On 08/12/2017 12:51, Peter Lieven wrote:
> we forgot to set the allocmap to invalid if an UNMAP call fails.
This is only needed for "a power loss, a medium error, or hardware
error" according to the spec, but I guess this is the safe thing to do.
Paolo
> Cc: qemu-sta...@nongnu.org
> Signed-off-by: Peter Lieven <p...@kamp.de>
> ---
> block/iscsi.c | 8 ++++----
> 1 file changed, 4 insertions(+), 4 deletions(-)
>
> diff --git a/block/iscsi.c b/block/iscsi.c
> index 4683f3b..c532ec7 100644
> --- a/block/iscsi.c
> +++ b/block/iscsi.c
> @@ -2,7 +2,7 @@
> * QEMU Block driver for iSCSI images
> *
> * Copyright (c) 2010-2011 Ronnie Sahlberg <ronniesahlb...@gmail.com>
> - * Copyright (c) 2012-2016 Peter Lieven <p...@kamp.de>
> + * Copyright (c) 2012-2017 Peter Lieven <p...@kamp.de>
> *
> * Permission is hereby granted, free of charge, to any person obtaining a
> copy
> * of this software and associated documentation files (the "Software"), to
> deal
> @@ -1128,6 +1128,9 @@ retry:
> goto retry;
> }
>
> + iscsi_allocmap_set_invalid(iscsilun, offset >> BDRV_SECTOR_BITS,
> + bytes >> BDRV_SECTOR_BITS);
> +
> if (iTask.status == SCSI_STATUS_CHECK_CONDITION) {
> /* the target might fail with a check condition if it
> is not happy with the alignment of the UNMAP request
> @@ -1140,9 +1143,6 @@ retry:
> goto out_unlock;
> }
>
> - iscsi_allocmap_set_invalid(iscsilun, offset >> BDRV_SECTOR_BITS,
> - bytes >> BDRV_SECTOR_BITS);
> -
> out_unlock:
> qemu_mutex_unlock(&iscsilun->mutex);
> return r;
>
