linux-kernel - [PATCH] x86/mm: Define PTRS_PER

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: <Z8oa8AUVyi2HWfo9@gmail.com>
Date: Thu, 6 Mar 2025 23:00:16 +0100
From: Ingo Molnar <mingo@...nel.org>
To: Linus Torvalds <torvalds@...ux-foundation.org>
Cc: linux-kernel@...r.kernel.org, linux-tip-commits@...r.kernel.org,
	kernel test robot <lkp@...el.com>,
	Andy Shevchenko <andriy.shevchenko@...ux.intel.com>, x86@...nel.org
Subject: [PATCH] x86/mm: Define PTRS_PER_PMD for assembly code too


* Ingo Molnar <mingo@...nel.org> wrote:

> Separating out the assembler-compatible defines from the types 
> headers appears to be a bigger patch, since it's all mixed in with C 
> syntax:
> 
> <=-----------------------------------===============================
> typedef struct { pud_t pud; } pmd_t;
> 
> #define PMD_SHIFT       PUD_SHIFT
> #define PTRS_PER_PMD    1
> #define PMD_SIZE        (1UL << PMD_SHIFT)
> #define PMD_MASK        (~(PMD_SIZE-1))
> 
> /*
>  * The "pud_xxx()" functions here are trivial for a folded two-level
>  * setup: the pmd is never bad, and a pmd always exists (as it's folded
>  * into the pud entry)
>  */
> static inline int pud_none(pud_t pud)           { return 0; }
> static inline int pud_bad(pud_t pud)            { return 0; }
> static inline int pud_present(pud_t pud)        { return 1; }
> ================================================================>
> 
> In any case I've removed the commit for the time being until this all 
> is cleared up.

So there's a simple solution: define it on i386 too, via the patch 
below. It appears the double-definition doesn't create any warnings, on 
GCC at least.

But if it's an issue, we could do something like this in 
<asm-generic/pgtable-nopmd.h>:

 #if defined(PTRS_PER_PMD) && (PTRS_PER_PMD != 1)
 # error "mm: Wait a minute, that's a super confusing pagetable setup ..."
 #endif

?

Thanks,

	Ingo

=========================>
From: Ingo Molnar <mingo@...nel.org>
Date: Thu, 6 Mar 2025 22:53:49 +0100
Subject: [PATCH] x86/mm: Define PTRS_PER_PMD for assembly code too

Andy reported the following build warning from head_32.S:

  In file included from arch/x86/kernel/head_32.S:29:
  arch/x86/include/asm/pgtable_32.h:59:5: error: "PTRS_PER_PMD" is not defined, evaluates to 0 [-Werror=undef]
       59 | #if PTRS_PER_PMD > 1

The reason is that on 2-level i386 paging the folded in PMD's
PTRS_PER_PMD constant is not defined in assembly headers,
only in generic MM C headers.

Instead of trying to fish out the definition from the generic
headers, just define it - it even has a comment for it already...

Reported-by: Andy Shevchenko <andriy.shevchenko@...ux.intel.com>
Signed-off-by: Ingo Molnar <mingo@...nel.org>
---
 arch/x86/include/asm/pgtable-2level_types.h | 8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)

diff --git a/arch/x86/include/asm/pgtable-2level_types.h b/arch/x86/include/asm/pgtable-2level_types.h
index 7f6ccff0ba72..4a12c276b181 100644
--- a/arch/x86/include/asm/pgtable-2level_types.h
+++ b/arch/x86/include/asm/pgtable-2level_types.h
@@ -23,17 +23,17 @@ typedef union {
 #define ARCH_PAGE_TABLE_SYNC_MASK	PGTBL_PMD_MODIFIED
 
 /*
- * traditional i386 two-level paging structure:
+ * Traditional i386 two-level paging structure:
  */
 
 #define PGDIR_SHIFT	22
 #define PTRS_PER_PGD	1024
 
-
 /*
- * the i386 is two-level, so we don't really have any
- * PMD directory physically.
+ * The i386 is two-level, so we don't really have any
+ * PMD directory physically:
  */
+#define PTRS_PER_PMD	1
 
 #define PTRS_PER_PTE	1024