Skip to contentSkip to conversation

xen: arm: Split MMU code in preparation for MPU work (part 2)

PATCH Superseded Ayan Kumar Halder 5 participants 26 messages Aug 13, 2024 – Aug 20, 2024

v3 of 4 Superseded Latest is v4 →

lore
Share & tools
lore
#0 lore raw

v3 cover xen: arm: Split MMU code in preparation for MPU work (part 2)

Hi,

In https://patchew.org/Xen/[email protected]/, Henry has reorganized some of the code between the MMU specific and generic files.

In this patch serie, we address the remaining code reorg so that MMU specific code is cleanly separated and we have added stubs wherever necessary to avoid introducing if-def.

Ayan Kumar Halder (4):
xen: arm: Add a new helper update_boot_mapping()
xen: make VMAP only support in MMU system
xen: arm: Move the functions of domain_page to MMU specific
xen: arm: Enclose access to EL2 MMU specific registers under
CONFIG_MMU

 xen/arch/arm/Kconfig                 |  4 +++-
 xen/arch/arm/Makefile                |  1 -
 xen/arch/arm/arm64/mmu/mm.c          |  7 ++++++-
 xen/arch/arm/arm64/smpboot.c         |  6 +++---
 xen/arch/arm/include/asm/arm64/mm.h  |  2 +-
 xen/arch/arm/mmu/Makefile            |  1 +
 xen/arch/arm/{ => mmu}/domain_page.c |  0
 xen/arch/arm/setup.c                 |  2 ++
 xen/arch/arm/traps.c                 | 10 ++++++++++
 xen/arch/x86/Kconfig                 |  2 ++
 xen/common/Kconfig                   |  3 +++
 xen/include/xen/vmap.h               |  2 ++
 12 files changed, 33 insertions(+), 7 deletions(-)
 rename xen/arch/arm/{ => mmu}/domain_page.c (100%)
Signature · 1 line
2.25.1
#1 lore raw

PATCH v3 1/4 xen: arm: Add a new helper update_boot_mapping()

update_boot_mapping() invokes update_identity_mapping() for the MMU specific code.
Later when the MPU code is added, update_boot_mapping() would invoke the equivalent.

The common code now invokes update_boot_mapping() instead of update_identity_mapping(). So, that there is clear abstraction between the common and MMU/MPU specific logic.

This is in continuation to commit
f661a20aa880: "Extract MMU-specific MM code".

update_identity_mapping() is now marked as static as it is called within xen/arch/arm/arm64/mmu/mm.c only. Also, updated the prototype to update_boot_mapping() which is now invoked from other files.

---
Changes from :-

v1 - 1. Introduced update_boot_mapping() which invokes update_identity_mapping() in MMU specific code.

v2 - 1. Make update_identity_mapping() static and update the prototype.

 xen/arch/arm/arm64/mmu/mm.c         | 7 ++++++-
 xen/arch/arm/arm64/smpboot.c        | 6 +++---
 xen/arch/arm/include/asm/arm64/mm.h | 2 +-
 3 files changed, 10 insertions(+), 5 deletions(-)
xen/arch/arm/arm64/mmu/mm.c +6 -1
index 293acb67e0..1afbbeda5a 100644
@@ -111,7 +111,7 @@ void __init arch_setup_page_tables(void)
     prepare_runtime_identity_mapping();
 }
 
-void update_identity_mapping(bool enable)
+static void update_identity_mapping(bool enable)
 {
     paddr_t id_addr = virt_to_maddr(_start);
     int rc;
@@ -125,6 +125,11 @@ void update_identity_mapping(bool enable)
     BUG_ON(rc);
 }
 
+void update_boot_mapping(bool enable)
+{
+    update_identity_mapping(enable);
+}
+
 extern void switch_ttbr_id(uint64_t ttbr);
 
 typedef void (switch_ttbr_fn)(uint64_t ttbr);
xen/arch/arm/arm64/smpboot.c +3 -3
index a225fae64d..789f352ab6 100644
@@ -112,18 +112,18 @@ int arch_cpu_up(int cpu)
     if ( !smp_enable_ops[cpu].prepare_cpu )
         return -ENODEV;
 
-    update_identity_mapping(true);
+    update_boot_mapping(true);
 
     rc = smp_enable_ops[cpu].prepare_cpu(cpu);
     if ( rc )
-        update_identity_mapping(false);
+        update_boot_mapping(false);
 
     return rc;
 }
 
 void arch_cpu_up_finish(void)
 {
-    update_identity_mapping(false);
+    update_boot_mapping(false);
 }
 
 /*
xen/arch/arm/include/asm/arm64/mm.h +1 -1
index e0bd23a6ed..ac8d1f5c78 100644
@@ -21,7 +21,7 @@ void arch_setup_page_tables(void);
  * Note that nested call (e.g. enable=true, enable=true) is not
  * supported.
  */
-void update_identity_mapping(bool enable);
+void update_boot_mapping(bool enable);
 
 #endif /* __ARM_ARM64_MM_H__ */
 
Signature · 1 line
2.25.1
Reviewed-by on v3 1/4
#2 lore raw

Re: [PATCH v3 1/4] xen: arm: Add a new helper update_boot_mapping()

Hi Ayan,

Quoted 86 lines · Ayan Kumar Halder

update_boot_mapping() invokes update_identity_mapping() for the MMU specific code.
Later when the MPU code is added, update_boot_mapping() would invoke the equivalent.

The common code now invokes update_boot_mapping() instead of update_identity_mapping(). So, that there is clear abstraction between the common and MMU/MPU specific logic.

This is in continuation to commit
f661a20aa880: "Extract MMU-specific MM code".

update_identity_mapping() is now marked as static as it is called within xen/arch/arm/arm64/mmu/mm.c only. Also, updated the prototype to update_boot_mapping() which is now invoked from other files.

---
Changes from :-

v1 - 1. Introduced update_boot_mapping() which invokes update_identity_mapping() in MMU specific code.

v2 - 1. Make update_identity_mapping() static and update the prototype.

 xen/arch/arm/arm64/mmu/mm.c         | 7 ++++++-
 xen/arch/arm/arm64/smpboot.c        | 6 +++---
 xen/arch/arm/include/asm/arm64/mm.h | 2 +-
 3 files changed, 10 insertions(+), 5 deletions(-)
xen/arch/arm/arm64/mmu/mm.c +6 -1
index 293acb67e0..1afbbeda5a 100644
@@ -111,7 +111,7 @@ void __init arch_setup_page_tables(void)
     prepare_runtime_identity_mapping();
 }
 
-void update_identity_mapping(bool enable)
+static void update_identity_mapping(bool enable)
 {
     paddr_t id_addr = virt_to_maddr(_start);
     int rc;
@@ -125,6 +125,11 @@ void update_identity_mapping(bool enable)
     BUG_ON(rc);
 }
 
+void update_boot_mapping(bool enable)
+{
+    update_identity_mapping(enable);
+}
+
 extern void switch_ttbr_id(uint64_t ttbr);
 
 typedef void (switch_ttbr_fn)(uint64_t ttbr);
xen/arch/arm/arm64/smpboot.c +3 -3
index a225fae64d..789f352ab6 100644
@@ -112,18 +112,18 @@ int arch_cpu_up(int cpu)
     if ( !smp_enable_ops[cpu].prepare_cpu )
         return -ENODEV;
 
-    update_identity_mapping(true);
+    update_boot_mapping(true);
 
     rc = smp_enable_ops[cpu].prepare_cpu(cpu);
     if ( rc )
-        update_identity_mapping(false);
+        update_boot_mapping(false);
 
     return rc;
 }
 
 void arch_cpu_up_finish(void)
 {
-    update_identity_mapping(false);
+    update_boot_mapping(false);
 }
 
 /*
xen/arch/arm/include/asm/arm64/mm.h +0 -0
index e0bd23a6ed..ac8d1f5c78 100644

xen/arch/arm/include/asm/arm64/mm.h void arch_setup_page_tables(void); · line 24

Quoted 5 lines
xen/arch/arm/include/asm/arm64/mm.h +1 -1
@@ -21,7 +21,7 @@ void arch_setup_page_tables(void);
  * Note that nested call (e.g. enable=true, enable=true) is not
  * supported.
  */
-void update_identity_mapping(bool enable);
+void update_boot_mapping(bool enable);

The whole point of adding a wrapper was to avoid ambiguity \wrt identity mapping and MPU. You changed the name of a function but left the comment which is pretty MMU specific. I think it should be moved to update_identity_mapping.

Apart from that:

~Michal

#3 lore raw

Re: [PATCH v3 1/4] xen: arm: Add a new helper update_boot_mapping()

Quoted 1 line · Michal Orzel

Hi Ayan,

Hi Michal,

Quoted 82 lines · Ayan Kumar Halder
Quoted 80 lines

update_boot_mapping() invokes update_identity_mapping() for the MMU specific code.
Later when the MPU code is added, update_boot_mapping() would invoke the equivalent.

The common code now invokes update_boot_mapping() instead of update_identity_mapping(). So, that there is clear abstraction between the common and MMU/MPU specific logic.

This is in continuation to commit
f661a20aa880: "Extract MMU-specific MM code".

update_identity_mapping() is now marked as static as it is called within xen/arch/arm/arm64/mmu/mm.c only. Also, updated the prototype to update_boot_mapping() which is now invoked from other files.

---
Changes from :-

v1 - 1. Introduced update_boot_mapping() which invokes update_identity_mapping() in MMU specific code.

v2 - 1. Make update_identity_mapping() static and update the prototype.

  xen/arch/arm/arm64/mmu/mm.c         | 7 ++++++-
  xen/arch/arm/arm64/smpboot.c        | 6 +++---
  xen/arch/arm/include/asm/arm64/mm.h | 2 +-
  3 files changed, 10 insertions(+), 5 deletions(-)
xen/arch/arm/arm64/mmu/mm.c +6 -1
index 293acb67e0..1afbbeda5a 100644
@@ -111,7 +111,7 @@ void __init arch_setup_page_tables(void)
      prepare_runtime_identity_mapping();
  }
 
-void update_identity_mapping(bool enable)
+static void update_identity_mapping(bool enable)
  {
      paddr_t id_addr = virt_to_maddr(_start);
      int rc;
@@ -125,6 +125,11 @@ void update_identity_mapping(bool enable)
      BUG_ON(rc);
  }
 
+void update_boot_mapping(bool enable)
+{
+    update_identity_mapping(enable);
+}
+
  extern void switch_ttbr_id(uint64_t ttbr);
 
  typedef void (switch_ttbr_fn)(uint64_t ttbr);
xen/arch/arm/arm64/smpboot.c +3 -3
index a225fae64d..789f352ab6 100644
@@ -112,18 +112,18 @@ int arch_cpu_up(int cpu)
      if ( !smp_enable_ops[cpu].prepare_cpu )
          return -ENODEV;
 
-    update_identity_mapping(true);
+    update_boot_mapping(true);
 
      rc = smp_enable_ops[cpu].prepare_cpu(cpu);
      if ( rc )
-        update_identity_mapping(false);
+        update_boot_mapping(false);
 
      return rc;
  }
 
  void arch_cpu_up_finish(void)
  {
-    update_identity_mapping(false);
+    update_boot_mapping(false);
  }
 
Quoted 14 lines

xen/arch/arm/include/asm/arm64/mm.h void arch_setup_page_tables(void); · line 24

Quoted 11 lines
xen/arch/arm/arm64/smpboot.c +0 -0
  /*
xen/arch/arm/include/asm/arm64/mm.h +1 -1
index e0bd23a6ed..ac8d1f5c78 100644
@@ -21,7 +21,7 @@ void arch_setup_page_tables(void);
   * Note that nested call (e.g. enable=true, enable=true) is not
   * supported.
   */
-void update_identity_mapping(bool enable);
+void update_boot_mapping(bool enable);

The whole point of adding a wrapper was to avoid ambiguity \wrt identity mapping and MPU. You changed the name of a function but left the comment which is pretty MMU specific. I think it should be moved to update_identity_mapping.

Yes, you are correct. I will move the comment before

static void update_identity_mapping(bool enable)

{...}

Quoted 3 lines

Apart from that:

I will keep your R-b

- Ayan

#4 lore raw

PATCH v3 2/4 xen: make VMAP only support in MMU system

In reply to Ayan Kumar Halder

From: Penny Zheng <[email protected]>

Introduced CONFIG_VMAP which is selected by the architectures that use MMU. vm_init() does not do anything if CONFIG_VMAP is not enabled.

VMAP is widely used in ALTERNATIVE feature to remap a range of memory with new memory attributes. Since this is highly dependent on virtual address translation, we choose to fold VMAP in MMU system.

In this patch, we introduce a new Kconfig CONFIG_HAS_VMAP, and make it only support in MMU system on ARM architecture. And ALTERNATIVE now depends on VMAP.

HARDEN_BRANCH_PREDICTOR is now gated on HAS_VMAP as speculative
attacks are not possible on non MMU based systems (ie Cortex-R52, R82).
See https://developer.arm.com/Arm%20Security%20Center/Speculative%20Processor%20Vulnerability.

---
Changes from :-

v1 - 1. HARDEN_BRANCH_PREDICTOR is now gated on HAS_VMAP.
2. cpuerrata.c is not gated on HAS_VMAP.

v2 - 1. Introduced CONFIG_VMAP in common/Kconfig.
2. Architectures using MMU select this config.
3. vm_init() now uses CONFIG_VMAP.

 xen/arch/arm/Kconfig   | 4 +++-
 xen/arch/arm/setup.c   | 2 ++
 xen/arch/x86/Kconfig   | 2 ++
 xen/common/Kconfig     | 3 +++
 xen/include/xen/vmap.h | 2 ++
 5 files changed, 12 insertions(+), 1 deletion(-)
xen/arch/arm/Kconfig +3 -1
index 21d03d9f44..e30a7da186 100644
@@ -12,7 +12,7 @@ config ARM_64
 config ARM
 	def_bool y
 	select FUNCTION_ALIGNMENT_4B
-	select HAS_ALTERNATIVE
+	select HAS_ALTERNATIVE if HAS_VMAP
 	select HAS_DEVICE_TREE
 	select HAS_PASSTHROUGH
 	select HAS_UBSAN
@@ -61,6 +61,7 @@ config PADDR_BITS
 config MMU
 	def_bool y
 	select HAS_PMAP
+	select HAS_VMAP
 
 source "arch/Kconfig"
 
@@ -171,6 +172,7 @@ config ARM_SSBD
 
 config HARDEN_BRANCH_PREDICTOR
 	bool "Harden the branch predictor against aliasing attacks" if EXPERT
+	depends on HAS_VMAP
 	default y
 	help
 	  Speculation attacks against some high-performance processors rely on
xen/arch/arm/setup.c +2 -0
index cb2c0a16b8..7f686d2cca 100644
@@ -447,7 +447,9 @@ void asmlinkage __init start_xen(unsigned long boot_phys_offset,
      * It needs to be called after do_initcalls to be able to use
      * stop_machine (tasklets initialized via an initcall).
      */
+#ifdef CONFIG_HAS_ALTERNATIVE
     apply_alternatives_all();
+#endif
     enable_errata_workarounds();
     enable_cpu_features();
 
xen/arch/x86/Kconfig +2 -0
index 7ef5c8bc48..32be057978 100644
@@ -1,6 +1,7 @@
 config X86_64
 	def_bool y
 	select 64BIT
+	select HAS_VMAP
 
 config X86
 	def_bool y
@@ -31,6 +32,7 @@ config X86
 	select HAS_UBSAN
 	select HAS_VPCI if HVM
 	select NEEDS_LIBELF
+	select HAS_VMAP
 
 config ARCH_DEFCONFIG
 	string
xen/common/Kconfig +3 -0
index 565ceda741..188918ec5c 100644
@@ -77,6 +77,9 @@ config HAS_PIRQ
 config HAS_PMAP
 	bool
 
+config HAS_VMAP
+	bool
+
 config HAS_SCHED_GRANULARITY
 	bool
 
xen/include/xen/vmap.h +2 -0
index fdae37e950..c1dd7ac22f 100644
@@ -141,7 +141,9 @@ void *arch_vmap_virt_end(void);
 /* Initialises the VMAP_DEFAULT virtual range */
 static inline void vm_init(void)
 {
+#ifdef CONFIG_HAS_VMAP
     vm_init_type(VMAP_DEFAULT, (void *)VMAP_VIRT_START, arch_vmap_virt_end());
+#endif
 }
 
 #endif /* __XEN_VMAP_H__ */
Signature · 1 line
2.25.1
#5 lore raw

Re: [PATCH v3 2/4] xen: make VMAP only support in MMU system

Quoted 9 lines · Ayan Kumar Halder

From: Penny Zheng <[email protected]>

Introduced CONFIG_VMAP which is selected by the architectures that use MMU. vm_init() does not do anything if CONFIG_VMAP is not enabled.

VMAP is widely used in ALTERNATIVE feature to remap a range of memory with new memory attributes. Since this is highly dependent on virtual address translation, we choose to fold VMAP in MMU system.

Quoted 7 lines

In this patch, we introduce a new Kconfig CONFIG_HAS_VMAP, and make it only support in MMU system on ARM architecture. And ALTERNATIVE now depends on VMAP.

HARDEN_BRANCH_PREDICTOR is now gated on HAS_VMAP as speculative
attacks are not possible on non MMU based systems (ie Cortex-R52, R82).
See https://developer.arm.com/Arm%20Security%20Center/Speculative%20Processor%20Vulnerability.

While I'm not an Arm expert and hence I'm likely missing aspects, I question the one (Spectre-BHB) vulnerability there to be sufficient to draw a conclusion towards the usefulness of branch hardening. I would advise against encoding such a connection in the Kconfig dependencies.

Quoted 11 lines
xen/arch/x86/Kconfig +1 -0
@@ -1,6 +1,7 @@
 config X86_64
 	def_bool y
 	select 64BIT
+	select HAS_VMAP
 
 config X86
 	def_bool y

xen/arch/x86/Kconfig config X86 · line 35

Quoted 4 lines
xen/arch/x86/Kconfig +1 -0
@@ -31,6 +32,7 @@ config X86
 	select HAS_UBSAN
 	select HAS_VPCI if HVM
 	select NEEDS_LIBELF
+	select HAS_VMAP

Why in two places? Also please respect alphabetic sorting here (if this hunk is kept, which may be the more consistent approach) ...

xen/common/Kconfig config HAS_PIRQ · line 84

Quoted 11 lines
xen/common/Kconfig +3 -0
@@ -77,6 +77,9 @@ config HAS_PIRQ
 config HAS_PMAP
 	bool
 
+config HAS_VMAP
+	bool
+
 config HAS_SCHED_GRANULARITY
 	bool

... and here.

Jan

#6 lore raw

Re: [PATCH v3 2/4] xen: make VMAP only support in MMU system

Hi Jan,

Quoted 14 lines · Jan Beulich
Quoted 13 lines · Ayan Kumar Halder

From: Penny Zheng <[email protected]>

Introduced CONFIG_VMAP which is selected by the architectures that use MMU. vm_init() does not do anything if CONFIG_VMAP is not enabled.

VMAP is widely used in ALTERNATIVE feature to remap a range of memory with new memory attributes. Since this is highly dependent on virtual address translation, we choose to fold VMAP in MMU system.

In this patch, we introduce a new Kconfig CONFIG_HAS_VMAP, and make it only support in MMU system on ARM architecture. And ALTERNATIVE now depends on VMAP.

Quoted 7 lines
Quoted 3 lines

HARDEN_BRANCH_PREDICTOR is now gated on HAS_VMAP as speculative
attacks are not possible on non MMU based systems (ie Cortex-R52, R82).
See https://developer.arm.com/Arm%20Security%20Center/Speculative%20Processor%20Vulnerability.

While I'm not an Arm expert and hence I'm likely missing aspects, I question the one (Spectre-BHB) vulnerability there to be sufficient to draw a conclusion towards the usefulness of branch hardening. I would advise against encoding such a connection in the Kconfig dependencies.

AFAIU, to address 'Spectre' like vulnerabilities 'branch hardening' was added.

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

And from https://lists.linaro.org/archives/list/[email protected]/message/F4MGL4WT2R7T54NLRDGKFRQNSXF3DZGD/

Spectre is valid on MMU based systems.

Thus, I would make 'branch hardenining' valid on MMU based systems only.

Let me know your thoughts.

Quoted 9 lines
Quoted 9 lines
xen/arch/x86/Kconfig +1 -0
@@ -1,6 +1,7 @@
  config X86_64
  	def_bool y
  	select 64BIT
+	select HAS_VMAP
 
Quoted 9 lines

xen/arch/x86/Kconfig config X86 · line 35

Quoted 7 lines
xen/arch/x86/Kconfig +1 -0
  config X86
  	def_bool y
@@ -31,6 +32,7 @@ config X86
  	select HAS_UBSAN
  	select HAS_VPCI if HVM
  	select NEEDS_LIBELF
+	select HAS_VMAP

Why in two places? Also please respect alphabetic sorting here (if this hunk is kept, which may be the more consistent approach) ...

My mistake. I initially thought user could select one of the two. However, "vm_init()" is invoked only from xen/arch/x86/setup.c.

Thus, keeping 'HAS_VMAP' under 'config X86' is dufficient.

I agree that HAS_VMAP needs to go before HAS_VPCI.

Quoted 7 lines
Quoted 7 lines
xen/common/Kconfig +0 -0
@@ -77,6 +77,9 @@ config HAS_PIRQ
  config HAS_PMAP
  	bool
 
Quoted 6 lines

xen/common/Kconfig config HAS_PIRQ · line 84

Quoted 5 lines
xen/common/Kconfig +3 -0
+config HAS_VMAP
+	bool
+
  config HAS_SCHED_GRANULARITY
  	bool

... and here.

Yes, it needs to go after HAS_UBSAN.

- Ayan

Quoted 2 lines

Jan

#7 lore raw

Re: [PATCH v3 2/4] xen: make VMAP only support in MMU system

Quoted 28 lines · Ayan Kumar Halder

Hi Jan,

Quoted 14 lines · Jan Beulich
Quoted 13 lines · Ayan Kumar Halder

From: Penny Zheng <[email protected]>

Introduced CONFIG_VMAP which is selected by the architectures that use MMU. vm_init() does not do anything if CONFIG_VMAP is not enabled.

VMAP is widely used in ALTERNATIVE feature to remap a range of memory with new memory attributes. Since this is highly dependent on virtual address translation, we choose to fold VMAP in MMU system.

In this patch, we introduce a new Kconfig CONFIG_HAS_VMAP, and make it only support in MMU system on ARM architecture. And ALTERNATIVE now depends on VMAP.

Quoted 7 lines
Quoted 3 lines

HARDEN_BRANCH_PREDICTOR is now gated on HAS_VMAP as speculative
attacks are not possible on non MMU based systems (ie Cortex-R52, R82).
See https://developer.arm.com/Arm%20Security%20Center/Speculative%20Processor%20Vulnerability.

While I'm not an Arm expert and hence I'm likely missing aspects, I question the one (Spectre-BHB) vulnerability there to be sufficient to draw a conclusion towards the usefulness of branch hardening. I would advise against encoding such a connection in the Kconfig dependencies.

AFAIU, to address 'Spectre' like vulnerabilities 'branch hardening' was added.

Quoted 6 lines

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

And from
https://lists.linaro.org/archives/list/[email protected]/message/F4MGL4WT2R7T54NLRDGKFRQNSXF3DZGD/

Spectre is valid on MMU based systems.

Since then various other issues / flavors were found. I've been focusing on the x86 side of things, but I'd be very surprised if some didn't affect other architectures as well. Plus branch hardening can be a pre- cautionary measure, too, I think.

Jan

#8 lore raw

Re: [PATCH v3 2/4] xen: make VMAP only support in MMU system

Hi Jan,

Quoted 30 lines · Jan Beulich
Quoted 29 lines · Ayan Kumar Halder

Hi Jan,

Quoted 14 lines · Jan Beulich
Quoted 13 lines · Ayan Kumar Halder

From: Penny Zheng <[email protected]>

Introduced CONFIG_VMAP which is selected by the architectures that use MMU. vm_init() does not do anything if CONFIG_VMAP is not enabled.

VMAP is widely used in ALTERNATIVE feature to remap a range of memory with new memory attributes. Since this is highly dependent on virtual address translation, we choose to fold VMAP in MMU system.

In this patch, we introduce a new Kconfig CONFIG_HAS_VMAP, and make it only support in MMU system on ARM architecture. And ALTERNATIVE now depends on VMAP.

Quoted 7 lines
Quoted 3 lines

HARDEN_BRANCH_PREDICTOR is now gated on HAS_VMAP as speculative
attacks are not possible on non MMU based systems (ie Cortex-R52, R82).
See https://developer.arm.com/Arm%20Security%20Center/Speculative%20Processor%20Vulnerability.

While I'm not an Arm expert and hence I'm likely missing aspects, I question the one (Spectre-BHB) vulnerability there to be sufficient to draw a conclusion towards the usefulness of branch hardening. I would advise against encoding such a connection in the Kconfig dependencies.

AFAIU, to address 'Spectre' like vulnerabilities 'branch hardening' was added.

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

Quoted 7 lines
Quoted 4 lines

And from
https://lists.linaro.org/archives/list/[email protected]/message/F4MGL4WT2R7T54NLRDGKFRQNSXF3DZGD/

Spectre is valid on MMU based systems.

Since then various other issues / flavors were found. I've been focusing on the x86 side of things, but I'd be very surprised if some didn't affect other architectures as well.

We are talking about Arm here as "HARDEN_BRANCH_PREDICTOR" is specific to Arm.

https://developer.arm.com/Arm%20Security%20Center/Speculative%20Processor%20Vulnerability covers all the flavours and it does not include Cortex-R82 or R52.

It says the following :-

"Cortex-R cores typically use a closed software stack. In those environments, applications or processes are strictly controlled, and therefore not exploitable"

Quoted 2 lines

Plus branch hardening can be a pre-
cautionary measure, too, I think.

The first two Arm non MMU cores that we wish to support in the forthcoming series is Cortex-R82 and R52.

As seen in https://developer.arm.com/documentation/ka005109/latest/, it explicitly states the following about R82

The Cortex-R82 implements the faulting feature (FEAT_FPAC) but is not vulnerable. The Cortex-R82 behaves differently than vulnerable A-class CPUs when speculatively executing past an instruction that authenticates PAC, and that behavior does not allow the attack software to create the "oracle".

We can re-enable branch hardening if we know there is a valid non MMU Arm core which is vulnerable.

Let me know if you are ok with the rationale.

- Ayan

Quoted 2 lines

Jan

#9 lore raw

Re: [PATCH v3 2/4] xen: make VMAP only support in MMU system

Quoted 64 lines · Ayan Kumar Halder
Quoted 28 lines · Jan Beulich
Quoted 27 lines · Ayan Kumar Halder
Quoted 14 lines · Jan Beulich
Quoted 13 lines · Ayan Kumar Halder

From: Penny Zheng <[email protected]>

Introduced CONFIG_VMAP which is selected by the architectures that use MMU. vm_init() does not do anything if CONFIG_VMAP is not enabled.

VMAP is widely used in ALTERNATIVE feature to remap a range of memory with new memory attributes. Since this is highly dependent on virtual address translation, we choose to fold VMAP in MMU system.

In this patch, we introduce a new Kconfig CONFIG_HAS_VMAP, and make it only support in MMU system on ARM architecture. And ALTERNATIVE now depends on VMAP.

Quoted 7 lines
Quoted 3 lines

HARDEN_BRANCH_PREDICTOR is now gated on HAS_VMAP as speculative
attacks are not possible on non MMU based systems (ie Cortex-R52, R82).
See https://developer.arm.com/Arm%20Security%20Center/Speculative%20Processor%20Vulnerability.

While I'm not an Arm expert and hence I'm likely missing aspects, I question the one (Spectre-BHB) vulnerability there to be sufficient to draw a conclusion towards the usefulness of branch hardening. I would advise against encoding such a connection in the Kconfig dependencies.

AFAIU, to address 'Spectre' like vulnerabilities 'branch hardening' was added.

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

Quoted 7 lines
Quoted 4 lines

And from
https://lists.linaro.org/archives/list/[email protected]/message/F4MGL4WT2R7T54NLRDGKFRQNSXF3DZGD/

Spectre is valid on MMU based systems.

Since then various other issues / flavors were found. I've been focusing on the x86 side of things, but I'd be very surprised if some didn't affect other architectures as well.

We are talking about Arm here as "HARDEN_BRANCH_PREDICTOR" is specific to Arm.

https://developer.arm.com/Arm%20Security%20Center/Speculative%20Processor%20Vulnerability
covers all the flavours and it does not include Cortex-R82 or R52.

It says the following :-

"Cortex-R cores typically use a closed software stack. In those environments, applications or processes are strictly controlled, and therefore not exploitable"

Quoted 2 lines

Plus branch hardening can be a pre-
cautionary measure, too, I think.

The first two Arm non MMU cores that we wish to support in the forthcoming series is Cortex-R82 and R52.

As seen in https://developer.arm.com/documentation/ka005109/latest/, it explicitly states the following about R82

The Cortex-R82 implements the faulting feature (FEAT_FPAC) but is not vulnerable. The Cortex-R82 behaves differently than vulnerable A-class CPUs when speculatively executing past an instruction that authenticates PAC, and that behavior does not allow the attack software to create the "oracle".

Quoted 4 lines

We can re-enable branch hardening if we know there is a valid non MMU Arm core which is vulnerable.

Let me know if you are ok with the rationale.

I'm okay with any rationale. As indicated, I'm likely missing (large) parts of the picture here. The connection between the two Kconfig settings merely seemed questionable to me, hence the comment.

Jan

#10 lore raw

Re: [PATCH v3 2/4] xen: make VMAP only support in MMU system

In reply to Ayan Kumar Halder

Hi Ayan,

Quoted 68 lines · Ayan Kumar Halder

Hi Jan,

Quoted 30 lines · Jan Beulich
Quoted 29 lines · Ayan Kumar Halder

Hi Jan,

Quoted 14 lines · Jan Beulich
Quoted 13 lines · Ayan Kumar Halder

From: Penny Zheng <[email protected]>

Introduced CONFIG_VMAP which is selected by the architectures that use MMU. vm_init() does not do anything if CONFIG_VMAP is not enabled.

VMAP is widely used in ALTERNATIVE feature to remap a range of memory with new memory attributes. Since this is highly dependent on virtual address translation, we choose to fold VMAP in MMU system.

In this patch, we introduce a new Kconfig CONFIG_HAS_VMAP, and make it only support in MMU system on ARM architecture. And ALTERNATIVE now depends on VMAP.

Quoted 7 lines
Quoted 3 lines

HARDEN_BRANCH_PREDICTOR is now gated on HAS_VMAP as speculative
attacks are not possible on non MMU based systems (ie Cortex-R52, R82).
See https://developer.arm.com/Arm%20Security%20Center/Speculative%20Processor%20Vulnerability.

While I'm not an Arm expert and hence I'm likely missing aspects, I question the one (Spectre-BHB) vulnerability there to be sufficient to draw a conclusion towards the usefulness of branch hardening. I would advise against encoding such a connection in the Kconfig dependencies.

AFAIU, to address 'Spectre' like vulnerabilities 'branch hardening' was added.

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

Quoted 7 lines
Quoted 4 lines

And from
https://lists.linaro.org/archives/list/[email protected]/message/F4MGL4WT2R7T54NLRDGKFRQNSXF3DZGD/

Spectre is valid on MMU based systems.

Since then various other issues / flavors were found. I've been focusing on the x86 side of things, but I'd be very surprised if some didn't affect other architectures as well.

We are talking about Arm here as "HARDEN_BRANCH_PREDICTOR" is specific to Arm.

https://developer.arm.com/Arm%20Security%20Center/Speculative%20Processor%20Vulnerability
covers all the flavours and it does not include Cortex-R82 or R52.

It says the following :-

"Cortex-R cores typically use a closed software stack. In those environments, applications or processes are strictly controlled, and therefore not exploitable"

Quoted 2 lines

Plus branch hardening can be a pre-
cautionary measure, too, I think.

The first two Arm non MMU cores that we wish to support in the forthcoming series is Cortex-R82 and R52.

As seen in https://developer.arm.com/documentation/ka005109/latest/, it explicitly states the following about R82

The Cortex-R82 implements the faulting feature (FEAT_FPAC) but is not vulnerable. The Cortex-R82 behaves differently than vulnerable A-class CPUs when speculatively executing past an instruction that authenticates PAC, and that behavior does not allow the attack software to create the "oracle".

Quoted 4 lines

We can re-enable branch hardening if we know there is a valid non MMU Arm core which is vulnerable.

Let me know if you are ok with the rationale.

I'm ok with your rationale.

I have one question for this patch. Why can't we use CONFIG_HAS_VMAP to conditionally compile vmap.c, like:
obj-$(CONFIG_HAS_VMAP) += vmap.o
and get rid of VMAP_VIRT_START guard on an entire file?
With this config in place, it seems strange to use VMAP_VIRT_START as a guard.

~Michal

#11 lore raw

Re: [PATCH v3 2/4] xen: make VMAP only support in MMU system

Quoted 1 line · Michal Orzel

Hi Ayan,

Hi Michal,

Quoted 71 lines · Ayan Kumar Halder
Quoted 69 lines

Hi Jan,

Quoted 30 lines · Jan Beulich
Quoted 29 lines · Ayan Kumar Halder

Hi Jan,

Quoted 14 lines · Jan Beulich
Quoted 13 lines · Ayan Kumar Halder

From: Penny Zheng <[email protected]>

Introduced CONFIG_VMAP which is selected by the architectures that use MMU. vm_init() does not do anything if CONFIG_VMAP is not enabled.

VMAP is widely used in ALTERNATIVE feature to remap a range of memory with new memory attributes. Since this is highly dependent on virtual address translation, we choose to fold VMAP in MMU system.

In this patch, we introduce a new Kconfig CONFIG_HAS_VMAP, and make it only support in MMU system on ARM architecture. And ALTERNATIVE now depends on VMAP.

Quoted 7 lines
Quoted 3 lines

HARDEN_BRANCH_PREDICTOR is now gated on HAS_VMAP as speculative
attacks are not possible on non MMU based systems (ie Cortex-R52, R82).
See https://developer.arm.com/Arm%20Security%20Center/Speculative%20Processor%20Vulnerability.

While I'm not an Arm expert and hence I'm likely missing aspects, I question the one (Spectre-BHB) vulnerability there to be sufficient to draw a conclusion towards the usefulness of branch hardening. I would advise against encoding such a connection in the Kconfig dependencies.

AFAIU, to address 'Spectre' like vulnerabilities 'branch hardening' was added.

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

Quoted 7 lines
Quoted 4 lines

And from
https://lists.linaro.org/archives/list/[email protected]/message/F4MGL4WT2R7T54NLRDGKFRQNSXF3DZGD/

Spectre is valid on MMU based systems.

Since then various other issues / flavors were found. I've been focusing on the x86 side of things, but I'd be very surprised if some didn't affect other architectures as well.

We are talking about Arm here as "HARDEN_BRANCH_PREDICTOR" is specific to Arm.

https://developer.arm.com/Arm%20Security%20Center/Speculative%20Processor%20Vulnerability
covers all the flavours and it does not include Cortex-R82 or R52.

It says the following :-

"Cortex-R cores typically use a closed software stack. In those environments, applications or processes are strictly controlled, and therefore not exploitable"

Quoted 2 lines

Plus branch hardening can be a pre-
cautionary measure, too, I think.

The first two Arm non MMU cores that we wish to support in the forthcoming series is Cortex-R82 and R52.

As seen in https://developer.arm.com/documentation/ka005109/latest/, it explicitly states the following about R82

The Cortex-R82 implements the faulting feature (FEAT_FPAC) but is not vulnerable. The Cortex-R82 behaves differently than vulnerable A-class CPUs when speculatively executing past an instruction that authenticates PAC, and that behavior does not allow the attack software to create the "oracle".

We can re-enable branch hardening if we know there is a valid non MMU Arm core which is vulnerable.

Quoted 7 lines
Quoted 1 line

Let me know if you are ok with the rationale.

I'm ok with your rationale.

I have one question for this patch. Why can't we use CONFIG_HAS_VMAP to conditionally compile vmap.c, like:
obj-$(CONFIG_HAS_VMAP) += vmap.o
and get rid of VMAP_VIRT_START guard on an entire file?
With this config in place, it seems strange to use VMAP_VIRT_START as a guard.

I am fine with your suggestion.

However, this implies HAS_VMAP needs to be selected by "config PPC" as well (as it defines VMAP_VIRT_START in xen/arch/ppc/include/asm/config.h ).

This is fine as I could see PPC uses MMU as well.

- Ayan

#12 lore raw

Re: [PATCH v3 2/4] xen: make VMAP only support in MMU system

In reply to Ayan Kumar Halder

Hi Ayan,

Quoted 49 lines · Ayan Kumar Halder

Hi Jan,

Quoted 30 lines · Jan Beulich
Quoted 29 lines · Ayan Kumar Halder

Hi Jan,

Quoted 14 lines · Jan Beulich
Quoted 13 lines · Ayan Kumar Halder

From: Penny Zheng <[email protected]>

Introduced CONFIG_VMAP which is selected by the architectures that use MMU. vm_init() does not do anything if CONFIG_VMAP is not enabled.

VMAP is widely used in ALTERNATIVE feature to remap a range of memory with new memory attributes. Since this is highly dependent on virtual address translation, we choose to fold VMAP in MMU system.

In this patch, we introduce a new Kconfig CONFIG_HAS_VMAP, and make it only support in MMU system on ARM architecture. And ALTERNATIVE now depends on VMAP.

Quoted 7 lines
Quoted 3 lines

HARDEN_BRANCH_PREDICTOR is now gated on HAS_VMAP as speculative
attacks are not possible on non MMU based systems (ie Cortex-R52, R82).
See https://developer.arm.com/Arm%20Security%20Center/ Speculative%20Processor%20Vulnerability.

While I'm not an Arm expert and hence I'm likely missing aspects, I question the one (Spectre-BHB) vulnerability there to be sufficient to draw a conclusion towards the usefulness of branch hardening. I would advise against encoding such a connection in the Kconfig dependencies.

AFAIU, to address 'Spectre' like vulnerabilities 'branch hardening' was added.

See https://lore.kernel.org/all/E1fNadD-0000fz-9r@rmk- PC.armlinux.org.uk/

Quoted 7 lines
Quoted 4 lines

And from
https://lists.linaro.org/archives/list/linux-stable- [email protected]/message/F4MGL4WT2R7T54NLRDGKFRQNSXF3DZGD/

Spectre is valid on MMU based systems.

Since then various other issues / flavors were found. I've been focusing on the x86 side of things, but I'd be very surprised if some didn't affect other architectures as well.

We are talking about Arm here as "HARDEN_BRANCH_PREDICTOR" is specific to Arm.

https://developer.arm.com/Arm%20Security%20Center/ Speculative%20Processor%20Vulnerability covers all the flavours and it does not include Cortex-R82 or R52.

It says the following :-

"Cortex-R cores typically use a closed software stack. In those environments, applications or processes are strictly controlled, and therefore not exploitable"

Quoted 8 lines
Quoted 2 lines

Plus branch hardening can be a pre-
cautionary measure, too, I think.

The first two Arm non MMU cores that we wish to support in the forthcoming series is Cortex-R82 and R52.

As seen in https://developer.arm.com/documentation/ka005109/latest/, it explicitly states the following about R82

The Cortex-R82 implements the faulting feature (FEAT_FPAC) but is not vulnerable. The Cortex-R82 behaves differently than vulnerable A-class CPUs when speculatively executing past an instruction that authenticates PAC, and that behavior does not allow the attack software to create the "oracle".

I am confused. This is describing why R82 is not affected by PACMAN. But the Kconfig HARDEN_BRANCH_PREDICTOR is not for that (Xen doesn't yet use Pointer Authentification Codes). The Kconfig was introduced with XSA-254 which predates PACMAN by nearly 4 years.

Quoted 2 lines

We can re-enable branch hardening if we know there is a valid non MMU Arm core which is vulnerable.

Re-quoting what you wrote earlier:

"Cortex-R cores typically use a closed software stack. In those environments, applications or processes are strictly controlled, and therefore not exploitable"

It is quite subtle. This wording doesn't imply the cores are not vulnerable. It says that if they are, then it would be difficult to exploit because the software should be tightly controlled.

Now this would be really up to the user to decide whether they want to be extra cautious/futureproof or not.

As we are at the beginning of the MPU support, then I don't think we need to resolve issue right now. And it would be fine to gate it. But maybe ARCH_VMAP was an incorrect suggestion. It might be better to gate with the !MMU (IIRC this would imply MPU).

But before the feature can be marked as security support. We will need to agree on how the hypervisor is intended to be used on ARMv8-R. Maybe it would need a caveat "only trusted software can be run" which means we don't have to worry about speculation on Cortex-R. Although, it would be nice to have some defense in-depth :).

Cheers,

Signature · 1 line
Julien Grall
#13 lore raw

Re: [PATCH v3 2/4] xen: make VMAP only support in MMU system

Quoted 1 line · Julien Grall

Hi Ayan,

Hi Julien,

Quoted 54 lines · Ayan Kumar Halder
Quoted 52 lines

Hi Jan,

Quoted 30 lines · Jan Beulich
Quoted 29 lines · Ayan Kumar Halder

Hi Jan,

Quoted 14 lines · Jan Beulich
Quoted 13 lines · Ayan Kumar Halder

From: Penny Zheng <[email protected]>

Introduced CONFIG_VMAP which is selected by the architectures that use MMU. vm_init() does not do anything if CONFIG_VMAP is not enabled.

VMAP is widely used in ALTERNATIVE feature to remap a range of memory with new memory attributes. Since this is highly dependent on virtual address translation, we choose to fold VMAP in MMU system.

In this patch, we introduce a new Kconfig CONFIG_HAS_VMAP, and make it only support in MMU system on ARM architecture. And ALTERNATIVE now depends on VMAP.

Quoted 7 lines
Quoted 3 lines

HARDEN_BRANCH_PREDICTOR is now gated on HAS_VMAP as speculative
attacks are not possible on non MMU based systems (ie Cortex-R52, R82).
See https://developer.arm.com/Arm%20Security%20Center/ Speculative%20Processor%20Vulnerability.

While I'm not an Arm expert and hence I'm likely missing aspects, I question the one (Spectre-BHB) vulnerability there to be sufficient to draw a conclusion towards the usefulness of branch hardening. I would advise against encoding such a connection in the Kconfig dependencies.

AFAIU, to address 'Spectre' like vulnerabilities 'branch hardening' was added.

See https://lore.kernel.org/all/E1fNadD-0000fz-9r@rmk- PC.armlinux.org.uk/

Quoted 7 lines
Quoted 4 lines

And from
https://lists.linaro.org/archives/list/linux-stable- [email protected]/message/F4MGL4WT2R7T54NLRDGKFRQNSXF3DZGD/

Spectre is valid on MMU based systems.

Since then various other issues / flavors were found. I've been focusing on the x86 side of things, but I'd be very surprised if some didn't affect other architectures as well.

We are talking about Arm here as "HARDEN_BRANCH_PREDICTOR" is specific to Arm.

https://developer.arm.com/Arm%20Security%20Center/ Speculative%20Processor%20Vulnerability covers all the flavours and it does not include Cortex-R82 or R52.

It says the following :-

"Cortex-R cores typically use a closed software stack. In those environments, applications or processes are strictly controlled, and therefore not exploitable"

Quoted 2 lines

Plus branch hardening can be a pre-
cautionary measure, too, I think.

Quoted 7 lines
Quoted 5 lines

The first two Arm non MMU cores that we wish to support in the forthcoming series is Cortex-R82 and R52.

As seen in https://developer.arm.com/documentation/ka005109/latest/, it explicitly states the following about R82

The Cortex-R82 implements the faulting feature (FEAT_FPAC) but is not vulnerable. The Cortex-R82 behaves differently than vulnerable A-class CPUs when speculatively executing past an instruction that authenticates PAC, and that behavior does not allow the attack software to create the "oracle".

I am confused. This is describing why R82 is not affected by PACMAN. But the Kconfig HARDEN_BRANCH_PREDICTOR is not for that (Xen doesn't yet use Pointer Authentification Codes). The Kconfig was introduced with XSA-254 which predates PACMAN by nearly 4 years.

Yes, you are correct. I was somehow linking PACMAN to branch prediction attacks. So, by mistake I assumed that HARDEN_BRANCH_PREDICTOR is used for protection against these attacks.

Quoted 10 lines
Quoted 3 lines

We can re-enable branch hardening if we know there is a valid non MMU Arm core which is vulnerable.

Re-quoting what you wrote earlier:

"Cortex-R cores typically use a closed software stack. In those environments, applications or processes are strictly controlled, and therefore not exploitable"

Quoted 5 lines

It is quite subtle. This wording doesn't imply the cores are not vulnerable. It says that if they are, then it would be difficult to exploit because the software should be tightly controlled.

Now this would be really up to the user to decide whether they want to be extra cautious/futureproof or not.

As we are at the beginning of the MPU support, then I don't think we need to resolve issue right now. And it would be fine to gate it. But maybe ARCH_VMAP was an incorrect suggestion. It might be better to gate with the !MMU (IIRC this would imply MPU).

I am ok with this. This has the benefit that the change can be contained within arch/arm if we do the following :-

xen/arch/arm/setup.c +0 -0
index cb2c0a16b8..26f7406278 100644
@@ -329,7 +329,9 @@ void asmlinkage __init start_xen(unsigned long boot_phys_offset,
 

     setup_mm();

+#ifdef CONFIG_MMU
     vm_init();
+#endif

     /* Parse the ACPI tables for possible boot-time configuration */ acpi_boot_table_init();

Are we ok with this ?

The definition of vm_init() is in xen/include/xen/vmap.h. If I enclose it using any CONFIG_XXX (like I have done in the current patch), then I need to introduce it in common/Kconfig and define it for x86 and PPC. I would prefer to contain the change within arch/arm only if possible.

Quoted 2 lines

But before the feature can be marked as security support. We will need to agree on how the hypervisor is intended to be used on ARMv8-R. Maybe it would need a caveat "only trusted software can be run" which means we don't have to worry about speculation on Cortex-R. Although, it would be nice to have some defense in-depth :).

Yes, the support for ARMv8-R will be 'experimental' for some time. I am planning to add the minimal support required to just boot one or more DomUs.

- Ayan

#14 lore raw

Re: [PATCH v3 2/4] xen: make VMAP only support in MMU system

Hi Ayan,

Quoted 18 lines · Ayan Kumar Halder

I am ok with this. This has the benefit that the change can be contained within arch/arm if we do the following :-

xen/arch/arm/setup.c +2 -0
index cb2c0a16b8..26f7406278 100644
@@ -329,7 +329,9 @@ void asmlinkage __init start_xen(unsigned long 
 boot_phys_offset,
 
      setup_mm();
 
+#ifdef CONFIG_MMU
      vm_init();
+#endif
 
      /* Parse the ACPI tables for possible boot-time configuration */

     acpi_boot_table_init();

Quoted 3 lines

Are we ok with this ?

The definition of vm_init() is in xen/include/xen/vmap.h. If I enclose it using any CONFIG_XXX (like I have done in the current patch), then I need to introduce it in common/Kconfig and define it for x86 and PPC. I would prefer to contain the change within arch/arm only if possible.

Just to clarify, are you suggesting to just protect the call vm_init(). In other word, common/vmap.c would still be included in the final binary for the MPU?

If yes, then I think it would be a bit odd... Someone could still call vmap() and this would not break until runtime.

So I don't see how we could get away from modifying the common code.

Cheers,

Signature · 1 line
Julien Grall
#15 lore raw

Re: [PATCH v3 2/4] xen: make VMAP only support in MMU system

Quoted 23 lines · Julien Grall

Hi Ayan,

Quoted 20 lines · Ayan Kumar Halder

I am ok with this. This has the benefit that the change can be contained within arch/arm if we do the following :-

xen/arch/arm/setup.c +2 -0
index cb2c0a16b8..26f7406278 100644
@@ -329,7 +329,9 @@ void asmlinkage __init start_xen(unsigned long 
 boot_phys_offset,
 
       setup_mm();
 
+#ifdef CONFIG_MMU
       vm_init();
+#endif
 
       /* Parse the ACPI tables for possible boot-time configuration */

      acpi_boot_table_init();

Are we ok with this ?

Quoted 7 lines
Quoted 1 line

The definition of vm_init() is in xen/include/xen/vmap.h. If I enclose it using any CONFIG_XXX (like I have done in the current patch), then I need to introduce it in common/Kconfig and define it for x86 and PPC. I would prefer to contain the change within arch/arm only if possible.

Just to clarify, are you suggesting to just protect the call vm_init(). In other word, common/vmap.c would still be included in the final binary for the MPU?

If yes, then I think it would be a bit odd... Someone could still call vmap() and this would not break until runtime.

So I don't see how we could get away from modifying the common code.

Readying my previous reply again. I think the confusion comes from:

Quoted 1 line

But maybe ARCH_VMAP was an incorrect suggestion. It might be better to gate with the !MMU (IIRC this would imply MPU).

This was specifically referring to the branch predictor Kconfig. This was not a suggestion to avoid introducing ARCH_VMAP.

Cheers,

Signature · 1 line
Julien Grall
#16 lore raw

Re: [PATCH v3 2/4] xen: make VMAP only support in MMU system

Hi Julien/Jan,

Quoted 32 lines · Julien Grall
Quoted 29 lines · Julien Grall

Hi Ayan,

Quoted 18 lines · Ayan Kumar Halder

I am ok with this. This has the benefit that the change can be contained within arch/arm if we do the following :-

xen/arch/arm/setup.c +2 -0
index cb2c0a16b8..26f7406278 100644
@@ -329,7 +329,9 @@ void asmlinkage __init start_xen(unsigned long 
 boot_phys_offset,
 
       setup_mm();
 
+#ifdef CONFIG_MMU
       vm_init();
+#endif
 
       /* Parse the ACPI tables for possible boot-time configuration */

      acpi_boot_table_init();

Quoted 3 lines

Are we ok with this ?

The definition of vm_init() is in xen/include/xen/vmap.h. If I enclose it using any CONFIG_XXX (like I have done in the current patch), then I need to introduce it in common/Kconfig and define it for x86 and PPC. I would prefer to contain the change within arch/arm only if possible.

Just to clarify, are you suggesting to just protect the call vm_init(). In other word, common/vmap.c would still be included in the final binary for the MPU?

If yes, then I think it would be a bit odd... Someone could still call vmap() and this would not break until runtime.

Quoted 8 lines
Quoted 1 line

So I don't see how we could get away from modifying the common code.

Readying my previous reply again. I think the confusion comes from:

Quoted 1 line

But maybe ARCH_VMAP was an incorrect suggestion. It might be better

to gate with the !MMU (IIRC this would imply MPU).

This was specifically referring to the branch predictor Kconfig. This was not a suggestion to avoid introducing ARCH_VMAP.

Thanks for clarifying this. Yes, I misunderstood your previous comment. So I will do :-

1. HARDEN_BRANCH_PREDICTOR will depend on MMU.

2. ARCH_VMAP will be selected by PPC and RISCV. The reason is below.

3. xen/common/vmap.c will be conditionally compiled on ARCH_VMAP and the "#ifdef VMAP_VIRT_START .. endif" will be from removed within the file. As VMAP_VIRT_START is defined by RISCV and PPC, thus #2 is needed.

Julien, Jan :- Please let me know if you are ok with #3. This was in response to Michal's comment. While his suggestion makes sense, I am not sure if extending the changes to other architectures is the correct approach. Or do you prefer keeping xen/common/vmap.c unchanged.

Kind regards,

Ayan

Quoted 3 lines

Cheers,

#17 lore raw

Re: [PATCH v3 2/4] xen: make VMAP only support in MMU system

Quoted 6 lines · Ayan Kumar Halder

So I will do :-

1. HARDEN_BRANCH_PREDICTOR will depend on MMU.

2. ARCH_VMAP will be selected by PPC and RISCV. The reason is below.

Quoted 8 lines

3. xen/common/vmap.c will be conditionally compiled on ARCH_VMAP and the "#ifdef VMAP_VIRT_START .. endif" will be from removed within the file. As VMAP_VIRT_START is defined by RISCV and PPC, thus #2 is needed.

Julien, Jan :- Please let me know if you are ok with #3. This was in response to Michal's comment. While his suggestion makes sense, I am not sure if extending the changes to other architectures is the correct approach. Or do you prefer keeping xen/common/vmap.c unchanged.

No, the VMAP_VIRT_START thing was, from all I recall, merely to cover for the lack of a way to exclude building of the file from Makefile. After all this pre-dates the introduction of kconfig in Xen.

And yes, suitably covering PPC and RISC-V is The Right Thing To Do (tm); their maintainers can shout if they disagree.

Jan

#18 lore raw

Re: [PATCH v3 2/4] xen: make VMAP only support in MMU system

In reply to Ayan Kumar Halder

Guys,

Quoted 2 lines · Ayan Kumar Halder
Quoted 1 line · Julien Grall

On 14/08/2024 13:33, Ayan Kumar Halder wrote:

mind me asking why I continue to be on the To: list of this communication between the two of you?

Jan

#19 lore raw

Re: [PATCH v3 2/4] xen: make VMAP only support in MMU system

Quoted 8 lines · Jan Beulich

Guys,

Quoted 2 lines · Ayan Kumar Halder
Quoted 1 line · Julien Grall

On 14/08/2024 13:33, Ayan Kumar Halder wrote:

mind me asking why I continue to be on the To: list of this communication between the two of you?

You were involved in the review and AFAICT will still touch the common code. I am happy to remove you from the conversation if you are not interested :).

Cheers,

Signature · 1 line
Julien Grall
#20 lore raw

Re: [PATCH v3 2/4] xen: make VMAP only support in MMU system

Quoted 5 lines · Julien Grall
Quoted 4 lines · Jan Beulich

Guys,

Quoted 1 line · Ayan Kumar Halder

On 16/08/2024 17:40, Julien Grall wrote:

Quoted 8 lines
Quoted 4 lines
Quoted 1 line
Quoted 1 line · Ayan Kumar Halder

mind me asking why I continue to be on the To: list of this communication between the two of you?

You were involved in the review and AFAICT will still touch the common code. I am happy to remove you from the conversation if you are not interested :).

That wasn't my request though. It's fine to keep me Cc-ed.

Jan

#21 lore raw

Re: [PATCH v3 2/4] xen: make VMAP only support in MMU system

Hi Jan,

Quoted 8 lines · Jan Beulich
Quoted 7 lines · Julien Grall
Quoted 6 lines · Jan Beulich

Guys,

Quoted 2 lines · Ayan Kumar Halder
Quoted 1 line · Julien Grall

On 14/08/2024 13:33, Ayan Kumar Halder wrote:

Quoted 8 lines
Quoted 6 lines
Quoted 2 lines

mind me asking why I continue to be on the To: list of this communication between the two of you?

You were involved in the review and AFAICT will still touch the common code. I am happy to remove you from the conversation if you are not interested :).

That wasn't my request though. It's fine to keep me Cc-ed.

TBH, in this context, the difference between CC and TO is very blur.

For me, CC is for people that may be interested but not involved in the conversation. In this case, Ayan was suggesting whether we should use ARCH_VMAP or not.

Given you reviewed the rest, I feel you have your say on this decision. Hence why the TO makes sense. But I also understand why people may view it differently...

Personally, I don't make any filter on "cc" vs "to" because this is a matter of opinion whether you think someone should be actively involved (or not) in the discussion. So it is best if I look at everything even if this is just having a glancing look at on the content.

If you have some clear and unambiguous rules on when you want to be in the CC or TO, then I would be happy to follow them.

Cheers,

Signature · 1 line
Julien Grall
#22 lore raw

PATCH v3 3/4 xen: arm: Move the functions of domain_page to MMU specific

In reply to Ayan Kumar Halder

Moved init_domheap_mappings(), map_domain_page_global(), unmap_domain_page_global(), map_domain_page(), unmap_domain_page(), domain_page_map_to_mfn() to MMU specific folder.

---
Changes from :-

v1 - Moved domain_page.c to mmu/domain_page.c.

v2 - Updated arm/Makefile.

 xen/arch/arm/Makefile                | 1 -
 xen/arch/arm/mmu/Makefile            | 1 +
 xen/arch/arm/{ => mmu}/domain_page.c | 0
 3 files changed, 1 insertion(+), 1 deletion(-)
 rename xen/arch/arm/{ => mmu}/domain_page.c (100%)
xen/arch/arm/Makefile +0 -1
index da9c979dc4..7792bff597 100644
@@ -18,7 +18,6 @@ obj-$(CONFIG_IOREQ_SERVER) += dm.o
 obj-$(CONFIG_DOM0LESS_BOOT) += dom0less-build.init.o
 obj-y += domain.o
 obj-y += domain_build.init.o
-obj-$(CONFIG_ARCH_MAP_DOMAIN_PAGE) += domain_page.o
 obj-y += domctl.o
 obj-$(CONFIG_EARLY_PRINTK) += early_printk.o
 obj-y += efi/
xen/arch/arm/mmu/Makefile +1 -0
index 67475fcd80..2cb44b857d 100644
@@ -2,3 +2,4 @@ obj-y += p2m.o
 obj-y += pt.o
 obj-y += setup.o
 obj-y += smpboot.o
+obj-$(CONFIG_ARCH_MAP_DOMAIN_PAGE) += domain_page.o
xen/arch/arm/mmu/domain_page.c renamed from xen/arch/arm/domain_page.c +0 -0
similarity index 100%
rename from xen/arch/arm/domain_page.c
rename to xen/arch/arm/mmu/domain_page.c
Signature · 1 line
2.25.1
Reviewed-by on v3 3/4
#23 lore raw

Re: [PATCH v3 3/4] xen: arm: Move the functions of domain_page to MMU specific

Hi Ayan,

Quoted 3 lines · Ayan Kumar Halder

Moved init_domheap_mappings(), map_domain_page_global(), unmap_domain_page_global(), map_domain_page(), unmap_domain_page(), domain_page_map_to_mfn() to MMU specific folder.

Both the commit title and msg seem incorrect (you took them from v1, instead of v2).

With that changed (i.e. title and msg taken from v2):

~Michal

#24 lore raw

PATCH v3 4/4 xen: arm: Enclose access to EL2 MMU specific registers under CONFIG_MMU

In reply to Ayan Kumar Halder

All the EL2 MMU specific registers are enclosed within CONFIG_MMU.

---
Changes from :

v1 -
1. 'vttbr_el2' field is enclosed with ifdef.
2. No movement of code.

v2 -
1. Enclosed 'vttbr_el2' access in show_registers() and vcpu_show_registers().

 xen/arch/arm/traps.c | 10 ++++++++++
 1 file changed, 10 insertions(+)
xen/arch/arm/traps.c +10 -0
index aac6c599f8..737f4d65e3 100644
@@ -720,8 +720,10 @@ struct reg_ctxt {
     uint32_t ifsr32_el2;
 #endif
 
+#ifdef CONFIG_MMU
     /* Hypervisor-side state */
     uint64_t vttbr_el2;
+#endif
 };
 
 static const char *mode_string(register_t cpsr)
@@ -919,12 +921,16 @@ static void _show_registers(const struct cpu_user_regs *regs,
 #endif
     }
     printk("  VTCR_EL2: %"PRIregister"\n", READ_SYSREG(VTCR_EL2));
+#ifdef CONFIG_MMU
     printk(" VTTBR_EL2: %016"PRIx64"\n", ctxt->vttbr_el2);
+#endif
     printk("\n");
 
     printk(" SCTLR_EL2: %"PRIregister"\n", READ_SYSREG(SCTLR_EL2));
     printk("   HCR_EL2: %"PRIregister"\n", READ_SYSREG(HCR_EL2));
+#ifdef CONFIG_MMU
     printk(" TTBR0_EL2: %016"PRIx64"\n", READ_SYSREG64(TTBR0_EL2));
+#endif
     printk("\n");
     printk("   ESR_EL2: %"PRIregister"\n", regs->hsr);
     printk(" HPFAR_EL2: %"PRIregister"\n", READ_SYSREG(HPFAR_EL2));
@@ -956,7 +962,9 @@ void show_registers(const struct cpu_user_regs *regs)
     if ( guest_mode(regs) && is_32bit_domain(current->domain) )
         ctxt.ifsr32_el2 = READ_SYSREG(IFSR32_EL2);
 #endif
+#ifdef CONFIG_MMU
     ctxt.vttbr_el2 = READ_SYSREG64(VTTBR_EL2);
+#endif
 
     _show_registers(regs, &ctxt, guest_mode(regs), current);
 }
@@ -979,7 +987,9 @@ void vcpu_show_registers(const struct vcpu *v)
     ctxt.ifsr32_el2 = v->arch.ifsr;
 #endif
 
+#ifdef CONFIG_MMU
     ctxt.vttbr_el2 = v->domain->arch.p2m.vttbr;
+#endif
 
     _show_registers(&v->arch.cpu_info->guest_cpu_user_regs, &ctxt, 1, v);
 }
Signature · 1 line
2.25.1