X-Git-Url: http://git.ipfire.org/?p=thirdparty%2Fmdadm.git;a=blobdiff_plain;f=super-intel.c;h=cfc8904e2d2cca5df92a54933a3d50ca6ccc5dff;hp=17034bbd3edf93bd3a9fe8dcf7ef6173beb98f18;hb=676e87a806240b2bf32ad8738c59fe1c3b4b2b5b;hpb=d2bde6d3aa9468ddf0965f09907a666b92186e42 diff --git a/super-intel.c b/super-intel.c index 17034bbd..cfc8904e 100644 --- a/super-intel.c +++ b/super-intel.c @@ -95,15 +95,16 @@ #define IMSM_MAX_DEVICES 255 struct imsm_disk { __u8 serial[MAX_RAID_SERIAL_LEN];/* 0xD8 - 0xE7 ascii serial number */ - __u32 total_blocks; /* 0xE8 - 0xEB total blocks */ + __u32 total_blocks_lo; /* 0xE8 - 0xEB total blocks lo */ __u32 scsi_id; /* 0xEC - 0xEF scsi ID */ #define SPARE_DISK __cpu_to_le32(0x01) /* Spare */ #define CONFIGURED_DISK __cpu_to_le32(0x02) /* Member of some RaidDev */ #define FAILED_DISK __cpu_to_le32(0x04) /* Permanent failure */ __u32 status; /* 0xF0 - 0xF3 */ - __u32 owner_cfg_num; /* which config 0,1,2... owns this disk */ -#define IMSM_DISK_FILLERS 4 - __u32 filler[IMSM_DISK_FILLERS]; /* 0xF4 - 0x107 MPB_DISK_FILLERS for future expansion */ + __u32 owner_cfg_num; /* which config 0,1,2... owns this disk */ + __u32 total_blocks_hi; /* 0xF4 - 0xF5 total blocks hi */ +#define IMSM_DISK_FILLERS 3 + __u32 filler[IMSM_DISK_FILLERS]; /* 0xF5 - 0x107 MPB_DISK_FILLERS for future expansion */ }; /* map selector for map managment @@ -114,9 +115,9 @@ struct imsm_disk { /* RAID map configuration infos. */ struct imsm_map { - __u32 pba_of_lba0; /* start address of partition */ - __u32 blocks_per_member;/* blocks per member */ - __u32 num_data_stripes; /* number of data stripes */ + __u32 pba_of_lba0_lo; /* start address of partition */ + __u32 blocks_per_member_lo;/* blocks per member */ + __u32 num_data_stripes_lo; /* number of data stripes */ __u16 blocks_per_strip; __u8 map_state; /* Normal, Uninitialized, Degraded, Failed */ #define IMSM_T_STATE_NORMAL 0 @@ -131,7 +132,10 @@ struct imsm_map { __u8 num_domains; /* number of parity domains */ __u8 failed_disk_num; /* valid only when state is degraded */ __u8 ddf; - __u32 filler[7]; /* expansion area */ + __u32 pba_of_lba0_hi; + __u32 blocks_per_member_hi; + __u32 num_data_stripes_hi; + __u32 filler[4]; /* expansion area */ #define IMSM_ORD_REBUILD (1 << 24) __u32 disk_ord_tbl[1]; /* disk_ord_tbl[num_members], * top byte contains some flags @@ -231,7 +235,6 @@ struct bbm_log { struct bbm_log_entry mapped_block_entries[BBM_LOG_MAX_ENTRIES]; } __attribute__ ((__packed__)); - #ifndef MDASSEMBLE static char *map_state_str[] = { "normal", "uninitialized", "degraded", "failed" }; #endif @@ -245,7 +248,6 @@ static char *map_state_str[] = { "normal", "uninitialized", "degraded", "failed" * MIGR_REC_BUF_SIZE <= MIGR_REC_POSITION */ - #define UNIT_SRC_NORMAL 0 /* Source data for curr_migr_unit must * be recovered using srcMap */ #define UNIT_SRC_IN_CP_AREA 1 /* Source data for curr_migr_unit has @@ -291,7 +293,7 @@ struct md_list { struct md_list *next; }; -#define pr_vrb(fmt, arg...) (void) (verbose && fprintf(stderr, Name fmt, ##arg)) +#define pr_vrb(fmt, arg...) (void) (verbose && pr_err(fmt, ##arg)) static __u8 migr_type(struct imsm_dev *dev) { @@ -318,7 +320,7 @@ static void set_migr_type(struct imsm_dev *dev, __u8 migr_type) static unsigned int sector_count(__u32 bytes) { - return ((bytes + (512-1)) & (~(512-1))) / 512; + return ROUND_UP(bytes, 512) / 512; } static unsigned int mpb_sectors(struct imsm_super *mpb) @@ -353,12 +355,15 @@ struct intel_super { void *migr_rec_buf; /* buffer for I/O operations */ struct migr_record *migr_rec; /* migration record */ }; + int clean_migration_record_by_mdmon; /* when reshape is switched to next + array, it indicates that mdmon is allowed to clean migration + record */ size_t len; /* size of the 'buf' allocation */ void *next_buf; /* for realloc'ing buf from the manager */ size_t next_len; int updates_pending; /* count of pending updates for mdmon */ int current_vol; /* index of raid device undergoing creation */ - __u32 create_offset; /* common start for 'current_vol' */ + unsigned long long create_offset; /* common start for 'current_vol' */ __u32 random; /* random data for seeding new family numbers */ struct intel_dev *devlist; struct dl { @@ -398,6 +403,7 @@ struct extent { enum imsm_reshape_type { CH_TAKEOVER, CH_MIGRATION, + CH_ARRAY_SIZE, }; /* definition of messages passed to imsm_process_update */ @@ -411,6 +417,7 @@ enum imsm_update_type { update_reshape_migration, update_takeover, update_general_migration_checkpoint, + update_size_change, }; struct imsm_update_activate_spare { @@ -422,9 +429,9 @@ struct imsm_update_activate_spare { }; struct geo_params { - int dev_id; + char devnm[32]; char *dev_name; - long long size; + unsigned long long size; int level; int layout; int chunksize; @@ -463,6 +470,12 @@ struct imsm_update_reshape_migration { int new_disks[1]; /* new_raid_disks - old_raid_disks makedev number */ }; +struct imsm_update_size_change { + enum imsm_update_type type; + int subdev; + long long new_size; +}; + struct imsm_update_general_migration_checkpoint { enum imsm_update_type type; __u32 curr_migr_unit; @@ -493,11 +506,12 @@ struct imsm_update_add_remove_disk { enum imsm_update_type type; }; - static const char *_sys_dev_type[] = { [SYS_DEV_UNKNOWN] = "Unknown", [SYS_DEV_SAS] = "SAS", - [SYS_DEV_SATA] = "SATA" + [SYS_DEV_SATA] = "SATA", + [SYS_DEV_NVME] = "NVMe", + [SYS_DEV_VMD] = "VMD" }; const char *get_sys_dev_type(enum sys_dev_type type) @@ -510,20 +524,21 @@ const char *get_sys_dev_type(enum sys_dev_type type) static struct intel_hba * alloc_intel_hba(struct sys_dev *device) { - struct intel_hba *result = malloc(sizeof(*result)); - if (result) { - result->type = device->type; - result->path = strdup(device->path); - result->next = NULL; - if (result->path && (result->pci_id = strrchr(result->path, '/')) != NULL) - result->pci_id++; - } + struct intel_hba *result = xmalloc(sizeof(*result)); + + result->type = device->type; + result->path = xstrdup(device->path); + result->next = NULL; + if (result->path && (result->pci_id = strrchr(result->path, '/')) != NULL) + result->pci_id++; + return result; } static struct intel_hba * find_intel_hba(struct intel_hba *hba, struct sys_dev *device) { - struct intel_hba *result=NULL; + struct intel_hba *result; + for (result = hba; result; result = result->next) { if (result->type == device->type && strcmp(result->path, device->path) == 0) break; @@ -547,11 +562,21 @@ static int attach_hba_to_super(struct intel_super *super, struct sys_dev *device hba = super->hba; /* Intel metadata allows for all disks attached to the same type HBA. - * Do not sypport odf HBA types mixing + * Do not support HBA types mixing */ if (device->type != hba->type) return 2; + /* Always forbid spanning between VMD domains (seen as different controllers by mdadm) */ + if (device->type == SYS_DEV_VMD && !path_attached_to_hba(device->path, hba->path)) + return 2; + + /* Multiple same type HBAs can be used if they share the same OROM */ + const struct imsm_orom *device_orom = get_orom_by_device_id(device->dev_id); + + if (device_orom != super->orom) + return 2; + while (hba->next) hba = hba->next; @@ -561,7 +586,7 @@ static int attach_hba_to_super(struct intel_super *super, struct sys_dev *device static struct sys_dev* find_disk_attached_hba(int fd, const char *devname) { - struct sys_dev *list, *elem, *prev; + struct sys_dev *list, *elem; char *disk_path; if ((list = find_intel_devices()) == NULL) @@ -572,32 +597,19 @@ static struct sys_dev* find_disk_attached_hba(int fd, const char *devname) else disk_path = diskfd_to_devpath(fd); - if (!disk_path) { - free_sys_dev(&list); + if (!disk_path) return 0; - } - for (prev = NULL, elem = list; elem; prev = elem, elem = elem->next) { - if (path_attached_to_hba(disk_path, elem->path)) { - if (prev == NULL) - list = list->next; - else - prev->next = elem->next; - elem->next = NULL; - if (disk_path != devname) - free(disk_path); - free_sys_dev(&list); + for (elem = list; elem; elem = elem->next) + if (path_attached_to_hba(disk_path, elem->path)) return elem; - } - } + if (disk_path != devname) free(disk_path); - free_sys_dev(&list); return NULL; } - static int find_intel_hba_capability(int fd, struct intel_super *super, char *devname); @@ -610,11 +622,7 @@ static struct supertype *match_metadata_desc_imsm(char *arg) ) return NULL; - st = malloc(sizeof(*st)); - if (!st) - return NULL; - memset(st, 0, sizeof(*st)); - st->container_dev = NoMdDev; + st = xcalloc(1, sizeof(*st)); st->ss = &super_imsm; st->max_devs = IMSM_MAX_DEVICES; st->minor_version = 0; @@ -673,12 +681,12 @@ static __u32 __gen_imsm_checksum(struct imsm_super *mpb) __u32 *p = (__u32 *) mpb; __u32 sum = 0; - while (end--) { - sum += __le32_to_cpu(*p); + while (end--) { + sum += __le32_to_cpu(*p); p++; } - return sum - __le32_to_cpu(mpb->check_sum); + return sum - __le32_to_cpu(mpb->check_sum); } static size_t sizeof_imsm_map(struct imsm_map *map) @@ -867,6 +875,71 @@ static int count_memberships(struct dl *dl, struct intel_super *super) static __u32 imsm_min_reserved_sectors(struct intel_super *super); +static int split_ull(unsigned long long n, __u32 *lo, __u32 *hi) +{ + if (lo == 0 || hi == 0) + return 1; + *lo = __le32_to_cpu((unsigned)n); + *hi = __le32_to_cpu((unsigned)(n >> 32)); + return 0; +} + +static unsigned long long join_u32(__u32 lo, __u32 hi) +{ + return (unsigned long long)__le32_to_cpu(lo) | + (((unsigned long long)__le32_to_cpu(hi)) << 32); +} + +static unsigned long long total_blocks(struct imsm_disk *disk) +{ + if (disk == NULL) + return 0; + return join_u32(disk->total_blocks_lo, disk->total_blocks_hi); +} + +static unsigned long long pba_of_lba0(struct imsm_map *map) +{ + if (map == NULL) + return 0; + return join_u32(map->pba_of_lba0_lo, map->pba_of_lba0_hi); +} + +static unsigned long long blocks_per_member(struct imsm_map *map) +{ + if (map == NULL) + return 0; + return join_u32(map->blocks_per_member_lo, map->blocks_per_member_hi); +} + +#ifndef MDASSEMBLE +static unsigned long long num_data_stripes(struct imsm_map *map) +{ + if (map == NULL) + return 0; + return join_u32(map->num_data_stripes_lo, map->num_data_stripes_hi); +} + +static void set_total_blocks(struct imsm_disk *disk, unsigned long long n) +{ + split_ull(n, &disk->total_blocks_lo, &disk->total_blocks_hi); +} +#endif + +static void set_pba_of_lba0(struct imsm_map *map, unsigned long long n) +{ + split_ull(n, &map->pba_of_lba0_lo, &map->pba_of_lba0_hi); +} + +static void set_blocks_per_member(struct imsm_map *map, unsigned long long n) +{ + split_ull(n, &map->blocks_per_member_lo, &map->blocks_per_member_hi); +} + +static void set_num_data_stripes(struct imsm_map *map, unsigned long long n) +{ + split_ull(n, &map->num_data_stripes_lo, &map->num_data_stripes_hi); +} + static struct extent *get_extents(struct intel_super *super, struct dl *dl) { /* find a list of used extents on the given physical device */ @@ -884,9 +957,7 @@ static struct extent *get_extents(struct intel_super *super, struct dl *dl) else reservation = MPB_SECTOR_CNT + IMSM_RESERVED_SECTORS; - rv = malloc(sizeof(struct extent) * (memberships + 1)); - if (!rv) - return NULL; + rv = xcalloc(sizeof(struct extent), (memberships + 1)); e = rv; for (i = 0; i < super->anchor->num_raid_devs; i++) { @@ -894,24 +965,23 @@ static struct extent *get_extents(struct intel_super *super, struct dl *dl) struct imsm_map *map = get_imsm_map(dev, MAP_0); if (get_imsm_disk_slot(map, dl->index) >= 0) { - e->start = __le32_to_cpu(map->pba_of_lba0); - e->size = __le32_to_cpu(map->blocks_per_member); + e->start = pba_of_lba0(map); + e->size = blocks_per_member(map); e++; } } qsort(rv, memberships, sizeof(*rv), cmp_extent); - /* determine the start of the metadata + /* determine the start of the metadata * when no raid devices are defined use the default * ...otherwise allow the metadata to truncate the value * as is the case with older versions of imsm */ if (memberships) { struct extent *last = &rv[memberships - 1]; - __u32 remainder; + unsigned long long remainder; - remainder = __le32_to_cpu(dl->disk.total_blocks) - - (last->start + last->size); + remainder = total_blocks(&dl->disk) - (last->start + last->size); /* round down to 1k block to satisfy precision of the kernel * 'size' interface */ @@ -922,7 +992,7 @@ static struct extent *get_extents(struct intel_super *super, struct dl *dl) if (reservation > remainder) reservation = remainder; } - e->start = __le32_to_cpu(dl->disk.total_blocks) - reservation; + e->start = total_blocks(&dl->disk) - reservation; e->size = 0; return rv; } @@ -951,7 +1021,7 @@ static __u32 imsm_reserved_sectors(struct intel_super *super, struct dl *dl) for (i = 0; e[i].size; i++) continue; - rv = __le32_to_cpu(dl->disk.total_blocks) - e[i].start; + rv = total_blocks(&dl->disk) - e[i].start; free(e); @@ -981,7 +1051,8 @@ static __u32 imsm_min_reserved_sectors(struct intel_super *super) { struct extent *e; int i; - __u32 min_active, remainder; + unsigned long long min_active; + __u32 remainder; __u32 rv = MPB_SECTOR_CNT + IMSM_RESERVED_SECTORS; struct dl *dl, *dl_min = NULL; @@ -992,9 +1063,10 @@ static __u32 imsm_min_reserved_sectors(struct intel_super *super) for (dl = super->disks; dl; dl = dl->next) { if (dl->index < 0) continue; - if (dl->disk.total_blocks < min_active || min_active == 0) { + unsigned long long blocks = total_blocks(&dl->disk); + if (blocks < min_active || min_active == 0) { dl_min = dl; - min_active = dl->disk.total_blocks; + min_active = blocks; } } if (!dl_min) @@ -1112,13 +1184,13 @@ static void print_imsm_dev(struct intel_super *super, sz += __le32_to_cpu(dev->size_low); printf(" Array Size : %llu%s\n", (unsigned long long)sz, human_size(sz * 512)); - sz = __le32_to_cpu(map->blocks_per_member); + sz = blocks_per_member(map); printf(" Per Dev Size : %llu%s\n", (unsigned long long)sz, human_size(sz * 512)); - printf(" Sector Offset : %u\n", - __le32_to_cpu(map->pba_of_lba0)); - printf(" Num Stripes : %u\n", - __le32_to_cpu(map->num_data_stripes)); + printf(" Sector Offset : %llu\n", + pba_of_lba0(map)); + printf(" Num Stripes : %llu\n", + num_data_stripes(map)); printf(" Chunk Size : %u KiB", __le16_to_cpu(map->blocks_per_strip) / 2); if (map2) @@ -1179,7 +1251,7 @@ static void print_imsm_disk(struct imsm_disk *disk, int index, __u32 reserved) is_configured(disk) ? " active" : "", is_failed(disk) ? " failed" : ""); printf(" Id : %08x\n", __le32_to_cpu(disk->scsi_id)); - sz = __le32_to_cpu(disk->total_blocks) - reserved; + sz = total_blocks(disk) - reserved; printf(" Usable Size : %llu%s\n", (unsigned long long)sz, human_size(sz * 512)); } @@ -1244,12 +1316,12 @@ void examine_migr_rec_imsm(struct intel_super *super) /******************************************************************************* * function: imsm_check_attributes * Description: Function checks if features represented by attributes flags - * are supported by mdadm. + * are supported by mdadm. * Parameters: * attributes - Attributes read from metadata * Returns: - * 0 - passed attributes contains unsupported features flags - * 1 - all features are supported + * 0 - passed attributes contains unsupported features flags + * 1 - all features are supported ******************************************************************************/ static int imsm_check_attributes(__u32 attributes) { @@ -1260,7 +1332,7 @@ static int imsm_check_attributes(__u32 attributes) not_supported &= attributes; if (not_supported) { - fprintf(stderr, Name "(IMSM): Unsupported attributes : %x\n", + pr_err("(IMSM): Unsupported attributes : %x\n", (unsigned)__le32_to_cpu(not_supported)); if (not_supported & MPB_ATTRIB_CHECKSUM_VERIFY) { dprintf("\t\tMPB_ATTRIB_CHECKSUM_VERIFY \n"); @@ -1320,7 +1392,7 @@ static int imsm_check_attributes(__u32 attributes) } if (not_supported) - dprintf(Name "(IMSM): Unknown attributes : %x\n", not_supported); + dprintf("(IMSM): Unknown attributes : %x\n", not_supported); ret_val = 0; } @@ -1456,6 +1528,59 @@ static void export_examine_super_imsm(struct supertype *st) printf("MD_DEVICES=%u\n", mpb->num_disks); } +static int copy_metadata_imsm(struct supertype *st, int from, int to) +{ + /* The second last 512byte sector of the device contains + * the "struct imsm_super" metadata. + * This contains mpb_size which is the size in bytes of the + * extended metadata. This is located immediately before + * the imsm_super. + * We want to read all that, plus the last sector which + * may contain a migration record, and write it all + * to the target. + */ + void *buf; + unsigned long long dsize, offset; + int sectors; + struct imsm_super *sb; + int written = 0; + + if (posix_memalign(&buf, 4096, 4096) != 0) + return 1; + + if (!get_dev_size(from, NULL, &dsize)) + goto err; + + if (lseek64(from, dsize-1024, 0) < 0) + goto err; + if (read(from, buf, 512) != 512) + goto err; + sb = buf; + if (strncmp((char*)sb->sig, MPB_SIGNATURE, MPB_SIG_LEN) != 0) + goto err; + + sectors = mpb_sectors(sb) + 2; + offset = dsize - sectors * 512; + if (lseek64(from, offset, 0) < 0 || + lseek64(to, offset, 0) < 0) + goto err; + while (written < sectors * 512) { + int n = sectors*512 - written; + if (n > 4096) + n = 4096; + if (read(from, buf, n) != n) + goto err; + if (write(to, buf, n) != n) + goto err; + written += n; + } + free(buf); + return 0; +err: + free(buf); + return 1; +} + static void detail_super_imsm(struct supertype *st, char *homehost) { struct mdinfo info; @@ -1491,8 +1616,8 @@ static int ahci_enumerate_ports(const char *hba_path, int port_count, int host_b unsigned long port_mask = (1 << port_count) - 1; if (port_count > (int)sizeof(port_mask) * 8) { - if (verbose) - fprintf(stderr, Name ": port_count %d out of range\n", port_count); + if (verbose > 0) + pr_err("port_count %d out of range\n", port_count); return 2; } @@ -1500,7 +1625,10 @@ static int ahci_enumerate_ports(const char *hba_path, int port_count, int host_b * this hba */ dir = opendir("/sys/dev/block"); - for (ent = dir ? readdir(dir) : NULL; ent; ent = readdir(dir)) { + if (!dir) + return 1; + + for (ent = readdir(dir); ent; ent = readdir(dir)) { int fd; char model[64]; char vendor[64]; @@ -1524,15 +1652,15 @@ static int ahci_enumerate_ports(const char *hba_path, int port_count, int host_b /* retrieve the scsi device type */ if (asprintf(&device, "/sys/dev/block/%d:%d/device/xxxxxxx", major, minor) < 0) { - if (verbose) - fprintf(stderr, Name ": failed to allocate 'device'\n"); + if (verbose > 0) + pr_err("failed to allocate 'device'\n"); err = 2; break; } sprintf(device, "/sys/dev/block/%d:%d/device/type", major, minor); - if (load_sys(device, buf) != 0) { - if (verbose) - fprintf(stderr, Name ": failed to read device type for %s\n", + if (load_sys(device, buf, sizeof(buf)) != 0) { + if (verbose > 0) + pr_err("failed to read device type for %s\n", path); err = 2; free(device); @@ -1545,7 +1673,7 @@ static int ahci_enumerate_ports(const char *hba_path, int port_count, int host_b vendor[0] = '\0'; model[0] = '\0'; sprintf(device, "/sys/dev/block/%d:%d/device/vendor", major, minor); - if (load_sys(device, buf) == 0) { + if (load_sys(device, buf, sizeof(buf)) == 0) { strncpy(vendor, buf, sizeof(vendor)); vendor[sizeof(vendor) - 1] = '\0'; c = (char *) &vendor[sizeof(vendor) - 1]; @@ -1554,7 +1682,7 @@ static int ahci_enumerate_ports(const char *hba_path, int port_count, int host_b } sprintf(device, "/sys/dev/block/%d:%d/device/model", major, minor); - if (load_sys(device, buf) == 0) { + if (load_sys(device, buf, sizeof(buf)) == 0) { strncpy(model, buf, sizeof(model)); model[sizeof(model) - 1] = '\0'; c = (char *) &model[sizeof(model) - 1]; @@ -1584,18 +1712,19 @@ static int ahci_enumerate_ports(const char *hba_path, int port_count, int host_b /* chop device path to 'host%d' and calculate the port number */ c = strchr(&path[hba_len], '/'); if (!c) { - if (verbose) - fprintf(stderr, Name ": %s - invalid path name\n", path + hba_len); + if (verbose > 0) + pr_err("%s - invalid path name\n", path + hba_len); err = 2; break; } *c = '\0'; - if (sscanf(&path[hba_len], "host%d", &port) == 1) + if ((sscanf(&path[hba_len], "ata%d", &port) == 1) || + ((sscanf(&path[hba_len], "host%d", &port) == 1))) port -= host_base; else { - if (verbose) { + if (verbose > 0) { *c = '/'; /* repair the full string */ - fprintf(stderr, Name ": failed to determine port number for %s\n", + pr_err("failed to determine port number for %s\n", path); } err = 2; @@ -1641,15 +1770,77 @@ static int ahci_enumerate_ports(const char *hba_path, int port_count, int host_b return err; } +static int print_vmd_attached_devs(struct sys_dev *hba) +{ + struct dirent *ent; + DIR *dir; + char path[292]; + char link[256]; + char *c, *rp; + + if (hba->type != SYS_DEV_VMD) + return 1; + + /* scroll through /sys/dev/block looking for devices attached to + * this hba + */ + dir = opendir("/sys/bus/pci/drivers/nvme"); + if (!dir) + return 1; + + for (ent = readdir(dir); ent; ent = readdir(dir)) { + int n; + + /* is 'ent' a device? check that the 'subsystem' link exists and + * that its target matches 'bus' + */ + sprintf(path, "/sys/bus/pci/drivers/nvme/%s/subsystem", + ent->d_name); + n = readlink(path, link, sizeof(link)); + if (n < 0 || n >= (int)sizeof(link)) + continue; + link[n] = '\0'; + c = strrchr(link, '/'); + if (!c) + continue; + if (strncmp("pci", c+1, strlen("pci")) != 0) + continue; + + sprintf(path, "/sys/bus/pci/drivers/nvme/%s", ent->d_name); + /* if not a intel NVMe - skip it*/ + if (devpath_to_vendor(path) != 0x8086) + continue; + + rp = realpath(path, NULL); + if (!rp) + continue; + + if (path_attached_to_hba(rp, hba->path)) { + printf(" NVMe under VMD : %s\n", rp); + } + free(rp); + } + + closedir(dir); + return 0; +} + static void print_found_intel_controllers(struct sys_dev *elem) { for (; elem; elem = elem->next) { - fprintf(stderr, Name ": found Intel(R) "); + pr_err("found Intel(R) "); if (elem->type == SYS_DEV_SATA) fprintf(stderr, "SATA "); else if (elem->type == SYS_DEV_SAS) fprintf(stderr, "SAS "); - fprintf(stderr, "RAID controller"); + else if (elem->type == SYS_DEV_NVME) + fprintf(stderr, "NVMe "); + + if (elem->type == SYS_DEV_VMD) + fprintf(stderr, "VMD domain"); + else + fprintf(stderr, "RAID controller"); + if (elem->pci_id) fprintf(stderr, " at %s", elem->pci_id); fprintf(stderr, ".\n"); @@ -1670,7 +1861,8 @@ static int ahci_get_port_count(const char *hba_path, int *port_count) for (ent = readdir(dir); ent; ent = readdir(dir)) { int host; - if (sscanf(ent->d_name, "host%d", &host) != 1) + if ((sscanf(ent->d_name, "ata%d", &host) != 1) && + ((sscanf(ent->d_name, "host%d", &host) != 1))) continue; if (*port_count == 0) host_base = host; @@ -1686,9 +1878,15 @@ static int ahci_get_port_count(const char *hba_path, int *port_count) static void print_imsm_capability(const struct imsm_orom *orom) { - printf(" Platform : Intel(R) Matrix Storage Manager\n"); - printf(" Version : %d.%d.%d.%d\n", orom->major_ver, orom->minor_ver, - orom->hotfix_ver, orom->build); + printf(" Platform : Intel(R) "); + if (orom->capabilities == 0 && orom->driver_features == 0) + printf("Matrix Storage Manager\n"); + else + printf("Rapid Storage Technology%s\n", + imsm_orom_is_enterprise(orom) ? " enterprise" : ""); + if (orom->major_ver || orom->minor_ver || orom->hotfix_ver || orom->build) + printf(" Version : %d.%d.%d.%d\n", orom->major_ver, + orom->minor_ver, orom->hotfix_ver, orom->build); printf(" RAID Levels :%s%s%s%s%s\n", imsm_orom_has_raid0(orom) ? " raid0" : "", imsm_orom_has_raid1(orom) ? " raid1" : "", @@ -1712,12 +1910,54 @@ static void print_imsm_capability(const struct imsm_orom *orom) imsm_orom_has_chunk(orom, 1024*16) ? " 16M" : "", imsm_orom_has_chunk(orom, 1024*32) ? " 32M" : "", imsm_orom_has_chunk(orom, 1024*64) ? " 64M" : ""); + printf(" 2TB volumes :%s supported\n", + (orom->attr & IMSM_OROM_ATTR_2TB)?"":" not"); + printf(" 2TB disks :%s supported\n", + (orom->attr & IMSM_OROM_ATTR_2TB_DISK)?"":" not"); printf(" Max Disks : %d\n", orom->tds); - printf(" Max Volumes : %d\n", orom->vpa); + printf(" Max Volumes : %d per array, %d per %s\n", + orom->vpa, orom->vphba, + imsm_orom_is_nvme(orom) ? "platform" : "controller"); return; } -static int detail_platform_imsm(int verbose, int enumerate_only) +static void print_imsm_capability_export(const struct imsm_orom *orom) +{ + printf("MD_FIRMWARE_TYPE=imsm\n"); + if (orom->major_ver || orom->minor_ver || orom->hotfix_ver || orom->build) + printf("IMSM_VERSION=%d.%d.%d.%d\n", orom->major_ver, orom->minor_ver, + orom->hotfix_ver, orom->build); + printf("IMSM_SUPPORTED_RAID_LEVELS=%s%s%s%s%s\n", + imsm_orom_has_raid0(orom) ? "raid0 " : "", + imsm_orom_has_raid1(orom) ? "raid1 " : "", + imsm_orom_has_raid1e(orom) ? "raid1e " : "", + imsm_orom_has_raid5(orom) ? "raid10 " : "", + imsm_orom_has_raid10(orom) ? "raid5 " : ""); + printf("IMSM_SUPPORTED_CHUNK_SIZES=%s%s%s%s%s%s%s%s%s%s%s%s%s%s%s%s\n", + imsm_orom_has_chunk(orom, 2) ? "2k " : "", + imsm_orom_has_chunk(orom, 4) ? "4k " : "", + imsm_orom_has_chunk(orom, 8) ? "8k " : "", + imsm_orom_has_chunk(orom, 16) ? "16k " : "", + imsm_orom_has_chunk(orom, 32) ? "32k " : "", + imsm_orom_has_chunk(orom, 64) ? "64k " : "", + imsm_orom_has_chunk(orom, 128) ? "128k " : "", + imsm_orom_has_chunk(orom, 256) ? "256k " : "", + imsm_orom_has_chunk(orom, 512) ? "512k " : "", + imsm_orom_has_chunk(orom, 1024*1) ? "1M " : "", + imsm_orom_has_chunk(orom, 1024*2) ? "2M " : "", + imsm_orom_has_chunk(orom, 1024*4) ? "4M " : "", + imsm_orom_has_chunk(orom, 1024*8) ? "8M " : "", + imsm_orom_has_chunk(orom, 1024*16) ? "16M " : "", + imsm_orom_has_chunk(orom, 1024*32) ? "32M " : "", + imsm_orom_has_chunk(orom, 1024*64) ? "64M " : ""); + printf("IMSM_2TB_VOLUMES=%s\n",(orom->attr & IMSM_OROM_ATTR_2TB) ? "yes" : "no"); + printf("IMSM_2TB_DISKS=%s\n",(orom->attr & IMSM_OROM_ATTR_2TB_DISK) ? "yes" : "no"); + printf("IMSM_MAX_DISKS=%d\n",orom->tds); + printf("IMSM_MAX_VOLUMES_PER_ARRAY=%d\n",orom->vpa); + printf("IMSM_MAX_VOLUMES_PER_CONTROLLER=%d\n",orom->vphba); +} + +static int detail_platform_imsm(int verbose, int enumerate_only, char *controller_path) { /* There are two components to imsm platform support, the ahci SATA * controller and the option-rom. To find the SATA controller we @@ -1730,11 +1970,10 @@ static int detail_platform_imsm(int verbose, int enumerate_only) * platform capabilities. If raid support is disabled in the BIOS the * option-rom capability structure will not be available. */ - const struct imsm_orom *orom; struct sys_dev *list, *hba; int host_base = 0; int port_count = 0; - int result=0; + int result=1; if (enumerate_only) { if (check_env("IMSM_NO_PLATFORM")) @@ -1743,53 +1982,136 @@ static int detail_platform_imsm(int verbose, int enumerate_only) if (!list) return 2; for (hba = list; hba; hba = hba->next) { - orom = find_imsm_capability(hba->type); - if (!orom) { - result = 2; + if (find_imsm_capability(hba)) { + result = 0; break; } + else + result = 2; } - free_sys_dev(&list); return result; } list = find_intel_devices(); if (!list) { - if (verbose) - fprintf(stderr, Name ": no active Intel(R) RAID " - "controller found.\n"); - free_sys_dev(&list); + if (verbose > 0) + pr_err("no active Intel(R) RAID controller found.\n"); return 2; - } else if (verbose) + } else if (verbose > 0) print_found_intel_controllers(list); for (hba = list; hba; hba = hba->next) { - orom = find_imsm_capability(hba->type); - if (!orom) - fprintf(stderr, Name ": imsm capabilities not found for controller: %s (type %s)\n", + if (controller_path && (compare_paths(hba->path, controller_path) != 0)) + continue; + if (!find_imsm_capability(hba)) { + char buf[PATH_MAX]; + pr_err("imsm capabilities not found for controller: %s (type %s)\n", + hba->type == SYS_DEV_VMD ? vmd_domain_to_controller(hba, buf) : hba->path, + get_sys_dev_type(hba->type)); + continue; + } + result = 0; + } + + if (controller_path && result == 1) { + pr_err("no active Intel(R) RAID controller found under %s\n", + controller_path); + return result; + } + + const struct orom_entry *entry; + + for (entry = orom_entries; entry; entry = entry->next) { + if (entry->type == SYS_DEV_VMD) { + for (hba = list; hba; hba = hba->next) { + if (hba->type == SYS_DEV_VMD) { + char buf[PATH_MAX]; + print_imsm_capability(&entry->orom); + printf(" I/O Controller : %s (%s)\n", + vmd_domain_to_controller(hba, buf), get_sys_dev_type(hba->type)); + if (print_vmd_attached_devs(hba)) { + if (verbose > 0) + pr_err("failed to get devices attached to VMD domain.\n"); + result |= 2; + } + printf("\n"); + } + } + continue; + } + + print_imsm_capability(&entry->orom); + if (entry->type == SYS_DEV_NVME) { + for (hba = list; hba; hba = hba->next) { + if (hba->type == SYS_DEV_NVME) + printf(" NVMe Device : %s\n", hba->path); + } + printf("\n"); + continue; + } + + struct devid_list *devid; + for (devid = entry->devid_list; devid; devid = devid->next) { + hba = device_by_id(devid->devid); + if (!hba) + continue; + + printf(" I/O Controller : %s (%s)\n", hba->path, get_sys_dev_type(hba->type)); - else - print_imsm_capability(orom); + if (hba->type == SYS_DEV_SATA) { + host_base = ahci_get_port_count(hba->path, &port_count); + if (ahci_enumerate_ports(hba->path, port_count, host_base, verbose)) { + if (verbose > 0) + pr_err("failed to enumerate ports on SATA controller at %s.\n", hba->pci_id); + result |= 2; + } + } + } + printf("\n"); + } + + return result; +} + +static int export_detail_platform_imsm(int verbose, char *controller_path) +{ + struct sys_dev *list, *hba; + int result=1; + + list = find_intel_devices(); + if (!list) { + if (verbose > 0) + pr_err("IMSM_DETAIL_PLATFORM_ERROR=NO_INTEL_DEVICES\n"); + result = 2; + return result; } for (hba = list; hba; hba = hba->next) { - printf(" I/O Controller : %s (%s)\n", - hba->path, get_sys_dev_type(hba->type)); - - if (hba->type == SYS_DEV_SATA) { - host_base = ahci_get_port_count(hba->path, &port_count); - if (ahci_enumerate_ports(hba->path, port_count, host_base, verbose)) { - if (verbose) - fprintf(stderr, Name ": failed to enumerate " - "ports on SATA controller at %s.", hba->pci_id); - result |= 2; - } + if (controller_path && (compare_paths(hba->path,controller_path) != 0)) + continue; + if (!find_imsm_capability(hba) && verbose > 0) { + char buf[PATH_MAX]; + pr_err("IMSM_DETAIL_PLATFORM_ERROR=NO_IMSM_CAPABLE_DEVICE_UNDER_%s\n", + hba->type == SYS_DEV_VMD ? vmd_domain_to_controller(hba, buf) : hba->path); + } + else + result = 0; + } + + const struct orom_entry *entry; + + for (entry = orom_entries; entry; entry = entry->next) { + if (entry->type == SYS_DEV_VMD) { + for (hba = list; hba; hba = hba->next) + print_imsm_capability_export(&entry->orom); + continue; } + print_imsm_capability_export(&entry->orom); } - free_sys_dev(&list); return result; } + #endif static int match_home_imsm(struct supertype *st, char *homehost) @@ -1816,7 +2138,7 @@ static void uuid_from_super_imsm(struct supertype *st, int uuid[4]) * 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. @@ -1956,13 +2278,15 @@ static __u8 imsm_num_data_members(struct imsm_dev *dev, int second_map) switch (get_imsm_raid_level(map)) { case 0: + return map->num_members; + break; case 1: case 10: - return map->num_members; + return map->num_members/2; case 5: return map->num_members - 1; default: - dprintf("%s: unsupported raid level\n", __func__); + dprintf("unsupported raid level\n"); return 0; } } @@ -2095,16 +2419,14 @@ static int read_imsm_migr_rec(int fd, struct intel_super *super) get_dev_size(fd, NULL, &dsize); if (lseek64(fd, dsize - MIGR_REC_POSITION, SEEK_SET) < 0) { - fprintf(stderr, - Name ": Cannot seek to anchor block: %s\n", - strerror(errno)); + pr_err("Cannot seek to anchor block: %s\n", + strerror(errno)); goto out; } if (read(fd, super->migr_rec_buf, MIGR_REC_BUF_SIZE) != MIGR_REC_BUF_SIZE) { - fprintf(stderr, - Name ": Cannot read migr record block: %s\n", - strerror(errno)); + pr_err("Cannot read migr record block: %s\n", + strerror(errno)); goto out; } ret_val = 0; @@ -2141,12 +2463,12 @@ static struct imsm_dev *imsm_get_device_during_migration( static int load_imsm_migr_rec(struct intel_super *super, struct mdinfo *info) { struct mdinfo *sd; - struct dl *dl = NULL; + struct dl *dl; char nm[30]; int retval = -1; int fd = -1; struct imsm_dev *dev; - struct imsm_map *map = NULL; + struct imsm_map *map; int slot = -1; /* find map under migration */ @@ -2221,7 +2543,7 @@ static int imsm_create_metadata_checkpoint_update( int update_memory_size = 0; - dprintf("imsm_create_metadata_checkpoint_update(enter)\n"); + dprintf("(enter)\n"); if (u == NULL) return 0; @@ -2231,21 +2553,18 @@ static int imsm_create_metadata_checkpoint_update( update_memory_size = sizeof(struct imsm_update_general_migration_checkpoint); - *u = calloc(1, update_memory_size); + *u = xcalloc(1, update_memory_size); if (*u == NULL) { - dprintf("error: cannot get memory for " - "imsm_create_metadata_checkpoint_update update\n"); + dprintf("error: cannot get memory\n"); return 0; } (*u)->type = update_general_migration_checkpoint; (*u)->curr_migr_unit = __le32_to_cpu(super->migr_rec->curr_migr_unit); - dprintf("imsm_create_metadata_checkpoint_update: prepared for %u\n", - (*u)->curr_migr_unit); + dprintf("prepared for %u\n", (*u)->curr_migr_unit); return update_memory_size; } - static void imsm_update_metadata_locally(struct supertype *st, void *buf, int len); @@ -2270,7 +2589,7 @@ static int write_imsm_migr_rec(struct supertype *st) int len; struct imsm_update_general_migration_checkpoint *u; struct imsm_dev *dev; - struct imsm_map *map = NULL; + struct imsm_map *map; /* find map under migration */ dev = imsm_get_device_during_migration(super); @@ -2301,16 +2620,14 @@ static int write_imsm_migr_rec(struct supertype *st) continue; get_dev_size(fd, NULL, &dsize); if (lseek64(fd, dsize - MIGR_REC_POSITION, SEEK_SET) < 0) { - fprintf(stderr, - Name ": Cannot seek to anchor block: %s\n", - strerror(errno)); + pr_err("Cannot seek to anchor block: %s\n", + strerror(errno)); goto out; } if (write(fd, super->migr_rec_buf, MIGR_REC_BUF_SIZE) != MIGR_REC_BUF_SIZE) { - fprintf(stderr, - Name ": Cannot write migr record block: %s\n", - strerror(errno)); + pr_err("Cannot write migr record block: %s\n", + strerror(errno)); goto out; } close(fd); @@ -2368,6 +2685,30 @@ int imsm_reshape_blocks_arrays_changes(struct intel_super *super) } return rv; } +static unsigned long long imsm_component_size_aligment_check(int level, + int chunk_size, + unsigned long long component_size) +{ + unsigned int component_size_alligment; + + /* check component size aligment + */ + component_size_alligment = component_size % (chunk_size/512); + + dprintf("(Level: %i, chunk_size = %i, component_size = %llu), component_size_alligment = %u\n", + level, chunk_size, component_size, + component_size_alligment); + + if (component_size_alligment && (level != 1) && (level != UnSet)) { + dprintf("imsm: reported component size alligned from %llu ", + component_size); + component_size -= component_size_alligment; + dprintf_cont("to %llu (%i).\n", + component_size, component_size_alligment); + } + + return component_size; +} static void getinfo_super_imsm_volume(struct supertype *st, struct mdinfo *info, char *dmap) { @@ -2378,8 +2719,6 @@ static void getinfo_super_imsm_volume(struct supertype *st, struct mdinfo *info, struct imsm_map *prev_map = get_imsm_map(dev, MAP_1); struct imsm_map *map_to_analyse = map; struct dl *dl; - char *devname; - unsigned int component_size_alligment; int map_disks = info->array.raid_disks; memset(info, 0, sizeof(*info)); @@ -2459,23 +2798,13 @@ static void getinfo_super_imsm_volume(struct supertype *st, struct mdinfo *info, dl->index); } - info->data_offset = __le32_to_cpu(map_to_analyse->pba_of_lba0); - info->component_size = - __le32_to_cpu(map_to_analyse->blocks_per_member); + info->data_offset = pba_of_lba0(map_to_analyse); + info->component_size = blocks_per_member(map_to_analyse); - /* check component size aligment - */ - component_size_alligment = - info->component_size % (info->array.chunk_size/512); - - if (component_size_alligment && - (info->array.level != 1) && (info->array.level != UnSet)) { - dprintf("imsm: reported component size alligned from %llu ", - info->component_size); - info->component_size -= component_size_alligment; - dprintf("to %llu (%i).\n", - info->component_size, component_size_alligment); - } + info->component_size = imsm_component_size_aligment_check( + info->array.level, + info->array.chunk_size, + info->component_size); memset(info->uuid, 0, sizeof(info->uuid)); info->recovery_start = MaxSector; @@ -2514,15 +2843,14 @@ static void getinfo_super_imsm_volume(struct supertype *st, struct mdinfo *info, info->reshape_progress = blocks_per_unit * units; - dprintf("IMSM: General Migration checkpoint : %llu " - "(%llu) -> read reshape progress : %llu\n", + dprintf("IMSM: General Migration checkpoint : %llu (%llu) -> read reshape progress : %llu\n", (unsigned long long)units, (unsigned long long)blocks_per_unit, info->reshape_progress); used_disks = imsm_num_data_members(dev, MAP_1); if (used_disks > 0) { - array_blocks = map->blocks_per_member * + array_blocks = blocks_per_member(map) * used_disks; /* round array size down to closest MB */ @@ -2551,11 +2879,7 @@ static void getinfo_super_imsm_volume(struct supertype *st, struct mdinfo *info, info->array.major_version = -1; info->array.minor_version = -2; - devname = devnum2devname(st->container_dev); - *info->text_version = '\0'; - if (devname) - sprintf(info->text_version, "/%s/%d", devname, info->container_member); - free(devname); + sprintf(info->text_version, "/%s/%d", st->container_devnm, info->container_member); info->safe_mode_delay = 4000; /* 4 secs like the Matrix driver */ uuid_from_super_imsm(st, info->uuid); @@ -2580,7 +2904,6 @@ static __u8 imsm_check_degraded(struct intel_super *super, struct imsm_dev *dev, static int imsm_count_failed(struct intel_super *super, struct imsm_dev *dev, int look_in_map); - #ifndef MDASSEMBLE static void manage_second_map(struct intel_super *super, struct imsm_dev *dev) { @@ -2631,7 +2954,7 @@ static void getinfo_super_imsm(struct supertype *st, struct mdinfo *info, char * info->array.level = LEVEL_CONTAINER; info->array.layout = 0; info->array.md_minor = -1; - info->array.ctime = 0; /* N/A for imsm */ + info->array.ctime = 0; /* N/A for imsm */ info->array.utime = 0; info->array.chunk_size = 0; @@ -2699,14 +3022,14 @@ static void getinfo_super_imsm(struct supertype *st, struct mdinfo *info, char * */ max_enough = max(max_enough, enough); } - dprintf("%s: enough: %d\n", __func__, max_enough); + dprintf("enough: %d\n", max_enough); info->container_enough = max_enough; if (super->disks) { __u32 reserved = imsm_reserved_sectors(super, super->disks); disk = &super->disks->disk; - info->data_offset = __le32_to_cpu(disk->total_blocks) - reserved; + info->data_offset = total_blocks(&super->disks->disk) - reserved; info->component_size = reserved; info->disk.state = is_configured(disk) ? (1 << MD_DISK_ACTIVE) : 0; /* we don't change info->disk.raid_disk here because @@ -2738,7 +3061,7 @@ static void getinfo_super_imsm(struct supertype *st, struct mdinfo *info, char * * for each disk in array */ struct mdinfo *getinfo_super_disks_imsm(struct supertype *st) { - struct mdinfo *mddev = NULL; + struct mdinfo *mddev; struct intel_super *super = st->sb; struct imsm_disk *disk; int count = 0; @@ -2746,23 +3069,11 @@ struct mdinfo *getinfo_super_disks_imsm(struct supertype *st) if (!super || !super->disks) return NULL; dl = super->disks; - mddev = malloc(sizeof(*mddev)); - if (!mddev) { - fprintf(stderr, Name ": Failed to allocate memory.\n"); - return NULL; - } - memset(mddev, 0, sizeof(*mddev)); + mddev = xcalloc(1, sizeof(*mddev)); while (dl) { struct mdinfo *tmp; disk = &dl->disk; - tmp = malloc(sizeof(*tmp)); - if (!tmp) { - fprintf(stderr, Name ": Failed to allocate memory.\n"); - if (mddev) - sysfs_free(mddev); - return NULL; - } - memset(tmp, 0, sizeof(*tmp)); + tmp = xcalloc(1, sizeof(*tmp)); if (mddev->devs) tmp->next = mddev->devs; mddev->devs = tmp; @@ -2868,7 +3179,8 @@ static size_t disks_to_mpb_size(int disks) return size; } -static __u64 avail_size_imsm(struct supertype *st, __u64 devsize) +static __u64 avail_size_imsm(struct supertype *st, __u64 devsize, + unsigned long long data_offset) { if (devsize < (MPB_SECTOR_CNT + IMSM_RESERVED_SECTORS)) return 0; @@ -2905,21 +3217,27 @@ static int compare_super_imsm(struct supertype *st, struct supertype *tst) struct intel_super *first = st->sb; struct intel_super *sec = tst->sb; - if (!first) { - st->sb = tst->sb; - tst->sb = NULL; - return 0; - } + if (!first) { + st->sb = tst->sb; + tst->sb = NULL; + return 0; + } /* in platform dependent environment test if the disks * use the same Intel hba + * If not on Intel hba at all, allow anything. */ - if (!check_env("IMSM_NO_PLATFORM")) { - if (!first->hba || !sec->hba || - (first->hba->type != sec->hba->type)) { + if (!check_env("IMSM_NO_PLATFORM") && first->hba && sec->hba) { + if (first->hba->type != sec->hba->type) { + fprintf(stderr, + "HBAs of devices do not match %s != %s\n", + get_sys_dev_type(first->hba->type), + get_sys_dev_type(sec->hba->type)); + return 3; + } + if (first->orom != sec->orom) { fprintf(stderr, - "HBAs of devices does not match %s != %s\n", - first->hba ? get_sys_dev_type(first->hba->type) : NULL, - sec->hba ? get_sys_dev_type(sec->hba->type) : NULL); + "HBAs of devices do not match %s != %s\n", + first->hba->pci_id, sec->hba->pci_id); return 3; } } @@ -2950,7 +3268,6 @@ static int compare_super_imsm(struct supertype *st, struct supertype *tst) } - /* if 'first' is a spare promote it to a populated mpb with sec's * family number */ @@ -2964,14 +3281,8 @@ static int compare_super_imsm(struct supertype *st, struct supertype *tst) * fails here we don't associate the spare */ for (i = 0; i < sec->anchor->num_raid_devs; i++) { - dv = malloc(sizeof(*dv)); - if (!dv) - break; - dev = malloc(sizeof_imsm_dev(get_imsm_dev(sec, i), 1)); - if (!dev) { - free(dv); - break; - } + dv = xmalloc(sizeof(*dv)); + dev = xmalloc(sizeof_imsm_dev(get_imsm_dev(sec, i), 1)); dv->dev = dev; dv->index = i; dv->next = first->devlist; @@ -2980,7 +3291,7 @@ static int compare_super_imsm(struct supertype *st, struct supertype *tst) if (i < sec->anchor->num_raid_devs) { /* allocation failure */ free_devlist(first); - fprintf(stderr, "imsm: failed to associate spare\n"); + pr_err("imsm: failed to associate spare\n"); return 3; } first->anchor->num_raid_devs = sec->anchor->num_raid_devs; @@ -3046,18 +3357,16 @@ static int imsm_read_serial(int fd, char *devname, if (rv != 0) { if (devname) - fprintf(stderr, - Name ": Failed to retrieve serial for %s\n", - devname); + pr_err("Failed to retrieve serial for %s\n", + devname); return rv; } rsp_len = scsi_serial[3]; if (!rsp_len) { if (devname) - fprintf(stderr, - Name ": Failed to retrieve serial for %s\n", - devname); + pr_err("Failed to retrieve serial for %s\n", + devname); return 2; } rsp_buf = (char *) &scsi_serial[4]; @@ -3146,14 +3455,7 @@ load_imsm_disk(int fd, struct intel_super *super, char *devname, int keep_fd) if (rv != 0) return 2; - dl = calloc(1, sizeof(*dl)); - if (!dl) { - if (devname) - fprintf(stderr, - Name ": failed to allocate disk buffer for %s\n", - devname); - return 2; - } + dl = xcalloc(1, sizeof(*dl)); fstat(fd, &stb); dl->major = major(stb.st_rdev); @@ -3167,9 +3469,9 @@ load_imsm_disk(int fd, struct intel_super *super, char *devname, int keep_fd) dl->e = NULL; fd2devname(fd, name); if (devname) - dl->devname = strdup(devname); + dl->devname = xstrdup(devname); else - dl->devname = strdup(name); + dl->devname = xstrdup(name); /* look up this disk's index in the current anchor */ disk = __serial_to_disk(dl->serial, super->anchor, &dl->index); @@ -3296,18 +3598,12 @@ static int parse_raid_devices(struct intel_super *super) if (len_migr > len) space_needed += len_migr - len; - dv = malloc(sizeof(*dv)); - if (!dv) - return 1; + dv = xmalloc(sizeof(*dv)); if (max_len < len_migr) max_len = len_migr; if (max_len > len_migr) space_needed += max_len - len_migr; - dev_new = malloc(max_len); - if (!dev_new) { - free(dv); - return 1; - } + dev_new = xmalloc(max_len); imsm_copy_dev(dev_new, dev_iter); dv->dev = dev_new; dv->index = i; @@ -3343,7 +3639,7 @@ struct bbm_log *__get_imsm_bbm_log(struct imsm_super *mpb) if (__le32_to_cpu(mpb->bbm_log_size)) { ptr = mpb; ptr += mpb->mpb_size - __le32_to_cpu(mpb->bbm_log_size); - } + } return ptr; } @@ -3374,7 +3670,7 @@ int check_mpb_migr_compatibility(struct intel_super *super) /* This device is migrating */ map0 = get_imsm_map(dev_iter, MAP_0); map1 = get_imsm_map(dev_iter, MAP_1); - if (map0->pba_of_lba0 != map1->pba_of_lba0) + if (pba_of_lba0(map0) != pba_of_lba0(map1)) /* migration optimization area was used */ return -1; if (migr_rec->ascending_migr == 0 @@ -3402,40 +3698,34 @@ static int load_imsm_mpb(int fd, struct intel_super *super, char *devname) get_dev_size(fd, NULL, &dsize); if (dsize < 1024) { if (devname) - fprintf(stderr, - Name ": %s: device to small for imsm\n", - devname); + pr_err("%s: device to small for imsm\n", + devname); return 1; } if (lseek64(fd, dsize - (512 * 2), SEEK_SET) < 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 (posix_memalign((void**)&anchor, 512, 512) != 0) { if (devname) - fprintf(stderr, - Name ": Failed to allocate imsm anchor buffer" - " on %s\n", devname); + pr_err("Failed to allocate imsm anchor buffer on %s\n", devname); return 1; } if (read(fd, 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)); free(anchor); return 1; } if (strncmp((char *) anchor->sig, MPB_SIGNATURE, MPB_SIG_LEN) != 0) { if (devname) - fprintf(stderr, - Name ": no IMSM anchor on %s\n", devname); + pr_err("no IMSM anchor on %s\n", devname); free(anchor); return 2; } @@ -3448,9 +3738,8 @@ static int load_imsm_mpb(int fd, struct intel_super *super, char *devname) super->len = ROUND_UP(anchor->mpb_size, 512); if (posix_memalign(&super->buf, 512, super->len) != 0) { if (devname) - fprintf(stderr, - Name ": unable to allocate %zu byte mpb buffer\n", - super->len); + pr_err("unable to allocate %zu byte mpb buffer\n", + super->len); free(anchor); return 2; } @@ -3460,21 +3749,20 @@ static int load_imsm_mpb(int fd, struct intel_super *super, char *devname) free(anchor); if (posix_memalign(&super->migr_rec_buf, 512, MIGR_REC_BUF_SIZE) != 0) { - fprintf(stderr, Name - ": %s could not allocate migr_rec buffer\n", __func__); + pr_err("could not allocate migr_rec buffer\n"); free(super->buf); return 2; } + super->clean_migration_record_by_mdmon = 0; if (!sectors) { check_sum = __gen_imsm_checksum(super->anchor); if (check_sum != __le32_to_cpu(super->anchor->check_sum)) { if (devname) - fprintf(stderr, - Name ": IMSM checksum %x != %x on %s\n", - check_sum, - __le32_to_cpu(super->anchor->check_sum), - devname); + pr_err("IMSM checksum %x != %x on %s\n", + check_sum, + __le32_to_cpu(super->anchor->check_sum), + devname); return 2; } @@ -3484,27 +3772,24 @@ static int load_imsm_mpb(int fd, struct intel_super *super, char *devname) /* read the extended mpb */ if (lseek64(fd, dsize - (512 * (2 + sectors)), SEEK_SET) < 0) { if (devname) - fprintf(stderr, - Name ": Cannot seek to extended mpb on %s: %s\n", - devname, strerror(errno)); + pr_err("Cannot seek to extended mpb on %s: %s\n", + devname, strerror(errno)); return 1; } if ((unsigned)read(fd, super->buf + 512, super->len - 512) != super->len - 512) { if (devname) - fprintf(stderr, - Name ": Cannot read extended mpb on %s: %s\n", - devname, strerror(errno)); + pr_err("Cannot read extended mpb on %s: %s\n", + devname, strerror(errno)); return 2; } check_sum = __gen_imsm_checksum(super->anchor); if (check_sum != __le32_to_cpu(super->anchor->check_sum)) { if (devname) - fprintf(stderr, - Name ": IMSM checksum %x != %x on %s\n", - check_sum, __le32_to_cpu(super->anchor->check_sum), - devname); + pr_err("IMSM checksum %x != %x on %s\n", + check_sum, __le32_to_cpu(super->anchor->check_sum), + devname); return 3; } @@ -3519,6 +3804,32 @@ static int load_imsm_mpb(int fd, struct intel_super *super, char *devname) static int read_imsm_migr_rec(int fd, struct intel_super *super); +/* clears hi bits in metadata if MPB_ATTRIB_2TB_DISK not set */ +static void clear_hi(struct intel_super *super) +{ + struct imsm_super *mpb = super->anchor; + int i, n; + if (mpb->attributes & MPB_ATTRIB_2TB_DISK) + return; + for (i = 0; i < mpb->num_disks; ++i) { + struct imsm_disk *disk = &mpb->disk[i]; + disk->total_blocks_hi = 0; + } + for (i = 0; i < mpb->num_raid_devs; ++i) { + struct imsm_dev *dev = get_imsm_dev(super, i); + if (!dev) + return; + for (n = 0; n < 2; ++n) { + struct imsm_map *map = get_imsm_map(dev, n); + if (!map) + continue; + map->pba_of_lba0_hi = 0; + map->blocks_per_member_hi = 0; + map->num_data_stripes_hi = 0; + } + } +} + static int load_and_parse_mpb(int fd, struct intel_super *super, char *devname, int keep_fd) { @@ -3531,7 +3842,7 @@ load_and_parse_mpb(int fd, struct intel_super *super, char *devname, int keep_fd if (err) return err; err = parse_raid_devices(super); - + clear_hi(super); return err; } @@ -3617,13 +3928,10 @@ static void free_super_imsm(struct supertype *st) static struct intel_super *alloc_super(void) { - struct intel_super *super = malloc(sizeof(*super)); + struct intel_super *super = xcalloc(1, sizeof(*super)); - if (super) { - memset(super, 0, sizeof(*super)); - super->current_vol = -1; - super->create_offset = ~((__u32 ) 0); - } + super->current_vol = -1; + super->create_offset = ~((unsigned long long) 0); return super; } @@ -3643,9 +3951,8 @@ static int find_intel_hba_capability(int fd, struct intel_super *super, char *de hba_name = find_disk_attached_hba(fd, NULL); if (!hba_name) { if (devname) - fprintf(stderr, - Name ": %s is not attached to Intel(R) RAID controller.\n", - devname); + pr_err("%s is not attached to Intel(R) RAID controller.\n", + devname); return 1; } rv = attach_hba_to_super(super, hba_name); @@ -3653,14 +3960,14 @@ static int find_intel_hba_capability(int fd, struct intel_super *super, char *de if (devname) { struct intel_hba *hba = super->hba; - fprintf(stderr, Name ": %s is attached to Intel(R) %s RAID " - "controller (%s),\n" - " but the container is assigned to Intel(R) " - "%s RAID controller (", + pr_err("%s is attached to Intel(R) %s %s (%s),\n" + " but the container is assigned to Intel(R) %s %s (", devname, - hba_name->path, + get_sys_dev_type(hba_name->type), + hba_name->type == SYS_DEV_VMD ? "domain" : "RAID controller", hba_name->pci_id ? : "Err!", - get_sys_dev_type(hba_name->type)); + get_sys_dev_type(super->hba->type), + hba->type == SYS_DEV_VMD ? "domain" : "RAID controller"); while (hba) { fprintf(stderr, "%s", hba->pci_id ? : "Err!"); @@ -3668,18 +3975,16 @@ static int find_intel_hba_capability(int fd, struct intel_super *super, char *de fprintf(stderr, ", "); hba = hba->next; } - fprintf(stderr, ").\n" - " Mixing devices attached to different controllers " - "is not allowed.\n"); + " Mixing devices attached to different %s is not allowed.\n", + hba_name->type == SYS_DEV_VMD ? "VMD domains" : "controllers"); } - free_sys_dev(&hba_name); return 2; } - super->orom = find_imsm_capability(hba_name->type); - free_sys_dev(&hba_name); + super->orom = find_imsm_capability(hba_name); if (!super->orom) return 3; + return 0; } @@ -3700,13 +4005,11 @@ static int find_missing(struct intel_super *super) if (dl) continue; - dl = malloc(sizeof(*dl)); - if (!dl) - return 1; + dl = xmalloc(sizeof(*dl)); dl->major = 0; dl->minor = 0; dl->fd = -1; - dl->devname = strdup("missing"); + dl->devname = xstrdup("missing"); dl->index = i; serialcpy(dl->serial, disk->serial); dl->disk = *disk; @@ -3746,8 +4049,8 @@ static int __prep_thunderdome(struct intel_super **table, int tbl_size, if (tbl_mpb->family_num == mpb->family_num) { if (tbl_mpb->check_sum == mpb->check_sum) { - dprintf("%s: mpb from %d:%d matches %d:%d\n", - __func__, super->disks->major, + dprintf("mpb from %d:%d matches %d:%d\n", + super->disks->major, super->disks->minor, table[i]->disks->major, table[i]->disks->minor); @@ -3764,8 +4067,8 @@ static int __prep_thunderdome(struct intel_super **table, int tbl_size, */ struct intel_disk *idisk; - dprintf("%s: mpb from %d:%d replaces %d:%d\n", - __func__, super->disks->major, + dprintf("mpb from %d:%d replaces %d:%d\n", + super->disks->major, super->disks->minor, table[i]->disks->major, table[i]->disks->minor); @@ -3793,8 +4096,8 @@ static int __prep_thunderdome(struct intel_super **table, int tbl_size, idisk->disk.status |= CONFIGURED_DISK; } - dprintf("%s: mpb from %d:%d prefer %d:%d\n", - __func__, super->disks->major, + dprintf("mpb from %d:%d prefer %d:%d\n", + super->disks->major, super->disks->minor, table[i]->disks->major, table[i]->disks->minor); @@ -3821,9 +4124,7 @@ static int __prep_thunderdome(struct intel_super **table, int tbl_size, is_failed(&idisk->disk)) idisk->disk.status &= ~(SPARE_DISK); } else { - idisk = calloc(1, sizeof(*idisk)); - if (!idisk) - return -1; + idisk = xcalloc(1, sizeof(*idisk)); idisk->owner = IMSM_UNKNOWN_OWNER; idisk->disk = *disk; idisk->next = *disk_list; @@ -3855,12 +4156,12 @@ validate_members(struct intel_super *super, struct intel_disk *disk_list, idisk->owner == IMSM_UNKNOWN_OWNER) ok_count++; else - dprintf("%s: '%.16s' owner %d != %d\n", - __func__, disk->serial, idisk->owner, + dprintf("'%.16s' owner %d != %d\n", + disk->serial, idisk->owner, owner); } else { - dprintf("%s: unknown disk %x [%d]: %.16s\n", - __func__, __le32_to_cpu(mpb->family_num), i, + dprintf("unknown disk %x [%d]: %.16s\n", + __le32_to_cpu(mpb->family_num), i, disk->serial); break; } @@ -3878,7 +4179,7 @@ static void show_conflicts(__u32 family_num, struct intel_super *super_list) for (s = super_list; s; s = s->next) { if (family_num != s->anchor->family_num) continue; - fprintf(stderr, "Conflict, offlining family %#x on '%s'\n", + pr_err("Conflict, offlining family %#x on '%s'\n", __le32_to_cpu(family_num), s->disks->devname); } } @@ -3916,8 +4217,8 @@ imsm_thunderdome(struct intel_super **super_list, int len) s = NULL; if (!s) - dprintf("%s: marking family: %#x from %d:%d offline\n", - __func__, mpb->family_num, + dprintf("marking family: %#x from %d:%d offline\n", + mpb->family_num, super_table[i]->disks->major, super_table[i]->disks->minor); super_table[i] = s; @@ -3964,8 +4265,7 @@ imsm_thunderdome(struct intel_super **super_list, int len) champion = s; if (conflict) - fprintf(stderr, "Chose family %#x on '%s', " - "assemble conflicts to new container with '--update=uuid'\n", + pr_err("Chose family %#x on '%s', assemble conflicts to new container with '--update=uuid'\n", __le32_to_cpu(s->anchor->family_num), s->disks->devname); /* collect all dl's onto 'champion', and update them to @@ -3978,6 +4278,8 @@ imsm_thunderdome(struct intel_super **super_list, int len) if (s == champion) continue; + mpb->attributes |= s->anchor->attributes & MPB_ATTRIB_2TB_DISK; + for (i = 0; i < mpb->num_disks; i++) { struct imsm_disk *disk; @@ -4035,16 +4337,14 @@ imsm_thunderdome(struct intel_super **super_list, int len) return champion; } - static int get_sra_super_block(int fd, struct intel_super **super_list, char *devname, int *max, int keep_fd); -static int get_super_block(struct intel_super **super_list, int devnum, char *devname, +static int get_super_block(struct intel_super **super_list, char *devnm, char *devname, int major, int minor, int keep_fd); static int get_devlist_super_block(struct md_list *devlist, struct intel_super **super_list, int *max, int keep_fd); - static int load_super_imsm_all(struct supertype *st, int fd, void **sbp, char *devname, struct md_list *devlist, int keep_fd) @@ -4087,7 +4387,7 @@ static int load_super_imsm_all(struct supertype *st, int fd, void **sbp, /* Check migration compatibility */ if ((err == 0) && (check_mpb_migr_compatibility(super) != 0)) { - fprintf(stderr, Name ": Unsupported migration detected"); + pr_err("Unsupported migration detected"); if (devname) fprintf(stderr, " on %s\n", devname); else @@ -4107,15 +4407,14 @@ static int load_super_imsm_all(struct supertype *st, int fd, void **sbp, free_imsm(s); } - if (err) return err; *sbp = super; if (fd >= 0) - st->container_dev = fd2devnum(fd); + strcpy(st->container_devnm, fd2devnm(fd)); else - st->container_dev = NoMdDev; + st->container_devnm[0] = 0; if (err == 0 && st->ss == NULL) { st->ss = &super_imsm; st->minor_version = 0; @@ -4124,7 +4423,6 @@ static int load_super_imsm_all(struct supertype *st, int fd, void **sbp, return 0; } - static int get_devlist_super_block(struct md_list *devlist, struct intel_super **super_list, int *max, int keep_fd) @@ -4140,7 +4438,7 @@ get_devlist_super_block(struct md_list *devlist, struct intel_super **super_list int lmax = 0; int fd = dev_open(tmpdev->devname, O_RDONLY|O_EXCL); if (fd < 0) { - fprintf(stderr, Name ": cannot open device %s: %s\n", + pr_err("cannot open device %s: %s\n", tmpdev->devname, strerror(errno)); err = 8; goto error; @@ -4158,7 +4456,7 @@ get_devlist_super_block(struct md_list *devlist, struct intel_super **super_list int major = major(tmpdev->st_rdev); int minor = minor(tmpdev->st_rdev); err = get_super_block(super_list, - -1, + NULL, tmpdev->devname, major, minor, keep_fd); @@ -4174,13 +4472,12 @@ get_devlist_super_block(struct md_list *devlist, struct intel_super **super_list return err; } -static int get_super_block(struct intel_super **super_list, int devnum, char *devname, +static int get_super_block(struct intel_super **super_list, char *devnm, char *devname, int major, int minor, int keep_fd) { - struct intel_super*s = NULL; + struct intel_super *s; char nm[32]; int dfd = -1; - int rv; int err = 0; int retry; @@ -4197,17 +4494,11 @@ static int get_super_block(struct intel_super **super_list, int devnum, char *de goto error; } - rv = find_intel_hba_capability(dfd, s, devname); - /* no orom/efi or non-intel hba of the disk */ - if (rv != 0) { - err = 4; - goto error; - } - + find_intel_hba_capability(dfd, s, devname); err = load_and_parse_mpb(dfd, s, NULL, keep_fd); /* retry the load if we might have raced against mdmon */ - if (err == 3 && (devnum != -1) && mdmon_running(devnum)) + if (err == 3 && devnm && mdmon_running(devnm)) for (retry = 0; retry < 3; retry++) { usleep(3000); err = load_and_parse_mpb(dfd, s, NULL, keep_fd); @@ -4221,7 +4512,7 @@ static int get_super_block(struct intel_super **super_list, int devnum, char *de } else { if (s) free(s); - if (dfd) + if (dfd >= 0) close(dfd); } if ((dfd >= 0) && (!keep_fd)) @@ -4234,11 +4525,11 @@ static int get_sra_super_block(int fd, struct intel_super **super_list, char *devname, int *max, int keep_fd) { struct mdinfo *sra; - int devnum; + char *devnm; struct mdinfo *sd; int err = 0; int i = 0; - sra = sysfs_read(fd, 0, GET_LEVEL|GET_VERSION|GET_DEVS|GET_STATE); + sra = sysfs_read(fd, NULL, GET_LEVEL|GET_VERSION|GET_DEVS|GET_STATE); if (!sra) return 1; @@ -4249,9 +4540,9 @@ get_sra_super_block(int fd, struct intel_super **super_list, char *devname, int goto error; } /* load all mpbs */ - devnum = fd2devnum(fd); + devnm = fd2devnm(fd); for (sd = sra->devs, i = 0; sd; sd = sd->next, i++) { - if (get_super_block(super_list, devnum, devname, + if (get_super_block(super_list, devnm, devname, sd->disk.major, sd->disk.minor, keep_fd) != 0) { err = 7; goto error; @@ -4273,6 +4564,7 @@ static int load_super_imsm(struct supertype *st, int fd, char *devname) { struct intel_super *super; int rv; + int retry; if (test_partition(fd)) /* IMSM not allowed on partitions */ @@ -4281,12 +4573,6 @@ static int load_super_imsm(struct supertype *st, int fd, char *devname) free_super_imsm(st); super = alloc_super(); - if (!super) { - fprintf(stderr, - Name ": malloc of %zu failed.\n", - sizeof(*super)); - return 1; - } /* Load hba and capabilities if they exist. * But do not preclude loading metadata in case capabilities or hba are * non-compliant and ignore_hw_compat is set. @@ -4295,18 +4581,35 @@ static int load_super_imsm(struct supertype *st, int fd, char *devname) /* no orom/efi or non-intel hba of the disk */ if ((rv != 0) && (st->ignore_hw_compat == 0)) { if (devname) - fprintf(stderr, - Name ": No OROM/EFI properties for %s\n", devname); + pr_err("No OROM/EFI properties for %s\n", devname); free_imsm(super); return 2; } rv = load_and_parse_mpb(fd, super, devname, 0); + /* retry the load if we might have raced against mdmon */ + if (rv == 3) { + struct mdstat_ent *mdstat = NULL; + char *name = fd2kname(fd); + + if (name) + mdstat = mdstat_by_component(name); + + if (mdstat && mdmon_running(mdstat->devnm) && getpid() != mdmon_pid(mdstat->devnm)) { + for (retry = 0; retry < 3; retry++) { + usleep(3000); + rv = load_and_parse_mpb(fd, super, devname, 0); + if (rv != 3) + break; + } + } + + free_mdstat(mdstat); + } + 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_imsm(super); return rv; } @@ -4322,8 +4625,7 @@ static int load_super_imsm(struct supertype *st, int fd, char *devname) if (load_imsm_migr_rec(super, NULL) == 0) { /* Check for unsupported migration features */ if (check_mpb_migr_compatibility(super) != 0) { - fprintf(stderr, - Name ": Unsupported migration detected"); + pr_err("Unsupported migration detected"); if (devname) fprintf(stderr, " on %s\n", devname); else @@ -4342,22 +4644,13 @@ static __u16 info_to_blocks_per_strip(mdu_array_info_t *info) return info->chunk_size >> 9; } -static __u32 info_to_num_data_stripes(mdu_array_info_t *info, int num_domains) -{ - __u32 num_stripes; - - num_stripes = (info->size * 2) / info_to_blocks_per_strip(info); - num_stripes /= num_domains; - - return num_stripes; -} - -static __u32 info_to_blocks_per_member(mdu_array_info_t *info) +static unsigned long long info_to_blocks_per_member(mdu_array_info_t *info, + unsigned long long size) { if (info->level == 1) - return info->size * 2; + return size * 2; else - return (info->size * 2) & ~(info_to_blocks_per_strip(info) - 1); + return (size * 2) & ~(info_to_blocks_per_strip(info) - 1); } static void imsm_update_version_info(struct intel_super *super) @@ -4426,14 +4719,15 @@ static int check_name(struct intel_super *super, char *name, int quiet) } if (reason && !quiet) - fprintf(stderr, Name ": imsm volume name %s\n", reason); + pr_err("imsm volume name %s\n", reason); return !reason; } static int init_super_imsm_volume(struct supertype *st, mdu_array_info_t *info, unsigned long long size, char *name, - char *homehost, int *uuid) + char *homehost, int *uuid, + long long data_offset) { /* We are creating a volume inside a pre-existing container. * so st->sb is already set. @@ -4448,11 +4742,10 @@ static int init_super_imsm_volume(struct supertype *st, mdu_array_info_t *info, int i; unsigned long long array_blocks; size_t size_old, size_new; - __u32 num_data_stripes; + unsigned long long num_data_stripes; if (super->orom && mpb->num_raid_devs >= super->orom->vpa) { - fprintf(stderr, Name": This imsm-container already has the " - "maximum of %d volumes\n", super->orom->vpa); + pr_err("This imsm-container already has the maximum of %d volumes\n", super->orom->vpa); return 0; } @@ -4464,14 +4757,12 @@ static int init_super_imsm_volume(struct supertype *st, mdu_array_info_t *info, size_t size_round = ROUND_UP(size_new, 512); if (posix_memalign(&mpb_new, 512, size_round) != 0) { - fprintf(stderr, Name": could not allocate new mpb\n"); + pr_err("could not allocate new mpb\n"); return 0; } if (posix_memalign(&super->migr_rec_buf, 512, MIGR_REC_BUF_SIZE) != 0) { - fprintf(stderr, Name - ": %s could not allocate migr_rec buffer\n", - __func__); + pr_err("could not allocate migr_rec buffer\n"); free(super->buf); free(super); free(mpb_new); @@ -4515,32 +4806,19 @@ static int init_super_imsm_volume(struct supertype *st, mdu_array_info_t *info, for (d = super->missing; d; d = d->next) missing++; if (info->failed_disks > missing) { - fprintf(stderr, Name": unable to add 'missing' disk to container\n"); + pr_err("unable to add 'missing' disk to container\n"); return 0; } } if (!check_name(super, name, 0)) return 0; - dv = malloc(sizeof(*dv)); - if (!dv) { - fprintf(stderr, Name ": failed to allocate device list entry\n"); - return 0; - } - dev = calloc(1, sizeof(*dev) + sizeof(__u32) * (info->raid_disks - 1)); - if (!dev) { - free(dv); - fprintf(stderr, Name": could not allocate raid device\n"); - return 0; - } - + dv = xmalloc(sizeof(*dv)); + dev = xcalloc(1, sizeof(*dev) + sizeof(__u32) * (info->raid_disks - 1)); strncpy((char *) dev->volume, name, MAX_RAID_SERIAL_LEN); - if (info->level == 1) - array_blocks = info_to_blocks_per_member(info); - else - array_blocks = calc_array_size(info->level, info->raid_disks, + array_blocks = calc_array_size(info->level, info->raid_disks, info->layout, info->chunk_size, - info->size*2); + size * 2); /* round array size down to closest MB */ array_blocks = (array_blocks >> SECT_PER_MB_SHIFT) << SECT_PER_MB_SHIFT; @@ -4553,12 +4831,13 @@ static int init_super_imsm_volume(struct supertype *st, mdu_array_info_t *info, vol->dirty = !info->state; vol->curr_migr_unit = 0; map = get_imsm_map(dev, MAP_0); - map->pba_of_lba0 = __cpu_to_le32(super->create_offset); - map->blocks_per_member = __cpu_to_le32(info_to_blocks_per_member(info)); + set_pba_of_lba0(map, super->create_offset); + set_blocks_per_member(map, info_to_blocks_per_member(info, size)); map->blocks_per_strip = __cpu_to_le16(info_to_blocks_per_strip(info)); map->failed_disk_num = ~0; if (info->level > 0) - map->map_state = IMSM_T_STATE_UNINITIALIZED; + map->map_state = (info->state ? IMSM_T_STATE_NORMAL + : IMSM_T_STATE_UNINITIALIZED); else map->map_state = info->failed_disks ? IMSM_T_STATE_FAILED : IMSM_T_STATE_NORMAL; @@ -4567,8 +4846,7 @@ static int init_super_imsm_volume(struct supertype *st, mdu_array_info_t *info, if (info->level == 1 && info->raid_disks > 2) { free(dev); free(dv); - fprintf(stderr, Name": imsm does not support more than 2 disks" - "in a raid1 volume\n"); + pr_err("imsm does not support more than 2 disksin a raid1 volume\n"); return 0; } @@ -4581,8 +4859,10 @@ static int init_super_imsm_volume(struct supertype *st, mdu_array_info_t *info, else map->num_domains = 1; - num_data_stripes = info_to_num_data_stripes(info, map->num_domains); - map->num_data_stripes = __cpu_to_le32(num_data_stripes); + /* info->size is only int so use the 'size' parameter instead */ + num_data_stripes = (size * 2) / info_to_blocks_per_strip(info); + num_data_stripes /= map->num_domains; + set_num_data_stripes(map, num_data_stripes); map->num_members = info->raid_disks; for (i = 0; i < map->num_members; i++) { @@ -4603,7 +4883,8 @@ static int init_super_imsm_volume(struct supertype *st, mdu_array_info_t *info, static int init_super_imsm(struct supertype *st, mdu_array_info_t *info, unsigned long long size, char *name, - char *homehost, int *uuid) + char *homehost, 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 @@ -4618,8 +4899,14 @@ static int init_super_imsm(struct supertype *st, mdu_array_info_t *info, size_t mpb_size; char *version; + if (data_offset != INVALID_SECTORS) { + pr_err("data-offset not supported by imsm\n"); + return 0; + } + if (st->sb) - return init_super_imsm_volume(st, info, size, name, homehost, uuid); + return init_super_imsm_volume(st, info, size, name, homehost, uuid, + data_offset); if (info) mpb_size = disks_to_mpb_size(info->nr_disks); @@ -4632,13 +4919,11 @@ static int init_super_imsm(struct supertype *st, mdu_array_info_t *info, super = NULL; } if (!super) { - fprintf(stderr, Name - ": %s could not allocate superblock\n", __func__); + pr_err("could not allocate superblock\n"); return 0; } if (posix_memalign(&super->migr_rec_buf, 512, MIGR_REC_BUF_SIZE) != 0) { - fprintf(stderr, Name - ": %s could not allocate migr_rec buffer\n", __func__); + pr_err("could not allocate migr_rec buffer\n"); free(super->buf); free(super); return 0; @@ -4679,7 +4964,7 @@ static int add_to_super_imsm_volume(struct supertype *st, mdu_disk_info_t *dk, map = get_imsm_map(dev, MAP_0); if (! (dk->state & (1<index); if (slot >= 0 && (get_imsm_ord_tbl_ent(dev, slot, MAP_X) & IMSM_ORD_REBUILD) == 0) { - fprintf(stderr, Name ": %s has been included in this array twice\n", + pr_err("%s has been included in this array twice\n", devname); return 1; } @@ -4725,8 +5010,8 @@ static int add_to_super_imsm_volume(struct supertype *st, mdu_disk_info_t *dk, */ if (super->current_vol == 0) { for (df = super->missing; df; df = df->next) { - if (dl->disk.total_blocks > df->disk.total_blocks) - df->disk.total_blocks = dl->disk.total_blocks; + if (total_blocks(&dl->disk) > total_blocks(&df->disk)) + set_total_blocks(&df->disk, total_blocks(&dl->disk)); _disk = __get_imsm_disk(mpb, df->index); *_disk = df->disk; } @@ -4768,7 +5053,7 @@ static int add_to_super_imsm_volume(struct supertype *st, mdu_disk_info_t *dk, _disk = __get_imsm_disk(mpb, dl->index); if (!_dev || !_disk) { - fprintf(stderr, Name ": BUG mpb setup error\n"); + pr_err("BUG mpb setup error\n"); return 1; } *_dev = *dev; @@ -4815,7 +5100,8 @@ int mark_spare(struct dl *disk) } static int add_to_super_imsm(struct supertype *st, mdu_disk_info_t *dk, - int fd, char *devname) + int fd, char *devname, + unsigned long long data_offset) { struct intel_super *super = st->sb; struct dl *dd; @@ -4841,31 +5127,36 @@ static int add_to_super_imsm(struct supertype *st, mdu_disk_info_t *dk, return add_to_super_imsm_volume(st, dk, fd, devname); fstat(fd, &stb); - dd = malloc(sizeof(*dd)); - if (!dd) { - fprintf(stderr, - Name ": malloc failed %s:%d.\n", __func__, __LINE__); - return 1; - } - memset(dd, 0, sizeof(*dd)); + dd = xcalloc(sizeof(*dd), 1); dd->major = major(stb.st_rdev); dd->minor = minor(stb.st_rdev); - dd->devname = devname ? strdup(devname) : NULL; + dd->devname = devname ? xstrdup(devname) : NULL; dd->fd = fd; dd->e = NULL; dd->action = DISK_ADD; rv = imsm_read_serial(fd, devname, dd->serial); if (rv) { - fprintf(stderr, - Name ": failed to retrieve scsi serial, aborting\n"); + pr_err("failed to retrieve scsi serial, aborting\n"); free(dd); abort(); } get_dev_size(fd, NULL, &size); + /* clear migr_rec when adding disk to container */ + memset(super->migr_rec_buf, 0, MIGR_REC_BUF_SIZE); + if (lseek64(fd, size - MIGR_REC_POSITION, SEEK_SET) >= 0) { + if (write(fd, super->migr_rec_buf, + MIGR_REC_BUF_SIZE) != MIGR_REC_BUF_SIZE) + perror("Write migr_rec failed"); + } + size /= 512; serialcpy(dd->disk.serial, dd->serial); - dd->disk.total_blocks = __cpu_to_le32(size); + set_total_blocks(&dd->disk, size); + if (__le32_to_cpu(dd->disk.total_blocks_hi) > 0) { + struct imsm_super *mpb = super->anchor; + mpb->attributes |= MPB_ATTRIB_2TB_DISK; + } mark_spare(dd); if (sysfs_disk_to_scsi_id(fd, &id) == 0) dd->disk.scsi_id = __cpu_to_le32(id); @@ -4884,7 +5175,6 @@ static int add_to_super_imsm(struct supertype *st, mdu_disk_info_t *dk, return 0; } - static int remove_from_super_imsm(struct supertype *st, mdu_disk_info_t *dk) { struct intel_super *super = st->sb; @@ -4895,18 +5185,10 @@ static int remove_from_super_imsm(struct supertype *st, mdu_disk_info_t *dk) * is prepared. */ if (!st->update_tail) { - fprintf(stderr, - Name ": %s shall be used in mdmon context only" - "(line %d).\n", __func__, __LINE__); + pr_err("shall be used in mdmon context only\n"); return 1; } - dd = malloc(sizeof(*dd)); - if (!dd) { - fprintf(stderr, - Name ": malloc failed %s:%d.\n", __func__, __LINE__); - return 1; - } - memset(dd, 0, sizeof(*dd)); + dd = xcalloc(1, sizeof(*dd)); dd->major = dk->major; dd->minor = dk->minor; dd->fd = -1; @@ -4916,7 +5198,6 @@ static int remove_from_super_imsm(struct supertype *st, mdu_disk_info_t *dk) dd->next = super->disk_mgmt_list; super->disk_mgmt_list = dd; - return 0; } @@ -4937,13 +5218,13 @@ static int write_super_imsm_spares(struct intel_super *super, int doclose) __u32 sum; struct dl *d; - spare->mpb_size = __cpu_to_le32(sizeof(struct imsm_super)), - spare->generation_num = __cpu_to_le32(1UL), + spare->mpb_size = __cpu_to_le32(sizeof(struct imsm_super)); + spare->generation_num = __cpu_to_le32(1UL); spare->attributes = MPB_ATTRIB_CHECKSUM_VERIFY; - spare->num_disks = 1, - spare->num_raid_devs = 0, - spare->cache_size = mpb->cache_size, - spare->pwr_cycle_count = __cpu_to_le32(1), + spare->num_disks = 1; + spare->num_raid_devs = 0; + spare->cache_size = mpb->cache_size; + spare->pwr_cycle_count = __cpu_to_le32(1); snprintf((char *) spare->sig, MAX_SIGNATURE_LENGTH, MPB_SIGNATURE MPB_VERSION_RAID0); @@ -4953,6 +5234,9 @@ static int write_super_imsm_spares(struct intel_super *super, int doclose) continue; spare->disk[0] = d->disk; + if (__le32_to_cpu(d->disk.total_blocks_hi) > 0) + spare->attributes |= MPB_ATTRIB_2TB_DISK; + sum = __gen_imsm_checksum(spare); spare->family_num = __cpu_to_le32(sum); spare->orig_family_num = 0; @@ -4960,8 +5244,8 @@ static int write_super_imsm_spares(struct intel_super *super, int doclose) spare->check_sum = __cpu_to_le32(sum); if (store_imsm_mpb(d->fd, spare)) { - fprintf(stderr, "%s: failed for device %d:%d %s\n", - __func__, d->major, d->minor, strerror(errno)); + pr_err("failed for device %d:%d %s\n", + d->major, d->minor, strerror(errno)); return 1; } if (doclose) { @@ -5029,6 +5313,10 @@ static int write_super_imsm(struct supertype *st, int doclose) sum = __gen_imsm_checksum(mpb); mpb->check_sum = __cpu_to_le32(sum); + if (super->clean_migration_record_by_mdmon) { + clear_migration_record = 1; + super->clean_migration_record_by_mdmon = 0; + } if (clear_migration_record) memset(super->migr_rec_buf, 0, MIGR_REC_BUF_SIZE); @@ -5036,9 +5324,6 @@ static int write_super_imsm(struct supertype *st, int doclose) for (d = super->disks; d ; d = d->next) { if (d->index < 0 || is_failed(&d->disk)) continue; - if (store_imsm_mpb(d->fd, mpb)) - fprintf(stderr, "%s: failed for device %d:%d (fd: %d)%s\n", - __func__, d->major, d->minor, d->fd, strerror(errno)); if (clear_migration_record) { unsigned long long dsize; @@ -5050,6 +5335,13 @@ static int write_super_imsm(struct supertype *st, int doclose) perror("Write migr_rec failed"); } } + + if (store_imsm_mpb(d->fd, mpb)) + fprintf(stderr, + "failed for device %d:%d (fd: %d)%s\n", + d->major, d->minor, + d->fd, strerror(errno)); + if (doclose) { close(d->fd); d->fd = -1; @@ -5062,7 +5354,6 @@ static int write_super_imsm(struct supertype *st, int doclose) return 0; } - static int create_array(struct supertype *st, int dev_idx) { size_t len; @@ -5076,13 +5367,7 @@ static int create_array(struct supertype *st, int dev_idx) len = sizeof(*u) - sizeof(*dev) + sizeof_imsm_dev(dev, 0) + sizeof(*inf) * map->num_members; - u = malloc(len); - if (!u) { - fprintf(stderr, "%s: failed to allocate update buffer\n", - __func__); - return 1; - } - + u = xmalloc(len); u->type = update_create_array; u->dev_idx = dev_idx; imsm_copy_dev(&u->dev, dev); @@ -5091,6 +5376,8 @@ static int create_array(struct supertype *st, int dev_idx) int idx = get_imsm_disk_idx(dev, i, MAP_X); disk = get_imsm_disk(super, idx); + if (!disk) + disk = get_imsm_missing(super, idx); serialcpy(inf[i].serial, disk->serial); } append_metadata_update(st, u, len); @@ -5108,13 +5395,7 @@ static int mgmt_disk(struct supertype *st) return 0; len = sizeof(*u); - u = malloc(len); - if (!u) { - fprintf(stderr, "%s: failed to allocate update buffer\n", - __func__); - return 1; - } - + u = xmalloc(len); u->type = update_add_remove_disk; append_metadata_update(st, u, len); @@ -5147,7 +5428,7 @@ static int write_init_super_imsm(struct supertype *st) } else { struct dl *d; for (d = super->disks; d; d = d->next) - Kill(d->devname, NULL, 0, 1, 1); + Kill(d->devname, NULL, 0, -1, 1); return write_super_imsm(st, 1); } } @@ -5176,13 +5457,15 @@ static int imsm_bbm_log_size(struct imsm_super *mpb) #ifndef MDASSEMBLE static int validate_geometry_imsm_container(struct supertype *st, int level, int layout, int raiddisks, int chunk, - unsigned long long size, char *dev, + unsigned long long size, + unsigned long long data_offset, + char *dev, unsigned long long *freesize, int verbose) { int fd; unsigned long long ldsize; - struct intel_super *super=NULL; + struct intel_super *super; int rv = 0; if (level != LEVEL_CONTAINER) @@ -5192,8 +5475,8 @@ static int validate_geometry_imsm_container(struct supertype *st, int level, fd = open(dev, O_RDONLY|O_EXCL, 0); if (fd < 0) { - if (verbose) - fprintf(stderr, Name ": imsm: Cannot open %s: %s\n", + if (verbose > 0) + pr_err("imsm: Cannot open %s: %s\n", dev, strerror(errno)); return 0; } @@ -5206,20 +5489,12 @@ static int validate_geometry_imsm_container(struct supertype *st, int level, * note that there is no fd for the disks in array. */ super = alloc_super(); - if (!super) { - fprintf(stderr, - Name ": malloc of %zu failed.\n", - sizeof(*super)); - close(fd); - return 0; - } - - rv = find_intel_hba_capability(fd, super, verbose ? dev : NULL); + rv = find_intel_hba_capability(fd, super, verbose > 0 ? dev : NULL); if (rv != 0) { #if DEBUG char str[256]; fd2devname(fd, str); - dprintf("validate_geometry_imsm_container: fd: %d %s orom: %p rv: %d raiddisk: %d\n", + dprintf("fd: %d %s orom: %p rv: %d raiddisk: %d\n", fd, str, super->orom, rv, raiddisks); #endif /* no orom/efi or non-intel hba of the disk */ @@ -5228,17 +5503,24 @@ static int validate_geometry_imsm_container(struct supertype *st, int level, return 0; } close(fd); - if (super->orom && raiddisks > super->orom->tds) { - if (verbose) - fprintf(stderr, Name ": %d exceeds maximum number of" - " platform supported disks: %d\n", - raiddisks, super->orom->tds); - - free_imsm(super); - return 0; + if (super->orom) { + if (raiddisks > super->orom->tds) { + if (verbose) + pr_err("%d exceeds maximum number of platform supported disks: %d\n", + raiddisks, super->orom->tds); + free_imsm(super); + return 0; + } + if ((super->orom->attr & IMSM_OROM_ATTR_2TB_DISK) == 0 && + (ldsize >> 9) >> 32 > 0) { + if (verbose) + pr_err("%s exceeds maximum platform supported size\n", dev); + free_imsm(super); + return 0; + } } - *freesize = avail_size_imsm(st, ldsize >> 9); + *freesize = avail_size_imsm(st, ldsize >> 9, data_offset); free_imsm(super); return 1; @@ -5277,7 +5559,7 @@ static unsigned long long merge_extents(struct intel_super *super, int sum_exten * 'maxsize' given the "all disks in an array must share a common start * offset" constraint */ - struct extent *e = calloc(sum_extents, sizeof(*e)); + struct extent *e = xcalloc(sum_extents, sizeof(*e)); struct dl *dl; int i, j; int start_extent; @@ -5286,9 +5568,6 @@ static unsigned long long merge_extents(struct intel_super *super, int sum_exten unsigned long long maxsize; unsigned long reserve; - if (!e) - return 0; - /* coalesce and sort all extents. also, check to see if we need to * reserve space between member arrays */ @@ -5344,7 +5623,7 @@ static unsigned long long merge_extents(struct intel_super *super, int sum_exten if (maxsize < reserve) return 0; - super->create_offset = ~((__u32) 0); + super->create_offset = ~((unsigned long long) 0); if (start + reserve > super->create_offset) return 0; /* start overflows create_offset */ super->create_offset = start + reserve; @@ -5374,16 +5653,15 @@ static int is_raid_level_supported(const struct imsm_orom *orom, int level, int return 0; } - static int active_arrays_by_format(char *name, char* hba, struct md_list **devlist, int dpa, int verbose) { struct mdstat_ent *mdstat = mdstat_read(0, 0); - struct mdstat_ent *memb = NULL; + struct mdstat_ent *memb; int count = 0; int num = 0; - struct md_list *dv = NULL; + struct md_list *dv; int found; for (memb = mdstat ; memb ; memb = memb->next) { @@ -5395,17 +5673,15 @@ active_arrays_by_format(char *name, char* hba, struct md_list **devlist, struct dev_member *dev = memb->members; int fd = -1; while(dev && (fd < 0)) { - char *path = malloc(strlen(dev->name) + strlen("/dev/") + 1); - if (path) { - num = sprintf(path, "%s%s", "/dev/", dev->name); - if (num > 0) - fd = open(path, O_RDONLY, 0); - if ((num <= 0) || (fd < 0)) { - pr_vrb(": Cannot open %s: %s\n", - dev->name, strerror(errno)); - } - free(path); + char *path = xmalloc(strlen(dev->name) + strlen("/dev/") + 1); + num = sprintf(path, "%s%s", "/dev/", dev->name); + if (num > 0) + fd = open(path, O_RDONLY, 0); + if ((num <= 0) || (fd < 0)) { + pr_vrb("Cannot open %s: %s\n", + dev->name, strerror(errno)); } + free(path); dev = dev->next; } found = 0; @@ -5414,26 +5690,19 @@ active_arrays_by_format(char *name, char* hba, struct md_list **devlist, for (vol = mdstat ; vol ; vol = vol->next) { if ((vol->active > 0) && vol->metadata_version && - is_container_member(vol, memb->dev)) { + is_container_member(vol, memb->devnm)) { found++; count++; } } if (*devlist && (found < dpa)) { - dv = calloc(1, sizeof(*dv)); - if (dv == NULL) - fprintf(stderr, Name ": calloc failed\n"); - else { - dv->devname = malloc(strlen(memb->dev) + strlen("/dev/") + 1); - if (dv->devname != NULL) { - sprintf(dv->devname, "%s%s", "/dev/", memb->dev); - dv->found = found; - dv->used = 0; - dv->next = *devlist; - *devlist = dv; - } else - free(dv); - } + dv = xcalloc(1, sizeof(*dv)); + dv->devname = xmalloc(strlen(memb->devnm) + strlen("/dev/") + 1); + sprintf(dv->devname, "%s%s", "/dev/", memb->devnm); + dv->found = found; + dv->used = 0; + dv->next = *devlist; + *devlist = dv; } } if (fd >= 0) @@ -5450,20 +5719,11 @@ get_loop_devices(void) { int i; struct md_list *devlist = NULL; - struct md_list *dv = NULL; + struct md_list *dv; for(i = 0; i < 12; i++) { - dv = calloc(1, sizeof(*dv)); - if (dv == NULL) { - fprintf(stderr, Name ": calloc failed\n"); - break; - } - dv->devname = malloc(40); - if (dv->devname == NULL) { - fprintf(stderr, Name ": malloc failed\n"); - free(dv); - break; - } + dv = xcalloc(1, sizeof(*dv)); + dv->devname = xmalloc(40); sprintf(dv->devname, "/dev/loop%d", i); dv->next = devlist; devlist = dv; @@ -5476,7 +5736,7 @@ static struct md_list* get_devices(const char *hba_path) { struct md_list *devlist = NULL; - struct md_list *dv = NULL; + struct md_list *dv; struct dirent *ent; DIR *dir; int err = 0; @@ -5511,25 +5771,13 @@ get_devices(const char *hba_path) fd2devname(fd, buf); close(fd); } else { - fprintf(stderr, Name ": cannot open device: %s\n", + pr_err("cannot open device: %s\n", ent->d_name); continue; } - - dv = calloc(1, sizeof(*dv)); - if (dv == NULL) { - fprintf(stderr, Name ": malloc failed\n"); - err = 1; - break; - } - dv->devname = strdup(buf); - if (dv->devname == NULL) { - fprintf(stderr, Name ": malloc failed\n"); - err = 1; - free(dv); - break; - } + dv = xcalloc(1, sizeof(*dv)); + dv->devname = xstrdup(buf); dv->next = devlist; devlist = dv; } @@ -5541,6 +5789,7 @@ get_devices(const char *hba_path) free(dv); } } + closedir(dir); return devlist; } @@ -5550,7 +5799,7 @@ count_volumes_list(struct md_list *devlist, char *homehost, { struct md_list *tmpdev; int count = 0; - struct supertype *st = NULL; + struct supertype *st; /* first walk the list of devices to find a consistent set * that match the criterea, if that is possible. @@ -5559,7 +5808,7 @@ count_volumes_list(struct md_list *devlist, char *homehost, *found = 0; st = match_metadata_desc_imsm("imsm"); if (st == NULL) { - pr_vrb(": cannot allocate memory for imsm supertype\n"); + pr_vrb("cannot allocate memory for imsm supertype\n"); return 0; } @@ -5572,33 +5821,33 @@ count_volumes_list(struct md_list *devlist, char *homehost, continue; tst = dup_super(st); if (tst == NULL) { - pr_vrb(": cannot allocate memory for imsm supertype\n"); + pr_vrb("cannot allocate memory for imsm supertype\n"); goto err_1; } tmpdev->container = 0; dfd = dev_open(devname, O_RDONLY|O_EXCL); if (dfd < 0) { - dprintf(": cannot open device %s: %s\n", + dprintf("cannot open device %s: %s\n", devname, strerror(errno)); tmpdev->used = 2; } else if (fstat(dfd, &stb)< 0) { /* Impossible! */ - dprintf(": fstat failed for %s: %s\n", + dprintf("fstat failed for %s: %s\n", devname, strerror(errno)); tmpdev->used = 2; } else if ((stb.st_mode & S_IFMT) != S_IFBLK) { - dprintf(": %s is not a block device.\n", + dprintf("%s is not a block device.\n", devname); tmpdev->used = 2; } else if (must_be_container(dfd)) { struct supertype *cst; cst = super_by_fd(dfd, NULL); if (cst == NULL) { - dprintf(": cannot recognize container type %s\n", + dprintf("cannot recognize container type %s\n", devname); tmpdev->used = 2; } else if (tst->ss != st->ss) { - dprintf(": non-imsm container - ignore it: %s\n", + dprintf("non-imsm container - ignore it: %s\n", devname); tmpdev->used = 2; } else if (!tst->ss->load_container || @@ -5612,11 +5861,11 @@ count_volumes_list(struct md_list *devlist, char *homehost, } else { tmpdev->st_rdev = stb.st_rdev; if (tst->ss->load_super(tst,dfd, NULL)) { - dprintf(": no RAID superblock on %s\n", + dprintf("no RAID superblock on %s\n", devname); tmpdev->used = 2; } else if (tst->ss->compare_super == NULL) { - dprintf(": Cannot assemble %s metadata on %s\n", + dprintf("Cannot assemble %s metadata on %s\n", tst->ss->name, devname); tmpdev->used = 2; } @@ -5653,7 +5902,7 @@ count_volumes_list(struct md_list *devlist, char *homehost, * Or, if we are auto assembling, we just ignore the second * for now. */ - dprintf(": superblock on %s doesn't match others - assembly aborted\n", + dprintf("superblock on %s doesn't match others - assembly aborted\n", devname); goto loop; } @@ -5675,7 +5924,7 @@ count_volumes_list(struct md_list *devlist, char *homehost, if (iter->array.state & (1<text_version); } else count++; @@ -5683,11 +5932,11 @@ count_volumes_list(struct md_list *devlist, char *homehost, sysfs_free(head); } else { - dprintf(" no valid super block on device list: err: %d %p\n", + dprintf("No valid super block on device list: err: %d %p\n", err, st->sb); } } else { - dprintf(" no more devices to examin\n"); + dprintf("no more devices to examine\n"); } for (tmpdev = devlist; tmpdev; tmpdev = tmpdev->next) { @@ -5708,39 +5957,76 @@ count_volumes_list(struct md_list *devlist, char *homehost, return count; } - static int -count_volumes(char *hba, int dpa, int verbose) +count_volumes(struct intel_hba *hba, int dpa, int verbose) { - struct md_list *devlist = NULL; + struct sys_dev *idev, *intel_devices = find_intel_devices(); int count = 0; - int found = 0;; + const struct orom_entry *entry; + struct devid_list *dv, *devid_list; - devlist = get_devices(hba); - /* if no intel devices return zero volumes */ - if (devlist == NULL) + if (!hba || !hba->path) return 0; - count = active_arrays_by_format("imsm", hba, &devlist, dpa, verbose); - dprintf(" path: %s active arrays: %d\n", hba, count); - if (devlist == NULL) + for (idev = intel_devices; idev; idev = idev->next) { + if (strstr(idev->path, hba->path)) + break; + } + + if (!idev || !idev->dev_id) return 0; - do { - found = 0; - count += count_volumes_list(devlist, - NULL, - verbose, - &found); - dprintf("found %d count: %d\n", found, count); - } while (found); - - dprintf("path: %s total number of volumes: %d\n", hba, count); - - while(devlist) { - struct md_list *dv = devlist; - devlist = devlist->next; - free(dv->devname); - free(dv); + + entry = get_orom_entry_by_device_id(idev->dev_id); + + if (!entry || !entry->devid_list) + return 0; + + devid_list = entry->devid_list; + for (dv = devid_list; dv; dv = dv->next) { + struct md_list *devlist; + struct sys_dev *device = device_by_id(dv->devid); + char *hba_path; + int found = 0; + + if (device) + hba_path = device->path; + else + return 0; + + /* VMD has one orom entry for all domain, but spanning is not allowed. + * VMD arrays should be counted per domain (controller), so skip + * domains that are not the given one. + */ + if ((hba->type == SYS_DEV_VMD) && + (strncmp(device->path, hba->path, strlen(device->path)) != 0)) + continue; + + devlist = get_devices(hba_path); + /* if no intel devices return zero volumes */ + if (devlist == NULL) + return 0; + + count += active_arrays_by_format("imsm", hba_path, &devlist, dpa, verbose); + dprintf("path: %s active arrays: %d\n", hba_path, count); + if (devlist == NULL) + return 0; + do { + found = 0; + count += count_volumes_list(devlist, + NULL, + verbose, + &found); + dprintf("found %d count: %d\n", found, count); + } while (found); + + dprintf("path: %s total number of volumes: %d\n", hba_path, count); + + while (devlist) { + struct md_list *dv = devlist; + devlist = devlist->next; + free(dv->devname); + free(dv); + } } return count; } @@ -5757,50 +6043,58 @@ static int imsm_default_chunk(const struct imsm_orom *orom) static int validate_geometry_imsm_orom(struct intel_super *super, int level, int layout, - int raiddisks, int *chunk, int verbose) + int raiddisks, int *chunk, unsigned long long size, int verbose) { /* check/set platform and metadata limits/defaults */ if (super->orom && raiddisks > super->orom->dpa) { - pr_vrb(": platform supports a maximum of %d disks per array\n", + pr_vrb("platform supports a maximum of %d disks per array\n", super->orom->dpa); return 0; } - /* capabilities of OROM tested - copied from validate_geometry_imsm_volume */ + /* capabilities of OROM tested - copied from validate_geometry_imsm_volume */ if (!is_raid_level_supported(super->orom, level, raiddisks)) { - pr_vrb(": platform does not support raid%d with %d disk%s\n", + pr_vrb("platform does not support raid%d with %d disk%s\n", level, raiddisks, raiddisks > 1 ? "s" : ""); return 0; } - if (chunk && (*chunk == 0 || *chunk == UnSet)) + if (*chunk == 0 || *chunk == UnSet) *chunk = imsm_default_chunk(super->orom); - if (super->orom && chunk && !imsm_orom_has_chunk(super->orom, *chunk)) { - pr_vrb(": platform does not support a chunk size of: " - "%d\n", *chunk); + if (super->orom && !imsm_orom_has_chunk(super->orom, *chunk)) { + pr_vrb("platform does not support a chunk size of: %d\n", *chunk); return 0; } if (layout != imsm_level_to_layout(level)) { if (level == 5) - pr_vrb(": imsm raid 5 only supports the left-asymmetric layout\n"); + pr_vrb("imsm raid 5 only supports the left-asymmetric layout\n"); else if (level == 10) - pr_vrb(": imsm raid 10 only supports the n2 layout\n"); + pr_vrb("imsm raid 10 only supports the n2 layout\n"); else - pr_vrb(": imsm unknown layout %#x for this raid level %d\n", + pr_vrb("imsm unknown layout %#x for this raid level %d\n", layout, level); return 0; } + + if (super->orom && (super->orom->attr & IMSM_OROM_ATTR_2TB) == 0 && + (calc_array_size(level, raiddisks, layout, *chunk, size) >> 32) > 0) { + pr_vrb("platform does not support a volume size over 2TB\n"); + return 0; + } + return 1; } -/* validate_geometry_imsm_volume - lifted from validate_geometry_ddf_bvd +/* validate_geometry_imsm_volume - lifted from validate_geometry_ddf_bvd * FIX ME add ahci details */ static int validate_geometry_imsm_volume(struct supertype *st, int level, int layout, int raiddisks, int *chunk, - unsigned long long size, char *dev, + unsigned long long size, + unsigned long long data_offset, + char *dev, unsigned long long *freesize, int verbose) { @@ -5819,9 +6113,8 @@ static int validate_geometry_imsm_volume(struct supertype *st, int level, mpb = super->anchor; - if (!validate_geometry_imsm_orom(super, level, layout, raiddisks, chunk, verbose)) { - fprintf(stderr, Name ": RAID gemetry validation failed. " - "Cannot proceed with the action(s).\n"); + if (!validate_geometry_imsm_orom(super, level, layout, raiddisks, chunk, size, verbose)) { + pr_err("RAID gemetry validation failed. Cannot proceed with the action(s).\n"); return 0; } if (!dev) { @@ -5862,9 +6155,7 @@ static int validate_geometry_imsm_volume(struct supertype *st, int level, } if (dcnt < raiddisks) { if (verbose) - fprintf(stderr, Name ": imsm: Not enough " - "devices with space for this array " - "(%d < %d)\n", + pr_err("imsm: Not enough devices with space for this array (%d < %d)\n", dcnt, raiddisks); return 0; } @@ -5883,8 +6174,7 @@ static int validate_geometry_imsm_volume(struct supertype *st, int level, } if (!dl) { if (verbose) - fprintf(stderr, Name ": %s is not in the " - "same imsm set\n", dev); + pr_err("%s is not in the same imsm set\n", dev); return 0; } else if (super->orom && dl->index < 0 && mpb->num_raid_devs) { /* If a volume is present then the current creation attempt @@ -5892,15 +6182,12 @@ static int validate_geometry_imsm_volume(struct supertype *st, int level, * understand this configuration (all member disks must be * members of each array in the container). */ - fprintf(stderr, Name ": %s is a spare and a volume" - " is already defined for this container\n", dev); - fprintf(stderr, Name ": The option-rom requires all member" - " disks to be a member of all volumes\n"); + pr_err("%s is a spare and a volume is already defined for this container\n", dev); + pr_err("The option-rom requires all member disks to be a member of all volumes\n"); return 0; } else if (super->orom && mpb->num_raid_devs > 0 && mpb->num_disks != raiddisks) { - fprintf(stderr, Name ": The option-rom requires all member" - " disks to be a member of all volumes\n"); + pr_err("The option-rom requires all member disks to be a member of all volumes\n"); return 0; } @@ -5922,13 +6209,13 @@ static int validate_geometry_imsm_volume(struct supertype *st, int level, dl->extent_cnt = i; } else { if (verbose) - fprintf(stderr, Name ": unable to determine free space for: %s\n", + pr_err("unable to determine free space for: %s\n", dev); return 0; } if (maxsize < size) { if (verbose) - fprintf(stderr, Name ": %s not enough space (%llu < %llu)\n", + pr_err("%s not enough space (%llu < %llu)\n", dev, maxsize, size); return 0; } @@ -5943,21 +6230,16 @@ static int validate_geometry_imsm_volume(struct supertype *st, int level, if (!check_env("IMSM_NO_PLATFORM") && mpb->num_raid_devs > 0 && size && size != maxsize) { - fprintf(stderr, Name ": attempting to create a second " - "volume with size less then remaining space. " - "Aborting...\n"); + pr_err("attempting to create a second volume with size less then remaining space. Aborting...\n"); return 0; } if (maxsize < size || maxsize == 0) { if (verbose) { if (maxsize == 0) - fprintf(stderr, Name ": no free space" - " left on device. Aborting...\n"); + pr_err("no free space left on device. Aborting...\n"); else - fprintf(stderr, Name ": not enough space" - " to create volume of given size" - " (%llu < %llu). Aborting...\n", + pr_err("not enough space to create volume of given size (%llu < %llu). Aborting...\n", maxsize, size); } return 0; @@ -5966,10 +6248,10 @@ static int validate_geometry_imsm_volume(struct supertype *st, int level, *freesize = maxsize; if (super->orom) { - int count = count_volumes(super->hba->path, + int count = count_volumes(super->hba, super->orom->dpa, verbose); if (super->orom->vphba <= count) { - pr_vrb(": platform does not support more then %d raid volumes.\n", + pr_vrb("platform does not support more than %d raid volumes.\n", super->orom->vphba); return 0; } @@ -5977,7 +6259,7 @@ static int validate_geometry_imsm_volume(struct supertype *st, int level, return 1; } -static int reserve_space(struct supertype *st, int raiddisks, +static int imsm_get_free_size(struct supertype *st, int raiddisks, unsigned long long size, int chunk, unsigned long long *freesize) { @@ -6029,7 +6311,7 @@ static int reserve_space(struct supertype *st, int raiddisks, (super->orom && used && used != raiddisks) || maxsize < minsize || maxsize == 0) { - 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 */ } @@ -6043,9 +6325,7 @@ static int reserve_space(struct supertype *st, int raiddisks, } if (!check_env("IMSM_NO_PLATFORM") && mpb->num_raid_devs > 0 && size && size != maxsize) { - fprintf(stderr, Name ": attempting to create a second " - "volume with size less then remaining space. " - "Aborting...\n"); + pr_err("attempting to create a second volume with size less then remaining space. Aborting...\n"); return 0; } cnt = 0; @@ -6055,11 +6335,35 @@ static int reserve_space(struct supertype *st, int raiddisks, *freesize = size; + dprintf("imsm: imsm_get_free_size() returns : %llu\n", size); + return 1; } +static int reserve_space(struct supertype *st, int raiddisks, + unsigned long long size, int chunk, + unsigned long long *freesize) +{ + struct intel_super *super = st->sb; + struct dl *dl; + int cnt; + int rv = 0; + + rv = imsm_get_free_size(st, raiddisks, size, chunk, freesize); + if (rv) { + cnt = 0; + for (dl = super->disks; dl; dl = dl->next) + if (dl->e) + dl->raiddisk = cnt++; + rv = 1; + } + + return rv; +} + static int validate_geometry_imsm(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) { @@ -6075,7 +6379,8 @@ static int validate_geometry_imsm(struct supertype *st, int level, int layout, /* Must be a fresh device to add to a container */ return validate_geometry_imsm_container(st, level, layout, raiddisks, - chunk?*chunk:0, size, + *chunk, + size, data_offset, dev, freesize, verbose); } @@ -6084,7 +6389,7 @@ static int validate_geometry_imsm(struct supertype *st, int level, int layout, if (st->sb) { struct intel_super *super = st->sb; if (!validate_geometry_imsm_orom(st->sb, level, layout, - raiddisks, chunk, + raiddisks, chunk, size, verbose)) return 0; /* we are being asked to automatically layout a @@ -6099,18 +6404,17 @@ static int validate_geometry_imsm(struct supertype *st, int level, int layout, created */ if (super->orom && freesize) { int count; - count = count_volumes(super->hba->path, + count = count_volumes(super->hba, super->orom->dpa, verbose); if (super->orom->vphba <= count) { - pr_vrb(": platform does not support more" - "then %d raid volumes.\n", + pr_vrb("platform does not support more than %d raid volumes.\n", super->orom->vphba); return 0; } } if (freesize) return reserve_space(st, raiddisks, size, - chunk?*chunk:0, freesize); + *chunk, freesize); } return 1; } @@ -6118,6 +6422,7 @@ static int validate_geometry_imsm(struct supertype *st, int level, int layout, /* creating in a given container */ return validate_geometry_imsm_volume(st, level, layout, raiddisks, chunk, size, + data_offset, dev, freesize, verbose); } @@ -6125,15 +6430,14 @@ static int validate_geometry_imsm(struct supertype *st, int level, int layout, fd = open(dev, O_RDONLY|O_EXCL, 0); if (fd >= 0) { if (verbose) - fprintf(stderr, - Name ": Cannot create this array on device %s\n", - dev); + pr_err("Cannot create this array on device %s\n", + dev); close(fd); return 0; } if (errno != EBUSY || (fd = open(dev, O_RDONLY, 0)) < 0) { if (verbose) - fprintf(stderr, Name ": Cannot open %s: %s\n", + pr_err("Cannot open %s: %s\n", dev, strerror(errno)); return 0; } @@ -6142,11 +6446,11 @@ static int validate_geometry_imsm(struct supertype *st, int level, int layout, close(fd); if (cfd < 0) { if (verbose) - fprintf(stderr, Name ": Cannot use %s: It is busy\n", + pr_err("Cannot use %s: It is busy\n", dev); return 0; } - sra = sysfs_read(cfd, 0, GET_VERSION); + sra = sysfs_read(cfd, NULL, GET_VERSION); if (sra && sra->array.major_version == -1 && strcmp(sra->text_version, "imsm") == 0) is_member = 1; @@ -6159,18 +6463,18 @@ static int validate_geometry_imsm(struct supertype *st, int level, int layout, if (load_super_imsm_all(st, cfd, (void **) &super, NULL, NULL, 1) == 0) { st->sb = super; - st->container_dev = fd2devnum(cfd); + strcpy(st->container_devnm, fd2devnm(cfd)); close(cfd); return validate_geometry_imsm_volume(st, level, layout, raiddisks, chunk, - size, dev, + size, data_offset, dev, freesize, 1) ? 1 : -1; } } if (verbose) - fprintf(stderr, Name ": failed container membership check\n"); + pr_err("failed container membership check\n"); close(cfd); return 0; @@ -6217,20 +6521,17 @@ static int kill_subarray_imsm(struct supertype *st) if (i < current_vol) continue; sprintf(subarray, "%u", i); - if (is_subarray_active(subarray, st->devname)) { - fprintf(stderr, - Name ": deleting subarray-%d would change the UUID of active subarray-%d, aborting\n", - current_vol, i); + if (is_subarray_active(subarray, st->devnm)) { + pr_err("deleting subarray-%d would change the UUID of active subarray-%d, aborting\n", + current_vol, i); return 2; } } if (st->update_tail) { - struct imsm_update_kill_array *u = malloc(sizeof(*u)); + struct imsm_update_kill_array *u = xmalloc(sizeof(*u)); - if (!u) - return 2; u->type = update_kill_array; u->dev_idx = current_vol; append_metadata_update(st, u, sizeof(*u)); @@ -6276,9 +6577,8 @@ static int update_subarray_imsm(struct supertype *st, char *subarray, char *ep; int vol; - if (is_subarray_active(subarray, st->devname)) { - fprintf(stderr, - Name ": Unable to update name of active subarray\n"); + if (is_subarray_active(subarray, st->devnm)) { + pr_err("Unable to update name of active subarray\n"); return 2; } @@ -6290,10 +6590,8 @@ static int update_subarray_imsm(struct supertype *st, char *subarray, return 2; if (st->update_tail) { - struct imsm_update_rename_array *u = malloc(sizeof(*u)); + struct imsm_update_rename_array *u = xmalloc(sizeof(*u)); - if (!u) - return 2; u->type = update_rename_array; u->dev_idx = vol; snprintf((char *) u->name, MAX_RAID_SERIAL_LEN, "%s", name); @@ -6393,7 +6691,7 @@ static void update_recovery_start(struct intel_super *super, * IMSM_ORD_REBUILD, so assume they are missing and the * disk_ord_tbl was not correctly updated */ - dprintf("%s: failed to locate out-of-sync disk\n", __func__); + dprintf("failed to locate out-of-sync disk\n"); return; } @@ -6427,18 +6725,15 @@ static struct mdinfo *container_content_imsm(struct supertype *st, char *subarra /* do not assemble arrays when not all attributes are supported */ if (imsm_check_attributes(mpb->attributes) == 0) { sb_errors = 1; - fprintf(stderr, Name ": Unsupported attributes in IMSM metadata." - "Arrays activation is blocked.\n"); + pr_err("Unsupported attributes in IMSM metadata.Arrays activation is blocked.\n"); } /* check for bad blocks */ if (imsm_bbm_log_size(super->anchor)) { - fprintf(stderr, Name ": BBM log found in IMSM metadata." - "Arrays activation is blocked.\n"); + pr_err("BBM log found in IMSM metadata.Arrays activation is blocked.\n"); sb_errors = 1; } - /* count spare devices, not used in maps */ for (d = super->disks; d; d = d->next) @@ -6469,8 +6764,7 @@ static struct mdinfo *container_content_imsm(struct supertype *st, char *subarra */ if (dev->vol.migr_state && (migr_type(dev) == MIGR_STATE_CHANGE)) { - fprintf(stderr, Name ": cannot assemble volume '%.16s':" - " unsupported migration in progress\n", + pr_err("cannot assemble volume '%.16s': unsupported migration in progress\n", dev->volume); continue; } @@ -6478,12 +6772,7 @@ static struct mdinfo *container_content_imsm(struct supertype *st, char *subarra * OROM/EFI */ - this = malloc(sizeof(*this)); - if (!this) { - fprintf(stderr, Name ": failed to allocate %zu bytes\n", - sizeof(*this)); - break; - } + this = xmalloc(sizeof(*this)); super->current_vol = i; getinfo_super_imsm_volume(st, this, NULL); @@ -6495,10 +6784,9 @@ static struct mdinfo *container_content_imsm(struct supertype *st, char *subarra get_imsm_raid_level(map), /* RAID level */ imsm_level_to_layout(get_imsm_raid_level(map)), map->num_members, /* raid disks */ - &chunk, + &chunk, join_u32(dev->size_low, dev->size_high), 1 /* verbose */)) { - fprintf(stderr, Name ": IMSM RAID geometry validation" - " failed. Array %s activation is blocked.\n", + pr_err("IMSM RAID geometry validation failed. Array %s activation is blocked.\n", dev->volume); this->array.state |= (1<volume); - info_d = this->devs; - while (info_d) { - struct mdinfo *d = info_d->next; - - free(info_d); - info_d = d; - } - free(this); - this = rest; - break; - } + info_d = xcalloc(1, sizeof(*info_d)); info_d->next = this->devs; this->devs = info_d; @@ -6585,8 +6859,8 @@ static struct mdinfo *container_content_imsm(struct supertype *st, char *subarra this->array.working_disks++; info_d->events = __le32_to_cpu(mpb->generation_num); - info_d->data_offset = __le32_to_cpu(map->pba_of_lba0); - info_d->component_size = __le32_to_cpu(map->blocks_per_member); + info_d->data_offset = pba_of_lba0(map); + info_d->component_size = blocks_per_member(map); } /* now that the disk list is up-to-date fixup recovery_start */ update_recovery_start(super, dev, this); @@ -6603,7 +6877,6 @@ static struct mdinfo *container_content_imsm(struct supertype *st, char *subarra return rest; } - static __u8 imsm_check_degraded(struct intel_super *super, struct imsm_dev *dev, int failed, int look_in_map) { @@ -6612,7 +6885,7 @@ static __u8 imsm_check_degraded(struct intel_super *super, struct imsm_dev *dev, map = get_imsm_map(dev, look_in_map); if (!failed) - return map->map_state == IMSM_T_STATE_UNINITIALIZED ? + return map->map_state == IMSM_T_STATE_UNINITIALIZED ? IMSM_T_STATE_UNINITIALIZED : IMSM_T_STATE_NORMAL; switch (get_imsm_raid_level(map)) { @@ -6642,7 +6915,7 @@ static __u8 imsm_check_degraded(struct intel_super *super, struct imsm_dev *dev, struct imsm_disk *disk; /* reset the potential in-sync count on even-numbered - * slots. num_copies is always 2 for imsm raid10 + * slots. num_copies is always 2 for imsm raid10 */ if ((i & 1) == 0) insync = 2; @@ -6735,8 +7008,7 @@ static int imsm_open_new(struct supertype *c, struct active_array *a, struct imsm_super *mpb = super->anchor; if (atoi(inst) >= mpb->num_raid_devs) { - fprintf(stderr, "%s: subarry index %d, out of range\n", - __func__, atoi(inst)); + pr_err("subarry index %d, out of range\n", atoi(inst)); return -ENODEV; } @@ -6833,6 +7105,12 @@ static void handle_missing(struct intel_super *super, struct imsm_dev *dev) if (!super->missing) return; + /* When orom adds replacement for missing disk it does + * not remove entry of missing disk, but just updates map with + * new added disk. So it is not enough just to test if there is + * any missing disk, we have to look if there are any failed disks + * in map to stop migration */ + dprintf("imsm: mark missing\n"); /* end process for initialization and rebuild only */ @@ -6843,14 +7121,16 @@ static void handle_missing(struct intel_super *super, struct imsm_dev *dev) failed = imsm_count_failed(super, dev, MAP_0); map_state = imsm_check_degraded(super, dev, failed, MAP_0); - end_migration(dev, super, map_state); + if (failed) + end_migration(dev, super, map_state); } for (dl = super->missing; dl; dl = dl->next) mark_missing(dev, &dl->disk, dl->index); super->updates_pending++; } -static unsigned long long imsm_set_array_size(struct imsm_dev *dev) +static unsigned long long imsm_set_array_size(struct imsm_dev *dev, + long long new_size) { int used_disks = imsm_num_data_members(dev, MAP_0); unsigned long long array_blocks; @@ -6869,8 +7149,17 @@ static unsigned long long imsm_set_array_size(struct imsm_dev *dev) /* set array size in metadata */ - map = get_imsm_map(dev, MAP_0); - array_blocks = map->blocks_per_member * used_disks; + if (new_size <= 0) { + /* OLCE size change is caused by added disks + */ + map = get_imsm_map(dev, MAP_0); + array_blocks = blocks_per_member(map) * used_disks; + } else { + /* Online Volume Size Change + * Using available free space + */ + array_blocks = new_size; + } /* round array size down to closest MB */ @@ -6927,7 +7216,8 @@ static void imsm_progress_container_reshape(struct intel_super *super) memcpy(map2, map, copy_map_size); map2->num_members = prev_num_members; - imsm_set_array_size(dev); + imsm_set_array_size(dev, -1); + super->clean_migration_record_by_mdmon = 1; super->updates_pending++; } } @@ -6984,7 +7274,7 @@ static int imsm_set_array_state(struct active_array *a, int consistent) used_disks = imsm_num_data_members(dev, MAP_0); if (used_disks > 0) { array_blocks = - map->blocks_per_member * + blocks_per_member(map) * used_disks; /* round array size down to closest MB */ @@ -7090,12 +7380,14 @@ static void imsm_set_disk(struct active_array *a, int n, int state) struct imsm_dev *dev = get_imsm_dev(super, inst); struct imsm_map *map = get_imsm_map(dev, MAP_0); struct imsm_disk *disk; + struct mdinfo *mdi; + int recovery_not_finished = 0; int failed; __u32 ord; __u8 map_state; if (n > map->num_members) - fprintf(stderr, "imsm: set_disk %d out of range 0..%d\n", + pr_err("imsm: set_disk %d out of range 0..%d\n", n, map->num_members - 1); if (n < 0) @@ -7129,7 +7421,22 @@ static void imsm_set_disk(struct active_array *a, int n, int state) case IMSM_T_STATE_NORMAL: /* transition to normal state */ dprintf("normal: "); if (is_rebuilding(dev)) { - dprintf("while rebuilding"); + dprintf_cont("while rebuilding"); + /* check if recovery is really finished */ + for (mdi = a->info.devs; mdi ; mdi = mdi->next) + if (mdi->recovery_start != MaxSector) { + recovery_not_finished = 1; + break; + } + if (recovery_not_finished) { + dprintf_cont("\n"); + dprintf("Rebuild has not finished yet, state not changed"); + if (a->last_checkpoint < mdi->recovery_start) { + a->last_checkpoint = mdi->recovery_start; + super->updates_pending++; + } + break; + } end_migration(dev, super, map_state); map = get_imsm_map(dev, MAP_0); map->failed_disk_num = ~0; @@ -7138,7 +7445,7 @@ static void imsm_set_disk(struct active_array *a, int n, int state) break; } if (is_gen_migration(dev)) { - dprintf("while general migration"); + dprintf_cont("while general migration"); if (a->last_checkpoint >= a->info.component_size) end_migration(dev, super, map_state); else @@ -7150,26 +7457,26 @@ static void imsm_set_disk(struct active_array *a, int n, int state) } break; case IMSM_T_STATE_DEGRADED: /* transition to degraded state */ - dprintf("degraded: "); + dprintf_cont("degraded: "); if ((map->map_state != map_state) && !dev->vol.migr_state) { - dprintf("mark degraded"); + dprintf_cont("mark degraded"); map->map_state = map_state; super->updates_pending++; a->last_checkpoint = 0; break; } if (is_rebuilding(dev)) { - dprintf("while rebuilding."); + dprintf_cont("while rebuilding."); if (map->map_state != map_state) { - dprintf(" Map state change"); + dprintf_cont(" Map state change"); end_migration(dev, super, map_state); super->updates_pending++; } break; } if (is_gen_migration(dev)) { - dprintf("while general migration"); + dprintf_cont("while general migration"); if (a->last_checkpoint >= a->info.component_size) end_migration(dev, super, map_state); else { @@ -7180,22 +7487,22 @@ static void imsm_set_disk(struct active_array *a, int n, int state) break; } if (is_initializing(dev)) { - dprintf("while initialization."); + dprintf_cont("while initialization."); map->map_state = map_state; super->updates_pending++; break; } break; case IMSM_T_STATE_FAILED: /* transition to failed state */ - dprintf("failed: "); + dprintf_cont("failed: "); if (is_gen_migration(dev)) { - dprintf("while general migration"); + dprintf_cont("while general migration"); map->map_state = map_state; super->updates_pending++; break; } if (map->map_state != map_state) { - dprintf("mark failed"); + dprintf_cont("mark failed"); end_migration(dev, super, map_state); super->updates_pending++; a->last_checkpoint = 0; @@ -7203,10 +7510,9 @@ static void imsm_set_disk(struct active_array *a, int n, int state) } break; default: - dprintf("state %i\n", map_state); + dprintf_cont("state %i\n", map_state); } - dprintf("\n"); - + dprintf_cont("\n"); } static int store_imsm_mpb(int fd, struct imsm_super *mpb) @@ -7268,7 +7574,7 @@ static struct dl *imsm_readd(struct intel_super *super, int idx, struct active_a dl = NULL; if (dl) - dprintf("%s: found %x:%x\n", __func__, dl->major, dl->minor); + dprintf("found %x:%x\n", dl->major, dl->minor); return dl; } @@ -7351,9 +7657,9 @@ static struct dl *imsm_add_spare(struct intel_super *super, int slot, found = 0; j = 0; pos = 0; - array_start = __le32_to_cpu(map->pba_of_lba0); + array_start = pba_of_lba0(map); array_end = array_start + - __le32_to_cpu(map->blocks_per_member) - 1; + blocks_per_member(map) - 1; do { /* check that we can start at pba_of_lba0 with @@ -7384,7 +7690,6 @@ static struct dl *imsm_add_spare(struct intel_super *super, int slot, return dl; } - static int imsm_rebuild_allowed(struct supertype *cont, int dev_idx, int failed) { struct imsm_dev *dev2; @@ -7470,8 +7775,7 @@ static struct mdinfo *imsm_activate_spare(struct active_array *a, /* Cannot activate another spare if rebuild is in progress already */ if (is_rebuilding(dev)) { - dprintf("imsm: No spare activation allowed. " - "Rebuild in progress already.\n"); + dprintf("imsm: No spare activation allowed. Rebuild in progress already.\n"); return NULL; } @@ -7485,14 +7789,18 @@ static struct mdinfo *imsm_activate_spare(struct active_array *a, IMSM_T_STATE_DEGRADED) return NULL; + if (get_imsm_map(dev, MAP_0)->map_state == IMSM_T_STATE_UNINITIALIZED) { + dprintf("imsm: No spare activation allowed. Volume is not initialized.\n"); + return NULL; + } + /* * If there are any failed disks check state of the other volume. * Block rebuild if the another one is failed until failed disks * are removed from container. */ if (failed) { - dprintf("found failed disks in %.*s, check if there another" - "failed sub-array.\n", + dprintf("found failed disks in %.*s, check if there anotherfailed sub-array.\n", MAX_RAID_SERIAL_LEN, dev->volume); /* check if states of the other volumes allow for rebuild */ for (i = 0; i < super->anchor->num_raid_devs; i++) { @@ -7528,12 +7836,9 @@ static struct mdinfo *imsm_activate_spare(struct active_array *a, dl = imsm_add_spare(super, i, a, 1, rv); if (!dl) continue; - + /* found a usable disk with enough space */ - di = malloc(sizeof(*di)); - if (!di) - continue; - memset(di, 0, sizeof(*di)); + di = xcalloc(1, sizeof(*di)); /* dl->index will be -1 in the case we are activating a * pristine spare. imsm_process_update() will create a @@ -7553,7 +7858,7 @@ static struct mdinfo *imsm_activate_spare(struct active_array *a, di->disk.minor = dl->minor; di->disk.state = 0; di->recovery_start = 0; - di->data_offset = __le32_to_cpu(map->pba_of_lba0); + di->data_offset = pba_of_lba0(map); di->component_size = a->info.component_size; di->container_member = inst; super->random = random32(); @@ -7571,24 +7876,9 @@ static struct mdinfo *imsm_activate_spare(struct active_array *a, * Create a metadata_update record to update the * disk_ord_tbl for the array */ - mu = malloc(sizeof(*mu)); - if (mu) { - mu->buf = malloc(sizeof(struct imsm_update_activate_spare) * num_spares); - if (mu->buf == NULL) { - free(mu); - mu = NULL; - } - } - if (!mu) { - while (rv) { - struct mdinfo *n = rv->next; - - free(rv); - rv = n; - } - return NULL; - } - + mu = xmalloc(sizeof(*mu)); + mu->buf = xcalloc(num_spares, + sizeof(struct imsm_update_activate_spare)); mu->space = NULL; mu->space_list = NULL; mu->len = sizeof(struct imsm_update_activate_spare) * num_spares; @@ -7630,10 +7920,10 @@ static int disks_overlap(struct intel_super *super, int idx, struct imsm_update_ return 0; } - static struct dl *get_disk_super(struct intel_super *super, int major, int minor) { - struct dl *dl = NULL; + struct dl *dl; + for (dl = super->disks; dl; dl = dl->next) if ((dl->major == major) && (dl->minor == minor)) return dl; @@ -7642,7 +7932,7 @@ static struct dl *get_disk_super(struct intel_super *super, int major, int minor static int remove_disk_super(struct intel_super *super, int major, int minor) { - struct dl *prev = NULL; + struct dl *prev; struct dl *dl; prev = NULL; @@ -7655,8 +7945,7 @@ static int remove_disk_super(struct intel_super *super, int major, int minor) super->disks = dl->next; dl->next = NULL; __free_imsm_disk(dl); - dprintf("%s: removed %x:%x\n", - __func__, major, minor); + dprintf("removed %x:%x\n", major, minor); break; } prev = dl; @@ -7669,7 +7958,8 @@ static void imsm_delete(struct intel_super *super, struct dl **dlp, unsigned ind static int add_remove_disk_update(struct intel_super *super) { int check_degraded = 0; - struct dl *disk = NULL; + struct dl *disk; + /* add/remove some spares to/from the metadata/contrainer */ while (super->disk_mgmt_list) { struct dl *disk_cfg; @@ -7682,9 +7972,8 @@ static int add_remove_disk_update(struct intel_super *super) disk_cfg->next = super->disks; super->disks = disk_cfg; check_degraded = 1; - dprintf("%s: added %x:%x\n", - __func__, disk_cfg->major, - disk_cfg->minor); + dprintf("added %x:%x\n", + disk_cfg->major, disk_cfg->minor); } else if (disk_cfg->action == DISK_REMOVE) { dprintf("Disk remove action processed: %x.%x\n", disk_cfg->major, disk_cfg->minor); @@ -7708,7 +7997,6 @@ static int add_remove_disk_update(struct intel_super *super) return check_degraded; } - static int apply_reshape_migration_update(struct imsm_update_reshape_migration *u, struct intel_super *super, void ***space_list) @@ -7717,7 +8005,7 @@ static int apply_reshape_migration_update(struct imsm_update_reshape_migration * void **tofree = NULL; int ret_val = 0; - dprintf("apply_reshape_migration_update()\n"); + dprintf("(enter)\n"); if ((u->subdev < 0) || (u->subdev > 1)) { dprintf("imsm: Error: Wrong subdev: %i\n", u->subdev); @@ -7795,8 +8083,7 @@ static int apply_reshape_migration_update(struct imsm_update_reshape_migration * new_disk = get_disk_super(super, major(u->new_disks[0]), minor(u->new_disks[0])); - dprintf("imsm: new disk for reshape is: %i:%i " - "(%p, index = %i)\n", + dprintf("imsm: new disk for reshape is: %i:%i (%p, index = %i)\n", major(u->new_disks[0]), minor(u->new_disks[0]), new_disk, new_disk->index); @@ -7816,7 +8103,7 @@ skip_disk_add: *tofree = *space_list; /* calculate new size */ - imsm_set_array_size(new_dev); + imsm_set_array_size(new_dev, -1); ret_val = 1; } @@ -7831,6 +8118,42 @@ error_disk_add: return ret_val; } +static int apply_size_change_update(struct imsm_update_size_change *u, + struct intel_super *super) +{ + struct intel_dev *id; + int ret_val = 0; + + dprintf("(enter)\n"); + if ((u->subdev < 0) || + (u->subdev > 1)) { + dprintf("imsm: Error: Wrong subdev: %i\n", u->subdev); + return ret_val; + } + + for (id = super->devlist ; id; id = id->next) { + if (id->index == (unsigned)u->subdev) { + struct imsm_dev *dev = get_imsm_dev(super, u->subdev); + struct imsm_map *map = get_imsm_map(dev, MAP_0); + int used_disks = imsm_num_data_members(dev, MAP_0); + unsigned long long blocks_per_member; + + /* calculate new size + */ + blocks_per_member = u->new_size / used_disks; + dprintf("(size: %llu, blocks per member: %llu)\n", + u->new_size, blocks_per_member); + set_blocks_per_member(map, blocks_per_member); + imsm_set_array_size(dev, u->new_size); + + ret_val = 1; + break; + } + } + + return ret_val; +} + static int apply_update_activate_spare(struct imsm_update_activate_spare *u, struct intel_super *super, struct active_array *active_array) @@ -7860,8 +8183,7 @@ static int apply_update_activate_spare(struct imsm_update_activate_spare *u, break; if (!dl) { - fprintf(stderr, "error: imsm_activate_spare passed " - "an unknown disk (index: %d)\n", + pr_err("error: imsm_activate_spare passed an unknown disk (index: %d)\n", u->dl->index); return 0; } @@ -7961,15 +8283,14 @@ static int apply_reshape_container_disks_update(struct imsm_update_reshape *u, int ret_val = 0; unsigned int dev_id; - dprintf("imsm: apply_reshape_container_disks_update()\n"); + dprintf("(enter)\n"); /* enable spares to use in array */ for (i = 0; i < delta_disks; i++) { new_disk = get_disk_super(super, major(u->new_disks[i]), minor(u->new_disks[i])); - dprintf("imsm: new disk for reshape is: %i:%i " - "(%p, index = %i)\n", + dprintf("imsm: new disk for reshape is: %i:%i (%p, index = %i)\n", major(u->new_disks[i]), minor(u->new_disks[i]), new_disk, new_disk->index); if ((new_disk == NULL) || @@ -8030,7 +8351,7 @@ static int apply_reshape_container_disks_update(struct imsm_update_reshape *u, newmap = get_imsm_map(newdev, MAP_1); memcpy(newmap, oldmap, sizeof_imsm_map(oldmap)); - imsm_set_array_size(newdev); + imsm_set_array_size(newdev, -1); } sp = (void **)id->dev; @@ -8166,9 +8487,9 @@ static void imsm_process_update(struct supertype *st, * the arrays for general migration and convert selected spares * into active devices. * update_activate_spare - a spare device has replaced a failed - * device in an array, update the disk_ord_tbl. If this disk is - * present in all member arrays then also clear the SPARE_DISK - * flag + * device in an array, update the disk_ord_tbl. If this disk is + * present in all member arrays then also clear the SPARE_DISK + * flag * update_create_array * update_kill_array * update_rename_array @@ -8202,8 +8523,7 @@ static void imsm_process_update(struct supertype *st, struct imsm_update_general_migration_checkpoint *u = (void *)update->buf; - dprintf("imsm: process_update() " - "for update_general_migration_checkpoint called\n"); + dprintf("called for update_general_migration_checkpoint\n"); /* find device under general migration */ for (id = super->devlist ; id; id = id->next) { @@ -8238,8 +8558,14 @@ static void imsm_process_update(struct supertype *st, super->updates_pending++; break; } + case update_size_change: { + struct imsm_update_size_change *u = (void *)update->buf; + if (apply_size_change_update(u, super)) + super->updates_pending++; + break; + } case update_activate_spare: { - struct imsm_update_activate_spare *u = (void *) update->buf; + struct imsm_update_activate_spare *u = (void *) update->buf; if (apply_update_activate_spare(u, super, st->arrays)) super->updates_pending++; break; @@ -8266,21 +8592,20 @@ static void imsm_process_update(struct supertype *st, /* handle racing creates: first come first serve */ if (u->dev_idx < mpb->num_raid_devs) { - dprintf("%s: subarray %d already defined\n", - __func__, u->dev_idx); + dprintf("subarray %d already defined\n", u->dev_idx); goto create_error; } /* check update is next in sequence */ if (u->dev_idx != mpb->num_raid_devs) { - dprintf("%s: can not create array %d expected index %d\n", - __func__, u->dev_idx, mpb->num_raid_devs); + dprintf("can not create array %d expected index %d\n", + u->dev_idx, mpb->num_raid_devs); goto create_error; } new_map = get_imsm_map(&u->dev, MAP_0); - new_start = __le32_to_cpu(new_map->pba_of_lba0); - new_end = new_start + __le32_to_cpu(new_map->blocks_per_member); + new_start = pba_of_lba0(new_map); + new_end = new_start + blocks_per_member(new_map); inf = get_disk_info(u); /* handle activate_spare versus create race: @@ -8290,8 +8615,8 @@ static void imsm_process_update(struct supertype *st, for (i = 0; i < mpb->num_raid_devs; i++) { dev = get_imsm_dev(super, i); map = get_imsm_map(dev, MAP_0); - start = __le32_to_cpu(map->pba_of_lba0); - end = start + __le32_to_cpu(map->blocks_per_member); + start = pba_of_lba0(map); + end = start + blocks_per_member(map); if ((new_start >= start && new_start <= end) || (start >= new_start && start <= new_end)) /* overlap */; @@ -8299,14 +8624,14 @@ static void imsm_process_update(struct supertype *st, continue; if (disks_overlap(super, i, u)) { - dprintf("%s: arrays overlap\n", __func__); + dprintf("arrays overlap\n"); goto create_error; } } /* check that prepare update was successful */ if (!update->space) { - dprintf("%s: prepare update failed\n", __func__); + dprintf("prepare update failed\n"); goto create_error; } @@ -8318,7 +8643,7 @@ static void imsm_process_update(struct supertype *st, for (i = 0; i < new_map->num_members; i++) { dl = serial_to_dl(inf[i].serial, super); if (!dl) { - dprintf("%s: disk disappeared\n", __func__); + dprintf("disk disappeared\n"); goto create_error; } } @@ -8424,7 +8749,7 @@ static void imsm_process_update(struct supertype *st, } case update_add_remove_disk: { /* we may be able to repair some arrays if disks are - * being added, check teh status of add_remove_disk + * being added, check the status of add_remove_disk * if discs has been added. */ if (add_remove_disk_update(super)) { @@ -8437,15 +8762,14 @@ static void imsm_process_update(struct supertype *st, break; } default: - fprintf(stderr, "error: unsuported process update type:" - "(type: %d)\n", type); + pr_err("error: unsuported process update type:(type: %d)\n", type); } } static struct mdinfo *get_spares_for_grow(struct supertype *st); -static void imsm_prepare_update(struct supertype *st, - struct metadata_update *update) +static int imsm_prepare_update(struct supertype *st, + struct metadata_update *update) { /** * Allocate space to hold new disk entries, raid-device entries or a new @@ -8454,19 +8778,27 @@ static void imsm_prepare_update(struct supertype *st, * integrated by the monitor thread without worrying about live pointers * in the manager thread. */ - enum imsm_update_type type = *(enum imsm_update_type *) update->buf; + enum imsm_update_type type; struct intel_super *super = st->sb; struct imsm_super *mpb = super->anchor; size_t buf_len; size_t len = 0; + if (update->len < (int)sizeof(type)) + return 0; + + type = *(enum imsm_update_type *) update->buf; + switch (type) { case update_general_migration_checkpoint: - dprintf("imsm: prepare_update() " - "for update_general_migration_checkpoint called\n"); + if (update->len < (int)sizeof(struct imsm_update_general_migration_checkpoint)) + return 0; + dprintf("called for update_general_migration_checkpoint\n"); break; case update_takeover: { struct imsm_update_takeover *u = (void *)update->buf; + if (update->len < (int)sizeof(*u)) + return 0; if (u->direction == R0_TO_R10) { void **tail = (void **)&update->space_list; struct imsm_dev *dev = get_imsm_dev(super, u->subarray); @@ -8474,15 +8806,10 @@ static void imsm_prepare_update(struct supertype *st, int num_members = map->num_members; void *space; int size, i; - int err = 0; /* allocate memory for added disks */ for (i = 0; i < num_members; i++) { size = sizeof(struct dl); - space = malloc(size); - if (!space) { - err++; - break; - } + space = xmalloc(size); *tail = space; tail = space; *tail = NULL; @@ -8490,24 +8817,11 @@ static void imsm_prepare_update(struct supertype *st, /* allocate memory for new device */ size = sizeof_imsm_dev(super->devlist->dev, 0) + (num_members * sizeof(__u32)); - space = malloc(size); - if (!space) - err++; - else { - *tail = space; - tail = space; - *tail = NULL; - } - if (!err) { - len = disks_to_mpb_size(num_members * 2); - } else { - /* if allocation didn't success, free buffer */ - while (update->space_list) { - void **sp = update->space_list; - update->space_list = *sp; - free(sp); - } - } + space = xmalloc(size); + *tail = space; + tail = space; + *tail = NULL; + len = disks_to_mpb_size(num_members * 2); } break; @@ -8525,7 +8839,10 @@ static void imsm_prepare_update(struct supertype *st, struct intel_dev *dl; void **space_tail = (void**)&update->space_list; - dprintf("imsm: imsm_prepare_update() for update_reshape\n"); + if (update->len < (int)sizeof(*u)) + return 0; + + dprintf("for update_reshape\n"); for (dl = super->devlist; dl; dl = dl->next) { int size = sizeof_imsm_dev(dl->dev, 1); @@ -8533,9 +8850,7 @@ static void imsm_prepare_update(struct supertype *st, if (u->new_raid_disks > u->old_raid_disks) size += sizeof(__u32)*2* (u->new_raid_disks - u->old_raid_disks); - s = malloc(size); - if (!s) - break; + s = xmalloc(size); *space_tail = s; space_tail = s; *space_tail = NULL; @@ -8559,7 +8874,10 @@ static void imsm_prepare_update(struct supertype *st, void *s; int current_level = -1; - dprintf("imsm: imsm_prepare_update() for update_reshape\n"); + if (update->len < (int)sizeof(*u)) + return 0; + + dprintf("for update_reshape\n"); /* add space for bigger array in update */ @@ -8569,9 +8887,7 @@ static void imsm_prepare_update(struct supertype *st, if (u->new_raid_disks > u->old_raid_disks) size += sizeof(__u32)*2* (u->new_raid_disks - u->old_raid_disks); - s = malloc(size); - if (!s) - break; + s = xmalloc(size); *space_tail = s; space_tail = s; *space_tail = NULL; @@ -8584,12 +8900,7 @@ static void imsm_prepare_update(struct supertype *st, /* add space for disk in update */ size = sizeof(struct dl); - s = malloc(size); - if (!s) { - free(update->space_list); - update->space_list = NULL; - break; - } + s = xmalloc(size); *space_tail = s; space_tail = s; *space_tail = NULL; @@ -8632,6 +8943,16 @@ static void imsm_prepare_update(struct supertype *st, dprintf("New anchor length is %llu\n", (unsigned long long)len); break; } + case update_size_change: { + if (update->len < (int)sizeof(struct imsm_update_size_change)) + return 0; + break; + } + case update_activate_spare: { + if (update->len < (int)sizeof(struct imsm_update_activate_spare)) + return 0; + break; + } case update_create_array: { struct imsm_update_create_array *u = (void *) update->buf; struct intel_dev *dv; @@ -8642,19 +8963,15 @@ static void imsm_prepare_update(struct supertype *st, int i; int activate = 0; + if (update->len < (int)sizeof(*u)) + return 0; + inf = get_disk_info(u); len = sizeof_imsm_dev(dev, 1); /* allocate a new super->devlist entry */ - dv = malloc(sizeof(*dv)); - if (dv) { - dv->dev = malloc(len); - if (dv->dev) - update->space = dv; - else { - free(dv); - update->space = NULL; - } - } + dv = xmalloc(sizeof(*dv)); + dv->dev = xmalloc(len); + update->space = dv; /* count how many spares will be converted to members */ for (i = 0; i < map->num_members; i++) { @@ -8670,9 +8987,22 @@ static void imsm_prepare_update(struct supertype *st, } len += activate * sizeof(struct imsm_disk); break; - default: + } + case update_kill_array: { + if (update->len < (int)sizeof(struct imsm_update_kill_array)) + return 0; break; } + case update_rename_array: { + if (update->len < (int)sizeof(struct imsm_update_rename_array)) + return 0; + break; + } + case update_add_remove_disk: + /* no update->len needed */ + break; + default: + return 0; } /* check if we need a larger metadata buffer */ @@ -8696,6 +9026,7 @@ static void imsm_prepare_update(struct supertype *st, else super->next_buf = NULL; } + return 1; } /* must be called while manager is quiesced */ @@ -8708,8 +9039,7 @@ static void imsm_delete(struct intel_super *super, struct dl **dlp, unsigned ind int i, j, num_members; __u32 ord; - dprintf("%s: deleting device[%d] from imsm_super\n", - __func__, index); + dprintf("deleting device[%d] from imsm_super\n", index); /* shift all indexes down one */ for (iter = super->disks; iter; iter = iter->next) @@ -8771,6 +9101,7 @@ static int imsm_get_allowed_degradation(int level, int raid_disks, struct imsm_dev *dev) { switch (level) { + case 1: case 10:{ int ret_val = 0; struct imsm_map *map; @@ -8818,7 +9149,6 @@ static int imsm_get_allowed_degradation(int level, int raid_disks, } } - /******************************************************************************* * Function: open_backup_targets * Description: Function opens file descriptors for all devices given in @@ -8859,7 +9189,7 @@ int open_backup_targets(struct mdinfo *info, int raid_disks, int *raid_fds, sd->disk.minor, 1); raid_fds[sd->disk.raid_disk] = dev_open(dn, O_RDWR); if (raid_fds[sd->disk.raid_disk] < 0) { - fprintf(stderr, "cannot open component\n"); + pr_err("cannot open component\n"); continue; } opened++; @@ -8870,13 +9200,90 @@ int open_backup_targets(struct mdinfo *info, int raid_disks, int *raid_fds, imsm_get_allowed_degradation(info->new_level, raid_disks, super, dev)) { - fprintf(stderr, "Not enough disks can be opened.\n"); + pr_err("Not enough disks can be opened.\n"); close_targets(raid_fds, raid_disks); return -2; } return 0; } +/******************************************************************************* + * Function: validate_container_imsm + * Description: This routine validates container after assemble, + * eg. if devices in container are under the same controller. + * + * Parameters: + * info : linked list with info about devices used in array + * Returns: + * 1 : HBA mismatch + * 0 : Success + ******************************************************************************/ +int validate_container_imsm(struct mdinfo *info) +{ + if (check_env("IMSM_NO_PLATFORM")) + return 0; + + struct sys_dev *idev; + struct sys_dev *hba = NULL; + struct sys_dev *intel_devices = find_intel_devices(); + char *dev_path = devt_to_devpath(makedev(info->disk.major, + info->disk.minor)); + + for (idev = intel_devices; idev; idev = idev->next) { + if (dev_path && strstr(dev_path, idev->path)) { + hba = idev; + break; + } + } + if (dev_path) + free(dev_path); + + if (!hba) { + pr_err("WARNING - Cannot detect HBA for device %s!\n", + devid2kname(makedev(info->disk.major, info->disk.minor))); + return 1; + } + + const struct imsm_orom *orom = get_orom_by_device_id(hba->dev_id); + struct mdinfo *dev; + + for (dev = info->next; dev; dev = dev->next) { + dev_path = devt_to_devpath(makedev(dev->disk.major, dev->disk.minor)); + + struct sys_dev *hba2 = NULL; + for (idev = intel_devices; idev; idev = idev->next) { + if (dev_path && strstr(dev_path, idev->path)) { + hba2 = idev; + break; + } + } + if (dev_path) + free(dev_path); + + const struct imsm_orom *orom2 = hba2 == NULL ? NULL : + get_orom_by_device_id(hba2->dev_id); + + if (hba2 && hba->type != hba2->type) { + pr_err("WARNING - HBAs of devices do not match %s != %s\n", + get_sys_dev_type(hba->type), get_sys_dev_type(hba2->type)); + return 1; + } + + if ((orom != orom2) || ((hba->type == SYS_DEV_VMD) && (hba != hba2))) { + pr_err("WARNING - IMSM container assembled with disks under different HBAs!\n" + " This operation is not supported and can lead to data loss.\n"); + return 1; + } + + if (!orom) { + pr_err("WARNING - IMSM container assembled with disks under HBAs without IMSM platform support!\n" + " This operation is not supported and can lead to data loss.\n"); + return 1; + } + } + + return 0; +} #ifndef MDASSEMBLE /******************************************************************************* * Function: init_migr_record_imsm @@ -8912,7 +9319,8 @@ void init_migr_record_imsm(struct supertype *st, struct imsm_dev *dev, migr_rec->dest_depth_per_unit = GEN_MIGR_AREA_SIZE / max(map_dest->blocks_per_strip, map_src->blocks_per_strip); - migr_rec->dest_depth_per_unit *= map_dest->blocks_per_strip; + migr_rec->dest_depth_per_unit *= + max(map_dest->blocks_per_strip, map_src->blocks_per_strip); new_data_disks = imsm_num_data_members(dev, MAP_0); migr_rec->blocks_per_unit = __cpu_to_le32(migr_rec->dest_depth_per_unit * new_data_disks); @@ -8929,7 +9337,6 @@ void init_migr_record_imsm(struct supertype *st, struct imsm_dev *dev, migr_rec->post_migr_vol_cap = dev->size_low; migr_rec->post_migr_vol_cap_hi = dev->size_high; - /* Find the smallest dev */ for (sd = info->devs ; sd ; sd = sd->next) { sprintf(nm, "%d:%d", sd->disk.major, sd->disk.minor); @@ -8974,8 +9381,8 @@ int save_backup_imsm(struct supertype *st, { int rv = -1; struct intel_super *super = st->sb; - unsigned long long *target_offsets = NULL; - int *targets = NULL; + unsigned long long *target_offsets; + int *targets; int i; struct imsm_map *map_dest = get_imsm_map(dev, MAP_0); int new_disks = map_dest->num_members; @@ -8984,16 +9391,12 @@ int save_backup_imsm(struct supertype *st, unsigned long long start; int data_disks = imsm_num_data_members(dev, MAP_0); - targets = malloc(new_disks * sizeof(int)); - if (!targets) - goto abort; + targets = xmalloc(new_disks * sizeof(int)); for (i = 0; i < new_disks; i++) targets[i] = -1; - target_offsets = malloc(new_disks * sizeof(unsigned long long)); - if (!target_offsets) - goto abort; + target_offsets = xcalloc(new_disks, sizeof(unsigned long long)); start = info->reshape_progress * 512; for (i = 0; i < new_disks; i++) { @@ -9024,7 +9427,7 @@ int save_backup_imsm(struct supertype *st, start, length, buf) != 0) { - fprintf(stderr, Name ": Error restoring stripes\n"); + pr_err("Error restoring stripes\n"); goto abort; } @@ -9060,8 +9463,7 @@ int save_checkpoint_imsm(struct supertype *st, struct mdinfo *info, int state) unsigned long long curr_migr_unit; if (load_imsm_migr_rec(super, info) != 0) { - dprintf("imsm: ERROR: Cannot read migration record " - "for checkpoint save.\n"); + dprintf("imsm: ERROR: Cannot read migration record for checkpoint save.\n"); return 1; } @@ -9085,8 +9487,7 @@ int save_checkpoint_imsm(struct supertype *st, struct mdinfo *info, int state) __cpu_to_le32(curr_migr_unit * __le32_to_cpu(super->migr_rec->dest_depth_per_unit)); if (write_imsm_migr_rec(st) < 0) { - dprintf("imsm: Cannot write migration record " - "outside backup area\n"); + dprintf("imsm: Cannot write migration record outside backup area\n"); return 1; } @@ -9108,7 +9509,7 @@ int recover_backup_imsm(struct supertype *st, struct mdinfo *info) { struct intel_super *super = st->sb; struct migr_record *migr_rec = super->migr_rec; - struct imsm_map *map_dest = NULL; + struct imsm_map *map_dest; struct intel_dev *id = NULL; unsigned long long read_offset; unsigned long long write_offset; @@ -9150,18 +9551,15 @@ int recover_backup_imsm(struct supertype *st, struct mdinfo *info) write_offset = ((unsigned long long) __le32_to_cpu(migr_rec->dest_1st_member_lba) + - __le32_to_cpu(map_dest->pba_of_lba0)) * 512; + pba_of_lba0(map_dest)) * 512; unit_len = __le32_to_cpu(migr_rec->dest_depth_per_unit) * 512; if (posix_memalign((void **)&buf, 512, unit_len) != 0) goto abort; - targets = malloc(new_disks * sizeof(int)); - if (!targets) - goto abort; + targets = xcalloc(new_disks, sizeof(int)); if (open_backup_targets(info, new_disks, targets, super, id->dev)) { - fprintf(stderr, - Name ": Cannot open some devices belonging to array.\n"); + pr_err("Cannot open some devices belonging to array.\n"); goto abort; } @@ -9171,30 +9569,26 @@ int recover_backup_imsm(struct supertype *st, struct mdinfo *info) continue; } if (lseek64(targets[i], read_offset, SEEK_SET) < 0) { - fprintf(stderr, - Name ": Cannot seek to block: %s\n", - strerror(errno)); + pr_err("Cannot seek to block: %s\n", + strerror(errno)); skipped_disks++; continue; } if ((unsigned)read(targets[i], buf, unit_len) != unit_len) { - fprintf(stderr, - Name ": Cannot read copy area block: %s\n", - strerror(errno)); + pr_err("Cannot read copy area block: %s\n", + strerror(errno)); skipped_disks++; continue; } if (lseek64(targets[i], write_offset, SEEK_SET) < 0) { - fprintf(stderr, - Name ": Cannot seek to block: %s\n", - strerror(errno)); + pr_err("Cannot seek to block: %s\n", + strerror(errno)); skipped_disks++; continue; } if ((unsigned)write(targets[i], buf, unit_len) != unit_len) { - fprintf(stderr, - Name ": Cannot restore block: %s\n", - strerror(errno)); + pr_err("Cannot restore block: %s\n", + strerror(errno)); skipped_disks++; continue; } @@ -9204,17 +9598,14 @@ int recover_backup_imsm(struct supertype *st, struct mdinfo *info) new_disks, super, id->dev)) { - fprintf(stderr, - Name ": Cannot restore data from backup." - " Too many failed disks\n"); + pr_err("Cannot restore data from backup. Too many failed disks\n"); goto abort; } if (save_checkpoint_imsm(st, info, UNIT_SRC_NORMAL)) { /* ignore error == 2, this can mean end of reshape here */ - dprintf("imsm: Cannot write checkpoint to " - "migration record (UNIT_SRC_NORMAL) during restart\n"); + dprintf("imsm: Cannot write checkpoint to migration record (UNIT_SRC_NORMAL) during restart\n"); } else retval = 0; @@ -9237,11 +9628,11 @@ static const char *imsm_get_disk_controller_domain(const char *path) char *drv=NULL; struct stat st; - strncpy(disk_path, disk_by_path, PATH_MAX - 1); + strcpy(disk_path, disk_by_path); strncat(disk_path, path, PATH_MAX - strlen(disk_path) - 1); if (stat(disk_path, &st) == 0) { struct sys_dev* hba; - char *path=NULL; + char *path; path = devt_to_devpath(st.st_rdev); if (path == NULL) @@ -9251,35 +9642,35 @@ static const char *imsm_get_disk_controller_domain(const char *path) drv = "isci"; else if (hba && hba->type == SYS_DEV_SATA) drv = "ahci"; - else + else drv = "unknown"; dprintf("path: %s hba: %s attached: %s\n", path, (hba) ? hba->path : "NULL", drv); free(path); - if (hba) - free_sys_dev(&hba); } return drv; } -static int imsm_find_array_minor_by_subdev(int subdev, int container, int *minor) +static char *imsm_find_array_devnm_by_subdev(int subdev, char *container) { + static char devnm[32]; char subdev_name[20]; struct mdstat_ent *mdstat; sprintf(subdev_name, "%d", subdev); mdstat = mdstat_by_subdev(subdev_name, container); if (!mdstat) - return -1; + return NULL; - *minor = mdstat->devnum; + strcpy(devnm, mdstat->devnm); free_mdstat(mdstat); - return 0; + return devnm; } static int imsm_reshape_is_allowed_on_container(struct supertype *st, struct geo_params *geo, - int *old_raid_disks) + int *old_raid_disks, + int direction) { /* currently we only support increasing the number of devices * for a container. This increases the number of device for each @@ -9289,24 +9680,25 @@ static int imsm_reshape_is_allowed_on_container(struct supertype *st, struct mdinfo *info, *member; int devices_that_can_grow = 0; - dprintf("imsm: imsm_reshape_is_allowed_on_container(ENTER): " - "st->devnum = (%i)\n", - st->devnum); + dprintf("imsm: imsm_reshape_is_allowed_on_container(ENTER): st->devnm = (%s)\n", st->devnm); - if (geo->size != -1 || + if (geo->size > 0 || geo->level != UnSet || geo->layout != UnSet || geo->chunksize != 0 || geo->raid_disks == UnSet) { - dprintf("imsm: Container operation is allowed for " - "raid disks number change only.\n"); + dprintf("imsm: Container operation is allowed for raid disks number change only.\n"); + return ret_val; + } + + if (direction == ROLLBACK_METADATA_CHANGES) { + dprintf("imsm: Metadata changes rollback is not supported for container operation.\n"); return ret_val; } info = container_content_imsm(st, NULL); for (member = info; member; member = member->next) { - int result; - int minor; + char *result; dprintf("imsm: checking device_num: %i\n", member->container_member); @@ -9315,8 +9707,7 @@ static int imsm_reshape_is_allowed_on_container(struct supertype *st, /* we work on container for Online Capacity Expansion * only so raid_disks has to grow */ - dprintf("imsm: for container operation raid disks " - "increase is required\n"); + dprintf("imsm: for container operation raid disks increase is required\n"); break; } @@ -9324,8 +9715,7 @@ static int imsm_reshape_is_allowed_on_container(struct supertype *st, (info->array.level != 5)) { /* we cannot use this container with other raid level */ - dprintf("imsm: for container operation wrong" - " raid level (%i) detected\n", + dprintf("imsm: for container operation wrong raid level (%i) detected\n", info->array.level); break; } else { @@ -9336,8 +9726,7 @@ static int imsm_reshape_is_allowed_on_container(struct supertype *st, if (!is_raid_level_supported(super->orom, member->array.level, geo->raid_disks)) { - dprintf("platform does not support raid%d with" - " %d disk%s\n", + dprintf("platform does not support raid%d with %d disk%s\n", info->array.level, geo->raid_disks, geo->raid_disks > 1 ? "s" : ""); @@ -9347,8 +9736,7 @@ static int imsm_reshape_is_allowed_on_container(struct supertype *st, */ if (info->component_size % (info->array.chunk_size/512)) { - dprintf("Component size is not aligned to " - "chunk size\n"); + dprintf("Component size is not aligned to chunk size\n"); break; } } @@ -9363,10 +9751,9 @@ static int imsm_reshape_is_allowed_on_container(struct supertype *st, * so they need to be assembled. We have already * checked that no recovery etc is happening. */ - result = imsm_find_array_minor_by_subdev(member->container_member, - st->container_dev, - &minor); - if (result < 0) { + result = imsm_find_array_devnm_by_subdev(member->container_member, + st->container_devnm); + if (result == NULL) { dprintf("imsm: cannot find array\n"); break; } @@ -9377,19 +9764,19 @@ static int imsm_reshape_is_allowed_on_container(struct supertype *st, ret_val = 1; if (ret_val) - dprintf("\tContainer operation allowed\n"); + dprintf("Container operation allowed\n"); else - dprintf("\tError: %i\n", ret_val); + dprintf("Error: %i\n", ret_val); return ret_val; } /* Function: get_spares_for_grow * Description: Allocates memory and creates list of spare devices - * avaliable in container. Checks if spare drive size is acceptable. + * avaliable in container. Checks if spare drive size is acceptable. * Parameters: Pointer to the supertype structure * Returns: Pointer to the list of spare devices (mdinfo structure) on success, - * NULL if fail + * NULL if fail */ static struct mdinfo *get_spares_for_grow(struct supertype *st) { @@ -9410,15 +9797,14 @@ static int imsm_create_metadata_update_for_reshape( { struct intel_super *super = st->sb; struct imsm_super *mpb = super->anchor; - int update_memory_size = 0; - struct imsm_update_reshape *u = NULL; - struct mdinfo *spares = NULL; + int update_memory_size; + struct imsm_update_reshape *u; + struct mdinfo *spares; int i; - int delta_disks = 0; + int delta_disks; struct mdinfo *dev; - dprintf("imsm_update_metadata_for_reshape(enter) raid_disks = %i\n", - geo->raid_disks); + dprintf("(enter) raid_disks = %i\n", geo->raid_disks); delta_disks = geo->raid_disks - old_raid_disks; @@ -9428,12 +9814,7 @@ static int imsm_create_metadata_update_for_reshape( /* now add space for spare disks that we need to add. */ update_memory_size += sizeof(u->new_disks[0]) * (delta_disks - 1); - u = calloc(1, update_memory_size); - if (u == NULL) { - dprintf("error: " - "cannot get memory for imsm_update_reshape update\n"); - return 0; - } + u = xcalloc(1, update_memory_size); u->type = update_reshape_container_disks; u->old_raid_disks = old_raid_disks; u->new_raid_disks = geo->raid_disks; @@ -9444,8 +9825,7 @@ static int imsm_create_metadata_update_for_reshape( if (spares == NULL || delta_disks > spares->array.spare_disks) { - fprintf(stderr, Name ": imsm: ERROR: Cannot get spare devices " - "for %s.\n", geo->dev_name); + pr_err("imsm: ERROR: Cannot get spare devices for %s.\n", geo->dev_name); i = -1; goto abort; } @@ -9477,16 +9857,46 @@ abort: dprintf("imsm: reshape update preparation :"); if (i == delta_disks) { - dprintf(" OK\n"); + dprintf_cont(" OK\n"); *updatep = u; return update_memory_size; } free(u); - dprintf(" Error\n"); + dprintf_cont(" Error\n"); return 0; } +/****************************************************************************** + * function: imsm_create_metadata_update_for_size_change() + * Creates update for IMSM array for array size change. + * + ******************************************************************************/ +static int imsm_create_metadata_update_for_size_change( + struct supertype *st, + struct geo_params *geo, + struct imsm_update_size_change **updatep) +{ + struct intel_super *super = st->sb; + int update_memory_size; + struct imsm_update_size_change *u; + + dprintf("(enter) New size = %llu\n", geo->size); + + /* size of all update data without anchor */ + update_memory_size = sizeof(struct imsm_update_size_change); + + u = xcalloc(1, update_memory_size); + u->type = update_size_change; + u->subdev = super->current_vol; + u->new_size = geo->size; + + dprintf("imsm: reshape update preparation : OK\n"); + *updatep = u; + + return update_memory_size; +} + /****************************************************************************** * function: imsm_create_metadata_update_for_migration() * Creates update for IMSM array. @@ -9498,23 +9908,17 @@ static int imsm_create_metadata_update_for_migration( struct imsm_update_reshape_migration **updatep) { struct intel_super *super = st->sb; - int update_memory_size = 0; - struct imsm_update_reshape_migration *u = NULL; + int update_memory_size; + struct imsm_update_reshape_migration *u; struct imsm_dev *dev; int previous_level = -1; - dprintf("imsm_create_metadata_update_for_migration(enter)" - " New Level = %i\n", geo->level); + dprintf("(enter) New Level = %i\n", geo->level); /* size of all update data without anchor */ update_memory_size = sizeof(struct imsm_update_reshape_migration); - u = calloc(1, update_memory_size); - if (u == NULL) { - dprintf("error: cannot get memory for " - "imsm_create_metadata_update_for_migration\n"); - return 0; - } + u = xcalloc(1, update_memory_size); u->type = update_reshape_migration; u->subdev = super->current_vol; u->new_level = geo->level; @@ -9534,8 +9938,7 @@ static int imsm_create_metadata_update_for_migration( if (geo->chunksize != current_chunk_size) { u->new_chunksize = geo->chunksize / 1024; - dprintf("imsm: " - "chunk size change from %i to %i\n", + dprintf("imsm: chunk size change from %i to %i\n", current_chunk_size, u->new_chunksize); } previous_level = map->raid_level; @@ -9550,8 +9953,7 @@ static int imsm_create_metadata_update_for_migration( free(u); sysfs_free(spares); update_memory_size = 0; - dprintf("error: cannot get spare device " - "for requested migration"); + dprintf("error: cannot get spare device for requested migration"); return 0; } sysfs_free(spares); @@ -9572,8 +9974,8 @@ static void imsm_update_metadata_locally(struct supertype *st, mu.space = NULL; mu.space_list = NULL; mu.next = NULL; - imsm_prepare_update(st, &mu); - imsm_process_update(st, &mu); + if (imsm_prepare_update(st, &mu)) + imsm_process_update(st, &mu); while (mu.space_list) { void **space = mu.space_list; @@ -9585,12 +9987,14 @@ static void imsm_update_metadata_locally(struct supertype *st, /*************************************************************************** * Function: imsm_analyze_change * Description: Function analyze change for single volume -* and validate if transition is supported -* Parameters: Geometry parameters, supertype structure +* and validate if transition is supported +* Parameters: Geometry parameters, supertype structure, +* metadata change direction (apply/rollback) * Returns: Operation type code on success, -1 if fail ****************************************************************************/ enum imsm_reshape_type imsm_analyze_change(struct supertype *st, - struct geo_params *geo) + struct geo_params *geo, + int direction) { struct mdinfo info; int change = -1; @@ -9600,6 +10004,13 @@ enum imsm_reshape_type imsm_analyze_change(struct supertype *st, int devNumChange = 0; /* imsm compatible layout value for array geometry verification */ int imsm_layout = -1; + int data_disks; + struct imsm_dev *dev; + struct intel_super *super; + unsigned long long current_size; + unsigned long long free_size; + unsigned long long max_size; + int rv; getinfo_super_imsm_volume(st, &info, NULL); if ((geo->level != info.array.level) && @@ -9610,10 +10021,7 @@ enum imsm_reshape_type imsm_analyze_change(struct supertype *st, if (geo->level == 5) { change = CH_MIGRATION; if (geo->layout != ALGORITHM_LEFT_ASYMMETRIC) { - fprintf(stderr, - Name " Error. Requested Layout " - "not supported (left-asymmetric layout " - "is supported only)!\n"); + pr_err("Error. Requested Layout not supported (left-asymmetric layout is supported only)!\n"); change = -1; goto analyse_change_exit; } @@ -9638,10 +10046,8 @@ enum imsm_reshape_type imsm_analyze_change(struct supertype *st, break; } if (change == -1) { - fprintf(stderr, - Name " Error. Level Migration from %d to %d " - "not supported!\n", - info.array.level, geo->level); + pr_err("Error. Level Migration from %d to %d not supported!\n", + info.array.level, geo->level); goto analyse_change_exit; } } else @@ -9661,10 +10067,8 @@ enum imsm_reshape_type imsm_analyze_change(struct supertype *st, geo->layout = 0; geo->level = 5; } else { - fprintf(stderr, - Name " Error. Layout Migration from %d to %d " - "not supported!\n", - info.array.layout, geo->layout); + pr_err("Error. Layout Migration from %d to %d not supported!\n", + info.array.layout, geo->layout); change = -1; goto analyse_change_exit; } @@ -9681,12 +10085,98 @@ enum imsm_reshape_type imsm_analyze_change(struct supertype *st, geo->chunksize = info.array.chunk_size; chunk = geo->chunksize / 1024; + + super = st->sb; + dev = get_imsm_dev(super, super->current_vol); + data_disks = imsm_num_data_members(dev , MAP_0); + /* compute current size per disk member + */ + current_size = info.custom_array_size / data_disks; + + if ((geo->size > 0) && (geo->size != MAX_SIZE)) { + /* align component size + */ + geo->size = imsm_component_size_aligment_check( + get_imsm_raid_level(dev->vol.map), + chunk * 1024, + geo->size * 2); + if (geo->size == 0) { + pr_err("Error. Size expansion is supported only (current size is %llu, requested size /rounded/ is 0).\n", + current_size); + goto analyse_change_exit; + } + } + + if ((current_size != geo->size) && (geo->size > 0)) { + if (change != -1) { + pr_err("Error. Size change should be the only one at a time.\n"); + change = -1; + goto analyse_change_exit; + } + if ((super->current_vol + 1) != super->anchor->num_raid_devs) { + pr_err("Error. The last volume in container can be expanded only (%i/%s).\n", + super->current_vol, st->devnm); + goto analyse_change_exit; + } + /* check the maximum available size + */ + rv = imsm_get_free_size(st, dev->vol.map->num_members, + 0, chunk, &free_size); + if (rv == 0) + /* Cannot find maximum available space + */ + max_size = 0; + else { + max_size = free_size + current_size; + /* align component size + */ + max_size = imsm_component_size_aligment_check( + get_imsm_raid_level(dev->vol.map), + chunk * 1024, + max_size); + } + if (geo->size == MAX_SIZE) { + /* requested size change to the maximum available size + */ + if (max_size == 0) { + pr_err("Error. Cannot find maximum available space.\n"); + change = -1; + goto analyse_change_exit; + } else + geo->size = max_size; + } + + if ((direction == ROLLBACK_METADATA_CHANGES)) { + /* accept size for rollback only + */ + } else { + /* round size due to metadata compatibility + */ + geo->size = (geo->size >> SECT_PER_MB_SHIFT) + << SECT_PER_MB_SHIFT; + dprintf("Prepare update for size change to %llu\n", + geo->size ); + if (current_size >= geo->size) { + pr_err("Error. Size expansion is supported only (current size is %llu, requested size /rounded/ is %llu).\n", + current_size, geo->size); + goto analyse_change_exit; + } + if (max_size && geo->size > max_size) { + pr_err("Error. Requested size is larger than maximum available size (maximum available size is %llu, requested size /rounded/ is %llu).\n", + max_size, geo->size); + goto analyse_change_exit; + } + } + geo->size *= data_disks; + geo->raid_disks = dev->vol.map->num_members; + change = CH_ARRAY_SIZE; + } if (!validate_geometry_imsm(st, geo->level, imsm_layout, geo->raid_disks + devNumChange, &chunk, - geo->size, + geo->size, INVALID_SECTORS, 0, 0, 1)) change = -1; @@ -9695,17 +10185,18 @@ enum imsm_reshape_type imsm_analyze_change(struct supertype *st, struct imsm_super *mpb = super->anchor; if (mpb->num_raid_devs > 1) { - fprintf(stderr, - Name " Error. Cannot perform operation on %s" - "- for this operation it MUST be single " - "array in container\n", - geo->dev_name); + pr_err("Error. Cannot perform operation on %s- for this operation it MUST be single array in container\n", + geo->dev_name); change = -1; } } analyse_change_exit: - + if ((direction == ROLLBACK_METADATA_CHANGES) && + ((change == CH_MIGRATION) || (change == CH_TAKEOVER))) { + dprintf("imsm: Metadata changes rollback is not supported for migration and takeover operations.\n"); + change = -1; + } return change; } @@ -9714,9 +10205,7 @@ int imsm_takeover(struct supertype *st, struct geo_params *geo) struct intel_super *super = st->sb; struct imsm_update_takeover *u; - u = malloc(sizeof(struct imsm_update_takeover)); - if (u == NULL) - return 1; + u = xmalloc(sizeof(struct imsm_update_takeover)); u->type = update_takeover; u->subarray = super->current_vol; @@ -9742,20 +10231,21 @@ int imsm_takeover(struct supertype *st, struct geo_params *geo) return 0; } -static int imsm_reshape_super(struct supertype *st, long long size, int level, +static int imsm_reshape_super(struct supertype *st, unsigned long long size, + int level, int layout, int chunksize, int raid_disks, int delta_disks, char *backup, char *dev, - int verbose) + int direction, int verbose) { int ret_val = 1; struct geo_params geo; - dprintf("imsm: reshape_super called.\n"); + dprintf("(enter)\n"); memset(&geo, 0, sizeof(struct geo_params)); geo.dev_name = dev; - geo.dev_id = st->devnum; + strcpy(geo.devnm, st->devnm); geo.size = size; geo.level = level; geo.layout = layout; @@ -9764,19 +10254,19 @@ static int imsm_reshape_super(struct supertype *st, long long size, int level, if (delta_disks != UnSet) geo.raid_disks += delta_disks; - dprintf("\tfor level : %i\n", geo.level); - dprintf("\tfor raid_disks : %i\n", geo.raid_disks); + dprintf("for level : %i\n", geo.level); + dprintf("for raid_disks : %i\n", geo.raid_disks); if (experimental() == 0) return ret_val; - if (st->container_dev == st->devnum) { + if (strcmp(st->container_devnm, st->devnm) == 0) { /* On container level we can only increase number of devices. */ dprintf("imsm: info: Container operation\n"); int old_raid_disks = 0; if (imsm_reshape_is_allowed_on_container( - st, &geo, &old_raid_disks)) { + st, &geo, &old_raid_disks, direction)) { struct imsm_update_reshape *u = NULL; int len; @@ -9798,8 +10288,7 @@ static int imsm_reshape_super(struct supertype *st, long long size, int level, free(u); } else { - fprintf(stderr, Name ": (imsm) Operation " - "is not allowed on this container\n"); + pr_err("(imsm) Operation is not allowed on this container\n"); } } else { /* On volume level we support following operations @@ -9809,23 +10298,24 @@ static int imsm_reshape_super(struct supertype *st, long long size, int level, */ struct intel_super *super = st->sb; struct intel_dev *dev = super->devlist; - int change, devnum; + int change; dprintf("imsm: info: Volume operation\n"); /* find requested device */ while (dev) { - if (imsm_find_array_minor_by_subdev( - dev->index, st->container_dev, &devnum) == 0 - && devnum == geo.dev_id) + char *devnm = + imsm_find_array_devnm_by_subdev( + dev->index, st->container_devnm); + if (devnm && strcmp(devnm, geo.devnm) == 0) break; dev = dev->next; } if (dev == NULL) { - fprintf(stderr, Name " Cannot find %s (%i) subarray\n", - geo.dev_name, geo.dev_id); + pr_err("Cannot find %s (%s) subarray\n", + geo.dev_name, geo.devnm); goto exit_imsm_reshape_super; } super->current_vol = dev->index; - change = imsm_analyze_change(st, &geo); + change = imsm_analyze_change(st, &geo, direction); switch (change) { case CH_TAKEOVER: ret_val = imsm_takeover(st, &geo); @@ -9836,8 +10326,26 @@ static int imsm_reshape_super(struct supertype *st, long long size, int level, imsm_create_metadata_update_for_migration( st, &geo, &u); if (len < 1) { - dprintf("imsm: " - "Cannot prepare update\n"); + dprintf("imsm: Cannot prepare update\n"); + break; + } + ret_val = 0; + /* update metadata locally */ + imsm_update_metadata_locally(st, u, len); + /* and possibly remotely */ + if (st->update_tail) + append_metadata_update(st, u, len); + else + free(u); + } + break; + case CH_ARRAY_SIZE: { + struct imsm_update_size_change *u = NULL; + int len = + imsm_create_metadata_update_for_size_change( + st, &geo, &u); + if (len < 1) { + dprintf("imsm: Cannot prepare update\n"); break; } ret_val = 0; @@ -9860,6 +10368,33 @@ exit_imsm_reshape_super: return ret_val; } +#define COMPLETED_OK 0 +#define COMPLETED_NONE 1 +#define COMPLETED_DELAYED 2 + +static int read_completed(int fd, unsigned long long *val) +{ + int ret; + char buf[50]; + + ret = sysfs_fd_get_str(fd, buf, 50); + if (ret < 0) + return ret; + + ret = COMPLETED_OK; + if (strncmp(buf, "none", 4) == 0) { + ret = COMPLETED_NONE; + } else if (strncmp(buf, "delayed", 7) == 0) { + ret = COMPLETED_DELAYED; + } else { + char *ep; + *val = strtoull(buf, &ep, 0); + if (ep == buf || (*ep != 0 && *ep != '\n' && *ep != ' ')) + ret = -1; + } + return ret; +} + /******************************************************************************* * Function: wait_for_reshape_imsm * Description: Function writes new sync_max value and waits until @@ -9874,62 +10409,71 @@ exit_imsm_reshape_super: ******************************************************************************/ int wait_for_reshape_imsm(struct mdinfo *sra, int ndata) { - int fd = sysfs_get_fd(sra, NULL, "reshape_position"); + int fd = sysfs_get_fd(sra, NULL, "sync_completed"); + int retry = 3; unsigned long long completed; /* to_complete : new sync_max position */ unsigned long long to_complete = sra->reshape_progress; unsigned long long position_to_set = to_complete / ndata; if (fd < 0) { - dprintf("imsm: wait_for_reshape_imsm() " - "cannot open reshape_position\n"); + dprintf("cannot open reshape_position\n"); return 1; } - if (sysfs_fd_get_ll(fd, &completed) < 0) { - dprintf("imsm: wait_for_reshape_imsm() " - "cannot read reshape_position (no reshape in progres)\n"); - close(fd); - return 0; - } + do { + if (sysfs_fd_get_ll(fd, &completed) < 0) { + if (!retry) { + dprintf("cannot read reshape_position (no reshape in progres)\n"); + close(fd); + return 1; + } + usleep(30000); + } else + break; + } while (retry--); - if (completed > to_complete) { - dprintf("imsm: wait_for_reshape_imsm() " - "wrong next position to set %llu (%llu)\n", - to_complete, completed); + if (completed > position_to_set) { + dprintf("wrong next position to set %llu (%llu)\n", + to_complete, position_to_set); close(fd); return -1; } dprintf("Position set: %llu\n", position_to_set); if (sysfs_set_num(sra, NULL, "sync_max", position_to_set) != 0) { - dprintf("imsm: wait_for_reshape_imsm() " - "cannot set reshape position to %llu\n", + dprintf("cannot set reshape position to %llu\n", position_to_set); close(fd); return -1; } do { + int rc; char action[20]; - fd_set rfds; - FD_ZERO(&rfds); - FD_SET(fd, &rfds); - select(fd+1, &rfds, NULL, NULL, NULL); + int timeout = 3000; + + sysfs_wait(fd, &timeout); if (sysfs_get_str(sra, NULL, "sync_action", action, 20) > 0 && - strncmp(action, "reshape", 7) != 0) - break; - if (sysfs_fd_get_ll(fd, &completed) < 0) { - dprintf("imsm: wait_for_reshape_imsm() " - "cannot read reshape_position (in loop)\n"); + strncmp(action, "reshape", 7) != 0) { + if (strncmp(action, "idle", 4) == 0) + break; close(fd); - return 1; + return -1; } - } while (completed < to_complete); + + rc = read_completed(fd, &completed); + if (rc < 0) { + dprintf("cannot read reshape_position (in loop)\n"); + close(fd); + return 1; + } else if (rc == COMPLETED_NONE) + break; + } while (completed < position_to_set); + close(fd); return 0; - } /******************************************************************************* @@ -9947,8 +10491,10 @@ int check_degradation_change(struct mdinfo *info, int degraded) { unsigned long long new_degraded; - sysfs_get_ll(info, NULL, "degraded", &new_degraded); - if (new_degraded != (unsigned long long)degraded) { + int rv; + + rv = sysfs_get_ll(info, NULL, "degraded", &new_degraded); + if ((rv == -1) || (new_degraded != (unsigned long long)degraded)) { /* check each device to ensure it is still working */ struct mdinfo *sd; new_degraded = 0; @@ -9983,7 +10529,7 @@ int check_degradation_change(struct mdinfo *info, * Function: imsm_manage_reshape * Description: Function finds array under reshape and it manages reshape * process. It creates stripes backups (if required) and sets - * checheckpoits. + * checkpoints. * Parameters: * afd : Backup handle (nattive) - not used * sra : general array info @@ -10007,7 +10553,7 @@ static int imsm_manage_reshape( { int ret_val = 0; struct intel_super *super = st->sb; - struct intel_dev *dv = NULL; + struct intel_dev *dv; struct imsm_dev *dev = NULL; struct imsm_map *map_src; int migr_vol_qan = 0; @@ -10025,7 +10571,10 @@ static int imsm_manage_reshape( int degraded = 0; int source_layout = 0; - if (!fds || !offsets || !sra) + if (!sra) + return ret_val; + + if (!fds || !offsets) goto abort; /* Find volume during the reshape */ @@ -10038,7 +10587,7 @@ static int imsm_manage_reshape( } /* Only one volume can migrate at the same time */ if (migr_vol_qan != 1) { - fprintf(stderr, Name " : %s", migr_vol_qan ? + pr_err("%s", migr_vol_qan ? "Number of migrating volumes greater than 1\n" : "There is no volume during migrationg\n"); goto abort; @@ -10061,8 +10610,17 @@ static int imsm_manage_reshape( init_migr_record_imsm(st, dev, sra); else { if (__le32_to_cpu(migr_rec->rec_status) != UNIT_SRC_NORMAL) { - dprintf("imsm: cannot restart migration when data " - "are present in copy area.\n"); + dprintf("imsm: cannot restart migration when data are present in copy area.\n"); + goto abort; + } + /* Save checkpoint to update migration record for current + * reshape position (in md). It can be farther than current + * reshape position in metadata. + */ + if (save_checkpoint_imsm(st, sra, UNIT_SRC_NORMAL) == 1) { + /* ignore error == 2, this can mean end of reshape here + */ + dprintf("imsm: Cannot write checkpoint to migration record (UNIT_SRC_NORMAL, initial save)\n"); goto abort; } } @@ -10093,8 +10651,7 @@ static int imsm_manage_reshape( */ degraded = check_degradation_change(sra, fds, degraded); if (degraded > 1) { - dprintf("imsm: Abort reshape due to degradation" - " level (%i)\n", degraded); + dprintf("imsm: Abort reshape due to degradation level (%i)\n", degraded); goto abort; } @@ -10105,7 +10662,7 @@ static int imsm_manage_reshape( start = current_position * 512; - /* allign reading start to old geometry */ + /* align reading start to old geometry */ start_buf_shift = start % old_data_stripe_length; start_src = start - start_buf_shift; @@ -10119,7 +10676,7 @@ static int imsm_manage_reshape( * to backup alligned to source array * [bytes] */ - unsigned long long next_step_filler = 0; + unsigned long long next_step_filler; unsigned long long copy_length = next_step * 512; /* allign copy area length to stripe in old geometry */ @@ -10128,10 +10685,7 @@ static int imsm_manage_reshape( if (next_step_filler) next_step_filler = (old_data_stripe_length - next_step_filler); - dprintf("save_stripes() parameters: start = %llu," - "\tstart_src = %llu,\tnext_step*512 = %llu," - "\tstart_in_buf_shift = %llu," - "\tnext_step_filler = %llu\n", + dprintf("save_stripes() parameters: start = %llu,\tstart_src = %llu,\tnext_step*512 = %llu,\tstart_in_buf_shift = %llu,\tnext_step_filler = %llu\n", start, start_src, copy_length, start_buf_shift, next_step_filler); @@ -10141,8 +10695,7 @@ static int imsm_manage_reshape( copy_length + next_step_filler + start_buf_shift, buf)) { - dprintf("imsm: Cannot save stripes" - " to buffer\n"); + dprintf("imsm: Cannot save stripes to buffer\n"); goto abort; } /* Convert data to destination format and store it @@ -10150,14 +10703,12 @@ static int imsm_manage_reshape( */ if (save_backup_imsm(st, dev, sra, buf + start_buf_shift, copy_length)) { - dprintf("imsm: Cannot save stripes to " - "target devices\n"); + dprintf("imsm: Cannot save stripes to target devices\n"); goto abort; } if (save_checkpoint_imsm(st, sra, UNIT_SRC_IN_CP_AREA)) { - dprintf("imsm: Cannot write checkpoint to " - "migration record (UNIT_SRC_IN_CP_AREA)\n"); + dprintf("imsm: Cannot write checkpoint to migration record (UNIT_SRC_IN_CP_AREA)\n"); goto abort; } } else { @@ -10178,29 +10729,52 @@ static int imsm_manage_reshape( sra->reshape_progress = next_step; /* wait until reshape finish */ - if (wait_for_reshape_imsm(sra, ndata) < 0) { + if (wait_for_reshape_imsm(sra, ndata)) { dprintf("wait_for_reshape_imsm returned error!\n"); goto abort; } + if (sigterm) + goto abort; if (save_checkpoint_imsm(st, sra, UNIT_SRC_NORMAL) == 1) { /* ignore error == 2, this can mean end of reshape here */ - dprintf("imsm: Cannot write checkpoint to " - "migration record (UNIT_SRC_NORMAL)\n"); + dprintf("imsm: Cannot write checkpoint to migration record (UNIT_SRC_NORMAL)\n"); goto abort; } } + /* clear migr_rec on disks after successful migration */ + struct dl *d; + + memset(super->migr_rec_buf, 0, MIGR_REC_BUF_SIZE); + for (d = super->disks; d; d = d->next) { + if (d->index < 0 || is_failed(&d->disk)) + continue; + unsigned long long dsize; + + get_dev_size(d->fd, NULL, &dsize); + if (lseek64(d->fd, dsize - MIGR_REC_POSITION, + SEEK_SET) >= 0) { + if (write(d->fd, super->migr_rec_buf, + MIGR_REC_BUF_SIZE) != MIGR_REC_BUF_SIZE) + perror("Write migr_rec failed"); + } + } + /* return '1' if done */ ret_val = 1; abort: free(buf); - abort_reshape(sra); + /* See Grow.c: abort_reshape() for further explanation */ + sysfs_set_num(sra, NULL, "suspend_lo", 0x7FFFFFFFFFFFFFFFULL); + sysfs_set_num(sra, NULL, "suspend_hi", 0); + sysfs_set_num(sra, NULL, "suspend_lo", 0); return ret_val; } + #endif /* MDASSEMBLE */ struct superswitch super_imsm = { @@ -10216,6 +10790,7 @@ struct superswitch super_imsm = { .add_to_super = add_to_super_imsm, .remove_from_super = remove_from_super_imsm, .detail_platform = detail_platform_imsm, + .export_detail_platform = export_detail_platform_imsm, .kill_subarray = kill_subarray_imsm, .update_subarray = update_subarray_imsm, .load_container = load_container_imsm, @@ -10224,6 +10799,7 @@ struct superswitch super_imsm = { .reshape_super = imsm_reshape_super, .manage_reshape = imsm_manage_reshape, .recover_backup = recover_backup_imsm, + .copy_metadata = copy_metadata_imsm, #endif .match_home = match_home_imsm, .uuid_from_super= uuid_from_super_imsm, @@ -10242,7 +10818,7 @@ struct superswitch super_imsm = { .free_super = free_super_imsm, .match_metadata_desc = match_metadata_desc_imsm, .container_content = container_content_imsm, - + .validate_container = validate_container_imsm, .external = 1, .name = "imsm",