qcow2: Implement zstd cluster compression method
v6 cover qcow2: Implement zstd cluster compression method
v6:
* "block/qcow2-threads: fix qcow2_decompress" is removed from the series since it has been accepted by Max already
* add compile time checking for Qcow2Header to be a multiple of 8 [Max, Alberto]
* report error on qcow2 amending when the compression type is actually chnged [Max]
* remove the extra space and the extra new line [Max]
* re-arrange acks and signed-off-s [Vladimir]
v5:
* replace -ENOTSUP with abort in qcow2_co_decompress [Vladimir]
* set cluster size for all test cases in the beginning of the 287 test
v4:
* the series is rebased on top of 01 "block/qcow2-threads: fix qcow2_decompress"
* 01 is just a no-change resend to avoid extra dependencies. Still, it may be merged in separate
v3:
* remove redundant max compression type value check [Vladimir, Eric] (the switch below checks everything)
* prevent compression type changing on "qemu-img amend" [Vladimir]
* remove zstd config setting, since it has been added already by "migration" patches [Vladimir]
* change the compression type error message [Vladimir]
* fix alignment and 80-chars exceeding [Vladimir]
v2:
* rework compression type setting [Vladimir]
* squash iotest changes to the compression type introduction patch [Vladimir, Eric]
* fix zstd availability checking in zstd iotest [Vladimir]
* remove unnecessry casting [Eric]
* remove rudundant checks [Eric]
* fix compressed cluster layout in qcow2 spec [Vladimir]
* fix wording [Eric, Vladimir]
* fix compression type filtering in iotests [Eric]
v1:
the initial series
Denis Plotnikov (4):
qcow2: introduce compression type feature
qcow2: rework the cluster compression routine
qcow2: add zstd cluster compression
iotests: 287: add qcow2 compression type test
docs/interop/qcow2.txt | 20 ++++ configure | 2 +- qapi/block-core.json | 23 +++- block/qcow2.h | 20 +++- include/block/block_int.h | 1 + block/qcow2-threads.c | 195 +++++++++++++++++++++++++++++-- block/qcow2.c | 132 +++++++++++++++++++++ tests/qemu-iotests/031.out | 14 +-- tests/qemu-iotests/036.out | 4 +- tests/qemu-iotests/049.out | 102 ++++++++-------- tests/qemu-iotests/060.out | 1 + tests/qemu-iotests/061.out | 34 +++--- tests/qemu-iotests/065 | 28 +++-- tests/qemu-iotests/080 | 2 +- tests/qemu-iotests/144.out | 4 +- tests/qemu-iotests/182.out | 2 +- tests/qemu-iotests/242.out | 5 + tests/qemu-iotests/255.out | 8 +- tests/qemu-iotests/287 | 128 ++++++++++++++++++++ tests/qemu-iotests/287.out | 43 +++++++ tests/qemu-iotests/common.filter | 3 +- tests/qemu-iotests/group | 1 + 22 files changed, 664 insertions(+), 108 deletions(-) create mode 100755 tests/qemu-iotests/287 create mode 100644 tests/qemu-iotests/287.out
Signature · 1 line
2.17.0
PATCH v6 1/4 qcow2: introduce compression type feature
The patch adds some preparation parts for incompatible compression type feature to qcow2 allowing the use different compression methods for image clusters (de)compressing.
It is implied that the compression type is set on the image creation and can be changed only later by image conversion, thus compression type defines the only compression algorithm used for the image, and thus, for all image clusters.
The goal of the feature is to add support of other compression methods to qcow2. For example, ZSTD which is more effective on compression than ZLIB.
The default compression is ZLIB. Images created with ZLIB compression type are backward compatible with older qemu versions.
Adding of the compression type breaks a number of tests because now the compression type is reported on image creation and there are some changes in the qcow2 header in size and offsets.
The tests are fixed in the following ways:
* filter out compression_type for all the tests
* fix header size, feature table size and backing file offset
affected tests: 031, 036, 061, 080
header_size +=8: 1 byte compression type
7 bytes padding
feature_table += 48: incompatible feture compression type
backing_file_offset += 56 (8 + 48 -> header_change + fature_table_change)
* add "compression type" for test output matching when it isn't filtered
affected tests: 049, 060, 061, 065, 144, 182, 242, 255--- qapi/block-core.json | 22 +++++- block/qcow2.h | 20 ++++- include/block/block_int.h | 1 + block/qcow2.c | 121 +++++++++++++++++++++++++++++++ tests/qemu-iotests/031.out | 14 ++-- tests/qemu-iotests/036.out | 4 +- tests/qemu-iotests/049.out | 102 +++++++++++++------------- tests/qemu-iotests/060.out | 1 + tests/qemu-iotests/061.out | 34 +++++---- tests/qemu-iotests/065 | 28 ++++--- tests/qemu-iotests/080 | 2 +- tests/qemu-iotests/144.out | 4 +- tests/qemu-iotests/182.out | 2 +- tests/qemu-iotests/242.out | 5 ++ tests/qemu-iotests/255.out | 8 +- tests/qemu-iotests/common.filter | 3 +- 16 files changed, 275 insertions(+), 96 deletions(-)
qapi/block-core.json +20 -2
@@ -78,6 +78,8 @@ # # @bitmaps: A list of qcow2 bitmap details (since 4.0) # +# @compression-type: the image cluster compression method (since 5.0) +# # Since: 1.7 ## { 'struct': 'ImageInfoSpecificQCow2',
@@ -89,7 +91,8 @@ '*corrupt': 'bool', 'refcount-bits': 'int', '*encrypt': 'ImageInfoSpecificQCow2Encryption', - '*bitmaps': ['Qcow2BitmapInfo'] + '*bitmaps': ['Qcow2BitmapInfo'], + 'compression-type': 'Qcow2CompressionType' } } ##
@@ -4392,6 +4395,18 @@ 'data': [ 'v2', 'v3' ] } +## +# @Qcow2CompressionType: +# +# Compression type used in qcow2 image file +# +# @zlib: zlib compression, see <http://zlib.net/> +# +# Since: 5.0 +## +{ 'enum': 'Qcow2CompressionType', + 'data': [ 'zlib' ] } + ## # @BlockdevCreateOptionsQcow2: #
@@ -4415,6 +4430,8 @@ # allowed values: off, falloc, full, metadata) # @lazy-refcounts: True if refcounts may be updated lazily (default: off) # @refcount-bits: Width of reference counts in bits (default: 16) +# @compression-type: The image cluster compression method +# (default: zlib, since 5.0) # # Since: 2.12 ##
@@ -4430,7 +4447,8 @@ '*cluster-size': 'size', '*preallocation': 'PreallocMode', '*lazy-refcounts': 'bool', - '*refcount-bits': 'int' } } + '*refcount-bits': 'int', + '*compression-type':'Qcow2CompressionType' } } ## # @BlockdevCreateOptionsQed:
block/qcow2.h +19 -1
@@ -146,8 +146,16 @@ typedef struct QCowHeader {uint32_t refcount_order;uint32_t header_length; + +/* Additional fields */ +uint8_t compression_type; + +/* header must be a multiple of 8 */ +uint8_t padding[7 ]; }QEMU_PACKED QCowHeader; +QEMU_BUILD_BUG_ON (sizeof (QCowHeader) %8 !=0 ); + typedef struct QEMU_PACKED QCowSnapshotHeader {/* header is 8 byte aligned */ uint64_t l1_table_offset;
@@ -216,13 +224,16 @@ enum {QCOW2_INCOMPAT_DIRTY_BITNR =0 ,QCOW2_INCOMPAT_CORRUPT_BITNR =1 ,QCOW2_INCOMPAT_DATA_FILE_BITNR =2 , +QCOW2_INCOMPAT_COMPRESSION_BITNR =3 ,QCOW2_INCOMPAT_DIRTY =1 <<QCOW2_INCOMPAT_DIRTY_BITNR ,QCOW2_INCOMPAT_CORRUPT =1 <<QCOW2_INCOMPAT_CORRUPT_BITNR ,QCOW2_INCOMPAT_DATA_FILE =1 <<QCOW2_INCOMPAT_DATA_FILE_BITNR , +QCOW2_INCOMPAT_COMPRESSION =1 <<QCOW2_INCOMPAT_COMPRESSION_BITNR ,QCOW2_INCOMPAT_MASK =QCOW2_INCOMPAT_DIRTY |QCOW2_INCOMPAT_CORRUPT - | QCOW2_INCOMPAT_DATA_FILE, + |QCOW2_INCOMPAT_DATA_FILE + | QCOW2_INCOMPAT_COMPRESSION, };/* Compatible feature bits */
@@ -369,6 +380,13 @@ typedef struct BDRVQcow2State {bool metadata_preallocation_checked;bool metadata_preallocation; +/* +* Compression type used for the image. Default: 0 - ZLIB +* The image compression type is set on image creation. +* The only way to change the compression type is to convert the image +* with the desired compression type set +*/ +Qcow2CompressionType compression_type; } BDRVQcow2State; typedef struct Qcow2COWRegion {
include/block/block_int.h +1 -0
@@ -57,6 +57,7 @@#define BLOCK_OPT_REFCOUNT_BITS "refcount_bits"#define BLOCK_OPT_DATA_FILE "data_file"#define BLOCK_OPT_DATA_FILE_RAW "data_file_raw" +#define BLOCK_OPT_COMPRESSION_TYPE "compression_type"#define BLOCK_PROBE_BUF_SIZE 512
block/qcow2.c +121 -0
199 lines not on this page · raw message
tests/qemu-iotests/031.out +7 -7
@@ -113,11 +113,11 @@ incompatible_features [] compatible_features [] autoclear_features [] refcount_order 4 -header_length 104 +header_length 112 Header extension: magic 0x6803f857 -length 192 +length 240 data <binary> Header extension:
@@ -146,11 +146,11 @@ incompatible_features [] compatible_features [] autoclear_features [] refcount_order 4 -header_length 104 +header_length 112 Header extension: magic 0x6803f857 -length 192 +length 240 data <binary> Header extension:
@@ -164,7 +164,7 @@ No errors were found on the image. magic 0x514649fb version 3 -backing_file_offset 0x178 +backing_file_offset 0x1b0 backing_file_size 0x17 cluster_bits 16 size 67108864
@@ -179,7 +179,7 @@ incompatible_features [] compatible_features [] autoclear_features [] refcount_order 4 -header_length 104 +header_length 112 Header extension: magic 0xe2792aca
@@ -188,7 +188,7 @@ data 'host_device' Header extension: magic 0x6803f857 -length 192 +length 240 data <binary> Header extension:
tests/qemu-iotests/036.out +2 -2
@@ -26,7 +26,7 @@ compatible_features [] autoclear_features [63] Header extension: magic 0x6803f857 -length 192 +length 240 data <binary>
@@ -38,7 +38,7 @@ compatible_features [] autoclear_features [] Header extension: magic 0x6803f857 -length 192 +length 240 data <binary> *** done
tests/qemu-iotests/049.out +51 -51
225 lines not on this page · raw message
tests/qemu-iotests/060.out +1 -0
@@ -17,6 +17,7 @@ virtual size: 64 MiB (67108864 bytes) cluster_size: 65536 Format specific information: compat: 1.1 + compression type: zlib lazy refcounts: false refcount bits: 16 corrupt: true
tests/qemu-iotests/061.out +20 -14
@@ -22,11 +22,11 @@ incompatible_features [] compatible_features [0] autoclear_features [] refcount_order 4 -header_length 104 +header_length 112 Header extension: magic 0x6803f857 -length 192 +length 240 data <binary> magic 0x514649fb
@@ -80,11 +80,11 @@ incompatible_features [] compatible_features [0] autoclear_features [] refcount_order 4 -header_length 104 +header_length 112 Header extension: magic 0x6803f857 -length 192 +length 240 data <binary> magic 0x514649fb
@@ -136,11 +136,11 @@ incompatible_features [0] compatible_features [0] autoclear_features [] refcount_order 4 -header_length 104 +header_length 112 Header extension: magic 0x6803f857 -length 192 +length 240 data <binary> ERROR cluster 5 refcount=0 reference=1
@@ -191,11 +191,11 @@ incompatible_features [] compatible_features [42] autoclear_features [42] refcount_order 4 -header_length 104 +header_length 112 Header extension: magic 0x6803f857 -length 192 +length 240 data <binary> magic 0x514649fb
@@ -260,11 +260,11 @@ incompatible_features [] compatible_features [0] autoclear_features [] refcount_order 4 -header_length 104 +header_length 112 Header extension: magic 0x6803f857 -length 192 +length 240 data <binary> read 65536/65536 bytes at offset 44040192
@@ -294,11 +294,11 @@ incompatible_features [0] compatible_features [0] autoclear_features [] refcount_order 4 -header_length 104 +header_length 112 Header extension: magic 0x6803f857 -length 192 +length 240 data <binary> ERROR cluster 5 refcount=0 reference=1
@@ -323,11 +323,11 @@ incompatible_features [] compatible_features [] autoclear_features [] refcount_order 4 -header_length 104 +header_length 112 Header extension: magic 0x6803f857 -length 192 +length 240 data <binary> read 131072/131072 bytes at offset 0
@@ -491,6 +491,7 @@ virtual size: 64 MiB (67108864 bytes) cluster_size: 65536 Format specific information: compat: 1.1 + compression type: zlib lazy refcounts: false refcount bits: 16 data file: TEST_DIR/t.IMGFMT.data
@@ -511,6 +512,7 @@ virtual size: 64 MiB (67108864 bytes) cluster_size: 65536 Format specific information: compat: 1.1 + compression type: zlib lazy refcounts: false refcount bits: 16 data file: foo
@@ -524,6 +526,7 @@ virtual size: 64 MiB (67108864 bytes) cluster_size: 65536 Format specific information: compat: 1.1 + compression type: zlib lazy refcounts: false refcount bits: 16 data file raw: false
@@ -538,6 +541,7 @@ virtual size: 64 MiB (67108864 bytes) cluster_size: 65536 Format specific information: compat: 1.1 + compression type: zlib lazy refcounts: false refcount bits: 16 data file: TEST_DIR/t.IMGFMT.data
@@ -550,6 +554,7 @@ virtual size: 64 MiB (67108864 bytes) cluster_size: 65536 Format specific information: compat: 1.1 + compression type: zlib lazy refcounts: false refcount bits: 16 data file: TEST_DIR/t.IMGFMT.data
@@ -563,6 +568,7 @@ virtual size: 64 MiB (67108864 bytes) cluster_size: 65536 Format specific information: compat: 1.1 + compression type: zlib lazy refcounts: false refcount bits: 16 data file: TEST_DIR/t.IMGFMT.data
tests/qemu-iotests/065 +18 -10
@@ -88,24 +88,30 @@ class TestQMP(TestImageInfoSpecific): class TestQCow2(TestQemuImgInfo): '''Testing a qcow2 version 2 image''' img_options = 'compat=0.10' - json_compare = { 'compat': '0.10', 'refcount-bits': 16 } - human_compare = [ 'compat: 0.10', 'refcount bits: 16' ] + json_compare = { 'compat': '0.10', 'refcount-bits': 16, + 'compression-type': 'zlib' } + human_compare = [ 'compat: 0.10', 'compression type: zlib', + 'refcount bits: 16' ] class TestQCow3NotLazy(TestQemuImgInfo): '''Testing a qcow2 version 3 image with lazy refcounts disabled''' img_options = 'compat=1.1,lazy_refcounts=off' json_compare = { 'compat': '1.1', 'lazy-refcounts': False, - 'refcount-bits': 16, 'corrupt': False } - human_compare = [ 'compat: 1.1', 'lazy refcounts: false', - 'refcount bits: 16', 'corrupt: false' ] + 'refcount-bits': 16, 'corrupt': False, + 'compression-type': 'zlib' } + human_compare = [ 'compat: 1.1', 'compression type: zlib', + 'lazy refcounts: false', 'refcount bits: 16', + 'corrupt: false' ] class TestQCow3Lazy(TestQemuImgInfo): '''Testing a qcow2 version 3 image with lazy refcounts enabled''' img_options = 'compat=1.1,lazy_refcounts=on' json_compare = { 'compat': '1.1', 'lazy-refcounts': True, - 'refcount-bits': 16, 'corrupt': False } - human_compare = [ 'compat: 1.1', 'lazy refcounts: true', - 'refcount bits: 16', 'corrupt: false' ] + 'refcount-bits': 16, 'corrupt': False, + 'compression-type': 'zlib' } + human_compare = [ 'compat: 1.1', 'compression type: zlib', + 'lazy refcounts: true', 'refcount bits: 16', + 'corrupt: false' ] class TestQCow3NotLazyQMP(TestQMP): '''Testing a qcow2 version 3 image with lazy refcounts disabled, opening
@@ -113,7 +119,8 @@ class TestQCow3NotLazyQMP(TestQMP): img_options = 'compat=1.1,lazy_refcounts=off' qemu_options = 'lazy-refcounts=on' compare = { 'compat': '1.1', 'lazy-refcounts': False, - 'refcount-bits': 16, 'corrupt': False } + 'refcount-bits': 16, 'corrupt': False, + 'compression-type': 'zlib' } class TestQCow3LazyQMP(TestQMP):
@@ -122,7 +129,8 @@ class TestQCow3LazyQMP(TestQMP): img_options = 'compat=1.1,lazy_refcounts=on' qemu_options = 'lazy-refcounts=off' compare = { 'compat': '1.1', 'lazy-refcounts': True, - 'refcount-bits': 16, 'corrupt': False } + 'refcount-bits': 16, 'corrupt': False, + 'compression-type': 'zlib' } TestImageInfoSpecific = None TestQemuImgInfo = None
tests/qemu-iotests/080 +1 -1
@@ -45,7 +45,7 @@ _supported_os Linux # - This is generally a test for compat=1.1 images _unsupported_imgopts 'refcount_bits=1[^0-9]' data_file 'compat=0.10' -header_size=104 +header_size=112 offset_backing_file_offset=8 offset_backing_file_size=16
tests/qemu-iotests/144.out +2 -2
@@ -9,7 +9,7 @@ Formatting 'TEST_DIR/t.IMGFMT', fmt=IMGFMT size=536870912 { 'execute': 'qmp_capabilities' } {"return": {}} { 'execute': 'blockdev-snapshot-sync', 'arguments': { 'device': 'virtio0', 'snapshot-file':'TEST_DIR/tmp.IMGFMT', 'format': 'IMGFMT' } } -Formatting 'TEST_DIR/tmp.qcow2', fmt=qcow2 size=536870912 backing_file=TEST_DIR/t.qcow2 backing_fmt=qcow2 cluster_size=65536 lazy_refcounts=off refcount_bits=16 +Formatting 'TEST_DIR/tmp.qcow2', fmt=qcow2 size=536870912 backing_file=TEST_DIR/t.qcow2 backing_fmt=qcow2 cluster_size=65536 lazy_refcounts=off refcount_bits=16 compression_type=zlib {"return": {}} === Performing block-commit on active layer ===
@@ -31,6 +31,6 @@ Formatting 'TEST_DIR/tmp.qcow2', fmt=qcow2 size=536870912 backing_file=TEST_DIR/ === Performing Live Snapshot 2 === { 'execute': 'blockdev-snapshot-sync', 'arguments': { 'device': 'virtio0', 'snapshot-file':'TEST_DIR/tmp2.IMGFMT', 'format': 'IMGFMT' } } -Formatting 'TEST_DIR/tmp2.qcow2', fmt=qcow2 size=536870912 backing_file=TEST_DIR/t.qcow2 backing_fmt=qcow2 cluster_size=65536 lazy_refcounts=off refcount_bits=16 +Formatting 'TEST_DIR/tmp2.qcow2', fmt=qcow2 size=536870912 backing_file=TEST_DIR/t.qcow2 backing_fmt=qcow2 cluster_size=65536 lazy_refcounts=off refcount_bits=16 compression_type=zlib {"return": {}} *** done
tests/qemu-iotests/182.out +1 -1
@@ -13,7 +13,7 @@ Is another process using the image [TEST_DIR/t.qcow2]? {'execute': 'blockdev-add', 'arguments': { 'node-name': 'node0', 'driver': 'file', 'filename': 'TEST_DIR/t.IMGFMT', 'locking': 'on' } } {"return": {}} {'execute': 'blockdev-snapshot-sync', 'arguments': { 'node-name': 'node0', 'snapshot-file': 'TEST_DIR/t.IMGFMT.overlay', 'snapshot-node-name': 'node1' } } -Formatting 'TEST_DIR/t.qcow2.overlay', fmt=qcow2 size=197120 backing_file=TEST_DIR/t.qcow2 backing_fmt=file cluster_size=65536 lazy_refcounts=off refcount_bits=16 +Formatting 'TEST_DIR/t.qcow2.overlay', fmt=qcow2 size=197120 backing_file=TEST_DIR/t.qcow2 backing_fmt=file cluster_size=65536 lazy_refcounts=off refcount_bits=16 compression_type=zlib {"return": {}} {'execute': 'blockdev-add', 'arguments': { 'node-name': 'node1', 'driver': 'file', 'filename': 'TEST_DIR/t.IMGFMT', 'locking': 'on' } } {"return": {}}
tests/qemu-iotests/242.out +5 -0
@@ -12,6 +12,7 @@ virtual size: 1 MiB (1048576 bytes) cluster_size: 65536 Format specific information: compat: 1.1 + compression type: zlib lazy refcounts: false refcount bits: 16 corrupt: false
@@ -32,6 +33,7 @@ virtual size: 1 MiB (1048576 bytes) cluster_size: 65536 Format specific information: compat: 1.1 + compression type: zlib lazy refcounts: false bitmaps: [0]:
@@ -64,6 +66,7 @@ virtual size: 1 MiB (1048576 bytes) cluster_size: 65536 Format specific information: compat: 1.1 + compression type: zlib lazy refcounts: false bitmaps: [0]:
@@ -104,6 +107,7 @@ virtual size: 1 MiB (1048576 bytes) cluster_size: 65536 Format specific information: compat: 1.1 + compression type: zlib lazy refcounts: false bitmaps: [0]:
@@ -153,6 +157,7 @@ virtual size: 1 MiB (1048576 bytes) cluster_size: 65536 Format specific information: compat: 1.1 + compression type: zlib lazy refcounts: false bitmaps: [0]:
tests/qemu-iotests/255.out +4 -4
@@ -3,9 +3,9 @@ Finishing a commit job with background reads === Create backing chain and start VM === -Formatting 'TEST_DIR/PID-t.qcow2.mid', fmt=qcow2 size=134217728 cluster_size=65536 lazy_refcounts=off refcount_bits=16 +Formatting 'TEST_DIR/PID-t.qcow2.mid', fmt=qcow2 size=134217728 cluster_size=65536 lazy_refcounts=off refcount_bits=16 compression_type=zlib -Formatting 'TEST_DIR/PID-t.qcow2', fmt=qcow2 size=134217728 cluster_size=65536 lazy_refcounts=off refcount_bits=16 +Formatting 'TEST_DIR/PID-t.qcow2', fmt=qcow2 size=134217728 cluster_size=65536 lazy_refcounts=off refcount_bits=16 compression_type=zlib === Start background read requests ===
@@ -23,9 +23,9 @@ Closing the VM while a job is being cancelled === Create images and start VM === -Formatting 'TEST_DIR/PID-src.qcow2', fmt=qcow2 size=134217728 cluster_size=65536 lazy_refcounts=off refcount_bits=16 +Formatting 'TEST_DIR/PID-src.qcow2', fmt=qcow2 size=134217728 cluster_size=65536 lazy_refcounts=off refcount_bits=16 compression_type=zlib -Formatting 'TEST_DIR/PID-dst.qcow2', fmt=qcow2 size=134217728 cluster_size=65536 lazy_refcounts=off refcount_bits=16 +Formatting 'TEST_DIR/PID-dst.qcow2', fmt=qcow2 size=134217728 cluster_size=65536 lazy_refcounts=off refcount_bits=16 compression_type=zlib wrote 1048576/1048576 bytes at offset 0 1 MiB, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec)
tests/qemu-iotests/common.filter +2 -1
@@ -152,7 +152,8 @@ _filter_img_create() -e "s# refcount_bits=[0-9]\\+##g" \ -e "s# key-secret=[a-zA-Z0-9]\\+##g" \ -e "s# iter-time=[0-9]\\+##g" \ - -e "s# force_size=\\(on\\|off\\)##g" + -e "s# force_size=\\(on\\|off\\)##g" \ + -e "s# compression_type=[a-zA-Z0-9]\\+##g" } _filter_img_info()
Signature · 1 line
2.17.0
Re: [PATCH v6 1/4] qcow2: introduce compression type feature
Quoted 15 lines · Denis Plotnikov
The patch adds some preparation parts for incompatible compression type feature to qcow2 allowing the use different compression methods for image clusters (de)compressing.
It is implied that the compression type is set on the image creation and can be changed only later by image conversion, thus compression type defines the only compression algorithm used for the image, and thus, for all image clusters.
The goal of the feature is to add support of other compression methods to qcow2. For example, ZSTD which is more effective on compression than ZLIB.
The default compression is ZLIB. Images created with ZLIB compression type are backward compatible with older qemu versions.
Quoted 6 lines
Adding of the compression type breaks a number of tests because now the compression type is reported on image creation and there are some changes in the qcow2 header in size and offsets.
The tests are fixed in the following ways:
* filter outcompression_typefor all the tests
Presumably this filter is optional, and we will not use it on the specific new tests that prove zstd compression works - but that should be later in the series, so for this patch it is okay.
Quoted 5 lines
* fix header size, feature table size and backing file offset
affected tests: 031, 036, 061, 080
header_size +=8: 1 byte compression type
7 bytes padding
feature_table += 48: incompatible feture compression typefeature
Quoted 1 line
backing_file_offset += 56 (8 + 48 -> header_change + fature_table_change)
feature
(interesting that you have two different changed spellings ;)
Quoted 2 lines
* add "compression type" for test output matching when it isn't filtered affected tests: 049, 060, 061, 065, 144, 182, 242, 255
Or maybe the comment above should be changed to "many tests" rather than "all the tests".
Quoted 25 lines
Signed-off-by Denis Plotnikov <[email protected]>Reviewed-by Vladimir Sementsov-Ogievskiy <[email protected]>--- qapi/block-core.json | 22 +++++- block/qcow2.h | 20 ++++- include/block/block_int.h | 1 + block/qcow2.c | 121 +++++++++++++++++++++++++++++++ tests/qemu-iotests/031.out | 14 ++-- tests/qemu-iotests/036.out | 4 +- tests/qemu-iotests/049.out | 102 +++++++++++++------------- tests/qemu-iotests/060.out | 1 + tests/qemu-iotests/061.out | 34 +++++---- tests/qemu-iotests/065 | 28 ++++--- tests/qemu-iotests/080 | 2 +- tests/qemu-iotests/144.out | 4 +- tests/qemu-iotests/182.out | 2 +- tests/qemu-iotests/242.out | 5 ++ tests/qemu-iotests/255.out | 8 +- tests/qemu-iotests/common.filter | 3 +- 16 files changed, 275 insertions(+), 96 deletions(-)+++
b/block/qcow2.h
block/qcow2.h typedef struct QCowHeader { · line 154
Quoted 9 lines
block/qcow2.h +6
-0@@ -146,8 +146,16 @@ typedef struct QCowHeader {uint32_t refcount_order;uint32_t header_length; + +/* Additional fields */ +uint8_t compression_type; + +/* header must be a multiple of 8 */ +uint8_t padding[7 ];
Why two spaces after uint8_t (twice)?
block/qcow2.h typedef struct BDRVQcow2State { · line 387
Quoted 9 lines
block/qcow2.h +5
-0@@ -369,6 +380,13 @@ typedef struct BDRVQcow2State {bool metadata_preallocation_checked;bool metadata_preallocation; +/* +* Compression type used for the image. Default: 0 - ZLIB +* The image compression type is set on image creation. +* The only way to change the compression type is to convert the image +* with the desired compression type set
Missing trailing '.'. Maybe someday we can get 'qemu-img amend' to also adjust the compression type in-place; if that's something we think we might do, then this could be better worded as "For now, the only way to change...".
Quoted 5 lines
block/qcow2.c +1
-0+++ b/block/qcow2.c@@ -1242,6 +1242,48 @@ static int qcow2_update_options(BlockDriverState *bs, QDict *options,return ret; }
block/qcow2.c static int qcow2_update_options(BlockDriverState *bs, QDict *options, · line 1254
Quoted 10 lines
block/qcow2.c +9
-0+static int validate_compression_type (BDRVQcow2State *s,Error **errp) + +static int qcow2_compression_type_from_format(const char *ct) +{ +if (g_str_equal (ct,"zlib" )) { +return QCOW2_COMPRESSION_TYPE_ZLIB ; + }else { +return -EINVAL ; + }
Why are you open-coding this?
qapi_enum_parse(&Qcow2CompressionType_lookup, ct, -1, errp)
should do what you use this for, and automatically updates itself when you add zstd to the qapi enum later.
block/qcow2.c qcow2_co_create(BlockdevCreateOptions *create_options, Error **errp) · line 3496
Quoted 5 lines
block/qcow2.c +1
-0@@ -3401,6 +3493,8 @@ qcow2_co_create(BlockdevCreateOptions *create_options, Error **errp) .refcount_table_offset = cpu_to_be64(cluster_size), .refcount_table_clusters = cpu_to_be32(1), .refcount_order = cpu_to_be32(refcount_order), +/* don't deal with endians since compression_type is 1 byte long */
endianness
block/qcow2.c static QemuOptsList qcow2_create_opts = { · line 5637
Quoted 12 lines
block/qcow2.c +5
-0+ .compression_type = compression_type, .header_length = cpu_to_be32(sizeof(*header)), };@@ -5516,6 +5631,12 @@ static QemuOptsList qcow2_create_opts = { .help = "Width of a reference count entry in bits", .def_value_str ="16" }, + { + .name = BLOCK_OPT_COMPRESSION_TYPE, + .type = QEMU_OPT_STRING, + .help = "Compression method used for image clusters compression",
s/clusters/cluster/
Overall, getting closer. I assume this would go through Max's tree, if he has time to include it before soft freeze...
Signature · 3 lines
Eric Blake, Principal Software Engineer Red Hat, Inc. +1-919-301-3226 Virtualization: qemu.org | libvirt.org
Re: [PATCH v6 1/4] qcow2: introduce compression type feature
Thanks for the comments.
I'll make the fixes accordingly and re-sent the series shortly.
Denis
Quoted 143 lines · Eric Blake
Quoted 15 lines · Denis Plotnikov
The patch adds some preparation parts for incompatible compression type feature to qcow2 allowing the use different compression methods for image clusters (de)compressing.
It is implied that the compression type is set on the image creation and can be changed only later by image conversion, thus compression type defines the only compression algorithm used for the image, and thus, for all image clusters.
The goal of the feature is to add support of other compression methods to qcow2. For example, ZSTD which is more effective on compression than ZLIB.
The default compression is ZLIB. Images created with ZLIB compression type are backward compatible with older qemu versions.
Quoted 6 lines
Adding of the compression type breaks a number of tests because now the compression type is reported on image creation and there are some changes in the qcow2 header in size and offsets.
The tests are fixed in the following ways:
* filter outcompression_typefor all the testsPresumably this filter is optional, and we will not use it on the specific new tests that prove zstd compression works - but that should be later in the series, so for this patch it is okay.
Quoted 5 lines
* fix header size, feature table size and backing file offset affected tests: 031, 036, 061, 080
header_size+=8: 1 byte compression type
7 bytes padding
feature_table+= 48: incompatible feture compression typefeature
Quoted 1 line
backing_file_offset+= 56 (8 + 48 ->header_change+fature_table_change)feature
(interesting that you have two different changed spellings ;)
Quoted 2 lines
* add "compression type" for test output matching when it isn't filtered affected tests: 049, 060, 061, 065, 144, 182, 242, 255
Or maybe the comment above should be changed to "many tests" rather than "all the tests".
Quoted 25 lines
Signed-off-by Denis Plotnikov <[email protected]>Reviewed-by Vladimir Sementsov-Ogievskiy <[email protected]>---
qapi/block-core.json | 22 +++++- block/qcow2.h | 20 ++++- include/block/block_int.h | 1 +
block/qcow2.c | 121 +++++++++++++++++++++++++++++++tests/qemu-iotests/031.out | 14 ++-- tests/qemu-iotests/036.out | 4 +- tests/qemu-iotests/049.out | 102 +++++++++++++------------- tests/qemu-iotests/060.out | 1 + tests/qemu-iotests/061.out | 34 +++++---- tests/qemu-iotests/065 | 28 ++++--- tests/qemu-iotests/080 | 2 +- tests/qemu-iotests/144.out | 4 +- tests/qemu-iotests/182.out | 2 +- tests/qemu-iotests/242.out | 5 ++ tests/qemu-iotests/255.out | 8 +- tests/qemu-iotests/common.filter | 3 +- 16 files changed, 275 insertions(+), 96 deletions(-)+++
b/block/qcow2.hblock/qcow2.h typedef struct QCowHeader { · line 153
Quoted 8 lines
block/qcow2.h +6
-0@@ -146,8 +146,16 @@ typedef struct QCowHeader { uint32_t refcount_order; uint32_t header_length; + + /* Additional fields */ + uint8_t compression_type; + + /* header must be a multiple of 8 */ + uint8_t padding[7];Why two spaces after
uint8_t(twice)?Quoted 9 lines
block/qcow2.h +4
-0@@ -369,6 +380,13 @@ typedef struct BDRVQcow2State { bool metadata_preallocation_checked; bool metadata_preallocation; + /* +* Compression type used for the image. Default: 0 - ZLIB +* The image compression type is set on image creation. +* The only way to change the compression type is to convert the image
+ * with the desired compression type setMissing trailing '.'. Maybe someday we can get 'qemu-img amend' to also adjust the compression type in-place; if that's something we think we might do, then this could be better worded as "For now, the only way to change...".
Quoted 5 lines
+++
b/block/qcow2.cblock/qcow2.c +0
-0@@ -1242,6 +1242,48 @@ static int qcow2_update_options(BlockDriverState *bs, QDict *options, return ret; }block/qcow2.c static int · line 1254
Quoted 10 lines
block/qcow2.c +8
-0+static int validate_compression_type(BDRVQcow2State *s, Error **errp) + +static int qcow2_compression_type_from_format(const char *ct) +{ + if (g_str_equal(ct, "zlib")) { + return QCOW2_COMPRESSION_TYPE_ZLIB; + } else { + return -EINVAL; + }Why are you open-coding this?
qapi_enum_parse(&Qcow2CompressionType_lookup, ct, -1, errp)should do what you use this for, and automatically updates itself when you add zstd to the qapi enum later.
Quoted 7 lines
block/qcow2.c +1
-0@@ -3401,6 +3493,8 @@ qcow2_co_create(BlockdevCreateOptions *create_options, Error **errp) .refcount_table_offset = cpu_to_be64(cluster_size), .refcount_table_clusters = cpu_to_be32(1), .refcount_order = cpu_to_be32(refcount_order), + /* don't deal with endians since compression_type is 1 bytelong */
endianness
Quoted 4 lines
+ .compression_type =
compression_type, .header_length = cpu_to_be32(sizeof(*header)), };
@@ -5516,6 +5631,12 @@ static QemuOptsListqcow2_create_opts= {Quoted 8 lines
.help = "Width of a reference count entry in bits", .def_value_str = "16"
},
+ {
+ .name =BLOCK_OPT_COMPRESSION_TYPE,
+ .type =QEMU_OPT_STRING,
+ .help = "Compression method used for image clusters compression",s/clusters/cluster/
Overall, getting closer. I assume this would go through Max's tree, if he has time to include it before soft freeze...
PATCH v6 2/4 qcow2: rework the cluster compression routine
In reply to Denis PlotnikovThe patch enables processing the image compression type defined for the image and chooses an appropriate method for image clusters (de)compression.
--- block/qcow2-threads.c | 71 ++++++++++++++++++++++++++++++++++++------- 1 file changed, 60 insertions(+), 11 deletions(-)
block/qcow2-threads.c +60 -11
@@ -74,7 +74,9 @@ typedef struct Qcow2CompressData { } Qcow2CompressData;/* -* qcow2_compress() +* qcow2_zlib_compress() +* +* Compress @src_size bytes of data using zlib compression method * * @dest - destination buffer, @dest_size bytes * @src - source buffer, @src_size bytes
@@ -83,8 +85,8 @@ typedef struct Qcow2CompressData {* -ENOMEM destination buffer is not enough to store compressed data * -EIO on any other error */ -static ssize_t qcow2_compress (void *dest,size_t dest_size, -const void *src,size_t src_size) +static ssize_t qcow2_zlib_compress (void *dest,size_t dest_size, +const void *src,size_t src_size) {ssize_t ret;z_stream strm;
@@ -119,10 +121,10 @@ static ssize_t qcow2_compress(void *dest, size_t dest_size, }/* -* qcow2_decompress() +* qcow2_zlib_decompress() * * Decompress some data (not more than @src_size bytes) to produce exactly -* @dest_size bytes. +* @dest_size bytes using zlib compression method * * @dest - destination buffer, @dest_size bytes * @src - source buffer, @src_size bytes
@@ -130,8 +132,8 @@ static ssize_t qcow2_compress(void *dest, size_t dest_size,* Returns: 0 on success * -EIO on fail */ -static ssize_t qcow2_decompress (void *dest,size_t dest_size, -const void *src,size_t src_size) +static ssize_t qcow2_zlib_decompress (void *dest,size_t dest_size, +const void *src,size_t src_size) {int ret;z_stream strm;
@@ -191,20 +193,67 @@ qcow2_co_do_compress(BlockDriverState *bs, void *dest, size_t dest_size,return arg.ret; } +/* +* qcow2_co_compress() +* +* Compress @src_size bytes of data using the compression +* method defined by the image compression type +* +* @dest - destination buffer, @dest_size bytes +* @src - source buffer, @src_size bytes +* +* Returns: compressed size on success +* a negative error code on failure +*/ ssize_t coroutine_fnqcow2_co_compress (BlockDriverState *bs,void *dest,size_t dest_size,const void *src,size_t src_size) { -return qcow2_co_do_compress (bs, dest, dest_size, src, src_size, - qcow2_compress); +BDRVQcow2State *s = bs->opaque; +Qcow2CompressFunc fn; + +switch (s->compression_type) { +case QCOW2_COMPRESSION_TYPE_ZLIB : + fn = qcow2_zlib_compress; +break ; + +default : +abort (); + } + +return qcow2_co_do_compress (bs, dest, dest_size, src, src_size, fn); } +/* +* qcow2_co_decompress() +* +* Decompress some data (not more than @src_size bytes) to produce exactly +* @dest_size bytes using the compression method defined by the image +* compression type +* +* @dest - destination buffer, @dest_size bytes +* @src - source buffer, @src_size bytes +* +* Returns: 0 on success +* a negative error code on failure +*/ ssize_t coroutine_fnqcow2_co_decompress (BlockDriverState *bs,void *dest,size_t dest_size,const void *src,size_t src_size) { -return qcow2_co_do_compress (bs, dest, dest_size, src, src_size, - qcow2_decompress); +BDRVQcow2State *s = bs->opaque; +Qcow2CompressFunc fn; + +switch (s->compression_type) { +case QCOW2_COMPRESSION_TYPE_ZLIB : + fn = qcow2_zlib_decompress; +break ; + +default : +abort (); + } + +return qcow2_co_do_compress (bs, dest, dest_size, src, src_size, fn); }
Signature · 1 line
2.17.0
PATCH v6 3/4 qcow2: add zstd cluster compression
In reply to Denis Plotnikovzstd significantly reduces cluster compression time. It provides better compression performance maintaining the same level of the compression ratio in comparison with zlib, which, at the moment, is the only compression method available.
The performance test results:
Test compresses and decompresses qemu qcow2 image with just installed rhel-7.6 guest.
Image cluster size: 64K. Image on disk size: 2.2G
The test was conducted with brd disk to reduce the influence of disk subsystem to the test results.
The results is given in seconds.
compress cmd:
time ./qemu-img convert -O qcow2 -c -o compression_type=[zlib|zstd]
src.img [zlib|zstd]_compressed.img
decompress cmd
time ./qemu-img convert -O qcow2
[zlib|zstd]_compressed.img uncompressed.img
compression decompression
zlib zstd zlib zstd
------------------------------------------------------------
real 65.5 16.3 (-75 %) 1.9 1.6 (-16 %) user 65.0 15.8 5.3 2.5 sys 3.3 0.2 2.0 2.0
Both ZLIB and ZSTD gave the same compression ratio: 1.57 compressed image size in both cases: 1.4G
QAPI part:
--- docs/interop/qcow2.txt | 20 +++++++ configure | 2 +- qapi/block-core.json | 3 +- block/qcow2-threads.c | 124 +++++++++++++++++++++++++++++++++++++++++ block/qcow2.c | 11 ++++ 5 files changed, 158 insertions(+), 2 deletions(-)
docs/interop/qcow2.txt +20 -0
@@ -208,6 +208,7 @@ version 2. Available compression type values: 0: zlib <https://www.zlib.net/> + 1: zstd <http://github.com/facebook/zstd> === Header padding ===
@@ -575,11 +576,30 @@ Compressed Clusters Descriptor (x = 62 - (cluster_bits - 8)): Another compressed cluster may map to the tail of the final sector used by this compressed cluster. + The layout of the compressed data depends on the compression + type used for the image (see compressed cluster layout). + If a cluster is unallocated, read requests shall read the data from the backing file (except if bit 0 in the Standard Cluster Descriptor is set). If there is no backing file or the backing file is smaller than the image, they shall read zeros for all parts that are not covered by the backing file. +=== Compressed Cluster Layout === + +The compressed cluster data has a layout depending on the compression +type used for the image, as follows: + +Compressed data layout for the available compression types: +data_space_lenght - data chunk length available to store a compressed cluster. +(for more details see "Compressed Clusters Descriptor") +x = data_space_length - 1 + + 0: (default) zlib <http://zlib.net/>: + Byte 0 - x: the compressed data content + all the space provided used for compressed data + 1: zstd <http://github.com/facebook/zstd>: + Byte 0 - 3: the length of compressed data in bytes + 4 - x: the compressed data content == Snapshots ==
configure +1 -1
@@ -1835,7 +1835,7 @@ disabled with --disable-FEATURE, default is enabled if available: lzfse support of lzfse compression library (for reading lzfse-compressed dmg images) zstd support for zstd compression library - (for migration compression) + (for migration compression and qcow2 cluster compression) seccomp seccomp support coroutine-pool coroutine freelist (better performance) glusterfs GlusterFS backend
qapi/block-core.json +2 -1
@@ -4401,11 +4401,12 @@ # Compression type used in qcow2 image file # # @zlib: zlib compression, see <http://zlib.net/> +# @zstd: zstd compression, see <http://github.com/facebook/zstd> # # Since: 5.0 ## { 'enum': 'Qcow2CompressionType', - 'data': [ 'zlib' ] } + 'data': [ 'zlib', { 'name': 'zstd', 'if': 'defined(CONFIG_ZSTD)' } ] } ## # @BlockdevCreateOptionsQcow2:
block/qcow2-threads.c +124 -0
148 lines not on this page · raw message
block/qcow2.c +11 -0
@@ -1246,6 +1246,9 @@ static int validate_compression_type(BDRVQcow2State *s, Error **errp) {switch (s->compression_type) {case QCOW2_COMPRESSION_TYPE_ZLIB : +#ifdef CONFIG_ZSTD +case QCOW2_COMPRESSION_TYPE_ZSTD : +#endif break ;default :
@@ -1279,6 +1282,10 @@ static int qcow2_compression_type_from_format(const char *ct) {if (g_str_equal (ct,"zlib" )) {return QCOW2_COMPRESSION_TYPE_ZLIB ; +#ifdef CONFIG_ZSTD + }else if (g_str_equal (ct,"zstd" )) { +return QCOW2_COMPRESSION_TYPE_ZSTD ; +#endif }else {return -EINVAL ; }
@@ -3463,6 +3470,10 @@ qcow2_co_create(BlockdevCreateOptions *create_options, Error **errp) }switch (qcow2_opts->compression_type) { +#ifdef CONFIG_ZSTD +case QCOW2_COMPRESSION_TYPE_ZSTD : +break ; +#endif default :error_setg (errp,"Unknown compression type" ); goto out;
Signature · 1 line
2.17.0
Re: [PATCH v6 3/4] qcow2: add zstd cluster compression
Quoted 25 lines · Denis Plotnikov
zstd significantly reduces cluster compression time. It provides better compression performance maintaining the same level of the compression ratio in comparison with zlib, which, at the moment, is the only compression method available.
+++
b/docs/interop/qcow2.txtdocs/interop/qcow2.txt +4
-0@@ -208,6 +208,7 @@ version 2. Available compression type values: 0: zlib <https://www.zlib.net/> + 1: zstd <http://github.com/facebook/zstd> === Header padding ===@@ -575,11 +576,30 @@ Compressed Clusters Descriptor (x = 62 - (cluster_bits - 8)): Another compressed cluster may map to the tail of the final sector used by this compressed cluster. + The layout of the compressed data depends on the compression + type used for the image (see compressed cluster layout). + If a cluster is unallocated, read requests shall read the data from the backing file (except if bit 0 in the Standard Cluster Descriptor is set). If there is
docs/interop/qcow2.txt Compressed Clusters Descriptor (x = 62 - (cluster_bits - 8)): · line 593
Quoted 10 lines
docs/interop/qcow2.txt +7
-0no backing file or the backing file is smaller than the image, they shall read zeros for all parts that are not covered by the backing file. +=== Compressed Cluster Layout === + +The compressed cluster data has a layout depending on the compression +type used for the image, as follows: + +Compressed data layout for the available compression types: +data_space_lenght - data chunk length available to store a compressed cluster.
length
docs/interop/qcow2.txt Compressed Clusters Descriptor (x = 62 - (cluster_bits - 8)): · line 595
Quoted 2 lines
docs/interop/qcow2.txt +2
-0+(for more details see "Compressed Clusters Descriptor") +x = data_space_length - 1
If I understand correctly, data_space_length is really an upper bounds on the length available, because it is computed by rounding UP to the next 512-byte boundary (that is, the L2 descriptor lists the number of additional sectors used in storing the compressed data). Which really means that we have the following, where + is cluster boundaries, S and E are the start and end of the compressed data, and D is the offset determined by data_space_length:
+-------+-------+------+
S============E...D
docs/interop/qcow2.txt Compressed Clusters Descriptor (x = 62 - (cluster_bits - 8)): · line 599
Quoted 4 lines
docs/interop/qcow2.txt +4
-0+ + 0: (default) zlib <http://zlib.net/>: + Byte 0 - x: the compressed data content + all the space provided used for compressed data
For zlib, we have byte 0-E are compressed data, and bytes (E+1)-D (if any) are ignored. There is no way to tell how many bytes between E and D exist, because zlib doesn't care (the compression stream itself ensures that decompression stops when input reaches E because the output reached a cluster boundary at that point).
docs/interop/qcow2.txt Compressed Clusters Descriptor (x = 62 - (cluster_bits - 8)): · line 602
Quoted 3 lines
docs/interop/qcow2.txt +3
-0+ 1: zstd <http://github.com/facebook/zstd>: + Byte 0 - 3: the length of compressed data in bytes + 4 - x: the compressed data content
Whereas for zstd, the decompression MUST know the actual location of E, rather than passing in the slop between E and D; bytes 0-3 give us that information.
But your description is not very accurate: if 'x' is point E, then it is NOT data_space_length - 1, but rather data_space_length - slop, where slop can be up to 511 bytes (the number of bytes from (E+1) to D). And if 'x' is point E, then the real layout for zlib is:
byte 0 - E: the compressed data content
byte E+1 - x: ignored slop (E is implied solely by the compressed data)
and for zstd is:
byte 0 - 3: the length of the compressed data byte 4 - E: the compressed data (E computed from byte 0-3) byte E+1 - x: ignored
I'm not sure what the best way is to document this.
Quoted 14 lines
block/qcow2-threads.c +13
-0+++ b/block/qcow2-threads.c +static ssize_t qcow2_zstd_compress (void *dest,size_t dest_size, +const void *src,size_t src_size) +{ +size_t ret; + +/* +* steal ZSTD_LEN_BUF bytes in the very beginning of the buffer +* to store compressed chunk size +*/ +char *d_buf = ((char *) dest) +ZSTD_LEN_BUF ; + +/*
block/qcow2-threads.c static ssize_t qcow2_zlib_decompress(void *dest, size_t dest_size, · line 212
Quoted 10 lines
block/qcow2-threads.c +10
-0+* sanity check that we can store the compressed data length, +* and there is some space left for the compressor buffer +*/ +if (dest_size <=ZSTD_LEN_BUF ) { +return -ENOMEM ; + } + + dest_size -=ZSTD_LEN_BUF ; + + ret =ZSTD_compress (d_buf, dest_size, src, src_size,5 );
Where does the magic number 5 come from?
Quoted 6 lines
block/qcow2-threads.c +6
-0+ +if (ZSTD_isError (ret)) { +if (ZSTD_getErrorCode (ret) == ZSTD_error_dstSize_tooSmall) { +return -ENOMEM ; + }else { +return -EIO ;
block/qcow2-threads.c static ssize_t qcow2_zlib_decompress(void *dest, size_t dest_size, · line 228
Quoted 10 lines
block/qcow2-threads.c +10
-0+ } + } + +/* +* paranoid sanity check that we can store +* the compressed size in the first 4 bytes +*/ +if (ret >UINT32_MAX ) { +return -ENOMEM ; + }
The if is awkward. I'd prefer to change this to:
/*
* Our largest cluster is 2M, and we insist that compression
* actually compressed things.
*/
assert(ret < UINT32_MAX);or even tighten to assert(ret <= dest_size)
Quoted 37 lines
block/qcow2-threads.c +37
-0+ +/* store the compressed chunk size in the very beginning of the buffer */ +stl_be_p (dest, ret); + +return ret +ZSTD_LEN_BUF ; +} + +/* +* qcow2_zstd_decompress() +* +* Decompress some data (not more than @src_size bytes) to produce exactly +* @dest_size bytes using zstd compression method +* +* @dest - destination buffer, @dest_size bytes +* @src - source buffer, @src_size bytes +* +* Returns: 0 on success +* -EIO on any error +*/ +static ssize_t qcow2_zstd_decompress (void *dest,size_t dest_size, +const void *src,size_t src_size) +{ +/* +* zstd decompress wants to know the exact length of the data. +* For that purpose, on compression, the length is stored in +* the very beginning of the compressed buffer +*/ +size_t s_size; +const char *s_buf = ((const char *) src) +ZSTD_LEN_BUF ; + +/* +* sanity check that we can read 4 byte the content length and +* and there is some content to decompress +*/ +if (src_size <=ZSTD_LEN_BUF ) { +return -EIO ; + }
block/qcow2-threads.c static ssize_t qcow2_zlib_decompress(void *dest, size_t dest_size, · line 275
Quoted 10 lines
block/qcow2-threads.c +10
-0+ + s_size =ldl_be_p (src); + +/* sanity check that the buffer is big enough to read the content from */ +if (src_size -ZSTD_LEN_BUF < s_size) { +return -EIO ; + } + +if (ZSTD_isError ( +ZSTD_decompress (dest, dest_size, s_buf, s_size))) {
You are correct that ZSTD_decompress() is picky that it must be given the exact size of the compressed buffer it is decompressing. But the ZSTD manual mentions that if an exact size is not known in advance, that the streaming API can be used instead:
https://facebook.github.io/zstd/zstd_manual.html#Chapter9
In other words, would it be possible to NOT have to prepend four bytes of exact size information, by instead setting up decompression via the streaming API where the input is (usually) oversized, but the output buffer limited to exactly one cluster is sufficient to consume the exact compressed data and ignore the slop, just as we do in zlib?
The rest of this patch looks okay.
Signature · 3 lines
Eric Blake, Principal Software Engineer Red Hat, Inc. +1-919-301-3226 Virtualization: qemu.org | libvirt.org
Re: [PATCH v6 3/4] qcow2: add zstd cluster compression
Quoted 90 lines · Eric Blake
Quoted 32 lines · Denis Plotnikov
zstd significantly reduces cluster compression time. It provides better compression performance maintaining the same level of the compression ratio in comparison with zlib, which, at the moment, is the only compression method available.
+++
b/docs/interop/qcow2.txt(quoted diff) +0
-0@@ -208,6 +208,7 @@ version 2.Available compression type values:
0: zlib <https://www.zlib.net/>
+ 1: zstd <http://github.com/facebook/zstd>
=== Header padding ===(quoted diff) +0
-0@@ -575,11 +576,30 @@ Compressed Clusters Descriptor (x = 62 -(
cluster_bits- 8)):
Another compressed cluster may map to the tail of the final sector used by this compressed cluster.
+ The layout of the compressed data depends on the compression + type used for the image (see compressed cluster layout).
+
If a cluster is unallocated, read requests shall read the data from the backing file (except if bit 0 in the Standard Cluster Descriptor is set). If there is no backing file or the backing file is smaller than the image, they shall read zeros for all parts that are not covered by the backing file.
+=== Compressed Cluster Layout ===docs/interop/qcow2.txt +6
-0+ +The compressed cluster data has a layout depending on the compression +type used for the image, as follows: + +Compressed data layout for the available compression types: +data_space_lenght - data chunk length available to store aQuoted 1 line
compressed cluster.
length
docs/interop/qcow2.txt Compressed Clusters Descriptor (x = 62 - (cluster_bits - 8)): · line 595
Quoted 2 lines
docs/interop/qcow2.txt +2
-0+(for more details see "Compressed Clusters Descriptor") +x = data_space_length - 1If I understand correctly,
data_space_lengthis really an upper bounds on the length available, because it is computed by rounding UP to the next 512-byte boundary (that is, the L2 descriptor lists the number of additional sectors used in storing the compressed data). Which really means that we have the following, where + is cluster boundaries, S and E are the start and end of the compressed data, and D is the offset determined bydata_space_length:+-------+-------+------+
S============E...DQuoted 5 lines
+
+ 0: (default) zlib <http://zlib.net/>:
+ Byte 0 - x: the compressed data content + all the space provided used for compressed dataFor zlib, we have byte 0-E are compressed data, and bytes (E+1)-D (if any) are ignored. There is no way to tell how many bytes between E and D exist, because zlib doesn't care (the compression stream itself ensures that decompression stops when input reaches E because the output reached a cluster boundary at that point).
Quoted 3 lines
+ 1: zstd <http://github.com/facebook/zstd>:
+ Byte 0 - 3: the length of compressed data in bytes + 4 - x: the compressed data contentWhereas for zstd, the decompression MUST know the actual location of E, rather than passing in the slop between E and D; bytes 0-3 give us that information.
But your description is not very accurate: if 'x' is point E, then it is NOT
data_space_length- 1, but ratherdata_space_length- slop, where slop can be up to 511 bytes (the number of bytes from (E+1) to D). And if 'x' is point E, then the real layout for zlib is:byte 0 - E: the compressed data content
byte E+1 - x: ignored slop (E is implied solely by the compressed data)and for zstd is:
byte 0 - 3: the length of the compressed data byte 4 - E: the compressed data (E computed from byte 0-3) byte E+1 - x: ignored
I'm not sure what the best way is to document this.
Quoted 18 lines
+++
b/block/qcow2-threads.cblock/qcow2-threads.c +16
-0+static ssize_t qcow2_zstd_compress (void *dest,size_t dest_size, + const void *src, size_t src_size) +{ + size_t ret; + + /* + * steal ZSTD_LEN_BUF bytes in the very beginning of the buffer + * to store compressed chunk size + */ + char *d_buf = ((char *) dest) + ZSTD_LEN_BUF; + + /* + * sanity check that we can store the compressed data length, + * and there is some space left for the compressor buffer + */ + if (dest_size <= ZSTD_LEN_BUF) {
Quoted 8 lines
block/qcow2-threads.c static ssize_t qcow2_zlib_decompress(void *dest, size_t dest_size, · line 212
Quoted 6 lines
+ return
-ENOMEM;
+ }block/qcow2-threads.c +4
-0+ + dest_size -= ZSTD_LEN_BUF; + + ret = ZSTD_compress(d_buf, dest_size, src, src_size, 5);Where does the magic number 5 come from?
I did some tests to get the same compression ratio as zlib but do it faster than zlib.
ZLIB also used hardcoded "compression ratio". Changing of the compression ratios in both compression types is something that can be changed with later patches.
Quoted 82 lines
Quoted 8 lines
block/qcow2-threads.c +6
-0+ + if (ZSTD_isError(ret)) { + if (ZSTD_getErrorCode(ret) == ZSTD_error_dstSize_tooSmall) { + return -ENOMEM; + } else { + return -EIO;block/qcow2-threads.c static ssize_t qcow2_zlib_decompress(void *dest, size_t dest_size, · line 228
Quoted 10 lines
block/qcow2-threads.c +10
-0+ } + } + + /* + * paranoid sanity check that we can store + * the compressed size in the first 4 bytes + */ + if (ret > UINT32_MAX) { + return -ENOMEM; + }The if is awkward. I'd prefer to change this to:
/*
* Our largest cluster is 2M, and we insist that compression
* actually compressed things.
*/
assert(ret < UINT32_MAX);or even tighten to
assert(ret <= dest_size)Quoted 43 lines
+
+ /* store the compressed chunk size in the very beginning of the buffer */
+stl_be_p(dest, ret);block/qcow2-threads.c +8
-0+ + return ret + ZSTD_LEN_BUF; +} + +/* +* qcow2_zstd_decompress() +* +* Decompress some data (not more than @src_size bytes) to produce exactly
+ * @dest_size bytes using zstd compression method
+ *
+ * @dest - destination buffer, @dest_size bytes
+ * @src - source buffer, @src_size bytes
+ *
+ * Returns: 0 on success
+ *-EIOon any error
+ */block/qcow2-threads.c +22
-0+static ssize_t qcow2_zstd_decompress (void *dest,size_t dest_size, + const void *src, size_t src_size) +{ + /* + * zstd decompress wants to know the exact length of the data. + * For that purpose, on compression, the length is stored in + * the very beginning of the compressed buffer + */ + size_t s_size; + const char *s_buf = ((const char *) src) + ZSTD_LEN_BUF; + + /* + * sanity check that we can read 4 byte the content length and + * and there is some content to decompress + */ + if (src_size <= ZSTD_LEN_BUF) { + return -EIO; + } + + s_size = ldl_be_p(src); + + /* sanity check that the buffer is big enough to read theQuoted 7 lines
content from */
+ if (src_size-ZSTD_LEN_BUF<s_size) {
+ return-EIO;
+ }
+
+ if (ZSTD_isError(
+ZSTD_decompress(dest, dest_size, s_buf, s_size))) {You are correct that
ZSTD_decompress()is picky that it must be given the exact size of the compressed buffer it is decompressing. But the ZSTD manual mentions that if an exact size is not known in advance, that the streaming API can be used instead:
To be honest, I didn't find where they mentioned that explicitly. Could you please point where exactly?
But I found the following:
Calling ZSTD_compressStream2() with ZSTD_e_end instructs to finish a frame.
It will perform a flush and write frame epilogue.
The epilogue is required for decoders to consider a frame completed.
flush operation is the same, and follows same rules as calling ZSTD_compressStream2() with ZSTD_e_flush. You must continue calling ZSTD_compressStream2() with ZSTD_e_end until it returns 0, at which point you are free to start a new frame
I think in the epilogue they store the same information that I did and potentially (I didn't check) some more to finish the frame. So we didn't win any space. Additionally, using streaming API will make the code more complex.
So I decided to stick with more simple version.
Quoted 5 lines
In other words, would it be possible to NOT have to prepend four bytes of exact size information, by instead setting up decompression via the streaming API where the input is (usually) oversized, but the output buffer limited to exactly one cluster is sufficient to consume the exact compressed data and ignore the slop, just as we do in zlib?
The rest of this patch looks okay.
PATCH v6 4/4 iotests: 287: add qcow2 compression type test
In reply to Denis PlotnikovThe test checks fulfilling qcow2 requiriements for the compression type feature and zstd compression type operability.
--- tests/qemu-iotests/287 | 128 +++++++++++++++++++++++++++++++++++++ tests/qemu-iotests/287.out | 43 +++++++++++++ tests/qemu-iotests/group | 1 + 3 files changed, 172 insertions(+) create mode 100755 tests/qemu-iotests/287 create mode 100644 tests/qemu-iotests/287.out
tests/qemu-iotests/287 new file +128 -0
@@ -0,0 +1,128 @@ +#!/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 [email protected] + +seq="$(basename $0)" +echo "QA output created by $seq" + +status=1 # failure is the default! + +_cleanup() +{ + _cleanup_test_img +} +trap "_cleanup; exit \$status" 0 1 2 3 15 + +# standard environment +. ./common.rc +. ./common.filter + +# This tests qocw2-specific low-level functionality +_supported_fmt qcow2 +_supported_proto file +_supported_os Linux + +# for all the cases +CLUSTER_SIZE=65536 + +# Check if we can run this test. + +IMGOPTS='compression_type=zstd' _make_test_img 64M | grep "Invalid parameter 'zstd'" 2>&1 1>/dev/null + +ZSTD_SUPPORTED=$? + +if (($ZSTD_SUPPORTED==0)); then + _notrun "ZSTD is disabled" +fi + +# Test: when compression is zlib the incompatible bit is unset +echo +echo "=== Testing compression type incompatible bit setting for zlib ===" +echo + +IMGOPTS='compression_type=zlib' _make_test_img 64M +$PYTHON qcow2.py "$TEST_IMG" dump-header | grep incompatible_features + +# Test: when compression differs from zlib the incompatible bit is set +echo +echo "=== Testing compression type incompatible bit setting for zstd ===" +echo + +IMGOPTS='compression_type=zstd' _make_test_img 64M +$PYTHON qcow2.py "$TEST_IMG" dump-header | grep incompatible_features + +# Test: an image can't be openned if compression type is zlib and +# incompatible feature compression type is set +echo +echo "=== Testing zlib with incompatible bit set ===" +echo + +IMGOPTS='compression_type=zlib' _make_test_img 64M +$PYTHON qcow2.py "$TEST_IMG" set-feature-bit incompatible 3 +# to make sure the bit was actually set +$PYTHON qcow2.py "$TEST_IMG" dump-header | grep incompatible_features +$QEMU_IMG info "$TEST_IMG" 2>1 1>/dev/null +if (($?==0)); then + echo "Error: The image openned successfully. The image must not be openned" +fi + +# Test: an image can't be openned if compression type is NOT zlib and +# incompatible feature compression type is UNSET +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 +$QEMU_IMG info "$TEST_IMG" 2>1 1>/dev/null +if (($?==0)); then + echo "Error: The image openned successfully. The image must not be openned" +fi +# Test: check compression type values +echo +echo "=== Testing compression type values ===" +echo +# zlib=0 +IMGOPTS='compression_type=zlib' _make_test_img 64M +od -j104 -N1 -An -vtu1 "$TEST_IMG" + +# zstd=1 +IMGOPTS='compression_type=zstd' _make_test_img 64M +od -j104 -N1 -An -vtu1 "$TEST_IMG" + +# Test: using zstd compression, write to and read from an image +echo +echo "=== Testing reading and writing with zstd ===" +echo + +IMGOPTS='compression_type=zstd' _make_test_img 64M +$QEMU_IO -c "write -c -P 0xAC 65536 64k " "$TEST_IMG" | _filter_qemu_io +$QEMU_IO -c "read -P 0xAC 65536 65536 " "$TEST_IMG" | _filter_qemu_io +$QEMU_IO -c "read -v 131070 8 " "$TEST_IMG" | _filter_qemu_io +$QEMU_IO -c "read -v 65534 8" "$TEST_IMG" | _filter_qemu_io + +# success, all done +echo "*** done" +rm -f $seq.full +status=0
tests/qemu-iotests/287.out new file +43 -0
43 lines not on this page · raw message
tests/qemu-iotests/group +1 -0
@@ -293,3 +293,4 @@ 283 auto quick 284 rw 286 rw quick +287 auto quick
Signature · 1 line
2.17.0