/* First the virtual disk. We have a slightly fake header */
len = sizeof(struct virtual_disk) + sizeof(struct virtual_entry);
- vd = malloc(len);
+ vd = xmalloc(len);
*vd = *ddf->virt;
vd->entries[0] = ddf->virt->entries[currentconf->vcnum];
vd->populated_vdes = __cpu_to_be16(currentconf->vcnum);
/* Then the vd_config */
len = ddf->conf_rec_len * 512;
- vc = malloc(len);
+ vc = xmalloc(len);
memcpy(vc, ¤tconf->conf, len);
append_metadata_update(st, vc, len);
/* FIXME I need to close the fds! */
return 0;
- } else {
+ } else {
struct dl *d;
for (d = ddf->dlist; d; d=d->next)
- while (Kill(d->devname, NULL, 0, 1, 1) == 0);
- return __write_init_super_ddf(st, 1);
+ while (Kill(d->devname, NULL, 0, -1, 1) == 0);
+ return __write_init_super_ddf(st);
}
}
#endif
-static __u64 avail_size_ddf(struct supertype *st, __u64 devsize)
+static __u64 avail_size_ddf(struct supertype *st, __u64 devsize,
+ unsigned long long data_offset)
{
/* We must reserve the last 32Meg */
if (devsize <= 32*1024*2)
int cnt = 0;
for (dl = ddf->dlist; dl ; dl=dl->next) {
- dl->raiddisk = -1;
+ dl->raiddisk = -1;
dl->esize = 0;
}
/* Now find largest extent on each device */
free(e);
}
if (cnt < raiddisks) {
- fprintf(stderr, Name ": not enough devices with space to create array.\n");
+ pr_err("not enough devices with space to create array.\n");
return 0; /* No enough free spaces large enough */
}
if (size == 0) {
continue;
/* This is bigger than 'size', see if there are enough */
cnt = 0;
- for (dl2 = dl; dl2 ; dl2=dl2->next)
+ for (dl2 = ddf->dlist; dl2 ; dl2=dl2->next)
if (dl2->esize >= dl->esize)
cnt++;
if (cnt >= raiddisks)
}
*freesize = size;
if (size < 32) {
- fprintf(stderr, Name ": not enough spare devices to create array.\n");
+ pr_err("not enough spare devices to create array.\n");
return 0;
}
}
for (dl = ddf->dlist ; dl && cnt < raiddisks ; dl=dl->next) {
if (dl->esize < size)
continue;
-
+
dl->raiddisk = cnt;
cnt++;
}
return 1;
}
-
-
static int
validate_geometry_ddf_container(struct supertype *st,
int level, int layout, int raiddisks,
int chunk, unsigned long long size,
+ unsigned long long data_offset,
char *dev, unsigned long long *freesize,
int verbose);
static int validate_geometry_ddf_bvd(struct supertype *st,
int level, int layout, int raiddisks,
- int chunk, unsigned long long size,
+ int *chunk, unsigned long long size,
+ unsigned long long data_offset,
char *dev, unsigned long long *freesize,
int verbose);
static int validate_geometry_ddf(struct supertype *st,
int level, int layout, int raiddisks,
- int chunk, unsigned long long size,
+ int *chunk, unsigned long long size,
+ unsigned long long data_offset,
char *dev, unsigned long long *freesize,
int verbose)
{
* If given BVDs, we make an SVD, changing all the GUIDs in the process.
*/
+ if (chunk && *chunk == UnSet)
+ *chunk = DEFAULT_CHUNK;
+
+ if (level == -1000000) level = LEVEL_CONTAINER;
if (level == LEVEL_CONTAINER) {
/* Must be a fresh device to add to a container */
return validate_geometry_ddf_container(st, level, layout,
- raiddisks, chunk,
- size, dev, freesize,
+ raiddisks, chunk?*chunk:0,
+ size, data_offset, dev,
+ freesize,
verbose);
}
break;
if (ddf_level_num[i].num1 == MAXINT) {
if (verbose)
- fprintf(stderr, Name ": DDF does not support level %d arrays\n",
- level);
+ pr_err("DDF does not support level %d arrays\n",
+ level);
return 0;
}
/* Should check layout? etc */
* chosen so that add_to_super/getinfo_super
* can return them.
*/
- return reserve_space(st, raiddisks, size, chunk, freesize);
+ return reserve_space(st, raiddisks, size, chunk?*chunk:0, freesize);
}
return 1;
}
* Should make a distinction one day.
*/
return validate_geometry_ddf_bvd(st, level, layout, raiddisks,
- chunk, size, dev, freesize,
+ chunk, size, data_offset, dev,
+ freesize,
verbose);
}
/* This is the first device for the array.
*/
fd = open(dev, O_RDONLY|O_EXCL, 0);
if (fd >= 0) {
- sra = sysfs_read(fd, 0, GET_VERSION);
+ sra = sysfs_read(fd, NULL, GET_VERSION);
close(fd);
if (sra && sra->array.major_version == -1 &&
strcmp(sra->text_version, "ddf") == 0) {
}
if (verbose)
- fprintf(stderr,
- Name ": ddf: Cannot create this array "
- "on device %s - a container is required.\n",
- dev);
+ pr_err("ddf: Cannot create this array "
+ "on device %s - a container is required.\n",
+ dev);
return 0;
}
if (errno != EBUSY || (fd = open(dev, O_RDONLY, 0)) < 0) {
if (verbose)
- fprintf(stderr, Name ": ddf: Cannot open %s: %s\n",
- dev, strerror(errno));
+ pr_err("ddf: Cannot open %s: %s\n",
+ dev, strerror(errno));
return 0;
}
/* Well, it is in use by someone, maybe a 'ddf' container. */
if (cfd < 0) {
close(fd);
if (verbose)
- fprintf(stderr, Name ": ddf: Cannot use %s: %s\n",
- dev, strerror(EBUSY));
+ pr_err("ddf: Cannot use %s: %s\n",
+ dev, strerror(EBUSY));
return 0;
}
- sra = sysfs_read(cfd, 0, GET_VERSION);
+ sra = sysfs_read(cfd, NULL, GET_VERSION);
close(fd);
if (sra && sra->array.major_version == -1 &&
strcmp(sra->text_version, "ddf") == 0) {
* and try to create a bvd
*/
struct ddf_super *ddf;
- if (load_super_ddf_all(st, cfd, (void **)&ddf, NULL, 1) == 0) {
+ if (load_super_ddf_all(st, cfd, (void **)&ddf, NULL) == 0) {
st->sb = ddf;
- st->container_dev = fd2devnum(cfd);
+ strcpy(st->container_devnm, fd2devnm(cfd));
close(cfd);
return validate_geometry_ddf_bvd(st, level, layout,
raiddisks, chunk, size,
+ data_offset,
dev, freesize,
verbose);
}
validate_geometry_ddf_container(struct supertype *st,
int level, int layout, int raiddisks,
int chunk, unsigned long long size,
+ unsigned long long data_offset,
char *dev, unsigned long long *freesize,
int verbose)
{
fd = open(dev, O_RDONLY|O_EXCL, 0);
if (fd < 0) {
if (verbose)
- fprintf(stderr, Name ": ddf: Cannot open %s: %s\n",
- dev, strerror(errno));
+ pr_err("ddf: Cannot open %s: %s\n",
+ dev, strerror(errno));
return 0;
}
if (!get_dev_size(fd, dev, &ldsize)) {
}
close(fd);
- *freesize = avail_size_ddf(st, ldsize >> 9);
+ *freesize = avail_size_ddf(st, ldsize >> 9, INVALID_SECTORS);
if (*freesize == 0)
return 0;
static int validate_geometry_ddf_bvd(struct supertype *st,
int level, int layout, int raiddisks,
- int chunk, unsigned long long size,
+ int *chunk, unsigned long long size,
+ unsigned long long data_offset,
char *dev, unsigned long long *freesize,
int verbose)
{
/* ddf/bvd supports lots of things, but not containers */
if (level == LEVEL_CONTAINER) {
if (verbose)
- fprintf(stderr, Name ": DDF cannot create a container within an container\n");
+ pr_err("DDF cannot create a container within an container\n");
return 0;
}
/* We must have the container info already read in. */
}
if (dcnt < raiddisks) {
if (verbose)
- fprintf(stderr,
- Name ": ddf: Not enough devices with "
- "space for this array (%d < %d)\n",
- dcnt, raiddisks);
+ pr_err("ddf: Not enough devices with "
+ "space for this array (%d < %d)\n",
+ dcnt, raiddisks);
return 0;
}
return 1;
if ((S_IFMT & stb.st_mode) != S_IFBLK)
return 0;
for (dl = ddf->dlist ; dl ; dl = dl->next) {
- if (dl->major == major(stb.st_rdev) &&
- dl->minor == minor(stb.st_rdev))
+ if (dl->major == (int)major(stb.st_rdev) &&
+ dl->minor == (int)minor(stb.st_rdev))
break;
}
if (!dl) {
if (verbose)
- fprintf(stderr, Name ": ddf: %s is not in the "
- "same DDF set\n",
- dev);
+ pr_err("ddf: %s is not in the "
+ "same DDF set\n",
+ dev);
return 0;
}
e = get_extents(ddf, dl);
maxsize = 0;
i = 0;
if (e) do {
- unsigned long long esize;
- esize = e[i].start - pos;
- if (esize >= maxsize)
- maxsize = esize;
- pos = e[i].start + e[i].size;
- i++;
- } while (e[i-1].size);
+ unsigned long long esize;
+ esize = e[i].start - pos;
+ if (esize >= maxsize)
+ maxsize = esize;
+ pos = e[i].start + e[i].size;
+ i++;
+ } while (e[i-1].size);
*freesize = maxsize;
// FIXME here I am
}
static int load_super_ddf_all(struct supertype *st, int fd,
- void **sbp, char *devname, int keep_fd)
+ void **sbp, char *devname)
{
struct mdinfo *sra;
struct ddf_super *super;
int seq;
char nm[20];
int dfd;
- int devnum = fd2devnum(fd);
- enum sysfs_read_flags flags;
-
- flags = GET_LEVEL|GET_VERSION|GET_DEVS|GET_STATE;
- if (mdmon_running(devnum))
- flags |= SKIP_GONE_DEVS;
- sra = sysfs_read(fd, 0, flags);
+ sra = sysfs_read(fd, 0, GET_LEVEL|GET_VERSION|GET_DEVS|GET_STATE);
if (!sra)
return 1;
if (sra->array.major_version != -1 ||
int rv;
sprintf(nm, "%d:%d", sd->disk.major, sd->disk.minor);
- dfd = dev_open(nm, keep_fd? O_RDWR : O_RDONLY);
+ dfd = dev_open(nm, O_RDWR);
if (dfd < 0)
return 2;
rv = load_ddf_headers(dfd, super, NULL);
if (rv == 0)
- rv = load_ddf_local(dfd, super, NULL, keep_fd);
- if (!keep_fd) close(dfd);
+ rv = load_ddf_local(dfd, super, NULL, 1);
if (rv)
return 1;
}
- if (st->subarray[0]) {
- unsigned long val;
- struct vcl *v;
- char *ep;
-
- val = strtoul(st->subarray, &ep, 10);
- if (*ep != '\0') {
- free(super);
- return 1;
- }
-
- for (v = super->conflist; v; v = v->next)
- if (v->vcnum == val)
- super->currentconf = v;
- if (!super->currentconf) {
- free(super);
- return 1;
- }
- }
*sbp = super;
if (st->ss == NULL) {
st->ss = &super_ddf;
st->minor_version = 0;
st->max_devs = 512;
- st->container_dev = fd2devnum(fd);
}
- st->loaded_container = 1;
+ strcpy(st->container_devnm, fd2devnm(fd));
return 0;
}
+
+static int load_container_ddf(struct supertype *st, int fd,
+ char *devname)
+{
+ return load_super_ddf_all(st, fd, &st->sb, devname);
+}
+
#endif /* MDASSEMBLE */
-static struct mdinfo *container_content_ddf(struct supertype *st)
+static int check_secondary(const struct vcl *vc)
+{
+ const struct vd_config *conf = &vc->conf;
+ int i;
+
+ /* The only DDF secondary RAID level md can support is
+ * RAID 10, if the stripe sizes and Basic volume sizes
+ * are all equal.
+ * Other configurations could in theory be supported by exposing
+ * the BVDs to user space and using device mapper for the secondary
+ * mapping. So far we don't support that.
+ */
+
+ __u64 sec_elements[4] = {0, 0, 0, 0};
+#define __set_sec_seen(n) (sec_elements[(n)>>6] |= (1<<((n)&63)))
+#define __was_sec_seen(n) ((sec_elements[(n)>>6] & (1<<((n)&63))) != 0)
+
+ if (vc->other_bvds == NULL) {
+ pr_err("No BVDs for secondary RAID found\n");
+ return -1;
+ }
+ if (conf->prl != DDF_RAID1) {
+ pr_err("Secondary RAID level only supported for mirrored BVD\n");
+ return -1;
+ }
+ if (conf->srl != DDF_2STRIPED && conf->srl != DDF_2SPANNED) {
+ pr_err("Secondary RAID level %d is unsupported\n",
+ conf->srl);
+ return -1;
+ }
+ __set_sec_seen(conf->sec_elmnt_seq);
+ for (i = 0; i < conf->sec_elmnt_count-1; i++) {
+ const struct vd_config *bvd = vc->other_bvds[i];
+ if (bvd == NULL) {
+ pr_err("BVD %d is missing\n", i+1);
+ return -1;
+ }
+ if (bvd->srl != conf->srl) {
+ pr_err("Inconsistent secondary RAID level across BVDs\n");
+ return -1;
+ }
+ if (bvd->prl != conf->prl) {
+ pr_err("Different RAID levels for BVDs are unsupported\n");
+ return -1;
+ }
+ if (bvd->prim_elmnt_count != conf->prim_elmnt_count) {
+ pr_err("All BVDs must have the same number of primary elements\n");
+ return -1;
+ }
+ if (bvd->chunk_shift != conf->chunk_shift) {
+ pr_err("Different strip sizes for BVDs are unsupported\n");
+ return -1;
+ }
+ if (bvd->array_blocks != conf->array_blocks) {
+ pr_err("Different BVD sizes are unsupported\n");
+ return -1;
+ }
+ __set_sec_seen(bvd->sec_elmnt_seq);
+ }
+ for (i = 0; i < conf->sec_elmnt_count; i++) {
+ if (!__was_sec_seen(i)) {
+ pr_err("BVD %d is missing\n", i);
+ return -1;
+ }
+ }
+ return 0;
+}
+
+#define NO_SUCH_REFNUM (0xFFFFFFFF)
+static unsigned int get_pd_index_from_refnum(const struct vcl *vc,
+ __u32 refnum, unsigned int nmax,
+ const struct vd_config **bvd,
+ unsigned int *idx)
+{
+ unsigned int i, j, n, sec, cnt;
+
+ cnt = __be16_to_cpu(vc->conf.prim_elmnt_count);
+ sec = (vc->conf.sec_elmnt_count == 1 ? 0 : vc->conf.sec_elmnt_seq);
+
+ for (i = 0, j = 0 ; i < nmax ; i++) {
+ /* j counts valid entries for this BVD */
+ if (vc->conf.phys_refnum[i] != 0xffffffff)
+ j++;
+ if (vc->conf.phys_refnum[i] == refnum) {
+ *bvd = &vc->conf;
+ *idx = i;
+ return sec * cnt + j - 1;
+ }
+ }
+ if (vc->other_bvds == NULL)
+ goto bad;
+
+ for (n = 1; n < vc->conf.sec_elmnt_count; n++) {
+ struct vd_config *vd = vc->other_bvds[n-1];
+ if (vd == NULL)
+ continue;
+ sec = vd->sec_elmnt_seq;
+ for (i = 0, j = 0 ; i < nmax ; i++) {
+ if (vd->phys_refnum[i] != 0xffffffff)
+ j++;
+ if (vd->phys_refnum[i] == refnum) {
+ *bvd = vd;
+ *idx = i;
+ return sec * cnt + j - 1;
+ }
+ }
+ }
+bad:
+ *bvd = NULL;
+ return NO_SUCH_REFNUM;
+}
+
+static struct mdinfo *container_content_ddf(struct supertype *st, char *subarray)
{
/* Given a container loaded by load_super_ddf_all,
* extract information about all the arrays into
for (vc = ddf->conflist ; vc ; vc=vc->next)
{
- int i;
- int j;
+ unsigned int i;
+ unsigned int j;
struct mdinfo *this;
- this = malloc(sizeof(*this));
- memset(this, 0, sizeof(*this));
+ char *ep;
+ __u32 *cptr;
+ unsigned int pd;
+
+ if (subarray &&
+ (strtoul(subarray, &ep, 10) != vc->vcnum ||
+ *ep != '\0'))
+ continue;
+
+ if (vc->conf.sec_elmnt_count > 1) {
+ if (check_secondary(vc) != 0)
+ continue;
+ }
+
+ this = xcalloc(1, sizeof(*this));
this->next = rest;
rest = this;
- this->array.level = map_num1(ddf_level_num, vc->conf.prl);
- this->array.raid_disks =
- __be16_to_cpu(vc->conf.prim_elmnt_count);
- this->array.layout = rlq_to_layout(vc->conf.rlq, vc->conf.prl,
- this->array.raid_disks);
+ if (vc->conf.sec_elmnt_count == 1) {
+ this->array.level = map_num1(ddf_level_num,
+ vc->conf.prl);
+ this->array.raid_disks =
+ __be16_to_cpu(vc->conf.prim_elmnt_count);
+ this->array.layout =
+ rlq_to_layout(vc->conf.rlq, vc->conf.prl,
+ this->array.raid_disks);
+ } else {
+ /* The only supported layout is RAID 10.
+ * Compatibility has been checked in check_secondary()
+ * above.
+ */
+ this->array.level = 10;
+ this->array.raid_disks =
+ __be16_to_cpu(vc->conf.prim_elmnt_count)
+ * vc->conf.sec_elmnt_count;
+ this->array.layout = 0x100 |
+ __be16_to_cpu(vc->conf.prim_elmnt_count);
+ }
this->array.md_minor = -1;
this->array.major_version = -1;
this->array.minor_version = -2;
- this->array.ctime = DECADE +
- __be32_to_cpu(*(__u32*)(vc->conf.guid+16));
+ cptr = (__u32 *)(vc->conf.guid + 16);
+ this->array.ctime = DECADE + __be32_to_cpu(*cptr);
this->array.utime = DECADE +
__be32_to_cpu(vc->conf.timestamp);
this->array.chunk_size = 512 << vc->conf.chunk_shift;
ddf->currentconf = NULL;
sprintf(this->text_version, "/%s/%d",
- devnum2devname(st->container_dev),
- this->container_member);
+ st->container_devnm, this->container_member);
- for (i=0 ; i < ddf->mppe ; i++) {
+ for (pd = 0; pd < __be16_to_cpu(ddf->phys->used_pdes); pd++) {
struct mdinfo *dev;
struct dl *d;
+ const struct vd_config *bvd;
+ unsigned int iphys;
+ __u64 *lba_offset;
+ int stt;
- if (vc->conf.phys_refnum[i] == 0xFFFFFFFF)
+ if (ddf->phys->entries[pd].refnum == 0xFFFFFFFF)
+ continue;
+
+ stt = __be16_to_cpu(ddf->phys->entries[pd].state);
+ if ((stt & (DDF_Online|DDF_Failed|DDF_Rebuilding))
+ != DDF_Online)
+ continue;
+
+ i = get_pd_index_from_refnum(
+ vc, ddf->phys->entries[pd].refnum,
+ ddf->mppe, &bvd, &iphys);
+ if (i == NO_SUCH_REFNUM)
continue;
this->array.working_disks++;
for (d = ddf->dlist; d ; d=d->next)
- if (d->disk.refnum == vc->conf.phys_refnum[i])
+ if (d->disk.refnum ==
+ ddf->phys->entries[pd].refnum)
break;
if (d == NULL)
/* Haven't found that one yet, maybe there are others */
continue;
- dev = malloc(sizeof(*dev));
- memset(dev, 0, sizeof(*dev));
+ dev = xcalloc(1, sizeof(*dev));
dev->next = this->devs;
this->devs = dev;
dev->recovery_start = MaxSector;
dev->events = __be32_to_cpu(ddf->primary.seq);
- dev->data_offset = __be64_to_cpu(vc->lba_offset[i]);
- dev->component_size = __be64_to_cpu(vc->conf.blocks);
+ lba_offset = (__u64 *)&bvd->phys_refnum[ddf->mppe];
+ dev->data_offset = __be64_to_cpu(lba_offset[iphys]);
+ dev->component_size = __be64_to_cpu(bvd->blocks);
if (d->devname)
strcpy(dev->name, d->devname);
}
*/
struct ddf_super *first = st->sb;
struct ddf_super *second = tst->sb;
+ struct dl *dl1, *dl2;
+ struct vcl *vl1, *vl2;
+ unsigned int max_vds, max_pds, pd, vd;
if (!first) {
st->sb = tst->sb;
if (memcmp(first->anchor.guid, second->anchor.guid, DDF_GUID_LEN) != 0)
return 2;
+ if (first->anchor.seq != second->anchor.seq) {
+ dprintf("%s: sequence number mismatch %u/%u\n", __func__,
+ __be32_to_cpu(first->anchor.seq),
+ __be32_to_cpu(second->anchor.seq));
+ return 3;
+ }
+ if (first->max_part != second->max_part ||
+ first->phys->used_pdes != second->phys->used_pdes ||
+ first->virt->populated_vdes != second->virt->populated_vdes) {
+ dprintf("%s: PD/VD number mismatch\n", __func__);
+ return 3;
+ }
+
+ max_pds = __be16_to_cpu(first->phys->used_pdes);
+ for (dl2 = second->dlist; dl2; dl2 = dl2->next) {
+ for (pd = 0; pd < max_pds; pd++)
+ if (first->phys->entries[pd].refnum == dl2->disk.refnum)
+ break;
+ if (pd == max_pds) {
+ dprintf("%s: no match for disk %08x\n", __func__,
+ __be32_to_cpu(dl2->disk.refnum));
+ return 3;
+ }
+ }
+
+ max_vds = __be16_to_cpu(first->active->max_vd_entries);
+ for (vl2 = second->conflist; vl2; vl2 = vl2->next) {
+ if (vl2->conf.magic != DDF_VD_CONF_MAGIC)
+ continue;
+ for (vd = 0; vd < max_vds; vd++)
+ if (!memcmp(first->virt->entries[vd].guid,
+ vl2->conf.guid, DDF_GUID_LEN))
+ break;
+ if (vd == max_vds) {
+ dprintf("%s: no match for VD config\n", __func__);
+ return 3;
+ }
+ }
/* FIXME should I look at anything else? */
+
+ /*
+ At this point we are fairly sure that the meta data matches.
+ But the new disk may contain additional local data.
+ Add it to the super block.
+ */
+ for (vl2 = second->conflist; vl2; vl2 = vl2->next) {
+ for (vl1 = first->conflist; vl1; vl1 = vl1->next)
+ if (!memcmp(vl1->conf.guid, vl2->conf.guid,
+ DDF_GUID_LEN))
+ break;
+ if (vl1) {
+ if (vl1->other_bvds != NULL &&
+ vl1->conf.sec_elmnt_seq !=
+ vl2->conf.sec_elmnt_seq) {
+ dprintf("%s: adding BVD %u\n", __func__,
+ vl2->conf.sec_elmnt_seq);
+ add_other_bvd(vl1, &vl2->conf,
+ first->conf_rec_len*512);
+ }
+ continue;
+ }
+
+ if (posix_memalign((void **)&vl1, 512,
+ (first->conf_rec_len*512 +
+ offsetof(struct vcl, conf))) != 0) {
+ pr_err("%s could not allocate vcl buf\n",
+ __func__);
+ return 3;
+ }
+
+ vl1->next = first->conflist;
+ vl1->block_sizes = NULL;
+ if (vl1->conf.sec_elmnt_count > 1) {
+ vl1->other_bvds = xcalloc(vl2->conf.sec_elmnt_count - 1,
+ sizeof(struct vd_config *));
+ } else
+ vl1->other_bvds = NULL;
+ memcpy(&vl1->conf, &vl2->conf, first->conf_rec_len*512);
+ vl1->lba_offset = (__u64 *)
+ &vl1->conf.phys_refnum[first->mppe];
+ for (vd = 0; vd < max_vds; vd++)
+ if (!memcmp(first->virt->entries[vd].guid,
+ vl1->conf.guid, DDF_GUID_LEN))
+ break;
+ vl1->vcnum = vd;
+ dprintf("%s: added config for VD %u\n", __func__, vl1->vcnum);
+ first->conflist = vl1;
+ }
+
+ for (dl2 = second->dlist; dl2; dl2 = dl2->next) {
+ for (dl1 = first->dlist; dl1; dl1 = dl1->next)
+ if (dl1->disk.refnum == dl2->disk.refnum)
+ break;
+ if (dl1)
+ continue;
+
+ if (posix_memalign((void **)&dl1, 512,
+ sizeof(*dl1) + (first->max_part) * sizeof(dl1->vlist[0]))
+ != 0) {
+ pr_err("%s could not allocate disk info buffer\n",
+ __func__);
+ return 3;
+ }
+ memcpy(dl1, dl2, sizeof(*dl1));
+ dl1->mdupdate = NULL;
+ dl1->next = first->dlist;
+ dl1->fd = -1;
+ for (pd = 0; pd < max_pds; pd++)
+ if (first->phys->entries[pd].refnum == dl1->disk.refnum)
+ break;
+ dl1->pdnum = pd;
+ if (dl2->spare) {
+ if (posix_memalign((void **)&dl1->spare, 512,
+ first->conf_rec_len*512) != 0) {
+ pr_err("%s could not allocate spare info buf\n",
+ __func__);
+ return 3;
+ }
+ memcpy(dl1->spare, dl2->spare, first->conf_rec_len*512);
+ }
+ for (vd = 0 ; vd < first->max_part ; vd++) {
+ if (!dl2->vlist[vd]) {
+ dl1->vlist[vd] = NULL;
+ continue;
+ }
+ for (vl1 = first->conflist; vl1; vl1 = vl1->next) {
+ if (!memcmp(vl1->conf.guid,
+ dl2->vlist[vd]->conf.guid,
+ DDF_GUID_LEN))
+ break;
+ dl1->vlist[vd] = vl1;
+ }
+ }
+ first->dlist = dl1;
+ dprintf("%s: added disk %d: %08x\n", __func__, dl1->pdnum,
+ dl1->disk.refnum);
+ }
+
return 0;
}
else
ddf->virt->entries[inst].state |= DDF_state_inconsistent;
if (old != ddf->virt->entries[inst].state)
- ddf->updates_pending = 1;
+ ddf_set_updates_pending(ddf);
old = ddf->virt->entries[inst].init_state;
ddf->virt->entries[inst].init_state &= ~DDF_initstate_mask;
else
ddf->virt->entries[inst].init_state |= DDF_init_quick;
if (old != ddf->virt->entries[inst].init_state)
- ddf->updates_pending = 1;
+ ddf_set_updates_pending(ddf);
dprintf("ddf mark %d %s %llu\n", inst, consistent?"clean":"dirty",
a->info.resync_start);
return consistent;
}
+#define container_of(ptr, type, member) ({ \
+ const typeof( ((type *)0)->member ) *__mptr = (ptr); \
+ (type *)( (char *)__mptr - offsetof(type,member) );})
/*
* The state of each disk is stored in the global phys_disk structure
* in phys_disk.entries[n].state.
static void ddf_set_disk(struct active_array *a, int n, int state)
{
struct ddf_super *ddf = a->container->sb;
- int inst = a->info.container_member;
+ unsigned int inst = a->info.container_member;
struct vd_config *vc = find_vdcr(ddf, inst);
int pd = find_phys(ddf, vc->phys_refnum[n]);
int i, st, working;
+ struct mdinfo *mdi;
+ struct dl *dl;
if (vc == NULL) {
dprintf("ddf: cannot find instance %d!!\n", inst);
return;
}
- if (pd < 0) {
- /* disk doesn't currently exist. If it is now in_sync,
- * insert it. */
+ /* Find the matching slot in 'info'. */
+ for (mdi = a->info.devs; mdi; mdi = mdi->next)
+ if (mdi->disk.raid_disk == n)
+ break;
+ if (!mdi)
+ return;
+
+ /* and find the 'dl' entry corresponding to that. */
+ for (dl = ddf->dlist; dl; dl = dl->next)
+ if (mdi->state_fd >= 0 &&
+ mdi->disk.major == dl->major &&
+ mdi->disk.minor == dl->minor)
+ break;
+ if (!dl)
+ return;
+
+ if (pd < 0 || pd != dl->pdnum) {
+ /* disk doesn't currently exist or has changed.
+ * If it is now in_sync, insert it. */
if ((state & DS_INSYNC) && ! (state & DS_FAULTY)) {
- /* Find dev 'n' in a->info->devs, determine the
- * ddf refnum, and set vc->phys_refnum and update
- * phys->entries[]
- */
- /* FIXME */
+ struct vcl *vcl;
+ pd = dl->pdnum;
+ vc->phys_refnum[n] = dl->disk.refnum;
+ vcl = container_of(vc, struct vcl, conf);
+ vcl->lba_offset[n] = mdi->data_offset;
+ ddf->phys->entries[pd].type &=
+ ~__cpu_to_be16(DDF_Global_Spare);
+ ddf->phys->entries[pd].type |=
+ __cpu_to_be16(DDF_Active_in_VD);
+ ddf_set_updates_pending(ddf);
}
} else {
int old = ddf->phys->entries[pd].state;
ddf->phys->entries[pd].state &= __cpu_to_be16(~DDF_Rebuilding);
}
if (old != ddf->phys->entries[pd].state)
- ddf->updates_pending = 1;
+ ddf_set_updates_pending(ddf);
}
dprintf("ddf: set_disk %d to %x\n", n, state);
if (working == a->info.array.raid_disks)
state = DDF_state_optimal;
else switch(vc->prl) {
- case DDF_RAID0:
- case DDF_CONCAT:
- case DDF_JBOD:
- state = DDF_state_failed;
- break;
- case DDF_RAID1:
- if (working == 0)
- state = DDF_state_failed;
- break;
- case DDF_RAID4:
- case DDF_RAID5:
- if (working < a->info.array.raid_disks-1)
- state = DDF_state_failed;
- break;
- case DDF_RAID6:
- if (working < a->info.array.raid_disks-2)
+ case DDF_RAID0:
+ case DDF_CONCAT:
+ case DDF_JBOD:
state = DDF_state_failed;
- else if (working == a->info.array.raid_disks-1)
- state = DDF_state_part_optimal;
- break;
- }
+ break;
+ case DDF_RAID1:
+ if (working == 0)
+ state = DDF_state_failed;
+ else if (working == 2 && state == DDF_state_degraded)
+ state = DDF_state_part_optimal;
+ break;
+ case DDF_RAID4:
+ case DDF_RAID5:
+ if (working < a->info.array.raid_disks-1)
+ state = DDF_state_failed;
+ break;
+ case DDF_RAID6:
+ if (working < a->info.array.raid_disks-2)
+ state = DDF_state_failed;
+ else if (working == a->info.array.raid_disks-1)
+ state = DDF_state_part_optimal;
+ break;
+ }
if (ddf->virt->entries[inst].state !=
((ddf->virt->entries[inst].state & ~DDF_state_mask)
ddf->virt->entries[inst].state =
(ddf->virt->entries[inst].state & ~DDF_state_mask)
| state;
- ddf->updates_pending = 1;
+ ddf_set_updates_pending(ddf);
}
}
if (!ddf->updates_pending)
return;
ddf->updates_pending = 0;
- __write_init_super_ddf(st, 0);
+ __write_init_super_ddf(st);
dprintf("ddf: sync_metadata\n");
}
* our actions.
* Possible update are:
* DDF_PHYS_RECORDS_MAGIC
- * Add a new physical device. Changes to this record
- * only happen implicitly.
+ * Add a new physical device or remove an old one.
+ * Changes to this record only happen implicitly.
* used_pdes is the device number.
* DDF_VIRT_RECORDS_MAGIC
* Add a new VD. Possibly also change the 'access' bits.
struct vd_config *vc;
struct vcl *vcl;
struct dl *dl;
- int mppe;
- int ent;
+ unsigned int mppe;
+ unsigned int ent;
+ unsigned int pdnum, pd2;
dprintf("Process update %x\n", *magic);
ent = __be16_to_cpu(pd->used_pdes);
if (ent >= __be16_to_cpu(ddf->phys->max_pdes))
return;
+ if (pd->entries[0].state & __cpu_to_be16(DDF_Missing)) {
+ struct dl **dlp;
+ /* removing this disk. */
+ ddf->phys->entries[ent].state |= __cpu_to_be16(DDF_Missing);
+ for (dlp = &ddf->dlist; *dlp; dlp = &(*dlp)->next) {
+ struct dl *dl = *dlp;
+ if (dl->pdnum == (signed)ent) {
+ close(dl->fd);
+ dl->fd = -1;
+ /* FIXME this doesn't free
+ * dl->devname */
+ update->space = dl;
+ *dlp = dl->next;
+ break;
+ }
+ }
+ ddf_set_updates_pending(ddf);
+ return;
+ }
if (!all_ff(ddf->phys->entries[ent].guid))
return;
ddf->phys->entries[ent] = pd->entries[0];
ddf->phys->used_pdes = __cpu_to_be16(1 +
- __be16_to_cpu(ddf->phys->used_pdes));
- ddf->updates_pending = 1;
+ __be16_to_cpu(ddf->phys->used_pdes));
+ ddf_set_updates_pending(ddf);
if (ddf->add_list) {
struct active_array *a;
struct dl *al = ddf->add_list;
return;
ddf->virt->entries[ent] = vd->entries[0];
ddf->virt->populated_vdes = __cpu_to_be16(1 +
- __be16_to_cpu(ddf->virt->populated_vdes));
- ddf->updates_pending = 1;
+ __be16_to_cpu(ddf->virt->populated_vdes));
+ ddf_set_updates_pending(ddf);
break;
case DDF_VD_CONF_MAGIC:
dprintf("len %d %d\n", update->len, ddf->conf_rec_len);
mppe = __be16_to_cpu(ddf->anchor.max_primary_element_entries);
- if (update->len != ddf->conf_rec_len * 512)
+ if ((unsigned)update->len != ddf->conf_rec_len * 512)
return;
vc = (struct vd_config*)update->buf;
for (vcl = ddf->conflist; vcl ; vcl = vcl->next)
memcpy(&vcl->conf, vc, update->len);
vcl->lba_offset = (__u64*)
&vcl->conf.phys_refnum[mppe];
+ for (ent = 0;
+ ent < __be16_to_cpu(ddf->virt->populated_vdes);
+ ent++)
+ if (memcmp(vc->guid, ddf->virt->entries[ent].guid,
+ DDF_GUID_LEN) == 0) {
+ vcl->vcnum = ent;
+ break;
+ }
ddf->conflist = vcl;
}
+ /* Set DDF_Transition on all Failed devices - to help
+ * us detect those that are no longer in use
+ */
+ for (pdnum = 0; pdnum < __be16_to_cpu(ddf->phys->used_pdes); pdnum++)
+ if (ddf->phys->entries[pdnum].state
+ & __be16_to_cpu(DDF_Failed))
+ ddf->phys->entries[pdnum].state
+ |= __be16_to_cpu(DDF_Transition);
/* Now make sure vlist is correct for each dl. */
for (dl = ddf->dlist; dl; dl = dl->next) {
- int dn;
- int vn = 0;
+ unsigned int dn;
+ unsigned int vn = 0;
+ int in_degraded = 0;
for (vcl = ddf->conflist; vcl ; vcl = vcl->next)
for (dn=0; dn < ddf->mppe ; dn++)
if (vcl->conf.phys_refnum[dn] ==
dl->disk.refnum) {
+ int vstate;
dprintf("dev %d has %p at %d\n",
dl->pdnum, vcl, vn);
+ /* Clear the Transition flag */
+ if (ddf->phys->entries[dl->pdnum].state
+ & __be16_to_cpu(DDF_Failed))
+ ddf->phys->entries[dl->pdnum].state &=
+ ~__be16_to_cpu(DDF_Transition);
+
dl->vlist[vn++] = vcl;
+ vstate = ddf->virt->entries[vcl->vcnum].state
+ & DDF_state_mask;
+ if (vstate == DDF_state_degraded ||
+ vstate == DDF_state_part_optimal)
+ in_degraded = 1;
break;
}
while (vn < ddf->max_part)
if (dl->vlist[0]) {
ddf->phys->entries[dl->pdnum].type &=
~__cpu_to_be16(DDF_Global_Spare);
- ddf->phys->entries[dl->pdnum].type |=
- __cpu_to_be16(DDF_Active_in_VD);
+ if (!(ddf->phys->entries[dl->pdnum].type &
+ __cpu_to_be16(DDF_Active_in_VD))) {
+ ddf->phys->entries[dl->pdnum].type |=
+ __cpu_to_be16(DDF_Active_in_VD);
+ if (in_degraded)
+ ddf->phys->entries[dl->pdnum].state |=
+ __cpu_to_be16(DDF_Rebuilding);
+ }
}
if (dl->spare) {
ddf->phys->entries[dl->pdnum].type &=
DDF_Active_in_VD);
}
}
- ddf->updates_pending = 1;
+
+ /* Now remove any 'Failed' devices that are not part
+ * of any VD. They will have the Transition flag set.
+ * Once done, we need to update all dl->pdnum numbers.
+ */
+ pd2 = 0;
+ for (pdnum = 0; pdnum < __be16_to_cpu(ddf->phys->used_pdes); pdnum++)
+ if ((ddf->phys->entries[pdnum].state
+ & __be16_to_cpu(DDF_Failed))
+ && (ddf->phys->entries[pdnum].state
+ & __be16_to_cpu(DDF_Transition)))
+ /* skip this one */;
+ else if (pdnum == pd2)
+ pd2++;
+ else {
+ ddf->phys->entries[pd2] = ddf->phys->entries[pdnum];
+ for (dl = ddf->dlist; dl; dl = dl->next)
+ if (dl->pdnum == (int)pdnum)
+ dl->pdnum = pd2;
+ pd2++;
+ }
+ ddf->phys->used_pdes = __cpu_to_be16(pd2);
+ while (pd2 < pdnum) {
+ memset(ddf->phys->entries[pd2].guid, 0xff, DDF_GUID_LEN);
+ pd2++;
+ }
+
+ ddf_set_updates_pending(ddf);
break;
case DDF_SPARE_ASSIGN_MAGIC:
default: break;
__u32 *magic = (__u32*)update->buf;
if (*magic == DDF_VD_CONF_MAGIC)
if (posix_memalign(&update->space, 512,
- offsetof(struct vcl, conf)
- + ddf->conf_rec_len * 512) != 0)
+ offsetof(struct vcl, conf)
+ + ddf->conf_rec_len * 512) != 0)
update->space = NULL;
}
for (d = a->info.devs ; d ; d = d->next) {
if ((d->curr_state & DS_FAULTY) &&
- d->state_fd >= 0)
+ d->state_fd >= 0)
/* wait for Removal to happen */
return NULL;
if (d->state_fd >= 0)
int is_global = 0;
int is_dedicated = 0;
struct extent *ex;
- int j;
+ unsigned int j;
/* If in this array, skip */
for (d2 = a->info.devs ; d2 ; d2 = d2->next)
- if (d2->disk.major == dl->major &&
+ if (d2->state_fd >= 0 &&
+ d2->disk.major == dl->major &&
d2->disk.minor == dl->minor) {
dprintf("%x:%x already in array\n", dl->major, dl->minor);
break;
} else if (ddf->phys->entries[dl->pdnum].type &
__cpu_to_be16(DDF_Global_Spare)) {
is_global = 1;
+ } else if (!(ddf->phys->entries[dl->pdnum].state &
+ __cpu_to_be16(DDF_Failed))) {
+ /* we can possibly use some of this */
+ is_global = 1;
}
if ( ! (is_dedicated ||
(is_global && global_ok))) {
dprintf("%x:%x not suitable: %d %d\n", dl->major, dl->minor,
- is_dedicated, is_global);
+ is_dedicated, is_global);
continue;
}
esize = ex[j].start - pos;
if (esize >= a->info.component_size)
break;
- pos = ex[i].start + ex[i].size;
- i++;
- } while (ex[i-1].size);
+ pos = ex[j].start + ex[j].size;
+ j++;
+ } while (ex[j-1].size);
free(ex);
if (esize < a->info.component_size) {
- dprintf("%x:%x has no room: %llu %llu\n", dl->major, dl->minor,
+ dprintf("%x:%x has no room: %llu %llu\n",
+ dl->major, dl->minor,
esize, a->info.component_size);
/* No room */
continue;
}
/* Cool, we have a device with some space at pos */
- di = malloc(sizeof(*di));
- if (!di)
- continue;
- memset(di, 0, sizeof(*di));
+ di = xcalloc(1, sizeof(*di));
di->disk.number = i;
di->disk.raid_disk = i;
di->disk.major = dl->major;
* Create a metadata_update record to update the
* phys_refnum and lba_offset values
*/
- mu = malloc(sizeof(*mu));
- if (mu && posix_memalign(&mu->space, 512, sizeof(struct vcl)) != 0) {
+ mu = xmalloc(sizeof(*mu));
+ if (posix_memalign(&mu->space, 512, sizeof(struct vcl)) != 0) {
free(mu);
mu = NULL;
}
- if (!mu) {
- while (rv) {
- struct mdinfo *n = rv->next;
-
- free(rv);
- rv = n;
- }
- return NULL;
- }
-
- mu->buf = malloc(ddf->conf_rec_len * 512);
- mu->len = ddf->conf_rec_len;
+ mu->buf = xmalloc(ddf->conf_rec_len * 512);
+ mu->len = ddf->conf_rec_len * 512;
+ mu->space = NULL;
+ mu->space_list = NULL;
mu->next = *updates;
vc = find_vdcr(ddf, a->info.container_member);
memcpy(mu->buf, vc, ddf->conf_rec_len * 512);
}
}
+static void default_geometry_ddf(struct supertype *st, int *level, int *layout, int *chunk)
+{
+ if (level && *level == UnSet)
+ *level = LEVEL_CONTAINER;
+
+ if (level && layout && *layout == UnSet)
+ *layout = ddf_level_to_layout(*level);
+}
+
struct superswitch super_ddf = {
#ifndef MDASSEMBLE
.examine_super = examine_super_ddf,
.validate_geometry = validate_geometry_ddf,
.write_init_super = write_init_super_ddf,
.add_to_super = add_to_super_ddf,
+ .remove_from_super = remove_from_super_ddf,
+ .load_container = load_container_ddf,
#endif
.match_home = match_home_ddf,
.uuid_from_super= uuid_from_super_ddf,
.free_super = free_super_ddf,
.match_metadata_desc = match_metadata_desc_ddf,
.container_content = container_content_ddf,
- .default_layout = ddf_level_to_layout,
+ .default_geometry = default_geometry_ddf,
.external = 1,
projects / thirdparty / mdadm.git / blobdiff