diff options
| author |   Konrad Rzeszutek Wilk <konrad.wilk@oracle.com> | 2011-09-16 15:15:14 -0400 |
|---|---|---|
| committer | Konrad Rzeszutek Wilk <konrad.wilk@oracle.com> | 2011-10-13 09:48:35 -0400 |
| commit | d11e6158307bed3f178399a4e6216eec67d16200 (patch) | |
| tree | c5fa6989bcba7d39b3924628c60e386e3f24edcc /drivers/block | |
| parent | xen-blkback: use kzalloc() in favor of kmalloc()+memset() (diff) | |
| download | linux-dev-d11e6158307bed3f178399a4e6216eec67d16200.tar.xz linux-dev-d11e6158307bed3f178399a4e6216eec67d16200.zip | |
xen-blkfront: If no barrier or flush is supported, use invalid operation.
Guard against issuing BLKIF_OP_WRITE_BARRIER or BLKIF_OP_FLUSH_CACHE
by checking whether we successfully negotiated with the backend.
The negotiation with the backend also sets the q->flush_flags which
fortunately for us is also used when submitting an bio to us. If
we don't support barriers or flushes it would be set to zero so
we should never end up having to deal with REQ_FLUSH | REQ_FUA.
However, other third party implementations of __make_request that
might be stacked on top of us might not be so smart, so lets fix this up.
Acked-by: Jan Beulich <jbeulich@suse.com>
Signed-off-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Diffstat (limited to 'drivers/block')
| -rw-r--r-- | drivers/block/xen-blkfront.c | 4 |
1 files changed, 3 insertions, 1 deletions
diff --git a/drivers/block/xen-blkfront.c b/drivers/block/xen-blkfront.c index 7d1487760283..8bf0cb7b2565 100644 --- a/drivers/block/xen-blkfront.c +++ b/drivers/block/xen-blkfront.c @@ -380,7 +380,9 @@ static void do_blkif_request(struct request_queue *rq) blk_start_request(req); - if (req->cmd_type != REQ_TYPE_FS) { + if ((req->cmd_type != REQ_TYPE_FS) || + ((req->cmd_flags & (REQ_FLUSH | REQ_FUA)) && + !info->flush_op)) { __blk_end_request_all(req, -EIO); continue; } |