[PATCH v4 00/41] Kernel userspace access/execution prevention with hash translation

[PATCH v4 00/41] Kernel userspace access/execution prevention with hash translation

[Aneesh Kumar K.V]

This patch series implements KUAP and KUEP with hash translation mode using
memory keys. The kernel now uses memory protection key 3 to control access
to the kernel. Kernel page table entries are now configured with key 3.
Access to locations configured with any other key value is denied when in
kernel mode (MSR_PR=0). This includes userspace which is by default configured
with key 0.

Changes from v3:
* Fix build error reported by kernel test robot <lkp@intel.com>

Changes from v2:
* Rebase to the latest kernel.
* Fixed a bug with disabling KUEP/KUAP on kernel command line
* Added a patch to make kup key dynamic.

Changes from V1:
* Rebased on latest kernel

Aneesh Kumar K.V (41):
  powerpc/book3s64/pkeys: Fixup bit numbering
  powerpc/book3s64/pkeys: pkeys are supported only on hash on book3s.
  powerpc/book3s64/pkeys: Move pkey related bits in the linux page table
  powerpc/book3s64/pkeys: Explain key 1 reservation details
  powerpc/book3s64/pkeys: Simplify the key initialization
  powerpc/book3s64/pkeys: Prevent key 1 modification from userspace.
  powerpc/book3s64/pkeys: kill cpu feature key CPU_FTR_PKEY
  powerpc/book3s64/pkeys: Convert execute key support to static key
  powerpc/book3s64/pkeys: Simplify pkey disable branch
  powerpc/book3s64/pkeys: Convert pkey_total to max_pkey
  powerpc/book3s64/pkeys: Make initial_allocation_mask static
  powerpc/book3s64/pkeys: Mark all the pkeys above max pkey as reserved
  powerpc/book3s64/pkeys: Enable MMU_FTR_PKEY
  powerpc/book3s64/kuep: Add MMU_FTR_KUEP
  powerpc/book3s64/pkeys: Use execute_pkey_disable static key
  powerpc/book3s64/pkeys: Use MMU_FTR_PKEY instead of pkey_disabled
    static key
  powerpc/book3s64/kuap: Move KUAP related function outside radix
  powerpc/book3s64/kuep: Move KUEP related function outside radix
  powerpc/book3s64/kuap: Rename MMU_FTR_RADIX_KUAP to MMU_FTR_KUAP
  powerpc/book3s64/kuap/kuep: Make KUAP and KUEP a subfeature of
    PPC_MEM_KEYS
  powerpc/book3s64/kuap: Move UAMOR setup to key init function
  powerpc/book3s64/kuap: Use Key 3 for kernel mapping with hash
    translation
  powerpc/exec: Set thread.regs early during exec
  powerpc/book3s64/pkeys: Store/restore userspace AMR correctly on entry
    and exit from kernel
  powerpc/book3s64/kuep: Store/restore userspace IAMR correctly on entry
    and exit from kernel
  powerpc/book3s64/pkeys: Inherit correctly on fork.
  powerpc/book3s64/pkeys: Reset userspace AMR correctly on exec
  powerpc/ptrace-view: Use pt_regs values instead of thread_struct based
    one.
  powerpc/book3s64/pkeys: Don't update SPRN_AMR when in kernel mode.
  powerpc/book3s64/kuap: Restrict access to userspace based on userspace
    AMR
  powerpc/book3s64/kuap: Improve error reporting with KUAP
  powerpc/book3s64/kuap: Use Key 3 to implement KUAP with hash
    translation.
  powerpc/book3s64/kuep: Use Key 3 to implement KUEP with hash
    translation.
  powerpc/book3s64/hash/kuap: Enable kuap on hash
  powerpc/book3s64/hash/kuep: Enable KUEP on hash
  powerpc/book3s64/keys: Print information during boot.
  powerpc/selftest/ptrave-pkey: Rename variables to make it easier to
    follow code
  powerpc/selftest/ptrace-pkey: Update the test to mark an invalid pkey
    correctly
  powerpc/selftest/ptrace-pkey: IAMR and uamor cannot be updated by
    ptrace
  powerpc/book3s64/keys/kuap: Reset AMR/IAMR values on kexec
  powerpc/book3s64/hash/kup: Don't hardcode kup key

 arch/powerpc/include/asm/book3s/32/kup.h      |   4 +-
 arch/powerpc/include/asm/book3s/64/hash-4k.h  |  21 +-
 arch/powerpc/include/asm/book3s/64/hash-64k.h |  12 +-
 .../powerpc/include/asm/book3s/64/hash-pkey.h |  24 +
 arch/powerpc/include/asm/book3s/64/hash.h     |   3 +-
 .../powerpc/include/asm/book3s/64/kup-radix.h | 185 --------
 arch/powerpc/include/asm/book3s/64/kup.h      | 391 +++++++++++++++++
 arch/powerpc/include/asm/book3s/64/mmu-hash.h |   9 +-
 arch/powerpc/include/asm/book3s/64/mmu.h      |   6 +
 arch/powerpc/include/asm/book3s/64/pgtable.h  |  17 +-
 arch/powerpc/include/asm/book3s/64/pkeys.h    |  25 ++
 arch/powerpc/include/asm/cputable.h           |  13 +-
 arch/powerpc/include/asm/kup.h                |  20 +-
 arch/powerpc/include/asm/mmu.h                |  17 +-
 arch/powerpc/include/asm/mmu_context.h        |   2 +-
 arch/powerpc/include/asm/nohash/32/kup-8xx.h  |   4 +-
 arch/powerpc/include/asm/pkeys.h              |  65 +--
 arch/powerpc/include/asm/processor.h          |   5 -
 arch/powerpc/include/asm/ptrace.h             |   8 +-
 arch/powerpc/include/asm/thread_info.h        |   2 -
 arch/powerpc/kernel/asm-offsets.c             |   6 +-
 arch/powerpc/kernel/dt_cpu_ftrs.c             |   6 -
 arch/powerpc/kernel/entry_64.S                |   6 +-
 arch/powerpc/kernel/exceptions-64s.S          |   4 +-
 arch/powerpc/kernel/misc_64.S                 |  14 -
 arch/powerpc/kernel/process.c                 |  56 ++-
 arch/powerpc/kernel/prom.c                    |   5 +
 arch/powerpc/kernel/ptrace/ptrace-view.c      |  23 +-
 arch/powerpc/kernel/smp.c                     |   5 +
 arch/powerpc/kernel/syscall_64.c              |  32 +-
 arch/powerpc/kernel/traps.c                   |   6 -
 arch/powerpc/kexec/core_64.c                  |   3 +
 arch/powerpc/mm/book3s64/hash_4k.c            |   2 +-
 arch/powerpc/mm/book3s64/hash_64k.c           |   4 +-
 arch/powerpc/mm/book3s64/hash_hugepage.c      |   2 +-
 arch/powerpc/mm/book3s64/hash_hugetlbpage.c   |   2 +-
 arch/powerpc/mm/book3s64/hash_pgtable.c       |   2 +-
 arch/powerpc/mm/book3s64/hash_utils.c         |  10 +-
 arch/powerpc/mm/book3s64/pgtable.c            |   3 +
 arch/powerpc/mm/book3s64/pkeys.c              | 412 +++++++++++-------
 arch/powerpc/mm/book3s64/radix_pgtable.c      |  36 --
 arch/powerpc/mm/fault.c                       |   2 +-
 arch/powerpc/platforms/Kconfig.cputype        |   4 +-
 .../selftests/powerpc/ptrace/ptrace-pkey.c    |  53 +--
 44 files changed, 939 insertions(+), 592 deletions(-)
 create mode 100644 arch/powerpc/include/asm/book3s/64/hash-pkey.h
 delete mode 100644 arch/powerpc/include/asm/book3s/64/kup-radix.h
 create mode 100644 arch/powerpc/include/asm/book3s/64/kup.h
 create mode 100644 arch/powerpc/include/asm/book3s/64/pkeys.h

-- 
2.26.2

[PATCH v4 01/41] powerpc/book3s64/pkeys: Fixup bit numbering

[Aneesh Kumar K.V]

This number the pkey bit such that it is easy to follow. PKEY_BIT0 is
the lower order bit. This makes further changes easy to follow.

No functional change in this patch other than linux page table for
hash translation now maps pkeys differently.

Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
---
 arch/powerpc/include/asm/book3s/64/hash-4k.h  |  9 +++----
 arch/powerpc/include/asm/book3s/64/hash-64k.h |  8 +++----
 arch/powerpc/include/asm/book3s/64/mmu-hash.h |  8 +++----
 arch/powerpc/include/asm/pkeys.h              | 24 +++++++++----------
 4 files changed, 25 insertions(+), 24 deletions(-)

diff --git a/arch/powerpc/include/asm/book3s/64/hash-4k.h b/arch/powerpc/include/asm/book3s/64/hash-4k.h
index 3f9ae3585ab9..f889d56bf8cf 100644
--- a/arch/powerpc/include/asm/book3s/64/hash-4k.h
+++ b/arch/powerpc/include/asm/book3s/64/hash-4k.h
@@ -57,11 +57,12 @@
 #define H_PMD_FRAG_NR	(PAGE_SIZE >> H_PMD_FRAG_SIZE_SHIFT)
 
 /* memory key bits, only 8 keys supported */
-#define H_PTE_PKEY_BIT0	0
-#define H_PTE_PKEY_BIT1	0
+#define H_PTE_PKEY_BIT4	0
+#define H_PTE_PKEY_BIT3	0
 #define H_PTE_PKEY_BIT2	_RPAGE_RSV3
-#define H_PTE_PKEY_BIT3	_RPAGE_RSV4
-#define H_PTE_PKEY_BIT4	_RPAGE_RSV5
+#define H_PTE_PKEY_BIT1	_RPAGE_RSV4
+#define H_PTE_PKEY_BIT0	_RPAGE_RSV5
+
 
 /*
  * On all 4K setups, remap_4k_pfn() equates to remap_pfn_range()
diff --git a/arch/powerpc/include/asm/book3s/64/hash-64k.h b/arch/powerpc/include/asm/book3s/64/hash-64k.h
index 0729c034e56f..0a15fd14cf72 100644
--- a/arch/powerpc/include/asm/book3s/64/hash-64k.h
+++ b/arch/powerpc/include/asm/book3s/64/hash-64k.h
@@ -36,11 +36,11 @@
 #define H_PAGE_HASHPTE	_RPAGE_RPN43	/* PTE has associated HPTE */
 
 /* memory key bits. */
-#define H_PTE_PKEY_BIT0	_RPAGE_RSV1
-#define H_PTE_PKEY_BIT1	_RPAGE_RSV2
+#define H_PTE_PKEY_BIT4	_RPAGE_RSV1
+#define H_PTE_PKEY_BIT3	_RPAGE_RSV2
 #define H_PTE_PKEY_BIT2	_RPAGE_RSV3
-#define H_PTE_PKEY_BIT3	_RPAGE_RSV4
-#define H_PTE_PKEY_BIT4	_RPAGE_RSV5
+#define H_PTE_PKEY_BIT1	_RPAGE_RSV4
+#define H_PTE_PKEY_BIT0	_RPAGE_RSV5
 
 /*
  * We need to differentiate between explicit huge page and THP huge
diff --git a/arch/powerpc/include/asm/book3s/64/mmu-hash.h b/arch/powerpc/include/asm/book3s/64/mmu-hash.h
index 3fa1b962dc27..58fcc959f9d5 100644
--- a/arch/powerpc/include/asm/book3s/64/mmu-hash.h
+++ b/arch/powerpc/include/asm/book3s/64/mmu-hash.h
@@ -86,8 +86,8 @@
 #define HPTE_R_PP0		ASM_CONST(0x8000000000000000)
 #define HPTE_R_TS		ASM_CONST(0x4000000000000000)
 #define HPTE_R_KEY_HI		ASM_CONST(0x3000000000000000)
-#define HPTE_R_KEY_BIT0		ASM_CONST(0x2000000000000000)
-#define HPTE_R_KEY_BIT1		ASM_CONST(0x1000000000000000)
+#define HPTE_R_KEY_BIT4		ASM_CONST(0x2000000000000000)
+#define HPTE_R_KEY_BIT3		ASM_CONST(0x1000000000000000)
 #define HPTE_R_RPN_SHIFT	12
 #define HPTE_R_RPN		ASM_CONST(0x0ffffffffffff000)
 #define HPTE_R_RPN_3_0		ASM_CONST(0x01fffffffffff000)
@@ -103,8 +103,8 @@
 #define HPTE_R_R		ASM_CONST(0x0000000000000100)
 #define HPTE_R_KEY_LO		ASM_CONST(0x0000000000000e00)
 #define HPTE_R_KEY_BIT2		ASM_CONST(0x0000000000000800)
-#define HPTE_R_KEY_BIT3		ASM_CONST(0x0000000000000400)
-#define HPTE_R_KEY_BIT4		ASM_CONST(0x0000000000000200)
+#define HPTE_R_KEY_BIT1		ASM_CONST(0x0000000000000400)
+#define HPTE_R_KEY_BIT0		ASM_CONST(0x0000000000000200)
 #define HPTE_R_KEY		(HPTE_R_KEY_LO | HPTE_R_KEY_HI)
 
 #define HPTE_V_1TB_SEG		ASM_CONST(0x4000000000000000)
diff --git a/arch/powerpc/include/asm/pkeys.h b/arch/powerpc/include/asm/pkeys.h
index 20ebf153c871..f8f4d0793789 100644
--- a/arch/powerpc/include/asm/pkeys.h
+++ b/arch/powerpc/include/asm/pkeys.h
@@ -35,11 +35,11 @@ static inline u64 vmflag_to_pte_pkey_bits(u64 vm_flags)
 	if (static_branch_likely(&pkey_disabled))
 		return 0x0UL;
 
-	return (((vm_flags & VM_PKEY_BIT0) ? H_PTE_PKEY_BIT4 : 0x0UL) |
-		((vm_flags & VM_PKEY_BIT1) ? H_PTE_PKEY_BIT3 : 0x0UL) |
+	return (((vm_flags & VM_PKEY_BIT0) ? H_PTE_PKEY_BIT0 : 0x0UL) |
+		((vm_flags & VM_PKEY_BIT1) ? H_PTE_PKEY_BIT1 : 0x0UL) |
 		((vm_flags & VM_PKEY_BIT2) ? H_PTE_PKEY_BIT2 : 0x0UL) |
-		((vm_flags & VM_PKEY_BIT3) ? H_PTE_PKEY_BIT1 : 0x0UL) |
-		((vm_flags & VM_PKEY_BIT4) ? H_PTE_PKEY_BIT0 : 0x0UL));
+		((vm_flags & VM_PKEY_BIT3) ? H_PTE_PKEY_BIT3 : 0x0UL) |
+		((vm_flags & VM_PKEY_BIT4) ? H_PTE_PKEY_BIT4 : 0x0UL));
 }
 
 static inline int vma_pkey(struct vm_area_struct *vma)
@@ -53,20 +53,20 @@ static inline int vma_pkey(struct vm_area_struct *vma)
 
 static inline u64 pte_to_hpte_pkey_bits(u64 pteflags)
 {
-	return (((pteflags & H_PTE_PKEY_BIT0) ? HPTE_R_KEY_BIT0 : 0x0UL) |
-		((pteflags & H_PTE_PKEY_BIT1) ? HPTE_R_KEY_BIT1 : 0x0UL) |
-		((pteflags & H_PTE_PKEY_BIT2) ? HPTE_R_KEY_BIT2 : 0x0UL) |
+	return (((pteflags & H_PTE_PKEY_BIT4) ? HPTE_R_KEY_BIT4 : 0x0UL) |
 		((pteflags & H_PTE_PKEY_BIT3) ? HPTE_R_KEY_BIT3 : 0x0UL) |
-		((pteflags & H_PTE_PKEY_BIT4) ? HPTE_R_KEY_BIT4 : 0x0UL));
+		((pteflags & H_PTE_PKEY_BIT2) ? HPTE_R_KEY_BIT2 : 0x0UL) |
+		((pteflags & H_PTE_PKEY_BIT1) ? HPTE_R_KEY_BIT1 : 0x0UL) |
+		((pteflags & H_PTE_PKEY_BIT0) ? HPTE_R_KEY_BIT0 : 0x0UL));
 }
 
 static inline u16 pte_to_pkey_bits(u64 pteflags)
 {
-	return (((pteflags & H_PTE_PKEY_BIT0) ? 0x10 : 0x0UL) |
-		((pteflags & H_PTE_PKEY_BIT1) ? 0x8 : 0x0UL) |
+	return (((pteflags & H_PTE_PKEY_BIT4) ? 0x10 : 0x0UL) |
+		((pteflags & H_PTE_PKEY_BIT3) ? 0x8 : 0x0UL) |
 		((pteflags & H_PTE_PKEY_BIT2) ? 0x4 : 0x0UL) |
-		((pteflags & H_PTE_PKEY_BIT3) ? 0x2 : 0x0UL) |
-		((pteflags & H_PTE_PKEY_BIT4) ? 0x1 : 0x0UL));
+		((pteflags & H_PTE_PKEY_BIT1) ? 0x2 : 0x0UL) |
+		((pteflags & H_PTE_PKEY_BIT0) ? 0x1 : 0x0UL));
 }
 
 #define pkey_alloc_mask(pkey) (0x1 << pkey)
-- 
2.26.2

[PATCH v4 02/41] powerpc/book3s64/pkeys: pkeys are supported only on hash on book3s.

[Aneesh Kumar K.V]

Move them to hash specific file and add BUG() for radix path.
---
 .../powerpc/include/asm/book3s/64/hash-pkey.h | 32 ++++++++++++++++
 arch/powerpc/include/asm/book3s/64/pkeys.h    | 25 +++++++++++++
 arch/powerpc/include/asm/pkeys.h              | 37 ++++---------------
 3 files changed, 64 insertions(+), 30 deletions(-)
 create mode 100644 arch/powerpc/include/asm/book3s/64/hash-pkey.h
 create mode 100644 arch/powerpc/include/asm/book3s/64/pkeys.h

diff --git a/arch/powerpc/include/asm/book3s/64/hash-pkey.h b/arch/powerpc/include/asm/book3s/64/hash-pkey.h
new file mode 100644
index 000000000000..795010897e5d
--- /dev/null
+++ b/arch/powerpc/include/asm/book3s/64/hash-pkey.h
@@ -0,0 +1,32 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+#ifndef _ASM_POWERPC_BOOK3S_64_HASH_PKEY_H
+#define _ASM_POWERPC_BOOK3S_64_HASH_PKEY_H
+
+static inline u64 hash__vmflag_to_pte_pkey_bits(u64 vm_flags)
+{
+	return (((vm_flags & VM_PKEY_BIT0) ? H_PTE_PKEY_BIT0 : 0x0UL) |
+		((vm_flags & VM_PKEY_BIT1) ? H_PTE_PKEY_BIT1 : 0x0UL) |
+		((vm_flags & VM_PKEY_BIT2) ? H_PTE_PKEY_BIT2 : 0x0UL) |
+		((vm_flags & VM_PKEY_BIT3) ? H_PTE_PKEY_BIT3 : 0x0UL) |
+		((vm_flags & VM_PKEY_BIT4) ? H_PTE_PKEY_BIT4 : 0x0UL));
+}
+
+static inline u64 pte_to_hpte_pkey_bits(u64 pteflags)
+{
+	return (((pteflags & H_PTE_PKEY_BIT4) ? HPTE_R_KEY_BIT4 : 0x0UL) |
+		((pteflags & H_PTE_PKEY_BIT3) ? HPTE_R_KEY_BIT3 : 0x0UL) |
+		((pteflags & H_PTE_PKEY_BIT2) ? HPTE_R_KEY_BIT2 : 0x0UL) |
+		((pteflags & H_PTE_PKEY_BIT1) ? HPTE_R_KEY_BIT1 : 0x0UL) |
+		((pteflags & H_PTE_PKEY_BIT0) ? HPTE_R_KEY_BIT0 : 0x0UL));
+}
+
+static inline u16 hash__pte_to_pkey_bits(u64 pteflags)
+{
+	return (((pteflags & H_PTE_PKEY_BIT4) ? 0x10 : 0x0UL) |
+		((pteflags & H_PTE_PKEY_BIT3) ? 0x8 : 0x0UL) |
+		((pteflags & H_PTE_PKEY_BIT2) ? 0x4 : 0x0UL) |
+		((pteflags & H_PTE_PKEY_BIT1) ? 0x2 : 0x0UL) |
+		((pteflags & H_PTE_PKEY_BIT0) ? 0x1 : 0x0UL));
+}
+
+#endif
diff --git a/arch/powerpc/include/asm/book3s/64/pkeys.h b/arch/powerpc/include/asm/book3s/64/pkeys.h
new file mode 100644
index 000000000000..8174662a9173
--- /dev/null
+++ b/arch/powerpc/include/asm/book3s/64/pkeys.h
@@ -0,0 +1,25 @@
+/* SPDX-License-Identifier: GPL-2.0+ */
+
+#ifndef _ASM_POWERPC_BOOK3S_64_PKEYS_H
+#define _ASM_POWERPC_BOOK3S_64_PKEYS_H
+
+#include <asm/book3s/64/hash-pkey.h>
+
+static inline u64 vmflag_to_pte_pkey_bits(u64 vm_flags)
+{
+	if (static_branch_likely(&pkey_disabled))
+		return 0x0UL;
+
+	if (radix_enabled())
+		BUG();
+	return hash__vmflag_to_pte_pkey_bits(vm_flags);
+}
+
+static inline u16 pte_to_pkey_bits(u64 pteflags)
+{
+	if (radix_enabled())
+		BUG();
+	return hash__pte_to_pkey_bits(pteflags);
+}
+
+#endif /*_ASM_POWERPC_KEYS_H */
diff --git a/arch/powerpc/include/asm/pkeys.h b/arch/powerpc/include/asm/pkeys.h
index f8f4d0793789..5dd0a79d1809 100644
--- a/arch/powerpc/include/asm/pkeys.h
+++ b/arch/powerpc/include/asm/pkeys.h
@@ -25,23 +25,18 @@ extern u32 reserved_allocation_mask; /* bits set for reserved keys */
 				PKEY_DISABLE_WRITE  | \
 				PKEY_DISABLE_EXECUTE)
 
+#ifdef CONFIG_PPC_BOOK3S_64
+#include <asm/book3s/64/pkeys.h>
+#else
+#error "Not supported"
+#endif
+
+
 static inline u64 pkey_to_vmflag_bits(u16 pkey)
 {
 	return (((u64)pkey << VM_PKEY_SHIFT) & ARCH_VM_PKEY_FLAGS);
 }
 
-static inline u64 vmflag_to_pte_pkey_bits(u64 vm_flags)
-{
-	if (static_branch_likely(&pkey_disabled))
-		return 0x0UL;
-
-	return (((vm_flags & VM_PKEY_BIT0) ? H_PTE_PKEY_BIT0 : 0x0UL) |
-		((vm_flags & VM_PKEY_BIT1) ? H_PTE_PKEY_BIT1 : 0x0UL) |
-		((vm_flags & VM_PKEY_BIT2) ? H_PTE_PKEY_BIT2 : 0x0UL) |
-		((vm_flags & VM_PKEY_BIT3) ? H_PTE_PKEY_BIT3 : 0x0UL) |
-		((vm_flags & VM_PKEY_BIT4) ? H_PTE_PKEY_BIT4 : 0x0UL));
-}
-
 static inline int vma_pkey(struct vm_area_struct *vma)
 {
 	if (static_branch_likely(&pkey_disabled))
@@ -51,24 +46,6 @@ static inline int vma_pkey(struct vm_area_struct *vma)
 
 #define arch_max_pkey() pkeys_total
 
-static inline u64 pte_to_hpte_pkey_bits(u64 pteflags)
-{
-	return (((pteflags & H_PTE_PKEY_BIT4) ? HPTE_R_KEY_BIT4 : 0x0UL) |
-		((pteflags & H_PTE_PKEY_BIT3) ? HPTE_R_KEY_BIT3 : 0x0UL) |
-		((pteflags & H_PTE_PKEY_BIT2) ? HPTE_R_KEY_BIT2 : 0x0UL) |
-		((pteflags & H_PTE_PKEY_BIT1) ? HPTE_R_KEY_BIT1 : 0x0UL) |
-		((pteflags & H_PTE_PKEY_BIT0) ? HPTE_R_KEY_BIT0 : 0x0UL));
-}
-
-static inline u16 pte_to_pkey_bits(u64 pteflags)
-{
-	return (((pteflags & H_PTE_PKEY_BIT4) ? 0x10 : 0x0UL) |
-		((pteflags & H_PTE_PKEY_BIT3) ? 0x8 : 0x0UL) |
-		((pteflags & H_PTE_PKEY_BIT2) ? 0x4 : 0x0UL) |
-		((pteflags & H_PTE_PKEY_BIT1) ? 0x2 : 0x0UL) |
-		((pteflags & H_PTE_PKEY_BIT0) ? 0x1 : 0x0UL));
-}
-
 #define pkey_alloc_mask(pkey) (0x1 << pkey)
 
 #define mm_pkey_allocation_map(mm) (mm->context.pkey_allocation_map)
-- 
2.26.2

[PATCH v4 03/41] powerpc/book3s64/pkeys: Move pkey related bits in the linux page table

[Aneesh Kumar K.V]

To keep things simple, all the pkey related bits are kept together
in linux page table for 64K config with hash translation. With hash-4k
kernel requires 4 bits to store slots details. This is done by overloading
some of the RPN bits for storing the slot details. Due to this PKEY_BIT0 on
the 4K config is used for storing hash slot details.

64K before

|....|RSV1| RSV2| RSV3 | RSV4 | RPN44| RPN43   |.... | RSV5|
|....| P4 |  P3 |  P2  |  P1  | Busy | HASHPTE |.... |  P0 |

after

|....|RSV1| RSV2| RSV3 | RSV4 | RPN44 | RPN43   |.... | RSV5 |
|....| P4 |  P3 |  P2  |  P1  | P0    | HASHPTE |.... | Busy |

4k before

|....| RSV1 | RSV2     | RSV3 | RSV4 | RPN44| RPN43.... | RSV5|
|....| Busy |  HASHPTE |  P2  |  P1  | F_SEC| F_GIX.... |  P0 |

after

|....| RSV1    | RSV2| RSV3 | RSV4 | Free | RPN43.... | RSV5 |
|....| HASHPTE |  P2 |  P1  |  P0  | F_SEC| F_GIX.... | BUSY |

Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
---
 arch/powerpc/include/asm/book3s/64/hash-4k.h  | 16 ++++++++--------
 arch/powerpc/include/asm/book3s/64/hash-64k.h | 12 ++++++------
 arch/powerpc/include/asm/book3s/64/pgtable.h  | 17 ++++++++---------
 3 files changed, 22 insertions(+), 23 deletions(-)

diff --git a/arch/powerpc/include/asm/book3s/64/hash-4k.h b/arch/powerpc/include/asm/book3s/64/hash-4k.h
index f889d56bf8cf..082b98808701 100644
--- a/arch/powerpc/include/asm/book3s/64/hash-4k.h
+++ b/arch/powerpc/include/asm/book3s/64/hash-4k.h
@@ -34,11 +34,11 @@
 #define H_PUD_TABLE_SIZE	(sizeof(pud_t) << H_PUD_INDEX_SIZE)
 #define H_PGD_TABLE_SIZE	(sizeof(pgd_t) << H_PGD_INDEX_SIZE)
 
-#define H_PAGE_F_GIX_SHIFT	53
-#define H_PAGE_F_SECOND	_RPAGE_RPN44	/* HPTE is in 2ndary HPTEG */
-#define H_PAGE_F_GIX	(_RPAGE_RPN43 | _RPAGE_RPN42 | _RPAGE_RPN41)
-#define H_PAGE_BUSY	_RPAGE_RSV1     /* software: PTE & hash are busy */
-#define H_PAGE_HASHPTE	_RPAGE_RSV2     /* software: PTE & hash are busy */
+#define H_PAGE_F_GIX_SHIFT	_PAGE_PA_MAX
+#define H_PAGE_F_SECOND		_RPAGE_PKEY_BIT0 /* HPTE is in 2ndary HPTEG */
+#define H_PAGE_F_GIX		(_RPAGE_RPN43 | _RPAGE_RPN42 | _RPAGE_RPN41)
+#define H_PAGE_BUSY		_RPAGE_RSV1
+#define H_PAGE_HASHPTE		_RPAGE_PKEY_BIT4
 
 /* PTE flags to conserve for HPTE identification */
 #define _PAGE_HPTEFLAGS (H_PAGE_BUSY | H_PAGE_HASHPTE | \
@@ -59,9 +59,9 @@
 /* memory key bits, only 8 keys supported */
 #define H_PTE_PKEY_BIT4	0
 #define H_PTE_PKEY_BIT3	0
-#define H_PTE_PKEY_BIT2	_RPAGE_RSV3
-#define H_PTE_PKEY_BIT1	_RPAGE_RSV4
-#define H_PTE_PKEY_BIT0	_RPAGE_RSV5
+#define H_PTE_PKEY_BIT2	_RPAGE_PKEY_BIT3
+#define H_PTE_PKEY_BIT1	_RPAGE_PKEY_BIT2
+#define H_PTE_PKEY_BIT0	_RPAGE_PKEY_BIT1
 
 
 /*
diff --git a/arch/powerpc/include/asm/book3s/64/hash-64k.h b/arch/powerpc/include/asm/book3s/64/hash-64k.h
index 0a15fd14cf72..f20de1149ebe 100644
--- a/arch/powerpc/include/asm/book3s/64/hash-64k.h
+++ b/arch/powerpc/include/asm/book3s/64/hash-64k.h
@@ -32,15 +32,15 @@
  */
 #define H_PAGE_COMBO	_RPAGE_RPN0 /* this is a combo 4k page */
 #define H_PAGE_4K_PFN	_RPAGE_RPN1 /* PFN is for a single 4k page */
-#define H_PAGE_BUSY	_RPAGE_RPN44     /* software: PTE & hash are busy */
+#define H_PAGE_BUSY	_RPAGE_RSV1     /* software: PTE & hash are busy */
 #define H_PAGE_HASHPTE	_RPAGE_RPN43	/* PTE has associated HPTE */
 
 /* memory key bits. */
-#define H_PTE_PKEY_BIT4	_RPAGE_RSV1
-#define H_PTE_PKEY_BIT3	_RPAGE_RSV2
-#define H_PTE_PKEY_BIT2	_RPAGE_RSV3
-#define H_PTE_PKEY_BIT1	_RPAGE_RSV4
-#define H_PTE_PKEY_BIT0	_RPAGE_RSV5
+#define H_PTE_PKEY_BIT4		_RPAGE_PKEY_BIT4
+#define H_PTE_PKEY_BIT3		_RPAGE_PKEY_BIT3
+#define H_PTE_PKEY_BIT2		_RPAGE_PKEY_BIT2
+#define H_PTE_PKEY_BIT1		_RPAGE_PKEY_BIT1
+#define H_PTE_PKEY_BIT0		_RPAGE_PKEY_BIT0
 
 /*
  * We need to differentiate between explicit huge page and THP huge
diff --git a/arch/powerpc/include/asm/book3s/64/pgtable.h b/arch/powerpc/include/asm/book3s/64/pgtable.h
index f17442c3a092..b7c0ba977d6a 100644
--- a/arch/powerpc/include/asm/book3s/64/pgtable.h
+++ b/arch/powerpc/include/asm/book3s/64/pgtable.h
@@ -32,11 +32,13 @@
 #define _RPAGE_SW1		0x00800
 #define _RPAGE_SW2		0x00400
 #define _RPAGE_SW3		0x00200
-#define _RPAGE_RSV1		0x1000000000000000UL
-#define _RPAGE_RSV2		0x0800000000000000UL
-#define _RPAGE_RSV3		0x0400000000000000UL
-#define _RPAGE_RSV4		0x0200000000000000UL
-#define _RPAGE_RSV5		0x00040UL
+#define _RPAGE_RSV1		0x00040UL
+
+#define _RPAGE_PKEY_BIT4	0x1000000000000000UL
+#define _RPAGE_PKEY_BIT3	0x0800000000000000UL
+#define _RPAGE_PKEY_BIT2	0x0400000000000000UL
+#define _RPAGE_PKEY_BIT1	0x0200000000000000UL
+#define _RPAGE_PKEY_BIT0	0x0100000000000000UL
 
 #define _PAGE_PTE		0x4000000000000000UL	/* distinguishes PTEs from pointers */
 #define _PAGE_PRESENT		0x8000000000000000UL	/* pte contains a translation */
@@ -58,13 +60,12 @@
  */
 #define _RPAGE_RPN0		0x01000
 #define _RPAGE_RPN1		0x02000
-#define _RPAGE_RPN44		0x0100000000000000UL
 #define _RPAGE_RPN43		0x0080000000000000UL
 #define _RPAGE_RPN42		0x0040000000000000UL
 #define _RPAGE_RPN41		0x0020000000000000UL
 
 /* Max physical address bit as per radix table */
-#define _RPAGE_PA_MAX		57
+#define _RPAGE_PA_MAX		56
 
 /*
  * Max physical address bit we will use for now.
@@ -125,8 +126,6 @@
 			 _PAGE_ACCESSED | _PAGE_SPECIAL | _PAGE_PTE |	\
 			 _PAGE_SOFT_DIRTY | _PAGE_DEVMAP)
 
-#define H_PTE_PKEY  (H_PTE_PKEY_BIT0 | H_PTE_PKEY_BIT1 | H_PTE_PKEY_BIT2 | \
-		     H_PTE_PKEY_BIT3 | H_PTE_PKEY_BIT4)
 /*
  * We define 2 sets of base prot bits, one for basic pages (ie,
  * cacheable kernel and user pages) and one for non cacheable
-- 
2.26.2

[PATCH v4 04/41] powerpc/book3s64/pkeys: Explain key 1 reservation details

[Aneesh Kumar K.V]

This explains the details w.r.t key 1.

Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
---
 arch/powerpc/mm/book3s64/pkeys.c | 5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)

diff --git a/arch/powerpc/mm/book3s64/pkeys.c b/arch/powerpc/mm/book3s64/pkeys.c
index 1199fc2bfaec..d60e6bfa3e03 100644
--- a/arch/powerpc/mm/book3s64/pkeys.c
+++ b/arch/powerpc/mm/book3s64/pkeys.c
@@ -124,7 +124,10 @@ static int pkey_initialize(void)
 #else
 	os_reserved = 0;
 #endif
-	/* Bits are in LE format. */
+	/*
+	 * key 1 is recommended not to be used. PowerISA(3.0) page 1015,
+	 * programming note.
+	 */
 	reserved_allocation_mask = (0x1 << 1) | (0x1 << execute_only_key);
 
 	/* register mask is in BE format */
-- 
2.26.2

[PATCH v4 05/41] powerpc/book3s64/pkeys: Simplify the key initialization

[Aneesh Kumar K.V]

Add documentation explaining the execute_only_key. The reservation and initialization mask
details are also explained in this patch.

No functional change in this patch.

Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
---
 arch/powerpc/mm/book3s64/pkeys.c | 186 ++++++++++++++++++-------------
 1 file changed, 107 insertions(+), 79 deletions(-)

diff --git a/arch/powerpc/mm/book3s64/pkeys.c b/arch/powerpc/mm/book3s64/pkeys.c
index d60e6bfa3e03..3db0b3cfc322 100644
--- a/arch/powerpc/mm/book3s64/pkeys.c
+++ b/arch/powerpc/mm/book3s64/pkeys.c
@@ -15,48 +15,71 @@
 DEFINE_STATIC_KEY_TRUE(pkey_disabled);
 int  pkeys_total;		/* Total pkeys as per device tree */
 u32  initial_allocation_mask;   /* Bits set for the initially allocated keys */
-u32  reserved_allocation_mask;  /* Bits set for reserved keys */
+/*
+ *  Keys marked in the reservation list cannot be allocated by  userspace
+ */
+u32  reserved_allocation_mask;
 static bool pkey_execute_disable_supported;
-static bool pkeys_devtree_defined;	/* property exported by device tree */
-static u64 pkey_amr_mask;		/* Bits in AMR not to be touched */
-static u64 pkey_iamr_mask;		/* Bits in AMR not to be touched */
-static u64 pkey_uamor_mask;		/* Bits in UMOR not to be touched */
+static u64 default_amr;
+static u64 default_iamr;
+/* Allow all keys to be modified by default */
+static u64 default_uamor = ~0x0UL;
+/*
+ * Key used to implement PROT_EXEC mmap. Denies READ/WRITE
+ * We pick key 2 because 0 is special key and 1 is reserved as per ISA.
+ */
 static int execute_only_key = 2;
 
+
 #define AMR_BITS_PER_PKEY 2
 #define AMR_RD_BIT 0x1UL
 #define AMR_WR_BIT 0x2UL
 #define IAMR_EX_BIT 0x1UL
-#define PKEY_REG_BITS (sizeof(u64)*8)
+#define PKEY_REG_BITS (sizeof(u64) * 8)
 #define pkeyshift(pkey) (PKEY_REG_BITS - ((pkey+1) * AMR_BITS_PER_PKEY))
 
-static void scan_pkey_feature(void)
+static int scan_pkey_feature(void)
 {
 	u32 vals[2];
+	int pkeys_total = 0;
 	struct device_node *cpu;
 
+	/*
+	 * Pkey is not supported with Radix translation.
+	 */
+	if (radix_enabled())
+		return 0;
+
 	cpu = of_find_node_by_type(NULL, "cpu");
 	if (!cpu)
-		return;
+		return 0;
 
 	if (of_property_read_u32_array(cpu,
-			"ibm,processor-storage-keys", vals, 2))
-		return;
+				       "ibm,processor-storage-keys", vals, 2) == 0) {
+		/*
+		 * Since any pkey can be used for data or execute, we will
+		 * just treat all keys as equal and track them as one entity.
+		 */
+		pkeys_total = vals[0];
+		/*  Should we check for IAMR support FIXME!! */
+	} else {
+		/*
+		 * Let's assume 32 pkeys on P8 bare metal, if its not defined by device
+		 * tree. We make this exception since skiboot forgot to expose this
+		 * property on power8.
+		 */
+		if (!firmware_has_feature(FW_FEATURE_LPAR) &&
+		    cpu_has_feature(CPU_FTRS_POWER8))
+			pkeys_total = 32;
+	}
 
 	/*
-	 * Since any pkey can be used for data or execute, we will just treat
-	 * all keys as equal and track them as one entity.
+	 * Adjust the upper limit, based on the number of bits supported by
+	 * arch-neutral code.
 	 */
-	pkeys_total = vals[0];
-	pkeys_devtree_defined = true;
-}
-
-static inline bool pkey_mmu_enabled(void)
-{
-	if (firmware_has_feature(FW_FEATURE_LPAR))
-		return pkeys_total;
-	else
-		return cpu_has_feature(CPU_FTR_PKEY);
+	pkeys_total = min_t(int, pkeys_total,
+			    ((ARCH_VM_PKEY_FLAGS >> VM_PKEY_SHIFT) + 1));
+	return pkeys_total;
 }
 
 static int pkey_initialize(void)
@@ -80,31 +103,13 @@ static int pkey_initialize(void)
 				!= (sizeof(u64) * BITS_PER_BYTE));
 
 	/* scan the device tree for pkey feature */
-	scan_pkey_feature();
-
-	/*
-	 * Let's assume 32 pkeys on P8 bare metal, if its not defined by device
-	 * tree. We make this exception since skiboot forgot to expose this
-	 * property on power8.
-	 */
-	if (!pkeys_devtree_defined && !firmware_has_feature(FW_FEATURE_LPAR) &&
-			cpu_has_feature(CPU_FTRS_POWER8))
-		pkeys_total = 32;
-
-	/*
-	 * Adjust the upper limit, based on the number of bits supported by
-	 * arch-neutral code.
-	 */
-	pkeys_total = min_t(int, pkeys_total,
-			((ARCH_VM_PKEY_FLAGS >> VM_PKEY_SHIFT)+1));
-
-	if (!pkey_mmu_enabled() || radix_enabled() || !pkeys_total)
-		static_branch_enable(&pkey_disabled);
-	else
+	pkeys_total = scan_pkey_feature();
+	if (pkeys_total)
 		static_branch_disable(&pkey_disabled);
-
-	if (static_branch_likely(&pkey_disabled))
+	else {
+		static_branch_enable(&pkey_disabled);
 		return 0;
+	}
 
 	/*
 	 * The device tree cannot be relied to indicate support for
@@ -118,48 +123,71 @@ static int pkey_initialize(void)
 #ifdef CONFIG_PPC_4K_PAGES
 	/*
 	 * The OS can manage only 8 pkeys due to its inability to represent them
-	 * in the Linux 4K PTE.
+	 * in the Linux 4K PTE. Mark all other keys reserved.
 	 */
 	os_reserved = pkeys_total - 8;
 #else
 	os_reserved = 0;
 #endif
-	/*
-	 * key 1 is recommended not to be used. PowerISA(3.0) page 1015,
-	 * programming note.
-	 */
-	reserved_allocation_mask = (0x1 << 1) | (0x1 << execute_only_key);
-
-	/* register mask is in BE format */
-	pkey_amr_mask = ~0x0ul;
-	pkey_amr_mask &= ~(0x3ul << pkeyshift(0));
-
-	pkey_iamr_mask = ~0x0ul;
-	pkey_iamr_mask &= ~(0x3ul << pkeyshift(0));
-	pkey_iamr_mask &= ~(0x3ul << pkeyshift(execute_only_key));
-
-	pkey_uamor_mask = ~0x0ul;
-	pkey_uamor_mask &= ~(0x3ul << pkeyshift(0));
-	pkey_uamor_mask &= ~(0x3ul << pkeyshift(execute_only_key));
-
-	/* mark the rest of the keys as reserved and hence unavailable */
-	for (i = (pkeys_total - os_reserved); i < pkeys_total; i++) {
-		reserved_allocation_mask |= (0x1 << i);
-		pkey_uamor_mask &= ~(0x3ul << pkeyshift(i));
-	}
-	initial_allocation_mask = reserved_allocation_mask | (0x1 << 0);
 
 	if (unlikely((pkeys_total - os_reserved) <= execute_only_key)) {
 		/*
 		 * Insufficient number of keys to support
 		 * execute only key. Mark it unavailable.
-		 * Any AMR, UAMOR, IAMR bit set for
-		 * this key is irrelevant since this key
-		 * can never be allocated.
 		 */
 		execute_only_key = -1;
+	} else {
+		/*
+		 * Mark the execute_only_pkey as not available for
+		 * user allocation via pkey_alloc.
+		 */
+		reserved_allocation_mask |= (0x1 << execute_only_key);
+
+		/*
+		 * Deny READ/WRITE for execute_only_key.
+		 * Allow execute in IAMR.
+		 */
+		default_amr  |= (0x3ul << pkeyshift(execute_only_key));
+		default_iamr &= ~(0x3ul << pkeyshift(execute_only_key));
+
+		/*
+		 * Clear the uamor bits for this key.
+		 */
+		default_uamor &= ~(0x3ul << pkeyshift(execute_only_key));
 	}
 
+	/*
+	 * Allow access for only key 0. And prevent any other modification.
+	 */
+	default_amr   &= ~(0x3ul << pkeyshift(0));
+	default_iamr  &= ~(0x3ul << pkeyshift(0));
+	default_uamor &= ~(0x3ul << pkeyshift(0));
+	/*
+	 * key 0 is special in that we want to consider it an allocated
+	 * key which is preallocated. We don't allow changing AMR bits
+	 * w.r.t key 0. But one can pkey_free(key0)
+	 */
+	initial_allocation_mask |= (0x1 << 0);
+
+	/*
+	 * key 1 is recommended not to be used. PowerISA(3.0) page 1015,
+	 * programming note.
+	 */
+	reserved_allocation_mask |= (0x1 << 1);
+
+	/*
+	 * Prevent the usage of OS reserved the keys. Update UAMOR
+	 * for those keys.
+	 */
+	for (i = (pkeys_total - os_reserved); i < pkeys_total; i++) {
+		reserved_allocation_mask |= (0x1 << i);
+		default_uamor &= ~(0x3ul << pkeyshift(i));
+	}
+	/*
+	 * Prevent the allocation of reserved keys too.
+	 */
+	initial_allocation_mask |= reserved_allocation_mask;
+
 	return 0;
 }
 
@@ -301,13 +329,13 @@ void thread_pkey_regs_init(struct thread_struct *thread)
 	if (static_branch_likely(&pkey_disabled))
 		return;
 
-	thread->amr = pkey_amr_mask;
-	thread->iamr = pkey_iamr_mask;
-	thread->uamor = pkey_uamor_mask;
+	thread->amr   = default_amr;
+	thread->iamr  = default_iamr;
+	thread->uamor = default_uamor;
 
-	write_uamor(pkey_uamor_mask);
-	write_amr(pkey_amr_mask);
-	write_iamr(pkey_iamr_mask);
+	write_amr(default_amr);
+	write_iamr(default_iamr);
+	write_uamor(default_uamor);
 }
 
 int __execute_only_pkey(struct mm_struct *mm)
-- 
2.26.2

[PATCH v4 06/41] powerpc/book3s64/pkeys: Prevent key 1 modification from userspace.

[Aneesh Kumar K.V]

Key 1 is marked reserved by ISA. Setup uamor to prevent userspace modification
of the same.

Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
---
 arch/powerpc/mm/book3s64/pkeys.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/arch/powerpc/mm/book3s64/pkeys.c b/arch/powerpc/mm/book3s64/pkeys.c
index 3db0b3cfc322..9e68a08799ee 100644
--- a/arch/powerpc/mm/book3s64/pkeys.c
+++ b/arch/powerpc/mm/book3s64/pkeys.c
@@ -174,6 +174,7 @@ static int pkey_initialize(void)
 	 * programming note.
 	 */
 	reserved_allocation_mask |= (0x1 << 1);
+	default_uamor &= ~(0x3ul << pkeyshift(1));
 
 	/*
 	 * Prevent the usage of OS reserved the keys. Update UAMOR
-- 
2.26.2

[PATCH v4 07/41] powerpc/book3s64/pkeys: kill cpu feature key CPU_FTR_PKEY

[Aneesh Kumar K.V]

We don't use CPU_FTR_PKEY anymore. Remove the feature bit and mark it
free.

Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
---
 arch/powerpc/include/asm/cputable.h | 13 ++++++-------
 arch/powerpc/kernel/dt_cpu_ftrs.c   |  6 ------
 2 files changed, 6 insertions(+), 13 deletions(-)

diff --git a/arch/powerpc/include/asm/cputable.h b/arch/powerpc/include/asm/cputable.h
index bac2252c839e..dd0a2e77a695 100644
--- a/arch/powerpc/include/asm/cputable.h
+++ b/arch/powerpc/include/asm/cputable.h
@@ -198,7 +198,7 @@ static inline void cpu_feature_keys_init(void) { }
 #define CPU_FTR_STCX_CHECKS_ADDRESS	LONG_ASM_CONST(0x0000000080000000)
 #define CPU_FTR_POPCNTB			LONG_ASM_CONST(0x0000000100000000)
 #define CPU_FTR_POPCNTD			LONG_ASM_CONST(0x0000000200000000)
-#define CPU_FTR_PKEY			LONG_ASM_CONST(0x0000000400000000)
+/* LONG_ASM_CONST(0x0000000400000000) Free */
 #define CPU_FTR_VMX_COPY		LONG_ASM_CONST(0x0000000800000000)
 #define CPU_FTR_TM			LONG_ASM_CONST(0x0000001000000000)
 #define CPU_FTR_CFAR			LONG_ASM_CONST(0x0000002000000000)
@@ -438,7 +438,7 @@ static inline void cpu_feature_keys_init(void) { }
 	    CPU_FTR_DSCR | CPU_FTR_SAO  | CPU_FTR_ASYM_SMT | \
 	    CPU_FTR_STCX_CHECKS_ADDRESS | CPU_FTR_POPCNTB | CPU_FTR_POPCNTD | \
 	    CPU_FTR_CFAR | CPU_FTR_HVMODE | \
-	    CPU_FTR_VMX_COPY | CPU_FTR_HAS_PPR | CPU_FTR_DABRX | CPU_FTR_PKEY)
+	    CPU_FTR_VMX_COPY | CPU_FTR_HAS_PPR | CPU_FTR_DABRX )
 #define CPU_FTRS_POWER8 (CPU_FTR_LWSYNC | \
 	    CPU_FTR_PPCAS_ARCH_V2 | CPU_FTR_CTRL | CPU_FTR_ARCH_206 |\
 	    CPU_FTR_MMCRA | CPU_FTR_SMT | \
@@ -448,7 +448,7 @@ static inline void cpu_feature_keys_init(void) { }
 	    CPU_FTR_STCX_CHECKS_ADDRESS | CPU_FTR_POPCNTB | CPU_FTR_POPCNTD | \
 	    CPU_FTR_CFAR | CPU_FTR_HVMODE | CPU_FTR_VMX_COPY | \
 	    CPU_FTR_DBELL | CPU_FTR_HAS_PPR | CPU_FTR_DAWR | \
-	    CPU_FTR_ARCH_207S | CPU_FTR_TM_COMP | CPU_FTR_PKEY)
+	    CPU_FTR_ARCH_207S | CPU_FTR_TM_COMP )
 #define CPU_FTRS_POWER8E (CPU_FTRS_POWER8 | CPU_FTR_PMAO_BUG)
 #define CPU_FTRS_POWER9 (CPU_FTR_LWSYNC | \
 	    CPU_FTR_PPCAS_ARCH_V2 | CPU_FTR_CTRL | CPU_FTR_ARCH_206 |\
@@ -459,8 +459,8 @@ static inline void cpu_feature_keys_init(void) { }
 	    CPU_FTR_STCX_CHECKS_ADDRESS | CPU_FTR_POPCNTB | CPU_FTR_POPCNTD | \
 	    CPU_FTR_CFAR | CPU_FTR_HVMODE | CPU_FTR_VMX_COPY | \
 	    CPU_FTR_DBELL | CPU_FTR_HAS_PPR | CPU_FTR_ARCH_207S | \
-	    CPU_FTR_TM_COMP | CPU_FTR_ARCH_300 | CPU_FTR_PKEY | \
-	    CPU_FTR_P9_TLBIE_STQ_BUG | CPU_FTR_P9_TLBIE_ERAT_BUG | CPU_FTR_P9_TIDR)
+	    CPU_FTR_TM_COMP | CPU_FTR_ARCH_300 | CPU_FTR_P9_TLBIE_STQ_BUG | \
+	    CPU_FTR_P9_TLBIE_ERAT_BUG | CPU_FTR_P9_TIDR)
 #define CPU_FTRS_POWER9_DD2_0 (CPU_FTRS_POWER9 | CPU_FTR_P9_RADIX_PREFETCH_BUG)
 #define CPU_FTRS_POWER9_DD2_1 (CPU_FTRS_POWER9 | \
 			       CPU_FTR_P9_RADIX_PREFETCH_BUG | \
@@ -477,8 +477,7 @@ static inline void cpu_feature_keys_init(void) { }
 	    CPU_FTR_STCX_CHECKS_ADDRESS | CPU_FTR_POPCNTB | CPU_FTR_POPCNTD | \
 	    CPU_FTR_CFAR | CPU_FTR_HVMODE | CPU_FTR_VMX_COPY | \
 	    CPU_FTR_DBELL | CPU_FTR_HAS_PPR | CPU_FTR_ARCH_207S | \
-	    CPU_FTR_TM_COMP | CPU_FTR_ARCH_300 | CPU_FTR_PKEY | \
-	    CPU_FTR_ARCH_31)
+	    CPU_FTR_TM_COMP | CPU_FTR_ARCH_300 | CPU_FTR_ARCH_31)
 #define CPU_FTRS_CELL	(CPU_FTR_LWSYNC | \
 	    CPU_FTR_PPCAS_ARCH_V2 | CPU_FTR_CTRL | \
 	    CPU_FTR_ALTIVEC_COMP | CPU_FTR_MMCRA | CPU_FTR_SMT | \
diff --git a/arch/powerpc/kernel/dt_cpu_ftrs.c b/arch/powerpc/kernel/dt_cpu_ftrs.c
index 3a409517c031..0acec481d4d1 100644
--- a/arch/powerpc/kernel/dt_cpu_ftrs.c
+++ b/arch/powerpc/kernel/dt_cpu_ftrs.c
@@ -776,12 +776,6 @@ static __init void cpufeatures_cpu_quirks(void)
 	}
 
 	update_tlbie_feature_flag(version);
-	/*
-	 * PKEY was not in the initial base or feature node
-	 * specification, but it should become optional in the next
-	 * cpu feature version sequence.
-	 */
-	cur_cpu_spec->cpu_features |= CPU_FTR_PKEY;
 }
 
 static void __init cpufeatures_setup_finished(void)
-- 
2.26.2

[PATCH v4 08/41] powerpc/book3s64/pkeys: Convert execute key support to static key

[Aneesh Kumar K.V]

Convert the bool to a static key like pkey_disabled.

Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
---
 arch/powerpc/mm/book3s64/pkeys.c | 12 +++++-------
 1 file changed, 5 insertions(+), 7 deletions(-)

diff --git a/arch/powerpc/mm/book3s64/pkeys.c b/arch/powerpc/mm/book3s64/pkeys.c
index 9e68a08799ee..7d400d5a4076 100644
--- a/arch/powerpc/mm/book3s64/pkeys.c
+++ b/arch/powerpc/mm/book3s64/pkeys.c
@@ -13,13 +13,13 @@
 #include <linux/of_device.h>
 
 DEFINE_STATIC_KEY_TRUE(pkey_disabled);
+DEFINE_STATIC_KEY_FALSE(execute_pkey_disabled);
 int  pkeys_total;		/* Total pkeys as per device tree */
 u32  initial_allocation_mask;   /* Bits set for the initially allocated keys */
 /*
  *  Keys marked in the reservation list cannot be allocated by  userspace
  */
 u32  reserved_allocation_mask;
-static bool pkey_execute_disable_supported;
 static u64 default_amr;
 static u64 default_iamr;
 /* Allow all keys to be modified by default */
@@ -116,9 +116,7 @@ static int pkey_initialize(void)
 	 * execute_disable support. Instead we use a PVR check.
 	 */
 	if (pvr_version_is(PVR_POWER7) || pvr_version_is(PVR_POWER7p))
-		pkey_execute_disable_supported = false;
-	else
-		pkey_execute_disable_supported = true;
+		static_branch_enable(&execute_pkey_disabled);
 
 #ifdef CONFIG_PPC_4K_PAGES
 	/*
@@ -214,7 +212,7 @@ static inline void write_amr(u64 value)
 
 static inline u64 read_iamr(void)
 {
-	if (!likely(pkey_execute_disable_supported))
+	if (static_branch_unlikely(&execute_pkey_disabled))
 		return 0x0UL;
 
 	return mfspr(SPRN_IAMR);
@@ -222,7 +220,7 @@ static inline u64 read_iamr(void)
 
 static inline void write_iamr(u64 value)
 {
-	if (!likely(pkey_execute_disable_supported))
+	if (static_branch_unlikely(&execute_pkey_disabled))
 		return;
 
 	mtspr(SPRN_IAMR, value);
@@ -282,7 +280,7 @@ int __arch_set_user_pkey_access(struct task_struct *tsk, int pkey,
 		return -EINVAL;
 
 	if (init_val & PKEY_DISABLE_EXECUTE) {
-		if (!pkey_execute_disable_supported)
+		if (static_branch_unlikely(&execute_pkey_disabled))
 			return -EINVAL;
 		new_iamr_bits |= IAMR_EX_BIT;
 	}
-- 
2.26.2

[PATCH v4 09/41] powerpc/book3s64/pkeys: Simplify pkey disable branch

[Aneesh Kumar K.V]

Make the default value FALSE (pkey enabled) and set to TRUE when we
find the total number of keys supported to be zero.

Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
---
 arch/powerpc/include/asm/pkeys.h | 2 +-
 arch/powerpc/mm/book3s64/pkeys.c | 7 +++----
 2 files changed, 4 insertions(+), 5 deletions(-)

diff --git a/arch/powerpc/include/asm/pkeys.h b/arch/powerpc/include/asm/pkeys.h
index 5dd0a79d1809..75d2a2c19c04 100644
--- a/arch/powerpc/include/asm/pkeys.h
+++ b/arch/powerpc/include/asm/pkeys.h
@@ -11,7 +11,7 @@
 #include <linux/jump_label.h>
 #include <asm/firmware.h>
 
-DECLARE_STATIC_KEY_TRUE(pkey_disabled);
+DECLARE_STATIC_KEY_FALSE(pkey_disabled);
 extern int pkeys_total; /* total pkeys as per device tree */
 extern u32 initial_allocation_mask; /*  bits set for the initially allocated keys */
 extern u32 reserved_allocation_mask; /* bits set for reserved keys */
diff --git a/arch/powerpc/mm/book3s64/pkeys.c b/arch/powerpc/mm/book3s64/pkeys.c
index 7d400d5a4076..87d882a9aaf2 100644
--- a/arch/powerpc/mm/book3s64/pkeys.c
+++ b/arch/powerpc/mm/book3s64/pkeys.c
@@ -12,7 +12,7 @@
 #include <linux/pkeys.h>
 #include <linux/of_device.h>
 
-DEFINE_STATIC_KEY_TRUE(pkey_disabled);
+DEFINE_STATIC_KEY_FALSE(pkey_disabled);
 DEFINE_STATIC_KEY_FALSE(execute_pkey_disabled);
 int  pkeys_total;		/* Total pkeys as per device tree */
 u32  initial_allocation_mask;   /* Bits set for the initially allocated keys */
@@ -104,9 +104,8 @@ static int pkey_initialize(void)
 
 	/* scan the device tree for pkey feature */
 	pkeys_total = scan_pkey_feature();
-	if (pkeys_total)
-		static_branch_disable(&pkey_disabled);
-	else {
+	if (!pkeys_total) {
+		/* No support for pkey. Mark it disabled */
 		static_branch_enable(&pkey_disabled);
 		return 0;
 	}
-- 
2.26.2

[PATCH v4 10/41] powerpc/book3s64/pkeys: Convert pkey_total to max_pkey

[Aneesh Kumar K.V]

max_pkey now represents max key value that userspace can allocate.

Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
---
 arch/powerpc/include/asm/pkeys.h |  7 +++++--
 arch/powerpc/mm/book3s64/pkeys.c | 14 +++++++-------
 2 files changed, 12 insertions(+), 9 deletions(-)

diff --git a/arch/powerpc/include/asm/pkeys.h b/arch/powerpc/include/asm/pkeys.h
index 75d2a2c19c04..652bad7334f3 100644
--- a/arch/powerpc/include/asm/pkeys.h
+++ b/arch/powerpc/include/asm/pkeys.h
@@ -12,7 +12,7 @@
 #include <asm/firmware.h>
 
 DECLARE_STATIC_KEY_FALSE(pkey_disabled);
-extern int pkeys_total; /* total pkeys as per device tree */
+extern int max_pkey;
 extern u32 initial_allocation_mask; /*  bits set for the initially allocated keys */
 extern u32 reserved_allocation_mask; /* bits set for reserved keys */
 
@@ -44,7 +44,10 @@ static inline int vma_pkey(struct vm_area_struct *vma)
 	return (vma->vm_flags & ARCH_VM_PKEY_FLAGS) >> VM_PKEY_SHIFT;
 }
 
-#define arch_max_pkey() pkeys_total
+static inline int arch_max_pkey(void)
+{
+	return max_pkey;
+}
 
 #define pkey_alloc_mask(pkey) (0x1 << pkey)
 
diff --git a/arch/powerpc/mm/book3s64/pkeys.c b/arch/powerpc/mm/book3s64/pkeys.c
index 87d882a9aaf2..a4d7287082a8 100644
--- a/arch/powerpc/mm/book3s64/pkeys.c
+++ b/arch/powerpc/mm/book3s64/pkeys.c
@@ -14,7 +14,7 @@
 
 DEFINE_STATIC_KEY_FALSE(pkey_disabled);
 DEFINE_STATIC_KEY_FALSE(execute_pkey_disabled);
-int  pkeys_total;		/* Total pkeys as per device tree */
+int  max_pkey;			/* Maximum key value supported */
 u32  initial_allocation_mask;   /* Bits set for the initially allocated keys */
 /*
  *  Keys marked in the reservation list cannot be allocated by  userspace
@@ -84,7 +84,7 @@ static int scan_pkey_feature(void)
 
 static int pkey_initialize(void)
 {
-	int os_reserved, i;
+	int pkeys_total, i;
 
 	/*
 	 * We define PKEY_DISABLE_EXECUTE in addition to the arch-neutral
@@ -122,12 +122,12 @@ static int pkey_initialize(void)
 	 * The OS can manage only 8 pkeys due to its inability to represent them
 	 * in the Linux 4K PTE. Mark all other keys reserved.
 	 */
-	os_reserved = pkeys_total - 8;
+	max_pkey = min(8, pkeys_total);
 #else
-	os_reserved = 0;
+	max_pkey = pkeys_total;
 #endif
 
-	if (unlikely((pkeys_total - os_reserved) <= execute_only_key)) {
+	if (unlikely(max_pkey <= execute_only_key)) {
 		/*
 		 * Insufficient number of keys to support
 		 * execute only key. Mark it unavailable.
@@ -174,10 +174,10 @@ static int pkey_initialize(void)
 	default_uamor &= ~(0x3ul << pkeyshift(1));
 
 	/*
-	 * Prevent the usage of OS reserved the keys. Update UAMOR
+	 * Prevent the usage of OS reserved keys. Update UAMOR
 	 * for those keys.
 	 */
-	for (i = (pkeys_total - os_reserved); i < pkeys_total; i++) {
+	for (i = max_pkey; i < pkeys_total; i++) {
 		reserved_allocation_mask |= (0x1 << i);
 		default_uamor &= ~(0x3ul << pkeyshift(i));
 	}
-- 
2.26.2

[PATCH v4 11/41] powerpc/book3s64/pkeys: Make initial_allocation_mask static

[Aneesh Kumar K.V]

initial_allocation_mask is not used outside this file.

Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
---
 arch/powerpc/include/asm/pkeys.h | 1 -
 arch/powerpc/mm/book3s64/pkeys.c | 2 +-
 2 files changed, 1 insertion(+), 2 deletions(-)

diff --git a/arch/powerpc/include/asm/pkeys.h b/arch/powerpc/include/asm/pkeys.h
index 652bad7334f3..47c81d41ea9a 100644
--- a/arch/powerpc/include/asm/pkeys.h
+++ b/arch/powerpc/include/asm/pkeys.h
@@ -13,7 +13,6 @@
 
 DECLARE_STATIC_KEY_FALSE(pkey_disabled);
 extern int max_pkey;
-extern u32 initial_allocation_mask; /*  bits set for the initially allocated keys */
 extern u32 reserved_allocation_mask; /* bits set for reserved keys */
 
 #define ARCH_VM_PKEY_FLAGS (VM_PKEY_BIT0 | VM_PKEY_BIT1 | VM_PKEY_BIT2 | \
diff --git a/arch/powerpc/mm/book3s64/pkeys.c b/arch/powerpc/mm/book3s64/pkeys.c
index a4d7287082a8..73b5ef1490c8 100644
--- a/arch/powerpc/mm/book3s64/pkeys.c
+++ b/arch/powerpc/mm/book3s64/pkeys.c
@@ -15,11 +15,11 @@
 DEFINE_STATIC_KEY_FALSE(pkey_disabled);
 DEFINE_STATIC_KEY_FALSE(execute_pkey_disabled);
 int  max_pkey;			/* Maximum key value supported */
-u32  initial_allocation_mask;   /* Bits set for the initially allocated keys */
 /*
  *  Keys marked in the reservation list cannot be allocated by  userspace
  */
 u32  reserved_allocation_mask;
+static u32  initial_allocation_mask;   /* Bits set for the initially allocated keys */
 static u64 default_amr;
 static u64 default_iamr;
 /* Allow all keys to be modified by default */
-- 
2.26.2

[PATCH v4 12/41] powerpc/book3s64/pkeys: Mark all the pkeys above max pkey as reserved

[Aneesh Kumar K.V]

The hypervisor can return less than max allowed pkey (for ex: 31) instead
of 32. We should mark all the pkeys above max allowed as reserved so
that we avoid the allocation of the wrong pkey(for ex: key 31 in the above
case) by userspace.

Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
---
 arch/powerpc/mm/book3s64/pkeys.c | 5 +++--
 1 file changed, 3 insertions(+), 2 deletions(-)

diff --git a/arch/powerpc/mm/book3s64/pkeys.c b/arch/powerpc/mm/book3s64/pkeys.c
index 73b5ef1490c8..0ff59acdbb84 100644
--- a/arch/powerpc/mm/book3s64/pkeys.c
+++ b/arch/powerpc/mm/book3s64/pkeys.c
@@ -175,9 +175,10 @@ static int pkey_initialize(void)
 
 	/*
 	 * Prevent the usage of OS reserved keys. Update UAMOR
-	 * for those keys.
+	 * for those keys. Also mark the rest of the bits in the
+	 * 32 bit mask as reserved.
 	 */
-	for (i = max_pkey; i < pkeys_total; i++) {
+	for (i = max_pkey; i < 32 ; i++) {
 		reserved_allocation_mask |= (0x1 << i);
 		default_uamor &= ~(0x3ul << pkeyshift(i));
 	}
-- 
2.26.2

[PATCH v4 13/41] powerpc/book3s64/pkeys: Enable MMU_FTR_PKEY

[Aneesh Kumar K.V]

Parse storage keys related device tree entry in early_init_devtree
and enable MMU feature MMU_FTR_PKEY if pkeys are supported.

MMU feature is used instead of CPU feature because this enables us
to group MMU_FTR_KUAP and MMU_FTR_PKEY in asm feature fixup code.

Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
---
 arch/powerpc/include/asm/book3s/64/mmu.h |  6 +++
 arch/powerpc/include/asm/mmu.h           |  6 +++
 arch/powerpc/kernel/prom.c               |  5 +++
 arch/powerpc/mm/book3s64/pkeys.c         | 54 ++++++++++++++----------
 4 files changed, 48 insertions(+), 23 deletions(-)

diff --git a/arch/powerpc/include/asm/book3s/64/mmu.h b/arch/powerpc/include/asm/book3s/64/mmu.h
index 5393a535240c..3371ea05b7d3 100644
--- a/arch/powerpc/include/asm/book3s/64/mmu.h
+++ b/arch/powerpc/include/asm/book3s/64/mmu.h
@@ -209,6 +209,12 @@ extern int mmu_io_psize;
 void mmu_early_init_devtree(void);
 void hash__early_init_devtree(void);
 void radix__early_init_devtree(void);
+#ifdef CONFIG_PPC_MEM_KEYS
+void pkey_early_init_devtree(void);
+#else
+static inline void pkey_early_init_devtree(void) {}
+#endif
+
 extern void hash__early_init_mmu(void);
 extern void radix__early_init_mmu(void);
 static inline void __init early_init_mmu(void)
diff --git a/arch/powerpc/include/asm/mmu.h b/arch/powerpc/include/asm/mmu.h
index f4ac25d4df05..72966d3d8f64 100644
--- a/arch/powerpc/include/asm/mmu.h
+++ b/arch/powerpc/include/asm/mmu.h
@@ -23,6 +23,7 @@
 
 /* Radix page table supported and enabled */
 #define MMU_FTR_TYPE_RADIX		ASM_CONST(0x00000040)
+#define MMU_FTR_PKEY			ASM_CONST(0x00000080)
 
 /*
  * Individual features below.
@@ -177,6 +178,9 @@ enum {
 		MMU_FTR_RADIX_KUAP |
 #endif /* CONFIG_PPC_KUAP */
 #endif /* CONFIG_PPC_RADIX_MMU */
+#ifdef CONFIG_PPC_MEM_KEYS
+	MMU_FTR_PKEY |
+#endif
 		0,
 };
 
@@ -356,6 +360,8 @@ extern void setup_initial_memory_limit(phys_addr_t first_memblock_base,
 				       phys_addr_t first_memblock_size);
 static inline void mmu_early_init_devtree(void) { }
 
+static inline void pkey_early_init_devtree(void) {}
+
 extern void *abatron_pteptrs[2];
 #endif /* __ASSEMBLY__ */
 #endif
diff --git a/arch/powerpc/kernel/prom.c b/arch/powerpc/kernel/prom.c
index 6a3bac357e24..6d70797352d8 100644
--- a/arch/powerpc/kernel/prom.c
+++ b/arch/powerpc/kernel/prom.c
@@ -815,6 +815,11 @@ void __init early_init_devtree(void *params)
 	/* Now try to figure out if we are running on LPAR and so on */
 	pseries_probe_fw_features();
 
+	/*
+	 * Initialize pkey features and default AMR/IAMR values
+	 */
+	pkey_early_init_devtree();
+
 #ifdef CONFIG_PPC_PS3
 	/* Identify PS3 firmware */
 	if (of_flat_dt_is_compatible(of_get_flat_dt_root(), "sony,ps3"))
diff --git a/arch/powerpc/mm/book3s64/pkeys.c b/arch/powerpc/mm/book3s64/pkeys.c
index 0ff59acdbb84..bbba9c601e14 100644
--- a/arch/powerpc/mm/book3s64/pkeys.c
+++ b/arch/powerpc/mm/book3s64/pkeys.c
@@ -10,7 +10,8 @@
 #include <asm/mmu.h>
 #include <asm/setup.h>
 #include <linux/pkeys.h>
-#include <linux/of_device.h>
+#include <linux/of_fdt.h>
+
 
 DEFINE_STATIC_KEY_FALSE(pkey_disabled);
 DEFINE_STATIC_KEY_FALSE(execute_pkey_disabled);
@@ -38,38 +39,45 @@ static int execute_only_key = 2;
 #define PKEY_REG_BITS (sizeof(u64) * 8)
 #define pkeyshift(pkey) (PKEY_REG_BITS - ((pkey+1) * AMR_BITS_PER_PKEY))
 
+static int __init dt_scan_storage_keys(unsigned long node,
+				       const char *uname, int depth,
+				       void *data)
+{
+	const char *type = of_get_flat_dt_prop(node, "device_type", NULL);
+	const __be32 *prop;
+	int pkeys_total;
+
+	/* We are scanning "cpu" nodes only */
+	if (type == NULL || strcmp(type, "cpu") != 0)
+		return 0;
+
+	prop = of_get_flat_dt_prop(node, "ibm,processor-storage-keys", NULL);
+	if (!prop)
+		return 0;
+	pkeys_total = be32_to_cpu(prop[0]);
+	return pkeys_total;
+}
+
 static int scan_pkey_feature(void)
 {
-	u32 vals[2];
-	int pkeys_total = 0;
-	struct device_node *cpu;
+	int pkeys_total;
 
 	/*
 	 * Pkey is not supported with Radix translation.
 	 */
-	if (radix_enabled())
+	if (early_radix_enabled())
 		return 0;
 
-	cpu = of_find_node_by_type(NULL, "cpu");
-	if (!cpu)
-		return 0;
+	pkeys_total = of_scan_flat_dt(dt_scan_storage_keys, NULL);
+	if (pkeys_total == 0) {
 
-	if (of_property_read_u32_array(cpu,
-				       "ibm,processor-storage-keys", vals, 2) == 0) {
-		/*
-		 * Since any pkey can be used for data or execute, we will
-		 * just treat all keys as equal and track them as one entity.
-		 */
-		pkeys_total = vals[0];
-		/*  Should we check for IAMR support FIXME!! */
-	} else {
 		/*
 		 * Let's assume 32 pkeys on P8 bare metal, if its not defined by device
 		 * tree. We make this exception since skiboot forgot to expose this
 		 * property on power8.
 		 */
 		if (!firmware_has_feature(FW_FEATURE_LPAR) &&
-		    cpu_has_feature(CPU_FTRS_POWER8))
+		    early_cpu_has_feature(CPU_FTRS_POWER8))
 			pkeys_total = 32;
 	}
 
@@ -82,7 +90,7 @@ static int scan_pkey_feature(void)
 	return pkeys_total;
 }
 
-static int pkey_initialize(void)
+void __init pkey_early_init_devtree(void)
 {
 	int pkeys_total, i;
 
@@ -107,9 +115,11 @@ static int pkey_initialize(void)
 	if (!pkeys_total) {
 		/* No support for pkey. Mark it disabled */
 		static_branch_enable(&pkey_disabled);
-		return 0;
+		return;
 	}
 
+	cur_cpu_spec->mmu_features |= MMU_FTR_PKEY;
+
 	/*
 	 * The device tree cannot be relied to indicate support for
 	 * execute_disable support. Instead we use a PVR check.
@@ -187,11 +197,9 @@ static int pkey_initialize(void)
 	 */
 	initial_allocation_mask |= reserved_allocation_mask;
 
-	return 0;
+	return;
 }
 
-arch_initcall(pkey_initialize);
-
 void pkey_mm_init(struct mm_struct *mm)
 {
 	if (static_branch_likely(&pkey_disabled))
-- 
2.26.2

[PATCH v4 14/41] powerpc/book3s64/kuep: Add MMU_FTR_KUEP

[Aneesh Kumar K.V]

This will be used to enable/disable Kernel Userspace Execution
Prevention (KUEP).

Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
---
 arch/powerpc/include/asm/mmu.h | 5 +++++
 1 file changed, 5 insertions(+)

diff --git a/arch/powerpc/include/asm/mmu.h b/arch/powerpc/include/asm/mmu.h
index 72966d3d8f64..94435f85e3bc 100644
--- a/arch/powerpc/include/asm/mmu.h
+++ b/arch/powerpc/include/asm/mmu.h
@@ -24,6 +24,7 @@
 /* Radix page table supported and enabled */
 #define MMU_FTR_TYPE_RADIX		ASM_CONST(0x00000040)
 #define MMU_FTR_PKEY			ASM_CONST(0x00000080)
+#define MMU_FTR_KUEP			ASM_CONST(0x00000100)
 
 /*
  * Individual features below.
@@ -181,6 +182,10 @@ enum {
 #ifdef CONFIG_PPC_MEM_KEYS
 	MMU_FTR_PKEY |
 #endif
+#ifdef CONFIG_PPC_KUEP
+	MMU_FTR_KUEP |
+#endif /* CONFIG_PPC_KUAP */
+
 		0,
 };
 
-- 
2.26.2

[PATCH v4 15/41] powerpc/book3s64/pkeys: Use execute_pkey_disable static key

[Aneesh Kumar K.V]

Use execute_pkey_disabled static key to check for execute key support instead
of pkey_disabled.

Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
---
 arch/powerpc/include/asm/pkeys.h | 10 +---------
 arch/powerpc/mm/book3s64/pkeys.c |  5 ++++-
 2 files changed, 5 insertions(+), 10 deletions(-)

diff --git a/arch/powerpc/include/asm/pkeys.h b/arch/powerpc/include/asm/pkeys.h
index 47c81d41ea9a..09fbaa409ac4 100644
--- a/arch/powerpc/include/asm/pkeys.h
+++ b/arch/powerpc/include/asm/pkeys.h
@@ -126,15 +126,7 @@ static inline int mm_pkey_free(struct mm_struct *mm, int pkey)
  * Try to dedicate one of the protection keys to be used as an
  * execute-only protection key.
  */
-extern int __execute_only_pkey(struct mm_struct *mm);
-static inline int execute_only_pkey(struct mm_struct *mm)
-{
-	if (static_branch_likely(&pkey_disabled))
-		return -1;
-
-	return __execute_only_pkey(mm);
-}
-
+extern int execute_only_pkey(struct mm_struct *mm);
 extern int __arch_override_mprotect_pkey(struct vm_area_struct *vma,
 					 int prot, int pkey);
 static inline int arch_override_mprotect_pkey(struct vm_area_struct *vma,
diff --git a/arch/powerpc/mm/book3s64/pkeys.c b/arch/powerpc/mm/book3s64/pkeys.c
index bbba9c601e14..fed4f159011b 100644
--- a/arch/powerpc/mm/book3s64/pkeys.c
+++ b/arch/powerpc/mm/book3s64/pkeys.c
@@ -345,8 +345,11 @@ void thread_pkey_regs_init(struct thread_struct *thread)
 	write_uamor(default_uamor);
 }
 
-int __execute_only_pkey(struct mm_struct *mm)
+int execute_only_pkey(struct mm_struct *mm)
 {
+	if (static_branch_likely(&execute_pkey_disabled))
+		return -1;
+
 	return mm->context.execute_only_pkey;
 }
 
-- 
2.26.2

[PATCH v4 16/41] powerpc/book3s64/pkeys: Use MMU_FTR_PKEY instead of pkey_disabled static key

[Aneesh Kumar K.V]

Instead of pkey_disabled static key use mmu feature MMU_FTR_PKEY.

Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
---
 arch/powerpc/include/asm/book3s/64/pkeys.h |  2 +-
 arch/powerpc/include/asm/pkeys.h           | 14 ++++++--------
 arch/powerpc/mm/book3s64/pkeys.c           | 16 +++++++---------
 3 files changed, 14 insertions(+), 18 deletions(-)

diff --git a/arch/powerpc/include/asm/book3s/64/pkeys.h b/arch/powerpc/include/asm/book3s/64/pkeys.h
index 8174662a9173..5b178139f3c0 100644
--- a/arch/powerpc/include/asm/book3s/64/pkeys.h
+++ b/arch/powerpc/include/asm/book3s/64/pkeys.h
@@ -7,7 +7,7 @@
 
 static inline u64 vmflag_to_pte_pkey_bits(u64 vm_flags)
 {
-	if (static_branch_likely(&pkey_disabled))
+	if (!mmu_has_feature(MMU_FTR_PKEY))
 		return 0x0UL;
 
 	if (radix_enabled())
diff --git a/arch/powerpc/include/asm/pkeys.h b/arch/powerpc/include/asm/pkeys.h
index 09fbaa409ac4..b1d448c53209 100644
--- a/arch/powerpc/include/asm/pkeys.h
+++ b/arch/powerpc/include/asm/pkeys.h
@@ -11,7 +11,6 @@
 #include <linux/jump_label.h>
 #include <asm/firmware.h>
 
-DECLARE_STATIC_KEY_FALSE(pkey_disabled);
 extern int max_pkey;
 extern u32 reserved_allocation_mask; /* bits set for reserved keys */
 
@@ -38,7 +37,7 @@ static inline u64 pkey_to_vmflag_bits(u16 pkey)
 
 static inline int vma_pkey(struct vm_area_struct *vma)
 {
-	if (static_branch_likely(&pkey_disabled))
+	if (!mmu_has_feature(MMU_FTR_PKEY))
 		return 0;
 	return (vma->vm_flags & ARCH_VM_PKEY_FLAGS) >> VM_PKEY_SHIFT;
 }
@@ -93,9 +92,8 @@ static inline int mm_pkey_alloc(struct mm_struct *mm)
 	u32 all_pkeys_mask = (u32)(~(0x0));
 	int ret;
 
-	if (static_branch_likely(&pkey_disabled))
+	if (!mmu_has_feature(MMU_FTR_PKEY))
 		return -1;
-
 	/*
 	 * Are we out of pkeys? We must handle this specially because ffz()
 	 * behavior is undefined if there are no zeros.
@@ -111,7 +109,7 @@ static inline int mm_pkey_alloc(struct mm_struct *mm)
 
 static inline int mm_pkey_free(struct mm_struct *mm, int pkey)
 {
-	if (static_branch_likely(&pkey_disabled))
+	if (!mmu_has_feature(MMU_FTR_PKEY))
 		return -1;
 
 	if (!mm_pkey_is_allocated(mm, pkey))
@@ -132,7 +130,7 @@ extern int __arch_override_mprotect_pkey(struct vm_area_struct *vma,
 static inline int arch_override_mprotect_pkey(struct vm_area_struct *vma,
 					      int prot, int pkey)
 {
-	if (static_branch_likely(&pkey_disabled))
+	if (!mmu_has_feature(MMU_FTR_PKEY))
 		return 0;
 
 	/*
@@ -150,7 +148,7 @@ extern int __arch_set_user_pkey_access(struct task_struct *tsk, int pkey,
 static inline int arch_set_user_pkey_access(struct task_struct *tsk, int pkey,
 					    unsigned long init_val)
 {
-	if (static_branch_likely(&pkey_disabled))
+	if (!mmu_has_feature(MMU_FTR_PKEY))
 		return -EINVAL;
 
 	/*
@@ -167,7 +165,7 @@ static inline int arch_set_user_pkey_access(struct task_struct *tsk, int pkey,
 
 static inline bool arch_pkeys_enabled(void)
 {
-	return !static_branch_likely(&pkey_disabled);
+	return mmu_has_feature(MMU_FTR_PKEY);
 }
 
 extern void pkey_mm_init(struct mm_struct *mm);
diff --git a/arch/powerpc/mm/book3s64/pkeys.c b/arch/powerpc/mm/book3s64/pkeys.c
index fed4f159011b..810118123e70 100644
--- a/arch/powerpc/mm/book3s64/pkeys.c
+++ b/arch/powerpc/mm/book3s64/pkeys.c
@@ -13,7 +13,6 @@
 #include <linux/of_fdt.h>
 
 
-DEFINE_STATIC_KEY_FALSE(pkey_disabled);
 DEFINE_STATIC_KEY_FALSE(execute_pkey_disabled);
 int  max_pkey;			/* Maximum key value supported */
 /*
@@ -114,7 +113,6 @@ void __init pkey_early_init_devtree(void)
 	pkeys_total = scan_pkey_feature();
 	if (!pkeys_total) {
 		/* No support for pkey. Mark it disabled */
-		static_branch_enable(&pkey_disabled);
 		return;
 	}
 
@@ -202,7 +200,7 @@ void __init pkey_early_init_devtree(void)
 
 void pkey_mm_init(struct mm_struct *mm)
 {
-	if (static_branch_likely(&pkey_disabled))
+	if (!mmu_has_feature(MMU_FTR_PKEY))
 		return;
 	mm_pkey_allocation_map(mm) = initial_allocation_mask;
 	mm->context.execute_only_pkey = execute_only_key;
@@ -306,7 +304,7 @@ int __arch_set_user_pkey_access(struct task_struct *tsk, int pkey,
 
 void thread_pkey_regs_save(struct thread_struct *thread)
 {
-	if (static_branch_likely(&pkey_disabled))
+	if (!mmu_has_feature(MMU_FTR_PKEY))
 		return;
 
 	/*
@@ -320,7 +318,7 @@ void thread_pkey_regs_save(struct thread_struct *thread)
 void thread_pkey_regs_restore(struct thread_struct *new_thread,
 			      struct thread_struct *old_thread)
 {
-	if (static_branch_likely(&pkey_disabled))
+	if (!mmu_has_feature(MMU_FTR_PKEY))
 		return;
 
 	if (old_thread->amr != new_thread->amr)
@@ -333,7 +331,7 @@ void thread_pkey_regs_restore(struct thread_struct *new_thread,
 
 void thread_pkey_regs_init(struct thread_struct *thread)
 {
-	if (static_branch_likely(&pkey_disabled))
+	if (!mmu_has_feature(MMU_FTR_PKEY))
 		return;
 
 	thread->amr   = default_amr;
@@ -408,7 +406,7 @@ static bool pkey_access_permitted(int pkey, bool write, bool execute)
 
 bool arch_pte_access_permitted(u64 pte, bool write, bool execute)
 {
-	if (static_branch_likely(&pkey_disabled))
+	if (!mmu_has_feature(MMU_FTR_PKEY))
 		return true;
 
 	return pkey_access_permitted(pte_to_pkey_bits(pte), write, execute);
@@ -425,7 +423,7 @@ bool arch_pte_access_permitted(u64 pte, bool write, bool execute)
 bool arch_vma_access_permitted(struct vm_area_struct *vma, bool write,
 			       bool execute, bool foreign)
 {
-	if (static_branch_likely(&pkey_disabled))
+	if (!mmu_has_feature(MMU_FTR_PKEY))
 		return true;
 	/*
 	 * Do not enforce our key-permissions on a foreign vma.
@@ -438,7 +436,7 @@ bool arch_vma_access_permitted(struct vm_area_struct *vma, bool write,
 
 void arch_dup_pkeys(struct mm_struct *oldmm, struct mm_struct *mm)
 {
-	if (static_branch_likely(&pkey_disabled))
+	if (!mmu_has_feature(MMU_FTR_PKEY))
 		return;
 
 	/* Duplicate the oldmm pkey state in mm: */
-- 
2.26.2

[PATCH v4 17/41] powerpc/book3s64/kuap: Move KUAP related function outside radix

[Aneesh Kumar K.V]

The next set of patches adds support for kuap with hash translation.
In preparation for that rename/move kuap related functions to
non radix names.

Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
---
 .../asm/book3s/64/{kup-radix.h => kup.h}      |  6 +++---
 arch/powerpc/include/asm/kup.h                |  2 +-
 arch/powerpc/kernel/syscall_64.c              |  2 +-
 arch/powerpc/mm/book3s64/pkeys.c              | 19 +++++++++++++++++++
 arch/powerpc/mm/book3s64/radix_pgtable.c      | 18 ------------------
 5 files changed, 24 insertions(+), 23 deletions(-)
 rename arch/powerpc/include/asm/book3s/64/{kup-radix.h => kup.h} (97%)

diff --git a/arch/powerpc/include/asm/book3s/64/kup-radix.h b/arch/powerpc/include/asm/book3s/64/kup.h
similarity index 97%
rename from arch/powerpc/include/asm/book3s/64/kup-radix.h
rename to arch/powerpc/include/asm/book3s/64/kup.h
index 3ee1ec60be84..dff1fef765fa 100644
--- a/arch/powerpc/include/asm/book3s/64/kup-radix.h
+++ b/arch/powerpc/include/asm/book3s/64/kup.h
@@ -1,6 +1,6 @@
 /* SPDX-License-Identifier: GPL-2.0 */
-#ifndef _ASM_POWERPC_BOOK3S_64_KUP_RADIX_H
-#define _ASM_POWERPC_BOOK3S_64_KUP_RADIX_H
+#ifndef _ASM_POWERPC_BOOK3S_64_KUP_H
+#define _ASM_POWERPC_BOOK3S_64_KUP_H
 
 #include <linux/const.h>
 #include <asm/reg.h>
@@ -182,4 +182,4 @@ static inline unsigned long kuap_get_and_check_amr(void)
 
 #endif /* __ASSEMBLY__ */
 
-#endif /* _ASM_POWERPC_BOOK3S_64_KUP_RADIX_H */
+#endif /* _ASM_POWERPC_BOOK3S_64_KUP_H */
diff --git a/arch/powerpc/include/asm/kup.h b/arch/powerpc/include/asm/kup.h
index c745ee41ad66..015f51b02741 100644
--- a/arch/powerpc/include/asm/kup.h
+++ b/arch/powerpc/include/asm/kup.h
@@ -15,7 +15,7 @@
 #define KUAP_CURRENT		(KUAP_CURRENT_READ | KUAP_CURRENT_WRITE)
 
 #ifdef CONFIG_PPC64
-#include <asm/book3s/64/kup-radix.h>
+#include <asm/book3s/64/kup.h>
 #endif
 #ifdef CONFIG_PPC_8xx
 #include <asm/nohash/32/kup-8xx.h>
diff --git a/arch/powerpc/kernel/syscall_64.c b/arch/powerpc/kernel/syscall_64.c
index 79edba3ab312..7e560a01afa4 100644
--- a/arch/powerpc/kernel/syscall_64.c
+++ b/arch/powerpc/kernel/syscall_64.c
@@ -2,7 +2,7 @@
 
 #include <linux/err.h>
 #include <asm/asm-prototypes.h>
-#include <asm/book3s/64/kup-radix.h>
+#include <asm/book3s/64/kup.h>
 #include <asm/cputime.h>
 #include <asm/hw_irq.h>
 #include <asm/kprobes.h>
diff --git a/arch/powerpc/mm/book3s64/pkeys.c b/arch/powerpc/mm/book3s64/pkeys.c
index 810118123e70..cdd200441b58 100644
--- a/arch/powerpc/mm/book3s64/pkeys.c
+++ b/arch/powerpc/mm/book3s64/pkeys.c
@@ -12,6 +12,7 @@
 #include <linux/pkeys.h>
 #include <linux/of_fdt.h>
 
+#include <asm/smp.h>
 
 DEFINE_STATIC_KEY_FALSE(execute_pkey_disabled);
 int  max_pkey;			/* Maximum key value supported */
@@ -198,6 +199,24 @@ void __init pkey_early_init_devtree(void)
 	return;
 }
 
+#ifdef CONFIG_PPC_KUAP
+void __init setup_kuap(bool disabled)
+{
+	if (disabled || !early_radix_enabled())
+		return;
+
+	if (smp_processor_id() == boot_cpuid) {
+		pr_info("Activating Kernel Userspace Access Prevention\n");
+		cur_cpu_spec->mmu_features |= MMU_FTR_RADIX_KUAP;
+	}
+
+	/* Make sure userspace can't change the AMR */
+	mtspr(SPRN_UAMOR, 0);
+	mtspr(SPRN_AMR, AMR_KUAP_BLOCKED);
+	isync();
+}
+#endif
+
 void pkey_mm_init(struct mm_struct *mm)
 {
 	if (!mmu_has_feature(MMU_FTR_PKEY))
diff --git a/arch/powerpc/mm/book3s64/radix_pgtable.c b/arch/powerpc/mm/book3s64/radix_pgtable.c
index 8acb96de0e48..7ebaf35ec21d 100644
--- a/arch/powerpc/mm/book3s64/radix_pgtable.c
+++ b/arch/powerpc/mm/book3s64/radix_pgtable.c
@@ -532,24 +532,6 @@ void setup_kuep(bool disabled)
 }
 #endif
 
-#ifdef CONFIG_PPC_KUAP
-void setup_kuap(bool disabled)
-{
-	if (disabled || !early_radix_enabled())
-		return;
-
-	if (smp_processor_id() == boot_cpuid) {
-		pr_info("Activating Kernel Userspace Access Prevention\n");
-		cur_cpu_spec->mmu_features |= MMU_FTR_RADIX_KUAP;
-	}
-
-	/* Make sure userspace can't change the AMR */
-	mtspr(SPRN_UAMOR, 0);
-	mtspr(SPRN_AMR, AMR_KUAP_BLOCKED);
-	isync();
-}
-#endif
-
 void __init radix__early_init_mmu(void)
 {
 	unsigned long lpcr;
-- 
2.26.2

[PATCH v4 18/41] powerpc/book3s64/kuep: Move KUEP related function outside radix

[Aneesh Kumar K.V]

The next set of patches adds support for kuep with hash translation.
In preparation for that rename/move kuap related functions to
non radix names.

Also set MMU_FTR_KUEP and add the missing isync().

Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
---
 arch/powerpc/include/asm/book3s/64/kup.h |  1 +
 arch/powerpc/mm/book3s64/pkeys.c         | 21 +++++++++++++++++++++
 arch/powerpc/mm/book3s64/radix_pgtable.c | 18 ------------------
 3 files changed, 22 insertions(+), 18 deletions(-)

diff --git a/arch/powerpc/include/asm/book3s/64/kup.h b/arch/powerpc/include/asm/book3s/64/kup.h
index dff1fef765fa..bbac8b47df13 100644
--- a/arch/powerpc/include/asm/book3s/64/kup.h
+++ b/arch/powerpc/include/asm/book3s/64/kup.h
@@ -7,6 +7,7 @@
 
 #define AMR_KUAP_BLOCK_READ	UL(0x4000000000000000)
 #define AMR_KUAP_BLOCK_WRITE	UL(0x8000000000000000)
+#define AMR_KUEP_BLOCKED	(1UL << 62)
 #define AMR_KUAP_BLOCKED	(AMR_KUAP_BLOCK_READ | AMR_KUAP_BLOCK_WRITE)
 #define AMR_KUAP_SHIFT		62
 
diff --git a/arch/powerpc/mm/book3s64/pkeys.c b/arch/powerpc/mm/book3s64/pkeys.c
index cdd200441b58..3c048a8d238f 100644
--- a/arch/powerpc/mm/book3s64/pkeys.c
+++ b/arch/powerpc/mm/book3s64/pkeys.c
@@ -199,6 +199,27 @@ void __init pkey_early_init_devtree(void)
 	return;
 }
 
+#ifdef CONFIG_PPC_KUEP
+void __init setup_kuep(bool disabled)
+{
+	if (disabled || !early_radix_enabled())
+		return;
+
+	if (smp_processor_id() == boot_cpuid) {
+		pr_info("Activating Kernel Userspace Execution Prevention\n");
+		cur_cpu_spec->mmu_features |= MMU_FTR_KUEP;
+	}
+
+	/*
+	 * Radix always uses key0 of the IAMR to determine if an access is
+	 * allowed. We set bit 0 (IBM bit 1) of key0, to prevent instruction
+	 * fetch.
+	 */
+	mtspr(SPRN_IAMR, AMR_KUEP_BLOCKED);
+	isync();
+}
+#endif
+
 #ifdef CONFIG_PPC_KUAP
 void __init setup_kuap(bool disabled)
 {
diff --git a/arch/powerpc/mm/book3s64/radix_pgtable.c b/arch/powerpc/mm/book3s64/radix_pgtable.c
index 7ebaf35ec21d..2f641e1d7e82 100644
--- a/arch/powerpc/mm/book3s64/radix_pgtable.c
+++ b/arch/powerpc/mm/book3s64/radix_pgtable.c
@@ -514,24 +514,6 @@ static void radix_init_amor(void)
 	mtspr(SPRN_AMOR, (3ul << 62));
 }
 
-#ifdef CONFIG_PPC_KUEP
-void setup_kuep(bool disabled)
-{
-	if (disabled || !early_radix_enabled())
-		return;
-
-	if (smp_processor_id() == boot_cpuid)
-		pr_info("Activating Kernel Userspace Execution Prevention\n");
-
-	/*
-	 * Radix always uses key0 of the IAMR to determine if an access is
-	 * allowed. We set bit 0 (IBM bit 1) of key0, to prevent instruction
-	 * fetch.
-	 */
-	mtspr(SPRN_IAMR, (1ul << 62));
-}
-#endif
-
 void __init radix__early_init_mmu(void)
 {
 	unsigned long lpcr;
-- 
2.26.2

[PATCH v4 19/41] powerpc/book3s64/kuap: Rename MMU_FTR_RADIX_KUAP to MMU_FTR_KUAP

[Aneesh Kumar K.V]

The next set of patches adds support for kuap with hash translation.
In preparation for that rename/move kuap related functions to
non radix names.

Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
---
 arch/powerpc/include/asm/book3s/64/kup.h | 18 +++++++++---------
 arch/powerpc/include/asm/mmu.h           |  6 +++---
 arch/powerpc/mm/book3s64/pkeys.c         |  2 +-
 3 files changed, 13 insertions(+), 13 deletions(-)

diff --git a/arch/powerpc/include/asm/book3s/64/kup.h b/arch/powerpc/include/asm/book3s/64/kup.h
index bbac8b47df13..476bcd7b0e8b 100644
--- a/arch/powerpc/include/asm/book3s/64/kup.h
+++ b/arch/powerpc/include/asm/book3s/64/kup.h
@@ -24,7 +24,7 @@
 	mtspr	SPRN_AMR, \gpr2
 	/* No isync required, see kuap_restore_amr() */
 998:
-	END_MMU_FTR_SECTION_NESTED_IFSET(MMU_FTR_RADIX_KUAP, 67)
+	END_MMU_FTR_SECTION_NESTED_IFSET(MMU_FTR_KUAP, 67)
 #endif
 .endm
 
@@ -36,7 +36,7 @@
 	sldi	\gpr2, \gpr2, AMR_KUAP_SHIFT
 999:	tdne	\gpr1, \gpr2
 	EMIT_BUG_ENTRY 999b, __FILE__, __LINE__, (BUGFLAG_WARNING | BUGFLAG_ONCE)
-	END_MMU_FTR_SECTION_NESTED_IFSET(MMU_FTR_RADIX_KUAP, 67)
+	END_MMU_FTR_SECTION_NESTED_IFSET(MMU_FTR_KUAP, 67)
 #endif
 .endm
 
@@ -56,7 +56,7 @@
 	mtspr	SPRN_AMR, \gpr2
 	isync
 99:
-	END_MMU_FTR_SECTION_NESTED_IFSET(MMU_FTR_RADIX_KUAP, 67)
+	END_MMU_FTR_SECTION_NESTED_IFSET(MMU_FTR_KUAP, 67)
 #endif
 .endm
 
@@ -69,7 +69,7 @@
 
 static inline void kuap_restore_amr(struct pt_regs *regs, unsigned long amr)
 {
-	if (mmu_has_feature(MMU_FTR_RADIX_KUAP) && unlikely(regs->kuap != amr)) {
+	if (mmu_has_feature(MMU_FTR_KUAP) && unlikely(regs->kuap != amr)) {
 		isync();
 		mtspr(SPRN_AMR, regs->kuap);
 		/*
@@ -82,7 +82,7 @@ static inline void kuap_restore_amr(struct pt_regs *regs, unsigned long amr)
 
 static inline unsigned long kuap_get_and_check_amr(void)
 {
-	if (mmu_has_feature(MMU_FTR_RADIX_KUAP)) {
+	if (mmu_has_feature(MMU_FTR_KUAP)) {
 		unsigned long amr = mfspr(SPRN_AMR);
 		if (IS_ENABLED(CONFIG_PPC_KUAP_DEBUG)) /* kuap_check_amr() */
 			WARN_ON_ONCE(amr != AMR_KUAP_BLOCKED);
@@ -93,7 +93,7 @@ static inline unsigned long kuap_get_and_check_amr(void)
 
 static inline void kuap_check_amr(void)
 {
-	if (IS_ENABLED(CONFIG_PPC_KUAP_DEBUG) && mmu_has_feature(MMU_FTR_RADIX_KUAP))
+	if (IS_ENABLED(CONFIG_PPC_KUAP_DEBUG) && mmu_has_feature(MMU_FTR_KUAP))
 		WARN_ON_ONCE(mfspr(SPRN_AMR) != AMR_KUAP_BLOCKED);
 }
 
@@ -104,7 +104,7 @@ static inline void kuap_check_amr(void)
 
 static inline unsigned long get_kuap(void)
 {
-	if (!early_mmu_has_feature(MMU_FTR_RADIX_KUAP))
+	if (!early_mmu_has_feature(MMU_FTR_KUAP))
 		return 0;
 
 	return mfspr(SPRN_AMR);
@@ -112,7 +112,7 @@ static inline unsigned long get_kuap(void)
 
 static inline void set_kuap(unsigned long value)
 {
-	if (!early_mmu_has_feature(MMU_FTR_RADIX_KUAP))
+	if (!early_mmu_has_feature(MMU_FTR_KUAP))
 		return;
 
 	/*
@@ -162,7 +162,7 @@ static inline void restore_user_access(unsigned long flags)
 static inline bool
 bad_kuap_fault(struct pt_regs *regs, unsigned long address, bool is_write)
 {
-	return WARN(mmu_has_feature(MMU_FTR_RADIX_KUAP) &&
+	return WARN(mmu_has_feature(MMU_FTR_KUAP) &&
 		    (regs->kuap & (is_write ? AMR_KUAP_BLOCK_WRITE : AMR_KUAP_BLOCK_READ)),
 		    "Bug: %s fault blocked by AMR!", is_write ? "Write" : "Read");
 }
diff --git a/arch/powerpc/include/asm/mmu.h b/arch/powerpc/include/asm/mmu.h
index 94435f85e3bc..14d7e6803453 100644
--- a/arch/powerpc/include/asm/mmu.h
+++ b/arch/powerpc/include/asm/mmu.h
@@ -112,7 +112,7 @@
 /*
  * Supports KUAP (key 0 controlling userspace addresses) on radix
  */
-#define MMU_FTR_RADIX_KUAP		ASM_CONST(0x80000000)
+#define MMU_FTR_KUAP			ASM_CONST(0x80000000)
 
 /* MMU feature bit sets for various CPUs */
 #define MMU_FTRS_DEFAULT_HPTE_ARCH_V2	\
@@ -175,10 +175,10 @@ enum {
 #endif
 #ifdef CONFIG_PPC_RADIX_MMU
 		MMU_FTR_TYPE_RADIX |
+#endif /* CONFIG_PPC_RADIX_MMU */
 #ifdef CONFIG_PPC_KUAP
-		MMU_FTR_RADIX_KUAP |
+	MMU_FTR_KUAP |
 #endif /* CONFIG_PPC_KUAP */
-#endif /* CONFIG_PPC_RADIX_MMU */
 #ifdef CONFIG_PPC_MEM_KEYS
 	MMU_FTR_PKEY |
 #endif
diff --git a/arch/powerpc/mm/book3s64/pkeys.c b/arch/powerpc/mm/book3s64/pkeys.c
index 3c048a8d238f..ad3348b7e25c 100644
--- a/arch/powerpc/mm/book3s64/pkeys.c
+++ b/arch/powerpc/mm/book3s64/pkeys.c
@@ -228,7 +228,7 @@ void __init setup_kuap(bool disabled)
 
 	if (smp_processor_id() == boot_cpuid) {
 		pr_info("Activating Kernel Userspace Access Prevention\n");
-		cur_cpu_spec->mmu_features |= MMU_FTR_RADIX_KUAP;
+		cur_cpu_spec->mmu_features |= MMU_FTR_KUAP;
 	}
 
 	/* Make sure userspace can't change the AMR */
-- 
2.26.2

[PATCH v4 20/41] powerpc/book3s64/kuap/kuep: Make KUAP and KUEP a subfeature of PPC_MEM_KEYS

[Aneesh Kumar K.V]

The next set of patches adds support for kuap with hash translation.
Hence make KUAP a BOOK3S_64 feature. Also make it a subfeature of
PPC_MEM_KEYS. Hash translation is going to use pkeys to support
KUAP/KUEP. Adding this dependency reduces the code complexity and
enables us to move some of the initialization code to pkeys.c

Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
---
 arch/powerpc/include/asm/book3s/64/kup.h | 33 ++++++++++++++----------
 arch/powerpc/include/asm/ptrace.h        |  2 +-
 arch/powerpc/kernel/asm-offsets.c        |  2 +-
 arch/powerpc/platforms/Kconfig.cputype   |  4 +--
 4 files changed, 23 insertions(+), 18 deletions(-)

diff --git a/arch/powerpc/include/asm/book3s/64/kup.h b/arch/powerpc/include/asm/book3s/64/kup.h
index 476bcd7b0e8b..aa2787c8b12a 100644
--- a/arch/powerpc/include/asm/book3s/64/kup.h
+++ b/arch/powerpc/include/asm/book3s/64/kup.h
@@ -62,7 +62,7 @@
 
 #else /* !__ASSEMBLY__ */
 
-#ifdef CONFIG_PPC_KUAP
+#ifdef CONFIG_PPC_MEM_KEYS
 
 #include <asm/mmu.h>
 #include <asm/ptrace.h>
@@ -97,6 +97,24 @@ static inline void kuap_check_amr(void)
 		WARN_ON_ONCE(mfspr(SPRN_AMR) != AMR_KUAP_BLOCKED);
 }
 
+#else /* CONFIG_PPC_MEM_KEYS */
+
+static inline void kuap_restore_amr(struct pt_regs *regs, unsigned long amr)
+{
+}
+
+static inline void kuap_check_amr(void)
+{
+}
+
+static inline unsigned long kuap_get_and_check_amr(void)
+{
+	return 0;
+}
+#endif /* CONFIG_PPC_MEM_KEYS */
+
+
+#ifdef CONFIG_PPC_KUAP
 /*
  * We support individually allowing read or write, but we don't support nesting
  * because that would require an expensive read/modify write of the AMR.
@@ -166,19 +184,6 @@ bad_kuap_fault(struct pt_regs *regs, unsigned long address, bool is_write)
 		    (regs->kuap & (is_write ? AMR_KUAP_BLOCK_WRITE : AMR_KUAP_BLOCK_READ)),
 		    "Bug: %s fault blocked by AMR!", is_write ? "Write" : "Read");
 }
-#else /* CONFIG_PPC_KUAP */
-static inline void kuap_restore_amr(struct pt_regs *regs, unsigned long amr)
-{
-}
-
-static inline void kuap_check_amr(void)
-{
-}
-
-static inline unsigned long kuap_get_and_check_amr(void)
-{
-	return 0;
-}
 #endif /* CONFIG_PPC_KUAP */
 
 #endif /* __ASSEMBLY__ */
diff --git a/arch/powerpc/include/asm/ptrace.h b/arch/powerpc/include/asm/ptrace.h
index ac3970fff0d5..1a6cadf63d14 100644
--- a/arch/powerpc/include/asm/ptrace.h
+++ b/arch/powerpc/include/asm/ptrace.h
@@ -53,7 +53,7 @@ struct pt_regs
 #ifdef CONFIG_PPC64
 			unsigned long ppr;
 #endif
-#ifdef CONFIG_PPC_KUAP
+#ifdef CONFIG_PPC_HAVE_KUAP
 			unsigned long kuap;
 #endif
 		};
diff --git a/arch/powerpc/kernel/asm-offsets.c b/arch/powerpc/kernel/asm-offsets.c
index 9b9cde07e396..1694c4f531b9 100644
--- a/arch/powerpc/kernel/asm-offsets.c
+++ b/arch/powerpc/kernel/asm-offsets.c
@@ -354,7 +354,7 @@ int main(void)
 	STACK_PT_REGS_OFFSET(_PPR, ppr);
 #endif /* CONFIG_PPC64 */
 
-#ifdef CONFIG_PPC_KUAP
+#ifdef CONFIG_PPC_HAVE_KUAP
 	STACK_PT_REGS_OFFSET(STACK_REGS_KUAP, kuap);
 #endif
 
diff --git a/arch/powerpc/platforms/Kconfig.cputype b/arch/powerpc/platforms/Kconfig.cputype
index d349603fb889..053c46aecf80 100644
--- a/arch/powerpc/platforms/Kconfig.cputype
+++ b/arch/powerpc/platforms/Kconfig.cputype
@@ -99,6 +99,8 @@ config PPC_BOOK3S_64
 	select ARCH_SUPPORTS_NUMA_BALANCING
 	select IRQ_WORK
 	select PPC_MM_SLICES
+	select PPC_HAVE_KUAP if PPC_MEM_KEYS
+	select PPC_HAVE_KUEP if PPC_MEM_KEYS
 
 config PPC_BOOK3E_64
 	bool "Embedded processors"
@@ -350,8 +352,6 @@ config PPC_RADIX_MMU
 	bool "Radix MMU Support"
 	depends on PPC_BOOK3S_64
 	select ARCH_HAS_GIGANTIC_PAGE
-	select PPC_HAVE_KUEP
-	select PPC_HAVE_KUAP
 	default y
 	help
 	  Enable support for the Power ISA 3.0 Radix style MMU. Currently this
-- 
2.26.2

[PATCH v4 21/41] powerpc/book3s64/kuap: Move UAMOR setup to key init function

[Aneesh Kumar K.V]

With hash translation, the kernel will use key 3 for implementing
KUAP feature. Hence the default UAMOR value depends on what other
keys are marked reserved. Move the UAMOR initialization to pkeys init.

Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
---
 arch/powerpc/include/asm/book3s/64/kup.h |  2 ++
 arch/powerpc/kernel/smp.c                |  5 +++++
 arch/powerpc/mm/book3s64/pkeys.c         | 25 +++++++++++++++++++-----
 3 files changed, 27 insertions(+), 5 deletions(-)

diff --git a/arch/powerpc/include/asm/book3s/64/kup.h b/arch/powerpc/include/asm/book3s/64/kup.h
index aa2787c8b12a..e6ee1c34842f 100644
--- a/arch/powerpc/include/asm/book3s/64/kup.h
+++ b/arch/powerpc/include/asm/book3s/64/kup.h
@@ -67,6 +67,8 @@
 #include <asm/mmu.h>
 #include <asm/ptrace.h>
 
+extern u64 default_uamor;
+
 static inline void kuap_restore_amr(struct pt_regs *regs, unsigned long amr)
 {
 	if (mmu_has_feature(MMU_FTR_KUAP) && unlikely(regs->kuap != amr)) {
diff --git a/arch/powerpc/kernel/smp.c b/arch/powerpc/kernel/smp.c
index c820c95162ff..eec40082599f 100644
--- a/arch/powerpc/kernel/smp.c
+++ b/arch/powerpc/kernel/smp.c
@@ -59,6 +59,7 @@
 #include <asm/asm-prototypes.h>
 #include <asm/cpu_has_feature.h>
 #include <asm/ftrace.h>
+#include <asm/kup.h>
 
 #ifdef DEBUG
 #include <asm/udbg.h>
@@ -1256,6 +1257,10 @@ void start_secondary(void *unused)
 	mmgrab(&init_mm);
 	current->active_mm = &init_mm;
 
+#ifdef CONFIG_PPC_MEM_KEYS
+	mtspr(SPRN_UAMOR, default_uamor);
+#endif
+
 	smp_store_cpu_info(cpu);
 	set_dec(tb_ticks_per_jiffy);
 	preempt_disable();
diff --git a/arch/powerpc/mm/book3s64/pkeys.c b/arch/powerpc/mm/book3s64/pkeys.c
index ad3348b7e25c..4a7d05c1a9c7 100644
--- a/arch/powerpc/mm/book3s64/pkeys.c
+++ b/arch/powerpc/mm/book3s64/pkeys.c
@@ -24,7 +24,7 @@ static u32  initial_allocation_mask;   /* Bits set for the initially allocated k
 static u64 default_amr;
 static u64 default_iamr;
 /* Allow all keys to be modified by default */
-static u64 default_uamor = ~0x0UL;
+u64 default_uamor = ~0x0UL;
 /*
  * Key used to implement PROT_EXEC mmap. Denies READ/WRITE
  * We pick key 2 because 0 is special key and 1 is reserved as per ISA.
@@ -113,8 +113,16 @@ void __init pkey_early_init_devtree(void)
 	/* scan the device tree for pkey feature */
 	pkeys_total = scan_pkey_feature();
 	if (!pkeys_total) {
-		/* No support for pkey. Mark it disabled */
-		return;
+		/*
+		 * No key support but on radix we can use key 0
+		 * to implement kuap.
+		 */
+		if (early_radix_enabled())
+			/*
+			 * Make sure userspace can't change the AMR
+			 */
+			default_uamor = 0;
+		goto err_out;
 	}
 
 	cur_cpu_spec->mmu_features |= MMU_FTR_PKEY;
@@ -196,6 +204,12 @@ void __init pkey_early_init_devtree(void)
 	 */
 	initial_allocation_mask |= reserved_allocation_mask;
 
+err_out:
+	/*
+	 * Setup uamor on boot cpu
+	 */
+	mtspr(SPRN_UAMOR, default_uamor);
+
 	return;
 }
 
@@ -231,8 +245,9 @@ void __init setup_kuap(bool disabled)
 		cur_cpu_spec->mmu_features |= MMU_FTR_KUAP;
 	}
 
-	/* Make sure userspace can't change the AMR */
-	mtspr(SPRN_UAMOR, 0);
+	/*
+	 * Set the default kernel AMR values on all cpus.
+	 */
 	mtspr(SPRN_AMR, AMR_KUAP_BLOCKED);
 	isync();
 }
-- 
2.26.2

[PATCH v4 22/41] powerpc/book3s64/kuap: Use Key 3 for kernel mapping with hash translation

[Aneesh Kumar K.V]

This patch updates kernel hash page table entries to use storage key 3
for its mapping. This implies all kernel access will now use key 3 to
control READ/WRITE. The patch also prevents the allocation of key 3 from
userspace and UAMOR value is updated such that userspace cannot modify key 3.

Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
---
 .../powerpc/include/asm/book3s/64/hash-pkey.h | 24 ++++++++++++++-----
 arch/powerpc/include/asm/book3s/64/hash.h     |  3 ++-
 arch/powerpc/include/asm/book3s/64/mmu-hash.h |  1 +
 arch/powerpc/include/asm/mmu_context.h        |  2 +-
 arch/powerpc/mm/book3s64/hash_4k.c            |  2 +-
 arch/powerpc/mm/book3s64/hash_64k.c           |  4 ++--
 arch/powerpc/mm/book3s64/hash_hugepage.c      |  2 +-
 arch/powerpc/mm/book3s64/hash_hugetlbpage.c   |  2 +-
 arch/powerpc/mm/book3s64/hash_pgtable.c       |  2 +-
 arch/powerpc/mm/book3s64/hash_utils.c         | 10 ++++----
 arch/powerpc/mm/book3s64/pkeys.c              |  4 ++++
 11 files changed, 38 insertions(+), 18 deletions(-)

diff --git a/arch/powerpc/include/asm/book3s/64/hash-pkey.h b/arch/powerpc/include/asm/book3s/64/hash-pkey.h
index 795010897e5d..9f44e208f036 100644
--- a/arch/powerpc/include/asm/book3s/64/hash-pkey.h
+++ b/arch/powerpc/include/asm/book3s/64/hash-pkey.h
@@ -2,6 +2,9 @@
 #ifndef _ASM_POWERPC_BOOK3S_64_HASH_PKEY_H
 #define _ASM_POWERPC_BOOK3S_64_HASH_PKEY_H
 
+/*  We use key 3 for KERNEL */
+#define HASH_DEFAULT_KERNEL_KEY (HPTE_R_KEY_BIT0 | HPTE_R_KEY_BIT1)
+
 static inline u64 hash__vmflag_to_pte_pkey_bits(u64 vm_flags)
 {
 	return (((vm_flags & VM_PKEY_BIT0) ? H_PTE_PKEY_BIT0 : 0x0UL) |
@@ -11,13 +14,22 @@ static inline u64 hash__vmflag_to_pte_pkey_bits(u64 vm_flags)
 		((vm_flags & VM_PKEY_BIT4) ? H_PTE_PKEY_BIT4 : 0x0UL));
 }
 
-static inline u64 pte_to_hpte_pkey_bits(u64 pteflags)
+static inline u64 pte_to_hpte_pkey_bits(u64 pteflags, unsigned long flags)
 {
-	return (((pteflags & H_PTE_PKEY_BIT4) ? HPTE_R_KEY_BIT4 : 0x0UL) |
-		((pteflags & H_PTE_PKEY_BIT3) ? HPTE_R_KEY_BIT3 : 0x0UL) |
-		((pteflags & H_PTE_PKEY_BIT2) ? HPTE_R_KEY_BIT2 : 0x0UL) |
-		((pteflags & H_PTE_PKEY_BIT1) ? HPTE_R_KEY_BIT1 : 0x0UL) |
-		((pteflags & H_PTE_PKEY_BIT0) ? HPTE_R_KEY_BIT0 : 0x0UL));
+	unsigned long pte_pkey;
+
+	pte_pkey = (((pteflags & H_PTE_PKEY_BIT4) ? HPTE_R_KEY_BIT4 : 0x0UL) |
+		    ((pteflags & H_PTE_PKEY_BIT3) ? HPTE_R_KEY_BIT3 : 0x0UL) |
+		    ((pteflags & H_PTE_PKEY_BIT2) ? HPTE_R_KEY_BIT2 : 0x0UL) |
+		    ((pteflags & H_PTE_PKEY_BIT1) ? HPTE_R_KEY_BIT1 : 0x0UL) |
+		    ((pteflags & H_PTE_PKEY_BIT0) ? HPTE_R_KEY_BIT0 : 0x0UL));
+
+	if (mmu_has_feature(MMU_FTR_KUAP) || mmu_has_feature(MMU_FTR_KUEP)) {
+		if ((pte_pkey == 0) && (flags & HPTE_USE_KERNEL_KEY))
+			return HASH_DEFAULT_KERNEL_KEY;
+	}
+
+	return pte_pkey;
 }
 
 static inline u16 hash__pte_to_pkey_bits(u64 pteflags)
diff --git a/arch/powerpc/include/asm/book3s/64/hash.h b/arch/powerpc/include/asm/book3s/64/hash.h
index 73ad038ed10b..a3c0f93d2768 100644
--- a/arch/powerpc/include/asm/book3s/64/hash.h
+++ b/arch/powerpc/include/asm/book3s/64/hash.h
@@ -145,7 +145,8 @@ extern void hash__mark_initmem_nx(void);
 
 extern void hpte_need_flush(struct mm_struct *mm, unsigned long addr,
 			    pte_t *ptep, unsigned long pte, int huge);
-extern unsigned long htab_convert_pte_flags(unsigned long pteflags);
+extern unsigned long htab_convert_pte_flags(unsigned long pteflags,
+					    unsigned long flags);
 /* Atomic PTE updates */
 static inline unsigned long hash__pte_update(struct mm_struct *mm,
 					 unsigned long addr,
diff --git a/arch/powerpc/include/asm/book3s/64/mmu-hash.h b/arch/powerpc/include/asm/book3s/64/mmu-hash.h
index 58fcc959f9d5..eb9950043b78 100644
--- a/arch/powerpc/include/asm/book3s/64/mmu-hash.h
+++ b/arch/powerpc/include/asm/book3s/64/mmu-hash.h
@@ -452,6 +452,7 @@ static inline unsigned long hpt_hash(unsigned long vpn,
 
 #define HPTE_LOCAL_UPDATE	0x1
 #define HPTE_NOHPTE_UPDATE	0x2
+#define HPTE_USE_KERNEL_KEY	0x4
 
 extern int __hash_page_4K(unsigned long ea, unsigned long access,
 			  unsigned long vsid, pte_t *ptep, unsigned long trap,
diff --git a/arch/powerpc/include/asm/mmu_context.h b/arch/powerpc/include/asm/mmu_context.h
index 1a474f6b1992..2d85e0ea5f1c 100644
--- a/arch/powerpc/include/asm/mmu_context.h
+++ b/arch/powerpc/include/asm/mmu_context.h
@@ -286,7 +286,7 @@ static inline bool arch_vma_access_permitted(struct vm_area_struct *vma,
 #define thread_pkey_regs_init(thread)
 #define arch_dup_pkeys(oldmm, mm)
 
-static inline u64 pte_to_hpte_pkey_bits(u64 pteflags)
+static inline u64 pte_to_hpte_pkey_bits(u64 pteflags, unsigned long flags)
 {
 	return 0x0UL;
 }
diff --git a/arch/powerpc/mm/book3s64/hash_4k.c b/arch/powerpc/mm/book3s64/hash_4k.c
index 22e787123cdf..7de1a8a0c62a 100644
--- a/arch/powerpc/mm/book3s64/hash_4k.c
+++ b/arch/powerpc/mm/book3s64/hash_4k.c
@@ -54,7 +54,7 @@ int __hash_page_4K(unsigned long ea, unsigned long access, unsigned long vsid,
 	 * PP bits. _PAGE_USER is already PP bit 0x2, so we only
 	 * need to add in 0x1 if it's a read-only user page
 	 */
-	rflags = htab_convert_pte_flags(new_pte);
+	rflags = htab_convert_pte_flags(new_pte, flags);
 	rpte = __real_pte(__pte(old_pte), ptep, PTRS_PER_PTE);
 
 	if (cpu_has_feature(CPU_FTR_NOEXECUTE) &&
diff --git a/arch/powerpc/mm/book3s64/hash_64k.c b/arch/powerpc/mm/book3s64/hash_64k.c
index 7084ce2951e6..998c6817ed47 100644
--- a/arch/powerpc/mm/book3s64/hash_64k.c
+++ b/arch/powerpc/mm/book3s64/hash_64k.c
@@ -72,7 +72,7 @@ int __hash_page_4K(unsigned long ea, unsigned long access, unsigned long vsid,
 	 * Handle the subpage protection bits
 	 */
 	subpg_pte = new_pte & ~subpg_prot;
-	rflags = htab_convert_pte_flags(subpg_pte);
+	rflags = htab_convert_pte_flags(subpg_pte, flags);
 
 	if (cpu_has_feature(CPU_FTR_NOEXECUTE) &&
 	    !cpu_has_feature(CPU_FTR_COHERENT_ICACHE)) {
@@ -260,7 +260,7 @@ int __hash_page_64K(unsigned long ea, unsigned long access,
 			new_pte |= _PAGE_DIRTY;
 	} while (!pte_xchg(ptep, __pte(old_pte), __pte(new_pte)));
 
-	rflags = htab_convert_pte_flags(new_pte);
+	rflags = htab_convert_pte_flags(new_pte, flags);
 	rpte = __real_pte(__pte(old_pte), ptep, PTRS_PER_PTE);
 
 	if (cpu_has_feature(CPU_FTR_NOEXECUTE) &&
diff --git a/arch/powerpc/mm/book3s64/hash_hugepage.c b/arch/powerpc/mm/book3s64/hash_hugepage.c
index 440823797de7..c0fabe6c5a12 100644
--- a/arch/powerpc/mm/book3s64/hash_hugepage.c
+++ b/arch/powerpc/mm/book3s64/hash_hugepage.c
@@ -57,7 +57,7 @@ int __hash_page_thp(unsigned long ea, unsigned long access, unsigned long vsid,
 	if (!(old_pmd & (H_PAGE_THP_HUGE | _PAGE_DEVMAP)))
 		return 0;
 
-	rflags = htab_convert_pte_flags(new_pmd);
+	rflags = htab_convert_pte_flags(new_pmd, flags);
 
 #if 0
 	if (!cpu_has_feature(CPU_FTR_COHERENT_ICACHE)) {
diff --git a/arch/powerpc/mm/book3s64/hash_hugetlbpage.c b/arch/powerpc/mm/book3s64/hash_hugetlbpage.c
index eefa89c6117b..8ce8fc327af4 100644
--- a/arch/powerpc/mm/book3s64/hash_hugetlbpage.c
+++ b/arch/powerpc/mm/book3s64/hash_hugetlbpage.c
@@ -72,7 +72,7 @@ int __hash_page_huge(unsigned long ea, unsigned long access, unsigned long vsid,
 	if (old_pte & (H_PAGE_THP_HUGE | _PAGE_DEVMAP))
 		return 0;
 
-	rflags = htab_convert_pte_flags(new_pte);
+	rflags = htab_convert_pte_flags(new_pte, flags);
 	if (unlikely(mmu_psize == MMU_PAGE_16G))
 		offset = PTRS_PER_PUD;
 	else
diff --git a/arch/powerpc/mm/book3s64/hash_pgtable.c b/arch/powerpc/mm/book3s64/hash_pgtable.c
index 8b4b0a602158..6a213e3ddd35 100644
--- a/arch/powerpc/mm/book3s64/hash_pgtable.c
+++ b/arch/powerpc/mm/book3s64/hash_pgtable.c
@@ -445,7 +445,7 @@ void hash__mark_initmem_nx(void)
 	start = (unsigned long)__init_begin;
 	end = (unsigned long)__init_end;
 
-	pp = htab_convert_pte_flags(pgprot_val(PAGE_KERNEL));
+	pp = htab_convert_pte_flags(pgprot_val(PAGE_KERNEL), HPTE_USE_KERNEL_KEY);
 
 	WARN_ON(!hash__change_memory_range(start, end, pp));
 }
diff --git a/arch/powerpc/mm/book3s64/hash_utils.c b/arch/powerpc/mm/book3s64/hash_utils.c
index 0124003e60d0..9686caf36be9 100644
--- a/arch/powerpc/mm/book3s64/hash_utils.c
+++ b/arch/powerpc/mm/book3s64/hash_utils.c
@@ -186,7 +186,7 @@ static struct mmu_psize_def mmu_psize_defaults_gp[] = {
  *    - We make sure R is always set and never lost
  *    - C is _PAGE_DIRTY, and *should* always be set for a writeable mapping
  */
-unsigned long htab_convert_pte_flags(unsigned long pteflags)
+unsigned long htab_convert_pte_flags(unsigned long pteflags, unsigned long flags)
 {
 	unsigned long rflags = 0;
 
@@ -240,7 +240,7 @@ unsigned long htab_convert_pte_flags(unsigned long pteflags)
 		 */
 		rflags |= HPTE_R_M;
 
-	rflags |= pte_to_hpte_pkey_bits(pteflags);
+	rflags |= pte_to_hpte_pkey_bits(pteflags, flags);
 	return rflags;
 }
 
@@ -255,7 +255,7 @@ int htab_bolt_mapping(unsigned long vstart, unsigned long vend,
 	shift = mmu_psize_defs[psize].shift;
 	step = 1 << shift;
 
-	prot = htab_convert_pte_flags(prot);
+	prot = htab_convert_pte_flags(prot, HPTE_USE_KERNEL_KEY);
 
 	DBG("htab_bolt_mapping(%lx..%lx -> %lx (%lx,%d,%d)\n",
 	    vstart, vend, pstart, prot, psize, ssize);
@@ -1301,12 +1301,14 @@ int hash_page_mm(struct mm_struct *mm, unsigned long ea,
 		vsid = get_kernel_vsid(ea, mmu_kernel_ssize);
 		psize = mmu_vmalloc_psize;
 		ssize = mmu_kernel_ssize;
+		flags |= HPTE_USE_KERNEL_KEY;
 		break;
 
 	case IO_REGION_ID:
 		vsid = get_kernel_vsid(ea, mmu_kernel_ssize);
 		psize = mmu_io_psize;
 		ssize = mmu_kernel_ssize;
+		flags |= HPTE_USE_KERNEL_KEY;
 		break;
 	default:
 		/*
@@ -1864,7 +1866,7 @@ static void kernel_map_linear_page(unsigned long vaddr, unsigned long lmi)
 	unsigned long hash;
 	unsigned long vsid = get_kernel_vsid(vaddr, mmu_kernel_ssize);
 	unsigned long vpn = hpt_vpn(vaddr, vsid, mmu_kernel_ssize);
-	unsigned long mode = htab_convert_pte_flags(pgprot_val(PAGE_KERNEL));
+	unsigned long mode = htab_convert_pte_flags(pgprot_val(PAGE_KERNEL), HPTE_USE_KERNEL_KEY);
 	long ret;
 
 	hash = hpt_hash(vpn, PAGE_SHIFT, mmu_kernel_ssize);
diff --git a/arch/powerpc/mm/book3s64/pkeys.c b/arch/powerpc/mm/book3s64/pkeys.c
index 4a7d05c1a9c7..7980b013d161 100644
--- a/arch/powerpc/mm/book3s64/pkeys.c
+++ b/arch/powerpc/mm/book3s64/pkeys.c
@@ -190,6 +190,10 @@ void __init pkey_early_init_devtree(void)
 	reserved_allocation_mask |= (0x1 << 1);
 	default_uamor &= ~(0x3ul << pkeyshift(1));
 
+	/*  handle key 3 which is used by kernel for KAUP */
+	reserved_allocation_mask |= (0x1 << 3);
+	default_uamor &= ~(0x3ul << pkeyshift(3));
+
 	/*
 	 * Prevent the usage of OS reserved keys. Update UAMOR
 	 * for those keys. Also mark the rest of the bits in the
-- 
2.26.2

[PATCH v4 23/41] powerpc/exec: Set thread.regs early during exec

[Aneesh Kumar K.V]

In later patches during exec, we would like to access default regs.kuap to
control access to the user mapping. Having thread.regs set early makes the
code changes simpler.

Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
---
 arch/powerpc/include/asm/thread_info.h |  2 --
 arch/powerpc/kernel/process.c          | 37 +++++++++++++++++---------
 2 files changed, 25 insertions(+), 14 deletions(-)

diff --git a/arch/powerpc/include/asm/thread_info.h b/arch/powerpc/include/asm/thread_info.h
index ca6c97025704..9418dff1cfe1 100644
--- a/arch/powerpc/include/asm/thread_info.h
+++ b/arch/powerpc/include/asm/thread_info.h
@@ -77,10 +77,8 @@ struct thread_info {
 /* how to get the thread information struct from C */
 extern int arch_dup_task_struct(struct task_struct *dst, struct task_struct *src);
 
-#ifdef CONFIG_PPC_BOOK3S_64
 void arch_setup_new_exec(void);
 #define arch_setup_new_exec arch_setup_new_exec
-#endif
 
 #endif /* __ASSEMBLY__ */
 
diff --git a/arch/powerpc/kernel/process.c b/arch/powerpc/kernel/process.c
index 048d64c4e115..14599c7e4a37 100644
--- a/arch/powerpc/kernel/process.c
+++ b/arch/powerpc/kernel/process.c
@@ -1473,10 +1473,32 @@ void flush_thread(void)
 #ifdef CONFIG_PPC_BOOK3S_64
 void arch_setup_new_exec(void)
 {
-	if (radix_enabled())
-		return;
-	hash__setup_new_exec();
+	if (!radix_enabled())
+		hash__setup_new_exec();
+
+	/*
+	 * If we exec out of a kernel thread then thread.regs will not be
+	 * set.  Do it now.
+	 */
+	if (!current->thread.regs) {
+		struct pt_regs *regs = task_stack_page(current) + THREAD_SIZE;
+		current->thread.regs = regs - 1;
+	}
+
 }
+#else
+void arch_setup_new_exec(void)
+{
+	/*
+	 * If we exec out of a kernel thread then thread.regs will not be
+	 * set.  Do it now.
+	 */
+	if (!current->thread.regs) {
+		struct pt_regs *regs = task_stack_page(current) + THREAD_SIZE;
+		current->thread.regs = regs - 1;
+	}
+}
+
 #endif
 
 #ifdef CONFIG_PPC64
@@ -1704,15 +1726,6 @@ void start_thread(struct pt_regs *regs, unsigned long start, unsigned long sp)
 #endif
 #endif
 
-	/*
-	 * If we exec out of a kernel thread then thread.regs will not be
-	 * set.  Do it now.
-	 */
-	if (!current->thread.regs) {
-		struct pt_regs *regs = task_stack_page(current) + THREAD_SIZE;
-		current->thread.regs = regs - 1;
-	}
-
 #ifdef CONFIG_PPC_TRANSACTIONAL_MEM
 	/*
 	 * Clear any transactional state, we're exec()ing. The cause is
-- 
2.26.2

[PATCH v4 24/41] powerpc/book3s64/pkeys: Store/restore userspace AMR correctly on entry and exit from kernel

[Aneesh Kumar K.V]

This prepare kernel to operate with a different value than userspace AMR.
For this, AMR needs to be saved and restored on entry and return from the
kernel.

With KUAP we modify kernel AMR when accessing user address from the kernel
via copy_to/from_user interfaces.

If MMU_FTR_KEY is enabled we always use the key mechanism to implement KUAP
feature. If MMU_FTR_KEY is not supported and if we support MMU_FTR_KUAP
(radix translation on POWER9), we can skip restoring AMR on return
to userspace. Userspace won't be using AMR in that specific config.

Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
---
 arch/powerpc/include/asm/book3s/64/kup.h | 141 ++++++++++++++++++-----
 arch/powerpc/kernel/entry_64.S           |   6 +-
 arch/powerpc/kernel/exceptions-64s.S     |   4 +-
 arch/powerpc/kernel/syscall_64.c         |  26 ++++-
 4 files changed, 144 insertions(+), 33 deletions(-)

diff --git a/arch/powerpc/include/asm/book3s/64/kup.h b/arch/powerpc/include/asm/book3s/64/kup.h
index e6ee1c34842f..6979cd1a0003 100644
--- a/arch/powerpc/include/asm/book3s/64/kup.h
+++ b/arch/powerpc/include/asm/book3s/64/kup.h
@@ -13,18 +13,47 @@
 
 #ifdef __ASSEMBLY__
 
-.macro kuap_restore_amr	gpr1, gpr2
-#ifdef CONFIG_PPC_KUAP
+.macro kuap_restore_user_amr gpr1
+#if defined(CONFIG_PPC_MEM_KEYS)
 	BEGIN_MMU_FTR_SECTION_NESTED(67)
-	mfspr	\gpr1, SPRN_AMR
+	/*
+	 * AMR is going to be different when
+	 * returning to userspace.
+	 */
+	ld	\gpr1, STACK_REGS_KUAP(r1)
+	isync
+	mtspr	SPRN_AMR, \gpr1
+
+	/* No isync required, see kuap_restore_user_amr() */
+	END_MMU_FTR_SECTION_NESTED_IFSET(MMU_FTR_PKEY , 67)
+#endif
+.endm
+
+.macro kuap_restore_kernel_amr	gpr1, gpr2
+#if defined(CONFIG_PPC_MEM_KEYS)
+	BEGIN_MMU_FTR_SECTION_NESTED(67)
+	b	99f  // handle_pkey_restore_amr
+	END_MMU_FTR_SECTION_NESTED_IFSET(MMU_FTR_PKEY , 67)
+
+	BEGIN_MMU_FTR_SECTION_NESTED(68)
+	b	99f  // handle_kuap_restore_amr
+	MMU_FTR_SECTION_ELSE_NESTED(68)
+	b	100f  // skip_restore_amr
+	ALT_MMU_FTR_SECTION_END_NESTED_IFSET(MMU_FTR_KUAP, 68)
+
+99:
+	/*
+	 * AMR is going to be mostly the same since we are
+	 * returning to the kernel. Compare and do a mtspr.
+	 */
 	ld	\gpr2, STACK_REGS_KUAP(r1)
+	mfspr	\gpr1, SPRN_AMR
 	cmpd	\gpr1, \gpr2
-	beq	998f
+	beq	100f
 	isync
 	mtspr	SPRN_AMR, \gpr2
 	/* No isync required, see kuap_restore_amr() */
-998:
-	END_MMU_FTR_SECTION_NESTED_IFSET(MMU_FTR_KUAP, 67)
+100:  // skip_restore_amr
 #endif
 .endm
 
@@ -40,23 +69,58 @@
 #endif
 .endm
 
+/*
+ * MMU_FTR_PKEY and MMU_FTR_KUAP can both be enabled on a platform. We prefer
+ * PKEY over KUAP if both can be enabled on the platform.
+ *
+ * With KUAP only enabled on exception if we are coming from userspace we don't
+ * save the AMR at all, because the expectation is that userspace can't change
+ * the AMR if KUAP feature is enabled.
+ */
 .macro kuap_save_amr_and_lock gpr1, gpr2, use_cr, msr_pr_cr
-#ifdef CONFIG_PPC_KUAP
+#if defined(CONFIG_PPC_MEM_KEYS)
+
 	BEGIN_MMU_FTR_SECTION_NESTED(67)
+	b	101f   // handle_pkey_save_amr
+        END_MMU_FTR_SECTION_NESTED_IFSET(MMU_FTR_PKEY , 67)
+
+	BEGIN_MMU_FTR_SECTION_NESTED(68)
+	b	99f  // handle_kuap_save_amr
+	MMU_FTR_SECTION_ELSE_NESTED(68)
+	b	100f  // skip_save_amr
+	ALT_MMU_FTR_SECTION_END_NESTED_IFSET(MMU_FTR_KUAP, 68)
+
+
+99: // handle_kuap_save_amr
 	.ifnb \msr_pr_cr
-	bne	\msr_pr_cr, 99f
+	/*
+	 * We avoid changing AMR outside the kernel
+	 * hence skip this completely.
+	 */
+	bne	\msr_pr_cr, 100f  // from userspace
 	.endif
+
+101:   // handle_pkey_save_amr
 	mfspr	\gpr1, SPRN_AMR
 	std	\gpr1, STACK_REGS_KUAP(r1)
-	li	\gpr2, (AMR_KUAP_BLOCKED >> AMR_KUAP_SHIFT)
-	sldi	\gpr2, \gpr2, AMR_KUAP_SHIFT
+
+	/*
+	 * update kernel AMR with AMR_KUAP_BLOCKED only
+	 * if KUAP feature is enabled
+	 */
+	BEGIN_MMU_FTR_SECTION_NESTED(69)
+	LOAD_REG_IMMEDIATE(\gpr2, AMR_KUAP_BLOCKED)
 	cmpd	\use_cr, \gpr1, \gpr2
-	beq	\use_cr, 99f
-	// We don't isync here because we very recently entered via rfid
+	beq	\use_cr, 102f
+	/*
+	 * We don't isync here because we very recently entered via an interrupt
+	 */
 	mtspr	SPRN_AMR, \gpr2
 	isync
-99:
-	END_MMU_FTR_SECTION_NESTED_IFSET(MMU_FTR_KUAP, 67)
+102:
+	END_MMU_FTR_SECTION_NESTED_IFSET(MMU_FTR_KUAP, 69)
+
+100: // skip_save_amr
 #endif
 .endm
 
@@ -69,22 +133,40 @@
 
 extern u64 default_uamor;
 
-static inline void kuap_restore_amr(struct pt_regs *regs, unsigned long amr)
+static inline void kuap_restore_user_amr(struct pt_regs *regs)
 {
-	if (mmu_has_feature(MMU_FTR_KUAP) && unlikely(regs->kuap != amr)) {
-		isync();
-		mtspr(SPRN_AMR, regs->kuap);
-		/*
-		 * No isync required here because we are about to RFI back to
-		 * previous context before any user accesses would be made,
-		 * which is a CSI.
-		 */
+	if (!mmu_has_feature(MMU_FTR_PKEY))
+		return;
+
+	isync();
+	mtspr(SPRN_AMR, regs->kuap);
+	/*
+	 * No isync required here because we are about to rfi
+	 * back to previous context before any user accesses
+	 * would be made, which is a CSI.
+	 */
+}
+
+static inline void kuap_restore_kernel_amr(struct pt_regs *regs,
+					   unsigned long amr)
+{
+	if (mmu_has_feature(MMU_FTR_KUAP) || mmu_has_feature(MMU_FTR_PKEY)) {
+
+		if (unlikely(regs->kuap != amr)) {
+			isync();
+			mtspr(SPRN_AMR, regs->kuap);
+			/*
+			 * No isync required here because we are about to rfi
+			 * back to previous context before any user accesses
+			 * would be made, which is a CSI.
+			 */
+		}
 	}
 }
 
 static inline unsigned long kuap_get_and_check_amr(void)
 {
-	if (mmu_has_feature(MMU_FTR_KUAP)) {
+	if (mmu_has_feature(MMU_FTR_KUAP) || mmu_has_feature(MMU_FTR_PKEY)) {
 		unsigned long amr = mfspr(SPRN_AMR);
 		if (IS_ENABLED(CONFIG_PPC_KUAP_DEBUG)) /* kuap_check_amr() */
 			WARN_ON_ONCE(amr != AMR_KUAP_BLOCKED);
@@ -95,13 +177,18 @@ static inline unsigned long kuap_get_and_check_amr(void)
 
 static inline void kuap_check_amr(void)
 {
-	if (IS_ENABLED(CONFIG_PPC_KUAP_DEBUG) && mmu_has_feature(MMU_FTR_KUAP))
+	if (IS_ENABLED(CONFIG_PPC_KUAP_DEBUG) &&
+	    (mmu_has_feature(MMU_FTR_KUAP) || mmu_has_feature(MMU_FTR_PKEY)))
 		WARN_ON_ONCE(mfspr(SPRN_AMR) != AMR_KUAP_BLOCKED);
 }
 
 #else /* CONFIG_PPC_MEM_KEYS */
 
-static inline void kuap_restore_amr(struct pt_regs *regs, unsigned long amr)
+static inline void kuap_restore_user_amr(struct pt_regs *regs)
+{
+}
+
+static inline void kuap_restore_kernel_amr(struct pt_regs *regs, unsigned long amr)
 {
 }
 
@@ -113,6 +200,7 @@ static inline unsigned long kuap_get_and_check_amr(void)
 {
 	return 0;
 }
+
 #endif /* CONFIG_PPC_MEM_KEYS */
 
 
@@ -187,7 +275,6 @@ bad_kuap_fault(struct pt_regs *regs, unsigned long address, bool is_write)
 		    "Bug: %s fault blocked by AMR!", is_write ? "Write" : "Read");
 }
 #endif /* CONFIG_PPC_KUAP */
-
 #endif /* __ASSEMBLY__ */
 
 #endif /* _ASM_POWERPC_BOOK3S_64_KUP_H */
diff --git a/arch/powerpc/kernel/entry_64.S b/arch/powerpc/kernel/entry_64.S
index 9d49338e0c85..a087cbe0b17d 100644
--- a/arch/powerpc/kernel/entry_64.S
+++ b/arch/powerpc/kernel/entry_64.S
@@ -481,8 +481,8 @@ _ASM_NOKPROBE_SYMBOL(fast_interrupt_return)
 	kuap_check_amr r3, r4
 	ld	r5,_MSR(r1)
 	andi.	r0,r5,MSR_PR
-	bne	.Lfast_user_interrupt_return
-	kuap_restore_amr r3, r4
+	bne	.Lfast_user_interrupt_return_amr
+	kuap_restore_kernel_amr r3, r4
 	andi.	r0,r5,MSR_RI
 	li	r3,0 /* 0 return value, no EMULATE_STACK_STORE */
 	bne+	.Lfast_kernel_interrupt_return
@@ -502,6 +502,8 @@ _ASM_NOKPROBE_SYMBOL(interrupt_return)
 	cmpdi	r3,0
 	bne-	.Lrestore_nvgprs
 
+.Lfast_user_interrupt_return_amr:
+	kuap_restore_user_amr r3
 .Lfast_user_interrupt_return:
 	ld	r11,_NIP(r1)
 	ld	r12,_MSR(r1)
diff --git a/arch/powerpc/kernel/exceptions-64s.S b/arch/powerpc/kernel/exceptions-64s.S
index e70ebb5c318c..8226af444d77 100644
--- a/arch/powerpc/kernel/exceptions-64s.S
+++ b/arch/powerpc/kernel/exceptions-64s.S
@@ -971,7 +971,7 @@ EXC_COMMON_BEGIN(system_reset_common)
 	ld	r10,SOFTE(r1)
 	stb	r10,PACAIRQSOFTMASK(r13)
 
-	kuap_restore_amr r9, r10
+	kuap_restore_kernel_amr r9, r10
 	EXCEPTION_RESTORE_REGS
 	RFI_TO_USER_OR_KERNEL
 
@@ -2784,7 +2784,7 @@ EXC_COMMON_BEGIN(soft_nmi_common)
 	ld	r10,SOFTE(r1)
 	stb	r10,PACAIRQSOFTMASK(r13)
 
-	kuap_restore_amr r9, r10
+	kuap_restore_kernel_amr r9, r10
 	EXCEPTION_RESTORE_REGS hsrr=0
 	RFI_TO_KERNEL
 
diff --git a/arch/powerpc/kernel/syscall_64.c b/arch/powerpc/kernel/syscall_64.c
index 7e560a01afa4..fded67982fbe 100644
--- a/arch/powerpc/kernel/syscall_64.c
+++ b/arch/powerpc/kernel/syscall_64.c
@@ -35,7 +35,21 @@ notrace long system_call_exception(long r3, long r4, long r5,
 	BUG_ON(!FULL_REGS(regs));
 	BUG_ON(regs->softe != IRQS_ENABLED);
 
-	kuap_check_amr();
+#ifdef CONFIG_PPC_MEM_KEYS
+	if (mmu_has_feature(MMU_FTR_PKEY)) {
+		unsigned long amr;
+		/*
+		 * When entering from userspace we mostly have the AMR
+		 * different from kernel default values. Hence don't compare.
+		 */
+		amr = mfspr(SPRN_AMR);
+		regs->kuap = amr;
+		if (mmu_has_feature(MMU_FTR_KUAP))
+			mtspr(SPRN_AMR, AMR_KUAP_BLOCKED);
+		isync();
+	} else
+#endif
+		kuap_check_amr();
 
 	account_cpu_user_entry();
 
@@ -222,6 +236,10 @@ notrace unsigned long syscall_exit_prepare(unsigned long r3,
 
 	account_cpu_user_exit();
 
+	/*
+	 * We do this at the end so that we do context switch with KERNEL AMR
+	 */
+	kuap_restore_user_amr(regs);
 	return ret;
 }
 
@@ -306,6 +324,10 @@ notrace unsigned long interrupt_exit_user_prepare(struct pt_regs *regs, unsigned
 
 	account_cpu_user_exit();
 
+	/*
+	 * We do this at the end so that we do context switch with KERNEL AMR
+	 */
+	kuap_restore_user_amr(regs);
 	return ret;
 }
 
@@ -376,7 +398,7 @@ notrace unsigned long interrupt_exit_kernel_prepare(struct pt_regs *regs, unsign
 	 * which would cause Read-After-Write stalls. Hence, we take the AMR
 	 * value from the check above.
 	 */
-	kuap_restore_amr(regs, amr);
+	kuap_restore_kernel_amr(regs, amr);
 
 	return ret;
 }
-- 
2.26.2

Re: [PATCH v4 24/41] powerpc/book3s64/pkeys: Store/restore userspace AMR correctly on entry and exit from kernel

[Nicholas Piggin]

Excerpts from Aneesh Kumar K.V's message of June 15, 2020 4:14 pm:
> This prepare kernel to operate with a different value than userspace AMR.
> For this, AMR needs to be saved and restored on entry and return from the
> kernel.
> 
> With KUAP we modify kernel AMR when accessing user address from the kernel
> via copy_to/from_user interfaces.
> 
> If MMU_FTR_KEY is enabled we always use the key mechanism to implement KUAP
> feature. If MMU_FTR_KEY is not supported and if we support MMU_FTR_KUAP
> (radix translation on POWER9), we can skip restoring AMR on return
> to userspace. Userspace won't be using AMR in that specific config.
> 
> Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
> ---
>  arch/powerpc/include/asm/book3s/64/kup.h | 141 ++++++++++++++++++-----
>  arch/powerpc/kernel/entry_64.S           |   6 +-
>  arch/powerpc/kernel/exceptions-64s.S     |   4 +-
>  arch/powerpc/kernel/syscall_64.c         |  26 ++++-
>  4 files changed, 144 insertions(+), 33 deletions(-)
> 
> diff --git a/arch/powerpc/include/asm/book3s/64/kup.h b/arch/powerpc/include/asm/book3s/64/kup.h
> index e6ee1c34842f..6979cd1a0003 100644
> --- a/arch/powerpc/include/asm/book3s/64/kup.h
> +++ b/arch/powerpc/include/asm/book3s/64/kup.h
> @@ -13,18 +13,47 @@
>  
>  #ifdef __ASSEMBLY__
>  
> -.macro kuap_restore_amr	gpr1, gpr2
> -#ifdef CONFIG_PPC_KUAP
> +.macro kuap_restore_user_amr gpr1
> +#if defined(CONFIG_PPC_MEM_KEYS)
>  	BEGIN_MMU_FTR_SECTION_NESTED(67)
> -	mfspr	\gpr1, SPRN_AMR
> +	/*
> +	 * AMR is going to be different when
> +	 * returning to userspace.
> +	 */
> +	ld	\gpr1, STACK_REGS_KUAP(r1)
> +	isync
> +	mtspr	SPRN_AMR, \gpr1
> +
> +	/* No isync required, see kuap_restore_user_amr() */
> +	END_MMU_FTR_SECTION_NESTED_IFSET(MMU_FTR_PKEY , 67)
> +#endif
> +.endm
> +
> +.macro kuap_restore_kernel_amr	gpr1, gpr2
> +#if defined(CONFIG_PPC_MEM_KEYS)
> +	BEGIN_MMU_FTR_SECTION_NESTED(67)
> +	b	99f  // handle_pkey_restore_amr
> +	END_MMU_FTR_SECTION_NESTED_IFSET(MMU_FTR_PKEY , 67)
> +
> +	BEGIN_MMU_FTR_SECTION_NESTED(68)
> +	b	99f  // handle_kuap_restore_amr
> +	MMU_FTR_SECTION_ELSE_NESTED(68)
> +	b	100f  // skip_restore_amr
> +	ALT_MMU_FTR_SECTION_END_NESTED_IFSET(MMU_FTR_KUAP, 68)
> +
> +99:
> +	/*
> +	 * AMR is going to be mostly the same since we are
> +	 * returning to the kernel. Compare and do a mtspr.
> +	 */
>  	ld	\gpr2, STACK_REGS_KUAP(r1)
> +	mfspr	\gpr1, SPRN_AMR
>  	cmpd	\gpr1, \gpr2
> -	beq	998f
> +	beq	100f
>  	isync
>  	mtspr	SPRN_AMR, \gpr2
>  	/* No isync required, see kuap_restore_amr() */
> -998:
> -	END_MMU_FTR_SECTION_NESTED_IFSET(MMU_FTR_KUAP, 67)
> +100:  // skip_restore_amr

Can't you code it like this? (_IFCLR requires none of the bits to be 
set)

BEGIN_MMU_FTR_SECTION_NESTED(67)
	b	99f	// nothing using AMR, no need to restore
END_MMU_FTR_SECTION_NESTED_IFCLR(MMU_FTR_PKEY | MMU_FTR_KUAP, 67)

That saves you a branch in the common case of using AMR. Similar
for others.

> @@ -69,22 +133,40 @@
>  
>  extern u64 default_uamor;
>  
> -static inline void kuap_restore_amr(struct pt_regs *regs, unsigned long amr)
> +static inline void kuap_restore_user_amr(struct pt_regs *regs)
>  {
> -	if (mmu_has_feature(MMU_FTR_KUAP) && unlikely(regs->kuap != amr)) {
> -		isync();
> -		mtspr(SPRN_AMR, regs->kuap);
> -		/*
> -		 * No isync required here because we are about to RFI back to
> -		 * previous context before any user accesses would be made,
> -		 * which is a CSI.
> -		 */
> +	if (!mmu_has_feature(MMU_FTR_PKEY))
> +		return;

If you have PKEY but not KUAP, do you still have to restore?

> +
> +	isync();
> +	mtspr(SPRN_AMR, regs->kuap);
> +	/*
> +	 * No isync required here because we are about to rfi
> +	 * back to previous context before any user accesses
> +	 * would be made, which is a CSI.
> +	 */
> +}
> +
> +static inline void kuap_restore_kernel_amr(struct pt_regs *regs,
> +					   unsigned long amr)
> +{
> +	if (mmu_has_feature(MMU_FTR_KUAP) || mmu_has_feature(MMU_FTR_PKEY)) {
> +
> +		if (unlikely(regs->kuap != amr)) {
> +			isync();
> +			mtspr(SPRN_AMR, regs->kuap);
> +			/*
> +			 * No isync required here because we are about to rfi
> +			 * back to previous context before any user accesses
> +			 * would be made, which is a CSI.
> +			 */
> +		}
>  	}
>  }
>  
>  static inline unsigned long kuap_get_and_check_amr(void)
>  {
> -	if (mmu_has_feature(MMU_FTR_KUAP)) {
> +	if (mmu_has_feature(MMU_FTR_KUAP) || mmu_has_feature(MMU_FTR_PKEY)) {
>  		unsigned long amr = mfspr(SPRN_AMR);
>  		if (IS_ENABLED(CONFIG_PPC_KUAP_DEBUG)) /* kuap_check_amr() */
>  			WARN_ON_ONCE(amr != AMR_KUAP_BLOCKED);

We could do a static key that's based on this condition, but that can 
wait for another day.

>  
>  static inline void kuap_check_amr(void)
>  {
> -	if (IS_ENABLED(CONFIG_PPC_KUAP_DEBUG) && mmu_has_feature(MMU_FTR_KUAP))
> +	if (IS_ENABLED(CONFIG_PPC_KUAP_DEBUG) &&
> +	    (mmu_has_feature(MMU_FTR_KUAP) || mmu_has_feature(MMU_FTR_PKEY)))
>  		WARN_ON_ONCE(mfspr(SPRN_AMR) != AMR_KUAP_BLOCKED);
>  }
>  
>  #else /* CONFIG_PPC_MEM_KEYS */
>  
> -static inline void kuap_restore_amr(struct pt_regs *regs, unsigned long amr)
> +static inline void kuap_restore_user_amr(struct pt_regs *regs)
> +{
> +}
> +
> +static inline void kuap_restore_kernel_amr(struct pt_regs *regs, unsigned long amr)
>  {
>  }
>  
> @@ -113,6 +200,7 @@ static inline unsigned long kuap_get_and_check_amr(void)
>  {
>  	return 0;
>  }
> +
>  #endif /* CONFIG_PPC_MEM_KEYS */
>  
>  
> @@ -187,7 +275,6 @@ bad_kuap_fault(struct pt_regs *regs, unsigned long address, bool is_write)
>  		    "Bug: %s fault blocked by AMR!", is_write ? "Write" : "Read");
>  }
>  #endif /* CONFIG_PPC_KUAP */
> -
>  #endif /* __ASSEMBLY__ */
>  
>  #endif /* _ASM_POWERPC_BOOK3S_64_KUP_H */

Unnecessary hunks.

> diff --git a/arch/powerpc/kernel/entry_64.S b/arch/powerpc/kernel/entry_64.S
> index 9d49338e0c85..a087cbe0b17d 100644
> --- a/arch/powerpc/kernel/entry_64.S
> +++ b/arch/powerpc/kernel/entry_64.S
> @@ -481,8 +481,8 @@ _ASM_NOKPROBE_SYMBOL(fast_interrupt_return)
>  	kuap_check_amr r3, r4
>  	ld	r5,_MSR(r1)
>  	andi.	r0,r5,MSR_PR
> -	bne	.Lfast_user_interrupt_return
> -	kuap_restore_amr r3, r4
> +	bne	.Lfast_user_interrupt_return_amr
> +	kuap_restore_kernel_amr r3, r4
>  	andi.	r0,r5,MSR_RI
>  	li	r3,0 /* 0 return value, no EMULATE_STACK_STORE */
>  	bne+	.Lfast_kernel_interrupt_return
> @@ -502,6 +502,8 @@ _ASM_NOKPROBE_SYMBOL(interrupt_return)
>  	cmpdi	r3,0
>  	bne-	.Lrestore_nvgprs
>  
> +.Lfast_user_interrupt_return_amr:
> +	kuap_restore_user_amr r3
>  .Lfast_user_interrupt_return:
>  	ld	r11,_NIP(r1)
>  	ld	r12,_MSR(r1)
> diff --git a/arch/powerpc/kernel/exceptions-64s.S b/arch/powerpc/kernel/exceptions-64s.S
> index e70ebb5c318c..8226af444d77 100644
> --- a/arch/powerpc/kernel/exceptions-64s.S
> +++ b/arch/powerpc/kernel/exceptions-64s.S
> @@ -971,7 +971,7 @@ EXC_COMMON_BEGIN(system_reset_common)
>  	ld	r10,SOFTE(r1)
>  	stb	r10,PACAIRQSOFTMASK(r13)
>  
> -	kuap_restore_amr r9, r10
> +	kuap_restore_kernel_amr r9, r10
>  	EXCEPTION_RESTORE_REGS
>  	RFI_TO_USER_OR_KERNEL
>  
> @@ -2784,7 +2784,7 @@ EXC_COMMON_BEGIN(soft_nmi_common)
>  	ld	r10,SOFTE(r1)
>  	stb	r10,PACAIRQSOFTMASK(r13)
>  
> -	kuap_restore_amr r9, r10
> +	kuap_restore_kernel_amr r9, r10
>  	EXCEPTION_RESTORE_REGS hsrr=0
>  	RFI_TO_KERNEL
>  
> diff --git a/arch/powerpc/kernel/syscall_64.c b/arch/powerpc/kernel/syscall_64.c
> index 7e560a01afa4..fded67982fbe 100644
> --- a/arch/powerpc/kernel/syscall_64.c
> +++ b/arch/powerpc/kernel/syscall_64.c
> @@ -35,7 +35,21 @@ notrace long system_call_exception(long r3, long r4, long r5,
>  	BUG_ON(!FULL_REGS(regs));
>  	BUG_ON(regs->softe != IRQS_ENABLED);
>  
> -	kuap_check_amr();
> +#ifdef CONFIG_PPC_MEM_KEYS
> +	if (mmu_has_feature(MMU_FTR_PKEY)) {
> +		unsigned long amr;
> +		/*
> +		 * When entering from userspace we mostly have the AMR
> +		 * different from kernel default values. Hence don't compare.
> +		 */
> +		amr = mfspr(SPRN_AMR);
> +		regs->kuap = amr;
> +		if (mmu_has_feature(MMU_FTR_KUAP))
> +			mtspr(SPRN_AMR, AMR_KUAP_BLOCKED);
> +		isync();

isync should be inside the if(). Again do pkeys need to save this if 
KUAP is not being used? I haven't really looked at how all that works, 
but what's changing for the PKEY && !KUAP case?

This would be nice if it could all go into a wrapper function rather 
than ifdef.

> +	} else
> +#endif
> +		kuap_check_amr();
>  
>  	account_cpu_user_entry();
>  
> @@ -222,6 +236,10 @@ notrace unsigned long syscall_exit_prepare(unsigned long r3,
>  
>  	account_cpu_user_exit();
>  
> +	/*
> +	 * We do this at the end so that we do context switch with KERNEL AMR
> +	 */
> +	kuap_restore_user_amr(regs);
>  	return ret;

Comment doesn't make sense, newline required before return.

>  }
>  
> @@ -306,6 +324,10 @@ notrace unsigned long interrupt_exit_user_prepare(struct pt_regs *regs, unsigned
>  
>  	account_cpu_user_exit();
>  
> +	/*
> +	 * We do this at the end so that we do context switch with KERNEL AMR
> +	 */
> +	kuap_restore_user_amr(regs);

Duplicated comments I prefer to just have like this instead of trying to 
keep them in sync. Can complete the circular reference by having a

* similarly in interrupt_exit_user_prepare

in the main comment, but if they come close to one another in the same 
file it's not so important to keep them together.

 +	kuap_restore_user_amr(regs); /* see syscall_exit_prepare */

>  	return ret;
>  }
>  
> @@ -376,7 +398,7 @@ notrace unsigned long interrupt_exit_kernel_prepare(struct pt_regs *regs, unsign
>  	 * which would cause Read-After-Write stalls. Hence, we take the AMR
>  	 * value from the check above.
>  	 */
> -	kuap_restore_amr(regs, amr);
> +	kuap_restore_kernel_amr(regs, amr);
>  
>  	return ret;
>  }
> -- 
> 2.26.2
> 
> 

Re: [PATCH v4 24/41] powerpc/book3s64/pkeys: Store/restore userspace AMR correctly on entry and exit from kernel

[Aneesh Kumar K.V]

On 7/3/20 2:48 PM, Nicholas Piggin wrote:
> Excerpts from Aneesh Kumar K.V's message of June 15, 2020 4:14 pm:
>> This prepare kernel to operate with a different value than userspace AMR.
>> For this, AMR needs to be saved and restored on entry and return from the
>> kernel.
>>
>> With KUAP we modify kernel AMR when accessing user address from the kernel
>> via copy_to/from_user interfaces.
>>
>> If MMU_FTR_KEY is enabled we always use the key mechanism to implement KUAP
>> feature. If MMU_FTR_KEY is not supported and if we support MMU_FTR_KUAP
>> (radix translation on POWER9), we can skip restoring AMR on return
>> to userspace. Userspace won't be using AMR in that specific config.
>>
>> Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
>> ---
>>   arch/powerpc/include/asm/book3s/64/kup.h | 141 ++++++++++++++++++-----
>>   arch/powerpc/kernel/entry_64.S           |   6 +-
>>   arch/powerpc/kernel/exceptions-64s.S     |   4 +-
>>   arch/powerpc/kernel/syscall_64.c         |  26 ++++-
>>   4 files changed, 144 insertions(+), 33 deletions(-)
>>
>> diff --git a/arch/powerpc/include/asm/book3s/64/kup.h b/arch/powerpc/include/asm/book3s/64/kup.h
>> index e6ee1c34842f..6979cd1a0003 100644
>> --- a/arch/powerpc/include/asm/book3s/64/kup.h
>> +++ b/arch/powerpc/include/asm/book3s/64/kup.h
>> @@ -13,18 +13,47 @@
>>   
>>   #ifdef __ASSEMBLY__
>>   
>> -.macro kuap_restore_amr	gpr1, gpr2
>> -#ifdef CONFIG_PPC_KUAP
>> +.macro kuap_restore_user_amr gpr1
>> +#if defined(CONFIG_PPC_MEM_KEYS)
>>   	BEGIN_MMU_FTR_SECTION_NESTED(67)
>> -	mfspr	\gpr1, SPRN_AMR
>> +	/*
>> +	 * AMR is going to be different when
>> +	 * returning to userspace.
>> +	 */
>> +	ld	\gpr1, STACK_REGS_KUAP(r1)
>> +	isync
>> +	mtspr	SPRN_AMR, \gpr1
>> +
>> +	/* No isync required, see kuap_restore_user_amr() */
>> +	END_MMU_FTR_SECTION_NESTED_IFSET(MMU_FTR_PKEY , 67)
>> +#endif
>> +.endm
>> +
>> +.macro kuap_restore_kernel_amr	gpr1, gpr2
>> +#if defined(CONFIG_PPC_MEM_KEYS)
>> +	BEGIN_MMU_FTR_SECTION_NESTED(67)
>> +	b	99f  // handle_pkey_restore_amr
>> +	END_MMU_FTR_SECTION_NESTED_IFSET(MMU_FTR_PKEY , 67)
>> +
>> +	BEGIN_MMU_FTR_SECTION_NESTED(68)
>> +	b	99f  // handle_kuap_restore_amr
>> +	MMU_FTR_SECTION_ELSE_NESTED(68)
>> +	b	100f  // skip_restore_amr
>> +	ALT_MMU_FTR_SECTION_END_NESTED_IFSET(MMU_FTR_KUAP, 68)
>> +
>> +99:
>> +	/*
>> +	 * AMR is going to be mostly the same since we are
>> +	 * returning to the kernel. Compare and do a mtspr.
>> +	 */
>>   	ld	\gpr2, STACK_REGS_KUAP(r1)
>> +	mfspr	\gpr1, SPRN_AMR
>>   	cmpd	\gpr1, \gpr2
>> -	beq	998f
>> +	beq	100f
>>   	isync
>>   	mtspr	SPRN_AMR, \gpr2
>>   	/* No isync required, see kuap_restore_amr() */
>> -998:
>> -	END_MMU_FTR_SECTION_NESTED_IFSET(MMU_FTR_KUAP, 67)
>> +100:  // skip_restore_amr
> 
> Can't you code it like this? (_IFCLR requires none of the bits to be
> set)
> 
> BEGIN_MMU_FTR_SECTION_NESTED(67)
> 	b	99f	// nothing using AMR, no need to restore
> END_MMU_FTR_SECTION_NESTED_IFCLR(MMU_FTR_PKEY | MMU_FTR_KUAP, 67)
> 
> That saves you a branch in the common case of using AMR. Similar
> for others.



Yes i could switch to that. The code is taking extra 200 cycles even 
with KUAP/KUEP disabled and no keys being used on hash. I am yet to 
analyze this closely. So will rework things based on that analysis.

> 
>> @@ -69,22 +133,40 @@
>>   
>>   extern u64 default_uamor;
>>   
>> -static inline void kuap_restore_amr(struct pt_regs *regs, unsigned long amr)
>> +static inline void kuap_restore_user_amr(struct pt_regs *regs)
>>   {
>> -	if (mmu_has_feature(MMU_FTR_KUAP) && unlikely(regs->kuap != amr)) {
>> -		isync();
>> -		mtspr(SPRN_AMR, regs->kuap);
>> -		/*
>> -		 * No isync required here because we are about to RFI back to
>> -		 * previous context before any user accesses would be made,
>> -		 * which is a CSI.
>> -		 */
>> +	if (!mmu_has_feature(MMU_FTR_PKEY))
>> +		return;
> 
> If you have PKEY but not KUAP, do you still have to restore?
> 


Yes, because user space pkey is now set on the exit path. This is needed 
to handle things like exec/fork().


>> +
>> +	isync();
>> +	mtspr(SPRN_AMR, regs->kuap);
>> +	/*
>> +	 * No isync required here because we are about to rfi
>> +	 * back to previous context before any user accesses
>> +	 * would be made, which is a CSI.
>> +	 */
>> +}
>> +
>> +static inline void kuap_restore_kernel_amr(struct pt_regs *regs,
>> +					   unsigned long amr)
>> +{
>> +	if (mmu_has_feature(MMU_FTR_KUAP) || mmu_has_feature(MMU_FTR_PKEY)) {
>> +
>> +		if (unlikely(regs->kuap != amr)) {
>> +			isync();
>> +			mtspr(SPRN_AMR, regs->kuap);
>> +			/*
>> +			 * No isync required here because we are about to rfi
>> +			 * back to previous context before any user accesses
>> +			 * would be made, which is a CSI.
>> +			 */
>> +		}
>>   	}
>>   }
>>   
>>   static inline unsigned long kuap_get_and_check_amr(void)
>>   {
>> -	if (mmu_has_feature(MMU_FTR_KUAP)) {
>> +	if (mmu_has_feature(MMU_FTR_KUAP) || mmu_has_feature(MMU_FTR_PKEY)) {
>>   		unsigned long amr = mfspr(SPRN_AMR);
>>   		if (IS_ENABLED(CONFIG_PPC_KUAP_DEBUG)) /* kuap_check_amr() */
>>   			WARN_ON_ONCE(amr != AMR_KUAP_BLOCKED);
> 
> We could do a static key that's based on this condition, but that can
> wait for another day.
> 
>>   
>>   static inline void kuap_check_amr(void)
>>   {
>> -	if (IS_ENABLED(CONFIG_PPC_KUAP_DEBUG) && mmu_has_feature(MMU_FTR_KUAP))
>> +	if (IS_ENABLED(CONFIG_PPC_KUAP_DEBUG) &&
>> +	    (mmu_has_feature(MMU_FTR_KUAP) || mmu_has_feature(MMU_FTR_PKEY)))
>>   		WARN_ON_ONCE(mfspr(SPRN_AMR) != AMR_KUAP_BLOCKED);
>>   }
>>   
>>   #else /* CONFIG_PPC_MEM_KEYS */
>>   
>> -static inline void kuap_restore_amr(struct pt_regs *regs, unsigned long amr)
>> +static inline void kuap_restore_user_amr(struct pt_regs *regs)
>> +{
>> +}
>> +
>> +static inline void kuap_restore_kernel_amr(struct pt_regs *regs, unsigned long amr)
>>   {
>>   }
>>   
>> @@ -113,6 +200,7 @@ static inline unsigned long kuap_get_and_check_amr(void)
>>   {
>>   	return 0;
>>   }
>> +
>>   #endif /* CONFIG_PPC_MEM_KEYS */
>>   
>>   
>> @@ -187,7 +275,6 @@ bad_kuap_fault(struct pt_regs *regs, unsigned long address, bool is_write)
>>   		    "Bug: %s fault blocked by AMR!", is_write ? "Write" : "Read");
>>   }
>>   #endif /* CONFIG_PPC_KUAP */
>> -
>>   #endif /* __ASSEMBLY__ */
>>   
>>   #endif /* _ASM_POWERPC_BOOK3S_64_KUP_H */
> 
> Unnecessary hunks.


will remove

> 
>> diff --git a/arch/powerpc/kernel/entry_64.S b/arch/powerpc/kernel/entry_64.S
>> index 9d49338e0c85..a087cbe0b17d 100644
>> --- a/arch/powerpc/kernel/entry_64.S
>> +++ b/arch/powerpc/kernel/entry_64.S
>> @@ -481,8 +481,8 @@ _ASM_NOKPROBE_SYMBOL(fast_interrupt_return)
>>   	kuap_check_amr r3, r4
>>   	ld	r5,_MSR(r1)
>>   	andi.	r0,r5,MSR_PR
>> -	bne	.Lfast_user_interrupt_return
>> -	kuap_restore_amr r3, r4
>> +	bne	.Lfast_user_interrupt_return_amr
>> +	kuap_restore_kernel_amr r3, r4
>>   	andi.	r0,r5,MSR_RI
>>   	li	r3,0 /* 0 return value, no EMULATE_STACK_STORE */
>>   	bne+	.Lfast_kernel_interrupt_return
>> @@ -502,6 +502,8 @@ _ASM_NOKPROBE_SYMBOL(interrupt_return)
>>   	cmpdi	r3,0
>>   	bne-	.Lrestore_nvgprs
>>   
>> +.Lfast_user_interrupt_return_amr:
>> +	kuap_restore_user_amr r3
>>   .Lfast_user_interrupt_return:
>>   	ld	r11,_NIP(r1)
>>   	ld	r12,_MSR(r1)
>> diff --git a/arch/powerpc/kernel/exceptions-64s.S b/arch/powerpc/kernel/exceptions-64s.S
>> index e70ebb5c318c..8226af444d77 100644
>> --- a/arch/powerpc/kernel/exceptions-64s.S
>> +++ b/arch/powerpc/kernel/exceptions-64s.S
>> @@ -971,7 +971,7 @@ EXC_COMMON_BEGIN(system_reset_common)
>>   	ld	r10,SOFTE(r1)
>>   	stb	r10,PACAIRQSOFTMASK(r13)
>>   
>> -	kuap_restore_amr r9, r10
>> +	kuap_restore_kernel_amr r9, r10
>>   	EXCEPTION_RESTORE_REGS
>>   	RFI_TO_USER_OR_KERNEL
>>   
>> @@ -2784,7 +2784,7 @@ EXC_COMMON_BEGIN(soft_nmi_common)
>>   	ld	r10,SOFTE(r1)
>>   	stb	r10,PACAIRQSOFTMASK(r13)
>>   
>> -	kuap_restore_amr r9, r10
>> +	kuap_restore_kernel_amr r9, r10
>>   	EXCEPTION_RESTORE_REGS hsrr=0
>>   	RFI_TO_KERNEL
>>   
>> diff --git a/arch/powerpc/kernel/syscall_64.c b/arch/powerpc/kernel/syscall_64.c
>> index 7e560a01afa4..fded67982fbe 100644
>> --- a/arch/powerpc/kernel/syscall_64.c
>> +++ b/arch/powerpc/kernel/syscall_64.c
>> @@ -35,7 +35,21 @@ notrace long system_call_exception(long r3, long r4, long r5,
>>   	BUG_ON(!FULL_REGS(regs));
>>   	BUG_ON(regs->softe != IRQS_ENABLED);
>>   
>> -	kuap_check_amr();
>> +#ifdef CONFIG_PPC_MEM_KEYS
>> +	if (mmu_has_feature(MMU_FTR_PKEY)) {
>> +		unsigned long amr;
>> +		/*
>> +		 * When entering from userspace we mostly have the AMR
>> +		 * different from kernel default values. Hence don't compare.
>> +		 */
>> +		amr = mfspr(SPRN_AMR);
>> +		regs->kuap = amr;
>> +		if (mmu_has_feature(MMU_FTR_KUAP))
>> +			mtspr(SPRN_AMR, AMR_KUAP_BLOCKED);
>> +		isync();
> 
> isync should be inside the if(). Again do pkeys need to save this if
> KUAP is not being used? I haven't really looked at how all that works,
> but what's changing for the PKEY && !KUAP case?
> 

There is no SPR switch in context switch now and all the AMR/IAMR 
handling is now in the exit to userspace.


> This would be nice if it could all go into a wrapper function rather
> than ifdef.
> 

will do

>> +	} else
>> +#endif
>> +		kuap_check_amr();
>>   
>>   	account_cpu_user_entry();
>>   
>> @@ -222,6 +236,10 @@ notrace unsigned long syscall_exit_prepare(unsigned long r3,
>>   
>>   	account_cpu_user_exit();
>>   
>> +	/*
>> +	 * We do this at the end so that we do context switch with KERNEL AMR
>> +	 */
>> +	kuap_restore_user_amr(regs);
>>   	return ret;
> 
> Comment doesn't make sense, newline required before return.


Ok the detail there was we need to make sure we restore AMR towrads the 
end and make sure all the kernel code continue to run with KERNEL AMR 
value. There is a schedule() call in there with _TIF_NEED_RESCHED. But 
those details are not really relevant. That was me tracking down some 
issues and writing comment around that part of the code. The only real 
detail is switch to userspace AMR in the end.



> 
>>   }
>>   
>> @@ -306,6 +324,10 @@ notrace unsigned long interrupt_exit_user_prepare(struct pt_regs *regs, unsigned
>>   
>>   	account_cpu_user_exit();
>>   
>> +	/*
>> +	 * We do this at the end so that we do context switch with KERNEL AMR
>> +	 */
>> +	kuap_restore_user_amr(regs);
> 
> Duplicated comments I prefer to just have like this instead of trying to
> keep them in sync. Can complete the circular reference by having a
> 
> * similarly in interrupt_exit_user_prepare
> 
> in the main comment, but if they come close to one another in the same
> file it's not so important to keep them together.
> 
>   +	kuap_restore_user_amr(regs); /* see syscall_exit_prepare */
> 
>>   	return ret;
>>   }
>>   
>> @@ -376,7 +398,7 @@ notrace unsigned long interrupt_exit_kernel_prepare(struct pt_regs *regs, unsign
>>   	 * which would cause Read-After-Write stalls. Hence, we take the AMR
>>   	 * value from the check above.
>>   	 */
>> -	kuap_restore_amr(regs, amr);
>> +	kuap_restore_kernel_amr(regs, amr);
>>   
>>   	return ret;
>>   }
>> -- 
>> 2.26.2
>>
>>


-aneesh

Re: [PATCH v4 24/41] powerpc/book3s64/pkeys: Store/restore userspace AMR correctly on entry and exit from kernel

[Nicholas Piggin]

Excerpts from Aneesh Kumar K.V's message of July 3, 2020 7:30 pm:
> On 7/3/20 2:48 PM, Nicholas Piggin wrote:
>> Excerpts from Aneesh Kumar K.V's message of June 15, 2020 4:14 pm:
>>> This prepare kernel to operate with a different value than userspace AMR.
>>> For this, AMR needs to be saved and restored on entry and return from the
>>> kernel.
>>>
>>> With KUAP we modify kernel AMR when accessing user address from the kernel
>>> via copy_to/from_user interfaces.
>>>
>>> If MMU_FTR_KEY is enabled we always use the key mechanism to implement KUAP
>>> feature. If MMU_FTR_KEY is not supported and if we support MMU_FTR_KUAP
>>> (radix translation on POWER9), we can skip restoring AMR on return
>>> to userspace. Userspace won't be using AMR in that specific config.
>>>
>>> Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
>>> ---
>>>   arch/powerpc/include/asm/book3s/64/kup.h | 141 ++++++++++++++++++-----
>>>   arch/powerpc/kernel/entry_64.S           |   6 +-
>>>   arch/powerpc/kernel/exceptions-64s.S     |   4 +-
>>>   arch/powerpc/kernel/syscall_64.c         |  26 ++++-
>>>   4 files changed, 144 insertions(+), 33 deletions(-)
>>>
>>> diff --git a/arch/powerpc/include/asm/book3s/64/kup.h b/arch/powerpc/include/asm/book3s/64/kup.h
>>> index e6ee1c34842f..6979cd1a0003 100644
>>> --- a/arch/powerpc/include/asm/book3s/64/kup.h
>>> +++ b/arch/powerpc/include/asm/book3s/64/kup.h
>>> @@ -13,18 +13,47 @@
>>>   
>>>   #ifdef __ASSEMBLY__
>>>   
>>> -.macro kuap_restore_amr	gpr1, gpr2
>>> -#ifdef CONFIG_PPC_KUAP
>>> +.macro kuap_restore_user_amr gpr1
>>> +#if defined(CONFIG_PPC_MEM_KEYS)
>>>   	BEGIN_MMU_FTR_SECTION_NESTED(67)
>>> -	mfspr	\gpr1, SPRN_AMR
>>> +	/*
>>> +	 * AMR is going to be different when
>>> +	 * returning to userspace.
>>> +	 */
>>> +	ld	\gpr1, STACK_REGS_KUAP(r1)
>>> +	isync
>>> +	mtspr	SPRN_AMR, \gpr1
>>> +
>>> +	/* No isync required, see kuap_restore_user_amr() */
>>> +	END_MMU_FTR_SECTION_NESTED_IFSET(MMU_FTR_PKEY , 67)
>>> +#endif
>>> +.endm
>>> +
>>> +.macro kuap_restore_kernel_amr	gpr1, gpr2
>>> +#if defined(CONFIG_PPC_MEM_KEYS)
>>> +	BEGIN_MMU_FTR_SECTION_NESTED(67)
>>> +	b	99f  // handle_pkey_restore_amr
>>> +	END_MMU_FTR_SECTION_NESTED_IFSET(MMU_FTR_PKEY , 67)
>>> +
>>> +	BEGIN_MMU_FTR_SECTION_NESTED(68)
>>> +	b	99f  // handle_kuap_restore_amr
>>> +	MMU_FTR_SECTION_ELSE_NESTED(68)
>>> +	b	100f  // skip_restore_amr
>>> +	ALT_MMU_FTR_SECTION_END_NESTED_IFSET(MMU_FTR_KUAP, 68)
>>> +
>>> +99:
>>> +	/*
>>> +	 * AMR is going to be mostly the same since we are
>>> +	 * returning to the kernel. Compare and do a mtspr.
>>> +	 */
>>>   	ld	\gpr2, STACK_REGS_KUAP(r1)
>>> +	mfspr	\gpr1, SPRN_AMR
>>>   	cmpd	\gpr1, \gpr2
>>> -	beq	998f
>>> +	beq	100f
>>>   	isync
>>>   	mtspr	SPRN_AMR, \gpr2
>>>   	/* No isync required, see kuap_restore_amr() */
>>> -998:
>>> -	END_MMU_FTR_SECTION_NESTED_IFSET(MMU_FTR_KUAP, 67)
>>> +100:  // skip_restore_amr
>> 
>> Can't you code it like this? (_IFCLR requires none of the bits to be
>> set)
>> 
>> BEGIN_MMU_FTR_SECTION_NESTED(67)
>> 	b	99f	// nothing using AMR, no need to restore
>> END_MMU_FTR_SECTION_NESTED_IFCLR(MMU_FTR_PKEY | MMU_FTR_KUAP, 67)
>> 
>> That saves you a branch in the common case of using AMR. Similar
>> for others.
> 
> 
> 
> Yes i could switch to that. The code is taking extra 200 cycles even 
> with KUAP/KUEP disabled and no keys being used on hash. I am yet to 
> analyze this closely. So will rework things based on that analysis.
> 
>> 
>>> @@ -69,22 +133,40 @@
>>>   
>>>   extern u64 default_uamor;
>>>   
>>> -static inline void kuap_restore_amr(struct pt_regs *regs, unsigned long amr)
>>> +static inline void kuap_restore_user_amr(struct pt_regs *regs)
>>>   {
>>> -	if (mmu_has_feature(MMU_FTR_KUAP) && unlikely(regs->kuap != amr)) {
>>> -		isync();
>>> -		mtspr(SPRN_AMR, regs->kuap);
>>> -		/*
>>> -		 * No isync required here because we are about to RFI back to
>>> -		 * previous context before any user accesses would be made,
>>> -		 * which is a CSI.
>>> -		 */
>>> +	if (!mmu_has_feature(MMU_FTR_PKEY))
>>> +		return;
>> 
>> If you have PKEY but not KUAP, do you still have to restore?
>> 
> 
> 
> Yes, because user space pkey is now set on the exit path. This is needed 
> to handle things like exec/fork().
> 
> 
>>> +
>>> +	isync();
>>> +	mtspr(SPRN_AMR, regs->kuap);
>>> +	/*
>>> +	 * No isync required here because we are about to rfi
>>> +	 * back to previous context before any user accesses
>>> +	 * would be made, which is a CSI.
>>> +	 */
>>> +}
>>> +
>>> +static inline void kuap_restore_kernel_amr(struct pt_regs *regs,
>>> +					   unsigned long amr)
>>> +{
>>> +	if (mmu_has_feature(MMU_FTR_KUAP) || mmu_has_feature(MMU_FTR_PKEY)) {
>>> +
>>> +		if (unlikely(regs->kuap != amr)) {
>>> +			isync();
>>> +			mtspr(SPRN_AMR, regs->kuap);
>>> +			/*
>>> +			 * No isync required here because we are about to rfi
>>> +			 * back to previous context before any user accesses
>>> +			 * would be made, which is a CSI.
>>> +			 */
>>> +		}
>>>   	}
>>>   }
>>>   
>>>   static inline unsigned long kuap_get_and_check_amr(void)
>>>   {
>>> -	if (mmu_has_feature(MMU_FTR_KUAP)) {
>>> +	if (mmu_has_feature(MMU_FTR_KUAP) || mmu_has_feature(MMU_FTR_PKEY)) {
>>>   		unsigned long amr = mfspr(SPRN_AMR);
>>>   		if (IS_ENABLED(CONFIG_PPC_KUAP_DEBUG)) /* kuap_check_amr() */
>>>   			WARN_ON_ONCE(amr != AMR_KUAP_BLOCKED);
>> 
>> We could do a static key that's based on this condition, but that can
>> wait for another day.
>> 
>>>   
>>>   static inline void kuap_check_amr(void)
>>>   {
>>> -	if (IS_ENABLED(CONFIG_PPC_KUAP_DEBUG) && mmu_has_feature(MMU_FTR_KUAP))
>>> +	if (IS_ENABLED(CONFIG_PPC_KUAP_DEBUG) &&
>>> +	    (mmu_has_feature(MMU_FTR_KUAP) || mmu_has_feature(MMU_FTR_PKEY)))
>>>   		WARN_ON_ONCE(mfspr(SPRN_AMR) != AMR_KUAP_BLOCKED);
>>>   }
>>>   
>>>   #else /* CONFIG_PPC_MEM_KEYS */
>>>   
>>> -static inline void kuap_restore_amr(struct pt_regs *regs, unsigned long amr)
>>> +static inline void kuap_restore_user_amr(struct pt_regs *regs)
>>> +{
>>> +}
>>> +
>>> +static inline void kuap_restore_kernel_amr(struct pt_regs *regs, unsigned long amr)
>>>   {
>>>   }
>>>   
>>> @@ -113,6 +200,7 @@ static inline unsigned long kuap_get_and_check_amr(void)
>>>   {
>>>   	return 0;
>>>   }
>>> +
>>>   #endif /* CONFIG_PPC_MEM_KEYS */
>>>   
>>>   
>>> @@ -187,7 +275,6 @@ bad_kuap_fault(struct pt_regs *regs, unsigned long address, bool is_write)
>>>   		    "Bug: %s fault blocked by AMR!", is_write ? "Write" : "Read");
>>>   }
>>>   #endif /* CONFIG_PPC_KUAP */
>>> -
>>>   #endif /* __ASSEMBLY__ */
>>>   
>>>   #endif /* _ASM_POWERPC_BOOK3S_64_KUP_H */
>> 
>> Unnecessary hunks.
> 
> 
> will remove
> 
>> 
>>> diff --git a/arch/powerpc/kernel/entry_64.S b/arch/powerpc/kernel/entry_64.S
>>> index 9d49338e0c85..a087cbe0b17d 100644
>>> --- a/arch/powerpc/kernel/entry_64.S
>>> +++ b/arch/powerpc/kernel/entry_64.S
>>> @@ -481,8 +481,8 @@ _ASM_NOKPROBE_SYMBOL(fast_interrupt_return)
>>>   	kuap_check_amr r3, r4
>>>   	ld	r5,_MSR(r1)
>>>   	andi.	r0,r5,MSR_PR
>>> -	bne	.Lfast_user_interrupt_return
>>> -	kuap_restore_amr r3, r4
>>> +	bne	.Lfast_user_interrupt_return_amr
>>> +	kuap_restore_kernel_amr r3, r4
>>>   	andi.	r0,r5,MSR_RI
>>>   	li	r3,0 /* 0 return value, no EMULATE_STACK_STORE */
>>>   	bne+	.Lfast_kernel_interrupt_return
>>> @@ -502,6 +502,8 @@ _ASM_NOKPROBE_SYMBOL(interrupt_return)
>>>   	cmpdi	r3,0
>>>   	bne-	.Lrestore_nvgprs
>>>   
>>> +.Lfast_user_interrupt_return_amr:
>>> +	kuap_restore_user_amr r3
>>>   .Lfast_user_interrupt_return:
>>>   	ld	r11,_NIP(r1)
>>>   	ld	r12,_MSR(r1)
>>> diff --git a/arch/powerpc/kernel/exceptions-64s.S b/arch/powerpc/kernel/exceptions-64s.S
>>> index e70ebb5c318c..8226af444d77 100644
>>> --- a/arch/powerpc/kernel/exceptions-64s.S
>>> +++ b/arch/powerpc/kernel/exceptions-64s.S
>>> @@ -971,7 +971,7 @@ EXC_COMMON_BEGIN(system_reset_common)
>>>   	ld	r10,SOFTE(r1)
>>>   	stb	r10,PACAIRQSOFTMASK(r13)
>>>   
>>> -	kuap_restore_amr r9, r10
>>> +	kuap_restore_kernel_amr r9, r10
>>>   	EXCEPTION_RESTORE_REGS
>>>   	RFI_TO_USER_OR_KERNEL
>>>   
>>> @@ -2784,7 +2784,7 @@ EXC_COMMON_BEGIN(soft_nmi_common)
>>>   	ld	r10,SOFTE(r1)
>>>   	stb	r10,PACAIRQSOFTMASK(r13)
>>>   
>>> -	kuap_restore_amr r9, r10
>>> +	kuap_restore_kernel_amr r9, r10
>>>   	EXCEPTION_RESTORE_REGS hsrr=0
>>>   	RFI_TO_KERNEL
>>>   
>>> diff --git a/arch/powerpc/kernel/syscall_64.c b/arch/powerpc/kernel/syscall_64.c
>>> index 7e560a01afa4..fded67982fbe 100644
>>> --- a/arch/powerpc/kernel/syscall_64.c
>>> +++ b/arch/powerpc/kernel/syscall_64.c
>>> @@ -35,7 +35,21 @@ notrace long system_call_exception(long r3, long r4, long r5,
>>>   	BUG_ON(!FULL_REGS(regs));
>>>   	BUG_ON(regs->softe != IRQS_ENABLED);
>>>   
>>> -	kuap_check_amr();
>>> +#ifdef CONFIG_PPC_MEM_KEYS
>>> +	if (mmu_has_feature(MMU_FTR_PKEY)) {
>>> +		unsigned long amr;
>>> +		/*
>>> +		 * When entering from userspace we mostly have the AMR
>>> +		 * different from kernel default values. Hence don't compare.
>>> +		 */
>>> +		amr = mfspr(SPRN_AMR);
>>> +		regs->kuap = amr;
>>> +		if (mmu_has_feature(MMU_FTR_KUAP))
>>> +			mtspr(SPRN_AMR, AMR_KUAP_BLOCKED);
>>> +		isync();
>> 
>> isync should be inside the if(). Again do pkeys need to save this if
>> KUAP is not being used? I haven't really looked at how all that works,
>> but what's changing for the PKEY && !KUAP case?
>> 
> 
> There is no SPR switch in context switch now and all the AMR/IAMR 
> handling is now in the exit to userspace.

If we have pkeys and no kuap, we could keep the switch in context 
switch?

If you don't think it's worth bothering to optimise that case because we 
expect KUAP to be used, that's probably okay although maybe an 
adjustment to the comment (we don't expect userspace to have different
from kernel values if kernel is not using it for KUAP).

>>> +	/*
>>> +	 * We do this at the end so that we do context switch with KERNEL AMR
>>> +	 */
>>> +	kuap_restore_user_amr(regs);
>>>   	return ret;
>> 
>> Comment doesn't make sense, newline required before return.
> 
> 
> Ok the detail there was we need to make sure we restore AMR towrads the 
> end and make sure all the kernel code continue to run with KERNEL AMR 
> value. There is a schedule() call in there with _TIF_NEED_RESCHED. But 
> those details are not really relevant. That was me tracking down some 
> issues and writing comment around that part of the code. The only real 
> detail is switch to userspace AMR in the end.

Yep, I don't think that comment is needed at all. A space before the
return would be nice. I guess after the account_cpu_user_exit is fine,
that thing's a pain anyway that needs to be changed to avoid an SPR
stall I think so I'll look at that afterward anyway.

Thanks,
Nick

[PATCH v4 25/41] powerpc/book3s64/kuep: Store/restore userspace IAMR correctly on entry and exit from kernel

[Aneesh Kumar K.V]

This prepare kernel to operate with a different value than userspace IAMR.
For this, IAMR needs to be saved and restored on entry and return from the
kernel.

If MMU_FTR_KEY is enabled we always use the key mechanism to implement KUEP
feature. If MMU_FTR_KEY is not supported and if we support MMU_FTR_KUEP
(radix translation on POWER9), we can skip restoring IAMR on return
to userspace. Userspace won't be using IAMR in that specific config.

We don't need to save/restore IAMR on reentry into the kernel due to interrupt
because the kernel doesn't modify IAMR internally.

Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
---
 arch/powerpc/include/asm/book3s/64/kup.h | 46 ++++++++++++++++++++++--
 arch/powerpc/include/asm/ptrace.h        |  6 +++-
 arch/powerpc/kernel/asm-offsets.c        |  4 +++
 arch/powerpc/kernel/syscall_64.c         |  8 +++--
 4 files changed, 59 insertions(+), 5 deletions(-)

diff --git a/arch/powerpc/include/asm/book3s/64/kup.h b/arch/powerpc/include/asm/book3s/64/kup.h
index 6979cd1a0003..49182efea945 100644
--- a/arch/powerpc/include/asm/book3s/64/kup.h
+++ b/arch/powerpc/include/asm/book3s/64/kup.h
@@ -17,15 +17,26 @@
 #if defined(CONFIG_PPC_MEM_KEYS)
 	BEGIN_MMU_FTR_SECTION_NESTED(67)
 	/*
-	 * AMR is going to be different when
+	 * AMR and IAMR are going to be different when
 	 * returning to userspace.
 	 */
 	ld	\gpr1, STACK_REGS_KUAP(r1)
 	isync
 	mtspr	SPRN_AMR, \gpr1
+	/*
+	 * Restore IAMR only when returning to userspace
+	 */
+	ld	\gpr1, STACK_REGS_KUEP(r1)
+	mtspr	SPRN_IAMR, \gpr1
 
 	/* No isync required, see kuap_restore_user_amr() */
 	END_MMU_FTR_SECTION_NESTED_IFSET(MMU_FTR_PKEY , 67)
+
+	/*
+	 * We don't check KUEP feature here, because if FTR_PKEY
+	 * is not enabled we don't need to restore IAMR on
+	 * return to usespace.
+	 */
 #endif
 .endm
 
@@ -53,6 +64,9 @@
 	isync
 	mtspr	SPRN_AMR, \gpr2
 	/* No isync required, see kuap_restore_amr() */
+	/*
+	 * No need to restore IAMR when returning to kernel space.
+	 */
 100:  // skip_restore_amr
 #endif
 .endm
@@ -90,6 +104,12 @@
 	b	100f  // skip_save_amr
 	ALT_MMU_FTR_SECTION_END_NESTED_IFSET(MMU_FTR_KUAP, 68)
 
+	/*
+	 * We don't check KUEP feature here, because if FTR_PKEY
+	 * is not enabled we don't need to save IAMR on
+	 * entry from usespace. That is handled by either
+	 * handle_kuap_save_amr or skip_save_amr
+	 */
 
 99: // handle_kuap_save_amr
 	.ifnb \msr_pr_cr
@@ -120,6 +140,25 @@
 102:
 	END_MMU_FTR_SECTION_NESTED_IFSET(MMU_FTR_KUAP, 69)
 
+	.ifnb \msr_pr_cr
+	beq	\msr_pr_cr, 103f // from kernel space
+	mfspr	\gpr1, SPRN_IAMR
+	std	\gpr1, STACK_REGS_KUEP(r1)
+
+	/*
+	 * update kernel IAMR with AMR_KUEP_BLOCKED only
+	 * if KUEP feature is enabled
+	 */
+	BEGIN_MMU_FTR_SECTION_NESTED(70)
+	LOAD_REG_IMMEDIATE(\gpr2, AMR_KUEP_BLOCKED)
+	cmpd	\use_cr, \gpr1, \gpr2
+	beq	\use_cr, 103f
+	mtspr	SPRN_IAMR, \gpr2
+	isync
+103:
+	END_MMU_FTR_SECTION_NESTED_IFSET(MMU_FTR_KUEP, 70)
+	.endif
+
 100: // skip_save_amr
 #endif
 .endm
@@ -140,13 +179,13 @@ static inline void kuap_restore_user_amr(struct pt_regs *regs)
 
 	isync();
 	mtspr(SPRN_AMR, regs->kuap);
+	mtspr(SPRN_IAMR, regs->kuep);
 	/*
 	 * No isync required here because we are about to rfi
 	 * back to previous context before any user accesses
 	 * would be made, which is a CSI.
 	 */
 }
-
 static inline void kuap_restore_kernel_amr(struct pt_regs *regs,
 					   unsigned long amr)
 {
@@ -162,6 +201,9 @@ static inline void kuap_restore_kernel_amr(struct pt_regs *regs,
 			 */
 		}
 	}
+	/*
+	 * No need to restore IAMR when returning to kernel space.
+	 */
 }
 
 static inline unsigned long kuap_get_and_check_amr(void)
diff --git a/arch/powerpc/include/asm/ptrace.h b/arch/powerpc/include/asm/ptrace.h
index 1a6cadf63d14..2b610dc20426 100644
--- a/arch/powerpc/include/asm/ptrace.h
+++ b/arch/powerpc/include/asm/ptrace.h
@@ -56,8 +56,12 @@ struct pt_regs
 #ifdef CONFIG_PPC_HAVE_KUAP
 			unsigned long kuap;
 #endif
+#ifdef CONFIG_PPC_HAVE_KUEP
+			unsigned long kuep;
+#endif
+
 		};
-		unsigned long __pad[2];	/* Maintain 16 byte interrupt stack alignment */
+		unsigned long __pad[4];	/* Maintain 16 byte interrupt stack alignment */
 	};
 };
 #endif
diff --git a/arch/powerpc/kernel/asm-offsets.c b/arch/powerpc/kernel/asm-offsets.c
index 1694c4f531b9..5b68970a52a7 100644
--- a/arch/powerpc/kernel/asm-offsets.c
+++ b/arch/powerpc/kernel/asm-offsets.c
@@ -357,6 +357,10 @@ int main(void)
 #ifdef CONFIG_PPC_HAVE_KUAP
 	STACK_PT_REGS_OFFSET(STACK_REGS_KUAP, kuap);
 #endif
+#ifdef CONFIG_PPC_HAVE_KUEP
+	STACK_PT_REGS_OFFSET(STACK_REGS_KUEP, kuep);
+#endif
+
 
 #if defined(CONFIG_PPC32)
 #if defined(CONFIG_BOOKE) || defined(CONFIG_40x)
diff --git a/arch/powerpc/kernel/syscall_64.c b/arch/powerpc/kernel/syscall_64.c
index fded67982fbe..8229730c6cb5 100644
--- a/arch/powerpc/kernel/syscall_64.c
+++ b/arch/powerpc/kernel/syscall_64.c
@@ -37,15 +37,19 @@ notrace long system_call_exception(long r3, long r4, long r5,
 
 #ifdef CONFIG_PPC_MEM_KEYS
 	if (mmu_has_feature(MMU_FTR_PKEY)) {
-		unsigned long amr;
+		unsigned long amr, iamr;
 		/*
-		 * When entering from userspace we mostly have the AMR
+		 * When entering from userspace we mostly have the AMR/IAMR
 		 * different from kernel default values. Hence don't compare.
 		 */
 		amr = mfspr(SPRN_AMR);
+		iamr = mfspr(SPRN_IAMR);
 		regs->kuap = amr;
+		regs->kuep = iamr;
 		if (mmu_has_feature(MMU_FTR_KUAP))
 			mtspr(SPRN_AMR, AMR_KUAP_BLOCKED);
+		if (mmu_has_feature(MMU_FTR_KUEP))
+			mtspr(SPRN_IAMR, AMR_KUEP_BLOCKED);
 		isync();
 	} else
 #endif
-- 
2.26.2

[PATCH v4 26/41] powerpc/book3s64/pkeys: Inherit correctly on fork.

[Aneesh Kumar K.V]

Child thread.kuap value is inherited from the parent in copy_thread_tls. We still
need to make sure when the child returns from a fork in the kernel we start with the kernel
default AMR value.

Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
---
 arch/powerpc/kernel/process.c | 9 +++++++++
 1 file changed, 9 insertions(+)

diff --git a/arch/powerpc/kernel/process.c b/arch/powerpc/kernel/process.c
index 14599c7e4a37..dbce0b1daf2f 100644
--- a/arch/powerpc/kernel/process.c
+++ b/arch/powerpc/kernel/process.c
@@ -1705,6 +1705,15 @@ int copy_thread_tls(unsigned long clone_flags, unsigned long usp,
 		childregs->ppr = DEFAULT_PPR;
 
 	p->thread.tidr = 0;
+#endif
+	/*
+	 * Run with the current AMR value of the kernel
+	 */
+#if defined(CONFIG_PPC_MEM_KEYS)
+	if (mmu_has_feature(MMU_FTR_KUAP))
+		kregs->kuap = AMR_KUAP_BLOCKED;
+	if (mmu_has_feature(MMU_FTR_KUEP))
+		kregs->kuep = AMR_KUEP_BLOCKED;
 #endif
 	kregs->nip = ppc_function_entry(f);
 	return 0;
-- 
2.26.2

[PATCH v4 27/41] powerpc/book3s64/pkeys: Reset userspace AMR correctly on exec

[Aneesh Kumar K.V]

On fork, we inherit from the parent and on exec, we should switch to default_amr values.

Also, avoid changing the AMR register value within the kernel. The kernel now runs with
different AMR values.

Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
---
 arch/powerpc/include/asm/book3s/64/kup.h |  2 ++
 arch/powerpc/kernel/process.c            |  6 +++++-
 arch/powerpc/mm/book3s64/pkeys.c         | 18 ++----------------
 3 files changed, 9 insertions(+), 17 deletions(-)

diff --git a/arch/powerpc/include/asm/book3s/64/kup.h b/arch/powerpc/include/asm/book3s/64/kup.h
index 49182efea945..23d6f212d39c 100644
--- a/arch/powerpc/include/asm/book3s/64/kup.h
+++ b/arch/powerpc/include/asm/book3s/64/kup.h
@@ -171,6 +171,8 @@
 #include <asm/ptrace.h>
 
 extern u64 default_uamor;
+extern u64 default_amr;
+extern u64 default_iamr;
 
 static inline void kuap_restore_user_amr(struct pt_regs *regs)
 {
diff --git a/arch/powerpc/kernel/process.c b/arch/powerpc/kernel/process.c
index dbce0b1daf2f..340e473e8738 100644
--- a/arch/powerpc/kernel/process.c
+++ b/arch/powerpc/kernel/process.c
@@ -1485,6 +1485,11 @@ void arch_setup_new_exec(void)
 		current->thread.regs = regs - 1;
 	}
 
+#ifdef CONFIG_PPC_MEM_KEYS
+	current->thread.regs->kuap  = default_amr;
+	current->thread.regs->kuep  = default_iamr;
+#endif
+
 }
 #else
 void arch_setup_new_exec(void)
@@ -1839,7 +1844,6 @@ void start_thread(struct pt_regs *regs, unsigned long start, unsigned long sp)
 	current->thread.load_tm = 0;
 #endif /* CONFIG_PPC_TRANSACTIONAL_MEM */
 
-	thread_pkey_regs_init(&current->thread);
 }
 EXPORT_SYMBOL(start_thread);
 
diff --git a/arch/powerpc/mm/book3s64/pkeys.c b/arch/powerpc/mm/book3s64/pkeys.c
index 7980b013d161..95e972d6bbfa 100644
--- a/arch/powerpc/mm/book3s64/pkeys.c
+++ b/arch/powerpc/mm/book3s64/pkeys.c
@@ -21,8 +21,8 @@ int  max_pkey;			/* Maximum key value supported */
  */
 u32  reserved_allocation_mask;
 static u32  initial_allocation_mask;   /* Bits set for the initially allocated keys */
-static u64 default_amr;
-static u64 default_iamr;
+u64 default_amr;
+u64 default_iamr;
 /* Allow all keys to be modified by default */
 u64 default_uamor = ~0x0UL;
 /*
@@ -388,20 +388,6 @@ void thread_pkey_regs_restore(struct thread_struct *new_thread,
 		write_uamor(new_thread->uamor);
 }
 
-void thread_pkey_regs_init(struct thread_struct *thread)
-{
-	if (!mmu_has_feature(MMU_FTR_PKEY))
-		return;
-
-	thread->amr   = default_amr;
-	thread->iamr  = default_iamr;
-	thread->uamor = default_uamor;
-
-	write_amr(default_amr);
-	write_iamr(default_iamr);
-	write_uamor(default_uamor);
-}
-
 int execute_only_pkey(struct mm_struct *mm)
 {
 	if (static_branch_likely(&execute_pkey_disabled))
-- 
2.26.2

[PATCH v4 28/41] powerpc/ptrace-view: Use pt_regs values instead of thread_struct based one.

[Aneesh Kumar K.V]

We will remove thread.amr/iamr/uamor in a later patch

Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
---
 arch/powerpc/kernel/ptrace/ptrace-view.c | 23 +++++++++++++++++------
 1 file changed, 17 insertions(+), 6 deletions(-)

diff --git a/arch/powerpc/kernel/ptrace/ptrace-view.c b/arch/powerpc/kernel/ptrace/ptrace-view.c
index caeb5822a8f4..878764d0ee4c 100644
--- a/arch/powerpc/kernel/ptrace/ptrace-view.c
+++ b/arch/powerpc/kernel/ptrace/ptrace-view.c
@@ -488,14 +488,25 @@ static int pkey_active(struct task_struct *target, const struct user_regset *reg
 static int pkey_get(struct task_struct *target, const struct user_regset *regset,
 		    unsigned int pos, unsigned int count, void *kbuf, void __user *ubuf)
 {
-	BUILD_BUG_ON(TSO(amr) + sizeof(unsigned long) != TSO(iamr));
-	BUILD_BUG_ON(TSO(iamr) + sizeof(unsigned long) != TSO(uamor));
+	int ret;
 
 	if (!arch_pkeys_enabled())
 		return -ENODEV;
 
-	return user_regset_copyout(&pos, &count, &kbuf, &ubuf, &target->thread.amr,
-				   0, ELF_NPKEY * sizeof(unsigned long));
+	ret = user_regset_copyout(&pos, &count, &kbuf, &ubuf, &target->thread.regs->kuap,
+				  0, 1 * sizeof(unsigned long));
+	if (ret)
+		goto err_out;
+
+	ret = user_regset_copyout(&pos, &count, &kbuf, &ubuf, &target->thread.regs->kuep,
+				  1 * sizeof(unsigned long), 2 * sizeof(unsigned long));
+	if (ret)
+		goto err_out;
+
+	ret = user_regset_copyout(&pos, &count, &kbuf, &ubuf, &default_uamor,
+				  2 * sizeof(unsigned long), 3 * sizeof(unsigned long));
+err_out:
+	return ret;
 }
 
 static int pkey_set(struct task_struct *target, const struct user_regset *regset,
@@ -518,8 +529,8 @@ static int pkey_set(struct task_struct *target, const struct user_regset *regset
 		return ret;
 
 	/* UAMOR determines which bits of the AMR can be set from userspace. */
-	target->thread.amr = (new_amr & target->thread.uamor) |
-			     (target->thread.amr & ~target->thread.uamor);
+	target->thread.regs->kuap = (new_amr & default_uamor) |
+		(target->thread.regs->kuap & ~default_uamor);
 
 	return 0;
 }
-- 
2.26.2

[PATCH v4 29/41] powerpc/book3s64/pkeys: Don't update SPRN_AMR when in kernel mode.

[Aneesh Kumar K.V]

Now that kernel correctly store/restore userspace AMR/IAMR values, avoid
manipulating AMR and IAMR from the kernel on behalf of userspace.

Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
---
 arch/powerpc/include/asm/book3s/64/kup.h | 23 ++++++++
 arch/powerpc/include/asm/processor.h     |  5 --
 arch/powerpc/kernel/process.c            |  4 --
 arch/powerpc/kernel/traps.c              |  6 --
 arch/powerpc/mm/book3s64/pkeys.c         | 71 ++++--------------------
 5 files changed, 34 insertions(+), 75 deletions(-)

diff --git a/arch/powerpc/include/asm/book3s/64/kup.h b/arch/powerpc/include/asm/book3s/64/kup.h
index 23d6f212d39c..557dc8b66291 100644
--- a/arch/powerpc/include/asm/book3s/64/kup.h
+++ b/arch/powerpc/include/asm/book3s/64/kup.h
@@ -174,6 +174,29 @@ extern u64 default_uamor;
 extern u64 default_amr;
 extern u64 default_iamr;
 
+/*
+ * For kernel thread that doesn't have thread.regs return
+ * default AMR/IAMR values.
+ */
+static inline u64 current_thread_amr(void)
+{
+	if (current->thread.regs)
+		return current->thread.regs->kuap;
+	return AMR_KUAP_BLOCKED;
+}
+
+static inline u64 current_thread_iamr(void)
+{
+	if (current->thread.regs)
+		return current->thread.regs->kuep;
+	return AMR_KUEP_BLOCKED;
+}
+
+static inline u64 read_uamor(void)
+{
+	return default_uamor;
+}
+
 static inline void kuap_restore_user_amr(struct pt_regs *regs)
 {
 	if (!mmu_has_feature(MMU_FTR_PKEY))
diff --git a/arch/powerpc/include/asm/processor.h b/arch/powerpc/include/asm/processor.h
index 52a67835057a..bac4258a34b1 100644
--- a/arch/powerpc/include/asm/processor.h
+++ b/arch/powerpc/include/asm/processor.h
@@ -234,11 +234,6 @@ struct thread_struct {
 	struct thread_vr_state ckvr_state; /* Checkpointed VR state */
 	unsigned long	ckvrsave; /* Checkpointed VRSAVE */
 #endif /* CONFIG_PPC_TRANSACTIONAL_MEM */
-#ifdef CONFIG_PPC_MEM_KEYS
-	unsigned long	amr;
-	unsigned long	iamr;
-	unsigned long	uamor;
-#endif
 #ifdef CONFIG_KVM_BOOK3S_32_HANDLER
 	void*		kvm_shadow_vcpu; /* KVM internal data */
 #endif /* CONFIG_KVM_BOOK3S_32_HANDLER */
diff --git a/arch/powerpc/kernel/process.c b/arch/powerpc/kernel/process.c
index 340e473e8738..7eb6598375f1 100644
--- a/arch/powerpc/kernel/process.c
+++ b/arch/powerpc/kernel/process.c
@@ -585,7 +585,6 @@ static void save_all(struct task_struct *tsk)
 		__giveup_spe(tsk);
 
 	msr_check_and_clear(msr_all_available);
-	thread_pkey_regs_save(&tsk->thread);
 }
 
 void flush_all_to_thread(struct task_struct *tsk)
@@ -1109,8 +1108,6 @@ static inline void save_sprs(struct thread_struct *t)
 		t->tar = mfspr(SPRN_TAR);
 	}
 #endif
-
-	thread_pkey_regs_save(t);
 }
 
 static inline void restore_sprs(struct thread_struct *old_thread,
@@ -1151,7 +1148,6 @@ static inline void restore_sprs(struct thread_struct *old_thread,
 		mtspr(SPRN_TIDR, new_thread->tidr);
 #endif
 
-	thread_pkey_regs_restore(new_thread, old_thread);
 }
 
 struct task_struct *__switch_to(struct task_struct *prev,
diff --git a/arch/powerpc/kernel/traps.c b/arch/powerpc/kernel/traps.c
index 067e501f2202..e441e8eacfbc 100644
--- a/arch/powerpc/kernel/traps.c
+++ b/arch/powerpc/kernel/traps.c
@@ -348,12 +348,6 @@ static bool exception_common(int signr, struct pt_regs *regs, int code,
 
 	current->thread.trap_nr = code;
 
-	/*
-	 * Save all the pkey registers AMR/IAMR/UAMOR. Eg: Core dumps need
-	 * to capture the content, if the task gets killed.
-	 */
-	thread_pkey_regs_save(&current->thread);
-
 	return true;
 }
 
diff --git a/arch/powerpc/mm/book3s64/pkeys.c b/arch/powerpc/mm/book3s64/pkeys.c
index 95e972d6bbfa..b260e33e4e23 100644
--- a/arch/powerpc/mm/book3s64/pkeys.c
+++ b/arch/powerpc/mm/book3s64/pkeys.c
@@ -265,40 +265,17 @@ void pkey_mm_init(struct mm_struct *mm)
 	mm->context.execute_only_pkey = execute_only_key;
 }
 
-static inline u64 read_amr(void)
+static inline void update_current_thread_amr(u64 value)
 {
-	return mfspr(SPRN_AMR);
+	current->thread.regs->kuap = value;
 }
 
-static inline void write_amr(u64 value)
-{
-	mtspr(SPRN_AMR, value);
-}
-
-static inline u64 read_iamr(void)
-{
-	if (static_branch_unlikely(&execute_pkey_disabled))
-		return 0x0UL;
-
-	return mfspr(SPRN_IAMR);
-}
-
-static inline void write_iamr(u64 value)
+static inline void update_current_thread_iamr(u64 value)
 {
 	if (static_branch_unlikely(&execute_pkey_disabled))
 		return;
 
-	mtspr(SPRN_IAMR, value);
-}
-
-static inline u64 read_uamor(void)
-{
-	return mfspr(SPRN_UAMOR);
-}
-
-static inline void write_uamor(u64 value)
-{
-	mtspr(SPRN_UAMOR, value);
+	current->thread.regs->kuep = value;
 }
 
 static bool is_pkey_enabled(int pkey)
@@ -315,20 +292,21 @@ static bool is_pkey_enabled(int pkey)
 	return !!(uamor_pkey_bits);
 }
 
+/*  FIXME!! what happens to other threads AMR value? */
 static inline void init_amr(int pkey, u8 init_bits)
 {
 	u64 new_amr_bits = (((u64)init_bits & 0x3UL) << pkeyshift(pkey));
-	u64 old_amr = read_amr() & ~((u64)(0x3ul) << pkeyshift(pkey));
+	u64 old_amr = current_thread_amr() & ~((u64)(0x3ul) << pkeyshift(pkey));
 
-	write_amr(old_amr | new_amr_bits);
+	update_current_thread_amr(old_amr | new_amr_bits);
 }
 
 static inline void init_iamr(int pkey, u8 init_bits)
 {
 	u64 new_iamr_bits = (((u64)init_bits & 0x1UL) << pkeyshift(pkey));
-	u64 old_iamr = read_iamr() & ~((u64)(0x1ul) << pkeyshift(pkey));
+	u64 old_iamr = current_thread_iamr() & ~((u64)(0x1ul) << pkeyshift(pkey));
 
-	write_iamr(old_iamr | new_iamr_bits);
+	update_current_thread_iamr(old_iamr | new_iamr_bits);
 }
 
 /*
@@ -361,33 +339,6 @@ int __arch_set_user_pkey_access(struct task_struct *tsk, int pkey,
 	return 0;
 }
 
-void thread_pkey_regs_save(struct thread_struct *thread)
-{
-	if (!mmu_has_feature(MMU_FTR_PKEY))
-		return;
-
-	/*
-	 * TODO: Skip saving registers if @thread hasn't used any keys yet.
-	 */
-	thread->amr = read_amr();
-	thread->iamr = read_iamr();
-	thread->uamor = read_uamor();
-}
-
-void thread_pkey_regs_restore(struct thread_struct *new_thread,
-			      struct thread_struct *old_thread)
-{
-	if (!mmu_has_feature(MMU_FTR_PKEY))
-		return;
-
-	if (old_thread->amr != new_thread->amr)
-		write_amr(new_thread->amr);
-	if (old_thread->iamr != new_thread->iamr)
-		write_iamr(new_thread->iamr);
-	if (old_thread->uamor != new_thread->uamor)
-		write_uamor(new_thread->uamor);
-}
-
 int execute_only_pkey(struct mm_struct *mm)
 {
 	if (static_branch_likely(&execute_pkey_disabled))
@@ -441,10 +392,10 @@ static bool pkey_access_permitted(int pkey, bool write, bool execute)
 		return true;
 
 	pkey_shift = pkeyshift(pkey);
-	if (execute && !(read_iamr() & (IAMR_EX_BIT << pkey_shift)))
+	if (execute && !(current_thread_iamr() & (IAMR_EX_BIT << pkey_shift)))
 		return true;
 
-	amr = read_amr(); /* Delay reading amr until absolutely needed */
+	amr = current_thread_amr();
 	return ((!write && !(amr & (AMR_RD_BIT << pkey_shift))) ||
 		(write &&  !(amr & (AMR_WR_BIT << pkey_shift))));
 }
-- 
2.26.2

[PATCH v4 30/41] powerpc/book3s64/kuap: Restrict access to userspace based on userspace AMR

[Aneesh Kumar K.V]

If an application has configured address protection such that read/write is
denied using pkey even the kernel should receive a FAULT on accessing the same.

This patch use user AMR value stored in pt_regs.kuap to achieve the same.

Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
---
 arch/powerpc/include/asm/book3s/64/kup.h | 12 +++++++++---
 1 file changed, 9 insertions(+), 3 deletions(-)

diff --git a/arch/powerpc/include/asm/book3s/64/kup.h b/arch/powerpc/include/asm/book3s/64/kup.h
index 557dc8b66291..fcb1d5eb54dc 100644
--- a/arch/powerpc/include/asm/book3s/64/kup.h
+++ b/arch/powerpc/include/asm/book3s/64/kup.h
@@ -302,14 +302,20 @@ static inline void set_kuap(unsigned long value)
 static __always_inline void allow_user_access(void __user *to, const void __user *from,
 					      unsigned long size, unsigned long dir)
 {
+	unsigned long thread_amr = 0;
+
 	// This is written so we can resolve to a single case at build time
 	BUILD_BUG_ON(!__builtin_constant_p(dir));
+
+	if (mmu_has_feature(MMU_FTR_PKEY))
+		thread_amr = current_thread_amr();
+
 	if (dir == KUAP_READ)
-		set_kuap(AMR_KUAP_BLOCK_WRITE);
+		set_kuap(thread_amr | AMR_KUAP_BLOCK_WRITE);
 	else if (dir == KUAP_WRITE)
-		set_kuap(AMR_KUAP_BLOCK_READ);
+		set_kuap(thread_amr | AMR_KUAP_BLOCK_READ);
 	else if (dir == KUAP_READ_WRITE)
-		set_kuap(0);
+		set_kuap(thread_amr);
 	else
 		BUILD_BUG();
 }
-- 
2.26.2

[PATCH v4 31/41] powerpc/book3s64/kuap: Improve error reporting with KUAP

[Aneesh Kumar K.V]

With hash translation use DSISR_KEYFAULT to identify a wrong access.
With Radix we look at the AMR value and type of fault.

Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
---
 arch/powerpc/include/asm/book3s/32/kup.h     |  4 +--
 arch/powerpc/include/asm/book3s/64/kup.h     | 28 ++++++++++++++++----
 arch/powerpc/include/asm/kup.h               |  4 +--
 arch/powerpc/include/asm/nohash/32/kup-8xx.h |  4 +--
 arch/powerpc/mm/fault.c                      |  2 +-
 5 files changed, 30 insertions(+), 12 deletions(-)

diff --git a/arch/powerpc/include/asm/book3s/32/kup.h b/arch/powerpc/include/asm/book3s/32/kup.h
index 32fd4452e960..b18cd931e325 100644
--- a/arch/powerpc/include/asm/book3s/32/kup.h
+++ b/arch/powerpc/include/asm/book3s/32/kup.h
@@ -177,8 +177,8 @@ static inline void restore_user_access(unsigned long flags)
 		allow_user_access(to, to, end - addr, KUAP_READ_WRITE);
 }
 
-static inline bool
-bad_kuap_fault(struct pt_regs *regs, unsigned long address, bool is_write)
+static inline bool bad_kuap_fault(struct pt_regs *regs, unsigned long address,
+				  bool is_write, unsigned long error_code)
 {
 	unsigned long begin = regs->kuap & 0xf0000000;
 	unsigned long end = regs->kuap << 28;
diff --git a/arch/powerpc/include/asm/book3s/64/kup.h b/arch/powerpc/include/asm/book3s/64/kup.h
index fcb1d5eb54dc..0fcdd63d0de8 100644
--- a/arch/powerpc/include/asm/book3s/64/kup.h
+++ b/arch/powerpc/include/asm/book3s/64/kup.h
@@ -340,14 +340,32 @@ static inline void restore_user_access(unsigned long flags)
 	set_kuap(flags);
 }
 
-static inline bool
-bad_kuap_fault(struct pt_regs *regs, unsigned long address, bool is_write)
+#define RADIX_KUAP_BLOCK_READ	UL(0x4000000000000000)
+#define RADIX_KUAP_BLOCK_WRITE	UL(0x8000000000000000)
+
+static inline bool bad_kuap_fault(struct pt_regs *regs, unsigned long address,
+				  bool is_write, unsigned long error_code)
 {
-	return WARN(mmu_has_feature(MMU_FTR_KUAP) &&
-		    (regs->kuap & (is_write ? AMR_KUAP_BLOCK_WRITE : AMR_KUAP_BLOCK_READ)),
-		    "Bug: %s fault blocked by AMR!", is_write ? "Write" : "Read");
+	if (!mmu_has_feature(MMU_FTR_KUAP))
+		return false;
+
+	if (radix_enabled()) {
+		/*
+		 * Will be a storage protection fault.
+		 * Only check the details of AMR[0]
+		 */
+		return WARN((regs->kuap & (is_write ? RADIX_KUAP_BLOCK_WRITE : RADIX_KUAP_BLOCK_READ)),
+			    "Bug: %s fault blocked by AMR!", is_write ? "Write" : "Read");
+	}
+	/*
+	 * We don't want to WARN here because userspace can setup
+	 * keys such that a kernel access to user address can cause
+	 * fault
+	 */
+	return !!(error_code & DSISR_KEYFAULT);
 }
 #endif /* CONFIG_PPC_KUAP */
+
 #endif /* __ASSEMBLY__ */
 
 #endif /* _ASM_POWERPC_BOOK3S_64_KUP_H */
diff --git a/arch/powerpc/include/asm/kup.h b/arch/powerpc/include/asm/kup.h
index 015f51b02741..a29f69bbf6ec 100644
--- a/arch/powerpc/include/asm/kup.h
+++ b/arch/powerpc/include/asm/kup.h
@@ -59,8 +59,8 @@ static inline void prevent_user_access(void __user *to, const void __user *from,
 				       unsigned long size, unsigned long dir) { }
 static inline unsigned long prevent_user_access_return(void) { return 0UL; }
 static inline void restore_user_access(unsigned long flags) { }
-static inline bool
-bad_kuap_fault(struct pt_regs *regs, unsigned long address, bool is_write)
+static inline bool bad_kuap_fault(struct pt_regs *regs, unsigned long address,
+				  bool is_write, unsigned long error_code)
 {
 	return false;
 }
diff --git a/arch/powerpc/include/asm/nohash/32/kup-8xx.h b/arch/powerpc/include/asm/nohash/32/kup-8xx.h
index 85ed2390fb99..c401e4e404d4 100644
--- a/arch/powerpc/include/asm/nohash/32/kup-8xx.h
+++ b/arch/powerpc/include/asm/nohash/32/kup-8xx.h
@@ -60,8 +60,8 @@ static inline void restore_user_access(unsigned long flags)
 	mtspr(SPRN_MD_AP, flags);
 }
 
-static inline bool
-bad_kuap_fault(struct pt_regs *regs, unsigned long address, bool is_write)
+static inline bool bad_kuap_fault(struct pt_regs *regs, unsigned long address,
+				  bool is_write, unsigned long error_code)
 {
 	return WARN(!((regs->kuap ^ MD_APG_KUAP) & 0xf0000000),
 		    "Bug: fault blocked by AP register !");
diff --git a/arch/powerpc/mm/fault.c b/arch/powerpc/mm/fault.c
index 2393ed9d84bb..785c3e32c4e7 100644
--- a/arch/powerpc/mm/fault.c
+++ b/arch/powerpc/mm/fault.c
@@ -260,7 +260,7 @@ static bool bad_kernel_fault(struct pt_regs *regs, unsigned long error_code,
 
 	// Read/write fault in a valid region (the exception table search passed
 	// above), but blocked by KUAP is bad, it can never succeed.
-	if (bad_kuap_fault(regs, address, is_write))
+	if (bad_kuap_fault(regs, address, is_write, error_code))
 		return true;
 
 	// What's left? Kernel fault on user in well defined regions (extable
-- 
2.26.2

[PATCH v4 32/41] powerpc/book3s64/kuap: Use Key 3 to implement KUAP with hash translation.

[Aneesh Kumar K.V]

Radix use AMR Key 0 and hash translation use AMR key 3.

Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
---
 arch/powerpc/include/asm/book3s/64/kup.h | 9 ++++-----
 1 file changed, 4 insertions(+), 5 deletions(-)

diff --git a/arch/powerpc/include/asm/book3s/64/kup.h b/arch/powerpc/include/asm/book3s/64/kup.h
index 0fcdd63d0de8..f6ce7ba63df6 100644
--- a/arch/powerpc/include/asm/book3s/64/kup.h
+++ b/arch/powerpc/include/asm/book3s/64/kup.h
@@ -5,11 +5,10 @@
 #include <linux/const.h>
 #include <asm/reg.h>
 
-#define AMR_KUAP_BLOCK_READ	UL(0x4000000000000000)
-#define AMR_KUAP_BLOCK_WRITE	UL(0x8000000000000000)
+#define AMR_KUAP_BLOCK_READ	UL(0x5455555555555555)
+#define AMR_KUAP_BLOCK_WRITE	UL(0xa8aaaaaaaaaaaaaa)
 #define AMR_KUEP_BLOCKED	(1UL << 62)
 #define AMR_KUAP_BLOCKED	(AMR_KUAP_BLOCK_READ | AMR_KUAP_BLOCK_WRITE)
-#define AMR_KUAP_SHIFT		62
 
 #ifdef __ASSEMBLY__
 
@@ -75,8 +74,8 @@
 #ifdef CONFIG_PPC_KUAP_DEBUG
 	BEGIN_MMU_FTR_SECTION_NESTED(67)
 	mfspr	\gpr1, SPRN_AMR
-	li	\gpr2, (AMR_KUAP_BLOCKED >> AMR_KUAP_SHIFT)
-	sldi	\gpr2, \gpr2, AMR_KUAP_SHIFT
+	/* Prevent access to userspace using any key values */
+	LOAD_REG_IMMEDIATE(\gpr2, AMR_KUAP_BLOCKED)
 999:	tdne	\gpr1, \gpr2
 	EMIT_BUG_ENTRY 999b, __FILE__, __LINE__, (BUGFLAG_WARNING | BUGFLAG_ONCE)
 	END_MMU_FTR_SECTION_NESTED_IFSET(MMU_FTR_KUAP, 67)
-- 
2.26.2

[PATCH v4 33/41] powerpc/book3s64/kuep: Use Key 3 to implement KUEP with hash translation.

[Aneesh Kumar K.V]

Radix use IAMR Key 0 and hash translation use IAMR key 3.

Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
---
 arch/powerpc/include/asm/book3s/64/kup.h | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/arch/powerpc/include/asm/book3s/64/kup.h b/arch/powerpc/include/asm/book3s/64/kup.h
index f6ce7ba63df6..78b3f67ae8c5 100644
--- a/arch/powerpc/include/asm/book3s/64/kup.h
+++ b/arch/powerpc/include/asm/book3s/64/kup.h
@@ -7,7 +7,7 @@
 
 #define AMR_KUAP_BLOCK_READ	UL(0x5455555555555555)
 #define AMR_KUAP_BLOCK_WRITE	UL(0xa8aaaaaaaaaaaaaa)
-#define AMR_KUEP_BLOCKED	(1UL << 62)
+#define AMR_KUEP_BLOCKED	UL(0x5455555555555555)
 #define AMR_KUAP_BLOCKED	(AMR_KUAP_BLOCK_READ | AMR_KUAP_BLOCK_WRITE)
 
 #ifdef __ASSEMBLY__
-- 
2.26.2

[PATCH v4 34/41] powerpc/book3s64/hash/kuap: Enable kuap on hash

[Aneesh Kumar K.V]

Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
---
 arch/powerpc/mm/book3s64/pkeys.c | 7 ++++++-
 1 file changed, 6 insertions(+), 1 deletion(-)

diff --git a/arch/powerpc/mm/book3s64/pkeys.c b/arch/powerpc/mm/book3s64/pkeys.c
index b260e33e4e23..0d6fd69708f2 100644
--- a/arch/powerpc/mm/book3s64/pkeys.c
+++ b/arch/powerpc/mm/book3s64/pkeys.c
@@ -241,7 +241,12 @@ void __init setup_kuep(bool disabled)
 #ifdef CONFIG_PPC_KUAP
 void __init setup_kuap(bool disabled)
 {
-	if (disabled || !early_radix_enabled())
+	if (disabled)
+		return;
+	/*
+	 * On hash if PKEY feature is not enabled, disable KUAP too.
+	 */
+	if (!early_radix_enabled() && !early_mmu_has_feature(MMU_FTR_PKEY))
 		return;
 
 	if (smp_processor_id() == boot_cpuid) {
-- 
2.26.2

[PATCH v4 35/41] powerpc/book3s64/hash/kuep: Enable KUEP on hash

[Aneesh Kumar K.V]

Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
---
 arch/powerpc/mm/book3s64/pkeys.c | 7 ++++++-
 1 file changed, 6 insertions(+), 1 deletion(-)

diff --git a/arch/powerpc/mm/book3s64/pkeys.c b/arch/powerpc/mm/book3s64/pkeys.c
index 0d6fd69708f2..9ab8fdaf7146 100644
--- a/arch/powerpc/mm/book3s64/pkeys.c
+++ b/arch/powerpc/mm/book3s64/pkeys.c
@@ -220,7 +220,12 @@ void __init pkey_early_init_devtree(void)
 #ifdef CONFIG_PPC_KUEP
 void __init setup_kuep(bool disabled)
 {
-	if (disabled || !early_radix_enabled())
+	if (disabled)
+		return;
+	/*
+	 * On hash if PKEY feature is not enabled, disable KUAP too.
+	 */
+	if (!early_radix_enabled() && !early_mmu_has_feature(MMU_FTR_PKEY))
 		return;
 
 	if (smp_processor_id() == boot_cpuid) {
-- 
2.26.2

[PATCH v4 36/41] powerpc/book3s64/keys: Print information during boot.

[Aneesh Kumar K.V]

Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
---
 arch/powerpc/mm/book3s64/pkeys.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/arch/powerpc/mm/book3s64/pkeys.c b/arch/powerpc/mm/book3s64/pkeys.c
index 9ab8fdaf7146..65dbb335f0a0 100644
--- a/arch/powerpc/mm/book3s64/pkeys.c
+++ b/arch/powerpc/mm/book3s64/pkeys.c
@@ -208,6 +208,7 @@ void __init pkey_early_init_devtree(void)
 	 */
 	initial_allocation_mask |= reserved_allocation_mask;
 
+	pr_info("Enabling Memory keys with max key count %d", max_pkey);
 err_out:
 	/*
 	 * Setup uamor on boot cpu
-- 
2.26.2

[PATCH v4 37/41] powerpc/selftest/ptrave-pkey: Rename variables to make it easier to follow code

[Aneesh Kumar K.V]

Rename variable to indicate that they are invalid values which we will use to
test ptrace update of pkeys.

Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
---
 .../selftests/powerpc/ptrace/ptrace-pkey.c    | 26 +++++++++----------
 1 file changed, 13 insertions(+), 13 deletions(-)

diff --git a/tools/testing/selftests/powerpc/ptrace/ptrace-pkey.c b/tools/testing/selftests/powerpc/ptrace/ptrace-pkey.c
index bdbbbe8431e0..f9216c7a1829 100644
--- a/tools/testing/selftests/powerpc/ptrace/ptrace-pkey.c
+++ b/tools/testing/selftests/powerpc/ptrace/ptrace-pkey.c
@@ -44,7 +44,7 @@ struct shared_info {
 	unsigned long amr2;
 
 	/* AMR value that ptrace should refuse to write to the child. */
-	unsigned long amr3;
+	unsigned long invalid_amr;
 
 	/* IAMR value the parent expects to read from the child. */
 	unsigned long expected_iamr;
@@ -57,8 +57,8 @@ struct shared_info {
 	 * (even though they're valid ones) because userspace doesn't have
 	 * access to those registers.
 	 */
-	unsigned long new_iamr;
-	unsigned long new_uamor;
+	unsigned long invalid_iamr;
+	unsigned long invalid_uamor;
 };
 
 static int sys_pkey_alloc(unsigned long flags, unsigned long init_access_rights)
@@ -100,7 +100,7 @@ static int child(struct shared_info *info)
 
 	info->amr1 |= 3ul << pkeyshift(pkey1);
 	info->amr2 |= 3ul << pkeyshift(pkey2);
-	info->amr3 |= info->amr2 | 3ul << pkeyshift(pkey3);
+	info->invalid_amr |= info->amr2 | 3ul << pkeyshift(pkey3);
 
 	if (disable_execute)
 		info->expected_iamr |= 1ul << pkeyshift(pkey1);
@@ -111,8 +111,8 @@ static int child(struct shared_info *info)
 
 	info->expected_uamor |= 3ul << pkeyshift(pkey1) |
 				3ul << pkeyshift(pkey2);
-	info->new_iamr |= 1ul << pkeyshift(pkey1) | 1ul << pkeyshift(pkey2);
-	info->new_uamor |= 3ul << pkeyshift(pkey1);
+	info->invalid_iamr |= 1ul << pkeyshift(pkey1) | 1ul << pkeyshift(pkey2);
+	info->invalid_uamor |= 3ul << pkeyshift(pkey1);
 
 	/*
 	 * We won't use pkey3. We just want a plausible but invalid key to test
@@ -196,9 +196,9 @@ static int parent(struct shared_info *info, pid_t pid)
 	PARENT_SKIP_IF_UNSUPPORTED(ret, &info->child_sync);
 	PARENT_FAIL_IF(ret, &info->child_sync);
 
-	info->amr1 = info->amr2 = info->amr3 = regs[0];
-	info->expected_iamr = info->new_iamr = regs[1];
-	info->expected_uamor = info->new_uamor = regs[2];
+	info->amr1 = info->amr2 = info->invalid_amr = regs[0];
+	info->expected_iamr = info->invalid_iamr = regs[1];
+	info->expected_uamor = info->invalid_uamor = regs[2];
 
 	/* Wake up child so that it can set itself up. */
 	ret = prod_child(&info->child_sync);
@@ -234,10 +234,10 @@ static int parent(struct shared_info *info, pid_t pid)
 		return ret;
 
 	/* Write invalid AMR value in child. */
-	ret = ptrace_write_regs(pid, NT_PPC_PKEY, &info->amr3, 1);
+	ret = ptrace_write_regs(pid, NT_PPC_PKEY, &info->invalid_amr, 1);
 	PARENT_FAIL_IF(ret, &info->child_sync);
 
-	printf("%-30s AMR: %016lx\n", ptrace_write_running, info->amr3);
+	printf("%-30s AMR: %016lx\n", ptrace_write_running, info->invalid_amr);
 
 	/* Wake up child so that it can verify it didn't change. */
 	ret = prod_child(&info->child_sync);
@@ -249,7 +249,7 @@ static int parent(struct shared_info *info, pid_t pid)
 
 	/* Try to write to IAMR. */
 	regs[0] = info->amr1;
-	regs[1] = info->new_iamr;
+	regs[1] = info->invalid_iamr;
 	ret = ptrace_write_regs(pid, NT_PPC_PKEY, regs, 2);
 	PARENT_FAIL_IF(!ret, &info->child_sync);
 
@@ -257,7 +257,7 @@ static int parent(struct shared_info *info, pid_t pid)
 	       ptrace_write_running, regs[0], regs[1]);
 
 	/* Try to write to IAMR and UAMOR. */
-	regs[2] = info->new_uamor;
+	regs[2] = info->invalid_uamor;
 	ret = ptrace_write_regs(pid, NT_PPC_PKEY, regs, 3);
 	PARENT_FAIL_IF(!ret, &info->child_sync);
 
-- 
2.26.2

[PATCH v4 38/41] powerpc/selftest/ptrace-pkey: Update the test to mark an invalid pkey correctly

[Aneesh Kumar K.V]

Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
---
 .../selftests/powerpc/ptrace/ptrace-pkey.c    | 30 ++++++++-----------
 1 file changed, 12 insertions(+), 18 deletions(-)

diff --git a/tools/testing/selftests/powerpc/ptrace/ptrace-pkey.c b/tools/testing/selftests/powerpc/ptrace/ptrace-pkey.c
index f9216c7a1829..bc33d748d95b 100644
--- a/tools/testing/selftests/powerpc/ptrace/ptrace-pkey.c
+++ b/tools/testing/selftests/powerpc/ptrace/ptrace-pkey.c
@@ -66,11 +66,6 @@ static int sys_pkey_alloc(unsigned long flags, unsigned long init_access_rights)
 	return syscall(__NR_pkey_alloc, flags, init_access_rights);
 }
 
-static int sys_pkey_free(int pkey)
-{
-	return syscall(__NR_pkey_free, pkey);
-}
-
 static int child(struct shared_info *info)
 {
 	unsigned long reg;
@@ -100,7 +95,11 @@ static int child(struct shared_info *info)
 
 	info->amr1 |= 3ul << pkeyshift(pkey1);
 	info->amr2 |= 3ul << pkeyshift(pkey2);
-	info->invalid_amr |= info->amr2 | 3ul << pkeyshift(pkey3);
+	/*
+	 * invalid amr value where we try to force write
+	 * things which are deined by a uamor setting.
+	 */
+	info->invalid_amr = info->amr2 | (~0x0UL & ~info->expected_uamor);
 
 	if (disable_execute)
 		info->expected_iamr |= 1ul << pkeyshift(pkey1);
@@ -111,17 +110,12 @@ static int child(struct shared_info *info)
 
 	info->expected_uamor |= 3ul << pkeyshift(pkey1) |
 				3ul << pkeyshift(pkey2);
-	info->invalid_iamr |= 1ul << pkeyshift(pkey1) | 1ul << pkeyshift(pkey2);
-	info->invalid_uamor |= 3ul << pkeyshift(pkey1);
-
 	/*
-	 * We won't use pkey3. We just want a plausible but invalid key to test
-	 * whether ptrace will let us write to AMR bits we are not supposed to.
-	 *
-	 * This also tests whether the kernel restores the UAMOR permissions
-	 * after a key is freed.
+	 * Create an IAMR value different from expected value.
+	 * Kernel will reject an IAMR and UAMOR change.
 	 */
-	sys_pkey_free(pkey3);
+	info->invalid_iamr = info->expected_iamr | (1ul << pkeyshift(pkey1) | 1ul << pkeyshift(pkey2));
+	info->invalid_uamor = info->expected_uamor & ~(0x3ul << pkeyshift(pkey1));
 
 	printf("%-30s AMR: %016lx pkey1: %d pkey2: %d pkey3: %d\n",
 	       user_write, info->amr1, pkey1, pkey2, pkey3);
@@ -196,9 +190,9 @@ static int parent(struct shared_info *info, pid_t pid)
 	PARENT_SKIP_IF_UNSUPPORTED(ret, &info->child_sync);
 	PARENT_FAIL_IF(ret, &info->child_sync);
 
-	info->amr1 = info->amr2 = info->invalid_amr = regs[0];
-	info->expected_iamr = info->invalid_iamr = regs[1];
-	info->expected_uamor = info->invalid_uamor = regs[2];
+	info->amr1 = info->amr2 = regs[0];
+	info->expected_iamr = regs[1];
+	info->expected_uamor = regs[2];
 
 	/* Wake up child so that it can set itself up. */
 	ret = prod_child(&info->child_sync);
-- 
2.26.2

[PATCH v4 39/41] powerpc/selftest/ptrace-pkey: IAMR and uamor cannot be updated by ptrace

[Aneesh Kumar K.V]

Both IAMR and uamor are privileged and cannot be updated by userspace. Hence
we also don't allow ptrace interface to update them. Don't update them in the
test. Also expected_iamr is only changed if we can allocate a  DISABLE_EXECUTE
pkey.

Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
---
 tools/testing/selftests/powerpc/ptrace/ptrace-pkey.c | 9 +++------
 1 file changed, 3 insertions(+), 6 deletions(-)

diff --git a/tools/testing/selftests/powerpc/ptrace/ptrace-pkey.c b/tools/testing/selftests/powerpc/ptrace/ptrace-pkey.c
index bc33d748d95b..5c3c8222de46 100644
--- a/tools/testing/selftests/powerpc/ptrace/ptrace-pkey.c
+++ b/tools/testing/selftests/powerpc/ptrace/ptrace-pkey.c
@@ -101,15 +101,12 @@ static int child(struct shared_info *info)
 	 */
 	info->invalid_amr = info->amr2 | (~0x0UL & ~info->expected_uamor);
 
+	/*
+	 * if PKEY_DISABLE_EXECUTE succeeded we should update the expected_iamr
+	 */
 	if (disable_execute)
 		info->expected_iamr |= 1ul << pkeyshift(pkey1);
-	else
-		info->expected_iamr &= ~(1ul << pkeyshift(pkey1));
-
-	info->expected_iamr &= ~(1ul << pkeyshift(pkey2) | 1ul << pkeyshift(pkey3));
 
-	info->expected_uamor |= 3ul << pkeyshift(pkey1) |
-				3ul << pkeyshift(pkey2);
 	/*
 	 * Create an IAMR value different from expected value.
 	 * Kernel will reject an IAMR and UAMOR change.
-- 
2.26.2

[PATCH v4 40/41] powerpc/book3s64/keys/kuap: Reset AMR/IAMR values on kexec

[Aneesh Kumar K.V]

We can kexec into a kernel that doesn't use memory keys for kernel
mapping (such as an older kernel which doesn't support kuap/kuep with hash
translation). We need to make sure we reset the AMR/IAMR value on kexec
otherwise, the new kernel will use key 0 for kernel mapping and the old
AMR value prevents access to key 0.

This patch also removes reset if IAMR and AMOR in kexec_sequence. Reset of AMOR
is not needed and the IAMR reset is partial (it doesn't do the reset
on secondary cpus) and is redundant with this patch.

Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
---
 arch/powerpc/include/asm/book3s/64/kup.h | 20 ++++++++++++++++++++
 arch/powerpc/include/asm/kup.h           | 14 ++++++++++++++
 arch/powerpc/kernel/misc_64.S            | 14 --------------
 arch/powerpc/kexec/core_64.c             |  3 +++
 arch/powerpc/mm/book3s64/pgtable.c       |  3 +++
 5 files changed, 40 insertions(+), 14 deletions(-)

diff --git a/arch/powerpc/include/asm/book3s/64/kup.h b/arch/powerpc/include/asm/book3s/64/kup.h
index 78b3f67ae8c5..be04733c89f3 100644
--- a/arch/powerpc/include/asm/book3s/64/kup.h
+++ b/arch/powerpc/include/asm/book3s/64/kup.h
@@ -365,6 +365,26 @@ static inline bool bad_kuap_fault(struct pt_regs *regs, unsigned long address,
 }
 #endif /* CONFIG_PPC_KUAP */
 
+#define reset_kuap reset_kuap
+static inline void reset_kuap(void)
+{
+	if (mmu_has_feature(MMU_FTR_KUAP)) {
+		mtspr(SPRN_AMR, 0);
+		/*  Do we need isync()? We are going via a kexec reset */
+		isync();
+	}
+}
+
+#define reset_kuep reset_kuep
+static inline void reset_kuep(void)
+{
+	if (mmu_has_feature(MMU_FTR_KUEP)) {
+		mtspr(SPRN_IAMR, 0);
+		/*  Do we need isync()? We are going via a kexec reset */
+		isync();
+	}
+}
+
 #endif /* __ASSEMBLY__ */
 
 #endif /* _ASM_POWERPC_BOOK3S_64_KUP_H */
diff --git a/arch/powerpc/include/asm/kup.h b/arch/powerpc/include/asm/kup.h
index a29f69bbf6ec..c7ab7310f230 100644
--- a/arch/powerpc/include/asm/kup.h
+++ b/arch/powerpc/include/asm/kup.h
@@ -113,6 +113,20 @@ static inline void prevent_current_write_to_user(void)
 	prevent_user_access(NULL, NULL, ~0UL, KUAP_CURRENT_WRITE);
 }
 
+#ifndef reset_kuap
+#define reset_kuap reset_kuap
+static inline void reset_kuap(void)
+{
+}
+#endif
+
+#ifndef reset_kuep
+#define reset_kuep reset_kuep
+static inline void reset_kuep(void)
+{
+}
+#endif
+
 #endif /* !__ASSEMBLY__ */
 
 #endif /* _ASM_POWERPC_KUAP_H_ */
diff --git a/arch/powerpc/kernel/misc_64.S b/arch/powerpc/kernel/misc_64.S
index 1864605eca29..7bb46ad98207 100644
--- a/arch/powerpc/kernel/misc_64.S
+++ b/arch/powerpc/kernel/misc_64.S
@@ -413,20 +413,6 @@ _GLOBAL(kexec_sequence)
 	li	r0,0
 	std	r0,16(r1)
 
-BEGIN_FTR_SECTION
-	/*
-	 * This is the best time to turn AMR/IAMR off.
-	 * key 0 is used in radix for supervisor<->user
-	 * protection, but on hash key 0 is reserved
-	 * ideally we want to enter with a clean state.
-	 * NOTE, we rely on r0 being 0 from above.
-	 */
-	mtspr	SPRN_IAMR,r0
-BEGIN_FTR_SECTION_NESTED(42)
-	mtspr	SPRN_AMOR,r0
-END_FTR_SECTION_NESTED_IFSET(CPU_FTR_HVMODE, 42)
-END_FTR_SECTION_IFSET(CPU_FTR_ARCH_300)
-
 	/* save regs for local vars on new stack.
 	 * yes, we won't go back, but ...
 	 */
diff --git a/arch/powerpc/kexec/core_64.c b/arch/powerpc/kexec/core_64.c
index b4184092172a..a124715f33ea 100644
--- a/arch/powerpc/kexec/core_64.c
+++ b/arch/powerpc/kexec/core_64.c
@@ -152,6 +152,9 @@ static void kexec_smp_down(void *arg)
 	if (ppc_md.kexec_cpu_down)
 		ppc_md.kexec_cpu_down(0, 1);
 
+	reset_kuap();
+	reset_kuep();
+
 	kexec_smp_wait();
 	/* NOTREACHED */
 }
diff --git a/arch/powerpc/mm/book3s64/pgtable.c b/arch/powerpc/mm/book3s64/pgtable.c
index c58ad1049909..9673f4b74c9a 100644
--- a/arch/powerpc/mm/book3s64/pgtable.c
+++ b/arch/powerpc/mm/book3s64/pgtable.c
@@ -165,6 +165,9 @@ void mmu_cleanup_all(void)
 		radix__mmu_cleanup_all();
 	else if (mmu_hash_ops.hpte_clear_all)
 		mmu_hash_ops.hpte_clear_all();
+
+	reset_kuap();
+	reset_kuep();
 }
 
 #ifdef CONFIG_MEMORY_HOTPLUG
-- 
2.26.2

[PATCH v4 41/41] powerpc/book3s64/hash/kup: Don't hardcode kup key

[Aneesh Kumar K.V]

Make KUAP/KUEP key a variable and also check whether the platform
limit the max key such that we can't use the key for KUAP/KEUP.

Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
---
 .../powerpc/include/asm/book3s/64/hash-pkey.h | 22 +--------
 arch/powerpc/include/asm/book3s/64/kup.h      |  1 +
 arch/powerpc/mm/book3s64/pkeys.c              | 46 +++++++++++++++++--
 3 files changed, 43 insertions(+), 26 deletions(-)

diff --git a/arch/powerpc/include/asm/book3s/64/hash-pkey.h b/arch/powerpc/include/asm/book3s/64/hash-pkey.h
index 9f44e208f036..ff9907c72ee3 100644
--- a/arch/powerpc/include/asm/book3s/64/hash-pkey.h
+++ b/arch/powerpc/include/asm/book3s/64/hash-pkey.h
@@ -2,9 +2,7 @@
 #ifndef _ASM_POWERPC_BOOK3S_64_HASH_PKEY_H
 #define _ASM_POWERPC_BOOK3S_64_HASH_PKEY_H
 
-/*  We use key 3 for KERNEL */
-#define HASH_DEFAULT_KERNEL_KEY (HPTE_R_KEY_BIT0 | HPTE_R_KEY_BIT1)
-
+u64 pte_to_hpte_pkey_bits(u64 pteflags, unsigned long flags);
 static inline u64 hash__vmflag_to_pte_pkey_bits(u64 vm_flags)
 {
 	return (((vm_flags & VM_PKEY_BIT0) ? H_PTE_PKEY_BIT0 : 0x0UL) |
@@ -14,24 +12,6 @@ static inline u64 hash__vmflag_to_pte_pkey_bits(u64 vm_flags)
 		((vm_flags & VM_PKEY_BIT4) ? H_PTE_PKEY_BIT4 : 0x0UL));
 }
 
-static inline u64 pte_to_hpte_pkey_bits(u64 pteflags, unsigned long flags)
-{
-	unsigned long pte_pkey;
-
-	pte_pkey = (((pteflags & H_PTE_PKEY_BIT4) ? HPTE_R_KEY_BIT4 : 0x0UL) |
-		    ((pteflags & H_PTE_PKEY_BIT3) ? HPTE_R_KEY_BIT3 : 0x0UL) |
-		    ((pteflags & H_PTE_PKEY_BIT2) ? HPTE_R_KEY_BIT2 : 0x0UL) |
-		    ((pteflags & H_PTE_PKEY_BIT1) ? HPTE_R_KEY_BIT1 : 0x0UL) |
-		    ((pteflags & H_PTE_PKEY_BIT0) ? HPTE_R_KEY_BIT0 : 0x0UL));
-
-	if (mmu_has_feature(MMU_FTR_KUAP) || mmu_has_feature(MMU_FTR_KUEP)) {
-		if ((pte_pkey == 0) && (flags & HPTE_USE_KERNEL_KEY))
-			return HASH_DEFAULT_KERNEL_KEY;
-	}
-
-	return pte_pkey;
-}
-
 static inline u16 hash__pte_to_pkey_bits(u64 pteflags)
 {
 	return (((pteflags & H_PTE_PKEY_BIT4) ? 0x10 : 0x0UL) |
diff --git a/arch/powerpc/include/asm/book3s/64/kup.h b/arch/powerpc/include/asm/book3s/64/kup.h
index be04733c89f3..42ed67ec4e64 100644
--- a/arch/powerpc/include/asm/book3s/64/kup.h
+++ b/arch/powerpc/include/asm/book3s/64/kup.h
@@ -172,6 +172,7 @@
 extern u64 default_uamor;
 extern u64 default_amr;
 extern u64 default_iamr;
+extern int kup_key;
 
 /*
  * For kernel thread that doesn't have thread.regs return
diff --git a/arch/powerpc/mm/book3s64/pkeys.c b/arch/powerpc/mm/book3s64/pkeys.c
index 65dbb335f0a0..9fb90ceef08b 100644
--- a/arch/powerpc/mm/book3s64/pkeys.c
+++ b/arch/powerpc/mm/book3s64/pkeys.c
@@ -30,6 +30,10 @@ u64 default_uamor = ~0x0UL;
  * We pick key 2 because 0 is special key and 1 is reserved as per ISA.
  */
 static int execute_only_key = 2;
+/*
+ * key used to implement KUAP/KUEP with hash translation.
+ */
+int kup_key = 3;
 
 
 #define AMR_BITS_PER_PKEY 2
@@ -170,6 +174,18 @@ void __init pkey_early_init_devtree(void)
 		default_uamor &= ~(0x3ul << pkeyshift(execute_only_key));
 	}
 
+	if (unlikely(max_pkey <= kup_key)) {
+		/*
+		 * Insufficient number of keys to support
+		 * KUAP/KUEP feature.
+		 */
+		kup_key = -1;
+	} else {
+		/*  handle key which is used by kernel for KAUP */
+		reserved_allocation_mask |= (0x1 << kup_key);
+		default_uamor &= ~(0x3ul << pkeyshift(kup_key));
+	}
+
 	/*
 	 * Allow access for only key 0. And prevent any other modification.
 	 */
@@ -190,9 +206,6 @@ void __init pkey_early_init_devtree(void)
 	reserved_allocation_mask |= (0x1 << 1);
 	default_uamor &= ~(0x3ul << pkeyshift(1));
 
-	/*  handle key 3 which is used by kernel for KAUP */
-	reserved_allocation_mask |= (0x1 << 3);
-	default_uamor &= ~(0x3ul << pkeyshift(3));
 
 	/*
 	 * Prevent the usage of OS reserved keys. Update UAMOR
@@ -221,7 +234,7 @@ void __init pkey_early_init_devtree(void)
 #ifdef CONFIG_PPC_KUEP
 void __init setup_kuep(bool disabled)
 {
-	if (disabled)
+	if (disabled || kup_key == -1)
 		return;
 	/*
 	 * On hash if PKEY feature is not enabled, disable KUAP too.
@@ -247,7 +260,7 @@ void __init setup_kuep(bool disabled)
 #ifdef CONFIG_PPC_KUAP
 void __init setup_kuap(bool disabled)
 {
-	if (disabled)
+	if (disabled || kup_key == -1)
 		return;
 	/*
 	 * On hash if PKEY feature is not enabled, disable KUAP too.
@@ -450,3 +463,26 @@ void arch_dup_pkeys(struct mm_struct *oldmm, struct mm_struct *mm)
 	mm_pkey_allocation_map(mm) = mm_pkey_allocation_map(oldmm);
 	mm->context.execute_only_pkey = oldmm->context.execute_only_pkey;
 }
+
+u64 pte_to_hpte_pkey_bits(u64 pteflags, unsigned long flags)
+{
+	unsigned long pte_pkey;
+
+	pte_pkey = (((pteflags & H_PTE_PKEY_BIT4) ? HPTE_R_KEY_BIT4 : 0x0UL) |
+		    ((pteflags & H_PTE_PKEY_BIT3) ? HPTE_R_KEY_BIT3 : 0x0UL) |
+		    ((pteflags & H_PTE_PKEY_BIT2) ? HPTE_R_KEY_BIT2 : 0x0UL) |
+		    ((pteflags & H_PTE_PKEY_BIT1) ? HPTE_R_KEY_BIT1 : 0x0UL) |
+		    ((pteflags & H_PTE_PKEY_BIT0) ? HPTE_R_KEY_BIT0 : 0x0UL));
+
+	if (mmu_has_feature(MMU_FTR_KUAP) || mmu_has_feature(MMU_FTR_KUEP)) {
+		if ((pte_pkey == 0) &&
+		    (flags & HPTE_USE_KERNEL_KEY) && (kup_key != -1)) {
+			u64 vm_flag = pkey_to_vmflag_bits(kup_key);
+			u64 pte_flag = hash__vmflag_to_pte_pkey_bits(vm_flag);
+			return pte_to_hpte_pkey_bits(pte_flag, 0);
+		}
+	}
+
+	return pte_pkey;
+}
+
-- 
2.26.2