arm64: Disable TTBR0_EL1 during normal kernel execution

When the TTBR0 PAN feature is enabled, the kernel entry points need to
disable access to TTBR0_EL1. The PAN status of the interrupted context
is stored as part of the saved pstate, reusing the PSR_PAN_BIT (22).
Restoring access to TTBR0_EL1 is done on exception return if returning
to user or returning to a context where PAN was disabled.

Context switching via switch_mm() must defer the update of TTBR0_EL1
until a return to user or an explicit uaccess_enable() call.

Special care needs to be taken for two cases where TTBR0_EL1 is set
outside the normal kernel context switch operation: EFI run-time
services (via efi_set_pgd) and CPU suspend (via cpu_(un)install_idmap).
Code has been added to avoid deferred TTBR0_EL1 switching as in
switch_mm() and restore the reserved TTBR0_EL1 when uninstalling the
special TTBR0_EL1.

User cache maintenance (user_cache_maint_handler and
__flush_cache_user_range) needs the TTBR0_EL1 re-instated since the
operations are performed by user virtual address.

This patch also removes a stale comment on the switch_mm() function.

Cc: Will Deacon <will.deacon@arm.com>
Cc: James Morse <james.morse@arm.com>
Cc: Kees Cook <keescook@chromium.org>
Cc: Mark Rutland <mark.rutland@arm.com>
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>

arch/arm64/include/asm/efi.h

 #ifndef _ASM_EFI_H
 #define _ASM_EFI_H
 
+#include <asm/cpufeature.h>
 #include <asm/io.h>
 #include <asm/mmu_context.h>
 #include <asm/neon.h>
@@ -75,7 +76,30 @@ static inline void efifb_setup_from_dmi(struct screen_info *si, const char *opt)
 
 static inline void efi_set_pgd(struct mm_struct *mm)
 {
-	switch_mm(NULL, mm, NULL);
+	__switch_mm(mm);
+
+	if (system_uses_ttbr0_pan()) {
+		if (mm != current->active_mm) {
+			/*
+			 * Update the current thread's saved ttbr0 since it is
+			 * restored as part of a return from exception. Set
+			 * the hardware TTBR0_EL1 using cpu_switch_mm()
+			 * directly to enable potential errata workarounds.
+			 */
+			update_saved_ttbr0(current, mm);
+			cpu_switch_mm(mm->pgd, mm);
+		} else {
+			/*
+			 * Defer the switch to the current thread's TTBR0_EL1
+			 * until uaccess_enable(). Restore the current
+			 * thread's saved ttbr0 corresponding to its active_mm
+			 * (if different from init_mm).
+			 */
+			cpu_set_reserved_ttbr0();
+			if (current->active_mm != &init_mm)
+				update_saved_ttbr0(current, current->active_mm);
+		}
+	}
 }
 
 void efi_virtmap_load(void);

arch/arm64/include/asm/mmu_context.h

@@ -23,6 +23,7 @@
 #include <linux/sched.h>
 
 #include <asm/cacheflush.h>
+#include <asm/cpufeature.h>
 #include <asm/proc-fns.h>
 #include <asm-generic/mm_hooks.h>
 #include <asm/cputype.h>
@@ -103,7 +104,7 @@ static inline void cpu_uninstall_idmap(void)
 	local_flush_tlb_all();
 	cpu_set_default_tcr_t0sz();
 
-	if (mm != &init_mm)
+	if (mm != &init_mm && !system_uses_ttbr0_pan())
 		cpu_switch_mm(mm->pgd, mm);
 }
 
@@ -163,20 +164,26 @@ enter_lazy_tlb(struct mm_struct *mm, struct task_struct *tsk)
 {
 }
 
-/*
- * This is the actual mm switch as far as the scheduler
- * is concerned.  No registers are touched.  We avoid
- * calling the CPU specific function when the mm hasn't
- * actually changed.
- */
-static inline void
-switch_mm(struct mm_struct *prev, struct mm_struct *next,
-	  struct task_struct *tsk)
+#ifdef CONFIG_ARM64_SW_TTBR0_PAN
+static inline void update_saved_ttbr0(struct task_struct *tsk,
+				      struct mm_struct *mm)
 {
-	unsigned int cpu = smp_processor_id();
+	if (system_uses_ttbr0_pan()) {
+		BUG_ON(mm->pgd == swapper_pg_dir);
+		task_thread_info(tsk)->ttbr0 =
+			virt_to_phys(mm->pgd) | ASID(mm) << 48;
+	}
+}
+#else
+static inline void update_saved_ttbr0(struct task_struct *tsk,
+				      struct mm_struct *mm)
+{
+}
+#endif
 
-	if (prev == next)
-		return;
+static inline void __switch_mm(struct mm_struct *next)
+{
+	unsigned int cpu = smp_processor_id();
 
 	/*
 	 * init_mm.pgd does not contain any user mappings and it is always
@@ -190,8 +197,26 @@ switch_mm(struct mm_struct *prev, struct mm_struct *next,
 	check_and_switch_context(next, cpu);
 }
 
+static inline void
+switch_mm(struct mm_struct *prev, struct mm_struct *next,
+	  struct task_struct *tsk)
+{
+	if (prev != next)
+		__switch_mm(next);
+
+	/*
+	 * Update the saved TTBR0_EL1 of the scheduled-in task as the previous
+	 * value may have not been initialised yet (activate_mm caller) or the
+	 * ASID has changed since the last run (following the context switch
+	 * of another thread of the same process). Avoid setting the reserved
+	 * TTBR0_EL1 to swapper_pg_dir (init_mm; e.g. via idle_task_exit).
+	 */
+	if (next != &init_mm)
+		update_saved_ttbr0(tsk, next);
+}
+
 #define deactivate_mm(tsk,mm)	do { } while (0)
-#define activate_mm(prev,next)	switch_mm(prev, next, NULL)
+#define activate_mm(prev,next)	switch_mm(prev, next, current)
 
 void verify_cpu_asid_bits(void);
 

arch/arm64/kernel/entry.S

@@ -29,7 +29,9 @@
 #include <asm/esr.h>
 #include <asm/irq.h>
 #include <asm/memory.h>
+#include <asm/ptrace.h>
 #include <asm/thread_info.h>
+#include <asm/uaccess.h>
 #include <asm/unistd.h>
 
 /*
@@ -108,6 +110,32 @@
 	mrs	x22, elr_el1
 	mrs	x23, spsr_el1
 	stp	lr, x21, [sp, #S_LR]
+
+#ifdef CONFIG_ARM64_SW_TTBR0_PAN
+	/*
+	 * Set the TTBR0 PAN bit in SPSR. When the exception is taken from
+	 * EL0, there is no need to check the state of TTBR0_EL1 since
+	 * accesses are always enabled.
+	 * Note that the meaning of this bit differs from the ARMv8.1 PAN
+	 * feature as all TTBR0_EL1 accesses are disabled, not just those to
+	 * user mappings.
+	 */
+alternative_if ARM64_HAS_PAN
+	b	1f				// skip TTBR0 PAN
+alternative_else_nop_endif
+
+	.if	\el != 0
+	mrs	x21, ttbr0_el1
+	tst	x21, #0xffff << 48		// Check for the reserved ASID
+	orr	x23, x23, #PSR_PAN_BIT		// Set the emulated PAN in the saved SPSR
+	b.eq	1f				// TTBR0 access already disabled
+	and	x23, x23, #~PSR_PAN_BIT		// Clear the emulated PAN in the saved SPSR
+	.endif
+
+	__uaccess_ttbr0_disable x21
+1:
+#endif
+
 	stp	x22, x23, [sp, #S_PC]
 
 	/*
@@ -146,6 +174,40 @@
 	ldp	x21, x22, [sp, #S_PC]		// load ELR, SPSR
 	.if	\el == 0
 	ct_user_enter
+	.endif
+
+#ifdef CONFIG_ARM64_SW_TTBR0_PAN
+	/*
+	 * Restore access to TTBR0_EL1. If returning to EL0, no need for SPSR
+	 * PAN bit checking.
+	 */
+alternative_if ARM64_HAS_PAN
+	b	2f				// skip TTBR0 PAN
+alternative_else_nop_endif
+
+	.if	\el != 0
+	tbnz	x22, #22, 1f			// Skip re-enabling TTBR0 access if the PSR_PAN_BIT is set
+	.endif
+
+	__uaccess_ttbr0_enable x0
+
+	.if	\el == 0
+	/*
+	 * Enable errata workarounds only if returning to user. The only
+	 * workaround currently required for TTBR0_EL1 changes are for the
+	 * Cavium erratum 27456 (broadcast TLBI instructions may cause I-cache
+	 * corruption).
+	 */
+	post_ttbr0_update_workaround
+	.endif
+1:
+	.if	\el != 0
+	and	x22, x22, #~PSR_PAN_BIT		// ARMv8.0 CPUs do not understand this bit
+	.endif
+2:
+#endif
+
+	.if	\el == 0
 	ldr	x23, [sp, #S_SP]		// load return stack pointer
 	msr	sp_el0, x23
 #ifdef CONFIG_ARM64_ERRATUM_845719
@@ -161,6 +223,7 @@ alternative_if ARM64_WORKAROUND_845719
 alternative_else_nop_endif
 #endif
 	.endif
+
 	msr	elr_el1, x21			// set up the return data
 	msr	spsr_el1, x22
 	ldp	x0, x1, [sp, #16 * 0]

arch/arm64/kernel/setup.c

@@ -291,6 +291,15 @@ void __init setup_arch(char **cmdline_p)
 	smp_init_cpus();
 	smp_build_mpidr_hash();
 
+#ifdef CONFIG_ARM64_SW_TTBR0_PAN
+	/*
+	 * Make sure init_thread_info.ttbr0 always generates translation
+	 * faults in case uaccess_enable() is inadvertently called by the init
+	 * thread.
+	 */
+	init_task.thread_info.ttbr0 = virt_to_phys(empty_zero_page);
+#endif
+
 #ifdef CONFIG_VT
 #if defined(CONFIG_VGA_CONSOLE)
 	conswitchp = &vga_con;

arch/arm64/kernel/traps.c

@@ -440,9 +440,10 @@ int cpu_enable_cache_maint_trap(void *__unused)
 }
 
 #define __user_cache_maint(insn, address, res)			\
-	if (untagged_addr(address) >= user_addr_max())		\
+	if (untagged_addr(address) >= user_addr_max()) {	\
 		res = -EFAULT;					\
-	else							\
+	} else {						\
+		uaccess_ttbr0_enable();				\
 		asm volatile (					\
 			"1:	" insn ", %1\n"			\
 			"	mov	%w0, #0\n"		\
@@ -454,7 +455,9 @@ int cpu_enable_cache_maint_trap(void *__unused)
 			"	.popsection\n"			\
 			_ASM_EXTABLE(1b, 3b)			\
 			: "=r" (res)				\
-			: "r" (address), "i" (-EFAULT) )
+			: "r" (address), "i" (-EFAULT));	\
+		uaccess_ttbr0_disable();			\
+	}
 
 static void user_cache_maint_handler(unsigned int esr, struct pt_regs *regs)
 {

arch/arm64/mm/cache.S

@@ -23,6 +23,7 @@
 #include <asm/assembler.h>
 #include <asm/cpufeature.h>
 #include <asm/alternative.h>
+#include <asm/uaccess.h>
 
 /*
  *	flush_icache_range(start,end)
@@ -48,6 +49,7 @@ ENTRY(flush_icache_range)
  *	- end     - virtual end address of region
  */
 ENTRY(__flush_cache_user_range)
+	uaccess_ttbr0_enable x2, x3
 	dcache_line_size x2, x3
 	sub	x3, x2, #1
 	bic	x4, x0, x3
@@ -69,10 +71,12 @@ USER(9f, ic	ivau, x4	)		// invalidate I line PoU
 	dsb	ish
 	isb
 	mov	x0, #0
+1:
+	uaccess_ttbr0_disable x1
 	ret
 9:
 	mov	x0, #-EFAULT
-	ret
+	b	1b
 ENDPROC(flush_icache_range)
 ENDPROC(__flush_cache_user_range)
 

arch/arm64/mm/context.c

@@ -221,7 +221,12 @@ void check_and_switch_context(struct mm_struct *mm, unsigned int cpu)
 	raw_spin_unlock_irqrestore(&cpu_asid_lock, flags);
 
 switch_mm_fastpath:
-	cpu_switch_mm(mm->pgd, mm);
+	/*
+	 * Defer TTBR0_EL1 setting for user threads to uaccess_enable() when
+	 * emulating PAN.
+	 */
+	if (!system_uses_ttbr0_pan())
+		cpu_switch_mm(mm->pgd, mm);
 }
 
 static int asids_init(void)