Re: [PATCH RFC 01/11] mm: introduce hw_pte_t for PTE table storage

On 31/07/2026 6:49 am, Anshuman Khandual wrote:

On Mon, Jul 27, 2026 at 05:46:52PM +0100, Muhammad Usama Anjum wrote:

quoted

pte_t is used both for logical PTE values and for entries stored in a PTE
table, so pte_t * does not distinguish a pointer to a copied value from a
pointer to table storage.

Should we rather call them as "Linux PTE" and "HW PTE" instead ?

Some like "Logical PTE" does not really sound very clear on what it represents.

Yeah, I should make distinction more visible in the cover letter.

quoted

Introduce hw_pte_t as the generic name for a PTE table element. Define it
as an alias of pte_t by default so architectures retain their existing
representation and behavior until they opt in to a distinct storage type.

Makes sense.

quoted

Include asm/page.h so pte_t is defined before the generic alias.

Signed-off-by: Muhammad Usama Anjum <redacted>
---
 MAINTAINERS                   |  1 +
 include/linux/pgtable_types.h | 11 +++++++++++
 2 files changed, 12 insertions(+)
 create mode 100644 include/linux/pgtable_types.h

diff --git a/MAINTAINERS b/MAINTAINERS
index 15dd00c7ffec5..78aac05949829 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS

@@ -16981,6 +16981,7 @@ F:	include/linux/mmu_notifier.h
 F:	include/linux/pagewalk.h
 F:	include/linux/pgalloc.h
 F:	include/linux/pgtable.h
+F:	include/linux/pgtable_types.h

s/pgtable_types.h/pgtable_hw_types.h ?

I prefer shorter file names.

quoted

 F:	include/linux/ptdump.h
 F:	include/linux/vmpressure.h
 F:	include/linux/vmstat.h

diff --git a/include/linux/pgtable_types.h b/include/linux/pgtable_types.h
new file mode 100644
index 0000000000000..ed5dba42a4f8c
--- /dev/null
+++ b/include/linux/pgtable_types.h

@@ -0,0 +1,11 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+#ifndef _LINUX_PGTABLE_TYPES_H
+#define _LINUX_PGTABLE_TYPES_H
+
+#include <asm/page.h>
+
+#ifndef __HAVE_ARCH_HW_PTE_T

__HAVE_ARCH_HW_PTE_T macro is a better approach here or a config ARCH_HAS_HW_PTE_T
based selective platform subscription should be introduced instead ?

Moving to CONFIG_ARCH_HAS_PTE_T from the next version as suggested here [1].

[1] https://lore.kernel.org/all/78a0402d-5436-4aab-a704-90fab56eec50@kernel.org (local) 

-- 
Thanks,
Usama