qemu-devel
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [PATCH v19 4/4] iotests: 287: add qcow2 compression type test


From: Denis Plotnikov
Subject: Re: [PATCH v19 4/4] iotests: 287: add qcow2 compression type test
Date: Mon, 20 Apr 2020 17:30:04 +0300
User-agent: Mozilla/5.0 (X11; Linux x86_64; rv:68.0) Gecko/20100101 Thunderbird/68.4.1



On 20.04.2020 15:56, Vladimir Sementsov-Ogievskiy wrote:
20.04.2020 12:17, Denis Plotnikov wrote:
The test checks fulfilling qcow2 requirements for the compression
type feature and zstd compression type operability.

Signed-off-by: Denis Plotnikov <address@hidden>
---
  tests/qemu-iotests/287     | 153 +++++++++++++++++++++++++++++++++++++
  tests/qemu-iotests/287.out |  67 ++++++++++++++++
  tests/qemu-iotests/group   |   1 +
  3 files changed, 221 insertions(+)
  create mode 100755 tests/qemu-iotests/287
  create mode 100644 tests/qemu-iotests/287.out

diff --git a/tests/qemu-iotests/287 b/tests/qemu-iotests/287
new file mode 100755
index 0000000000..44988bc14b
--- /dev/null
+++ b/tests/qemu-iotests/287
@@ -0,0 +1,153 @@
+#!/usr/bin/env bash
+#
+# Test case for an image using zstd compression
+#
+# Copyright (c) 2020 Virtuozzo International GmbH
+#
+# This program is free software; you can redistribute it and/or modify
+# it under the terms of the GNU General Public License as published by
+# the Free Software Foundation; either version 2 of the License, or
+# (at your option) any later version.
+#
+# This program is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+# GNU General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with this program.  If not, see <http://www.gnu.org/licenses/>.
+#
+
+# creator
+owner=address@hidden
+
+seq="$(basename $0)"
+echo "QA output created by $seq"
+
+status=1    # failure is the default!
+
+# standard environment
+. ./common.rc
+. ./common.filter
+
+# This tests qocw2-specific low-level functionality
+_supported_fmt qcow2
+_supported_proto file
+_supported_os Linux
+
+# Sanity check: our use of $RAND_FILE fails if $TEST_DIR contains spaces
+# or other problems

Hmm, but I see, you now don't use writing from file, but instead, rewrite rand_file by 0xFA pattern. So, we don't have now any restrictions on RAND_FILE now, and than, we should drop this check.

+case "$TEST_DIR" in
+    *[^-_a-zA-Z0-9/]*)
+        _notrun "Suspicious TEST_DIR='$TEST_DIR', cowardly refusing to run" ;;
+esac
+
+COMPR_IMG="$TEST_IMG.compressed"
+RAND_FILE="$TEST_DIR/rand_data"
+
+_cleanup()
+{
+    _cleanup_test_img
+    rm -f "$COMPR_IMG"
+    rm -f "$RAND_FILE"
+}
+trap "_cleanup; exit \$status" 0 1 2 3 15
+
+# for all the cases
+CLUSTER_SIZE=65536
+
+# Check if we can run this test.
+if IMGOPTS='compression_type=zstd' _make_test_img 64M |
+    grep "Invalid parameter 'zstd'"; then
+    _notrun "ZSTD is disabled"
+fi
+
+echo
+echo "=== Testing compression type incompatible bit setting for zstd ==="
+echo
+# use the zstd image created on the test runnable check above.

Hmm, but you don't? Should you also drop the following line?

+IMGOPTS='compression_type=zstd' _make_test_img 64M
+$PYTHON qcow2.py "$TEST_IMG" dump-header | grep incompatible_features
+
+echo
+echo "=== Testing compression type incompatible bit setting for zlib ==="
+echo
+IMGOPTS='compression_type=zlib' _make_test_img 64M
+$PYTHON qcow2.py "$TEST_IMG" dump-header | grep incompatible_features
+
+echo
+echo "=== Testing zlib with incompatible bit set ==="
+echo
+IMGOPTS='compression_type=zlib' _make_test_img 64M
+$PYTHON qcow2.py "$TEST_IMG" set-feature-bit incompatible 3
+# to make sure the bit was actually set
+$PYTHON qcow2.py "$TEST_IMG" dump-header | grep incompatible_features
+
+if $QEMU_IMG info "$TEST_IMG" >/dev/null 2>&1 ; then
+    echo "Error: The image opened successfully. The image must not be opened."
+fi
+
+echo
+echo "=== Testing zstd with incompatible bit unset ==="
+echo
+IMGOPTS='compression_type=zstd' _make_test_img 64M
+$PYTHON qcow2.py "$TEST_IMG" set-header incompatible_features 0
+# to make sure the bit was actually unset
+$PYTHON qcow2.py "$TEST_IMG" dump-header | grep incompatible_features
+
+if $QEMU_IMG info "$TEST_IMG" >/dev/null 2>&1 ; then
+    echo "Error: The image opened successfully. The image must not be opened."
+fi
+
+echo
+echo "=== Testing compression type values ==="
+echo
+# zlib=0
+IMGOPTS='compression_type=zlib' _make_test_img 64M
+peek_file_be "$TEST_IMG" 104 1
+echo
+
+# zstd=1
+IMGOPTS='compression_type=zstd' _make_test_img 64M
+peek_file_be "$TEST_IMG" 104 1
+echo
+
+echo
+echo "=== Testing simple reading and writing with zstd ==="
+echo
+IMGOPTS='compression_type=zstd' _make_test_img 64M
+$QEMU_IO -c "write -c -P 0xAC 64K 64K " "$TEST_IMG" | _filter_qemu_io
+$QEMU_IO -c "read -P 0xAC 64K 64K " "$TEST_IMG" | _filter_qemu_io
+# read on the cluster boundaries
+$QEMU_IO -c "read -v 131070 8 " "$TEST_IMG" | _filter_qemu_io
+$QEMU_IO -c "read -v 65534 8" "$TEST_IMG" | _filter_qemu_io
+
+echo
+echo "=== Testing adjacent clusters reading and writing with zstd ==="
+echo
+IMGOPTS='compression_type=zstd' _make_test_img 64M
+$QEMU_IO -c "write -c -P 0xAB 0 64K " "$TEST_IMG" | _filter_qemu_io
+$QEMU_IO -c "write -c -P 0xAC 64K 64K " "$TEST_IMG" | _filter_qemu_io
+$QEMU_IO -c "write -c -P 0xAD 128K 64K " "$TEST_IMG" | _filter_qemu_io
+
+$QEMU_IO -c "read -P 0xAB 0 64k " "$TEST_IMG" | _filter_qemu_io
+$QEMU_IO -c "read -P 0xAC 64K 64k " "$TEST_IMG" | _filter_qemu_io
+$QEMU_IO -c "read -P 0xAD 128K 64k " "$TEST_IMG" | _filter_qemu_io
+
+echo
+echo "=== Testing incompressible cluster processing with zstd ==="
+echo
+# create a 2M imgae and fill it with 1M likely incompressible data

image

+# and 1M compressible data
+dd if=/dev/urandom of="$RAND_FILE" bs=2M count=1
+# rewrite the last 1M with compressible data
+$QEMU_IO -f raw -c "write -P 0xFA 1M 1M" "$RAND_FILE" | _filter_qemu_io

So, we avoid problems with whitespaces  in RAND_FILE.

Still, you may improve it a bit, by not generating extra 1M of random data, for example, create 2M raw image, then write to it b dd only 1M, then the other megabyte by pattern.
Would it be any significant improvement? This example 2 lines of code. Your example is 3 lines (create-dd-write). You example demand more output in test.out file also because of 3 command instead of 2. We won't win anything significant in performance by reducing 2M rand data to 1M.
So, I don't really know

Also, I believe, that it works as is, but passing two -f options to qemu-io is not good thing, I think, actually, we'd better forbid it in qemu-io option parsing (Haha, I'm sure, a lot of iotests will need a fix:).. So, more correct is something like

QEMU_IO_OPTIONS=$QEMU_IO_OPTIONS_NO_FMT $QEMU_IO -f raw [...]







reply via email to

[Prev in Thread] Current Thread [Next in Thread]