Re: [PATCH v1 3/7] xen: xenstore: add possibility to preserve owner

From: Volodymyr Babchuk <Volodymyr_Babchuk@epam.com>
To: David Woodhouse <dwmw2@infradead.org>
Cc: "qemu-devel@nongnu.org" <qemu-devel@nongnu.org>,
	Paul Durrant <paul@xen.org>,
	"Michael S. Tsirkin" <mst@redhat.com>,
	Marcel Apfelbaum <marcel.apfelbaum@gmail.com>,
	Paolo Bonzini <pbonzini@redhat.com>,
	Richard Henderson <richard.henderson@linaro.org>,
	Eduardo Habkost <eduardo@habkost.net>,
	Stefano Stabellini <sstabellini@kernel.org>,
	Anthony Perard <anthony.perard@citrix.com>,
	"open  list:X86 Xen CPUs" <xen-devel@lists.xenproject.org>
Subject: Re: [PATCH v1 3/7] xen: xenstore: add possibility to preserve owner
Date: Mon, 13 Nov 2023 13:02:22 +0000	[thread overview]
Message-ID: <8734x97qn6.fsf@epam.com> (raw)
In-Reply-To: <9cdd033411f5c15920762808891f278001073c6b.camel@infradead.org>

Hi David,

David Woodhouse <dwmw2@infradead.org> writes:

> [[S/MIME Signed Part:Undecided]]
> On Sat, 2023-11-11 at 11:01 +0000, David Woodhouse wrote:
>> 
>> > --- a/hw/xen/xen-operations.c
>> > +++ b/hw/xen/xen-operations.c
>> > @@ -300,6 +300,18 @@ static bool libxenstore_create(struct qemu_xs_handle *h, xs_transaction_t t,
>> >          return false;
>> >      }
>> >  
>> > +    if (owner == XS_PRESERVE_OWNER) {
>> > +        struct xs_permissions *tmp;
>> > +        unsigned int num;
>> > +
>> > +        tmp = xs_get_permissions(h->xsh, 0, path, &num);
>> > +        if (tmp == NULL) {
>> > +            return false;
>> > +        }
>> > +        perms_list[0].id = tmp[0].id;
>> > +        free(tmp);
>> > +    }
>> > +
>> 
>> Don't see what saves you from someone else changing it at this point on
>> true Xen though. Which is why I'd prefer XenStore to do it natively.
>
> I suppose maybe you could do it in a transaction *if* the transaction_t
> you're passed in isn't already XBT_NULL?

Yes, I need to pass "t" to xs_get_permissions(), of course.

> One might argue that the mkdir+set_perms in libxenstore_create() ought
> to have been within the same transaction *anyway*? 

Yes, all operations should be performed inside one transaction.

-- 
WBR, Volodymyr