perf, x86: Enable not tagged retired instruction counting on P4s

This should turn on instruction counting on P4s, which was missing in
the first version of the new PMU driver.

It's inaccurate for now, we still need dependant event to tag mops
before we can count them precisely. The result is that the number of
instruction may be lifted up.
Signed-off-by: Cyrill Gorcunov <gorcunov@openvz.org>
Signed-off-by: Lin Ming <ming.m.lin@intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>
LKML-Reference: <1268629102.3355.11.camel@minggr.sh.intel.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>

arch/x86/include/asm/perf_event_p4.h

@@ -324,8 +324,8 @@ static inline u32 p4_default_escr_conf(int cpu, int exclude_os, int exclude_usr)
 
 #define P4_SSE_INPUT_ASSIST		P4_EVENT_PACK(0x34, 0x01)
 	/*
-	 * MSR_P4_FIRM_ESCR:	8, 9
-	 * MSR_P4_FIRM_ESCR:	10, 11
+	 * MSR_P4_FIRM_ESCR0:	8, 9
+	 * MSR_P4_FIRM_ESCR1:	10, 11
 	 */
 
 #define P4_PACKED_SP_UOP		P4_EVENT_PACK(0x08, 0x01)
@@ -462,8 +462,8 @@ static inline u32 p4_default_escr_conf(int cpu, int exclude_os, int exclude_usr)
 
 #define P4_INSTR_RETIRED		P4_EVENT_PACK(0x02, 0x04)
 	/*
-	 * MSR_P4_CRU_ESCR2:	12, 13, 16
-	 * MSR_P4_CRU_ESCR3:	14, 15, 17
+	 * MSR_P4_CRU_ESCR0:	12, 13, 16
+	 * MSR_P4_CRU_ESCR1:	14, 15, 17
 	 */
 
 #define P4_UOPS_RETIRED			P4_EVENT_PACK(0x01, 0x04)

arch/x86/kernel/cpu/perf_event_p4.c

@@ -60,13 +60,11 @@ struct p4_event_template p4_templates[] = {
 	[2] = {
 		.opcode	= P4_INSTR_RETIRED,
 		.config	= 0,
-		.dep	= 0, /* needs front-end tagging */
+		.dep	= -1, /* needs front-end tagging */
 		.emask	=
 			P4_EVENT_ATTR(P4_INSTR_RETIRED, NBOGUSNTAG)	|
-			P4_EVENT_ATTR(P4_INSTR_RETIRED, NBOGUSTAG)	|
-			P4_EVENT_ATTR(P4_INSTR_RETIRED, BOGUSNTAG)	|
-			P4_EVENT_ATTR(P4_INSTR_RETIRED, BOGUSTAG),
-		.escr_msr	= { MSR_P4_CRU_ESCR2, MSR_P4_CRU_ESCR3 },
+			P4_EVENT_ATTR(P4_INSTR_RETIRED, BOGUSNTAG),
+		.escr_msr	= { MSR_P4_CRU_ESCR0, MSR_P4_CRU_ESCR1 },
 		.cntr		= { 12, 14 },
 	},
 	[3] = {