From: Anlai Lu <agicy@qq.com>
To: Jean-Philippe Brucker <jpb@kernel.org>,
Joerg Roedel <joro@8bytes.org>, Will Deacon <will@kernel.org>
Cc: Robin Murphy <robin.murphy@arm.com>,
virtualization@lists.linux.dev, iommu@lists.linux.dev,
linux-kernel@vger.kernel.org, Anlai Lu <agicy@qq.com>
Subject: [PATCH 5/6] iommu/virtio: read the endpoint count under the lock in the iotlb paths
Date: Sun, 4 Oct 2026 11:57:08 +0000 [thread overview]
Message-ID: <tencent_C7BA3FB54A956BDCEC9B48C7741F06566708@qq.com> (raw)
In-Reply-To: <tencent_85AC6E749F6A912B7BB9FF360F8ECB3F5C09@qq.com>
viommu_iotlb_sync_map() and viommu_flush_iotlb_all() read nr_endpoints
without the lock its writers take: the one place left after the endpoint
helpers were introduced. Order a sync only when the device holds the
domain, and read the count under the same lock as the writers.
Signed-off-by: Anlai Lu <agicy@qq.com>
---
drivers/iommu/virtio-iommu.c | 19 +++++++++++++++++--
1 file changed, 17 insertions(+), 2 deletions(-)
diff --git a/drivers/iommu/virtio-iommu.c b/drivers/iommu/virtio-iommu.c
index ba1d78e4daa0..4e578d4255f6 100644
--- a/drivers/iommu/virtio-iommu.c
+++ b/drivers/iommu/virtio-iommu.c
@@ -131,6 +131,19 @@ static bool viommu_device_live(struct viommu_dev *viommu)
return !viommu->removed;
}
+/* Does the device hold this domain for at least one endpoint? */
+static bool viommu_domain_has_endpoint(struct viommu_domain *vdomain)
+{
+ unsigned long flags;
+ bool has;
+
+ spin_lock_irqsave(&vdomain->mappings_lock, flags);
+ has = vdomain->nr_endpoints != 0;
+ spin_unlock_irqrestore(&vdomain->mappings_lock, flags);
+
+ return has;
+}
+
static int viommu_get_req_errno(void *buf, size_t len)
{
struct virtio_iommu_req_tail *tail = buf + len - sizeof(*tail);
@@ -1273,8 +1286,10 @@ static int viommu_iotlb_sync_map(struct iommu_domain *domain,
* May be called before the viommu is initialized including
* while creating direct mapping
*/
- if (!vdomain->nr_endpoints)
+ if (!viommu_domain_has_endpoint(vdomain))
return 0;
+
+ /* Wait for this batch's MAPs, whose outcome nobody looks at */
return viommu_sync_req(vdomain->viommu);
}
@@ -1286,7 +1301,7 @@ static void viommu_flush_iotlb_all(struct iommu_domain *domain)
* May be called before the viommu is initialized including
* while creating direct mapping
*/
- if (!vdomain->nr_endpoints)
+ if (!viommu_domain_has_endpoint(vdomain))
return;
viommu_sync_req(vdomain->viommu);
}
--
2.55.0
next prev parent reply other threads:[~2026-10-04 11:57 UTC|newest]
Thread overview: 7+ messages / expand[flat|nested] mbox.gz Atom feed top
2026-10-04 11:55 [PATCH 0/6] iommu/virtio: batch the mapping replay Anlai Lu
2026-10-04 11:57 ` [PATCH 1/6] iommu/virtio: publish the endpoint before replaying it Anlai Lu
2026-10-04 11:57 ` [PATCH 2/6] iommu/virtio: queue an UNMAP for every removed mapping Anlai Lu
2026-10-04 11:57 ` [PATCH 3/6] iommu/virtio: allocate the UNMAP request with the mapping Anlai Lu
2026-10-04 11:57 ` [PATCH 4/6] iommu/virtio: stop queueing and draining once the device is removed Anlai Lu
2026-10-04 11:57 ` Anlai Lu [this message]
2026-10-04 11:57 ` [PATCH 6/6] iommu/virtio: batch the mapping replay on domain attach Anlai Lu
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=tencent_C7BA3FB54A956BDCEC9B48C7741F06566708@qq.com \
--to=agicy@qq.com \
--cc=iommu@lists.linux.dev \
--cc=joro@8bytes.org \
--cc=jpb@kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=robin.murphy@arm.com \
--cc=virtualization@lists.linux.dev \
--cc=will@kernel.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
all inboxes | Powered by JetHome®