block: add image streaming block job

Signed-off-by: Stefan Hajnoczi <stefanha@linux.vnet.ibm.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>

4 files changed, 141 insertions, 0 deletions

diff --git a/Makefile.objs b/Makefile.objs
index 48d14777a0..06a147b0b0 100644
--- a/Makefile.objs
+++ b/Makefile.objs
@@ -35,6 +35,7 @@ block-nested-y += qcow2.o qcow2-refcount.o qcow2-cluster.o qcow2-snapshot.o qcow
 block-nested-y += qed.o qed-gencb.o qed-l2-cache.o qed-table.o qed-cluster.o
 block-nested-y += qed-check.o
 block-nested-y += parallels.o nbd.o blkdebug.o sheepdog.o blkverify.o
+block-nested-y += stream.o
 block-nested-$(CONFIG_WIN32) += raw-win32.o
 block-nested-$(CONFIG_POSIX) += raw-posix.o
 block-nested-$(CONFIG_LIBISCSI) += iscsi.o
diff --git a/block/stream.c b/block/stream.c
new file mode 100644
index 0000000000..3a9909480e
--- /dev/null
+++ b/block/stream.c
@@ -0,0 +1,133 @@
+/*
+ * Image streaming
+ *
+ * Copyright IBM, Corp. 2011
+ *
+ * Authors:
+ *  Stefan Hajnoczi   <stefanha@linux.vnet.ibm.com>
+ *
+ * This work is licensed under the terms of the GNU LGPL, version 2 or later.
+ * See the COPYING.LIB file in the top-level directory.
+ *
+ */
+
+#include "trace.h"
+#include "block_int.h"
+
+enum {
+    /*
+     * Size of data buffer for populating the image file.  This should be large
+     * enough to process multiple clusters in a single call, so that populating
+     * contiguous regions of the image is efficient.
+     */
+    STREAM_BUFFER_SIZE = 512 * 1024, /* in bytes */
+};
+
+typedef struct StreamBlockJob {
+    BlockJob common;
+    BlockDriverState *base;
+} StreamBlockJob;
+
+static int coroutine_fn stream_populate(BlockDriverState *bs,
+                                        int64_t sector_num, int nb_sectors,
+                                        void *buf)
+{
+    struct iovec iov = {
+        .iov_base = buf,
+        .iov_len  = nb_sectors * BDRV_SECTOR_SIZE,
+    };
+    QEMUIOVector qiov;
+
+    qemu_iovec_init_external(&qiov, &iov, 1);
+
+    /* Copy-on-read the unallocated clusters */
+    return bdrv_co_copy_on_readv(bs, sector_num, nb_sectors, &qiov);
+}
+
+static void coroutine_fn stream_run(void *opaque)
+{
+    StreamBlockJob *s = opaque;
+    BlockDriverState *bs = s->common.bs;
+    int64_t sector_num, end;
+    int ret = 0;
+    int n;
+    void *buf;
+
+    s->common.len = bdrv_getlength(bs);
+    if (s->common.len < 0) {
+        block_job_complete(&s->common, s->common.len);
+        return;
+    }
+
+    end = s->common.len >> BDRV_SECTOR_BITS;
+    buf = qemu_blockalign(bs, STREAM_BUFFER_SIZE);
+
+    /* Turn on copy-on-read for the whole block device so that guest read
+     * requests help us make progress.  Only do this when copying the entire
+     * backing chain since the copy-on-read operation does not take base into
+     * account.
+     */
+    if (!base) {
+        bdrv_enable_copy_on_read(bs);
+    }
+
+    for (sector_num = 0; sector_num < end; sector_num += n) {
+        if (block_job_is_cancelled(&s->common)) {
+            break;
+        }
+
+        /* TODO rate-limit */
+        /* Note that even when no rate limit is applied we need to yield with
+         * no pending I/O here so that qemu_aio_flush() is able to return.
+         */
+        co_sleep_ns(rt_clock, 0);
+
+        ret = bdrv_co_is_allocated(bs, sector_num,
+                                   STREAM_BUFFER_SIZE / BDRV_SECTOR_SIZE, &n);
+        trace_stream_one_iteration(s, sector_num, n, ret);
+        if (ret == 0) {
+            ret = stream_populate(bs, sector_num, n, buf);
+        }
+        if (ret < 0) {
+            break;
+        }
+
+        /* Publish progress */
+        s->common.offset += n * BDRV_SECTOR_SIZE;
+    }
+
+    if (!base) {
+        bdrv_disable_copy_on_read(bs);
+    }
+
+    if (sector_num == end && ret == 0) {
+        ret = bdrv_change_backing_file(bs, NULL, NULL);
+    }
+
+    qemu_vfree(buf);
+    block_job_complete(&s->common, ret);
+}
+
+static BlockJobType stream_job_type = {
+    .instance_size = sizeof(StreamBlockJob),
+    .job_type      = "stream",
+};
+
+int stream_start(BlockDriverState *bs, BlockDriverState *base,
+                 BlockDriverCompletionFunc *cb, void *opaque)
+{
+    StreamBlockJob *s;
+    Coroutine *co;
+
+    s = block_job_create(&stream_job_type, bs, cb, opaque);
+    if (!s) {
+        return -EBUSY; /* bs must already be in use */
+    }
+
+    s->base = base;
+
+    co = qemu_coroutine_create(stream_run);
+    trace_stream_start(bs, base, s, co, opaque);
+    qemu_coroutine_enter(co, s);
+    return 0;
+}
diff --git a/block_int.h b/block_int.h
index 3566acb894..4f638705e0 100644
--- a/block_int.h
+++ b/block_int.h
@@ -327,4 +327,7 @@ int block_job_set_speed(BlockJob *job, int64_t value);
 void block_job_cancel(BlockJob *job);
 bool block_job_is_cancelled(BlockJob *job);
 
+int stream_start(BlockDriverState *bs, BlockDriverState *base,
+                 BlockDriverCompletionFunc *cb, void *opaque);
+
 #endif /* BLOCK_INT_H */
diff --git a/trace-events b/trace-events
index 035d08fe7c..f6a3cd1aaf 100644
--- a/trace-events
+++ b/trace-events
@@ -70,6 +70,10 @@ bdrv_co_writev(void *bs, int64_t sector_num, int nb_sector) "bs %p sector_num %"
 bdrv_co_io_em(void *bs, int64_t sector_num, int nb_sectors, int is_write, void *acb) "bs %p sector_num %"PRId64" nb_sectors %d is_write %d acb %p"
 bdrv_co_do_copy_on_readv(void *bs, int64_t sector_num, int nb_sectors, int64_t cluster_sector_num, int cluster_nb_sectors) "bs %p sector_num %"PRId64" nb_sectors %d cluster_sector_num %"PRId64" cluster_nb_sectors %d"
 
+# block/stream.c
+stream_one_iteration(void *s, int64_t sector_num, int nb_sectors, int is_allocated) "s %p sector_num %"PRId64" nb_sectors %d is_allocated %d"
+stream_start(void *bs, void *base, void *s, void *co, void *opaque) "bs %p base %p s %p co %p opaque %p"
+
 # hw/virtio-blk.c
 virtio_blk_req_complete(void *req, int status) "req %p status %d"
 virtio_blk_rw_complete(void *req, int ret) "req %p ret %d"