From: Max Zhen <Max.Zhen@Sun.COM>
To: xen-devel@lists.xensource.com
Subject: problem of the permissions system in xenstore
Date: Wed, 01 Nov 2006 18:32:30 +0800 [thread overview]
Message-ID: <454877BE.2040204@Sun.COM> (raw)
Hi,
I encountered a problem while testing non-dom0 backend driver. I'm using
3.0.3-0.
I create backend and frontend vbd devices as below:
frontend path - /local/domain/4/device/vbd/1
backend path - /local/domain/3/backend/vbd/4/1
When I try to destroy dom 4 (frontend dom), the frontend path is removed
in a transaction.
In the end of that transaction, xenstored will try to fire watches for
what has been removed.
Since backend is watching /local/domain/4/device/vbd/1/state, the watch
should be fired for it, which is exactly the case for a dom0 backend device.
But, for a domU backend device (which is my case), the watch is not
fired. The reason is that the permission of frontend path for dom 3
(backend dom) is already gone during the transaction. So, xenstored will
check the parent directory (/local/domain/4/device/vbd) to determine if
dom 3 has the permission to know that the frontend path is gone.
Unfortunately, although backend has the read right for
/local/domain/4/device/vbd/1 and set watch on
/local/domain/4/device/vbd/1/state successfully, it has *no* read
permission for the parent directory (/local/domain/4/device/vbd), which
cause the watch-firing to fail.
(As for a dom0 backend device, it has all permissions automatically, so
the watch will be fired for a dom0 backend device driver correctly.)
Is it a bug that a domU has no read permission to a path while has read
permission to a path under it?
Thanks,
Max
next reply other threads:[~2006-11-01 10:32 UTC|newest]
Thread overview: 2+ messages / expand[flat|nested] mbox.gz Atom feed top
2006-11-01 10:32 Max Zhen [this message]
2006-11-01 11:03 ` problem of the permissions system in xenstore Keir Fraser
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=454877BE.2040204@Sun.COM \
--to=max.zhen@sun.com \
--cc=xen-devel@lists.xensource.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.