Re: [PATCH v5 2/8] block-backend: Introduce blk_get_guest_block_size()

From: Yongji Xie <xieyongji@bytedance.com>
To: Stefan Hajnoczi <stefanha@redhat.com>
Cc: "Michael S. Tsirkin" <mst@redhat.com>,
	Jason Wang <jasowang@redhat.com>,
	 Stefano Garzarella <sgarzare@redhat.com>,
	Kevin Wolf <kwolf@redhat.com>,
	mreitz@redhat.com,  mlureau@redhat.com, jsnow@redhat.com,
	Eric Blake <eblake@redhat.com>,
	 Coiby.Xu@gmail.com, hreitz@redhat.com, qemu-block@nongnu.org,
	 qemu-devel@nongnu.org
Subject: Re: [PATCH v5 2/8] block-backend: Introduce blk_get_guest_block_size()
Date: Thu, 19 May 2022 16:03:22 +0800	[thread overview]
Message-ID: <CACycT3skLBKJnWhCQVKyHrqX8-RkTSG_jsFP9Mrgpc4ZTRde1A@mail.gmail.com> (raw)
In-Reply-To: <YoTxwwS97POt79PQ@stefanha-x1.localdomain>

On Wed, May 18, 2022 at 9:17 PM Stefan Hajnoczi <stefanha@redhat.com> wrote:
>
> On Wed, May 04, 2022 at 03:40:45PM +0800, Xie Yongji wrote:
> > Support getting the guest block size for the block backend.
> > It's needed for the following commit.
> >
> > Signed-off-by: Xie Yongji <xieyongji@bytedance.com>
> > ---
> >  block/block-backend.c             | 6 ++++++
> >  include/sysemu/block-backend-io.h | 1 +
> >  2 files changed, 7 insertions(+)
> >
> > diff --git a/block/block-backend.c b/block/block-backend.c
> > index 35457a6a1d..1582ff81c9 100644
> > --- a/block/block-backend.c
> > +++ b/block/block-backend.c
> > @@ -2106,6 +2106,12 @@ void blk_set_guest_block_size(BlockBackend *blk, int align)
> >      blk->guest_block_size = align;
> >  }
> >
> > +int blk_get_guest_block_size(BlockBackend *blk)
> > +{
> > +    IO_CODE();
> > +    return blk->guest_block_size;
>
> I have sent a patch to remove blk->guest_block_size because this field
> is currently unused.
>

OK.

> I think there is a cleaner way for this patch series to store the guest
> logical_block_size (see next patch). Stashing it in BlockBackend was
> attractive because virtio-blk-handler.c lacks a struct to store its
> parameters (writable, serial, logical_block_size), but if such a struct
> is introduced then there's no need to stash it in BlockBackend.
>

Looks good to me.

Thanks,
Yongji