Merge remote-tracking branch 'kwolf/for-anthony' into staging

* kwolf/for-anthony:
  qemu-img: use QemuOpts instead of QEMUOptionParameter in resize function
  qemu-iotests: Be more flexible with image creation options
  qemu-iotests: add 039 qcow2 lazy refcounts test
  qemu-io: add "abort" command to simulate program crash
  qcow2: implement lazy refcounts
  qemu-iotests: ignore qemu-img create lazy_refcounts output
  docs: add lazy refcounts bit to qcow2 specification
  qcow2: introduce dirty bit
  docs: add dirty bit to qcow2 specification
  qemu-iotests: add qed.py image manipulation utility
  qapi: generalize documentation of streaming commands
  ide scsi: Mess with geometry only for hard disk devices

block/qcow2-cluster.c

@@ -662,7 +662,10 @@ int qcow2_alloc_cluster_link_l2(BlockDriverState *bs, QCowL2Meta *m)
         qcow2_cache_depends_on_flush(s->l2_table_cache);
     }
 
-    qcow2_cache_set_dependency(bs, s->l2_table_cache, s->refcount_block_cache);
+    if (qcow2_need_accurate_refcounts(s)) {
+        qcow2_cache_set_dependency(bs, s->l2_table_cache,
+                                   s->refcount_block_cache);
+    }
     ret = get_cluster_table(bs, m->offset, &l2_table, &l2_index);
     if (ret < 0) {
         goto err;

block/qcow2.c

@@ -214,6 +214,62 @@ static void report_unsupported_feature(BlockDriverState *bs,
     }
 }
 
+/*
+ * Sets the dirty bit and flushes afterwards if necessary.
+ *
+ * The incompatible_features bit is only set if the image file header was
+ * updated successfully.  Therefore it is not required to check the return
+ * value of this function.
+ */
+static int qcow2_mark_dirty(BlockDriverState *bs)
+{
+    BDRVQcowState *s = bs->opaque;
+    uint64_t val;
+    int ret;
+
+    assert(s->qcow_version >= 3);
+
+    if (s->incompatible_features & QCOW2_INCOMPAT_DIRTY) {
+        return 0; /* already dirty */
+    }
+
+    val = cpu_to_be64(s->incompatible_features | QCOW2_INCOMPAT_DIRTY);
+    ret = bdrv_pwrite(bs->file, offsetof(QCowHeader, incompatible_features),
+                      &val, sizeof(val));
+    if (ret < 0) {
+        return ret;
+    }
+    ret = bdrv_flush(bs->file);
+    if (ret < 0) {
+        return ret;
+    }
+
+    /* Only treat image as dirty if the header was updated successfully */
+    s->incompatible_features |= QCOW2_INCOMPAT_DIRTY;
+    return 0;
+}
+
+/*
+ * Clears the dirty bit and flushes before if necessary.  Only call this
+ * function when there are no pending requests, it does not guard against
+ * concurrent requests dirtying the image.
+ */
+static int qcow2_mark_clean(BlockDriverState *bs)
+{
+    BDRVQcowState *s = bs->opaque;
+
+    if (s->incompatible_features & QCOW2_INCOMPAT_DIRTY) {
+        int ret = bdrv_flush(bs);
+        if (ret < 0) {
+            return ret;
+        }
+
+        s->incompatible_features &= ~QCOW2_INCOMPAT_DIRTY;
+        return qcow2_update_header(bs);
+    }
+    return 0;
+}
+
 static int qcow2_open(BlockDriverState *bs, int flags)
 {
     BDRVQcowState *s = bs->opaque;
@@ -287,12 +343,13 @@ static int qcow2_open(BlockDriverState *bs, int flags)
     s->compatible_features      = header.compatible_features;
     s->autoclear_features       = header.autoclear_features;
 
-    if (s->incompatible_features != 0) {
+    if (s->incompatible_features & ~QCOW2_INCOMPAT_MASK) {
         void *feature_table = NULL;
         qcow2_read_extensions(bs, header.header_length, ext_end,
                               &feature_table);
         report_unsupported_feature(bs, feature_table,
-                                   s->incompatible_features);
+                                   s->incompatible_features &
+                                   ~QCOW2_INCOMPAT_MASK);
         ret = -ENOTSUP;
         goto fail;
     }
@@ -412,6 +469,22 @@ static int qcow2_open(BlockDriverState *bs, int flags)
     /* Initialise locks */
     qemu_co_mutex_init(&s->lock);
 
+    /* Repair image if dirty */
+    if ((s->incompatible_features & QCOW2_INCOMPAT_DIRTY) &&
+        !bs->read_only) {
+        BdrvCheckResult result = {0};
+
+        ret = qcow2_check_refcounts(bs, &result, BDRV_FIX_ERRORS);
+        if (ret < 0) {
+            goto fail;
+        }
+
+        ret = qcow2_mark_clean(bs);
+        if (ret < 0) {
+            goto fail;
+        }
+    }
+
 #ifdef DEBUG_ALLOC
     {
         BdrvCheckResult result = {0};
@@ -714,6 +787,11 @@ static coroutine_fn int qcow2_co_writev(BlockDriverState *bs,
             goto fail;
         }
 
+        if (l2meta.nb_clusters > 0 &&
+            (s->compatible_features & QCOW2_COMPAT_LAZY_REFCOUNTS)) {
+            qcow2_mark_dirty(bs);
+        }
+
         cluster_offset = l2meta.cluster_offset;
         assert((cluster_offset & 511) == 0);
 
@@ -785,6 +863,8 @@ static void qcow2_close(BlockDriverState *bs)
     qcow2_cache_flush(bs, s->l2_table_cache);
     qcow2_cache_flush(bs, s->refcount_block_cache);
 
+    qcow2_mark_clean(bs);
+
     qcow2_cache_destroy(bs, s->l2_table_cache);
     qcow2_cache_destroy(bs, s->refcount_block_cache);
 
@@ -949,7 +1029,16 @@ int qcow2_update_header(BlockDriverState *bs)
 
     /* Feature table */
     Qcow2Feature features[] = {
-        /* no feature defined yet */
+        {
+            .type = QCOW2_FEAT_TYPE_INCOMPATIBLE,
+            .bit  = QCOW2_INCOMPAT_DIRTY_BITNR,
+            .name = "dirty bit",
+        },
+        {
+            .type = QCOW2_FEAT_TYPE_COMPATIBLE,
+            .bit  = QCOW2_COMPAT_LAZY_REFCOUNTS_BITNR,
+            .name = "lazy refcounts",
+        },
     };
 
     ret = header_ext_add(buf, QCOW2_EXT_MAGIC_FEATURE_TABLE,
@@ -1132,6 +1221,11 @@ static int qcow2_create2(const char *filename, int64_t total_size,
         header.crypt_method = cpu_to_be32(QCOW_CRYPT_NONE);
     }
 
+    if (flags & BLOCK_FLAG_LAZY_REFCOUNTS) {
+        header.compatible_features |=
+            cpu_to_be64(QCOW2_COMPAT_LAZY_REFCOUNTS);
+    }
+
     ret = bdrv_pwrite(bs, 0, &header, sizeof(header));
     if (ret < 0) {
         goto out;
@@ -1245,6 +1339,8 @@ static int qcow2_create(const char *filename, QEMUOptionParameter *options)
                     options->value.s);
                 return -EINVAL;
             }
+        } else if (!strcmp(options->name, BLOCK_OPT_LAZY_REFCOUNTS)) {
+            flags |= options->value.n ? BLOCK_FLAG_LAZY_REFCOUNTS : 0;
         }
         options++;
     }
@@ -1255,6 +1351,12 @@ static int qcow2_create(const char *filename, QEMUOptionParameter *options)
         return -EINVAL;
     }
 
+    if (version < 3 && (flags & BLOCK_FLAG_LAZY_REFCOUNTS)) {
+        fprintf(stderr, "Lazy refcounts only supported with compatibility "
+                "level 1.1 and above (use compat=1.1 or greater)\n");
+        return -EINVAL;
+    }
+
     return qcow2_create2(filename, sectors, backing_file, backing_fmt, flags,
                          cluster_size, prealloc, options, version);
 }
@@ -1441,10 +1543,12 @@ static coroutine_fn int qcow2_co_flush_to_os(BlockDriverState *bs)
         return ret;
     }
 
-    ret = qcow2_cache_flush(bs, s->refcount_block_cache);
-    if (ret < 0) {
-        qemu_co_mutex_unlock(&s->lock);
-        return ret;
+    if (qcow2_need_accurate_refcounts(s)) {
+        ret = qcow2_cache_flush(bs, s->refcount_block_cache);
+        if (ret < 0) {
+            qemu_co_mutex_unlock(&s->lock);
+            return ret;
+        }
     }
     qemu_co_mutex_unlock(&s->lock);
 
@@ -1559,6 +1663,11 @@ static QEMUOptionParameter qcow2_create_options[] = {
         .type = OPT_STRING,
         .help = "Preallocation mode (allowed values: off, metadata)"
     },
+    {
+        .name = BLOCK_OPT_LAZY_REFCOUNTS,
+        .type = OPT_FLAG,
+        .help = "Postpone refcount updates",
+    },
     { NULL }
 };
 

block/qcow2.h

@@ -110,6 +110,22 @@ enum {
     QCOW2_FEAT_TYPE_AUTOCLEAR       = 2,
 };
 
+/* Incompatible feature bits */
+enum {
+    QCOW2_INCOMPAT_DIRTY_BITNR   = 0,
+    QCOW2_INCOMPAT_DIRTY         = 1 << QCOW2_INCOMPAT_DIRTY_BITNR,
+
+    QCOW2_INCOMPAT_MASK          = QCOW2_INCOMPAT_DIRTY,
+};
+
+/* Compatible feature bits */
+enum {
+    QCOW2_COMPAT_LAZY_REFCOUNTS_BITNR = 0,
+    QCOW2_COMPAT_LAZY_REFCOUNTS       = 1 << QCOW2_COMPAT_LAZY_REFCOUNTS_BITNR,
+
+    QCOW2_COMPAT_FEAT_MASK            = QCOW2_COMPAT_LAZY_REFCOUNTS,
+};
+
 typedef struct Qcow2Feature {
     uint8_t type;
     uint8_t bit;
@@ -237,6 +253,11 @@ static inline int qcow2_get_cluster_type(uint64_t l2_entry)
     }
 }
 
+/* Check whether refcounts are eager or lazy */
+static inline bool qcow2_need_accurate_refcounts(BDRVQcowState *s)
+{
+    return !(s->incompatible_features & QCOW2_INCOMPAT_DIRTY);
+}
 
 // FIXME Need qcow2_ prefix to global functions
 

block_int.h

@@ -31,8 +31,9 @@
 #include "qemu-timer.h"
 #include "qapi-types.h"
 
-#define BLOCK_FLAG_ENCRYPT	1
-#define BLOCK_FLAG_COMPAT6	4
+#define BLOCK_FLAG_ENCRYPT          1
+#define BLOCK_FLAG_COMPAT6          4
+#define BLOCK_FLAG_LAZY_REFCOUNTS   8
 
 #define BLOCK_IO_LIMIT_READ     0
 #define BLOCK_IO_LIMIT_WRITE    1
@@ -41,16 +42,17 @@
 #define BLOCK_IO_SLICE_TIME     100000000
 #define NANOSECONDS_PER_SECOND  1000000000.0
 
-#define BLOCK_OPT_SIZE          "size"
-#define BLOCK_OPT_ENCRYPT       "encryption"
-#define BLOCK_OPT_COMPAT6       "compat6"
-#define BLOCK_OPT_BACKING_FILE  "backing_file"
-#define BLOCK_OPT_BACKING_FMT   "backing_fmt"
-#define BLOCK_OPT_CLUSTER_SIZE  "cluster_size"
-#define BLOCK_OPT_TABLE_SIZE    "table_size"
-#define BLOCK_OPT_PREALLOC      "preallocation"
-#define BLOCK_OPT_SUBFMT        "subformat"
-#define BLOCK_OPT_COMPAT_LEVEL  "compat"
+#define BLOCK_OPT_SIZE              "size"
+#define BLOCK_OPT_ENCRYPT           "encryption"
+#define BLOCK_OPT_COMPAT6           "compat6"
+#define BLOCK_OPT_BACKING_FILE      "backing_file"
+#define BLOCK_OPT_BACKING_FMT       "backing_fmt"
+#define BLOCK_OPT_CLUSTER_SIZE      "cluster_size"
+#define BLOCK_OPT_TABLE_SIZE        "table_size"
+#define BLOCK_OPT_PREALLOC          "preallocation"
+#define BLOCK_OPT_SUBFMT            "subformat"
+#define BLOCK_OPT_COMPAT_LEVEL      "compat"
+#define BLOCK_OPT_LAZY_REFCOUNTS    "lazy_refcounts"
 
 typedef struct BdrvTrackedRequest BdrvTrackedRequest;
 

docs/specs/qcow2.txt

@@ -75,13 +75,23 @@ in the description of a field.
                     Bitmask of incompatible features. An implementation must
                     fail to open an image if an unknown bit is set.
 
-                    Bits 0-63:  Reserved (set to 0)
+                    Bit 0:      Dirty bit.  If this bit is set then refcounts
+                                may be inconsistent, make sure to scan L1/L2
+                                tables to repair refcounts before accessing the
+                                image.
+
+                    Bits 1-63:  Reserved (set to 0)
 
          80 -  87:  compatible_features
                     Bitmask of compatible features. An implementation can
                     safely ignore any unknown bits that are set.
 
-                    Bits 0-63:  Reserved (set to 0)
+                    Bit 0:      Lazy refcounts bit.  If this bit is set then
+                                lazy refcount updates can be used.  This means
+                                marking the image file dirty and postponing
+                                refcount metadata updates.
+
+                    Bits 1-63:  Reserved (set to 0)
 
          88 -  95:  autoclear_features
                     Bitmask of auto-clear features. An implementation may only

hmp-commands.hx

@@ -101,7 +101,7 @@ ETEXI
         .name       = "block_job_cancel",
         .args_type  = "device:B",
         .params     = "device",
-        .help       = "stop an active block streaming operation",
+        .help       = "stop an active background block operation",
         .mhandler.cmd = hmp_block_job_cancel,
     },
 

hw/ide/qdev.c

@@ -149,7 +149,8 @@ static int ide_dev_initfn(IDEDevice *dev, IDEDriveKind kind)
     }
 
     blkconf_serial(&dev->conf, &dev->serial);
-    if (blkconf_geometry(&dev->conf, &dev->chs_trans, 65536, 16, 255) < 0) {
+    if (kind != IDE_CD
+        && blkconf_geometry(&dev->conf, &dev->chs_trans, 65536, 16, 255) < 0) {
         return -1;
     }
 

hw/scsi-disk.c

@@ -1958,7 +1958,8 @@ static int scsi_initfn(SCSIDevice *dev)
     }
 
     blkconf_serial(&s->qdev.conf, &s->serial);
-    if (blkconf_geometry(&dev->conf, NULL, 65535, 255, 255) < 0) {
+    if (dev->type == TYPE_DISK
+        && blkconf_geometry(&dev->conf, NULL, 65535, 255, 255) < 0) {
         return -1;
     }
 

qapi-schema.json

@@ -1660,7 +1660,7 @@
 # Returns: Nothing on success
 #          If the job type does not support throttling, NotSupported
 #          If the speed value is invalid, InvalidParameter
-#          If streaming is not active on this device, DeviceNotActive
+#          If no background operation is active on this device, DeviceNotActive
 #
 # Since: 1.1
 ##
@@ -1670,9 +1670,9 @@
 ##
 # @block-job-cancel:
 #
-# Stop an active block streaming operation.
+# Stop an active background block operation.
 #
-# This command returns immediately after marking the active block streaming
+# This command returns immediately after marking the active background block
 # operation for cancellation.  It is an error to call this command if no
 # operation is in progress.
 #
@@ -1680,16 +1680,15 @@
 # BLOCK_JOB_CANCELLED event.  Before that happens the job is still visible when
 # enumerated using query-block-jobs.
 #
-# The image file retains its backing file unless the streaming operation happens
-# to complete just as it is being cancelled.
-#
-# A new block streaming operation can be started at a later time to finish
-# copying all data from the backing file.
+# For streaming, the image file retains its backing file unless the streaming
+# operation happens to complete just as it is being cancelled.  A new streaming
+# operation can be started at a later time to finish copying all data from the
+# backing file.
 #
 # @device: the device name
 #
 # Returns: Nothing on success
-#          If streaming is not active on this device, DeviceNotActive
+#          If no background operation is active on this device, DeviceNotActive
 #          If cancellation already in progress, DeviceInUse
 #
 # Since: 1.1

qemu-img.c

@@ -1567,14 +1567,19 @@ static int img_resize(int argc, char **argv)
     const char *filename, *fmt, *size;
     int64_t n, total_size;
     BlockDriverState *bs = NULL;
-    QEMUOptionParameter *param;
-    QEMUOptionParameter resize_options[] = {
-        {
-            .name = BLOCK_OPT_SIZE,
-            .type = OPT_SIZE,
-            .help = "Virtual disk size"
+    QemuOpts *param;
+    static QemuOptsList resize_options = {
+        .name = "resize_options",
+        .head = QTAILQ_HEAD_INITIALIZER(resize_options.head),
+        .desc = {
+            {
+                .name = BLOCK_OPT_SIZE,
+                .type = QEMU_OPT_SIZE,
+                .help = "Virtual disk size"
+            }, {
+                /* end of list */
+            }
         },
-        { NULL }
     };
 
     /* Remove size from argv manually so that negative numbers are not treated
@@ -1624,14 +1629,15 @@ static int img_resize(int argc, char **argv)
     }
 
     /* Parse size */
-    param = parse_option_parameters("", resize_options, NULL);
-    if (set_option_parameter(param, BLOCK_OPT_SIZE, size)) {
+    param = qemu_opts_create(&resize_options, NULL, 0, NULL);
+    if (qemu_opt_set(param, BLOCK_OPT_SIZE, size)) {
         /* Error message already printed when size parsing fails */
         ret = -1;
+        qemu_opts_del(param);
         goto out;
     }
-    n = get_option_parameter(param, BLOCK_OPT_SIZE)->value.n;
-    free_option_parameters(param);
+    n = qemu_opt_get_size(param, BLOCK_OPT_SIZE, 0);
+    qemu_opts_del(param);
 
     bs = bdrv_new_open(filename, fmt, BDRV_O_FLAGS | BDRV_O_RDWR);
     if (!bs) {

qemu-io.c

@@ -1652,6 +1652,17 @@ static const cmdinfo_t map_cmd = {
        .oneline        = "prints the allocated areas of a file",
 };
 
+static int abort_f(int argc, char **argv)
+{
+    abort();
+}
+
+static const cmdinfo_t abort_cmd = {
+       .name           = "abort",
+       .cfunc          = abort_f,
+       .flags          = CMD_NOFILE_OK,
+       .oneline        = "simulate a program crash using abort(3)",
+};
 
 static int close_f(int argc, char **argv)
 {
@@ -1905,6 +1916,7 @@ int main(int argc, char **argv)
     add_command(&discard_cmd);
     add_command(&alloc_cmd);
     add_command(&map_cmd);
+    add_command(&abort_cmd);
 
     add_args_command(init_args_command);
     add_check_command(init_check_command);

tests/qemu-iotests/031.out

@@ -54,8 +54,8 @@ header_length             72
 
 Header extension:
 magic                     0x6803f857
-length                    0
-data                      ''
+length                    96
+data                      <binary>
 
 Header extension:
 magic                     0x12345678
@@ -68,7 +68,7 @@ No errors were found on the image.
 
 magic                     0x514649fb
 version                   2
-backing_file_offset       0x98
+backing_file_offset       0xf8
 backing_file_size         0x17
 cluster_bits              16
 size                      67108864
@@ -92,8 +92,8 @@ data                      'host_device'
 
 Header extension:
 magic                     0x6803f857
-length                    0
-data                      ''
+length                    96
+data                      <binary>
 
 Header extension:
 magic                     0x12345678
@@ -155,8 +155,8 @@ header_length             104
 
 Header extension:
 magic                     0x6803f857
-length                    0
-data                      ''
+length                    96
+data                      <binary>
 
 Header extension:
 magic                     0x12345678
@@ -169,7 +169,7 @@ No errors were found on the image.
 
 magic                     0x514649fb
 version                   3
-backing_file_offset       0xb8
+backing_file_offset       0x118
 backing_file_size         0x17
 cluster_bits              16
 size                      67108864
@@ -193,8 +193,8 @@ data                      'host_device'
 
 Header extension:
 magic                     0x6803f857
-length                    0
-data                      ''
+length                    96
+data                      <binary>
 
 Header extension:
 magic                     0x12345678

tests/qemu-iotests/036.out

@@ -46,7 +46,7 @@ header_length             104
 
 Header extension:
 magic                     0x6803f857
-length                    0
-data                      ''
+length                    96
+data                      <binary>
 
 *** done

tests/qemu-iotests/039

+#!/bin/bash
+#
+# Test qcow2 lazy refcounts
+#
+# Copyright (C) 2012 Red Hat, Inc.
+# Copyright IBM, Corp. 2010
+#
+# Based on test 038.
+#
+# 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=stefanha@linux.vnet.ibm.com
+
+seq=`basename $0`
+echo "QA output created by $seq"
+
+here=`pwd`
+tmp=/tmp/$$
+status=1	# failure is the default!
+
+_cleanup()
+{
+	_cleanup_test_img
+}
+trap "_cleanup; exit \$status" 0 1 2 3 15
+
+# get standard environment, filters and checks
+. ./common.rc
+. ./common.filter
+
+_supported_fmt qcow2
+_supported_proto generic
+_supported_os Linux
+
+size=128M
+
+echo
+echo "== Checking that image is clean on shutdown =="
+
+IMGOPTS="compat=1.1,lazy_refcounts=on"
+_make_test_img $size
+
+$QEMU_IO -c "write -P 0x5a 0 512" $TEST_IMG | _filter_qemu_io
+
+# The dirty bit must not be set
+./qcow2.py $TEST_IMG dump-header | grep incompatible_features
+_check_test_img
+
+echo
+echo "== Creating a dirty image file =="
+
+IMGOPTS="compat=1.1,lazy_refcounts=on"
+_make_test_img $size
+
+old_ulimit=$(ulimit -c)
+ulimit -c 0 # do not produce a core dump on abort(3)
+$QEMU_IO -c "write -P 0x5a 0 512" -c "abort" $TEST_IMG | _filter_qemu_io
+ulimit -c "$old_ulimit"
+
+# The dirty bit must be set
+./qcow2.py $TEST_IMG dump-header | grep incompatible_features
+_check_test_img
+
+echo
+echo "== Read-only access must still work =="
+
+$QEMU_IO -r -c "read -P 0x5a 0 512" $TEST_IMG | _filter_qemu_io
+
+# The dirty bit must be set
+./qcow2.py $TEST_IMG dump-header | grep incompatible_features
+
+echo
+echo "== Repairing the image file must succeed =="
+
+$QEMU_IMG check -r all $TEST_IMG
+
+# The dirty bit must not be set
+./qcow2.py $TEST_IMG dump-header | grep incompatible_features
+
+echo
+echo "== Data should still be accessible after repair =="
+
+$QEMU_IO -c "read -P 0x5a 0 512" $TEST_IMG | _filter_qemu_io
+
+echo
+echo "== Opening a dirty image read/write should repair it =="
+
+IMGOPTS="compat=1.1,lazy_refcounts=on"
+_make_test_img $size
+