From: Darrick J. Wong <darrick.wong@xxxxxxxxxx> libxfs_writebuf is not a well named function -- it marks the buffer dirty and then releases the caller's reference. The actual write comes when the cache is flushed, either because someone explicitly told the cache to flush or because we started buffer reclaim. Make the buffer release explicit in the callers and rename the function to say what it actually does -- it marks the buffer dirty outside of transaction context. Signed-off-by: Darrick J. Wong <darrick.wong@xxxxxxxxxx> --- libxfs/libxfs_io.h | 10 +++++----- libxfs/rdwr.c | 21 ++++++++++++++------- libxfs/trans.c | 3 ++- mkfs/proto.c | 6 ++++-- mkfs/xfs_mkfs.c | 21 ++++++++++++++------- repair/attr_repair.c | 17 +++++++++-------- repair/da_util.c | 12 ++++++++---- repair/dino_chunks.c | 6 ++++-- repair/dinode.c | 12 ++++++++---- repair/dir2.c | 9 ++++++--- repair/phase3.c | 6 ++++-- repair/phase5.c | 42 ++++++++++++++++++++++++++++-------------- repair/rmap.c | 3 ++- repair/scan.c | 21 ++++++++++++++------- repair/xfs_repair.c | 3 ++- 15 files changed, 124 insertions(+), 68 deletions(-) diff --git a/libxfs/libxfs_io.h b/libxfs/libxfs_io.h index 6598dba7..ad8acc1e 100644 --- a/libxfs/libxfs_io.h +++ b/libxfs/libxfs_io.h @@ -138,8 +138,8 @@ extern struct cache_operations libxfs_bcache_operations; #define libxfs_readbuf_map(dev, map, nmaps, flags, ops) \ libxfs_trace_readbuf_map(__FUNCTION__, __FILE__, __LINE__, \ (dev), (map), (nmaps), (flags), (ops)) -#define libxfs_writebuf(buf, flags) \ - libxfs_trace_writebuf(__FUNCTION__, __FILE__, __LINE__, \ +#define libxfs_buf_dirty(buf, flags) \ + libxfs_trace_dirtybuf(__FUNCTION__, __FILE__, __LINE__, \ (buf), (flags)) #define libxfs_buf_get(dev, daddr, len) \ libxfs_trace_getbuf(__FUNCTION__, __FILE__, __LINE__, \ @@ -159,8 +159,8 @@ struct xfs_buf *libxfs_trace_readbuf(const char *func, const char *file, extern xfs_buf_t *libxfs_trace_readbuf_map(const char *, const char *, int, struct xfs_buftarg *, struct xfs_buf_map *, int, int, const struct xfs_buf_ops *); -extern int libxfs_trace_writebuf(const char *, const char *, int, - xfs_buf_t *, int); +void libxfs_trace_dirtybuf(const char *func, const char *file, int line, + struct xfs_buf *bp, int flags); struct xfs_buf *libxfs_trace_getbuf(const char *func, const char *file, int line, struct xfs_buftarg *btp, xfs_daddr_t daddr, size_t len); @@ -175,7 +175,7 @@ extern void libxfs_trace_putbuf (const char *, const char *, int, extern xfs_buf_t *libxfs_readbuf_map(struct xfs_buftarg *, struct xfs_buf_map *, int, int, const struct xfs_buf_ops *); -extern int libxfs_writebuf(xfs_buf_t *, int); +void libxfs_buf_dirty(struct xfs_buf *bp, int flags); extern xfs_buf_t *libxfs_getbuf_map(struct xfs_buftarg *, struct xfs_buf_map *, int, int); extern xfs_buf_t *libxfs_getbuf_flags(struct xfs_buftarg *, xfs_daddr_t, diff --git a/libxfs/rdwr.c b/libxfs/rdwr.c index 20a8b0ce..af363bef 100644 --- a/libxfs/rdwr.c +++ b/libxfs/rdwr.c @@ -425,11 +425,16 @@ libxfs_trace_readbuf( return bp; } -int -libxfs_trace_writebuf(const char *func, const char *file, int line, xfs_buf_t *bp, int flags) +void +libxfs_trace_dirtybuf( + const char *func, + const char *file, + int line, + struct xfs_buf *bp, + int flags) { __add_trace(bp, func, file, line); - return libxfs_writebuf(bp, flags); + libxfs_buf_dirty(bp, flags); } struct xfs_buf * @@ -1233,8 +1238,12 @@ libxfs_writebuf_int(xfs_buf_t *bp, int flags) return 0; } -int -libxfs_writebuf( +/* + * Mark a buffer dirty. The dirty data will be written out when the cache + * is flushed (or at release time if the buffer is uncached). + */ +void +libxfs_buf_dirty( struct xfs_buf *bp, int flags) { @@ -1256,8 +1265,6 @@ libxfs_writebuf( bp->b_error = 0; bp->b_flags &= ~LIBXFS_B_STALE; bp->b_flags |= bflags; - libxfs_buf_relse(bp); - return 0; } void diff --git a/libxfs/trans.c b/libxfs/trans.c index 59cb897f..4c208422 100644 --- a/libxfs/trans.c +++ b/libxfs/trans.c @@ -843,7 +843,8 @@ inode_item_done( goto free; } - libxfs_writebuf(bp, 0); + libxfs_buf_dirty(bp, 0); + libxfs_buf_relse(bp); free: xfs_inode_item_put(iip); } diff --git a/mkfs/proto.c b/mkfs/proto.c index 0025fa08..de5ae306 100644 --- a/mkfs/proto.c +++ b/mkfs/proto.c @@ -261,8 +261,10 @@ newfile( memset((char *)bp->b_addr + len, 0, bp->b_bcount - len); if (logit) libxfs_trans_log_buf(tp, bp, 0, bp->b_bcount - 1); - else - libxfs_writebuf(bp, LIBXFS_WRITEBUF_FAIL_EXIT); + else { + libxfs_buf_dirty(bp, LIBXFS_WRITEBUF_FAIL_EXIT); + libxfs_buf_relse(bp); + } } ip->i_d.di_size = len; return flags; diff --git a/mkfs/xfs_mkfs.c b/mkfs/xfs_mkfs.c index f58f235d..b50b8b3a 100644 --- a/mkfs/xfs_mkfs.c +++ b/mkfs/xfs_mkfs.c @@ -3462,7 +3462,8 @@ prepare_devices( buf = get_write_buf(mp->m_ddev_targp, (xi->dsize - whack_blks), whack_blks); memset(buf->b_addr, 0, WHACK_SIZE); - libxfs_writebuf(buf, LIBXFS_WRITEBUF_FAIL_EXIT); + libxfs_buf_dirty(buf, LIBXFS_WRITEBUF_FAIL_EXIT); + libxfs_buf_relse(buf); /* * Now zero out the beginning of the device, to obliterate any old @@ -3472,7 +3473,8 @@ prepare_devices( */ buf = get_write_buf(mp->m_ddev_targp, 0, whack_blks); memset(buf->b_addr, 0, WHACK_SIZE); - libxfs_writebuf(buf, LIBXFS_WRITEBUF_FAIL_EXIT); + libxfs_buf_dirty(buf, LIBXFS_WRITEBUF_FAIL_EXIT); + libxfs_buf_relse(buf); /* OK, now write the superblock... */ buf = get_write_buf(mp->m_ddev_targp, XFS_SB_DADDR, @@ -3480,7 +3482,8 @@ prepare_devices( buf->b_ops = &xfs_sb_buf_ops; memset(buf->b_addr, 0, cfg->sectorsize); libxfs_sb_to_disk(buf->b_addr, sbp); - libxfs_writebuf(buf, LIBXFS_WRITEBUF_FAIL_EXIT); + libxfs_buf_dirty(buf, LIBXFS_WRITEBUF_FAIL_EXIT); + libxfs_buf_relse(buf); /* ...and zero the log.... */ lsunit = sbp->sb_logsunit; @@ -3499,7 +3502,8 @@ prepare_devices( XFS_FSB_TO_BB(mp, cfg->rtblocks - 1LL), BTOBB(cfg->blocksize)); memset(buf->b_addr, 0, cfg->blocksize); - libxfs_writebuf(buf, LIBXFS_WRITEBUF_FAIL_EXIT); + libxfs_buf_dirty(buf, LIBXFS_WRITEBUF_FAIL_EXIT); + libxfs_buf_relse(buf); } } @@ -3591,7 +3595,8 @@ rewrite_secondary_superblocks( XFS_FSS_TO_BB(mp, 1), LIBXFS_READBUF_FAIL_EXIT, &xfs_sb_buf_ops); XFS_BUF_TO_SBP(buf)->sb_rootino = cpu_to_be64(mp->m_sb.sb_rootino); - libxfs_writebuf(buf, LIBXFS_WRITEBUF_FAIL_EXIT); + libxfs_buf_dirty(buf, LIBXFS_WRITEBUF_FAIL_EXIT); + libxfs_buf_relse(buf); /* and one in the middle for luck if there's enough AGs for that */ if (mp->m_sb.sb_agcount <= 2) @@ -3603,7 +3608,8 @@ rewrite_secondary_superblocks( XFS_FSS_TO_BB(mp, 1), LIBXFS_READBUF_FAIL_EXIT, &xfs_sb_buf_ops); XFS_BUF_TO_SBP(buf)->sb_rootino = cpu_to_be64(mp->m_sb.sb_rootino); - libxfs_writebuf(buf, LIBXFS_WRITEBUF_FAIL_EXIT); + libxfs_buf_dirty(buf, LIBXFS_WRITEBUF_FAIL_EXIT); + libxfs_buf_relse(buf); } static void @@ -3951,7 +3957,8 @@ main( if (!buf || buf->b_error) exit(1); (XFS_BUF_TO_SBP(buf))->sb_inprogress = 0; - libxfs_writebuf(buf, LIBXFS_WRITEBUF_FAIL_EXIT); + libxfs_buf_dirty(buf, LIBXFS_WRITEBUF_FAIL_EXIT); + libxfs_buf_relse(buf); /* Make sure our new fs made it to stable storage. */ libxfs_flush_devices(mp, &d, &l, &r); diff --git a/repair/attr_repair.c b/repair/attr_repair.c index cc20b1a1..fb64b0be 100644 --- a/repair/attr_repair.c +++ b/repair/attr_repair.c @@ -835,8 +835,10 @@ process_leaf_attr_level(xfs_mount_t *mp, if (!no_modify && bp->b_error == -EFSBADCRC) repair++; - if (repair && !no_modify) - libxfs_writebuf(bp, 0); + if (repair && !no_modify) { + libxfs_buf_dirty(bp, 0); + libxfs_buf_relse(bp); + } else libxfs_buf_relse(bp); } while (da_bno != 0); @@ -999,9 +1001,8 @@ _("would clear forw/back pointers in block 0 for attributes in inode %" PRIu64 " *repair = *repair || repairlinks; if (*repair && !no_modify) - libxfs_writebuf(bp, 0); - else - libxfs_buf_relse(bp); + libxfs_buf_dirty(bp, 0); + libxfs_buf_relse(bp); return 0; } @@ -1043,9 +1044,9 @@ _("would clear forw/back pointers in block 0 for attributes in inode %" PRIu64 " /* must do this now, to release block 0 before the traversal */ if ((*repair || repairlinks) && !no_modify) { *repair = 1; - libxfs_writebuf(bp, 0); - } else - libxfs_buf_relse(bp); + libxfs_buf_dirty(bp, 0); + } + libxfs_buf_relse(bp); error = process_node_attr(mp, ino, dip, blkmap); /* + repair */ if (error) *repair = 0; diff --git a/repair/da_util.c b/repair/da_util.c index c02d621c..d1e17ec3 100644 --- a/repair/da_util.c +++ b/repair/da_util.c @@ -403,8 +403,10 @@ _("would correct bad hashval in non-leaf %s block\n" ASSERT(cursor->level[this_level].dirty == 0 || (cursor->level[this_level].dirty && !no_modify)); - if (cursor->level[this_level].dirty && !no_modify) - libxfs_writebuf(cursor->level[this_level].bp, 0); + if (cursor->level[this_level].dirty && !no_modify) { + libxfs_buf_dirty(cursor->level[this_level].bp, 0); + libxfs_buf_relse(cursor->level[this_level].bp); + } else libxfs_buf_relse(cursor->level[this_level].bp); @@ -619,8 +621,10 @@ _("bad level %d in %s block %u for inode %" PRIu64 "\n"), cursor->level[this_level].bp->b_error == -EFSBADCRC) cursor->level[this_level].dirty = 1; - if (cursor->level[this_level].dirty && !no_modify) - libxfs_writebuf(cursor->level[this_level].bp, 0); + if (cursor->level[this_level].dirty && !no_modify) { + libxfs_buf_dirty(cursor->level[this_level].bp, 0); + libxfs_buf_relse(cursor->level[this_level].bp); + } else libxfs_buf_relse(cursor->level[this_level].bp); diff --git a/repair/dino_chunks.c b/repair/dino_chunks.c index 76e9f773..1378fc59 100644 --- a/repair/dino_chunks.c +++ b/repair/dino_chunks.c @@ -939,8 +939,10 @@ process_inode_chunk( bplist[bp_index], (long long) XFS_BUF_ADDR(bplist[bp_index]), agno); - if (dirty && !no_modify) - libxfs_writebuf(bplist[bp_index], 0); + if (dirty && !no_modify) { + libxfs_buf_dirty(bplist[bp_index], 0); + libxfs_buf_relse(bplist[bp_index]); + } else libxfs_buf_relse(bplist[bp_index]); } diff --git a/repair/dinode.c b/repair/dinode.c index 848eac09..276dd2e1 100644 --- a/repair/dinode.c +++ b/repair/dinode.c @@ -1231,8 +1231,10 @@ _("cannot read inode %" PRIu64 ", file block %" PRIu64 ", disk block %" PRIu64 " } } - if (writebuf && !no_modify) - libxfs_writebuf(bp, 0); + if (writebuf && !no_modify) { + libxfs_buf_dirty(bp, 0); + libxfs_buf_relse(bp); + } else libxfs_buf_relse(bp); } @@ -1328,8 +1330,10 @@ _("bad symlink header ino %" PRIu64 ", file block %d, disk block %" PRIu64 "\n") offset += byte_cnt; i++; - if (badcrc && !no_modify) - libxfs_writebuf(bp, 0); + if (badcrc && !no_modify) { + libxfs_buf_dirty(bp, 0); + libxfs_buf_relse(bp); + } else libxfs_buf_relse(bp); } diff --git a/repair/dir2.c b/repair/dir2.c index 769e341c..a8b8de58 100644 --- a/repair/dir2.c +++ b/repair/dir2.c @@ -1010,7 +1010,8 @@ _("bad directory block magic # %#x in block %u for directory inode %" PRIu64 "\n dirty = 1; if (dirty && !no_modify) { *repair = 1; - libxfs_writebuf(bp, 0); + libxfs_buf_dirty(bp, 0); + libxfs_buf_relse(bp); } else libxfs_buf_relse(bp); return rval; @@ -1180,7 +1181,8 @@ _("bad sibling back pointer for block %u in directory inode %" PRIu64 "\n"), ASSERT(buf_dirty == 0 || (buf_dirty && !no_modify)); if (buf_dirty && !no_modify) { *repair = 1; - libxfs_writebuf(bp, 0); + libxfs_buf_dirty(bp, 0); + libxfs_buf_relse(bp); } else libxfs_buf_relse(bp); } while (da_bno != 0); @@ -1339,7 +1341,8 @@ _("bad directory block magic # %#x in block %" PRIu64 " for directory inode %" P } if (dirty && !no_modify) { *repair = 1; - libxfs_writebuf(bp, 0); + libxfs_buf_dirty(bp, 0); + libxfs_buf_relse(bp); } else libxfs_buf_relse(bp); } diff --git a/repair/phase3.c b/repair/phase3.c index 886acd1f..396743a4 100644 --- a/repair/phase3.c +++ b/repair/phase3.c @@ -46,8 +46,10 @@ process_agi_unlinked( } } - if (agi_dirty) - libxfs_writebuf(bp, 0); + if (agi_dirty) { + libxfs_buf_dirty(bp, 0); + libxfs_buf_relse(bp); + } else libxfs_buf_relse(bp); } diff --git a/repair/phase5.c b/repair/phase5.c index cdbf6697..561a6b3f 100644 --- a/repair/phase5.c +++ b/repair/phase5.c @@ -321,9 +321,11 @@ write_cursor(bt_status_t *curs) fprintf(stderr, "writing bt prev block %u\n", curs->level[i].prev_agbno); #endif - libxfs_writebuf(curs->level[i].prev_buf_p, 0); + libxfs_buf_dirty(curs->level[i].prev_buf_p, 0); + libxfs_buf_relse(curs->level[i].prev_buf_p); } - libxfs_writebuf(curs->level[i].buf_p, 0); + libxfs_buf_dirty(curs->level[i].buf_p, 0); + libxfs_buf_relse(curs->level[i].buf_p); } } @@ -681,7 +683,8 @@ prop_freespace_cursor(xfs_mount_t *mp, xfs_agnumber_t agno, #endif if (lptr->prev_agbno != NULLAGBLOCK) { ASSERT(lptr->prev_buf_p != NULL); - libxfs_writebuf(lptr->prev_buf_p, 0); + libxfs_buf_dirty(lptr->prev_buf_p, 0); + libxfs_buf_relse(lptr->prev_buf_p); } lptr->prev_agbno = lptr->agbno;; lptr->prev_buf_p = lptr->buf_p; @@ -870,7 +873,8 @@ build_freespace_tree(xfs_mount_t *mp, xfs_agnumber_t agno, lptr->prev_agbno); #endif ASSERT(lptr->prev_agbno != NULLAGBLOCK); - libxfs_writebuf(lptr->prev_buf_p, 0); + libxfs_buf_dirty(lptr->prev_buf_p, 0); + libxfs_buf_relse(lptr->prev_buf_p); } lptr->prev_buf_p = lptr->buf_p; lptr->prev_agbno = lptr->agbno; @@ -1046,7 +1050,8 @@ prop_ino_cursor(xfs_mount_t *mp, xfs_agnumber_t agno, bt_status_t *btree_curs, #endif if (lptr->prev_agbno != NULLAGBLOCK) { ASSERT(lptr->prev_buf_p != NULL); - libxfs_writebuf(lptr->prev_buf_p, 0); + libxfs_buf_dirty(lptr->prev_buf_p, 0); + libxfs_buf_relse(lptr->prev_buf_p); } lptr->prev_agbno = lptr->agbno;; lptr->prev_buf_p = lptr->buf_p; @@ -1137,7 +1142,8 @@ build_agi(xfs_mount_t *mp, xfs_agnumber_t agno, bt_status_t *btree_curs, agi->agi_free_level = cpu_to_be32(finobt_curs->num_levels); } - libxfs_writebuf(agi_buf, 0); + libxfs_buf_dirty(agi_buf, 0); + libxfs_buf_relse(agi_buf); } /* @@ -1299,7 +1305,8 @@ build_ino_tree(xfs_mount_t *mp, xfs_agnumber_t agno, lptr->prev_agbno); #endif ASSERT(lptr->prev_agbno != NULLAGBLOCK); - libxfs_writebuf(lptr->prev_buf_p, 0); + libxfs_buf_dirty(lptr->prev_buf_p, 0); + libxfs_buf_relse(lptr->prev_buf_p); } lptr->prev_buf_p = lptr->buf_p; lptr->prev_agbno = lptr->agbno; @@ -1451,7 +1458,8 @@ prop_rmap_cursor( #endif if (lptr->prev_agbno != NULLAGBLOCK) { ASSERT(lptr->prev_buf_p != NULL); - libxfs_writebuf(lptr->prev_buf_p, 0); + libxfs_buf_dirty(lptr->prev_buf_p, 0); + libxfs_buf_relse(lptr->prev_buf_p); } lptr->prev_agbno = lptr->agbno; lptr->prev_buf_p = lptr->buf_p; @@ -1661,7 +1669,8 @@ _("Insufficient memory to construct reverse-map cursor.")); lptr->prev_agbno); #endif ASSERT(lptr->prev_agbno != NULLAGBLOCK); - libxfs_writebuf(lptr->prev_buf_p, 0); + libxfs_buf_dirty(lptr->prev_buf_p, 0); + libxfs_buf_relse(lptr->prev_buf_p); } lptr->prev_buf_p = lptr->buf_p; lptr->prev_agbno = lptr->agbno; @@ -1801,7 +1810,8 @@ prop_refc_cursor( #endif if (lptr->prev_agbno != NULLAGBLOCK) { ASSERT(lptr->prev_buf_p != NULL); - libxfs_writebuf(lptr->prev_buf_p, 0); + libxfs_buf_dirty(lptr->prev_buf_p, 0); + libxfs_buf_relse(lptr->prev_buf_p); } lptr->prev_agbno = lptr->agbno; lptr->prev_buf_p = lptr->buf_p; @@ -1954,7 +1964,8 @@ _("Insufficient memory to construct refcount cursor.")); lptr->prev_agbno); #endif ASSERT(lptr->prev_agbno != NULLAGBLOCK); - libxfs_writebuf(lptr->prev_buf_p, 0); + libxfs_buf_dirty(lptr->prev_buf_p, 0); + libxfs_buf_relse(lptr->prev_buf_p); } lptr->prev_buf_p = lptr->buf_p; lptr->prev_agbno = lptr->agbno; @@ -2142,7 +2153,8 @@ _("Insufficient memory saving lost blocks.\n")); agf->agf_flcount = 0; } - libxfs_writebuf(agfl_buf, 0); + libxfs_buf_dirty(agfl_buf, 0); + libxfs_buf_relse(agfl_buf); ext_ptr = findbiggest_bcnt_extent(agno); agf->agf_longest = cpu_to_be32((ext_ptr != NULL) ? @@ -2155,7 +2167,8 @@ _("Insufficient memory saving lost blocks.\n")); ASSERT(be32_to_cpu(agf->agf_refcount_root) != be32_to_cpu(agf->agf_roots[XFS_BTNUM_CNTi])); - libxfs_writebuf(agf_buf, 0); + libxfs_buf_dirty(agf_buf, 0); + libxfs_buf_relse(agf_buf); /* * now fix up the free list appropriately @@ -2189,7 +2202,8 @@ sync_sb(xfs_mount_t *mp) update_sb_version(mp); libxfs_sb_to_disk(XFS_BUF_TO_SBP(bp), &mp->m_sb); - libxfs_writebuf(bp, 0); + libxfs_buf_dirty(bp, 0); + libxfs_buf_relse(bp); } /* diff --git a/repair/rmap.c b/repair/rmap.c index bc53e6c0..c8c851a6 100644 --- a/repair/rmap.c +++ b/repair/rmap.c @@ -1231,7 +1231,8 @@ _("setting reflink flag on inode %"PRIu64"\n"), else dino->di_flags2 &= cpu_to_be64(~XFS_DIFLAG2_REFLINK); libxfs_dinode_calc_crc(mp, dino); - libxfs_writebuf(buf, 0); + libxfs_buf_dirty(buf, 0); + libxfs_buf_relse(buf); return 0; } diff --git a/repair/scan.c b/repair/scan.c index f4e4fef5..8b91b27e 100644 --- a/repair/scan.c +++ b/repair/scan.c @@ -151,8 +151,10 @@ scan_lbtree( ASSERT(dirty == 0 || (dirty && !no_modify)); - if ((dirty || badcrc) && !no_modify) - libxfs_writebuf(bp, 0); + if ((dirty || badcrc) && !no_modify) { + libxfs_buf_dirty(bp, 0); + libxfs_buf_relse(bp); + } else libxfs_buf_relse(bp); @@ -2429,13 +2431,17 @@ scan_ag( sb_dirty += (sbbuf->b_error == -EFSBADCRC); } - if (agi_dirty && !no_modify) - libxfs_writebuf(agibuf, 0); + if (agi_dirty && !no_modify) { + libxfs_buf_dirty(agibuf, 0); + libxfs_buf_relse(agibuf); + } else libxfs_buf_relse(agibuf); - if (agf_dirty && !no_modify) - libxfs_writebuf(agfbuf, 0); + if (agf_dirty && !no_modify) { + libxfs_buf_dirty(agfbuf, 0); + libxfs_buf_relse(agfbuf); + } else libxfs_buf_relse(agfbuf); @@ -2443,7 +2449,8 @@ scan_ag( if (agno == 0) memcpy(&mp->m_sb, sb, sizeof(xfs_sb_t)); libxfs_sb_to_disk(XFS_BUF_TO_SBP(sbbuf), sb); - libxfs_writebuf(sbbuf, 0); + libxfs_buf_dirty(sbbuf, 0); + libxfs_buf_relse(sbbuf); } else libxfs_buf_relse(sbbuf); free(sb); diff --git a/repair/xfs_repair.c b/repair/xfs_repair.c index 8642c5cd..2189c9fd 100644 --- a/repair/xfs_repair.c +++ b/repair/xfs_repair.c @@ -1126,7 +1126,8 @@ _("Note - stripe unit (%d) and width (%d) were copied from a backup superblock.\ be32_to_cpu(dsb->sb_unit), be32_to_cpu(dsb->sb_width)); } - libxfs_writebuf(sbp, 0); + libxfs_buf_dirty(sbp, 0); + libxfs_buf_relse(sbp); /* * Done. Flush all cached buffers and inodes first to ensure all