Merge tag 'arm64-upstream' of git://git.kernel.org/pub/scm/linux/kernel/git/arm64/linux

Pull second set of arm64 updates from Catalin Marinas:
 "A second pull request for this merging window, mainly with fixes and
  docs clarification:

   - Documentation clarification on CPU topology and booting
     requirements
   - Additional cache flushing during boot (needed in the presence of
     external caches or under virtualisation)
   - DMA range invalidation fix for non cache line aligned buffers
   - Build failure fix with !COMPAT
   - Kconfig update for STRICT_DEVMEM"

* tag 'arm64-upstream' of git://git.kernel.org/pub/scm/linux/kernel/git/arm64/linux:
  arm64: Fix DMA range invalidation for cache line unaligned buffers
  arm64: Add missing Kconfig for CONFIG_STRICT_DEVMEM
  arm64: fix !CONFIG_COMPAT build failures
  Revert "arm64: virt: ensure visibility of __boot_cpu_mode"
  arm64: Relax the kernel cache requirements for boot
  arm64: Update the TCR_EL1 translation granule definitions for 16K pages
  ARM: topology: Make it clear that all CPUs need to be described

8 files changed, 87 insertions, 31 deletions

diff --git a/arch/arm64/Kconfig.debug b/arch/arm64/Kconfig.debug
index 835c559786bd..d10ec334c93b 100644
--- a/arch/arm64/Kconfig.debug
+++ b/arch/arm64/Kconfig.debug
@@ -6,6 +6,20 @@ config FRAME_POINTER
         bool
         default y
 
+config STRICT_DEVMEM
+        bool "Filter access to /dev/mem"
+        depends on MMU
+        help
+          If this option is disabled, you allow userspace (root) access to all
+          of memory, including kernel and userspace memory. Accidental
+          access to this is obviously disastrous, but specific access can
+          be used by people debugging the kernel.
+
+          If this option is switched on, the /dev/mem file only allows
+          userspace access to memory mapped peripherals.
+
+          If in doubt, say Y.
+
 config EARLY_PRINTK
         bool "Early printk support"
         default y
diff --git a/arch/arm64/include/asm/pgtable-hwdef.h b/arch/arm64/include/asm/pgtable-hwdef.h
index f7af66b54cb2..5fc8a66c3924 100644
--- a/arch/arm64/include/asm/pgtable-hwdef.h
+++ b/arch/arm64/include/asm/pgtable-hwdef.h
@@ -120,8 +120,12 @@
 #define TCR_ORGN_WBnWA          ((UL(3) << 10) | (UL(3) << 26))
 #define TCR_ORGN_MASK           ((UL(3) << 10) | (UL(3) << 26))
 #define TCR_SHARED              ((UL(3) << 12) | (UL(3) << 28))
+#define TCR_TG0_4K              (UL(0) << 14)
 #define TCR_TG0_64K             (UL(1) << 14)
-#define TCR_TG1_64K             (UL(1) << 30)
+#define TCR_TG0_16K             (UL(2) << 14)
+#define TCR_TG1_16K             (UL(1) << 30)
+#define TCR_TG1_4K              (UL(2) << 30)
+#define TCR_TG1_64K             (UL(3) << 30)
 #define TCR_ASID16              (UL(1) << 36)
 #define TCR_TBI0                (UL(1) << 37)
 
diff --git a/arch/arm64/include/asm/virt.h b/arch/arm64/include/asm/virt.h
index 130e2be952cf..215ad4649dd7 100644
--- a/arch/arm64/include/asm/virt.h
+++ b/arch/arm64/include/asm/virt.h
@@ -22,7 +22,6 @@
 #define BOOT_CPU_MODE_EL2       (0xe12)
 
 #ifndef __ASSEMBLY__
-#include <asm/cacheflush.h>
 
 /*
  * __boot_cpu_mode records what mode CPUs were booted in.
@@ -38,20 +37,9 @@ extern u32 __boot_cpu_mode[2];
 void __hyp_set_vectors(phys_addr_t phys_vector_base);
 phys_addr_t __hyp_get_vectors(void);
 
-static inline void sync_boot_mode(void)
-{
-        /*
-         * As secondaries write to __boot_cpu_mode with caches disabled, we
-         * must flush the corresponding cache entries to ensure the visibility
-         * of their writes.
-         */
-        __flush_dcache_area(__boot_cpu_mode, sizeof(__boot_cpu_mode));
-}
-
 /* Reports the availability of HYP mode */
 static inline bool is_hyp_mode_available(void)
 {
-        sync_boot_mode();
         return (__boot_cpu_mode[0] == BOOT_CPU_MODE_EL2 &&
                 __boot_cpu_mode[1] == BOOT_CPU_MODE_EL2);
 }
@@ -59,7 +47,6 @@ static inline bool is_hyp_mode_available(void)
 /* Check if the bootloader has booted CPUs in different modes */
 static inline bool is_hyp_mode_mismatched(void)
 {
-        sync_boot_mode();
         return __boot_cpu_mode[0] != __boot_cpu_mode[1];
 }
 
diff --git a/arch/arm64/kernel/head.S b/arch/arm64/kernel/head.S
index 1fe5d8d2bdfd..0fd565000772 100644
--- a/arch/arm64/kernel/head.S
+++ b/arch/arm64/kernel/head.S
@@ -26,6 +26,7 @@
 #include <asm/assembler.h>
 #include <asm/ptrace.h>
 #include <asm/asm-offsets.h>
+#include <asm/cache.h>
 #include <asm/cputype.h>
 #include <asm/memory.h>
 #include <asm/thread_info.h>
@@ -229,7 +230,11 @@ ENTRY(set_cpu_boot_mode_flag)
         cmp     w20, #BOOT_CPU_MODE_EL2
         b.ne    1f
         add     x1, x1, #4
-1:      str     w20, [x1]                       // This CPU has booted in EL1
+1:      dc      cvac, x1                        // Clean potentially dirty cache line
+        dsb     sy
+        str     w20, [x1]                       // This CPU has booted in EL1
+        dc      civac, x1                       // Clean&invalidate potentially stale cache line
+        dsb     sy
         ret
 ENDPROC(set_cpu_boot_mode_flag)
 
@@ -240,8 +245,9 @@ ENDPROC(set_cpu_boot_mode_flag)
  * This is not in .bss, because we set it sufficiently early that the boot-time
  * zeroing of .bss would clobber it.
  */
-        .pushsection    .data
+        .pushsection    .data..cacheline_aligned
 ENTRY(__boot_cpu_mode)
+        .align  L1_CACHE_SHIFT
         .long   BOOT_CPU_MODE_EL2
         .long   0
         .popsection
@@ -408,6 +414,15 @@ ENDPROC(__calc_phys_offset)
  */
 __create_page_tables:
         pgtbl   x25, x26, x24                   // idmap_pg_dir and swapper_pg_dir addresses
+        mov     x27, lr
+
+        /*
+         * Invalidate the idmap and swapper page tables to avoid potential
+         * dirty cache lines being evicted.
+         */
+        mov     x0, x25
+        add     x1, x26, #SWAPPER_DIR_SIZE
+        bl      __inval_cache_range
 
         /*
          * Clear the idmap and swapper page tables.
@@ -467,6 +482,17 @@ __create_page_tables:
         ldr     x5, =FIXADDR_TOP                // Fixed mapping virtual address
         add     x0, x26, #2 * PAGE_SIZE         // section table address
         create_pgd_entry x26, x0, x5, x6, x7
+
+        /*
+         * Since the page tables have been populated with non-cacheable
+         * accesses (MMU disabled), invalidate the idmap and swapper page
+         * tables again to remove any speculatively loaded cache lines.
+         */
+        mov     x0, x25
+        add     x1, x26, #SWAPPER_DIR_SIZE
+        bl      __inval_cache_range
+
+        mov     lr, x27
         ret
 ENDPROC(__create_page_tables)
         .ltorg
diff --git a/arch/arm64/kernel/perf_event.c b/arch/arm64/kernel/perf_event.c
index e868c72a7938..baf5afb7e6a0 100644
--- a/arch/arm64/kernel/perf_event.c
+++ b/arch/arm64/kernel/perf_event.c
@@ -1386,6 +1386,7 @@ user_backtrace(struct frame_tail __user *tail,
         return buftail.fp;
 }
 
+#ifdef CONFIG_COMPAT
 /*
  * The registers we're interested in are at the end of the variable
  * length saved register structure. The fp points at the end of this
@@ -1430,6 +1431,7 @@ compat_user_backtrace(struct compat_frame_tail __user *tail,
 
         return (struct compat_frame_tail __user *)compat_ptr(buftail.fp) - 1;
 }
+#endif /* CONFIG_COMPAT */
 
 void perf_callchain_user(struct perf_callchain_entry *entry,
                          struct pt_regs *regs)
@@ -1451,6 +1453,7 @@ void perf_callchain_user(struct perf_callchain_entry *entry,
                        tail && !((unsigned long)tail & 0xf))
                         tail = user_backtrace(tail, entry);
         } else {
+#ifdef CONFIG_COMPAT
                 /* AARCH32 compat mode */
                 struct compat_frame_tail __user *tail;
 
@@ -1459,6 +1462,7 @@ void perf_callchain_user(struct perf_callchain_entry *entry,
                 while ((entry->nr < PERF_MAX_STACK_DEPTH) &&
                         tail && !((unsigned long)tail & 0x3))
                         tail = compat_user_backtrace(tail, entry);
+#endif
         }
 }
 
diff --git a/arch/arm64/kernel/perf_regs.c b/arch/arm64/kernel/perf_regs.c
index f2d6f0a36d63..422ebd63b619 100644
--- a/arch/arm64/kernel/perf_regs.c
+++ b/arch/arm64/kernel/perf_regs.c
@@ -2,6 +2,8 @@
 #include <linux/kernel.h>
 #include <linux/perf_event.h>
 #include <linux/bug.h>
+
+#include <asm/compat.h>
 #include <asm/perf_regs.h>
 #include <asm/ptrace.h>
 
diff --git a/arch/arm64/mm/cache.S b/arch/arm64/mm/cache.S
index c46f48b33c14..fda756875fa6 100644
--- a/arch/arm64/mm/cache.S
+++ b/arch/arm64/mm/cache.S
@@ -168,6 +168,14 @@ ENTRY(__flush_dcache_area)
 ENDPROC(__flush_dcache_area)
 
 /*
+ *      __inval_cache_range(start, end)
+ *      - start   - start address of region
+ *      - end     - end address of region
+ */
+ENTRY(__inval_cache_range)
+        /* FALLTHROUGH */
+
+/*
  *      __dma_inv_range(start, end)
  *      - start   - virtual start address of region
  *      - end     - virtual end address of region
@@ -175,14 +183,22 @@ ENDPROC(__flush_dcache_area)
 __dma_inv_range:
         dcache_line_size x2, x3
         sub     x3, x2, #1
-        bic     x0, x0, x3
+        tst     x1, x3                          // end cache line aligned?
         bic     x1, x1, x3
-1:      dc      ivac, x0                        // invalidate D / U line
-        add     x0, x0, x2
+        b.eq    1f
+        dc      civac, x1                       // clean & invalidate D / U line
+1:      tst     x0, x3                          // start cache line aligned?
+        bic     x0, x0, x3
+        b.eq    2f
+        dc      civac, x0                       // clean & invalidate D / U line
+        b       3f
+2:      dc      ivac, x0                        // invalidate D / U line
+3:      add     x0, x0, x2
         cmp     x0, x1
-        b.lo    1b
+        b.lo    2b
         dsb     sy
         ret
+ENDPROC(__inval_cache_range)
 ENDPROC(__dma_inv_range)
 
 /*
diff --git a/arch/arm64/mm/proc.S b/arch/arm64/mm/proc.S
index e085ee6ef4e2..9042aff5e9e3 100644
--- a/arch/arm64/mm/proc.S
+++ b/arch/arm64/mm/proc.S
@@ -28,14 +28,21 @@
 
 #include "proc-macros.S"
 
-#ifndef CONFIG_SMP
-/* PTWs cacheable, inner/outer WBWA not shareable */
-#define TCR_FLAGS       TCR_IRGN_WBWA | TCR_ORGN_WBWA
+#ifdef CONFIG_ARM64_64K_PAGES
+#define TCR_TG_FLAGS    TCR_TG0_64K | TCR_TG1_64K
+#else
+#define TCR_TG_FLAGS    TCR_TG0_4K | TCR_TG1_4K
+#endif
+
+#ifdef CONFIG_SMP
+#define TCR_SMP_FLAGS   TCR_SHARED
 #else
-/* PTWs cacheable, inner/outer WBWA shareable */
-#define TCR_FLAGS       TCR_IRGN_WBWA | TCR_ORGN_WBWA | TCR_SHARED
+#define TCR_SMP_FLAGS   0
 #endif
 
+/* PTWs cacheable, inner/outer WBWA */
+#define TCR_CACHE_FLAGS TCR_IRGN_WBWA | TCR_ORGN_WBWA
+
 #define MAIR(attr, mt)  ((attr) << ((mt) * 8))
 
 /*
@@ -209,18 +216,14 @@ ENTRY(__cpu_setup)
          * Set/prepare TCR and TTBR. We use 512GB (39-bit) address range for
          * both user and kernel.
          */
-        ldr     x10, =TCR_TxSZ(VA_BITS) | TCR_FLAGS | \
-                      TCR_ASID16 | TCR_TBI0 | (1 << 31)
+        ldr     x10, =TCR_TxSZ(VA_BITS) | TCR_CACHE_FLAGS | TCR_SMP_FLAGS | \
+                        TCR_TG_FLAGS | TCR_ASID16 | TCR_TBI0
         /*
          * Read the PARange bits from ID_AA64MMFR0_EL1 and set the IPS bits in
          * TCR_EL1.
          */
         mrs     x9, ID_AA64MMFR0_EL1
         bfi     x10, x9, #32, #3
-#ifdef CONFIG_ARM64_64K_PAGES
-        orr     x10, x10, TCR_TG0_64K
-        orr     x10, x10, TCR_TG1_64K
-#endif
         msr     tcr_el1, x10
         ret                                     // return to head.S
 ENDPROC(__cpu_setup)