s390/nmi: s390: New low level handling for machine check happening in guest

Add the logic to check if the machine check happens when the guest is
running. If yes, set the exit reason -EINTR in the machine check's
interrupt handler. Refactor s390_do_machine_check to avoid panicing
the host for some kinds of machine checks which happen
when guest is running.
Reinject the instruction processing damage's machine checks including
Delayed Access Exception instead of damaging the host if it happens
in the guest because it could be caused by improper update on TLB entry
or other software case and impacts the guest only.

Signed-off-by: QingFeng Hao <haoqf@linux.vnet.ibm.com>
Acked-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Acked-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Acked-by: Christian Borntraeger <borntraeger@de.ibm.com>
Signed-off-by: Christian Borntraeger <borntraeger@de.ibm.com>

1 files changed, 40 insertions, 10 deletions

diff --git a/arch/s390/kernel/nmi.c b/arch/s390/kernel/nmi.c
index 985589523970..958cc3352faa 100644
--- a/arch/s390/kernel/nmi.c
+++ b/arch/s390/kernel/nmi.c
@@ -25,6 +25,7 @@
 #include <asm/crw.h>
 #include <asm/switch_to.h>
 #include <asm/ctl_reg.h>
+#include <asm/asm-offsets.h>
 
 struct mcck_struct {
 	unsigned int kill_task : 1;
@@ -280,6 +281,8 @@ static int notrace s390_validate_registers(union mci mci, int umode)
 #define ED_STP_ISLAND	6	/* External damage STP island check */
 #define ED_STP_SYNC	7	/* External damage STP sync check */
 
+#define MCCK_CODE_NO_GUEST	(MCCK_CODE_CP | MCCK_CODE_EXT_DAMAGE)
+
 /*
  * machine check handler.
  */
@@ -291,6 +294,7 @@ void notrace s390_do_machine_check(struct pt_regs *regs)
 	struct mcck_struct *mcck;
 	unsigned long long tmp;
 	union mci mci;
+	unsigned long mcck_dam_code;
 
 	nmi_enter();
 	inc_irq_stat(NMI_NMI);
@@ -301,7 +305,13 @@ void notrace s390_do_machine_check(struct pt_regs *regs)
 		/* System damage -> stopping machine */
 		s390_handle_damage();
 	}
-	if (mci.pd) {
+
+	/*
+	 * Reinject the instruction processing damages' machine checks
+	 * including Delayed Access Exception into the guest
+	 * instead of damaging the host if they happen in the guest.
+	 */
+	if (mci.pd && !test_cpu_flag(CIF_MCCK_GUEST)) {
 		if (mci.b) {
 			/* Processing backup -> verify if we can survive this */
 			u64 z_mcic, o_mcic, t_mcic;
@@ -358,15 +368,22 @@ void notrace s390_do_machine_check(struct pt_regs *regs)
 		if (mcck->stp_queue)
 			set_cpu_flag(CIF_MCCK_PENDING);
 	}
-	if (mci.se)
-		/* Storage error uncorrected */
-		s390_handle_damage();
-	if (mci.ke)
-		/* Storage key-error uncorrected */
-		s390_handle_damage();
-	if (mci.ds && mci.fa)
-		/* Storage degradation */
-		s390_handle_damage();
+
+	/*
+	 * Reinject storage related machine checks into the guest if they
+	 * happen when the guest is running.
+	 */
+	if (!test_cpu_flag(CIF_MCCK_GUEST)) {
+		if (mci.se)
+			/* Storage error uncorrected */
+			s390_handle_damage();
+		if (mci.ke)
+			/* Storage key-error uncorrected */
+			s390_handle_damage();
+		if (mci.ds && mci.fa)
+			/* Storage degradation */
+			s390_handle_damage();
+	}
 	if (mci.cp) {
 		/* Channel report word pending */
 		mcck->channel_report = 1;
@@ -377,6 +394,19 @@ void notrace s390_do_machine_check(struct pt_regs *regs)
 		mcck->warning = 1;
 		set_cpu_flag(CIF_MCCK_PENDING);
 	}
+
+	/*
+	 * If there are only Channel Report Pending and External Damage
+	 * machine checks, they will not be reinjected into the guest
+	 * because they refer to host conditions only.
+	 */
+	mcck_dam_code = (mci.val & MCIC_SUBCLASS_MASK);
+	if (test_cpu_flag(CIF_MCCK_GUEST) &&
+	(mcck_dam_code & MCCK_CODE_NO_GUEST) != mcck_dam_code) {
+		/* Set exit reason code for host's later handling */
+		*((long *)(regs->gprs[15] + __SF_SIE_REASON)) = -EINTR;
+	}
+	clear_cpu_flag(CIF_MCCK_GUEST);
 	nmi_exit();
 }