lists.openwall.net   lists  /  announce  owl-users  owl-dev  john-users  john-dev  passwdqc-users  yescrypt  popa3d-users  /  oss-security  kernel-hardening  musl  sabotage  tlsify  passwords  /  crypt-dev  xvendor  /  Bugtraq  Full-Disclosure  linux-kernel  linux-netdev  linux-ext4  linux-hardening  linux-cve-announce  PHC 
Open Source and information security mailing list archives
 
Hash Suite: Windows password security audit tool. GUI, reports in PDF.
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-Id: <20221018145132.998866-4-schnelle@linux.ibm.com>
Date:   Tue, 18 Oct 2022 16:51:30 +0200
From:   Niklas Schnelle <schnelle@...ux.ibm.com>
To:     Matthew Rosato <mjrosato@...ux.ibm.com>, iommu@...ts.linux.dev,
        Joerg Roedel <joro@...tes.org>, Will Deacon <will@...nel.org>,
        Robin Murphy <robin.murphy@....com>,
        Jason Gunthorpe <jgg@...dia.com>
Cc:     Gerd Bayer <gbayer@...ux.ibm.com>,
        Pierre Morel <pmorel@...ux.ibm.com>,
        linux-s390@...r.kernel.org, borntraeger@...ux.ibm.com,
        hca@...ux.ibm.com, gor@...ux.ibm.com,
        gerald.schaefer@...ux.ibm.com, agordeev@...ux.ibm.com,
        svens@...ux.ibm.com, linux-kernel@...r.kernel.org
Subject: [PATCH 3/5] iommu/s390: Use RCU to allow concurrent domain_list iteration

The s390_domain->devices list is only added to when new devices are
attached but is iterated through in read-only fashion for every mapping
operation as well as for I/O TLB flushes and thus in performance
critical code causing contention on the s390_domain->list_lock.
Fortunately such a read-mostly linked list is a standard use case for
RCU. This change closely follows the example fpr RCU protected list
given in Documentation/RCU/listRCU.rst.

Signed-off-by: Niklas Schnelle <schnelle@...ux.ibm.com>
---
 arch/s390/include/asm/pci.h |  1 +
 arch/s390/pci/pci.c         |  2 +-
 drivers/iommu/s390-iommu.c  | 31 ++++++++++++++++---------------
 3 files changed, 18 insertions(+), 16 deletions(-)

diff --git a/arch/s390/include/asm/pci.h b/arch/s390/include/asm/pci.h
index 07361e2fd8c5..e4c3e4e04d30 100644
--- a/arch/s390/include/asm/pci.h
+++ b/arch/s390/include/asm/pci.h
@@ -119,6 +119,7 @@ struct zpci_dev {
 	struct list_head entry;		/* list of all zpci_devices, needed for hotplug, etc. */
 	struct list_head iommu_list;
 	struct kref kref;
+	struct rcu_head rcu;
 	struct hotplug_slot hotplug_slot;
 
 	enum zpci_state state;
diff --git a/arch/s390/pci/pci.c b/arch/s390/pci/pci.c
index a703dcd94a68..ef38b1514c77 100644
--- a/arch/s390/pci/pci.c
+++ b/arch/s390/pci/pci.c
@@ -996,7 +996,7 @@ void zpci_release_device(struct kref *kref)
 		break;
 	}
 	zpci_dbg(3, "rem fid:%x\n", zdev->fid);
-	kfree(zdev);
+	kfree_rcu(zdev, rcu);
 }
 
 int zpci_report_error(struct pci_dev *pdev,
diff --git a/drivers/iommu/s390-iommu.c b/drivers/iommu/s390-iommu.c
index a4c2e9bc6d83..4e90987be387 100644
--- a/drivers/iommu/s390-iommu.c
+++ b/drivers/iommu/s390-iommu.c
@@ -10,6 +10,8 @@
 #include <linux/iommu.h>
 #include <linux/iommu-helper.h>
 #include <linux/sizes.h>
+#include <linux/rculist.h>
+#include <linux/rcupdate.h>
 #include <asm/pci_dma.h>
 
 static const struct iommu_ops s390_iommu_ops;
@@ -61,7 +63,7 @@ static struct iommu_domain *s390_domain_alloc(unsigned domain_type)
 
 	spin_lock_init(&s390_domain->dma_table_lock);
 	spin_lock_init(&s390_domain->list_lock);
-	INIT_LIST_HEAD(&s390_domain->devices);
+	INIT_LIST_HEAD_RCU(&s390_domain->devices);
 
 	return &s390_domain->domain;
 }
@@ -70,7 +72,9 @@ static void s390_domain_free(struct iommu_domain *domain)
 {
 	struct s390_domain *s390_domain = to_s390_domain(domain);
 
+	rcu_read_lock();
 	WARN_ON(!list_empty(&s390_domain->devices));
+	rcu_read_unlock();
 	dma_cleanup_tables(s390_domain->dma_table);
 	kfree(s390_domain);
 }
@@ -84,7 +88,7 @@ static void __s390_iommu_detach_device(struct zpci_dev *zdev)
 		return;
 
 	spin_lock_irqsave(&s390_domain->list_lock, flags);
-	list_del_init(&zdev->iommu_list);
+	list_del_rcu(&zdev->iommu_list);
 	spin_unlock_irqrestore(&s390_domain->list_lock, flags);
 
 	zpci_unregister_ioat(zdev, 0);
@@ -127,7 +131,7 @@ static int s390_iommu_attach_device(struct iommu_domain *domain,
 	zdev->s390_domain = s390_domain;
 
 	spin_lock_irqsave(&s390_domain->list_lock, flags);
-	list_add(&zdev->iommu_list, &s390_domain->devices);
+	list_add_rcu(&zdev->iommu_list, &s390_domain->devices);
 	spin_unlock_irqrestore(&s390_domain->list_lock, flags);
 
 	return 0;
@@ -203,17 +207,16 @@ static void s390_iommu_flush_iotlb_all(struct iommu_domain *domain)
 {
 	struct s390_domain *s390_domain = to_s390_domain(domain);
 	struct zpci_dev *zdev;
-	unsigned long flags;
 	int rc;
 
-	spin_lock_irqsave(&s390_domain->list_lock, flags);
-	list_for_each_entry(zdev, &s390_domain->devices, iommu_list) {
+	rcu_read_lock();
+	list_for_each_entry_rcu(zdev, &s390_domain->devices, iommu_list) {
 		rc = zpci_refresh_trans((u64)zdev->fh << 32, zdev->start_dma,
 					zdev->end_dma - zdev->start_dma + 1);
 		if (rc)
 			break;
 	}
-	spin_unlock_irqrestore(&s390_domain->list_lock, flags);
+	rcu_read_unlock();
 }
 
 static void s390_iommu_iotlb_sync(struct iommu_domain *domain,
@@ -222,21 +225,20 @@ static void s390_iommu_iotlb_sync(struct iommu_domain *domain,
 	struct s390_domain *s390_domain = to_s390_domain(domain);
 	size_t size = gather->end - gather->start + 1;
 	struct zpci_dev *zdev;
-	unsigned long flags;
 	int rc;
 
 	/* If gather was never added to there is nothing to flush */
 	if (gather->start == ULONG_MAX)
 		return;
 
-	spin_lock_irqsave(&s390_domain->list_lock, flags);
-	list_for_each_entry(zdev, &s390_domain->devices, iommu_list) {
+	rcu_read_lock();
+	list_for_each_entry_rcu(zdev, &s390_domain->devices, iommu_list) {
 		rc = zpci_refresh_trans((u64)zdev->fh << 32, gather->start,
 					size);
 		if (rc)
 			break;
 	}
-	spin_unlock_irqrestore(&s390_domain->list_lock, flags);
+	rcu_read_unlock();
 }
 
 static void s390_iommu_iotlb_sync_map(struct iommu_domain *domain,
@@ -244,11 +246,10 @@ static void s390_iommu_iotlb_sync_map(struct iommu_domain *domain,
 {
 	struct s390_domain *s390_domain = to_s390_domain(domain);
 	struct zpci_dev *zdev;
-	unsigned long flags;
 	int rc;
 
-	spin_lock_irqsave(&s390_domain->list_lock, flags);
-	list_for_each_entry(zdev, &s390_domain->devices, iommu_list) {
+	rcu_read_lock();
+	list_for_each_entry_rcu(zdev, &s390_domain->devices, iommu_list) {
 		if (!zdev->tlb_refresh)
 			continue;
 		rc = zpci_refresh_trans((u64)zdev->fh << 32,
@@ -256,7 +257,7 @@ static void s390_iommu_iotlb_sync_map(struct iommu_domain *domain,
 		if (rc)
 			break;
 	}
-	spin_unlock_irqrestore(&s390_domain->list_lock, flags);
+	rcu_read_unlock();
 }
 
 static int s390_iommu_update_trans(struct s390_domain *s390_domain,
-- 
2.34.1

Powered by blists - more mailing lists

Powered by Openwall GNU/*/Linux Powered by OpenVZ