This is an automated email from the git hooks/post-receive script.
andyp pushed a commit to branch andyp-extents in repository gfs2-utils.
commit 8747d9003ae4219ec6a2b8790075ec3d5fd0aa13 Author: Andreas Gruenbacher agruenba@redhat.com AuthorDate: Wed Apr 21 20:52:31 2021 +0200
fsck.gfs2: Add partial extent-based inode support
Signed-off-by: Andreas Gruenbacher agruenba@redhat.com --- gfs2/fsck/metawalk.c | 185 ++++++++++++++++++++++++++++++++++++++------------- 1 file changed, 140 insertions(+), 45 deletions(-)
diff --git a/gfs2/fsck/metawalk.c b/gfs2/fsck/metawalk.c index 8b721015..62476ab8 100644 --- a/gfs2/fsck/metawalk.c +++ b/gfs2/fsck/metawalk.c @@ -1446,38 +1446,93 @@ static int metawalk_check_data(struct gfs2_inode *ip, struct metawalk_fxns *pass struct gfs2_buffer_head *bh, unsigned int height, uint64_t *blks_checked, struct error_block *error_blk) { - int error = 0, rc = 0; + int error = 0, rc; uint64_t block; - __be64 *ptr_start = (uint64_t *)(bh->b_data + hdr_size(bh, height)); - __be64 *ptr_end = (uint64_t *)(bh->b_data + ip->i_sbd->bsize); - __be64 *ptr; uint64_t metablock = bh->b_blocknr;
/* If there isn't much pointer corruption check the pointers */ log_debug("Processing data blocks for inode 0x%llx, metadata block 0x%llx.\n", (unsigned long long)ip->i_di.di_num.no_addr, (unsigned long long)metablock); - for (ptr = ptr_start ; ptr < ptr_end && !fsck_abort; ptr++) { - if (!*ptr) - continue;
- if (skip_this_pass || fsck_abort) - return error; - block = be64_to_cpu(*ptr); - /* It's important that we don't call valid_block() and - bypass calling check_data on invalid blocks because that - would defeat the rangecheck_block related functions in - pass1. Therefore the individual check_data functions - should do a range check. */ - rc = pass->check_data(ip, metablock, block, pass->private, - bh, ptr); - if (rc && (!error || (rc < error))) { - report_data_error(metablock, (char *)ptr - bh->b_data, block, error_blk, rc, error); - error = rc; + /* It's important that we don't call valid_block() and + bypass calling check_data on invalid blocks because that + would defeat the rangecheck_block related functions in + pass1. Therefore the individual check_data functions + should do a range check. */ + + if (ip->i_di.di_flags & GFS2_DIF_EXTENTS) { + unsigned int max_extents = (ip->i_sbd->bsize - hdr_size(bh, height)) / + sizeof(struct gfs2_extent); + struct gfs2_extent_header *eh = (struct gfs2_extent_header *) + (bh->b_data + hdr_size(bh, height)); + struct gfs2_extent *first_ex = (struct gfs2_extent *)(eh + 1); + unsigned int extents = be16_to_cpu(eh->eh_entries); + struct gfs2_extent *ex; + + if (extents > max_extents) { + log_err(_("Inode %"PRId64" (0x%"PRIx64") has a " + "number of extents greater than %u.\n"), + ip->i_di.di_num.no_addr, + ip->i_di.di_num.no_addr, + max_extents); + return META_ERROR; + } + + for (ex = first_ex; ex != first_ex + extents && !fsck_abort; ex++) { + uint64_t first_block = be64_to_cpu(ex->ex_addr); + unsigned int len = be16_to_cpu(ex->ex_len); + + /* if (!ex->ex_addr || !ex->ex_len || too_big(ex->ex_len)) */ + + for (block = first_block; block < first_block + len; block++) { + __be64 dummy = cpu_to_be64(-1); + + /* + * FIXME: We cannot relocate or free blocks in + * the middle of an extent without rewriting + * the extent tree, which may require an + * allocation; handling this in clone_data() + * won't work. + */ + + if (skip_this_pass || fsck_abort) + return error; + rc = pass->check_data(ip, metablock, block, pass->private, + bh, &dummy); + if (rc && (!error || (rc < error))) { + report_data_error(metablock, (char *)ex - bh->b_data, + block, error_blk, rc, error); + error = rc; + } + if (rc < 0) + return rc; + (*blks_checked)++; + } + } + } else { + __be64 *ptr_start = (uint64_t *)(bh->b_data + hdr_size(bh, height)); + __be64 *ptr_end = (uint64_t *)(bh->b_data + ip->i_sbd->bsize); + __be64 *ptr; + + for (ptr = ptr_start ; ptr < ptr_end && !fsck_abort; ptr++) { + if (!*ptr) + continue; + + if (skip_this_pass || fsck_abort) + return error; + block = be64_to_cpu(*ptr); + rc = pass->check_data(ip, metablock, block, pass->private, + bh, ptr); + if (rc && (!error || (rc < error))) { + report_data_error(metablock, (char *)ptr - bh->b_data, + block, error_blk, rc, error); + error = rc; + } + if (rc < 0) + return rc; + (*blks_checked)++; } - if (rc < 0) - return rc; - (*blks_checked)++; } return error; } @@ -1515,28 +1570,54 @@ static int undo_check_data(struct gfs2_inode *ip, struct metawalk_fxns *pass, struct gfs2_buffer_head *bh, unsigned int height, struct error_block *error_blk, int error) { - __be64 *ptr_start = (uint64_t *)(bh->b_data + hdr_size(bh, height)); - __be64 *ptr_end = (uint64_t *)(bh->b_data + ip->i_sbd->bsize); - __be64 *ptr; uint64_t metablock = bh->b_blocknr; - int rc = 0; + int rc; uint64_t block; int found_error_blk = 0;
/* If there isn't much pointer corruption check the pointers */ - for (ptr = ptr_start ; ptr < ptr_end && !fsck_abort; ptr++) { - if (!*ptr) - continue;
- if (skip_this_pass || fsck_abort) - return 1; - block = be64_to_cpu(*ptr); - if (report_undo_data_error(metablock, (char *)ptr - bh->b_data, - block, error_blk, &found_error_blk, error)) - return 1; - rc = pass->undo_check_data(ip, block, pass->private); - if (rc < 0) - return rc; + if (ip->i_di.di_flags & GFS2_DIF_EXTENTS) { + struct gfs2_extent_header *eh = (struct gfs2_extent_header *) + (bh->b_data + hdr_size(bh, height)); + struct gfs2_extent *first_ex = (struct gfs2_extent *)(eh + 1); + unsigned int extents = be16_to_cpu(eh->eh_entries); + struct gfs2_extent *ex; + + for (ex = first_ex; ex != first_ex + extents && !fsck_abort; ex++) { + uint64_t first_block = be64_to_cpu(ex->ex_addr); + unsigned int len = be16_to_cpu(ex->ex_len); + + for (block = first_block; block < first_block + len; block++) { + if (skip_this_pass || fsck_abort) + return 1; + if (report_undo_data_error(metablock, (char *)ex - bh->b_data, + block, error_blk, &found_error_blk, error)) + return 1; + rc = pass->undo_check_data(ip, block, pass->private); + if (rc < 0) + return rc; + } + } + } else { + __be64 *ptr_start = (uint64_t *)(bh->b_data + hdr_size(bh, height)); + __be64 *ptr_end = (uint64_t *)(bh->b_data + ip->i_sbd->bsize); + __be64 *ptr; + + for (ptr = ptr_start ; ptr < ptr_end && !fsck_abort; ptr++) { + if (!*ptr) + continue; + + if (skip_this_pass || fsck_abort) + return 1; + block = be64_to_cpu(*ptr); + if (report_undo_data_error(metablock, (char *)ptr - bh->b_data, + block, error_blk, &found_error_blk, error)) + return 1; + rc = pass->undo_check_data(ip, block, pass->private); + if (rc < 0) + return rc; + } } return found_error_blk; } @@ -1562,7 +1643,7 @@ int check_metatree(struct gfs2_inode *ip, struct metawalk_fxns *pass) struct gfs2_buffer_head *bh; unsigned int i; uint64_t blks_checked = 0; - int error, rc; + int error = 0, rc; int metadata_clean = 0; struct error_block error_blk = {0, 0, 0}; int hit_error_blk = 0; @@ -1574,11 +1655,25 @@ int check_metatree(struct gfs2_inode *ip, struct metawalk_fxns *pass) for (i = 0; i <= height; i++) osi_list_init(&metalist[i]);
- /* create and check the metadata list for each height */ - error = build_and_check_metalist(ip, metalist, pass); - if (error) { - stack; - goto undo_metalist; + if (ip->i_di.di_flags & GFS2_DIF_EXTENTS) { + if (height > 1) { + log_err(_("Inode %"PRId64" (0x%"PRIx64") has an " + "extent tree height greater than %u.\n"), + ip->i_di.di_num.no_addr, + ip->i_di.di_num.no_addr, + 1); + error = META_ERROR; + goto undo_metalist; + } + /* make sure we check the inode below */ + osi_list_add(&ip->i_bh->b_altlist, metalist); + } else { + /* create and check the metadata list for each height */ + error = build_and_check_metalist(ip, metalist, pass); + if (error) { + stack; + goto undo_metalist; + } }
metadata_clean = 1;
cluster-commits@lists.fedorahosted.org
-
pagure＠pagure.io