[PATCH 2/4] powerpc/64s: system reset interrupt preserve HSRRs

Nicholas Piggin npiggin at gmail.com
Tue Jan 22 17:46:16 AEDT 2019


The system reset interrupt may use HSRR registers (e.g., to call in to
OPAL), but code that uses HSRR registers is not required to clear
MSR[RI] by convention.

Rather than introduce that requirement, have system reset interrupt
save HSRRs before they might be used.

Signed-off-by: Nicholas Piggin <npiggin at gmail.com>
---
 arch/powerpc/kernel/traps.c | 25 ++++++++++++++++++++++++-
 1 file changed, 24 insertions(+), 1 deletion(-)

diff --git a/arch/powerpc/kernel/traps.c b/arch/powerpc/kernel/traps.c
index b429b2264a1f..d54459152a2b 100644
--- a/arch/powerpc/kernel/traps.c
+++ b/arch/powerpc/kernel/traps.c
@@ -438,14 +438,32 @@ void hv_nmi_check_nonrecoverable(struct pt_regs *regs)
 
 void system_reset_exception(struct pt_regs *regs)
 {
+	unsigned long hsrr0, hsrr1;
+	bool hsrrs_saved = false;
+	bool nested = in_nmi();
+
 	/*
 	 * Avoid crashes in case of nested NMI exceptions. Recoverability
 	 * is determined by RI and in_nmi
 	 */
-	bool nested = in_nmi();
 	if (!nested)
 		nmi_enter();
 
+	/*
+	 * System reset can interrupt a region where HSRRs are live and
+	 * MSR[RI]=1, and it may clobber HSRRs itself (e.g., to call OPAL),
+	 * so save them before doing anything.
+	 *
+	 * Machine checks should be okay to avoid this, as the real mode
+	 * handler is careful to avoid HSRRs, and the virt code is not
+	 * delivered as an NMI.
+	 */
+	if (cpu_has_feature(CPU_FTR_HVMODE)) {
+		hsrrs_saved = true;
+		hsrr0 = mfspr(SPRN_HSRR0);
+		hsrr1 = mfspr(SPRN_HSRR1);
+	}
+
 	hv_nmi_check_nonrecoverable(regs);
 
 	__this_cpu_inc(irq_stat.sreset_irqs);
@@ -495,6 +513,11 @@ void system_reset_exception(struct pt_regs *regs)
 	if (!(regs->msr & MSR_RI))
 		nmi_panic(regs, "Unrecoverable System Reset");
 
+	if (hsrrs_saved) {
+		mtspr(SPRN_HSRR0, hsrr0);
+		mtspr(SPRN_HSRR1, hsrr1);
+	}
+
 	if (!nested)
 		nmi_exit();
 
-- 
2.18.0



More information about the Linuxppc-dev mailing list