tests/generic/692

#! /bin/bash
# SPDX-License-Identifier: GPL-2.0
# Copyright (c) 2022 Meta, Inc.  All Rights Reserved.
#
# FS QA Test 692
#
# fs-verity requires the filesystem to decide how it stores the Merkle tree,
# which can be quite large.
# It is convenient to treat the Merkle tree as past EOF, and ext4, f2fs, and
# btrfs do so in at least some fashion. This leads to an edge case where a
# large file can be under the file system file size limit, but trigger EFBIG
# on enabling fs-verity. Test enabling verity on some large files to exercise
# EFBIG logic for filesystems with fs-verity specific limits.
#
. ./common/preamble
_begin_fstest auto quick verity

# Override the default cleanup function.
_cleanup()
{
	cd /
	_restore_fsverity_signatures
	rm -f $tmp.*
}

# Import common functions.
. ./common/filter
. ./common/verity

_require_test
_require_math
_require_scratch_verity
_require_fsverity_max_file_size_limit
_disable_fsverity_signatures

_scratch_mkfs_verity &>> $seqres.full
_scratch_mount

fsv_file=$SCRATCH_MNT/file.fsv

max_sz=$(_get_max_file_size $SCRATCH_MNT)
_fsv_scratch_begin_subtest "way too big: fail on first merkle block"
truncate -s $max_sz $fsv_file
_fsv_enable $fsv_file |& _filter_scratch

# The goal of this second test is to make a big enough file that we trip the
# EFBIG codepath, but not so big that we hit it immediately when writing the
# first Merkle leaf.
#
# The Merkle tree is stored with the leaf node level (L0) last, but it is
# written first.  To get an interesting overflow, we need the maximum file size
# ($max_sz) to be in the middle of L0 -- ideally near the beginning of L0 so
# that we don't have to write many blocks before getting an error.

bs=$FSV_BLOCK_SIZE
hash_size=32   # SHA-256
hashes_per_block=$(echo "scale=30; $bs/$hash_size" | $BC -q)

# Compute the proportion of the original file size that the non-leaf levels of
# the Merkle tree take up.  Ignoring padding, this is 1/($hashes_per_block^2) +
# 1/($hashes_per_block^3) + 1/($hashes_per_block^4) + ...  Compute it using the
# formula for the sum of a geometric series: \sum_{k=0}^{\infty} ar^k = a/(1-r).
a=$(echo "scale=30; 1/($hashes_per_block^2)" | $BC -q)
r=$(echo "scale=30; 1/$hashes_per_block" | $BC -q)
nonleaves_relative_size=$(echo "scale=30; $a/(1-$r)" | $BC -q)

# Compute the original file size where the leaf level L0 starts at $max_sz.
# Padding is still ignored, so the real value is slightly smaller than this.
sz=$(echo "$max_sz/(1+$nonleaves_relative_size)" | $BC -q)

# Finally, to get a file size where $max_sz occurs just after the start of L0
# rather than *at* the start of L0, subtract an overestimate of the padding: 64K
# after the file contents, then $bs per level for 11 levels.  11 levels is the
# most levels that can ever be needed, assuming the block size is at least 1K.
sz=$(echo "$sz - 65536 - $bs*11" | $BC -q)

_fsv_scratch_begin_subtest "still too big: fail on first invalid merkle block"
truncate -s $sz $fsv_file
_fsv_enable $fsv_file |& _filter_scratch

# success, all done
status=0
exit