On Mon, May 11, 2020 at 10:45:24AM +0800, Zheng Bin wrote: > Construct an img like this: > > dd if=/dev/zero of=xfs.img bs=1M count=20 > mkfs.xfs -d agcount=1 xfs.img > xfs_db -x xfs.img > sb 0 > write fdblocks 0 > agf 0 > write freeblks 0 > write longest 0 > quit > > mount it, df -h /mnt(xfs mount point), will show this: > Filesystem Size Used Avail Use% Mounted on > /dev/loop0 17M -64Z -32K 100% /mnt > > Reported-by: Hulk Robot <hulkci@xxxxxxxxxx> > Signed-off-by: Zheng Bin <zhengbin13@xxxxxxxxxx> > --- Seems reasonable to me: Reviewed-by: Brian Foster <bfoster@xxxxxxxxxx> > fs/xfs/xfs_super.c | 3 ++- > 1 file changed, 2 insertions(+), 1 deletion(-) > > diff --git a/fs/xfs/xfs_super.c b/fs/xfs/xfs_super.c > index e80bd2c4c279..aae469f73efe 100644 > --- a/fs/xfs/xfs_super.c > +++ b/fs/xfs/xfs_super.c > @@ -807,7 +807,8 @@ xfs_fs_statfs( > statp->f_blocks = sbp->sb_dblocks - lsize; > spin_unlock(&mp->m_sb_lock); > > - statp->f_bfree = fdblocks - mp->m_alloc_set_aside; > + /* make sure statp->f_bfree does not underflow */ > + statp->f_bfree = max_t(int64_t, fdblocks - mp->m_alloc_set_aside, 0); > statp->f_bavail = statp->f_bfree; > > fakeinos = XFS_FSB_TO_INO(mp, statp->f_bfree); > -- > 2.26.0.106.g9fadedd >
