arm64: hibernate: Support DEBUG_PAGEALLOC

DEBUG_PAGEALLOC removes the valid bit of page table entries to prevent
any access to unallocated memory. Hibernate uses this as a hint that those
pages don't need to be saved/restored. This patch adds the
kernel_page_present() function it uses.

hibernate.c copies the resume kernel's linear map for use during restore.
Add _copy_pte() to fill-in the holes made by DEBUG_PAGEALLOC in the resume
kernel, so we can restore data the original kernel had at these addresses.

Finally, DEBUG_PAGEALLOC means the linear-map alias of KERNEL_START to
KERNEL_END may have holes in it, so we can't lazily clean this whole
area to the PoC. Only clean the new mmuoff region, and the kernel/kvm
idmaps.

This reverts commit da24eb1f.
Reported-by: Will Deacon <will.deacon@arm.com>
Signed-off-by: James Morse <james.morse@arm.com>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Reviewed-by: Catalin Marinas <catalin.marinas@arm.com>
Signed-off-by: Will Deacon <will.deacon@arm.com>

arch/arm64/Kconfig

@@ -607,7 +607,6 @@ source kernel/Kconfig.preempt
 source kernel/Kconfig.hz
 
 config ARCH_SUPPORTS_DEBUG_PAGEALLOC
-	depends on !HIBERNATION
 	def_bool y
 
 config ARCH_HAS_HOLES_MEMORYMODEL

arch/arm64/include/asm/pgtable.h

@@ -155,6 +155,16 @@ static inline pte_t pte_mknoncont(pte_t pte)
 	return clear_pte_bit(pte, __pgprot(PTE_CONT));
 }
 
+static inline pte_t pte_clear_rdonly(pte_t pte)
+{
+	return clear_pte_bit(pte, __pgprot(PTE_RDONLY));
+}
+
+static inline pte_t pte_mkpresent(pte_t pte)
+{
+	return set_pte_bit(pte, __pgprot(PTE_VALID));
+}
+
 static inline pmd_t pmd_mkcont(pmd_t pmd)
 {
 	return __pmd(pmd_val(pmd) | PMD_SECT_CONT);

arch/arm64/kernel/hibernate.c

@@ -235,6 +235,7 @@ static int create_safe_exec_page(void *src_start, size_t length,
 	return rc;
 }
 
+#define dcache_clean_range(start, end)	__flush_dcache_area(start, (end - start))
 
 int swsusp_arch_suspend(void)
 {
@@ -252,8 +253,13 @@ int swsusp_arch_suspend(void)
 	if (__cpu_suspend_enter(&state)) {
 		ret = swsusp_save();
 	} else {
-		/* Clean kernel to PoC for secondary core startup */
-		__flush_dcache_area(LMADDR(KERNEL_START), KERNEL_END - KERNEL_START);
+		/* Clean kernel core startup/idle code to PoC*/
+		dcache_clean_range(__mmuoff_data_start, __mmuoff_data_end);
+		dcache_clean_range(__idmap_text_start, __idmap_text_end);
+
+		/* Clean kvm setup code to PoC? */
+		if (el2_reset_needed())
+			dcache_clean_range(__hyp_idmap_text_start, __hyp_idmap_text_end);
 
 		/*
 		 * Tell the hibernation core that we've just restored
@@ -269,6 +275,33 @@ int swsusp_arch_suspend(void)
 	return ret;
 }
 
+static void _copy_pte(pte_t *dst_pte, pte_t *src_pte, unsigned long addr)
+{
+	pte_t pte = *src_pte;
+
+	if (pte_valid(pte)) {
+		/*
+		 * Resume will overwrite areas that may be marked
+		 * read only (code, rodata). Clear the RDONLY bit from
+		 * the temporary mappings we use during restore.
+		 */
+		set_pte(dst_pte, pte_clear_rdonly(pte));
+	} else if (debug_pagealloc_enabled() && !pte_none(pte)) {
+		/*
+		 * debug_pagealloc will removed the PTE_VALID bit if
+		 * the page isn't in use by the resume kernel. It may have
+		 * been in use by the original kernel, in which case we need
+		 * to put it back in our copy to do the restore.
+		 *
+		 * Before marking this entry valid, check the pfn should
+		 * be mapped.
+		 */
+		BUG_ON(!pfn_valid(pte_pfn(pte)));
+
+		set_pte(dst_pte, pte_mkpresent(pte_clear_rdonly(pte)));
+	}
+}
+
 static int copy_pte(pmd_t *dst_pmd, pmd_t *src_pmd, unsigned long start,
 		    unsigned long end)
 {
@@ -284,13 +317,7 @@ static int copy_pte(pmd_t *dst_pmd, pmd_t *src_pmd, unsigned long start,
 
 	src_pte = pte_offset_kernel(src_pmd, start);
 	do {
-		if (!pte_none(*src_pte))
-			/*
-			 * Resume will overwrite areas that may be marked
-			 * read only (code, rodata). Clear the RDONLY bit from
-			 * the temporary mappings we use during restore.
-			 */
-			set_pte(dst_pte, __pte(pte_val(*src_pte) & ~PTE_RDONLY));
+		_copy_pte(dst_pte, src_pte, addr);
 	} while (dst_pte++, src_pte++, addr += PAGE_SIZE, addr != end);
 
 	return 0;

arch/arm64/mm/pageattr.c

@@ -139,4 +139,43 @@ void __kernel_map_pages(struct page *page, int numpages, int enable)
 					__pgprot(0),
 					__pgprot(PTE_VALID));
 }
-#endif
+#ifdef CONFIG_HIBERNATION
+/*
+ * When built with CONFIG_DEBUG_PAGEALLOC and CONFIG_HIBERNATION, this function
+ * is used to determine if a linear map page has been marked as not-valid by
+ * CONFIG_DEBUG_PAGEALLOC. Walk the page table and check the PTE_VALID bit.
+ * This is based on kern_addr_valid(), which almost does what we need.
+ *
+ * Because this is only called on the kernel linear map,  p?d_sect() implies
+ * p?d_present(). When debug_pagealloc is enabled, sections mappings are
+ * disabled.
+ */
+bool kernel_page_present(struct page *page)
+{
+	pgd_t *pgd;
+	pud_t *pud;
+	pmd_t *pmd;
+	pte_t *pte;
+	unsigned long addr = (unsigned long)page_address(page);
+
+	pgd = pgd_offset_k(addr);
+	if (pgd_none(*pgd))
+		return false;
+
+	pud = pud_offset(pgd, addr);
+	if (pud_none(*pud))
+		return false;
+	if (pud_sect(*pud))
+		return true;
+
+	pmd = pmd_offset(pud, addr);
+	if (pmd_none(*pmd))
+		return false;
+	if (pmd_sect(*pmd))
+		return true;
+
+	pte = pte_offset_kernel(pmd, addr);
+	return pte_valid(*pte);
+}
+#endif /* CONFIG_HIBERNATION */
+#endif /* CONFIG_DEBUG_PAGEALLOC */