[Qemu-devel] [PULL 25/29] raw_bsd: add offset and size options

[Kevin Wolf <kwolf@redhat.com>]

From: Tomáš Golembiovský <tgolembi@redhat.com>

Added two new options 'offset' and 'size'. This makes it possible to use
only part of the file as a device. This can be used e.g. to limit the
access only to single partition in a disk image or use a disk inside a
tar archive (like OVA).

When 'size' is specified we do our best to honour it.

Signed-off-by: Tomáš Golembiovský <tgolembi@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
---
 block/raw_bsd.c      | 217 ++++++++++++++++++++++++++++++++++++++++++++++++++-
 qapi/block-core.json |  16 +++-
 2 files changed, 229 insertions(+), 4 deletions(-)

diff --git a/block/raw_bsd.c b/block/raw_bsd.c
index fc16ec1..7c9bebb 100644
--- a/block/raw_bsd.c
+++ b/block/raw_bsd.c
@@ -31,6 +31,30 @@
 #include "qapi/error.h"
 #include "qemu/option.h"
 
+typedef struct BDRVRawState {
+    uint64_t offset;
+    uint64_t size;
+    bool has_size;
+} BDRVRawState;
+
+static QemuOptsList raw_runtime_opts = {
+    .name = "raw",
+    .head = QTAILQ_HEAD_INITIALIZER(raw_runtime_opts.head),
+    .desc = {
+        {
+            .name = "offset",
+            .type = QEMU_OPT_SIZE,
+            .help = "offset in the disk where the image starts",
+        },
+        {
+            .name = "size",
+            .type = QEMU_OPT_SIZE,
+            .help = "virtual disk size",
+        },
+        { /* end of list */ }
+    },
+};
+
 static QemuOptsList raw_create_opts = {
     .name = "raw-create-opts",
     .head = QTAILQ_HEAD_INITIALIZER(raw_create_opts.head),
@@ -44,16 +68,108 @@ static QemuOptsList raw_create_opts = {
     }
 };
 
+static int raw_read_options(QDict *options, BlockDriverState *bs,
+    BDRVRawState *s, Error **errp)
+{
+    Error *local_err = NULL;
+    QemuOpts *opts = NULL;
+    int64_t real_size = 0;
+    int ret;
+
+    real_size = bdrv_getlength(bs->file->bs);
+    if (real_size < 0) {
+        error_setg_errno(errp, -real_size, "Could not get image size");
+        return real_size;
+    }
+
+    opts = qemu_opts_create(&raw_runtime_opts, NULL, 0, &error_abort);
+    qemu_opts_absorb_qdict(opts, options, &local_err);
+    if (local_err) {
+        error_propagate(errp, local_err);
+        ret = -EINVAL;
+        goto end;
+    }
+
+    s->offset = qemu_opt_get_size(opts, "offset", 0);
+    if (qemu_opt_find(opts, "size") != NULL) {
+        s->size = qemu_opt_get_size(opts, "size", 0);
+        s->has_size = true;
+    } else {
+        s->has_size = false;
+        s->size = real_size - s->offset;
+    }
+
+    /* Check size and offset */
+    if (real_size < s->offset || (real_size - s->offset) < s->size) {
+        error_setg(errp, "The sum of offset (%" PRIu64 ") and size "
+            "(%" PRIu64 ") has to be smaller or equal to the "
+            " actual size of the containing file (%" PRId64 ")",
+            s->offset, s->size, real_size);
+        ret = -EINVAL;
+        goto end;
+    }
+
+    /* Make sure size is multiple of BDRV_SECTOR_SIZE to prevent rounding
+     * up and leaking out of the specified area. */
+    if (!QEMU_IS_ALIGNED(s->size, BDRV_SECTOR_SIZE)) {
+        error_setg(errp, "Specified size is not multiple of %llu",
+            BDRV_SECTOR_SIZE);
+        ret = -EINVAL;
+        goto end;
+    }
+
+    ret = 0;
+
+end:
+
+    qemu_opts_del(opts);
+
+    return ret;
+}
+
 static int raw_reopen_prepare(BDRVReopenState *reopen_state,
                               BlockReopenQueue *queue, Error **errp)
 {
-    return 0;
+    assert(reopen_state != NULL);
+    assert(reopen_state->bs != NULL);
+
+    reopen_state->opaque = g_new0(BDRVRawState, 1);
+
+    return raw_read_options(
+        reopen_state->options,
+        reopen_state->bs,
+        reopen_state->opaque,
+        errp);
+}
+
+static void raw_reopen_commit(BDRVReopenState *state)
+{
+    BDRVRawState *new_s = state->opaque;
+    BDRVRawState *s = state->bs->opaque;
+
+    memcpy(s, new_s, sizeof(BDRVRawState));
+
+    g_free(state->opaque);
+    state->opaque = NULL;
+}
+
+static void raw_reopen_abort(BDRVReopenState *state)
+{
+    g_free(state->opaque);
+    state->opaque = NULL;
 }
 
 static int coroutine_fn raw_co_preadv(BlockDriverState *bs, uint64_t offset,
                                       uint64_t bytes, QEMUIOVector *qiov,
                                       int flags)
 {
+    BDRVRawState *s = bs->opaque;
+
+    if (offset > UINT64_MAX - s->offset) {
+        return -EINVAL;
+    }
+    offset += s->offset;
+
     BLKDBG_EVENT(bs->file, BLKDBG_READ_AIO);
     return bdrv_co_preadv(bs->file, offset, bytes, qiov, flags);
 }
@@ -62,11 +178,23 @@ static int coroutine_fn raw_co_pwritev(BlockDriverState *bs, uint64_t offset,
                                        uint64_t bytes, QEMUIOVector *qiov,
                                        int flags)
 {
+    BDRVRawState *s = bs->opaque;
     void *buf = NULL;
     BlockDriver *drv;
     QEMUIOVector local_qiov;
     int ret;
 
+    if (s->has_size && (offset > s->size || bytes > (s->size - offset))) {
+        /* There's not enough space for the data. Don't write anything and just
+         * fail to prevent leaking out of the size specified in options. */
+        return -ENOSPC;
+    }
+
+    if (offset > UINT64_MAX - s->offset) {
+        ret = -EINVAL;
+        goto fail;
+    }
+
     if (bs->probed && offset < BLOCK_PROBE_BUF_SIZE && bytes) {
         /* Handling partial writes would be a pain - so we just
          * require that guests have 512-byte request alignment if
@@ -101,6 +229,8 @@ static int coroutine_fn raw_co_pwritev(BlockDriverState *bs, uint64_t offset,
         qiov = &local_qiov;
     }
 
+    offset += s->offset;
+
     BLKDBG_EVENT(bs->file, BLKDBG_WRITE_AIO);
     ret = bdrv_co_pwritev(bs->file, offset, bytes, qiov, flags);
 
@@ -117,8 +247,10 @@ static int64_t coroutine_fn raw_co_get_block_status(BlockDriverState *bs,
                                             int nb_sectors, int *pnum,
                                             BlockDriverState **file)
 {
+    BDRVRawState *s = bs->opaque;
     *pnum = nb_sectors;
     *file = bs->file->bs;
+    sector_num += s->offset / BDRV_SECTOR_SIZE;
     return BDRV_BLOCK_RAW | BDRV_BLOCK_OFFSET_VALID | BDRV_BLOCK_DATA |
            (sector_num << BDRV_SECTOR_BITS);
 }
@@ -127,18 +259,49 @@ static int coroutine_fn raw_co_pwrite_zeroes(BlockDriverState *bs,
                                              int64_t offset, int count,
                                              BdrvRequestFlags flags)
 {
+    BDRVRawState *s = bs->opaque;
+    if (offset > UINT64_MAX - s->offset) {
+        return -EINVAL;
+    }
+    offset += s->offset;
     return bdrv_co_pwrite_zeroes(bs->file, offset, count, flags);
 }
 
 static int coroutine_fn raw_co_pdiscard(BlockDriverState *bs,
                                         int64_t offset, int count)
 {
+    BDRVRawState *s = bs->opaque;
+    if (offset > UINT64_MAX - s->offset) {
+        return -EINVAL;
+    }
+    offset += s->offset;
     return bdrv_co_pdiscard(bs->file->bs, offset, count);
 }
 
 static int64_t raw_getlength(BlockDriverState *bs)
 {
-    return bdrv_getlength(bs->file->bs);
+    int64_t len;
+    BDRVRawState *s = bs->opaque;
+
+    /* Update size. It should not change unless the file was externally
+     * modified. */
+    len = bdrv_getlength(bs->file->bs);
+    if (len < 0) {
+        return len;
+    }
+
+    if (len < s->offset) {
+        s->size = 0;
+    } else {
+        if (s->has_size) {
+            /* Try to honour the size */
+            s->size = MIN(s->size, len - s->offset);
+        } else {
+            s->size = len - s->offset;
+        }
+    }
+
+    return s->size;
 }
 
 static int raw_get_info(BlockDriverState *bs, BlockDriverInfo *bdi)
@@ -158,6 +321,18 @@ static void raw_refresh_limits(BlockDriverState *bs, Error **errp)
 
 static int raw_truncate(BlockDriverState *bs, int64_t offset)
 {
+    BDRVRawState *s = bs->opaque;
+
+    if (s->has_size) {
+        return -ENOTSUP;
+    }
+
+    if (INT64_MAX - offset < s->offset) {
+        return -EINVAL;
+    }
+
+    s->size = offset;
+    offset += s->offset;
     return bdrv_truncate(bs->file->bs, offset);
 }
 
@@ -178,6 +353,10 @@ static void raw_lock_medium(BlockDriverState *bs, bool locked)
 
 static int raw_co_ioctl(BlockDriverState *bs, unsigned long int req, void *buf)
 {
+    BDRVRawState *s = bs->opaque;
+    if (s->offset || s->has_size) {
+        return -ENOTSUP;
+    }
     return bdrv_co_ioctl(bs->file->bs, req, buf);
 }
 
@@ -194,6 +373,9 @@ static int raw_create(const char *filename, QemuOpts *opts, Error **errp)
 static int raw_open(BlockDriverState *bs, QDict *options, int flags,
                     Error **errp)
 {
+    BDRVRawState *s = bs->opaque;
+    int ret;
+
     bs->sg = bs->file->bs->sg;
     bs->supported_write_flags = BDRV_REQ_FUA &
         bs->file->bs->supported_write_flags;
@@ -211,6 +393,16 @@ static int raw_open(BlockDriverState *bs, QDict *options, int flags,
                 bs->file->bs->filename);
     }
 
+    ret = raw_read_options(options, bs, s, errp);
+    if (ret < 0) {
+        return ret;
+    }
+
+    if (bs->sg && (s->offset || s->has_size)) {
+        error_setg(errp, "Cannot use offset/size with SCSI generic devices");
+        return -EINVAL;
+    }
+
     return 0;
 }
 
@@ -228,18 +420,37 @@ static int raw_probe(const uint8_t *buf, int buf_size, const char *filename)
 
 static int raw_probe_blocksizes(BlockDriverState *bs, BlockSizes *bsz)
 {
-    return bdrv_probe_blocksizes(bs->file->bs, bsz);
+    BDRVRawState *s = bs->opaque;
+    int ret;
+
+    ret = bdrv_probe_blocksizes(bs->file->bs, bsz);
+    if (ret < 0) {
+        return ret;
+    }
+
+    if (!QEMU_IS_ALIGNED(s->offset, MAX(bsz->log, bsz->phys))) {
+        return -ENOTSUP;
+    }
+
+    return 0;
 }
 
 static int raw_probe_geometry(BlockDriverState *bs, HDGeometry *geo)
 {
+    BDRVRawState *s = bs->opaque;
+    if (s->offset || s->has_size) {
+        return -ENOTSUP;
+    }
     return bdrv_probe_geometry(bs->file->bs, geo);
 }
 
 BlockDriver bdrv_raw = {
     .format_name          = "raw",
+    .instance_size        = sizeof(BDRVRawState),
     .bdrv_probe           = &raw_probe,
     .bdrv_reopen_prepare  = &raw_reopen_prepare,
+    .bdrv_reopen_commit   = &raw_reopen_commit,
+    .bdrv_reopen_abort    = &raw_reopen_abort,
     .bdrv_open            = &raw_open,
     .bdrv_close           = &raw_close,
     .bdrv_create          = &raw_create,
diff --git a/qapi/block-core.json b/qapi/block-core.json
index 3592a9d..df3b987 100644
--- a/qapi/block-core.json
+++ b/qapi/block-core.json
@@ -2270,6 +2270,20 @@
             '*tls-creds': 'str' } }
 
 ##
+# @BlockdevOptionsRaw
+#
+# Driver specific block device options for the raw driver.
+#
+# @offset:      #optional position where the block device starts
+# @size:        #optional the assumed size of the device
+#
+# Since: 2.8
+##
+{ 'struct': 'BlockdevOptionsRaw',
+  'base': 'BlockdevOptionsGenericFormat',
+  'data': { '*offset': 'int', '*size': 'int' } }
+
+##
 # @BlockdevOptions
 #
 # Options for creating a block device.  Many options are available for all
@@ -2323,7 +2337,7 @@
       'qcow':       'BlockdevOptionsGenericCOWFormat',
       'qed':        'BlockdevOptionsGenericCOWFormat',
       'quorum':     'BlockdevOptionsQuorum',
-      'raw':        'BlockdevOptionsGenericFormat',
+      'raw':        'BlockdevOptionsRaw',
 # TODO rbd: Wait for structured options
       'replication':'BlockdevOptionsReplication',
 # TODO sheepdog: Wait for structured options
-- 
1.8.3.1