* [OpenRISC] [PATCH V3 00/30] mm/mmap: Drop protection_map[] and platform's __SXXX/__PXXX requirements
@ 2022-02-28 10:47 Anshuman Khandual
2022-02-28 10:47 ` [OpenRISC] [PATCH V3 01/30] mm/debug_vm_pgtable: Drop protection_map[] usage Anshuman Khandual
` (29 more replies)
0 siblings, 30 replies; 49+ messages in thread
From: Anshuman Khandual @ 2022-02-28 10:47 UTC (permalink / raw)
To: openrisc
protection_map[] is an array based construct that translates given vm_flags
combination. This array contains page protection map, which is populated by
the platform via [__S000 .. __S111] and [__P000 .. __P111] exported macros.
Primary usage for protection_map[] is for vm_get_page_prot(), which is used
to determine page protection value for a given vm_flags. vm_get_page_prot()
implementation, could again call platform overrides arch_vm_get_page_prot()
and arch_filter_pgprot(). Some platforms override protection_map[] that was
originally built with __SXXX/__PXXX with different runtime values.
Currently there are multiple layers of abstraction i.e __SXXX/__PXXX macros
, protection_map[], arch_vm_get_page_prot() and arch_filter_pgprot() built
between the platform and generic MM, finally defining vm_get_page_prot().
Hence this series proposes to drop all these abstraction levels and instead
just move the responsibility of defining vm_get_page_prot() to the platform
itself making it clean and simple.
This first introduces ARCH_HAS_VM_GET_PAGE_PROT which enables the platforms
to define custom vm_get_page_prot(). This starts converting platforms that
either change protection_map[] or define the overrides arch_filter_pgprot()
or arch_vm_get_page_prot() which enables for those constructs to be dropped
off completely. This series then converts remaining platforms which enables
for __SXXX/__PXXX constructs to be dropped off completely. Finally it drops
the generic vm_get_page_prot() and then ARCH_HAS_VM_GET_PAGE_PROT as every
platform now defines their own vm_get_page_prot().
The series has been inspired from an earlier discuss with Christoph Hellwig
https://lore.kernel.org/all/1632712920-8171-1-git-send-email-anshuman.khandual at arm.com/
This series applies on 5.17-rc6 after the following patch.
https://lore.kernel.org/all/1643004823-16441-1-git-send-email-anshuman.khandual at arm.com/
This series has been cross built for multiple platforms.
- Anshuman
Cc: Christoph Hellwig <hch@infradead.org>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: linuxppc-dev at lists.ozlabs.org
Cc: linux-arm-kernel at lists.infradead.org
Cc: sparclinux at vger.kernel.org
Cc: linux-mips at vger.kernel.org
Cc: linux-m68k at lists.linux-m68k.org
Cc: linux-s390 at vger.kernel.org
Cc: linux-riscv at lists.infradead.org
Cc: linux-alpha at vger.kernel.org
Cc: linux-sh at vger.kernel.org
Cc: linux-snps-arc at lists.infradead.org
Cc: linux-csky at vger.kernel.org
Cc: linux-xtensa at linux-xtensa.org
Cc: linux-parisc at vger.kernel.org
Cc: openrisc at lists.librecores.org
Cc: linux-um at lists.infradead.org
Cc: linux-hexagon at vger.kernel.org
Cc: linux-ia64 at vger.kernel.org
Cc: linux-mm at kvack.org
Cc: linux-arch at vger.kernel.org
Cc: linux-kernel at vger.kernel.org
Changes in V3:
- Dropped variable 'i' from sme_early_init() on x86 platform
- Moved CONFIG_COLDFIRE vm_get_page_prot() inside arch/m68k/mm/mcfmmu.c
- Moved CONFIG_SUN3 vm_get_page_prot() inside arch/m68k/mm/sun3mmu.c
- Dropped cachebits for vm_get_page_prot() inside arch/m68k/mm/motorola.c
- Dropped PAGE_XXX_C definitions from arch/m68k/include/asm/motorola_pgtable.h
- Used PAGE_XXX instead for vm_get_page_prot() inside arch/m68k/mm/motorola.c
- Dropped all references to protection_map[] in the tree
- Replaced s/extensa/xtensa/ on the patch title
- Moved left over comments from pgtable.h into init.c on nios2 platform
Changes in V2:
https://lore.kernel.org/all/1645425519-9034-1-git-send-email-anshuman.khandual at arm.com/
- Dropped the entire comment block in [PATCH 30/30] per Geert
- Replaced __P010 (although commented) with __PAGE_COPY on arm platform
- Replaced __P101 with PAGE_READONLY on um platform
Changes in V1:
https://lore.kernel.org/all/1644805853-21338-1-git-send-email-anshuman.khandual at arm.com/
- Add white spaces around the | operators
- Moved powerpc_vm_get_page_prot() near vm_get_page_prot() on powerpc
- Moved arm64_vm_get_page_prot() near vm_get_page_prot() on arm64
- Moved sparc_vm_get_page_prot() near vm_get_page_prot() on sparc
- Compacted vm_get_page_prot() switch cases on all platforms
- _PAGE_CACHE040 inclusion is dependent on CPU_IS_040_OR_060
- VM_SHARED case should return PAGE_NONE (not PAGE_COPY) on SH platform
- Reorganized VM_SHARED, VM_EXEC, VM_WRITE, VM_READ
- Dropped the last patch [RFC V1 31/31] which added macros for vm_flags combinations
https://lore.kernel.org/all/1643029028-12710-32-git-send-email-anshuman.khandual at arm.com/
Changes in RFC:
https://lore.kernel.org/all/1643029028-12710-1-git-send-email-anshuman.khandual at arm.com/
Anshuman Khandual (29):
mm/debug_vm_pgtable: Drop protection_map[] usage
mm/mmap: Clarify protection_map[] indices
mm/mmap: Add new config ARCH_HAS_VM_GET_PAGE_PROT
powerpc/mm: Enable ARCH_HAS_VM_GET_PAGE_PROT
arm64/mm: Enable ARCH_HAS_VM_GET_PAGE_PROT
sparc/mm: Enable ARCH_HAS_VM_GET_PAGE_PROT
mips/mm: Enable ARCH_HAS_VM_GET_PAGE_PROT
m68k/mm: Enable ARCH_HAS_VM_GET_PAGE_PROT
arm/mm: Enable ARCH_HAS_VM_GET_PAGE_PROT
mm/mmap: Drop protection_map[]
mm/mmap: Drop arch_filter_pgprot()
mm/mmap: Drop arch_vm_get_page_pgprot()
s390/mm: Enable ARCH_HAS_VM_GET_PAGE_PROT
riscv/mm: Enable ARCH_HAS_VM_GET_PAGE_PROT
alpha/mm: Enable ARCH_HAS_VM_GET_PAGE_PROT
sh/mm: Enable ARCH_HAS_VM_GET_PAGE_PROT
arc/mm: Enable ARCH_HAS_VM_GET_PAGE_PROT
csky/mm: Enable ARCH_HAS_VM_GET_PAGE_PROT
xtensa/mm: Enable ARCH_HAS_VM_GET_PAGE_PROT
parisc/mm: Enable ARCH_HAS_VM_GET_PAGE_PROT
openrisc/mm: Enable ARCH_HAS_VM_GET_PAGE_PROT
um/mm: Enable ARCH_HAS_VM_GET_PAGE_PROT
microblaze/mm: Enable ARCH_HAS_VM_GET_PAGE_PROT
nios2/mm: Enable ARCH_HAS_VM_GET_PAGE_PROT
hexagon/mm: Enable ARCH_HAS_VM_GET_PAGE_PROT
nds32/mm: Enable ARCH_HAS_VM_GET_PAGE_PROT
ia64/mm: Enable ARCH_HAS_VM_GET_PAGE_PROT
mm/mmap: Drop generic vm_get_page_prot()
mm/mmap: Drop ARCH_HAS_VM_GET_PAGE_PROT
Christoph Hellwig (1):
x86/mm: Enable ARCH_HAS_VM_GET_PAGE_PROT
arch/alpha/include/asm/pgtable.h | 17 -----
arch/alpha/mm/init.c | 37 +++++++++++
arch/arc/include/asm/pgtable-bits-arcv2.h | 17 -----
arch/arc/mm/mmap.c | 41 ++++++++++++
arch/arm/include/asm/pgtable.h | 20 +-----
arch/arm/lib/uaccess_with_memcpy.c | 2 +-
arch/arm/mm/mmu.c | 44 +++++++++++--
arch/arm64/Kconfig | 1 -
arch/arm64/include/asm/memory.h | 3 +-
arch/arm64/include/asm/mman.h | 24 -------
arch/arm64/include/asm/pgtable-prot.h | 18 ------
arch/arm64/include/asm/pgtable.h | 11 ----
arch/arm64/mm/mmap.c | 78 +++++++++++++++++++++++
arch/csky/include/asm/pgtable.h | 18 ------
arch/csky/mm/init.c | 32 ++++++++++
arch/hexagon/include/asm/pgtable.h | 24 -------
arch/hexagon/mm/init.c | 67 +++++++++++++++++++
arch/ia64/include/asm/pgtable.h | 17 -----
arch/ia64/mm/init.c | 41 +++++++++++-
arch/m68k/include/asm/mcf_pgtable.h | 59 -----------------
arch/m68k/include/asm/motorola_pgtable.h | 29 ---------
arch/m68k/include/asm/sun3_pgtable.h | 22 -------
arch/m68k/mm/mcfmmu.c | 59 +++++++++++++++++
arch/m68k/mm/motorola.c | 43 ++++++++++++-
arch/m68k/mm/sun3mmu.c | 39 ++++++++++++
arch/microblaze/include/asm/pgtable.h | 17 -----
arch/microblaze/mm/init.c | 41 ++++++++++++
arch/mips/include/asm/pgtable.h | 22 -------
arch/mips/mm/cache.c | 60 +++++++++--------
arch/nds32/include/asm/pgtable.h | 17 -----
arch/nds32/mm/mmap.c | 37 +++++++++++
arch/nios2/include/asm/pgtable.h | 24 -------
arch/nios2/mm/init.c | 47 ++++++++++++++
arch/openrisc/include/asm/pgtable.h | 18 ------
arch/openrisc/mm/init.c | 41 ++++++++++++
arch/parisc/include/asm/pgtable.h | 20 ------
arch/parisc/mm/init.c | 40 ++++++++++++
arch/powerpc/include/asm/mman.h | 12 ----
arch/powerpc/include/asm/pgtable.h | 19 ------
arch/powerpc/mm/mmap.c | 59 +++++++++++++++++
arch/riscv/include/asm/pgtable.h | 16 -----
arch/riscv/mm/init.c | 42 ++++++++++++
arch/s390/include/asm/pgtable.h | 17 -----
arch/s390/mm/mmap.c | 33 ++++++++++
arch/sh/include/asm/pgtable.h | 17 -----
arch/sh/mm/mmap.c | 38 +++++++++++
arch/sparc/include/asm/mman.h | 6 --
arch/sparc/include/asm/pgtable_32.h | 19 ------
arch/sparc/include/asm/pgtable_64.h | 19 ------
arch/sparc/mm/init_32.c | 35 ++++++++++
arch/sparc/mm/init_64.c | 70 ++++++++++++++------
arch/um/include/asm/pgtable.h | 17 -----
arch/um/kernel/mem.c | 35 ++++++++++
arch/x86/Kconfig | 1 -
arch/x86/include/asm/pgtable.h | 5 --
arch/x86/include/asm/pgtable_types.h | 19 ------
arch/x86/include/uapi/asm/mman.h | 14 ----
arch/x86/mm/Makefile | 2 +-
arch/x86/mm/mem_encrypt_amd.c | 6 --
arch/x86/mm/pgprot.c | 71 +++++++++++++++++++++
arch/x86/um/mem_32.c | 2 +-
arch/xtensa/include/asm/pgtable.h | 19 ------
arch/xtensa/mm/init.c | 35 ++++++++++
drivers/gpu/drm/drm_vm.c | 4 +-
include/linux/mm.h | 6 --
include/linux/mman.h | 4 --
mm/Kconfig | 3 -
mm/debug_vm_pgtable.c | 31 +++++----
mm/mmap.c | 44 +------------
69 files changed, 1138 insertions(+), 729 deletions(-)
create mode 100644 arch/x86/mm/pgprot.c
--
2.25.1
^ permalink raw reply [flat|nested] 49+ messages in thread
* [OpenRISC] [PATCH V3 01/30] mm/debug_vm_pgtable: Drop protection_map[] usage
2022-02-28 10:47 [OpenRISC] [PATCH V3 00/30] mm/mmap: Drop protection_map[] and platform's __SXXX/__PXXX requirements Anshuman Khandual
@ 2022-02-28 10:47 ` Anshuman Khandual
2022-02-28 10:47 ` [OpenRISC] [PATCH V3 02/30] mm/mmap: Clarify protection_map[] indices Anshuman Khandual
` (28 subsequent siblings)
29 siblings, 0 replies; 49+ messages in thread
From: Anshuman Khandual @ 2022-02-28 10:47 UTC (permalink / raw)
To: openrisc
Although protection_map[] contains the platform defined page protection map
for a given vm_flags combination, vm_get_page_prot() is the right interface
to use. This will also reduce dependency on protection_map[] which is going
to be dropped off completely later on.
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: linux-mm at kvack.org
Cc: linux-kernel at vger.kernel.org
Signed-off-by: Anshuman Khandual <anshuman.khandual@arm.com>
---
mm/debug_vm_pgtable.c | 31 +++++++++++++++++++------------
1 file changed, 19 insertions(+), 12 deletions(-)
diff --git a/mm/debug_vm_pgtable.c b/mm/debug_vm_pgtable.c
index db2abd9e415b..30fd11a2ed32 100644
--- a/mm/debug_vm_pgtable.c
+++ b/mm/debug_vm_pgtable.c
@@ -93,7 +93,7 @@ struct pgtable_debug_args {
static void __init pte_basic_tests(struct pgtable_debug_args *args, int idx)
{
- pgprot_t prot = protection_map[idx];
+ pgprot_t prot = vm_get_page_prot(idx);
pte_t pte = pfn_pte(args->fixed_pte_pfn, prot);
unsigned long val = idx, *ptr = &val;
@@ -101,7 +101,7 @@ static void __init pte_basic_tests(struct pgtable_debug_args *args, int idx)
/*
* This test needs to be executed after the given page table entry
- * is created with pfn_pte() to make sure that protection_map[idx]
+ * is created with pfn_pte() to make sure that vm_get_page_prot(idx)
* does not have the dirty bit enabled from the beginning. This is
* important for platforms like arm64 where (!PTE_RDONLY) indicate
* dirty bit being set.
@@ -190,7 +190,7 @@ static void __init pte_savedwrite_tests(struct pgtable_debug_args *args)
#ifdef CONFIG_TRANSPARENT_HUGEPAGE
static void __init pmd_basic_tests(struct pgtable_debug_args *args, int idx)
{
- pgprot_t prot = protection_map[idx];
+ pgprot_t prot = vm_get_page_prot(idx);
unsigned long val = idx, *ptr = &val;
pmd_t pmd;
@@ -202,7 +202,7 @@ static void __init pmd_basic_tests(struct pgtable_debug_args *args, int idx)
/*
* This test needs to be executed after the given page table entry
- * is created with pfn_pmd() to make sure that protection_map[idx]
+ * is created with pfn_pmd() to make sure that vm_get_page_prot(idx)
* does not have the dirty bit enabled from the beginning. This is
* important for platforms like arm64 where (!PTE_RDONLY) indicate
* dirty bit being set.
@@ -325,7 +325,7 @@ static void __init pmd_savedwrite_tests(struct pgtable_debug_args *args)
#ifdef CONFIG_HAVE_ARCH_TRANSPARENT_HUGEPAGE_PUD
static void __init pud_basic_tests(struct pgtable_debug_args *args, int idx)
{
- pgprot_t prot = protection_map[idx];
+ pgprot_t prot = vm_get_page_prot(idx);
unsigned long val = idx, *ptr = &val;
pud_t pud;
@@ -337,7 +337,7 @@ static void __init pud_basic_tests(struct pgtable_debug_args *args, int idx)
/*
* This test needs to be executed after the given page table entry
- * is created with pfn_pud() to make sure that protection_map[idx]
+ * is created with pfn_pud() to make sure that vm_get_page_prot(idx)
* does not have the dirty bit enabled from the beginning. This is
* important for platforms like arm64 where (!PTE_RDONLY) indicate
* dirty bit being set.
@@ -1106,14 +1106,14 @@ static int __init init_args(struct pgtable_debug_args *args)
/*
* Initialize the debugging data.
*
- * protection_map[0] (or even protection_map[8]) will help create
- * page table entries with PROT_NONE permission as required for
- * pxx_protnone_tests().
+ * vm_get_page_prot(VM_NONE) or vm_get_page_prot(VM_SHARED|VM_NONE)
+ * will help create page table entries with PROT_NONE permission as
+ * required for pxx_protnone_tests().
*/
memset(args, 0, sizeof(*args));
args->vaddr = get_random_vaddr();
args->page_prot = vm_get_page_prot(VMFLAGS);
- args->page_prot_none = protection_map[0];
+ args->page_prot_none = vm_get_page_prot(VM_NONE);
args->is_contiguous_page = false;
args->pud_pfn = ULONG_MAX;
args->pmd_pfn = ULONG_MAX;
@@ -1248,12 +1248,19 @@ static int __init debug_vm_pgtable(void)
return ret;
/*
- * Iterate over the protection_map[] to make sure that all
+ * Iterate over each possible vm_flags to make sure that all
* the basic page table transformation validations just hold
* true irrespective of the starting protection value for a
* given page table entry.
+ *
+ * Protection based vm_flags combinatins are always linear
+ * and increasing i.e starting from VM_NONE and going upto
+ * (VM_SHARED | READ | WRITE | EXEC).
*/
- for (idx = 0; idx < ARRAY_SIZE(protection_map); idx++) {
+#define VM_FLAGS_START (VM_NONE)
+#define VM_FLAGS_END (VM_SHARED | VM_EXEC | VM_WRITE | VM_READ)
+
+ for (idx = VM_FLAGS_START; idx <= VM_FLAGS_END; idx++) {
pte_basic_tests(&args, idx);
pmd_basic_tests(&args, idx);
pud_basic_tests(&args, idx);
--
2.25.1
^ permalink raw reply related [flat|nested] 49+ messages in thread
* [OpenRISC] [PATCH V3 02/30] mm/mmap: Clarify protection_map[] indices
2022-02-28 10:47 [OpenRISC] [PATCH V3 00/30] mm/mmap: Drop protection_map[] and platform's __SXXX/__PXXX requirements Anshuman Khandual
2022-02-28 10:47 ` [OpenRISC] [PATCH V3 01/30] mm/debug_vm_pgtable: Drop protection_map[] usage Anshuman Khandual
@ 2022-02-28 10:47 ` Anshuman Khandual
2022-02-28 10:47 ` [OpenRISC] [PATCH V3 03/30] mm/mmap: Add new config ARCH_HAS_VM_GET_PAGE_PROT Anshuman Khandual
` (27 subsequent siblings)
29 siblings, 0 replies; 49+ messages in thread
From: Anshuman Khandual @ 2022-02-28 10:47 UTC (permalink / raw)
To: openrisc
protection_map[] maps vm_flags access combinations into page protection
value as defined by the platform via __PXXX and __SXXX macros. The array
indices in protection_map[], represents vm_flags access combinations but
it's not very intuitive to derive. This makes it clear and explicit.
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: linux-mm at kvack.org
Cc: linux-kernel at vger.kernel.org
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Anshuman Khandual <anshuman.khandual@arm.com>
---
mm/mmap.c | 18 ++++++++++++++++--
1 file changed, 16 insertions(+), 2 deletions(-)
diff --git a/mm/mmap.c b/mm/mmap.c
index d445c1b9d606..34000a7d7efa 100644
--- a/mm/mmap.c
+++ b/mm/mmap.c
@@ -102,8 +102,22 @@ static void unmap_region(struct mm_struct *mm,
* x: (yes) yes
*/
pgprot_t protection_map[16] __ro_after_init = {
- __P000, __P001, __P010, __P011, __P100, __P101, __P110, __P111,
- __S000, __S001, __S010, __S011, __S100, __S101, __S110, __S111
+ [VM_NONE] = __P000,
+ [VM_READ] = __P001,
+ [VM_WRITE] = __P010,
+ [VM_WRITE | VM_READ] = __P011,
+ [VM_EXEC] = __P100,
+ [VM_EXEC | VM_READ] = __P101,
+ [VM_EXEC | VM_WRITE] = __P110,
+ [VM_EXEC | VM_WRITE | VM_READ] = __P111,
+ [VM_SHARED] = __S000,
+ [VM_SHARED | VM_READ] = __S001,
+ [VM_SHARED | VM_WRITE] = __S010,
+ [VM_SHARED | VM_WRITE | VM_READ] = __S011,
+ [VM_SHARED | VM_EXEC] = __S100,
+ [VM_SHARED | VM_EXEC | VM_READ] = __S101,
+ [VM_SHARED | VM_EXEC | VM_WRITE] = __S110,
+ [VM_SHARED | VM_EXEC | VM_WRITE | VM_READ] = __S111
};
#ifndef CONFIG_ARCH_HAS_FILTER_PGPROT
--
2.25.1
^ permalink raw reply related [flat|nested] 49+ messages in thread
* [OpenRISC] [PATCH V3 03/30] mm/mmap: Add new config ARCH_HAS_VM_GET_PAGE_PROT
2022-02-28 10:47 [OpenRISC] [PATCH V3 00/30] mm/mmap: Drop protection_map[] and platform's __SXXX/__PXXX requirements Anshuman Khandual
2022-02-28 10:47 ` [OpenRISC] [PATCH V3 01/30] mm/debug_vm_pgtable: Drop protection_map[] usage Anshuman Khandual
2022-02-28 10:47 ` [OpenRISC] [PATCH V3 02/30] mm/mmap: Clarify protection_map[] indices Anshuman Khandual
@ 2022-02-28 10:47 ` Anshuman Khandual
2022-02-28 10:47 ` [OpenRISC] [PATCH V3 04/30] powerpc/mm: Enable ARCH_HAS_VM_GET_PAGE_PROT Anshuman Khandual
` (26 subsequent siblings)
29 siblings, 0 replies; 49+ messages in thread
From: Anshuman Khandual @ 2022-02-28 10:47 UTC (permalink / raw)
To: openrisc
Add a new config ARCH_HAS_VM_GET_PAGE_PROT, which when subscribed enables a
given platform to define its own vm_get_page_prot(). This framework will
help remove protection_map[] dependency going forward.
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: linux-mm at kvack.org
Cc: linux-kernel at vger.kernel.org
Suggested-by: Christoph Hellwig <hch@infradead.org>
Signed-off-by: Anshuman Khandual <anshuman.khandual@arm.com>
---
mm/Kconfig | 3 +++
mm/mmap.c | 2 ++
2 files changed, 5 insertions(+)
diff --git a/mm/Kconfig b/mm/Kconfig
index 257ed9c86de3..fa436478a94c 100644
--- a/mm/Kconfig
+++ b/mm/Kconfig
@@ -747,6 +747,9 @@ config ARCH_HAS_CACHE_LINE_SIZE
config ARCH_HAS_FILTER_PGPROT
bool
+config ARCH_HAS_VM_GET_PAGE_PROT
+ bool
+
config ARCH_HAS_PTE_DEVMAP
bool
diff --git a/mm/mmap.c b/mm/mmap.c
index 34000a7d7efa..f2310f6e7466 100644
--- a/mm/mmap.c
+++ b/mm/mmap.c
@@ -81,6 +81,7 @@ static void unmap_region(struct mm_struct *mm,
struct vm_area_struct *vma, struct vm_area_struct *prev,
unsigned long start, unsigned long end);
+#ifndef CONFIG_ARCH_HAS_VM_GET_PAGE_PROT
/* description of effects of mapping type and prot in current implementation.
* this is due to the limited x86 page protection hardware. The expected
* behavior is in parens:
@@ -136,6 +137,7 @@ pgprot_t vm_get_page_prot(unsigned long vm_flags)
return arch_filter_pgprot(ret);
}
EXPORT_SYMBOL(vm_get_page_prot);
+#endif /* CONFIG_ARCH_HAS_VM_GET_PAGE_PROT */
static pgprot_t vm_pgprot_modify(pgprot_t oldprot, unsigned long vm_flags)
{
--
2.25.1
^ permalink raw reply related [flat|nested] 49+ messages in thread
* [OpenRISC] [PATCH V3 04/30] powerpc/mm: Enable ARCH_HAS_VM_GET_PAGE_PROT
2022-02-28 10:47 [OpenRISC] [PATCH V3 00/30] mm/mmap: Drop protection_map[] and platform's __SXXX/__PXXX requirements Anshuman Khandual
` (2 preceding siblings ...)
2022-02-28 10:47 ` [OpenRISC] [PATCH V3 03/30] mm/mmap: Add new config ARCH_HAS_VM_GET_PAGE_PROT Anshuman Khandual
@ 2022-02-28 10:47 ` Anshuman Khandual
2022-03-02 5:23 ` Michael Ellerman
2022-02-28 10:47 ` [OpenRISC] [PATCH V3 05/30] arm64/mm: " Anshuman Khandual
` (25 subsequent siblings)
29 siblings, 1 reply; 49+ messages in thread
From: Anshuman Khandual @ 2022-02-28 10:47 UTC (permalink / raw)
To: openrisc
This defines and exports a platform specific custom vm_get_page_prot() via
subscribing ARCH_HAS_VM_GET_PAGE_PROT. Subsequently all __SXXX and __PXXX
macros can be dropped which are no longer needed. While here, this also
localizes arch_vm_get_page_prot() as powerpc_vm_get_page_prot() and moves
it near vm_get_page_prot().
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Paul Mackerras <paulus@samba.org>
Cc: linuxppc-dev at lists.ozlabs.org
Cc: linux-kernel at vger.kernel.org
Signed-off-by: Anshuman Khandual <anshuman.khandual@arm.com>
---
arch/powerpc/Kconfig | 1 +
arch/powerpc/include/asm/mman.h | 12 ------
arch/powerpc/include/asm/pgtable.h | 19 ----------
arch/powerpc/mm/mmap.c | 59 ++++++++++++++++++++++++++++++
4 files changed, 60 insertions(+), 31 deletions(-)
diff --git a/arch/powerpc/Kconfig b/arch/powerpc/Kconfig
index b779603978e1..ddb4a3687c05 100644
--- a/arch/powerpc/Kconfig
+++ b/arch/powerpc/Kconfig
@@ -135,6 +135,7 @@ config PPC
select ARCH_HAS_TICK_BROADCAST if GENERIC_CLOCKEVENTS_BROADCAST
select ARCH_HAS_UACCESS_FLUSHCACHE
select ARCH_HAS_UBSAN_SANITIZE_ALL
+ select ARCH_HAS_VM_GET_PAGE_PROT
select ARCH_HAVE_NMI_SAFE_CMPXCHG
select ARCH_KEEP_MEMBLOCK
select ARCH_MIGHT_HAVE_PC_PARPORT
diff --git a/arch/powerpc/include/asm/mman.h b/arch/powerpc/include/asm/mman.h
index 7cb6d18f5cd6..1b024e64c8ec 100644
--- a/arch/powerpc/include/asm/mman.h
+++ b/arch/powerpc/include/asm/mman.h
@@ -24,18 +24,6 @@ static inline unsigned long arch_calc_vm_prot_bits(unsigned long prot,
}
#define arch_calc_vm_prot_bits(prot, pkey) arch_calc_vm_prot_bits(prot, pkey)
-static inline pgprot_t arch_vm_get_page_prot(unsigned long vm_flags)
-{
-#ifdef CONFIG_PPC_MEM_KEYS
- return (vm_flags & VM_SAO) ?
- __pgprot(_PAGE_SAO | vmflag_to_pte_pkey_bits(vm_flags)) :
- __pgprot(0 | vmflag_to_pte_pkey_bits(vm_flags));
-#else
- return (vm_flags & VM_SAO) ? __pgprot(_PAGE_SAO) : __pgprot(0);
-#endif
-}
-#define arch_vm_get_page_prot(vm_flags) arch_vm_get_page_prot(vm_flags)
-
static inline bool arch_validate_prot(unsigned long prot, unsigned long addr)
{
if (prot & ~(PROT_READ | PROT_WRITE | PROT_EXEC | PROT_SEM | PROT_SAO))
diff --git a/arch/powerpc/include/asm/pgtable.h b/arch/powerpc/include/asm/pgtable.h
index d564d0ecd4cd..3cbb6de20f9d 100644
--- a/arch/powerpc/include/asm/pgtable.h
+++ b/arch/powerpc/include/asm/pgtable.h
@@ -20,25 +20,6 @@ struct mm_struct;
#include <asm/nohash/pgtable.h>
#endif /* !CONFIG_PPC_BOOK3S */
-/* Note due to the way vm flags are laid out, the bits are XWR */
-#define __P000 PAGE_NONE
-#define __P001 PAGE_READONLY
-#define __P010 PAGE_COPY
-#define __P011 PAGE_COPY
-#define __P100 PAGE_READONLY_X
-#define __P101 PAGE_READONLY_X
-#define __P110 PAGE_COPY_X
-#define __P111 PAGE_COPY_X
-
-#define __S000 PAGE_NONE
-#define __S001 PAGE_READONLY
-#define __S010 PAGE_SHARED
-#define __S011 PAGE_SHARED
-#define __S100 PAGE_READONLY_X
-#define __S101 PAGE_READONLY_X
-#define __S110 PAGE_SHARED_X
-#define __S111 PAGE_SHARED_X
-
#ifndef __ASSEMBLY__
#ifndef MAX_PTRS_PER_PGD
diff --git a/arch/powerpc/mm/mmap.c b/arch/powerpc/mm/mmap.c
index c475cf810aa8..ee275937fe19 100644
--- a/arch/powerpc/mm/mmap.c
+++ b/arch/powerpc/mm/mmap.c
@@ -254,3 +254,62 @@ void arch_pick_mmap_layout(struct mm_struct *mm, struct rlimit *rlim_stack)
mm->get_unmapped_area = arch_get_unmapped_area_topdown;
}
}
+
+static inline pgprot_t __vm_get_page_prot(unsigned long vm_flags)
+{
+ switch (vm_flags & (VM_READ | VM_WRITE | VM_EXEC | VM_SHARED)) {
+ case VM_NONE:
+ return PAGE_NONE;
+ case VM_READ:
+ return PAGE_READONLY;
+ case VM_WRITE:
+ case VM_WRITE | VM_READ:
+ return PAGE_COPY;
+ case VM_EXEC:
+ case VM_EXEC | VM_READ:
+ return PAGE_READONLY_X;
+ case VM_EXEC | VM_WRITE:
+ case VM_EXEC | VM_WRITE | VM_READ:
+ return PAGE_COPY_X;
+ case VM_SHARED:
+ return PAGE_NONE;
+ case VM_SHARED | VM_READ:
+ return PAGE_READONLY;
+ case VM_SHARED | VM_WRITE:
+ case VM_SHARED | VM_WRITE | VM_READ:
+ return PAGE_SHARED;
+ case VM_SHARED | VM_EXEC:
+ case VM_SHARED | VM_EXEC | VM_READ:
+ return PAGE_READONLY_X;
+ case VM_SHARED | VM_EXEC | VM_WRITE:
+ case VM_SHARED | VM_EXEC | VM_WRITE | VM_READ:
+ return PAGE_SHARED_X;
+ default:
+ BUILD_BUG();
+ }
+}
+
+#ifdef CONFIG_PPC64
+static pgprot_t powerpc_vm_get_page_prot(unsigned long vm_flags)
+{
+#ifdef CONFIG_PPC_MEM_KEYS
+ return (vm_flags & VM_SAO) ?
+ __pgprot(_PAGE_SAO | vmflag_to_pte_pkey_bits(vm_flags)) :
+ __pgprot(0 | vmflag_to_pte_pkey_bits(vm_flags));
+#else
+ return (vm_flags & VM_SAO) ? __pgprot(_PAGE_SAO) : __pgprot(0);
+#endif
+}
+#else
+static pgprot_t powerpc_vm_get_page_prot(unsigned long vm_flags)
+{
+ return __pgprot(0);
+}
+#endif /* CONFIG_PPC64 */
+
+pgprot_t vm_get_page_prot(unsigned long vm_flags)
+{
+ return __pgprot(pgprot_val(__vm_get_page_prot(vm_flags)) |
+ pgprot_val(powerpc_vm_get_page_prot(vm_flags)));
+}
+EXPORT_SYMBOL(vm_get_page_prot);
--
2.25.1
^ permalink raw reply related [flat|nested] 49+ messages in thread
* [OpenRISC] [PATCH V3 05/30] arm64/mm: Enable ARCH_HAS_VM_GET_PAGE_PROT
2022-02-28 10:47 [OpenRISC] [PATCH V3 00/30] mm/mmap: Drop protection_map[] and platform's __SXXX/__PXXX requirements Anshuman Khandual
` (3 preceding siblings ...)
2022-02-28 10:47 ` [OpenRISC] [PATCH V3 04/30] powerpc/mm: Enable ARCH_HAS_VM_GET_PAGE_PROT Anshuman Khandual
@ 2022-02-28 10:47 ` Anshuman Khandual
2022-03-03 15:28 ` Catalin Marinas
2022-02-28 10:47 ` [OpenRISC] [PATCH V3 06/30] sparc/mm: " Anshuman Khandual
` (24 subsequent siblings)
29 siblings, 1 reply; 49+ messages in thread
From: Anshuman Khandual @ 2022-02-28 10:47 UTC (permalink / raw)
To: openrisc
This defines and exports a platform specific custom vm_get_page_prot() via
subscribing ARCH_HAS_VM_GET_PAGE_PROT. Subsequently all __SXXX and __PXXX
macros can be dropped which are no longer needed. This also localizes both
arch_filter_pgprot and arch_vm_get_page_prot() helpers, unsubscribing from
ARCH_HAS_FILTER_PGPROT as well. Moved both these localized functions near
vm_get_page_prot().
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Will Deacon <will@kernel.org>
Cc: linux-arm-kernel at lists.infradead.org
Cc: linux-kernel at vger.kernel.org
Signed-off-by: Anshuman Khandual <anshuman.khandual@arm.com>
---
arch/arm64/Kconfig | 2 +-
arch/arm64/include/asm/memory.h | 3 +-
arch/arm64/include/asm/mman.h | 24 ---------
arch/arm64/include/asm/pgtable-prot.h | 18 -------
arch/arm64/include/asm/pgtable.h | 11 ----
arch/arm64/mm/mmap.c | 78 +++++++++++++++++++++++++++
6 files changed, 80 insertions(+), 56 deletions(-)
diff --git a/arch/arm64/Kconfig b/arch/arm64/Kconfig
index 2e5d2eac6fc6..7153d5fff603 100644
--- a/arch/arm64/Kconfig
+++ b/arch/arm64/Kconfig
@@ -23,7 +23,6 @@ config ARM64
select ARCH_HAS_DMA_PREP_COHERENT
select ARCH_HAS_ACPI_TABLE_UPGRADE if ACPI
select ARCH_HAS_FAST_MULTIPLIER
- select ARCH_HAS_FILTER_PGPROT
select ARCH_HAS_FORTIFY_SOURCE
select ARCH_HAS_GCOV_PROFILE_ALL
select ARCH_HAS_GIGANTIC_PAGE
@@ -44,6 +43,7 @@ config ARM64
select ARCH_HAS_SYSCALL_WRAPPER
select ARCH_HAS_TEARDOWN_DMA_OPS if IOMMU_SUPPORT
select ARCH_HAS_TICK_BROADCAST if GENERIC_CLOCKEVENTS_BROADCAST
+ select ARCH_HAS_VM_GET_PAGE_PROT
select ARCH_HAS_ZONE_DMA_SET if EXPERT
select ARCH_HAVE_ELF_PROT
select ARCH_HAVE_NMI_SAFE_CMPXCHG
diff --git a/arch/arm64/include/asm/memory.h b/arch/arm64/include/asm/memory.h
index 0af70d9abede..64a613a0349b 100644
--- a/arch/arm64/include/asm/memory.h
+++ b/arch/arm64/include/asm/memory.h
@@ -126,8 +126,7 @@
* Memory types available.
*
* IMPORTANT: MT_NORMAL must be index 0 since vm_get_page_prot() may 'or' in
- * the MT_NORMAL_TAGGED memory type for PROT_MTE mappings. Note
- * that protection_map[] only contains MT_NORMAL attributes.
+ * the MT_NORMAL_TAGGED memory type for PROT_MTE mappings.
*/
#define MT_NORMAL 0
#define MT_NORMAL_TAGGED 1
diff --git a/arch/arm64/include/asm/mman.h b/arch/arm64/include/asm/mman.h
index e3e28f7daf62..5966ee4a6154 100644
--- a/arch/arm64/include/asm/mman.h
+++ b/arch/arm64/include/asm/mman.h
@@ -35,30 +35,6 @@ static inline unsigned long arch_calc_vm_flag_bits(unsigned long flags)
}
#define arch_calc_vm_flag_bits(flags) arch_calc_vm_flag_bits(flags)
-static inline pgprot_t arch_vm_get_page_prot(unsigned long vm_flags)
-{
- pteval_t prot = 0;
-
- if (vm_flags & VM_ARM64_BTI)
- prot |= PTE_GP;
-
- /*
- * There are two conditions required for returning a Normal Tagged
- * memory type: (1) the user requested it via PROT_MTE passed to
- * mmap() or mprotect() and (2) the corresponding vma supports MTE. We
- * register (1) as VM_MTE in the vma->vm_flags and (2) as
- * VM_MTE_ALLOWED. Note that the latter can only be set during the
- * mmap() call since mprotect() does not accept MAP_* flags.
- * Checking for VM_MTE only is sufficient since arch_validate_flags()
- * does not permit (VM_MTE & !VM_MTE_ALLOWED).
- */
- if (vm_flags & VM_MTE)
- prot |= PTE_ATTRINDX(MT_NORMAL_TAGGED);
-
- return __pgprot(prot);
-}
-#define arch_vm_get_page_prot(vm_flags) arch_vm_get_page_prot(vm_flags)
-
static inline bool arch_validate_prot(unsigned long prot,
unsigned long addr __always_unused)
{
diff --git a/arch/arm64/include/asm/pgtable-prot.h b/arch/arm64/include/asm/pgtable-prot.h
index 7032f04c8ac6..d8ee0aa7886d 100644
--- a/arch/arm64/include/asm/pgtable-prot.h
+++ b/arch/arm64/include/asm/pgtable-prot.h
@@ -88,24 +88,6 @@ extern bool arm64_use_ng_mappings;
#define PAGE_READONLY_EXEC __pgprot(_PAGE_DEFAULT | PTE_USER | PTE_RDONLY | PTE_NG | PTE_PXN)
#define PAGE_EXECONLY __pgprot(_PAGE_DEFAULT | PTE_RDONLY | PTE_NG | PTE_PXN)
-#define __P000 PAGE_NONE
-#define __P001 PAGE_READONLY
-#define __P010 PAGE_READONLY
-#define __P011 PAGE_READONLY
-#define __P100 PAGE_EXECONLY
-#define __P101 PAGE_READONLY_EXEC
-#define __P110 PAGE_READONLY_EXEC
-#define __P111 PAGE_READONLY_EXEC
-
-#define __S000 PAGE_NONE
-#define __S001 PAGE_READONLY
-#define __S010 PAGE_SHARED
-#define __S011 PAGE_SHARED
-#define __S100 PAGE_EXECONLY
-#define __S101 PAGE_READONLY_EXEC
-#define __S110 PAGE_SHARED_EXEC
-#define __S111 PAGE_SHARED_EXEC
-
#endif /* __ASSEMBLY__ */
#endif /* __ASM_PGTABLE_PROT_H */
diff --git a/arch/arm64/include/asm/pgtable.h b/arch/arm64/include/asm/pgtable.h
index c4ba047a82d2..94e147e5456c 100644
--- a/arch/arm64/include/asm/pgtable.h
+++ b/arch/arm64/include/asm/pgtable.h
@@ -1017,17 +1017,6 @@ static inline bool arch_wants_old_prefaulted_pte(void)
}
#define arch_wants_old_prefaulted_pte arch_wants_old_prefaulted_pte
-static inline pgprot_t arch_filter_pgprot(pgprot_t prot)
-{
- if (cpus_have_const_cap(ARM64_HAS_EPAN))
- return prot;
-
- if (pgprot_val(prot) != pgprot_val(PAGE_EXECONLY))
- return prot;
-
- return PAGE_READONLY_EXEC;
-}
-
static inline bool pud_sect_supported(void)
{
return PAGE_SIZE == SZ_4K;
diff --git a/arch/arm64/mm/mmap.c b/arch/arm64/mm/mmap.c
index a38f54cd638c..bd0233d376a2 100644
--- a/arch/arm64/mm/mmap.c
+++ b/arch/arm64/mm/mmap.c
@@ -10,6 +10,7 @@
#include <linux/types.h>
#include <asm/page.h>
+#include <asm/mman.h>
/*
* You really shouldn't be using read() or write() on /dev/mem. This might go
@@ -38,3 +39,80 @@ int valid_mmap_phys_addr_range(unsigned long pfn, size_t size)
{
return !(((pfn << PAGE_SHIFT) + size) & ~PHYS_MASK);
}
+
+static inline pgprot_t __vm_get_page_prot(unsigned long vm_flags)
+{
+ switch (vm_flags & (VM_READ | VM_WRITE | VM_EXEC | VM_SHARED)) {
+ case VM_NONE:
+ return PAGE_NONE;
+ case VM_READ:
+ case VM_WRITE:
+ case VM_WRITE | VM_READ:
+ return PAGE_READONLY;
+ case VM_EXEC:
+ return PAGE_EXECONLY;
+ case VM_EXEC | VM_READ:
+ case VM_EXEC | VM_WRITE:
+ case VM_EXEC | VM_WRITE | VM_READ:
+ return PAGE_READONLY_EXEC;
+ case VM_SHARED:
+ return PAGE_NONE;
+ case VM_SHARED | VM_READ:
+ return PAGE_READONLY;
+ case VM_SHARED | VM_WRITE:
+ case VM_SHARED | VM_WRITE | VM_READ:
+ return PAGE_SHARED;
+ case VM_SHARED | VM_EXEC:
+ return PAGE_EXECONLY;
+ case VM_SHARED | VM_EXEC | VM_READ:
+ return PAGE_READONLY_EXEC;
+ case VM_SHARED | VM_EXEC | VM_WRITE:
+ case VM_SHARED | VM_EXEC | VM_WRITE | VM_READ:
+ return PAGE_SHARED_EXEC;
+ default:
+ BUILD_BUG();
+ }
+}
+
+static pgprot_t arm64_arch_filter_pgprot(pgprot_t prot)
+{
+ if (cpus_have_const_cap(ARM64_HAS_EPAN))
+ return prot;
+
+ if (pgprot_val(prot) != pgprot_val(PAGE_EXECONLY))
+ return prot;
+
+ return PAGE_READONLY_EXEC;
+}
+
+static pgprot_t arm64_arch_vm_get_page_prot(unsigned long vm_flags)
+{
+ pteval_t prot = 0;
+
+ if (vm_flags & VM_ARM64_BTI)
+ prot |= PTE_GP;
+
+ /*
+ * There are two conditions required for returning a Normal Tagged
+ * memory type: (1) the user requested it via PROT_MTE passed to
+ * mmap() or mprotect() and (2) the corresponding vma supports MTE. We
+ * register (1) as VM_MTE in the vma->vm_flags and (2) as
+ * VM_MTE_ALLOWED. Note that the latter can only be set during the
+ * mmap() call since mprotect() does not accept MAP_* flags.
+ * Checking for VM_MTE only is sufficient since arch_validate_flags()
+ * does not permit (VM_MTE & !VM_MTE_ALLOWED).
+ */
+ if (vm_flags & VM_MTE)
+ prot |= PTE_ATTRINDX(MT_NORMAL_TAGGED);
+
+ return __pgprot(prot);
+}
+
+pgprot_t vm_get_page_prot(unsigned long vm_flags)
+{
+ pgprot_t ret = __pgprot(pgprot_val(__vm_get_page_prot(vm_flags)) |
+ pgprot_val(arm64_arch_vm_get_page_prot(vm_flags)));
+
+ return arm64_arch_filter_pgprot(ret);
+}
+EXPORT_SYMBOL(vm_get_page_prot);
--
2.25.1
^ permalink raw reply related [flat|nested] 49+ messages in thread
* [OpenRISC] [PATCH V3 06/30] sparc/mm: Enable ARCH_HAS_VM_GET_PAGE_PROT
2022-02-28 10:47 [OpenRISC] [PATCH V3 00/30] mm/mmap: Drop protection_map[] and platform's __SXXX/__PXXX requirements Anshuman Khandual
` (4 preceding siblings ...)
2022-02-28 10:47 ` [OpenRISC] [PATCH V3 05/30] arm64/mm: " Anshuman Khandual
@ 2022-02-28 10:47 ` Anshuman Khandual
2022-02-28 10:47 ` [OpenRISC] [PATCH V3 07/30] mips/mm: " Anshuman Khandual
` (23 subsequent siblings)
29 siblings, 0 replies; 49+ messages in thread
From: Anshuman Khandual @ 2022-02-28 10:47 UTC (permalink / raw)
To: openrisc
This defines and exports a platform specific custom vm_get_page_prot() via
subscribing ARCH_HAS_VM_GET_PAGE_PROT. Subsequently all __SXXX and __PXXX
macros can be dropped which are no longer needed. This also localizes the
helper arch_vm_get_page_prot() as sparc_vm_get_page_prot() and moves near
vm_get_page_prot().
Cc: "David S. Miller" <davem@davemloft.net>
Cc: Khalid Aziz <khalid.aziz@oracle.com>
Cc: sparclinux at vger.kernel.org
Cc: linux-kernel at vger.kernel.org
Reviewed-by: Khalid Aziz <khalid.aziz@oracle.com>
Acked-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Anshuman Khandual <anshuman.khandual@arm.com>
---
arch/sparc/Kconfig | 2 +
arch/sparc/include/asm/mman.h | 6 ---
arch/sparc/include/asm/pgtable_32.h | 19 --------
arch/sparc/include/asm/pgtable_64.h | 19 --------
arch/sparc/mm/init_32.c | 35 +++++++++++++++
arch/sparc/mm/init_64.c | 70 +++++++++++++++++++++--------
6 files changed, 88 insertions(+), 63 deletions(-)
diff --git a/arch/sparc/Kconfig b/arch/sparc/Kconfig
index 1cab1b284f1a..ff29156f2380 100644
--- a/arch/sparc/Kconfig
+++ b/arch/sparc/Kconfig
@@ -59,6 +59,7 @@ config SPARC32
select HAVE_UID16
select OLD_SIGACTION
select ZONE_DMA
+ select ARCH_HAS_VM_GET_PAGE_PROT
config SPARC64
def_bool 64BIT
@@ -84,6 +85,7 @@ config SPARC64
select PERF_USE_VMALLOC
select ARCH_HAVE_NMI_SAFE_CMPXCHG
select HAVE_C_RECORDMCOUNT
+ select ARCH_HAS_VM_GET_PAGE_PROT
select HAVE_ARCH_AUDITSYSCALL
select ARCH_SUPPORTS_ATOMIC_RMW
select ARCH_SUPPORTS_DEBUG_PAGEALLOC
diff --git a/arch/sparc/include/asm/mman.h b/arch/sparc/include/asm/mman.h
index 274217e7ed70..af9c10c83dc5 100644
--- a/arch/sparc/include/asm/mman.h
+++ b/arch/sparc/include/asm/mman.h
@@ -46,12 +46,6 @@ static inline unsigned long sparc_calc_vm_prot_bits(unsigned long prot)
}
}
-#define arch_vm_get_page_prot(vm_flags) sparc_vm_get_page_prot(vm_flags)
-static inline pgprot_t sparc_vm_get_page_prot(unsigned long vm_flags)
-{
- return (vm_flags & VM_SPARC_ADI) ? __pgprot(_PAGE_MCD_4V) : __pgprot(0);
-}
-
#define arch_validate_prot(prot, addr) sparc_validate_prot(prot, addr)
static inline int sparc_validate_prot(unsigned long prot, unsigned long addr)
{
diff --git a/arch/sparc/include/asm/pgtable_32.h b/arch/sparc/include/asm/pgtable_32.h
index ffccfe3b22ed..060a435f96d6 100644
--- a/arch/sparc/include/asm/pgtable_32.h
+++ b/arch/sparc/include/asm/pgtable_32.h
@@ -64,25 +64,6 @@ void paging_init(void);
extern unsigned long ptr_in_current_pgd;
-/* xwr */
-#define __P000 PAGE_NONE
-#define __P001 PAGE_READONLY
-#define __P010 PAGE_COPY
-#define __P011 PAGE_COPY
-#define __P100 PAGE_READONLY
-#define __P101 PAGE_READONLY
-#define __P110 PAGE_COPY
-#define __P111 PAGE_COPY
-
-#define __S000 PAGE_NONE
-#define __S001 PAGE_READONLY
-#define __S010 PAGE_SHARED
-#define __S011 PAGE_SHARED
-#define __S100 PAGE_READONLY
-#define __S101 PAGE_READONLY
-#define __S110 PAGE_SHARED
-#define __S111 PAGE_SHARED
-
/* First physical page can be anywhere, the following is needed so that
* va-->pa and vice versa conversions work properly without performance
* hit for all __pa()/__va() operations.
diff --git a/arch/sparc/include/asm/pgtable_64.h b/arch/sparc/include/asm/pgtable_64.h
index 4679e45c8348..a779418ceba9 100644
--- a/arch/sparc/include/asm/pgtable_64.h
+++ b/arch/sparc/include/asm/pgtable_64.h
@@ -187,25 +187,6 @@ bool kern_addr_valid(unsigned long addr);
#define _PAGE_SZHUGE_4U _PAGE_SZ4MB_4U
#define _PAGE_SZHUGE_4V _PAGE_SZ4MB_4V
-/* These are actually filled in at boot time by sun4{u,v}_pgprot_init() */
-#define __P000 __pgprot(0)
-#define __P001 __pgprot(0)
-#define __P010 __pgprot(0)
-#define __P011 __pgprot(0)
-#define __P100 __pgprot(0)
-#define __P101 __pgprot(0)
-#define __P110 __pgprot(0)
-#define __P111 __pgprot(0)
-
-#define __S000 __pgprot(0)
-#define __S001 __pgprot(0)
-#define __S010 __pgprot(0)
-#define __S011 __pgprot(0)
-#define __S100 __pgprot(0)
-#define __S101 __pgprot(0)
-#define __S110 __pgprot(0)
-#define __S111 __pgprot(0)
-
#ifndef __ASSEMBLY__
pte_t mk_pte_io(unsigned long, pgprot_t, int, unsigned long);
diff --git a/arch/sparc/mm/init_32.c b/arch/sparc/mm/init_32.c
index 1e9f577f084d..348cbfe08b60 100644
--- a/arch/sparc/mm/init_32.c
+++ b/arch/sparc/mm/init_32.c
@@ -302,3 +302,38 @@ void sparc_flush_page_to_ram(struct page *page)
__flush_page_to_ram(vaddr);
}
EXPORT_SYMBOL(sparc_flush_page_to_ram);
+
+pgprot_t vm_get_page_prot(unsigned long vm_flags)
+{
+ switch (vm_flags & (VM_READ | VM_WRITE | VM_EXEC | VM_SHARED)) {
+ case VM_NONE:
+ return PAGE_NONE;
+ case VM_READ:
+ return PAGE_READONLY;
+ case VM_WRITE:
+ case VM_WRITE | VM_READ:
+ return PAGE_COPY;
+ case VM_EXEC:
+ case VM_EXEC | VM_READ:
+ return PAGE_READONLY;
+ case VM_EXEC | VM_WRITE:
+ case VM_EXEC | VM_WRITE | VM_READ:
+ return PAGE_COPY;
+ case VM_SHARED:
+ return PAGE_NONE;
+ case VM_SHARED | VM_READ:
+ return PAGE_READONLY;
+ case VM_SHARED | VM_WRITE:
+ case VM_SHARED | VM_WRITE | VM_READ:
+ return PAGE_SHARED;
+ case VM_SHARED | VM_EXEC:
+ case VM_SHARED | VM_EXEC | VM_READ:
+ return PAGE_READONLY;
+ case VM_SHARED | VM_EXEC | VM_WRITE:
+ case VM_SHARED | VM_EXEC | VM_WRITE | VM_READ:
+ return PAGE_SHARED;
+ default:
+ BUILD_BUG();
+ }
+}
+EXPORT_SYMBOL(vm_get_page_prot);
diff --git a/arch/sparc/mm/init_64.c b/arch/sparc/mm/init_64.c
index 1b23639e2fcd..a390116d371b 100644
--- a/arch/sparc/mm/init_64.c
+++ b/arch/sparc/mm/init_64.c
@@ -50,6 +50,7 @@
#include <asm/cpudata.h>
#include <asm/setup.h>
#include <asm/irq.h>
+#include <asm/mman.h>
#include "init_64.h"
@@ -2641,29 +2642,13 @@ static void prot_init_common(unsigned long page_none,
{
PAGE_COPY = __pgprot(page_copy);
PAGE_SHARED = __pgprot(page_shared);
-
- protection_map[0x0] = __pgprot(page_none);
- protection_map[0x1] = __pgprot(page_readonly & ~page_exec_bit);
- protection_map[0x2] = __pgprot(page_copy & ~page_exec_bit);
- protection_map[0x3] = __pgprot(page_copy & ~page_exec_bit);
- protection_map[0x4] = __pgprot(page_readonly);
- protection_map[0x5] = __pgprot(page_readonly);
- protection_map[0x6] = __pgprot(page_copy);
- protection_map[0x7] = __pgprot(page_copy);
- protection_map[0x8] = __pgprot(page_none);
- protection_map[0x9] = __pgprot(page_readonly & ~page_exec_bit);
- protection_map[0xa] = __pgprot(page_shared & ~page_exec_bit);
- protection_map[0xb] = __pgprot(page_shared & ~page_exec_bit);
- protection_map[0xc] = __pgprot(page_readonly);
- protection_map[0xd] = __pgprot(page_readonly);
- protection_map[0xe] = __pgprot(page_shared);
- protection_map[0xf] = __pgprot(page_shared);
}
+static unsigned long page_none, page_shared, page_copy, page_readonly;
+static unsigned long page_exec_bit;
+
static void __init sun4u_pgprot_init(void)
{
- unsigned long page_none, page_shared, page_copy, page_readonly;
- unsigned long page_exec_bit;
int i;
PAGE_KERNEL = __pgprot (_PAGE_PRESENT_4U | _PAGE_VALID |
@@ -3183,3 +3168,50 @@ void copy_highpage(struct page *to, struct page *from)
}
}
EXPORT_SYMBOL(copy_highpage);
+
+static inline pgprot_t __vm_get_page_prot(unsigned long vm_flags)
+{
+ switch (vm_flags & (VM_READ | VM_WRITE | VM_EXEC | VM_SHARED)) {
+ case VM_NONE:
+ return __pgprot(page_none);
+ case VM_READ:
+ return __pgprot(page_readonly & ~page_exec_bit);
+ case VM_WRITE:
+ case VM_WRITE | VM_READ:
+ return __pgprot(page_copy & ~page_exec_bit);
+ case VM_EXEC:
+ case VM_EXEC | VM_READ:
+ return __pgprot(page_readonly);
+ case VM_EXEC | VM_WRITE:
+ case VM_EXEC | VM_WRITE | VM_READ:
+ return __pgprot(page_copy);
+ case VM_SHARED:
+ return __pgprot(page_none);
+ case VM_SHARED | VM_READ:
+ return __pgprot(page_readonly & ~page_exec_bit);
+ case VM_SHARED | VM_WRITE:
+ case VM_SHARED | VM_WRITE | VM_READ:
+ return __pgprot(page_shared & ~page_exec_bit);
+ case VM_SHARED | VM_EXEC:
+ case VM_SHARED | VM_EXEC | VM_READ:
+ return __pgprot(page_readonly);
+ case VM_SHARED | VM_EXEC | VM_WRITE:
+ case VM_SHARED | VM_EXEC | VM_WRITE | VM_READ:
+ return __pgprot(page_shared);
+ default:
+ BUILD_BUG();
+ }
+}
+
+static pgprot_t sparc_vm_get_page_prot(unsigned long vm_flags)
+{
+ return (vm_flags & VM_SPARC_ADI) ? __pgprot(_PAGE_MCD_4V) : __pgprot(0);
+}
+
+pgprot_t vm_get_page_prot(unsigned long vm_flags)
+{
+ return __pgprot(pgprot_val(__vm_get_page_prot(vm_flags)) |
+ pgprot_val(sparc_vm_get_page_prot(vm_flags)));
+
+}
+EXPORT_SYMBOL(vm_get_page_prot);
--
2.25.1
^ permalink raw reply related [flat|nested] 49+ messages in thread
* [OpenRISC] [PATCH V3 07/30] mips/mm: Enable ARCH_HAS_VM_GET_PAGE_PROT
2022-02-28 10:47 [OpenRISC] [PATCH V3 00/30] mm/mmap: Drop protection_map[] and platform's __SXXX/__PXXX requirements Anshuman Khandual
` (5 preceding siblings ...)
2022-02-28 10:47 ` [OpenRISC] [PATCH V3 06/30] sparc/mm: " Anshuman Khandual
@ 2022-02-28 10:47 ` Anshuman Khandual
2022-02-28 10:47 ` [OpenRISC] [PATCH V3 08/30] m68k/mm: " Anshuman Khandual
` (22 subsequent siblings)
29 siblings, 0 replies; 49+ messages in thread
From: Anshuman Khandual @ 2022-02-28 10:47 UTC (permalink / raw)
To: openrisc
This defines and exports a platform specific custom vm_get_page_prot() via
subscribing ARCH_HAS_VM_GET_PAGE_PROT. Subsequently all __SXXX and __PXXX
macros can be dropped which are no longer needed.
Cc: Thomas Bogendoerfer <tsbogend@alpha.franken.de>
Cc: linux-mips at vger.kernel.org
Cc: linux-kernel at vger.kernel.org
Acked-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de>
Signed-off-by: Anshuman Khandual <anshuman.khandual@arm.com>
---
arch/mips/Kconfig | 1 +
arch/mips/include/asm/pgtable.h | 22 ------------
arch/mips/mm/cache.c | 60 +++++++++++++++++++--------------
3 files changed, 36 insertions(+), 47 deletions(-)
diff --git a/arch/mips/Kconfig b/arch/mips/Kconfig
index 058446f01487..fcbfc52a1567 100644
--- a/arch/mips/Kconfig
+++ b/arch/mips/Kconfig
@@ -13,6 +13,7 @@ config MIPS
select ARCH_HAS_STRNLEN_USER
select ARCH_HAS_TICK_BROADCAST if GENERIC_CLOCKEVENTS_BROADCAST
select ARCH_HAS_UBSAN_SANITIZE_ALL
+ select ARCH_HAS_VM_GET_PAGE_PROT
select ARCH_HAS_GCOV_PROFILE_ALL
select ARCH_KEEP_MEMBLOCK
select ARCH_SUPPORTS_UPROBES
diff --git a/arch/mips/include/asm/pgtable.h b/arch/mips/include/asm/pgtable.h
index 7b8037f25d9e..bf193ad4f195 100644
--- a/arch/mips/include/asm/pgtable.h
+++ b/arch/mips/include/asm/pgtable.h
@@ -41,28 +41,6 @@ struct vm_area_struct;
* by reasonable means..
*/
-/*
- * Dummy values to fill the table in mmap.c
- * The real values will be generated at runtime
- */
-#define __P000 __pgprot(0)
-#define __P001 __pgprot(0)
-#define __P010 __pgprot(0)
-#define __P011 __pgprot(0)
-#define __P100 __pgprot(0)
-#define __P101 __pgprot(0)
-#define __P110 __pgprot(0)
-#define __P111 __pgprot(0)
-
-#define __S000 __pgprot(0)
-#define __S001 __pgprot(0)
-#define __S010 __pgprot(0)
-#define __S011 __pgprot(0)
-#define __S100 __pgprot(0)
-#define __S101 __pgprot(0)
-#define __S110 __pgprot(0)
-#define __S111 __pgprot(0)
-
extern unsigned long _page_cachable_default;
extern void __update_cache(unsigned long address, pte_t pte);
diff --git a/arch/mips/mm/cache.c b/arch/mips/mm/cache.c
index 830ab91e574f..9f33ce4fb105 100644
--- a/arch/mips/mm/cache.c
+++ b/arch/mips/mm/cache.c
@@ -159,30 +159,6 @@ EXPORT_SYMBOL(_page_cachable_default);
#define PM(p) __pgprot(_page_cachable_default | (p))
-static inline void setup_protection_map(void)
-{
- protection_map[0] = PM(_PAGE_PRESENT | _PAGE_NO_EXEC | _PAGE_NO_READ);
- protection_map[1] = PM(_PAGE_PRESENT | _PAGE_NO_EXEC);
- protection_map[2] = PM(_PAGE_PRESENT | _PAGE_NO_EXEC | _PAGE_NO_READ);
- protection_map[3] = PM(_PAGE_PRESENT | _PAGE_NO_EXEC);
- protection_map[4] = PM(_PAGE_PRESENT);
- protection_map[5] = PM(_PAGE_PRESENT);
- protection_map[6] = PM(_PAGE_PRESENT);
- protection_map[7] = PM(_PAGE_PRESENT);
-
- protection_map[8] = PM(_PAGE_PRESENT | _PAGE_NO_EXEC | _PAGE_NO_READ);
- protection_map[9] = PM(_PAGE_PRESENT | _PAGE_NO_EXEC);
- protection_map[10] = PM(_PAGE_PRESENT | _PAGE_NO_EXEC | _PAGE_WRITE |
- _PAGE_NO_READ);
- protection_map[11] = PM(_PAGE_PRESENT | _PAGE_NO_EXEC | _PAGE_WRITE);
- protection_map[12] = PM(_PAGE_PRESENT);
- protection_map[13] = PM(_PAGE_PRESENT);
- protection_map[14] = PM(_PAGE_PRESENT | _PAGE_WRITE);
- protection_map[15] = PM(_PAGE_PRESENT | _PAGE_WRITE);
-}
-
-#undef PM
-
void cpu_cache_init(void)
{
if (cpu_has_3k_cache) {
@@ -206,6 +182,40 @@ void cpu_cache_init(void)
octeon_cache_init();
}
+}
- setup_protection_map();
+pgprot_t vm_get_page_prot(unsigned long vm_flags)
+{
+ switch (vm_flags & (VM_READ | VM_WRITE | VM_EXEC | VM_SHARED)) {
+ case VM_NONE:
+ return PM(_PAGE_PRESENT | _PAGE_NO_EXEC | _PAGE_NO_READ);
+ case VM_READ:
+ return PM(_PAGE_PRESENT | _PAGE_NO_EXEC);
+ case VM_WRITE:
+ return PM(_PAGE_PRESENT | _PAGE_NO_EXEC | _PAGE_NO_READ);
+ case VM_WRITE | VM_READ:
+ return PM(_PAGE_PRESENT | _PAGE_NO_EXEC);
+ case VM_EXEC:
+ case VM_EXEC | VM_READ:
+ case VM_EXEC | VM_WRITE:
+ case VM_EXEC | VM_WRITE | VM_READ:
+ return PM(_PAGE_PRESENT);
+ case VM_SHARED:
+ return PM(_PAGE_PRESENT | _PAGE_NO_EXEC | _PAGE_NO_READ);
+ case VM_SHARED | VM_READ:
+ return PM(_PAGE_PRESENT | _PAGE_NO_EXEC);
+ case VM_SHARED | VM_WRITE:
+ return PM(_PAGE_PRESENT | _PAGE_NO_EXEC | _PAGE_WRITE | _PAGE_NO_READ);
+ case VM_SHARED | VM_WRITE | VM_READ:
+ return PM(_PAGE_PRESENT | _PAGE_NO_EXEC | _PAGE_WRITE);
+ case VM_SHARED | VM_EXEC:
+ case VM_SHARED | VM_EXEC | VM_READ:
+ return PM(_PAGE_PRESENT);
+ case VM_SHARED | VM_EXEC | VM_WRITE:
+ case VM_SHARED | VM_EXEC | VM_WRITE | VM_READ:
+ return PM(_PAGE_PRESENT | _PAGE_WRITE);
+ default:
+ BUILD_BUG();
+ }
}
+EXPORT_SYMBOL(vm_get_page_prot);
--
2.25.1
^ permalink raw reply related [flat|nested] 49+ messages in thread
* [OpenRISC] [PATCH V3 08/30] m68k/mm: Enable ARCH_HAS_VM_GET_PAGE_PROT
2022-02-28 10:47 [OpenRISC] [PATCH V3 00/30] mm/mmap: Drop protection_map[] and platform's __SXXX/__PXXX requirements Anshuman Khandual
` (6 preceding siblings ...)
2022-02-28 10:47 ` [OpenRISC] [PATCH V3 07/30] mips/mm: " Anshuman Khandual
@ 2022-02-28 10:47 ` Anshuman Khandual
2022-02-28 10:47 ` [OpenRISC] [PATCH V3 09/30] arm/mm: " Anshuman Khandual
` (21 subsequent siblings)
29 siblings, 0 replies; 49+ messages in thread
From: Anshuman Khandual @ 2022-02-28 10:47 UTC (permalink / raw)
To: openrisc
This defines and exports a platform specific custom vm_get_page_prot() via
subscribing ARCH_HAS_VM_GET_PAGE_PROT. Subsequently all __SXXX and __PXXX
macros can be dropped which are no longer needed.
Cc: Thomas Bogendoerfer <tsbogend@alpha.franken.de>
Cc: linux-m68k at lists.linux-m68k.org
Cc: linux-kernel at vger.kernel.org
Signed-off-by: Anshuman Khandual <anshuman.khandual@arm.com>
---
arch/m68k/Kconfig | 1 +
arch/m68k/include/asm/mcf_pgtable.h | 59 ------------------------
arch/m68k/include/asm/motorola_pgtable.h | 29 ------------
arch/m68k/include/asm/sun3_pgtable.h | 22 ---------
arch/m68k/mm/mcfmmu.c | 59 ++++++++++++++++++++++++
arch/m68k/mm/motorola.c | 43 +++++++++++++++--
arch/m68k/mm/sun3mmu.c | 39 ++++++++++++++++
7 files changed, 139 insertions(+), 113 deletions(-)
diff --git a/arch/m68k/Kconfig b/arch/m68k/Kconfig
index 936e1803c7c7..114e65164692 100644
--- a/arch/m68k/Kconfig
+++ b/arch/m68k/Kconfig
@@ -11,6 +11,7 @@ config M68K
select ARCH_NO_PREEMPT if !COLDFIRE
select ARCH_USE_MEMTEST if MMU_MOTOROLA
select ARCH_WANT_IPC_PARSE_VERSION
+ select ARCH_HAS_VM_GET_PAGE_PROT
select BINFMT_FLAT_ARGVP_ENVP_ON_STACK
select DMA_DIRECT_REMAP if HAS_DMA && MMU && !COLDFIRE
select GENERIC_ATOMIC64
diff --git a/arch/m68k/include/asm/mcf_pgtable.h b/arch/m68k/include/asm/mcf_pgtable.h
index 6f2b87d7a50d..dc5c8ab6aa57 100644
--- a/arch/m68k/include/asm/mcf_pgtable.h
+++ b/arch/m68k/include/asm/mcf_pgtable.h
@@ -86,65 +86,6 @@
| CF_PAGE_READABLE \
| CF_PAGE_DIRTY)
-/*
- * Page protections for initialising protection_map. See mm/mmap.c
- * for use. In general, the bit positions are xwr, and P-items are
- * private, the S-items are shared.
- */
-#define __P000 PAGE_NONE
-#define __P001 __pgprot(CF_PAGE_VALID \
- | CF_PAGE_ACCESSED \
- | CF_PAGE_READABLE)
-#define __P010 __pgprot(CF_PAGE_VALID \
- | CF_PAGE_ACCESSED \
- | CF_PAGE_WRITABLE)
-#define __P011 __pgprot(CF_PAGE_VALID \
- | CF_PAGE_ACCESSED \
- | CF_PAGE_READABLE \
- | CF_PAGE_WRITABLE)
-#define __P100 __pgprot(CF_PAGE_VALID \
- | CF_PAGE_ACCESSED \
- | CF_PAGE_EXEC)
-#define __P101 __pgprot(CF_PAGE_VALID \
- | CF_PAGE_ACCESSED \
- | CF_PAGE_READABLE \
- | CF_PAGE_EXEC)
-#define __P110 __pgprot(CF_PAGE_VALID \
- | CF_PAGE_ACCESSED \
- | CF_PAGE_WRITABLE \
- | CF_PAGE_EXEC)
-#define __P111 __pgprot(CF_PAGE_VALID \
- | CF_PAGE_ACCESSED \
- | CF_PAGE_READABLE \
- | CF_PAGE_WRITABLE \
- | CF_PAGE_EXEC)
-
-#define __S000 PAGE_NONE
-#define __S001 __pgprot(CF_PAGE_VALID \
- | CF_PAGE_ACCESSED \
- | CF_PAGE_READABLE)
-#define __S010 PAGE_SHARED
-#define __S011 __pgprot(CF_PAGE_VALID \
- | CF_PAGE_ACCESSED \
- | CF_PAGE_SHARED \
- | CF_PAGE_READABLE)
-#define __S100 __pgprot(CF_PAGE_VALID \
- | CF_PAGE_ACCESSED \
- | CF_PAGE_EXEC)
-#define __S101 __pgprot(CF_PAGE_VALID \
- | CF_PAGE_ACCESSED \
- | CF_PAGE_READABLE \
- | CF_PAGE_EXEC)
-#define __S110 __pgprot(CF_PAGE_VALID \
- | CF_PAGE_ACCESSED \
- | CF_PAGE_SHARED \
- | CF_PAGE_EXEC)
-#define __S111 __pgprot(CF_PAGE_VALID \
- | CF_PAGE_ACCESSED \
- | CF_PAGE_SHARED \
- | CF_PAGE_READABLE \
- | CF_PAGE_EXEC)
-
#define PTE_MASK PAGE_MASK
#define CF_PAGE_CHG_MASK (PTE_MASK | CF_PAGE_ACCESSED | CF_PAGE_DIRTY)
diff --git a/arch/m68k/include/asm/motorola_pgtable.h b/arch/m68k/include/asm/motorola_pgtable.h
index 022c3abc280d..dcbb856f567e 100644
--- a/arch/m68k/include/asm/motorola_pgtable.h
+++ b/arch/m68k/include/asm/motorola_pgtable.h
@@ -76,35 +76,6 @@ extern unsigned long mm_cachebits;
#define PAGE_READONLY __pgprot(_PAGE_PRESENT | _PAGE_RONLY | _PAGE_ACCESSED | mm_cachebits)
#define PAGE_KERNEL __pgprot(_PAGE_PRESENT | _PAGE_DIRTY | _PAGE_ACCESSED | mm_cachebits)
-/* Alternate definitions that are compile time constants, for
- initializing protection_map. The cachebits are fixed later. */
-#define PAGE_NONE_C __pgprot(_PAGE_PROTNONE | _PAGE_ACCESSED)
-#define PAGE_SHARED_C __pgprot(_PAGE_PRESENT | _PAGE_ACCESSED)
-#define PAGE_COPY_C __pgprot(_PAGE_PRESENT | _PAGE_RONLY | _PAGE_ACCESSED)
-#define PAGE_READONLY_C __pgprot(_PAGE_PRESENT | _PAGE_RONLY | _PAGE_ACCESSED)
-
-/*
- * The m68k can't do page protection for execute, and considers that the same are read.
- * Also, write permissions imply read permissions. This is the closest we can get..
- */
-#define __P000 PAGE_NONE_C
-#define __P001 PAGE_READONLY_C
-#define __P010 PAGE_COPY_C
-#define __P011 PAGE_COPY_C
-#define __P100 PAGE_READONLY_C
-#define __P101 PAGE_READONLY_C
-#define __P110 PAGE_COPY_C
-#define __P111 PAGE_COPY_C
-
-#define __S000 PAGE_NONE_C
-#define __S001 PAGE_READONLY_C
-#define __S010 PAGE_SHARED_C
-#define __S011 PAGE_SHARED_C
-#define __S100 PAGE_READONLY_C
-#define __S101 PAGE_READONLY_C
-#define __S110 PAGE_SHARED_C
-#define __S111 PAGE_SHARED_C
-
#define pmd_pgtable(pmd) ((pgtable_t)pmd_page_vaddr(pmd))
/*
diff --git a/arch/m68k/include/asm/sun3_pgtable.h b/arch/m68k/include/asm/sun3_pgtable.h
index 5b24283a0a42..086fabdd8d4c 100644
--- a/arch/m68k/include/asm/sun3_pgtable.h
+++ b/arch/m68k/include/asm/sun3_pgtable.h
@@ -66,28 +66,6 @@
| SUN3_PAGE_SYSTEM \
| SUN3_PAGE_NOCACHE)
-/*
- * Page protections for initialising protection_map. The sun3 has only two
- * protection settings, valid (implying read and execute) and writeable. These
- * are as close as we can get...
- */
-#define __P000 PAGE_NONE
-#define __P001 PAGE_READONLY
-#define __P010 PAGE_COPY
-#define __P011 PAGE_COPY
-#define __P100 PAGE_READONLY
-#define __P101 PAGE_READONLY
-#define __P110 PAGE_COPY
-#define __P111 PAGE_COPY
-
-#define __S000 PAGE_NONE
-#define __S001 PAGE_READONLY
-#define __S010 PAGE_SHARED
-#define __S011 PAGE_SHARED
-#define __S100 PAGE_READONLY
-#define __S101 PAGE_READONLY
-#define __S110 PAGE_SHARED
-#define __S111 PAGE_SHARED
/* Use these fake page-protections on PMDs. */
#define SUN3_PMD_VALID (0x00000001)
diff --git a/arch/m68k/mm/mcfmmu.c b/arch/m68k/mm/mcfmmu.c
index 6f1f25125294..795ead15d1d8 100644
--- a/arch/m68k/mm/mcfmmu.c
+++ b/arch/m68k/mm/mcfmmu.c
@@ -234,3 +234,62 @@ void steal_context(void)
destroy_context(mm);
}
+/*
+ * In general, the bit positions are xwr, and P-items are private,
+ * the S-items are shared.
+ */
+pgprot_t vm_get_page_prot(unsigned long vm_flags)
+{
+ switch (vm_flags & (VM_READ | VM_WRITE | VM_EXEC | VM_SHARED)) {
+ case VM_NONE:
+ return PAGE_NONE;
+ case VM_READ:
+ return __pgprot(CF_PAGE_VALID | CF_PAGE_ACCESSED |
+ CF_PAGE_READABLE);
+ case VM_WRITE:
+ return __pgprot(CF_PAGE_VALID | CF_PAGE_ACCESSED |
+ CF_PAGE_WRITABLE);
+ case VM_WRITE | VM_READ:
+ return __pgprot(CF_PAGE_VALID | CF_PAGE_ACCESSED |
+ CF_PAGE_READABLE | CF_PAGE_WRITABLE);
+ case VM_EXEC:
+ return __pgprot(CF_PAGE_VALID | CF_PAGE_ACCESSED |
+ CF_PAGE_EXEC);
+ case VM_EXEC | VM_READ:
+ return __pgprot(CF_PAGE_VALID | CF_PAGE_ACCESSED |
+ CF_PAGE_READABLE | CF_PAGE_EXEC);
+ case VM_EXEC | VM_WRITE:
+ return __pgprot(CF_PAGE_VALID | CF_PAGE_ACCESSED |
+ CF_PAGE_WRITABLE | CF_PAGE_EXEC);
+ case VM_EXEC | VM_WRITE | VM_READ:
+ return __pgprot(CF_PAGE_VALID | CF_PAGE_ACCESSED |
+ CF_PAGE_READABLE | CF_PAGE_WRITABLE |
+ CF_PAGE_EXEC);
+ case VM_SHARED:
+ return PAGE_NONE;
+ case VM_SHARED | VM_READ:
+ return __pgprot(CF_PAGE_VALID | CF_PAGE_ACCESSED |
+ CF_PAGE_READABLE);
+ case VM_SHARED | VM_WRITE:
+ return PAGE_SHARED;
+ case VM_SHARED | VM_WRITE | VM_READ:
+ return __pgprot(CF_PAGE_VALID | CF_PAGE_ACCESSED |
+ CF_PAGE_READABLE | CF_PAGE_SHARED);
+ case VM_SHARED | VM_EXEC:
+ return __pgprot(CF_PAGE_VALID | CF_PAGE_ACCESSED |
+ CF_PAGE_EXEC);
+ case VM_SHARED | VM_EXEC | VM_READ:
+ return __pgprot(CF_PAGE_VALID | CF_PAGE_ACCESSED |
+ CF_PAGE_READABLE | CF_PAGE_EXEC);
+ case VM_SHARED | VM_EXEC | VM_WRITE:
+ return __pgprot(CF_PAGE_VALID | CF_PAGE_ACCESSED |
+ CF_PAGE_SHARED | CF_PAGE_EXEC);
+ case VM_SHARED | VM_EXEC | VM_WRITE | VM_READ:
+ return __pgprot(CF_PAGE_VALID | CF_PAGE_ACCESSED |
+ CF_PAGE_READABLE | CF_PAGE_SHARED |
+ CF_PAGE_EXEC);
+ default:
+ BUILD_BUG();
+ }
+}
+EXPORT_SYMBOL(vm_get_page_prot);
diff --git a/arch/m68k/mm/motorola.c b/arch/m68k/mm/motorola.c
index ecbe948f4c1a..c6d43319fe1e 100644
--- a/arch/m68k/mm/motorola.c
+++ b/arch/m68k/mm/motorola.c
@@ -400,12 +400,9 @@ void __init paging_init(void)
/* Fix the cache mode in the page descriptors for the 680[46]0. */
if (CPU_IS_040_OR_060) {
- int i;
#ifndef mm_cachebits
mm_cachebits = _PAGE_CACHE040;
#endif
- for (i = 0; i < 16; i++)
- pgprot_val(protection_map[i]) |= _PAGE_CACHE040;
}
min_addr = m68k_memory[0].addr;
@@ -483,3 +480,43 @@ void __init paging_init(void)
max_zone_pfn[ZONE_DMA] = memblock_end_of_DRAM();
free_area_init(max_zone_pfn);
}
+
+/*
+ * The m68k can't do page protection for execute, and considers that
+ * the same are read. Also, write permissions imply read permissions.
+ * This is the closest we can get..
+ */
+pgprot_t vm_get_page_prot(unsigned long vm_flags)
+{
+ switch (vm_flags & (VM_READ | VM_WRITE | VM_EXEC | VM_SHARED)) {
+ case VM_NONE:
+ return PAGE_NONE;
+ case VM_READ:
+ return PAGE_READONLY;
+ case VM_WRITE:
+ case VM_WRITE | VM_READ:
+ return PAGE_COPY;
+ case VM_EXEC:
+ case VM_EXEC | VM_READ:
+ return PAGE_READONLY;
+ case VM_EXEC | VM_WRITE:
+ case VM_EXEC | VM_WRITE | VM_READ:
+ return PAGE_COPY;
+ case VM_SHARED:
+ return PAGE_NONE;
+ case VM_SHARED | VM_READ:
+ return PAGE_READONLY;
+ case VM_SHARED | VM_WRITE:
+ case VM_SHARED | VM_WRITE | VM_READ:
+ return PAGE_SHARED;
+ case VM_SHARED | VM_EXEC:
+ case VM_SHARED | VM_EXEC | VM_READ:
+ return PAGE_READONLY;
+ case VM_SHARED | VM_EXEC | VM_WRITE:
+ case VM_SHARED | VM_EXEC | VM_WRITE | VM_READ:
+ return PAGE_SHARED;
+ default:
+ BUILD_BUG();
+ }
+}
+EXPORT_SYMBOL(vm_get_page_prot);
diff --git a/arch/m68k/mm/sun3mmu.c b/arch/m68k/mm/sun3mmu.c
index dad494224497..2072630099f3 100644
--- a/arch/m68k/mm/sun3mmu.c
+++ b/arch/m68k/mm/sun3mmu.c
@@ -95,3 +95,42 @@ void __init paging_init(void)
}
+
+/*
+ * The sun3 has only two protection settings, valid (implying read and execute)
+ * and writeable. These are as close as we can get...
+ */
+pgprot_t vm_get_page_prot(unsigned long vm_flags)
+{
+ switch (vm_flags & (VM_READ | VM_WRITE | VM_EXEC | VM_SHARED)) {
+ case VM_NONE:
+ return PAGE_NONE;
+ case VM_READ:
+ return PAGE_READONLY;
+ case VM_WRITE:
+ case VM_WRITE | VM_READ:
+ return PAGE_COPY;
+ case VM_EXEC:
+ case VM_EXEC | VM_READ:
+ return PAGE_READONLY;
+ case VM_EXEC | VM_WRITE:
+ case VM_EXEC | VM_WRITE | VM_READ:
+ return PAGE_COPY;
+ case VM_SHARED:
+ return PAGE_NONE;
+ case VM_SHARED | VM_READ:
+ return PAGE_READONLY;
+ case VM_SHARED | VM_WRITE:
+ case VM_SHARED | VM_WRITE | VM_READ:
+ return PAGE_SHARED;
+ case VM_SHARED | VM_EXEC:
+ case VM_SHARED | VM_EXEC | VM_READ:
+ return PAGE_READONLY;
+ case VM_SHARED | VM_EXEC | VM_WRITE:
+ case VM_SHARED | VM_EXEC | VM_WRITE | VM_READ:
+ return PAGE_SHARED;
+ default:
+ BUILD_BUG();
+ }
+}
+EXPORT_SYMBOL(vm_get_page_prot);
--
2.25.1
^ permalink raw reply related [flat|nested] 49+ messages in thread
* [OpenRISC] [PATCH V3 09/30] arm/mm: Enable ARCH_HAS_VM_GET_PAGE_PROT
2022-02-28 10:47 [OpenRISC] [PATCH V3 00/30] mm/mmap: Drop protection_map[] and platform's __SXXX/__PXXX requirements Anshuman Khandual
` (7 preceding siblings ...)
2022-02-28 10:47 ` [OpenRISC] [PATCH V3 08/30] m68k/mm: " Anshuman Khandual
@ 2022-02-28 10:47 ` Anshuman Khandual
2022-02-28 10:57 ` Russell King
2022-02-28 10:47 ` [OpenRISC] [PATCH V3 10/30] x86/mm: " Anshuman Khandual
` (20 subsequent siblings)
29 siblings, 1 reply; 49+ messages in thread
From: Anshuman Khandual @ 2022-02-28 10:47 UTC (permalink / raw)
To: openrisc
This defines and exports a platform specific custom vm_get_page_prot() via
subscribing ARCH_HAS_VM_GET_PAGE_PROT. Subsequently all __SXXX and __PXXX
macros can be dropped which are no longer needed.
Cc: Russell King <linux@armlinux.org.uk>
Cc: Arnd Bergmann <arnd@arndb.de>
Cc: linux-arm-kernel at lists.infradead.org
Cc: linux-kernel at vger.kernel.org
Signed-off-by: Anshuman Khandual <anshuman.khandual@arm.com>
---
arch/arm/Kconfig | 1 +
arch/arm/include/asm/pgtable.h | 20 +-------------
arch/arm/lib/uaccess_with_memcpy.c | 2 +-
arch/arm/mm/mmu.c | 44 ++++++++++++++++++++++++++----
4 files changed, 41 insertions(+), 26 deletions(-)
diff --git a/arch/arm/Kconfig b/arch/arm/Kconfig
index 4c97cb40eebb..87b2e89ef3d6 100644
--- a/arch/arm/Kconfig
+++ b/arch/arm/Kconfig
@@ -23,6 +23,7 @@ config ARM
select ARCH_HAS_SYNC_DMA_FOR_CPU if SWIOTLB || !MMU
select ARCH_HAS_TEARDOWN_DMA_OPS if MMU
select ARCH_HAS_TICK_BROADCAST if GENERIC_CLOCKEVENTS_BROADCAST
+ select ARCH_HAS_VM_GET_PAGE_PROT
select ARCH_HAVE_CUSTOM_GPIO_H
select ARCH_HAVE_NMI_SAFE_CMPXCHG if CPU_V7 || CPU_V7M || CPU_V6K
select ARCH_HAS_GCOV_PROFILE_ALL
diff --git a/arch/arm/include/asm/pgtable.h b/arch/arm/include/asm/pgtable.h
index cd1f84bb40ae..64711716cd84 100644
--- a/arch/arm/include/asm/pgtable.h
+++ b/arch/arm/include/asm/pgtable.h
@@ -70,7 +70,7 @@ extern void __pgd_error(const char *file, int line, pgd_t);
#endif
/*
- * The pgprot_* and protection_map entries will be fixed up in runtime
+ * The pgprot_* entries will be fixed up in runtime in vm_get_page_prot()
* to include the cachable and bufferable bits based on memory policy,
* as well as any architecture dependent bits like global/ASID and SMP
* shared mapping bits.
@@ -137,24 +137,6 @@ extern pgprot_t phys_mem_access_prot(struct file *file, unsigned long pfn,
* 2) If we could do execute protection, then read is implied
* 3) write implies read permissions
*/
-#define __P000 __PAGE_NONE
-#define __P001 __PAGE_READONLY
-#define __P010 __PAGE_COPY
-#define __P011 __PAGE_COPY
-#define __P100 __PAGE_READONLY_EXEC
-#define __P101 __PAGE_READONLY_EXEC
-#define __P110 __PAGE_COPY_EXEC
-#define __P111 __PAGE_COPY_EXEC
-
-#define __S000 __PAGE_NONE
-#define __S001 __PAGE_READONLY
-#define __S010 __PAGE_SHARED
-#define __S011 __PAGE_SHARED
-#define __S100 __PAGE_READONLY_EXEC
-#define __S101 __PAGE_READONLY_EXEC
-#define __S110 __PAGE_SHARED_EXEC
-#define __S111 __PAGE_SHARED_EXEC
-
#ifndef __ASSEMBLY__
/*
* ZERO_PAGE is a global shared page that is always zero: used
diff --git a/arch/arm/lib/uaccess_with_memcpy.c b/arch/arm/lib/uaccess_with_memcpy.c
index 106f83a5ea6d..12d8d9794a28 100644
--- a/arch/arm/lib/uaccess_with_memcpy.c
+++ b/arch/arm/lib/uaccess_with_memcpy.c
@@ -247,7 +247,7 @@ static int __init test_size_treshold(void)
if (!dst_page)
goto no_dst;
kernel_ptr = page_address(src_page);
- user_ptr = vmap(&dst_page, 1, VM_IOREMAP, __pgprot(__P010));
+ user_ptr = vmap(&dst_page, 1, VM_IOREMAP, __pgprot(__PAGE_COPY));
if (!user_ptr)
goto no_vmap;
diff --git a/arch/arm/mm/mmu.c b/arch/arm/mm/mmu.c
index 274e4f73fd33..9cdf45da57de 100644
--- a/arch/arm/mm/mmu.c
+++ b/arch/arm/mm/mmu.c
@@ -403,6 +403,8 @@ void __set_fixmap(enum fixed_addresses idx, phys_addr_t phys, pgprot_t prot)
local_flush_tlb_kernel_range(vaddr, vaddr + PAGE_SIZE);
}
+static pteval_t user_pgprot;
+
/*
* Adjust the PMD section entries according to the CPU in use.
*/
@@ -410,7 +412,7 @@ static void __init build_mem_type_table(void)
{
struct cachepolicy *cp;
unsigned int cr = get_cr();
- pteval_t user_pgprot, kern_pgprot, vecs_pgprot;
+ pteval_t kern_pgprot, vecs_pgprot;
int cpu_arch = cpu_architecture();
int i;
@@ -627,11 +629,6 @@ static void __init build_mem_type_table(void)
user_pgprot |= PTE_EXT_PXN;
#endif
- for (i = 0; i < 16; i++) {
- pteval_t v = pgprot_val(protection_map[i]);
- protection_map[i] = __pgprot(v | user_pgprot);
- }
-
mem_types[MT_LOW_VECTORS].prot_pte |= vecs_pgprot;
mem_types[MT_HIGH_VECTORS].prot_pte |= vecs_pgprot;
@@ -670,6 +667,41 @@ static void __init build_mem_type_table(void)
}
}
+pgprot_t vm_get_page_prot(unsigned long vm_flags)
+{
+ switch (vm_flags & (VM_READ | VM_WRITE | VM_EXEC | VM_SHARED)) {
+ case VM_NONE:
+ return __pgprot(pgprot_val(__PAGE_NONE) | user_pgprot);
+ case VM_READ:
+ return __pgprot(pgprot_val(__PAGE_READONLY) | user_pgprot);
+ case VM_WRITE:
+ case VM_WRITE | VM_READ:
+ return __pgprot(pgprot_val(__PAGE_COPY) | user_pgprot);
+ case VM_EXEC:
+ case VM_EXEC | VM_READ:
+ return __pgprot(pgprot_val(__PAGE_READONLY_EXEC) | user_pgprot);
+ case VM_EXEC | VM_WRITE:
+ case VM_EXEC | VM_WRITE | VM_READ:
+ return __pgprot(pgprot_val(__PAGE_COPY_EXEC) | user_pgprot);
+ case VM_SHARED:
+ return __pgprot(pgprot_val(__PAGE_NONE) | user_pgprot);
+ case VM_SHARED | VM_READ:
+ return __pgprot(pgprot_val(__PAGE_READONLY) | user_pgprot);
+ case VM_SHARED | VM_WRITE:
+ case VM_SHARED | VM_WRITE | VM_READ:
+ return __pgprot(pgprot_val(__PAGE_SHARED) | user_pgprot);
+ case VM_SHARED | VM_EXEC:
+ case VM_SHARED | VM_EXEC | VM_READ:
+ return __pgprot(pgprot_val(__PAGE_READONLY_EXEC) | user_pgprot);
+ case VM_SHARED | VM_EXEC | VM_WRITE:
+ case VM_SHARED | VM_EXEC | VM_WRITE | VM_READ:
+ return __pgprot(pgprot_val(__PAGE_SHARED_EXEC) | user_pgprot);
+ default:
+ BUILD_BUG();
+ }
+}
+EXPORT_SYMBOL(vm_get_page_prot);
+
#ifdef CONFIG_ARM_DMA_MEM_BUFFERABLE
pgprot_t phys_mem_access_prot(struct file *file, unsigned long pfn,
unsigned long size, pgprot_t vma_prot)
--
2.25.1
^ permalink raw reply related [flat|nested] 49+ messages in thread
* [OpenRISC] [PATCH V3 10/30] x86/mm: Enable ARCH_HAS_VM_GET_PAGE_PROT
2022-02-28 10:47 [OpenRISC] [PATCH V3 00/30] mm/mmap: Drop protection_map[] and platform's __SXXX/__PXXX requirements Anshuman Khandual
` (8 preceding siblings ...)
2022-02-28 10:47 ` [OpenRISC] [PATCH V3 09/30] arm/mm: " Anshuman Khandual
@ 2022-02-28 10:47 ` Anshuman Khandual
2022-02-28 10:47 ` [OpenRISC] [PATCH V3 11/30] mm/mmap: Drop protection_map[] Anshuman Khandual
` (19 subsequent siblings)
29 siblings, 0 replies; 49+ messages in thread
From: Anshuman Khandual @ 2022-02-28 10:47 UTC (permalink / raw)
To: openrisc
From: Christoph Hellwig <hch@infradead.org>
This defines and exports a platform specific custom vm_get_page_prot() via
subscribing ARCH_HAS_VM_GET_PAGE_PROT. Subsequently all __SXXX and __PXXX
macros can be dropped which are no longer needed. This also unsubscribes
from ARCH_HAS_FILTER_PGPROT, after dropping off arch_filter_pgprot() and
arch_vm_get_page_prot().
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: linux-kernel at vger.kernel.org
Signed-off-by: Christoph Hellwig <hch@infradead.org>
Signed-off-by: Anshuman Khandual <anshuman.khandual@arm.com>
---
arch/x86/Kconfig | 2 +-
arch/x86/include/asm/pgtable.h | 5 --
arch/x86/include/asm/pgtable_types.h | 19 --------
arch/x86/include/uapi/asm/mman.h | 14 ------
arch/x86/mm/Makefile | 2 +-
arch/x86/mm/mem_encrypt_amd.c | 6 ---
arch/x86/mm/pgprot.c | 71 ++++++++++++++++++++++++++++
7 files changed, 73 insertions(+), 46 deletions(-)
create mode 100644 arch/x86/mm/pgprot.c
diff --git a/arch/x86/Kconfig b/arch/x86/Kconfig
index b1ce75d0ab0c..b2ea06c87708 100644
--- a/arch/x86/Kconfig
+++ b/arch/x86/Kconfig
@@ -75,7 +75,6 @@ config X86
select ARCH_HAS_EARLY_DEBUG if KGDB
select ARCH_HAS_ELF_RANDOMIZE
select ARCH_HAS_FAST_MULTIPLIER
- select ARCH_HAS_FILTER_PGPROT
select ARCH_HAS_FORTIFY_SOURCE
select ARCH_HAS_GCOV_PROFILE_ALL
select ARCH_HAS_KCOV if X86_64
@@ -94,6 +93,7 @@ config X86
select ARCH_HAS_SYNC_CORE_BEFORE_USERMODE
select ARCH_HAS_SYSCALL_WRAPPER
select ARCH_HAS_UBSAN_SANITIZE_ALL
+ select ARCH_HAS_VM_GET_PAGE_PROT
select ARCH_HAS_DEBUG_WX
select ARCH_HAS_ZONE_DMA_SET if EXPERT
select ARCH_HAVE_NMI_SAFE_CMPXCHG
diff --git a/arch/x86/include/asm/pgtable.h b/arch/x86/include/asm/pgtable.h
index 8a9432fb3802..985e1b823691 100644
--- a/arch/x86/include/asm/pgtable.h
+++ b/arch/x86/include/asm/pgtable.h
@@ -648,11 +648,6 @@ static inline pgprot_t pgprot_modify(pgprot_t oldprot, pgprot_t newprot)
#define canon_pgprot(p) __pgprot(massage_pgprot(p))
-static inline pgprot_t arch_filter_pgprot(pgprot_t prot)
-{
- return canon_pgprot(prot);
-}
-
static inline int is_new_memtype_allowed(u64 paddr, unsigned long size,
enum page_cache_mode pcm,
enum page_cache_mode new_pcm)
diff --git a/arch/x86/include/asm/pgtable_types.h b/arch/x86/include/asm/pgtable_types.h
index 40497a9020c6..1a9dd933088e 100644
--- a/arch/x86/include/asm/pgtable_types.h
+++ b/arch/x86/include/asm/pgtable_types.h
@@ -228,25 +228,6 @@ enum page_cache_mode {
#endif /* __ASSEMBLY__ */
-/* xwr */
-#define __P000 PAGE_NONE
-#define __P001 PAGE_READONLY
-#define __P010 PAGE_COPY
-#define __P011 PAGE_COPY
-#define __P100 PAGE_READONLY_EXEC
-#define __P101 PAGE_READONLY_EXEC
-#define __P110 PAGE_COPY_EXEC
-#define __P111 PAGE_COPY_EXEC
-
-#define __S000 PAGE_NONE
-#define __S001 PAGE_READONLY
-#define __S010 PAGE_SHARED
-#define __S011 PAGE_SHARED
-#define __S100 PAGE_READONLY_EXEC
-#define __S101 PAGE_READONLY_EXEC
-#define __S110 PAGE_SHARED_EXEC
-#define __S111 PAGE_SHARED_EXEC
-
/*
* early identity mapping pte attrib macros.
*/
diff --git a/arch/x86/include/uapi/asm/mman.h b/arch/x86/include/uapi/asm/mman.h
index d4a8d0424bfb..775dbd3aff73 100644
--- a/arch/x86/include/uapi/asm/mman.h
+++ b/arch/x86/include/uapi/asm/mman.h
@@ -5,20 +5,6 @@
#define MAP_32BIT 0x40 /* only give out 32bit addresses */
#ifdef CONFIG_X86_INTEL_MEMORY_PROTECTION_KEYS
-/*
- * Take the 4 protection key bits out of the vma->vm_flags
- * value and turn them in to the bits that we can put in
- * to a pte.
- *
- * Only override these if Protection Keys are available
- * (which is only on 64-bit).
- */
-#define arch_vm_get_page_prot(vm_flags) __pgprot( \
- ((vm_flags) & VM_PKEY_BIT0 ? _PAGE_PKEY_BIT0 : 0) | \
- ((vm_flags) & VM_PKEY_BIT1 ? _PAGE_PKEY_BIT1 : 0) | \
- ((vm_flags) & VM_PKEY_BIT2 ? _PAGE_PKEY_BIT2 : 0) | \
- ((vm_flags) & VM_PKEY_BIT3 ? _PAGE_PKEY_BIT3 : 0))
-
#define arch_calc_vm_prot_bits(prot, key) ( \
((key) & 0x1 ? VM_PKEY_BIT0 : 0) | \
((key) & 0x2 ? VM_PKEY_BIT1 : 0) | \
diff --git a/arch/x86/mm/Makefile b/arch/x86/mm/Makefile
index fe3d3061fc11..fb6b41a48ae5 100644
--- a/arch/x86/mm/Makefile
+++ b/arch/x86/mm/Makefile
@@ -20,7 +20,7 @@ CFLAGS_REMOVE_mem_encrypt_identity.o = -pg
endif
obj-y := init.o init_$(BITS).o fault.o ioremap.o extable.o mmap.o \
- pgtable.o physaddr.o setup_nx.o tlb.o cpu_entry_area.o maccess.o
+ pgtable.o physaddr.o setup_nx.o tlb.o cpu_entry_area.o maccess.o pgprot.o
obj-y += pat/
diff --git a/arch/x86/mm/mem_encrypt_amd.c b/arch/x86/mm/mem_encrypt_amd.c
index 2b2d018ea345..95f94eaac9a0 100644
--- a/arch/x86/mm/mem_encrypt_amd.c
+++ b/arch/x86/mm/mem_encrypt_amd.c
@@ -179,8 +179,6 @@ void __init sme_map_bootdata(char *real_mode_data)
void __init sme_early_init(void)
{
- unsigned int i;
-
if (!sme_me_mask)
return;
@@ -188,10 +186,6 @@ void __init sme_early_init(void)
__supported_pte_mask = __sme_set(__supported_pte_mask);
- /* Update the protection map with memory encryption mask */
- for (i = 0; i < ARRAY_SIZE(protection_map); i++)
- protection_map[i] = pgprot_encrypted(protection_map[i]);
-
if (cc_platform_has(CC_ATTR_GUEST_MEM_ENCRYPT))
swiotlb_force = SWIOTLB_FORCE;
}
diff --git a/arch/x86/mm/pgprot.c b/arch/x86/mm/pgprot.c
new file mode 100644
index 000000000000..5f2f029ce4fa
--- /dev/null
+++ b/arch/x86/mm/pgprot.c
@@ -0,0 +1,71 @@
+// SPDX-License-Identifier: GPL-2.0
+
+#include <linux/export.h>
+#include <linux/mm.h>
+#include <asm/pgtable.h>
+
+static inline pgprot_t __vm_get_page_prot(unsigned long vm_flags)
+{
+ switch (vm_flags & (VM_READ | VM_WRITE | VM_EXEC | VM_SHARED)) {
+ case VM_NONE:
+ return PAGE_NONE;
+ case VM_READ:
+ return PAGE_READONLY;
+ case VM_WRITE:
+ return PAGE_COPY;
+ case VM_WRITE | VM_READ:
+ return PAGE_COPY;
+ case VM_EXEC:
+ case VM_EXEC | VM_READ:
+ return PAGE_READONLY_EXEC;
+ case VM_EXEC | VM_WRITE:
+ case VM_EXEC | VM_WRITE | VM_READ:
+ return PAGE_COPY_EXEC;
+ case VM_SHARED:
+ return PAGE_NONE;
+ case VM_SHARED | VM_READ:
+ return PAGE_READONLY;
+ case VM_SHARED | VM_WRITE:
+ case VM_SHARED | VM_WRITE | VM_READ:
+ return PAGE_SHARED;
+ case VM_SHARED | VM_EXEC:
+ case VM_SHARED | VM_EXEC | VM_READ:
+ return PAGE_READONLY_EXEC;
+ case VM_SHARED | VM_EXEC | VM_WRITE:
+ case VM_SHARED | VM_EXEC | VM_WRITE | VM_READ:
+ return PAGE_SHARED_EXEC;
+ default:
+ BUILD_BUG();
+ return PAGE_NONE;
+ }
+}
+
+
+pgprot_t vm_get_page_prot(unsigned long vm_flags)
+{
+ unsigned long val = pgprot_val(__vm_get_page_prot(vm_flags));
+
+#ifdef CONFIG_X86_INTEL_MEMORY_PROTECTION_KEYS
+ /*
+ * Take the 4 protection key bits out of the vma->vm_flags value and
+ * turn them in to the bits that we can put in to a pte.
+ *
+ * Only override these if Protection Keys are available (which is only
+ * on 64-bit).
+ */
+ if (vm_flags & VM_PKEY_BIT0)
+ val |= _PAGE_PKEY_BIT0;
+ if (vm_flags & VM_PKEY_BIT1)
+ val |= _PAGE_PKEY_BIT1;
+ if (vm_flags & VM_PKEY_BIT2)
+ val |= _PAGE_PKEY_BIT2;
+ if (vm_flags & VM_PKEY_BIT3)
+ val |= _PAGE_PKEY_BIT3;
+#endif
+
+ val = __sme_set(val);
+ if (val & _PAGE_PRESENT)
+ val &= __supported_pte_mask;
+ return __pgprot(val);
+}
+EXPORT_SYMBOL(vm_get_page_prot);
--
2.25.1
^ permalink raw reply related [flat|nested] 49+ messages in thread
* [OpenRISC] [PATCH V3 11/30] mm/mmap: Drop protection_map[]
2022-02-28 10:47 [OpenRISC] [PATCH V3 00/30] mm/mmap: Drop protection_map[] and platform's __SXXX/__PXXX requirements Anshuman Khandual
` (9 preceding siblings ...)
2022-02-28 10:47 ` [OpenRISC] [PATCH V3 10/30] x86/mm: " Anshuman Khandual
@ 2022-02-28 10:47 ` Anshuman Khandual
2022-02-28 10:47 ` [OpenRISC] [PATCH V3 12/30] mm/mmap: Drop arch_filter_pgprot() Anshuman Khandual
` (18 subsequent siblings)
29 siblings, 0 replies; 49+ messages in thread
From: Anshuman Khandual @ 2022-02-28 10:47 UTC (permalink / raw)
To: openrisc
There are no other users for protection_map[]. Hence just drop this array
construct and instead define __vm_get_page_prot() which will provide page
protection map based on vm_flags combination switch.
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: linux-mm at kvack.org
Cc: linux-kernel at vger.kernel.org
Signed-off-by: Anshuman Khandual <anshuman.khandual@arm.com>
---
drivers/gpu/drm/drm_vm.c | 4 +--
include/linux/mm.h | 6 ----
mm/mmap.c | 63 ++++++++++++++++++++++++++--------------
3 files changed, 44 insertions(+), 29 deletions(-)
diff --git a/drivers/gpu/drm/drm_vm.c b/drivers/gpu/drm/drm_vm.c
index e957d4851dc0..14862df7532f 100644
--- a/drivers/gpu/drm/drm_vm.c
+++ b/drivers/gpu/drm/drm_vm.c
@@ -482,7 +482,7 @@ static int drm_mmap_dma(struct file *filp, struct vm_area_struct *vma)
#else
/* Ye gads this is ugly. With more thought
we could move this up higher and use
- `protection_map' instead. */
+ `vm_get_page_prot()' instead. */
vma->vm_page_prot =
__pgprot(pte_val
(pte_wrprotect
@@ -566,7 +566,7 @@ static int drm_mmap_locked(struct file *filp, struct vm_area_struct *vma)
#else
/* Ye gads this is ugly. With more thought
we could move this up higher and use
- `protection_map' instead. */
+ `vm_get_page_prot()' instead. */
vma->vm_page_prot =
__pgprot(pte_val
(pte_wrprotect
diff --git a/include/linux/mm.h b/include/linux/mm.h
index 213cc569b192..ff74bd2d7850 100644
--- a/include/linux/mm.h
+++ b/include/linux/mm.h
@@ -418,12 +418,6 @@ extern unsigned int kobjsize(const void *objp);
#endif
#define VM_FLAGS_CLEAR (ARCH_VM_PKEY_FLAGS | VM_ARCH_CLEAR)
-/*
- * mapping from the currently active vm_flags protection bits (the
- * low four bits) to a page protection mask..
- */
-extern pgprot_t protection_map[16];
-
/*
* The default fault flags that should be used by most of the
* arch-specific page fault handlers.
diff --git a/mm/mmap.c b/mm/mmap.c
index f2310f6e7466..78eeac277a80 100644
--- a/mm/mmap.c
+++ b/mm/mmap.c
@@ -102,24 +102,6 @@ static void unmap_region(struct mm_struct *mm,
* w: (no) no
* x: (yes) yes
*/
-pgprot_t protection_map[16] __ro_after_init = {
- [VM_NONE] = __P000,
- [VM_READ] = __P001,
- [VM_WRITE] = __P010,
- [VM_WRITE | VM_READ] = __P011,
- [VM_EXEC] = __P100,
- [VM_EXEC | VM_READ] = __P101,
- [VM_EXEC | VM_WRITE] = __P110,
- [VM_EXEC | VM_WRITE | VM_READ] = __P111,
- [VM_SHARED] = __S000,
- [VM_SHARED | VM_READ] = __S001,
- [VM_SHARED | VM_WRITE] = __S010,
- [VM_SHARED | VM_WRITE | VM_READ] = __S011,
- [VM_SHARED | VM_EXEC] = __S100,
- [VM_SHARED | VM_EXEC | VM_READ] = __S101,
- [VM_SHARED | VM_EXEC | VM_WRITE] = __S110,
- [VM_SHARED | VM_EXEC | VM_WRITE | VM_READ] = __S111
-};
#ifndef CONFIG_ARCH_HAS_FILTER_PGPROT
static inline pgprot_t arch_filter_pgprot(pgprot_t prot)
@@ -128,10 +110,49 @@ static inline pgprot_t arch_filter_pgprot(pgprot_t prot)
}
#endif
+static inline pgprot_t __vm_get_page_prot(unsigned long vm_flags)
+{
+ switch (vm_flags & (VM_READ | VM_WRITE | VM_EXEC | VM_SHARED)) {
+ case VM_NONE:
+ return __P000;
+ case VM_READ:
+ return __P001;
+ case VM_WRITE:
+ return __P010;
+ case VM_READ | VM_WRITE:
+ return __P011;
+ case VM_EXEC:
+ return __P100;
+ case VM_EXEC | VM_READ:
+ return __P101;
+ case VM_EXEC | VM_WRITE:
+ return __P110;
+ case VM_EXEC | VM_READ | VM_WRITE:
+ return __P111;
+ case VM_SHARED:
+ return __S000;
+ case VM_SHARED | VM_READ:
+ return __S001;
+ case VM_SHARED | VM_WRITE:
+ return __S010;
+ case VM_SHARED | VM_READ | VM_WRITE:
+ return __S011;
+ case VM_SHARED | VM_EXEC:
+ return __S100;
+ case VM_SHARED | VM_EXEC | VM_READ:
+ return __S101;
+ case VM_SHARED | VM_EXEC | VM_WRITE:
+ return __S110;
+ case VM_SHARED | VM_EXEC | VM_READ | VM_WRITE:
+ return __S111;
+ default:
+ BUILD_BUG();
+ }
+}
+
pgprot_t vm_get_page_prot(unsigned long vm_flags)
{
- pgprot_t ret = __pgprot(pgprot_val(protection_map[vm_flags &
- (VM_READ|VM_WRITE|VM_EXEC|VM_SHARED)]) |
+ pgprot_t ret = __pgprot(pgprot_val(__vm_get_page_prot(vm_flags)) |
pgprot_val(arch_vm_get_page_prot(vm_flags)));
return arch_filter_pgprot(ret);
@@ -1684,7 +1705,7 @@ SYSCALL_DEFINE1(old_mmap, struct mmap_arg_struct __user *, arg)
/*
* Some shared mappings will want the pages marked read-only
* to track write events. If so, we'll downgrade vm_page_prot
- * to the private version (using protection_map[] without the
+ * to the private version (using vm_get_page_prot() without the
* VM_SHARED bit).
*/
int vma_wants_writenotify(struct vm_area_struct *vma, pgprot_t vm_page_prot)
--
2.25.1
^ permalink raw reply related [flat|nested] 49+ messages in thread
* [OpenRISC] [PATCH V3 12/30] mm/mmap: Drop arch_filter_pgprot()
2022-02-28 10:47 [OpenRISC] [PATCH V3 00/30] mm/mmap: Drop protection_map[] and platform's __SXXX/__PXXX requirements Anshuman Khandual
` (10 preceding siblings ...)
2022-02-28 10:47 ` [OpenRISC] [PATCH V3 11/30] mm/mmap: Drop protection_map[] Anshuman Khandual
@ 2022-02-28 10:47 ` Anshuman Khandual
2022-02-28 10:47 ` [OpenRISC] [PATCH V3 13/30] mm/mmap: Drop arch_vm_get_page_pgprot() Anshuman Khandual
` (17 subsequent siblings)
29 siblings, 0 replies; 49+ messages in thread
From: Anshuman Khandual @ 2022-02-28 10:47 UTC (permalink / raw)
To: openrisc
There are no platforms left which subscribe ARCH_HAS_FILTER_PGPROT. Hence
just drop arch_filter_pgprot() and also the config ARCH_HAS_FILTER_PGPROT.
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: linux-mm at kvack.org
Cc: linux-kernel at vger.kernel.org
Signed-off-by: Anshuman Khandual <anshuman.khandual@arm.com>
---
mm/Kconfig | 3 ---
mm/mmap.c | 10 +---------
2 files changed, 1 insertion(+), 12 deletions(-)
diff --git a/mm/Kconfig b/mm/Kconfig
index fa436478a94c..212fb6e1ddaa 100644
--- a/mm/Kconfig
+++ b/mm/Kconfig
@@ -744,9 +744,6 @@ config IDLE_PAGE_TRACKING
config ARCH_HAS_CACHE_LINE_SIZE
bool
-config ARCH_HAS_FILTER_PGPROT
- bool
-
config ARCH_HAS_VM_GET_PAGE_PROT
bool
diff --git a/mm/mmap.c b/mm/mmap.c
index 78eeac277a80..c8fd8f06bf7c 100644
--- a/mm/mmap.c
+++ b/mm/mmap.c
@@ -102,14 +102,6 @@ static void unmap_region(struct mm_struct *mm,
* w: (no) no
* x: (yes) yes
*/
-
-#ifndef CONFIG_ARCH_HAS_FILTER_PGPROT
-static inline pgprot_t arch_filter_pgprot(pgprot_t prot)
-{
- return prot;
-}
-#endif
-
static inline pgprot_t __vm_get_page_prot(unsigned long vm_flags)
{
switch (vm_flags & (VM_READ | VM_WRITE | VM_EXEC | VM_SHARED)) {
@@ -155,7 +147,7 @@ pgprot_t vm_get_page_prot(unsigned long vm_flags)
pgprot_t ret = __pgprot(pgprot_val(__vm_get_page_prot(vm_flags)) |
pgprot_val(arch_vm_get_page_prot(vm_flags)));
- return arch_filter_pgprot(ret);
+ return ret;
}
EXPORT_SYMBOL(vm_get_page_prot);
#endif /* CONFIG_ARCH_HAS_VM_GET_PAGE_PROT */
--
2.25.1
^ permalink raw reply related [flat|nested] 49+ messages in thread
* [OpenRISC] [PATCH V3 13/30] mm/mmap: Drop arch_vm_get_page_pgprot()
2022-02-28 10:47 [OpenRISC] [PATCH V3 00/30] mm/mmap: Drop protection_map[] and platform's __SXXX/__PXXX requirements Anshuman Khandual
` (11 preceding siblings ...)
2022-02-28 10:47 ` [OpenRISC] [PATCH V3 12/30] mm/mmap: Drop arch_filter_pgprot() Anshuman Khandual
@ 2022-02-28 10:47 ` Anshuman Khandual
2022-02-28 10:47 ` [OpenRISC] [PATCH V3 14/30] s390/mm: Enable ARCH_HAS_VM_GET_PAGE_PROT Anshuman Khandual
` (16 subsequent siblings)
29 siblings, 0 replies; 49+ messages in thread
From: Anshuman Khandual @ 2022-02-28 10:47 UTC (permalink / raw)
To: openrisc
There are no platforms left which use arch_vm_get_page_prot(). Just drop
arch_vm_get_page_prot() construct and simplify remaining code.
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: linux-mm at kvack.org
Cc: linux-kernel at vger.kernel.org
Signed-off-by: Anshuman Khandual <anshuman.khandual@arm.com>
---
include/linux/mman.h | 4 ----
mm/mmap.c | 10 +---------
2 files changed, 1 insertion(+), 13 deletions(-)
diff --git a/include/linux/mman.h b/include/linux/mman.h
index b66e91b8176c..58b3abd457a3 100644
--- a/include/linux/mman.h
+++ b/include/linux/mman.h
@@ -93,10 +93,6 @@ static inline void vm_unacct_memory(long pages)
#define arch_calc_vm_flag_bits(flags) 0
#endif
-#ifndef arch_vm_get_page_prot
-#define arch_vm_get_page_prot(vm_flags) __pgprot(0)
-#endif
-
#ifndef arch_validate_prot
/*
* This is called from mprotect(). PROT_GROWSDOWN and PROT_GROWSUP have
diff --git a/mm/mmap.c b/mm/mmap.c
index c8fd8f06bf7c..c13dd9c37866 100644
--- a/mm/mmap.c
+++ b/mm/mmap.c
@@ -102,7 +102,7 @@ static void unmap_region(struct mm_struct *mm,
* w: (no) no
* x: (yes) yes
*/
-static inline pgprot_t __vm_get_page_prot(unsigned long vm_flags)
+pgprot_t vm_get_page_prot(unsigned long vm_flags)
{
switch (vm_flags & (VM_READ | VM_WRITE | VM_EXEC | VM_SHARED)) {
case VM_NONE:
@@ -141,14 +141,6 @@ static inline pgprot_t __vm_get_page_prot(unsigned long vm_flags)
BUILD_BUG();
}
}
-
-pgprot_t vm_get_page_prot(unsigned long vm_flags)
-{
- pgprot_t ret = __pgprot(pgprot_val(__vm_get_page_prot(vm_flags)) |
- pgprot_val(arch_vm_get_page_prot(vm_flags)));
-
- return ret;
-}
EXPORT_SYMBOL(vm_get_page_prot);
#endif /* CONFIG_ARCH_HAS_VM_GET_PAGE_PROT */
--
2.25.1
^ permalink raw reply related [flat|nested] 49+ messages in thread
* [OpenRISC] [PATCH V3 14/30] s390/mm: Enable ARCH_HAS_VM_GET_PAGE_PROT
2022-02-28 10:47 [OpenRISC] [PATCH V3 00/30] mm/mmap: Drop protection_map[] and platform's __SXXX/__PXXX requirements Anshuman Khandual
` (12 preceding siblings ...)
2022-02-28 10:47 ` [OpenRISC] [PATCH V3 13/30] mm/mmap: Drop arch_vm_get_page_pgprot() Anshuman Khandual
@ 2022-02-28 10:47 ` Anshuman Khandual
2022-02-28 10:47 ` [OpenRISC] [PATCH V3 15/30] riscv/mm: " Anshuman Khandual
` (15 subsequent siblings)
29 siblings, 0 replies; 49+ messages in thread
From: Anshuman Khandual @ 2022-02-28 10:47 UTC (permalink / raw)
To: openrisc
This defines and exports a platform specific custom vm_get_page_prot() via
subscribing ARCH_HAS_VM_GET_PAGE_PROT. Subsequently all __SXXX and __PXXX
macros can be dropped which are no longer needed.
Cc: Heiko Carstens <hca@linux.ibm.com>
Cc: Vasily Gorbik <gor@linux.ibm.com>
Cc: linux-s390 at vger.kernel.org
Cc: linux-kernel at vger.kernel.org
Acked-by: Sven Schnelle <svens@linux.ibm.com>
Acked-by: Alexander Gordeev <agordeev@linux.ibm.com>
Signed-off-by: Anshuman Khandual <anshuman.khandual@arm.com>
---
arch/s390/Kconfig | 1 +
arch/s390/include/asm/pgtable.h | 17 -----------------
arch/s390/mm/mmap.c | 33 +++++++++++++++++++++++++++++++++
3 files changed, 34 insertions(+), 17 deletions(-)
diff --git a/arch/s390/Kconfig b/arch/s390/Kconfig
index be9f39fd06df..cb1b487e8201 100644
--- a/arch/s390/Kconfig
+++ b/arch/s390/Kconfig
@@ -78,6 +78,7 @@ config S390
select ARCH_HAS_SYSCALL_WRAPPER
select ARCH_HAS_UBSAN_SANITIZE_ALL
select ARCH_HAS_VDSO_DATA
+ select ARCH_HAS_VM_GET_PAGE_PROT
select ARCH_HAVE_NMI_SAFE_CMPXCHG
select ARCH_INLINE_READ_LOCK
select ARCH_INLINE_READ_LOCK_BH
diff --git a/arch/s390/include/asm/pgtable.h b/arch/s390/include/asm/pgtable.h
index 008a6c856fa4..3893ef64b439 100644
--- a/arch/s390/include/asm/pgtable.h
+++ b/arch/s390/include/asm/pgtable.h
@@ -422,23 +422,6 @@ static inline int is_module_addr(void *addr)
* implies read permission.
*/
/*xwr*/
-#define __P000 PAGE_NONE
-#define __P001 PAGE_RO
-#define __P010 PAGE_RO
-#define __P011 PAGE_RO
-#define __P100 PAGE_RX
-#define __P101 PAGE_RX
-#define __P110 PAGE_RX
-#define __P111 PAGE_RX
-
-#define __S000 PAGE_NONE
-#define __S001 PAGE_RO
-#define __S010 PAGE_RW
-#define __S011 PAGE_RW
-#define __S100 PAGE_RX
-#define __S101 PAGE_RX
-#define __S110 PAGE_RWX
-#define __S111 PAGE_RWX
/*
* Segment entry (large page) protection definitions.
diff --git a/arch/s390/mm/mmap.c b/arch/s390/mm/mmap.c
index e54f928503c5..e99c198aa5de 100644
--- a/arch/s390/mm/mmap.c
+++ b/arch/s390/mm/mmap.c
@@ -188,3 +188,36 @@ void arch_pick_mmap_layout(struct mm_struct *mm, struct rlimit *rlim_stack)
mm->get_unmapped_area = arch_get_unmapped_area_topdown;
}
}
+
+pgprot_t vm_get_page_prot(unsigned long vm_flags)
+{
+ switch (vm_flags & (VM_READ | VM_WRITE | VM_EXEC | VM_SHARED)) {
+ case VM_NONE:
+ return PAGE_NONE;
+ case VM_READ:
+ case VM_WRITE:
+ case VM_WRITE | VM_READ:
+ return PAGE_RO;
+ case VM_EXEC:
+ case VM_EXEC | VM_READ:
+ case VM_EXEC | VM_WRITE:
+ case VM_EXEC | VM_WRITE | VM_READ:
+ return PAGE_RX;
+ case VM_SHARED:
+ return PAGE_NONE;
+ case VM_SHARED | VM_READ:
+ return PAGE_RO;
+ case VM_SHARED | VM_WRITE:
+ case VM_SHARED | VM_WRITE | VM_READ:
+ return PAGE_RW;
+ case VM_SHARED | VM_EXEC:
+ case VM_SHARED | VM_EXEC | VM_READ:
+ return PAGE_RX;
+ case VM_SHARED | VM_EXEC | VM_WRITE:
+ case VM_SHARED | VM_EXEC | VM_WRITE | VM_READ:
+ return PAGE_RWX;
+ default:
+ BUILD_BUG();
+ }
+}
+EXPORT_SYMBOL(vm_get_page_prot);
--
2.25.1
^ permalink raw reply related [flat|nested] 49+ messages in thread
* [OpenRISC] [PATCH V3 15/30] riscv/mm: Enable ARCH_HAS_VM_GET_PAGE_PROT
2022-02-28 10:47 [OpenRISC] [PATCH V3 00/30] mm/mmap: Drop protection_map[] and platform's __SXXX/__PXXX requirements Anshuman Khandual
` (13 preceding siblings ...)
2022-02-28 10:47 ` [OpenRISC] [PATCH V3 14/30] s390/mm: Enable ARCH_HAS_VM_GET_PAGE_PROT Anshuman Khandual
@ 2022-02-28 10:47 ` Anshuman Khandual
2022-02-28 10:47 ` [OpenRISC] [PATCH V3 16/30] alpha/mm: " Anshuman Khandual
` (14 subsequent siblings)
29 siblings, 0 replies; 49+ messages in thread
From: Anshuman Khandual @ 2022-02-28 10:47 UTC (permalink / raw)
To: openrisc
This defines and exports a platform specific custom vm_get_page_prot() via
subscribing ARCH_HAS_VM_GET_PAGE_PROT. Subsequently all __SXXX and __PXXX
macros can be dropped which are no longer needed.
Cc: Paul Walmsley <paul.walmsley@sifive.com>
Cc: Palmer Dabbelt <palmer@dabbelt.com>
Cc: linux-riscv at lists.infradead.org
Cc: linux-kernel at vger.kernel.org
Signed-off-by: Anshuman Khandual <anshuman.khandual@arm.com>
---
arch/riscv/Kconfig | 1 +
arch/riscv/include/asm/pgtable.h | 16 ------------
arch/riscv/mm/init.c | 42 ++++++++++++++++++++++++++++++++
3 files changed, 43 insertions(+), 16 deletions(-)
diff --git a/arch/riscv/Kconfig b/arch/riscv/Kconfig
index 5adcbd9b5e88..9391742f9286 100644
--- a/arch/riscv/Kconfig
+++ b/arch/riscv/Kconfig
@@ -31,6 +31,7 @@ config RISCV
select ARCH_HAS_STRICT_MODULE_RWX if MMU && !XIP_KERNEL
select ARCH_HAS_TICK_BROADCAST if GENERIC_CLOCKEVENTS_BROADCAST
select ARCH_HAS_UBSAN_SANITIZE_ALL
+ select ARCH_HAS_VM_GET_PAGE_PROT
select ARCH_OPTIONAL_KERNEL_RWX if ARCH_HAS_STRICT_KERNEL_RWX
select ARCH_OPTIONAL_KERNEL_RWX_DEFAULT
select ARCH_STACKWALK
diff --git a/arch/riscv/include/asm/pgtable.h b/arch/riscv/include/asm/pgtable.h
index 7e949f25c933..d2bb14cac28b 100644
--- a/arch/riscv/include/asm/pgtable.h
+++ b/arch/riscv/include/asm/pgtable.h
@@ -183,24 +183,8 @@ extern struct pt_alloc_ops pt_ops __initdata;
extern pgd_t swapper_pg_dir[];
/* MAP_PRIVATE permissions: xwr (copy-on-write) */
-#define __P000 PAGE_NONE
-#define __P001 PAGE_READ
-#define __P010 PAGE_COPY
-#define __P011 PAGE_COPY
-#define __P100 PAGE_EXEC
-#define __P101 PAGE_READ_EXEC
-#define __P110 PAGE_COPY_EXEC
-#define __P111 PAGE_COPY_READ_EXEC
/* MAP_SHARED permissions: xwr */
-#define __S000 PAGE_NONE
-#define __S001 PAGE_READ
-#define __S010 PAGE_SHARED
-#define __S011 PAGE_SHARED
-#define __S100 PAGE_EXEC
-#define __S101 PAGE_READ_EXEC
-#define __S110 PAGE_SHARED_EXEC
-#define __S111 PAGE_SHARED_EXEC
#ifdef CONFIG_TRANSPARENT_HUGEPAGE
static inline int pmd_present(pmd_t pmd)
diff --git a/arch/riscv/mm/init.c b/arch/riscv/mm/init.c
index c27294128e18..8cb5d1eeb287 100644
--- a/arch/riscv/mm/init.c
+++ b/arch/riscv/mm/init.c
@@ -1050,3 +1050,45 @@ int __meminit vmemmap_populate(unsigned long start, unsigned long end, int node,
return vmemmap_populate_basepages(start, end, node, NULL);
}
#endif
+
+#ifdef CONFIG_MMU
+pgprot_t vm_get_page_prot(unsigned long vm_flags)
+{
+ switch (vm_flags & (VM_READ | VM_WRITE | VM_EXEC | VM_SHARED)) {
+ /* MAP_PRIVATE permissions: xwr (copy-on-write) */
+ case VM_NONE:
+ return PAGE_NONE;
+ case VM_READ:
+ return PAGE_READ;
+ case VM_WRITE:
+ case VM_WRITE | VM_READ:
+ return PAGE_COPY;
+ case VM_EXEC:
+ return PAGE_EXEC;
+ case VM_EXEC | VM_READ:
+ return PAGE_READ_EXEC;
+ case VM_EXEC | VM_WRITE:
+ return PAGE_COPY_EXEC;
+ case VM_EXEC | VM_WRITE | VM_READ:
+ return PAGE_COPY_READ_EXEC;
+ /* MAP_SHARED permissions: xwr */
+ case VM_SHARED:
+ return PAGE_NONE;
+ case VM_SHARED | VM_READ:
+ return PAGE_READ;
+ case VM_SHARED | VM_WRITE:
+ case VM_SHARED | VM_WRITE | VM_READ:
+ return PAGE_SHARED;
+ case VM_SHARED | VM_EXEC:
+ return PAGE_EXEC;
+ case VM_SHARED | VM_EXEC | VM_READ:
+ return PAGE_READ_EXEC;
+ case VM_SHARED | VM_EXEC | VM_WRITE:
+ case VM_SHARED | VM_EXEC | VM_WRITE | VM_READ:
+ return PAGE_SHARED_EXEC;
+ default:
+ BUILD_BUG();
+ }
+}
+EXPORT_SYMBOL(vm_get_page_prot);
+#endif /* CONFIG_MMU */
--
2.25.1
^ permalink raw reply related [flat|nested] 49+ messages in thread
* [OpenRISC] [PATCH V3 16/30] alpha/mm: Enable ARCH_HAS_VM_GET_PAGE_PROT
2022-02-28 10:47 [OpenRISC] [PATCH V3 00/30] mm/mmap: Drop protection_map[] and platform's __SXXX/__PXXX requirements Anshuman Khandual
` (14 preceding siblings ...)
2022-02-28 10:47 ` [OpenRISC] [PATCH V3 15/30] riscv/mm: " Anshuman Khandual
@ 2022-02-28 10:47 ` Anshuman Khandual
2022-02-28 10:47 ` [OpenRISC] [PATCH V3 17/30] sh/mm: " Anshuman Khandual
` (13 subsequent siblings)
29 siblings, 0 replies; 49+ messages in thread
From: Anshuman Khandual @ 2022-02-28 10:47 UTC (permalink / raw)
To: openrisc
This defines and exports a platform specific custom vm_get_page_prot() via
subscribing ARCH_HAS_VM_GET_PAGE_PROT. Subsequently all __SXXX and __PXXX
macros can be dropped which are no longer needed.
Cc: Richard Henderson <rth@twiddle.net>
Cc: linux-alpha at vger.kernel.org
Cc: linux-kernel at vger.kernel.org
Signed-off-by: Anshuman Khandual <anshuman.khandual@arm.com>
---
arch/alpha/Kconfig | 1 +
arch/alpha/include/asm/pgtable.h | 17 ---------------
arch/alpha/mm/init.c | 37 ++++++++++++++++++++++++++++++++
3 files changed, 38 insertions(+), 17 deletions(-)
diff --git a/arch/alpha/Kconfig b/arch/alpha/Kconfig
index 4e87783c90ad..73e82fe5c770 100644
--- a/arch/alpha/Kconfig
+++ b/arch/alpha/Kconfig
@@ -2,6 +2,7 @@
config ALPHA
bool
default y
+ select ARCH_HAS_VM_GET_PAGE_PROT
select ARCH_32BIT_USTAT_F_TINODE
select ARCH_MIGHT_HAVE_PC_PARPORT
select ARCH_MIGHT_HAVE_PC_SERIO
diff --git a/arch/alpha/include/asm/pgtable.h b/arch/alpha/include/asm/pgtable.h
index 02f0429f1068..9fb5e9d10bb6 100644
--- a/arch/alpha/include/asm/pgtable.h
+++ b/arch/alpha/include/asm/pgtable.h
@@ -116,23 +116,6 @@ struct vm_area_struct;
* arch/alpha/mm/fault.c)
*/
/* xwr */
-#define __P000 _PAGE_P(_PAGE_FOE | _PAGE_FOW | _PAGE_FOR)
-#define __P001 _PAGE_P(_PAGE_FOE | _PAGE_FOW)
-#define __P010 _PAGE_P(_PAGE_FOE)
-#define __P011 _PAGE_P(_PAGE_FOE)
-#define __P100 _PAGE_P(_PAGE_FOW | _PAGE_FOR)
-#define __P101 _PAGE_P(_PAGE_FOW)
-#define __P110 _PAGE_P(0)
-#define __P111 _PAGE_P(0)
-
-#define __S000 _PAGE_S(_PAGE_FOE | _PAGE_FOW | _PAGE_FOR)
-#define __S001 _PAGE_S(_PAGE_FOE | _PAGE_FOW)
-#define __S010 _PAGE_S(_PAGE_FOE)
-#define __S011 _PAGE_S(_PAGE_FOE)
-#define __S100 _PAGE_S(_PAGE_FOW | _PAGE_FOR)
-#define __S101 _PAGE_S(_PAGE_FOW)
-#define __S110 _PAGE_S(0)
-#define __S111 _PAGE_S(0)
/*
* pgprot_noncached() is only for infiniband pci support, and a real
diff --git a/arch/alpha/mm/init.c b/arch/alpha/mm/init.c
index f6114d03357c..2e78008b2553 100644
--- a/arch/alpha/mm/init.c
+++ b/arch/alpha/mm/init.c
@@ -280,3 +280,40 @@ mem_init(void)
high_memory = (void *) __va(max_low_pfn * PAGE_SIZE);
memblock_free_all();
}
+
+pgprot_t vm_get_page_prot(unsigned long vm_flags)
+{
+ switch (vm_flags & (VM_READ | VM_WRITE | VM_EXEC | VM_SHARED)) {
+ case VM_NONE:
+ return _PAGE_P(_PAGE_FOE | _PAGE_FOW | _PAGE_FOR);
+ case VM_READ:
+ return _PAGE_P(_PAGE_FOE | _PAGE_FOW);
+ case VM_WRITE:
+ case VM_WRITE | VM_READ:
+ return _PAGE_P(_PAGE_FOE);
+ case VM_EXEC:
+ return _PAGE_P(_PAGE_FOW | _PAGE_FOR);
+ case VM_EXEC | VM_READ:
+ return _PAGE_P(_PAGE_FOW);
+ case VM_EXEC | VM_WRITE:
+ case VM_EXEC | VM_WRITE | VM_READ:
+ return _PAGE_P(0);
+ case VM_SHARED:
+ return _PAGE_S(_PAGE_FOE | _PAGE_FOW | _PAGE_FOR);
+ case VM_SHARED | VM_READ:
+ return _PAGE_S(_PAGE_FOE | _PAGE_FOW);
+ case VM_SHARED | VM_WRITE:
+ case VM_SHARED | VM_WRITE | VM_READ:
+ return _PAGE_S(_PAGE_FOE);
+ case VM_SHARED | VM_EXEC:
+ return _PAGE_S(_PAGE_FOW | _PAGE_FOR);
+ case VM_SHARED | VM_EXEC | VM_READ:
+ return _PAGE_S(_PAGE_FOW);
+ case VM_SHARED | VM_EXEC | VM_WRITE:
+ case VM_SHARED | VM_EXEC | VM_WRITE | VM_READ:
+ return _PAGE_S(0);
+ default:
+ BUILD_BUG();
+ }
+}
+EXPORT_SYMBOL(vm_get_page_prot);
--
2.25.1
^ permalink raw reply related [flat|nested] 49+ messages in thread
* [OpenRISC] [PATCH V3 17/30] sh/mm: Enable ARCH_HAS_VM_GET_PAGE_PROT
2022-02-28 10:47 [OpenRISC] [PATCH V3 00/30] mm/mmap: Drop protection_map[] and platform's __SXXX/__PXXX requirements Anshuman Khandual
` (15 preceding siblings ...)
2022-02-28 10:47 ` [OpenRISC] [PATCH V3 16/30] alpha/mm: " Anshuman Khandual
@ 2022-02-28 10:47 ` Anshuman Khandual
2022-02-28 10:47 ` [OpenRISC] [PATCH V3 18/30] arc/mm: " Anshuman Khandual
` (12 subsequent siblings)
29 siblings, 0 replies; 49+ messages in thread
From: Anshuman Khandual @ 2022-02-28 10:47 UTC (permalink / raw)
To: openrisc
This defines and exports a platform specific custom vm_get_page_prot() via
subscribing ARCH_HAS_VM_GET_PAGE_PROT. Subsequently all __SXXX and __PXXX
macros can be dropped which are no longer needed.
Cc: Yoshinori Sato <ysato@users.sourceforge.jp>
Cc: Rich Felker <dalias@libc.org>
Cc: linux-sh at vger.kernel.org
Cc: linux-kernel at vger.kernel.org
Signed-off-by: Anshuman Khandual <anshuman.khandual@arm.com>
---
arch/sh/Kconfig | 1 +
arch/sh/include/asm/pgtable.h | 17 ----------------
arch/sh/mm/mmap.c | 38 +++++++++++++++++++++++++++++++++++
3 files changed, 39 insertions(+), 17 deletions(-)
diff --git a/arch/sh/Kconfig b/arch/sh/Kconfig
index 2474a04ceac4..f3fcd1c5e002 100644
--- a/arch/sh/Kconfig
+++ b/arch/sh/Kconfig
@@ -11,6 +11,7 @@ config SUPERH
select ARCH_HAS_GCOV_PROFILE_ALL
select ARCH_HAS_PTE_SPECIAL
select ARCH_HAS_TICK_BROADCAST if GENERIC_CLOCKEVENTS_BROADCAST
+ select ARCH_HAS_VM_GET_PAGE_PROT
select ARCH_HIBERNATION_POSSIBLE if MMU
select ARCH_MIGHT_HAVE_PC_PARPORT
select ARCH_WANT_IPC_PARSE_VERSION
diff --git a/arch/sh/include/asm/pgtable.h b/arch/sh/include/asm/pgtable.h
index d7ddb1ec86a0..6fb9ec54cf9b 100644
--- a/arch/sh/include/asm/pgtable.h
+++ b/arch/sh/include/asm/pgtable.h
@@ -89,23 +89,6 @@ static inline unsigned long phys_addr_mask(void)
* completely separate permission bits for user and kernel space.
*/
/*xwr*/
-#define __P000 PAGE_NONE
-#define __P001 PAGE_READONLY
-#define __P010 PAGE_COPY
-#define __P011 PAGE_COPY
-#define __P100 PAGE_EXECREAD
-#define __P101 PAGE_EXECREAD
-#define __P110 PAGE_COPY
-#define __P111 PAGE_COPY
-
-#define __S000 PAGE_NONE
-#define __S001 PAGE_READONLY
-#define __S010 PAGE_WRITEONLY
-#define __S011 PAGE_SHARED
-#define __S100 PAGE_EXECREAD
-#define __S101 PAGE_EXECREAD
-#define __S110 PAGE_RWX
-#define __S111 PAGE_RWX
typedef pte_t *pte_addr_t;
diff --git a/arch/sh/mm/mmap.c b/arch/sh/mm/mmap.c
index 6a1a1297baae..cad14af6c8e6 100644
--- a/arch/sh/mm/mmap.c
+++ b/arch/sh/mm/mmap.c
@@ -162,3 +162,41 @@ int valid_mmap_phys_addr_range(unsigned long pfn, size_t size)
{
return 1;
}
+
+#ifdef CONFIG_MMU
+pgprot_t vm_get_page_prot(unsigned long vm_flags)
+{
+ switch (vm_flags & (VM_READ | VM_WRITE | VM_EXEC | VM_SHARED)) {
+ case VM_NONE:
+ return PAGE_NONE;
+ case VM_READ:
+ return PAGE_READONLY;
+ case VM_WRITE:
+ case VM_WRITE | VM_READ:
+ return PAGE_COPY;
+ case VM_EXEC:
+ case VM_EXEC | VM_READ:
+ return PAGE_EXECREAD;
+ case VM_EXEC | VM_WRITE:
+ case VM_EXEC | VM_WRITE | VM_READ:
+ return PAGE_COPY;
+ case VM_SHARED:
+ return PAGE_NONE;
+ case VM_SHARED | VM_READ:
+ return PAGE_READONLY;
+ case VM_SHARED | VM_WRITE:
+ return PAGE_WRITEONLY;
+ case VM_SHARED | VM_WRITE | VM_READ:
+ return PAGE_SHARED;
+ case VM_SHARED | VM_EXEC:
+ case VM_SHARED | VM_EXEC | VM_READ:
+ return PAGE_EXECREAD;
+ case VM_SHARED | VM_EXEC | VM_WRITE:
+ case VM_SHARED | VM_EXEC | VM_WRITE | VM_READ:
+ return PAGE_RWX;
+ default:
+ BUILD_BUG();
+ }
+}
+EXPORT_SYMBOL(vm_get_page_prot);
+#endif
--
2.25.1
^ permalink raw reply related [flat|nested] 49+ messages in thread
* [OpenRISC] [PATCH V3 18/30] arc/mm: Enable ARCH_HAS_VM_GET_PAGE_PROT
2022-02-28 10:47 [OpenRISC] [PATCH V3 00/30] mm/mmap: Drop protection_map[] and platform's __SXXX/__PXXX requirements Anshuman Khandual
` (16 preceding siblings ...)
2022-02-28 10:47 ` [OpenRISC] [PATCH V3 17/30] sh/mm: " Anshuman Khandual
@ 2022-02-28 10:47 ` Anshuman Khandual
2022-02-28 10:47 ` [OpenRISC] [PATCH V3 19/30] csky/mm: " Anshuman Khandual
` (11 subsequent siblings)
29 siblings, 0 replies; 49+ messages in thread
From: Anshuman Khandual @ 2022-02-28 10:47 UTC (permalink / raw)
To: openrisc
This defines and exports a platform specific custom vm_get_page_prot() via
subscribing ARCH_HAS_VM_GET_PAGE_PROT. Subsequently all __SXXX and __PXXX
macros can be dropped which are no longer needed.
Cc: Vineet Gupta <vgupta@kernel.org>
Cc: linux-snps-arc at lists.infradead.org
Cc: linux-kernel at vger.kernel.org
Signed-off-by: Anshuman Khandual <anshuman.khandual@arm.com>
---
arch/arc/Kconfig | 1 +
arch/arc/include/asm/pgtable-bits-arcv2.h | 17 ----------
arch/arc/mm/mmap.c | 41 +++++++++++++++++++++++
3 files changed, 42 insertions(+), 17 deletions(-)
diff --git a/arch/arc/Kconfig b/arch/arc/Kconfig
index 3c2a4753d09b..78ff0644b343 100644
--- a/arch/arc/Kconfig
+++ b/arch/arc/Kconfig
@@ -13,6 +13,7 @@ config ARC
select ARCH_HAS_SETUP_DMA_OPS
select ARCH_HAS_SYNC_DMA_FOR_CPU
select ARCH_HAS_SYNC_DMA_FOR_DEVICE
+ select ARCH_HAS_VM_GET_PAGE_PROT
select ARCH_SUPPORTS_ATOMIC_RMW if ARC_HAS_LLSC
select ARCH_32BIT_OFF_T
select BUILDTIME_TABLE_SORT
diff --git a/arch/arc/include/asm/pgtable-bits-arcv2.h b/arch/arc/include/asm/pgtable-bits-arcv2.h
index 183d23bc1e00..798308f4dbad 100644
--- a/arch/arc/include/asm/pgtable-bits-arcv2.h
+++ b/arch/arc/include/asm/pgtable-bits-arcv2.h
@@ -72,23 +72,6 @@
* This is to enable COW mechanism
*/
/* xwr */
-#define __P000 PAGE_U_NONE
-#define __P001 PAGE_U_R
-#define __P010 PAGE_U_R /* Pvt-W => !W */
-#define __P011 PAGE_U_R /* Pvt-W => !W */
-#define __P100 PAGE_U_X_R /* X => R */
-#define __P101 PAGE_U_X_R
-#define __P110 PAGE_U_X_R /* Pvt-W => !W and X => R */
-#define __P111 PAGE_U_X_R /* Pvt-W => !W */
-
-#define __S000 PAGE_U_NONE
-#define __S001 PAGE_U_R
-#define __S010 PAGE_U_W_R /* W => R */
-#define __S011 PAGE_U_W_R
-#define __S100 PAGE_U_X_R /* X => R */
-#define __S101 PAGE_U_X_R
-#define __S110 PAGE_U_X_W_R /* X => R */
-#define __S111 PAGE_U_X_W_R
#ifndef __ASSEMBLY__
diff --git a/arch/arc/mm/mmap.c b/arch/arc/mm/mmap.c
index 722d26b94307..d286894d7359 100644
--- a/arch/arc/mm/mmap.c
+++ b/arch/arc/mm/mmap.c
@@ -74,3 +74,44 @@ arch_get_unmapped_area(struct file *filp, unsigned long addr,
info.align_offset = pgoff << PAGE_SHIFT;
return vm_unmapped_area(&info);
}
+
+pgprot_t vm_get_page_prot(unsigned long vm_flags)
+{
+ switch (vm_flags & (VM_READ | VM_WRITE | VM_EXEC | VM_SHARED)) {
+ case VM_NONE:
+ return PAGE_U_NONE;
+ case VM_READ:
+ /* Pvt-W => !W */
+ case VM_WRITE:
+ /* Pvt-W => !W */
+ case VM_WRITE | VM_READ:
+ return PAGE_U_R;
+ /* X => R */
+ case VM_EXEC:
+ case VM_EXEC | VM_READ:
+ /* Pvt-W => !W and X => R */
+ case VM_EXEC | VM_WRITE:
+ /* Pvt-W => !W */
+ case VM_EXEC | VM_WRITE | VM_READ:
+ return PAGE_U_X_R;
+ case VM_SHARED:
+ return PAGE_U_NONE;
+ case VM_SHARED | VM_READ:
+ return PAGE_U_R;
+ /* W => R */
+ case VM_SHARED | VM_WRITE:
+ case VM_SHARED | VM_WRITE | VM_READ:
+ return PAGE_U_W_R;
+ /* X => R */
+ case VM_SHARED | VM_EXEC:
+ case VM_SHARED | VM_EXEC | VM_READ:
+ return PAGE_U_X_R;
+ /* X => R */
+ case VM_SHARED | VM_EXEC | VM_WRITE:
+ case VM_SHARED | VM_EXEC | VM_WRITE | VM_READ:
+ return PAGE_U_X_W_R;
+ default:
+ BUILD_BUG();
+ }
+}
+EXPORT_SYMBOL(vm_get_page_prot);
--
2.25.1
^ permalink raw reply related [flat|nested] 49+ messages in thread
* [OpenRISC] [PATCH V3 19/30] csky/mm: Enable ARCH_HAS_VM_GET_PAGE_PROT
2022-02-28 10:47 [OpenRISC] [PATCH V3 00/30] mm/mmap: Drop protection_map[] and platform's __SXXX/__PXXX requirements Anshuman Khandual
` (17 preceding siblings ...)
2022-02-28 10:47 ` [OpenRISC] [PATCH V3 18/30] arc/mm: " Anshuman Khandual
@ 2022-02-28 10:47 ` Anshuman Khandual
2022-03-01 14:00 ` Guo Ren
2022-02-28 10:47 ` [OpenRISC] [PATCH V3 20/30] xtensa/mm: " Anshuman Khandual
` (10 subsequent siblings)
29 siblings, 1 reply; 49+ messages in thread
From: Anshuman Khandual @ 2022-02-28 10:47 UTC (permalink / raw)
To: openrisc
This defines and exports a platform specific custom vm_get_page_prot() via
subscribing ARCH_HAS_VM_GET_PAGE_PROT. Subsequently all __SXXX and __PXXX
macros can be dropped which are no longer needed.
Cc: Geert Uytterhoeven <geert@linux-m68k.org>
Cc: linux-csky at vger.kernel.org
Cc: linux-kernel at vger.kernel.org
Signed-off-by: Anshuman Khandual <anshuman.khandual@arm.com>
---
arch/csky/Kconfig | 1 +
arch/csky/include/asm/pgtable.h | 18 ------------------
arch/csky/mm/init.c | 32 ++++++++++++++++++++++++++++++++
3 files changed, 33 insertions(+), 18 deletions(-)
diff --git a/arch/csky/Kconfig b/arch/csky/Kconfig
index 132f43f12dd8..209dac5686dd 100644
--- a/arch/csky/Kconfig
+++ b/arch/csky/Kconfig
@@ -6,6 +6,7 @@ config CSKY
select ARCH_HAS_GCOV_PROFILE_ALL
select ARCH_HAS_SYNC_DMA_FOR_CPU
select ARCH_HAS_SYNC_DMA_FOR_DEVICE
+ select ARCH_HAS_VM_GET_PAGE_PROT
select ARCH_USE_BUILTIN_BSWAP
select ARCH_USE_QUEUED_RWLOCKS
select ARCH_WANT_FRAME_POINTERS if !CPU_CK610 && $(cc-option,-mbacktrace)
diff --git a/arch/csky/include/asm/pgtable.h b/arch/csky/include/asm/pgtable.h
index 151607ed5158..2c6b1cfb1cce 100644
--- a/arch/csky/include/asm/pgtable.h
+++ b/arch/csky/include/asm/pgtable.h
@@ -76,24 +76,6 @@
#define MAX_SWAPFILES_CHECK() \
BUILD_BUG_ON(MAX_SWAPFILES_SHIFT != 5)
-#define __P000 PAGE_NONE
-#define __P001 PAGE_READ
-#define __P010 PAGE_READ
-#define __P011 PAGE_READ
-#define __P100 PAGE_READ
-#define __P101 PAGE_READ
-#define __P110 PAGE_READ
-#define __P111 PAGE_READ
-
-#define __S000 PAGE_NONE
-#define __S001 PAGE_READ
-#define __S010 PAGE_WRITE
-#define __S011 PAGE_WRITE
-#define __S100 PAGE_READ
-#define __S101 PAGE_READ
-#define __S110 PAGE_WRITE
-#define __S111 PAGE_WRITE
-
extern unsigned long empty_zero_page[PAGE_SIZE / sizeof(unsigned long)];
#define ZERO_PAGE(vaddr) (virt_to_page(empty_zero_page))
diff --git a/arch/csky/mm/init.c b/arch/csky/mm/init.c
index bf2004aa811a..f9babbed17d4 100644
--- a/arch/csky/mm/init.c
+++ b/arch/csky/mm/init.c
@@ -197,3 +197,35 @@ void __init fixaddr_init(void)
vaddr = __fix_to_virt(__end_of_fixed_addresses - 1) & PMD_MASK;
fixrange_init(vaddr, vaddr + PMD_SIZE, swapper_pg_dir);
}
+
+pgprot_t vm_get_page_prot(unsigned long vm_flags)
+{
+ switch (vm_flags & (VM_READ | VM_WRITE | VM_EXEC | VM_SHARED)) {
+ case VM_NONE:
+ return PAGE_NONE;
+ case VM_READ:
+ case VM_WRITE:
+ case VM_WRITE | VM_READ:
+ case VM_EXEC:
+ case VM_EXEC | VM_READ:
+ case VM_EXEC | VM_WRITE:
+ case VM_EXEC | VM_WRITE | VM_READ:
+ return PAGE_READ;
+ case VM_SHARED:
+ return PAGE_NONE;
+ case VM_SHARED | VM_READ:
+ return PAGE_READ;
+ case VM_SHARED | VM_WRITE:
+ case VM_SHARED | VM_WRITE | VM_READ:
+ return PAGE_WRITE;
+ case VM_SHARED | VM_EXEC:
+ case VM_SHARED | VM_EXEC | VM_READ:
+ return PAGE_READ;
+ case VM_SHARED | VM_EXEC | VM_WRITE:
+ case VM_SHARED | VM_EXEC | VM_WRITE | VM_READ:
+ return PAGE_WRITE;
+ default:
+ BUILD_BUG();
+ }
+}
+EXPORT_SYMBOL(vm_get_page_prot);
--
2.25.1
^ permalink raw reply related [flat|nested] 49+ messages in thread
* [OpenRISC] [PATCH V3 20/30] xtensa/mm: Enable ARCH_HAS_VM_GET_PAGE_PROT
2022-02-28 10:47 [OpenRISC] [PATCH V3 00/30] mm/mmap: Drop protection_map[] and platform's __SXXX/__PXXX requirements Anshuman Khandual
` (18 preceding siblings ...)
2022-02-28 10:47 ` [OpenRISC] [PATCH V3 19/30] csky/mm: " Anshuman Khandual
@ 2022-02-28 10:47 ` Anshuman Khandual
2022-02-28 10:47 ` [OpenRISC] [PATCH V3 21/30] parisc/mm: " Anshuman Khandual
` (9 subsequent siblings)
29 siblings, 0 replies; 49+ messages in thread
From: Anshuman Khandual @ 2022-02-28 10:47 UTC (permalink / raw)
To: openrisc
This defines and exports a platform specific custom vm_get_page_prot() via
subscribing ARCH_HAS_VM_GET_PAGE_PROT. Subsequently all __SXXX and __PXXX
macros can be dropped which are no longer needed.
Cc: Chris Zankel <chris@zankel.net>
Cc: Guo Ren <guoren@kernel.org>
Cc: linux-xtensa at linux-xtensa.org
Cc: linux-csky at vger.kernel.org
Cc: linux-kernel at vger.kernel.org
Signed-off-by: Anshuman Khandual <anshuman.khandual@arm.com>
---
arch/xtensa/Kconfig | 1 +
arch/xtensa/include/asm/pgtable.h | 19 -----------------
arch/xtensa/mm/init.c | 35 +++++++++++++++++++++++++++++++
3 files changed, 36 insertions(+), 19 deletions(-)
diff --git a/arch/xtensa/Kconfig b/arch/xtensa/Kconfig
index 8ac599aa6d99..1608f7517546 100644
--- a/arch/xtensa/Kconfig
+++ b/arch/xtensa/Kconfig
@@ -9,6 +9,7 @@ config XTENSA
select ARCH_HAS_DMA_SET_UNCACHED if MMU
select ARCH_HAS_STRNCPY_FROM_USER if !KASAN
select ARCH_HAS_STRNLEN_USER
+ select ARCH_HAS_VM_GET_PAGE_PROT
select ARCH_USE_MEMTEST
select ARCH_USE_QUEUED_RWLOCKS
select ARCH_USE_QUEUED_SPINLOCKS
diff --git a/arch/xtensa/include/asm/pgtable.h b/arch/xtensa/include/asm/pgtable.h
index bd5aeb795567..509f765281d8 100644
--- a/arch/xtensa/include/asm/pgtable.h
+++ b/arch/xtensa/include/asm/pgtable.h
@@ -198,26 +198,7 @@
* the MMU can't do page protection for execute, and considers that the same as
* read. Also, write permissions may imply read permissions.
* What follows is the closest we can get by reasonable means..
- * See linux/mm/mmap.c for protection_map[] array that uses these definitions.
*/
-#define __P000 PAGE_NONE /* private --- */
-#define __P001 PAGE_READONLY /* private --r */
-#define __P010 PAGE_COPY /* private -w- */
-#define __P011 PAGE_COPY /* private -wr */
-#define __P100 PAGE_READONLY_EXEC /* private x-- */
-#define __P101 PAGE_READONLY_EXEC /* private x-r */
-#define __P110 PAGE_COPY_EXEC /* private xw- */
-#define __P111 PAGE_COPY_EXEC /* private xwr */
-
-#define __S000 PAGE_NONE /* shared --- */
-#define __S001 PAGE_READONLY /* shared --r */
-#define __S010 PAGE_SHARED /* shared -w- */
-#define __S011 PAGE_SHARED /* shared -wr */
-#define __S100 PAGE_READONLY_EXEC /* shared x-- */
-#define __S101 PAGE_READONLY_EXEC /* shared x-r */
-#define __S110 PAGE_SHARED_EXEC /* shared xw- */
-#define __S111 PAGE_SHARED_EXEC /* shared xwr */
-
#ifndef __ASSEMBLY__
#define pte_ERROR(e) \
diff --git a/arch/xtensa/mm/init.c b/arch/xtensa/mm/init.c
index 6a32b2cf2718..5f090749e9e0 100644
--- a/arch/xtensa/mm/init.c
+++ b/arch/xtensa/mm/init.c
@@ -216,3 +216,38 @@ static int __init parse_memmap_opt(char *str)
return 0;
}
early_param("memmap", parse_memmap_opt);
+
+pgprot_t vm_get_page_prot(unsigned long vm_flags)
+{
+ switch (vm_flags & (VM_READ | VM_WRITE | VM_EXEC | VM_SHARED)) {
+ case VM_NONE:
+ return PAGE_NONE;
+ case VM_READ:
+ return PAGE_READONLY;
+ case VM_WRITE:
+ case VM_WRITE | VM_READ:
+ return PAGE_COPY;
+ case VM_EXEC:
+ case VM_EXEC | VM_READ:
+ return PAGE_READONLY_EXEC;
+ case VM_EXEC | VM_WRITE:
+ case VM_EXEC | VM_WRITE | VM_READ:
+ return PAGE_COPY_EXEC;
+ case VM_SHARED:
+ return PAGE_NONE;
+ case VM_SHARED | VM_READ:
+ return PAGE_READONLY;
+ case VM_SHARED | VM_WRITE:
+ case VM_SHARED | VM_WRITE | VM_READ:
+ return PAGE_SHARED;
+ case VM_SHARED | VM_EXEC:
+ case VM_SHARED | VM_EXEC | VM_READ:
+ return PAGE_READONLY_EXEC;
+ case VM_SHARED | VM_EXEC | VM_WRITE:
+ case VM_SHARED | VM_EXEC | VM_WRITE | VM_READ:
+ return PAGE_SHARED_EXEC;
+ default:
+ BUILD_BUG();
+ }
+}
+EXPORT_SYMBOL(vm_get_page_prot);
--
2.25.1
^ permalink raw reply related [flat|nested] 49+ messages in thread
* [OpenRISC] [PATCH V3 21/30] parisc/mm: Enable ARCH_HAS_VM_GET_PAGE_PROT
2022-02-28 10:47 [OpenRISC] [PATCH V3 00/30] mm/mmap: Drop protection_map[] and platform's __SXXX/__PXXX requirements Anshuman Khandual
` (19 preceding siblings ...)
2022-02-28 10:47 ` [OpenRISC] [PATCH V3 20/30] xtensa/mm: " Anshuman Khandual
@ 2022-02-28 10:47 ` Anshuman Khandual
2022-02-28 10:47 ` [OpenRISC] [PATCH V3 22/30] openrisc/mm: " Anshuman Khandual
` (8 subsequent siblings)
29 siblings, 0 replies; 49+ messages in thread
From: Anshuman Khandual @ 2022-02-28 10:47 UTC (permalink / raw)
To: openrisc
This defines and exports a platform specific custom vm_get_page_prot() via
subscribing ARCH_HAS_VM_GET_PAGE_PROT. Subsequently all __SXXX and __PXXX
macros can be dropped which are no longer needed.
Cc: "James E.J. Bottomley" <James.Bottomley@HansenPartnership.com>
Cc: linux-parisc at vger.kernel.org
Cc: linux-kernel at vger.kernel.org
Signed-off-by: Anshuman Khandual <anshuman.khandual@arm.com>
---
arch/parisc/Kconfig | 1 +
arch/parisc/include/asm/pgtable.h | 20 ----------------
arch/parisc/mm/init.c | 40 +++++++++++++++++++++++++++++++
3 files changed, 41 insertions(+), 20 deletions(-)
diff --git a/arch/parisc/Kconfig b/arch/parisc/Kconfig
index 43c1c880def6..de512f120b50 100644
--- a/arch/parisc/Kconfig
+++ b/arch/parisc/Kconfig
@@ -10,6 +10,7 @@ config PARISC
select ARCH_HAS_ELF_RANDOMIZE
select ARCH_HAS_STRICT_KERNEL_RWX
select ARCH_HAS_UBSAN_SANITIZE_ALL
+ select ARCH_HAS_VM_GET_PAGE_PROT
select ARCH_NO_SG_CHAIN
select ARCH_SUPPORTS_HUGETLBFS if PA20
select ARCH_SUPPORTS_MEMORY_FAILURE
diff --git a/arch/parisc/include/asm/pgtable.h b/arch/parisc/include/asm/pgtable.h
index 3e7cf882639f..80d99b2b5913 100644
--- a/arch/parisc/include/asm/pgtable.h
+++ b/arch/parisc/include/asm/pgtable.h
@@ -269,26 +269,6 @@ extern void __update_cache(pte_t pte);
* pages.
*/
- /*xwr*/
-#define __P000 PAGE_NONE
-#define __P001 PAGE_READONLY
-#define __P010 __P000 /* copy on write */
-#define __P011 __P001 /* copy on write */
-#define __P100 PAGE_EXECREAD
-#define __P101 PAGE_EXECREAD
-#define __P110 __P100 /* copy on write */
-#define __P111 __P101 /* copy on write */
-
-#define __S000 PAGE_NONE
-#define __S001 PAGE_READONLY
-#define __S010 PAGE_WRITEONLY
-#define __S011 PAGE_SHARED
-#define __S100 PAGE_EXECREAD
-#define __S101 PAGE_EXECREAD
-#define __S110 PAGE_RWX
-#define __S111 PAGE_RWX
-
-
extern pgd_t swapper_pg_dir[]; /* declared in init_task.c */
/* initial page tables for 0-8MB for kernel */
diff --git a/arch/parisc/mm/init.c b/arch/parisc/mm/init.c
index 1dc2e88e7b04..f9e841f874a8 100644
--- a/arch/parisc/mm/init.c
+++ b/arch/parisc/mm/init.c
@@ -865,3 +865,43 @@ void flush_tlb_all(void)
spin_unlock(&sid_lock);
}
#endif
+
+pgprot_t vm_get_page_prot(unsigned long vm_flags)
+{
+ switch (vm_flags & (VM_READ | VM_WRITE | VM_EXEC | VM_SHARED)) {
+ case VM_NONE:
+ return PAGE_NONE;
+ case VM_READ:
+ return PAGE_READONLY;
+ /* copy on write */
+ case VM_WRITE:
+ return PAGE_NONE;
+ /* copy on write */
+ case VM_WRITE | VM_READ:
+ return PAGE_READONLY;
+ case VM_EXEC:
+ case VM_EXEC | VM_READ:
+ /* copy on write */
+ case VM_EXEC | VM_WRITE:
+ /* copy on write */
+ case VM_EXEC | VM_WRITE | VM_READ:
+ return PAGE_EXECREAD;
+ case VM_SHARED:
+ return PAGE_NONE;
+ case VM_SHARED | VM_READ:
+ return PAGE_READONLY;
+ case VM_SHARED | VM_WRITE:
+ return PAGE_WRITEONLY;
+ case VM_SHARED | VM_WRITE | VM_READ:
+ return PAGE_SHARED;
+ case VM_SHARED | VM_EXEC:
+ case VM_SHARED | VM_EXEC | VM_READ:
+ return PAGE_EXECREAD;
+ case VM_SHARED | VM_EXEC | VM_WRITE:
+ case VM_SHARED | VM_EXEC | VM_WRITE | VM_READ:
+ return PAGE_RWX;
+ default:
+ BUILD_BUG();
+ }
+}
+EXPORT_SYMBOL(vm_get_page_prot);
--
2.25.1
^ permalink raw reply related [flat|nested] 49+ messages in thread
* [OpenRISC] [PATCH V3 22/30] openrisc/mm: Enable ARCH_HAS_VM_GET_PAGE_PROT
2022-02-28 10:47 [OpenRISC] [PATCH V3 00/30] mm/mmap: Drop protection_map[] and platform's __SXXX/__PXXX requirements Anshuman Khandual
` (20 preceding siblings ...)
2022-02-28 10:47 ` [OpenRISC] [PATCH V3 21/30] parisc/mm: " Anshuman Khandual
@ 2022-02-28 10:47 ` Anshuman Khandual
2022-02-28 10:47 ` [OpenRISC] [PATCH V3 23/30] um/mm: " Anshuman Khandual
` (7 subsequent siblings)
29 siblings, 0 replies; 49+ messages in thread
From: Anshuman Khandual @ 2022-02-28 10:47 UTC (permalink / raw)
To: openrisc
This defines and exports a platform specific custom vm_get_page_prot() via
subscribing ARCH_HAS_VM_GET_PAGE_PROT. Subsequently all __SXXX and __PXXX
macros can be dropped which are no longer needed.
Cc: Jonas Bonn <jonas@southpole.se>
Cc: openrisc at lists.librecores.org
Cc: linux-kernel at vger.kernel.org
Acked-by: Stafford Horne <shorne@gmail.com>
Signed-off-by: Anshuman Khandual <anshuman.khandual@arm.com>
---
arch/openrisc/Kconfig | 1 +
arch/openrisc/include/asm/pgtable.h | 18 -------------
arch/openrisc/mm/init.c | 41 +++++++++++++++++++++++++++++
3 files changed, 42 insertions(+), 18 deletions(-)
diff --git a/arch/openrisc/Kconfig b/arch/openrisc/Kconfig
index f724b3f1aeed..842a61426816 100644
--- a/arch/openrisc/Kconfig
+++ b/arch/openrisc/Kconfig
@@ -10,6 +10,7 @@ config OPENRISC
select ARCH_HAS_DMA_SET_UNCACHED
select ARCH_HAS_DMA_CLEAR_UNCACHED
select ARCH_HAS_SYNC_DMA_FOR_DEVICE
+ select ARCH_HAS_VM_GET_PAGE_PROT
select COMMON_CLK
select OF
select OF_EARLY_FLATTREE
diff --git a/arch/openrisc/include/asm/pgtable.h b/arch/openrisc/include/asm/pgtable.h
index cdd657f80bfa..fe686c4b7065 100644
--- a/arch/openrisc/include/asm/pgtable.h
+++ b/arch/openrisc/include/asm/pgtable.h
@@ -176,24 +176,6 @@ extern void paging_init(void);
__pgprot(_PAGE_ALL | _PAGE_SRE | _PAGE_SWE \
| _PAGE_SHARED | _PAGE_DIRTY | _PAGE_EXEC | _PAGE_CI)
-#define __P000 PAGE_NONE
-#define __P001 PAGE_READONLY_X
-#define __P010 PAGE_COPY
-#define __P011 PAGE_COPY_X
-#define __P100 PAGE_READONLY
-#define __P101 PAGE_READONLY_X
-#define __P110 PAGE_COPY
-#define __P111 PAGE_COPY_X
-
-#define __S000 PAGE_NONE
-#define __S001 PAGE_READONLY_X
-#define __S010 PAGE_SHARED
-#define __S011 PAGE_SHARED_X
-#define __S100 PAGE_READONLY
-#define __S101 PAGE_READONLY_X
-#define __S110 PAGE_SHARED
-#define __S111 PAGE_SHARED_X
-
/* zero page used for uninitialized stuff */
extern unsigned long empty_zero_page[2048];
#define ZERO_PAGE(vaddr) (virt_to_page(empty_zero_page))
diff --git a/arch/openrisc/mm/init.c b/arch/openrisc/mm/init.c
index 97305bde1b16..0d06e3ebef5d 100644
--- a/arch/openrisc/mm/init.c
+++ b/arch/openrisc/mm/init.c
@@ -210,3 +210,44 @@ void __init mem_init(void)
mem_init_done = 1;
return;
}
+
+pgprot_t vm_get_page_prot(unsigned long vm_flags)
+{
+ switch (vm_flags & (VM_READ | VM_WRITE | VM_EXEC | VM_SHARED)) {
+ case VM_NONE:
+ return PAGE_NONE;
+ case VM_READ:
+ return PAGE_READONLY_X;
+ case VM_WRITE:
+ return PAGE_COPY;
+ case VM_WRITE | VM_READ:
+ return PAGE_COPY_X;
+ case VM_EXEC:
+ return PAGE_READONLY;
+ case VM_EXEC | VM_READ:
+ return PAGE_READONLY_X;
+ case VM_EXEC | VM_WRITE:
+ return PAGE_COPY;
+ case VM_EXEC | VM_WRITE | VM_READ:
+ return PAGE_COPY_X;
+ case VM_SHARED:
+ return PAGE_NONE;
+ case VM_SHARED | VM_READ:
+ return PAGE_READONLY_X;
+ case VM_SHARED | VM_WRITE:
+ return PAGE_SHARED;
+ case VM_SHARED | VM_WRITE | VM_READ:
+ return PAGE_SHARED_X;
+ case VM_SHARED | VM_EXEC:
+ return PAGE_READONLY;
+ case VM_SHARED | VM_EXEC | VM_READ:
+ return PAGE_READONLY_X;
+ case VM_SHARED | VM_EXEC | VM_WRITE:
+ return PAGE_SHARED;
+ case VM_SHARED | VM_EXEC | VM_WRITE | VM_READ:
+ return PAGE_SHARED_X;
+ default:
+ BUILD_BUG();
+ }
+}
+EXPORT_SYMBOL(vm_get_page_prot);
--
2.25.1
^ permalink raw reply related [flat|nested] 49+ messages in thread
* [OpenRISC] [PATCH V3 23/30] um/mm: Enable ARCH_HAS_VM_GET_PAGE_PROT
2022-02-28 10:47 [OpenRISC] [PATCH V3 00/30] mm/mmap: Drop protection_map[] and platform's __SXXX/__PXXX requirements Anshuman Khandual
` (21 preceding siblings ...)
2022-02-28 10:47 ` [OpenRISC] [PATCH V3 22/30] openrisc/mm: " Anshuman Khandual
@ 2022-02-28 10:47 ` Anshuman Khandual
2022-02-28 10:47 ` [OpenRISC] [PATCH V3 24/30] microblaze/mm: " Anshuman Khandual
` (6 subsequent siblings)
29 siblings, 0 replies; 49+ messages in thread
From: Anshuman Khandual @ 2022-02-28 10:47 UTC (permalink / raw)
To: openrisc
This defines and exports a platform specific custom vm_get_page_prot() via
subscribing ARCH_HAS_VM_GET_PAGE_PROT. Subsequently all __SXXX and __PXXX
macros can be dropped which are no longer needed.
Cc: Jeff Dike <jdike@addtoit.com>
Cc: linux-um at lists.infradead.org
Cc: linux-kernel at vger.kernel.org
Signed-off-by: Anshuman Khandual <anshuman.khandual@arm.com>
---
arch/um/Kconfig | 1 +
arch/um/include/asm/pgtable.h | 17 -----------------
arch/um/kernel/mem.c | 35 +++++++++++++++++++++++++++++++++++
arch/x86/um/mem_32.c | 2 +-
4 files changed, 37 insertions(+), 18 deletions(-)
diff --git a/arch/um/Kconfig b/arch/um/Kconfig
index 4d398b80aea8..5836296868a8 100644
--- a/arch/um/Kconfig
+++ b/arch/um/Kconfig
@@ -9,6 +9,7 @@ config UML
select ARCH_HAS_KCOV
select ARCH_HAS_STRNCPY_FROM_USER
select ARCH_HAS_STRNLEN_USER
+ select ARCH_HAS_VM_GET_PAGE_PROT
select ARCH_NO_PREEMPT
select HAVE_ARCH_AUDITSYSCALL
select HAVE_ARCH_SECCOMP_FILTER
diff --git a/arch/um/include/asm/pgtable.h b/arch/um/include/asm/pgtable.h
index b9e20bbe2f75..d982622c0708 100644
--- a/arch/um/include/asm/pgtable.h
+++ b/arch/um/include/asm/pgtable.h
@@ -68,23 +68,6 @@ extern unsigned long end_iomem;
* Also, write permissions imply read permissions. This is the closest we can
* get..
*/
-#define __P000 PAGE_NONE
-#define __P001 PAGE_READONLY
-#define __P010 PAGE_COPY
-#define __P011 PAGE_COPY
-#define __P100 PAGE_READONLY
-#define __P101 PAGE_READONLY
-#define __P110 PAGE_COPY
-#define __P111 PAGE_COPY
-
-#define __S000 PAGE_NONE
-#define __S001 PAGE_READONLY
-#define __S010 PAGE_SHARED
-#define __S011 PAGE_SHARED
-#define __S100 PAGE_READONLY
-#define __S101 PAGE_READONLY
-#define __S110 PAGE_SHARED
-#define __S111 PAGE_SHARED
/*
* ZERO_PAGE is a global shared page that is always zero: used
diff --git a/arch/um/kernel/mem.c b/arch/um/kernel/mem.c
index 15295c3237a0..37c6c7b9dadc 100644
--- a/arch/um/kernel/mem.c
+++ b/arch/um/kernel/mem.c
@@ -197,3 +197,38 @@ void *uml_kmalloc(int size, int flags)
{
return kmalloc(size, flags);
}
+
+pgprot_t vm_get_page_prot(unsigned long vm_flags)
+{
+ switch (vm_flags & (VM_READ | VM_WRITE | VM_EXEC | VM_SHARED)) {
+ case VM_NONE:
+ return PAGE_NONE;
+ case VM_READ:
+ return PAGE_READONLY;
+ case VM_WRITE:
+ case VM_WRITE | VM_READ:
+ return PAGE_COPY;
+ case VM_EXEC:
+ case VM_EXEC | VM_READ:
+ return PAGE_READONLY;
+ case VM_EXEC | VM_WRITE:
+ case VM_EXEC | VM_WRITE | VM_READ:
+ return PAGE_COPY;
+ case VM_SHARED:
+ return PAGE_NONE;
+ case VM_SHARED | VM_READ:
+ return PAGE_READONLY;
+ case VM_SHARED | VM_WRITE:
+ case VM_SHARED | VM_WRITE | VM_READ:
+ return PAGE_SHARED;
+ case VM_SHARED | VM_EXEC:
+ case VM_SHARED | VM_EXEC | VM_READ:
+ return PAGE_READONLY;
+ case VM_SHARED | VM_EXEC | VM_WRITE:
+ case VM_SHARED | VM_EXEC | VM_WRITE | VM_READ:
+ return PAGE_SHARED;
+ default:
+ BUILD_BUG();
+ }
+}
+EXPORT_SYMBOL(vm_get_page_prot);
diff --git a/arch/x86/um/mem_32.c b/arch/x86/um/mem_32.c
index 19c5dbd46770..cafd01f730da 100644
--- a/arch/x86/um/mem_32.c
+++ b/arch/x86/um/mem_32.c
@@ -17,7 +17,7 @@ static int __init gate_vma_init(void)
gate_vma.vm_start = FIXADDR_USER_START;
gate_vma.vm_end = FIXADDR_USER_END;
gate_vma.vm_flags = VM_READ | VM_MAYREAD | VM_EXEC | VM_MAYEXEC;
- gate_vma.vm_page_prot = __P101;
+ gate_vma.vm_page_prot = PAGE_READONLY;
return 0;
}
--
2.25.1
^ permalink raw reply related [flat|nested] 49+ messages in thread
* [OpenRISC] [PATCH V3 24/30] microblaze/mm: Enable ARCH_HAS_VM_GET_PAGE_PROT
2022-02-28 10:47 [OpenRISC] [PATCH V3 00/30] mm/mmap: Drop protection_map[] and platform's __SXXX/__PXXX requirements Anshuman Khandual
` (22 preceding siblings ...)
2022-02-28 10:47 ` [OpenRISC] [PATCH V3 23/30] um/mm: " Anshuman Khandual
@ 2022-02-28 10:47 ` Anshuman Khandual
2022-02-28 10:47 ` [OpenRISC] [PATCH V3 25/30] nios2/mm: " Anshuman Khandual
` (5 subsequent siblings)
29 siblings, 0 replies; 49+ messages in thread
From: Anshuman Khandual @ 2022-02-28 10:47 UTC (permalink / raw)
To: openrisc
This defines and exports a platform specific custom vm_get_page_prot() via
subscribing ARCH_HAS_VM_GET_PAGE_PROT. Subsequently all __SXXX and __PXXX
macros can be dropped which are no longer needed.
Cc: Michal Simek <monstr@monstr.eu>
Cc: linux-kernel at vger.kernel.org
Signed-off-by: Anshuman Khandual <anshuman.khandual@arm.com>
---
arch/microblaze/Kconfig | 1 +
arch/microblaze/include/asm/pgtable.h | 17 -----------
arch/microblaze/mm/init.c | 41 +++++++++++++++++++++++++++
3 files changed, 42 insertions(+), 17 deletions(-)
diff --git a/arch/microblaze/Kconfig b/arch/microblaze/Kconfig
index 59798e43cdb0..f2c25ba8621e 100644
--- a/arch/microblaze/Kconfig
+++ b/arch/microblaze/Kconfig
@@ -7,6 +7,7 @@ config MICROBLAZE
select ARCH_HAS_GCOV_PROFILE_ALL
select ARCH_HAS_SYNC_DMA_FOR_CPU
select ARCH_HAS_SYNC_DMA_FOR_DEVICE
+ select ARCH_HAS_VM_GET_PAGE_PROT
select ARCH_MIGHT_HAVE_PC_PARPORT
select ARCH_WANT_IPC_PARSE_VERSION
select BUILDTIME_TABLE_SORT
diff --git a/arch/microblaze/include/asm/pgtable.h b/arch/microblaze/include/asm/pgtable.h
index c136a01e467e..6df373077ff2 100644
--- a/arch/microblaze/include/asm/pgtable.h
+++ b/arch/microblaze/include/asm/pgtable.h
@@ -204,23 +204,6 @@ extern pte_t *va_to_pte(unsigned long address);
* We consider execute permission the same as read.
* Also, write permissions imply read permissions.
*/
-#define __P000 PAGE_NONE
-#define __P001 PAGE_READONLY_X
-#define __P010 PAGE_COPY
-#define __P011 PAGE_COPY_X
-#define __P100 PAGE_READONLY
-#define __P101 PAGE_READONLY_X
-#define __P110 PAGE_COPY
-#define __P111 PAGE_COPY_X
-
-#define __S000 PAGE_NONE
-#define __S001 PAGE_READONLY_X
-#define __S010 PAGE_SHARED
-#define __S011 PAGE_SHARED_X
-#define __S100 PAGE_READONLY
-#define __S101 PAGE_READONLY_X
-#define __S110 PAGE_SHARED
-#define __S111 PAGE_SHARED_X
#ifndef __ASSEMBLY__
/*
diff --git a/arch/microblaze/mm/init.c b/arch/microblaze/mm/init.c
index 952f35b335b2..68faf7d04faf 100644
--- a/arch/microblaze/mm/init.c
+++ b/arch/microblaze/mm/init.c
@@ -280,3 +280,44 @@ void * __ref zalloc_maybe_bootmem(size_t size, gfp_t mask)
return p;
}
+
+pgprot_t vm_get_page_prot(unsigned long vm_flags)
+{
+ switch (vm_flags & (VM_READ | VM_WRITE | VM_EXEC | VM_SHARED)) {
+ case VM_NONE:
+ return PAGE_NONE;
+ case VM_READ:
+ return PAGE_READONLY_X;
+ case VM_WRITE:
+ return PAGE_COPY;
+ case VM_WRITE | VM_READ:
+ return PAGE_COPY_X;
+ case VM_EXEC:
+ return PAGE_READONLY;
+ case VM_EXEC | VM_READ:
+ return PAGE_READONLY_X;
+ case VM_EXEC | VM_WRITE:
+ return PAGE_COPY;
+ case VM_EXEC | VM_WRITE | VM_READ:
+ return PAGE_COPY_X;
+ case VM_SHARED:
+ return PAGE_NONE;
+ case VM_SHARED | VM_READ:
+ return PAGE_READONLY_X;
+ case VM_SHARED | VM_WRITE:
+ return PAGE_SHARED;
+ case VM_SHARED | VM_WRITE | VM_READ:
+ return PAGE_SHARED_X;
+ case VM_SHARED | VM_EXEC:
+ return PAGE_READONLY;
+ case VM_SHARED | VM_EXEC | VM_READ:
+ return PAGE_READONLY_X;
+ case VM_SHARED | VM_EXEC | VM_WRITE:
+ return PAGE_SHARED;
+ case VM_SHARED | VM_EXEC | VM_WRITE | VM_READ:
+ return PAGE_SHARED_X;
+ default:
+ BUILD_BUG();
+ }
+}
+EXPORT_SYMBOL(vm_get_page_prot);
--
2.25.1
^ permalink raw reply related [flat|nested] 49+ messages in thread
* [OpenRISC] [PATCH V3 25/30] nios2/mm: Enable ARCH_HAS_VM_GET_PAGE_PROT
2022-02-28 10:47 [OpenRISC] [PATCH V3 00/30] mm/mmap: Drop protection_map[] and platform's __SXXX/__PXXX requirements Anshuman Khandual
` (23 preceding siblings ...)
2022-02-28 10:47 ` [OpenRISC] [PATCH V3 24/30] microblaze/mm: " Anshuman Khandual
@ 2022-02-28 10:47 ` Anshuman Khandual
2022-02-28 10:47 ` [OpenRISC] [PATCH V3 26/30] hexagon/mm: " Anshuman Khandual
` (4 subsequent siblings)
29 siblings, 0 replies; 49+ messages in thread
From: Anshuman Khandual @ 2022-02-28 10:47 UTC (permalink / raw)
To: openrisc
This defines and exports a platform specific custom vm_get_page_prot() via
subscribing ARCH_HAS_VM_GET_PAGE_PROT. Subsequently all __SXXX and __PXXX
macros can be dropped which are no longer needed.
Cc: Dinh Nguyen <dinguyen@kernel.org>
Cc: linux-kernel at vger.kernel.org
Acked-by: Dinh Nguyen <dinguyen@kernel.org>
Signed-off-by: Anshuman Khandual <anshuman.khandual@arm.com>
---
arch/nios2/Kconfig | 1 +
arch/nios2/include/asm/pgtable.h | 24 ----------------
arch/nios2/mm/init.c | 47 ++++++++++++++++++++++++++++++++
3 files changed, 48 insertions(+), 24 deletions(-)
diff --git a/arch/nios2/Kconfig b/arch/nios2/Kconfig
index 33fd06f5fa41..85a58a357a3b 100644
--- a/arch/nios2/Kconfig
+++ b/arch/nios2/Kconfig
@@ -6,6 +6,7 @@ config NIOS2
select ARCH_HAS_SYNC_DMA_FOR_CPU
select ARCH_HAS_SYNC_DMA_FOR_DEVICE
select ARCH_HAS_DMA_SET_UNCACHED
+ select ARCH_HAS_VM_GET_PAGE_PROT
select ARCH_NO_SWAP
select COMMON_CLK
select TIMER_OF
diff --git a/arch/nios2/include/asm/pgtable.h b/arch/nios2/include/asm/pgtable.h
index 4a995fa628ee..ba3f9822c0b3 100644
--- a/arch/nios2/include/asm/pgtable.h
+++ b/arch/nios2/include/asm/pgtable.h
@@ -34,30 +34,6 @@ struct mm_struct;
((x) ? _PAGE_EXEC : 0) | \
((r) ? _PAGE_READ : 0) | \
((w) ? _PAGE_WRITE : 0))
-/*
- * These are the macros that generic kernel code needs
- * (to populate protection_map[])
- */
-
-/* Remove W bit on private pages for COW support */
-#define __P000 MKP(0, 0, 0)
-#define __P001 MKP(0, 0, 1)
-#define __P010 MKP(0, 0, 0) /* COW */
-#define __P011 MKP(0, 0, 1) /* COW */
-#define __P100 MKP(1, 0, 0)
-#define __P101 MKP(1, 0, 1)
-#define __P110 MKP(1, 0, 0) /* COW */
-#define __P111 MKP(1, 0, 1) /* COW */
-
-/* Shared pages can have exact HW mapping */
-#define __S000 MKP(0, 0, 0)
-#define __S001 MKP(0, 0, 1)
-#define __S010 MKP(0, 1, 0)
-#define __S011 MKP(0, 1, 1)
-#define __S100 MKP(1, 0, 0)
-#define __S101 MKP(1, 0, 1)
-#define __S110 MKP(1, 1, 0)
-#define __S111 MKP(1, 1, 1)
/* Used all over the kernel */
#define PAGE_KERNEL __pgprot(_PAGE_PRESENT | _PAGE_CACHED | _PAGE_READ | \
diff --git a/arch/nios2/mm/init.c b/arch/nios2/mm/init.c
index 613fcaa5988a..e867f5d85580 100644
--- a/arch/nios2/mm/init.c
+++ b/arch/nios2/mm/init.c
@@ -124,3 +124,50 @@ const char *arch_vma_name(struct vm_area_struct *vma)
{
return (vma->vm_start == KUSER_BASE) ? "[kuser]" : NULL;
}
+
+pgprot_t vm_get_page_prot(unsigned long vm_flags)
+{
+ switch (vm_flags & (VM_READ | VM_WRITE | VM_EXEC | VM_SHARED)) {
+ /* Remove W bit on private pages for COW support */
+ case VM_NONE:
+ return MKP(0, 0, 0);
+ case VM_READ:
+ return MKP(0, 0, 1);
+ /* COW */
+ case VM_WRITE:
+ return MKP(0, 0, 0);
+ /* COW */
+ case VM_WRITE | VM_READ:
+ return MKP(0, 0, 1);
+ case VM_EXEC:
+ return MKP(1, 0, 0);
+ case VM_EXEC | VM_READ:
+ return MKP(1, 0, 1);
+ /* COW */
+ case VM_EXEC | VM_WRITE:
+ return MKP(1, 0, 0);
+ /* COW */
+ case VM_EXEC | VM_WRITE | VM_READ:
+ return MKP(1, 0, 1);
+ /* Shared pages can have exact HW mapping */
+ case VM_SHARED:
+ return MKP(0, 0, 0);
+ case VM_SHARED | VM_READ:
+ return MKP(0, 0, 1);
+ case VM_SHARED | VM_WRITE:
+ return MKP(0, 1, 0);
+ case VM_SHARED | VM_WRITE | VM_READ:
+ return MKP(0, 1, 1);
+ case VM_SHARED | VM_EXEC:
+ return MKP(1, 0, 0);
+ case VM_SHARED | VM_EXEC | VM_READ:
+ return MKP(1, 0, 1);
+ case VM_SHARED | VM_EXEC | VM_WRITE:
+ return MKP(1, 1, 0);
+ case VM_SHARED | VM_EXEC | VM_WRITE | VM_READ:
+ return MKP(1, 1, 1);
+ default:
+ BUILD_BUG();
+ }
+}
+EXPORT_SYMBOL(vm_get_page_prot);
--
2.25.1
^ permalink raw reply related [flat|nested] 49+ messages in thread
* [OpenRISC] [PATCH V3 26/30] hexagon/mm: Enable ARCH_HAS_VM_GET_PAGE_PROT
2022-02-28 10:47 [OpenRISC] [PATCH V3 00/30] mm/mmap: Drop protection_map[] and platform's __SXXX/__PXXX requirements Anshuman Khandual
` (24 preceding siblings ...)
2022-02-28 10:47 ` [OpenRISC] [PATCH V3 25/30] nios2/mm: " Anshuman Khandual
@ 2022-02-28 10:47 ` Anshuman Khandual
2022-02-28 10:47 ` [OpenRISC] [PATCH V3 27/30] nds32/mm: " Anshuman Khandual
` (3 subsequent siblings)
29 siblings, 0 replies; 49+ messages in thread
From: Anshuman Khandual @ 2022-02-28 10:47 UTC (permalink / raw)
To: openrisc
This defines and exports a platform specific custom vm_get_page_prot() via
subscribing ARCH_HAS_VM_GET_PAGE_PROT. Subsequently all __SXXX and __PXXX
macros can be dropped which are no longer needed.
Cc: Brian Cain <bcain@codeaurora.org>
Cc: linux-hexagon at vger.kernel.org
Cc: linux-kernel at vger.kernel.org
Signed-off-by: Anshuman Khandual <anshuman.khandual@arm.com>
---
arch/hexagon/Kconfig | 1 +
arch/hexagon/include/asm/pgtable.h | 24 -----------
arch/hexagon/mm/init.c | 67 ++++++++++++++++++++++++++++++
3 files changed, 68 insertions(+), 24 deletions(-)
diff --git a/arch/hexagon/Kconfig b/arch/hexagon/Kconfig
index 15dd8f38b698..cdc5df32a1e3 100644
--- a/arch/hexagon/Kconfig
+++ b/arch/hexagon/Kconfig
@@ -6,6 +6,7 @@ config HEXAGON
def_bool y
select ARCH_32BIT_OFF_T
select ARCH_HAS_SYNC_DMA_FOR_DEVICE
+ select ARCH_HAS_VM_GET_PAGE_PROT
select ARCH_NO_PREEMPT
select DMA_GLOBAL_POOL
# Other pending projects/to-do items.
diff --git a/arch/hexagon/include/asm/pgtable.h b/arch/hexagon/include/asm/pgtable.h
index 18cd6ea9ab23..5eceddfe013d 100644
--- a/arch/hexagon/include/asm/pgtable.h
+++ b/arch/hexagon/include/asm/pgtable.h
@@ -127,31 +127,7 @@ extern unsigned long _dflt_cache_att;
#define CACHEDEF (CACHE_DEFAULT << 6)
/* Private (copy-on-write) page protections. */
-#define __P000 __pgprot(_PAGE_PRESENT | _PAGE_USER | CACHEDEF)
-#define __P001 __pgprot(_PAGE_PRESENT | _PAGE_USER | _PAGE_READ | CACHEDEF)
-#define __P010 __P000 /* Write-only copy-on-write */
-#define __P011 __P001 /* Read/Write copy-on-write */
-#define __P100 __pgprot(_PAGE_PRESENT | _PAGE_USER | \
- _PAGE_EXECUTE | CACHEDEF)
-#define __P101 __pgprot(_PAGE_PRESENT | _PAGE_USER | _PAGE_EXECUTE | \
- _PAGE_READ | CACHEDEF)
-#define __P110 __P100 /* Write/execute copy-on-write */
-#define __P111 __P101 /* Read/Write/Execute, copy-on-write */
-
/* Shared page protections. */
-#define __S000 __P000
-#define __S001 __P001
-#define __S010 __pgprot(_PAGE_PRESENT | _PAGE_USER | \
- _PAGE_WRITE | CACHEDEF)
-#define __S011 __pgprot(_PAGE_PRESENT | _PAGE_USER | _PAGE_READ | \
- _PAGE_WRITE | CACHEDEF)
-#define __S100 __pgprot(_PAGE_PRESENT | _PAGE_USER | \
- _PAGE_EXECUTE | CACHEDEF)
-#define __S101 __P101
-#define __S110 __pgprot(_PAGE_PRESENT | _PAGE_USER | \
- _PAGE_EXECUTE | _PAGE_WRITE | CACHEDEF)
-#define __S111 __pgprot(_PAGE_PRESENT | _PAGE_USER | _PAGE_READ | \
- _PAGE_EXECUTE | _PAGE_WRITE | CACHEDEF)
extern pgd_t swapper_pg_dir[PTRS_PER_PGD]; /* located in head.S */
diff --git a/arch/hexagon/mm/init.c b/arch/hexagon/mm/init.c
index f01e91e10d95..b53595fc4103 100644
--- a/arch/hexagon/mm/init.c
+++ b/arch/hexagon/mm/init.c
@@ -236,3 +236,70 @@ void __init setup_arch_memory(void)
* which is called by start_kernel() later on in the process
*/
}
+
+pgprot_t vm_get_page_prot(unsigned long vm_flags)
+{
+ switch (vm_flags & (VM_READ | VM_WRITE | VM_EXEC | VM_SHARED)) {
+ case VM_NONE:
+ return __pgprot(_PAGE_PRESENT | _PAGE_USER |
+ CACHEDEF);
+ case VM_READ:
+ return __pgprot(_PAGE_PRESENT | _PAGE_USER |
+ _PAGE_READ | CACHEDEF);
+ /* Write-only copy-on-write */
+ case VM_WRITE:
+ return __pgprot(_PAGE_PRESENT | _PAGE_USER |
+ CACHEDEF);
+ /* Read/Write copy-on-write */
+ case VM_WRITE | VM_READ:
+ return __pgprot(_PAGE_PRESENT | _PAGE_USER |
+ _PAGE_READ | CACHEDEF);
+ case VM_EXEC:
+ return __pgprot(_PAGE_PRESENT | _PAGE_USER |
+ _PAGE_EXECUTE | CACHEDEF);
+ case VM_EXEC | VM_READ:
+ return __pgprot(_PAGE_PRESENT | _PAGE_USER |
+ _PAGE_EXECUTE | _PAGE_READ |
+ CACHEDEF);
+ /* Write/execute copy-on-write */
+ case VM_EXEC | VM_WRITE:
+ return __pgprot(_PAGE_PRESENT | _PAGE_USER |
+ _PAGE_EXECUTE | CACHEDEF);
+ /* Read/Write/Execute, copy-on-write */
+ case VM_EXEC | VM_WRITE | VM_READ:
+ return __pgprot(_PAGE_PRESENT | _PAGE_USER |
+ _PAGE_EXECUTE | _PAGE_READ |
+ CACHEDEF);
+ case VM_SHARED:
+ return __pgprot(_PAGE_PRESENT | _PAGE_USER |
+ CACHEDEF);
+ case VM_SHARED | VM_READ:
+ return __pgprot(_PAGE_PRESENT | _PAGE_USER |
+ _PAGE_READ | CACHEDEF);
+ case VM_SHARED | VM_WRITE:
+ return __pgprot(_PAGE_PRESENT | _PAGE_USER |
+ _PAGE_WRITE | CACHEDEF);
+ case VM_SHARED | VM_WRITE | VM_READ:
+ return __pgprot(_PAGE_PRESENT | _PAGE_USER |
+ _PAGE_READ | _PAGE_WRITE |
+ CACHEDEF);
+ case VM_SHARED | VM_EXEC:
+ return __pgprot(_PAGE_PRESENT | _PAGE_USER |
+ _PAGE_EXECUTE | CACHEDEF);
+ case VM_SHARED | VM_EXEC | VM_READ:
+ return __pgprot(_PAGE_PRESENT | _PAGE_USER |
+ _PAGE_EXECUTE | _PAGE_READ |
+ CACHEDEF);
+ case VM_SHARED | VM_EXEC | VM_WRITE:
+ return __pgprot(_PAGE_PRESENT | _PAGE_USER |
+ _PAGE_EXECUTE | _PAGE_WRITE |
+ CACHEDEF);
+ case VM_SHARED | VM_EXEC | VM_WRITE | VM_READ:
+ return __pgprot(_PAGE_PRESENT | _PAGE_USER |
+ _PAGE_READ | _PAGE_EXECUTE |
+ _PAGE_WRITE | CACHEDEF);
+ default:
+ BUILD_BUG();
+ }
+}
+EXPORT_SYMBOL(vm_get_page_prot);
--
2.25.1
^ permalink raw reply related [flat|nested] 49+ messages in thread
* [OpenRISC] [PATCH V3 27/30] nds32/mm: Enable ARCH_HAS_VM_GET_PAGE_PROT
2022-02-28 10:47 [OpenRISC] [PATCH V3 00/30] mm/mmap: Drop protection_map[] and platform's __SXXX/__PXXX requirements Anshuman Khandual
` (25 preceding siblings ...)
2022-02-28 10:47 ` [OpenRISC] [PATCH V3 26/30] hexagon/mm: " Anshuman Khandual
@ 2022-02-28 10:47 ` Anshuman Khandual
2022-02-28 10:47 ` [OpenRISC] [PATCH V3 28/30] ia64/mm: " Anshuman Khandual
` (2 subsequent siblings)
29 siblings, 0 replies; 49+ messages in thread
From: Anshuman Khandual @ 2022-02-28 10:47 UTC (permalink / raw)
To: openrisc
This defines and exports a platform specific custom vm_get_page_prot() via
subscribing ARCH_HAS_VM_GET_PAGE_PROT. Subsequently all __SXXX and __PXXX
macros can be dropped which are no longer needed.
Cc: Nick Hu <nickhu@andestech.com>
Cc: linux-kernel at vger.kernel.org
Signed-off-by: Anshuman Khandual <anshuman.khandual@arm.com>
---
arch/nds32/Kconfig | 1 +
arch/nds32/include/asm/pgtable.h | 17 ---------------
arch/nds32/mm/mmap.c | 37 ++++++++++++++++++++++++++++++++
3 files changed, 38 insertions(+), 17 deletions(-)
diff --git a/arch/nds32/Kconfig b/arch/nds32/Kconfig
index 4d1421b18734..576e05479925 100644
--- a/arch/nds32/Kconfig
+++ b/arch/nds32/Kconfig
@@ -10,6 +10,7 @@ config NDS32
select ARCH_HAS_DMA_PREP_COHERENT
select ARCH_HAS_SYNC_DMA_FOR_CPU
select ARCH_HAS_SYNC_DMA_FOR_DEVICE
+ select ARCH_HAS_VM_GET_PAGE_PROT
select ARCH_WANT_FRAME_POINTERS if FTRACE
select CLKSRC_MMIO
select CLONE_BACKWARDS
diff --git a/arch/nds32/include/asm/pgtable.h b/arch/nds32/include/asm/pgtable.h
index 419f984eef70..79f64ed734cb 100644
--- a/arch/nds32/include/asm/pgtable.h
+++ b/arch/nds32/include/asm/pgtable.h
@@ -152,23 +152,6 @@ extern void __pgd_error(const char *file, int line, unsigned long val);
#endif /* __ASSEMBLY__ */
/* xwr */
-#define __P000 (PAGE_NONE | _PAGE_CACHE_SHRD)
-#define __P001 (PAGE_READ | _PAGE_CACHE_SHRD)
-#define __P010 (PAGE_COPY | _PAGE_CACHE_SHRD)
-#define __P011 (PAGE_COPY | _PAGE_CACHE_SHRD)
-#define __P100 (PAGE_EXEC | _PAGE_CACHE_SHRD)
-#define __P101 (PAGE_READ | _PAGE_E | _PAGE_CACHE_SHRD)
-#define __P110 (PAGE_COPY | _PAGE_E | _PAGE_CACHE_SHRD)
-#define __P111 (PAGE_COPY | _PAGE_E | _PAGE_CACHE_SHRD)
-
-#define __S000 (PAGE_NONE | _PAGE_CACHE_SHRD)
-#define __S001 (PAGE_READ | _PAGE_CACHE_SHRD)
-#define __S010 (PAGE_RDWR | _PAGE_CACHE_SHRD)
-#define __S011 (PAGE_RDWR | _PAGE_CACHE_SHRD)
-#define __S100 (PAGE_EXEC | _PAGE_CACHE_SHRD)
-#define __S101 (PAGE_READ | _PAGE_E | _PAGE_CACHE_SHRD)
-#define __S110 (PAGE_RDWR | _PAGE_E | _PAGE_CACHE_SHRD)
-#define __S111 (PAGE_RDWR | _PAGE_E | _PAGE_CACHE_SHRD)
#ifndef __ASSEMBLY__
/*
diff --git a/arch/nds32/mm/mmap.c b/arch/nds32/mm/mmap.c
index 1bdf5e7d1b43..0399b928948d 100644
--- a/arch/nds32/mm/mmap.c
+++ b/arch/nds32/mm/mmap.c
@@ -71,3 +71,40 @@ arch_get_unmapped_area(struct file *filp, unsigned long addr,
info.align_offset = pgoff << PAGE_SHIFT;
return vm_unmapped_area(&info);
}
+
+pgprot_t vm_get_page_prot(unsigned long vm_flags)
+{
+ switch (vm_flags & (VM_READ | VM_WRITE | VM_EXEC | VM_SHARED)) {
+ case VM_NONE:
+ return (PAGE_NONE | _PAGE_CACHE_SHRD);
+ case VM_READ:
+ return (PAGE_READ | _PAGE_CACHE_SHRD);
+ case VM_WRITE:
+ case VM_WRITE | VM_READ:
+ return (PAGE_COPY | _PAGE_CACHE_SHRD);
+ case VM_EXEC:
+ return (PAGE_EXEC | _PAGE_CACHE_SHRD);
+ case VM_EXEC | VM_READ:
+ return (PAGE_READ | _PAGE_E | _PAGE_CACHE_SHRD);
+ case VM_EXEC | VM_WRITE:
+ case VM_EXEC | VM_WRITE | VM_READ:
+ return (PAGE_COPY | _PAGE_E | _PAGE_CACHE_SHRD);
+ case VM_SHARED:
+ return (PAGE_NONE | _PAGE_CACHE_SHRD);
+ case VM_SHARED | VM_READ:
+ return (PAGE_READ | _PAGE_CACHE_SHRD);
+ case VM_SHARED | VM_WRITE:
+ case VM_SHARED | VM_WRITE | VM_READ:
+ return (PAGE_RDWR | _PAGE_CACHE_SHRD);
+ case VM_SHARED | VM_EXEC:
+ return (PAGE_EXEC | _PAGE_CACHE_SHRD);
+ case VM_SHARED | VM_EXEC | VM_READ:
+ return (PAGE_READ | _PAGE_E | _PAGE_CACHE_SHRD);
+ case VM_SHARED | VM_EXEC | VM_WRITE:
+ case VM_SHARED | VM_EXEC | VM_WRITE | VM_READ:
+ return (PAGE_RDWR | _PAGE_E | _PAGE_CACHE_SHRD);
+ default:
+ BUILD_BUG();
+ }
+}
+EXPORT_SYMBOL(vm_get_page_prot);
--
2.25.1
^ permalink raw reply related [flat|nested] 49+ messages in thread
* [OpenRISC] [PATCH V3 28/30] ia64/mm: Enable ARCH_HAS_VM_GET_PAGE_PROT
2022-02-28 10:47 [OpenRISC] [PATCH V3 00/30] mm/mmap: Drop protection_map[] and platform's __SXXX/__PXXX requirements Anshuman Khandual
` (26 preceding siblings ...)
2022-02-28 10:47 ` [OpenRISC] [PATCH V3 27/30] nds32/mm: " Anshuman Khandual
@ 2022-02-28 10:47 ` Anshuman Khandual
2022-02-28 10:47 ` [OpenRISC] [PATCH V3 29/30] mm/mmap: Drop generic vm_get_page_prot() Anshuman Khandual
2022-02-28 10:47 ` [OpenRISC] [PATCH V3 30/30] mm/mmap: Drop ARCH_HAS_VM_GET_PAGE_PROT Anshuman Khandual
29 siblings, 0 replies; 49+ messages in thread
From: Anshuman Khandual @ 2022-02-28 10:47 UTC (permalink / raw)
To: openrisc
This defines and exports a platform specific custom vm_get_page_prot() via
subscribing ARCH_HAS_VM_GET_PAGE_PROT. Subsequently all __SXXX and __PXXX
macros can be dropped which are no longer needed.
Cc: linux-ia64 at vger.kernel.org
Cc: linux-kernel at vger.kernel.org
Signed-off-by: Anshuman Khandual <anshuman.khandual@arm.com>
---
arch/ia64/Kconfig | 1 +
arch/ia64/include/asm/pgtable.h | 17 --------------
arch/ia64/mm/init.c | 41 ++++++++++++++++++++++++++++++++-
3 files changed, 41 insertions(+), 18 deletions(-)
diff --git a/arch/ia64/Kconfig b/arch/ia64/Kconfig
index a7e01573abd8..0ab15e8d5783 100644
--- a/arch/ia64/Kconfig
+++ b/arch/ia64/Kconfig
@@ -11,6 +11,7 @@ config IA64
select ARCH_HAS_DMA_MARK_CLEAN
select ARCH_HAS_STRNCPY_FROM_USER
select ARCH_HAS_STRNLEN_USER
+ select ARCH_HAS_VM_GET_PAGE_PROT
select ARCH_MIGHT_HAVE_PC_PARPORT
select ARCH_MIGHT_HAVE_PC_SERIO
select ACPI
diff --git a/arch/ia64/include/asm/pgtable.h b/arch/ia64/include/asm/pgtable.h
index 9584b2c5f394..8154c78bba56 100644
--- a/arch/ia64/include/asm/pgtable.h
+++ b/arch/ia64/include/asm/pgtable.h
@@ -161,23 +161,6 @@
* attempts to write to the page.
*/
/* xwr */
-#define __P000 PAGE_NONE
-#define __P001 PAGE_READONLY
-#define __P010 PAGE_READONLY /* write to priv pg -> copy & make writable */
-#define __P011 PAGE_READONLY /* ditto */
-#define __P100 __pgprot(__ACCESS_BITS | _PAGE_PL_3 | _PAGE_AR_X_RX)
-#define __P101 __pgprot(__ACCESS_BITS | _PAGE_PL_3 | _PAGE_AR_RX)
-#define __P110 PAGE_COPY_EXEC
-#define __P111 PAGE_COPY_EXEC
-
-#define __S000 PAGE_NONE
-#define __S001 PAGE_READONLY
-#define __S010 PAGE_SHARED /* we don't have (and don't need) write-only */
-#define __S011 PAGE_SHARED
-#define __S100 __pgprot(__ACCESS_BITS | _PAGE_PL_3 | _PAGE_AR_X_RX)
-#define __S101 __pgprot(__ACCESS_BITS | _PAGE_PL_3 | _PAGE_AR_RX)
-#define __S110 __pgprot(__ACCESS_BITS | _PAGE_PL_3 | _PAGE_AR_RWX)
-#define __S111 __pgprot(__ACCESS_BITS | _PAGE_PL_3 | _PAGE_AR_RWX)
#define pgd_ERROR(e) printk("%s:%d: bad pgd %016lx.\n", __FILE__, __LINE__, pgd_val(e))
#if CONFIG_PGTABLE_LEVELS == 4
diff --git a/arch/ia64/mm/init.c b/arch/ia64/mm/init.c
index 5d165607bf35..2a922883e30f 100644
--- a/arch/ia64/mm/init.c
+++ b/arch/ia64/mm/init.c
@@ -273,7 +273,7 @@ static int __init gate_vma_init(void)
gate_vma.vm_start = FIXADDR_USER_START;
gate_vma.vm_end = FIXADDR_USER_END;
gate_vma.vm_flags = VM_READ | VM_MAYREAD | VM_EXEC | VM_MAYEXEC;
- gate_vma.vm_page_prot = __P101;
+ gate_vma.vm_page_prot = __pgprot(__ACCESS_BITS | _PAGE_PL_3 | _PAGE_AR_RX);
return 0;
}
@@ -492,3 +492,42 @@ void arch_remove_memory(u64 start, u64 size, struct vmem_altmap *altmap)
__remove_pages(start_pfn, nr_pages, altmap);
}
#endif
+
+pgprot_t vm_get_page_prot(unsigned long vm_flags)
+{
+ switch (vm_flags & (VM_READ | VM_WRITE | VM_EXEC | VM_SHARED)) {
+ case VM_NONE:
+ return PAGE_NONE;
+ case VM_READ:
+ /* write to priv pg -> copy & make writable */
+ case VM_WRITE:
+ /* write to priv pg -> copy & make writable */
+ case VM_WRITE | VM_READ:
+ return PAGE_READONLY;
+ case VM_EXEC:
+ return __pgprot(__ACCESS_BITS | _PAGE_PL_3 | _PAGE_AR_X_RX);
+ case VM_EXEC | VM_READ:
+ return __pgprot(__ACCESS_BITS | _PAGE_PL_3 | _PAGE_AR_RX);
+ case VM_EXEC | VM_WRITE:
+ case VM_EXEC | VM_WRITE | VM_READ:
+ return PAGE_COPY_EXEC;
+ case VM_SHARED:
+ return PAGE_NONE;
+ case VM_SHARED | VM_READ:
+ return PAGE_READONLY;
+ /* we don't have (and don't need) write-only */
+ case VM_SHARED | VM_WRITE:
+ case VM_SHARED | VM_WRITE | VM_READ:
+ return PAGE_SHARED;
+ case VM_SHARED | VM_EXEC:
+ return __pgprot(__ACCESS_BITS | _PAGE_PL_3 | _PAGE_AR_X_RX);
+ case VM_SHARED | VM_EXEC | VM_READ:
+ return __pgprot(__ACCESS_BITS | _PAGE_PL_3 | _PAGE_AR_RX);
+ case VM_SHARED | VM_EXEC | VM_WRITE:
+ case VM_SHARED | VM_EXEC | VM_WRITE | VM_READ:
+ return __pgprot(__ACCESS_BITS | _PAGE_PL_3 | _PAGE_AR_RWX);
+ default:
+ BUILD_BUG();
+ }
+}
+EXPORT_SYMBOL(vm_get_page_prot);
--
2.25.1
^ permalink raw reply related [flat|nested] 49+ messages in thread
* [OpenRISC] [PATCH V3 29/30] mm/mmap: Drop generic vm_get_page_prot()
2022-02-28 10:47 [OpenRISC] [PATCH V3 00/30] mm/mmap: Drop protection_map[] and platform's __SXXX/__PXXX requirements Anshuman Khandual
` (27 preceding siblings ...)
2022-02-28 10:47 ` [OpenRISC] [PATCH V3 28/30] ia64/mm: " Anshuman Khandual
@ 2022-02-28 10:47 ` Anshuman Khandual
2022-02-28 10:47 ` [OpenRISC] [PATCH V3 30/30] mm/mmap: Drop ARCH_HAS_VM_GET_PAGE_PROT Anshuman Khandual
29 siblings, 0 replies; 49+ messages in thread
From: Anshuman Khandual @ 2022-02-28 10:47 UTC (permalink / raw)
To: openrisc
All available platforms export their own vm_get_page_prot() implementation
via ARCH_HAS_VM_GET_PAGE_PROT. Hence a generic implementation is no longer
needed.
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: linux-mm at kvack.org
Cc: linux-kernel at vger.kernel.org
Signed-off-by: Anshuman Khandual <anshuman.khandual@arm.com>
---
mm/mmap.c | 40 ----------------------------------------
1 file changed, 40 deletions(-)
diff --git a/mm/mmap.c b/mm/mmap.c
index c13dd9c37866..ff343f76a825 100644
--- a/mm/mmap.c
+++ b/mm/mmap.c
@@ -102,46 +102,6 @@ static void unmap_region(struct mm_struct *mm,
* w: (no) no
* x: (yes) yes
*/
-pgprot_t vm_get_page_prot(unsigned long vm_flags)
-{
- switch (vm_flags & (VM_READ | VM_WRITE | VM_EXEC | VM_SHARED)) {
- case VM_NONE:
- return __P000;
- case VM_READ:
- return __P001;
- case VM_WRITE:
- return __P010;
- case VM_READ | VM_WRITE:
- return __P011;
- case VM_EXEC:
- return __P100;
- case VM_EXEC | VM_READ:
- return __P101;
- case VM_EXEC | VM_WRITE:
- return __P110;
- case VM_EXEC | VM_READ | VM_WRITE:
- return __P111;
- case VM_SHARED:
- return __S000;
- case VM_SHARED | VM_READ:
- return __S001;
- case VM_SHARED | VM_WRITE:
- return __S010;
- case VM_SHARED | VM_READ | VM_WRITE:
- return __S011;
- case VM_SHARED | VM_EXEC:
- return __S100;
- case VM_SHARED | VM_EXEC | VM_READ:
- return __S101;
- case VM_SHARED | VM_EXEC | VM_WRITE:
- return __S110;
- case VM_SHARED | VM_EXEC | VM_READ | VM_WRITE:
- return __S111;
- default:
- BUILD_BUG();
- }
-}
-EXPORT_SYMBOL(vm_get_page_prot);
#endif /* CONFIG_ARCH_HAS_VM_GET_PAGE_PROT */
static pgprot_t vm_pgprot_modify(pgprot_t oldprot, unsigned long vm_flags)
--
2.25.1
^ permalink raw reply related [flat|nested] 49+ messages in thread
* [OpenRISC] [PATCH V3 30/30] mm/mmap: Drop ARCH_HAS_VM_GET_PAGE_PROT
2022-02-28 10:47 [OpenRISC] [PATCH V3 00/30] mm/mmap: Drop protection_map[] and platform's __SXXX/__PXXX requirements Anshuman Khandual
` (28 preceding siblings ...)
2022-02-28 10:47 ` [OpenRISC] [PATCH V3 29/30] mm/mmap: Drop generic vm_get_page_prot() Anshuman Khandual
@ 2022-02-28 10:47 ` Anshuman Khandual
29 siblings, 0 replies; 49+ messages in thread
From: Anshuman Khandual @ 2022-02-28 10:47 UTC (permalink / raw)
To: openrisc
All platforms now define their own vm_get_page_prot() and also there is no
generic version left to fallback on. Hence drop ARCH_HAS_GET_PAGE_PROT.
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: linux-mm at kvack.org
Cc: linux-kernel at vger.kernel.org
Signed-off-by: Anshuman Khandual <anshuman.khandual@arm.com>
---
arch/alpha/Kconfig | 1 -
arch/arc/Kconfig | 1 -
arch/arm/Kconfig | 1 -
arch/arm64/Kconfig | 1 -
arch/csky/Kconfig | 1 -
arch/hexagon/Kconfig | 1 -
arch/ia64/Kconfig | 1 -
arch/m68k/Kconfig | 1 -
arch/microblaze/Kconfig | 1 -
arch/mips/Kconfig | 1 -
arch/nds32/Kconfig | 1 -
arch/nios2/Kconfig | 1 -
arch/openrisc/Kconfig | 1 -
arch/parisc/Kconfig | 1 -
arch/powerpc/Kconfig | 1 -
arch/riscv/Kconfig | 1 -
arch/s390/Kconfig | 1 -
arch/sh/Kconfig | 1 -
arch/sparc/Kconfig | 2 --
arch/um/Kconfig | 1 -
arch/x86/Kconfig | 1 -
arch/xtensa/Kconfig | 1 -
mm/Kconfig | 3 ---
mm/mmap.c | 23 -----------------------
24 files changed, 49 deletions(-)
diff --git a/arch/alpha/Kconfig b/arch/alpha/Kconfig
index 73e82fe5c770..4e87783c90ad 100644
--- a/arch/alpha/Kconfig
+++ b/arch/alpha/Kconfig
@@ -2,7 +2,6 @@
config ALPHA
bool
default y
- select ARCH_HAS_VM_GET_PAGE_PROT
select ARCH_32BIT_USTAT_F_TINODE
select ARCH_MIGHT_HAVE_PC_PARPORT
select ARCH_MIGHT_HAVE_PC_SERIO
diff --git a/arch/arc/Kconfig b/arch/arc/Kconfig
index 78ff0644b343..3c2a4753d09b 100644
--- a/arch/arc/Kconfig
+++ b/arch/arc/Kconfig
@@ -13,7 +13,6 @@ config ARC
select ARCH_HAS_SETUP_DMA_OPS
select ARCH_HAS_SYNC_DMA_FOR_CPU
select ARCH_HAS_SYNC_DMA_FOR_DEVICE
- select ARCH_HAS_VM_GET_PAGE_PROT
select ARCH_SUPPORTS_ATOMIC_RMW if ARC_HAS_LLSC
select ARCH_32BIT_OFF_T
select BUILDTIME_TABLE_SORT
diff --git a/arch/arm/Kconfig b/arch/arm/Kconfig
index 87b2e89ef3d6..4c97cb40eebb 100644
--- a/arch/arm/Kconfig
+++ b/arch/arm/Kconfig
@@ -23,7 +23,6 @@ config ARM
select ARCH_HAS_SYNC_DMA_FOR_CPU if SWIOTLB || !MMU
select ARCH_HAS_TEARDOWN_DMA_OPS if MMU
select ARCH_HAS_TICK_BROADCAST if GENERIC_CLOCKEVENTS_BROADCAST
- select ARCH_HAS_VM_GET_PAGE_PROT
select ARCH_HAVE_CUSTOM_GPIO_H
select ARCH_HAVE_NMI_SAFE_CMPXCHG if CPU_V7 || CPU_V7M || CPU_V6K
select ARCH_HAS_GCOV_PROFILE_ALL
diff --git a/arch/arm64/Kconfig b/arch/arm64/Kconfig
index 7153d5fff603..bfb92b98d5aa 100644
--- a/arch/arm64/Kconfig
+++ b/arch/arm64/Kconfig
@@ -43,7 +43,6 @@ config ARM64
select ARCH_HAS_SYSCALL_WRAPPER
select ARCH_HAS_TEARDOWN_DMA_OPS if IOMMU_SUPPORT
select ARCH_HAS_TICK_BROADCAST if GENERIC_CLOCKEVENTS_BROADCAST
- select ARCH_HAS_VM_GET_PAGE_PROT
select ARCH_HAS_ZONE_DMA_SET if EXPERT
select ARCH_HAVE_ELF_PROT
select ARCH_HAVE_NMI_SAFE_CMPXCHG
diff --git a/arch/csky/Kconfig b/arch/csky/Kconfig
index 209dac5686dd..132f43f12dd8 100644
--- a/arch/csky/Kconfig
+++ b/arch/csky/Kconfig
@@ -6,7 +6,6 @@ config CSKY
select ARCH_HAS_GCOV_PROFILE_ALL
select ARCH_HAS_SYNC_DMA_FOR_CPU
select ARCH_HAS_SYNC_DMA_FOR_DEVICE
- select ARCH_HAS_VM_GET_PAGE_PROT
select ARCH_USE_BUILTIN_BSWAP
select ARCH_USE_QUEUED_RWLOCKS
select ARCH_WANT_FRAME_POINTERS if !CPU_CK610 && $(cc-option,-mbacktrace)
diff --git a/arch/hexagon/Kconfig b/arch/hexagon/Kconfig
index cdc5df32a1e3..15dd8f38b698 100644
--- a/arch/hexagon/Kconfig
+++ b/arch/hexagon/Kconfig
@@ -6,7 +6,6 @@ config HEXAGON
def_bool y
select ARCH_32BIT_OFF_T
select ARCH_HAS_SYNC_DMA_FOR_DEVICE
- select ARCH_HAS_VM_GET_PAGE_PROT
select ARCH_NO_PREEMPT
select DMA_GLOBAL_POOL
# Other pending projects/to-do items.
diff --git a/arch/ia64/Kconfig b/arch/ia64/Kconfig
index 0ab15e8d5783..a7e01573abd8 100644
--- a/arch/ia64/Kconfig
+++ b/arch/ia64/Kconfig
@@ -11,7 +11,6 @@ config IA64
select ARCH_HAS_DMA_MARK_CLEAN
select ARCH_HAS_STRNCPY_FROM_USER
select ARCH_HAS_STRNLEN_USER
- select ARCH_HAS_VM_GET_PAGE_PROT
select ARCH_MIGHT_HAVE_PC_PARPORT
select ARCH_MIGHT_HAVE_PC_SERIO
select ACPI
diff --git a/arch/m68k/Kconfig b/arch/m68k/Kconfig
index 114e65164692..936e1803c7c7 100644
--- a/arch/m68k/Kconfig
+++ b/arch/m68k/Kconfig
@@ -11,7 +11,6 @@ config M68K
select ARCH_NO_PREEMPT if !COLDFIRE
select ARCH_USE_MEMTEST if MMU_MOTOROLA
select ARCH_WANT_IPC_PARSE_VERSION
- select ARCH_HAS_VM_GET_PAGE_PROT
select BINFMT_FLAT_ARGVP_ENVP_ON_STACK
select DMA_DIRECT_REMAP if HAS_DMA && MMU && !COLDFIRE
select GENERIC_ATOMIC64
diff --git a/arch/microblaze/Kconfig b/arch/microblaze/Kconfig
index f2c25ba8621e..59798e43cdb0 100644
--- a/arch/microblaze/Kconfig
+++ b/arch/microblaze/Kconfig
@@ -7,7 +7,6 @@ config MICROBLAZE
select ARCH_HAS_GCOV_PROFILE_ALL
select ARCH_HAS_SYNC_DMA_FOR_CPU
select ARCH_HAS_SYNC_DMA_FOR_DEVICE
- select ARCH_HAS_VM_GET_PAGE_PROT
select ARCH_MIGHT_HAVE_PC_PARPORT
select ARCH_WANT_IPC_PARSE_VERSION
select BUILDTIME_TABLE_SORT
diff --git a/arch/mips/Kconfig b/arch/mips/Kconfig
index fcbfc52a1567..058446f01487 100644
--- a/arch/mips/Kconfig
+++ b/arch/mips/Kconfig
@@ -13,7 +13,6 @@ config MIPS
select ARCH_HAS_STRNLEN_USER
select ARCH_HAS_TICK_BROADCAST if GENERIC_CLOCKEVENTS_BROADCAST
select ARCH_HAS_UBSAN_SANITIZE_ALL
- select ARCH_HAS_VM_GET_PAGE_PROT
select ARCH_HAS_GCOV_PROFILE_ALL
select ARCH_KEEP_MEMBLOCK
select ARCH_SUPPORTS_UPROBES
diff --git a/arch/nds32/Kconfig b/arch/nds32/Kconfig
index 576e05479925..4d1421b18734 100644
--- a/arch/nds32/Kconfig
+++ b/arch/nds32/Kconfig
@@ -10,7 +10,6 @@ config NDS32
select ARCH_HAS_DMA_PREP_COHERENT
select ARCH_HAS_SYNC_DMA_FOR_CPU
select ARCH_HAS_SYNC_DMA_FOR_DEVICE
- select ARCH_HAS_VM_GET_PAGE_PROT
select ARCH_WANT_FRAME_POINTERS if FTRACE
select CLKSRC_MMIO
select CLONE_BACKWARDS
diff --git a/arch/nios2/Kconfig b/arch/nios2/Kconfig
index 85a58a357a3b..33fd06f5fa41 100644
--- a/arch/nios2/Kconfig
+++ b/arch/nios2/Kconfig
@@ -6,7 +6,6 @@ config NIOS2
select ARCH_HAS_SYNC_DMA_FOR_CPU
select ARCH_HAS_SYNC_DMA_FOR_DEVICE
select ARCH_HAS_DMA_SET_UNCACHED
- select ARCH_HAS_VM_GET_PAGE_PROT
select ARCH_NO_SWAP
select COMMON_CLK
select TIMER_OF
diff --git a/arch/openrisc/Kconfig b/arch/openrisc/Kconfig
index 842a61426816..f724b3f1aeed 100644
--- a/arch/openrisc/Kconfig
+++ b/arch/openrisc/Kconfig
@@ -10,7 +10,6 @@ config OPENRISC
select ARCH_HAS_DMA_SET_UNCACHED
select ARCH_HAS_DMA_CLEAR_UNCACHED
select ARCH_HAS_SYNC_DMA_FOR_DEVICE
- select ARCH_HAS_VM_GET_PAGE_PROT
select COMMON_CLK
select OF
select OF_EARLY_FLATTREE
diff --git a/arch/parisc/Kconfig b/arch/parisc/Kconfig
index de512f120b50..43c1c880def6 100644
--- a/arch/parisc/Kconfig
+++ b/arch/parisc/Kconfig
@@ -10,7 +10,6 @@ config PARISC
select ARCH_HAS_ELF_RANDOMIZE
select ARCH_HAS_STRICT_KERNEL_RWX
select ARCH_HAS_UBSAN_SANITIZE_ALL
- select ARCH_HAS_VM_GET_PAGE_PROT
select ARCH_NO_SG_CHAIN
select ARCH_SUPPORTS_HUGETLBFS if PA20
select ARCH_SUPPORTS_MEMORY_FAILURE
diff --git a/arch/powerpc/Kconfig b/arch/powerpc/Kconfig
index ddb4a3687c05..b779603978e1 100644
--- a/arch/powerpc/Kconfig
+++ b/arch/powerpc/Kconfig
@@ -135,7 +135,6 @@ config PPC
select ARCH_HAS_TICK_BROADCAST if GENERIC_CLOCKEVENTS_BROADCAST
select ARCH_HAS_UACCESS_FLUSHCACHE
select ARCH_HAS_UBSAN_SANITIZE_ALL
- select ARCH_HAS_VM_GET_PAGE_PROT
select ARCH_HAVE_NMI_SAFE_CMPXCHG
select ARCH_KEEP_MEMBLOCK
select ARCH_MIGHT_HAVE_PC_PARPORT
diff --git a/arch/riscv/Kconfig b/arch/riscv/Kconfig
index 9391742f9286..5adcbd9b5e88 100644
--- a/arch/riscv/Kconfig
+++ b/arch/riscv/Kconfig
@@ -31,7 +31,6 @@ config RISCV
select ARCH_HAS_STRICT_MODULE_RWX if MMU && !XIP_KERNEL
select ARCH_HAS_TICK_BROADCAST if GENERIC_CLOCKEVENTS_BROADCAST
select ARCH_HAS_UBSAN_SANITIZE_ALL
- select ARCH_HAS_VM_GET_PAGE_PROT
select ARCH_OPTIONAL_KERNEL_RWX if ARCH_HAS_STRICT_KERNEL_RWX
select ARCH_OPTIONAL_KERNEL_RWX_DEFAULT
select ARCH_STACKWALK
diff --git a/arch/s390/Kconfig b/arch/s390/Kconfig
index cb1b487e8201..be9f39fd06df 100644
--- a/arch/s390/Kconfig
+++ b/arch/s390/Kconfig
@@ -78,7 +78,6 @@ config S390
select ARCH_HAS_SYSCALL_WRAPPER
select ARCH_HAS_UBSAN_SANITIZE_ALL
select ARCH_HAS_VDSO_DATA
- select ARCH_HAS_VM_GET_PAGE_PROT
select ARCH_HAVE_NMI_SAFE_CMPXCHG
select ARCH_INLINE_READ_LOCK
select ARCH_INLINE_READ_LOCK_BH
diff --git a/arch/sh/Kconfig b/arch/sh/Kconfig
index f3fcd1c5e002..2474a04ceac4 100644
--- a/arch/sh/Kconfig
+++ b/arch/sh/Kconfig
@@ -11,7 +11,6 @@ config SUPERH
select ARCH_HAS_GCOV_PROFILE_ALL
select ARCH_HAS_PTE_SPECIAL
select ARCH_HAS_TICK_BROADCAST if GENERIC_CLOCKEVENTS_BROADCAST
- select ARCH_HAS_VM_GET_PAGE_PROT
select ARCH_HIBERNATION_POSSIBLE if MMU
select ARCH_MIGHT_HAVE_PC_PARPORT
select ARCH_WANT_IPC_PARSE_VERSION
diff --git a/arch/sparc/Kconfig b/arch/sparc/Kconfig
index ff29156f2380..1cab1b284f1a 100644
--- a/arch/sparc/Kconfig
+++ b/arch/sparc/Kconfig
@@ -59,7 +59,6 @@ config SPARC32
select HAVE_UID16
select OLD_SIGACTION
select ZONE_DMA
- select ARCH_HAS_VM_GET_PAGE_PROT
config SPARC64
def_bool 64BIT
@@ -85,7 +84,6 @@ config SPARC64
select PERF_USE_VMALLOC
select ARCH_HAVE_NMI_SAFE_CMPXCHG
select HAVE_C_RECORDMCOUNT
- select ARCH_HAS_VM_GET_PAGE_PROT
select HAVE_ARCH_AUDITSYSCALL
select ARCH_SUPPORTS_ATOMIC_RMW
select ARCH_SUPPORTS_DEBUG_PAGEALLOC
diff --git a/arch/um/Kconfig b/arch/um/Kconfig
index 5836296868a8..4d398b80aea8 100644
--- a/arch/um/Kconfig
+++ b/arch/um/Kconfig
@@ -9,7 +9,6 @@ config UML
select ARCH_HAS_KCOV
select ARCH_HAS_STRNCPY_FROM_USER
select ARCH_HAS_STRNLEN_USER
- select ARCH_HAS_VM_GET_PAGE_PROT
select ARCH_NO_PREEMPT
select HAVE_ARCH_AUDITSYSCALL
select HAVE_ARCH_SECCOMP_FILTER
diff --git a/arch/x86/Kconfig b/arch/x86/Kconfig
index b2ea06c87708..013d8d6179e5 100644
--- a/arch/x86/Kconfig
+++ b/arch/x86/Kconfig
@@ -93,7 +93,6 @@ config X86
select ARCH_HAS_SYNC_CORE_BEFORE_USERMODE
select ARCH_HAS_SYSCALL_WRAPPER
select ARCH_HAS_UBSAN_SANITIZE_ALL
- select ARCH_HAS_VM_GET_PAGE_PROT
select ARCH_HAS_DEBUG_WX
select ARCH_HAS_ZONE_DMA_SET if EXPERT
select ARCH_HAVE_NMI_SAFE_CMPXCHG
diff --git a/arch/xtensa/Kconfig b/arch/xtensa/Kconfig
index 1608f7517546..8ac599aa6d99 100644
--- a/arch/xtensa/Kconfig
+++ b/arch/xtensa/Kconfig
@@ -9,7 +9,6 @@ config XTENSA
select ARCH_HAS_DMA_SET_UNCACHED if MMU
select ARCH_HAS_STRNCPY_FROM_USER if !KASAN
select ARCH_HAS_STRNLEN_USER
- select ARCH_HAS_VM_GET_PAGE_PROT
select ARCH_USE_MEMTEST
select ARCH_USE_QUEUED_RWLOCKS
select ARCH_USE_QUEUED_SPINLOCKS
diff --git a/mm/Kconfig b/mm/Kconfig
index 212fb6e1ddaa..3326ee3903f3 100644
--- a/mm/Kconfig
+++ b/mm/Kconfig
@@ -744,9 +744,6 @@ config IDLE_PAGE_TRACKING
config ARCH_HAS_CACHE_LINE_SIZE
bool
-config ARCH_HAS_VM_GET_PAGE_PROT
- bool
-
config ARCH_HAS_PTE_DEVMAP
bool
diff --git a/mm/mmap.c b/mm/mmap.c
index ff343f76a825..07d952793315 100644
--- a/mm/mmap.c
+++ b/mm/mmap.c
@@ -81,29 +81,6 @@ static void unmap_region(struct mm_struct *mm,
struct vm_area_struct *vma, struct vm_area_struct *prev,
unsigned long start, unsigned long end);
-#ifndef CONFIG_ARCH_HAS_VM_GET_PAGE_PROT
-/* description of effects of mapping type and prot in current implementation.
- * this is due to the limited x86 page protection hardware. The expected
- * behavior is in parens:
- *
- * map_type prot
- * PROT_NONE PROT_READ PROT_WRITE PROT_EXEC
- * MAP_SHARED r: (no) no r: (yes) yes r: (no) yes r: (no) yes
- * w: (no) no w: (no) no w: (yes) yes w: (no) no
- * x: (no) no x: (no) yes x: (no) yes x: (yes) yes
- *
- * MAP_PRIVATE r: (no) no r: (yes) yes r: (no) yes r: (no) yes
- * w: (no) no w: (no) no w: (copy) copy w: (no) no
- * x: (no) no x: (no) yes x: (no) yes x: (yes) yes
- *
- * On arm64, PROT_EXEC has the following behaviour for both MAP_SHARED and
- * MAP_PRIVATE (with Enhanced PAN supported):
- * r: (no) no
- * w: (no) no
- * x: (yes) yes
- */
-#endif /* CONFIG_ARCH_HAS_VM_GET_PAGE_PROT */
-
static pgprot_t vm_pgprot_modify(pgprot_t oldprot, unsigned long vm_flags)
{
return pgprot_modify(oldprot, vm_get_page_prot(vm_flags));
--
2.25.1
^ permalink raw reply related [flat|nested] 49+ messages in thread
* [OpenRISC] [PATCH V3 09/30] arm/mm: Enable ARCH_HAS_VM_GET_PAGE_PROT
2022-02-28 10:47 ` [OpenRISC] [PATCH V3 09/30] arm/mm: " Anshuman Khandual
@ 2022-02-28 10:57 ` Russell King
2022-02-28 13:49 ` Geert Uytterhoeven
2022-03-01 0:00 ` Anshuman Khandual
0 siblings, 2 replies; 49+ messages in thread
From: Russell King @ 2022-02-28 10:57 UTC (permalink / raw)
To: openrisc
On Mon, Feb 28, 2022 at 04:17:32PM +0530, Anshuman Khandual wrote:
> This defines and exports a platform specific custom vm_get_page_prot() via
> subscribing ARCH_HAS_VM_GET_PAGE_PROT. Subsequently all __SXXX and __PXXX
> macros can be dropped which are no longer needed.
What I would really like to know is why having to run _code_ to work out
what the page protections need to be is better than looking it up in a
table.
Not only is this more expensive in terms of CPU cycles, it also brings
additional code size with it.
I'm struggling to see what the benefit is.
--
RMK's Patch system: https://www.armlinux.org.uk/developer/patches/
FTTP is here! 40Mbps down 10Mbps up. Decent connectivity at last!
^ permalink raw reply [flat|nested] 49+ messages in thread
* [OpenRISC] [PATCH V3 09/30] arm/mm: Enable ARCH_HAS_VM_GET_PAGE_PROT
2022-02-28 10:57 ` Russell King
@ 2022-02-28 13:49 ` Geert Uytterhoeven
2022-03-01 0:00 ` Anshuman Khandual
1 sibling, 0 replies; 49+ messages in thread
From: Geert Uytterhoeven @ 2022-02-28 13:49 UTC (permalink / raw)
To: openrisc
Hi Russell,
On Mon, Feb 28, 2022 at 11:57 AM Russell King (Oracle)
<linux@armlinux.org.uk> wrote:
> On Mon, Feb 28, 2022 at 04:17:32PM +0530, Anshuman Khandual wrote:
> > This defines and exports a platform specific custom vm_get_page_prot() via
> > subscribing ARCH_HAS_VM_GET_PAGE_PROT. Subsequently all __SXXX and __PXXX
> > macros can be dropped which are no longer needed.
>
> What I would really like to know is why having to run _code_ to work out
> what the page protections need to be is better than looking it up in a
> table.
>
> Not only is this more expensive in terms of CPU cycles, it also brings
> additional code size with it.
>
> I'm struggling to see what the benefit is.
I was wondering about that as well. But at least for code size on
m68k, this didn't have much impact. Looking at the generated code,
the increase due to using code for the (few different) cases is offset
by a 16-bit jump table (which is to be credited to the compiler).
In terms of CPU cycles, it's indeed worse than before.
Gr{oetje,eeting}s,
Geert
--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert at linux-m68k.org
In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
-- Linus Torvalds
^ permalink raw reply [flat|nested] 49+ messages in thread
* [OpenRISC] [PATCH V3 09/30] arm/mm: Enable ARCH_HAS_VM_GET_PAGE_PROT
2022-02-28 10:57 ` Russell King
2022-02-28 13:49 ` Geert Uytterhoeven
@ 2022-03-01 0:00 ` Anshuman Khandual
2022-03-01 0:31 ` Russell King
1 sibling, 1 reply; 49+ messages in thread
From: Anshuman Khandual @ 2022-03-01 0:00 UTC (permalink / raw)
To: openrisc
On 2/28/22 4:27 PM, Russell King (Oracle) wrote:
> On Mon, Feb 28, 2022 at 04:17:32PM +0530, Anshuman Khandual wrote:
>> This defines and exports a platform specific custom vm_get_page_prot() via
>> subscribing ARCH_HAS_VM_GET_PAGE_PROT. Subsequently all __SXXX and __PXXX
>> macros can be dropped which are no longer needed.
>
> What I would really like to know is why having to run _code_ to work out
> what the page protections need to be is better than looking it up in a
> table.
>
> Not only is this more expensive in terms of CPU cycles, it also brings
> additional code size with it.
>
> I'm struggling to see what the benefit is.
>
Currently vm_get_page_prot() is also being _run_ to fetch required page
protection values. Although that is being run in the core MM and from a
platform perspective __SXXX, __PXXX are just being exported for a table.
Looking it up in a table (and applying more constructs there after) is
not much different than a clean switch case statement in terms of CPU
usage. So this is not more expensive in terms of CPU cycles.
--------------------------
pgprot_t protection_map[16] __ro_after_init = {
__P000, __P001, __P010, __P011, __P100, __P101, __P110, __P111,
__S000, __S001, __S010, __S011, __S100, __S101, __S110, __S111
};
#ifndef CONFIG_ARCH_HAS_FILTER_PGPROT
static inline pgprot_t arch_filter_pgprot(pgprot_t prot)
{
return prot;
}
#endif
pgprot_t vm_get_page_prot(unsigned long vm_flags)
{
pgprot_t ret = __pgprot(pgprot_val(protection_map[vm_flags &
(VM_READ|VM_WRITE|VM_EXEC|VM_SHARED)]) |
pgprot_val(arch_vm_get_page_prot(vm_flags)));
return arch_filter_pgprot(ret);
}
EXPORT_SYMBOL(vm_get_page_prot)
----------------------------
There will be a single vm_get_page_prot() instance on a given platform
just like before. So this also does not bring any additional code size
with it.
As mentioned earlier on a previous version.
Remove multiple 'core MM <--> platform' abstraction layers to map
vm_flags access permission combination into page protection.
From the cover letter ......
----------
Currently there are multiple layers of abstraction i.e __SXXX/__PXXX macros
, protection_map[], arch_vm_get_page_prot() and arch_filter_pgprot() built
between the platform and generic MM, finally defining vm_get_page_prot().
Hence this series proposes to drop all these abstraction levels and instead
just move the responsibility of defining vm_get_page_prot() to the platform
itself making it clean and simple.
----------
Benefits
1. For platforms using arch_vm_get_page_prot() and/or arch_filter_pgprot()
- A simplified vm_get_page_prot()
- Dropped arch_vm_get_page_prot() and arch_filter_pgprot()
- Dropped __SXXX, __PXXX macros
2. For platforms which just exported __SXXX, __PXXX
- A simplified vm_get_page_prot()
- Dropped __SXXX, __PXXX macros
3. For core MM
- Dropped a complex vm_get_page_prot() with multiple layers
of abstraction i.e __SXXX/__PXXX macros, protection_map[],
arch_vm_get_page_prot(), arch_filter_pgprot() etc.
- Anshuman
^ permalink raw reply [flat|nested] 49+ messages in thread
* [OpenRISC] [PATCH V3 09/30] arm/mm: Enable ARCH_HAS_VM_GET_PAGE_PROT
2022-03-01 0:00 ` Anshuman Khandual
@ 2022-03-01 0:31 ` Russell King
2022-03-01 8:16 ` Christophe Leroy
2022-03-02 3:15 ` Anshuman Khandual
0 siblings, 2 replies; 49+ messages in thread
From: Russell King @ 2022-03-01 0:31 UTC (permalink / raw)
To: openrisc
On Tue, Mar 01, 2022 at 05:30:41AM +0530, Anshuman Khandual wrote:
> On 2/28/22 4:27 PM, Russell King (Oracle) wrote:
> > On Mon, Feb 28, 2022 at 04:17:32PM +0530, Anshuman Khandual wrote:
> >> This defines and exports a platform specific custom vm_get_page_prot() via
> >> subscribing ARCH_HAS_VM_GET_PAGE_PROT. Subsequently all __SXXX and __PXXX
> >> macros can be dropped which are no longer needed.
> >
> > What I would really like to know is why having to run _code_ to work out
> > what the page protections need to be is better than looking it up in a
> > table.
> >
> > Not only is this more expensive in terms of CPU cycles, it also brings
> > additional code size with it.
> >
> > I'm struggling to see what the benefit is.
>
> Currently vm_get_page_prot() is also being _run_ to fetch required page
> protection values. Although that is being run in the core MM and from a
> platform perspective __SXXX, __PXXX are just being exported for a table.
> Looking it up in a table (and applying more constructs there after) is
> not much different than a clean switch case statement in terms of CPU
> usage. So this is not more expensive in terms of CPU cycles.
I disagree.
However, let's base this disagreement on some evidence. Here is the
present 32-bit ARM implementation:
00000048 <vm_get_page_prot>:
48: e200000f and r0, r0, #15
4c: e3003000 movw r3, #0
4c: R_ARM_MOVW_ABS_NC .LANCHOR1
50: e3403000 movt r3, #0
50: R_ARM_MOVT_ABS .LANCHOR1
54: e7930100 ldr r0, [r3, r0, lsl #2]
58: e12fff1e bx lr
That is five instructions long.
Please show that your new implementation is not more expensive on
32-bit ARM. Please do so by building a 32-bit kernel, and providing
the disassembly.
I think you will find way more than five instructions in your version -
the compiler will have to issue code to decode the protection bits,
probably using a table of branches or absolute PC values, or possibly
the worst case of using multiple comparisons and branches. It then has
to load constants that may be moved using movw on ARMv7, but on
older architectures would have to be created from multiple instructions
or loaded from the literal pool. Then there'll be instructions to load
the address of "user_pgprot", retrieve its value, and bitwise or that.
Therefore, I fail to see how your approach of getting rid of the table
is somehow "better" than what we currently have in terms of the effect
on the resulting code.
If you don't like the __P and __S stuff and two arch_* hooks, you could
move the table into arch code along with vm_get_page_prot() without the
additional unnecessary hooks, while keeping all the benefits of the
table lookup.
Thanks.
--
RMK's Patch system: https://www.armlinux.org.uk/developer/patches/
FTTP is here! 40Mbps down 10Mbps up. Decent connectivity at last!
^ permalink raw reply [flat|nested] 49+ messages in thread
* [OpenRISC] [PATCH V3 09/30] arm/mm: Enable ARCH_HAS_VM_GET_PAGE_PROT
2022-03-01 0:31 ` Russell King
@ 2022-03-01 8:16 ` Christophe Leroy
2022-03-02 3:22 ` Anshuman Khandual
2022-03-02 3:15 ` Anshuman Khandual
1 sibling, 1 reply; 49+ messages in thread
From: Christophe Leroy @ 2022-03-01 8:16 UTC (permalink / raw)
To: openrisc
Le 01/03/2022 à 01:31, Russell King (Oracle) a écrit :
> On Tue, Mar 01, 2022 at 05:30:41AM +0530, Anshuman Khandual wrote:
>> On 2/28/22 4:27 PM, Russell King (Oracle) wrote:
>>> On Mon, Feb 28, 2022 at 04:17:32PM +0530, Anshuman Khandual wrote:
>>>> This defines and exports a platform specific custom vm_get_page_prot() via
>>>> subscribing ARCH_HAS_VM_GET_PAGE_PROT. Subsequently all __SXXX and __PXXX
>>>> macros can be dropped which are no longer needed.
>>>
>>> What I would really like to know is why having to run _code_ to work out
>>> what the page protections need to be is better than looking it up in a
>>> table.
>>>
>>> Not only is this more expensive in terms of CPU cycles, it also brings
>>> additional code size with it.
>>>
>>> I'm struggling to see what the benefit is.
>>
>> Currently vm_get_page_prot() is also being _run_ to fetch required page
>> protection values. Although that is being run in the core MM and from a
>> platform perspective __SXXX, __PXXX are just being exported for a table.
>> Looking it up in a table (and applying more constructs there after) is
>> not much different than a clean switch case statement in terms of CPU
>> usage. So this is not more expensive in terms of CPU cycles.
>
> I disagree.
So do I.
>
> However, let's base this disagreement on some evidence. Here is the
> present 32-bit ARM implementation:
>
> 00000048 <vm_get_page_prot>:
> 48: e200000f and r0, r0, #15
> 4c: e3003000 movw r3, #0
> 4c: R_ARM_MOVW_ABS_NC .LANCHOR1
> 50: e3403000 movt r3, #0
> 50: R_ARM_MOVT_ABS .LANCHOR1
> 54: e7930100 ldr r0, [r3, r0, lsl #2]
> 58: e12fff1e bx lr
>
> That is five instructions long.
On ppc32 I get:
00000094 <vm_get_page_prot>:
94: 3d 20 00 00 lis r9,0
96: R_PPC_ADDR16_HA .data..ro_after_init
98: 54 84 16 ba rlwinm r4,r4,2,26,29
9c: 39 29 00 00 addi r9,r9,0
9e: R_PPC_ADDR16_LO .data..ro_after_init
a0: 7d 29 20 2e lwzx r9,r9,r4
a4: 91 23 00 00 stw r9,0(r3)
a8: 4e 80 00 20 blr
>
> Please show that your new implementation is not more expensive on
> 32-bit ARM. Please do so by building a 32-bit kernel, and providing
> the disassembly.
With your series I get:
00000000 <vm_get_page_prot>:
0: 3d 20 00 00 lis r9,0
2: R_PPC_ADDR16_HA .rodata
4: 39 29 00 00 addi r9,r9,0
6: R_PPC_ADDR16_LO .rodata
8: 54 84 16 ba rlwinm r4,r4,2,26,29
c: 7d 49 20 2e lwzx r10,r9,r4
10: 7d 4a 4a 14 add r10,r10,r9
14: 7d 49 03 a6 mtctr r10
18: 4e 80 04 20 bctr
1c: 39 20 03 15 li r9,789
20: 91 23 00 00 stw r9,0(r3)
24: 4e 80 00 20 blr
28: 39 20 01 15 li r9,277
2c: 91 23 00 00 stw r9,0(r3)
30: 4e 80 00 20 blr
34: 39 20 07 15 li r9,1813
38: 91 23 00 00 stw r9,0(r3)
3c: 4e 80 00 20 blr
40: 39 20 05 15 li r9,1301
44: 91 23 00 00 stw r9,0(r3)
48: 4e 80 00 20 blr
4c: 39 20 01 11 li r9,273
50: 4b ff ff d0 b 20 <vm_get_page_prot+0x20>
That is definitely more expensive, it implements a table of branches.
Christophe
^ permalink raw reply [flat|nested] 49+ messages in thread
* [OpenRISC] [PATCH V3 19/30] csky/mm: Enable ARCH_HAS_VM_GET_PAGE_PROT
2022-02-28 10:47 ` [OpenRISC] [PATCH V3 19/30] csky/mm: " Anshuman Khandual
@ 2022-03-01 14:00 ` Guo Ren
0 siblings, 0 replies; 49+ messages in thread
From: Guo Ren @ 2022-03-01 14:00 UTC (permalink / raw)
To: openrisc
Acked-by: Guo Ren <guoren@kernel.org>
On Mon, Feb 28, 2022 at 7:10 PM Anshuman Khandual
<anshuman.khandual@arm.com> wrote:
>
> This defines and exports a platform specific custom vm_get_page_prot() via
> subscribing ARCH_HAS_VM_GET_PAGE_PROT. Subsequently all __SXXX and __PXXX
> macros can be dropped which are no longer needed.
>
> Cc: Geert Uytterhoeven <geert@linux-m68k.org>
> Cc: linux-csky at vger.kernel.org
> Cc: linux-kernel at vger.kernel.org
> Signed-off-by: Anshuman Khandual <anshuman.khandual@arm.com>
> ---
> arch/csky/Kconfig | 1 +
> arch/csky/include/asm/pgtable.h | 18 ------------------
> arch/csky/mm/init.c | 32 ++++++++++++++++++++++++++++++++
> 3 files changed, 33 insertions(+), 18 deletions(-)
>
> diff --git a/arch/csky/Kconfig b/arch/csky/Kconfig
> index 132f43f12dd8..209dac5686dd 100644
> --- a/arch/csky/Kconfig
> +++ b/arch/csky/Kconfig
> @@ -6,6 +6,7 @@ config CSKY
> select ARCH_HAS_GCOV_PROFILE_ALL
> select ARCH_HAS_SYNC_DMA_FOR_CPU
> select ARCH_HAS_SYNC_DMA_FOR_DEVICE
> + select ARCH_HAS_VM_GET_PAGE_PROT
> select ARCH_USE_BUILTIN_BSWAP
> select ARCH_USE_QUEUED_RWLOCKS
> select ARCH_WANT_FRAME_POINTERS if !CPU_CK610 && $(cc-option,-mbacktrace)
> diff --git a/arch/csky/include/asm/pgtable.h b/arch/csky/include/asm/pgtable.h
> index 151607ed5158..2c6b1cfb1cce 100644
> --- a/arch/csky/include/asm/pgtable.h
> +++ b/arch/csky/include/asm/pgtable.h
> @@ -76,24 +76,6 @@
> #define MAX_SWAPFILES_CHECK() \
> BUILD_BUG_ON(MAX_SWAPFILES_SHIFT != 5)
>
> -#define __P000 PAGE_NONE
> -#define __P001 PAGE_READ
> -#define __P010 PAGE_READ
> -#define __P011 PAGE_READ
> -#define __P100 PAGE_READ
> -#define __P101 PAGE_READ
> -#define __P110 PAGE_READ
> -#define __P111 PAGE_READ
> -
> -#define __S000 PAGE_NONE
> -#define __S001 PAGE_READ
> -#define __S010 PAGE_WRITE
> -#define __S011 PAGE_WRITE
> -#define __S100 PAGE_READ
> -#define __S101 PAGE_READ
> -#define __S110 PAGE_WRITE
> -#define __S111 PAGE_WRITE
> -
> extern unsigned long empty_zero_page[PAGE_SIZE / sizeof(unsigned long)];
> #define ZERO_PAGE(vaddr) (virt_to_page(empty_zero_page))
>
> diff --git a/arch/csky/mm/init.c b/arch/csky/mm/init.c
> index bf2004aa811a..f9babbed17d4 100644
> --- a/arch/csky/mm/init.c
> +++ b/arch/csky/mm/init.c
> @@ -197,3 +197,35 @@ void __init fixaddr_init(void)
> vaddr = __fix_to_virt(__end_of_fixed_addresses - 1) & PMD_MASK;
> fixrange_init(vaddr, vaddr + PMD_SIZE, swapper_pg_dir);
> }
> +
> +pgprot_t vm_get_page_prot(unsigned long vm_flags)
> +{
> + switch (vm_flags & (VM_READ | VM_WRITE | VM_EXEC | VM_SHARED)) {
> + case VM_NONE:
> + return PAGE_NONE;
> + case VM_READ:
> + case VM_WRITE:
> + case VM_WRITE | VM_READ:
> + case VM_EXEC:
> + case VM_EXEC | VM_READ:
> + case VM_EXEC | VM_WRITE:
> + case VM_EXEC | VM_WRITE | VM_READ:
> + return PAGE_READ;
> + case VM_SHARED:
> + return PAGE_NONE;
> + case VM_SHARED | VM_READ:
> + return PAGE_READ;
> + case VM_SHARED | VM_WRITE:
> + case VM_SHARED | VM_WRITE | VM_READ:
> + return PAGE_WRITE;
> + case VM_SHARED | VM_EXEC:
> + case VM_SHARED | VM_EXEC | VM_READ:
> + return PAGE_READ;
> + case VM_SHARED | VM_EXEC | VM_WRITE:
> + case VM_SHARED | VM_EXEC | VM_WRITE | VM_READ:
> + return PAGE_WRITE;
> + default:
> + BUILD_BUG();
> + }
> +}
> +EXPORT_SYMBOL(vm_get_page_prot);
> --
> 2.25.1
>
--
Best Regards
Guo Ren
ML: https://lore.kernel.org/linux-csky/
^ permalink raw reply [flat|nested] 49+ messages in thread
* [OpenRISC] [PATCH V3 09/30] arm/mm: Enable ARCH_HAS_VM_GET_PAGE_PROT
2022-03-01 0:31 ` Russell King
2022-03-01 8:16 ` Christophe Leroy
@ 2022-03-02 3:15 ` Anshuman Khandual
1 sibling, 0 replies; 49+ messages in thread
From: Anshuman Khandual @ 2022-03-02 3:15 UTC (permalink / raw)
To: openrisc
On 3/1/22 6:01 AM, Russell King (Oracle) wrote:
> On Tue, Mar 01, 2022 at 05:30:41AM +0530, Anshuman Khandual wrote:
>> On 2/28/22 4:27 PM, Russell King (Oracle) wrote:
>>> On Mon, Feb 28, 2022 at 04:17:32PM +0530, Anshuman Khandual wrote:
>>>> This defines and exports a platform specific custom vm_get_page_prot() via
>>>> subscribing ARCH_HAS_VM_GET_PAGE_PROT. Subsequently all __SXXX and __PXXX
>>>> macros can be dropped which are no longer needed.
>>> What I would really like to know is why having to run _code_ to work out
>>> what the page protections need to be is better than looking it up in a
>>> table.
>>>
>>> Not only is this more expensive in terms of CPU cycles, it also brings
>>> additional code size with it.
>>>
>>> I'm struggling to see what the benefit is.
>> Currently vm_get_page_prot() is also being _run_ to fetch required page
>> protection values. Although that is being run in the core MM and from a
>> platform perspective __SXXX, __PXXX are just being exported for a table.
>> Looking it up in a table (and applying more constructs there after) is
>> not much different than a clean switch case statement in terms of CPU
>> usage. So this is not more expensive in terms of CPU cycles.
> I disagree.
>
> However, let's base this disagreement on some evidence. Here is the
> present 32-bit ARM implementation:
>
> 00000048 <vm_get_page_prot>:
> 48: e200000f and r0, r0, #15
> 4c: e3003000 movw r3, #0
> 4c: R_ARM_MOVW_ABS_NC .LANCHOR1
> 50: e3403000 movt r3, #0
> 50: R_ARM_MOVT_ABS .LANCHOR1
> 54: e7930100 ldr r0, [r3, r0, lsl #2]
> 58: e12fff1e bx lr
>
> That is five instructions long.
>
> Please show that your new implementation is not more expensive on
> 32-bit ARM. Please do so by building a 32-bit kernel, and providing
> the disassembly.
>
> I think you will find way more than five instructions in your version -
> the compiler will have to issue code to decode the protection bits,
> probably using a table of branches or absolute PC values, or possibly
> the worst case of using multiple comparisons and branches. It then has
> to load constants that may be moved using movw on ARMv7, but on
> older architectures would have to be created from multiple instructions
> or loaded from the literal pool. Then there'll be instructions to load
> the address of "user_pgprot", retrieve its value, and bitwise or that.
>
> Therefore, I fail to see how your approach of getting rid of the table
> is somehow "better" than what we currently have in terms of the effect
> on the resulting code.
>
> If you don't like the __P and __S stuff and two arch_* hooks, you could
> move the table into arch code along with vm_get_page_prot() without the
> additional unnecessary hooks, while keeping all the benefits of the
> table lookup.
Okay, will change the arm's vm_get_page_prot() implementation as suggested.
^ permalink raw reply [flat|nested] 49+ messages in thread
* [OpenRISC] [PATCH V3 09/30] arm/mm: Enable ARCH_HAS_VM_GET_PAGE_PROT
2022-03-01 8:16 ` Christophe Leroy
@ 2022-03-02 3:22 ` Anshuman Khandual
2022-03-02 7:05 ` Christophe Leroy
0 siblings, 1 reply; 49+ messages in thread
From: Anshuman Khandual @ 2022-03-02 3:22 UTC (permalink / raw)
To: openrisc
On 3/1/22 1:46 PM, Christophe Leroy wrote:
>
>
> Le 01/03/2022 à 01:31, Russell King (Oracle) a écrit :
>> On Tue, Mar 01, 2022 at 05:30:41AM +0530, Anshuman Khandual wrote:
>>> On 2/28/22 4:27 PM, Russell King (Oracle) wrote:
>>>> On Mon, Feb 28, 2022 at 04:17:32PM +0530, Anshuman Khandual wrote:
>>>>> This defines and exports a platform specific custom vm_get_page_prot() via
>>>>> subscribing ARCH_HAS_VM_GET_PAGE_PROT. Subsequently all __SXXX and __PXXX
>>>>> macros can be dropped which are no longer needed.
>>>>
>>>> What I would really like to know is why having to run _code_ to work out
>>>> what the page protections need to be is better than looking it up in a
>>>> table.
>>>>
>>>> Not only is this more expensive in terms of CPU cycles, it also brings
>>>> additional code size with it.
>>>>
>>>> I'm struggling to see what the benefit is.
>>>
>>> Currently vm_get_page_prot() is also being _run_ to fetch required page
>>> protection values. Although that is being run in the core MM and from a
>>> platform perspective __SXXX, __PXXX are just being exported for a table.
>>> Looking it up in a table (and applying more constructs there after) is
>>> not much different than a clean switch case statement in terms of CPU
>>> usage. So this is not more expensive in terms of CPU cycles.
>>
>> I disagree.
>
> So do I.
>
>>
>> However, let's base this disagreement on some evidence. Here is the
>> present 32-bit ARM implementation:
>>
>> 00000048 <vm_get_page_prot>:
>> 48: e200000f and r0, r0, #15
>> 4c: e3003000 movw r3, #0
>> 4c: R_ARM_MOVW_ABS_NC .LANCHOR1
>> 50: e3403000 movt r3, #0
>> 50: R_ARM_MOVT_ABS .LANCHOR1
>> 54: e7930100 ldr r0, [r3, r0, lsl #2]
>> 58: e12fff1e bx lr
>>
>> That is five instructions long.
>
> On ppc32 I get:
>
> 00000094 <vm_get_page_prot>:
> 94: 3d 20 00 00 lis r9,0
> 96: R_PPC_ADDR16_HA .data..ro_after_init
> 98: 54 84 16 ba rlwinm r4,r4,2,26,29
> 9c: 39 29 00 00 addi r9,r9,0
> 9e: R_PPC_ADDR16_LO .data..ro_after_init
> a0: 7d 29 20 2e lwzx r9,r9,r4
> a4: 91 23 00 00 stw r9,0(r3)
> a8: 4e 80 00 20 blr
>
>
>>
>> Please show that your new implementation is not more expensive on
>> 32-bit ARM. Please do so by building a 32-bit kernel, and providing
>> the disassembly.
>
> With your series I get:
>
> 00000000 <vm_get_page_prot>:
> 0: 3d 20 00 00 lis r9,0
> 2: R_PPC_ADDR16_HA .rodata
> 4: 39 29 00 00 addi r9,r9,0
> 6: R_PPC_ADDR16_LO .rodata
> 8: 54 84 16 ba rlwinm r4,r4,2,26,29
> c: 7d 49 20 2e lwzx r10,r9,r4
> 10: 7d 4a 4a 14 add r10,r10,r9
> 14: 7d 49 03 a6 mtctr r10
> 18: 4e 80 04 20 bctr
> 1c: 39 20 03 15 li r9,789
> 20: 91 23 00 00 stw r9,0(r3)
> 24: 4e 80 00 20 blr
> 28: 39 20 01 15 li r9,277
> 2c: 91 23 00 00 stw r9,0(r3)
> 30: 4e 80 00 20 blr
> 34: 39 20 07 15 li r9,1813
> 38: 91 23 00 00 stw r9,0(r3)
> 3c: 4e 80 00 20 blr
> 40: 39 20 05 15 li r9,1301
> 44: 91 23 00 00 stw r9,0(r3)
> 48: 4e 80 00 20 blr
> 4c: 39 20 01 11 li r9,273
> 50: 4b ff ff d0 b 20 <vm_get_page_prot+0x20>
>
>
> That is definitely more expensive, it implements a table of branches.
Okay, will split out the PPC32 implementation that retains existing
table look up method. Also planning to keep that inside same file
(arch/powerpc/mm/mmap.c), unless you have a difference preference.
^ permalink raw reply [flat|nested] 49+ messages in thread
* [OpenRISC] [PATCH V3 04/30] powerpc/mm: Enable ARCH_HAS_VM_GET_PAGE_PROT
2022-02-28 10:47 ` [OpenRISC] [PATCH V3 04/30] powerpc/mm: Enable ARCH_HAS_VM_GET_PAGE_PROT Anshuman Khandual
@ 2022-03-02 5:23 ` Michael Ellerman
0 siblings, 0 replies; 49+ messages in thread
From: Michael Ellerman @ 2022-03-02 5:23 UTC (permalink / raw)
To: openrisc
Anshuman Khandual <anshuman.khandual@arm.com> writes:
> This defines and exports a platform specific custom vm_get_page_prot() via
> subscribing ARCH_HAS_VM_GET_PAGE_PROT. Subsequently all __SXXX and __PXXX
> macros can be dropped which are no longer needed. While here, this also
> localizes arch_vm_get_page_prot() as powerpc_vm_get_page_prot() and moves
> it near vm_get_page_prot().
>
> Cc: Michael Ellerman <mpe@ellerman.id.au>
> Cc: Paul Mackerras <paulus@samba.org>
> Cc: linuxppc-dev at lists.ozlabs.org
> Cc: linux-kernel at vger.kernel.org
> Signed-off-by: Anshuman Khandual <anshuman.khandual@arm.com>
> ---
> arch/powerpc/Kconfig | 1 +
> arch/powerpc/include/asm/mman.h | 12 ------
> arch/powerpc/include/asm/pgtable.h | 19 ----------
> arch/powerpc/mm/mmap.c | 59 ++++++++++++++++++++++++++++++
> 4 files changed, 60 insertions(+), 31 deletions(-)
Acked-by: Michael Ellerman <mpe@ellerman.id.au> (powerpc)
cheers
^ permalink raw reply [flat|nested] 49+ messages in thread
* [OpenRISC] [PATCH V3 09/30] arm/mm: Enable ARCH_HAS_VM_GET_PAGE_PROT
2022-03-02 3:22 ` Anshuman Khandual
@ 2022-03-02 7:05 ` Christophe Leroy
2022-03-02 9:51 ` Anshuman Khandual
0 siblings, 1 reply; 49+ messages in thread
From: Christophe Leroy @ 2022-03-02 7:05 UTC (permalink / raw)
To: openrisc
Le 02/03/2022 à 04:22, Anshuman Khandual a écrit :
>
>
> On 3/1/22 1:46 PM, Christophe Leroy wrote:
>>
>>
>> Le 01/03/2022 à 01:31, Russell King (Oracle) a écrit :
>>> On Tue, Mar 01, 2022 at 05:30:41AM +0530, Anshuman Khandual wrote:
>>>> On 2/28/22 4:27 PM, Russell King (Oracle) wrote:
>>>>> On Mon, Feb 28, 2022 at 04:17:32PM +0530, Anshuman Khandual wrote:
>>>>>> This defines and exports a platform specific custom vm_get_page_prot() via
>>>>>> subscribing ARCH_HAS_VM_GET_PAGE_PROT. Subsequently all __SXXX and __PXXX
>>>>>> macros can be dropped which are no longer needed.
>>>>>
>>>>> What I would really like to know is why having to run _code_ to work out
>>>>> what the page protections need to be is better than looking it up in a
>>>>> table.
>>>>>
>>>>> Not only is this more expensive in terms of CPU cycles, it also brings
>>>>> additional code size with it.
>>>>>
>>>>> I'm struggling to see what the benefit is.
>>>>
>>>> Currently vm_get_page_prot() is also being _run_ to fetch required page
>>>> protection values. Although that is being run in the core MM and from a
>>>> platform perspective __SXXX, __PXXX are just being exported for a table.
>>>> Looking it up in a table (and applying more constructs there after) is
>>>> not much different than a clean switch case statement in terms of CPU
>>>> usage. So this is not more expensive in terms of CPU cycles.
>>>
>>> I disagree.
>>
>> So do I.
>>
>>>
>>> However, let's base this disagreement on some evidence. Here is the
>>> present 32-bit ARM implementation:
>>>
>>> 00000048 <vm_get_page_prot>:
>>> 48: e200000f and r0, r0, #15
>>> 4c: e3003000 movw r3, #0
>>> 4c: R_ARM_MOVW_ABS_NC .LANCHOR1
>>> 50: e3403000 movt r3, #0
>>> 50: R_ARM_MOVT_ABS .LANCHOR1
>>> 54: e7930100 ldr r0, [r3, r0, lsl #2]
>>> 58: e12fff1e bx lr
>>>
>>> That is five instructions long.
>>
>> On ppc32 I get:
>>
>> 00000094 <vm_get_page_prot>:
>> 94: 3d 20 00 00 lis r9,0
>> 96: R_PPC_ADDR16_HA .data..ro_after_init
>> 98: 54 84 16 ba rlwinm r4,r4,2,26,29
>> 9c: 39 29 00 00 addi r9,r9,0
>> 9e: R_PPC_ADDR16_LO .data..ro_after_init
>> a0: 7d 29 20 2e lwzx r9,r9,r4
>> a4: 91 23 00 00 stw r9,0(r3)
>> a8: 4e 80 00 20 blr
>>
>>
>>>
>>> Please show that your new implementation is not more expensive on
>>> 32-bit ARM. Please do so by building a 32-bit kernel, and providing
>>> the disassembly.
>>
>> With your series I get:
>>
>> 00000000 <vm_get_page_prot>:
>> 0: 3d 20 00 00 lis r9,0
>> 2: R_PPC_ADDR16_HA .rodata
>> 4: 39 29 00 00 addi r9,r9,0
>> 6: R_PPC_ADDR16_LO .rodata
>> 8: 54 84 16 ba rlwinm r4,r4,2,26,29
>> c: 7d 49 20 2e lwzx r10,r9,r4
>> 10: 7d 4a 4a 14 add r10,r10,r9
>> 14: 7d 49 03 a6 mtctr r10
>> 18: 4e 80 04 20 bctr
>> 1c: 39 20 03 15 li r9,789
>> 20: 91 23 00 00 stw r9,0(r3)
>> 24: 4e 80 00 20 blr
>> 28: 39 20 01 15 li r9,277
>> 2c: 91 23 00 00 stw r9,0(r3)
>> 30: 4e 80 00 20 blr
>> 34: 39 20 07 15 li r9,1813
>> 38: 91 23 00 00 stw r9,0(r3)
>> 3c: 4e 80 00 20 blr
>> 40: 39 20 05 15 li r9,1301
>> 44: 91 23 00 00 stw r9,0(r3)
>> 48: 4e 80 00 20 blr
>> 4c: 39 20 01 11 li r9,273
>> 50: 4b ff ff d0 b 20 <vm_get_page_prot+0x20>
>>
>>
>> That is definitely more expensive, it implements a table of branches.
>
> Okay, will split out the PPC32 implementation that retains existing
> table look up method. Also planning to keep that inside same file
> (arch/powerpc/mm/mmap.c), unless you have a difference preference.
My point was not to get something specific for PPC32, but to amplify on
Russell's objection.
As this is bad for ARM and bad for PPC32, do we have any evidence that
your change is good for any other architecture ?
I checked PPC64 and there is exactly the same drawback. With the current
implementation it is a small function performing table read then a few
adjustment. After your change it is a bigger function implementing a
table of branches.
So, as requested by Russell, could you look at the disassembly for other
architectures and show us that ARM and POWERPC are the only ones for
which your change is not optimal ?
See below the difference for POWERPC64.
Current implementation:
0000000000000a60 <.vm_get_page_prot>:
a60: 3d 42 00 00 addis r10,r2,0
a62: R_PPC64_TOC16_HA .data..ro_after_init
a64: 78 89 1e 68 rldic r9,r4,3,57
a68: 39 4a 00 00 addi r10,r10,0
a6a: R_PPC64_TOC16_LO .data..ro_after_init
a6c: 74 88 01 00 andis. r8,r4,256
a70: 7d 2a 48 2a ldx r9,r10,r9
a74: 41 82 00 1c beq a90 <.vm_get_page_prot+0x30>
a78: 60 00 00 00 nop
a7c: 60 00 00 00 nop
a80: 48 00 00 18 b a98 <.vm_get_page_prot+0x38>
a84: 60 00 00 00 nop
a88: 60 00 00 00 nop
a8c: 60 00 00 00 nop
a90: 60 00 00 00 nop
a94: 60 00 00 00 nop
a98: 0f e0 00 00 twui r0,0
a9c: 60 00 00 00 nop
aa0: 38 80 00 10 li r4,16
aa4: 7d 29 23 78 or r9,r9,r4
aa8: f9 23 00 00 std r9,0(r3)
aac: 4e 80 00 20 blr
ab0: 78 84 d9 04 rldicr r4,r4,27,4
ab4: 78 84 e8 c2 rldicl r4,r4,61,3
ab8: 60 84 00 10 ori r4,r4,16
abc: 4b ff ff e8 b aa4 <.vm_get_page_prot+0x44>
ac0: 78 84 d9 04 rldicr r4,r4,27,4
ac4: 78 84 e8 c2 rldicl r4,r4,61,3
ac8: 4b ff ff dc b aa4 <.vm_get_page_prot+0x44>
With your series:
00000000000005b0 <.vm_get_page_prot>:
5b0: 3d 22 00 00 addis r9,r2,0
5b2: R_PPC64_TOC16_HA .toc+0x10
5b4: e9 49 00 00 ld r10,0(r9)
5b6: R_PPC64_TOC16_LO_DS .toc+0x10
5b8: 78 89 16 a8 rldic r9,r4,2,58
5bc: 7d 2a 4a aa lwax r9,r10,r9
5c0: 7d 29 52 14 add r9,r9,r10
5c4: 7d 29 03 a6 mtctr r9
5c8: 3d 20 80 00 lis r9,-32768
5cc: 79 29 07 c6 rldicr r9,r9,32,31
5d0: 4e 80 04 20 bctr
5d4: 00 00 00 ec .long 0xec
5d8: 00 00 00 6c .long 0x6c
5dc: 00 00 00 6c .long 0x6c
5e0: 00 00 00 6c .long 0x6c
5e4: 00 00 00 4c .long 0x4c
5e8: 00 00 00 4c .long 0x4c
5ec: 00 00 00 4c .long 0x4c
5f0: 00 00 00 4c .long 0x4c
5f4: 00 00 00 ec .long 0xec
5f8: 00 00 00 6c .long 0x6c
5fc: 00 00 00 cc .long 0xcc
600: 00 00 00 cc .long 0xcc
604: 00 00 00 4c .long 0x4c
608: 00 00 00 4c .long 0x4c
60c: 00 00 00 dc .long 0xdc
610: 00 00 00 dc .long 0xdc
614: 60 00 00 00 nop
618: 60 00 00 00 nop
61c: 60 00 00 00 nop
620: 61 29 01 05 ori r9,r9,261
624: 74 8a 01 00 andis. r10,r4,256
628: 41 82 00 24 beq 64c <.vm_get_page_prot+0x9c>
62c: 60 00 00 00 nop
630: 60 00 00 00 nop
634: 0f e0 00 00 twui r0,0
638: 60 00 00 00 nop
63c: 60 00 00 00 nop
640: 74 8a 01 00 andis. r10,r4,256
644: 61 29 01 04 ori r9,r9,260
648: 40 82 ff e4 bne 62c <.vm_get_page_prot+0x7c>
64c: 60 00 00 00 nop
650: 60 00 00 00 nop
654: 4b ff ff e0 b 634 <.vm_get_page_prot+0x84>
658: 60 00 00 00 nop
65c: 60 00 00 00 nop
660: 78 84 d9 04 rldicr r4,r4,27,4
664: 78 84 e8 c2 rldicl r4,r4,61,3
668: 7d 29 23 78 or r9,r9,r4
66c: f9 23 00 00 std r9,0(r3)
670: 4e 80 00 20 blr
674: 60 00 00 00 nop
678: 60 00 00 00 nop
67c: 60 00 00 00 nop
680: 38 80 00 10 li r4,16
684: 4b ff ff e4 b 668 <.vm_get_page_prot+0xb8>
688: 60 00 00 00 nop
68c: 60 00 00 00 nop
690: 78 84 d9 04 rldicr r4,r4,27,4
694: 78 84 e8 c2 rldicl r4,r4,61,3
698: 60 84 00 10 ori r4,r4,16
69c: 4b ff ff cc b 668 <.vm_get_page_prot+0xb8>
6a0: 61 29 01 06 ori r9,r9,262
6a4: 4b ff ff 80 b 624 <.vm_get_page_prot+0x74>
6a8: 60 00 00 00 nop
6ac: 60 00 00 00 nop
6b0: 61 29 01 07 ori r9,r9,263
6b4: 4b ff ff 70 b 624 <.vm_get_page_prot+0x74>
6b8: 60 00 00 00 nop
6bc: 60 00 00 00 nop
6c0: 61 29 01 08 ori r9,r9,264
6c4: 4b ff ff 60 b 624 <.vm_get_page_prot+0x74>
Thanks
Christophe
^ permalink raw reply [flat|nested] 49+ messages in thread
* [OpenRISC] [PATCH V3 09/30] arm/mm: Enable ARCH_HAS_VM_GET_PAGE_PROT
2022-03-02 7:05 ` Christophe Leroy
@ 2022-03-02 9:51 ` Anshuman Khandual
2022-03-02 10:05 ` Geert Uytterhoeven
0 siblings, 1 reply; 49+ messages in thread
From: Anshuman Khandual @ 2022-03-02 9:51 UTC (permalink / raw)
To: openrisc
On 3/2/22 12:35 PM, Christophe Leroy wrote:
>
>
> Le 02/03/2022 à 04:22, Anshuman Khandual a écrit :
>>
>>
>> On 3/1/22 1:46 PM, Christophe Leroy wrote:
>>>
>>>
>>> Le 01/03/2022 à 01:31, Russell King (Oracle) a écrit :
>>>> On Tue, Mar 01, 2022 at 05:30:41AM +0530, Anshuman Khandual wrote:
>>>>> On 2/28/22 4:27 PM, Russell King (Oracle) wrote:
>>>>>> On Mon, Feb 28, 2022 at 04:17:32PM +0530, Anshuman Khandual wrote:
>>>>>>> This defines and exports a platform specific custom vm_get_page_prot() via
>>>>>>> subscribing ARCH_HAS_VM_GET_PAGE_PROT. Subsequently all __SXXX and __PXXX
>>>>>>> macros can be dropped which are no longer needed.
>>>>>>
>>>>>> What I would really like to know is why having to run _code_ to work out
>>>>>> what the page protections need to be is better than looking it up in a
>>>>>> table.
>>>>>>
>>>>>> Not only is this more expensive in terms of CPU cycles, it also brings
>>>>>> additional code size with it.
>>>>>>
>>>>>> I'm struggling to see what the benefit is.
>>>>>
>>>>> Currently vm_get_page_prot() is also being _run_ to fetch required page
>>>>> protection values. Although that is being run in the core MM and from a
>>>>> platform perspective __SXXX, __PXXX are just being exported for a table.
>>>>> Looking it up in a table (and applying more constructs there after) is
>>>>> not much different than a clean switch case statement in terms of CPU
>>>>> usage. So this is not more expensive in terms of CPU cycles.
>>>>
>>>> I disagree.
>>>
>>> So do I.
>>>
>>>>
>>>> However, let's base this disagreement on some evidence. Here is the
>>>> present 32-bit ARM implementation:
>>>>
>>>> 00000048 <vm_get_page_prot>:
>>>> 48: e200000f and r0, r0, #15
>>>> 4c: e3003000 movw r3, #0
>>>> 4c: R_ARM_MOVW_ABS_NC .LANCHOR1
>>>> 50: e3403000 movt r3, #0
>>>> 50: R_ARM_MOVT_ABS .LANCHOR1
>>>> 54: e7930100 ldr r0, [r3, r0, lsl #2]
>>>> 58: e12fff1e bx lr
>>>>
>>>> That is five instructions long.
>>>
>>> On ppc32 I get:
>>>
>>> 00000094 <vm_get_page_prot>:
>>> 94: 3d 20 00 00 lis r9,0
>>> 96: R_PPC_ADDR16_HA .data..ro_after_init
>>> 98: 54 84 16 ba rlwinm r4,r4,2,26,29
>>> 9c: 39 29 00 00 addi r9,r9,0
>>> 9e: R_PPC_ADDR16_LO .data..ro_after_init
>>> a0: 7d 29 20 2e lwzx r9,r9,r4
>>> a4: 91 23 00 00 stw r9,0(r3)
>>> a8: 4e 80 00 20 blr
>>>
>>>
>>>>
>>>> Please show that your new implementation is not more expensive on
>>>> 32-bit ARM. Please do so by building a 32-bit kernel, and providing
>>>> the disassembly.
>>>
>>> With your series I get:
>>>
>>> 00000000 <vm_get_page_prot>:
>>> 0: 3d 20 00 00 lis r9,0
>>> 2: R_PPC_ADDR16_HA .rodata
>>> 4: 39 29 00 00 addi r9,r9,0
>>> 6: R_PPC_ADDR16_LO .rodata
>>> 8: 54 84 16 ba rlwinm r4,r4,2,26,29
>>> c: 7d 49 20 2e lwzx r10,r9,r4
>>> 10: 7d 4a 4a 14 add r10,r10,r9
>>> 14: 7d 49 03 a6 mtctr r10
>>> 18: 4e 80 04 20 bctr
>>> 1c: 39 20 03 15 li r9,789
>>> 20: 91 23 00 00 stw r9,0(r3)
>>> 24: 4e 80 00 20 blr
>>> 28: 39 20 01 15 li r9,277
>>> 2c: 91 23 00 00 stw r9,0(r3)
>>> 30: 4e 80 00 20 blr
>>> 34: 39 20 07 15 li r9,1813
>>> 38: 91 23 00 00 stw r9,0(r3)
>>> 3c: 4e 80 00 20 blr
>>> 40: 39 20 05 15 li r9,1301
>>> 44: 91 23 00 00 stw r9,0(r3)
>>> 48: 4e 80 00 20 blr
>>> 4c: 39 20 01 11 li r9,273
>>> 50: 4b ff ff d0 b 20 <vm_get_page_prot+0x20>
>>>
>>>
>>> That is definitely more expensive, it implements a table of branches.
>>
>> Okay, will split out the PPC32 implementation that retains existing
>> table look up method. Also planning to keep that inside same file
>> (arch/powerpc/mm/mmap.c), unless you have a difference preference.
>
> My point was not to get something specific for PPC32, but to amplify on
> Russell's objection.
>
> As this is bad for ARM and bad for PPC32, do we have any evidence that
> your change is good for any other architecture ?
>
> I checked PPC64 and there is exactly the same drawback. With the current
> implementation it is a small function performing table read then a few
> adjustment. After your change it is a bigger function implementing a
> table of branches.
I am wondering if this would not be the case for any other switch case
statement on the platform ? Is there something specific/different just
on vm_get_page_prot() implementation ? Are you suggesting that switch
case statements should just be avoided instead ?
>
> So, as requested by Russell, could you look at the disassembly for other
> architectures and show us that ARM and POWERPC are the only ones for
> which your change is not optimal ?
But the primary purpose of this series is not to guarantee optimized
code on platform by platform basis, while migrating from a table based
look up method into a switch case statement.
But instead, the purposes is to remove current levels of unnecessary
abstraction while converting a vm_flags access combination into page
protection. The switch case statement for platform implementation of
vm_get_page_prot() just seemed logical enough. Christoph's original
suggestion patch for x86 had the same implementation as well.
But if the table look up is still better/preferred method on certain
platforms like arm or ppc32, will be happy to preserve that.
- Anshuman
^ permalink raw reply [flat|nested] 49+ messages in thread
* [OpenRISC] [PATCH V3 09/30] arm/mm: Enable ARCH_HAS_VM_GET_PAGE_PROT
2022-03-02 9:51 ` Anshuman Khandual
@ 2022-03-02 10:05 ` Geert Uytterhoeven
2022-03-02 11:06 ` Anshuman Khandual
0 siblings, 1 reply; 49+ messages in thread
From: Geert Uytterhoeven @ 2022-03-02 10:05 UTC (permalink / raw)
To: openrisc
Hi Anshuman,
On Wed, Mar 2, 2022 at 10:51 AM Anshuman Khandual
<anshuman.khandual@arm.com> wrote:
> On 3/2/22 12:35 PM, Christophe Leroy wrote:
> > Le 02/03/2022 à 04:22, Anshuman Khandual a écrit :
> >> On 3/1/22 1:46 PM, Christophe Leroy wrote:
> >>> Le 01/03/2022 à 01:31, Russell King (Oracle) a écrit :
> >>>> On Tue, Mar 01, 2022 at 05:30:41AM +0530, Anshuman Khandual wrote:
> >>>>> On 2/28/22 4:27 PM, Russell King (Oracle) wrote:
> >>>>>> On Mon, Feb 28, 2022 at 04:17:32PM +0530, Anshuman Khandual wrote:
> >>>>>>> This defines and exports a platform specific custom vm_get_page_prot() via
> >>>>>>> subscribing ARCH_HAS_VM_GET_PAGE_PROT. Subsequently all __SXXX and __PXXX
> >>>>>>> macros can be dropped which are no longer needed.
> >>>>>>
> >>>>>> What I would really like to know is why having to run _code_ to work out
> >>>>>> what the page protections need to be is better than looking it up in a
> >>>>>> table.
> >>>>>>
> >>>>>> Not only is this more expensive in terms of CPU cycles, it also brings
> >>>>>> additional code size with it.
> >>>>>>
> >>>>>> I'm struggling to see what the benefit is.
> >>>>>
> >>>>> Currently vm_get_page_prot() is also being _run_ to fetch required page
> >>>>> protection values. Although that is being run in the core MM and from a
> >>>>> platform perspective __SXXX, __PXXX are just being exported for a table.
> >>>>> Looking it up in a table (and applying more constructs there after) is
> >>>>> not much different than a clean switch case statement in terms of CPU
> >>>>> usage. So this is not more expensive in terms of CPU cycles.
> >>>>
> >>>> I disagree.
> >>>
> >>> So do I.
> >>>
> >>>>
> >>>> However, let's base this disagreement on some evidence. Here is the
> >>>> present 32-bit ARM implementation:
> >>>>
> >>>> 00000048 <vm_get_page_prot>:
> >>>> 48: e200000f and r0, r0, #15
> >>>> 4c: e3003000 movw r3, #0
> >>>> 4c: R_ARM_MOVW_ABS_NC .LANCHOR1
> >>>> 50: e3403000 movt r3, #0
> >>>> 50: R_ARM_MOVT_ABS .LANCHOR1
> >>>> 54: e7930100 ldr r0, [r3, r0, lsl #2]
> >>>> 58: e12fff1e bx lr
> >>>>
> >>>> That is five instructions long.
> >>>
> >>> On ppc32 I get:
> >>>
> >>> 00000094 <vm_get_page_prot>:
> >>> 94: 3d 20 00 00 lis r9,0
> >>> 96: R_PPC_ADDR16_HA .data..ro_after_init
> >>> 98: 54 84 16 ba rlwinm r4,r4,2,26,29
> >>> 9c: 39 29 00 00 addi r9,r9,0
> >>> 9e: R_PPC_ADDR16_LO .data..ro_after_init
> >>> a0: 7d 29 20 2e lwzx r9,r9,r4
> >>> a4: 91 23 00 00 stw r9,0(r3)
> >>> a8: 4e 80 00 20 blr
> >>>
> >>>
> >>>>
> >>>> Please show that your new implementation is not more expensive on
> >>>> 32-bit ARM. Please do so by building a 32-bit kernel, and providing
> >>>> the disassembly.
> >>>
> >>> With your series I get:
> >>>
> >>> 00000000 <vm_get_page_prot>:
> >>> 0: 3d 20 00 00 lis r9,0
> >>> 2: R_PPC_ADDR16_HA .rodata
> >>> 4: 39 29 00 00 addi r9,r9,0
> >>> 6: R_PPC_ADDR16_LO .rodata
> >>> 8: 54 84 16 ba rlwinm r4,r4,2,26,29
> >>> c: 7d 49 20 2e lwzx r10,r9,r4
> >>> 10: 7d 4a 4a 14 add r10,r10,r9
> >>> 14: 7d 49 03 a6 mtctr r10
> >>> 18: 4e 80 04 20 bctr
> >>> 1c: 39 20 03 15 li r9,789
> >>> 20: 91 23 00 00 stw r9,0(r3)
> >>> 24: 4e 80 00 20 blr
> >>> 28: 39 20 01 15 li r9,277
> >>> 2c: 91 23 00 00 stw r9,0(r3)
> >>> 30: 4e 80 00 20 blr
> >>> 34: 39 20 07 15 li r9,1813
> >>> 38: 91 23 00 00 stw r9,0(r3)
> >>> 3c: 4e 80 00 20 blr
> >>> 40: 39 20 05 15 li r9,1301
> >>> 44: 91 23 00 00 stw r9,0(r3)
> >>> 48: 4e 80 00 20 blr
> >>> 4c: 39 20 01 11 li r9,273
> >>> 50: 4b ff ff d0 b 20 <vm_get_page_prot+0x20>
> >>>
> >>>
> >>> That is definitely more expensive, it implements a table of branches.
> >>
> >> Okay, will split out the PPC32 implementation that retains existing
> >> table look up method. Also planning to keep that inside same file
> >> (arch/powerpc/mm/mmap.c), unless you have a difference preference.
> >
> > My point was not to get something specific for PPC32, but to amplify on
> > Russell's objection.
> >
> > As this is bad for ARM and bad for PPC32, do we have any evidence that
> > your change is good for any other architecture ?
> >
> > I checked PPC64 and there is exactly the same drawback. With the current
> > implementation it is a small function performing table read then a few
> > adjustment. After your change it is a bigger function implementing a
> > table of branches.
>
> I am wondering if this would not be the case for any other switch case
> statement on the platform ? Is there something specific/different just
> on vm_get_page_prot() implementation ? Are you suggesting that switch
> case statements should just be avoided instead ?
>
> >
> > So, as requested by Russell, could you look at the disassembly for other
> > architectures and show us that ARM and POWERPC are the only ones for
> > which your change is not optimal ?
>
> But the primary purpose of this series is not to guarantee optimized
> code on platform by platform basis, while migrating from a table based
> look up method into a switch case statement.
>
> But instead, the purposes is to remove current levels of unnecessary
> abstraction while converting a vm_flags access combination into page
> protection. The switch case statement for platform implementation of
> vm_get_page_prot() just seemed logical enough. Christoph's original
> suggestion patch for x86 had the same implementation as well.
>
> But if the table look up is still better/preferred method on certain
> platforms like arm or ppc32, will be happy to preserve that.
I doubt the switch() variant would give better code on any platform.
What about using tables everywhere, using designated initializers
to improve readability?
Gr{oetje,eeting}s,
Geert
--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert at linux-m68k.org
In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
-- Linus Torvalds
^ permalink raw reply [flat|nested] 49+ messages in thread
* [OpenRISC] [PATCH V3 09/30] arm/mm: Enable ARCH_HAS_VM_GET_PAGE_PROT
2022-03-02 10:05 ` Geert Uytterhoeven
@ 2022-03-02 11:06 ` Anshuman Khandual
2022-03-02 11:14 ` Geert Uytterhoeven
2022-03-02 11:19 ` Russell King
0 siblings, 2 replies; 49+ messages in thread
From: Anshuman Khandual @ 2022-03-02 11:06 UTC (permalink / raw)
To: openrisc
On 3/2/22 3:35 PM, Geert Uytterhoeven wrote:
> Hi Anshuman,
>
> On Wed, Mar 2, 2022 at 10:51 AM Anshuman Khandual
> <anshuman.khandual@arm.com> wrote:
>> On 3/2/22 12:35 PM, Christophe Leroy wrote:
>>> Le 02/03/2022 à 04:22, Anshuman Khandual a écrit :
>>>> On 3/1/22 1:46 PM, Christophe Leroy wrote:
>>>>> Le 01/03/2022 à 01:31, Russell King (Oracle) a écrit :
>>>>>> On Tue, Mar 01, 2022 at 05:30:41AM +0530, Anshuman Khandual wrote:
>>>>>>> On 2/28/22 4:27 PM, Russell King (Oracle) wrote:
>>>>>>>> On Mon, Feb 28, 2022 at 04:17:32PM +0530, Anshuman Khandual wrote:
>>>>>>>>> This defines and exports a platform specific custom vm_get_page_prot() via
>>>>>>>>> subscribing ARCH_HAS_VM_GET_PAGE_PROT. Subsequently all __SXXX and __PXXX
>>>>>>>>> macros can be dropped which are no longer needed.
>>>>>>>>
>>>>>>>> What I would really like to know is why having to run _code_ to work out
>>>>>>>> what the page protections need to be is better than looking it up in a
>>>>>>>> table.
>>>>>>>>
>>>>>>>> Not only is this more expensive in terms of CPU cycles, it also brings
>>>>>>>> additional code size with it.
>>>>>>>>
>>>>>>>> I'm struggling to see what the benefit is.
>>>>>>>
>>>>>>> Currently vm_get_page_prot() is also being _run_ to fetch required page
>>>>>>> protection values. Although that is being run in the core MM and from a
>>>>>>> platform perspective __SXXX, __PXXX are just being exported for a table.
>>>>>>> Looking it up in a table (and applying more constructs there after) is
>>>>>>> not much different than a clean switch case statement in terms of CPU
>>>>>>> usage. So this is not more expensive in terms of CPU cycles.
>>>>>>
>>>>>> I disagree.
>>>>>
>>>>> So do I.
>>>>>
>>>>>>
>>>>>> However, let's base this disagreement on some evidence. Here is the
>>>>>> present 32-bit ARM implementation:
>>>>>>
>>>>>> 00000048 <vm_get_page_prot>:
>>>>>> 48: e200000f and r0, r0, #15
>>>>>> 4c: e3003000 movw r3, #0
>>>>>> 4c: R_ARM_MOVW_ABS_NC .LANCHOR1
>>>>>> 50: e3403000 movt r3, #0
>>>>>> 50: R_ARM_MOVT_ABS .LANCHOR1
>>>>>> 54: e7930100 ldr r0, [r3, r0, lsl #2]
>>>>>> 58: e12fff1e bx lr
>>>>>>
>>>>>> That is five instructions long.
>>>>>
>>>>> On ppc32 I get:
>>>>>
>>>>> 00000094 <vm_get_page_prot>:
>>>>> 94: 3d 20 00 00 lis r9,0
>>>>> 96: R_PPC_ADDR16_HA .data..ro_after_init
>>>>> 98: 54 84 16 ba rlwinm r4,r4,2,26,29
>>>>> 9c: 39 29 00 00 addi r9,r9,0
>>>>> 9e: R_PPC_ADDR16_LO .data..ro_after_init
>>>>> a0: 7d 29 20 2e lwzx r9,r9,r4
>>>>> a4: 91 23 00 00 stw r9,0(r3)
>>>>> a8: 4e 80 00 20 blr
>>>>>
>>>>>
>>>>>>
>>>>>> Please show that your new implementation is not more expensive on
>>>>>> 32-bit ARM. Please do so by building a 32-bit kernel, and providing
>>>>>> the disassembly.
>>>>>
>>>>> With your series I get:
>>>>>
>>>>> 00000000 <vm_get_page_prot>:
>>>>> 0: 3d 20 00 00 lis r9,0
>>>>> 2: R_PPC_ADDR16_HA .rodata
>>>>> 4: 39 29 00 00 addi r9,r9,0
>>>>> 6: R_PPC_ADDR16_LO .rodata
>>>>> 8: 54 84 16 ba rlwinm r4,r4,2,26,29
>>>>> c: 7d 49 20 2e lwzx r10,r9,r4
>>>>> 10: 7d 4a 4a 14 add r10,r10,r9
>>>>> 14: 7d 49 03 a6 mtctr r10
>>>>> 18: 4e 80 04 20 bctr
>>>>> 1c: 39 20 03 15 li r9,789
>>>>> 20: 91 23 00 00 stw r9,0(r3)
>>>>> 24: 4e 80 00 20 blr
>>>>> 28: 39 20 01 15 li r9,277
>>>>> 2c: 91 23 00 00 stw r9,0(r3)
>>>>> 30: 4e 80 00 20 blr
>>>>> 34: 39 20 07 15 li r9,1813
>>>>> 38: 91 23 00 00 stw r9,0(r3)
>>>>> 3c: 4e 80 00 20 blr
>>>>> 40: 39 20 05 15 li r9,1301
>>>>> 44: 91 23 00 00 stw r9,0(r3)
>>>>> 48: 4e 80 00 20 blr
>>>>> 4c: 39 20 01 11 li r9,273
>>>>> 50: 4b ff ff d0 b 20 <vm_get_page_prot+0x20>
>>>>>
>>>>>
>>>>> That is definitely more expensive, it implements a table of branches.
>>>>
>>>> Okay, will split out the PPC32 implementation that retains existing
>>>> table look up method. Also planning to keep that inside same file
>>>> (arch/powerpc/mm/mmap.c), unless you have a difference preference.
>>>
>>> My point was not to get something specific for PPC32, but to amplify on
>>> Russell's objection.
>>>
>>> As this is bad for ARM and bad for PPC32, do we have any evidence that
>>> your change is good for any other architecture ?
>>>
>>> I checked PPC64 and there is exactly the same drawback. With the current
>>> implementation it is a small function performing table read then a few
>>> adjustment. After your change it is a bigger function implementing a
>>> table of branches.
>>
>> I am wondering if this would not be the case for any other switch case
>> statement on the platform ? Is there something specific/different just
>> on vm_get_page_prot() implementation ? Are you suggesting that switch
>> case statements should just be avoided instead ?
>>
>>>
>>> So, as requested by Russell, could you look at the disassembly for other
>>> architectures and show us that ARM and POWERPC are the only ones for
>>> which your change is not optimal ?
>>
>> But the primary purpose of this series is not to guarantee optimized
>> code on platform by platform basis, while migrating from a table based
>> look up method into a switch case statement.
>>
>> But instead, the purposes is to remove current levels of unnecessary
>> abstraction while converting a vm_flags access combination into page
>> protection. The switch case statement for platform implementation of
>> vm_get_page_prot() just seemed logical enough. Christoph's original
>> suggestion patch for x86 had the same implementation as well.
>>
>> But if the table look up is still better/preferred method on certain
>> platforms like arm or ppc32, will be happy to preserve that.
>
> I doubt the switch() variant would give better code on any platform.
>
> What about using tables everywhere, using designated initializers
> to improve readability?
Designated initializers ? Could you please be more specific. A table look
up on arm platform would be something like this and arm_protection_map[]
needs to be updated with user_pgprot like before. Just wondering how a
designated initializer will help here.
static pgprot_t arm_protection_map[16] __ro_after_init = {
[VM_NONE] = __PAGE_NONE,
[VM_READ] = __PAGE_READONLY,
[VM_WRITE] = __PAGE_COPY,
[VM_WRITE | VM_READ] = __PAGE_COPY,
[VM_EXEC] = __PAGE_READONLY_EXEC,
[VM_EXEC | VM_READ] = __PAGE_READONLY_EXEC,
[VM_EXEC | VM_WRITE] = __PAGE_COPY_EXEC,
[VM_EXEC | VM_WRITE | VM_READ] = __PAGE_COPY_EXEC,
[VM_SHARED] = __PAGE_NONE,
[VM_SHARED | VM_READ] = __PAGE_READONLY,
[VM_SHARED | VM_WRITE] = __PAGE_SHARED,
[VM_SHARED | VM_WRITE | VM_READ] = __PAGE_SHARED,
[VM_SHARED | VM_EXEC] = __PAGE_READONLY_EXEC,
[VM_SHARED | VM_EXEC | VM_READ] = __PAGE_READONLY_EXEC,
[VM_SHARED | VM_EXEC | VM_WRITE] = __PAGE_SHARED_EXEC,
[VM_SHARED | VM_EXEC | VM_WRITE | VM_READ] = __PAGE_SHARED_EXEC
};
pgprot_t vm_get_page_prot(unsigned long vm_flags)
{
return __pgprot(pgprot_val(arm_protection_map[vm_flags &
(VM_READ|VM_WRITE|VM_EXEC|VM_SHARED)]));
}
EXPORT_SYMBOL(vm_get_page_prot);
^ permalink raw reply [flat|nested] 49+ messages in thread
* [OpenRISC] [PATCH V3 09/30] arm/mm: Enable ARCH_HAS_VM_GET_PAGE_PROT
2022-03-02 11:06 ` Anshuman Khandual
@ 2022-03-02 11:14 ` Geert Uytterhoeven
2022-03-09 11:33 ` Anshuman Khandual
2022-03-02 11:19 ` Russell King
1 sibling, 1 reply; 49+ messages in thread
From: Geert Uytterhoeven @ 2022-03-02 11:14 UTC (permalink / raw)
To: openrisc
Hi Anshuman,
On Wed, Mar 2, 2022 at 12:07 PM Anshuman Khandual
<anshuman.khandual@arm.com> wrote:
> On 3/2/22 3:35 PM, Geert Uytterhoeven wrote:
> > On Wed, Mar 2, 2022 at 10:51 AM Anshuman Khandual
> > <anshuman.khandual@arm.com> wrote:
> >> On 3/2/22 12:35 PM, Christophe Leroy wrote:
> >>> Le 02/03/2022 à 04:22, Anshuman Khandual a écrit :
> >>>> On 3/1/22 1:46 PM, Christophe Leroy wrote:
> >>>>> Le 01/03/2022 à 01:31, Russell King (Oracle) a écrit :
> >>>>>> On Tue, Mar 01, 2022 at 05:30:41AM +0530, Anshuman Khandual wrote:
> >>>>>>> On 2/28/22 4:27 PM, Russell King (Oracle) wrote:
> >>>>>>>> On Mon, Feb 28, 2022 at 04:17:32PM +0530, Anshuman Khandual wrote:
> >>>>>>>>> This defines and exports a platform specific custom vm_get_page_prot() via
> >>>>>>>>> subscribing ARCH_HAS_VM_GET_PAGE_PROT. Subsequently all __SXXX and __PXXX
> >>>>>>>>> macros can be dropped which are no longer needed.
> >>>>>>>>
> >>>>>>>> What I would really like to know is why having to run _code_ to work out
> >>>>>>>> what the page protections need to be is better than looking it up in a
> >>>>>>>> table.
> >>>>>>>>
> >>>>>>>> Not only is this more expensive in terms of CPU cycles, it also brings
> >>>>>>>> additional code size with it.
> >>>>>>>>
> >>>>>>>> I'm struggling to see what the benefit is.
> >>>>>>>
> >>>>>>> Currently vm_get_page_prot() is also being _run_ to fetch required page
> >>>>>>> protection values. Although that is being run in the core MM and from a
> >>>>>>> platform perspective __SXXX, __PXXX are just being exported for a table.
> >>>>>>> Looking it up in a table (and applying more constructs there after) is
> >>>>>>> not much different than a clean switch case statement in terms of CPU
> >>>>>>> usage. So this is not more expensive in terms of CPU cycles.
> >>>>>>
> >>>>>> I disagree.
> >>>>>
> >>>>> So do I.
> >>>>>
> >>>>>>
> >>>>>> However, let's base this disagreement on some evidence. Here is the
> >>>>>> present 32-bit ARM implementation:
> >>>>>>
> >>>>>> 00000048 <vm_get_page_prot>:
> >>>>>> 48: e200000f and r0, r0, #15
> >>>>>> 4c: e3003000 movw r3, #0
> >>>>>> 4c: R_ARM_MOVW_ABS_NC .LANCHOR1
> >>>>>> 50: e3403000 movt r3, #0
> >>>>>> 50: R_ARM_MOVT_ABS .LANCHOR1
> >>>>>> 54: e7930100 ldr r0, [r3, r0, lsl #2]
> >>>>>> 58: e12fff1e bx lr
> >>>>>>
> >>>>>> That is five instructions long.
> >>>>>
> >>>>> On ppc32 I get:
> >>>>>
> >>>>> 00000094 <vm_get_page_prot>:
> >>>>> 94: 3d 20 00 00 lis r9,0
> >>>>> 96: R_PPC_ADDR16_HA .data..ro_after_init
> >>>>> 98: 54 84 16 ba rlwinm r4,r4,2,26,29
> >>>>> 9c: 39 29 00 00 addi r9,r9,0
> >>>>> 9e: R_PPC_ADDR16_LO .data..ro_after_init
> >>>>> a0: 7d 29 20 2e lwzx r9,r9,r4
> >>>>> a4: 91 23 00 00 stw r9,0(r3)
> >>>>> a8: 4e 80 00 20 blr
> >>>>>
> >>>>>
> >>>>>>
> >>>>>> Please show that your new implementation is not more expensive on
> >>>>>> 32-bit ARM. Please do so by building a 32-bit kernel, and providing
> >>>>>> the disassembly.
> >>>>>
> >>>>> With your series I get:
> >>>>>
> >>>>> 00000000 <vm_get_page_prot>:
> >>>>> 0: 3d 20 00 00 lis r9,0
> >>>>> 2: R_PPC_ADDR16_HA .rodata
> >>>>> 4: 39 29 00 00 addi r9,r9,0
> >>>>> 6: R_PPC_ADDR16_LO .rodata
> >>>>> 8: 54 84 16 ba rlwinm r4,r4,2,26,29
> >>>>> c: 7d 49 20 2e lwzx r10,r9,r4
> >>>>> 10: 7d 4a 4a 14 add r10,r10,r9
> >>>>> 14: 7d 49 03 a6 mtctr r10
> >>>>> 18: 4e 80 04 20 bctr
> >>>>> 1c: 39 20 03 15 li r9,789
> >>>>> 20: 91 23 00 00 stw r9,0(r3)
> >>>>> 24: 4e 80 00 20 blr
> >>>>> 28: 39 20 01 15 li r9,277
> >>>>> 2c: 91 23 00 00 stw r9,0(r3)
> >>>>> 30: 4e 80 00 20 blr
> >>>>> 34: 39 20 07 15 li r9,1813
> >>>>> 38: 91 23 00 00 stw r9,0(r3)
> >>>>> 3c: 4e 80 00 20 blr
> >>>>> 40: 39 20 05 15 li r9,1301
> >>>>> 44: 91 23 00 00 stw r9,0(r3)
> >>>>> 48: 4e 80 00 20 blr
> >>>>> 4c: 39 20 01 11 li r9,273
> >>>>> 50: 4b ff ff d0 b 20 <vm_get_page_prot+0x20>
> >>>>>
> >>>>>
> >>>>> That is definitely more expensive, it implements a table of branches.
> >>>>
> >>>> Okay, will split out the PPC32 implementation that retains existing
> >>>> table look up method. Also planning to keep that inside same file
> >>>> (arch/powerpc/mm/mmap.c), unless you have a difference preference.
> >>>
> >>> My point was not to get something specific for PPC32, but to amplify on
> >>> Russell's objection.
> >>>
> >>> As this is bad for ARM and bad for PPC32, do we have any evidence that
> >>> your change is good for any other architecture ?
> >>>
> >>> I checked PPC64 and there is exactly the same drawback. With the current
> >>> implementation it is a small function performing table read then a few
> >>> adjustment. After your change it is a bigger function implementing a
> >>> table of branches.
> >>
> >> I am wondering if this would not be the case for any other switch case
> >> statement on the platform ? Is there something specific/different just
> >> on vm_get_page_prot() implementation ? Are you suggesting that switch
> >> case statements should just be avoided instead ?
> >>
> >>>
> >>> So, as requested by Russell, could you look at the disassembly for other
> >>> architectures and show us that ARM and POWERPC are the only ones for
> >>> which your change is not optimal ?
> >>
> >> But the primary purpose of this series is not to guarantee optimized
> >> code on platform by platform basis, while migrating from a table based
> >> look up method into a switch case statement.
> >>
> >> But instead, the purposes is to remove current levels of unnecessary
> >> abstraction while converting a vm_flags access combination into page
> >> protection. The switch case statement for platform implementation of
> >> vm_get_page_prot() just seemed logical enough. Christoph's original
> >> suggestion patch for x86 had the same implementation as well.
> >>
> >> But if the table look up is still better/preferred method on certain
> >> platforms like arm or ppc32, will be happy to preserve that.
> >
> > I doubt the switch() variant would give better code on any platform.
> >
> > What about using tables everywhere, using designated initializers
> > to improve readability?
>
> Designated initializers ? Could you please be more specific. A table look
> up on arm platform would be something like this and arm_protection_map[]
> needs to be updated with user_pgprot like before. Just wondering how a
> designated initializer will help here.
It's more readable than the original:
pgprot_t protection_map[16] __ro_after_init = {
__P000, __P001, __P010, __P011, __P100, __P101, __P110, __P111,
__S000, __S001, __S010, __S011, __S100, __S101, __S110, __S111
};
>
> static pgprot_t arm_protection_map[16] __ro_after_init = {
> [VM_NONE] = __PAGE_NONE,
> [VM_READ] = __PAGE_READONLY,
> [VM_WRITE] = __PAGE_COPY,
> [VM_WRITE | VM_READ] = __PAGE_COPY,
> [VM_EXEC] = __PAGE_READONLY_EXEC,
> [VM_EXEC | VM_READ] = __PAGE_READONLY_EXEC,
> [VM_EXEC | VM_WRITE] = __PAGE_COPY_EXEC,
> [VM_EXEC | VM_WRITE | VM_READ] = __PAGE_COPY_EXEC,
> [VM_SHARED] = __PAGE_NONE,
> [VM_SHARED | VM_READ] = __PAGE_READONLY,
> [VM_SHARED | VM_WRITE] = __PAGE_SHARED,
> [VM_SHARED | VM_WRITE | VM_READ] = __PAGE_SHARED,
> [VM_SHARED | VM_EXEC] = __PAGE_READONLY_EXEC,
> [VM_SHARED | VM_EXEC | VM_READ] = __PAGE_READONLY_EXEC,
> [VM_SHARED | VM_EXEC | VM_WRITE] = __PAGE_SHARED_EXEC,
> [VM_SHARED | VM_EXEC | VM_WRITE | VM_READ] = __PAGE_SHARED_EXEC
> };
Yeah, like that.
Seems like you already made such a conversion in
https://lore.kernel.org/all/1645425519-9034-3-git-send-email-anshuman.khandual at arm.com/
Gr{oetje,eeting}s,
Geert
--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert at linux-m68k.org
In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
-- Linus Torvalds
^ permalink raw reply [flat|nested] 49+ messages in thread
* [OpenRISC] [PATCH V3 09/30] arm/mm: Enable ARCH_HAS_VM_GET_PAGE_PROT
2022-03-02 11:06 ` Anshuman Khandual
2022-03-02 11:14 ` Geert Uytterhoeven
@ 2022-03-02 11:19 ` Russell King
1 sibling, 0 replies; 49+ messages in thread
From: Russell King @ 2022-03-02 11:19 UTC (permalink / raw)
To: openrisc
On Wed, Mar 02, 2022 at 04:36:52PM +0530, Anshuman Khandual wrote:
> On 3/2/22 3:35 PM, Geert Uytterhoeven wrote:
> > I doubt the switch() variant would give better code on any platform.
> >
> > What about using tables everywhere, using designated initializers
> > to improve readability?
>
> Designated initializers ? Could you please be more specific. A table look
> up on arm platform would be something like this and arm_protection_map[]
> needs to be updated with user_pgprot like before.
There is *absolutely* nothing wrong with that. Updating it once during
boot is way more efficient than having to compute the value each time
vm_get_page_prot() gets called.
Thanks.
--
RMK's Patch system: https://www.armlinux.org.uk/developer/patches/
FTTP is here! 40Mbps down 10Mbps up. Decent connectivity at last!
^ permalink raw reply [flat|nested] 49+ messages in thread
* [OpenRISC] [PATCH V3 05/30] arm64/mm: Enable ARCH_HAS_VM_GET_PAGE_PROT
2022-02-28 10:47 ` [OpenRISC] [PATCH V3 05/30] arm64/mm: " Anshuman Khandual
@ 2022-03-03 15:28 ` Catalin Marinas
2022-03-09 11:31 ` Anshuman Khandual
0 siblings, 1 reply; 49+ messages in thread
From: Catalin Marinas @ 2022-03-03 15:28 UTC (permalink / raw)
To: openrisc
Hi Anshuman,
On Mon, Feb 28, 2022 at 04:17:28PM +0530, Anshuman Khandual wrote:
> +static inline pgprot_t __vm_get_page_prot(unsigned long vm_flags)
> +{
> + switch (vm_flags & (VM_READ | VM_WRITE | VM_EXEC | VM_SHARED)) {
> + case VM_NONE:
> + return PAGE_NONE;
> + case VM_READ:
> + case VM_WRITE:
> + case VM_WRITE | VM_READ:
> + return PAGE_READONLY;
> + case VM_EXEC:
> + return PAGE_EXECONLY;
> + case VM_EXEC | VM_READ:
> + case VM_EXEC | VM_WRITE:
> + case VM_EXEC | VM_WRITE | VM_READ:
> + return PAGE_READONLY_EXEC;
> + case VM_SHARED:
> + return PAGE_NONE;
> + case VM_SHARED | VM_READ:
> + return PAGE_READONLY;
> + case VM_SHARED | VM_WRITE:
> + case VM_SHARED | VM_WRITE | VM_READ:
> + return PAGE_SHARED;
> + case VM_SHARED | VM_EXEC:
> + return PAGE_EXECONLY;
> + case VM_SHARED | VM_EXEC | VM_READ:
> + return PAGE_READONLY_EXEC;
> + case VM_SHARED | VM_EXEC | VM_WRITE:
> + case VM_SHARED | VM_EXEC | VM_WRITE | VM_READ:
> + return PAGE_SHARED_EXEC;
> + default:
> + BUILD_BUG();
> + }
> +}
I'd say ack for trying to get of the extra arch_vm_get_page_prot() and
arch_filter_pgprot() but, TBH, I'm not so keen on the outcome. I haven't
built the code to see what's generated but I suspect it's no significant
improvement. As for the code readability, the arm64 parts don't look
much better either. The only advantage with this patch is that all
functions have been moved under arch/arm64.
I'd keep most architectures that don't have own arch_vm_get_page_prot()
or arch_filter_pgprot() unchanged and with a generic protection_map[]
array. For architectures that need fancier stuff, add a
CONFIG_ARCH_HAS_VM_GET_PAGE_PROT (as you do) and allow them to define
vm_get_page_prot() while getting rid of arch_vm_get_page_prot() and
arch_filter_pgprot(). I think you could also duplicate protection_map[]
for architectures with own vm_get_page_prot() (make it static) and
#ifdef it out in mm/mmap.c.
If later you have more complex needs or a switch statement generates
better code, go for it, but for this series I'd keep things simple, only
focus on getting rid of arch_vm_get_page_prot() and
arch_filter_pgprot().
If I grep'ed correctly, there are only 4 architectures that have own
arch_vm_get_page_prot() (arm64, powerpc, sparc, x86) and 2 that have own
arch_filter_pgprot() (arm64, x86). Try to only change these for the time
being, together with the other generic mm cleanups you have in this
series. I think there are a couple more that touch protection_map[]
(arm, m68k). You can leave the generic protection_map[] global if the
arch does not select ARCH_HAS_VM_GET_PAGE_PROT.
> +static pgprot_t arm64_arch_filter_pgprot(pgprot_t prot)
> +{
> + if (cpus_have_const_cap(ARM64_HAS_EPAN))
> + return prot;
> +
> + if (pgprot_val(prot) != pgprot_val(PAGE_EXECONLY))
> + return prot;
> +
> + return PAGE_READONLY_EXEC;
> +}
> +
> +static pgprot_t arm64_arch_vm_get_page_prot(unsigned long vm_flags)
> +{
> + pteval_t prot = 0;
> +
> + if (vm_flags & VM_ARM64_BTI)
> + prot |= PTE_GP;
> +
> + /*
> + * There are two conditions required for returning a Normal Tagged
> + * memory type: (1) the user requested it via PROT_MTE passed to
> + * mmap() or mprotect() and (2) the corresponding vma supports MTE. We
> + * register (1) as VM_MTE in the vma->vm_flags and (2) as
> + * VM_MTE_ALLOWED. Note that the latter can only be set during the
> + * mmap() call since mprotect() does not accept MAP_* flags.
> + * Checking for VM_MTE only is sufficient since arch_validate_flags()
> + * does not permit (VM_MTE & !VM_MTE_ALLOWED).
> + */
> + if (vm_flags & VM_MTE)
> + prot |= PTE_ATTRINDX(MT_NORMAL_TAGGED);
> +
> + return __pgprot(prot);
> +}
> +
> +pgprot_t vm_get_page_prot(unsigned long vm_flags)
> +{
> + pgprot_t ret = __pgprot(pgprot_val(__vm_get_page_prot(vm_flags)) |
> + pgprot_val(arm64_arch_vm_get_page_prot(vm_flags)));
> +
> + return arm64_arch_filter_pgprot(ret);
> +}
If we kept the array, we can have everything in a single function
(untested and with my own comments for future changes):
pgprot_t vm_get_page_prot(unsigned long vm_flags)
{
pgprot_t prot = __pgprot(pgprot_val(protection_map[vm_flags &
(VM_READ|VM_WRITE|VM_EXEC|VM_SHARED)]));
/*
* We could get rid of this test if we updated protection_map[]
* to turn exec-only into read-exec during boot.
*/
if (!cpus_have_const_cap(ARM64_HAS_EPAN) &&
pgprot_val(prot) == pgprot_val(PAGE_EXECONLY))
prot = PAGE_READONLY_EXEC;
if (vm_flags & VM_ARM64_BTI)
prot != PTE_GP;
/*
* We can get rid of the requirement for PROT_NORMAL to be 0
* since here we can mask out PTE_ATTRINDX_MASK.
*/
if (vm_flags & VM_MTE) {
prot &= ~PTE_ATTRINDX_MASK;
prot |= PTE_ATTRINDX(MT_NORMAL_TAGGED);
}
return prot;
}
--
Catalin
^ permalink raw reply [flat|nested] 49+ messages in thread
* [OpenRISC] [PATCH V3 05/30] arm64/mm: Enable ARCH_HAS_VM_GET_PAGE_PROT
2022-03-03 15:28 ` Catalin Marinas
@ 2022-03-09 11:31 ` Anshuman Khandual
0 siblings, 0 replies; 49+ messages in thread
From: Anshuman Khandual @ 2022-03-09 11:31 UTC (permalink / raw)
To: openrisc
On 3/3/22 20:58, Catalin Marinas wrote:
> Hi Anshuman,
>
> On Mon, Feb 28, 2022 at 04:17:28PM +0530, Anshuman Khandual wrote:
>> +static inline pgprot_t __vm_get_page_prot(unsigned long vm_flags)
>> +{
>> + switch (vm_flags & (VM_READ | VM_WRITE | VM_EXEC | VM_SHARED)) {
>> + case VM_NONE:
>> + return PAGE_NONE;
>> + case VM_READ:
>> + case VM_WRITE:
>> + case VM_WRITE | VM_READ:
>> + return PAGE_READONLY;
>> + case VM_EXEC:
>> + return PAGE_EXECONLY;
>> + case VM_EXEC | VM_READ:
>> + case VM_EXEC | VM_WRITE:
>> + case VM_EXEC | VM_WRITE | VM_READ:
>> + return PAGE_READONLY_EXEC;
>> + case VM_SHARED:
>> + return PAGE_NONE;
>> + case VM_SHARED | VM_READ:
>> + return PAGE_READONLY;
>> + case VM_SHARED | VM_WRITE:
>> + case VM_SHARED | VM_WRITE | VM_READ:
>> + return PAGE_SHARED;
>> + case VM_SHARED | VM_EXEC:
>> + return PAGE_EXECONLY;
>> + case VM_SHARED | VM_EXEC | VM_READ:
>> + return PAGE_READONLY_EXEC;
>> + case VM_SHARED | VM_EXEC | VM_WRITE:
>> + case VM_SHARED | VM_EXEC | VM_WRITE | VM_READ:
>> + return PAGE_SHARED_EXEC;
>> + default:
>> + BUILD_BUG();
>> + }
>> +}
>
> I'd say ack for trying to get of the extra arch_vm_get_page_prot() and
> arch_filter_pgprot() but, TBH, I'm not so keen on the outcome. I haven't
> built the code to see what's generated but I suspect it's no significant
> improvement. As for the code readability, the arm64 parts don't look
> much better either. The only advantage with this patch is that all
> functions have been moved under arch/arm64.
Got it.
>
> I'd keep most architectures that don't have own arch_vm_get_page_prot()
> or arch_filter_pgprot() unchanged and with a generic protection_map[]
> array. For architectures that need fancier stuff, add a
> CONFIG_ARCH_HAS_VM_GET_PAGE_PROT (as you do) and allow them to define
> vm_get_page_prot() while getting rid of arch_vm_get_page_prot() and
> arch_filter_pgprot(). I think you could also duplicate protection_map[]
> for architectures with own vm_get_page_prot() (make it static) and
> #ifdef it out in mm/mmap.c.
>
> If later you have more complex needs or a switch statement generates
> better code, go for it, but for this series I'd keep things simple, only
> focus on getting rid of arch_vm_get_page_prot() and
> arch_filter_pgprot().
Got it.
>
> If I grep'ed correctly, there are only 4 architectures that have own
> arch_vm_get_page_prot() (arm64, powerpc, sparc, x86) and 2 that have own
> arch_filter_pgprot() (arm64, x86). Try to only change these for the time
> being, together with the other generic mm cleanups you have in this
> series. I think there are a couple more that touch protection_map[]
> (arm, m68k). You can leave the generic protection_map[] global if the
> arch does not select ARCH_HAS_VM_GET_PAGE_PROT.
Okay, I will probably split the series into two parts.
- Drop arch_vm_get_page_prot() and arch_filter_pgprot() on relevant
platforms i.e arm64, powerpc, sparc and x86 via this new config
ARCH_HAS_VM_GET_PAGE_PROT, keeping the generic protection_map[]
since platform __SXXX/__PXX macros would be still around.
- Drop __SXXX/__PXXX across all platforms via just initializing
protection_map[] early during boot in the platform OR moving
both vm_get_page_prot() via ARCH_HAS_VM_GET_PAGE_PROT and the
generic protection_map[] inside the platform.
There were some objections with respect to switch case code in
comparison to the array based table look up.
>
>> +static pgprot_t arm64_arch_filter_pgprot(pgprot_t prot)
>> +{
>> + if (cpus_have_const_cap(ARM64_HAS_EPAN))
>> + return prot;
>> +
>> + if (pgprot_val(prot) != pgprot_val(PAGE_EXECONLY))
>> + return prot;
>> +
>> + return PAGE_READONLY_EXEC;
>> +}
>> +
>> +static pgprot_t arm64_arch_vm_get_page_prot(unsigned long vm_flags)
>> +{
>> + pteval_t prot = 0;
>> +
>> + if (vm_flags & VM_ARM64_BTI)
>> + prot |= PTE_GP;
>> +
>> + /*
>> + * There are two conditions required for returning a Normal Tagged
>> + * memory type: (1) the user requested it via PROT_MTE passed to
>> + * mmap() or mprotect() and (2) the corresponding vma supports MTE. We
>> + * register (1) as VM_MTE in the vma->vm_flags and (2) as
>> + * VM_MTE_ALLOWED. Note that the latter can only be set during the
>> + * mmap() call since mprotect() does not accept MAP_* flags.
>> + * Checking for VM_MTE only is sufficient since arch_validate_flags()
>> + * does not permit (VM_MTE & !VM_MTE_ALLOWED).
>> + */
>> + if (vm_flags & VM_MTE)
>> + prot |= PTE_ATTRINDX(MT_NORMAL_TAGGED);
>> +
>> + return __pgprot(prot);
>> +}
>> +
>> +pgprot_t vm_get_page_prot(unsigned long vm_flags)
>> +{
>> + pgprot_t ret = __pgprot(pgprot_val(__vm_get_page_prot(vm_flags)) |
>> + pgprot_val(arm64_arch_vm_get_page_prot(vm_flags)));
>> +
>> + return arm64_arch_filter_pgprot(ret);
>> +}
>
> If we kept the array, we can have everything in a single function
> (untested and with my own comments for future changes):
Got it.
>
> pgprot_t vm_get_page_prot(unsigned long vm_flags)
> {
> pgprot_t prot = __pgprot(pgprot_val(protection_map[vm_flags &
> (VM_READ|VM_WRITE|VM_EXEC|VM_SHARED)]));
>
> /*
> * We could get rid of this test if we updated protection_map[]
> * to turn exec-only into read-exec during boot.
> */
> if (!cpus_have_const_cap(ARM64_HAS_EPAN) &&
> pgprot_val(prot) == pgprot_val(PAGE_EXECONLY))
> prot = PAGE_READONLY_EXEC;
>
> if (vm_flags & VM_ARM64_BTI)
> prot != PTE_GP;
>
> /*
> * We can get rid of the requirement for PROT_NORMAL to be 0
> * since here we can mask out PTE_ATTRINDX_MASK.
> */
> if (vm_flags & VM_MTE) {
> prot &= ~PTE_ATTRINDX_MASK;
> prot |= PTE_ATTRINDX(MT_NORMAL_TAGGED);
> }
>
> return prot;
> }
>
^ permalink raw reply [flat|nested] 49+ messages in thread
* [OpenRISC] [PATCH V3 09/30] arm/mm: Enable ARCH_HAS_VM_GET_PAGE_PROT
2022-03-02 11:14 ` Geert Uytterhoeven
@ 2022-03-09 11:33 ` Anshuman Khandual
0 siblings, 0 replies; 49+ messages in thread
From: Anshuman Khandual @ 2022-03-09 11:33 UTC (permalink / raw)
To: openrisc
On 3/2/22 16:44, Geert Uytterhoeven wrote:
> Hi Anshuman,
>
> On Wed, Mar 2, 2022 at 12:07 PM Anshuman Khandual
> <anshuman.khandual@arm.com> wrote:
>> On 3/2/22 3:35 PM, Geert Uytterhoeven wrote:
>>> On Wed, Mar 2, 2022 at 10:51 AM Anshuman Khandual
>>> <anshuman.khandual@arm.com> wrote:
>>>> On 3/2/22 12:35 PM, Christophe Leroy wrote:
>>>>> Le 02/03/2022 à 04:22, Anshuman Khandual a écrit :
>>>>>> On 3/1/22 1:46 PM, Christophe Leroy wrote:
>>>>>>> Le 01/03/2022 à 01:31, Russell King (Oracle) a écrit :
>>>>>>>> On Tue, Mar 01, 2022 at 05:30:41AM +0530, Anshuman Khandual wrote:
>>>>>>>>> On 2/28/22 4:27 PM, Russell King (Oracle) wrote:
>>>>>>>>>> On Mon, Feb 28, 2022 at 04:17:32PM +0530, Anshuman Khandual wrote:
>>>>>>>>>>> This defines and exports a platform specific custom vm_get_page_prot() via
>>>>>>>>>>> subscribing ARCH_HAS_VM_GET_PAGE_PROT. Subsequently all __SXXX and __PXXX
>>>>>>>>>>> macros can be dropped which are no longer needed.
>>>>>>>>>>
>>>>>>>>>> What I would really like to know is why having to run _code_ to work out
>>>>>>>>>> what the page protections need to be is better than looking it up in a
>>>>>>>>>> table.
>>>>>>>>>>
>>>>>>>>>> Not only is this more expensive in terms of CPU cycles, it also brings
>>>>>>>>>> additional code size with it.
>>>>>>>>>>
>>>>>>>>>> I'm struggling to see what the benefit is.
>>>>>>>>>
>>>>>>>>> Currently vm_get_page_prot() is also being _run_ to fetch required page
>>>>>>>>> protection values. Although that is being run in the core MM and from a
>>>>>>>>> platform perspective __SXXX, __PXXX are just being exported for a table.
>>>>>>>>> Looking it up in a table (and applying more constructs there after) is
>>>>>>>>> not much different than a clean switch case statement in terms of CPU
>>>>>>>>> usage. So this is not more expensive in terms of CPU cycles.
>>>>>>>>
>>>>>>>> I disagree.
>>>>>>>
>>>>>>> So do I.
>>>>>>>
>>>>>>>>
>>>>>>>> However, let's base this disagreement on some evidence. Here is the
>>>>>>>> present 32-bit ARM implementation:
>>>>>>>>
>>>>>>>> 00000048 <vm_get_page_prot>:
>>>>>>>> 48: e200000f and r0, r0, #15
>>>>>>>> 4c: e3003000 movw r3, #0
>>>>>>>> 4c: R_ARM_MOVW_ABS_NC .LANCHOR1
>>>>>>>> 50: e3403000 movt r3, #0
>>>>>>>> 50: R_ARM_MOVT_ABS .LANCHOR1
>>>>>>>> 54: e7930100 ldr r0, [r3, r0, lsl #2]
>>>>>>>> 58: e12fff1e bx lr
>>>>>>>>
>>>>>>>> That is five instructions long.
>>>>>>>
>>>>>>> On ppc32 I get:
>>>>>>>
>>>>>>> 00000094 <vm_get_page_prot>:
>>>>>>> 94: 3d 20 00 00 lis r9,0
>>>>>>> 96: R_PPC_ADDR16_HA .data..ro_after_init
>>>>>>> 98: 54 84 16 ba rlwinm r4,r4,2,26,29
>>>>>>> 9c: 39 29 00 00 addi r9,r9,0
>>>>>>> 9e: R_PPC_ADDR16_LO .data..ro_after_init
>>>>>>> a0: 7d 29 20 2e lwzx r9,r9,r4
>>>>>>> a4: 91 23 00 00 stw r9,0(r3)
>>>>>>> a8: 4e 80 00 20 blr
>>>>>>>
>>>>>>>
>>>>>>>>
>>>>>>>> Please show that your new implementation is not more expensive on
>>>>>>>> 32-bit ARM. Please do so by building a 32-bit kernel, and providing
>>>>>>>> the disassembly.
>>>>>>>
>>>>>>> With your series I get:
>>>>>>>
>>>>>>> 00000000 <vm_get_page_prot>:
>>>>>>> 0: 3d 20 00 00 lis r9,0
>>>>>>> 2: R_PPC_ADDR16_HA .rodata
>>>>>>> 4: 39 29 00 00 addi r9,r9,0
>>>>>>> 6: R_PPC_ADDR16_LO .rodata
>>>>>>> 8: 54 84 16 ba rlwinm r4,r4,2,26,29
>>>>>>> c: 7d 49 20 2e lwzx r10,r9,r4
>>>>>>> 10: 7d 4a 4a 14 add r10,r10,r9
>>>>>>> 14: 7d 49 03 a6 mtctr r10
>>>>>>> 18: 4e 80 04 20 bctr
>>>>>>> 1c: 39 20 03 15 li r9,789
>>>>>>> 20: 91 23 00 00 stw r9,0(r3)
>>>>>>> 24: 4e 80 00 20 blr
>>>>>>> 28: 39 20 01 15 li r9,277
>>>>>>> 2c: 91 23 00 00 stw r9,0(r3)
>>>>>>> 30: 4e 80 00 20 blr
>>>>>>> 34: 39 20 07 15 li r9,1813
>>>>>>> 38: 91 23 00 00 stw r9,0(r3)
>>>>>>> 3c: 4e 80 00 20 blr
>>>>>>> 40: 39 20 05 15 li r9,1301
>>>>>>> 44: 91 23 00 00 stw r9,0(r3)
>>>>>>> 48: 4e 80 00 20 blr
>>>>>>> 4c: 39 20 01 11 li r9,273
>>>>>>> 50: 4b ff ff d0 b 20 <vm_get_page_prot+0x20>
>>>>>>>
>>>>>>>
>>>>>>> That is definitely more expensive, it implements a table of branches.
>>>>>>
>>>>>> Okay, will split out the PPC32 implementation that retains existing
>>>>>> table look up method. Also planning to keep that inside same file
>>>>>> (arch/powerpc/mm/mmap.c), unless you have a difference preference.
>>>>>
>>>>> My point was not to get something specific for PPC32, but to amplify on
>>>>> Russell's objection.
>>>>>
>>>>> As this is bad for ARM and bad for PPC32, do we have any evidence that
>>>>> your change is good for any other architecture ?
>>>>>
>>>>> I checked PPC64 and there is exactly the same drawback. With the current
>>>>> implementation it is a small function performing table read then a few
>>>>> adjustment. After your change it is a bigger function implementing a
>>>>> table of branches.
>>>>
>>>> I am wondering if this would not be the case for any other switch case
>>>> statement on the platform ? Is there something specific/different just
>>>> on vm_get_page_prot() implementation ? Are you suggesting that switch
>>>> case statements should just be avoided instead ?
>>>>
>>>>>
>>>>> So, as requested by Russell, could you look at the disassembly for other
>>>>> architectures and show us that ARM and POWERPC are the only ones for
>>>>> which your change is not optimal ?
>>>>
>>>> But the primary purpose of this series is not to guarantee optimized
>>>> code on platform by platform basis, while migrating from a table based
>>>> look up method into a switch case statement.
>>>>
>>>> But instead, the purposes is to remove current levels of unnecessary
>>>> abstraction while converting a vm_flags access combination into page
>>>> protection. The switch case statement for platform implementation of
>>>> vm_get_page_prot() just seemed logical enough. Christoph's original
>>>> suggestion patch for x86 had the same implementation as well.
>>>>
>>>> But if the table look up is still better/preferred method on certain
>>>> platforms like arm or ppc32, will be happy to preserve that.
>>>
>>> I doubt the switch() variant would give better code on any platform.
>>>
>>> What about using tables everywhere, using designated initializers
>>> to improve readability?
>>
>> Designated initializers ? Could you please be more specific. A table look
>> up on arm platform would be something like this and arm_protection_map[]
>> needs to be updated with user_pgprot like before. Just wondering how a
>> designated initializer will help here.
>
> It's more readable than the original:
>
> pgprot_t protection_map[16] __ro_after_init = {
> __P000, __P001, __P010, __P011, __P100, __P101, __P110, __P111,
> __S000, __S001, __S010, __S011, __S100, __S101, __S110, __S111
> };
>
>>
>> static pgprot_t arm_protection_map[16] __ro_after_init = {
>> [VM_NONE] = __PAGE_NONE,
>> [VM_READ] = __PAGE_READONLY,
>> [VM_WRITE] = __PAGE_COPY,
>> [VM_WRITE | VM_READ] = __PAGE_COPY,
>> [VM_EXEC] = __PAGE_READONLY_EXEC,
>> [VM_EXEC | VM_READ] = __PAGE_READONLY_EXEC,
>> [VM_EXEC | VM_WRITE] = __PAGE_COPY_EXEC,
>> [VM_EXEC | VM_WRITE | VM_READ] = __PAGE_COPY_EXEC,
>> [VM_SHARED] = __PAGE_NONE,
>> [VM_SHARED | VM_READ] = __PAGE_READONLY,
>> [VM_SHARED | VM_WRITE] = __PAGE_SHARED,
>> [VM_SHARED | VM_WRITE | VM_READ] = __PAGE_SHARED,
>> [VM_SHARED | VM_EXEC] = __PAGE_READONLY_EXEC,
>> [VM_SHARED | VM_EXEC | VM_READ] = __PAGE_READONLY_EXEC,
>> [VM_SHARED | VM_EXEC | VM_WRITE] = __PAGE_SHARED_EXEC,
>> [VM_SHARED | VM_EXEC | VM_WRITE | VM_READ] = __PAGE_SHARED_EXEC
>> };
>
> Yeah, like that.
>
> Seems like you already made such a conversion in
> https://lore.kernel.org/all/1645425519-9034-3-git-send-email-anshuman.khandual at arm.com/
Will rework the series in two different phases as mentioned on the other thread.
^ permalink raw reply [flat|nested] 49+ messages in thread
end of thread, other threads:[~2022-03-09 11:33 UTC | newest]
Thread overview: 49+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2022-02-28 10:47 [OpenRISC] [PATCH V3 00/30] mm/mmap: Drop protection_map[] and platform's __SXXX/__PXXX requirements Anshuman Khandual
2022-02-28 10:47 ` [OpenRISC] [PATCH V3 01/30] mm/debug_vm_pgtable: Drop protection_map[] usage Anshuman Khandual
2022-02-28 10:47 ` [OpenRISC] [PATCH V3 02/30] mm/mmap: Clarify protection_map[] indices Anshuman Khandual
2022-02-28 10:47 ` [OpenRISC] [PATCH V3 03/30] mm/mmap: Add new config ARCH_HAS_VM_GET_PAGE_PROT Anshuman Khandual
2022-02-28 10:47 ` [OpenRISC] [PATCH V3 04/30] powerpc/mm: Enable ARCH_HAS_VM_GET_PAGE_PROT Anshuman Khandual
2022-03-02 5:23 ` Michael Ellerman
2022-02-28 10:47 ` [OpenRISC] [PATCH V3 05/30] arm64/mm: " Anshuman Khandual
2022-03-03 15:28 ` Catalin Marinas
2022-03-09 11:31 ` Anshuman Khandual
2022-02-28 10:47 ` [OpenRISC] [PATCH V3 06/30] sparc/mm: " Anshuman Khandual
2022-02-28 10:47 ` [OpenRISC] [PATCH V3 07/30] mips/mm: " Anshuman Khandual
2022-02-28 10:47 ` [OpenRISC] [PATCH V3 08/30] m68k/mm: " Anshuman Khandual
2022-02-28 10:47 ` [OpenRISC] [PATCH V3 09/30] arm/mm: " Anshuman Khandual
2022-02-28 10:57 ` Russell King
2022-02-28 13:49 ` Geert Uytterhoeven
2022-03-01 0:00 ` Anshuman Khandual
2022-03-01 0:31 ` Russell King
2022-03-01 8:16 ` Christophe Leroy
2022-03-02 3:22 ` Anshuman Khandual
2022-03-02 7:05 ` Christophe Leroy
2022-03-02 9:51 ` Anshuman Khandual
2022-03-02 10:05 ` Geert Uytterhoeven
2022-03-02 11:06 ` Anshuman Khandual
2022-03-02 11:14 ` Geert Uytterhoeven
2022-03-09 11:33 ` Anshuman Khandual
2022-03-02 11:19 ` Russell King
2022-03-02 3:15 ` Anshuman Khandual
2022-02-28 10:47 ` [OpenRISC] [PATCH V3 10/30] x86/mm: " Anshuman Khandual
2022-02-28 10:47 ` [OpenRISC] [PATCH V3 11/30] mm/mmap: Drop protection_map[] Anshuman Khandual
2022-02-28 10:47 ` [OpenRISC] [PATCH V3 12/30] mm/mmap: Drop arch_filter_pgprot() Anshuman Khandual
2022-02-28 10:47 ` [OpenRISC] [PATCH V3 13/30] mm/mmap: Drop arch_vm_get_page_pgprot() Anshuman Khandual
2022-02-28 10:47 ` [OpenRISC] [PATCH V3 14/30] s390/mm: Enable ARCH_HAS_VM_GET_PAGE_PROT Anshuman Khandual
2022-02-28 10:47 ` [OpenRISC] [PATCH V3 15/30] riscv/mm: " Anshuman Khandual
2022-02-28 10:47 ` [OpenRISC] [PATCH V3 16/30] alpha/mm: " Anshuman Khandual
2022-02-28 10:47 ` [OpenRISC] [PATCH V3 17/30] sh/mm: " Anshuman Khandual
2022-02-28 10:47 ` [OpenRISC] [PATCH V3 18/30] arc/mm: " Anshuman Khandual
2022-02-28 10:47 ` [OpenRISC] [PATCH V3 19/30] csky/mm: " Anshuman Khandual
2022-03-01 14:00 ` Guo Ren
2022-02-28 10:47 ` [OpenRISC] [PATCH V3 20/30] xtensa/mm: " Anshuman Khandual
2022-02-28 10:47 ` [OpenRISC] [PATCH V3 21/30] parisc/mm: " Anshuman Khandual
2022-02-28 10:47 ` [OpenRISC] [PATCH V3 22/30] openrisc/mm: " Anshuman Khandual
2022-02-28 10:47 ` [OpenRISC] [PATCH V3 23/30] um/mm: " Anshuman Khandual
2022-02-28 10:47 ` [OpenRISC] [PATCH V3 24/30] microblaze/mm: " Anshuman Khandual
2022-02-28 10:47 ` [OpenRISC] [PATCH V3 25/30] nios2/mm: " Anshuman Khandual
2022-02-28 10:47 ` [OpenRISC] [PATCH V3 26/30] hexagon/mm: " Anshuman Khandual
2022-02-28 10:47 ` [OpenRISC] [PATCH V3 27/30] nds32/mm: " Anshuman Khandual
2022-02-28 10:47 ` [OpenRISC] [PATCH V3 28/30] ia64/mm: " Anshuman Khandual
2022-02-28 10:47 ` [OpenRISC] [PATCH V3 29/30] mm/mmap: Drop generic vm_get_page_prot() Anshuman Khandual
2022-02-28 10:47 ` [OpenRISC] [PATCH V3 30/30] mm/mmap: Drop ARCH_HAS_VM_GET_PAGE_PROT Anshuman Khandual
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).