riscv: entry: Split ret_from_fork() into user and kernel

This function was unified into a single function in commit ab9164d
("riscv: entry: Consolidate ret_from_kernel_thread into ret_from_fork").
However that imposed a performance degradation.

Partially reverting this commit to have ret_from_fork() split again,
results in a 1% increase on the number of times fork is able to be called
per second.

Signed-off-by: Charlie Jenkins <charlie@rivosinc.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Acked-by: Alexandre Ghiti <alexghiti@rivosinc.com>
Link: https://lore.kernel.org/all/20250320-riscv_optimize_entry-v6-2-63e187e26041@rivosinc.com

Author: charlie-rivos

arch/riscv/include/asm/asm-prototypes.h

@@ -52,7 +52,8 @@ DECLARE_DO_ERROR_INFO(do_trap_ecall_s);
 DECLARE_DO_ERROR_INFO(do_trap_ecall_m);
 DECLARE_DO_ERROR_INFO(do_trap_break);
 
-asmlinkage void ret_from_fork(void *fn_arg, int (*fn)(void *), struct pt_regs *regs);
+asmlinkage void ret_from_fork_kernel(void *fn_arg, int (*fn)(void *), struct pt_regs *regs);
+asmlinkage void ret_from_fork_user(struct pt_regs *regs);
 asmlinkage void handle_bad_stack(struct pt_regs *regs);
 asmlinkage void do_page_fault(struct pt_regs *regs);
 asmlinkage void do_irq(struct pt_regs *regs);

arch/riscv/kernel/entry.S

@@ -319,14 +319,21 @@ SYM_CODE_END(handle_kernel_stack_overflow)
 ASM_NOKPROBE(handle_kernel_stack_overflow)
 #endif
 
-SYM_CODE_START(ret_from_fork_asm)
+SYM_CODE_START(ret_from_fork_kernel_asm)
 	call schedule_tail
 	move a0, s1 /* fn_arg */
 	move a1, s0 /* fn */
 	move a2, sp /* pt_regs */
-	call ret_from_fork
+	call ret_from_fork_kernel
 	j ret_from_exception
-SYM_CODE_END(ret_from_fork_asm)
+SYM_CODE_END(ret_from_fork_kernel_asm)
+
+SYM_CODE_START(ret_from_fork_user_asm)
+	call schedule_tail
+	move a0, sp /* pt_regs */
+	call ret_from_fork_user
+	j ret_from_exception
+SYM_CODE_END(ret_from_fork_user_asm)
 
 #ifdef CONFIG_IRQ_STACKS
 /*

arch/riscv/kernel/process.c

@@ -38,7 +38,8 @@ unsigned long __stack_chk_guard __read_mostly;
 EXPORT_SYMBOL(__stack_chk_guard);
 #endif
 
-extern asmlinkage void ret_from_fork_asm(void);
+extern asmlinkage void ret_from_fork_kernel_asm(void);
+extern asmlinkage void ret_from_fork_user_asm(void);
 
 void noinstr arch_cpu_idle(void)
 {
@@ -208,14 +209,18 @@ int arch_dup_task_struct(struct task_struct *dst, struct task_struct *src)
 	return 0;
 }
 
-asmlinkage void ret_from_fork(void *fn_arg, int (*fn)(void *), struct pt_regs *regs)
+asmlinkage void ret_from_fork_kernel(void *fn_arg, int (*fn)(void *), struct pt_regs *regs)
 {
-	if (unlikely(fn))
-		fn(fn_arg);
+	fn(fn_arg);
 
 	syscall_exit_to_user_mode(regs);
 }
 
+asmlinkage void ret_from_fork_user(struct pt_regs *regs)
+{
+	syscall_exit_to_user_mode(regs);
+}
+
 int copy_thread(struct task_struct *p, const struct kernel_clone_args *args)
 {
 	unsigned long clone_flags = args->flags;
@@ -238,6 +243,7 @@ int copy_thread(struct task_struct *p, const struct kernel_clone_args *args)
 
 		p->thread.s[0] = (unsigned long)args->fn;
 		p->thread.s[1] = (unsigned long)args->fn_arg;
+		p->thread.ra = (unsigned long)ret_from_fork_kernel_asm;
 	} else {
 		*childregs = *(current_pt_regs());
 		/* Turn off status.VS */
@@ -247,12 +253,11 @@ int copy_thread(struct task_struct *p, const struct kernel_clone_args *args)
 		if (clone_flags & CLONE_SETTLS)
 			childregs->tp = tls;
 		childregs->a0 = 0; /* Return value of fork() */
-		p->thread.s[0] = 0;
+		p->thread.ra = (unsigned long)ret_from_fork_user_asm;
 	}
 	p->thread.riscv_v_flags = 0;
 	if (has_vector() || has_xtheadvector())
 		riscv_v_thread_alloc(p);
-	p->thread.ra = (unsigned long)ret_from_fork_asm;
 	p->thread.sp = (unsigned long)childregs; /* kernel sp */
 	return 0;
 }