Very large volumes (20TB) would cause an integer overflow in statfs() and display incorrect block counts.
Statfs structure's f_blocks, f_bfree and f_bavail are stored as a u64, but the promotion to 64-bit happens after the shift has been done. Fix this issue by promoting the value before shifting.
The problem can be reproduced by creating a 20TB volume for HFS+, mounting and running statfs() on the mounted volume.
Cc: stable@vger.kernel.org Cc: linux-fsdevel@vger.kernel.org Reviewed-by: Anton Altaparmakov anton@tuxera.com Signed-off-by: Mikael Heino mikael@tuxera.com --- fs/hfsplus/super.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/fs/hfsplus/super.c b/fs/hfsplus/super.c index 948b8aaee33e..00bb23b0ff7d 100644 --- a/fs/hfsplus/super.c +++ b/fs/hfsplus/super.c @@ -322,8 +322,8 @@ static int hfsplus_statfs(struct dentry *dentry, struct kstatfs *buf)
buf->f_type = HFSPLUS_SUPER_MAGIC; buf->f_bsize = sb->s_blocksize; - buf->f_blocks = sbi->total_blocks << sbi->fs_shift; - buf->f_bfree = sbi->free_blocks << sbi->fs_shift; + buf->f_blocks = (u64)sbi->total_blocks << sbi->fs_shift; + buf->f_bfree = (u64)sbi->free_blocks << sbi->fs_shift; buf->f_bavail = buf->f_bfree; buf->f_files = 0xFFFFFFFF; buf->f_ffree = 0xFFFFFFFF - sbi->next_cnid;