Skip to contentSkip to conversation

kvm: arm64: allow the VM to select DEVICE_* and NORMAL_NC for IO memory

PATCH Superseded [email protected] 6 participants 19 messages Feb 7, 2024 – Feb 9, 2024

Subsystems (6)
Memory management KVM MM core arm64 +2 more subsystems

v6 of 9 Superseded Latest is v9 →

lore
Share & tools
lore
#0 lore raw

v6 cover kvm: arm64: allow the VM to select DEVICE_* and NORMAL_NC for IO memory

From: Ankit Agrawal <[email protected]>

Currently, KVM for ARM64 maps at stage 2 memory that is considered device with DEVICE_nGnRE memory attributes; this setting overrides (per ARM architecture [1]) any device MMIO mapping present at stage 1, resulting in a set-up whereby a guest operating system cannot determine device MMIO mapping memory attributes on its own but it is always overridden by the KVM stage 2 default.

This set-up does not allow guest operating systems to select device memory attributes independently from KVM stage-2 mappings (refer to [1], "Combining stage 1 and stage 2 memory type attributes"), which turns out to be an issue in that guest operating systems (e.g. Linux) may request to map devices MMIO regions with memory attributes that guarantee better performance (e.g. gathering attribute - that for some devices can generate larger PCIe memory writes TLPs) and specific operations (e.g. unaligned transactions) such as the NormalNC memory type.

The default device stage 2 mapping was chosen in KVM for ARM64 since it was considered safer (i.e. it would not allow guests to trigger uncontained failures ultimately crashing the machine) but this turned out to be asynchronous (SError) defeating the purpose.

For these reasons, relax the KVM stage 2 device memory attributes from DEVICE_nGnRE to Normal-NC.

Generalizing to other devices may be problematic, however. E.g. GICv2 VCPU interface, which is effectively a shared peripheral, can allow a guest to affect another guest's interrupt distribution. Hence limit the change to VFIO PCI as caution. This is achieved by making the VFIO PCI core module set a flag that is tested by KVM to activate the code. This could be extended to other devices in the future once that is deemed safe.

[1] section D8.5 - DDI0487J_a_a-profile_architecture_reference_manual.pdf

Applied over v6.8-rc2.

History
=======
v5 -> v6
- Rebased to v6.8-rc2

v4 -> v5
- Moved the cover letter description text to patch 1/4.
- Cleaned up stage2_set_prot_attr() based on Marc Zyngier suggestions.
- Moved the mm header file changes to a separate patch.
- Rebased to v6.7-rc3.

v3 -> v4
- Moved the vfio-pci change to use the VM_VFIO_ALLOW_WC into separate patch.
- Added check to warn on the case NORMAL_NC and DEVICE are set simultaneously.
- Fixed miscellaneous nitpicks suggested in v3.

v2 -> v3
- Added a new patch (and converted to patch series) suggested by Catalin Marinas to ensure the code changes are restricted to VFIO PCI devices.
- Introduced VM_VFIO_ALLOW_WC flag for VFIO PCI to communicate with VMM.
- Reverted GIC mapping to DEVICE.

v1 -> v2
- Updated commit log to the one posted by
Lorenzo Pieralisi <[email protected]> (Thanks!)
- Added new flag to represent the NORMAL_NC setting. Updated stage2_set_prot_attr() to handle new flag.

v5 Link:
https://lore.kernel.org/all/[email protected]/

Ankit Agrawal (4):
kvm: arm64: introduce new flag for non-cacheable IO memory
mm: introduce new flag to indicate wc safe
kvm: arm64: set io memory s2 pte as normalnc for vfio pci device
vfio: convey kvm that the vfio-pci device is wc safe

 arch/arm64/include/asm/kvm_pgtable.h |  2 ++
 arch/arm64/include/asm/memory.h      |  2 ++
 arch/arm64/kvm/hyp/pgtable.c         | 23 ++++++++++++++++++-----
 arch/arm64/kvm/mmu.c                 | 18 ++++++++++++++----
 drivers/vfio/pci/vfio_pci_core.c     |  3 ++-
 include/linux/mm.h                   | 14 ++++++++++++++
 6 files changed, 52 insertions(+), 10 deletions(-)
Signature · 1 line
2.34.1
#1 lore raw

PATCH v6 1/4 kvm: arm64: introduce new flag for non-cacheable IO memory

From: Ankit Agrawal <[email protected]>

Currently, KVM for ARM64 maps at stage 2 memory that is considered device (i.e. it is not RAM) with DEVICE_nGnRE memory attributes; this setting overrides (as per the ARM architecture [1]) any device MMIO mapping present at stage 1, resulting in a set-up whereby a guest operating system cannot determine device MMIO mapping memory attributes on its own but it is always overridden by the KVM stage 2 default.

This set-up does not allow guest operating systems to select device memory attributes independently from KVM stage-2 mappings (refer to [1], "Combining stage 1 and stage 2 memory type attributes"), which turns out to be an issue in that guest operating systems (e.g. Linux) may request to map devices MMIO regions with memory attributes that guarantee better performance (e.g. gathering attribute - that for some devices can generate larger PCIe memory writes TLPs) and specific operations (e.g. unaligned transactions) such as the NormalNC memory type.

The default device stage 2 mapping was chosen in KVM for ARM64 since it was considered safer (i.e. it would not allow guests to trigger uncontained failures ultimately crashing the machine) but this turned out to be asynchronous (SError) defeating the purpose.

Failures containability is a property of the platform and is independent from the memory type used for MMIO device memory mappings.

Actually, DEVICE_nGnRE memory type is even more problematic than Normal-NC memory type in terms of faults containability in that e.g. aborts triggered on DEVICE_nGnRE loads cannot be made, architecturally, synchronous (i.e. that would imply that the processor should issue at most 1 load transaction at a time - it cannot pipeline them - otherwise the synchronous abort semantics would break the no-speculation attribute attached to DEVICE_XXX memory).

This means that regardless of the combined stage1+stage2 mappings a platform is safe if and only if device transactions cannot trigger uncontained failures and that in turn relies on platform capabilities and the device type being assigned (i.e. PCIe AER/DPC error containment and RAS architecture[3]); therefore the default KVM device stage 2 memory attributes play no role in making device assignment safer for a given platform (if the platform design adheres to design guidelines outlined in [3]) and therefore can be relaxed.

For all these reasons, relax the KVM stage 2 device memory attributes from DEVICE_nGnRE to Normal-NC.

The NormalNC was chosen over a different Normal memory type default at stage-2 (e.g. Normal Write-through) to avoid cache allocation/snooping.

Relaxing S2 KVM device MMIO mappings to Normal-NC is not expected to trigger any issue on guest device reclaim use cases either (i.e. device MMIO unmap followed by a device reset) at least for PCIe devices, in that in PCIe a device reset is architected and carried out through PCI config space transactions that are naturally ordered with respect to MMIO transactions according to the PCI ordering rules.

Having Normal-NC S2 default puts guests in control (thanks to stage1+stage2 combined memory attributes rules [1]) of device MMIO regions memory mappings, according to the rules described in [1] and summarized here ([(S1) - stage1], [(S2) - stage 2]):

S1           |  S2           | Result
NORMAL-WB    |  NORMAL-NC    | NORMAL-NC
NORMAL-WT    |  NORMAL-NC    | NORMAL-NC
NORMAL-NC    |  NORMAL-NC    | NORMAL-NC
DEVICE<attr> |  NORMAL-NC    | DEVICE<attr>

It is worth noting that currently, to map devices MMIO space to user space in a device pass-through use case the VFIO framework applies memory attributes derived from pgprot_noncached() settings applied to VMAs, which result in device-nGnRnE memory attributes for the stage-1 VMM mappings.

This means that a userspace mapping for device MMIO space carried out with the current VFIO framework and a guest OS mapping for the same MMIO space may result in a mismatched alias as described in [2].

Defaulting KVM device stage-2 mappings to Normal-NC attributes does not change anything in this respect, in that the mismatched aliases would only affect (refer to [2] for a detailed explanation) ordering between the userspace and GuestOS mappings resulting stream of transactions (i.e. it does not cause loss of property for either stream of transactions on its own), which is harmless given that the userspace and GuestOS access to the device is carried out through independent transactions streams.

A Normal-NC flag is not present today. So add a new kvm_pgtable_prot (KVM_PGTABLE_PROT_NORMAL_NC) flag for it, along with its corresponding PTE value 0x5 (0b101) determined from [1].

Lastly, adapt the stage2 PTE property setter function (stage2_set_prot_attr) to handle the NormalNC attribute.

[1] section D8.5.5 - DDI0487J_a_a-profile_architecture_reference_manual.pdf
[2] section B2.8 - DDI0487J_a_a-profile_architecture_reference_manual.pdf
[3] sections 1.7.7.3/1.8.5.2/appendix C - DEN0029H_SBSA_7.1.pdf

---
 arch/arm64/include/asm/kvm_pgtable.h |  2 ++
 arch/arm64/include/asm/memory.h      |  2 ++
 arch/arm64/kvm/hyp/pgtable.c         | 23 ++++++++++++++++++-----
 3 files changed, 22 insertions(+), 5 deletions(-)
arch/arm64/include/asm/kvm_pgtable.h +2 -0
index cfdf40f734b1..19278dfe7978 100644
@@ -197,6 +197,7 @@ enum kvm_pgtable_stage2_flags {
  * @KVM_PGTABLE_PROT_W:		Write permission.
  * @KVM_PGTABLE_PROT_R:		Read permission.
  * @KVM_PGTABLE_PROT_DEVICE:	Device attributes.
+ * @KVM_PGTABLE_PROT_NORMAL_NC:	Normal noncacheable attributes.
  * @KVM_PGTABLE_PROT_SW0:	Software bit 0.
  * @KVM_PGTABLE_PROT_SW1:	Software bit 1.
  * @KVM_PGTABLE_PROT_SW2:	Software bit 2.
@@ -208,6 +209,7 @@ enum kvm_pgtable_prot {
 	KVM_PGTABLE_PROT_R			= BIT(2),
 
 	KVM_PGTABLE_PROT_DEVICE			= BIT(3),
+	KVM_PGTABLE_PROT_NORMAL_NC		= BIT(4),
 
 	KVM_PGTABLE_PROT_SW0			= BIT(55),
 	KVM_PGTABLE_PROT_SW1			= BIT(56),
arch/arm64/include/asm/memory.h +2 -0
index d82305ab420f..449ca2ff1df6 100644
@@ -173,6 +173,7 @@
  * Memory types for Stage-2 translation
  */
 #define MT_S2_NORMAL		0xf
+#define MT_S2_NORMAL_NC		0x5
 #define MT_S2_DEVICE_nGnRE	0x1
 
 /*
@@ -180,6 +181,7 @@
  * Stage-2 enforces Normal-WB and Device-nGnRE
  */
 #define MT_S2_FWB_NORMAL	6
+#define MT_S2_FWB_NORMAL_NC	5
 #define MT_S2_FWB_DEVICE_nGnRE	1
 
 #ifdef CONFIG_ARM64_4K_PAGES
arch/arm64/kvm/hyp/pgtable.c +18 -5
index c651df904fe3..2a893724ee9b 100644
@@ -717,15 +717,28 @@ void kvm_tlb_flush_vmid_range(struct kvm_s2_mmu *mmu,
 static int stage2_set_prot_attr(struct kvm_pgtable *pgt, enum kvm_pgtable_prot prot,
 				kvm_pte_t *ptep)
 {
-	bool device = prot & KVM_PGTABLE_PROT_DEVICE;
-	kvm_pte_t attr = device ? KVM_S2_MEMATTR(pgt, DEVICE_nGnRE) :
-			    KVM_S2_MEMATTR(pgt, NORMAL);
+	kvm_pte_t attr;
 	u32 sh = KVM_PTE_LEAF_ATTR_LO_S2_SH_IS;
 
+	switch (prot & (KVM_PGTABLE_PROT_DEVICE |
+			KVM_PGTABLE_PROT_NORMAL_NC)) {
+	case 0:
+		attr = KVM_S2_MEMATTR(pgt, NORMAL);
+		break;
+	case KVM_PGTABLE_PROT_DEVICE:
+		if (prot & KVM_PGTABLE_PROT_X)
+			return -EINVAL;
+		attr = KVM_S2_MEMATTR(pgt, DEVICE_nGnRE);
+		break;
+	case KVM_PGTABLE_PROT_NORMAL_NC:
+		attr = KVM_S2_MEMATTR(pgt, NORMAL_NC);
+		break;
+	default:
+		WARN_ON_ONCE(1);
+	}
+
 	if (!(prot & KVM_PGTABLE_PROT_X))
 		attr |= KVM_PTE_LEAF_ATTR_HI_S2_XN;
-	else if (device)
-		return -EINVAL;
 
 	if (prot & KVM_PGTABLE_PROT_R)
 		attr |= KVM_PTE_LEAF_ATTR_LO_S2_S2AP_R;
Signature · 1 line
2.34.1
maintainer
#2 lore raw

Re: [PATCH v6 1/4] kvm: arm64: introduce new flag for non-cacheable IO memory

Quoted 17 lines · On Thu, Feb 08, 2024 at 02:16:49AM +0530, [email protected] wrote:
arch/arm64/kvm/hyp/pgtable.c +4 -3
index c651df904fe3..2a893724ee9b 100644
@@ -717,15 +717,28 @@ void kvm_tlb_flush_vmid_range(struct kvm_s2_mmu *mmu,
 static int stage2_set_prot_attr(struct kvm_pgtable *pgt, enum kvm_pgtable_prot prot,
 				kvm_pte_t *ptep)
 {
-	bool device = prot & KVM_PGTABLE_PROT_DEVICE;
-	kvm_pte_t attr = device ? KVM_S2_MEMATTR(pgt, DEVICE_nGnRE) :
-			    KVM_S2_MEMATTR(pgt, NORMAL);
+	kvm_pte_t attr;
 	u32 sh = KVM_PTE_LEAF_ATTR_LO_S2_SH_IS;
 
+	switch (prot & (KVM_PGTABLE_PROT_DEVICE |
+			KVM_PGTABLE_PROT_NORMAL_NC)) {
+	case 0:

arch/arm64/kvm/hyp/pgtable.c void kvm_tlb_flush_vmid_range(struct kvm_s2_mmu *mmu, · line 735

Quoted 10 lines
arch/arm64/kvm/hyp/pgtable.c +10 -0
+		attr = KVM_S2_MEMATTR(pgt, NORMAL);
+		break;
+	case KVM_PGTABLE_PROT_DEVICE:
+		if (prot & KVM_PGTABLE_PROT_X)
+			return -EINVAL;
+		attr = KVM_S2_MEMATTR(pgt, DEVICE_nGnRE);
+		break;
+	case KVM_PGTABLE_PROT_NORMAL_NC:
+		attr = KVM_S2_MEMATTR(pgt, NORMAL_NC);
+		break;

Does it make sense to allow executable here as well? I don't think it's harmful but not sure there's a use-case for it either.

arch/arm64/kvm/hyp/pgtable.c void kvm_tlb_flush_vmid_range(struct kvm_s2_mmu *mmu, · line 737

Quoted 2 lines
arch/arm64/kvm/hyp/pgtable.c +2 -0
+	default:
+		WARN_ON_ONCE(1);

Return -EINVAL?

Signature · 1 line
Catalin
#3 lore raw

Re: [PATCH v6 1/4] kvm: arm64: introduce new flag for non-cacheable IO memory

Quoted 15 lines · Catalin Marinas
Quoted 14 lines · On Thu, Feb 08, 2024 at 02:16:49AM +0530, [email protected] wrote:
arch/arm64/kvm/hyp/pgtable.c +1 -3
index c651df904fe3..2a893724ee9b 100644
@@ -717,15 +717,28 @@ void kvm_tlb_flush_vmid_range(struct kvm_s2_mmu *mmu,
 static int stage2_set_prot_attr(struct kvm_pgtable *pgt, enum kvm_pgtable_prot prot,
 				kvm_pte_t *ptep)
 {
-	bool device = prot & KVM_PGTABLE_PROT_DEVICE;
-	kvm_pte_t attr = device ? KVM_S2_MEMATTR(pgt, DEVICE_nGnRE) :
-			    KVM_S2_MEMATTR(pgt, NORMAL);
+	kvm_pte_t attr;
 	u32 sh = KVM_PTE_LEAF_ATTR_LO_S2_SH_IS;
 
Quoted 16 lines

arch/arm64/kvm/hyp/pgtable.c void kvm_tlb_flush_vmid_range(struct kvm_s2_mmu *mmu, · line 735

Quoted 13 lines
arch/arm64/kvm/hyp/pgtable.c +13 -0
+	switch (prot & (KVM_PGTABLE_PROT_DEVICE |
+			KVM_PGTABLE_PROT_NORMAL_NC)) {
+	case 0:
+		attr = KVM_S2_MEMATTR(pgt, NORMAL);
+		break;
+	case KVM_PGTABLE_PROT_DEVICE:
+		if (prot & KVM_PGTABLE_PROT_X)
+			return -EINVAL;
+		attr = KVM_S2_MEMATTR(pgt, DEVICE_nGnRE);
+		break;
+	case KVM_PGTABLE_PROT_NORMAL_NC:
+		attr = KVM_S2_MEMATTR(pgt, NORMAL_NC);
+		break;

Does it make sense to allow executable here as well? I don't think it's harmful but not sure there's a use-case for it either.

Ah, we should just return EINVAL for that too.

I get that the memory attribute itself is not problematic, but since we're only using this thing for MMIO it'd be a rather massive bug in KVM... We reject attempts to do this earlier in user_mem_abort().

If, for some reason, we wanted to do Normal-NC actual memory then we would need to make sure that KVM does the appropriate cache maintenance at map / unmap.

Signature · 2 lines
Thanks,
Oliver
#4 lore raw

Re: [PATCH v6 1/4] kvm: arm64: introduce new flag for non-cacheable IO memory

Quoted 4 lines

arch/arm64/kvm/hyp/pgtable.c void kvm_tlb_flush_vmid_range(struct kvm_s2_mmu *mmu, · line 737

Quoted 2 lines
arch/arm64/kvm/hyp/pgtable.c +2 -0
+	default:
+		WARN_ON_ONCE(1);

Return -EINVAL?

Sure.

Quoted 4 lines
Quoted 4 lines
Quoted 3 lines

+   case KVM_PGTABLE_PROT_NORMAL_NC:
+           attr = KVM_S2_MEMATTR(pgt, NORMAL_NC); +           break;

Quoted 8 lines
Quoted 2 lines

Does it make sense to allow executable here as well? I don't think it's harmful but not sure there's a use-case for it either.

Ah, we should just return EINVAL for that too.

I get that the memory attribute itself is not problematic, but since we're only using this thing for MMIO it'd be a rather massive bug in KVM... We reject attempts to do this earlier in user_mem_abort().

Ack, will change to test executable and return -EINVAL in that case.

maintainer
#5 lore raw

Re: [PATCH v6 1/4] kvm: arm64: introduce new flag for non-cacheable IO memory

In reply to [email protected]
Quoted 20 lines · On Thu, Feb 08, 2024 at 02:16:49AM +0530, [email protected] wrote:
arch/arm64/kvm/hyp/pgtable.c +7 -3
index c651df904fe3..2a893724ee9b 100644
@@ -717,15 +717,28 @@ void kvm_tlb_flush_vmid_range(struct kvm_s2_mmu *mmu,
 static int stage2_set_prot_attr(struct kvm_pgtable *pgt, enum kvm_pgtable_prot prot,
 				kvm_pte_t *ptep)
 {
-	bool device = prot & KVM_PGTABLE_PROT_DEVICE;
-	kvm_pte_t attr = device ? KVM_S2_MEMATTR(pgt, DEVICE_nGnRE) :
-			    KVM_S2_MEMATTR(pgt, NORMAL);
+	kvm_pte_t attr;
 	u32 sh = KVM_PTE_LEAF_ATTR_LO_S2_SH_IS;
 
+	switch (prot & (KVM_PGTABLE_PROT_DEVICE |
+			KVM_PGTABLE_PROT_NORMAL_NC)) {
+	case 0:
+		attr = KVM_S2_MEMATTR(pgt, NORMAL);
+		break;
+	case KVM_PGTABLE_PROT_DEVICE:

arch/arm64/kvm/hyp/pgtable.c void kvm_tlb_flush_vmid_range(struct kvm_s2_mmu *mmu, · line 738

Quoted 10 lines
arch/arm64/kvm/hyp/pgtable.c +10 -0
+		if (prot & KVM_PGTABLE_PROT_X)
+			return -EINVAL;
+		attr = KVM_S2_MEMATTR(pgt, DEVICE_nGnRE);
+		break;
+	case KVM_PGTABLE_PROT_NORMAL_NC:
+		attr = KVM_S2_MEMATTR(pgt, NORMAL_NC);
+		break;
+	default:
+		WARN_ON_ONCE(1);
+	}

Cosmetic nit, but I'd find this a little easier to read if the normal case was the default (i.e. drop 'case 0') and we returned an error for DEVICE | NC.

Will

#6 lore raw

Re: [PATCH v6 1/4] kvm: arm64: introduce new flag for non-cacheable IO memory

Quoted 18 lines
Quoted 17 lines

+     switch (prot & (KVM_PGTABLE_PROT_DEVICE | +                     KVM_PGTABLE_PROT_NORMAL_NC)) { +     case 0:
+             attr = KVM_S2_MEMATTR(pgt, NORMAL); +             break;
+     case KVM_PGTABLE_PROT_DEVICE:
+             if (prot & KVM_PGTABLE_PROT_X)
+                     return -EINVAL;
+             attr = KVM_S2_MEMATTR(pgt, DEVICE_nGnRE); +             break;
+     case KVM_PGTABLE_PROT_NORMAL_NC:
+             attr = KVM_S2_MEMATTR(pgt, NORMAL_NC); +             break;
+     default:
+             WARN_ON_ONCE(1);
+     }

Quoted 3 lines

Cosmetic nit, but I'd find this a little easier to read if the normal case was the default (i.e. drop 'case 0') and we returned an error for DEVICE | NC.

Makes sense, will update the logic accordingly.

#7 lore raw

PATCH v6 2/4 mm: introduce new flag to indicate wc safe

In reply to [email protected]

From: Ankit Agrawal <[email protected]>

Generalizing S2 setting from DEVICE_nGnRE to NormalNc for non PCI devices may be problematic. E.g. GICv2 vCPU interface, which is effectively a shared peripheral, can allow a guest to affect another guest's interrupt distribution. The issue may be solved by limiting the relaxation to mappings that have a user VMA. Still there is insufficient information and uncertainity in the behavior of non PCI drivers.

Add a new flag VM_VFIO_ALLOW_WC to indicate KVM that the device is WC capable and these S2 changes can be extended to it. KVM can use this flag to activate the code.

---
 include/linux/mm.h | 14 ++++++++++++++
 1 file changed, 14 insertions(+)
include/linux/mm.h +14 -0
index f5a97dec5169..884c068a79eb 100644
@@ -391,6 +391,20 @@ extern unsigned int kobjsize(const void *objp);
 # define VM_UFFD_MINOR		VM_NONE
 #endif /* CONFIG_HAVE_ARCH_USERFAULTFD_MINOR */
 
+/*
+ * This flag is used to connect VFIO to arch specific KVM code. It
+ * indicates that the memory under this VMA is safe for use with any
+ * non-cachable memory type inside KVM. Some VFIO devices, on some
+ * platforms, are thought to be unsafe and can cause machine crashes if
+ * KVM does not lock down the memory type.
+ */
+#ifdef CONFIG_64BIT
+#define VM_VFIO_ALLOW_WC_BIT	39
+#define VM_VFIO_ALLOW_WC	BIT(VM_VFIO_ALLOW_WC_BIT)
+#else
+#define VM_VFIO_ALLOW_WC	VM_NONE
+#endif
+
 /* Bits set in the VMA until the stack is in its final location */
 #define VM_STACK_INCOMPLETE_SETUP (VM_RAND_READ | VM_SEQ_READ | VM_STACK_EARLY)
 
Signature · 1 line
2.34.1
maintainer
#8 lore raw

Re: [PATCH v6 2/4] mm: introduce new flag to indicate wc safe

Quoted 11 lines · On Thu, Feb 08, 2024 at 02:16:50AM +0530, [email protected] wrote:
include/linux/mm.h +3 -0
index f5a97dec5169..884c068a79eb 100644
@@ -391,6 +391,20 @@ extern unsigned int kobjsize(const void *objp);
 # define VM_UFFD_MINOR		VM_NONE
 #endif /* CONFIG_HAVE_ARCH_USERFAULTFD_MINOR */
 
+/*
+ * This flag is used to connect VFIO to arch specific KVM code. It
+ * indicates that the memory under this VMA is safe for use with any

include/linux/mm.h extern unsigned int kobjsize(const void *objp); · line 406

Quoted 10 lines
include/linux/mm.h +10 -0
+ * non-cachable memory type inside KVM. Some VFIO devices, on some
+ * platforms, are thought to be unsafe and can cause machine crashes if
+ * KVM does not lock down the memory type.
+ */
+#ifdef CONFIG_64BIT
+#define VM_VFIO_ALLOW_WC_BIT	39
+#define VM_VFIO_ALLOW_WC	BIT(VM_VFIO_ALLOW_WC_BIT)
+#else
+#define VM_VFIO_ALLOW_WC	VM_NONE
+#endif

Adding David Hildenbrand to this thread as well since we briefly discussed potential alternatives (not sure we came to any conclusion).

Signature · 1 line
Catalin
Reviewed-by
#9 lore raw

Re: [PATCH v6 2/4] mm: introduce new flag to indicate wc safe

Quoted 9 lines · Catalin Marinas
Quoted 8 lines · On Thu, Feb 08, 2024 at 02:16:50AM +0530, [email protected] wrote:
include/linux/mm.h +0 -0
index f5a97dec5169..884c068a79eb 100644
@@ -391,6 +391,20 @@ extern unsigned int kobjsize(const void *objp);
 # define VM_UFFD_MINOR		VM_NONE
 #endif /* CONFIG_HAVE_ARCH_USERFAULTFD_MINOR */
 
Quoted 16 lines

include/linux/mm.h extern unsigned int kobjsize(const void *objp); · line 406

Quoted 13 lines
include/linux/mm.h +13 -0
+/*
+ * This flag is used to connect VFIO to arch specific KVM code. It
+ * indicates that the memory under this VMA is safe for use with any
+ * non-cachable memory type inside KVM. Some VFIO devices, on some
+ * platforms, are thought to be unsafe and can cause machine crashes if
+ * KVM does not lock down the memory type.
+ */
+#ifdef CONFIG_64BIT
+#define VM_VFIO_ALLOW_WC_BIT	39
+#define VM_VFIO_ALLOW_WC	BIT(VM_VFIO_ALLOW_WC_BIT)
+#else
+#define VM_VFIO_ALLOW_WC	VM_NONE
+#endif

Adding David Hildenbrand to this thread as well since we briefly discussed potential alternatives (not sure we came to any conclusion).

FWIW, with my mm hat on:

But I'm interested if David has an alternative. We don't have a shortage of bits here so I'm not sure it is worth much fuss.

Jason

#10 lore raw

PATCH v6 3/4 kvm: arm64: set io memory s2 pte as normalnc for vfio pci device

In reply to [email protected]

From: Ankit Agrawal <[email protected]>

To provide VM with the ability to get device IO memory with NormalNC property, map device MMIO in KVM for ARM64 at stage2 as NormalNC. Having NormalNC S2 default puts guests in control (based on [1], "Combining stage 1 and stage 2 memory type attributes") of device MMIO regions memory mappings. The rules are summarized below:
([(S1) - stage1], [(S2) - stage 2])

S1           |  S2           | Result
NORMAL-WB    |  NORMAL-NC    | NORMAL-NC
NORMAL-WT    |  NORMAL-NC    | NORMAL-NC
NORMAL-NC    |  NORMAL-NC    | NORMAL-NC
DEVICE<attr> |  NORMAL-NC    | DEVICE<attr>

Still this cannot be generalized to non PCI devices such as GICv2. There is insufficient information and uncertainity in the behavior of non PCI driver. A driver must indicate support using the new flag VM_VFIO_ALLOW_WC.

Adapt KVM to make use of the flag VM_VFIO_ALLOW_WC as indicator to activate the S2 setting to NormalNc.

[1] section D8.5.5 of DDI0487J_a_a-profile_architecture_reference_manual.pdf

---
 arch/arm64/kvm/mmu.c | 18 ++++++++++++++----
 1 file changed, 14 insertions(+), 4 deletions(-)
arch/arm64/kvm/mmu.c +14 -4
index d14504821b79..e1e6847a793b 100644
@@ -1381,7 +1381,7 @@ static int user_mem_abort(struct kvm_vcpu *vcpu, phys_addr_t fault_ipa,
 	int ret = 0;
 	bool write_fault, writable, force_pte = false;
 	bool exec_fault, mte_allowed;
-	bool device = false;
+	bool device = false, vfio_allow_wc = false;
 	unsigned long mmu_seq;
 	struct kvm *kvm = vcpu->kvm;
 	struct kvm_mmu_memory_cache *memcache = &vcpu->arch.mmu_page_cache;
@@ -1472,6 +1472,8 @@ static int user_mem_abort(struct kvm_vcpu *vcpu, phys_addr_t fault_ipa,
 	gfn = fault_ipa >> PAGE_SHIFT;
 	mte_allowed = kvm_vma_mte_allowed(vma);
 
+	vfio_allow_wc = (vma->vm_flags & VM_VFIO_ALLOW_WC);
+
 	/* Don't use the VMA after the unlock -- it may have vanished */
 	vma = NULL;
 
@@ -1557,10 +1559,18 @@ static int user_mem_abort(struct kvm_vcpu *vcpu, phys_addr_t fault_ipa,
 	if (exec_fault)
 		prot |= KVM_PGTABLE_PROT_X;
 
-	if (device)
-		prot |= KVM_PGTABLE_PROT_DEVICE;
-	else if (cpus_have_final_cap(ARM64_HAS_CACHE_DIC))
+	if (device) {
+		/*
+		 * To provide VM with the ability to get device IO memory
+		 * with NormalNC property, map device MMIO as NormalNC in S2.
+		 */
+		if (vfio_allow_wc)
+			prot |= KVM_PGTABLE_PROT_NORMAL_NC;
+		else
+			prot |= KVM_PGTABLE_PROT_DEVICE;
+	} else if (cpus_have_final_cap(ARM64_HAS_CACHE_DIC)) {
 		prot |= KVM_PGTABLE_PROT_X;
+	}
 
 	/*
 	 * Under the premise of getting a FSC_PERM fault, we just need to relax
Signature · 1 line
2.34.1
#11 lore raw

Re: [PATCH v6 3/4] kvm: arm64: set io memory s2 pte as normalnc for vfio pci device

arch/arm64/kvm/mmu.c static int user_mem_abort(struct kvm_vcpu *vcpu, phys_addr_t fault_ipa, · line 1566

Quoted 12 lines · On Thu, Feb 08, 2024 at 02:16:51AM +0530, [email protected] wrote:
arch/arm64/kvm/mmu.c +5 -3
@@ -1557,10 +1559,18 @@ static int user_mem_abort(struct kvm_vcpu *vcpu, phys_addr_t fault_ipa,
 	if (exec_fault)
 		prot |= KVM_PGTABLE_PROT_X;
 
-	if (device)
-		prot |= KVM_PGTABLE_PROT_DEVICE;
-	else if (cpus_have_final_cap(ARM64_HAS_CACHE_DIC))
+	if (device) {
+		/*
+		 * To provide VM with the ability to get device IO memory
+		 * with NormalNC property, map device MMIO as NormalNC in S2.
+		 */

nit: the comment doesn't provide anything of value, the logic is rather straightforward here.

Quoted 13 lines
arch/arm64/kvm/mmu.c +6 -0
+		if (vfio_allow_wc)
+			prot |= KVM_PGTABLE_PROT_NORMAL_NC;
+		else
+			prot |= KVM_PGTABLE_PROT_DEVICE;
+	} else if (cpus_have_final_cap(ARM64_HAS_CACHE_DIC)) {
 		prot |= KVM_PGTABLE_PROT_X;
+	}
 
 	/*
 	 * Under the premise of getting a FSC_PERM fault, we just need to relax
Signature · 1 line
2.34.1
Signature · 2 lines
Thanks,
Oliver
maintainer Reviewed-by on v6 3/4
#12 lore raw

Re: [PATCH v6 3/4] kvm: arm64: set io memory s2 pte as normalnc for vfio pci device

In reply to [email protected]
Quoted 14 lines · On Thu, Feb 08, 2024 at 02:16:51AM +0530, [email protected] wrote:
arch/arm64/kvm/mmu.c +1 -1
index d14504821b79..e1e6847a793b 100644
@@ -1381,7 +1381,7 @@ static int user_mem_abort(struct kvm_vcpu *vcpu, phys_addr_t fault_ipa,
 	int ret = 0;
 	bool write_fault, writable, force_pte = false;
 	bool exec_fault, mte_allowed;
-	bool device = false;
+	bool device = false, vfio_allow_wc = false;
 	unsigned long mmu_seq;
 	struct kvm *kvm = vcpu->kvm;
 	struct kvm_mmu_memory_cache *memcache = &vcpu->arch.mmu_page_cache;

arch/arm64/kvm/mmu.c static int user_mem_abort(struct kvm_vcpu *vcpu, phys_addr_t fault_ipa, · line 1475

Quoted 4 lines
arch/arm64/kvm/mmu.c +1 -0
@@ -1472,6 +1472,8 @@ static int user_mem_abort(struct kvm_vcpu *vcpu, phys_addr_t fault_ipa,
 	gfn = fault_ipa >> PAGE_SHIFT;
 	mte_allowed = kvm_vma_mte_allowed(vma);
 
+	vfio_allow_wc = (vma->vm_flags & VM_VFIO_ALLOW_WC);

Nitpick: no need for brackets, '=' has a pretty low precedence.

#13 lore raw

Re: [PATCH v6 3/4] kvm: arm64: set io memory s2 pte as normalnc for vfio pci device

Quoted 7 lines

arch/arm64/kvm/mmu.c static int user_mem_abort(struct kvm_vcpu *vcpu, phys_addr_t fault_ipa, · line 1566

Quoted 4 lines
arch/arm64/kvm/mmu.c +4 -0
+		/*
+		 * To provide VM with the ability to get device IO memory
+		 * with NormalNC property, map device MMIO as NormalNC in S2.
+		 */

nit: the comment doesn't provide anything of value, the logic is rather straightforward here.

Sure, will remove it.

Quoted 6 lines
Quoted 2 lines

+     vfio_allow_wc = (vma->vm_flags & VM_VFIO_ALLOW_WC);

Nitpick: no need for brackets, '=' has a pretty low precedence.

Will change it. Thanks for the review.

#14 lore raw

PATCH v6 4/4 vfio: convey kvm that the vfio-pci device is wc safe

In reply to [email protected]

From: Ankit Agrawal <[email protected]>

The code to map the MMIO in S2 as NormalNC is enabled when conveyed that the device is WC safe using a new flag VM_VFIO_ALLOW_WC.

Make vfio-pci set the VM_VFIO_ALLOW_WC flag.

This could be extended to other devices in the future once that is deemed safe.

---
 drivers/vfio/pci/vfio_pci_core.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)
drivers/vfio/pci/vfio_pci_core.c +2 -1
index 1cbc990d42e0..c3f95ec7fc3a 100644
@@ -1863,7 +1863,8 @@ int vfio_pci_core_mmap(struct vfio_device *core_vdev, struct vm_area_struct *vma
 	 * See remap_pfn_range(), called from vfio_pci_fault() but we can't
 	 * change vm_flags within the fault handler.  Set them now.
 	 */
-	vm_flags_set(vma, VM_IO | VM_PFNMAP | VM_DONTEXPAND | VM_DONTDUMP);
+	vm_flags_set(vma, VM_VFIO_ALLOW_WC | VM_IO | VM_PFNMAP |
+			VM_DONTEXPAND | VM_DONTDUMP);
 	vma->vm_ops = &vfio_pci_mmap_ops;
 
 	return 0;
Signature · 1 line
2.34.1
maintainer Acked-by on v6 4/4
#15 lore raw

Re: [PATCH v6 4/4] vfio: convey kvm that the vfio-pci device is wc safe

+ David H

Quoted 22 lines · On Thu, Feb 08, 2024 at 02:16:52AM +0530, [email protected] wrote:

From: Ankit Agrawal <[email protected]>

The code to map the MMIO in S2 as NormalNC is enabled when conveyed that the device is WC safe using a new flag VM_VFIO_ALLOW_WC.

Make vfio-pci set the VM_VFIO_ALLOW_WC flag.

This could be extended to other devices in the future once that is deemed safe.

---
 drivers/vfio/pci/vfio_pci_core.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)
drivers/vfio/pci/vfio_pci_core.c +0 -0
index 1cbc990d42e0..c3f95ec7fc3a 100644

drivers/vfio/pci/vfio_pci_core.c int vfio_pci_core_mmap(struct vfio_device *core_vdev, struct vm_area_struct *vma · line 1870

Quoted 9 lines
drivers/vfio/pci/vfio_pci_core.c +2 -1
@@ -1863,7 +1863,8 @@ int vfio_pci_core_mmap(struct vfio_device *core_vdev, struct vm_area_struct *vma
 	 * See remap_pfn_range(), called from vfio_pci_fault() but we can't
 	 * change vm_flags within the fault handler.  Set them now.
 	 */
-	vm_flags_set(vma, VM_IO | VM_PFNMAP | VM_DONTEXPAND | VM_DONTDUMP);
+	vm_flags_set(vma, VM_VFIO_ALLOW_WC | VM_IO | VM_PFNMAP |
+			VM_DONTEXPAND | VM_DONTDUMP);
 	vma->vm_ops = &vfio_pci_mmap_ops;
 
 	return 0;
#16 lore raw

Re: [PATCH v6 4/4] vfio: convey kvm that the vfio-pci device is wc safe

In reply to [email protected]
Quoted 22 lines · On Thu, 8 Feb 2024 02:16:52 +0530 <[email protected]> wrote:

From: Ankit Agrawal <[email protected]>

The code to map the MMIO in S2 as NormalNC is enabled when conveyed that the device is WC safe using a new flag VM_VFIO_ALLOW_WC.

Make vfio-pci set the VM_VFIO_ALLOW_WC flag.

This could be extended to other devices in the future once that is deemed safe.

---
 drivers/vfio/pci/vfio_pci_core.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)
drivers/vfio/pci/vfio_pci_core.c +0 -0
index 1cbc990d42e0..c3f95ec7fc3a 100644

drivers/vfio/pci/vfio_pci_core.c int vfio_pci_core_mmap(struct vfio_device *core_vdev, struct vm_area_struct *vma · line 1870

Quoted 9 lines
drivers/vfio/pci/vfio_pci_core.c +2 -1
@@ -1863,7 +1863,8 @@ int vfio_pci_core_mmap(struct vfio_device *core_vdev, struct vm_area_struct *vma
 	 * See remap_pfn_range(), called from vfio_pci_fault() but we can't
 	 * change vm_flags within the fault handler.  Set them now.
 	 */
-	vm_flags_set(vma, VM_IO | VM_PFNMAP | VM_DONTEXPAND | VM_DONTDUMP);
+	vm_flags_set(vma, VM_VFIO_ALLOW_WC | VM_IO | VM_PFNMAP |
+			VM_DONTEXPAND | VM_DONTDUMP);
 	vma->vm_ops = &vfio_pci_mmap_ops;
 
 	return 0;

The comment above this is justifying the flags as equivalent to those set by the remap_pfn_range() path. That's no longer the case and the additional flag needs to be described there.

I'm honestly surprised that a vm_flags bit named so specifically for a single driver has gotten this far. It seems like the vfio use case for this and associated FUD for other use cases could all be encompassed in the comment where the bit is defined and we could use a name like VM_ALLOW_ANY_UNCACHED or VM_IO_ANY. Thanks,

Alex

#17 lore raw

Re: [PATCH v6 4/4] vfio: convey kvm that the vfio-pci device is wc safe

Quoted 33 lines · Alex Williamson
Quoted 30 lines · On Thu, 8 Feb 2024 02:16:52 +0530 <[email protected]> wrote:

From: Ankit Agrawal <[email protected]>

The code to map the MMIO in S2 as NormalNC is enabled when conveyed that the device is WC safe using a new flag VM_VFIO_ALLOW_WC.

Make vfio-pci set the VM_VFIO_ALLOW_WC flag.

This could be extended to other devices in the future once that is deemed safe.

---
 drivers/vfio/pci/vfio_pci_core.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)
drivers/vfio/pci/vfio_pci_core.c +2 -1
index 1cbc990d42e0..c3f95ec7fc3a 100644
@@ -1863,7 +1863,8 @@ int vfio_pci_core_mmap(struct vfio_device *core_vdev, struct vm_area_struct *vma
 	 * See remap_pfn_range(), called from vfio_pci_fault() but we can't
 	 * change vm_flags within the fault handler.  Set them now.
 	 */
-	vm_flags_set(vma, VM_IO | VM_PFNMAP | VM_DONTEXPAND | VM_DONTDUMP);
+	vm_flags_set(vma, VM_VFIO_ALLOW_WC | VM_IO | VM_PFNMAP |
+			VM_DONTEXPAND | VM_DONTDUMP);
 	vma->vm_ops = &vfio_pci_mmap_ops;
 
Quoted 8 lines

drivers/vfio/pci/vfio_pci_core.c int vfio_pci_core_mmap(struct vfio_device *core_vdev, struct vm_area_struct *vma · line 1870

Quoted 1 line
drivers/vfio/pci/vfio_pci_core.c +0 -0
 	return 0;

The comment above this is justifying the flags as equivalent to those set by the remap_pfn_range() path. That's no longer the case and the additional flag needs to be described there.

I'm honestly surprised that a vm_flags bit named so specifically for a single driver has gotten this far.

IIRC there was a small bike shed and this is what we came up with. Realistically it should not be used by anything but VFIO and KVM together. Generic names do sometimes invite abuse :)

Quoted 3 lines

It seems like the vfio use case for
this and associated FUD for other use cases could all be encompassed in

I think Ankit is talking about vfio-platform drivers by "other devices".

This is largely why it exists at all, there is a fear that the non-PCI VFIO devices will not be implemented the same as the PCI devices. If any platform devices have workloads that require WC and have HW that is safe then they will set the flag somehow in the vfio platform drivers.

Quoted 2 lines

the comment where the bit is defined and we could use a name like VM_ALLOW_ANY_UNCACHED or VM_IO_ANY. Thanks,

I'd pick VM_ALLOW_ANY_UNCACHED of those two

Thanks,
Jason

#18 lore raw

Re: [PATCH v6 4/4] vfio: convey kvm that the vfio-pci device is wc safe

Quoted 3 lines

The comment above this is justifying the flags as equivalent to those set by the remap_pfn_range() path. That's no longer the case and the additional flag needs to be described there.

Ack.

Quoted 4 lines
Quoted 2 lines

the comment where the bit is defined and we could use a name like VM_ALLOW_ANY_UNCACHED or VM_IO_ANY.  Thanks,

I'd pick VM_ALLOW_ANY_UNCACHED of those two

If there is consensus on this name, I'll make the change s/VM_VFIO_ALLOW_WC/VM_ALLOW_ANY_UNCACHED.