[PATCH PTI v3 03/10] x86/pti/64: Fix ESPFIX64 user mapping

[Andy Lutomirski <luto@kernel.org>]

The ESPFIX64 user mapping belongs in pti.c just like all the other
user mappings.  Move it there and make it work correctly while we're
at it.

Signed-off-by: Andy Lutomirski <luto@kernel.org>
---
 arch/x86/kernel/espfix_64.c | 16 ----------------
 arch/x86/mm/pti.c           | 42 +++++++++++++++++++++++++++++++++++++-----
 init/main.c                 | 11 +++++++----
 3 files changed, 44 insertions(+), 25 deletions(-)

diff --git a/arch/x86/kernel/espfix_64.c b/arch/x86/kernel/espfix_64.c
index 1c44e72ed1bc..9c4e7ba6870c 100644
--- a/arch/x86/kernel/espfix_64.c
+++ b/arch/x86/kernel/espfix_64.c
@@ -129,22 +129,6 @@ void __init init_espfix_bsp(void)
 	p4d = p4d_alloc(&init_mm, pgd, ESPFIX_BASE_ADDR);
 	p4d_populate(&init_mm, p4d, espfix_pud_page);
 
-	/*
-	 * Just copy the top-level PGD that is mapping the espfix area to
-	 * ensure it is mapped into the user page tables.
-	 *
-	 * For 5-level paging, the espfix pgd was populated when
-	 * pti_init() pre-populated all the pgd entries.  The above
-	 * p4d_alloc() would never do anything and the p4d_populate() would
-	 * be done to a p4d already mapped in the userspace pgd.
-	 */
-#ifdef CONFIG_PAGE_TABLE_ISOLATION
-	if (CONFIG_PGTABLE_LEVELS <= 4) {
-		set_pgd(kernel_to_user_pgdp(pgd),
-			__pgd(_KERNPG_TABLE | (p4d_pfn(*p4d) << PAGE_SHIFT)));
-	}
-#endif
-
 	/* Randomize the locations */
 	init_espfix_random();
 
diff --git a/arch/x86/mm/pti.c b/arch/x86/mm/pti.c
index f48645d2f3fd..e01c4aa3ec73 100644
--- a/arch/x86/mm/pti.c
+++ b/arch/x86/mm/pti.c
@@ -68,14 +68,12 @@ void __init pti_check_boottime_disable(void)
  * Walk the user copy of the page tables (optionally) trying to allocate
  * page table pages on the way down.
  *
- * Returns a pointer to a PMD on success, or NULL on failure.
+ * Returns a pointer to a P4D on success, or NULL on failure.
  */
-static pmd_t *pti_user_pagetable_walk_pmd(unsigned long address)
+static p4d_t *pti_user_pagetable_walk_p4d(unsigned long address)
 {
 	pgd_t *pgd = kernel_to_user_pgdp(pgd_offset_k(address));
 	gfp_t gfp = (GFP_KERNEL | __GFP_NOTRACK | __GFP_ZERO);
-	pud_t *pud;
-	p4d_t *p4d;
 
 	if (address < PAGE_OFFSET) {
 		WARN_ONCE(1, "attempt to walk user address\n");
@@ -96,7 +94,21 @@ static pmd_t *pti_user_pagetable_walk_pmd(unsigned long address)
 	}
 	BUILD_BUG_ON(pgd_large(*pgd) != 0);
 
-	p4d = p4d_offset(pgd, address);
+	return p4d_offset(pgd, address);
+}
+
+/*
+ * Walk the user copy of the page tables (optionally) trying to allocate
+ * page table pages on the way down.
+ *
+ * Returns a pointer to a PMD on success, or NULL on failure.
+ */
+static pmd_t *pti_user_pagetable_walk_pmd(unsigned long address)
+{
+	gfp_t gfp = (GFP_KERNEL | __GFP_NOTRACK | __GFP_ZERO);
+	p4d_t *p4d = pti_user_pagetable_walk_p4d(address);
+	pud_t *pud;
+
 	BUILD_BUG_ON(p4d_large(*p4d) != 0);
 	if (p4d_none(*p4d)) {
 		unsigned long new_pud_page = __get_free_page(gfp);
@@ -174,6 +186,25 @@ pti_clone_pmds(unsigned long start, unsigned long end, pmdval_t clear)
 	}
 }
 
+static void __init pti_setup_espfix64(void)
+{
+#ifdef CONFIG_X86_ESPFIX64
+	/*
+	 * ESPFIX64 uses a single p4d (i.e. a top-level entry on 4-level
+	 * systems and a next-level entry on 5-level systems.  Share that
+	 * entry between the user and kernel pagetables.
+	 */
+	pgd_t *kernel_pgd;
+	p4d_t *kernel_p4d, *user_p4d;
+
+	pr_err("espfix64 base = %lx\n", ESPFIX_BASE_ADDR);
+	user_p4d = pti_user_pagetable_walk_p4d(ESPFIX_BASE_ADDR);
+	kernel_pgd = pgd_offset_k(ESPFIX_BASE_ADDR);
+	kernel_p4d = p4d_offset(kernel_pgd, ESPFIX_BASE_ADDR);
+	*user_p4d = *kernel_p4d;
+#endif
+}
+
 /*
  * Clone the populated PMDs of the user shared fixmaps into the user space
  * visible page table.
@@ -212,4 +243,5 @@ void __init pti_init(void)
 
 	pti_clone_user_shared();
 	pti_clone_entry_text();
+	pti_setup_espfix64();
 }
diff --git a/init/main.c b/init/main.c
index 64b00b89e9e1..ce18b938c382 100644
--- a/init/main.c
+++ b/init/main.c
@@ -505,6 +505,13 @@ static void __init mm_init(void)
 	pgtable_init();
 	vmalloc_init();
 	ioremap_huge_init();
+
+#ifdef CONFIG_X86_ESPFIX64
+	/* Should be run before the first non-init thread is created */
+	init_espfix_bsp();
+#endif
+
+	/* Should be run after espfix64 is set up. */
 	pti_init();
 }
 
@@ -676,10 +683,6 @@ asmlinkage __visible void __init start_kernel(void)
 	if (efi_enabled(EFI_RUNTIME_SERVICES))
 		efi_enter_virtual_mode();
 #endif
-#ifdef CONFIG_X86_ESPFIX64
-	/* Should be run before the first non-init thread is created */
-	init_espfix_bsp();
-#endif
 	thread_stack_cache_init();
 	cred_init();
 	fork_init();
-- 
2.13.6