Skip to contentSkip to conversation

iommu: Enhance IOMMU default DMA mode build options

PATCH Applied John Garry 2 participants 8 messages Jul 12, 2021 – Jul 26, 2021

Subsystems (4)
lore
Share & tools
lore

John Garry

author
#0 lore raw

v15 cover iommu: Enhance IOMMU default DMA mode build options

This is a reboot of Zhen Lei's series from a couple of years ago, which never made it across the line.

I still think that it has some value, so taking up the mantle.

Motivation:
Allow lazy mode be default mode for DMA domains for all ARCHs, and not only those who hardcode it (to be lazy). For ARM64, currently we must use a kernel command line parameter to use lazy mode, which is less than ideal.

I have now included the print for strict/lazy mode, which I originally sent in:
https://lore.kernel.org/linux-iommu/[email protected]/

There was some concern there about drivers and their custom prints conflicting with the print in that patch, but I think that it should be ok.

Based on v5.14-rc1

Differences to v14:
- Add RB tags (Thanks!)
- Rebase

Differences to v13:
- Improve strict mode deprecation messages and cut out some kernel-parameters.txt legacy description
- Add tag in 1/6
- use pr_info_once() for vt-d message about VM and caching

Differences to v12:
- Add Robin's RB tags (thanks!)
- Add a patch to mark x86 strict cmdline params as deprecated
- Improve wording in Kconfig change and tweak iommu_dma_strict declaration

John Garry (3):
iommu: Deprecate Intel and AMD cmdline methods to enable strict mode
iommu: Print strict or lazy mode at init time
iommu: Remove mode argument from iommu_set_dma_strict()

Zhen Lei (3):
iommu: Enhance IOMMU default DMA mode build options iommu/vt-d: Add support for IOMMU default DMA mode build options iommu/amd: Add support for IOMMU default DMA mode build options

 .../admin-guide/kernel-parameters.txt         | 12 ++----
 drivers/iommu/Kconfig                         | 41 +++++++++++++++++++
 drivers/iommu/amd/amd_iommu_types.h           |  6 ---
 drivers/iommu/amd/init.c                      |  7 ++--
 drivers/iommu/amd/iommu.c                     |  6 ---
 drivers/iommu/intel/iommu.c                   | 16 ++++----
 drivers/iommu/iommu.c                         | 12 ++++--
 include/linux/iommu.h                         |  2 +-
 8 files changed, 65 insertions(+), 37 deletions(-)
Signature · 6 lines
2.26.2

_______________________________________________
iommu mailing list
[email protected]
https://lists.linuxfoundation.org/mailman/listinfo/iommu
author
#1 lore raw

PATCH v15 1/6 iommu: Deprecate Intel and AMD cmdline methods to enable strict mode

Now that the x86 drivers support iommu.strict, deprecate the custom methods.

---
 Documentation/admin-guide/kernel-parameters.txt | 9 ++-------
 drivers/iommu/amd/init.c                        | 4 +++-
 drivers/iommu/intel/iommu.c                     | 1 +
 3 files changed, 6 insertions(+), 8 deletions(-)
Documentation/admin-guide/kernel-parameters.txt +2 -7
index bdb22006f713..a04d2748c99a 100644
@@ -290,10 +290,7 @@
 	amd_iommu=	[HW,X86-64]
 			Pass parameters to the AMD IOMMU driver in the system.
 			Possible values are:
-			fullflush - enable flushing of IO/TLB entries when
-				    they are unmapped. Otherwise they are
-				    flushed before they will be reused, which
-				    is a lot of faster
+			fullflush - Deprecated, equivalent to iommu.strict=1
 			off	  - do not initialize any AMD IOMMU found in
 				    the system
 			force_isolation - Force device isolation for all
@@ -1944,9 +1941,7 @@
 			this case, gfx device will use physical address for
 			DMA.
 		strict [Default Off]
-			With this option on every unmap_single operation will
-			result in a hardware IOTLB flush operation as opposed
-			to batching them for performance.
+			Deprecated, equivalent to iommu.strict=1.
 		sp_off [Default Off]
 			By default, super page will be supported if Intel IOMMU
 			has the capability. With this option, super page will
drivers/iommu/amd/init.c +3 -1
index 46280e6e1535..3a2fb805f11e 100644
@@ -3098,8 +3098,10 @@ static int __init parse_amd_iommu_intr(char *str)
 static int __init parse_amd_iommu_options(char *str)
 {
 	for (; *str; ++str) {
-		if (strncmp(str, "fullflush", 9) == 0)
+		if (strncmp(str, "fullflush", 9) == 0) {
+			pr_warn("amd_iommu=fullflush deprecated; use iommu.strict=1 instead\n");
 			amd_iommu_unmap_flush = true;
+		}
 		if (strncmp(str, "force_enable", 12) == 0)
 			amd_iommu_force_enable = true;
 		if (strncmp(str, "off", 3) == 0)
drivers/iommu/intel/iommu.c +1 -0
index a6a07d985709..67c52b60f8de 100644
@@ -454,6 +454,7 @@ static int __init intel_iommu_setup(char *str)
 			pr_warn("intel_iommu=forcedac deprecated; use iommu.forcedac instead\n");
 			iommu_dma_forcedac = true;
 		} else if (!strncmp(str, "strict", 6)) {
+			pr_warn("intel_iommu=strict deprecated; use iommu.strict=1 instead\n");
 			pr_info("Disable batched IOTLB flush\n");
 			intel_iommu_strict = 1;
 		} else if (!strncmp(str, "sp_off", 6)) {
Signature · 6 lines
2.26.2

_______________________________________________
iommu mailing list
[email protected]
https://lists.linuxfoundation.org/mailman/listinfo/iommu
author
#2 lore raw

PATCH v15 2/6 iommu: Print strict or lazy mode at init time

In reply to John Garry

As well as the default domain type, it's useful to know whether strict or lazy for DMA domains, so add this info in a separate print.

The (stict/lazy) mode may be also set via iommu.strict earlyparm, but this will be processed prior to iommu_subsys_init(), so that print will be accurate for drivers which don't set the mode via custom means.

For the drivers which set the mode via custom means - AMD and Intel drivers
- they maintain prints to inform a change in policy or that custom cmdline methods to change policy are deprecated.

---
 drivers/iommu/iommu.c | 5 +++++
 1 file changed, 5 insertions(+)
drivers/iommu/iommu.c +5 -0
index 5419c4b9f27a..cf58949cc2f3 100644
@@ -138,6 +138,11 @@ static int __init iommu_subsys_init(void)
 		(iommu_cmd_line & IOMMU_CMD_LINE_DMA_API) ?
 			"(set via kernel command line)" : "");
 
+	pr_info("DMA domain TLB invalidation policy: %s mode %s\n",
+		iommu_dma_strict ? "strict" : "lazy",
+		(iommu_cmd_line & IOMMU_CMD_LINE_STRICT) ?
+			"(set via kernel command line)" : "");
+
 	return 0;
 }
 subsys_initcall(iommu_subsys_init);
Signature · 6 lines
2.26.2

_______________________________________________
iommu mailing list
[email protected]
https://lists.linuxfoundation.org/mailman/listinfo/iommu
author
#3 lore raw

PATCH v15 3/6 iommu: Enhance IOMMU default DMA mode build options

In reply to John Garry

From: Zhen Lei <[email protected]>

First, add build options IOMMU_DEFAULT_{LAZY|STRICT}, so that we have the opportunity to set {lazy|strict} mode as default at build time. Then put the two config options in an choice, as they are mutually exclusive.

[jpg: Make choice between strict and lazy only (and not passthrough)]
---
 .../admin-guide/kernel-parameters.txt         |  3 +-
 drivers/iommu/Kconfig                         | 40 +++++++++++++++++++
 drivers/iommu/iommu.c                         |  2 +-
 3 files changed, 43 insertions(+), 2 deletions(-)
Documentation/admin-guide/kernel-parameters.txt +2 -1
index a04d2748c99a..90b525cf0ec2 100644
@@ -2042,9 +2042,10 @@
 			  throughput at the cost of reduced device isolation.
 			  Will fall back to strict mode if not supported by
 			  the relevant IOMMU driver.
-			1 - Strict mode (default).
+			1 - Strict mode.
 			  DMA unmap operations invalidate IOMMU hardware TLBs
 			  synchronously.
+			unset - Use value of CONFIG_IOMMU_DEFAULT_{LAZY,STRICT}.
 			Note: on x86, the default behaviour depends on the
 			equivalent driver-specific parameters, but a strict
 			mode explicitly specified by either method takes
drivers/iommu/Kconfig +40 -0
index 07b7c25cbed8..9cd5d7afc766 100644
@@ -90,6 +90,46 @@ config IOMMU_DEFAULT_PASSTHROUGH
 
 	  If unsure, say N here.
 
+choice
+	prompt "IOMMU default DMA IOTLB invalidation mode"
+	depends on IOMMU_DMA
+
+	default IOMMU_DEFAULT_STRICT
+	help
+	  This option allows an IOMMU DMA IOTLB invalidation mode to be
+	  chosen at build time, to override the default mode of each ARCH,
+	  removing the need to pass in kernel parameters through command line.
+	  It is still possible to provide common boot params to override this
+	  config.
+
+	  If unsure, keep the default.
+
+config IOMMU_DEFAULT_STRICT
+	bool "strict"
+	help
+	  For every IOMMU DMA unmap operation, the flush operation of IOTLB and
+	  the free operation of IOVA are guaranteed to be done in the unmap
+	  function.
+
+config IOMMU_DEFAULT_LAZY
+	bool "lazy"
+	help
+	  Support lazy mode, where for every IOMMU DMA unmap operation, the
+	  flush operation of IOTLB and the free operation of IOVA are deferred.
+	  They are only guaranteed to be done before the related IOVA will be
+	  reused.
+
+	  The isolation provided in this mode is not as secure as STRICT mode,
+	  such that a vulnerable time window may be created between the DMA
+	  unmap and the mappings cached in the IOMMU IOTLB or device TLB
+	  finally being invalidated, where the device could still access the
+	  memory which has already been unmapped by the device driver.
+	  However this mode may provide better performance in high throughput
+	  scenarios, and is still considerably more secure than passthrough
+	  mode or no IOMMU.
+
+endchoice
+
 config OF_IOMMU
 	def_bool y
 	depends on OF && IOMMU_API
drivers/iommu/iommu.c +1 -1
index cf58949cc2f3..60b1ec42e73b 100644
@@ -29,7 +29,7 @@ static struct kset *iommu_group_kset;
 static DEFINE_IDA(iommu_group_ida);
 
 static unsigned int iommu_def_domain_type __read_mostly;
-static bool iommu_dma_strict __read_mostly = true;
+static bool iommu_dma_strict __read_mostly = IS_ENABLED(CONFIG_IOMMU_DEFAULT_STRICT);
 static u32 iommu_cmd_line __read_mostly;
 
 struct iommu_group {
Signature · 6 lines
2.26.2

_______________________________________________
iommu mailing list
[email protected]
https://lists.linuxfoundation.org/mailman/listinfo/iommu
author
#4 lore raw

PATCH v15 4/6 iommu/vt-d: Add support for IOMMU default DMA mode build options

In reply to John Garry

From: Zhen Lei <[email protected]>

Make IOMMU_DEFAULT_LAZY default for when INTEL_IOMMU config is set, as is current behaviour.

Also delete global flag intel_iommu_strict:
- In intel_iommu_setup(), call iommu_set_dma_strict(true) directly. Also remove the print, as iommu_subsys_init() prints the mode and we have already marked this param as deprecated.

- For cap_caching_mode() check in intel_iommu_setup(), call iommu_set_dma_strict(true) directly; also reword the accompanying print with a level downgrade and also add the missing '\n'.

- For Ironlake GPU, again call iommu_set_dma_strict(true) directly and keep the accompanying print.

[jpg: Remove intel_iommu_strict]
---
 drivers/iommu/Kconfig       |  1 +
 drivers/iommu/intel/iommu.c | 15 ++++++---------
 2 files changed, 7 insertions(+), 9 deletions(-)
drivers/iommu/Kconfig +1 -0
index 9cd5d7afc766..265d7a6c9d3a 100644
@@ -94,6 +94,7 @@ choice
 	prompt "IOMMU default DMA IOTLB invalidation mode"
 	depends on IOMMU_DMA
 
+	default IOMMU_DEFAULT_LAZY if INTEL_IOMMU
 	default IOMMU_DEFAULT_STRICT
 	help
 	  This option allows an IOMMU DMA IOTLB invalidation mode to be
drivers/iommu/intel/iommu.c +6 -9
index 67c52b60f8de..d06e8f71c259 100644
@@ -361,7 +361,6 @@ int intel_iommu_enabled = 0;
 EXPORT_SYMBOL_GPL(intel_iommu_enabled);
 
 static int dmar_map_gfx = 1;
-static int intel_iommu_strict;
 static int intel_iommu_superpage = 1;
 static int iommu_identity_mapping;
 static int iommu_skip_te_disable;
@@ -455,8 +454,7 @@ static int __init intel_iommu_setup(char *str)
 			iommu_dma_forcedac = true;
 		} else if (!strncmp(str, "strict", 6)) {
 			pr_warn("intel_iommu=strict deprecated; use iommu.strict=1 instead\n");
-			pr_info("Disable batched IOTLB flush\n");
-			intel_iommu_strict = 1;
+			iommu_set_dma_strict(true);
 		} else if (!strncmp(str, "sp_off", 6)) {
 			pr_info("Disable supported super page\n");
 			intel_iommu_superpage = 0;
@@ -4382,9 +4380,9 @@ int __init intel_iommu_init(void)
 		 * is likely to be much lower than the overhead of synchronizing
 		 * the virtual and physical IOMMU page-tables.
 		 */
-		if (!intel_iommu_strict && cap_caching_mode(iommu->cap)) {
-			pr_warn("IOMMU batching is disabled due to virtualization");
-			intel_iommu_strict = 1;
+		if (cap_caching_mode(iommu->cap)) {
+			pr_info_once("IOMMU batching disallowed due to virtualization\n");
+			iommu_set_dma_strict(true);
 		}
 		iommu_device_sysfs_add(&iommu->iommu, NULL,
 				       intel_iommu_groups,
@@ -4393,7 +4391,6 @@ int __init intel_iommu_init(void)
 	}
 	up_read(&dmar_global_lock);
 
-	iommu_set_dma_strict(intel_iommu_strict);
 	bus_set_iommu(&pci_bus_type, &intel_iommu_ops);
 	if (si_domain && !hw_pass_through)
 		register_memory_notifier(&intel_iommu_memory_nb);
@@ -5703,8 +5700,8 @@ static void quirk_calpella_no_shadow_gtt(struct pci_dev *dev)
 	} else if (dmar_map_gfx) {
 		/* we have to ensure the gfx device is idle before we flush */
 		pci_info(dev, "Disabling batched IOTLB flush on Ironlake\n");
-		intel_iommu_strict = 1;
-       }
+		iommu_set_dma_strict(true);
+	}
 }
 DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_INTEL, 0x0040, quirk_calpella_no_shadow_gtt);
 DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_INTEL, 0x0044, quirk_calpella_no_shadow_gtt);
Signature · 6 lines
2.26.2

_______________________________________________
iommu mailing list
[email protected]
https://lists.linuxfoundation.org/mailman/listinfo/iommu
author
#5 lore raw

PATCH v15 5/6 iommu/amd: Add support for IOMMU default DMA mode build options

In reply to John Garry

From: Zhen Lei <[email protected]>

Make IOMMU_DEFAULT_LAZY default for when AMD_IOMMU config is set, which matches current behaviour.

For "fullflush" param, just call iommu_set_dma_strict(true) directly.

Since we get a strict vs lazy mode print already in iommu_subsys_init(), and maintain a deprecation print when "fullflush" param is passed, drop the prints in amd_iommu_init_dma_ops().

Finally drop global flag amd_iommu_unmap_flush, as it has no longer has any purpose.

[jpg: Rebase for relocated file and drop amd_iommu_unmap_flush]
---
 drivers/iommu/Kconfig               | 2 +-
 drivers/iommu/amd/amd_iommu_types.h | 6 ------
 drivers/iommu/amd/init.c            | 3 +--
 drivers/iommu/amd/iommu.c           | 6 ------
 4 files changed, 2 insertions(+), 15 deletions(-)
drivers/iommu/Kconfig +1 -1
index 265d7a6c9d3a..c84da8205be7 100644
@@ -94,7 +94,7 @@ choice
 	prompt "IOMMU default DMA IOTLB invalidation mode"
 	depends on IOMMU_DMA
 
-	default IOMMU_DEFAULT_LAZY if INTEL_IOMMU
+	default IOMMU_DEFAULT_LAZY if (AMD_IOMMU || INTEL_IOMMU)
 	default IOMMU_DEFAULT_STRICT
 	help
 	  This option allows an IOMMU DMA IOTLB invalidation mode to be
drivers/iommu/amd/amd_iommu_types.h +0 -6
index 94c1a7a9876d..8dbe61e2b3c1 100644
@@ -779,12 +779,6 @@ extern u16 amd_iommu_last_bdf;
 /* allocation bitmap for domain ids */
 extern unsigned long *amd_iommu_pd_alloc_bitmap;
 
-/*
- * If true, the addresses will be flushed on unmap time, not when
- * they are reused
- */
-extern bool amd_iommu_unmap_flush;
-
 /* Smallest max PASID supported by any IOMMU in the system */
 extern u32 amd_iommu_max_pasid;
 
drivers/iommu/amd/init.c +1 -2
index 3a2fb805f11e..1e641cb6dddc 100644
@@ -161,7 +161,6 @@ u16 amd_iommu_last_bdf;			/* largest PCI device id we have
 					   to handle */
 LIST_HEAD(amd_iommu_unity_map);		/* a list of required unity mappings
 					   we find in ACPI */
-bool amd_iommu_unmap_flush;		/* if true, flush on every unmap */
 
 LIST_HEAD(amd_iommu_list);		/* list of all AMD IOMMUs in the
 					   system */
@@ -3100,7 +3099,7 @@ static int __init parse_amd_iommu_options(char *str)
 	for (; *str; ++str) {
 		if (strncmp(str, "fullflush", 9) == 0) {
 			pr_warn("amd_iommu=fullflush deprecated; use iommu.strict=1 instead\n");
-			amd_iommu_unmap_flush = true;
+			iommu_set_dma_strict(true);
 		}
 		if (strncmp(str, "force_enable", 12) == 0)
 			amd_iommu_force_enable = true;
drivers/iommu/amd/iommu.c +0 -6
index 811a49a95d04..52fe2326042a 100644
@@ -1775,12 +1775,6 @@ void amd_iommu_domain_update(struct protection_domain *domain)
 static void __init amd_iommu_init_dma_ops(void)
 {
 	swiotlb = (iommu_default_passthrough() || sme_me_mask) ? 1 : 0;
-
-	if (amd_iommu_unmap_flush)
-		pr_info("IO/TLB flush on unmap enabled\n");
-	else
-		pr_info("Lazy IO/TLB flushing enabled\n");
-	iommu_set_dma_strict(amd_iommu_unmap_flush);
 }
 
 int __init amd_iommu_init_api(void)
Signature · 6 lines
2.26.2

_______________________________________________
iommu mailing list
[email protected]
https://lists.linuxfoundation.org/mailman/listinfo/iommu
author
#6 lore raw

PATCH v15 6/6 iommu: Remove mode argument from iommu_set_dma_strict()

In reply to John Garry

We only ever now set strict mode enabled in iommu_set_dma_strict(), so just remove the argument.

---
 drivers/iommu/amd/init.c    | 2 +-
 drivers/iommu/intel/iommu.c | 6 +++---
 drivers/iommu/iommu.c       | 5 ++---
 include/linux/iommu.h       | 2 +-
 4 files changed, 7 insertions(+), 8 deletions(-)
drivers/iommu/amd/init.c +1 -1
index 1e641cb6dddc..6e12a615117b 100644
@@ -3099,7 +3099,7 @@ static int __init parse_amd_iommu_options(char *str)
 	for (; *str; ++str) {
 		if (strncmp(str, "fullflush", 9) == 0) {
 			pr_warn("amd_iommu=fullflush deprecated; use iommu.strict=1 instead\n");
-			iommu_set_dma_strict(true);
+			iommu_set_dma_strict();
 		}
 		if (strncmp(str, "force_enable", 12) == 0)
 			amd_iommu_force_enable = true;
drivers/iommu/intel/iommu.c +3 -3
index d06e8f71c259..6666089c2607 100644
@@ -454,7 +454,7 @@ static int __init intel_iommu_setup(char *str)
 			iommu_dma_forcedac = true;
 		} else if (!strncmp(str, "strict", 6)) {
 			pr_warn("intel_iommu=strict deprecated; use iommu.strict=1 instead\n");
-			iommu_set_dma_strict(true);
+			iommu_set_dma_strict();
 		} else if (!strncmp(str, "sp_off", 6)) {
 			pr_info("Disable supported super page\n");
 			intel_iommu_superpage = 0;
@@ -4382,7 +4382,7 @@ int __init intel_iommu_init(void)
 		 */
 		if (cap_caching_mode(iommu->cap)) {
 			pr_info_once("IOMMU batching disallowed due to virtualization\n");
-			iommu_set_dma_strict(true);
+			iommu_set_dma_strict();
 		}
 		iommu_device_sysfs_add(&iommu->iommu, NULL,
 				       intel_iommu_groups,
@@ -5700,7 +5700,7 @@ static void quirk_calpella_no_shadow_gtt(struct pci_dev *dev)
 	} else if (dmar_map_gfx) {
 		/* we have to ensure the gfx device is idle before we flush */
 		pci_info(dev, "Disabling batched IOTLB flush on Ironlake\n");
-		iommu_set_dma_strict(true);
+		iommu_set_dma_strict();
 	}
 }
 DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_INTEL, 0x0040, quirk_calpella_no_shadow_gtt);
drivers/iommu/iommu.c +2 -3
index 60b1ec42e73b..ff221d3ddcbc 100644
@@ -349,10 +349,9 @@ static int __init iommu_dma_setup(char *str)
 }
 early_param("iommu.strict", iommu_dma_setup);
 
-void iommu_set_dma_strict(bool strict)
+void iommu_set_dma_strict(void)
 {
-	if (strict || !(iommu_cmd_line & IOMMU_CMD_LINE_STRICT))
-		iommu_dma_strict = strict;
+	iommu_dma_strict = true;
 }
 
 bool iommu_get_dma_strict(struct iommu_domain *domain)
include/linux/iommu.h +1 -1
index 32d448050bf7..754f67d6dd90 100644
@@ -476,7 +476,7 @@ int iommu_enable_nesting(struct iommu_domain *domain);
 int iommu_set_pgtable_quirks(struct iommu_domain *domain,
 		unsigned long quirks);
 
-void iommu_set_dma_strict(bool val);
+void iommu_set_dma_strict(void);
 bool iommu_get_dma_strict(struct iommu_domain *domain);
 
 extern int report_iommu_fault(struct iommu_domain *domain, struct device *dev,
Signature · 6 lines
2.26.2

_______________________________________________
iommu mailing list
[email protected]
https://lists.linuxfoundation.org/mailman/listinfo/iommu
maintainer Applied
#7 lore raw
In reply to John Garry
Quoted 9 lines · John Garry

John Garry (3):
iommu: Deprecate Intel and AMD cmdline methods to enable strict mode
iommu: Print strict or lazy mode at init time
iommu: Remove mode argument from iommu_set_dma_strict()

Zhen Lei (3):
iommu: Enhance IOMMU default DMA mode build options iommu/vt-d: Add support for IOMMU default DMA mode build options iommu/amd: Add support for IOMMU default DMA mode build options

Applied to iommu/core, thanks.