Skip to contentSkip to conversation

arm64: mte: allow async MTE to be upgraded to sync on a per-CPU basis

PATCH Superseded Peter Collingbourne 3 participants 13 messages Jul 13, 2021 – Jul 27, 2021

Subsystems (2)

v10 of 11 Superseded Latest is v11 →

lore
Share & tools
lore
#0 lore raw

v10 cover arm64: mte: allow async MTE to be upgraded to sync on a per-CPU basis

On some CPUs the performance of MTE in synchronous mode is similar to that of asynchronous mode. This makes it worthwhile to enable synchronous mode on those CPUs when asynchronous mode is requested, in order to gain the error detection benefits of synchronous mode without the performance downsides. Therefore, make it possible for user programs to opt into upgrading to synchronous mode on those CPUs.

This is done by introducing a notion of a preferred TCF mode, which is controlled on a per-CPU basis by a sysfs node. The existing SYNC and ASYNC TCF settings are repurposed as bitfields that specify a set of possible modes. If the preferred TCF mode for a particular CPU is in the user-provided mode set (this will always be the case for mode sets containing more than one mode because the kernel only supports two tag checking modes, but future kernels may support more modes) then that mode is used when running on that CPU, otherwise one of the modes in the task's mode set will be selected in a currently unspecified manner.

v8:
- split into multiple patches
- remove MTE_CTRL_TCF_NONE
- improve documentation
- disable preemption and add comment to mte_update_sctlr_user
- bring back PR_MTE_TCF_SHIFT for source compatibility
- address formatting nit

v7:
- switch to new API proposed on list

v6:
- switch to strings in sysfs nodes instead of TCF values

v5:
- updated documentation
- address some nits in mte.c

v4:
- switch to new mte_ctrl field
- make register_mte_upgrade_async_sysctl return an int
- change the sysctl to take 0 or 1 instead of raw TCF values
- "same as" -> "similar to"

v3:
- drop the device tree support
- add documentation
- add static_assert to ensure no overlap with real HW bits
- move per-CPU variable initialization to mte.c
- use smp_call_function_single instead of stop_machine

v2:
- make it an opt-in behavior
- change the format of the device tree node
- also allow controlling the feature via sysfs

Peter Collingbourne (5):
arm64: mte: rename gcr_user_excl to mte_ctrl
arm64: mte: change ASYNC and SYNC TCF settings into bitfields
arm64: move preemption disablement to prctl handlers
arm64: mte: introduce a per-CPU tag checking mode preference
Documentation: document the preferred tag checking mode feature

 .../ABI/testing/sysfs-devices-system-cpu      |  18 +++
 .../arm64/memory-tagging-extension.rst        |  48 +++++-
 arch/arm64/include/asm/pointer_auth.h         |  12 +-
 arch/arm64/include/asm/processor.h            |  10 +-
 arch/arm64/kernel/asm-offsets.c               |   2 +-
 arch/arm64/kernel/entry.S                     |   4 +-
 arch/arm64/kernel/mte.c                       | 139 ++++++++++++------
 arch/arm64/kernel/pointer_auth.c              |  10 +-
 arch/arm64/kernel/process.c                   |  21 +--
 include/uapi/linux/prctl.h                    |  11 +-
 10 files changed, 189 insertions(+), 86 deletions(-)
Signature · 7 lines
2.32.0.93.g670b81a890-goog


_______________________________________________
linux-arm-kernel mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
#1 lore raw

PATCH v10 1/5 arm64: mte: rename gcr_user_excl to mte_ctrl

We are going to use this field to store more data. To prepare for that, rename it and change the users to rely on the bit position of gcr_user_excl in mte_ctrl.

---
 arch/arm64/include/asm/processor.h |  5 ++++-
 arch/arm64/kernel/asm-offsets.c    |  2 +-
 arch/arm64/kernel/entry.S          |  4 ++--
 arch/arm64/kernel/mte.c            | 14 ++++++++------
 4 files changed, 15 insertions(+), 10 deletions(-)
arch/arm64/include/asm/processor.h +4 -1
index 9df3feeee890..6322fb1714d5 100644
@@ -16,6 +16,9 @@
  */
 #define NET_IP_ALIGN	0
 
+#define MTE_CTRL_GCR_USER_EXCL_SHIFT	0
+#define MTE_CTRL_GCR_USER_EXCL_MASK	0xffff
+
 #ifndef __ASSEMBLY__
 
 #include <linux/build_bug.h>
@@ -151,7 +154,7 @@ struct thread_struct {
 	struct ptrauth_keys_kernel	keys_kernel;
 #endif
 #ifdef CONFIG_ARM64_MTE
-	u64			gcr_user_excl;
+	u64			mte_ctrl;
 #endif
 	u64			sctlr_user;
 };
arch/arm64/kernel/asm-offsets.c +1 -1
index 0cb34ccb6e73..63d02cd67b44 100644
@@ -49,7 +49,7 @@ int main(void)
   DEFINE(THREAD_KEYS_KERNEL,	offsetof(struct task_struct, thread.keys_kernel));
 #endif
 #ifdef CONFIG_ARM64_MTE
-  DEFINE(THREAD_GCR_EL1_USER,	offsetof(struct task_struct, thread.gcr_user_excl));
+  DEFINE(THREAD_MTE_CTRL,	offsetof(struct task_struct, thread.mte_ctrl));
 #endif
   BLANK();
   DEFINE(S_X0,			offsetof(struct pt_regs, regs[0]));
arch/arm64/kernel/entry.S +2 -2
index 3513984a88bd..ce59280355c5 100644
@@ -182,7 +182,7 @@ alternative_else_nop_endif
 	 * the RRND (bit[16]) setting.
 	 */
 	mrs_s	\tmp2, SYS_GCR_EL1
-	bfi	\tmp2, \tmp, #0, #16
+	bfxil	\tmp2, \tmp, #MTE_CTRL_GCR_USER_EXCL_SHIFT, #16
 	msr_s	SYS_GCR_EL1, \tmp2
 #endif
 	.endm
@@ -205,7 +205,7 @@ alternative_else_nop_endif
 alternative_if_not ARM64_MTE
 	b	1f
 alternative_else_nop_endif
-	ldr	\tmp, [\tsk, #THREAD_GCR_EL1_USER]
+	ldr	\tmp, [\tsk, #THREAD_MTE_CTRL]
 
 	mte_set_gcr \tmp, \tmp2
 1:
arch/arm64/kernel/mte.c +8 -6
index 125a10e413e9..d3884d09513d 100644
@@ -199,7 +199,7 @@ static void update_gcr_el1_excl(u64 excl)
 
 static void set_gcr_el1_excl(u64 excl)
 {
-	current->thread.gcr_user_excl = excl;
+	current->thread.mte_ctrl = excl;
 
 	/*
 	 * SYS_GCR_EL1 will be set to current->thread.gcr_user_excl value
@@ -263,8 +263,8 @@ void mte_suspend_exit(void)
 long set_mte_ctrl(struct task_struct *task, unsigned long arg)
 {
 	u64 sctlr = task->thread.sctlr_user & ~SCTLR_EL1_TCF0_MASK;
-	u64 gcr_excl = ~((arg & PR_MTE_TAG_MASK) >> PR_MTE_TAG_SHIFT) &
-		       SYS_GCR_EL1_EXCL_MASK;
+	u64 mte_ctrl = (~((arg & PR_MTE_TAG_MASK) >> PR_MTE_TAG_SHIFT) &
+			SYS_GCR_EL1_EXCL_MASK) << MTE_CTRL_GCR_USER_EXCL_SHIFT;
 
 	if (!system_supports_mte())
 		return 0;
@@ -285,10 +285,10 @@ long set_mte_ctrl(struct task_struct *task, unsigned long arg)
 
 	if (task != current) {
 		task->thread.sctlr_user = sctlr;
-		task->thread.gcr_user_excl = gcr_excl;
+		task->thread.mte_ctrl = mte_ctrl;
 	} else {
 		set_task_sctlr_el1(sctlr);
-		set_gcr_el1_excl(gcr_excl);
+		set_gcr_el1_excl(mte_ctrl);
 	}
 
 	return 0;
@@ -297,7 +297,9 @@ long set_mte_ctrl(struct task_struct *task, unsigned long arg)
 long get_mte_ctrl(struct task_struct *task)
 {
 	unsigned long ret;
-	u64 incl = ~task->thread.gcr_user_excl & SYS_GCR_EL1_EXCL_MASK;
+	u64 mte_ctrl = task->thread.mte_ctrl;
+	u64 incl = (~mte_ctrl >> MTE_CTRL_GCR_USER_EXCL_SHIFT) &
+		   SYS_GCR_EL1_EXCL_MASK;
 
 	if (!system_supports_mte())
 		return 0;
Signature · 7 lines
2.32.0.93.g670b81a890-goog


_______________________________________________
linux-arm-kernel mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
maintainer Acked-by on v10 1/5
#2 lore raw

Re: [PATCH v10 1/5] arm64: mte: rename gcr_user_excl to mte_ctrl

Quoted 7 lines · Peter Collingbourne

We are going to use this field to store more data. To prepare for that, rename it and change the users to rely on the bit position of gcr_user_excl in mte_ctrl.

Quoted 6 lines
---
 arch/arm64/include/asm/processor.h |  5 ++++-
 arch/arm64/kernel/asm-offsets.c    |  2 +-
 arch/arm64/kernel/entry.S          |  4 ++--
 arch/arm64/kernel/mte.c            | 14 ++++++++------
 4 files changed, 15 insertions(+), 10 deletions(-)

I acked this one already, but here it is again:

Will

_______________________________________________
linux-arm-kernel mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

#3 lore raw

PATCH v10 2/5 arm64: mte: change ASYNC and SYNC TCF settings into bitfields

In reply to Peter Collingbourne

Allow the user program to specify both ASYNC and SYNC TCF modes by repurposing the existing constants as bitfields. This will allow the kernel to select one of the modes on behalf of the user program. With this patch the kernel will always select async mode, but a subsequent patch will make this configurable.

---
v9:
- make mte_update_sctlr_user static

 arch/arm64/include/asm/processor.h |  3 ++
 arch/arm64/kernel/mte.c            | 70 ++++++++++++------------------
 include/uapi/linux/prctl.h         | 11 ++---
 3 files changed, 37 insertions(+), 47 deletions(-)
arch/arm64/include/asm/processor.h +3 -0
index 6322fb1714d5..80ceb9cbdd60 100644
@@ -19,6 +19,9 @@
 #define MTE_CTRL_GCR_USER_EXCL_SHIFT	0
 #define MTE_CTRL_GCR_USER_EXCL_MASK	0xffff
 
+#define MTE_CTRL_TCF_SYNC		(1UL << 16)
+#define MTE_CTRL_TCF_ASYNC		(1UL << 17)
+
 #ifndef __ASSEMBLY__
 
 #include <linux/build_bug.h>
arch/arm64/kernel/mte.c +28 -42

108 lines not on this page · raw message

include/uapi/linux/prctl.h +6 -5
index 18a9f59dc067..d3a5afb4c1ae 100644
@@ -234,14 +234,15 @@ struct prctl_mm_map {
 #define PR_GET_TAGGED_ADDR_CTRL		56
 # define PR_TAGGED_ADDR_ENABLE		(1UL << 0)
 /* MTE tag check fault modes */
-# define PR_MTE_TCF_SHIFT		1
-# define PR_MTE_TCF_NONE		(0UL << PR_MTE_TCF_SHIFT)
-# define PR_MTE_TCF_SYNC		(1UL << PR_MTE_TCF_SHIFT)
-# define PR_MTE_TCF_ASYNC		(2UL << PR_MTE_TCF_SHIFT)
-# define PR_MTE_TCF_MASK		(3UL << PR_MTE_TCF_SHIFT)
+# define PR_MTE_TCF_NONE		0
+# define PR_MTE_TCF_SYNC		(1UL << 1)
+# define PR_MTE_TCF_ASYNC		(1UL << 2)
+# define PR_MTE_TCF_MASK		(PR_MTE_TCF_SYNC | PR_MTE_TCF_ASYNC)
 /* MTE tag inclusion mask */
 # define PR_MTE_TAG_SHIFT		3
 # define PR_MTE_TAG_MASK		(0xffffUL << PR_MTE_TAG_SHIFT)
+/* Unused; kept only for source compatibility */
+# define PR_MTE_TCF_SHIFT		1
 
 /* Control reclaim behavior when allocating memory */
 #define PR_SET_IO_FLUSHER		57
Signature · 7 lines
2.32.0.93.g670b81a890-goog


_______________________________________________
linux-arm-kernel mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
maintainer Acked-by on v10 2/5
#4 lore raw

Re: [PATCH v10 2/5] arm64: mte: change ASYNC and SYNC TCF settings into bitfields

Quoted 9 lines · Peter Collingbourne

Allow the user program to specify both ASYNC and SYNC TCF modes by repurposing the existing constants as bitfields. This will allow the kernel to select one of the modes on behalf of the user program. With this patch the kernel will always select async mode, but a subsequent patch will make this configurable.

Quoted 8 lines

---
v9:
- make mte_update_sctlr_user static

 arch/arm64/include/asm/processor.h |  3 ++
 arch/arm64/kernel/mte.c            | 70 ++++++++++++------------------
 include/uapi/linux/prctl.h         | 11 ++---
 3 files changed, 37 insertions(+), 47 deletions(-)

Will

_______________________________________________
linux-arm-kernel mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

#5 lore raw

PATCH v10 3/5 arm64: move preemption disablement to prctl handlers

In reply to Peter Collingbourne

In the next patch, we will start reading sctlr_user from mte_update_sctlr_user and subsequently writing a new value based on the task's TCF setting and potentially the per-CPU TCF preference. This means that we need to be careful to disable preemption around any code sequences that read from sctlr_user and subsequently write to sctlr_user and/or SCTLR_EL1, so that we don't end up writing a stale value (based on the previous CPU's TCF preference) to either of them.

We currently have four such sequences, in the prctl handlers for PR_SET_TAGGED_ADDR_CTRL and PR_PAC_SET_ENABLED_KEYS, as well as in the task initialization code that resets the prctl settings. Change the prctl handlers to disable preemption in the handlers themselves rather than the functions that they call, and change the task initialization code to call the respective prctl handlers instead of setting sctlr_user directly.

As a result of this change, we no longer need the helper function set_task_sctlr_el1, nor does its behavior make sense any more, so remove it.

---
 arch/arm64/include/asm/pointer_auth.h | 12 ++++++------
 arch/arm64/include/asm/processor.h    |  2 +-
 arch/arm64/kernel/mte.c               |  8 ++++----
 arch/arm64/kernel/pointer_auth.c      | 10 ++++++----
 arch/arm64/kernel/process.c           | 21 +++++++--------------
 5 files changed, 24 insertions(+), 29 deletions(-)
arch/arm64/include/asm/pointer_auth.h +6 -6
index d50416be99be..592968f0bc22 100644
@@ -10,6 +10,9 @@
 #include <asm/memory.h>
 #include <asm/sysreg.h>
 
+#define PR_PAC_ENABLED_KEYS_MASK                                               \
+	(PR_PAC_APIAKEY | PR_PAC_APIBKEY | PR_PAC_APDAKEY | PR_PAC_APDBKEY)
+
 #ifdef CONFIG_ARM64_PTR_AUTH
 /*
  * Each key is a 128-bit quantity which is split across a pair of 64-bit
@@ -113,9 +116,9 @@ static __always_inline void ptrauth_enable(void)
 									       \
 		/* enable all keys */                                          \
 		if (system_supports_address_auth())                            \
-			set_task_sctlr_el1(current->thread.sctlr_user |        \
-					   SCTLR_ELx_ENIA | SCTLR_ELx_ENIB |   \
-					   SCTLR_ELx_ENDA | SCTLR_ELx_ENDB);   \
+			ptrauth_set_enabled_keys(current,                      \
+						 PR_PAC_ENABLED_KEYS_MASK,     \
+						 PR_PAC_ENABLED_KEYS_MASK);    \
 	} while (0)
 
 #define ptrauth_thread_switch_user(tsk)                                        \
@@ -139,7 +142,4 @@ static __always_inline void ptrauth_enable(void)
 #define ptrauth_thread_switch_kernel(tsk)
 #endif /* CONFIG_ARM64_PTR_AUTH */
 
-#define PR_PAC_ENABLED_KEYS_MASK                                               \
-	(PR_PAC_APIAKEY | PR_PAC_APIBKEY | PR_PAC_APDAKEY | PR_PAC_APDBKEY)
-
 #endif /* __ASM_POINTER_AUTH_H */
arch/arm64/include/asm/processor.h +1 -1
index 80ceb9cbdd60..ebb3b1aefed7 100644
@@ -257,7 +257,7 @@ extern void release_thread(struct task_struct *);
 
 unsigned long get_wchan(struct task_struct *p);
 
-void set_task_sctlr_el1(u64 sctlr);
+void update_sctlr_el1(u64 sctlr);
 
 /* Thread switching */
 extern struct task_struct *cpu_switch_to(struct task_struct *prev,
arch/arm64/kernel/mte.c +4 -4
index 53d89915029d..432d9b641e9c 100644
@@ -222,9 +222,7 @@ void mte_thread_init_user(void)
 	write_sysreg_s(0, SYS_TFSRE0_EL1);
 	clear_thread_flag(TIF_MTE_ASYNC_FAULT);
 	/* disable tag checking and reset tag generation mask */
-	current->thread.mte_ctrl = MTE_CTRL_GCR_USER_EXCL_MASK;
-	mte_update_sctlr_user(current);
-	set_task_sctlr_el1(current->thread.sctlr_user);
+	set_mte_ctrl(current, 0);
 }
 
 void mte_thread_switch(struct task_struct *next)
@@ -281,8 +279,10 @@ long set_mte_ctrl(struct task_struct *task, unsigned long arg)
 
 	task->thread.mte_ctrl = mte_ctrl;
 	if (task == current) {
+		preempt_disable();
 		mte_update_sctlr_user(task);
-		set_task_sctlr_el1(task->thread.sctlr_user);
+		update_sctlr_el1(task->thread.sctlr_user);
+		preempt_enable();
 	}
 
 	return 0;
arch/arm64/kernel/pointer_auth.c +6 -4
index 60901ab0a7fe..2708b620b4ae 100644
@@ -67,7 +67,7 @@ static u64 arg_to_enxx_mask(unsigned long arg)
 int ptrauth_set_enabled_keys(struct task_struct *tsk, unsigned long keys,
 			     unsigned long enabled)
 {
-	u64 sctlr = tsk->thread.sctlr_user;
+	u64 sctlr;
 
 	if (!system_supports_address_auth())
 		return -EINVAL;
@@ -78,12 +78,14 @@ int ptrauth_set_enabled_keys(struct task_struct *tsk, unsigned long keys,
 	if ((keys & ~PR_PAC_ENABLED_KEYS_MASK) || (enabled & ~keys))
 		return -EINVAL;
 
+	preempt_disable();
+	sctlr = tsk->thread.sctlr_user;
 	sctlr &= ~arg_to_enxx_mask(keys);
 	sctlr |= arg_to_enxx_mask(enabled);
+	tsk->thread.sctlr_user = sctlr;
 	if (tsk == current)
-		set_task_sctlr_el1(sctlr);
-	else
-		tsk->thread.sctlr_user = sctlr;
+		update_sctlr_el1(sctlr);
+	preempt_enable();
 
 	return 0;
 }
arch/arm64/kernel/process.c +7 -14
index b4bb67f17a2c..c548eec87810 100644
@@ -527,7 +527,13 @@ static void erratum_1418040_thread_switch(struct task_struct *prev,
 	write_sysreg(val, cntkctl_el1);
 }
 
-static void update_sctlr_el1(u64 sctlr)
+/*
+ * __switch_to() checks current->thread.sctlr_user as an optimisation. Therefore
+ * this function must be called with preemption disabled and the update to
+ * sctlr_user must be made in the same preemption disabled block so that
+ * __switch_to() does not see the variable update before the SCTLR_EL1 one.
+ */
+void update_sctlr_el1(u64 sctlr)
 {
 	/*
 	 * EnIA must not be cleared while in the kernel as this is necessary for
@@ -539,19 +545,6 @@ static void update_sctlr_el1(u64 sctlr)
 	isb();
 }
 
-void set_task_sctlr_el1(u64 sctlr)
-{
-	/*
-	 * __switch_to() checks current->thread.sctlr as an
-	 * optimisation. Disable preemption so that it does not see
-	 * the variable update before the SCTLR_EL1 one.
-	 */
-	preempt_disable();
-	current->thread.sctlr_user = sctlr;
-	update_sctlr_el1(sctlr);
-	preempt_enable();
-}
-
 /*
  * Thread switching.
  */
Signature · 7 lines
2.32.0.93.g670b81a890-goog


_______________________________________________
linux-arm-kernel mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
maintainer Acked-by on v10 3/5
#6 lore raw

Re: [PATCH v10 3/5] arm64: move preemption disablement to prctl handlers

Quoted 123 lines · Peter Collingbourne

In the next patch, we will start reading sctlr_user from mte_update_sctlr_user and subsequently writing a new value based on the task's TCF setting and potentially the per-CPU TCF preference. This means that we need to be careful to disable preemption around any code sequences that read from sctlr_user and subsequently write to sctlr_user and/or SCTLR_EL1, so that we don't end up writing a stale value (based on the previous CPU's TCF preference) to either of them.

We currently have four such sequences, in the prctl handlers for PR_SET_TAGGED_ADDR_CTRL and PR_PAC_SET_ENABLED_KEYS, as well as in the task initialization code that resets the prctl settings. Change the prctl handlers to disable preemption in the handlers themselves rather than the functions that they call, and change the task initialization code to call the respective prctl handlers instead of setting sctlr_user directly.

As a result of this change, we no longer need the helper function set_task_sctlr_el1, nor does its behavior make sense any more, so remove it.

---
 arch/arm64/include/asm/pointer_auth.h | 12 ++++++------
 arch/arm64/include/asm/processor.h    |  2 +-
 arch/arm64/kernel/mte.c               |  8 ++++----
 arch/arm64/kernel/pointer_auth.c      | 10 ++++++----
 arch/arm64/kernel/process.c           | 21 +++++++--------------
 5 files changed, 24 insertions(+), 29 deletions(-)
arch/arm64/include/asm/pointer_auth.h +6 -6
index d50416be99be..592968f0bc22 100644
@@ -10,6 +10,9 @@
 #include <asm/memory.h>
 #include <asm/sysreg.h>
 
+#define PR_PAC_ENABLED_KEYS_MASK                                               \
+	(PR_PAC_APIAKEY | PR_PAC_APIBKEY | PR_PAC_APDAKEY | PR_PAC_APDBKEY)
+
 #ifdef CONFIG_ARM64_PTR_AUTH
 /*
  * Each key is a 128-bit quantity which is split across a pair of 64-bit
@@ -113,9 +116,9 @@ static __always_inline void ptrauth_enable(void)
 									       \
 		/* enable all keys */                                          \
 		if (system_supports_address_auth())                            \
-			set_task_sctlr_el1(current->thread.sctlr_user |        \
-					   SCTLR_ELx_ENIA | SCTLR_ELx_ENIB |   \
-					   SCTLR_ELx_ENDA | SCTLR_ELx_ENDB);   \
+			ptrauth_set_enabled_keys(current,                      \
+						 PR_PAC_ENABLED_KEYS_MASK,     \
+						 PR_PAC_ENABLED_KEYS_MASK);    \
 	} while (0)
 
 #define ptrauth_thread_switch_user(tsk)                                        \
@@ -139,7 +142,4 @@ static __always_inline void ptrauth_enable(void)
 #define ptrauth_thread_switch_kernel(tsk)
 #endif /* CONFIG_ARM64_PTR_AUTH */
 
-#define PR_PAC_ENABLED_KEYS_MASK                                               \
-	(PR_PAC_APIAKEY | PR_PAC_APIBKEY | PR_PAC_APDAKEY | PR_PAC_APDBKEY)
-
 #endif /* __ASM_POINTER_AUTH_H */
arch/arm64/include/asm/processor.h +1 -1
index 80ceb9cbdd60..ebb3b1aefed7 100644
@@ -257,7 +257,7 @@ extern void release_thread(struct task_struct *);
 
 unsigned long get_wchan(struct task_struct *p);
 
-void set_task_sctlr_el1(u64 sctlr);
+void update_sctlr_el1(u64 sctlr);
 
 /* Thread switching */
 extern struct task_struct *cpu_switch_to(struct task_struct *prev,
arch/arm64/kernel/mte.c +4 -4
index 53d89915029d..432d9b641e9c 100644
@@ -222,9 +222,7 @@ void mte_thread_init_user(void)
 	write_sysreg_s(0, SYS_TFSRE0_EL1);
 	clear_thread_flag(TIF_MTE_ASYNC_FAULT);
 	/* disable tag checking and reset tag generation mask */
-	current->thread.mte_ctrl = MTE_CTRL_GCR_USER_EXCL_MASK;
-	mte_update_sctlr_user(current);
-	set_task_sctlr_el1(current->thread.sctlr_user);
+	set_mte_ctrl(current, 0);
 }
 
 void mte_thread_switch(struct task_struct *next)
@@ -281,8 +279,10 @@ long set_mte_ctrl(struct task_struct *task, unsigned long arg)
 
 	task->thread.mte_ctrl = mte_ctrl;
 	if (task == current) {
+		preempt_disable();
 		mte_update_sctlr_user(task);
-		set_task_sctlr_el1(task->thread.sctlr_user);
+		update_sctlr_el1(task->thread.sctlr_user);
+		preempt_enable();
 	}
 
 	return 0;
arch/arm64/kernel/pointer_auth.c +2 -1
index 60901ab0a7fe..2708b620b4ae 100644
@@ -67,7 +67,7 @@ static u64 arg_to_enxx_mask(unsigned long arg)
 int ptrauth_set_enabled_keys(struct task_struct *tsk, unsigned long keys,
 			     unsigned long enabled)
 {
-	u64 sctlr = tsk->thread.sctlr_user;
+	u64 sctlr;
 
 	if (!system_supports_address_auth())
 		return -EINVAL;
@@ -78,12 +78,14 @@ int ptrauth_set_enabled_keys(struct task_struct *tsk, unsigned long keys,
 	if ((keys & ~PR_PAC_ENABLED_KEYS_MASK) || (enabled & ~keys))
 		return -EINVAL;
 
+	preempt_disable();

arch/arm64/kernel/pointer_auth.c int ptrauth_set_enabled_keys(struct task_struct *tsk, unsigned long keys, · line 88

Quoted 10 lines
arch/arm64/kernel/pointer_auth.c +4 -3
+	sctlr = tsk->thread.sctlr_user;
 	sctlr &= ~arg_to_enxx_mask(keys);
 	sctlr |= arg_to_enxx_mask(enabled);
+	tsk->thread.sctlr_user = sctlr;
 	if (tsk == current)
-		set_task_sctlr_el1(sctlr);
-	else
-		tsk->thread.sctlr_user = sctlr;
+		update_sctlr_el1(sctlr);
+	preempt_enable();

This should work, as long as tsk is stopped if tsk != current, otherwise I think there are a bunch of existing races between ptrauth_set_enabled_keys(), set_mte_ctrl() and ptrauth_get_enabled_keys(). That's (probably?) true for the ptrace case, which is the only scenario where tsk != current afaict.

I can't help but feel we should add _something_ (as a separate patch) to avoid tripping over this in future. Maybe a WARN_ON(!task_is_stopped(tsk)) if tsk != current? Failing that, we could handle the concurrency by adding a helper to update sctlr_user, which takes a clear and a set mask and does a cmpxchg_relaxed() under the hood? That might also eliminate the need for some of these preempt_disable() sections.

Anyway, that's a pre-existing issue, so for this patch:

Will

_______________________________________________
linux-arm-kernel mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

#7 lore raw

PATCH v10 4/5 arm64: mte: introduce a per-CPU tag checking mode preference

In reply to Peter Collingbourne

Add a per-CPU sysfs node, mte_tcf_preferred, that allows the preferred tag checking mode to be configured. The current possible values are async and sync.

---
v10:
- stop calling smp_call_function_single

 arch/arm64/kernel/mte.c | 65 +++++++++++++++++++++++++++++++++++++++--
 1 file changed, 63 insertions(+), 2 deletions(-)
arch/arm64/kernel/mte.c +63 -2
index 432d9b641e9c..50ae1a7435ae 100644
@@ -4,6 +4,7 @@
  */
 
 #include <linux/bitops.h>
+#include <linux/cpu.h>
 #include <linux/kernel.h>
 #include <linux/mm.h>
 #include <linux/prctl.h>
@@ -26,6 +27,8 @@ u64 gcr_kernel_excl __ro_after_init;
 
 static bool report_fault_once = true;
 
+static DEFINE_PER_CPU_READ_MOSTLY(u64, mte_tcf_preferred);
+
 #ifdef CONFIG_KASAN_HW_TAGS
 /* Whether the MTE asynchronous mode is enabled. */
 DEFINE_STATIC_KEY_FALSE(mte_async_mode);
@@ -199,11 +202,18 @@ static void update_gcr_el1_excl(u64 excl)
 
 static void mte_update_sctlr_user(struct task_struct *task)
 {
+	/*
+	 * This must be called with preemption disabled and can only be called
+	 * on the current or next task since the CPU must match where the thread
+	 * is going to run. The caller is responsible for calling
+	 * update_sctlr_el1() later in the same preemption disabled block.
+	 */
 	unsigned long sctlr = task->thread.sctlr_user;
-	unsigned long pref = MTE_CTRL_TCF_ASYNC;
 	unsigned long mte_ctrl = task->thread.mte_ctrl;
-	unsigned long resolved_mte_tcf = (mte_ctrl & pref) ? pref : mte_ctrl;
+	unsigned long pref, resolved_mte_tcf;
 
+	pref = __this_cpu_read(mte_tcf_preferred);
+	resolved_mte_tcf = (mte_ctrl & pref) ? pref : mte_ctrl;
 	sctlr &= ~SCTLR_EL1_TCF0_MASK;
 	if (resolved_mte_tcf & MTE_CTRL_TCF_ASYNC)
 		sctlr |= SCTLR_EL1_TCF0_ASYNC;
@@ -441,3 +451,54 @@ int mte_ptrace_copy_tags(struct task_struct *child, long request,
 
 	return ret;
 }
+
+static ssize_t mte_tcf_preferred_show(struct device *dev,
+				      struct device_attribute *attr, char *buf)
+{
+	switch (per_cpu(mte_tcf_preferred, dev->id)) {
+	case MTE_CTRL_TCF_ASYNC:
+		return sysfs_emit(buf, "async\n");
+	case MTE_CTRL_TCF_SYNC:
+		return sysfs_emit(buf, "sync\n");
+	default:
+		return sysfs_emit(buf, "???\n");
+	}
+}
+
+static ssize_t mte_tcf_preferred_store(struct device *dev,
+				       struct device_attribute *attr,
+				       const char *buf, size_t count)
+{
+	u64 tcf;
+
+	if (sysfs_streq(buf, "async"))
+		tcf = MTE_CTRL_TCF_ASYNC;
+	else if (sysfs_streq(buf, "sync"))
+		tcf = MTE_CTRL_TCF_SYNC;
+	else
+		return -EINVAL;
+
+	device_lock(dev);
+	per_cpu(mte_tcf_preferred, dev->id) = tcf;
+	device_unlock(dev);
+
+	return count;
+}
+static DEVICE_ATTR_RW(mte_tcf_preferred);
+
+static int register_mte_tcf_preferred_sysctl(void)
+{
+	unsigned int cpu;
+
+	if (!system_supports_mte())
+		return 0;
+
+	for_each_possible_cpu(cpu) {
+		per_cpu(mte_tcf_preferred, cpu) = MTE_CTRL_TCF_ASYNC;
+		device_create_file(get_cpu_device(cpu),
+				   &dev_attr_mte_tcf_preferred);
+	}
+
+	return 0;
+}
+subsys_initcall(register_mte_tcf_preferred_sysctl);
Signature · 7 lines
2.32.0.93.g670b81a890-goog


_______________________________________________
linux-arm-kernel mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
maintainer Acked-by on v10 4/5
#8 lore raw

Re: [PATCH v10 4/5] arm64: mte: introduce a per-CPU tag checking mode preference

Quoted 7 lines · Peter Collingbourne

Add a per-CPU sysfs node, mte_tcf_preferred, that allows the preferred tag checking mode to be configured. The current possible values are async and sync.

Quoted 6 lines

---
v10:
- stop calling smp_call_function_single

 arch/arm64/kernel/mte.c | 65 +++++++++++++++++++++++++++++++++++++++--
 1 file changed, 63 insertions(+), 2 deletions(-)

Will

_______________________________________________
linux-arm-kernel mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

#9 lore raw

PATCH v10 5/5 Documentation: document the preferred tag checking mode feature

In reply to Peter Collingbourne

Document the functionality added in the previous patches.

---
v10:
- document that setting the sysfs node may not take effect immediately
- unabbreviate month name

v9:
- add documentation for sysfs node under Documentation/ABI

 .../ABI/testing/sysfs-devices-system-cpu      | 18 +++++++
 .../arm64/memory-tagging-extension.rst        | 48 ++++++++++++++++---
 2 files changed, 59 insertions(+), 7 deletions(-)
Documentation/ABI/testing/sysfs-devices-system-cpu +18 -0
index fe13baa53c59..6e667338df9f 100644
@@ -640,3 +640,21 @@ Description:	SPURR ticks for cpuX when it was idle.
 
 		This sysfs interface exposes the number of SPURR ticks
 		for cpuX when it was idle.
+
+What: 		/sys/devices/system/cpu/cpuX/mte_tcf_preferred
+Date:		July 2021
+Contact:	Linux ARM Kernel Mailing list <[email protected]>
+Description:	Preferred MTE tag checking mode
+
+		When a user program specifies more than one MTE tag checking
+		mode, this sysfs node is used to specify which mode should
+		be preferred when running on that CPU. Possible values:
+
+		================  ==============================================
+		"sync"	  	  Prefer synchronous mode
+		"async"	  	  Prefer asynchronous mode
+		================  ==============================================
+
+		Changes to this sysfs node may not take effect immediately.
+
+		See also: Documentation/arm64/memory-tagging-extension.rst
Documentation/arm64/memory-tagging-extension.rst +41 -7

79 lines not on this page · raw message

Signature · 7 lines
2.32.0.93.g670b81a890-goog


_______________________________________________
linux-arm-kernel mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
maintainer Acked-by on v10 5/5
#10 lore raw

Re: [PATCH v10 5/5] Documentation: document the preferred tag checking mode feature

Quoted 26 lines · Peter Collingbourne

Document the functionality added in the previous patches.

---
v10:
- document that setting the sysfs node may not take effect immediately
- unabbreviate month name

v9:
- add documentation for sysfs node under Documentation/ABI

 .../ABI/testing/sysfs-devices-system-cpu      | 18 +++++++
 .../arm64/memory-tagging-extension.rst        | 48 ++++++++++++++++---
 2 files changed, 59 insertions(+), 7 deletions(-)
Documentation/ABI/testing/sysfs-devices-system-cpu +0 -0
index fe13baa53c59..6e667338df9f 100644
@@ -640,3 +640,21 @@ Description:	SPURR ticks for cpuX when it was idle.
 
 		This sysfs interface exposes the number of SPURR ticks

Documentation/ABI/testing/sysfs-devices-system-cpu Description: SPURR ticks for cpuX when it was idle. · line 651

Quoted 10 lines
Documentation/ABI/testing/sysfs-devices-system-cpu +9 -0
 		for cpuX when it was idle.
+
+What: 		/sys/devices/system/cpu/cpuX/mte_tcf_preferred
+Date:		July 2021
+Contact:	Linux ARM Kernel Mailing list <[email protected]>
+Description:	Preferred MTE tag checking mode
+
+		When a user program specifies more than one MTE tag checking
+		mode, this sysfs node is used to specify which mode should
+		be preferred when running on that CPU. Possible values:

So I suppose we could replace "running" with "scheduling a task" and then remove:

Documentation/ABI/testing/sysfs-devices-system-cpu Description: SPURR ticks for cpuX when it was idle. · line 658

Quoted 1 line
Documentation/ABI/testing/sysfs-devices-system-cpu +1 -0
+		Changes to this sysfs node may not take effect immediately.

altogether, as it makes it sounds as though the change might just be ignored.

But I don't want to bikeshed this, so:

Will

_______________________________________________
linux-arm-kernel mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

maintainer
#11 lore raw
In reply to Peter Collingbourne

Hi Peter,

Quoted 6 lines · Peter Collingbourne

Peter Collingbourne (5):
arm64: mte: rename gcr_user_excl to mte_ctrl
arm64: mte: change ASYNC and SYNC TCF settings into bitfields
arm64: move preemption disablement to prctl handlers
arm64: mte: introduce a per-CPU tag checking mode preference
Documentation: document the preferred tag checking mode feature

Could you please rebase these patches on top of arm64 for-next/mte (based on 5.14-rc3)? They don't apply cleanly. You may want to update the other GCR_EL1 and double ISB patches as well in case there's a specific order in which they need to be applied.

Thanks.

Signature · 6 lines
Catalin

_______________________________________________
linux-arm-kernel mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
#12 lore raw
Quoted 11 lines · Catalin Marinas

Hi Peter,

Quoted 6 lines · Peter Collingbourne

Peter Collingbourne (5):
arm64: mte: rename gcr_user_excl to mte_ctrl
arm64: mte: change ASYNC and SYNC TCF settings into bitfields
arm64: move preemption disablement to prctl handlers
arm64: mte: introduce a per-CPU tag checking mode preference
Documentation: document the preferred tag checking mode feature

Quoted 4 lines

Could you please rebase these patches on top of arm64 for-next/mte (based on 5.14-rc3)? They don't apply cleanly. You may want to update the other GCR_EL1 and double ISB patches as well in case there's a specific order in which they need to be applied.

Okay, I rebased this series and the double ISB patch to 5.14-rc3 (the GCR_EL1 patch didn't need to be rebased). I also double checked that there aren't any conflicts with the TFSR patch that you've already taken. The GCR_EL1 patch needs to be applied after this series, and the double ISB patch doesn't need to be applied in any particular order.

Peter

_______________________________________________
linux-arm-kernel mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel