linux-kernel - [PATCH 1/1] x86/vmemmap: Use direct-mapped VA instead of vmemmap-based VA

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <20250217114133.400063-2-gwan-gyeong.mun@intel.com>
Date: Mon, 17 Feb 2025 13:41:33 +0200
From: Gwan-gyeong Mun <gwan-gyeong.mun@...el.com>
To: linux-kernel@...r.kernel.org
Cc: osalvador@...e.de,
	42.hyeyoo@...il.com,
	byungchul@...com,
	dave.hansen@...ux.intel.com,
	luto@...nel.org,
	peterz@...radead.org,
	akpm@...ux-foundation.org,
	max.byungchul.park@...com,
	max.byungchul.park@...il.com
Subject: [PATCH 1/1] x86/vmemmap: Use direct-mapped VA instead of vmemmap-based VA

When a process leads the addition of a struct page to vmemmap
(e.g. hot-plug), the page table update for the newly added vmemmap-based
virtual address is updated first in init_mm's page table and then
synchronized later.
If the vmemmap-based virtual address is accessed through the process's
page table before this sync, a page fault will occur.

This translates vmemmap-based virtual address to direct-mapped virtual
address and use it, if the current top-level page table is not init_mm's
page table when accessing a vmemmap-based virtual address before this sync.

Fixes: faf1c0008a33 ("x86/vmemmap: optimize for consecutive sections in partial populated PMDs")
Signed-off-by: Gwan-gyeong Mun <gwan-gyeong.mun@...el.com>
Cc: Oscar Salvador <osalvador@...e.de>
Cc: Hyeonggon Yoo <42.hyeyoo@...il.com>
Cc: Byungchul Park <byungchul@...com>
Cc: Dave Hansen <dave.hansen@...ux.intel.com>
Cc: Andy Lutomirski <luto@...nel.org>
Cc: Peter Zijlstra <peterz@...radead.org>
Cc: Andrew Morton <akpm@...ux-foundation.org>
---
 arch/x86/mm/init_64.c | 17 ++++++++++++++---
 1 file changed, 14 insertions(+), 3 deletions(-)

diff --git a/arch/x86/mm/init_64.c b/arch/x86/mm/init_64.c
index 01ea7c6df303..1cb26f692831 100644
--- a/arch/x86/mm/init_64.c
+++ b/arch/x86/mm/init_64.c
@@ -844,6 +844,17 @@ void __init paging_init(void)
  */
 static unsigned long unused_pmd_start __meminitdata;
 
+static void * __meminit vmemmap_va_to_kaddr(unsigned long vmemmap_va)
+{
+	void *kaddr = (void *)vmemmap_va;
+	pgd_t *pgd = __va(read_cr3_pa());
+
+	if (init_mm.pgd != pgd)
+		kaddr = __va(slow_virt_to_phys(kaddr));
+
+	return kaddr;
+}
+
 static void __meminit vmemmap_flush_unused_pmd(void)
 {
 	if (!unused_pmd_start)
@@ -851,7 +862,7 @@ static void __meminit vmemmap_flush_unused_pmd(void)
 	/*
 	 * Clears (unused_pmd_start, PMD_END]
 	 */
-	memset((void *)unused_pmd_start, PAGE_UNUSED,
+	memset(vmemmap_va_to_kaddr(unused_pmd_start), PAGE_UNUSED,
 	       ALIGN(unused_pmd_start, PMD_SIZE) - unused_pmd_start);
 	unused_pmd_start = 0;
 }
@@ -882,7 +893,7 @@ static void __meminit __vmemmap_use_sub_pmd(unsigned long start)
 	 * case the first memmap never gets initialized e.g., because the memory
 	 * block never gets onlined).
 	 */
-	memset((void *)start, 0, sizeof(struct page));
+	memset(vmemmap_va_to_kaddr(start), 0, sizeof(struct page));
 }
 
 static void __meminit vmemmap_use_sub_pmd(unsigned long start, unsigned long end)
@@ -924,7 +935,7 @@ static void __meminit vmemmap_use_new_sub_pmd(unsigned long start, unsigned long
 	 * Mark with PAGE_UNUSED the unused parts of the new memmap range
 	 */
 	if (!IS_ALIGNED(start, PMD_SIZE))
-		memset((void *)page, PAGE_UNUSED, start - page);
+		memset(vmemmap_va_to_kaddr(page), PAGE_UNUSED, start - page);
 
 	/*
 	 * We want to avoid memset(PAGE_UNUSED) when populating the vmemmap of
-- 
2.48.1