qemu-devel
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [PATCH 1/4] block: Add bdrv_co_get_lba_status


From: Claudio Fontana
Subject: Re: [PATCH 1/4] block: Add bdrv_co_get_lba_status
Date: Wed, 3 Jun 2020 16:46:33 +0200
User-agent: Mozilla/5.0 (X11; Linux x86_64; rv:68.0) Gecko/20100101 Thunderbird/68.4.1

Hi Lin,

just a couple of things, leaving the core aspects to more knowledgeable people 
in this area to address.

In general, you want all patches in the series to build correctly.

Applying only PATCH 1/4 breaks the build for me with:

/home/claudio/git/qemu/block/io.c:2190:1: error: no previous prototype for 
‘bdrv_co_get_lba_status’ [-Werror=missing-prototypes]
 bdrv_co_get_lba_status(BdrvChild *child, int64_t offset, int64_t bytes,
 ^~~~~~~~~~~~~~~~~~~~~~
/home/claudio/git/qemu/block/io.c: In function ‘bdrv_co_get_lba_status’:
/home/claudio/git/qemu/block/io.c:2194:9: error: ‘ret’ may be used 
uninitialized in this function [-Werror=maybe-uninitialized]
     int ret;
         ^~~
cc1: all warnings being treated as errors

On 6/2/20 9:41 AM, Lin Ma wrote:
> The get lba status wrapper based on the bdrv_block_status. The following
> patches will add GET LBA STATUS 16 support for scsi emulation layer.
> 
> Signed-off-by: Lin Ma <lma@suse.com>
> ---
>  block/io.c | 43 +++++++++++++++++++++++++++++++++++++++++++
>  1 file changed, 43 insertions(+)
> 
> diff --git a/block/io.c b/block/io.c
> index 121ce17a49..dacc3c2471 100644
> --- a/block/io.c
> +++ b/block/io.c
> @@ -2186,6 +2186,49 @@ int coroutine_fn bdrv_co_pwrite_zeroes(BdrvChild 
> *child, int64_t offset,
>                             BDRV_REQ_ZERO_WRITE | flags);
>  }
>  
> +int coroutine_fn
> +bdrv_co_get_lba_status(BdrvChild *child, int64_t offset, int64_t bytes,
> +                       uint32_t *num_blocks, uint32_t *is_deallocated)
> +{
> +    BlockDriverState *bs = child->bs;
> +    int ret;
> +    int64_t target_size, count = 0;
> +    bool first = true;
> +    uint8_t wanted_bit1 = 0;
> +
> +    target_size = bdrv_getlength(bs);
> +    if (target_size < 0) {
> +        return -EIO;
> +    }
> +
> +    if (offset < 0 || bytes < 0) {
> +        return -EIO;
> +    }
> +
> +    for ( ; offset <= target_size - bytes; offset += count) {
> +        ret = bdrv_block_status(bs, offset, bytes, &count, NULL, NULL);
> +        if (ret < 0) {
> +            goto out;
> +        }
> +        if (first) {
> +            if (ret & BDRV_BLOCK_ZERO) {
> +                wanted_bit1 = BDRV_BLOCK_ZERO >> 1;;

extra ; at the end of the line

> +                *is_deallocated = 1;
> +            } else {
> +                wanted_bit1 = 0;
> +            }
> +            first = false;
> +        }
> +        if ((ret & BDRV_BLOCK_ZERO) >> 1 == wanted_bit1) {
> +            (*num_blocks)++;
> +        } else {
> +            break;
> +        }
> +    }
> +out:
> +    return ret;
> +}
> +
>  /*
>   * Flush ALL BDSes regardless of if they are reachable via a BlkBackend or 
> not.
>   */
> 

Ciao,

CLaudio



reply via email to

[Prev in Thread] Current Thread [Next in Thread]