The code that tries to reopen a BlockDriverState in stream_start()
when the creation of a new block job fails crashes because it attempts
to dereference a pointer that is known to be NULL.

This is a regression introduced in a170a91fd3eab6155da39e740381867e,
likely because the code was copied from stream_complete().

Signed-off-by: Alberto Garcia <berto@igalia.com>
---
 block/stream.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/block/stream.c b/block/stream.c
index 0113710845..52d329f5c6 100644
--- a/block/stream.c
+++ b/block/stream.c
@@ -280,6 +280,6 @@ void stream_start(const char *job_id, BlockDriverState *bs,
 
 fail:
     if (orig_bs_flags != bdrv_get_flags(bs)) {
-        bdrv_reopen(bs, s->bs_flags, NULL);
+        bdrv_reopen(bs, orig_bs_flags, NULL);
     }
 }
-- 
2.11.0

Am 15.05.2017 um 11:34 hat Alberto Garcia geschrieben:
> The code that tries to reopen a BlockDriverState in stream_start()
> when the creation of a new block job fails crashes because it attempts
> to dereference a pointer that is known to be NULL.
> 
> This is a regression introduced in a170a91fd3eab6155da39e740381867e,
> likely because the code was copied from stream_complete().
> 
> Signed-off-by: Alberto Garcia <berto@igalia.com>

Cc: qemu-stable@nongnu.org

Thanks, applied to the block branch.

Kevin

On Mon, May 15, 2017 at 12:34:24PM +0300, Alberto Garcia wrote:
> The code that tries to reopen a BlockDriverState in stream_start()
> when the creation of a new block job fails crashes because it attempts
> to dereference a pointer that is known to be NULL.
> 
> This is a regression introduced in a170a91fd3eab6155da39e740381867e,
> likely because the code was copied from stream_complete().
> 
> Signed-off-by: Alberto Garcia <berto@igalia.com>
> ---
>  block/stream.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/block/stream.c b/block/stream.c
> index 0113710845..52d329f5c6 100644
> --- a/block/stream.c
> +++ b/block/stream.c
> @@ -280,6 +280,6 @@ void stream_start(const char *job_id, BlockDriverState *bs,
>  
>  fail:
>      if (orig_bs_flags != bdrv_get_flags(bs)) {
> -        bdrv_reopen(bs, s->bs_flags, NULL);
> +        bdrv_reopen(bs, orig_bs_flags, NULL);

Tested-by: Kashyap Chamarthy <kchamart@redhat.com>

Yep, this fixes it.

It gracefully throws an error when a 'job-id' was not specified:

-----------------------------------------------------------------------
(QEMU) block-stream device=#block890 base=disk1.qcow2
{
    "execute": "block-stream", 
    "arguments": {
        "device": "#block890", 
        "base": "disk1.qcow2"
    }
}
{
    "error": {
        "class": "GenericError", 
        "desc": "An explicit job ID is required for this node"
    }
}
(QEMU) block-stream device=#block890 base=disk1.qcow2 job-id=job0
{
    "execute": "block-stream", 
    "arguments": {
        "device": "#block890", 
        "job-id": "job0", 
        "base": "disk1.qcow2"
    }
}
{
    "return": {}
}
(QEMU) 
-----------------------------------------------------------------------

[...]

-- 
/kashyap