From 691c6ee1b6bb77bc44a5474d856771b0aec9882d Mon Sep 17 00:00:00 2001 From: NeilBrown Date: Thu, 29 Apr 2010 16:09:59 +1000 Subject: [PATCH 1/4] IMSM/DDF: don't recognised these metadata on partitions. These metadata are not expected on partitions, and they have no way of differentiation whether which is correct if they are found both on the device and on the last partition. So if the device is a partition, refuse to read the metadata. Signed-off-by: NeilBrown --- mdadm.h | 1 + super-ddf.c | 4 ++++ super-intel.c | 4 ++++ util.c | 25 +++++++++++++++++++++++++ 4 files changed, 34 insertions(+) diff --git a/mdadm.h b/mdadm.h index 1bf5ac0..d9d17b0 100644 --- a/mdadm.h +++ b/mdadm.h @@ -876,6 +876,7 @@ extern int enough(int level, int raid_disks, int layout, int clean, extern int ask(char *mesg); extern unsigned long long get_component_size(int fd); extern void remove_partitions(int fd); +extern int test_partition(int fd); extern unsigned long long calc_array_size(int level, int raid_disks, int layout, int chunksize, unsigned long long devsize); extern int flush_metadata_updates(struct supertype *st); diff --git a/super-ddf.c b/super-ddf.c index 3feea57..0e6f1e5 100644 --- a/super-ddf.c +++ b/super-ddf.c @@ -783,6 +783,10 @@ static int load_super_ddf(struct supertype *st, int fd, if (get_dev_size(fd, devname, &dsize) == 0) return 1; + if (test_partition(fd)) + /* DDF is not allowed on partitions */ + return 1; + /* 32M is a lower bound */ if (dsize <= 32*1024*1024) { if (devname) diff --git a/super-intel.c b/super-intel.c index 677396c..bdd7a96 100644 --- a/super-intel.c +++ b/super-intel.c @@ -2798,6 +2798,10 @@ static int load_super_imsm(struct supertype *st, int fd, char *devname) return 0; #endif + if (test_partition(fd)) + /* IMSM not allowed on partitions */ + return 1; + free_super_imsm(st); super = alloc_super(); diff --git a/util.c b/util.c index 79d2b0f..25f1e56 100644 --- a/util.c +++ b/util.c @@ -302,6 +302,31 @@ void remove_partitions(int fd) #endif } +int test_partition(int fd) +{ + /* Check if fd is a whole-disk or a partition. + * BLKPG will return EINVAL on a partition, and BLKPG_DEL_PARTITION + * will return ENXIO on an invalid partition number. + */ + struct blkpg_ioctl_arg a; + struct blkpg_partition p; + a.op = BLKPG_DEL_PARTITION; + a.data = (void*)&p; + a.datalen = sizeof(p); + a.flags = 0; + memset(a.data, 0, a.datalen); + p.pno = 1<<30; + if (ioctl(fd, BLKPG, &a) == 0) + /* Very unlikely, but not a partition */ + return 0; + if (errno == ENXIO) + /* not a partition */ + return 0; + + return 1; +} + + int enough(int level, int raid_disks, int layout, int clean, char *avail, int avail_disks) { From 200871adf9e15d5ad985f28c349fd89c386ef48a Mon Sep 17 00:00:00 2001 From: NeilBrown Date: Thu, 29 Apr 2010 16:14:30 +1000 Subject: [PATCH 2/4] Grow: avoid overflow of chunk sizes. Chunks aren't particularly big, but when you could them in bytes and multiply them together (as we do for calculating the backup size for 'grow') they can overflow a 32bit int. So group the division by 512 more closely with the chunk size so were would need 30Meg chunks to come close to overflowing 32bits. Signed-off-by: NeilBrown --- Grow.c | 48 ++++++++++++++++++++++++------------------------ 1 file changed, 24 insertions(+), 24 deletions(-) diff --git a/Grow.c b/Grow.c index 053a372..0916c5d 100644 --- a/Grow.c +++ b/Grow.c @@ -931,8 +931,8 @@ int Grow_reshape(char *devname, int fd, int quiet, char *backup_file, * old stripes and a whole number of new stripes. * So LCM for (chunksize*datadisks). */ - a = ochunk/512 * odata; - b = nchunk/512 * ndata; + a = (ochunk/512) * odata; + b = (nchunk/512) * ndata; /* Find GCD */ while (a != b) { if (a < b) @@ -941,7 +941,7 @@ int Grow_reshape(char *devname, int fd, int quiet, char *backup_file, a -= b; } /* LCM == product / GCD */ - blocks = ochunk/512 * nchunk/512 * odata * ndata / a; + blocks = (ochunk/512) * (nchunk/512) * odata * ndata / a; sysfs_free(sra); sra = sysfs_read(fd, 0, @@ -1284,7 +1284,7 @@ int grow_backup(struct mdinfo *sra, odata--; if (level == 6) odata--; - sysfs_set_num(sra, NULL, "suspend_hi", (offset + stripes * chunk/512) * odata); + sysfs_set_num(sra, NULL, "suspend_hi", (offset + stripes * (chunk/512)) * odata); /* Check that array hasn't become degraded, else we might backup the wrong data */ sysfs_get_ll(sra, NULL, "degraded", &new_degraded); if (new_degraded != *degraded) { @@ -1312,10 +1312,10 @@ int grow_backup(struct mdinfo *sra, } if (part) { bsb.arraystart2 = __cpu_to_le64(offset * odata); - bsb.length2 = __cpu_to_le64(stripes * chunk/512 * odata); + bsb.length2 = __cpu_to_le64(stripes * (chunk/512) * odata); } else { bsb.arraystart = __cpu_to_le64(offset * odata); - bsb.length = __cpu_to_le64(stripes * chunk/512 * odata); + bsb.length = __cpu_to_le64(stripes * (chunk/512) * odata); } if (part) bsb.magic[15] = '2'; @@ -1540,10 +1540,10 @@ static int child_grow(int afd, struct mdinfo *sra, unsigned long stripes, dests, destfd, destoffsets, 0, °raded, buf); validate(afd, destfd[0], destoffsets[0]); - wait_backup(sra, 0, stripes * chunk / 512, stripes * chunk / 512, + wait_backup(sra, 0, stripes * (chunk / 512), stripes * (chunk / 512), dests, destfd, destoffsets, 0); - sysfs_set_num(sra, NULL, "suspend_lo", (stripes * chunk/512) * data); + sysfs_set_num(sra, NULL, "suspend_lo", (stripes * (chunk/512)) * data); free(buf); /* FIXME this should probably be numeric */ sysfs_set_str(sra, NULL, "sync_max", "max"); @@ -1562,12 +1562,12 @@ static int child_shrink(int afd, struct mdinfo *sra, unsigned long stripes, if (posix_memalign((void**)&buf, 4096, disks * chunk)) return 0; - start = sra->component_size - stripes * chunk/512; + start = sra->component_size - stripes * (chunk/512); sysfs_set_num(sra, NULL, "sync_max", start); sysfs_set_str(sra, NULL, "sync_action", "reshape"); sysfs_set_num(sra, NULL, "suspend_lo", 0); sysfs_set_num(sra, NULL, "suspend_hi", 0); - rv = wait_backup(sra, 0, start - stripes * chunk/512, stripes * chunk/512, + rv = wait_backup(sra, 0, start - stripes * (chunk/512), stripes * (chunk/512), dests, destfd, destoffsets, 0); if (rv < 0) return 0; @@ -1577,9 +1577,9 @@ static int child_shrink(int afd, struct mdinfo *sra, unsigned long stripes, dests, destfd, destoffsets, 0, °raded, buf); validate(afd, destfd[0], destoffsets[0]); - wait_backup(sra, start, stripes*chunk/512, 0, + wait_backup(sra, start, stripes*(chunk/512), 0, dests, destfd, destoffsets, 0); - sysfs_set_num(sra, NULL, "suspend_lo", (stripes * chunk/512) * data); + sysfs_set_num(sra, NULL, "suspend_lo", (stripes * (chunk/512)) * data); free(buf); /* FIXME this should probably be numeric */ sysfs_set_str(sra, NULL, "sync_max", "max"); @@ -1614,7 +1614,7 @@ static int child_same_size(int afd, struct mdinfo *sra, unsigned long stripes, disks, chunk, level, layout, dests, destfd, destoffsets, 0, °raded, buf); - grow_backup(sra, (start + stripes) * chunk/512, stripes, + grow_backup(sra, (start + stripes) * (chunk/512), stripes, fds, offsets, disks, chunk, level, layout, dests, destfd, destoffsets, @@ -1624,16 +1624,16 @@ static int child_same_size(int afd, struct mdinfo *sra, unsigned long stripes, start += stripes * 2; /* where to read next */ size = sra->component_size / (chunk/512); while (start < size) { - if (wait_backup(sra, (start-stripes*2)*chunk/512, - stripes*chunk/512, 0, + if (wait_backup(sra, (start-stripes*2)*(chunk/512), + stripes*(chunk/512), 0, dests, destfd, destoffsets, part) < 0) return 0; - sysfs_set_num(sra, NULL, "suspend_lo", start*chunk/512 * data); + sysfs_set_num(sra, NULL, "suspend_lo", start*(chunk/512) * data); if (start + stripes > size) tailstripes = (size - start); - grow_backup(sra, start*chunk/512, tailstripes, + grow_backup(sra, start*(chunk/512), tailstripes, fds, offsets, disks, chunk, level, layout, dests, destfd, destoffsets, @@ -1642,15 +1642,15 @@ static int child_same_size(int afd, struct mdinfo *sra, unsigned long stripes, part = 1 - part; validate(afd, destfd[0], destoffsets[0]); } - if (wait_backup(sra, (start-stripes*2) * chunk/512, stripes * chunk/512, 0, + if (wait_backup(sra, (start-stripes*2) * (chunk/512), stripes * (chunk/512), 0, dests, destfd, destoffsets, part) < 0) return 0; - sysfs_set_num(sra, NULL, "suspend_lo", ((start-stripes)*chunk/512) * data); - wait_backup(sra, (start-stripes) * chunk/512, tailstripes * chunk/512, 0, + sysfs_set_num(sra, NULL, "suspend_lo", ((start-stripes)*(chunk/512)) * data); + wait_backup(sra, (start-stripes) * (chunk/512), tailstripes * (chunk/512), 0, dests, destfd, destoffsets, 1-part); - sysfs_set_num(sra, NULL, "suspend_lo", (size*chunk/512) * data); + sysfs_set_num(sra, NULL, "suspend_lo", (size*(chunk/512)) * data); sysfs_set_num(sra, NULL, "sync_speed_min", speed); free(buf); return 1; @@ -1991,8 +1991,8 @@ int Grow_continue(int mdfd, struct supertype *st, struct mdinfo *info, nchunk = info->new_chunk; - a = ochunk/512 * odata; - b = nchunk/512 * ndata; + a = (ochunk/512) * odata; + b = (nchunk/512) * ndata; /* Find GCD */ while (a != b) { if (a < b) @@ -2001,7 +2001,7 @@ int Grow_continue(int mdfd, struct supertype *st, struct mdinfo *info, a -= b; } /* LCM == product / GCD */ - blocks = ochunk/512 * nchunk/512 * odata * ndata / a; + blocks = (ochunk/512) * (nchunk/512) * odata * ndata / a; sra = sysfs_read(-1, devname2devnum(info->sys_name), GET_COMPONENT|GET_DEVS|GET_OFFSET|GET_STATE| From 33414a0182ae193150f65f7bca97a7e4d818a49e Mon Sep 17 00:00:00 2001 From: Dan Williams Date: Tue, 15 Jun 2010 17:55:41 -0700 Subject: [PATCH 3/4] Kill subarray v2 Support for deleting a subarray out of a container. When all subarrays are deleted the component devices are converted back into spares, a --zero-superblock is still needed to kill the remaining metadata at this point. This operation is blocked when the subarray is active and may also be blocked by the metadata handler when deleting the subarray might change the uuid of other active subarrays. For example, with imsm, deleting subarray 'n' may change the uuid of subarrays with indexes > n. Deleting a subarray needs to be a container wide event to ensure disks that record the modified subarray list perceive other disks that did not receive this change as out of date. Notes: The st->subarray parsing in super-intel.c and super-ddf.c is updated to be more strict now that we are reading user supplied subarray values. Offline container modification shares actions that mdmon typically handles so promote is_container_member() and version_to_superswitch() (formerly find_metadata_methods()) to generic utility functions for the cases where mdadm performs the operation. Signed-off-by: Dan Williams --- Kill.c | 78 ++++++++++++++++++++++ ReadMe.c | 1 + mdadm.8 | 10 +++ mdadm.c | 8 +++ mdadm.h | 8 +++ mdmon.c | 25 +------ super-ddf.c | 25 ++++++- super-intel.c | 179 ++++++++++++++++++++++++++++++++++++++++++++++---- util.c | 137 ++++++++++++++++++++++++++++++++++++++ 9 files changed, 431 insertions(+), 40 deletions(-) diff --git a/Kill.c b/Kill.c index e738978..b3344bd 100644 --- a/Kill.c +++ b/Kill.c @@ -79,3 +79,81 @@ int Kill(char *dev, struct supertype *st, int force, int quiet, int noexcl) close(fd); return rv; } + +int Kill_subarray(char *dev, char *subarray, int quiet) +{ + /* Delete a subarray out of a container, the subarry must be + * inactive. The subarray string must be a subarray index + * number. + * + * 0 = successfully deleted subarray from all container members + * 1 = failed to sync metadata to one or more devices + * 2 = failed to find the container, subarray, or other resource + * issue + */ + struct supertype supertype, *st = &supertype; + int fd, rv = 2; + + memset(st, 0, sizeof(*st)); + + if (snprintf(st->subarray, sizeof(st->subarray), "%s", subarray) >= + sizeof(st->subarray)) { + if (!quiet) + fprintf(stderr, + Name ": Input overflow for subarray '%s' > %zu bytes\n", + subarray, sizeof(st->subarray) - 1); + return 2; + } + + fd = open_subarray(dev, st, quiet); + if (fd < 0) + return 2; + + if (!st->ss->kill_subarray) { + if (!quiet) + fprintf(stderr, + Name ": Operation not supported for %s metadata\n", + st->ss->name); + goto free_super; + } + + if (is_subarray_active(subarray, st->devname)) { + if (!quiet) + fprintf(stderr, + Name ": Subarray-%s still active, aborting\n", + subarray); + goto free_super; + } + + if (mdmon_running(st->devnum)) + st->update_tail = &st->updates; + + /* ok we've found our victim, drop the axe */ + rv = st->ss->kill_subarray(st); + if (rv) { + if (!quiet) + fprintf(stderr, + Name ": Failed to delete subarray-%s from %s\n", + subarray, dev); + goto free_super; + } + + /* FIXME these routines do not report success/failure */ + if (st->update_tail) + flush_metadata_updates(st); + else + st->ss->sync_metadata(st); + + if (!quiet) + fprintf(stderr, + Name ": Deleted subarray-%s from %s, UUIDs may have changed\n", + subarray, dev); + + rv = 0; + + free_super: + st->ss->free_super(st); + close(fd); + + return rv; +} diff --git a/ReadMe.c b/ReadMe.c index 9d5a211..387ba6d 100644 --- a/ReadMe.c +++ b/ReadMe.c @@ -108,6 +108,7 @@ struct option long_options[] = { {"examine-bitmap", 0, 0, 'X'}, {"auto-detect", 0, 0, AutoDetect}, {"detail-platform", 0, 0, DetailPlatform}, + {"kill-subarray", 1, 0, KillSubarray}, /* synonyms */ {"monitor", 0, 0, 'F'}, diff --git a/mdadm.8 b/mdadm.8 index 4edfc41..784ba31 100644 --- a/mdadm.8 +++ b/mdadm.8 @@ -1172,6 +1172,16 @@ overwritten with zeros. With the block where the superblock would be is overwritten even if it doesn't appear to be valid. +.TP +.B \-\-kill\-subarray= +If the device is a container and the argument to \-\-kill\-subarray +specifies an inactive subarray in the container, then the subarray is +deleted. Deleting all subarrays will leave an 'empty-container' or +spare superblock on the drives. See \-\-zero\-superblock for completely +removing a superblock. Note that some formats depend on the subarray +index for generating a UUID, this command will fail if it would change +the UUID of an active subarray. + .TP .BR \-t ", " \-\-test When used with diff --git a/mdadm.c b/mdadm.c index d5e34c0..446fab8 100644 --- a/mdadm.c +++ b/mdadm.c @@ -103,6 +103,7 @@ int main(int argc, char *argv[]) int dosyslog = 0; int rebuild_map = 0; int auto_update_home = 0; + char *subarray = NULL; int print_help = 0; FILE *outf; @@ -216,6 +217,9 @@ int main(int argc, char *argv[]) case 'W': case Waitclean: case DetailPlatform: + case KillSubarray: + if (opt == KillSubarray) + subarray = optarg; case 'K': if (!mode) newmode = MISC; break; } if (mode && newmode == mode) { @@ -807,6 +811,7 @@ int main(int argc, char *argv[]) case O(MISC,'W'): case O(MISC, Waitclean): case O(MISC, DetailPlatform): + case O(MISC, KillSubarray): if (devmode && devmode != opt && (devmode == 'E' || (opt == 'E' && devmode != 'Q'))) { fprintf(stderr, Name ": --examine/-E cannot be given with -%c\n", @@ -1403,6 +1408,9 @@ int main(int argc, char *argv[]) rv |= Wait(dv->devname); continue; case Waitclean: rv |= WaitClean(dv->devname, -1, verbose-quiet); continue; + case KillSubarray: + rv |= Kill_subarray(dv->devname, subarray, quiet); + continue; } mdfd = open_mddev(dv->devname, 1); if (mdfd>=0) { diff --git a/mdadm.h b/mdadm.h index d9d17b0..f387477 100644 --- a/mdadm.h +++ b/mdadm.h @@ -273,6 +273,7 @@ enum special_options { AutoDetect, Waitclean, DetailPlatform, + KillSubarray, }; /* structures read from config file */ @@ -609,6 +610,8 @@ extern struct superswitch { struct mdinfo *(*container_content)(struct supertype *st); /* Allow a metadata handler to override mdadm's default layouts */ int (*default_layout)(int level); /* optional */ + /* Permit subarray's to be deleted from inactive containers */ + int (*kill_subarray)(struct supertype *st); /* optional */ /* for mdmon */ int (*open_new)(struct supertype *c, struct active_array *a, @@ -805,6 +808,7 @@ extern int Monitor(mddev_dev_t devlist, int dosyslog, int test, char *pidfile, int increments); extern int Kill(char *dev, struct supertype *st, int force, int quiet, int noexcl); +extern int Kill_subarray(char *dev, char *subarray, int quiet); extern int Wait(char *dev); extern int WaitClean(char *dev, int sock, int verbose); @@ -911,6 +915,10 @@ extern int create_mddev(char *dev, char *name, int autof, int trustworthy, #define METADATA 3 extern int open_mddev(char *dev, int report_errors); extern int open_container(int fd); +extern int is_container_member(struct mdstat_ent *ent, char *devname); +extern int is_subarray_active(char *subarray, char *devname); +extern int open_subarray(char *dev, struct supertype *st, int quiet); +extern struct superswitch *version_to_superswitch(char *vers); extern char *pid_dir; extern int mdmon_running(int devnum); diff --git a/mdmon.c b/mdmon.c index 69c320e..beb39cf 100644 --- a/mdmon.c +++ b/mdmon.c @@ -104,15 +104,6 @@ int __clone2(int (*fn)(void *), return mon_tid; } -static struct superswitch *find_metadata_methods(char *vers) -{ - if (strcmp(vers, "ddf") == 0) - return &super_ddf; - if (strcmp(vers, "imsm") == 0) - return &super_imsm; - return NULL; -} - static int make_pidfile(char *devname) { char path[100]; @@ -136,18 +127,6 @@ static int make_pidfile(char *devname) return 0; } -int is_container_member(struct mdstat_ent *mdstat, char *container) -{ - if (mdstat->metadata_version == NULL || - strncmp(mdstat->metadata_version, "external:", 9) != 0 || - !is_subarray(mdstat->metadata_version+9) || - strncmp(mdstat->metadata_version+10, container, strlen(container)) != 0 || - mdstat->metadata_version[10+strlen(container)] != '/') - return 0; - - return 1; -} - static void try_kill_monitor(pid_t pid, char *devname, int sock) { char buf[100]; @@ -414,9 +393,9 @@ static int mdmon(char *devname, int devnum, int must_fork, int takeover) exit(3); } - container->ss = find_metadata_methods(mdi->text_version); + container->ss = version_to_superswitch(mdi->text_version); if (container->ss == NULL) { - fprintf(stderr, "mdmon: %s uses unknown metadata: %s\n", + fprintf(stderr, "mdmon: %s uses unsupported metadata: %s\n", devname, mdi->text_version); exit(3); } diff --git a/super-ddf.c b/super-ddf.c index 0e6f1e5..736e07f 100644 --- a/super-ddf.c +++ b/super-ddf.c @@ -845,10 +845,18 @@ static int load_super_ddf(struct supertype *st, int fd, } if (st->subarray[0]) { + unsigned long val; struct vcl *v; + char *ep; + + val = strtoul(st->subarray, &ep, 10); + if (*ep != '\0') { + free(super); + return 1; + } for (v = super->conflist; v; v = v->next) - if (v->vcnum == atoi(st->subarray)) + if (v->vcnum == val) super->currentconf = v; if (!super->currentconf) { free(super); @@ -2870,14 +2878,25 @@ static int load_super_ddf_all(struct supertype *st, int fd, return 1; } if (st->subarray[0]) { + unsigned long val; struct vcl *v; + char *ep; + + val = strtoul(st->subarray, &ep, 10); + if (*ep != '\0') { + free(super); + return 1; + } for (v = super->conflist; v; v = v->next) - if (v->vcnum == atoi(st->subarray)) + if (v->vcnum == val) super->currentconf = v; - if (!super->currentconf) + if (!super->currentconf) { + free(super); return 1; + } } + *sbp = super; if (st->ss == NULL) { st->ss = &super_ddf; diff --git a/super-intel.c b/super-intel.c index bdd7a96..d81d620 100644 --- a/super-intel.c +++ b/super-intel.c @@ -282,6 +282,7 @@ struct extent { enum imsm_update_type { update_activate_spare, update_create_array, + update_kill_array, update_add_disk, }; @@ -303,6 +304,11 @@ struct imsm_update_create_array { struct imsm_dev dev; }; +struct imsm_update_kill_array { + enum imsm_update_type type; + int dev_idx; +}; + struct imsm_update_add_disk { enum imsm_update_type type; }; @@ -2753,11 +2759,20 @@ static int load_super_imsm_all(struct supertype *st, int fd, void **sbp, } if (st->subarray[0]) { - if (atoi(st->subarray) <= super->anchor->num_raid_devs) - super->current_vol = atoi(st->subarray); + unsigned long val; + char *ep; + + err = 1; + val = strtoul(st->subarray, &ep, 10); + if (*ep != '\0') { + free_imsm(super); + goto error; + } + + if (val < super->anchor->num_raid_devs) + super->current_vol = val; else { free_imsm(super); - err = 1; goto error; } } @@ -2824,8 +2839,17 @@ static int load_super_imsm(struct supertype *st, int fd, char *devname) } if (st->subarray[0]) { - if (atoi(st->subarray) <= super->anchor->num_raid_devs) - super->current_vol = atoi(st->subarray); + unsigned long val; + char *ep; + + val = strtoul(st->subarray, &ep, 10); + if (*ep != '\0') { + free_imsm(super); + return 1; + } + + if (val < super->anchor->num_raid_devs) + super->current_vol = val; else { free_imsm(super); return 1; @@ -4007,6 +4031,82 @@ static int validate_geometry_imsm(struct supertype *st, int level, int layout, close(cfd); return 0; } + +static void handle_missing(struct intel_super *super, struct imsm_dev *dev); + +static int kill_subarray_imsm(struct supertype *st) +{ + /* remove the subarray currently referenced by ->current_vol */ + __u8 i; + struct intel_dev **dp; + struct intel_super *super = st->sb; + __u8 current_vol = super->current_vol; + struct imsm_super *mpb = super->anchor; + + if (super->current_vol < 0) + return 2; + super->current_vol = -1; /* invalidate subarray cursor */ + + /* block deletions that would change the uuid of active subarrays + * + * FIXME when immutable ids are available, but note that we'll + * also need to fixup the invalidated/active subarray indexes in + * mdstat + */ + for (i = 0; i < mpb->num_raid_devs; i++) { + char subarray[4]; + + 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); + + return 2; + } + } + + if (st->update_tail) { + struct imsm_update_kill_array *u = malloc(sizeof(*u)); + + if (!u) + return 2; + u->type = update_kill_array; + u->dev_idx = current_vol; + append_metadata_update(st, u, sizeof(*u)); + + return 0; + } + + for (dp = &super->devlist; *dp;) + if ((*dp)->index == current_vol) { + *dp = (*dp)->next; + } else { + handle_missing(super, (*dp)->dev); + if ((*dp)->index > current_vol) + (*dp)->index--; + dp = &(*dp)->next; + } + + /* no more raid devices, all active components are now spares, + * but of course failed are still failed + */ + if (--mpb->num_raid_devs == 0) { + struct dl *d; + + for (d = super->disks; d; d = d->next) + if (d->index > -2) { + d->index = -1; + d->disk.status = SPARE_DISK; + } + } + + super->updates_pending++; + + return 0; +} #endif /* MDASSEMBLE */ static int is_rebuilding(struct imsm_dev *dev) @@ -4347,6 +4447,24 @@ static void mark_missing(struct imsm_dev *dev, struct imsm_disk *disk, int idx) memmove(&disk->serial[0], &disk->serial[1], MAX_RAID_SERIAL_LEN - 1); } +static void handle_missing(struct intel_super *super, struct imsm_dev *dev) +{ + __u8 map_state; + struct dl *dl; + int failed; + + if (!super->missing) + return; + failed = imsm_count_failed(super, dev); + map_state = imsm_check_degraded(super, dev, failed); + + dprintf("imsm: mark missing\n"); + end_migration(dev, map_state); + for (dl = super->missing; dl; dl = dl->next) + mark_missing(dev, &dl->disk, dl->index); + super->updates_pending++; +} + /* Handle dirty -> clean transititions and resync. Degraded and rebuild * states are handled in imsm_set_disk() with one exception, when a * resync is stopped due to a new failure this routine will set the @@ -4363,15 +4481,8 @@ static int imsm_set_array_state(struct active_array *a, int consistent) __u32 blocks_per_unit; /* before we activate this array handle any missing disks */ - if (consistent == 2 && super->missing) { - struct dl *dl; - - dprintf("imsm: mark missing\n"); - end_migration(dev, map_state); - for (dl = super->missing; dl; dl = dl->next) - mark_missing(dev, &dl->disk, dl->index); - super->updates_pending++; - } + if (consistent == 2) + handle_missing(super, dev); if (consistent == 2 && (!is_resync_complete(&a->info) || @@ -5067,6 +5178,45 @@ static void imsm_process_update(struct supertype *st, } break; } + case update_kill_array: { + struct imsm_update_kill_array *u = (void *) update->buf; + int victim = u->dev_idx; + struct active_array *a; + struct intel_dev **dp; + struct imsm_dev *dev; + + /* sanity check that we are not affecting the uuid of + * active arrays, or deleting an active array + * + * FIXME when immutable ids are available, but note that + * we'll also need to fixup the invalidated/active + * subarray indexes in mdstat + */ + for (a = st->arrays; a; a = a->next) + if (a->info.container_member >= victim) + break; + /* by definition if mdmon is running at least one array + * is active in the container, so checking + * mpb->num_raid_devs is just extra paranoia + */ + dev = get_imsm_dev(super, victim); + if (a || !dev || mpb->num_raid_devs == 1) { + dprintf("failed to delete subarray-%d\n", victim); + break; + } + + for (dp = &super->devlist; *dp;) + if ((*dp)->index == super->current_vol) { + *dp = (*dp)->next; + } else { + if ((*dp)->index > victim) + (*dp)->index--; + dp = &(*dp)->next; + } + mpb->num_raid_devs--; + super->updates_pending++; + break; + } case update_add_disk: /* we may be able to repair some arrays if disks are @@ -5242,6 +5392,7 @@ struct superswitch super_imsm = { .validate_geometry = validate_geometry_imsm, .add_to_super = add_to_super_imsm, .detail_platform = detail_platform_imsm, + .kill_subarray = kill_subarray_imsm, #endif .match_home = match_home_imsm, .uuid_from_super= uuid_from_super_imsm, diff --git a/util.c b/util.c index 25f1e56..66bf2f9 100644 --- a/util.c +++ b/util.c @@ -1392,6 +1392,143 @@ int open_container(int fd) return -1; } +struct superswitch *version_to_superswitch(char *vers) +{ + int i; + + for (i = 0; superlist[i]; i++) { + struct superswitch *ss = superlist[i]; + + if (strcmp(vers, ss->name) == 0) + return ss; + } + + return NULL; +} + +int is_container_member(struct mdstat_ent *mdstat, char *container) +{ + if (mdstat->metadata_version == NULL || + strncmp(mdstat->metadata_version, "external:", 9) != 0 || + !is_subarray(mdstat->metadata_version+9) || + strncmp(mdstat->metadata_version+10, container, strlen(container)) != 0 || + mdstat->metadata_version[10+strlen(container)] != '/') + return 0; + + return 1; +} + +int is_subarray_active(char *subarray, char *container) +{ + struct mdstat_ent *mdstat = mdstat_read(0, 0); + struct mdstat_ent *ent; + + for (ent = mdstat; ent; ent = ent->next) { + if (is_container_member(ent, container)) { + char *inst = &ent->metadata_version[10+strlen(container)+1]; + + if (strcmp(inst, subarray) == 0) + break; + } + } + + free_mdstat(mdstat); + + return ent != NULL; +} + +/* open_subarray - opens a subarray in a container + * @dev: container device name + * @st: supertype with only ->subarray set + * @quiet: block reporting errors flag + * + * On success returns an fd to a container and fills in *st + */ +int open_subarray(char *dev, struct supertype *st, int quiet) +{ + struct mdinfo *mdi; + int fd, err = 1; + + fd = open(dev, O_RDWR|O_EXCL); + if (fd < 0) { + if (!quiet) + fprintf(stderr, Name ": Couldn't open %s, aborting\n", + dev); + return 2; + } + + st->devnum = fd2devnum(fd); + if (st->devnum == NoMdDev) { + if (!quiet) + fprintf(stderr, + Name ": Failed to determine device number for %s\n", + dev); + goto close_fd; + } + + mdi = sysfs_read(fd, st->devnum, GET_VERSION|GET_LEVEL); + if (!mdi) { + if (!quiet) + fprintf(stderr, Name ": Failed to read sysfs for %s\n", + dev); + goto close_fd; + } + + if (mdi->array.level != UnSet) { + if (!quiet) + fprintf(stderr, Name ": %s is not a container\n", dev); + goto free_sysfs; + } + + st->ss = version_to_superswitch(mdi->text_version); + if (!st->ss) { + if (!quiet) + fprintf(stderr, + Name ": Operation not supported for %s metadata\n", + mdi->text_version); + goto free_sysfs; + } + + st->devname = devnum2devname(st->devnum); + if (!st->devname) { + if (!quiet) + fprintf(stderr, Name ": Failed to allocate device name\n"); + goto free_sysfs; + } + + if (st->ss->load_super(st, fd, NULL)) { + if (!quiet) + fprintf(stderr, Name ": Failed to find subarray-%s in %s\n", + st->subarray, dev); + goto free_name; + } + + if (!st->loaded_container) { + if (!quiet) + fprintf(stderr, Name ": %s is not a container\n", dev); + goto free_super; + } + + err = 0; + + free_super: + if (err) + st->ss->free_super(st); + free_name: + if (err) + free(st->devname); + free_sysfs: + sysfs_free(mdi); + close_fd: + if (err) + close(fd); + + if (err) + return -1; + else + return fd; +} + int add_disk(int mdfd, struct supertype *st, struct mdinfo *sra, struct mdinfo *info) { From aa534678baad80689a642ba1bd602a00a267ac03 Mon Sep 17 00:00:00 2001 From: Dan Williams Date: Tue, 22 Jun 2010 16:30:59 -0700 Subject: [PATCH 4/4] Rename subarray v2 Allow the name of the array stored in the metadata to be updated. In some cases the metadata format may not be able to support this rename without modifying the UUID. In these cases the request will be blocked. Otherwise we allow the rename to take place, even for active arrays. This assumes that the user understands the difference between the kernel node name, the device node symlink name, and the metadata specific name. Anticipating further need to modify subarrays in-place, introduce the ->update_subarray() superswitch method. A future potential use case is setting storage pool (spare-group) identifiers. Signed-off-by: Dan Williams --- Manage.c | 53 +++++++++++++++++++++++ ReadMe.c | 1 + mdadm.8 | 30 +++++++++++++ mdadm.c | 41 ++++++++++++++++-- mdadm.h | 7 ++- super-intel.c | 117 +++++++++++++++++++++++++++++++++++++++++++++----- 6 files changed, 235 insertions(+), 14 deletions(-) diff --git a/Manage.c b/Manage.c index f6fb3ef..cca1503 100644 --- a/Manage.c +++ b/Manage.c @@ -869,4 +869,57 @@ int autodetect(void) } return rv; } + +int Update_subarray(char *dev, char *subarray, char *update, mddev_ident_t ident, int quiet) +{ + struct supertype supertype, *st = &supertype; + int fd, rv = 2; + + memset(st, 0, sizeof(*st)); + if (snprintf(st->subarray, sizeof(st->subarray), "%s", subarray) >= + sizeof(st->subarray)) { + if (!quiet) + fprintf(stderr, + Name ": Input overflow for subarray '%s' > %zu bytes\n", + subarray, sizeof(st->subarray) - 1); + return 2; + } + + fd = open_subarray(dev, st, quiet); + if (fd < 0) + return 2; + + if (!st->ss->update_subarray) { + if (!quiet) + fprintf(stderr, + Name ": Operation not supported for %s metadata\n", + st->ss->name); + goto free_super; + } + + if (mdmon_running(st->devnum)) + st->update_tail = &st->updates; + + rv = st->ss->update_subarray(st, update, ident); + + if (rv) { + if (!quiet) + fprintf(stderr, Name ": Failed to update %s of subarray-%s in %s\n", + update, subarray, dev); + } else if (st->update_tail) + flush_metadata_updates(st); + else + st->ss->sync_metadata(st); + + if (rv == 0 && strcmp(update, "name") == 0 && !quiet) + fprintf(stderr, + Name ": Updated subarray-%s name from %s, UUIDs may have changed\n", + subarray, dev); + + free_super: + st->ss->free_super(st); + close(fd); + + return rv; +} #endif diff --git a/ReadMe.c b/ReadMe.c index 387ba6d..fa33310 100644 --- a/ReadMe.c +++ b/ReadMe.c @@ -109,6 +109,7 @@ struct option long_options[] = { {"auto-detect", 0, 0, AutoDetect}, {"detail-platform", 0, 0, DetailPlatform}, {"kill-subarray", 1, 0, KillSubarray}, + {"update-subarray", 1, 0, UpdateSubarray}, /* synonyms */ {"monitor", 0, 0, 'F'}, diff --git a/mdadm.8 b/mdadm.8 index 784ba31..1909819 100644 --- a/mdadm.8 +++ b/mdadm.8 @@ -1182,6 +1182,14 @@ removing a superblock. Note that some formats depend on the subarray index for generating a UUID, this command will fail if it would change the UUID of an active subarray. +.TP +.B \-\-update\-subarray= +If the device is a container and the argument to \-\-update\-subarray +specifies a subarray in the container, then attempt to update the given +superblock field in the subarray. See below in +.B MISC MODE +for details. + .TP .BR \-t ", " \-\-test When used with @@ -1760,6 +1768,28 @@ metadata is platform independent metadata failed to find its platform components on this system .RE +.TP +.B \-\-update\-subarray= +If the device is a container and the argument to \-\-update\-subarray +specifies a subarray in the container, then attempt to update the given +superblock field in the subarray. Similar to updating an array in +"assemble" mode, the field to update is selected by +.B \-U +or +.B \-\-update= +option. Currently only +.B name +is supported. + +The +.B name +option updates the subarray name in the metadata, it may not affect the +device node name or the device node symlink until the subarray is +re\-assembled. If updating +.B name +would change the UUID of an active subarray this operation is blocked, +and the command will end in an error. + .TP .B \-\-examine The device should be a component of an md array. diff --git a/mdadm.c b/mdadm.c index 446fab8..e7435fd 100644 --- a/mdadm.c +++ b/mdadm.c @@ -218,8 +218,14 @@ int main(int argc, char *argv[]) case Waitclean: case DetailPlatform: case KillSubarray: - if (opt == KillSubarray) + case UpdateSubarray: + if (opt == KillSubarray || opt == UpdateSubarray) { + if (subarray) { + fprintf(stderr, Name ": subarray can only be specified once\n"); + exit(2); + } subarray = optarg; + } case 'K': if (!mode) newmode = MISC; break; } if (mode && newmode == mode) { @@ -593,11 +599,16 @@ int main(int argc, char *argv[]) case O(CREATE,'N'): case O(ASSEMBLE,'N'): + case O(MISC,'N'): if (ident.name[0]) { fprintf(stderr, Name ": name cannot be set twice. " "Second value %s.\n", optarg); exit(2); } + if (mode == MISC && !subarray) { + fprintf(stderr, Name ": -N/--name only valid with --update-subarray in misc mode\n"); + exit(2); + } if (strlen(optarg) > 32) { fprintf(stderr, Name ": name '%s' is too long, 32 chars max.\n", optarg); @@ -624,11 +635,16 @@ int main(int argc, char *argv[]) continue; case O(ASSEMBLE,'U'): /* update the superblock */ + case O(MISC,'U'): if (update) { fprintf(stderr, Name ": Can only update one aspect of superblock, both %s and %s given.\n", update, optarg); exit(2); } + if (mode == MISC && !subarray) { + fprintf(stderr, Name ": Only subarrays can be updated in misc mode\n"); + exit(2); + } update = optarg; if (strcmp(update, "sparc2.2")==0) continue; @@ -812,10 +828,20 @@ int main(int argc, char *argv[]) case O(MISC, Waitclean): case O(MISC, DetailPlatform): case O(MISC, KillSubarray): + case O(MISC, UpdateSubarray): if (devmode && devmode != opt && (devmode == 'E' || (opt == 'E' && devmode != 'Q'))) { - fprintf(stderr, Name ": --examine/-E cannot be given with -%c\n", - devmode =='E'?opt:devmode); + fprintf(stderr, Name ": --examine/-E cannot be given with "); + if (devmode == 'E') { + if (option_index >= 0) + fprintf(stderr, "--%s\n", + long_options[option_index].name); + else + fprintf(stderr, "-%c\n", opt); + } else if (isalpha(devmode)) + fprintf(stderr, "-%c\n", devmode); + else + fprintf(stderr, "previous option\n"); exit(2); } devmode = opt; @@ -1411,6 +1437,15 @@ int main(int argc, char *argv[]) case KillSubarray: rv |= Kill_subarray(dv->devname, subarray, quiet); continue; + case UpdateSubarray: + if (update == NULL) { + fprintf(stderr, + Name ": -U/--update must be specified with --update-subarray\n"); + rv |= 1; + continue; + } + rv |= Update_subarray(dv->devname, subarray, update, &ident, quiet); + continue; } mdfd = open_mddev(dv->devname, 1); if (mdfd>=0) { diff --git a/mdadm.h b/mdadm.h index f387477..68d61a3 100644 --- a/mdadm.h +++ b/mdadm.h @@ -258,6 +258,7 @@ extern char Version[], Usage[], Help[], OptionHelp[], /* for option that don't have short equivilents, we assign arbitrary * small numbers. '1' means an undecorated option, so we start at '2'. + * (note we must stop before we get to 65 i.e. 'A') */ enum special_options { AssumeClean = 2, @@ -266,7 +267,7 @@ enum special_options { ReAdd, NoDegraded, Sparc22, - BackupFile, + BackupFile, /* 8 */ HomeHost, AutoHomeHost, Symlinks, @@ -274,6 +275,7 @@ enum special_options { Waitclean, DetailPlatform, KillSubarray, + UpdateSubarray, /* 16 */ }; /* structures read from config file */ @@ -612,6 +614,8 @@ extern struct superswitch { int (*default_layout)(int level); /* optional */ /* Permit subarray's to be deleted from inactive containers */ int (*kill_subarray)(struct supertype *st); /* optional */ + /* Permit subarray's to be modified */ + int (*update_subarray)(struct supertype *st, char *update, mddev_ident_t ident); /* optional */ /* for mdmon */ int (*open_new)(struct supertype *c, struct active_array *a, @@ -809,6 +813,7 @@ extern int Monitor(mddev_dev_t devlist, extern int Kill(char *dev, struct supertype *st, int force, int quiet, int noexcl); extern int Kill_subarray(char *dev, char *subarray, int quiet); +extern int Update_subarray(char *dev, char *subarray, char *update, mddev_ident_t ident, int quiet); extern int Wait(char *dev); extern int WaitClean(char *dev, int sock, int verbose); diff --git a/super-intel.c b/super-intel.c index d81d620..f0377b8 100644 --- a/super-intel.c +++ b/super-intel.c @@ -283,6 +283,7 @@ enum imsm_update_type { update_activate_spare, update_create_array, update_kill_array, + update_rename_array, update_add_disk, }; @@ -309,6 +310,12 @@ struct imsm_update_kill_array { int dev_idx; }; +struct imsm_update_rename_array { + enum imsm_update_type type; + __u8 name[MAX_RAID_SERIAL_LEN]; + int dev_idx; +}; + struct imsm_update_add_disk { enum imsm_update_type type; }; @@ -2939,6 +2946,30 @@ static void imsm_update_version_info(struct intel_super *super) } } +static int check_name(struct intel_super *super, char *name, int quiet) +{ + struct imsm_super *mpb = super->anchor; + char *reason = NULL; + int i; + + if (strlen(name) > MAX_RAID_SERIAL_LEN) + reason = "must be 16 characters or less"; + + for (i = 0; i < mpb->num_raid_devs; i++) { + struct imsm_dev *dev = get_imsm_dev(super, i); + + if (strncmp((char *) dev->volume, name, MAX_RAID_SERIAL_LEN) == 0) { + reason = "already exists"; + break; + } + } + + if (reason && !quiet) + fprintf(stderr, Name ": 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) @@ -2990,16 +3021,8 @@ static int init_super_imsm_volume(struct supertype *st, mdu_array_info_t *info, if (super->current_vol == 0) mpb->num_disks = 0; - for (i = 0; i < super->current_vol; i++) { - dev = get_imsm_dev(super, i); - if (strncmp((char *) dev->volume, name, - MAX_RAID_SERIAL_LEN) == 0) { - fprintf(stderr, Name": '%s' is already defined for this container\n", - name); - return 0; - } - } - + if (!check_name(super, name, 0)) + return 0; sprintf(st->subarray, "%d", idx); dv = malloc(sizeof(*dv)); if (!dv) { @@ -4107,6 +4130,54 @@ static int kill_subarray_imsm(struct supertype *st) return 0; } + +static int update_subarray_imsm(struct supertype *st, char *update, mddev_ident_t ident) +{ + /* update the subarray currently referenced by ->current_vol */ + struct intel_super *super = st->sb; + struct imsm_super *mpb = super->anchor; + + if (super->current_vol < 0) + return 2; + + if (strcmp(update, "name") == 0) { + char *name = ident->name; + + if (is_subarray_active(st->subarray, st->devname)) { + fprintf(stderr, + Name ": Unable to update name of active subarray\n"); + return 2; + } + + if (!check_name(super, name, 0)) + return 2; + + if (st->update_tail) { + struct imsm_update_rename_array *u = malloc(sizeof(*u)); + + if (!u) + return 2; + u->type = update_rename_array; + u->dev_idx = super->current_vol; + snprintf((char *) u->name, MAX_RAID_SERIAL_LEN, "%s", name); + append_metadata_update(st, u, sizeof(*u)); + } else { + struct imsm_dev *dev; + int i; + + dev = get_imsm_dev(super, super->current_vol); + snprintf((char *) dev->volume, MAX_RAID_SERIAL_LEN, "%s", name); + for (i = 0; i < mpb->num_raid_devs; i++) { + dev = get_imsm_dev(super, i); + handle_missing(super, dev); + } + super->updates_pending++; + } + } else + return 2; + + return 0; +} #endif /* MDASSEMBLE */ static int is_rebuilding(struct imsm_dev *dev) @@ -5217,6 +5288,31 @@ static void imsm_process_update(struct supertype *st, super->updates_pending++; break; } + case update_rename_array: { + struct imsm_update_rename_array *u = (void *) update->buf; + char name[MAX_RAID_SERIAL_LEN+1]; + int target = u->dev_idx; + struct active_array *a; + struct imsm_dev *dev; + + /* sanity check that we are not affecting the uuid of + * an active array + */ + snprintf(name, MAX_RAID_SERIAL_LEN, "%s", (char *) u->name); + name[MAX_RAID_SERIAL_LEN] = '\0'; + for (a = st->arrays; a; a = a->next) + if (a->info.container_member == target) + break; + dev = get_imsm_dev(super, u->dev_idx); + if (a || !dev || !check_name(super, name, 1)) { + dprintf("failed to rename subarray-%d\n", target); + break; + } + + snprintf((char *) dev->volume, MAX_RAID_SERIAL_LEN, name); + super->updates_pending++; + break; + } case update_add_disk: /* we may be able to repair some arrays if disks are @@ -5393,6 +5489,7 @@ struct superswitch super_imsm = { .add_to_super = add_to_super_imsm, .detail_platform = detail_platform_imsm, .kill_subarray = kill_subarray_imsm, + .update_subarray = update_subarray_imsm, #endif .match_home = match_home_imsm, .uuid_from_super= uuid_from_super_imsm,