Cyrille Chépélov
2011-Jan-25 18:46 UTC
full btrfs partition, became unmountable (+ a solution that thankfully worked for me)
Hello all,
Last Friday, the /var and /home partition on one of my appliances became
full. This should normally not be much of a problem, except that after
the incident, I had been unable to mount the partition back again.
The appliance runs 2.6.32 as provided by Debian during the last two
months.
The rescue computer runs 2.6.37; both exhibited the same behaviour at
mount: an infinite loop-and-abort cycle (I unfortunately did not write
down the exact messages, but in a nutshell, there was not enough free
space to replay the log, so it aborted).
After pulling the SD card (yes) to break the loop, I ended up with a
corrupt file system. Any attempt to mount, debug or fsck (using
btrfs-tools 0.19+20100601 as shipped by Debian, or compiled from git
1b444cd2e6ab8dcafdd) aborted with the following message:
btrfs-debug-tree: disk-io.c:741: open_ctree_fd: Assertion `!(!
tree_root->node)'' failed.
After much scavenging on the disk image, I finally managed to recover,
using the (dirty) patch attached here. Since apparently other people had
similar issues, I''m posting it in the hope it might be useful.
-- Cyrille
PS: Chris, if btrfs-images of "before" and "after" my
butcher fix would
be useful to you, just let me know.
Shawn Stricker
2011-Jan-26 04:38 UTC
Re: full btrfs partition, became unmountable (+ a solution that thankfully worked for me)
Not sure where you pulled your source from but a fresh checkout of either master or next of git.kernel.org/pub/scm/linux/kernel/git/mason/btrfs-progs-unstable.git does not compile properly. They both fail with cc1: warnings being treated as errors disk-io.c: In function ‘btrfs_read_dev_super’: disk-io.c:937: error: format ‘%lu’ expects type ‘long unsigned int’, but argument 4 has type ‘unsigned int’ disk-io.c:957: error: implicit declaration of function ‘uuid_unparse’ am I patching/compiling from the wrong source or is there something I am missing? On Jan 25, 2011, at 1:46 PM, Cyrille Chépélov wrote:> Hello all, > > Last Friday, the /var and /home partition on one of my appliances became > full. This should normally not be much of a problem, except that after > the incident, I had been unable to mount the partition back again. > > The appliance runs 2.6.32 as provided by Debian during the last two > months. > The rescue computer runs 2.6.37; both exhibited the same behaviour at > mount: an infinite loop-and-abort cycle (I unfortunately did not write > down the exact messages, but in a nutshell, there was not enough free > space to replay the log, so it aborted). > > After pulling the SD card (yes) to break the loop, I ended up with a > corrupt file system. Any attempt to mount, debug or fsck (using > btrfs-tools 0.19+20100601 as shipped by Debian, or compiled from git > 1b444cd2e6ab8dcafdd) aborted with the following message: > btrfs-debug-tree: disk-io.c:741: open_ctree_fd: Assertion `!(! > tree_root->node)'' failed. > > After much scavenging on the disk image, I finally managed to recover, > using the (dirty) patch attached here. Since apparently other people had > similar issues, I''m posting it in the hope it might be useful. > > -- Cyrille > > PS: Chris, if btrfs-images of "before" and "after" my butcher fix would > be useful to you, just let me know. > <scavenge.patch>-- To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
Cyrille Chépélov
2011-Jan-26 07:46 UTC
Re: full btrfs partition, became unmountable (+ a solution that thankfully worked for me)
Le mardi 25 janvier 2011 à 23:38 -0500, Shawn Stricker a écrit :> Not sure where you pulled your source from but a fresh checkout of either master or next of git.kernel.org/pub/scm/linux/kernel/git/mason/btrfs-progs-unstable.git does not compile properly. > They both fail with > > cc1: warnings being treated as errors > disk-io.c: In function ‘btrfs_read_dev_super’: > disk-io.c:937: error: format ‘%lu’ expects type ‘long unsigned int’, but argument 4 has type ‘unsigned int’ > disk-io.c:957: error: implicit declaration of function ‘uuid_unparse’ > > am I patching/compiling from the wrong source or is there something I am missing?uh, I had been compiling with CFLAGS=-g, where the makefile specifies "-O2 -Werror" -Werror causes warnings to be treated as errors, which is a good thing in a way (makes sure stuff as this gets caught :) ) fixes are: * line 937 (patched), should be %llu instead of %lu * line 957, there should be a prototype for uuid_unparse(), most certainly by including <uuid/uuid.h> please try this patch instead. Thanks for the feedback! -- Cyrille> On Jan 25, 2011, at 1:46 PM, Cyrille Chépélov wrote: > > > Hello all, > > > > Last Friday, the /var and /home partition on one of my appliances became > > full. This should normally not be much of a problem, except that after > > the incident, I had been unable to mount the partition back again. > > > > The appliance runs 2.6.32 as provided by Debian during the last two > > months. > > The rescue computer runs 2.6.37; both exhibited the same behaviour at > > mount: an infinite loop-and-abort cycle (I unfortunately did not write > > down the exact messages, but in a nutshell, there was not enough free > > space to replay the log, so it aborted). > > > > After pulling the SD card (yes) to break the loop, I ended up with a > > corrupt file system. Any attempt to mount, debug or fsck (using > > btrfs-tools 0.19+20100601 as shipped by Debian, or compiled from git > > 1b444cd2e6ab8dcafdd) aborted with the following message: > > btrfs-debug-tree: disk-io.c:741: open_ctree_fd: Assertion `!(! > > tree_root->node)'' failed. > > > > After much scavenging on the disk image, I finally managed to recover, > > using the (dirty) patch attached here. Since apparently other people had > > similar issues, I''m posting it in the hope it might be useful. > > > > -- Cyrille > > > > PS: Chris, if btrfs-images of "before" and "after" my butcher fix would > > be useful to you, just let me know. > > <scavenge.patch> >--=-TovPMUBbcwNB5a+bSJwV Content-Disposition: attachment; filename="scavenge-2.patch" Content-Type: text/x-patch; name="scavenge-2.patch"; charset="UTF-8" Content-Transfer-Encoding: 7bit diff --git a/btrfsck.c b/btrfsck.c index 63e44d1..1e6bc32 100644 --- a/btrfsck.c +++ b/btrfsck.c @@ -2823,13 +2823,17 @@ int main(int ac, char **av) u64 bytenr = 0; int ret; int num; + int emergency_super = 0; while(1) { int c; - c = getopt(ac, av, "s:"); + c = getopt(ac, av, "es:"); if (c < 0) break; switch(c) { + case ''e'': + emergency_super = 1; + break; case ''s'': num = atol(optarg); bytenr = btrfs_sb_offset(num); @@ -2861,6 +2865,12 @@ int main(int ac, char **av) if (root == NULL) return 1; + if (root->fs_info->emergency_root_tree && (!emergency_super)) { + printf("DANGEROUS: had to use a synthetic super. Please run with ''-e'' flag if you know why you do it.\n"); + printf(" ... and have perfect backups.\n"); + return 1; + } + ret = check_extents(root); if (ret) goto out; @@ -2869,6 +2879,27 @@ int main(int ac, char **av) goto out; ret = check_root_refs(root, &root_cache); + if (ret) + goto out; + + + if (root->fs_info->emergency_root_tree) { + printf("DANGEROUS: had to use a scavenged root. Apparently could figure out the primary trees ?\n" + " Now writing supers, knock wood.\n"); + free_root_recs(&root_cache); + close_ctree(root); + + cache_tree_init(&root_cache); + root = open_ctree(av[optind], bytenr, 1 /* WRITES! */); + + ret = write_all_supers(root); + if (ret) { + printf("Error writing superblocks.\n"); + goto out; + } + printf("Wrote back superblocks.\n"); + } + out: free_root_recs(&root_cache); close_ctree(root); @@ -2897,6 +2928,7 @@ out: printf("file data blocks allocated: %llu\n referenced %llu\n", (unsigned long long)data_bytes_allocated, (unsigned long long)data_bytes_referenced); + printf("%s\n", BTRFS_BUILD_VERSION); return ret; } diff --git a/ctree.h b/ctree.h index b79e238..7439d87 100644 --- a/ctree.h +++ b/ctree.h @@ -728,6 +728,7 @@ struct btrfs_fs_info { struct list_head space_info; int system_allocs; int readonly; + int emergency_root_tree; }; /* diff --git a/disk-io.c b/disk-io.c index a6e1000..0b4e7f9 100644 --- a/disk-io.c +++ b/disk-io.c @@ -23,6 +23,7 @@ #include <stdlib.h> #include <sys/types.h> #include <sys/stat.h> +#include <uuid/uuid.h> #include <fcntl.h> #include <unistd.h> #include "kerncompat.h" @@ -41,8 +42,11 @@ static int check_tree_block(struct btrfs_root *root, struct extent_buffer *buf) struct btrfs_fs_devices *fs_devices; int ret = 1; - if (buf->start != btrfs_header_bytenr(buf)) + if (buf->start != btrfs_header_bytenr(buf)) { + fprintf(stderr, "start place mismatch, buf says %llu btrfs_hdr says %llu\n", + buf->start, btrfs_header_bytenr(buf)); return ret; + } fs_devices = root->fs_info->fs_devices; while (fs_devices) { @@ -204,16 +208,26 @@ struct extent_buffer *read_tree_block(struct btrfs_root *root, u64 bytenr, eb->dev_bytenr = multi->stripes[0].physical; kfree(multi); ret = read_extent_from_disk(eb); - if (ret == 0 && check_tree_block(root, eb) == 0 && - csum_tree_block(root, eb, 1) == 0 && - verify_parent_transid(eb->tree, eb, parent_transid) == 0) { - btrfs_set_buffer_uptodate(eb); - return eb; + if (ret == 0) { + + if (check_tree_block(root, eb) == 0) { + /* fprintf(stderr, "checked tree block %p for %p -- %llu \n", root, eb, eb->start);*/ + if (csum_tree_block(root, eb, 1) == 0) { + /* fprintf(stderr, "tree block csum %p for %p--%llu is OK\n", root, eb, eb->start); */ + if (verify_parent_transid(eb->tree, eb, parent_transid) == 0) { + /*fprintf(stderr, "tree block %p--%llu has correct transid, setting uptodate\n", eb, eb->start); */ + + btrfs_set_buffer_uptodate(eb); + return eb; + } + } + } + } num_copies = btrfs_num_copies(&root->fs_info->mapping_tree, eb->start, eb->len); if (num_copies == 1) { - break; + //break; } mirror_num++; if (mirror_num > num_copies) { @@ -581,7 +595,7 @@ struct btrfs_root *open_ctree(const char *filename, u64 sb_bytenr, int writes) fp = open(filename, flags, 0600); if (fp < 0) { - fprintf (stderr, "Could not open %s\n", filename); + fprintf (stderr, "Could not open %s: %s\n", filename, strerror(errno)); return NULL; } root = open_ctree_fd(fp, filename, sb_bytenr, writes); @@ -590,6 +604,102 @@ struct btrfs_root *open_ctree(const char *filename, u64 sb_bytenr, int writes) return root; } +#define MAX_EMERG_BUCKETS 30 +static int try_emergency_tree_fixup(struct btrfs_super_block *disk_super, + struct btrfs_root *chunk_root, + struct btrfs_root *tree_root) +{ + struct btrfs_header* buf; + u64 ofs, oldofs; + u32 blocksize; + u64 generation; + int ret, i; + + u64 best_gen[MAX_EMERG_BUCKETS]; + u64 best_bytenr[MAX_EMERG_BUCKETS]; + + memset(&best_gen, 0, sizeof(best_gen)); + memset(&best_bytenr, 0, sizeof(best_bytenr)); + + blocksize = btrfs_level_size(tree_root, + btrfs_super_chunk_root_level(disk_super)); + generation = btrfs_super_chunk_root_generation(disk_super); + buf = (struct btrfs_header*)malloc(blocksize); + + oldofs = btrfs_super_bytenr(disk_super); + for (ofs = oldofs; + ofs < btrfs_super_total_bytes(disk_super); + ofs += blocksize) { + + ret = pread(chunk_root->node->fd, buf, blocksize, ofs); + if (ret == blocksize) { + u64 blockofs = le64_to_cpu(buf->bytenr); + if (blockofs != ofs) continue; + + char* src = chunk_root->node->data + (long)btrfs_header_fsid(chunk_root->node); + char* dst = (char*) &(buf->fsid); + if (memcmp(src, dst, sizeof(buf->fsid)) != 0) + continue; + + u64 blockgen = le64_to_cpu(buf->generation); + u64 blockowner = le64_to_cpu(buf->owner); + u32 blocknritems = le32_to_cpu(buf->nritems); + u8 blocklevel = le8_to_cpu(buf->level); + + /* + fprintf(stderr, + " found valid header at %llu(+%llu) -- gen=%llu owner=%llu nritems=%u level=%u ", + ofs, ofs-oldofs, blockgen, blockowner, blocknritems, blocklevel); + */ + + if ((blockowner < 0ull) && (blockowner > -11ull)) { blockowner += 30; /* hack */ } + + if ((blockowner >= 0) && (blockowner < MAX_EMERG_BUCKETS)) { + if (blockgen > best_gen[blockowner]) { + best_gen[blockowner] = blockgen; + best_bytenr[blockowner] = ofs; + + fprintf(stderr, + " found valid header at %llu(+%llu) -- gen=%llu owner=%llu nritems=%u level=%u ", + ofs, ofs-oldofs, blockgen, blockowner, blocknritems, blocklevel); + + fprintf(stderr, + " ... new best gen for ObjectID %llu at %llu\n", blockowner, ofs); + } else { + } + } else { + } + + oldofs = ofs; + } + } + fprintf(stderr,"*** done scanning, at offset %llu ***\n", ofs); + for ( i = 0; i < MAX_EMERG_BUCKETS; ++i) { + fprintf(stderr," for ObjectID: %d, max gen=%llu at %llu\n", i, best_gen[i], best_bytenr[i]); + } + + + free(buf); + +#if 1 + if (best_gen[BTRFS_CSUM_TREE_OBJECTID] == best_gen[BTRFS_ROOT_TREE_OBJECTID]) { + /* now we try to do the repair */ + fprintf(stderr, + " ATTEMPTING DANGEROUS REPAIR with root gen=%llu bytenr=%llu\n", + best_gen[BTRFS_ROOT_TREE_OBJECTID], best_bytenr[BTRFS_ROOT_TREE_OBJECTID]); + btrfs_set_super_generation(disk_super, best_gen[BTRFS_ROOT_TREE_OBJECTID]); + btrfs_set_super_root(disk_super, best_bytenr[BTRFS_ROOT_TREE_OBJECTID]); + + /* not actually changing the on-disk super. Debug should proceed, + fsck should eventually rewrite the super ?*/ + tree_root->fs_info->emergency_root_tree = 1; + return 0; + } +#endif + + return 1; /* nothing done */ +} + struct btrfs_root *open_ctree_fd(int fp, const char *path, u64 sb_bytenr, int writes) { @@ -736,7 +846,26 @@ struct btrfs_root *open_ctree_fd(int fp, const char *path, u64 sb_bytenr, tree_root->node = read_tree_block(tree_root, btrfs_super_root(disk_super), blocksize, generation); + if (!tree_root->node) { + fprintf(stderr,"*** did not find a root, about to abort ***\n"); + fprintf(stderr,"*** will attempt to find useful trees before bailing out anyway: ***\n"); + ret = try_emergency_tree_fixup(disk_super, chunk_root, tree_root); + if (!ret) { + fprintf(stderr,"a repair happened, trying again (once):\n"); + + generation = btrfs_super_generation(disk_super); + tree_root->node = read_tree_block(tree_root, + btrfs_super_root(disk_super), + blocksize, generation); + if (!tree_root->node) { + fprintf(stderr,"*** again, did not find a root, about to abort, for good. ***\n"); + } + } + } BUG_ON(!tree_root->node); + if (!tree_root->node) { + fprintf(stderr,"*** huh? ***\n"); + } ret = find_and_setup_root(tree_root, fs_info, BTRFS_EXTENT_TREE_OBJECTID, extent_root); BUG_ON(ret); @@ -774,12 +903,15 @@ struct btrfs_root *open_ctree_fd(int fp, const char *path, u64 sb_bytenr, int btrfs_read_dev_super(int fd, struct btrfs_super_block *sb, u64 sb_bytenr) { u8 fsid[BTRFS_FSID_SIZE]; + char up_fsid[37], up_ofsid[37]; struct btrfs_super_block buf; int i; int ret; u64 transid = 0; u64 bytenr; + + if (sb_bytenr != BTRFS_SUPER_INFO_OFFSET) { ret = pread64(fd, &buf, sizeof(buf), sb_bytenr); if (ret < sizeof(buf)) @@ -796,24 +928,63 @@ int btrfs_read_dev_super(int fd, struct btrfs_super_block *sb, u64 sb_bytenr) for (i = 0; i < BTRFS_SUPER_MIRROR_MAX; i++) { bytenr = btrfs_sb_offset(i); + fprintf(stderr, "trying potential super #%d at bytenr %llu \n", + i, bytenr); + + ret = pread64(fd, &buf, sizeof(buf), bytenr); - if (ret < sizeof(buf)) + if (ret < sizeof(buf)) { + fprintf(stderr, " got only %d bytes instead of %lu\n", + ret, sizeof(buf)); break; + } - if (btrfs_super_bytenr(&buf) != bytenr || - strncmp((char *)(&buf.magic), BTRFS_MAGIC, - sizeof(buf.magic))) + + if (btrfs_super_bytenr(&buf) != bytenr) { + fprintf(stderr, " misplaced block thinks it''s at %llu\n", + btrfs_super_bytenr(&buf)); + continue; + } + + if (strncmp((char *)(&buf.magic), BTRFS_MAGIC, + sizeof(buf.magic))) { + fprintf(stderr, " invalid magic\n"); continue; + } if (i == 0) memcpy(fsid, buf.fsid, sizeof(fsid)); - else if (memcmp(fsid, buf.fsid, sizeof(fsid))) + else if (memcmp(fsid, buf.fsid, sizeof(fsid))) { + uuid_unparse(fsid, up_fsid); + uuid_unparse(buf.fsid, up_ofsid); + fprintf(stderr, " wrong fsid %s expected %s \n", up_fsid, up_ofsid); + continue; + } + + if (btrfs_super_generation(&buf) < transid) { + fprintf(stderr, "super #%d at bytenr %llu has older generation %llu than %llu, skipping\n", + i, bytenr, btrfs_super_generation(&buf), transid); + continue; + } + + if (btrfs_super_generation(&buf) == transid) { + fprintf(stderr, "super #%d at bytenr %llu has same generation %llu than %llu, skipping\n", + i, bytenr, btrfs_super_generation(&buf), transid); - if (btrfs_super_generation(&buf) > transid) { - memcpy(sb, &buf, sizeof(*sb)); - transid = btrfs_super_generation(&buf); + if (memcmp(sb, &buf, sizeof(*sb))) { + fprintf(stderr, " warning: super #%d at bytenr %llu has different contents!\n", + i, bytenr); + } + continue; } + + /* btrfs_super_generation(&buf) > transid */ + fprintf(stderr, "super #%d at bytenr %llu has better generation %llu than %llu, using that\n", + i, bytenr, btrfs_super_generation(&buf), transid); + + memcpy(sb, &buf, sizeof(*sb)); + transid = btrfs_super_generation(&buf); } return transid > 0 ? 0 : -1; @@ -930,7 +1101,12 @@ static int close_all_devices(struct btrfs_fs_info *fs_info) struct list_head *next; struct btrfs_device *device; - return 0; + if (!fs_info->emergency_root_tree) { + /* huh? there was a "return 0" sitting here. Yes we leaked fd''s. + Leaving it on when not doing funky desperate things. + */ + return 0; + } list = &fs_info->fs_devices->devices; list_for_each(next, list) { diff --git a/disk-io.h b/disk-io.h index 49e5692..0af98b4 100644 --- a/disk-io.h +++ b/disk-io.h @@ -64,6 +64,7 @@ int btrfs_buffer_uptodate(struct extent_buffer *buf, u64 parent_transid); int btrfs_set_buffer_uptodate(struct extent_buffer *buf); int wait_on_tree_block_writeback(struct btrfs_root *root, struct extent_buffer *buf); +int write_all_supers(struct btrfs_root *root); u32 btrfs_csum_data(struct btrfs_root *root, char *data, u32 seed, size_t len); void btrfs_csum_final(u32 crc, char *result); --=-TovPMUBbcwNB5a+bSJwV-- -- To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
Shawn Stricker
2011-Jan-27 06:18 UTC
Re: full btrfs partition, became unmountable (+ a solution that thankfully worked for me)
any chance of getting a little more informative output? I started the command at about 2250 Eastern and now at 0117 Eastern the command is still running and all of the attached output happened in the first few minutes (under 5). On Jan 26, 2011, at 2:46 AM, Cyrille Chépélov wrote:> Le mardi 25 janvier 2011 à 23:38 -0500, Shawn Stricker a écrit : >> Not sure where you pulled your source from but a fresh checkout of either master or next of git.kernel.org/pub/scm/linux/kernel/git/mason/btrfs-progs-unstable.git does not compile properly. >> They both fail with >> >> cc1: warnings being treated as errors >> disk-io.c: In function ‘btrfs_read_dev_super’: >> disk-io.c:937: error: format ‘%lu’ expects type ‘long unsigned int’, but argument 4 has type ‘unsigned int’ >> disk-io.c:957: error: implicit declaration of function ‘uuid_unparse’ >> >> am I patching/compiling from the wrong source or is there something I am missing? > > uh, I had been compiling with CFLAGS=-g, where the makefile specifies > "-O2 -Werror" > > -Werror causes warnings to be treated as errors, which is a good thing > in a way (makes sure stuff as this gets caught :) ) > > fixes are: > * line 937 (patched), should be %llu instead of %lu > * line 957, there should be a prototype for uuid_unparse(), most > certainly by including <uuid/uuid.h> > > please try this patch instead. > > Thanks for the feedback! > > -- Cyrille > >> On Jan 25, 2011, at 1:46 PM, Cyrille Chépélov wrote: >> >>> Hello all, >>> >>> Last Friday, the /var and /home partition on one of my appliances became >>> full. This should normally not be much of a problem, except that after >>> the incident, I had been unable to mount the partition back again. >>> >>> The appliance runs 2.6.32 as provided by Debian during the last two >>> months. >>> The rescue computer runs 2.6.37; both exhibited the same behaviour at >>> mount: an infinite loop-and-abort cycle (I unfortunately did not write >>> down the exact messages, but in a nutshell, there was not enough free >>> space to replay the log, so it aborted). >>> >>> After pulling the SD card (yes) to break the loop, I ended up with a >>> corrupt file system. Any attempt to mount, debug or fsck (using >>> btrfs-tools 0.19+20100601 as shipped by Debian, or compiled from git >>> 1b444cd2e6ab8dcafdd) aborted with the following message: >>> btrfs-debug-tree: disk-io.c:741: open_ctree_fd: Assertion `!(! >>> tree_root->node)'' failed. >>> >>> After much scavenging on the disk image, I finally managed to recover, >>> using the (dirty) patch attached here. Since apparently other people had >>> similar issues, I''m posting it in the hope it might be useful. >>> >>> -- Cyrille >>> >>> PS: Chris, if btrfs-images of "before" and "after" my butcher fix would >>> be useful to you, just let me know. >>> <scavenge.patch> >> > > <scavenge-2.patch>
Cyrille Chépélov
2011-Jan-27 06:52 UTC
Re: full btrfs partition, became unmountable (+ a solution that thankfully worked for me)
Hello Shawn, it''s now performing a sequential read of the volume, which will probably take significantly more time for you than for me (where I was dealing with an image of a 16GB SD card, stored on a recent mechanical SATA disk). I''m a bit confused by what happens while reading the "potential supers". At first the blocks appear valid, then they are all "misplaced" (meaning the bytenr field != the bytenr from which the block has been read, IOW the block is most probably not part of btrfs structures, from what I understand). From the output before the "will attempt to find useful trees" messages, it seems btrfsck is now doing a sequential read not just of /dev/sde, but also every single block device ? disk-io.c: try_emergency_tree_fixup() is probably now a bit too silent for your use case at the moment. You might want to uncomment the commented out fprintf there; this will make it very verbose (an extra line per structure block) but will provide clues as to where on disk is it working. -- Cyrille Le jeudi 27 janvier 2011 à 01:18 -0500, Shawn Stricker a écrit :> any chance of getting a little more informative output? > I started the command at about 2250 Eastern and now at 0117 Eastern the command is still running and all of the attached output happened in the first few minutes (under 5). > On Jan 26, 2011, at 2:46 AM, Cyrille Chépélov wrote: > > > Le mardi 25 janvier 2011 à 23:38 -0500, Shawn Stricker a écrit : > >> Not sure where you pulled your source from but a fresh checkout of either master or next of git.kernel.org/pub/scm/linux/kernel/git/mason/btrfs-progs-unstable.git does not compile properly. > >> They both fail with > >> > >> cc1: warnings being treated as errors > >> disk-io.c: In function ‘btrfs_read_dev_super’: > >> disk-io.c:937: error: format ‘%lu’ expects type ‘long unsigned int’, but argument 4 has type ‘unsigned int’ > >> disk-io.c:957: error: implicit declaration of function ‘uuid_unparse’ > >> > >> am I patching/compiling from the wrong source or is there something I am missing? > > > > uh, I had been compiling with CFLAGS=-g, where the makefile specifies > > "-O2 -Werror" > > > > -Werror causes warnings to be treated as errors, which is a good thing > > in a way (makes sure stuff as this gets caught :) ) > > > > fixes are: > > * line 937 (patched), should be %llu instead of %lu > > * line 957, there should be a prototype for uuid_unparse(), most > > certainly by including <uuid/uuid.h> > > > > please try this patch instead. > > > > Thanks for the feedback! > > > > -- Cyrille > > > >> On Jan 25, 2011, at 1:46 PM, Cyrille Chépélov wrote: > >> > >>> Hello all, > >>> > >>> Last Friday, the /var and /home partition on one of my appliances became > >>> full. This should normally not be much of a problem, except that after > >>> the incident, I had been unable to mount the partition back again. > >>> > >>> The appliance runs 2.6.32 as provided by Debian during the last two > >>> months. > >>> The rescue computer runs 2.6.37; both exhibited the same behaviour at > >>> mount: an infinite loop-and-abort cycle (I unfortunately did not write > >>> down the exact messages, but in a nutshell, there was not enough free > >>> space to replay the log, so it aborted). > >>> > >>> After pulling the SD card (yes) to break the loop, I ended up with a > >>> corrupt file system. Any attempt to mount, debug or fsck (using > >>> btrfs-tools 0.19+20100601 as shipped by Debian, or compiled from git > >>> 1b444cd2e6ab8dcafdd) aborted with the following message: > >>> btrfs-debug-tree: disk-io.c:741: open_ctree_fd: Assertion `!(! > >>> tree_root->node)'' failed. > >>> > >>> After much scavenging on the disk image, I finally managed to recover, > >>> using the (dirty) patch attached here. Since apparently other people had > >>> similar issues, I''m posting it in the hope it might be useful. > >>> > >>> -- Cyrille > >>> > >>> PS: Chris, if btrfs-images of "before" and "after" my butcher fix would > >>> be useful to you, just let me know. > >>> <scavenge.patch> > >> > > > > <scavenge-2.patch> >-- To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html