s390/cpum_sf: Add raw data sampling to support the diagnostic-sampling function

Also support the diagnostic-sampling function in addition to the basic-sampling
function.  Diagnostic-sampling data entries contain hardware model specific
sampling data and additional programs are required to analyze the data.

To deliver diagnostic-sampling, as well, as basis-sampling data entries to user
space, introduce support for sampling "raw data".  If this particular perf
sampling type (PERF_SAMPLE_RAW) is used, sampling data entries are copied
to user space.  External programs can then analyze these data.
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

@@ -59,13 +59,15 @@ struct cpumf_ctr_info {
 /* QUERY SAMPLING INFORMATION block */
 struct hws_qsi_info_block {	    /* Bit(s) */
 	unsigned int b0_13:14;	    /* 0-13: zeros			 */
-	unsigned int as:1;	    /* 14: sampling authorisation control*/
-	unsigned int b15_21:7;	    /* 15-21: zeros			 */
-	unsigned int es:1;	    /* 22: sampling enable control	 */
-	unsigned int b23_29:7;	    /* 23-29: zeros			 */
-	unsigned int cs:1;	    /* 30: sampling activation control	 */
-	unsigned int:1; 	    /* 31: reserved			 */
-	unsigned int bsdes:16;	    /* 4-5: size of basic sampling entry      */
+	unsigned int as:1;	    /* 14: basic-sampling authorization	 */
+	unsigned int ad:1;	    /* 15: diag-sampling authorization	 */
+	unsigned int b16_21:6;	    /* 16-21: zeros			 */
+	unsigned int es:1;	    /* 22: basic-sampling enable control */
+	unsigned int ed:1;	    /* 23: diag-sampling enable control	 */
+	unsigned int b24_29:6;	    /* 24-29: zeros			 */
+	unsigned int cs:1;	    /* 30: basic-sampling activation control */
+	unsigned int cd:1;	    /* 31: diag-sampling activation control */
+	unsigned int bsdes:16;	    /* 4-5: size of basic sampling entry */
 	unsigned int dsdes:16;	    /* 6-7: size of diagnostic sampling entry */
 	unsigned long min_sampl_rate; /* 8-15: minimum sampling interval */
 	unsigned long max_sampl_rate; /* 16-23: maximum sampling interval*/
@@ -82,10 +84,11 @@ struct hws_lsctl_request_block {
 	unsigned int s:1;	    /* 0: maximum buffer indicator	 */
 	unsigned int h:1;	    /* 1: part. level reserved for VM use*/
 	unsigned long long b2_53:52;/* 2-53: zeros			 */
-	unsigned int es:1;	    /* 54: sampling enable control	 */
-	unsigned int b55_61:7;	    /* 55-61: - zeros			 */
-	unsigned int cs:1;	    /* 62: sampling activation control	 */
-	unsigned int b63:1;	    /* 63: zero 			 */
+	unsigned int es:1;	    /* 54: basic-sampling enable control */
+	unsigned int ed:1;	    /* 55: diag-sampling enable control	 */
+	unsigned int b56_61:6;	    /* 56-61: - zeros			 */
+	unsigned int cs:1;	    /* 62: basic-sampling activation control */
+	unsigned int cd:1;	    /* 63: diag-sampling activation control  */
 	unsigned long interval;     /* 8-15: sampling interval		 */
 	unsigned long tear;	    /* 16-23: TEAR contents		 */
 	unsigned long dear;	    /* 24-31: DEAR contents		 */
@@ -96,8 +99,7 @@ struct hws_lsctl_request_block {
 	unsigned long rsvrd4;	    /* reserved 			 */
 } __packed;
 
-
-struct hws_data_entry {
+struct hws_basic_entry {
 	unsigned int def:16;	    /* 0-15  Data Entry Format		 */
 	unsigned int R:4;	    /* 16-19 reserved			 */
 	unsigned int U:4;	    /* 20-23 Number of unique instruct.  */
@@ -114,6 +116,18 @@ struct hws_data_entry {
 	unsigned long long hpp;     /* Host Program Parameter		 */
 } __packed;
 
+struct hws_diag_entry {
+	unsigned int def:16;	    /* 0-15  Data Entry Format		 */
+	unsigned int R:14;	    /* 16-19 and 20-30 reserved		 */
+	unsigned int I:1;	    /* 31 entry valid or invalid	 */
+	u8	     data[];	    /* Machine-dependent sample data	 */
+} __packed;
+
+struct hws_combined_entry {
+	struct hws_basic_entry	basic;	/* Basic-sampling data entry */
+	struct hws_diag_entry	diag;	/* Diagnostic-sampling data entry */
+} __packed;
+
 struct hws_trailer_entry {
 	union {
 		struct {

arch/s390/include/asm/perf_event.h

@@ -52,15 +52,39 @@ struct perf_sf_sde_regs {
 #define PERF_CPUM_CF_MAX_CTR		256
 
 /* Perf PMU definitions for the sampling facility */
-#define PERF_CPUM_SF_MAX_CTR		1
-#define PERF_EVENT_CPUM_SF		0xB0000UL	/* Raw event ID */
+#define PERF_CPUM_SF_MAX_CTR		2
+#define PERF_EVENT_CPUM_SF		0xB0000UL /* Event: Basic-sampling */
+#define PERF_EVENT_CPUM_SF_DIAG		0xBD000UL /* Event: Combined-sampling */
+#define PERF_CPUM_SF_BASIC_MODE		0x0001	  /* Basic-sampling flag */
+#define PERF_CPUM_SF_DIAG_MODE		0x0002	  /* Diagnostic-sampling flag */
+#define PERF_CPUM_SF_MODE_MASK		(PERF_CPUM_SF_BASIC_MODE| \
+					 PERF_CPUM_SF_DIAG_MODE)
 
 #define REG_NONE		0
 #define REG_OVERFLOW		1
 #define OVERFLOW_REG(hwc)	((hwc)->extra_reg.config)
 #define SFB_ALLOC_REG(hwc)	((hwc)->extra_reg.alloc)
+#define RAWSAMPLE_REG(hwc)	((hwc)->config)
 #define TEAR_REG(hwc)		((hwc)->last_tag)
 #define SAMPL_RATE(hwc)		((hwc)->event_base)
+#define SAMPL_FLAGS(hwc)	((hwc)->config_base)
+#define SAMPL_DIAG_MODE(hwc)	(SAMPL_FLAGS(hwc) & PERF_CPUM_SF_DIAG_MODE)
+
+/* Structure for sampling data entries to be passed as perf raw sample data
+ * to user space.  Note that raw sample data must be aligned and, thus, might
+ * be padded with zeros.
+ */
+struct sf_raw_sample {
+#define SF_RAW_SAMPLE_BASIC	PERF_CPUM_SF_BASIC_MODE
+#define SF_RAW_SAMPLE_DIAG	PERF_CPUM_SF_DIAG_MODE
+	u64			format;
+	u32			 size;	  /* Size of sf_raw_sample */
+	u16			bsdes;	  /* Basic-sampling data entry size */
+	u16			dsdes;	  /* Diagnostic-sampling data entry size */
+	struct hws_basic_entry	basic;	  /* Basic-sampling data entry */
+	struct hws_diag_entry	 diag;	  /* Diagnostic-sampling data entry */
+	u8		    padding[];	  /* Padding to next multiple of 8 */
+} __packed;
 
 /* Perf hardware reserve and release functions */
 int perf_reserve_sampling(void);

arch/s390/kernel/perf_event.c

@@ -139,16 +139,21 @@ static void print_debug_sf(void)
 	int cpu = smp_processor_id();
 
 	memset(&si, 0, sizeof(si));
-	if (qsi(&si)) {
-		pr_err("CPU[%i]: CPM_SF: qsi failed\n");
+	if (qsi(&si))
 		return;
-	}
 
-	pr_info("CPU[%i]: CPM_SF: as=%i es=%i cs=%i bsdes=%i dsdes=%i"
-		" min=%i max=%i cpu_speed=%i tear=%p dear=%p\n",
-		cpu, si.as, si.es, si.cs, si.bsdes, si.dsdes,
-		si.min_sampl_rate, si.max_sampl_rate, si.cpu_speed,
-		si.tear, si.dear);
+	pr_info("CPU[%i] CPUM_SF: basic=%i diag=%i min=%i max=%i cpu_speed=%i\n",
+		cpu, si.as, si.ad, si.min_sampl_rate, si.max_sampl_rate,
+		si.cpu_speed);
+
+	if (si.as)
+		pr_info("CPU[%i] CPUM_SF: Basic-sampling: a=%i e=%i c=%i"
+			" bsdes=%i tear=%p dear=%p\n", cpu,
+			si.as, si.es, si.cs, si.bsdes, si.tear, si.dear);
+	if (si.ad)
+		pr_info("CPU[%i] CPUM_SF: Diagnostic-sampling: a=%i e=%i c=%i"
+			" dsdes=%i tear=%p dear=%p\n", cpu,
+			si.ad, si.ed, si.cd, si.dsdes, si.tear, si.dear);
 }
 
 void perf_event_print_debug(void)

arch/s390/oprofile/hwsampler.c

@@ -799,7 +799,7 @@ static void worker_on_interrupt(unsigned int cpu)
 static void add_samples_to_oprofile(unsigned int cpu, unsigned long *sdbt,
 		unsigned long *dear)
 {
-	struct hws_data_entry *sample_data_ptr;
+	struct hws_basic_entry *sample_data_ptr;
 	unsigned long *trailer;
 
 	trailer = trailer_entry_ptr(*sdbt);
@@ -809,7 +809,7 @@ static void add_samples_to_oprofile(unsigned int cpu, unsigned long *sdbt,
 		trailer = dear;
 	}
 
-	sample_data_ptr = (struct hws_data_entry *)(*sdbt);
+	sample_data_ptr = (struct hws_basic_entry *)(*sdbt);
 
 	while ((unsigned long *)sample_data_ptr < trailer) {
 		struct pt_regs *regs = NULL;