diff options
Diffstat (limited to 'fs/bcachefs/fs-io.c')
-rw-r--r-- | fs/bcachefs/fs-io.c | 28 |
1 files changed, 17 insertions, 11 deletions
diff --git a/fs/bcachefs/fs-io.c b/fs/bcachefs/fs-io.c index acb2135a3235..7823141ea98b 100644 --- a/fs/bcachefs/fs-io.c +++ b/fs/bcachefs/fs-io.c @@ -78,7 +78,13 @@ static inline struct folio_vec bio_iter_iovec_folio(struct bio *bio, #define bio_for_each_folio(bvl, bio, iter) \ __bio_for_each_folio(bvl, bio, iter, (bio)->bi_iter) -static inline loff_t folio_end_pos(struct folio *folio) +/* + * Use u64 for the end pos and sector helpers because if the folio covers the + * max supported range of the mapping, the start offset of the next folio + * overflows loff_t. This breaks much of the range based processing in the + * buffered write path. + */ +static inline u64 folio_end_pos(struct folio *folio) { return folio_pos(folio) + folio_size(folio); } @@ -93,7 +99,7 @@ static inline loff_t folio_sector(struct folio *folio) return folio_pos(folio) >> 9; } -static inline loff_t folio_end_sector(struct folio *folio) +static inline u64 folio_end_sector(struct folio *folio) { return folio_end_pos(folio) >> 9; } @@ -101,12 +107,12 @@ static inline loff_t folio_end_sector(struct folio *folio) typedef DARRAY(struct folio *) folios; static int filemap_get_contig_folios_d(struct address_space *mapping, - loff_t start, loff_t end, + loff_t start, u64 end, int fgp_flags, gfp_t gfp, folios *folios) { struct folio *f; - loff_t pos = start; + u64 pos = start; int ret = 0; while (pos < end) { @@ -1859,7 +1865,7 @@ static int __bch2_buffered_write(struct bch_inode_info *inode, folios folios; struct folio **fi, *f; unsigned copied = 0, f_offset; - loff_t end = pos + len, f_pos; + u64 end = pos + len, f_pos; loff_t last_folio_pos = inode->v.i_size; int ret = 0; @@ -1901,7 +1907,7 @@ static int __bch2_buffered_write(struct bch_inode_info *inode, f_offset = pos - folio_pos(darray_first(folios)); darray_for_each(folios, fi) { struct folio *f = *fi; - unsigned f_len = min(end, folio_end_pos(f)) - f_pos; + u64 f_len = min(end, folio_end_pos(f)) - f_pos; if (!bch2_folio_create(f, __GFP_NOFAIL)->uptodate) { ret = bch2_folio_set(c, inode_inum(inode), fi, @@ -1940,7 +1946,7 @@ static int __bch2_buffered_write(struct bch_inode_info *inode, f_offset = pos - folio_pos(darray_first(folios)); darray_for_each(folios, fi) { struct folio *f = *fi; - unsigned f_len = min(end, folio_end_pos(f)) - f_pos; + u64 f_len = min(end, folio_end_pos(f)) - f_pos; unsigned f_copied = copy_page_from_iter_atomic(&f->page, f_offset, f_len, iter); if (!f_copied) { @@ -1982,7 +1988,7 @@ static int __bch2_buffered_write(struct bch_inode_info *inode, f_offset = pos - folio_pos(darray_first(folios)); darray_for_each(folios, fi) { struct folio *f = *fi; - unsigned f_len = min(end, folio_end_pos(f)) - f_pos; + u64 f_len = min(end, folio_end_pos(f)) - f_pos; if (!folio_test_uptodate(f)) folio_mark_uptodate(f); @@ -2818,7 +2824,7 @@ static int __bch2_truncate_folio(struct bch_inode_info *inode, struct folio *folio; s64 i_sectors_delta = 0; int ret = 0; - loff_t end_pos; + u64 end_pos; folio = filemap_lock_folio(mapping, index); if (!folio) { @@ -2844,7 +2850,7 @@ static int __bch2_truncate_folio(struct bch_inode_info *inode, BUG_ON(end <= folio_pos(folio)); start_offset = max(start, folio_pos(folio)) - folio_pos(folio); - end_offset = min(end, folio_end_pos(folio)) - folio_pos(folio); + end_offset = min_t(u64, end, folio_end_pos(folio)) - folio_pos(folio); /* Folio boundary? Nothing to do */ if (start_offset == 0 && @@ -2895,7 +2901,7 @@ static int __bch2_truncate_folio(struct bch_inode_info *inode, WARN_ON_ONCE(folio_pos(folio) >= inode->v.i_size); end_pos = folio_end_pos(folio); if (inode->v.i_size > folio_pos(folio)) - end_pos = min(inode->v.i_size, end_pos); + end_pos = min_t(u64, inode->v.i_size, end_pos); ret = s->s[(end_pos - folio_pos(folio) - 1) >> 9].state >= SECTOR_dirty; folio_zero_segment(folio, start_offset, end_offset); |