Merge tag 'core_entry_for_v5.17_rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip

Pull thread_info flag accessor helper updates from Borislav Petkov:
 "Add a set of thread_info.flags accessors which snapshot it before
  accesing it in order to prevent any potential data races, and convert
  all users to those new accessors"

* tag 'core_entry_for_v5.17_rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  powerpc: Snapshot thread flags
  powerpc: Avoid discarding flags in system_call_exception()
  openrisc: Snapshot thread flags
  microblaze: Snapshot thread flags
  arm64: Snapshot thread flags
  ARM: Snapshot thread flags
  alpha: Snapshot thread flags
  sched: Snapshot thread flags
  entry: Snapshot thread flags
  x86: Snapshot thread flags
  thread_info: Add helpers to snapshot thread flags

arch/alpha/kernel/signal.c

@@ -535,6 +535,6 @@ do_work_pending(struct pt_regs *regs, unsigned long thread_flags,
 			}
 		}
 		local_irq_disable();
-		thread_flags = current_thread_info()->flags;
+		thread_flags = read_thread_flags();
 	} while (thread_flags & _TIF_WORK_MASK);
 }

arch/arm/kernel/signal.c

@@ -631,7 +631,7 @@ do_work_pending(struct pt_regs *regs, unsigned int thread_flags, int syscall)
 			}
 		}
 		local_irq_disable();
-		thread_flags = current_thread_info()->flags;
+		thread_flags = read_thread_flags();
 	} while (thread_flags & _TIF_WORK_MASK);
 	return 0;
 }

arch/arm/mm/alignment.c

@@ -990,7 +990,7 @@ do_alignment(unsigned long addr, unsigned int fsr, struct pt_regs *regs)
 		 * there is no work pending for this thread.
 		 */
 		raw_local_irq_disable();
-		if (!(current_thread_info()->flags & _TIF_WORK_MASK))
+		if (!(read_thread_flags() & _TIF_WORK_MASK))
 			set_cr(cr_no_alignment);
 	}
 

arch/arm64/kernel/entry-common.c

@@ -129,7 +129,7 @@ static __always_inline void prepare_exit_to_user_mode(struct pt_regs *regs)
 
 	local_daif_mask();
 
-	flags = READ_ONCE(current_thread_info()->flags);
+	flags = read_thread_flags();
 	if (unlikely(flags & _TIF_WORK_MASK))
 		do_notify_resume(regs, flags);
 }

arch/arm64/kernel/ptrace.c

@@ -1839,7 +1839,7 @@ static void tracehook_report_syscall(struct pt_regs *regs,
 
 int syscall_trace_enter(struct pt_regs *regs)
 {
-	unsigned long flags = READ_ONCE(current_thread_info()->flags);
+	unsigned long flags = read_thread_flags();
 
 	if (flags & (_TIF_SYSCALL_EMU | _TIF_SYSCALL_TRACE)) {
 		tracehook_report_syscall(regs, PTRACE_SYSCALL_ENTER);
@@ -1862,7 +1862,7 @@ int syscall_trace_enter(struct pt_regs *regs)
 
 void syscall_trace_exit(struct pt_regs *regs)
 {
-	unsigned long flags = READ_ONCE(current_thread_info()->flags);
+	unsigned long flags = read_thread_flags();
 
 	audit_syscall_exit(regs);
 

arch/arm64/kernel/signal.c

@@ -948,7 +948,7 @@ void do_notify_resume(struct pt_regs *regs, unsigned long thread_flags)
 		}
 
 		local_daif_mask();
-		thread_flags = READ_ONCE(current_thread_info()->flags);
+		thread_flags = read_thread_flags();
 	} while (thread_flags & _TIF_WORK_MASK);
 }
 

arch/arm64/kernel/syscall.c

@@ -81,7 +81,7 @@ void syscall_trace_exit(struct pt_regs *regs);
 static void el0_svc_common(struct pt_regs *regs, int scno, int sc_nr,
 			   const syscall_fn_t syscall_table[])
 {
-	unsigned long flags = current_thread_info()->flags;
+	unsigned long flags = read_thread_flags();
 
 	regs->orig_x0 = regs->regs[0];
 	regs->syscallno = scno;
@@ -148,7 +148,7 @@ static void el0_svc_common(struct pt_regs *regs, int scno, int sc_nr,
 	 */
 	if (!has_syscall_work(flags) && !IS_ENABLED(CONFIG_DEBUG_RSEQ)) {
 		local_daif_mask();
-		flags = current_thread_info()->flags;
+		flags = read_thread_flags();
 		if (!has_syscall_work(flags) && !(flags & _TIF_SINGLESTEP))
 			return;
 		local_daif_restore(DAIF_PROCCTX);

arch/microblaze/kernel/signal.c

@@ -283,7 +283,7 @@ static void do_signal(struct pt_regs *regs, int in_syscall)
 #ifdef DEBUG_SIG
 	pr_info("do signal: %p %d\n", regs, in_syscall);
 	pr_info("do signal2: %lx %lx %ld [%lx]\n", regs->pc, regs->r1,
-			regs->r12, current_thread_info()->flags);
+			regs->r12, read_thread_flags());
 #endif
 
 	if (get_signal(&ksig)) {

arch/openrisc/kernel/signal.c

@@ -313,7 +313,7 @@ do_work_pending(struct pt_regs *regs, unsigned int thread_flags, int syscall)
 			}
 		}
 		local_irq_disable();
-		thread_flags = current_thread_info()->flags;
+		thread_flags = read_thread_flags();
 	} while (thread_flags & _TIF_WORK_MASK);
 	return 0;
 }

arch/powerpc/kernel/interrupt.c

@@ -148,7 +148,7 @@ notrace long system_call_exception(long r3, long r4, long r5,
 	 */
 	if (IS_ENABLED(CONFIG_PPC_TRANSACTIONAL_MEM) &&
 			unlikely(MSR_TM_TRANSACTIONAL(regs->msr)))
-		current_thread_info()->flags |= _TIF_RESTOREALL;
+		set_bits(_TIF_RESTOREALL, &current_thread_info()->flags);
 
 	/*
 	 * If the system call was made with a transaction active, doom it and
@@ -181,7 +181,7 @@ notrace long system_call_exception(long r3, long r4, long r5,
 
 	local_irq_enable();
 
-	if (unlikely(current_thread_info()->flags & _TIF_SYSCALL_DOTRACE)) {
+	if (unlikely(read_thread_flags() & _TIF_SYSCALL_DOTRACE)) {
 		if (unlikely(trap_is_unsupported_scv(regs))) {
 			/* Unsupported scv vector */
 			_exception(SIGILL, regs, ILL_ILLOPC, regs->nip);
@@ -343,7 +343,7 @@ interrupt_exit_user_prepare_main(unsigned long ret, struct pt_regs *regs)
 	unsigned long ti_flags;
 
 again:
-	ti_flags = READ_ONCE(current_thread_info()->flags);
+	ti_flags = read_thread_flags();
 	while (unlikely(ti_flags & (_TIF_USER_WORK_MASK & ~_TIF_RESTORE_TM))) {
 		local_irq_enable();
 		if (ti_flags & _TIF_NEED_RESCHED) {
@@ -359,7 +359,7 @@ interrupt_exit_user_prepare_main(unsigned long ret, struct pt_regs *regs)
 			do_notify_resume(regs, ti_flags);
 		}
 		local_irq_disable();
-		ti_flags = READ_ONCE(current_thread_info()->flags);
+		ti_flags = read_thread_flags();
 	}
 
 	if (IS_ENABLED(CONFIG_PPC_BOOK3S_64) && IS_ENABLED(CONFIG_PPC_FPU)) {
@@ -437,7 +437,7 @@ notrace unsigned long syscall_exit_prepare(unsigned long r3,
 	/* Check whether the syscall is issued inside a restartable sequence */
 	rseq_syscall(regs);
 
-	ti_flags = current_thread_info()->flags;
+	ti_flags = read_thread_flags();
 
 	if (unlikely(r3 >= (unsigned long)-MAX_ERRNO) && is_not_scv) {
 		if (likely(!(ti_flags & (_TIF_NOERROR | _TIF_RESTOREALL)))) {
@@ -532,8 +532,7 @@ notrace unsigned long interrupt_exit_kernel_prepare(struct pt_regs *regs)
 	unsigned long flags;
 	unsigned long ret = 0;
 	unsigned long kuap;
-	bool stack_store = current_thread_info()->flags &
-						_TIF_EMULATE_STACK_STORE;
+	bool stack_store = read_thread_flags() & _TIF_EMULATE_STACK_STORE;
 
 	if (regs_is_unrecoverable(regs))
 		unrecoverable_exception(regs);
@@ -554,7 +553,7 @@ notrace unsigned long interrupt_exit_kernel_prepare(struct pt_regs *regs)
 again:
 		if (IS_ENABLED(CONFIG_PREEMPT)) {
 			/* Return to preemptible kernel context */
-			if (unlikely(current_thread_info()->flags & _TIF_NEED_RESCHED)) {
+			if (unlikely(read_thread_flags() & _TIF_NEED_RESCHED)) {
 				if (preempt_count() == 0)
 					preempt_schedule_irq();
 			}

arch/powerpc/kernel/ptrace/ptrace.c

@@ -260,8 +260,7 @@ long do_syscall_trace_enter(struct pt_regs *regs)
 {
 	u32 flags;
 
-	flags = READ_ONCE(current_thread_info()->flags) &
-		(_TIF_SYSCALL_EMU | _TIF_SYSCALL_TRACE);
+	flags = read_thread_flags() & (_TIF_SYSCALL_EMU | _TIF_SYSCALL_TRACE);
 
 	if (flags) {
 		int rc = tracehook_report_syscall_entry(regs);

arch/x86/kernel/process.c

@@ -365,7 +365,7 @@ void arch_setup_new_exec(void)
 		clear_thread_flag(TIF_SSBD);
 		task_clear_spec_ssb_disable(current);
 		task_clear_spec_ssb_noexec(current);
-		speculation_ctrl_update(task_thread_info(current)->flags);
+		speculation_ctrl_update(read_thread_flags());
 	}
 }
 
@@ -617,7 +617,7 @@ static unsigned long speculation_ctrl_update_tif(struct task_struct *tsk)
 			clear_tsk_thread_flag(tsk, TIF_SPEC_IB);
 	}
 	/* Return the updated threadinfo flags*/
-	return task_thread_info(tsk)->flags;
+	return read_task_thread_flags(tsk);
 }
 
 void speculation_ctrl_update(unsigned long tif)
@@ -653,8 +653,8 @@ void __switch_to_xtra(struct task_struct *prev_p, struct task_struct *next_p)
 {
 	unsigned long tifp, tifn;
 
-	tifn = READ_ONCE(task_thread_info(next_p)->flags);
-	tifp = READ_ONCE(task_thread_info(prev_p)->flags);
+	tifn = read_task_thread_flags(next_p);
+	tifp = read_task_thread_flags(prev_p);
 
 	switch_to_bitmap(tifp);
 

arch/x86/kernel/process.h

@@ -13,8 +13,8 @@ void __switch_to_xtra(struct task_struct *prev_p, struct task_struct *next_p);
 static inline void switch_to_extra(struct task_struct *prev,
 				   struct task_struct *next)
 {
-	unsigned long next_tif = task_thread_info(next)->flags;
-	unsigned long prev_tif = task_thread_info(prev)->flags;
+	unsigned long next_tif = read_task_thread_flags(next);
+	unsigned long prev_tif = read_task_thread_flags(prev);
 
 	if (IS_ENABLED(CONFIG_SMP)) {
 		/*

arch/x86/mm/tlb.c

@@ -361,7 +361,7 @@ static void l1d_flush_evaluate(unsigned long prev_mm, unsigned long next_mm,
 
 static unsigned long mm_mangle_tif_spec_bits(struct task_struct *next)
 {
-	unsigned long next_tif = task_thread_info(next)->flags;
+	unsigned long next_tif = read_task_thread_flags(next);
 	unsigned long spec_bits = (next_tif >> TIF_SPEC_IB) & LAST_USER_MM_SPEC_MASK;
 
 	/*

include/linux/entry-kvm.h

@@ -75,7 +75,7 @@ static inline void xfer_to_guest_mode_prepare(void)
  */
 static inline bool __xfer_to_guest_mode_work_pending(void)
 {
-	unsigned long ti_work = READ_ONCE(current_thread_info()->flags);
+	unsigned long ti_work = read_thread_flags();
 
 	return !!(ti_work & XFER_TO_GUEST_MODE_WORK);
 }

include/linux/thread_info.h

@@ -118,6 +118,15 @@ static inline int test_ti_thread_flag(struct thread_info *ti, int flag)
 	return test_bit(flag, (unsigned long *)&ti->flags);
 }
 
+/*
+ * This may be used in noinstr code, and needs to be __always_inline to prevent
+ * inadvertent instrumentation.
+ */
+static __always_inline unsigned long read_ti_thread_flags(struct thread_info *ti)
+{
+	return READ_ONCE(ti->flags);
+}
+
 #define set_thread_flag(flag) \
 	set_ti_thread_flag(current_thread_info(), flag)
 #define clear_thread_flag(flag) \
@@ -130,6 +139,11 @@ static inline int test_ti_thread_flag(struct thread_info *ti, int flag)
 	test_and_clear_ti_thread_flag(current_thread_info(), flag)
 #define test_thread_flag(flag) \
 	test_ti_thread_flag(current_thread_info(), flag)
+#define read_thread_flags() \
+	read_ti_thread_flags(current_thread_info())
+
+#define read_task_thread_flags(t) \
+	read_ti_thread_flags(task_thread_info(t))
 
 #ifdef CONFIG_GENERIC_ENTRY
 #define set_syscall_work(fl) \

kernel/entry/common.c

@@ -187,7 +187,7 @@ static unsigned long exit_to_user_mode_loop(struct pt_regs *regs,
 		/* Check if any of the above work has queued a deferred wakeup */
 		tick_nohz_user_enter_prepare();
 
-		ti_work = READ_ONCE(current_thread_info()->flags);
+		ti_work = read_thread_flags();
 	}
 
 	/* Return the latest work state for arch_exit_to_user_mode() */
@@ -196,7 +196,7 @@ static unsigned long exit_to_user_mode_loop(struct pt_regs *regs,
 
 static void exit_to_user_mode_prepare(struct pt_regs *regs)
 {
-	unsigned long ti_work = READ_ONCE(current_thread_info()->flags);
+	unsigned long ti_work = read_thread_flags();
 
 	lockdep_assert_irqs_disabled();
 

kernel/entry/kvm.c

@@ -26,7 +26,7 @@ static int xfer_to_guest_mode_work(struct kvm_vcpu *vcpu, unsigned long ti_work)
 		if (ret)
 			return ret;
 
-		ti_work = READ_ONCE(current_thread_info()->flags);
+		ti_work = read_thread_flags();
 	} while (ti_work & XFER_TO_GUEST_MODE_WORK || need_resched());
 	return 0;
 }
@@ -43,7 +43,7 @@ int xfer_to_guest_mode_handle_work(struct kvm_vcpu *vcpu)
 	 * disabled in the inner loop before going into guest mode. No need
 	 * to disable interrupts here.
 	 */
-	ti_work = READ_ONCE(current_thread_info()->flags);
+	ti_work = read_thread_flags();
 	if (!(ti_work & XFER_TO_GUEST_MODE_WORK))
 		return 0;
 

kernel/sched/core.c

@@ -8520,7 +8520,7 @@ void sched_show_task(struct task_struct *p)
 	rcu_read_unlock();
 	pr_cont(" stack:%5lu pid:%5d ppid:%6d flags:0x%08lx\n",
 		free, task_pid_nr(p), ppid,
-		(unsigned long)task_thread_info(p)->flags);
+		read_task_thread_flags(p));
 
 	print_worker_info(KERN_INFO, p);
 	print_stop_info(KERN_INFO, p);