qemu-block
[Top][All Lists]
Advanced

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

Re: [Qemu-block] [PATCH 1/2] block: pass bdrv_* methods to bs->file by d


From: Stefan Hajnoczi
Subject: Re: [Qemu-block] [PATCH 1/2] block: pass bdrv_* methods to bs->file by default
Date: Thu, 29 Jun 2017 15:00:35 +0100
User-agent: Mutt/1.8.0 (2017-02-23)

On Wed, Jun 28, 2017 at 01:46:15AM +0300, Manos Pitsidianakis wrote:
> @@ -3406,13 +3410,18 @@ int bdrv_truncate(BdrvChild *child, int64_t offset, 
> Error **errp)
>  
>      assert(child->perm & BLK_PERM_RESIZE);
>  
> -    if (!drv) {
> -        error_setg(errp, "No medium inserted");
> -        return -ENOMEDIUM;
> -    }
> -    if (!drv->bdrv_truncate) {
> -        error_setg(errp, "Image format driver does not support resize");
> -        return -ENOTSUP;
> +    if (!drv || !drv->bdrv_truncate) {

bs->drv == NULL means the BDS is closed.  We should not forward to
bs->file.  In practice bs->file will be NULL but please leave the
original if statement unchanged so it's clear that !drv results in
-ENOMEDIUM.

The same applies to the other changes below.

> @@ -3832,10 +3844,15 @@ void bdrv_get_backing_filename(BlockDriverState *bs,
>  int bdrv_get_info(BlockDriverState *bs, BlockDriverInfo *bdi)
>  {
>      BlockDriver *drv = bs->drv;
> -    if (!drv)
> -        return -ENOMEDIUM;
> -    if (!drv->bdrv_get_info)
> -        return -ENOTSUP;
> +    if (!drv || !drv->bdrv_get_info) {
> +        if (bs->file && bs->file->bs) {
> +            return bdrv_get_info(bs->file->bs, bdi);
> +        }
> +        if (!drv)
> +            return -ENOMEDIUM;

QEMU coding style always uses curly braces, even when the if statement
body is just 1 line.

> diff --git a/block/io.c b/block/io.c
> index c72d7015..c1b73226 100644
> --- a/block/io.c
> +++ b/block/io.c
> @@ -2403,6 +2403,10 @@ int bdrv_co_ioctl(BlockDriverState *bs, int req, void 
> *buf)
>  
>      bdrv_inc_in_flight(bs);
>      if (!drv || (!drv->bdrv_aio_ioctl && !drv->bdrv_co_ioctl)) {
> +        if (bs->file && bs->file->bs) {
> +            bdrv_dec_in_flight(bs);
> +            return bdrv_co_ioctl(bs->file->bs, req, buf);

bdrv_dec_in_flight() must happen after the request finishes, otherwise
bdrv_drain() will break.

Attachment: signature.asc
Description: PGP signature


reply via email to

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