Linux PCI subsystem development
 help / color / mirror / Atom feed
From: Ethan Zhao <haifeng.zhao@linux.intel.com>
To: bhelgaas@google.com, baolu.lu@linux.intel.com,
	dwmw2@infradead.org, will@kernel.org, robin.murphy@arm.com,
	lukas@wunner.de
Cc: linux-pci@vger.kernel.org, iommu@lists.linux.dev,
	linux-kernel@vger.kernel.org
Subject: [RFC PATCH v5 3/3] iommu/vt-d: abort the devTLB invalidation waiting if device is removed
Date: Fri, 22 Dec 2023 05:41:08 -0500	[thread overview]
Message-ID: <20231222104108.18499-4-haifeng.zhao@linux.intel.com> (raw)
In-Reply-To: <20231222104108.18499-1-haifeng.zhao@linux.intel.com>

Even the devTLB invalidation request is just submitted and waiting it
to be done/timeout in qi_submit_sync(), it is possible device is removed
or powered-off. try to break it out in such rare but possible case.

This patch is sent for more comment. not tested, only passed compiling.

Signed-off-by: Ethan Zhao <haifeng.zhao@linux.intel.com>
---
 drivers/iommu/intel/dmar.c  |  3 ++-
 drivers/iommu/intel/iommu.c | 36 ++++++++++++++++++++++++++++++++++++
 2 files changed, 38 insertions(+), 1 deletion(-)

diff --git a/drivers/iommu/intel/dmar.c b/drivers/iommu/intel/dmar.c
index 23cb80d62a9a..d8637ab93387 100644
--- a/drivers/iommu/intel/dmar.c
+++ b/drivers/iommu/intel/dmar.c
@@ -1422,7 +1422,8 @@ int qi_submit_sync(struct intel_iommu *iommu, struct qi_desc *desc,
 	 */
 	writel(qi->free_head << shift, iommu->reg + DMAR_IQT_REG);
 
-	while (qi->desc_status[wait_index] != QI_DONE) {
+	while (qi->desc_status[wait_index] != QI_DONE &&
+			qi->desc_status[wait_index] != QI_ABORT) {
 		/*
 		 * We will leave the interrupts disabled, to prevent interrupt
 		 * context to queue another cmd while a cmd is already submitted
diff --git a/drivers/iommu/intel/iommu.c b/drivers/iommu/intel/iommu.c
index 897159dba47d..33075d0688bc 100644
--- a/drivers/iommu/intel/iommu.c
+++ b/drivers/iommu/intel/iommu.c
@@ -4472,10 +4472,46 @@ static struct iommu_device *intel_iommu_probe_device(struct device *dev)
 	return &iommu->iommu;
 }
 
+static void intel_iommu_abort_devtlib_invalidate(struct device *dev)
+{
+	struct device_domain_info *info = dev_iommu_priv_get(dev);
+	struct intel_iommu *iommu = info->iommu;
+	struct q_inval *qi = iommu->qi;
+	struct qi_desc *desc, *idesc;
+	int index, offset, shift;
+	u16 sid, qdep, pfsid
+	unsigned long flags;
+
+	if (!dev_is_pci(info->dev) || !info->ats_enabled || !qi)
+		return;
+	if (!pci_dev_is_disconnected(to_pci_dev(dev)))
+		return;
+
+	sid = info->bus << 8 | info->devfn;
+	qdep = info->ats_qdep;
+	pfsid = info->pfsid;
+
+	raw_spin_lock_irqsave(&qi->q_lock, flags);
+	for (index = 1; index < QI_LENGTH; index++) {
+		offset = index << shift;
+		desc = qi->desc + offset;
+		if (desc->qw0 & QI_IWD_TYPE) {
+			offset = (index-1) << shift;
+			idesc = qi->desc + offset;
+			if (idesc->qw0 & QI_DEV_EIOTLB_SID(sid)) {
+				if (qi->desc_status[index] == QI_IN_USE)
+					qi->desc_status[index] = QI_ABORT;
+			}
+		}
+	}
+	raw_spin_unlock_irqrestore(&qi->q_lock, flags);
+
+}
 static void intel_iommu_release_device(struct device *dev)
 {
 	struct device_domain_info *info = dev_iommu_priv_get(dev);
 
+	intel_iommu_abort_devtlib_invalidate(dev);
 	dmar_remove_one_dev_info(dev);
 	intel_pasid_free_table(dev);
 	intel_iommu_debugfs_remove_dev(info);
-- 
2.31.1


  parent reply	other threads:[~2023-12-22 10:41 UTC|newest]

Thread overview: 6+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2023-12-22 10:41 [RFC PATCH v5 0/3] fix vt-d hard lockup when hotplug ATS capable device Ethan Zhao
2023-12-22 10:41 ` [RFC PATCH v5 1/3] PCI: make pci_dev_is_disconnected() helper public for other drivers Ethan Zhao
2023-12-22 10:41 ` [RFC PATCH v5 2/3] iommu/vt-d: don's issue devTLB flush request when device is disconnected Ethan Zhao
2023-12-22 10:41 ` Ethan Zhao [this message]
2023-12-22 23:35   ` [RFC PATCH v5 3/3] iommu/vt-d: abort the devTLB invalidation waiting if device is removed Ethan Zhao
2023-12-23  0:24     ` Ethan Zhao

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=20231222104108.18499-4-haifeng.zhao@linux.intel.com \
    --to=haifeng.zhao@linux.intel.com \
    --cc=baolu.lu@linux.intel.com \
    --cc=bhelgaas@google.com \
    --cc=dwmw2@infradead.org \
    --cc=iommu@lists.linux.dev \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-pci@vger.kernel.org \
    --cc=lukas@wunner.de \
    --cc=robin.murphy@arm.com \
    --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