mirror of https://lore.kernel.org/lkml/
 help / color / mirror / Atom feed
* [PATCH 0/3] x86: Make 5-level paging support unconditional for x86-64
@ 2024-06-21 16:44 Kirill A. Shutemov
  2024-06-21 16:44 ` [PATCH 1/3] x86/64/mm: Always use dynamic memory layout Kirill A. Shutemov
                   ` (2 more replies)
  0 siblings, 3 replies; 15+ messages in thread
From: Kirill A. Shutemov @ 2024-06-21 16:44 UTC (permalink / raw)
  To: Thomas Gleixner, Ingo Molnar, Borislav Petkov, Dave Hansen, x86,
	H. Peter Anvin
  Cc: Jonathan Corbet, Andy Lutomirski, Peter Zijlstra, Ard Biesheuvel,
	Jan Kiszka, Kieran Bingham, Kirill A. Shutemov, Michael Roth,
	Rick Edgecombe, Brijesh Singh, Sandipan Das, Juergen Gross,
	Tom Lendacky, linux-kernel, linux-doc, linux-efi, linux-mm

Both Intel and AMD CPUs support 5-level paging, which is expected to
become more widely adopted in the future.

Remove CONFIG_X86_5LEVEL.

In preparation to that remove CONFIG_DYNAMIC_MEMORY_LAYOUT and make
SPARSEMEM_VMEMMAP the only memory model.

Kirill A. Shutemov (3):
  x86/64/mm: Always use dynamic memory layout
  x86/64/mm: Make SPARSEMEM_VMEMMAP the only memory model
  x86/64/mm: Make 5-level paging support unconditional

 Documentation/arch/x86/cpuinfo.rst            |  8 ++---
 .../arch/x86/x86_64/5level-paging.rst         |  9 ------
 arch/x86/Kconfig                              | 32 ++-----------------
 arch/x86/boot/compressed/pgtable_64.c         | 11 ++-----
 arch/x86/boot/header.S                        |  4 ---
 arch/x86/include/asm/disabled-features.h      |  9 +-----
 arch/x86/include/asm/page_64.h                |  2 --
 arch/x86/include/asm/page_64_types.h          | 11 -------
 arch/x86/include/asm/pgtable_64_types.h       | 24 --------------
 arch/x86/kernel/alternative.c                 |  2 +-
 arch/x86/kernel/head64.c                      |  7 ----
 arch/x86/kernel/head_64.S                     |  2 --
 arch/x86/mm/init.c                            |  4 ---
 arch/x86/mm/init_64.c                         |  9 +-----
 arch/x86/mm/pgtable.c                         |  2 --
 drivers/firmware/efi/libstub/x86-5lvl.c       |  2 +-
 scripts/gdb/linux/pgtable.py                  |  4 +--
 .../arch/x86/include/asm/disabled-features.h  |  9 +-----
 18 files changed, 13 insertions(+), 138 deletions(-)

-- 
2.43.0


^ permalink raw reply	[flat|nested] 15+ messages in thread

* [PATCH 1/3] x86/64/mm: Always use dynamic memory layout
  2024-06-21 16:44 [PATCH 0/3] x86: Make 5-level paging support unconditional for x86-64 Kirill A. Shutemov
@ 2024-06-21 16:44 ` Kirill A. Shutemov
  2024-06-21 16:44 ` [PATCH 2/3] x86/64/mm: Make SPARSEMEM_VMEMMAP the only memory model Kirill A. Shutemov
  2024-06-21 16:44 ` [PATCH 3/3] x86/64/mm: Make 5-level paging support unconditional Kirill A. Shutemov
  2 siblings, 0 replies; 15+ messages in thread
From: Kirill A. Shutemov @ 2024-06-21 16:44 UTC (permalink / raw)
  To: Thomas Gleixner, Ingo Molnar, Borislav Petkov, Dave Hansen, x86,
	H. Peter Anvin
  Cc: Jonathan Corbet, Andy Lutomirski, Peter Zijlstra, Ard Biesheuvel,
	Jan Kiszka, Kieran Bingham, Kirill A. Shutemov, Michael Roth,
	Rick Edgecombe, Brijesh Singh, Sandipan Das, Juergen Gross,
	Tom Lendacky, linux-kernel, linux-doc, linux-efi, linux-mm

Dynamic memory layout is used by KASLR and 5-level paging.

CONFIG_X86_5LEVEL is going to be removed, making 5-level paging support
unconditional which requires unconditional support of dynamic memory
layout.

Remove CONFIG_DYNAMIC_MEMORY_LAYOUT.

Signed-off-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
---
 arch/x86/Kconfig                        | 8 --------
 arch/x86/include/asm/page_64_types.h    | 4 ----
 arch/x86/include/asm/pgtable_64_types.h | 6 ------
 arch/x86/kernel/head64.c                | 2 --
 scripts/gdb/linux/pgtable.py            | 4 +---
 5 files changed, 1 insertion(+), 23 deletions(-)

diff --git a/arch/x86/Kconfig b/arch/x86/Kconfig
index e30ea4129d2c..827928680ed6 100644
--- a/arch/x86/Kconfig
+++ b/arch/x86/Kconfig
@@ -1501,7 +1501,6 @@ config X86_PAE
 config X86_5LEVEL
 	bool "Enable 5-level page tables support"
 	default y
-	select DYNAMIC_MEMORY_LAYOUT
 	select SPARSEMEM_VMEMMAP
 	depends on X86_64
 	help
@@ -2237,17 +2236,10 @@ config PHYSICAL_ALIGN
 
 	  Don't change this unless you know what you are doing.
 
-config DYNAMIC_MEMORY_LAYOUT
-	bool
-	help
-	  This option makes base addresses of vmalloc and vmemmap as well as
-	  __PAGE_OFFSET movable during boot.
-
 config RANDOMIZE_MEMORY
 	bool "Randomize the kernel memory sections"
 	depends on X86_64
 	depends on RANDOMIZE_BASE
-	select DYNAMIC_MEMORY_LAYOUT
 	default RANDOMIZE_BASE
 	help
 	  Randomizes the base virtual address of kernel memory sections
diff --git a/arch/x86/include/asm/page_64_types.h b/arch/x86/include/asm/page_64_types.h
index 06ef25411d62..c2f3c50a2787 100644
--- a/arch/x86/include/asm/page_64_types.h
+++ b/arch/x86/include/asm/page_64_types.h
@@ -41,11 +41,7 @@
 #define __PAGE_OFFSET_BASE_L5	_AC(0xff11000000000000, UL)
 #define __PAGE_OFFSET_BASE_L4	_AC(0xffff888000000000, UL)
 
-#ifdef CONFIG_DYNAMIC_MEMORY_LAYOUT
 #define __PAGE_OFFSET           page_offset_base
-#else
-#define __PAGE_OFFSET           __PAGE_OFFSET_BASE_L4
-#endif /* CONFIG_DYNAMIC_MEMORY_LAYOUT */
 
 #define __START_KERNEL_map	_AC(0xffffffff80000000, UL)
 
diff --git a/arch/x86/include/asm/pgtable_64_types.h b/arch/x86/include/asm/pgtable_64_types.h
index 9053dfe9fa03..09df8939b997 100644
--- a/arch/x86/include/asm/pgtable_64_types.h
+++ b/arch/x86/include/asm/pgtable_64_types.h
@@ -130,15 +130,9 @@ extern unsigned int ptrs_per_p4d;
 #define __VMEMMAP_BASE_L4	0xffffea0000000000UL
 #define __VMEMMAP_BASE_L5	0xffd4000000000000UL
 
-#ifdef CONFIG_DYNAMIC_MEMORY_LAYOUT
 # define VMALLOC_START		vmalloc_base
 # define VMALLOC_SIZE_TB	(pgtable_l5_enabled() ? VMALLOC_SIZE_TB_L5 : VMALLOC_SIZE_TB_L4)
 # define VMEMMAP_START		vmemmap_base
-#else
-# define VMALLOC_START		__VMALLOC_BASE_L4
-# define VMALLOC_SIZE_TB	VMALLOC_SIZE_TB_L4
-# define VMEMMAP_START		__VMEMMAP_BASE_L4
-#endif /* CONFIG_DYNAMIC_MEMORY_LAYOUT */
 
 /*
  * End of the region for which vmalloc page tables are pre-allocated.
diff --git a/arch/x86/kernel/head64.c b/arch/x86/kernel/head64.c
index a817ed0724d1..ec36ad7117ae 100644
--- a/arch/x86/kernel/head64.c
+++ b/arch/x86/kernel/head64.c
@@ -60,14 +60,12 @@ unsigned int ptrs_per_p4d __ro_after_init = 1;
 EXPORT_SYMBOL(ptrs_per_p4d);
 #endif
 
-#ifdef CONFIG_DYNAMIC_MEMORY_LAYOUT
 unsigned long page_offset_base __ro_after_init = __PAGE_OFFSET_BASE_L4;
 EXPORT_SYMBOL(page_offset_base);
 unsigned long vmalloc_base __ro_after_init = __VMALLOC_BASE_L4;
 EXPORT_SYMBOL(vmalloc_base);
 unsigned long vmemmap_base __ro_after_init = __VMEMMAP_BASE_L4;
 EXPORT_SYMBOL(vmemmap_base);
-#endif
 
 static inline bool check_la57_support(void)
 {
diff --git a/scripts/gdb/linux/pgtable.py b/scripts/gdb/linux/pgtable.py
index 30d837f3dfae..09aac2421fb8 100644
--- a/scripts/gdb/linux/pgtable.py
+++ b/scripts/gdb/linux/pgtable.py
@@ -29,11 +29,9 @@ def page_mask(level=1):
         raise Exception(f'Unknown page level: {level}')
 
 
-#page_offset_base in case CONFIG_DYNAMIC_MEMORY_LAYOUT is disabled
-POB_NO_DYNAMIC_MEM_LAYOUT = '0xffff888000000000'
 def _page_offset_base():
     pob_symbol = gdb.lookup_global_symbol('page_offset_base')
-    pob = pob_symbol.name if pob_symbol else POB_NO_DYNAMIC_MEM_LAYOUT
+    pob = pob_symbol.name
     return gdb.parse_and_eval(pob)
 
 
-- 
2.43.0


^ permalink raw reply	[flat|nested] 15+ messages in thread

* [PATCH 2/3] x86/64/mm: Make SPARSEMEM_VMEMMAP the only memory model
  2024-06-21 16:44 [PATCH 0/3] x86: Make 5-level paging support unconditional for x86-64 Kirill A. Shutemov
  2024-06-21 16:44 ` [PATCH 1/3] x86/64/mm: Always use dynamic memory layout Kirill A. Shutemov
@ 2024-06-21 16:44 ` Kirill A. Shutemov
  2024-06-21 16:44 ` [PATCH 3/3] x86/64/mm: Make 5-level paging support unconditional Kirill A. Shutemov
  2 siblings, 0 replies; 15+ messages in thread
From: Kirill A. Shutemov @ 2024-06-21 16:44 UTC (permalink / raw)
  To: Thomas Gleixner, Ingo Molnar, Borislav Petkov, Dave Hansen, x86,
	H. Peter Anvin
  Cc: Jonathan Corbet, Andy Lutomirski, Peter Zijlstra, Ard Biesheuvel,
	Jan Kiszka, Kieran Bingham, Kirill A. Shutemov, Michael Roth,
	Rick Edgecombe, Brijesh Singh, Sandipan Das, Juergen Gross,
	Tom Lendacky, linux-kernel, linux-doc, linux-efi, linux-mm

5-level paging only supports SPARSEMEM_VMEMMAP. CONFIG_X86_5LEVEL is
being phased out, making 5-level paging support mandatory.

Make CONFIG_SPARSEMEM_VMEMMAP mandatory for x86-64 and eliminate
any associated conditional statements.

Signed-off-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
---
 arch/x86/Kconfig      | 2 +-
 arch/x86/mm/init_64.c | 9 +--------
 2 files changed, 2 insertions(+), 9 deletions(-)

diff --git a/arch/x86/Kconfig b/arch/x86/Kconfig
index 827928680ed6..54ad2462e9ef 100644
--- a/arch/x86/Kconfig
+++ b/arch/x86/Kconfig
@@ -1501,7 +1501,6 @@ config X86_PAE
 config X86_5LEVEL
 	bool "Enable 5-level page tables support"
 	default y
-	select SPARSEMEM_VMEMMAP
 	depends on X86_64
 	help
 	  5-level paging enables access to larger address space:
@@ -1625,6 +1624,7 @@ config ARCH_SPARSEMEM_ENABLE
 	depends on X86_64 || NUMA || X86_32 || X86_32_NON_STANDARD
 	select SPARSEMEM_STATIC if X86_32
 	select SPARSEMEM_VMEMMAP_ENABLE if X86_64
+	select SPARSEMEM_VMEMMAP if X86_64
 
 config ARCH_SPARSEMEM_DEFAULT
 	def_bool X86_64 || (NUMA && X86_32)
diff --git a/arch/x86/mm/init_64.c b/arch/x86/mm/init_64.c
index 28002cc7a37d..552a11d5829a 100644
--- a/arch/x86/mm/init_64.c
+++ b/arch/x86/mm/init_64.c
@@ -835,7 +835,6 @@ void __init paging_init(void)
 	zone_sizes_init();
 }
 
-#ifdef CONFIG_SPARSEMEM_VMEMMAP
 #define PAGE_UNUSED 0xFD
 
 /*
@@ -934,7 +933,6 @@ static void __meminit vmemmap_use_new_sub_pmd(unsigned long start, unsigned long
 	if (!IS_ALIGNED(end, PMD_SIZE))
 		unused_pmd_start = end;
 }
-#endif
 
 /*
  * Memory hotplug specific functions
@@ -1133,16 +1131,13 @@ remove_pmd_table(pmd_t *pmd_start, unsigned long addr, unsigned long end,
 				pmd_clear(pmd);
 				spin_unlock(&init_mm.page_table_lock);
 				pages++;
-			}
-#ifdef CONFIG_SPARSEMEM_VMEMMAP
-			else if (vmemmap_pmd_is_unused(addr, next)) {
+			} else if (vmemmap_pmd_is_unused(addr, next)) {
 					free_hugepage_table(pmd_page(*pmd),
 							    altmap);
 					spin_lock(&init_mm.page_table_lock);
 					pmd_clear(pmd);
 					spin_unlock(&init_mm.page_table_lock);
 			}
-#endif
 			continue;
 		}
 
@@ -1490,7 +1485,6 @@ unsigned long memory_block_size_bytes(void)
 	return memory_block_size_probed;
 }
 
-#ifdef CONFIG_SPARSEMEM_VMEMMAP
 /*
  * Initialise the sparsemem vmemmap using huge-pages at the PMD level.
  */
@@ -1639,4 +1633,3 @@ void __meminit vmemmap_populate_print_last(void)
 		node_start = 0;
 	}
 }
-#endif
-- 
2.43.0


^ permalink raw reply	[flat|nested] 15+ messages in thread

* [PATCH 3/3] x86/64/mm: Make 5-level paging support unconditional
  2024-06-21 16:44 [PATCH 0/3] x86: Make 5-level paging support unconditional for x86-64 Kirill A. Shutemov
  2024-06-21 16:44 ` [PATCH 1/3] x86/64/mm: Always use dynamic memory layout Kirill A. Shutemov
  2024-06-21 16:44 ` [PATCH 2/3] x86/64/mm: Make SPARSEMEM_VMEMMAP the only memory model Kirill A. Shutemov
@ 2024-06-21 16:44 ` Kirill A. Shutemov
  2024-06-25 23:21   ` kernel test robot
  2 siblings, 1 reply; 15+ messages in thread
From: Kirill A. Shutemov @ 2024-06-21 16:44 UTC (permalink / raw)
  To: Thomas Gleixner, Ingo Molnar, Borislav Petkov, Dave Hansen, x86,
	H. Peter Anvin
  Cc: Jonathan Corbet, Andy Lutomirski, Peter Zijlstra, Ard Biesheuvel,
	Jan Kiszka, Kieran Bingham, Kirill A. Shutemov, Michael Roth,
	Rick Edgecombe, Brijesh Singh, Sandipan Das, Juergen Gross,
	Tom Lendacky, linux-kernel, linux-doc, linux-efi, linux-mm

Both Intel and AMD CPUs support 5-level paging, which is expected to
become more widely adopted in the future.

Remove CONFIG_X86_5LEVEL and ifdeffery for it to make it more readable.

Signed-off-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Suggested-by: Borislav Petkov <bp@alien8.de>
---
 Documentation/arch/x86/cpuinfo.rst            |  8 +++----
 .../arch/x86/x86_64/5level-paging.rst         |  9 --------
 arch/x86/Kconfig                              | 22 +------------------
 arch/x86/boot/compressed/pgtable_64.c         | 11 ++--------
 arch/x86/boot/header.S                        |  4 ----
 arch/x86/include/asm/disabled-features.h      |  9 +-------
 arch/x86/include/asm/page_64.h                |  2 --
 arch/x86/include/asm/page_64_types.h          |  7 ------
 arch/x86/include/asm/pgtable_64_types.h       | 18 ---------------
 arch/x86/kernel/alternative.c                 |  2 +-
 arch/x86/kernel/head64.c                      |  5 -----
 arch/x86/kernel/head_64.S                     |  2 --
 arch/x86/mm/init.c                            |  4 ----
 arch/x86/mm/pgtable.c                         |  2 --
 drivers/firmware/efi/libstub/x86-5lvl.c       |  2 +-
 .../arch/x86/include/asm/disabled-features.h  |  9 +-------
 16 files changed, 10 insertions(+), 106 deletions(-)

diff --git a/Documentation/arch/x86/cpuinfo.rst b/Documentation/arch/x86/cpuinfo.rst
index 8895784d4784..0ea70924c89e 100644
--- a/Documentation/arch/x86/cpuinfo.rst
+++ b/Documentation/arch/x86/cpuinfo.rst
@@ -171,10 +171,10 @@ For example, when an old kernel is running on new hardware.
 
 c: The kernel disabled support for it at compile-time.
 ------------------------------------------------------
-For example, if 5-level-paging is not enabled when building (i.e.,
-CONFIG_X86_5LEVEL is not selected) the flag "la57" will not show up [#f1]_.
+For example, if Linear Address Masking (LAM) is not enabled when building (i.e.,
+CONFIG_ADDRESS_MASKING is not selected) the flag "lam" will not show up.
 Even though the feature will still be detected via CPUID, the kernel disables
-it by clearing via setup_clear_cpu_cap(X86_FEATURE_LA57).
+it by clearing via setup_clear_cpu_cap(X86_FEATURE_LAM).
 
 d: The feature is disabled at boot-time.
 ----------------------------------------
@@ -197,5 +197,3 @@ missing at runtime. For example, AVX flags will not show up if XSAVE feature
 is disabled since they depend on XSAVE feature. Another example would be broken
 CPUs and them missing microcode patches. Due to that, the kernel decides not to
 enable a feature.
-
-.. [#f1] 5-level paging uses linear address of 57 bits.
diff --git a/Documentation/arch/x86/x86_64/5level-paging.rst b/Documentation/arch/x86/x86_64/5level-paging.rst
index 71f882f4a173..ad7ddc13f79d 100644
--- a/Documentation/arch/x86/x86_64/5level-paging.rst
+++ b/Documentation/arch/x86/x86_64/5level-paging.rst
@@ -22,15 +22,6 @@ QEMU 2.9 and later support 5-level paging.
 Virtual memory layout for 5-level paging is described in
 Documentation/arch/x86/x86_64/mm.rst
 
-
-Enabling 5-level paging
-=======================
-CONFIG_X86_5LEVEL=y enables the feature.
-
-Kernel with CONFIG_X86_5LEVEL=y still able to boot on 4-level hardware.
-In this case additional page table level -- p4d -- will be folded at
-runtime.
-
 User-space and large virtual address space
 ==========================================
 On x86, 5-level paging enables 56-bit userspace virtual address space.
diff --git a/arch/x86/Kconfig b/arch/x86/Kconfig
index 54ad2462e9ef..f95a5048ad09 100644
--- a/arch/x86/Kconfig
+++ b/arch/x86/Kconfig
@@ -408,8 +408,7 @@ config DYNAMIC_PHYSICAL_MASK
 
 config PGTABLE_LEVELS
 	int
-	default 5 if X86_5LEVEL
-	default 4 if X86_64
+	default 5 if X86_64
 	default 3 if X86_PAE
 	default 2
 
@@ -1498,25 +1497,6 @@ config X86_PAE
 	  has the cost of more pagetable lookup overhead, and also
 	  consumes more pagetable space per process.
 
-config X86_5LEVEL
-	bool "Enable 5-level page tables support"
-	default y
-	depends on X86_64
-	help
-	  5-level paging enables access to larger address space:
-	  up to 128 PiB of virtual address space and 4 PiB of
-	  physical address space.
-
-	  It will be supported by future Intel CPUs.
-
-	  A kernel with the option enabled can be booted on machines that
-	  support 4- or 5-level paging.
-
-	  See Documentation/arch/x86/x86_64/5level-paging.rst for more
-	  information.
-
-	  Say N if unsure.
-
 config X86_DIRECT_GBPAGES
 	def_bool y
 	depends on X86_64
diff --git a/arch/x86/boot/compressed/pgtable_64.c b/arch/x86/boot/compressed/pgtable_64.c
index c882e1f67af0..61b9ca61bde1 100644
--- a/arch/x86/boot/compressed/pgtable_64.c
+++ b/arch/x86/boot/compressed/pgtable_64.c
@@ -10,12 +10,10 @@
 #define BIOS_START_MIN		0x20000U	/* 128K, less than this is insane */
 #define BIOS_START_MAX		0x9f000U	/* 640K, absolute maximum */
 
-#ifdef CONFIG_X86_5LEVEL
 /* __pgtable_l5_enabled needs to be in .data to avoid being cleared along with .bss */
 unsigned int __section(".data") __pgtable_l5_enabled;
 unsigned int __section(".data") pgdir_shift = 39;
 unsigned int __section(".data") ptrs_per_p4d = 1;
-#endif
 
 /* Buffer to preserve trampoline memory */
 static char trampoline_save[TRAMPOLINE_32BIT_SIZE];
@@ -113,18 +111,13 @@ asmlinkage void configure_5level_paging(struct boot_params *bp, void *pgtable)
 	 * Check if LA57 is desired and supported.
 	 *
 	 * There are several parts to the check:
-	 *   - if the kernel supports 5-level paging: CONFIG_X86_5LEVEL=y
 	 *   - if user asked to disable 5-level paging: no5lvl in cmdline
 	 *   - if the machine supports 5-level paging:
 	 *     + CPUID leaf 7 is supported
 	 *     + the leaf has the feature bit set
-	 *
-	 * That's substitute for boot_cpu_has() in early boot code.
 	 */
-	if (IS_ENABLED(CONFIG_X86_5LEVEL) &&
-			!cmdline_find_option_bool("no5lvl") &&
-			native_cpuid_eax(0) >= 7 &&
-			(native_cpuid_ecx(7) & (1 << (X86_FEATURE_LA57 & 31)))) {
+	if (!cmdline_find_option_bool("no5lvl") &&
+	    native_cpuid_eax(0) >= 7 && (native_cpuid_ecx(7) & BIT(16))) {
 		l5_required = true;
 
 		/* Initialize variables for 5-level paging */
diff --git a/arch/x86/boot/header.S b/arch/x86/boot/header.S
index b5c79f43359b..32361cef909e 100644
--- a/arch/x86/boot/header.S
+++ b/arch/x86/boot/header.S
@@ -361,12 +361,8 @@ xloadflags:
 #endif
 
 #ifdef CONFIG_X86_64
-#ifdef CONFIG_X86_5LEVEL
 #define XLF56 (XLF_5LEVEL|XLF_5LEVEL_ENABLED)
 #else
-#define XLF56 XLF_5LEVEL
-#endif
-#else
 #define XLF56 0
 #endif
 
diff --git a/arch/x86/include/asm/disabled-features.h b/arch/x86/include/asm/disabled-features.h
index c492bdc97b05..19cf1678fcaa 100644
--- a/arch/x86/include/asm/disabled-features.h
+++ b/arch/x86/include/asm/disabled-features.h
@@ -38,12 +38,6 @@
 # define DISABLE_OSPKE		(1<<(X86_FEATURE_OSPKE & 31))
 #endif /* CONFIG_X86_INTEL_MEMORY_PROTECTION_KEYS */
 
-#ifdef CONFIG_X86_5LEVEL
-# define DISABLE_LA57	0
-#else
-# define DISABLE_LA57	(1<<(X86_FEATURE_LA57 & 31))
-#endif
-
 #ifdef CONFIG_MITIGATION_PAGE_TABLE_ISOLATION
 # define DISABLE_PTI		0
 #else
@@ -149,8 +143,7 @@
 #define DISABLED_MASK13	0
 #define DISABLED_MASK14	0
 #define DISABLED_MASK15	0
-#define DISABLED_MASK16	(DISABLE_PKU|DISABLE_OSPKE|DISABLE_LA57|DISABLE_UMIP| \
-			 DISABLE_ENQCMD)
+#define DISABLED_MASK16	(DISABLE_PKU|DISABLE_OSPKE|DISABLE_UMIP|DISABLE_ENQCMD)
 #define DISABLED_MASK17	0
 #define DISABLED_MASK18	(DISABLE_IBT)
 #define DISABLED_MASK19	(DISABLE_SEV_SNP)
diff --git a/arch/x86/include/asm/page_64.h b/arch/x86/include/asm/page_64.h
index cc6b8e087192..3b8cb6a8b122 100644
--- a/arch/x86/include/asm/page_64.h
+++ b/arch/x86/include/asm/page_64.h
@@ -60,7 +60,6 @@ static inline void clear_page(void *page)
 
 void copy_page(void *to, void *from);
 
-#ifdef CONFIG_X86_5LEVEL
 /*
  * User space process size.  This is the first address outside the user range.
  * There are a few constraints that determine this:
@@ -91,7 +90,6 @@ static __always_inline unsigned long task_size_max(void)
 
 	return ret;
 }
-#endif	/* CONFIG_X86_5LEVEL */
 
 #endif	/* !__ASSEMBLY__ */
 
diff --git a/arch/x86/include/asm/page_64_types.h b/arch/x86/include/asm/page_64_types.h
index c2f3c50a2787..666a5d6ab910 100644
--- a/arch/x86/include/asm/page_64_types.h
+++ b/arch/x86/include/asm/page_64_types.h
@@ -48,14 +48,7 @@
 /* See Documentation/arch/x86/x86_64/mm.rst for a description of the memory map. */
 
 #define __PHYSICAL_MASK_SHIFT	52
-
-#ifdef CONFIG_X86_5LEVEL
 #define __VIRTUAL_MASK_SHIFT	(pgtable_l5_enabled() ? 56 : 47)
-/* See task_size_max() in <asm/page_64.h> */
-#else
-#define __VIRTUAL_MASK_SHIFT	47
-#define task_size_max()		((_AC(1,UL) << __VIRTUAL_MASK_SHIFT) - PAGE_SIZE)
-#endif
 
 #define TASK_SIZE_MAX		task_size_max()
 #define DEFAULT_MAP_WINDOW	((1UL << 47) - PAGE_SIZE)
diff --git a/arch/x86/include/asm/pgtable_64_types.h b/arch/x86/include/asm/pgtable_64_types.h
index 09df8939b997..2c77489ac86c 100644
--- a/arch/x86/include/asm/pgtable_64_types.h
+++ b/arch/x86/include/asm/pgtable_64_types.h
@@ -23,7 +23,6 @@ typedef struct { pmdval_t pmd; } pmd_t;
 
 extern unsigned int __pgtable_l5_enabled;
 
-#ifdef CONFIG_X86_5LEVEL
 #ifdef USE_EARLY_PGTABLE_L5
 /*
  * cpu_feature_enabled() is not available in early boot code.
@@ -37,10 +36,6 @@ static inline bool pgtable_l5_enabled(void)
 #define pgtable_l5_enabled() cpu_feature_enabled(X86_FEATURE_LA57)
 #endif /* USE_EARLY_PGTABLE_L5 */
 
-#else
-#define pgtable_l5_enabled() 0
-#endif /* CONFIG_X86_5LEVEL */
-
 extern unsigned int pgdir_shift;
 extern unsigned int ptrs_per_p4d;
 
@@ -48,8 +43,6 @@ extern unsigned int ptrs_per_p4d;
 
 #define SHARED_KERNEL_PMD	0
 
-#ifdef CONFIG_X86_5LEVEL
-
 /*
  * PGDIR_SHIFT determines what a top-level page table entry can map
  */
@@ -67,17 +60,6 @@ extern unsigned int ptrs_per_p4d;
 
 #define MAX_POSSIBLE_PHYSMEM_BITS	52
 
-#else /* CONFIG_X86_5LEVEL */
-
-/*
- * PGDIR_SHIFT determines what a top-level page table entry can map
- */
-#define PGDIR_SHIFT		39
-#define PTRS_PER_PGD		512
-#define MAX_PTRS_PER_P4D	1
-
-#endif /* CONFIG_X86_5LEVEL */
-
 /*
  * 3rd level page
  */
diff --git a/arch/x86/kernel/alternative.c b/arch/x86/kernel/alternative.c
index 37596a417094..f1c519abb925 100644
--- a/arch/x86/kernel/alternative.c
+++ b/arch/x86/kernel/alternative.c
@@ -457,7 +457,7 @@ void __init_or_module noinline apply_alternatives(struct alt_instr *start,
 	DPRINTK(ALT, "alt table %px, -> %px", start, end);
 
 	/*
-	 * In the case CONFIG_X86_5LEVEL=y, KASAN_SHADOW_START is defined using
+	 * KASAN_SHADOW_START is defined using
 	 * cpu_feature_enabled(X86_FEATURE_LA57) and is therefore patched here.
 	 * During the process, KASAN becomes confused seeing partial LA57
 	 * conversion and triggers a false-positive out-of-bound report.
diff --git a/arch/x86/kernel/head64.c b/arch/x86/kernel/head64.c
index ec36ad7117ae..ec3a7e2ea222 100644
--- a/arch/x86/kernel/head64.c
+++ b/arch/x86/kernel/head64.c
@@ -52,13 +52,11 @@ extern pmd_t early_dynamic_pgts[EARLY_DYNAMIC_PAGE_TABLES][PTRS_PER_PMD];
 static unsigned int __initdata next_early_pgt;
 pmdval_t early_pmd_flags = __PAGE_KERNEL_LARGE & ~(_PAGE_GLOBAL | _PAGE_NX);
 
-#ifdef CONFIG_X86_5LEVEL
 unsigned int __pgtable_l5_enabled __ro_after_init;
 unsigned int pgdir_shift __ro_after_init = 39;
 EXPORT_SYMBOL(pgdir_shift);
 unsigned int ptrs_per_p4d __ro_after_init = 1;
 EXPORT_SYMBOL(ptrs_per_p4d);
-#endif
 
 unsigned long page_offset_base __ro_after_init = __PAGE_OFFSET_BASE_L4;
 EXPORT_SYMBOL(page_offset_base);
@@ -69,9 +67,6 @@ EXPORT_SYMBOL(vmemmap_base);
 
 static inline bool check_la57_support(void)
 {
-	if (!IS_ENABLED(CONFIG_X86_5LEVEL))
-		return false;
-
 	/*
 	 * 5-level paging is detected and enabled at kernel decompression
 	 * stage. Only check if it has been enabled there.
diff --git a/arch/x86/kernel/head_64.S b/arch/x86/kernel/head_64.S
index 330922b328bf..4b2b2138c163 100644
--- a/arch/x86/kernel/head_64.S
+++ b/arch/x86/kernel/head_64.S
@@ -659,12 +659,10 @@ SYM_DATA_START_PTI_ALIGNED(init_top_pgt)
 SYM_DATA_END(init_top_pgt)
 #endif
 
-#ifdef CONFIG_X86_5LEVEL
 SYM_DATA_START_PAGE_ALIGNED(level4_kernel_pgt)
 	.fill	511,8,0
 	.quad	level3_kernel_pgt - __START_KERNEL_map + _PAGE_TABLE_NOENC
 SYM_DATA_END(level4_kernel_pgt)
-#endif
 
 SYM_DATA_START_PAGE_ALIGNED(level3_kernel_pgt)
 	.fill	L3_START_KERNEL,8,0
diff --git a/arch/x86/mm/init.c b/arch/x86/mm/init.c
index eb503f53c319..5a980a452f4c 100644
--- a/arch/x86/mm/init.c
+++ b/arch/x86/mm/init.c
@@ -173,11 +173,7 @@ __ref void *alloc_low_pages(unsigned int num)
  * randomization is enabled.
  */
 
-#ifndef CONFIG_X86_5LEVEL
-#define INIT_PGD_PAGE_TABLES    3
-#else
 #define INIT_PGD_PAGE_TABLES    4
-#endif
 
 #ifndef CONFIG_RANDOMIZE_MEMORY
 #define INIT_PGD_PAGE_COUNT      (2 * INIT_PGD_PAGE_TABLES)
diff --git a/arch/x86/mm/pgtable.c b/arch/x86/mm/pgtable.c
index 93e54ba91fbf..982775ef8b34 100644
--- a/arch/x86/mm/pgtable.c
+++ b/arch/x86/mm/pgtable.c
@@ -691,7 +691,6 @@ void native_set_fixmap(unsigned /* enum fixed_addresses */ idx,
 }
 
 #ifdef CONFIG_HAVE_ARCH_HUGE_VMAP
-#ifdef CONFIG_X86_5LEVEL
 /**
  * p4d_set_huge - setup kernel P4D mapping
  *
@@ -710,7 +709,6 @@ int p4d_set_huge(p4d_t *p4d, phys_addr_t addr, pgprot_t prot)
 void p4d_clear_huge(p4d_t *p4d)
 {
 }
-#endif
 
 /**
  * pud_set_huge - setup kernel PUD mapping
diff --git a/drivers/firmware/efi/libstub/x86-5lvl.c b/drivers/firmware/efi/libstub/x86-5lvl.c
index 77359e802181..f1c5fb45d5f7 100644
--- a/drivers/firmware/efi/libstub/x86-5lvl.c
+++ b/drivers/firmware/efi/libstub/x86-5lvl.c
@@ -62,7 +62,7 @@ efi_status_t efi_setup_5level_paging(void)
 
 void efi_5level_switch(void)
 {
-	bool want_la57 = IS_ENABLED(CONFIG_X86_5LEVEL) && !efi_no5lvl;
+	bool want_la57 = !efi_no5lvl;
 	bool have_la57 = native_read_cr4() & X86_CR4_LA57;
 	bool need_toggle = want_la57 ^ have_la57;
 	u64 *pgt = (void *)la57_toggle + PAGE_SIZE;
diff --git a/tools/arch/x86/include/asm/disabled-features.h b/tools/arch/x86/include/asm/disabled-features.h
index c492bdc97b05..19cf1678fcaa 100644
--- a/tools/arch/x86/include/asm/disabled-features.h
+++ b/tools/arch/x86/include/asm/disabled-features.h
@@ -38,12 +38,6 @@
 # define DISABLE_OSPKE		(1<<(X86_FEATURE_OSPKE & 31))
 #endif /* CONFIG_X86_INTEL_MEMORY_PROTECTION_KEYS */
 
-#ifdef CONFIG_X86_5LEVEL
-# define DISABLE_LA57	0
-#else
-# define DISABLE_LA57	(1<<(X86_FEATURE_LA57 & 31))
-#endif
-
 #ifdef CONFIG_MITIGATION_PAGE_TABLE_ISOLATION
 # define DISABLE_PTI		0
 #else
@@ -149,8 +143,7 @@
 #define DISABLED_MASK13	0
 #define DISABLED_MASK14	0
 #define DISABLED_MASK15	0
-#define DISABLED_MASK16	(DISABLE_PKU|DISABLE_OSPKE|DISABLE_LA57|DISABLE_UMIP| \
-			 DISABLE_ENQCMD)
+#define DISABLED_MASK16	(DISABLE_PKU|DISABLE_OSPKE|DISABLE_UMIP|DISABLE_ENQCMD)
 #define DISABLED_MASK17	0
 #define DISABLED_MASK18	(DISABLE_IBT)
 #define DISABLED_MASK19	(DISABLE_SEV_SNP)
-- 
2.43.0


^ permalink raw reply	[flat|nested] 15+ messages in thread

* Re: [PATCH 3/3] x86/64/mm: Make 5-level paging support unconditional
  2024-06-21 16:44 ` [PATCH 3/3] x86/64/mm: Make 5-level paging support unconditional Kirill A. Shutemov
@ 2024-06-25 23:21   ` kernel test robot
  2024-06-26 10:26     ` [PATCH] " Kirill A. Shutemov
  0 siblings, 1 reply; 15+ messages in thread
From: kernel test robot @ 2024-06-25 23:21 UTC (permalink / raw)
  To: Kirill A. Shutemov, Thomas Gleixner, Ingo Molnar,
	Borislav Petkov, Dave Hansen, x86, H. Peter Anvin
  Cc: oe-kbuild-all, Jonathan Corbet, Andy Lutomirski, Peter Zijlstra,
	Ard Biesheuvel, Jan Kiszka, Kieran Bingham, Kirill A. Shutemov,
	Michael Roth, Rick Edgecombe, Brijesh Singh, Sandipan Das,
	Juergen Gross, Tom Lendacky, linux-kernel, linux-doc, linux-efi,
	linux-mm

Hi Kirill,

kernel test robot noticed the following build errors:

[auto build test ERROR on tip/x86/core]
[also build test ERROR on tip/master linus/master tip/x86/mm v6.10-rc5 next-20240625]
[cannot apply to tip/auto-latest]
[If your patch is applied to the wrong git tree, kindly drop us a note.
And when submitting patch, we suggest to use '--base' as documented in
https://git-scm.com/docs/git-format-patch#_base_tree_information]

url:    https://github.com/intel-lab-lkp/linux/commits/Kirill-A-Shutemov/x86-64-mm-Always-use-dynamic-memory-layout/20240625-183159
base:   tip/x86/core
patch link:    https://lore.kernel.org/r/20240621164406.256314-4-kirill.shutemov%40linux.intel.com
patch subject: [PATCH 3/3] x86/64/mm: Make 5-level paging support unconditional
config: i386-buildonly-randconfig-004-20240626 (https://download.01.org/0day-ci/archive/20240626/202406260735.rkb4c8N7-lkp@intel.com/config)
compiler: gcc-13 (Ubuntu 13.2.0-4ubuntu3) 13.2.0
reproduce (this is a W=1 build): (https://download.01.org/0day-ci/archive/20240626/202406260735.rkb4c8N7-lkp@intel.com/reproduce)

If you fix the issue in a separate patch/commit (i.e. not just a new version of
the same patch/commit), kindly add following tags
| Reported-by: kernel test robot <lkp@intel.com>
| Closes: https://lore.kernel.org/oe-kbuild-all/202406260735.rkb4c8N7-lkp@intel.com/

All errors (new ones prefixed by >>):

>> arch/x86/mm/pgtable.c:699:5: error: redefinition of 'p4d_set_huge'
     699 | int p4d_set_huge(p4d_t *p4d, phys_addr_t addr, pgprot_t prot)
         |     ^~~~~~~~~~~~
   In file included from include/linux/mm.h:30,
                    from arch/x86/mm/pgtable.c:2:
   include/linux/pgtable.h:1665:19: note: previous definition of 'p4d_set_huge' with type 'int(p4d_t *, phys_addr_t,  pgprot_t)' {aka 'int(p4d_t *, long long unsigned int,  struct pgprot)'}
    1665 | static inline int p4d_set_huge(p4d_t *p4d, phys_addr_t addr, pgprot_t prot)
         |                   ^~~~~~~~~~~~
>> arch/x86/mm/pgtable.c:709:6: error: redefinition of 'p4d_clear_huge'
     709 | void p4d_clear_huge(p4d_t *p4d)
         |      ^~~~~~~~~~~~~~
   include/linux/pgtable.h:1669:20: note: previous definition of 'p4d_clear_huge' with type 'void(p4d_t *)'
    1669 | static inline void p4d_clear_huge(p4d_t *p4d) { }
         |                    ^~~~~~~~~~~~~~


vim +/p4d_set_huge +699 arch/x86/mm/pgtable.c

6b6378355b9250 Toshi Kani         2015-04-14  692  
6b6378355b9250 Toshi Kani         2015-04-14  693  #ifdef CONFIG_HAVE_ARCH_HUGE_VMAP
b8504058a06bd1 Kirill A. Shutemov 2017-03-30  694  /**
b8504058a06bd1 Kirill A. Shutemov 2017-03-30  695   * p4d_set_huge - setup kernel P4D mapping
b8504058a06bd1 Kirill A. Shutemov 2017-03-30  696   *
b8504058a06bd1 Kirill A. Shutemov 2017-03-30  697   * No 512GB pages yet -- always return 0
b8504058a06bd1 Kirill A. Shutemov 2017-03-30  698   */
b8504058a06bd1 Kirill A. Shutemov 2017-03-30 @699  int p4d_set_huge(p4d_t *p4d, phys_addr_t addr, pgprot_t prot)
b8504058a06bd1 Kirill A. Shutemov 2017-03-30  700  {
b8504058a06bd1 Kirill A. Shutemov 2017-03-30  701  	return 0;
b8504058a06bd1 Kirill A. Shutemov 2017-03-30  702  }
b8504058a06bd1 Kirill A. Shutemov 2017-03-30  703  
b8504058a06bd1 Kirill A. Shutemov 2017-03-30  704  /**
b8504058a06bd1 Kirill A. Shutemov 2017-03-30  705   * p4d_clear_huge - clear kernel P4D mapping when it is set
b8504058a06bd1 Kirill A. Shutemov 2017-03-30  706   *
b8504058a06bd1 Kirill A. Shutemov 2017-03-30  707   * No 512GB pages yet -- always return 0
b8504058a06bd1 Kirill A. Shutemov 2017-03-30  708   */
c8db8c2628afc7 Li kunyu           2022-05-12 @709  void p4d_clear_huge(p4d_t *p4d)
b8504058a06bd1 Kirill A. Shutemov 2017-03-30  710  {
b8504058a06bd1 Kirill A. Shutemov 2017-03-30  711  }
b8504058a06bd1 Kirill A. Shutemov 2017-03-30  712  

-- 
0-DAY CI Kernel Test Service
https://github.com/intel/lkp-tests/wiki

^ permalink raw reply	[flat|nested] 15+ messages in thread

* [PATCH] x86/64/mm: Make 5-level paging support unconditional
  2024-06-25 23:21   ` kernel test robot
@ 2024-06-26 10:26     ` Kirill A. Shutemov
  2024-06-26 10:46       ` Borislav Petkov
  0 siblings, 1 reply; 15+ messages in thread
From: Kirill A. Shutemov @ 2024-06-26 10:26 UTC (permalink / raw)
  To: lkp
  Cc: ardb, bp, brijesh.singh, corbet, dave.hansen, hpa, jan.kiszka,
	jgross, kbingham, kirill.shutemov, linux-doc, linux-efi,
	linux-kernel, linux-mm, luto, michael.roth, mingo, oe-kbuild-all,
	peterz, rick.p.edgecombe, sandipan.das, tglx, thomas.lendacky,
	x86

Both Intel and AMD CPUs support 5-level paging, which is expected to
become more widely adopted in the future.

Remove CONFIG_X86_5LEVEL and ifdeffery for it to make it more readable.

Signed-off-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Suggested-by: Borislav Petkov <bp@alien8.de>
---

 v2:
  - Fix 32-bit build by wrapping p4d_set_huge() and p4d_clear_huge() in
    #if CONFIG_PGTABLE_LEVELS > 4

---
 Documentation/arch/x86/cpuinfo.rst            |  8 +++----
 .../arch/x86/x86_64/5level-paging.rst         |  9 --------
 arch/x86/Kconfig                              | 22 +------------------
 arch/x86/boot/compressed/pgtable_64.c         | 11 ++--------
 arch/x86/boot/header.S                        |  4 ----
 arch/x86/include/asm/disabled-features.h      |  9 +-------
 arch/x86/include/asm/page_64.h                |  2 --
 arch/x86/include/asm/page_64_types.h          |  7 ------
 arch/x86/include/asm/pgtable_64_types.h       | 18 ---------------
 arch/x86/kernel/alternative.c                 |  2 +-
 arch/x86/kernel/head64.c                      |  5 -----
 arch/x86/kernel/head_64.S                     |  2 --
 arch/x86/mm/init.c                            |  4 ----
 arch/x86/mm/pgtable.c                         |  2 +-
 drivers/firmware/efi/libstub/x86-5lvl.c       |  2 +-
 .../arch/x86/include/asm/disabled-features.h  |  9 +-------
 16 files changed, 11 insertions(+), 105 deletions(-)

diff --git a/Documentation/arch/x86/cpuinfo.rst b/Documentation/arch/x86/cpuinfo.rst
index 8895784d4784..0ea70924c89e 100644
--- a/Documentation/arch/x86/cpuinfo.rst
+++ b/Documentation/arch/x86/cpuinfo.rst
@@ -171,10 +171,10 @@ For example, when an old kernel is running on new hardware.
 
 c: The kernel disabled support for it at compile-time.
 ------------------------------------------------------
-For example, if 5-level-paging is not enabled when building (i.e.,
-CONFIG_X86_5LEVEL is not selected) the flag "la57" will not show up [#f1]_.
+For example, if Linear Address Masking (LAM) is not enabled when building (i.e.,
+CONFIG_ADDRESS_MASKING is not selected) the flag "lam" will not show up.
 Even though the feature will still be detected via CPUID, the kernel disables
-it by clearing via setup_clear_cpu_cap(X86_FEATURE_LA57).
+it by clearing via setup_clear_cpu_cap(X86_FEATURE_LAM).
 
 d: The feature is disabled at boot-time.
 ----------------------------------------
@@ -197,5 +197,3 @@ missing at runtime. For example, AVX flags will not show up if XSAVE feature
 is disabled since they depend on XSAVE feature. Another example would be broken
 CPUs and them missing microcode patches. Due to that, the kernel decides not to
 enable a feature.
-
-.. [#f1] 5-level paging uses linear address of 57 bits.
diff --git a/Documentation/arch/x86/x86_64/5level-paging.rst b/Documentation/arch/x86/x86_64/5level-paging.rst
index 71f882f4a173..ad7ddc13f79d 100644
--- a/Documentation/arch/x86/x86_64/5level-paging.rst
+++ b/Documentation/arch/x86/x86_64/5level-paging.rst
@@ -22,15 +22,6 @@ QEMU 2.9 and later support 5-level paging.
 Virtual memory layout for 5-level paging is described in
 Documentation/arch/x86/x86_64/mm.rst
 
-
-Enabling 5-level paging
-=======================
-CONFIG_X86_5LEVEL=y enables the feature.
-
-Kernel with CONFIG_X86_5LEVEL=y still able to boot on 4-level hardware.
-In this case additional page table level -- p4d -- will be folded at
-runtime.
-
 User-space and large virtual address space
 ==========================================
 On x86, 5-level paging enables 56-bit userspace virtual address space.
diff --git a/arch/x86/Kconfig b/arch/x86/Kconfig
index 54ad2462e9ef..f95a5048ad09 100644
--- a/arch/x86/Kconfig
+++ b/arch/x86/Kconfig
@@ -408,8 +408,7 @@ config DYNAMIC_PHYSICAL_MASK
 
 config PGTABLE_LEVELS
 	int
-	default 5 if X86_5LEVEL
-	default 4 if X86_64
+	default 5 if X86_64
 	default 3 if X86_PAE
 	default 2
 
@@ -1498,25 +1497,6 @@ config X86_PAE
 	  has the cost of more pagetable lookup overhead, and also
 	  consumes more pagetable space per process.
 
-config X86_5LEVEL
-	bool "Enable 5-level page tables support"
-	default y
-	depends on X86_64
-	help
-	  5-level paging enables access to larger address space:
-	  up to 128 PiB of virtual address space and 4 PiB of
-	  physical address space.
-
-	  It will be supported by future Intel CPUs.
-
-	  A kernel with the option enabled can be booted on machines that
-	  support 4- or 5-level paging.
-
-	  See Documentation/arch/x86/x86_64/5level-paging.rst for more
-	  information.
-
-	  Say N if unsure.
-
 config X86_DIRECT_GBPAGES
 	def_bool y
 	depends on X86_64
diff --git a/arch/x86/boot/compressed/pgtable_64.c b/arch/x86/boot/compressed/pgtable_64.c
index c882e1f67af0..61b9ca61bde1 100644
--- a/arch/x86/boot/compressed/pgtable_64.c
+++ b/arch/x86/boot/compressed/pgtable_64.c
@@ -10,12 +10,10 @@
 #define BIOS_START_MIN		0x20000U	/* 128K, less than this is insane */
 #define BIOS_START_MAX		0x9f000U	/* 640K, absolute maximum */
 
-#ifdef CONFIG_X86_5LEVEL
 /* __pgtable_l5_enabled needs to be in .data to avoid being cleared along with .bss */
 unsigned int __section(".data") __pgtable_l5_enabled;
 unsigned int __section(".data") pgdir_shift = 39;
 unsigned int __section(".data") ptrs_per_p4d = 1;
-#endif
 
 /* Buffer to preserve trampoline memory */
 static char trampoline_save[TRAMPOLINE_32BIT_SIZE];
@@ -113,18 +111,13 @@ asmlinkage void configure_5level_paging(struct boot_params *bp, void *pgtable)
 	 * Check if LA57 is desired and supported.
 	 *
 	 * There are several parts to the check:
-	 *   - if the kernel supports 5-level paging: CONFIG_X86_5LEVEL=y
 	 *   - if user asked to disable 5-level paging: no5lvl in cmdline
 	 *   - if the machine supports 5-level paging:
 	 *     + CPUID leaf 7 is supported
 	 *     + the leaf has the feature bit set
-	 *
-	 * That's substitute for boot_cpu_has() in early boot code.
 	 */
-	if (IS_ENABLED(CONFIG_X86_5LEVEL) &&
-			!cmdline_find_option_bool("no5lvl") &&
-			native_cpuid_eax(0) >= 7 &&
-			(native_cpuid_ecx(7) & (1 << (X86_FEATURE_LA57 & 31)))) {
+	if (!cmdline_find_option_bool("no5lvl") &&
+	    native_cpuid_eax(0) >= 7 && (native_cpuid_ecx(7) & BIT(16))) {
 		l5_required = true;
 
 		/* Initialize variables for 5-level paging */
diff --git a/arch/x86/boot/header.S b/arch/x86/boot/header.S
index b5c79f43359b..32361cef909e 100644
--- a/arch/x86/boot/header.S
+++ b/arch/x86/boot/header.S
@@ -361,12 +361,8 @@ xloadflags:
 #endif
 
 #ifdef CONFIG_X86_64
-#ifdef CONFIG_X86_5LEVEL
 #define XLF56 (XLF_5LEVEL|XLF_5LEVEL_ENABLED)
 #else
-#define XLF56 XLF_5LEVEL
-#endif
-#else
 #define XLF56 0
 #endif
 
diff --git a/arch/x86/include/asm/disabled-features.h b/arch/x86/include/asm/disabled-features.h
index c492bdc97b05..19cf1678fcaa 100644
--- a/arch/x86/include/asm/disabled-features.h
+++ b/arch/x86/include/asm/disabled-features.h
@@ -38,12 +38,6 @@
 # define DISABLE_OSPKE		(1<<(X86_FEATURE_OSPKE & 31))
 #endif /* CONFIG_X86_INTEL_MEMORY_PROTECTION_KEYS */
 
-#ifdef CONFIG_X86_5LEVEL
-# define DISABLE_LA57	0
-#else
-# define DISABLE_LA57	(1<<(X86_FEATURE_LA57 & 31))
-#endif
-
 #ifdef CONFIG_MITIGATION_PAGE_TABLE_ISOLATION
 # define DISABLE_PTI		0
 #else
@@ -149,8 +143,7 @@
 #define DISABLED_MASK13	0
 #define DISABLED_MASK14	0
 #define DISABLED_MASK15	0
-#define DISABLED_MASK16	(DISABLE_PKU|DISABLE_OSPKE|DISABLE_LA57|DISABLE_UMIP| \
-			 DISABLE_ENQCMD)
+#define DISABLED_MASK16	(DISABLE_PKU|DISABLE_OSPKE|DISABLE_UMIP|DISABLE_ENQCMD)
 #define DISABLED_MASK17	0
 #define DISABLED_MASK18	(DISABLE_IBT)
 #define DISABLED_MASK19	(DISABLE_SEV_SNP)
diff --git a/arch/x86/include/asm/page_64.h b/arch/x86/include/asm/page_64.h
index cc6b8e087192..3b8cb6a8b122 100644
--- a/arch/x86/include/asm/page_64.h
+++ b/arch/x86/include/asm/page_64.h
@@ -60,7 +60,6 @@ static inline void clear_page(void *page)
 
 void copy_page(void *to, void *from);
 
-#ifdef CONFIG_X86_5LEVEL
 /*
  * User space process size.  This is the first address outside the user range.
  * There are a few constraints that determine this:
@@ -91,7 +90,6 @@ static __always_inline unsigned long task_size_max(void)
 
 	return ret;
 }
-#endif	/* CONFIG_X86_5LEVEL */
 
 #endif	/* !__ASSEMBLY__ */
 
diff --git a/arch/x86/include/asm/page_64_types.h b/arch/x86/include/asm/page_64_types.h
index c2f3c50a2787..666a5d6ab910 100644
--- a/arch/x86/include/asm/page_64_types.h
+++ b/arch/x86/include/asm/page_64_types.h
@@ -48,14 +48,7 @@
 /* See Documentation/arch/x86/x86_64/mm.rst for a description of the memory map. */
 
 #define __PHYSICAL_MASK_SHIFT	52
-
-#ifdef CONFIG_X86_5LEVEL
 #define __VIRTUAL_MASK_SHIFT	(pgtable_l5_enabled() ? 56 : 47)
-/* See task_size_max() in <asm/page_64.h> */
-#else
-#define __VIRTUAL_MASK_SHIFT	47
-#define task_size_max()		((_AC(1,UL) << __VIRTUAL_MASK_SHIFT) - PAGE_SIZE)
-#endif
 
 #define TASK_SIZE_MAX		task_size_max()
 #define DEFAULT_MAP_WINDOW	((1UL << 47) - PAGE_SIZE)
diff --git a/arch/x86/include/asm/pgtable_64_types.h b/arch/x86/include/asm/pgtable_64_types.h
index 09df8939b997..2c77489ac86c 100644
--- a/arch/x86/include/asm/pgtable_64_types.h
+++ b/arch/x86/include/asm/pgtable_64_types.h
@@ -23,7 +23,6 @@ typedef struct { pmdval_t pmd; } pmd_t;
 
 extern unsigned int __pgtable_l5_enabled;
 
-#ifdef CONFIG_X86_5LEVEL
 #ifdef USE_EARLY_PGTABLE_L5
 /*
  * cpu_feature_enabled() is not available in early boot code.
@@ -37,10 +36,6 @@ static inline bool pgtable_l5_enabled(void)
 #define pgtable_l5_enabled() cpu_feature_enabled(X86_FEATURE_LA57)
 #endif /* USE_EARLY_PGTABLE_L5 */
 
-#else
-#define pgtable_l5_enabled() 0
-#endif /* CONFIG_X86_5LEVEL */
-
 extern unsigned int pgdir_shift;
 extern unsigned int ptrs_per_p4d;
 
@@ -48,8 +43,6 @@ extern unsigned int ptrs_per_p4d;
 
 #define SHARED_KERNEL_PMD	0
 
-#ifdef CONFIG_X86_5LEVEL
-
 /*
  * PGDIR_SHIFT determines what a top-level page table entry can map
  */
@@ -67,17 +60,6 @@ extern unsigned int ptrs_per_p4d;
 
 #define MAX_POSSIBLE_PHYSMEM_BITS	52
 
-#else /* CONFIG_X86_5LEVEL */
-
-/*
- * PGDIR_SHIFT determines what a top-level page table entry can map
- */
-#define PGDIR_SHIFT		39
-#define PTRS_PER_PGD		512
-#define MAX_PTRS_PER_P4D	1
-
-#endif /* CONFIG_X86_5LEVEL */
-
 /*
  * 3rd level page
  */
diff --git a/arch/x86/kernel/alternative.c b/arch/x86/kernel/alternative.c
index 37596a417094..f1c519abb925 100644
--- a/arch/x86/kernel/alternative.c
+++ b/arch/x86/kernel/alternative.c
@@ -457,7 +457,7 @@ void __init_or_module noinline apply_alternatives(struct alt_instr *start,
 	DPRINTK(ALT, "alt table %px, -> %px", start, end);
 
 	/*
-	 * In the case CONFIG_X86_5LEVEL=y, KASAN_SHADOW_START is defined using
+	 * KASAN_SHADOW_START is defined using
 	 * cpu_feature_enabled(X86_FEATURE_LA57) and is therefore patched here.
 	 * During the process, KASAN becomes confused seeing partial LA57
 	 * conversion and triggers a false-positive out-of-bound report.
diff --git a/arch/x86/kernel/head64.c b/arch/x86/kernel/head64.c
index ec36ad7117ae..ec3a7e2ea222 100644
--- a/arch/x86/kernel/head64.c
+++ b/arch/x86/kernel/head64.c
@@ -52,13 +52,11 @@ extern pmd_t early_dynamic_pgts[EARLY_DYNAMIC_PAGE_TABLES][PTRS_PER_PMD];
 static unsigned int __initdata next_early_pgt;
 pmdval_t early_pmd_flags = __PAGE_KERNEL_LARGE & ~(_PAGE_GLOBAL | _PAGE_NX);
 
-#ifdef CONFIG_X86_5LEVEL
 unsigned int __pgtable_l5_enabled __ro_after_init;
 unsigned int pgdir_shift __ro_after_init = 39;
 EXPORT_SYMBOL(pgdir_shift);
 unsigned int ptrs_per_p4d __ro_after_init = 1;
 EXPORT_SYMBOL(ptrs_per_p4d);
-#endif
 
 unsigned long page_offset_base __ro_after_init = __PAGE_OFFSET_BASE_L4;
 EXPORT_SYMBOL(page_offset_base);
@@ -69,9 +67,6 @@ EXPORT_SYMBOL(vmemmap_base);
 
 static inline bool check_la57_support(void)
 {
-	if (!IS_ENABLED(CONFIG_X86_5LEVEL))
-		return false;
-
 	/*
 	 * 5-level paging is detected and enabled at kernel decompression
 	 * stage. Only check if it has been enabled there.
diff --git a/arch/x86/kernel/head_64.S b/arch/x86/kernel/head_64.S
index 330922b328bf..4b2b2138c163 100644
--- a/arch/x86/kernel/head_64.S
+++ b/arch/x86/kernel/head_64.S
@@ -659,12 +659,10 @@ SYM_DATA_START_PTI_ALIGNED(init_top_pgt)
 SYM_DATA_END(init_top_pgt)
 #endif
 
-#ifdef CONFIG_X86_5LEVEL
 SYM_DATA_START_PAGE_ALIGNED(level4_kernel_pgt)
 	.fill	511,8,0
 	.quad	level3_kernel_pgt - __START_KERNEL_map + _PAGE_TABLE_NOENC
 SYM_DATA_END(level4_kernel_pgt)
-#endif
 
 SYM_DATA_START_PAGE_ALIGNED(level3_kernel_pgt)
 	.fill	L3_START_KERNEL,8,0
diff --git a/arch/x86/mm/init.c b/arch/x86/mm/init.c
index eb503f53c319..5a980a452f4c 100644
--- a/arch/x86/mm/init.c
+++ b/arch/x86/mm/init.c
@@ -173,11 +173,7 @@ __ref void *alloc_low_pages(unsigned int num)
  * randomization is enabled.
  */
 
-#ifndef CONFIG_X86_5LEVEL
-#define INIT_PGD_PAGE_TABLES    3
-#else
 #define INIT_PGD_PAGE_TABLES    4
-#endif
 
 #ifndef CONFIG_RANDOMIZE_MEMORY
 #define INIT_PGD_PAGE_COUNT      (2 * INIT_PGD_PAGE_TABLES)
diff --git a/arch/x86/mm/pgtable.c b/arch/x86/mm/pgtable.c
index 93e54ba91fbf..a3ef75c06c58 100644
--- a/arch/x86/mm/pgtable.c
+++ b/arch/x86/mm/pgtable.c
@@ -691,7 +691,7 @@ void native_set_fixmap(unsigned /* enum fixed_addresses */ idx,
 }
 
 #ifdef CONFIG_HAVE_ARCH_HUGE_VMAP
-#ifdef CONFIG_X86_5LEVEL
+#if CONFIG_PGTABLE_LEVELS > 4
 /**
  * p4d_set_huge - setup kernel P4D mapping
  *
diff --git a/drivers/firmware/efi/libstub/x86-5lvl.c b/drivers/firmware/efi/libstub/x86-5lvl.c
index 77359e802181..f1c5fb45d5f7 100644
--- a/drivers/firmware/efi/libstub/x86-5lvl.c
+++ b/drivers/firmware/efi/libstub/x86-5lvl.c
@@ -62,7 +62,7 @@ efi_status_t efi_setup_5level_paging(void)
 
 void efi_5level_switch(void)
 {
-	bool want_la57 = IS_ENABLED(CONFIG_X86_5LEVEL) && !efi_no5lvl;
+	bool want_la57 = !efi_no5lvl;
 	bool have_la57 = native_read_cr4() & X86_CR4_LA57;
 	bool need_toggle = want_la57 ^ have_la57;
 	u64 *pgt = (void *)la57_toggle + PAGE_SIZE;
diff --git a/tools/arch/x86/include/asm/disabled-features.h b/tools/arch/x86/include/asm/disabled-features.h
index c492bdc97b05..19cf1678fcaa 100644
--- a/tools/arch/x86/include/asm/disabled-features.h
+++ b/tools/arch/x86/include/asm/disabled-features.h
@@ -38,12 +38,6 @@
 # define DISABLE_OSPKE		(1<<(X86_FEATURE_OSPKE & 31))
 #endif /* CONFIG_X86_INTEL_MEMORY_PROTECTION_KEYS */
 
-#ifdef CONFIG_X86_5LEVEL
-# define DISABLE_LA57	0
-#else
-# define DISABLE_LA57	(1<<(X86_FEATURE_LA57 & 31))
-#endif
-
 #ifdef CONFIG_MITIGATION_PAGE_TABLE_ISOLATION
 # define DISABLE_PTI		0
 #else
@@ -149,8 +143,7 @@
 #define DISABLED_MASK13	0
 #define DISABLED_MASK14	0
 #define DISABLED_MASK15	0
-#define DISABLED_MASK16	(DISABLE_PKU|DISABLE_OSPKE|DISABLE_LA57|DISABLE_UMIP| \
-			 DISABLE_ENQCMD)
+#define DISABLED_MASK16	(DISABLE_PKU|DISABLE_OSPKE|DISABLE_UMIP|DISABLE_ENQCMD)
 #define DISABLED_MASK17	0
 #define DISABLED_MASK18	(DISABLE_IBT)
 #define DISABLED_MASK19	(DISABLE_SEV_SNP)
-- 
2.43.0


^ permalink raw reply	[flat|nested] 15+ messages in thread

* Re: [PATCH] x86/64/mm: Make 5-level paging support unconditional
  2024-06-26 10:26     ` [PATCH] " Kirill A. Shutemov
@ 2024-06-26 10:46       ` Borislav Petkov
  2024-06-26 12:00         ` Kirill A. Shutemov
  0 siblings, 1 reply; 15+ messages in thread
From: Borislav Petkov @ 2024-06-26 10:46 UTC (permalink / raw)
  To: Kirill A. Shutemov
  Cc: lkp, ardb, brijesh.singh, corbet, dave.hansen, hpa, jan.kiszka,
	jgross, kbingham, linux-doc, linux-efi, linux-kernel, linux-mm,
	luto, michael.roth, mingo, oe-kbuild-all, peterz,
	rick.p.edgecombe, sandipan.das, tglx, thomas.lendacky, x86

On Wed, Jun 26, 2024 at 01:26:23PM +0300, Kirill A. Shutemov wrote:
> Both Intel and AMD CPUs support 5-level paging, which is expected to
> become more widely adopted in the future.
> 
> Remove CONFIG_X86_5LEVEL and ifdeffery for it to make it more readable.
> 
> Signed-off-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
> Suggested-by: Borislav Petkov <bp@alien8.de>
> ---
> 
>  v2:
>   - Fix 32-bit build by wrapping p4d_set_huge() and p4d_clear_huge() in
>     #if CONFIG_PGTABLE_LEVELS > 4

Is the 0day bot smart enough to detect separate v2 fixups like that and
re-test using them?

If not, you can push your tree to your repo so that it can chew on it again
and report any other potential breakages...

Thx.

-- 
Regards/Gruss,
    Boris.

https://people.kernel.org/tglx/notes-about-netiquette

^ permalink raw reply	[flat|nested] 15+ messages in thread

* Re: [PATCH] x86/64/mm: Make 5-level paging support unconditional
  2024-06-26 10:46       ` Borislav Petkov
@ 2024-06-26 12:00         ` Kirill A. Shutemov
  0 siblings, 0 replies; 15+ messages in thread
From: Kirill A. Shutemov @ 2024-06-26 12:00 UTC (permalink / raw)
  To: Borislav Petkov
  Cc: lkp, ardb, brijesh.singh, corbet, dave.hansen, hpa, jan.kiszka,
	jgross, kbingham, linux-doc, linux-efi, linux-kernel, linux-mm,
	luto, michael.roth, mingo, oe-kbuild-all, peterz,
	rick.p.edgecombe, sandipan.das, tglx, thomas.lendacky, x86

On Wed, Jun 26, 2024 at 12:46:48PM +0200, Borislav Petkov wrote:
> On Wed, Jun 26, 2024 at 01:26:23PM +0300, Kirill A. Shutemov wrote:
> > Both Intel and AMD CPUs support 5-level paging, which is expected to
> > become more widely adopted in the future.
> > 
> > Remove CONFIG_X86_5LEVEL and ifdeffery for it to make it more readable.
> > 
> > Signed-off-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
> > Suggested-by: Borislav Petkov <bp@alien8.de>
> > ---
> > 
> >  v2:
> >   - Fix 32-bit build by wrapping p4d_set_huge() and p4d_clear_huge() in
> >     #if CONFIG_PGTABLE_LEVELS > 4
> 
> Is the 0day bot smart enough to detect separate v2 fixups like that and
> re-test using them?
> 
> If not, you can push your tree to your repo so that it can chew on it again
> and report any other potential breakages...

Pushed:

https://git.kernel.org/pub/scm/linux/kernel/git/kas/linux.git x86/no5lvl-config

-- 
  Kiryl Shutsemau / Kirill A. Shutemov

^ permalink raw reply	[flat|nested] 15+ messages in thread

* Re: [PATCH 0/3] x86: Make 5-level paging support unconditional for x86-64
  2024-07-31 17:45     ` Shivank Garg
@ 2024-10-31 15:36       ` Dave Hansen
  0 siblings, 0 replies; 15+ messages in thread
From: Dave Hansen @ 2024-10-31 15:36 UTC (permalink / raw)
  To: Shivank Garg, Kirill A. Shutemov, Thomas Gleixner
  Cc: ardb, bp, brijesh.singh, corbet, dave.hansen, hpa, jan.kiszka,
	jgross, kbingham, linux-doc, linux-efi, linux-kernel, linux-mm,
	luto, michael.roth, mingo, peterz, rick.p.edgecombe,
	sandipan.das, thomas.lendacky, x86

On 7/31/24 10:45, Shivank Garg wrote:
> It would also be nice to get perf traces. Maybe it is purely SW issue.

Cycle counts aren't going to help much here.  For instance, if 5-level
paging makes *ALL* TLB misses slower, you would just see a regression in
any code that misses the TLB, which could show up all over.

On Intel we have some PMU events like this:

dtlb_store_misses.walk_active
       [Cycles when at least one PMH is busy
	with a page walk for a store]

(there's a load side one as well).  If a page walk gets more expensive,
you can see it there.  Note that this doesn't actually tell you how much
time the core spent _waiting_ for a page walk to complete.  If all the
speculation magic works perfectly in your favor, you could have the PMH
busy 100% of cycles but never had the core waiting on it.

So could we drill down a level in the "perf traces" please, and gather
some of the relevant performance counters and not just cycles?

^ permalink raw reply	[flat|nested] 15+ messages in thread

* Re: [PATCH 0/3] x86: Make 5-level paging support unconditional for x86-64
  2024-07-31 11:36   ` Kirill A. Shutemov
  2024-07-31 11:40     ` Peter Zijlstra
@ 2024-07-31 17:45     ` Shivank Garg
  2024-10-31 15:36       ` Dave Hansen
  1 sibling, 1 reply; 15+ messages in thread
From: Shivank Garg @ 2024-07-31 17:45 UTC (permalink / raw)
  To: Kirill A. Shutemov, Thomas Gleixner
  Cc: ardb, bp, brijesh.singh, corbet, dave.hansen, hpa, jan.kiszka,
	jgross, kbingham, linux-doc, linux-efi, linux-kernel, linux-mm,
	luto, michael.roth, mingo, peterz, rick.p.edgecombe,
	sandipan.das, thomas.lendacky, x86

On 7/31/2024 5:06 PM, Kirill A. Shutemov wrote:
> On Wed, Jul 31, 2024 at 11:15:05AM +0200, Thomas Gleixner wrote:
>> On Wed, Jul 31 2024 at 14:27, Shivank Garg wrote:
>>> lmbench:lat_pagefault: Metric- page-fault time (us) - Lower is better
>>>                 4-Level PT              5-Level PT		% Change
>>> THP-never       Mean:0.4068             Mean:0.4294		5.56
>>>                 95% CI:0.4057-0.4078    95% CI:0.4287-0.4302
>>>
>>> THP-Always      Mean: 0.4061            Mean: 0.4288		% Change
>>>                 95% CI: 0.4051-0.4071   95% CI: 0.4281-0.4295	5.59
>>>
>>> Inference:
>>> 5-level page table shows increase in page-fault latency but it does
>>> not significantly impact other benchmarks.
>>
>> 5% regression on lmbench is a NONO.
> 
> Yeah, that's a biggy.
> 
> In our testing (on Intel HW) we didn't see any significant difference
> between 4- and 5-level paging. But we were focused on TLB fill latency.
> In both bare metal and in VMs. Maybe something wrong in the fault path?
> 
> It requires a closer look.
> 
> Shivank, could you share how you run lat_pagefault? What file size? How
> parallel you run it?...

Hi Kirill,

I got lmbench from here:
https://github.com/foss-for-synopsys-dwc-arc-processors/lmbench/blob/master/src/lat_pagefault.c

and using this command:
numactl --membind=1 --cpunodebind=1 bin/x86_64-linux-gnu/lat_pagefault -N 100 1GB_dev_urandom_file

> 
> It would also be nice to get perf traces. Maybe it is purely SW issue.
> 

4-level-page-table:
      - 52.31% benchmark
         - 49.52% asm_exc_page_fault
            - 49.35% exc_page_fault
               - 48.36% do_user_addr_fault
                  - 46.15% handle_mm_fault
                     - 44.59% __handle_mm_fault
                        - 42.95% do_fault
                           - 40.89% filemap_map_pages
                              - 28.30% set_pte_range
                                 - 23.70% folio_add_file_rmap_ptes
                                    - 14.30% __lruvec_stat_mod_folio
                                       - 10.12% __mod_lruvec_state
                                          - 5.70% __mod_memcg_lruvec_state
                                               0.60% cgroup_rstat_updated
                                            1.06% __mod_node_page_state
                                      2.84% __rcu_read_unlock
                                      0.76% srso_alias_safe_ret
                                   0.84% set_ptes.isra.0
                              - 5.48% next_uptodate_folio
                                 - 1.19% xas_find
                                      0.96% xas_load
                                1.00% set_ptes.isra.0
                    1.22% lock_vma_under_rcu


5-level-page-table:
      - 52.75% benchmark
         - 50.04% asm_exc_page_fault
            - 49.90% exc_page_fault
               - 48.91% do_user_addr_fault
                  - 46.74% handle_mm_fault
                     - 45.27% __handle_mm_fault
                        - 43.30% do_fault
                           - 41.58% filemap_map_pages
                              - 28.04% set_pte_range
                                 - 22.77% folio_add_file_rmap_ptes
                                    - 17.74% __lruvec_stat_mod_folio
                                       - 10.89% __mod_lruvec_state
                                          - 5.97% __mod_memcg_lruvec_state
                                               1.94% cgroup_rstat_updated
                                            1.09% __mod_node_page_state
                                         0.56% __mod_node_page_state
                                      2.28% __rcu_read_unlock
                                   1.08% set_ptes.isra.0
                              - 5.94% next_uptodate_folio
                                 - 1.13% xas_find
                                      0.99% xas_load
                                1.13% srso_alias_safe_ret
                                0.52% set_ptes.isra.0
                    1.16% lock_vma_under_rcu

>> 5-level page tables add a cost in every hardware page table walk. That's
>> a matter of fact and there is absolutely no reason to inflict this cost
>> on everyone.
>>
>> The solution to this to make the 5-level mechanics smarter by evaluating
>> whether the machine has enough memory to require 5-level tables and
>> select the depth at boot time.
> 
> Let's understand the reason first.

Sure, please let me know how can I help in this debug.

Thanks,
Shivank

> 
> The risk with your proposal is that 5-level paging will not get any
> testing and rot over time.
> 
> I would like to keep it on, if possible.
> 


^ permalink raw reply	[flat|nested] 15+ messages in thread

* Re: [PATCH 0/3] x86: Make 5-level paging support unconditional for x86-64
  2024-07-31 11:36   ` Kirill A. Shutemov
@ 2024-07-31 11:40     ` Peter Zijlstra
  2024-07-31 17:45     ` Shivank Garg
  1 sibling, 0 replies; 15+ messages in thread
From: Peter Zijlstra @ 2024-07-31 11:40 UTC (permalink / raw)
  To: Kirill A. Shutemov
  Cc: Thomas Gleixner, Shivank Garg, ardb, bp, brijesh.singh, corbet,
	dave.hansen, hpa, jan.kiszka, jgross, kbingham, linux-doc,
	linux-efi, linux-kernel, linux-mm, luto, michael.roth, mingo,
	rick.p.edgecombe, sandipan.das, thomas.lendacky, x86

On Wed, Jul 31, 2024 at 02:36:47PM +0300, Kirill A. Shutemov wrote:
> The risk with your proposal is that 5-level paging will not get any
> testing and rot over time.
> 
> I would like to keep it on, if possible.

Well, if it is boot time, you just tell your CI to force enable 5level
and you're done, right? Then the rest of us use 4 and we all good.

^ permalink raw reply	[flat|nested] 15+ messages in thread

* Re: [PATCH 0/3] x86: Make 5-level paging support unconditional for x86-64
  2024-07-31  9:15 ` Thomas Gleixner
  2024-07-31 11:11   ` Peter Zijlstra
@ 2024-07-31 11:36   ` Kirill A. Shutemov
  2024-07-31 11:40     ` Peter Zijlstra
  2024-07-31 17:45     ` Shivank Garg
  1 sibling, 2 replies; 15+ messages in thread
From: Kirill A. Shutemov @ 2024-07-31 11:36 UTC (permalink / raw)
  To: Thomas Gleixner, Shivank Garg
  Cc: ardb, bp, brijesh.singh, corbet, dave.hansen, hpa, jan.kiszka,
	jgross, kbingham, linux-doc, linux-efi, linux-kernel, linux-mm,
	luto, michael.roth, mingo, peterz, rick.p.edgecombe,
	sandipan.das, thomas.lendacky, x86

On Wed, Jul 31, 2024 at 11:15:05AM +0200, Thomas Gleixner wrote:
> On Wed, Jul 31 2024 at 14:27, Shivank Garg wrote:
> > lmbench:lat_pagefault: Metric- page-fault time (us) - Lower is better
> >                 4-Level PT              5-Level PT		% Change
> > THP-never       Mean:0.4068             Mean:0.4294		5.56
> >                 95% CI:0.4057-0.4078    95% CI:0.4287-0.4302
> >
> > THP-Always      Mean: 0.4061            Mean: 0.4288		% Change
> >                 95% CI: 0.4051-0.4071   95% CI: 0.4281-0.4295	5.59
> >
> > Inference:
> > 5-level page table shows increase in page-fault latency but it does
> > not significantly impact other benchmarks.
> 
> 5% regression on lmbench is a NONO.

Yeah, that's a biggy.

In our testing (on Intel HW) we didn't see any significant difference
between 4- and 5-level paging. But we were focused on TLB fill latency.
In both bare metal and in VMs. Maybe something wrong in the fault path?

It requires a closer look.

Shivank, could you share how you run lat_pagefault? What file size? How
parallel you run it?...

It would also be nice to get perf traces. Maybe it is purely SW issue.

> 5-level page tables add a cost in every hardware page table walk. That's
> a matter of fact and there is absolutely no reason to inflict this cost
> on everyone.
>
> The solution to this to make the 5-level mechanics smarter by evaluating
> whether the machine has enough memory to require 5-level tables and
> select the depth at boot time.

Let's understand the reason first.

The risk with your proposal is that 5-level paging will not get any
testing and rot over time.

I would like to keep it on, if possible.

-- 
  Kiryl Shutsemau / Kirill A. Shutemov

^ permalink raw reply	[flat|nested] 15+ messages in thread

* Re: [PATCH 0/3] x86: Make 5-level paging support unconditional for x86-64
  2024-07-31  9:15 ` Thomas Gleixner
@ 2024-07-31 11:11   ` Peter Zijlstra
  2024-07-31 11:36   ` Kirill A. Shutemov
  1 sibling, 0 replies; 15+ messages in thread
From: Peter Zijlstra @ 2024-07-31 11:11 UTC (permalink / raw)
  To: Thomas Gleixner
  Cc: 20240621164406.256314-1-kirill.shutemov, kirill.shutemov, ardb,
	bp, brijesh.singh, corbet, dave.hansen, hpa, jan.kiszka, jgross,
	kbingham, linux-doc, linux-efi, linux-kernel, linux-mm, luto,
	michael.roth, mingo, rick.p.edgecombe, sandipan.das,
	thomas.lendacky, x86

On Wed, Jul 31, 2024 at 11:15:05AM +0200, Thomas Gleixner wrote:
> On Wed, Jul 31 2024 at 14:27, Shivank Garg wrote:
> > lmbench:lat_pagefault: Metric- page-fault time (us) - Lower is better
> >                 4-Level PT              5-Level PT		% Change
> > THP-never       Mean:0.4068             Mean:0.4294		5.56
> >                 95% CI:0.4057-0.4078    95% CI:0.4287-0.4302
> >
> > THP-Always      Mean: 0.4061            Mean: 0.4288		% Change
> >                 95% CI: 0.4051-0.4071   95% CI: 0.4281-0.4295	5.59
> >
> > Inference:
> > 5-level page table shows increase in page-fault latency but it does
> > not significantly impact other benchmarks.
> 
> 5% regression on lmbench is a NONO.
> 
> 5-level page tables add a cost in every hardware page table walk. That's
> a matter of fact and there is absolutely no reason to inflict this cost
> on everyone.
> 
> The solution to this to make the 5-level mechanics smarter by evaluating
> whether the machine has enough memory to require 5-level tables and
> select the depth at boot time.

I gotta mention (again) that its a pain we can't mix and match like
s390. They default run their userspace on 4 level, even if the kernel
runs 5. Only silly daft userspace that needs more than insane amounts of
memory get 5 level.

^ permalink raw reply	[flat|nested] 15+ messages in thread

* Re: [PATCH 0/3] x86: Make 5-level paging support unconditional for x86-64
  2024-07-31  8:57 [PATCH 0/3] x86: Make 5-level paging support unconditional for x86-64 Shivank Garg
@ 2024-07-31  9:15 ` Thomas Gleixner
  2024-07-31 11:11   ` Peter Zijlstra
  2024-07-31 11:36   ` Kirill A. Shutemov
  0 siblings, 2 replies; 15+ messages in thread
From: Thomas Gleixner @ 2024-07-31  9:15 UTC (permalink / raw)
  To: 20240621164406.256314-1-kirill.shutemov, kirill.shutemov
  Cc: ardb, bp, brijesh.singh, corbet, dave.hansen, hpa, jan.kiszka,
	jgross, kbingham, linux-doc, linux-efi, linux-kernel, linux-mm,
	luto, michael.roth, mingo, peterz, rick.p.edgecombe,
	sandipan.das, thomas.lendacky, x86

On Wed, Jul 31 2024 at 14:27, Shivank Garg wrote:
> lmbench:lat_pagefault: Metric- page-fault time (us) - Lower is better
>                 4-Level PT              5-Level PT		% Change
> THP-never       Mean:0.4068             Mean:0.4294		5.56
>                 95% CI:0.4057-0.4078    95% CI:0.4287-0.4302
>
> THP-Always      Mean: 0.4061            Mean: 0.4288		% Change
>                 95% CI: 0.4051-0.4071   95% CI: 0.4281-0.4295	5.59
>
> Inference:
> 5-level page table shows increase in page-fault latency but it does
> not significantly impact other benchmarks.

5% regression on lmbench is a NONO.

5-level page tables add a cost in every hardware page table walk. That's
a matter of fact and there is absolutely no reason to inflict this cost
on everyone.

The solution to this to make the 5-level mechanics smarter by evaluating
whether the machine has enough memory to require 5-level tables and
select the depth at boot time.

Thanks,

        tglx

^ permalink raw reply	[flat|nested] 15+ messages in thread

* Re: [PATCH 0/3] x86: Make 5-level paging support unconditional for x86-64
@ 2024-07-31  8:57 Shivank Garg
  2024-07-31  9:15 ` Thomas Gleixner
  0 siblings, 1 reply; 15+ messages in thread
From: Shivank Garg @ 2024-07-31  8:57 UTC (permalink / raw)
  To: kirill.shutemov
  Cc: ardb, bp, brijesh.singh, corbet, dave.hansen, hpa, jan.kiszka,
	jgross, kbingham, linux-doc, linux-efi, linux-kernel, linux-mm,
	luto, michael.roth, mingo, peterz, rick.p.edgecombe,
	sandipan.das, tglx, thomas.lendacky, x86

I did some experiments to understand the impact of making 5 level page tables
default.
Machine Info:  AMD Zen 4 EPYC server (2-socket system, 128 cores and 1 NUMA
node per socket, SMT Enabled)
Size of each NUMA node is approx 377 GB.

For experiments, I'm binding the benchmark to CPUs and memory nodes of single
socket for consistent results. Measured by enabling/disabling 5level Page
table using CONFIG_X86_5LEVEL.

% Change: (5L-4L)/4L*100
CoV (%):  Coefficient of Variation (%)

Results:

lmbench:lat_pagefault: Metric- page-fault time (us) - Lower is better
                4-Level PT              5-Level PT		% Change
THP-never       Mean:0.4068             Mean:0.4294		5.56
                95% CI:0.4057-0.4078    95% CI:0.4287-0.4302

THP-Always      Mean: 0.4061            Mean: 0.4288		% Change
                95% CI: 0.4051-0.4071   95% CI: 0.4281-0.4295	5.59


Btree (Thread:32): Metric- Time Taken (in seconds) - Lower is better 
                4-Level                 5-Level               
                Time Taken(s) CoV (%)   Time Taken(s) CoV(%)    % Change
THP Never       382.2         0.219     388.8         1.019     1.73
THP Madvise     383.0         0.261     384.8         0.809     0.47
THP Always      392.8         1.376     386.4         2.147     -1.63

Btree (Thread:256): Metric- Time Taken (in seconds) - Lower is better
                4-Level                 5-Level               
                Time Taken(s) CoV (%)   Time Taken(s) CoV(%)     % Change
THP Never       56.6          2.014     55.2          0.810     -2.47
THP Madvise     56.6          2.014     56.4          2.022     -0.35
THP Always      56.6          0.968     56.2          1.489     -0.71


Ebizzy: Metric- records/s - Higher is better
                4-Level                 5-Level
Threads         record/s    CoV (%)     record/s    CoV(%)      % Change
1               844         0.302       837         0.196       -0.85
256             10160       0.315       10288       1.081       1.26


XSBench (Thread:256, THP:Never) - Higher is better
Metric          4-Level         5-Level         % Change
Lookups/s       13720556        13396288        -2.36
CoV (%)         1.726           1.317


Hashjoin (Thread:256, THP:Never) - Lower is better
Metric          4-Level         5-Level         % Change
Time taken(s)   424.4           427.4           0.707
CoV (%)         0.394           0.209


Graph500(Thread:256, THP:Madvise) - Lower is better
Metric          4-Level         5-Level       % Change
Time Taken(s)   0.1879          0.1873        -0.32
CoV (%)         0.165           0.213


GUPS(Thread:128, THP:Madvise) - Higher is better
Metric          4-Level         5-Level       % Change
GUPS            1.3265          1.3252        -0.10
CoV (%)         0.037           0.027


pagerank(Thread:256, THP:Madvise) - Lower is better
Metric          4-Level         5-Level       % Change
Time taken(s)   143.67          143.67        0.00
CoV (%)         0.402           0.402


Redis(Thread:256, THP:Madvise) - Higher is better
Metric              4-Level     5-Level       % Change
Throughput(Ops/s)   141030744   139586376     -1.02
CoV (%)             0.372       0.561


memcached(Thread:256, THP:Madvise) - Higher is better
Metric              4-Level     5-Level       % Change
Throughput(Ops/s)   19916313    19743637      -0.87
CoV (%)             0.051       0.095


Inference:
5-level page table shows increase in page-fault latency but it does
not significantly impact other benchmarks.


Thanks,
Shivank

^ permalink raw reply	[flat|nested] 15+ messages in thread

end of thread, other threads:[~2024-10-31 15:36 UTC | newest]

Thread overview: 15+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2024-06-21 16:44 [PATCH 0/3] x86: Make 5-level paging support unconditional for x86-64 Kirill A. Shutemov
2024-06-21 16:44 ` [PATCH 1/3] x86/64/mm: Always use dynamic memory layout Kirill A. Shutemov
2024-06-21 16:44 ` [PATCH 2/3] x86/64/mm: Make SPARSEMEM_VMEMMAP the only memory model Kirill A. Shutemov
2024-06-21 16:44 ` [PATCH 3/3] x86/64/mm: Make 5-level paging support unconditional Kirill A. Shutemov
2024-06-25 23:21   ` kernel test robot
2024-06-26 10:26     ` [PATCH] " Kirill A. Shutemov
2024-06-26 10:46       ` Borislav Petkov
2024-06-26 12:00         ` Kirill A. Shutemov
2024-07-31  8:57 [PATCH 0/3] x86: Make 5-level paging support unconditional for x86-64 Shivank Garg
2024-07-31  9:15 ` Thomas Gleixner
2024-07-31 11:11   ` Peter Zijlstra
2024-07-31 11:36   ` Kirill A. Shutemov
2024-07-31 11:40     ` Peter Zijlstra
2024-07-31 17:45     ` Shivank Garg
2024-10-31 15:36       ` Dave Hansen

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox

all inboxes | Powered by JetHome®