[RISCV] fix: Use WRITE_ONCE() when setting page table entries

Documentation/features/sched/membarrier-sync-core/arch-support.txt

@@ -10,6 +10,22 @@
 # Rely on implicit context synchronization as a result of exception return
 # when returning from IPI handler, and when returning to user-space.
 #
+# * riscv
+#
+# riscv uses xRET as return from interrupt and to return to user-space.
+#
+# Given that xRET is not core serializing, we rely on FENCE.I for providing
+# core serialization:
+#
+#  - by calling sync_core_before_usermode() on return from interrupt (cf.
+#    ipi_sync_core()),
+#
+#  - via switch_mm() and sync_core_before_usermode() (respectively, for
+#    uthread->uthread and kthread->uthread transitions) before returning
+#    to user-space.
+#
+#  The serialization in switch_mm() is activated by prepare_sync_core_cmd().
+#
 # * x86
 #
 # x86-32 uses IRET as return from interrupt, which takes care of the IPI.
@@ -44,7 +60,7 @@
     |    openrisc: | TODO |
     |      parisc: | TODO |
     |     powerpc: |  ok  |
-    |       riscv: | TODO |
+    |       riscv: |  ok  |
     |        s390: |  ok  |
     |          sh: | TODO |
     |       sparc: | TODO |

Documentation/scheduler/index.rst

@@ -7,6 +7,7 @@ Scheduler
 
 
     completion
+    membarrier
     sched-arch
     sched-bwc
     sched-deadline

Documentation/scheduler/membarrier.rst

@@ -0,0 +1,39 @@
+.. SPDX-License-Identifier: GPL-2.0
+
+========================
+membarrier() System Call
+========================
+
+MEMBARRIER_CMD_{PRIVATE,GLOBAL}_EXPEDITED - Architecture requirements
+=====================================================================
+
+Memory barriers before updating rq->curr
+----------------------------------------
+
+The commands MEMBARRIER_CMD_PRIVATE_EXPEDITED and MEMBARRIER_CMD_GLOBAL_EXPEDITED
+require each architecture to have a full memory barrier after coming from
+user-space, before updating rq->curr.  This barrier is implied by the sequence
+rq_lock(); smp_mb__after_spinlock() in __schedule().  The barrier matches a full
+barrier in the proximity of the membarrier system call exit, cf.
+membarrier_{private,global}_expedited().
+
+Memory barriers after updating rq->curr
+---------------------------------------
+
+The commands MEMBARRIER_CMD_PRIVATE_EXPEDITED and MEMBARRIER_CMD_GLOBAL_EXPEDITED
+require each architecture to have a full memory barrier after updating rq->curr,
+before returning to user-space.  The schemes providing this barrier on the various
+architectures are as follows.
+
+ - alpha, arc, arm, hexagon, mips rely on the full barrier implied by
+   spin_unlock() in finish_lock_switch().
+
+ - arm64 relies on the full barrier implied by switch_to().
+
+ - powerpc, riscv, s390, sparc, x86 rely on the full barrier implied by
+   switch_mm(), if mm is not NULL; they rely on the full barrier implied
+   by mmdrop(), otherwise.  On powerpc and riscv, switch_mm() relies on
+   membarrier_arch_switch_mm().
+
+The barrier matches a full barrier in the proximity of the membarrier system call
+entry, cf. membarrier_{private,global}_expedited().

MAINTAINERS

@@ -13732,7 +13732,9 @@ M:	Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
 M:	"Paul E. McKenney" <paulmck@kernel.org>
 L:	linux-kernel@vger.kernel.org
 S:	Supported
-F:	arch/powerpc/include/asm/membarrier.h
+F:	Documentation/scheduler/membarrier.rst
+F:	arch/*/include/asm/membarrier.h
+F:	arch/*/include/asm/sync_core.h
 F:	include/uapi/linux/membarrier.h
 F:	kernel/sched/membarrier.c
 

arch/arm/include/asm/pgtable.h

@@ -151,6 +151,8 @@ extern pgprot_t phys_mem_access_prot(struct file *file, unsigned long pfn,
 
 extern pgd_t swapper_pg_dir[PTRS_PER_PGD];
 
+#define pgdp_get(pgpd)		READ_ONCE(*pgdp)
+
 #define pud_page(pud)		pmd_page(__pmd(pud_val(pud)))
 #define pud_write(pud)		pmd_write(__pmd(pud_val(pud)))
 

arch/riscv/Kconfig

@@ -27,14 +27,18 @@ config RISCV
 	select ARCH_HAS_GCOV_PROFILE_ALL
 	select ARCH_HAS_GIGANTIC_PAGE
 	select ARCH_HAS_KCOV
+	select ARCH_HAS_MEMBARRIER_CALLBACKS
+	select ARCH_HAS_MEMBARRIER_SYNC_CORE
 	select ARCH_HAS_MMIOWB
 	select ARCH_HAS_NON_OVERLAPPING_ADDRESS_SPACE
 	select ARCH_HAS_PMEM_API
+	select ARCH_HAS_PREPARE_SYNC_CORE_CMD
 	select ARCH_HAS_PTE_SPECIAL
 	select ARCH_HAS_SET_DIRECT_MAP if MMU
 	select ARCH_HAS_SET_MEMORY if MMU
 	select ARCH_HAS_STRICT_KERNEL_RWX if MMU && !XIP_KERNEL
 	select ARCH_HAS_STRICT_MODULE_RWX if MMU && !XIP_KERNEL
+	select ARCH_HAS_SYNC_CORE_BEFORE_USERMODE
 	select ARCH_HAS_SYSCALL_WRAPPER
 	select ARCH_HAS_TICK_BROADCAST if GENERIC_CLOCKEVENTS_BROADCAST
 	select ARCH_HAS_UBSAN_SANITIZE_ALL

arch/riscv/include/asm/kfence.h

@@ -18,9 +18,9 @@ static inline bool kfence_protect_page(unsigned long addr, bool protect)
 	pte_t *pte = virt_to_kpte(addr);
 
 	if (protect)
-		set_pte(pte, __pte(pte_val(*pte) & ~_PAGE_PRESENT));
+		set_pte(pte, __pte(pte_val(ptep_get(pte)) & ~_PAGE_PRESENT));
 	else
-		set_pte(pte, __pte(pte_val(*pte) | _PAGE_PRESENT));
+		set_pte(pte, __pte(pte_val(ptep_get(pte)) | _PAGE_PRESENT));
 
 	flush_tlb_kernel_range(addr, addr + PAGE_SIZE);
 

arch/riscv/include/asm/membarrier.h

@@ -0,0 +1,50 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+#ifndef _ASM_RISCV_MEMBARRIER_H
+#define _ASM_RISCV_MEMBARRIER_H
+
+static inline void membarrier_arch_switch_mm(struct mm_struct *prev,
+					     struct mm_struct *next,
+					     struct task_struct *tsk)
+{
+	/*
+	 * Only need the full barrier when switching between processes.
+	 * Barrier when switching from kernel to userspace is not
+	 * required here, given that it is implied by mmdrop(). Barrier
+	 * when switching from userspace to kernel is not needed after
+	 * store to rq->curr.
+	 */
+	if (IS_ENABLED(CONFIG_SMP) &&
+	    likely(!(atomic_read(&next->membarrier_state) &
+		     (MEMBARRIER_STATE_PRIVATE_EXPEDITED |
+		      MEMBARRIER_STATE_GLOBAL_EXPEDITED)) || !prev))
+		return;
+
+	/*
+	 * The membarrier system call requires a full memory barrier
+	 * after storing to rq->curr, before going back to user-space.
+	 *
+	 * This barrier is also needed for the SYNC_CORE command when
+	 * switching between processes; in particular, on a transition
+	 * from a thread belonging to another mm to a thread belonging
+	 * to the mm for which a membarrier SYNC_CORE is done on CPU0:
+	 *
+	 *   - [CPU0] sets all bits in the mm icache_stale_mask (in
+	 *     prepare_sync_core_cmd());
+	 *
+	 *   - [CPU1] stores to rq->curr (by the scheduler);
+	 *
+	 *   - [CPU0] loads rq->curr within membarrier and observes
+	 *     cpu_rq(1)->curr->mm != mm, so the IPI is skipped on
+	 *     CPU1; this means membarrier relies on switch_mm() to
+	 *     issue the sync-core;
+	 *
+	 *   - [CPU1] switch_mm() loads icache_stale_mask; if the bit
+	 *     is zero, switch_mm() may incorrectly skip the sync-core.
+	 *
+	 * Matches a full barrier in the proximity of the membarrier
+	 * system call entry.
+	 */
+	smp_mb();
+}
+
+#endif /* _ASM_RISCV_MEMBARRIER_H */

arch/riscv/include/asm/pgtable-64.h

@@ -198,7 +198,7 @@ static inline int pud_user(pud_t pud)
 
 static inline void set_pud(pud_t *pudp, pud_t pud)
 {
-	*pudp = pud;
+	WRITE_ONCE(*pudp, pud);
 }
 
 static inline void pud_clear(pud_t *pudp)
@@ -274,7 +274,7 @@ static inline unsigned long _pmd_pfn(pmd_t pmd)
 static inline void set_p4d(p4d_t *p4dp, p4d_t p4d)
 {
 	if (pgtable_l4_enabled)
-		*p4dp = p4d;
+		WRITE_ONCE(*p4dp, p4d);
 	else
 		set_pud((pud_t *)p4dp, (pud_t){ p4d_val(p4d) });
 }
@@ -336,18 +336,12 @@ static inline struct page *p4d_page(p4d_t p4d)
 #define pud_index(addr) (((addr) >> PUD_SHIFT) & (PTRS_PER_PUD - 1))
 
 #define pud_offset pud_offset
-static inline pud_t *pud_offset(p4d_t *p4d, unsigned long address)
-{
-	if (pgtable_l4_enabled)
-		return p4d_pgtable(*p4d) + pud_index(address);
-
-	return (pud_t *)p4d;
-}
+pud_t *pud_offset(p4d_t *p4d, unsigned long address);
 
 static inline void set_pgd(pgd_t *pgdp, pgd_t pgd)
 {
 	if (pgtable_l5_enabled)
-		*pgdp = pgd;
+		WRITE_ONCE(*pgdp, pgd);
 	else
 		set_p4d((p4d_t *)pgdp, (p4d_t){ pgd_val(pgd) });
 }
@@ -400,12 +394,6 @@ static inline struct page *pgd_page(pgd_t pgd)
 #define p4d_index(addr) (((addr) >> P4D_SHIFT) & (PTRS_PER_P4D - 1))
 
 #define p4d_offset p4d_offset
-static inline p4d_t *p4d_offset(pgd_t *pgd, unsigned long address)
-{
-	if (pgtable_l5_enabled)
-		return pgd_pgtable(*pgd) + p4d_index(address);
-
-	return (p4d_t *)pgd;
-}
+p4d_t *p4d_offset(pgd_t *pgd, unsigned long address);
 
 #endif /* _ASM_RISCV_PGTABLE_64_H */

arch/riscv/include/asm/pgtable.h

@@ -248,7 +248,7 @@ static inline int pmd_leaf(pmd_t pmd)
 
 static inline void set_pmd(pmd_t *pmdp, pmd_t pmd)
 {
-	*pmdp = pmd;
+	WRITE_ONCE(*pmdp, pmd);
 }
 
 static inline void pmd_clear(pmd_t *pmdp)
@@ -515,7 +515,7 @@ static inline int pte_same(pte_t pte_a, pte_t pte_b)
  */
 static inline void set_pte(pte_t *ptep, pte_t pteval)
 {
-	*ptep = pteval;
+	WRITE_ONCE(*ptep, pteval);
 }
 
 void flush_icache_pte(pte_t pte);
@@ -549,19 +549,12 @@ static inline void pte_clear(struct mm_struct *mm,
 	__set_pte_at(ptep, __pte(0));
 }
 
-#define __HAVE_ARCH_PTEP_SET_ACCESS_FLAGS
-static inline int ptep_set_access_flags(struct vm_area_struct *vma,
-					unsigned long address, pte_t *ptep,
-					pte_t entry, int dirty)
-{
-	if (!pte_same(*ptep, entry))
-		__set_pte_at(ptep, entry);
-	/*
-	 * update_mmu_cache will unconditionally execute, handling both
-	 * the case that the PTE changed and the spurious fault case.
-	 */
-	return true;
-}
+#define __HAVE_ARCH_PTEP_SET_ACCESS_FLAGS	/* defined in mm/pgtable.c */
+extern int ptep_set_access_flags(struct vm_area_struct *vma, unsigned long address,
+				 pte_t *ptep, pte_t entry, int dirty);
+#define __HAVE_ARCH_PTEP_TEST_AND_CLEAR_YOUNG	/* defined in mm/pgtable.c */
+extern int ptep_test_and_clear_young(struct vm_area_struct *vma, unsigned long address,
+				     pte_t *ptep);
 
 #define __HAVE_ARCH_PTEP_GET_AND_CLEAR
 static inline pte_t ptep_get_and_clear(struct mm_struct *mm,
@@ -574,16 +567,6 @@ static inline pte_t ptep_get_and_clear(struct mm_struct *mm,
 	return pte;
 }
 
-#define __HAVE_ARCH_PTEP_TEST_AND_CLEAR_YOUNG
-static inline int ptep_test_and_clear_young(struct vm_area_struct *vma,
-					    unsigned long address,
-					    pte_t *ptep)
-{
-	if (!pte_young(*ptep))
-		return 0;
-	return test_and_clear_bit(_PAGE_ACCESSED_OFFSET, &pte_val(*ptep));
-}
-
 #define __HAVE_ARCH_PTEP_SET_WRPROTECT
 static inline void ptep_set_wrprotect(struct mm_struct *mm,
 				      unsigned long address, pte_t *ptep)

arch/riscv/include/asm/sync_core.h

@@ -0,0 +1,29 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+#ifndef _ASM_RISCV_SYNC_CORE_H
+#define _ASM_RISCV_SYNC_CORE_H
+
+/*
+ * RISC-V implements return to user-space through an xRET instruction,
+ * which is not core serializing.
+ */
+static inline void sync_core_before_usermode(void)
+{
+	asm volatile ("fence.i" ::: "memory");
+}
+
+#ifdef CONFIG_SMP
+/*
+ * Ensure the next switch_mm() on every CPU issues a core serializing
+ * instruction for the given @mm.
+ */
+static inline void prepare_sync_core_cmd(struct mm_struct *mm)
+{
+	cpumask_setall(&mm->context.icache_stale_mask);
+}
+#else
+static inline void prepare_sync_core_cmd(struct mm_struct *mm)
+{
+}
+#endif /* CONFIG_SMP */
+
+#endif /* _ASM_RISCV_SYNC_CORE_H */

arch/riscv/kernel/efi.c

@@ -60,7 +60,7 @@ int __init efi_create_mapping(struct mm_struct *mm, efi_memory_desc_t *md)
 static int __init set_permissions(pte_t *ptep, unsigned long addr, void *data)
 {
 	efi_memory_desc_t *md = data;
-	pte_t pte = READ_ONCE(*ptep);
+	pte_t pte = ptep_get(ptep);
 	unsigned long val;
 
 	if (md->attribute & EFI_MEMORY_RO) {