powerpc/64s: avoid reloading (H)SRR registers if they are still valid

When an interrupt is taken, the SRR registers are set to return to where
it left off. Unless they are modified in the meantime, or the return
address or MSR are modified, there is no need to reload these registers
when returning from interrupt.

Introduce per-CPU flags that track the validity of SRR and HSRR
registers. These are cleared when returning from interrupt, when
using the registers for something else (e.g., OPAL calls), when
adjusting the return address or MSR of a context, and when context
switching (which changes the return address and MSR).

This improves the performance of interrupt returns.

Signed-off-by: Nicholas Piggin <[e-mail]>
[mpe: Fold in fixup patch from Nick]
Signed-off-by: Michael Ellerman <[e-mail]>
Link: https://lore.kernel.org/r/[e-mail]

arch/powerpc/Kconfig.debug

···
 config PPC_IRQ_SOFT_MASK_DEBUG
 	bool "Include extra checks for powerpc irq soft masking"
 
+config PPC_RFI_SRR_DEBUG
+	bool "Include extra checks for RFI SRR register validity"
+	depends on PPC_BOOK3S_64
+
 config XMON
 	bool "Include xmon kernel debugger"
 	depends on DEBUG_KERNEL

arch/powerpc/include/asm/hw_irq.h

···
 	return !(regs->msr & MSR_EE);
 }
 
-static inline void may_hard_irq_enable(void) { }
+static inline bool may_hard_irq_enable(void)
+{
+	return false;
+}
+
+static inline void do_hard_irq_enable(void)
+{
+	BUILD_BUG();
+}
 
 static inline void irq_soft_mask_regs_set_state(struct pt_regs *regs, unsigned long val)
 {

arch/powerpc/include/asm/interrupt.h

···
 #include <asm/kprobes.h>
 #include <asm/runlatch.h>
 
+#ifdef CONFIG_PPC_BOOK3S_64
+static inline void srr_regs_clobbered(void)
+{
+	local_paca->srr_valid = 0;
+	local_paca->hsrr_valid = 0;
+}
+#else
+static inline void srr_regs_clobbered(void)
+{
+}
+#endif
+
 static inline void nap_adjust_return(struct pt_regs *regs)
 {
 #ifdef CONFIG_PPC_970_NAP
 	if (unlikely(test_thread_local_flags(_TLF_NAPPING))) {
 		/* Can avoid a test-and-clear because NMIs do not call this */
 		clear_thread_local_flags(_TLF_NAPPING);
-		regs->nip = (unsigned long)power4_idle_nap_return;
+		regs_set_return_ip(regs, (unsigned long)power4_idle_nap_return);
 	}
 #endif
 }

arch/powerpc/include/asm/livepatch.h

···
 {
 	struct pt_regs *regs = ftrace_get_regs(fregs);
 
-	regs->nip = ip;
+	regs_set_return_ip(regs, ip);
 }
 
 #define klp_get_ftrace_location klp_get_ftrace_location

arch/powerpc/include/asm/paca.h

···
 #ifdef CONFIG_PPC_BOOK3E
 	u16 trap_save;			/* Used when bad stack is encountered */
 #endif
+#ifdef CONFIG_PPC_BOOK3S_64
+	u8 hsrr_valid;			/* HSRRs set for HRFID */
+	u8 srr_valid;			/* SRRs set for RFID */
+#endif
 	u8 irq_soft_mask;		/* mask for irq soft masking */
 	u8 irq_happened;		/* irq happened while soft-disabled */
 	u8 irq_work_pending;		/* IRQ_WORK interrupt while soft-disable */

arch/powerpc/include/asm/probes.h

···
 /* Enable single stepping for the current task */
 static inline void enable_single_step(struct pt_regs *regs)
 {
-	regs->msr |= MSR_SINGLESTEP;
+	regs_set_return_msr(regs, regs->msr | MSR_SINGLESTEP);
 #ifdef CONFIG_PPC_ADV_DEBUG_REGS
 	/*
 	 * We turn off Critical Input Exception(CE) to ensure that the single
 	 * step will be for the instruction we have the probe on; if we don't,
 	 * it is possible we'd get the single step reported for CE.
 	 */
-	regs->msr &= ~MSR_CE;
+	regs_set_return_msr(regs, regs->msr & ~MSR_CE);
 	mtspr(SPRN_DBCR0, mfspr(SPRN_DBCR0) | DBCR0_IC | DBCR0_IDM);
 #ifdef CONFIG_PPC_47x
 	isync();

arch/powerpc/include/asm/ptrace.h

···
 #endif /* __powerpc64__ */
 
 #ifndef __ASSEMBLY__
+#include <asm/paca.h>
+
+#ifdef CONFIG_SMP
+extern unsigned long profile_pc(struct pt_regs *regs);
+#else
+#define profile_pc(regs) instruction_pointer(regs)
+#endif
+
+long do_syscall_trace_enter(struct pt_regs *regs);
+void do_syscall_trace_leave(struct pt_regs *regs);
+
+static inline void regs_set_return_ip(struct pt_regs *regs, unsigned long ip)
+{
+	regs->nip = ip;
+#ifdef CONFIG_PPC_BOOK3S_64
+	local_paca->hsrr_valid = 0;
+	local_paca->srr_valid = 0;
+#endif
+}
+
+static inline void regs_set_return_msr(struct pt_regs *regs, unsigned long msr)
+{
+	regs->msr = msr;
+#ifdef CONFIG_PPC_BOOK3S_64
+	local_paca->hsrr_valid = 0;
+	local_paca->srr_valid = 0;
+#endif
+}
+
+static inline void set_return_regs_changed(void)
+{
+#ifdef CONFIG_PPC_BOOK3S_64
+	local_paca->hsrr_valid = 0;
+	local_paca->srr_valid = 0;
+#endif
+}
+
+static inline void regs_add_return_ip(struct pt_regs *regs, long offset)
+{
+	regs_set_return_ip(regs, regs->nip + offset);
+}
 
 static inline unsigned long instruction_pointer(struct pt_regs *regs)
 {
···
 static inline void instruction_pointer_set(struct pt_regs *regs,
 		unsigned long val)
 {
-	regs->nip = val;
+	regs_set_return_ip(regs, val);
 }
 
 static inline unsigned long user_stack_pointer(struct pt_regs *regs)
···
 {
 	return 0;
 }
-
-#ifdef CONFIG_SMP
-extern unsigned long profile_pc(struct pt_regs *regs);
-#else
-#define profile_pc(regs) instruction_pointer(regs)
-#endif
-
-long do_syscall_trace_enter(struct pt_regs *regs);
-void do_syscall_trace_leave(struct pt_regs *regs);
 
 #ifdef __powerpc64__
 #define user_mode(regs) ((((regs)->msr) >> MSR_PR_LG) & 0x1)

arch/powerpc/kernel/asm-offsets.c

···
 	OFFSET(PACATOC, paca_struct, kernel_toc);
 	OFFSET(PACAKBASE, paca_struct, kernelbase);
 	OFFSET(PACAKMSR, paca_struct, kernel_msr);
+#ifdef CONFIG_PPC_BOOK3S_64
+	OFFSET(PACAHSRR_VALID, paca_struct, hsrr_valid);
+	OFFSET(PACASRR_VALID, paca_struct, srr_valid);
+#endif
 	OFFSET(PACAIRQSOFTMASK, paca_struct, irq_soft_mask);
 	OFFSET(PACAIRQHAPPENED, paca_struct, irq_happened);
 	OFFSET(PACA_FTRACE_ENABLED, paca_struct, ftrace_enabled);

arch/powerpc/kernel/entry_64.S

···
 	.section	".text"
 	.align 7
 
+.macro DEBUG_SRR_VALID srr
+#ifdef CONFIG_PPC_RFI_SRR_DEBUG
+	.ifc \srr,srr
+	mfspr	r11,SPRN_SRR0
+	ld	r12,_NIP(r1)
+100:	tdne	r11,r12
+	EMIT_BUG_ENTRY 100b,__FILE__,__LINE__,(BUGFLAG_WARNING | BUGFLAG_ONCE)
+	mfspr	r11,SPRN_SRR1
+	ld	r12,_MSR(r1)
+100:	tdne	r11,r12
+	EMIT_BUG_ENTRY 100b,__FILE__,__LINE__,(BUGFLAG_WARNING | BUGFLAG_ONCE)
+	.else
+	mfspr	r11,SPRN_HSRR0
+	ld	r12,_NIP(r1)
+100:	tdne	r11,r12
+	EMIT_BUG_ENTRY 100b,__FILE__,__LINE__,(BUGFLAG_WARNING | BUGFLAG_ONCE)
+	mfspr	r11,SPRN_HSRR1
+	ld	r12,_MSR(r1)
+100:	tdne	r11,r12
+	EMIT_BUG_ENTRY 100b,__FILE__,__LINE__,(BUGFLAG_WARNING | BUGFLAG_ONCE)
+	.endif
+#endif
+.endm
+
 #ifdef CONFIG_PPC_BOOK3S
 .macro system_call_vectored name trapnr
 	.globl system_call_vectored_\name
···
 	ld	r11,exception_marker@toc(r2)
 	std	r11,-16(r10)		/* "regshere" marker */
 
+#ifdef CONFIG_PPC_BOOK3S
+	li	r11,1
+	stb	r11,PACASRR_VALID(r13)
+#endif
+
 	/*
 	 * We always enter kernel from userspace with irq soft-mask enabled and
 	 * nothing pending. system_call_exception() will call
···
 	bl	syscall_exit_prepare
 
 	ld	r2,_CCR(r1)
+	ld	r6,_LINK(r1)
+	mtlr	r6
+
+#ifdef CONFIG_PPC_BOOK3S
+	lbz	r4,PACASRR_VALID(r13)
+	cmpdi	r4,0
+	bne	1f
+	li	r4,0
+	stb	r4,PACASRR_VALID(r13)
+#endif
 	ld	r4,_NIP(r1)
 	ld	r5,_MSR(r1)
-	ld	r6,_LINK(r1)
+	mtspr	SPRN_SRR0,r4
+	mtspr	SPRN_SRR1,r5
+1:
+	DEBUG_SRR_VALID srr
 
 BEGIN_FTR_SECTION
 	stdcx.	r0,0,r1			/* to clear the reservation */
 END_FTR_SECTION_IFCLR(CPU_FTR_STCX_CHECKS_ADDRESS)
-
-	mtspr	SPRN_SRR0,r4
-	mtspr	SPRN_SRR1,r5
-	mtlr	r6
 
 	cmpdi	r3,0
 	bne	.Lsyscall_restore_regs
···
 	kuap_user_restore r3, r4
 #endif
 .Lfast_user_interrupt_return_\srr\():
-	ld	r11,_NIP(r1)
-	ld	r12,_MSR(r1)
+
 BEGIN_FTR_SECTION
 	ld	r10,_PPR(r1)
 	mtspr	SPRN_PPR,r10
 END_FTR_SECTION_IFSET(CPU_FTR_HAS_PPR)
+
+#ifdef CONFIG_PPC_BOOK3S
+	.ifc \srr,srr
+	lbz	r4,PACASRR_VALID(r13)
+	.else
+	lbz	r4,PACAHSRR_VALID(r13)
+	.endif
+	cmpdi	r4,0
+	li	r4,0
+	bne	1f
+#endif
+	ld	r11,_NIP(r1)
+	ld	r12,_MSR(r1)
 	.ifc \srr,srr
 	mtspr	SPRN_SRR0,r11
 	mtspr	SPRN_SRR1,r12
+1:
+#ifdef CONFIG_PPC_BOOK3S
+	stb	r4,PACASRR_VALID(r13)
+#endif
 	.else
 	mtspr	SPRN_HSRR0,r11
 	mtspr	SPRN_HSRR1,r12
+1:
+#ifdef CONFIG_PPC_BOOK3S
+	stb	r4,PACAHSRR_VALID(r13)
+#endif
 	.endif
+	DEBUG_SRR_VALID \srr
 
 BEGIN_FTR_SECTION
 	stdcx.	r0,0,r1		/* to clear the reservation */
···
 
 .Lfast_kernel_interrupt_return_\srr\():
 	cmpdi	cr1,r3,0
+#ifdef CONFIG_PPC_BOOK3S
+	.ifc \srr,srr
+	lbz	r4,PACASRR_VALID(r13)
+	.else
+	lbz	r4,PACAHSRR_VALID(r13)
+	.endif
+	cmpdi	r4,0
+	li	r4,0
+	bne	1f
+#endif
 	ld	r11,_NIP(r1)
 	ld	r12,_MSR(r1)
 	.ifc \srr,srr
 	mtspr	SPRN_SRR0,r11
 	mtspr	SPRN_SRR1,r12
+1:
+#ifdef CONFIG_PPC_BOOK3S
+	stb	r4,PACASRR_VALID(r13)
+#endif
 	.else
 	mtspr	SPRN_HSRR0,r11
 	mtspr	SPRN_HSRR1,r12
+1:
+#ifdef CONFIG_PPC_BOOK3S
+	stb	r4,PACAHSRR_VALID(r13)
+#endif
 	.endif
+	DEBUG_SRR_VALID \srr
 
 BEGIN_FTR_SECTION
 	stdcx.	r0,0,r1		/* to clear the reservation */

arch/powerpc/kernel/exceptions-64s.S

···
 	std	r0,GPR0(r1)		/* save r0 in stackframe	*/
 	std	r10,GPR1(r1)		/* save r1 in stackframe	*/
 
+	/* Mark our [H]SRRs valid for return */
+	li	r10,1
+	.if IHSRR_IF_HVMODE
+	BEGIN_FTR_SECTION
+	stb	r10,PACAHSRR_VALID(r13)
+	FTR_SECTION_ELSE
+	stb	r10,PACASRR_VALID(r13)
+	ALT_FTR_SECTION_END_IFSET(CPU_FTR_HVMODE | CPU_FTR_ARCH_206)
+	.elseif IHSRR
+	stb	r10,PACAHSRR_VALID(r13)
+	.else
+	stb	r10,PACASRR_VALID(r13)
+	.endif
+
 	.if ISET_RI
 	li	r10,MSR_RI
 	mtmsrd	r10,1			/* Set MSR_RI */
···
 .macro EXCEPTION_RESTORE_REGS hsrr=0
 	/* Move original SRR0 and SRR1 into the respective regs */
 	ld	r9,_MSR(r1)
+	li	r10,0
 	.if \hsrr
 	mtspr	SPRN_HSRR1,r9
+	stb	r10,PACAHSRR_VALID(r13)
 	.else
 	mtspr	SPRN_SRR1,r9
+	stb	r10,PACASRR_VALID(r13)
 	.endif
 	ld	r9,_NIP(r1)
 	.if \hsrr
···
 	 *
 	 * Be careful to avoid touching the kernel stack.
 	 */
+	li	r10,0
+	stb	r10,PACAHSRR_VALID(r13)
 	ld	r10,PACA_EXGEN+EX_CTR(r13)
 	mtctr	r10
 	mtcrf	0x80,r9
···
 	ld	r10,PACA_EXGEN+EX_CFAR(r13)
 	mtspr	SPRN_CFAR,r10
 END_FTR_SECTION_IFSET(CPU_FTR_CFAR)
+	li	r10,0
+	stb	r10,PACAHSRR_VALID(r13)
 	ld	r10,PACA_EXGEN+EX_R10(r13)
 	ld	r11,PACA_EXGEN+EX_R11(r13)
 	ld	r12,PACA_EXGEN+EX_R12(r13)
···
 	ori	r11,r11,PACA_IRQ_HARD_DIS
 	stb	r11,PACAIRQHAPPENED(r13)
 2:	/* done */
+	li	r10,0
+	.if \hsrr
+	stb	r10,PACAHSRR_VALID(r13)
+	.else
+	stb	r10,PACASRR_VALID(r13)
+	.endif
 	ld	r10,PACA_EXGEN+EX_CTR(r13)
 	mtctr	r10
 	mtcrf	0x80,r9

arch/powerpc/kernel/fpu.S

···
 	ori	r12,r12,MSR_FP
 	or	r12,r12,r4
 	std	r12,_MSR(r1)
+#ifdef CONFIG_PPC_BOOK3S_64
+	li	r4,0
+	stb	r4,PACASRR_VALID(r13)
+#endif
 #endif
 	li	r4,1
 	stb	r4,THREAD_LOAD_FP(r5)

arch/powerpc/kernel/hw_breakpoint.c

···
 	return;
 
 reset:
-	regs->msr &= ~MSR_SE;
+	regs_set_return_msr(regs, regs->msr & ~MSR_SE);
 	for (i = 0; i < nr_wp_slots(); i++) {
 		info = counter_arch_bp(__this_cpu_read(bp_per_reg[i]));
 		__set_breakpoint(i, info);
···
 			current->thread.last_hit_ubp[i] = bp[i];
 			info[i] = NULL;
 		}
-		regs->msr |= MSR_SE;
+		regs_set_return_msr(regs, regs->msr | MSR_SE);
 		return false;
 	}
 

arch/powerpc/kernel/kgdb.c

···
 		return 0;
 
 	if (*(u32 *)regs->nip == BREAK_INSTR)
-		regs->nip += BREAK_INSTR_SIZE;
+		regs_add_return_ip(regs, BREAK_INSTR_SIZE);
 
 	return 1;
 }
···
 
 void kgdb_arch_set_pc(struct pt_regs *regs, unsigned long pc)
 {
-	regs->nip = pc;
+	regs_set_return_ip(regs, pc);
 }
 
 /*
···
 	case 'c':
 		/* handle the optional parameter */
 		if (kgdb_hex2long(&ptr, &addr))
-			linux_regs->nip = addr;
+			regs_set_return_ip(linux_regs, addr);
 
 		atomic_set(&kgdb_cpu_doing_single_step, -1);
 		/* set the trace bit if we're stepping */
···
 #ifdef CONFIG_PPC_ADV_DEBUG_REGS
 			mtspr(SPRN_DBCR0,
 			      mfspr(SPRN_DBCR0) | DBCR0_IC | DBCR0_IDM);
-			linux_regs->msr |= MSR_DE;
+			regs_set_return_msr(linux_regs, linux_regs->msr | MSR_DE);
 #else
-			linux_regs->msr |= MSR_SE;
+			regs_set_return_msr(linux_regs, linux_regs->msr | MSR_SE);
 #endif
 			atomic_set(&kgdb_cpu_doing_single_step,
 				   raw_smp_processor_id());

arch/powerpc/kernel/kprobes-ftrace.c

···
 		 * On powerpc, NIP is *before* this instruction for the
 		 * pre handler
 		 */
-		regs->nip -= MCOUNT_INSN_SIZE;
+		regs_add_return_ip(regs, -MCOUNT_INSN_SIZE);
 
 		__this_cpu_write(current_kprobe, p);
 		kcb->kprobe_status = KPROBE_HIT_ACTIVE;
···
 			 * Emulate singlestep (and also recover regs->nip)
 			 * as if there is a nop
 			 */
-			regs->nip += MCOUNT_INSN_SIZE;
+			regs_add_return_ip(regs, MCOUNT_INSN_SIZE);
 			if (unlikely(p->post_handler)) {
 				kcb->kprobe_status = KPROBE_HIT_SSDONE;
 				p->post_handler(p, regs, 0);

arch/powerpc/kernel/kprobes.c

···
 	 * variant as values in regs could play a part in
 	 * if the trap is taken or not
 	 */
-	regs->nip = (unsigned long)p->ainsn.insn;
+	regs_set_return_ip(regs, (unsigned long)p->ainsn.insn);
 }
 
 static nokprobe_inline void save_previous_kprobe(struct kprobe_ctlblk *kcb)
···
 		kprobe_opcode_t insn = *p->ainsn.insn;
 		if (kcb->kprobe_status == KPROBE_HIT_SS && is_trap(insn)) {
 			/* Turn off 'trace' bits */
-			regs->msr &= ~MSR_SINGLESTEP;
-			regs->msr |= kcb->kprobe_saved_msr;
+			regs_set_return_msr(regs,
+				(regs->msr & ~MSR_SINGLESTEP) |
+				kcb->kprobe_saved_msr);
 			goto no_kprobe;
 		}
 
···
 	 * we end up emulating it in kprobe_handler(), which increments the nip
 	 * again.
 	 */
-	regs->nip = orig_ret_address - 4;
+	regs_set_return_ip(regs, orig_ret_address - 4);
 	regs->link = orig_ret_address;
 
 	return 0;
···
 	}
 
 	/* Adjust nip to after the single-stepped instruction */
-	regs->nip = (unsigned long)cur->addr + len;
-	regs->msr |= kcb->kprobe_saved_msr;
+	regs_set_return_ip(regs, (unsigned long)cur->addr + len);
+	regs_set_return_msr(regs, regs->msr | kcb->kprobe_saved_msr);
 
 	/*Restore back the original saved kprobes variables and continue. */
 	if (kcb->kprobe_status == KPROBE_REENTER) {
···
 		 * and allow the page fault handler to continue as a
 		 * normal page fault.
 		 */
-		regs->nip = (unsigned long)cur->addr;
-		regs->msr &= ~MSR_SINGLESTEP; /* Turn off 'trace' bits */
-		regs->msr |= kcb->kprobe_saved_msr;
+		regs_set_return_ip(regs, (unsigned long)cur->addr);
+		/* Turn off 'trace' bits */
+		regs_set_return_msr(regs,
+			(regs->msr & ~MSR_SINGLESTEP) |
+			kcb->kprobe_saved_msr);
 		if (kcb->kprobe_status == KPROBE_REENTER)
 			restore_previous_kprobe(kcb);
 		else
···
 		 * zero, try to fix up.
 		 */
 		if ((entry = search_exception_tables(regs->nip)) != NULL) {
-			regs->nip = extable_fixup(entry);
+			regs_set_return_ip(regs, extable_fixup(entry));
 			return 1;
 		}
 

arch/powerpc/kernel/mce.c

···
 	entry = search_kernel_exception_table(regs->nip);
 	if (entry) {
 		mce_err->ignore_event = true;
-		regs->nip = extable_fixup(entry);
+		regs_set_return_ip(regs, extable_fixup(entry));
 	}
 }
 

arch/powerpc/kernel/optprobes.c

···
 		kprobes_inc_nmissed_count(&op->kp);
 	} else {
 		__this_cpu_write(current_kprobe, &op->kp);
-		regs->nip = (unsigned long)op->kp.addr;
+		regs_set_return_ip(regs, (unsigned long)op->kp.addr);
 		get_kprobe_ctlblk()->kprobe_status = KPROBE_HIT_ACTIVE;
 		opt_pre_handler(&op->kp, regs);
 		__this_cpu_write(current_kprobe, NULL);

arch/powerpc/kernel/process.c

···
 	if (tsk == current && tsk->thread.regs &&
 	    MSR_TM_ACTIVE(tsk->thread.regs->msr) &&
 	    !test_thread_flag(TIF_RESTORE_TM)) {
-		tsk->thread.ckpt_regs.msr = tsk->thread.regs->msr;
+		regs_set_return_msr(&tsk->thread.ckpt_regs,
+						tsk->thread.regs->msr);
 		set_thread_flag(TIF_RESTORE_TM);
 	}
 }
···
 	msr &= ~(MSR_FP|MSR_FE0|MSR_FE1);
 	if (cpu_has_feature(CPU_FTR_VSX))
 		msr &= ~MSR_VSX;
-	tsk->thread.regs->msr = msr;
+	regs_set_return_msr(tsk->thread.regs, msr);
 }
 
 void giveup_fpu(struct task_struct *tsk)
···
 	msr &= ~MSR_VEC;
 	if (cpu_has_feature(CPU_FTR_VSX))
 		msr &= ~MSR_VSX;
-	tsk->thread.regs->msr = msr;
+	regs_set_return_msr(tsk->thread.regs, msr);
 }
 
 void giveup_altivec(struct task_struct *tsk)
···
 
 		msr_check_and_clear(new_msr);
 
-		regs->msr |= new_msr | fpexc_mode;
+		regs_set_return_msr(regs, regs->msr | new_msr | fpexc_mode);
 	}
 }
 #endif /* CONFIG_PPC_BOOK3S_64 */
···
 #endif
 	restore_math(regs);
 
-	regs->msr |= msr_diff;
+	regs_set_return_msr(regs, regs->msr | msr_diff);
 }
 
 #else /* !CONFIG_PPC_TRANSACTIONAL_MEM */
···
 	}
 
 	/*
-	 * Call restore_sprs() before calling _switch(). If we move it after
-	 * _switch() then we miss out on calling it for new tasks. The reason
-	 * for this is we manually create a stack frame for new tasks that
-	 * directly returns through ret_from_fork() or
+	 * Call restore_sprs() and set_return_regs_changed() before calling
+	 * _switch(). If we move it after _switch() then we miss out on calling
+	 * it for new tasks. The reason for this is we manually create a stack
+	 * frame for new tasks that directly returns through ret_from_fork() or
 	 * ret_from_kernel_thread(). See copy_thread() for details.
 	 */
 	restore_sprs(old_thread, new_thread);
+
+	set_return_regs_changed(); /* _switch changes stack (and regs) */
 
 #ifdef CONFIG_PPC32
 	kuap_assert_locked();
···
 			}
 			regs->gpr[2] = toc;
 		}
-		regs->nip = entry;
-		regs->msr = MSR_USER64;
+		regs_set_return_ip(regs, entry);
+		regs_set_return_msr(regs, MSR_USER64);
 	} else {
-		regs->nip = start;
 		regs->gpr[2] = 0;
-		regs->msr = MSR_USER32;
+		regs_set_return_ip(regs, start);
+		regs_set_return_msr(regs, MSR_USER32);
 	}
+
 #endif
 #ifdef CONFIG_VSX
 	current->thread.used_vsr = 0;
···
 	current->thread.tm_tfiar = 0;
 	current->thread.load_tm = 0;
 #endif /* CONFIG_PPC_TRANSACTIONAL_MEM */
-
 }
 EXPORT_SYMBOL(start_thread);
 
···
 	if (val > PR_FP_EXC_PRECISE)
 		return -EINVAL;
 	tsk->thread.fpexc_mode = __pack_fe01(val);
-	if (regs != NULL && (regs->msr & MSR_FP) != 0)
-		regs->msr = (regs->msr & ~(MSR_FE0|MSR_FE1))
-			| tsk->thread.fpexc_mode;
+	if (regs != NULL && (regs->msr & MSR_FP) != 0) {
+		regs_set_return_msr(regs, (regs->msr & ~(MSR_FE0|MSR_FE1))
+						| tsk->thread.fpexc_mode);
+	}
 	return 0;
 }
 
···
 		return -EINVAL;
 
 	if (val == PR_ENDIAN_BIG)
-		regs->msr &= ~MSR_LE;
+		regs_set_return_msr(regs, regs->msr & ~MSR_LE);
 	else if (val == PR_ENDIAN_LITTLE || val == PR_ENDIAN_PPC_LITTLE)
-		regs->msr |= MSR_LE;
+		regs_set_return_msr(regs, regs->msr | MSR_LE);
 	else
 		return -EINVAL;
 

arch/powerpc/kernel/prom_init.c

···
 #include <asm/rtas.h>
 #include <asm/page.h>
 #include <asm/processor.h>
+#include <asm/interrupt.h>
 #include <asm/irq.h>
 #include <asm/io.h>
 #include <asm/smp.h>
···
 	asm volatile("sc 1\n" : "=r" (arg1) :
 			"r" (arg1),
 			"r" (arg2) :);
+	srr_regs_clobbered();
+
 	return arg1;
 }
 

arch/powerpc/kernel/ptrace/ptrace-adv.c

···
 	if (regs != NULL) {
 		task->thread.debug.dbcr0 &= ~DBCR0_BT;
 		task->thread.debug.dbcr0 |= DBCR0_IDM | DBCR0_IC;
-		regs->msr |= MSR_DE;
+		regs_set_return_msr(regs, regs->msr | MSR_DE);
 	}
 	set_tsk_thread_flag(task, TIF_SINGLESTEP);
 }
···
 	if (regs != NULL) {
 		task->thread.debug.dbcr0 &= ~DBCR0_IC;
 		task->thread.debug.dbcr0 = DBCR0_IDM | DBCR0_BT;
-		regs->msr |= MSR_DE;
+		regs_set_return_msr(regs, regs->msr | MSR_DE);
 	}
 	set_tsk_thread_flag(task, TIF_SINGLESTEP);
 }
···
 			 * All debug events were off.....
 			 */
 			task->thread.debug.dbcr0 &= ~DBCR0_IDM;
-			regs->msr &= ~MSR_DE;
+			regs_set_return_msr(regs, regs->msr & ~MSR_DE);
 		}
 	}
 	clear_tsk_thread_flag(task, TIF_SINGLESTEP);
···
 
 int ptrace_set_debugreg(struct task_struct *task, unsigned long addr, unsigned long data)
 {
+	struct pt_regs *regs = task->thread.regs;
 #ifdef CONFIG_HAVE_HW_BREAKPOINT
 	int ret;
 	struct thread_struct *thread = &task->thread;
···
 		dbcr_dac(task) &= ~(DBCR_DAC1R | DBCR_DAC1W);
 		if (!DBCR_ACTIVE_EVENTS(task->thread.debug.dbcr0,
 					task->thread.debug.dbcr1)) {
-			task->thread.regs->msr &= ~MSR_DE;
+			regs_set_return_msr(regs, regs->msr & ~MSR_DE);
 			task->thread.debug.dbcr0 &= ~DBCR0_IDM;
 		}
 		return 0;
···
 		dbcr_dac(task) |= DBCR_DAC1R;
 	if (data & 0x2UL)
 		dbcr_dac(task) |= DBCR_DAC1W;
-	task->thread.regs->msr |= MSR_DE;
+	regs_set_return_msr(regs, regs->msr | MSR_DE);
 	return 0;
 }
 
···
 	}
 out:
 	child->thread.debug.dbcr0 |= DBCR0_IDM;
-	child->thread.regs->msr |= MSR_DE;
+	regs_set_return_msr(child->thread.regs, child->thread.regs->msr | MSR_DE);
 
 	return slot;
 }
···
 		return -ENOSPC;
 	}
 	child->thread.debug.dbcr0 |= DBCR0_IDM;
-	child->thread.regs->msr |= MSR_DE;
+	regs_set_return_msr(child->thread.regs, child->thread.regs->msr | MSR_DE);
 
 	return slot + 4;
 }
···
 		child->thread.debug.dbcr2  |= DBCR2_DAC12MX;
 	else	/* PPC_BREAKPOINT_MODE_MASK */
 		child->thread.debug.dbcr2  |= DBCR2_DAC12MM;
-	child->thread.regs->msr |= MSR_DE;
+	regs_set_return_msr(child->thread.regs, child->thread.regs->msr | MSR_DE);
 
 	return 5;
 }
···
 		if (!DBCR_ACTIVE_EVENTS(child->thread.debug.dbcr0,
 					child->thread.debug.dbcr1)) {
 			child->thread.debug.dbcr0 &= ~DBCR0_IDM;
-			child->thread.regs->msr &= ~MSR_DE;
+			regs_set_return_msr(child->thread.regs,
+					child->thread.regs->msr & ~MSR_DE);
 		}
 	}
 	return rc;

arch/powerpc/kernel/ptrace/ptrace-noadv.c

···
 {
 	struct pt_regs *regs = task->thread.regs;
 
-	if (regs != NULL) {
-		regs->msr &= ~MSR_BE;
-		regs->msr |= MSR_SE;
-	}
+	if (regs != NULL)
+		regs_set_return_msr(regs, (regs->msr & ~MSR_BE) | MSR_SE);
 	set_tsk_thread_flag(task, TIF_SINGLESTEP);
 }
 
···
 {
 	struct pt_regs *regs = task->thread.regs;
 
-	if (regs != NULL) {
-		regs->msr &= ~MSR_SE;
-		regs->msr |= MSR_BE;
-	}
+	if (regs != NULL)
+		regs_set_return_msr(regs, (regs->msr & ~MSR_SE) | MSR_BE);
 	set_tsk_thread_flag(task, TIF_SINGLESTEP);
 }
 
···
 	struct pt_regs *regs = task->thread.regs;
 
 	if (regs != NULL)
-		regs->msr &= ~(MSR_SE | MSR_BE);
+		regs_set_return_msr(regs, regs->msr & ~(MSR_SE | MSR_BE));
 
 	clear_tsk_thread_flag(task, TIF_SINGLESTEP);
 }

arch/powerpc/kernel/ptrace/ptrace-view.c

···
 
 static __always_inline int set_user_msr(struct task_struct *task, unsigned long msr)
 {
-	task->thread.regs->msr &= ~MSR_DEBUGCHANGE;
-	task->thread.regs->msr |= msr & MSR_DEBUGCHANGE;
+	unsigned long newmsr = (task->thread.regs->msr & ~MSR_DEBUGCHANGE) |
+				(msr & MSR_DEBUGCHANGE);
+	regs_set_return_msr(task->thread.regs, newmsr);
 	return 0;
 }
 

arch/powerpc/kernel/rtas.c

···
 #include <linux/reboot.h>
 #include <linux/syscalls.h>
 
+#include <asm/interrupt.h>
 #include <asm/prom.h>
 #include <asm/rtas.h>
 #include <asm/hvcall.h>
···
 
 /* This is here deliberately so it's only used in this file */
 void enter_rtas(unsigned long);
+
+static inline void do_enter_rtas(unsigned long args)
+{
+	enter_rtas(args);
+
+	srr_regs_clobbered(); /* rtas uses SRRs, invalidate */
+}
 
 struct rtas_t rtas = {
 	.lock = __ARCH_SPIN_LOCK_UNLOCKED
···
 	save_args = rtas.args;
 	rtas.args = err_args;
 
-	enter_rtas(__pa(&rtas.args));
+	do_enter_rtas(__pa(&rtas.args));
 
 	err_args = rtas.args;
 	rtas.args = save_args;
···
 	for (i = 0; i < nret; ++i)
 		args->rets[i] = 0;
 
-	enter_rtas(__pa(args));
+	do_enter_rtas(__pa(args));
 }
 
 void rtas_call_unlocked(struct rtas_args *args, int token, int nargs, int nret, ...)
···
 	flags = lock_rtas();
 
 	rtas.args = args;
-	enter_rtas(__pa(&rtas.args));
+	do_enter_rtas(__pa(&rtas.args));
 	args = rtas.args;
 
 	/* A -1 return code indicates that the last command couldn't

arch/powerpc/kernel/signal.c

···
 			regs->gpr[0] = __NR_restart_syscall;
 		else
 			regs->gpr[3] = regs->orig_gpr3;
-		regs->nip -= 4;
+		regs_add_return_ip(regs, -4);
 		regs->result = 0;
 	} else {
 		if (trap_is_scv(regs)) {
···
 	 * For signals taken in non-TM or suspended mode, we use the
 	 * normal/non-checkpointed stack pointer.
 	 */
-
-	unsigned long ret = tsk->thread.regs->gpr[1];
+	struct pt_regs *regs = tsk->thread.regs;
+	unsigned long ret = regs->gpr[1];
 
 #ifdef CONFIG_PPC_TRANSACTIONAL_MEM
 	BUG_ON(tsk != current);
 
-	if (MSR_TM_ACTIVE(tsk->thread.regs->msr)) {
+	if (MSR_TM_ACTIVE(regs->msr)) {
 		preempt_disable();
 		tm_reclaim_current(TM_CAUSE_SIGNAL);
-		if (MSR_TM_TRANSACTIONAL(tsk->thread.regs->msr))
+		if (MSR_TM_TRANSACTIONAL(regs->msr))
 			ret = tsk->thread.ckpt_regs.gpr[1];
 
 		/*
···
 		 * (tm_recheckpoint_new_task() would recheckpoint). Besides, we
 		 * enter the signal handler in non-transactional state.
 		 */
-		tsk->thread.regs->msr &= ~MSR_TS_MASK;
+		regs_set_return_msr(regs, regs->msr & ~MSR_TS_MASK);
 		preempt_enable();
 	}
 #endif

arch/powerpc/kernel/signal_32.c

···
 
 	/* if doing signal return, restore the previous little-endian mode */
 	if (sig)
-		regs->msr = (regs->msr & ~MSR_LE) | (msr & MSR_LE);
+		regs_set_return_msr(regs, (regs->msr & ~MSR_LE) | (msr & MSR_LE));
 
 #ifdef CONFIG_ALTIVEC
 	/*
 	 * Force the process to reload the altivec registers from
 	 * current->thread when it next does altivec instructions
 	 */
-	regs->msr &= ~MSR_VEC;
+	regs_set_return_msr(regs, regs->msr & ~MSR_VEC);
 	if (msr & MSR_VEC) {
 		/* restore altivec registers from the stack */
 		unsafe_copy_from_user(&current->thread.vr_state, &sr->mc_vregs,
···
 	 * Force the process to reload the VSX registers from
 	 * current->thread when it next does VSX instruction.
 	 */
-	regs->msr &= ~MSR_VSX;
+	regs_set_return_msr(regs, regs->msr & ~MSR_VSX);
 	if (msr & MSR_VSX) {
 		/*
 		 * Restore altivec registers from the stack to a local
···
 	 * force the process to reload the FP registers from
 	 * current->thread when it next does FP instructions
 	 */
-	regs->msr &= ~(MSR_FP | MSR_FE0 | MSR_FE1);
+	regs_set_return_msr(regs, regs->msr & ~(MSR_FP | MSR_FE0 | MSR_FE1));
 
 #ifdef CONFIG_SPE
 	/* force the process to reload the spe registers from
 	   current->thread when it next does spe instructions */
-	regs->msr &= ~MSR_SPE;
+	regs_set_return_msr(regs, regs->msr & ~MSR_SPE);
 	if (msr & MSR_SPE) {
 		/* restore spe registers from the stack */
 		unsafe_copy_from_user(current->thread.evr, &sr->mc_vregs,
···
 	unsafe_get_user(msr, &sr->mc_gregs[PT_MSR], failed);
 
 	/* Restore the previous little-endian mode */
-	regs->msr = (regs->msr & ~MSR_LE) | (msr & MSR_LE);
+	regs_set_return_msr(regs, (regs->msr & ~MSR_LE) | (msr & MSR_LE));
 
-	regs->msr &= ~MSR_VEC;
+	regs_set_return_msr(regs, regs->msr & ~MSR_VEC);
 	if (msr & MSR_VEC) {
 		/* restore altivec registers from the stack */
 		unsafe_copy_from_user(&current->thread.ckvr_state, &sr->mc_vregs,
···
 	if (cpu_has_feature(CPU_FTR_ALTIVEC))
 		mtspr(SPRN_VRSAVE, current->thread.ckvrsave);
 
-	regs->msr &= ~(MSR_FP | MSR_FE0 | MSR_FE1);
+	regs_set_return_msr(regs, regs->msr & ~(MSR_FP | MSR_FE0 | MSR_FE1));
 
 	unsafe_copy_fpr_from_user(current, &sr->mc_fregs, failed);
 
-	regs->msr &= ~MSR_VSX;
+	regs_set_return_msr(regs, regs->msr & ~MSR_VSX);
 	if (msr & MSR_VSX) {
 		/*
 		 * Restore altivec registers from the stack to a local
···
 	 *
 	 * Pull in the MSR TM bits from the user context
 	 */
-	regs->msr = (regs->msr & ~MSR_TS_MASK) | (msr_hi & MSR_TS_MASK);
+	regs_set_return_msr(regs, (regs->msr & ~MSR_TS_MASK) | (msr_hi & MSR_TS_MASK));
 	/* Now, recheckpoint.  This loads up all of the checkpointed (older)
 	 * registers, including FP and V[S]Rs.  After recheckpointing, the
 	 * transactional versions should be loaded.
···
 	msr_check_and_set(msr & (MSR_FP | MSR_VEC));
 	if (msr & MSR_FP) {
 		load_fp_state(&current->thread.fp_state);
-		regs->msr |= (MSR_FP | current->thread.fpexc_mode);
+		regs_set_return_msr(regs, regs->msr | (MSR_FP | current->thread.fpexc_mode));
 	}
 	if (msr & MSR_VEC) {
 		load_vr_state(&current->thread.vr_state);
-		regs->msr |= MSR_VEC;
+		regs_set_return_msr(regs, regs->msr | MSR_VEC);
 	}
 
 	preempt_enable();
···
 	regs->gpr[4] = (unsigned long)&frame->info;
 	regs->gpr[5] = (unsigned long)&frame->uc;
 	regs->gpr[6] = (unsigned long)frame;
-	regs->nip = (unsigned long) ksig->ka.sa.sa_handler;
+	regs_set_return_ip(regs, (unsigned long) ksig->ka.sa.sa_handler);
 	/* enter the signal handler in native-endian mode */
-	regs->msr &= ~MSR_LE;
-	regs->msr |= (MSR_KERNEL & MSR_LE);
+	regs_set_return_msr(regs, (regs->msr & ~MSR_LE) | (MSR_KERNEL & MSR_LE));
+
 	return 0;
 
 failed:
···
 	regs->gpr[1] = newsp;
 	regs->gpr[3] = ksig->sig;
 	regs->gpr[4] = (unsigned long) sc;
-	regs->nip = (unsigned long)ksig->ka.sa.sa_handler;
+	regs_set_return_ip(regs, (unsigned long) ksig->ka.sa.sa_handler);
 	/* enter the signal handler in native-endian mode */
-	regs->msr &= ~MSR_LE;
-	regs->msr |= (MSR_KERNEL & MSR_LE);
+	regs_set_return_msr(regs, (regs->msr & ~MSR_LE) | (MSR_KERNEL & MSR_LE));
+
 	return 0;
 
 failed:
···
 		 * set, and recheckpoint was not called. This avoid
 		 * hitting a TM Bad thing at RFID
 		 */
-		regs->msr &= ~MSR_TS_MASK;
+		regs_set_return_msr(regs, regs->msr & ~MSR_TS_MASK);
 	}
 	/* Fall through, for non-TM restore */
 #endif
···
 	   affect the contents of these registers.  After this point,
 	   failure is a problem, anyway, and it's very unlikely unless
 	   the user is really doing something wrong. */
-	regs->msr = new_msr;
+	regs_set_return_msr(regs, new_msr);
 #ifdef CONFIG_PPC_ADV_DEBUG_REGS
 	current->thread.debug.dbcr0 = new_dbcr0;
 #endif

arch/powerpc/kernel/signal_64.c

···
 	/* get MSR separately, transfer the LE bit if doing signal return */
 	unsafe_get_user(msr, &sc->gp_regs[PT_MSR], efault_out);
 	if (sig)
-		regs->msr = (regs->msr & ~MSR_LE) | (msr & MSR_LE);
+		regs_set_return_msr(regs, (regs->msr & ~MSR_LE) | (msr & MSR_LE));
 	unsafe_get_user(regs->orig_gpr3, &sc->gp_regs[PT_ORIG_R3], efault_out);
 	unsafe_get_user(regs->ctr, &sc->gp_regs[PT_CTR], efault_out);
 	unsafe_get_user(regs->link, &sc->gp_regs[PT_LNK], efault_out);
···
 	 * This has to be done before copying stuff into tsk->thread.fpr/vr
 	 * for the reasons explained in the previous comment.
 	 */
-	regs->msr &= ~(MSR_FP | MSR_FE0 | MSR_FE1 | MSR_VEC | MSR_VSX);
+	regs_set_return_msr(regs, regs->msr & ~(MSR_FP | MSR_FE0 | MSR_FE1 | MSR_VEC | MSR_VSX));
 
 #ifdef CONFIG_ALTIVEC
 	unsafe_get_user(v_regs, &sc->v_regs, efault_out);
···
 		return -EINVAL;
 
 	/* pull in MSR LE from user context */
-	regs->msr = (regs->msr & ~MSR_LE) | (msr & MSR_LE);
+	regs_set_return_msr(regs, (regs->msr & ~MSR_LE) | (msr & MSR_LE));
 
 	/* The following non-GPR non-FPR non-VR state is also checkpointed: */
 	err |= __get_user(regs->ctr, &tm_sc->gp_regs[PT_CTR]);
···
 	 * This has to be done before copying stuff into tsk->thread.fpr/vr
 	 * for the reasons explained in the previous comment.
 	 */
-	regs->msr &= ~(MSR_FP | MSR_FE0 | MSR_FE1 | MSR_VEC | MSR_VSX);
+	regs_set_return_msr(regs, regs->msr & ~(MSR_FP | MSR_FE0 | MSR_FE1 | MSR_VEC | MSR_VSX));
 
 #ifdef CONFIG_ALTIVEC
 	err |= __get_user(v_regs, &sc->v_regs);
···
 	preempt_disable();
 
 	/* pull in MSR TS bits from user context */
-	regs->msr |= msr & MSR_TS_MASK;
+	regs_set_return_msr(regs, regs->msr | (msr & MSR_TS_MASK));
 
 	/*
 	 * Ensure that TM is enabled in regs->msr before we leave the signal
···
 	 * to be de-scheduled with MSR[TS] set but without calling
 	 * tm_recheckpoint(). This can cause a bug.
 	 */
-	regs->msr |= MSR_TM;
+	regs_set_return_msr(regs, regs->msr | MSR_TM);
 
 	/* This loads the checkpointed FP/VEC state, if used */
 	tm_recheckpoint(&tsk->thread);
···
 	msr_check_and_set(msr & (MSR_FP | MSR_VEC));
 	if (msr & MSR_FP) {
 		load_fp_state(&tsk->thread.fp_state);
-		regs->msr |= (MSR_FP | tsk->thread.fpexc_mode);
+		regs_set_return_msr(regs, regs->msr | (MSR_FP | tsk->thread.fpexc_mode));
 	}
 	if (msr & MSR_VEC) {
 		load_vr_state(&tsk->thread.vr_state);
-		regs->msr |= MSR_VEC;
+		regs_set_return_msr(regs, regs->msr | MSR_VEC);
 	}
 
 	preempt_enable();
···
 
 	/* This returns like rt_sigreturn */
 	set_thread_flag(TIF_RESTOREALL);
+
 	return 0;
 
 efault_out:
···
 		 * the MSR[TS] that came from user context later, at
 		 * restore_tm_sigcontexts.
 		 */
-		regs->msr &= ~MSR_TS_MASK;
+		regs_set_return_msr(regs, regs->msr & ~MSR_TS_MASK);
 
 		if (__get_user(msr, &uc->uc_mcontext.gp_regs[PT_MSR]))
 			goto badframe;
···
 		 * MSR[TS] set, but without CPU in the proper state,
 		 * causing a TM bad thing.
 		 */
-		current->thread.regs->msr &= ~MSR_TS_MASK;
+		regs_set_return_msr(current->thread.regs,
+				current->thread.regs->msr & ~MSR_TS_MASK);
 		if (!user_read_access_begin(&uc->uc_mcontext, sizeof(uc->uc_mcontext)))
 			goto badframe;
 
···
 		goto badframe;
 
 	set_thread_flag(TIF_RESTOREALL);
+
 	return 0;
 
 badframe_block:
···
 
 	/* Set up to return from userspace. */
 	if (tsk->mm->context.vdso) {
-		regs->nip = VDSO64_SYMBOL(tsk->mm->context.vdso, sigtramp_rt64);
+		regs_set_return_ip(regs, VDSO64_SYMBOL(tsk->mm->context.vdso, sigtramp_rt64));
 	} else {
 		err |= setup_trampoline(__NR_rt_sigreturn, &frame->tramp[0]);
 		if (err)
 			goto badframe;
-		regs->nip = (unsigned long) &frame->tramp[0];
+		regs_set_return_ip(regs, (unsigned long) &frame->tramp[0]);
 	}
 
 	/* Allocate a dummy caller frame for the signal handler. */
···
 	}
 
 	/* enter the signal handler in native-endian mode */
-	regs->msr &= ~MSR_LE;
-	regs->msr |= (MSR_KERNEL & MSR_LE);
+	regs_set_return_msr(regs, (regs->msr & ~MSR_LE) | (MSR_KERNEL & MSR_LE));
 	regs->gpr[1] = newsp;
 	regs->gpr[3] = ksig->sig;
 	regs->result = 0;

arch/powerpc/kernel/syscalls.c

···
 {
 	struct thread_info *ti;
 
-	current->thread.regs->msr ^= MSR_LE;
+	regs_set_return_msr(current->thread.regs,
+				current->thread.regs->msr ^ MSR_LE);
 
 	/*
 	 * Set TIF_RESTOREALL so that r3 isn't clobbered on return to

arch/powerpc/kernel/traps.c

···
 	return;
 
 nonrecoverable:
-	regs->msr &= ~MSR_RI;
+	regs_set_return_msr(regs, regs->msr & ~MSR_RI);
 #endif
 }
 DEFINE_INTERRUPT_HANDLER_NMI(system_reset_exception)
···
 			printk(KERN_DEBUG "%s bad port %lx at %p\n",
 			       (*nip & 0x100)? "OUT to": "IN from",
 			       regs->gpr[rb] - _IO_BASE, nip);
-			regs->msr |= MSR_RI;
-			regs->nip = extable_fixup(entry);
+			regs_set_return_msr(regs, regs->msr | MSR_RI);
+			regs_set_return_ip(regs, extable_fixup(entry));
 			return 1;
 		}
 	}
···
 #define REASON_BOUNDARY		SRR1_BOUNDARY
 
 #define single_stepping(regs)	((regs)->msr & MSR_SE)
-#define clear_single_step(regs)	((regs)->msr &= ~MSR_SE)
-#define clear_br_trace(regs)	((regs)->msr &= ~MSR_BE)
+#define clear_single_step(regs)	(regs_set_return_msr((regs), (regs)->msr & ~MSR_SE))
+#define clear_br_trace(regs)	(regs_set_return_msr((regs), (regs)->msr & ~MSR_BE))
 #endif
 
 #define inst_length(reason)	(((reason) & REASON_PREFIXED) ? 8 : 4)
···
 #endif /* !__LITTLE_ENDIAN__ */
 
 	/* Go to next instruction */
-	regs->nip += 4;
+	regs_add_return_ip(regs, 4);
 }
 #endif /* CONFIG_VSX */
 
···
 
 		if (!(regs->msr & MSR_PR) &&  /* not user-mode */
 		    report_bug(bugaddr, regs) == BUG_TRAP_TYPE_WARN) {
-			regs->nip += 4;
+			regs_add_return_ip(regs, 4);
 			return;
 		}
 		_exception(SIGTRAP, regs, TRAP_BRKPT, regs->nip);
···
 	if (reason & (REASON_ILLEGAL | REASON_PRIVILEGED)) {
 		switch (emulate_instruction(regs)) {
 		case 0:
-			regs->nip += 4;
+			regs_add_return_ip(regs, 4);
 			emulate_single_step(regs);
 			return;
 		case -EFAULT:
···
  */
 DEFINE_INTERRUPT_HANDLER(emulation_assist_interrupt)
 {
-	regs->msr |= REASON_ILLEGAL;
+	regs_set_return_msr(regs, regs->msr | REASON_ILLEGAL);
 	do_program_check(regs);
 }
 
···
 
 	if (fixed == 1) {
 		/* skip over emulated instruction */
-		regs->nip += inst_length(reason);
+		regs_add_return_ip(regs, inst_length(reason));
 		emulate_single_step(regs);
 		return;
 	}
···
 #ifdef CONFIG_PPC_TRANSACTIONAL_MEM
 	if (user_mode(regs)) {
 		current->thread.load_tm++;
-		regs->msr |= MSR_TM;
+		regs_set_return_msr(regs, regs->msr | MSR_TM);
 		tm_enable();
 		tm_restore_sprs(&current->thread);
 		return;
···
 				pr_err("DSCR based mfspr emulation failed\n");
 				return;
 			}
-			regs->nip += 4;
+			regs_add_return_ip(regs, 4);
 			emulate_single_step(regs);
 		}
 		return;
···
 	 */
 	if (DBCR_ACTIVE_EVENTS(current->thread.debug.dbcr0,
 			       current->thread.debug.dbcr1))
-		regs->msr |= MSR_DE;
+		regs_set_return_msr(regs, regs->msr | MSR_DE);
 	else
 		/* Make sure the IDM flag is off */
 		current->thread.debug.dbcr0 &= ~DBCR0_IDM;
···
 	 * instead of stopping here when hitting a BT
 	 */
 	if (debug_status & DBSR_BT) {
-		regs->msr &= ~MSR_DE;
+		regs_set_return_msr(regs, regs->msr & ~MSR_DE);
 
 		/* Disable BT */
 		mtspr(SPRN_DBCR0, mfspr(SPRN_DBCR0) & ~DBCR0_BT);
···
 		if (user_mode(regs)) {
 			current->thread.debug.dbcr0 &= ~DBCR0_BT;
 			current->thread.debug.dbcr0 |= DBCR0_IDM | DBCR0_IC;
-			regs->msr |= MSR_DE;
+			regs_set_return_msr(regs, regs->msr | MSR_DE);
 			return;
 		}
 
···
 		if (debugger_sstep(regs))
 			return;
 	} else if (debug_status & DBSR_IC) { 	/* Instruction complete */
-		regs->msr &= ~MSR_DE;
+		regs_set_return_msr(regs, regs->msr & ~MSR_DE);
 
 		/* Disable instruction completion */
 		mtspr(SPRN_DBCR0, mfspr(SPRN_DBCR0) & ~DBCR0_IC);
···
 			current->thread.debug.dbcr0 &= ~DBCR0_IC;
 			if (DBCR_ACTIVE_EVENTS(current->thread.debug.dbcr0,
 					       current->thread.debug.dbcr1))
-				regs->msr |= MSR_DE;
+				regs_set_return_msr(regs, regs->msr | MSR_DE);
 			else
 				/* Make sure the IDM bit is off */
 				current->thread.debug.dbcr0 &= ~DBCR0_IDM;
···
 	PPC_WARN_EMULATED(altivec, regs);
 	err = emulate_altivec(regs);
 	if (err == 0) {
-		regs->nip += 4;		/* skip emulated instruction */
+		regs_add_return_ip(regs, 4); /* skip emulated instruction */
 		emulate_single_step(regs);
 		return;
 	}
···
 
 	err = do_spe_mathemu(regs);
 	if (err == 0) {
-		regs->nip += 4;		/* skip emulated instruction */
+		regs_add_return_ip(regs, 4); /* skip emulated instruction */
 		emulate_single_step(regs);
 		return;
 	}
···
 		giveup_spe(current);
 	preempt_enable();
 
-	regs->nip -= 4;
+	regs_add_return_ip(regs, -4);
 	err = speround_handler(regs);
 	if (err == 0) {
-		regs->nip += 4;		/* skip emulated instruction */
+		regs_add_return_ip(regs, 4); /* skip emulated instruction */
 		emulate_single_step(regs);
 		return;
 	}

arch/powerpc/kernel/uprobes.c

···
 
 	autask->saved_trap_nr = current->thread.trap_nr;
 	current->thread.trap_nr = UPROBE_TRAP_NR;
-	regs->nip = current->utask->xol_vaddr;
+	regs_set_return_ip(regs, current->utask->xol_vaddr);
 
 	user_enable_single_step(current);
 	return 0;
···
 	 * support doesn't exist and have to fix-up the next instruction
 	 * to be executed.
 	 */
-	regs->nip = (unsigned long)ppc_inst_next((void *)utask->vaddr, auprobe->insn);
+	regs_set_return_ip(regs, (unsigned long)ppc_inst_next((void *)utask->vaddr, auprobe->insn));
 
 	user_disable_single_step(current);
 	return 0;

arch/powerpc/kernel/vector.S

···
 	addi	r5,r4,THREAD		/* Get THREAD */
 	oris	r12,r12,MSR_VEC@h
 	std	r12,_MSR(r1)
+#ifdef CONFIG_PPC_BOOK3S_64
+	li	r4,0
+	stb	r4,PACASRR_VALID(r13)
+#endif
 #endif
 	li	r4,1
 	stb	r4,THREAD_LOAD_VEC(r5)
···
 	/* enable use of VSX after return */
 	oris	r12,r12,MSR_VSX@h
 	std	r12,_MSR(r1)
+	li	r4,0
+	stb	r4,PACASRR_VALID(r13)
 	b	fast_interrupt_return_srr
 
 #endif /* CONFIG_VSX */

arch/powerpc/kvm/book3s_hv.c

···
 
 	vcpu->arch.state = KVMPPC_VCPU_NOTREADY;
 	atomic_dec(&kvm->arch.vcpus_running);
+
+	srr_regs_clobbered();
+
 	return r;
 }
 

arch/powerpc/kvm/book3s_pr.c

···
 #include <asm/cputable.h>
 #include <asm/cacheflush.h>
 #include <linux/uaccess.h>
+#include <asm/interrupt.h>
 #include <asm/io.h>
 #include <asm/kvm_ppc.h>
 #include <asm/kvm_book3s.h>
···
 	/* Make sure we save the guest TAR/EBB/DSCR state */
 	kvmppc_giveup_fac(vcpu, FSCR_TAR_LG);
 
+	srr_regs_clobbered();
 out:
 	vcpu->mode = OUTSIDE_GUEST_MODE;
 	return ret;

arch/powerpc/lib/error-inject.c

···
 	 * function in the kernel/module, captured on a kprobe. We don't need
 	 * to worry about 32-bit userspace on a 64-bit kernel.
 	 */
-	regs->nip = regs->link;
+	regs_set_return_ip(regs, regs->link);
 }
 NOKPROBE_SYMBOL(override_function_with_return);

arch/powerpc/lib/sstep.c

···
 	default:
 		WARN_ON_ONCE(1);
 	}
-	regs->nip = next_pc;
+	regs_set_return_ip(regs, next_pc);
 }
 NOKPROBE_SYMBOL(emulate_update_regs);
 
···
 			/* can't step mtmsr[d] that would clear MSR_RI */
 			return -1;
 		/* here op.val is the mask of bits to change */
-		regs->msr = (regs->msr & ~op.val) | (val & op.val);
+		regs_set_return_msr(regs, (regs->msr & ~op.val) | (val & op.val));
 		goto instr_done;
 
 #ifdef CONFIG_PPC64
···
 		if (IS_ENABLED(CONFIG_PPC_FAST_ENDIAN_SWITCH) &&
 				cpu_has_feature(CPU_FTR_REAL_LE) &&
 				regs->gpr[0] == 0x1ebe) {
-			regs->msr ^= MSR_LE;
+			regs_set_return_msr(regs, regs->msr ^ MSR_LE);
 			goto instr_done;
 		}
 		regs->gpr[9] = regs->gpr[13];
···
 		regs->gpr[11] = regs->nip + 4;
 		regs->gpr[12] = regs->msr & MSR_MASK;
 		regs->gpr[13] = (unsigned long) get_paca();
-		regs->nip = (unsigned long) &system_call_common;
-		regs->msr = MSR_KERNEL;
+		regs_set_return_ip(regs, (unsigned long) &system_call_common);
+		regs_set_return_msr(regs, MSR_KERNEL);
 		return 1;
 
 #ifdef CONFIG_PPC_BOOK3S_64
···
 		regs->gpr[11] = regs->nip + 4;
 		regs->gpr[12] = regs->msr & MSR_MASK;
 		regs->gpr[13] = (unsigned long) get_paca();
-		regs->nip = (unsigned long) &system_call_vectored_emulate;
-		regs->msr = MSR_KERNEL;
+		regs_set_return_ip(regs, (unsigned long) &system_call_vectored_emulate);
+		regs_set_return_msr(regs, MSR_KERNEL);
 		return 1;
 #endif
 
···
 	return 0;
 
  instr_done:
-	regs->nip = truncate_if_32bit(regs->msr, regs->nip + GETLENGTH(op.type));
+	regs_set_return_ip(regs,
+		truncate_if_32bit(regs->msr, regs->nip + GETLENGTH(op.type)));
 	return 1;
 }
 NOKPROBE_SYMBOL(emulate_step);

arch/powerpc/lib/test_emulate_step.c

···
 	if (!regs || !ppc_inst_val(instr))
 		return -EINVAL;
 
+	/* This is not a return frame regs */
 	regs->nip = patch_site_addr(&patch__exec_instr);
 
 	analysed = analyse_instr(&op, regs, instr);

arch/powerpc/math-emu/math.c

···
 		break;
 	}
 
-	regs->nip += 4;
+	regs_add_return_ip(regs, 4);
 	return 0;
 
 illegal:

arch/powerpc/math-emu/math_efp.c

···
 illegal:
 	if (have_e500_cpu_a005_erratum) {
 		/* according to e500 cpu a005 erratum, reissue efp inst */
-		regs->nip -= 4;
+		regs_add_return_ip(regs, -4);
 		pr_debug("re-issue efp inst: %08lx\n", speinsn);
 		return 0;
 	}

arch/powerpc/platforms/embedded6xx/holly.c

···
 	/* Are we prepared to handle this fault */
 	if ((entry = search_exception_tables(regs->nip)) != NULL) {
 		tsi108_clear_pci_cfg_error();
-		regs->msr |= MSR_RI;
-		regs->nip = extable_fixup(entry);
+		regs_set_return_msr(regs, regs->msr | MSR_RI);
+		regs_set_return_ip(regs, extable_fixup(entry));
 		return 1;
 	}
 	return 0;

arch/powerpc/platforms/embedded6xx/mpc7448_hpc2.c

···
 	/* Are we prepared to handle this fault */
 	if ((entry = search_exception_tables(regs->nip)) != NULL) {
 		tsi108_clear_pci_cfg_error();
-		regs->msr |= MSR_RI;
-		regs->nip = extable_fixup(entry);
+		regs_set_return_msr(regs, regs->msr | MSR_RI);
+		regs_set_return_ip(regs, extable_fixup(entry));
 		return 1;
 	}
 	return 0;

arch/powerpc/platforms/pasemi/idle.c

···
 	 */
 
 	if (regs->msr & SRR1_WAKEMASK)
-		regs->nip = regs->link;
+		regs_set_return_ip(regs, regs->link);
 
 	switch (regs->msr & SRR1_WAKEMASK) {
 	case SRR1_WAKEDEC:
···
 	restore_astate(hard_smp_processor_id());
 
 	/* everything handled */
-	regs->msr |= MSR_RI;
+	regs_set_return_msr(regs, regs->msr | MSR_RI);
 	return 1;
 }
 

arch/powerpc/platforms/powernv/opal-call.c

···
 // SPDX-License-Identifier: GPL-2.0
 #include <linux/percpu.h>
 #include <linux/jump_label.h>
+#include <asm/interrupt.h>
 #include <asm/opal-api.h>
 #include <asm/trace.h>
 #include <asm/asm-prototypes.h>
···
 	unsigned long msr = mfmsr();
 	bool mmu = (msr & (MSR_IR|MSR_DR));
 	int64_t ret;
+
+	/* OPAL call / firmware may use SRR and/or HSRR */
+	srr_regs_clobbered();
 
 	msr &= ~MSR_EE;
 

arch/powerpc/platforms/powernv/opal.c

···
 	 * Setup regs->nip to rfi into fixup address.
 	 */
 	if (recover_addr)
-		regs->nip = recover_addr;
+		regs_set_return_ip(regs, recover_addr);
 
 out:
 	return !!recover_addr;

arch/powerpc/platforms/pseries/hvCall.S

···
 	mfcr	r0
 	stw	r0,8(r1)
 	HVSC				/* invoke the hypervisor */
+
+	li	r4,0
+	stb	r4,PACASRR_VALID(r13)
+
 	lwz	r0,8(r1)
 	mtcrf	0xff,r0
 	blr				/* return r3 = status */
···
 	HCALL_BRANCH(plpar_hcall_norets_trace)
 	HVSC				/* invoke the hypervisor */
 
+	li	r4,0
+	stb	r4,PACASRR_VALID(r13)
+
 	lwz	r0,8(r1)
 	mtcrf	0xff,r0
 	blr				/* return r3 = status */
···
 	HCALL_INST_PRECALL(R4)
 	HVSC
 	HCALL_INST_POSTCALL_NORETS
+
+	li	r4,0
+	stb	r4,PACASRR_VALID(r13)
+
 	lwz	r0,8(r1)
 	mtcrf	0xff,r0
 	blr
···
 	std	r6, 16(r12)
 	std	r7, 24(r12)
 
+	li	r4,0
+	stb	r4,PACASRR_VALID(r13)
+
 	lwz	r0,8(r1)
 	mtcrf	0xff,r0
 
···
 	std	r7,24(r12)
 
 	HCALL_INST_POSTCALL(r12)
+
+	li	r4,0
+	stb	r4,PACASRR_VALID(r13)
 
 	lwz	r0,8(r1)
 	mtcrf	0xff,r0
···
 	std	r5,  8(r12)
 	std	r6, 16(r12)
 	std	r7, 24(r12)
+
+	li	r4,0
+	stb	r4,PACASRR_VALID(r13)
 
 	lwz	r0,8(r1)
 	mtcrf	0xff,r0
···
 	std	r11,56(r12)
 	std	r0, 64(r12)
 
+	li	r4,0
+	stb	r4,PACASRR_VALID(r13)
+
 	lwz	r0,8(r1)
 	mtcrf	0xff,r0
 
···
 	std	r0,64(r12)
 
 	HCALL_INST_POSTCALL(r12)
+
+	li	r4,0
+	stb	r4,PACASRR_VALID(r13)
 
 	lwz	r0,8(r1)
 	mtcrf	0xff,r0
···
 	std	r10,48(r12)
 	std	r11,56(r12)
 	std	r0, 64(r12)
+
+	li	r4,0
+	stb	r4,PACASRR_VALID(r13)
 
 	lwz	r0,8(r1)
 	mtcrf	0xff,r0

arch/powerpc/platforms/pseries/ras.c

···
 	if ((be64_to_cpu(regs->msr) &
 			(MSR_LE|MSR_RI|MSR_DR|MSR_IR|MSR_ME|MSR_PR|
 			 MSR_ILE|MSR_HV|MSR_SF)) == (MSR_DR|MSR_SF)) {
-		regs->nip = be64_to_cpu((__be64)regs->nip);
-		regs->msr = 0;
+		regs_set_return_ip(regs, be64_to_cpu((__be64)regs->nip));
+		regs_set_return_msr(regs, 0);
 	}
 #endif
 

arch/powerpc/sysdev/fsl_pci.c

···
 			ret = get_kernel_nofault(inst, (void *)regs->nip);
 
 		if (!ret && mcheck_handle_load(regs, inst)) {
-			regs->nip += 4;
+			regs_add_return_ip(regs, 4);
 			return 1;
 		}
 	}

arch/powerpc/sysdev/fsl_rio.c

···
 				 __func__);
 			out_be32((u32 *)(rio_regs_win + RIO_LTLEDCSR),
 				 0);
-			regs->msr |= MSR_RI;
-			regs->nip = extable_fixup(entry);
+			regs_set_return_msr(regs, regs->msr | MSR_RI);
+			regs_set_return_ip(regs, extable_fixup(entry));
 			return 1;
 		}
 	}

arch/powerpc/xmon/xmon.c

···
 
 	bp = in_breakpoint_table(regs->nip, &offset);
 	if (bp != NULL) {
-		regs->nip = bp->address + offset;
+		regs_set_return_ip(regs, bp->address + offset);
 		atomic_dec(&bp->ref_count);
 	}
 
···
 	if (regs->msr & MSR_DE) {
 		bp = at_breakpoint(regs->nip);
 		if (bp != NULL) {
-			regs->nip = (unsigned long) &bp->instr[0];
+			regs_set_return_ip(regs, (unsigned long) &bp->instr[0]);
 			atomic_inc(&bp->ref_count);
 		}
 	}
···
 		if (bp != NULL) {
 			int stepped = emulate_step(regs, ppc_inst_read(bp->instr));
 			if (stepped == 0) {
-				regs->nip = (unsigned long) &bp->instr[0];
+				regs_set_return_ip(regs, (unsigned long) &bp->instr[0]);
 				atomic_inc(&bp->ref_count);
 			} else if (stepped < 0) {
 				printf("Couldn't single-step %s instruction\n",
···
 	/* Are we at the trap at bp->instr[1] for some bp? */
 	bp = in_breakpoint_table(regs->nip, &offset);
 	if (bp != NULL && (offset == 4 || offset == 8)) {
-		regs->nip = bp->address + offset;
+		regs_set_return_ip(regs, bp->address + offset);
 		atomic_dec(&bp->ref_count);
 		return 1;
 	}
···
 	if ((regs->msr & (MSR_IR|MSR_PR|MSR_64BIT)) == (MSR_IR|MSR_64BIT)) {
 		bp = in_breakpoint_table(regs->nip, &offset);
 		if (bp != NULL) {
-			regs->nip = bp->address + offset;
+			regs_set_return_ip(regs, bp->address + offset);
 			atomic_dec(&bp->ref_count);
 		}
 	}
···
 #ifdef CONFIG_BOOKE
 static int do_step(struct pt_regs *regs)
 {
-	regs->msr |= MSR_DE;
+	regs_set_return_msr(regs, regs->msr | MSR_DE);
 	mtspr(SPRN_DBCR0, mfspr(SPRN_DBCR0) | DBCR0_IC | DBCR0_IDM);
 	return 1;
 }
···
 			}
 		}
 	}
-	regs->msr |= MSR_SE;
+	regs_set_return_msr(regs, regs->msr | MSR_SE);
 	return 1;
 }
 #endif