From: "Roger Pau Monné" <roger.pau@citrix.com>
To: Demi Marie Obenour <demi@invisiblethingslab.com>
Cc: "Jens Axboe" <axboe@kernel.dk>,
"Alasdair Kergon" <agk@redhat.com>,
"Mike Snitzer" <snitzer@kernel.org>,
dm-devel@redhat.com,
"Marek Marczykowski-Górecki" <marmarek@invisiblethingslab.com>,
linux-block@vger.kernel.org, linux-kernel@vger.kernel.org,
xen-devel@lists.xenproject.org
Subject: Re: [PATCH v2 16/16] xen-blkback: Inform userspace that device has been opened
Date: Tue, 6 Jun 2023 11:15:37 +0200 [thread overview]
Message-ID: <ZH75OTMA6N3zYrH2@Air-de-Roger> (raw)
In-Reply-To: <20230530203116.2008-17-demi@invisiblethingslab.com>
On Tue, May 30, 2023 at 04:31:16PM -0400, Demi Marie Obenour wrote:
> Set "opened" to "0" before the hotplug script is called. Once the
> device node has been opened, set "opened" to "1".
>
> "opened" is used exclusively by userspace. It serves two purposes:
>
> 1. It tells userspace that the diskseq Xenstore entry is supported.
>
> 2. It tells userspace that it can wait for "opened" to be set to 1.
> Once "opened" is 1, blkback has a reference to the device, so
> userspace doesn't need to keep one.
>
> Together, these changes allow userspace to use block devices with
> delete-on-close behavior, such as loop devices with the autoclear flag
> set or device-mapper devices with the deferred-remove flag set.
There was some work in the past to allow reloading blkback as a
module, it's clear that using delete-on-close won't work if attempting
to reload blkback.
Isn't there some existing way to check whether a device is opened?
(stat syscall maybe?).
I would like to avoid adding more xenstore blkback state if such
information can be fetched from other methods.
> Signed-off-by: Demi Marie Obenour <demi@invisiblethingslab.com>
> ---
> drivers/block/xen-blkback/xenbus.c | 35 ++++++++++++++++++++++++++++++
> 1 file changed, 35 insertions(+)
>
> diff --git a/drivers/block/xen-blkback/xenbus.c b/drivers/block/xen-blkback/xenbus.c
> index 9c3eb148fbd802c74e626c3d7bcd69dcb09bd921..519a78aa9073d1faa1dce5c1b36e95ae58da534b 100644
> --- a/drivers/block/xen-blkback/xenbus.c
> +++ b/drivers/block/xen-blkback/xenbus.c
> @@ -3,6 +3,20 @@
> Copyright (C) 2005 Rusty Russell <rusty@rustcorp.com.au>
> Copyright (C) 2005 XenSource Ltd
>
> +In addition to the Xenstore nodes required by the Xen block device
> +specification, this implementation of blkback uses a new Xenstore
> +node: "opened". blkback sets "opened" to "0" before the hotplug script
> +is called. Once the device node has been opened, blkback sets "opened"
> +to "1".
> +
> +"opened" is read exclusively by userspace. It serves two purposes:
> +
> +1. It tells userspace that diskseq@major:minor syntax for "physical-device" is
> + supported.
> +
> +2. It tells userspace that it can wait for "opened" to be set to 1 after writing
> + "physical-device". Once "opened" is 1, blkback has a reference to the
> + device, so userspace doesn't need to keep one.
>
> */
>
> @@ -699,6 +713,14 @@ static int xen_blkbk_probe(struct xenbus_device *dev,
> if (err)
> pr_warn("%s write out 'max-ring-page-order' failed\n", __func__);
>
> + /*
> + * This informs userspace that the "opened" node will be set to "1" when
> + * the device has been opened successfully.
> + */
> + err = xenbus_write(XBT_NIL, dev->nodename, "opened", "0");
> + if (err)
> + goto fail;
> +
You would need to set "opened" before registering the xenstore backend
watch AFAICT, or else it could be racy.
Thanks, Roger.
next prev parent reply other threads:[~2023-06-06 9:15 UTC|newest]
Thread overview: 38+ messages / expand[flat|nested] mbox.gz Atom feed top
2023-05-30 20:31 [PATCH v2 00/16] Diskseq support in loop, device-mapper, and blkback Demi Marie Obenour
2023-05-30 20:31 ` [PATCH v2 01/16] device-mapper: Check that target specs are sufficiently aligned Demi Marie Obenour
2023-05-30 20:31 ` [PATCH v2 02/16] device-mapper: Avoid pointer arithmetic overflow Demi Marie Obenour
2023-05-30 20:31 ` [PATCH v2 03/16] device-mapper: do not allow targets to overlap 'struct dm_ioctl' Demi Marie Obenour
2023-05-30 20:31 ` [PATCH v2 04/16] device-mapper: Better error message for too-short target spec Demi Marie Obenour
2023-05-30 20:31 ` [PATCH v2 05/16] device-mapper: Target parameters must not overlap next " Demi Marie Obenour
2023-05-30 20:31 ` [PATCH v2 06/16] device-mapper: Avoid double-fetch of version Demi Marie Obenour
2023-05-30 20:31 ` [PATCH v2 07/16] device-mapper: Allow userspace to opt-in to strict parameter checks Demi Marie Obenour
2023-05-30 20:31 ` [PATCH v2 08/16] device-mapper: Allow userspace to provide expected diskseq Demi Marie Obenour
2023-05-30 20:31 ` [PATCH v2 09/16] device-mapper: Allow userspace to suppress uevent generation Demi Marie Obenour
2023-05-30 20:31 ` [PATCH v2 10/16] device-mapper: Refuse to create device named "control" Demi Marie Obenour
2023-05-30 20:31 ` [PATCH v2 11/16] device-mapper: "." and ".." are not valid symlink names Demi Marie Obenour
2023-05-30 20:31 ` [PATCH v2 12/16] device-mapper: inform caller about already-existing device Demi Marie Obenour
2023-05-30 20:31 ` [PATCH v2 13/16] xen-blkback: Implement diskseq checks Demi Marie Obenour
2023-06-06 8:25 ` Roger Pau Monné
2023-06-06 17:01 ` Demi Marie Obenour
2023-06-07 8:20 ` Roger Pau Monné
2023-06-07 16:14 ` Demi Marie Obenour
2023-06-08 8:29 ` Roger Pau Monné
2023-06-08 15:33 ` Demi Marie Obenour
2023-06-09 15:13 ` Roger Pau Monné
2023-06-09 16:55 ` Demi Marie Obenour
2023-06-12 8:09 ` Roger Pau Monné
2023-06-21 1:14 ` Demi Marie Obenour
2023-06-21 10:07 ` Roger Pau Monné
2023-05-30 20:31 ` [PATCH v2 14/16] block, loop: Increment diskseq when releasing a loop device Demi Marie Obenour
2023-05-30 20:31 ` [PATCH v2 15/16] xen-blkback: Minor cleanups Demi Marie Obenour
2023-06-06 8:36 ` Roger Pau Monné
2023-05-30 20:31 ` [PATCH v2 16/16] xen-blkback: Inform userspace that device has been opened Demi Marie Obenour
2023-06-06 9:15 ` Roger Pau Monné [this message]
2023-06-06 17:31 ` Demi Marie Obenour
2023-06-07 8:44 ` Roger Pau Monné
2023-06-07 16:29 ` Demi Marie Obenour
2023-06-08 9:11 ` Roger Pau Monné
2023-06-08 15:23 ` Demi Marie Obenour
2023-06-08 10:08 ` Roger Pau Monné
2023-06-08 15:24 ` Demi Marie Obenour
2023-05-31 13:06 ` [dm-devel] [PATCH v2 00/16] Diskseq support in loop, device-mapper, and blkback Christoph Hellwig
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=ZH75OTMA6N3zYrH2@Air-de-Roger \
--to=roger.pau@citrix.com \
--cc=agk@redhat.com \
--cc=axboe@kernel.dk \
--cc=demi@invisiblethingslab.com \
--cc=dm-devel@redhat.com \
--cc=linux-block@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=marmarek@invisiblethingslab.com \
--cc=snitzer@kernel.org \
--cc=xen-devel@lists.xenproject.org \
/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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox