powerpc: copy_thread differentiate kthreads and user mode threads

When copy_thread is given a kernel function to run in arg->fn, this
does not necessarily mean it is a kernel thread. User threads can be
created this way (e.g., kernel_init, see also x86's copy_thread()).
These threads run a kernel function which may call kernel_execve()
and return, which returns like a userspace exec(2) syscall.

Kernel threads are to  be differentiated with PF_KTHREAD, will always
have arg->fn set, and should never return from that function, instead
calling kthread_exit() to exit.

Create separate paths for the kthread and user kernel thread creation
logic. The kthread path will never exit and does not require a user
interrupt frame, so it gets a minimal stack frame.
Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://msgid.link/20230325122904.2375060-6-npiggin@gmail.com

arch/powerpc/include/asm/thread_info.h

@@ -176,9 +176,11 @@ static inline bool test_thread_local_flags(unsigned int flags)
 #ifdef CONFIG_COMPAT
 #define is_32bit_task()	(test_thread_flag(TIF_32BIT))
 #define is_tsk_32bit_task(tsk)	(test_tsk_thread_flag(tsk, TIF_32BIT))
+#define clear_tsk_compat_task(tsk) (clear_tsk_thread_flag(p, TIF_32BIT))
 #else
 #define is_32bit_task()	(IS_ENABLED(CONFIG_PPC32))
 #define is_tsk_32bit_task(tsk)	(IS_ENABLED(CONFIG_PPC32))
+#define clear_tsk_compat_task(tsk) do { } while (0)
 #endif
 
 #if defined(CONFIG_PPC64)

arch/powerpc/kernel/process.c

@@ -1738,10 +1738,7 @@ static void setup_ksp_vsid(struct task_struct *p, unsigned long sp)
  */
 int copy_thread(struct task_struct *p, const struct kernel_clone_args *args)
 {
-	unsigned long clone_flags = args->flags;
-	unsigned long usp = args->stack;
-	unsigned long tls = args->tls;
-	struct pt_regs *childregs, *kregs;
+	struct pt_regs *kregs; /* Switch frame regs */
 	extern void ret_from_fork(void);
 	extern void ret_from_fork_scv(void);
 	extern void ret_from_kernel_thread(void);
@@ -1754,45 +1751,76 @@ int copy_thread(struct task_struct *p, const struct kernel_clone_args *args)
 
 	klp_init_thread_info(p);
 
-	/* Create initial stack frame. */
-	sp -= STACK_USER_INT_FRAME_SIZE;
-	*(unsigned long *)(sp + STACK_INT_FRAME_MARKER) = STACK_FRAME_REGS_MARKER;
-
-	/* Copy registers */
-	childregs = (struct pt_regs *)(sp + STACK_INT_FRAME_REGS);
-	if (unlikely(args->fn)) {
+	if (unlikely(p->flags & PF_KTHREAD)) {
 		/* kernel thread */
+
+		/* Create initial minimum stack frame. */
+		sp -= STACK_FRAME_MIN_SIZE;
 		((unsigned long *)sp)[0] = 0;
-		memset(childregs, 0, sizeof(struct pt_regs));
-		childregs->gpr[1] = sp + STACK_USER_INT_FRAME_SIZE;
-#ifdef CONFIG_PPC64
-		clear_tsk_thread_flag(p, TIF_32BIT);
-		childregs->softe = IRQS_ENABLED;
-#endif
-		p->thread.regs = NULL;	/* no user register state */
-		ti->flags |= _TIF_RESTOREALL;
+
 		f = ret_from_kernel_thread;
+		p->thread.regs = NULL;	/* no user register state */
+		clear_tsk_compat_task(p);
 	} else {
 		/* user thread */
-		struct pt_regs *regs = current_pt_regs();
-		*childregs = *regs;
-		if (usp)
-			childregs->gpr[1] = usp;
-		((unsigned long *)sp)[0] = childregs->gpr[1];
-		p->thread.regs = childregs;
-		if (clone_flags & CLONE_SETTLS) {
-			if (!is_32bit_task())
-				childregs->gpr[13] = tls;
+		struct pt_regs *childregs;
+
+		/* Create initial user return stack frame. */
+		sp -= STACK_USER_INT_FRAME_SIZE;
+		*(unsigned long *)(sp + STACK_INT_FRAME_MARKER) = STACK_FRAME_REGS_MARKER;
+
+		childregs = (struct pt_regs *)(sp + STACK_INT_FRAME_REGS);
+
+		if (unlikely(args->fn)) {
+			/*
+			 * A user space thread, but it first runs a kernel
+			 * thread, and then returns as though it had called
+			 * execve rather than fork, so user regs will be
+			 * filled in (e.g., by kernel_execve()).
+			 */
+			((unsigned long *)sp)[0] = 0;
+			memset(childregs, 0, sizeof(struct pt_regs));
+#ifdef CONFIG_PPC64
+			childregs->softe = IRQS_ENABLED;
+#endif
+			ti->flags |= _TIF_RESTOREALL;
+			f = ret_from_kernel_thread;
+		} else {
+			struct pt_regs *regs = current_pt_regs();
+			unsigned long clone_flags = args->flags;
+			unsigned long usp = args->stack;
+
+			/* Copy registers */
+			*childregs = *regs;
+			if (usp)
+				childregs->gpr[1] = usp;
+			((unsigned long *)sp)[0] = childregs->gpr[1];
+#ifdef CONFIG_PPC_IRQ_SOFT_MASK_DEBUG
+			WARN_ON_ONCE(childregs->softe != IRQS_ENABLED);
+#endif
+			if (clone_flags & CLONE_SETTLS) {
+				unsigned long tls = args->tls;
+
+				if (!is_32bit_task())
+					childregs->gpr[13] = tls;
+				else
+					childregs->gpr[2] = tls;
+			}
+
+			if (trap_is_scv(regs))
+				f = ret_from_fork_scv;
 			else
-				childregs->gpr[2] = tls;
+				f = ret_from_fork;
 		}
 
-		if (trap_is_scv(regs))
-			f = ret_from_fork_scv;
-		else
-			f = ret_from_fork;
+#ifdef CONFIG_PPC64
+		if (cpu_has_feature(CPU_FTR_HAS_PPR))
+			childregs->ppr = DEFAULT_PPR;
+#endif
+
+		childregs->msr &= ~(MSR_FP|MSR_VEC|MSR_VSX);
+		p->thread.regs = childregs;
 	}
-	childregs->msr &= ~(MSR_FP|MSR_VEC|MSR_VSX);
 
 	/*
 	 * The way this works is that at some point in the future
@@ -1843,8 +1871,6 @@ int copy_thread(struct task_struct *p, const struct kernel_clone_args *args)
 		p->thread.dscr_inherit = current->thread.dscr_inherit;
 		p->thread.dscr = mfspr(SPRN_DSCR);
 	}
-	if (cpu_has_feature(CPU_FTR_HAS_PPR))
-		childregs->ppr = DEFAULT_PPR;
 
 	p->thread.tidr = 0;
 #endif