Re: [PATCH 0/2] vvfat: fix two crashes.

[Vladimir Sementsov-Ogievskiy]

24.05.2021 19:06, Programmingkid wrote:
> > On May 24, 2021, at 11:55 AM, Vladimir Sementsov-Ogievskiy 
> > <vsementsov@virtuozzo.com> wrote:
> >
> > 24.05.2021 18:41, Programmingkid wrote:
> > > > On May 24, 2021, at 6:12 AM, Vladimir Sementsov-Ogievskiy 
> > > > <vsementsov@virtuozzo.com> wrote:
> > > >
> > > > Hi!
> > > >
> > > > As reported by Programmingkid, command
> > > >
> > > > qemu-system-ppc -usb -device usb-storage,drive=fat16 -drive 
> > > > file=fat:rw:fat-type=16:"<path of a host folder>",id=fat16,format=raw,if=none
> > > >
> > > > crashes.
> > > >
> > > > I tested it with qemu-system-x86_64 and it reproduces for me. I even
> > > > kept "<path of a host folder>" as is :).
> > > >
> > > > So, here are two fixes.
> > > >
> > > > Vladimir Sementsov-Ogievskiy (2):
> > > >   block/vvfat: child_vvfat_qcow: add .get_parent_aio_context, fix crash
> > > >   block/vvfat: fix vvfat_child_perm crash
> > > >
> > > > include/block/block.h | 1 +
> > > > block.c               | 4 ++--
> > > > block/vvfat.c         | 8 +++-----
> > > > 3 files changed, 6 insertions(+), 7 deletions(-)
> > > >
> > > > --
> > > > 2.29.2
> > > I applied both of your patches to the most recent git commit today and they 
> > > worked. I was able to share files from the host on the guest.
> > > Thank you.
> > > Reviewed-by: John Arbuckle <programmingkidx@gmail.com>
> >
> > Thanks for testing! Didn't you mean "Tested-by: "? Or if you both reviewed and 
> > tested, having both marks makes sense.
>
> Yes, you are right. It should be: Tested-by: John Arbuckle 
> <programmingkidx@gmail.com>
>
> > Hmm, also, I think "Reported-by" in first patch should be fixed to your real 
> > name too for consistency.
>
> That should be fine but it isn't important.
>
> On a related topic would you know if it is possible to use fat32 instead of 
> fat16 for host folder sharing? I did try replacing the text fat16 with fat32 
> but it didn't appear to work.

No, I don't know..

Moreover, my quick look at the code of vvfat, this fixed bug (which is obviously not covered by 
tests), and also status of block/vvfat in MAINTAINERS file "Odd Fixes", all this leads to 
advice "don't use it if possible". May be Kevin can add something about it, he is 
maintainer..

Could you use for example NFS or Samba, or sshfs to share folders? Or you need 
exactly to make a host folder available in guest vm as usb drive?

--
Best regards,
Vladimir