From: Maurizio Lombardi <mlombard@redhat.com>
To: kbusch@kernel.org
Cc: mheyne@amazon.de, emilne@redhat.com, jmeneghi@redhat.com,
linux-nvme@lists.infradead.org, dwagner@suse.de,
mlombard@arkamax.eu, mkhalfella@purestorage.com,
chaitanyak@nvidia.com, hare@kernel.org, hch@lst.de
Subject: [PATCH V4 3/9] nvme: fix race condition between connected uevent and STARTED_ONCE flag
Date: Fri, 8 May 2026 15:33:29 +0200 [thread overview]
Message-ID: <20260508133335.98612-4-mlombard@redhat.com> (raw)
In-Reply-To: <20260508133335.98612-1-mlombard@redhat.com>
When a controller connects, nvme_start_ctrl() emits the
"NVME_EVENT=connected" uevent and sets the NVME_CTRL_STARTED_ONCE flag.
Currently, the uevent is emitted before the flag is set.
This creates a race condition for userspace tools (like udev rules)
that might rely on the "connected" event to configure sysfs attributes.
Specifically, if a udev rule attempts to set the newly introduced
`admin_timeout` attribute immediately after receiving the uevent,
the sysfs store function might evaluate the NVME_CTRL_STARTED_ONCE
bit before it is actually set, resulting in spurious -EBUSY error.
Swap the order of operations in nvme_start_ctrl() so that the
NVME_CTRL_STARTED_ONCE flag is set before the uevent is sent.
This guarantees that the admin_timeout can already be changed
when userspace is notified.
Signed-off-by: Maurizio Lombardi <mlombard@redhat.com>
---
drivers/nvme/host/core.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/drivers/nvme/host/core.c b/drivers/nvme/host/core.c
index b1bfcd0a0e5b..22be8cf5e982 100644
--- a/drivers/nvme/host/core.c
+++ b/drivers/nvme/host/core.c
@@ -5039,8 +5039,8 @@ void nvme_start_ctrl(struct nvme_ctrl *ctrl)
nvme_mpath_update(ctrl);
}
- nvme_change_uevent(ctrl, "NVME_EVENT=connected");
set_bit(NVME_CTRL_STARTED_ONCE, &ctrl->flags);
+ nvme_change_uevent(ctrl, "NVME_EVENT=connected");
}
EXPORT_SYMBOL_GPL(nvme_start_ctrl);
--
2.54.0
next prev parent reply other threads:[~2026-05-08 13:34 UTC|newest]
Thread overview: 46+ messages / expand[flat|nested] mbox.gz Atom feed top
2026-05-08 13:33 [PATCH V4 0/9] nvme: Refactor and expose per-controller timeout configuration Maurizio Lombardi
2026-05-08 13:33 ` [PATCH V4 1/9] nvme: Let the blocklayer set timeouts for requests Maurizio Lombardi
2026-05-11 9:37 ` Hannes Reinecke
2026-05-08 13:33 ` [PATCH V4 2/9] nvme: add sysfs attribute to change admin timeout per nvme controller Maurizio Lombardi
2026-05-08 16:57 ` Daniel Wagner
2026-05-10 22:10 ` Sagi Grimberg
2026-05-11 8:07 ` Christoph Hellwig
2026-05-11 11:29 ` Maurizio Lombardi
2026-05-11 12:31 ` Christoph Hellwig
2026-05-11 9:46 ` Hannes Reinecke
2026-05-11 10:05 ` Maurizio Lombardi
2026-05-08 13:33 ` Maurizio Lombardi [this message]
2026-05-08 16:57 ` [PATCH V4 3/9] nvme: fix race condition between connected uevent and STARTED_ONCE flag Daniel Wagner
2026-05-10 22:10 ` Sagi Grimberg
2026-05-11 8:07 ` Christoph Hellwig
2026-05-11 12:54 ` Sagi Grimberg
2026-05-11 15:09 ` Keith Busch
2026-05-11 15:45 ` Maurizio Lombardi
2026-05-11 17:10 ` Keith Busch
2026-05-11 8:08 ` Christoph Hellwig
2026-05-11 9:47 ` Hannes Reinecke
2026-05-08 13:33 ` [PATCH V4 4/9] nvme: pci: use admin queue timeout over NVME_ADMIN_TIMEOUT Maurizio Lombardi
2026-05-10 22:10 ` Sagi Grimberg
2026-05-11 8:08 ` Christoph Hellwig
2026-05-11 9:48 ` Hannes Reinecke
2026-05-08 13:33 ` [PATCH V4 5/9] nvme: add sysfs attribute to change IO timeout per controller Maurizio Lombardi
2026-05-08 17:08 ` Daniel Wagner
2026-05-10 22:12 ` Sagi Grimberg
2026-05-11 8:52 ` Maurizio Lombardi
2026-05-08 13:33 ` [PATCH V4 6/9] nvme: use per controller timeout waits over depending on global default Maurizio Lombardi
2026-05-11 8:10 ` Christoph Hellwig
2026-05-11 11:42 ` Maurizio Lombardi
2026-05-11 12:32 ` Christoph Hellwig
2026-05-11 9:50 ` Hannes Reinecke
2026-05-08 13:33 ` [PATCH V4 7/9] nvme-core: align fabrics_q teardown with admin_q in nvme_free_ctrl Maurizio Lombardi
2026-05-10 22:15 ` Sagi Grimberg
2026-05-11 8:11 ` Christoph Hellwig
2026-05-11 9:53 ` Hannes Reinecke
2026-05-11 9:57 ` Maurizio Lombardi
2026-05-08 13:33 ` [PATCH V4 8/9] nvmet-loop: do not alloc admin tag set during reset Maurizio Lombardi
2026-05-08 17:09 ` Daniel Wagner
2026-05-10 22:16 ` Sagi Grimberg
2026-05-11 8:12 ` Christoph Hellwig
2026-05-11 9:55 ` Hannes Reinecke
2026-05-08 13:33 ` [PATCH V4 9/9] nvme-core: warn on allocating admin tag set with existing queue Maurizio Lombardi
2026-05-10 22:16 ` Sagi Grimberg
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=20260508133335.98612-4-mlombard@redhat.com \
--to=mlombard@redhat.com \
--cc=chaitanyak@nvidia.com \
--cc=dwagner@suse.de \
--cc=emilne@redhat.com \
--cc=hare@kernel.org \
--cc=hch@lst.de \
--cc=jmeneghi@redhat.com \
--cc=kbusch@kernel.org \
--cc=linux-nvme@lists.infradead.org \
--cc=mheyne@amazon.de \
--cc=mkhalfella@purestorage.com \
--cc=mlombard@arkamax.eu \
/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