s390/cpum_sf: Atomically reset trailer entry fields of sample-data-blocks

Ensure to reset the sample-data-block full indicator and the overflow counter
at the same time.  This must be done atomically because the sampling hardware
is still active while full sample-data-block is processed.
Signed-off-by: Hendrik Brueckner <brueckner@linux.vnet.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>

arch/s390/include/asm/cpu_mf.h

@@ -115,10 +115,15 @@ struct hws_data_entry {
 } __packed;
 
 struct hws_trailer_entry {
-	unsigned int f:1;	    /* 0 - Block Full Indicator 	 */
-	unsigned int a:1;	    /* 1 - Alert request control	 */
-	unsigned int t:1;	    /* 2 - Timestamp format		 */
-	unsigned long long:61;	    /* 3 - 63: Reserved 		 */
+	union {
+		struct {
+			unsigned int f:1;	/* 0 - Block Full Indicator   */
+			unsigned int a:1;	/* 1 - Alert request control  */
+			unsigned int t:1;	/* 2 - Timestamp format	      */
+			unsigned long long:61;	/* 3 - 63: Reserved	      */
+		};
+		unsigned long long flags;	/* 0 - 63: All indicators     */
+	};
 	unsigned long long overflow;	 /* 64 - sample Overflow count	      */
 	unsigned long long timestamp;	 /* 16 - time-stamp		      */
 	unsigned long long timestamp1;	 /*				      */

arch/s390/kernel/perf_cpum_sf.c

@@ -953,7 +953,7 @@ static void hw_perf_event_update(struct perf_event *event, int flush_all)
 	struct hw_perf_event *hwc = &event->hw;
 	struct hws_trailer_entry *te;
 	unsigned long *sdbt;
-	unsigned long long event_overflow, sampl_overflow, num_sdb;
+	unsigned long long event_overflow, sampl_overflow, num_sdb, te_flags;
 	int done;
 
 	sdbt = (unsigned long *) TEAR_REG(hwc);
@@ -990,9 +990,13 @@ static void hw_perf_event_update(struct perf_event *event, int flush_all)
 		hw_collect_samples(event, sdbt, &event_overflow);
 		num_sdb++;
 
-		/* Reset trailer */
-		xchg(&te->overflow, 0);
-		xchg((unsigned char *) te, 0x40);
+		/* Reset trailer (using compare-double-and-swap) */
+		do {
+			te_flags = te->flags & ~SDB_TE_BUFFER_FULL_MASK;
+			te_flags |= SDB_TE_ALERT_REQ_MASK;
+		} while (!cmpxchg_double(&te->flags, &te->overflow,
+					 te->flags, te->overflow,
+					 te_flags, 0ULL));
 
 		/* Advance to next sample-data-block */
 		sdbt++;