[PATCH 0/6] perf, tools: Fixes and documentation updates

[PATCH 0/6] perf, tools: Fixes and documentation updates

[Robert Richter]

Some random fixes, esp. build and 32 bit fixes. Also some
documentation updates.

Robert Richter (6):
  perf, tools: Fix version file for perf documentation with OUTPUT
    variable set
  perf, tools: Fix lib/traceevent build dir with OUTPUT variable set
  perf, tools: Fix parsing of 64 bit raw config value for 32 bit
  perf, tools: Fix cast from pointer to integer for 32 bit
  perf, tools: Update documentation about raw event setup
  perf, tools: Document precise event sampling for AMD IBS

 tools/lib/traceevent/event-parse.c     |    3 +-
 tools/perf/Documentation/Makefile      |    6 ++--
 tools/perf/Documentation/perf-list.txt |   48 ++++++++++++++++++++++++--------
 tools/perf/Makefile                    |    8 ++--
 tools/perf/util/parse-events.c         |    6 ++--
 tools/perf/util/parse-events.h         |    6 ++--
 tools/perf/util/parse-events.l         |    4 +-
 tools/perf/util/parse-events.y         |   10 +++---
 8 files changed, 58 insertions(+), 33 deletions(-)

-- 
1.7.8.4

[PATCH 1/6] perf, tools: Fix version file for perf documentation with OUTPUT variable set

[Robert Richter]

Fixes the following:

 + make OUTPUT=/.../.build/perf-user/ DESTDIR=/.../.install/perf-user/ man install-man
 make -C Documentation man
 make[1]: Entering directory `/.../.source/linux.perf/tools/perf/Documentation'
 make[2]: Entering directory `/.../.source/linux.perf/tools/perf'
 make[2]: *** No rule to make target `PERF-VERSION-FILE'.  Stop.

Signed-off-by: Robert Richter <robert.richter@amd.com>
---
 tools/perf/Documentation/Makefile |    6 +++---
 1 files changed, 3 insertions(+), 3 deletions(-)

diff --git a/tools/perf/Documentation/Makefile b/tools/perf/Documentation/Makefile
index ca600e0..9f2e44f 100644
--- a/tools/perf/Documentation/Makefile
+++ b/tools/perf/Documentation/Makefile
@@ -195,10 +195,10 @@ install-pdf: pdf
 #install-html: html
 #	'$(SHELL_PATH_SQ)' ./install-webdoc.sh $(DESTDIR)$(htmldir)
 
-../PERF-VERSION-FILE: .FORCE-PERF-VERSION-FILE
-	$(QUIET_SUBDIR0)../ $(QUIET_SUBDIR1) PERF-VERSION-FILE
+$(OUTPUT)PERF-VERSION-FILE: .FORCE-PERF-VERSION-FILE
+	$(QUIET_SUBDIR0)../ $(QUIET_SUBDIR1) $(OUTPUT)PERF-VERSION-FILE
 
--include ../PERF-VERSION-FILE
+-include $(OUTPUT)PERF-VERSION-FILE
 
 #
 # Determine "include::" file references in asciidoc files.
-- 
1.7.8.4

[PATCH 2/6] perf, tools: Fix lib/traceevent build dir with OUTPUT variable set

[Robert Richter]

With the OUTPUT variable set the libtraceevent.a file is wrongly built
in the source directory:

 + make -d OUTPUT=/.../.build/perf-user/ DESTDIR=/.../.install/perf-user/
 ...
     Considering target file `../lib/traceevent//libtraceevent.a'.
      File `../lib/traceevent//libtraceevent.a' does not exist.
      Finished prerequisites of target file `../lib/traceevent//libtraceevent.a'.
     Must remake target `../lib/traceevent//libtraceevent.a'.
 Invoking recipe from Makefile:837 to update target `../lib/traceevent//libtraceevent.a'.
 Putting child 0x703850 (../lib/traceevent//libtraceevent.a) PID 8365 on the chain.
 Live child 0x703850 (../lib/traceevent//libtraceevent.a) PID 8365
     SUBDIR ../lib/traceevent/

 $ git clean -nxd
 Would remove tools/lib/traceevent/.event-parse.d
 Would remove tools/lib/traceevent/.parse-filter.d
 Would remove tools/lib/traceevent/.parse-utils.d
 Would remove tools/lib/traceevent/.trace-seq.d
 Would remove tools/lib/traceevent/event-parse.o
 Would remove tools/lib/traceevent/libtraceevent.a
 Would remove tools/lib/traceevent/parse-filter.o
 Would remove tools/lib/traceevent/parse-utils.o
 Would remove tools/lib/traceevent/trace-seq.o

This patch fixes this.

Note: Though this should already work with O=$outputdir we better use
the OUTPUT variable directly.

Signed-off-by: Robert Richter <robert.richter@amd.com>
---
 tools/perf/Makefile |    8 ++++----
 1 files changed, 4 insertions(+), 4 deletions(-)

diff --git a/tools/perf/Makefile b/tools/perf/Makefile
index 35655c3..2d4bf6e 100644
--- a/tools/perf/Makefile
+++ b/tools/perf/Makefile
@@ -186,10 +186,10 @@ SCRIPTS = $(patsubst %.sh,%,$(SCRIPT_SH))
 
 TRACE_EVENT_DIR = ../lib/traceevent/
 
-ifeq ("$(origin O)", "command line")
-	TE_PATH=$(OUTPUT)/
+ifneq ($(OUTPUT),)
+	TE_PATH=$(OUTPUT)
 else
-	TE_PATH=$(TRACE_EVENT_DIR)/
+	TE_PATH=$(TRACE_EVENT_DIR)
 endif
 
 LIBTRACEEVENT = $(TE_PATH)libtraceevent.a
@@ -842,7 +842,7 @@ $(LIB_FILE): $(LIB_OBJS)
 
 # libtraceevent.a
 $(LIBTRACEEVENT):
-	$(QUIET_SUBDIR0)$(TRACE_EVENT_DIR) $(QUIET_SUBDIR1) $(COMMAND_O) libtraceevent.a
+	$(QUIET_SUBDIR0)$(TRACE_EVENT_DIR) $(QUIET_SUBDIR1) O=$(OUTPUT) libtraceevent.a
 
 help:
 	@echo 'Perf make targets:'
-- 
1.7.8.4

[PATCH 3/6] perf, tools: Fix parsing of 64 bit raw config value for 32 bit

[Robert Richter]

perf record fails on 32 bit with:

 invalid or unsupported event: 'r40000F7E0'

Fixing this by parsing 64 bit num values.

Signed-off-by: Robert Richter <robert.richter@amd.com>
---
 tools/perf/util/parse-events.c |    6 +++---
 tools/perf/util/parse-events.h |    6 +++---
 tools/perf/util/parse-events.l |    4 ++--
 tools/perf/util/parse-events.y |   10 +++++-----
 4 files changed, 13 insertions(+), 13 deletions(-)

diff --git a/tools/perf/util/parse-events.c b/tools/perf/util/parse-events.c
index 74a5af4..8bdfa3e 100644
--- a/tools/perf/util/parse-events.c
+++ b/tools/perf/util/parse-events.c
@@ -551,7 +551,7 @@ static int config_attr(struct perf_event_attr *attr,
 }
 
 int parse_events_add_numeric(struct list_head **list, int *idx,
-			     unsigned long type, unsigned long config,
+			     u32 type, u64 config,
 			     struct list_head *head_config)
 {
 	struct perf_event_attr attr;
@@ -1005,7 +1005,7 @@ int parse_events__is_hardcoded_term(struct parse_events__term *term)
 
 static int new_term(struct parse_events__term **_term, int type_val,
 		    int type_term, char *config,
-		    char *str, long num)
+		    char *str, u64 num)
 {
 	struct parse_events__term *term;
 
@@ -1034,7 +1034,7 @@ static int new_term(struct parse_events__term **_term, int type_val,
 }
 
 int parse_events__term_num(struct parse_events__term **term,
-			   int type_term, char *config, long num)
+			   int type_term, char *config, u64 num)
 {
 	return new_term(term, PARSE_EVENTS__TERM_TYPE_NUM, type_term,
 			config, NULL, num);
diff --git a/tools/perf/util/parse-events.h b/tools/perf/util/parse-events.h
index ee9c218..163aad4 100644
--- a/tools/perf/util/parse-events.h
+++ b/tools/perf/util/parse-events.h
@@ -55,7 +55,7 @@ struct parse_events__term {
 	char *config;
 	union {
 		char *str;
-		long  num;
+		u64  num;
 	} val;
 	int type_val;
 	int type_term;
@@ -73,7 +73,7 @@ struct parse_events_data__terms {
 
 int parse_events__is_hardcoded_term(struct parse_events__term *term);
 int parse_events__term_num(struct parse_events__term **_term,
-			   int type_term, char *config, long num);
+			   int type_term, char *config, u64 num);
 int parse_events__term_str(struct parse_events__term **_term,
 			   int type_term, char *config, char *str);
 int parse_events__term_clone(struct parse_events__term **new,
@@ -83,7 +83,7 @@ int parse_events_modifier(struct list_head *list, char *str);
 int parse_events_add_tracepoint(struct list_head **list, int *idx,
 				char *sys, char *event);
 int parse_events_add_numeric(struct list_head **list, int *idx,
-			     unsigned long type, unsigned long config,
+			     u32 type, u64 config,
 			     struct list_head *head_config);
 int parse_events_add_cache(struct list_head **list, int *idx,
 			   char *type, char *op_result1, char *op_result2);
diff --git a/tools/perf/util/parse-events.l b/tools/perf/util/parse-events.l
index 384ca74..e4abdf2 100644
--- a/tools/perf/util/parse-events.l
+++ b/tools/perf/util/parse-events.l
@@ -15,10 +15,10 @@ YYSTYPE *parse_events_get_lval(yyscan_t yyscanner);
 
 static int __value(YYSTYPE *yylval, char *str, int base, int token)
 {
-	long num;
+	u64 num;
 
 	errno = 0;
-	num = strtoul(str, NULL, base);
+	num = strtoull(str, NULL, base);
 	if (errno)
 		return PE_ERROR;
 
diff --git a/tools/perf/util/parse-events.y b/tools/perf/util/parse-events.y
index 2bc5fbf..423d331 100644
--- a/tools/perf/util/parse-events.y
+++ b/tools/perf/util/parse-events.y
@@ -57,7 +57,7 @@ do { \
 %union
 {
 	char *str;
-	unsigned long num;
+	u64 num;
 	struct list_head *head;
 	struct parse_events__term *term;
 }
@@ -207,7 +207,7 @@ PE_VALUE ':' PE_VALUE
 	struct parse_events_data__events *data = _data;
 	struct list_head *list = NULL;
 
-	ABORT_ON(parse_events_add_numeric(&list, &data->idx, $1, $3, NULL));
+	ABORT_ON(parse_events_add_numeric(&list, &data->idx, (u32)$1, $3, NULL));
 	$$ = list;
 }
 
@@ -282,7 +282,7 @@ PE_TERM '=' PE_NAME
 {
 	struct parse_events__term *term;
 
-	ABORT_ON(parse_events__term_str(&term, $1, NULL, $3));
+	ABORT_ON(parse_events__term_str(&term, (int)$1, NULL, $3));
 	$$ = term;
 }
 |
@@ -290,7 +290,7 @@ PE_TERM '=' PE_VALUE
 {
 	struct parse_events__term *term;
 
-	ABORT_ON(parse_events__term_num(&term, $1, NULL, $3));
+	ABORT_ON(parse_events__term_num(&term, (int)$1, NULL, $3));
 	$$ = term;
 }
 |
@@ -298,7 +298,7 @@ PE_TERM
 {
 	struct parse_events__term *term;
 
-	ABORT_ON(parse_events__term_num(&term, $1, NULL, 1));
+	ABORT_ON(parse_events__term_num(&term, (int)$1, NULL, 1));
 	$$ = term;
 }
 
-- 
1.7.8.4

[PATCH 4/6] perf, tools: Fix cast from pointer to integer for 32 bit

[Robert Richter]

Fixing the integer cast reported by the following warning:

 tools/lib/traceevent/event-parse.c:3488:14: warning: cast from pointer to integer of different size [-Wpointer-to-int-cast]

Signed-off-by: Robert Richter <robert.richter@amd.com>
---
 tools/lib/traceevent/event-parse.c |    3 ++-
 1 files changed, 2 insertions(+), 1 deletions(-)

diff --git a/tools/lib/traceevent/event-parse.c b/tools/lib/traceevent/event-parse.c
index 5f34aa3..b7c2c49 100644
--- a/tools/lib/traceevent/event-parse.c
+++ b/tools/lib/traceevent/event-parse.c
@@ -31,6 +31,7 @@
 #include <stdarg.h>
 #include <ctype.h>
 #include <errno.h>
+#include <stdint.h>
 
 #include "event-parse.h"
 #include "event-utils.h"
@@ -3485,7 +3486,7 @@ process_defined_func(struct trace_seq *s, void *data, int size,
 			if (!string->str)
 				die("malloc str");
 
-			args[i] = (unsigned long long)string->str;
+			args[i] = (uintptr_t)string->str;
 			strings = string;
 			trace_seq_destroy(&str);
 			break;
-- 
1.7.8.4

[PATCH 5/6] perf, tools: Update documentation about raw event setup

[Robert Richter]

It was missing that only certain bit fields are passed to the config
value which confused users. Updating it.

Signed-off-by: Robert Richter <robert.richter@amd.com>
---
 tools/perf/Documentation/perf-list.txt |    6 ++++++
 1 files changed, 6 insertions(+), 0 deletions(-)

diff --git a/tools/perf/Documentation/perf-list.txt b/tools/perf/Documentation/perf-list.txt
index ddc2252..232be51 100644
--- a/tools/perf/Documentation/perf-list.txt
+++ b/tools/perf/Documentation/perf-list.txt
@@ -15,6 +15,7 @@ DESCRIPTION
 This command displays the symbolic event types which can be selected in the
 various perf commands with the -e option.
 
+[[EVENT_MODIFIERS]]
 EVENT MODIFIERS
 ---------------
 
@@ -44,6 +45,11 @@ layout of IA32_PERFEVTSELx MSRs (see [Intel® 64 and IA-32 Architectures Softwar
 of IA32_PERFEVTSELx MSRs) or AMD's PerfEvtSeln (see [AMD64 Architecture Programmer’s Manual Volume 2: System Programming], Page 344,
 Figure 13-7 Performance Event-Select Register (PerfEvtSeln)).
 
+Note: Only the following bit fields can be set in x86 counter
+registers: event, umask, edge, inv, cmask. Esp. guest/host only and
+OS/user mode flags must be setup using <<EVENT_MODIFIERS, EVENT
+MODIFIERS>>.
+
 Example:
 
 If the Intel docs for a QM720 Core i7 describe an event as:
-- 
1.7.8.4

[PATCH 6/6] perf, tools: Document precise event sampling for AMD IBS

[Robert Richter]

Updating man perf-list.

Signed-off-by: Robert Richter <robert.richter@amd.com>
---
 tools/perf/Documentation/perf-list.txt |   42 ++++++++++++++++++++++---------
 1 files changed, 30 insertions(+), 12 deletions(-)

diff --git a/tools/perf/Documentation/perf-list.txt b/tools/perf/Documentation/perf-list.txt
index 232be51..d1e39dc 100644
--- a/tools/perf/Documentation/perf-list.txt
+++ b/tools/perf/Documentation/perf-list.txt
@@ -20,20 +20,38 @@ EVENT MODIFIERS
 ---------------
 
 Events can optionally have a modifer by appending a colon and one or
-more modifiers.  Modifiers allow the user to restrict when events are
-counted with 'u' for user-space, 'k' for kernel, 'h' for hypervisor.
-Additional modifiers are 'G' for guest counting (in KVM guests) and 'H'
-for host counting (not in KVM guests).
+more modifiers. Modifiers allow the user to restrict the events to be
+counted. The following modifiers exist:
+
+ u - user-space counting
+ k - kernel counting
+ h - hypervisor counting
+ G - guest counting (in KVM guests)
+ H - host counting (not in KVM guests)
+ p - precise level
 
 The 'p' modifier can be used for specifying how precise the instruction
-address should be. The 'p' modifier is currently only implemented for
-Intel PEBS and can be specified multiple times:
-  0 - SAMPLE_IP can have arbitrary skid
-  1 - SAMPLE_IP must have constant skid
-  2 - SAMPLE_IP requested to have 0 skid
-  3 - SAMPLE_IP must have 0 skid
+address should be. The 'p' modifier can be specified multiple times:
+
+ 0 - SAMPLE_IP can have arbitrary skid
+ 1 - SAMPLE_IP must have constant skid
+ 2 - SAMPLE_IP requested to have 0 skid
+ 3 - SAMPLE_IP must have 0 skid
+
+For Intel systems precise event sampling is implemented with PEBS
+which supports up to precise-level 2.
+
+On AMD systems it is implemented using IBS (up to precise-level 2).
+The precise modifier works with event types 0x76 (cpu-cycles, CPU
+clocks not halted) and 0xC1 (micro-ops retired). Both events map to
+IBS execution sampling (IBS op) with the IBS Op Counter Control bit
+(IbsOpCntCtl) set respectively (see AMD64 Architecture Programmer’s
+Manual Volume 2: System Programming, 13.3 Instruction-Based
+Sampling). Examples to use IBS:
 
-The PEBS implementation now supports up to 2.
+ perf record -a -e cpu-cycles:p ...    # use ibs op counting cycles
+ perf record -a -e r076:p ...          # same as -e cpu-cycles:p
+ perf record -a -e r0C1:p ...          # use ibs op counting micro-ops
 
 RAW HARDWARE EVENT DESCRIPTOR
 -----------------------------
@@ -97,4 +115,4 @@ SEE ALSO
 linkperf:perf-stat[1], linkperf:perf-top[1],
 linkperf:perf-record[1],
 http://www.intel.com/Assets/PDF/manual/253669.pdf[Intel® 64 and IA-32 Architectures Software Developer's Manual Volume 3B: System Programming Guide],
-http://support.amd.com/us/Processor_TechDocs/24593.pdf[AMD64 Architecture Programmer’s Manual Volume 2: System Programming]
+http://support.amd.com/us/Processor_TechDocs/24593_APM_v2.pdf[AMD64 Architecture Programmer’s Manual Volume 2: System Programming]
-- 
1.7.8.4

[tip:perf/core] perf tools: Fix version file for perf documentation with OUTPUT variable set

[tip-bot for Robert Richter]

Commit-ID:  e6b978335424029d05edc01374744516b21b1ede
Gitweb:     http://git.kernel.org/tip/e6b978335424029d05edc01374744516b21b1ede
Author:     Robert Richter <robert.richter@amd.com>
AuthorDate: Tue, 7 Aug 2012 19:43:11 +0200
Committer:  Arnaldo Carvalho de Melo <acme@redhat.com>
CommitDate: Wed, 8 Aug 2012 12:36:49 -0300

perf tools: Fix version file for perf documentation with OUTPUT variable set

Fixes the following:

 + make OUTPUT=/.../.build/perf-user/ DESTDIR=/.../.install/perf-user/ man install-man
 make -C Documentation man
 make[1]: Entering directory `/.../.source/linux.perf/tools/perf/Documentation'
 make[2]: Entering directory `/.../.source/linux.perf/tools/perf'
 make[2]: *** No rule to make target `PERF-VERSION-FILE'.  Stop.

Signed-off-by: Robert Richter <robert.richter@amd.com>
Cc: Ingo Molnar <mingo@kernel.org>
Link: http://lkml.kernel.org/r/1344361396-7237-2-git-send-email-robert.richter@amd.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
---
 tools/perf/Documentation/Makefile |    6 +++---
 1 files changed, 3 insertions(+), 3 deletions(-)

diff --git a/tools/perf/Documentation/Makefile b/tools/perf/Documentation/Makefile
index ca600e0..9f2e44f 100644
--- a/tools/perf/Documentation/Makefile
+++ b/tools/perf/Documentation/Makefile
@@ -195,10 +195,10 @@ install-pdf: pdf
 #install-html: html
 #	'$(SHELL_PATH_SQ)' ./install-webdoc.sh $(DESTDIR)$(htmldir)
 
-../PERF-VERSION-FILE: .FORCE-PERF-VERSION-FILE
-	$(QUIET_SUBDIR0)../ $(QUIET_SUBDIR1) PERF-VERSION-FILE
+$(OUTPUT)PERF-VERSION-FILE: .FORCE-PERF-VERSION-FILE
+	$(QUIET_SUBDIR0)../ $(QUIET_SUBDIR1) $(OUTPUT)PERF-VERSION-FILE
 
--include ../PERF-VERSION-FILE
+-include $(OUTPUT)PERF-VERSION-FILE
 
 #
 # Determine "include::" file references in asciidoc files.

[tip:perf/core] perf tools: Fix lib/ traceevent build dir with OUTPUT variable set

[tip-bot for Robert Richter]

Commit-ID:  2ede8303db75ead3250f95c3390e6ba200cbe7d4
Gitweb:     http://git.kernel.org/tip/2ede8303db75ead3250f95c3390e6ba200cbe7d4
Author:     Robert Richter <robert.richter@amd.com>
AuthorDate: Tue, 7 Aug 2012 19:43:12 +0200
Committer:  Arnaldo Carvalho de Melo <acme@redhat.com>
CommitDate: Wed, 8 Aug 2012 12:37:41 -0300

perf tools: Fix lib/traceevent build dir with OUTPUT variable set

With the OUTPUT variable set the libtraceevent.a file is wrongly built
in the source directory:

 + make -d OUTPUT=/.../.build/perf-user/ DESTDIR=/.../.install/perf-user/
 ...
     Considering target file `../lib/traceevent//libtraceevent.a'.
      File `../lib/traceevent//libtraceevent.a' does not exist.
      Finished prerequisites of target file `../lib/traceevent//libtraceevent.a'.
     Must remake target `../lib/traceevent//libtraceevent.a'.
 Invoking recipe from Makefile:837 to update target `../lib/traceevent//libtraceevent.a'.
 Putting child 0x703850 (../lib/traceevent//libtraceevent.a) PID 8365 on the chain.
 Live child 0x703850 (../lib/traceevent//libtraceevent.a) PID 8365
     SUBDIR ../lib/traceevent/

 $ git clean -nxd
 Would remove tools/lib/traceevent/.event-parse.d
 Would remove tools/lib/traceevent/.parse-filter.d
 Would remove tools/lib/traceevent/.parse-utils.d
 Would remove tools/lib/traceevent/.trace-seq.d
 Would remove tools/lib/traceevent/event-parse.o
 Would remove tools/lib/traceevent/libtraceevent.a
 Would remove tools/lib/traceevent/parse-filter.o
 Would remove tools/lib/traceevent/parse-utils.o
 Would remove tools/lib/traceevent/trace-seq.o

This patch fixes this.

Note: Though this should already work with O=$outputdir we better use
the OUTPUT variable directly.

Signed-off-by: Robert Richter <robert.richter@amd.com>
Cc: Ingo Molnar <mingo@kernel.org>
Link: http://lkml.kernel.org/r/1344361396-7237-3-git-send-email-robert.richter@amd.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
---
 tools/perf/Makefile |    8 ++++----
 1 files changed, 4 insertions(+), 4 deletions(-)

diff --git a/tools/perf/Makefile b/tools/perf/Makefile
index 35655c3..2d4bf6e 100644
--- a/tools/perf/Makefile
+++ b/tools/perf/Makefile
@@ -186,10 +186,10 @@ SCRIPTS = $(patsubst %.sh,%,$(SCRIPT_SH))
 
 TRACE_EVENT_DIR = ../lib/traceevent/
 
-ifeq ("$(origin O)", "command line")
-	TE_PATH=$(OUTPUT)/
+ifneq ($(OUTPUT),)
+	TE_PATH=$(OUTPUT)
 else
-	TE_PATH=$(TRACE_EVENT_DIR)/
+	TE_PATH=$(TRACE_EVENT_DIR)
 endif
 
 LIBTRACEEVENT = $(TE_PATH)libtraceevent.a
@@ -842,7 +842,7 @@ $(LIB_FILE): $(LIB_OBJS)
 
 # libtraceevent.a
 $(LIBTRACEEVENT):
-	$(QUIET_SUBDIR0)$(TRACE_EVENT_DIR) $(QUIET_SUBDIR1) $(COMMAND_O) libtraceevent.a
+	$(QUIET_SUBDIR0)$(TRACE_EVENT_DIR) $(QUIET_SUBDIR1) O=$(OUTPUT) libtraceevent.a
 
 help:
 	@echo 'Perf make targets:'

[tip:perf/core] perf tools: Fix parsing of 64 bit raw config value for 32 bit

[tip-bot for Robert Richter]

Commit-ID:  b527bab59be7ca2154f644fcc4b9e3c267a6d855
Gitweb:     http://git.kernel.org/tip/b527bab59be7ca2154f644fcc4b9e3c267a6d855
Author:     Robert Richter <robert.richter@amd.com>
AuthorDate: Tue, 7 Aug 2012 19:43:13 +0200
Committer:  Arnaldo Carvalho de Melo <acme@redhat.com>
CommitDate: Wed, 8 Aug 2012 12:39:56 -0300

perf tools: Fix parsing of 64 bit raw config value for 32 bit

perf record fails on 32 bit with:

 invalid or unsupported event: 'r40000F7E0'

Fixing this by parsing 64 bit num values.

Signed-off-by: Robert Richter <robert.richter@amd.com>
Cc: Ingo Molnar <mingo@kernel.org>
Link: http://lkml.kernel.org/r/1344361396-7237-4-git-send-email-robert.richter@amd.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
---
 tools/perf/util/parse-events.c |    6 +++---
 tools/perf/util/parse-events.h |    6 +++---
 tools/perf/util/parse-events.l |    4 ++--
 tools/perf/util/parse-events.y |   10 +++++-----
 4 files changed, 13 insertions(+), 13 deletions(-)

diff --git a/tools/perf/util/parse-events.c b/tools/perf/util/parse-events.c
index 74a5af4..8bdfa3e 100644
--- a/tools/perf/util/parse-events.c
+++ b/tools/perf/util/parse-events.c
@@ -551,7 +551,7 @@ static int config_attr(struct perf_event_attr *attr,
 }
 
 int parse_events_add_numeric(struct list_head **list, int *idx,
-			     unsigned long type, unsigned long config,
+			     u32 type, u64 config,
 			     struct list_head *head_config)
 {
 	struct perf_event_attr attr;
@@ -1005,7 +1005,7 @@ int parse_events__is_hardcoded_term(struct parse_events__term *term)
 
 static int new_term(struct parse_events__term **_term, int type_val,
 		    int type_term, char *config,
-		    char *str, long num)
+		    char *str, u64 num)
 {
 	struct parse_events__term *term;
 
@@ -1034,7 +1034,7 @@ static int new_term(struct parse_events__term **_term, int type_val,
 }
 
 int parse_events__term_num(struct parse_events__term **term,
-			   int type_term, char *config, long num)
+			   int type_term, char *config, u64 num)
 {
 	return new_term(term, PARSE_EVENTS__TERM_TYPE_NUM, type_term,
 			config, NULL, num);
diff --git a/tools/perf/util/parse-events.h b/tools/perf/util/parse-events.h
index ee9c218..163aad4 100644
--- a/tools/perf/util/parse-events.h
+++ b/tools/perf/util/parse-events.h
@@ -55,7 +55,7 @@ struct parse_events__term {
 	char *config;
 	union {
 		char *str;
-		long  num;
+		u64  num;
 	} val;
 	int type_val;
 	int type_term;
@@ -73,7 +73,7 @@ struct parse_events_data__terms {
 
 int parse_events__is_hardcoded_term(struct parse_events__term *term);
 int parse_events__term_num(struct parse_events__term **_term,
-			   int type_term, char *config, long num);
+			   int type_term, char *config, u64 num);
 int parse_events__term_str(struct parse_events__term **_term,
 			   int type_term, char *config, char *str);
 int parse_events__term_clone(struct parse_events__term **new,
@@ -83,7 +83,7 @@ int parse_events_modifier(struct list_head *list, char *str);
 int parse_events_add_tracepoint(struct list_head **list, int *idx,
 				char *sys, char *event);
 int parse_events_add_numeric(struct list_head **list, int *idx,
-			     unsigned long type, unsigned long config,
+			     u32 type, u64 config,
 			     struct list_head *head_config);
 int parse_events_add_cache(struct list_head **list, int *idx,
 			   char *type, char *op_result1, char *op_result2);
diff --git a/tools/perf/util/parse-events.l b/tools/perf/util/parse-events.l
index 384ca74..e4abdf2 100644
--- a/tools/perf/util/parse-events.l
+++ b/tools/perf/util/parse-events.l
@@ -15,10 +15,10 @@ YYSTYPE *parse_events_get_lval(yyscan_t yyscanner);
 
 static int __value(YYSTYPE *yylval, char *str, int base, int token)
 {
-	long num;
+	u64 num;
 
 	errno = 0;
-	num = strtoul(str, NULL, base);
+	num = strtoull(str, NULL, base);
 	if (errno)
 		return PE_ERROR;
 
diff --git a/tools/perf/util/parse-events.y b/tools/perf/util/parse-events.y
index 2bc5fbf..423d331 100644
--- a/tools/perf/util/parse-events.y
+++ b/tools/perf/util/parse-events.y
@@ -57,7 +57,7 @@ do { \
 %union
 {
 	char *str;
-	unsigned long num;
+	u64 num;
 	struct list_head *head;
 	struct parse_events__term *term;
 }
@@ -207,7 +207,7 @@ PE_VALUE ':' PE_VALUE
 	struct parse_events_data__events *data = _data;
 	struct list_head *list = NULL;
 
-	ABORT_ON(parse_events_add_numeric(&list, &data->idx, $1, $3, NULL));
+	ABORT_ON(parse_events_add_numeric(&list, &data->idx, (u32)$1, $3, NULL));
 	$$ = list;
 }
 
@@ -282,7 +282,7 @@ PE_TERM '=' PE_NAME
 {
 	struct parse_events__term *term;
 
-	ABORT_ON(parse_events__term_str(&term, $1, NULL, $3));
+	ABORT_ON(parse_events__term_str(&term, (int)$1, NULL, $3));
 	$$ = term;
 }
 |
@@ -290,7 +290,7 @@ PE_TERM '=' PE_VALUE
 {
 	struct parse_events__term *term;
 
-	ABORT_ON(parse_events__term_num(&term, $1, NULL, $3));
+	ABORT_ON(parse_events__term_num(&term, (int)$1, NULL, $3));
 	$$ = term;
 }
 |
@@ -298,7 +298,7 @@ PE_TERM
 {
 	struct parse_events__term *term;
 
-	ABORT_ON(parse_events__term_num(&term, $1, NULL, 1));
+	ABORT_ON(parse_events__term_num(&term, (int)$1, NULL, 1));
 	$$ = term;
 }
 

[tip:perf/core] tools lib traceevent: Fix cast from pointer to integer for 32 bit

[tip-bot for Robert Richter]

Commit-ID:  0cf260131c52f681533d17db6fd07545a3dc184e
Gitweb:     http://git.kernel.org/tip/0cf260131c52f681533d17db6fd07545a3dc184e
Author:     Robert Richter <robert.richter@amd.com>
AuthorDate: Tue, 7 Aug 2012 19:43:14 +0200
Committer:  Arnaldo Carvalho de Melo <acme@redhat.com>
CommitDate: Wed, 8 Aug 2012 12:40:44 -0300

tools lib traceevent: Fix cast from pointer to integer for 32 bit

Fixing the integer cast reported by the following warning:

 tools/lib/traceevent/event-parse.c:3488:14: warning: cast from pointer to integer of different size [-Wpointer-to-int-cast]

Signed-off-by: Robert Richter <robert.richter@amd.com>
Cc: Ingo Molnar <mingo@kernel.org>
Link: http://lkml.kernel.org/r/1344361396-7237-5-git-send-email-robert.richter@amd.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
---
 tools/lib/traceevent/event-parse.c |    3 ++-
 1 files changed, 2 insertions(+), 1 deletions(-)

diff --git a/tools/lib/traceevent/event-parse.c b/tools/lib/traceevent/event-parse.c
index 5f34aa3..b7c2c49 100644
--- a/tools/lib/traceevent/event-parse.c
+++ b/tools/lib/traceevent/event-parse.c
@@ -31,6 +31,7 @@
 #include <stdarg.h>
 #include <ctype.h>
 #include <errno.h>
+#include <stdint.h>
 
 #include "event-parse.h"
 #include "event-utils.h"
@@ -3485,7 +3486,7 @@ process_defined_func(struct trace_seq *s, void *data, int size,
 			if (!string->str)
 				die("malloc str");
 
-			args[i] = (unsigned long long)string->str;
+			args[i] = (uintptr_t)string->str;
 			strings = string;
 			trace_seq_destroy(&str);
 			break;

[tip:perf/core] perf list: Update documentation about raw event setup

[tip-bot for Robert Richter]

Commit-ID:  75bc5ca89827fe3f2399321b2920a30bcf658049
Gitweb:     http://git.kernel.org/tip/75bc5ca89827fe3f2399321b2920a30bcf658049
Author:     Robert Richter <robert.richter@amd.com>
AuthorDate: Tue, 7 Aug 2012 19:43:15 +0200
Committer:  Arnaldo Carvalho de Melo <acme@redhat.com>
CommitDate: Wed, 8 Aug 2012 12:41:53 -0300

perf list: Update documentation about raw event setup

It was missing that only certain bit fields are passed to the config
value which confused users. Updating it.

Signed-off-by: Robert Richter <robert.richter@amd.com>
Cc: Ingo Molnar <mingo@kernel.org>
Link: http://lkml.kernel.org/r/1344361396-7237-6-git-send-email-robert.richter@amd.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
---
 tools/perf/Documentation/perf-list.txt |    6 ++++++
 1 files changed, 6 insertions(+), 0 deletions(-)

diff --git a/tools/perf/Documentation/perf-list.txt b/tools/perf/Documentation/perf-list.txt
index ddc2252..232be51 100644
--- a/tools/perf/Documentation/perf-list.txt
+++ b/tools/perf/Documentation/perf-list.txt
@@ -15,6 +15,7 @@ DESCRIPTION
 This command displays the symbolic event types which can be selected in the
 various perf commands with the -e option.
 
+[[EVENT_MODIFIERS]]
 EVENT MODIFIERS
 ---------------
 
@@ -44,6 +45,11 @@ layout of IA32_PERFEVTSELx MSRs (see [Intel® 64 and IA-32 Architectures Softwar
 of IA32_PERFEVTSELx MSRs) or AMD's PerfEvtSeln (see [AMD64 Architecture Programmer’s Manual Volume 2: System Programming], Page 344,
 Figure 13-7 Performance Event-Select Register (PerfEvtSeln)).
 
+Note: Only the following bit fields can be set in x86 counter
+registers: event, umask, edge, inv, cmask. Esp. guest/host only and
+OS/user mode flags must be setup using <<EVENT_MODIFIERS, EVENT
+MODIFIERS>>.
+
 Example:
 
 If the Intel docs for a QM720 Core i7 describe an event as:

[tip:perf/core] perf list: Document precise event sampling for AMD IBS

[tip-bot for Robert Richter]

Commit-ID:  2055fdaf8703d3101b12e0d9b7cbceaeabe35c17
Gitweb:     http://git.kernel.org/tip/2055fdaf8703d3101b12e0d9b7cbceaeabe35c17
Author:     Robert Richter <robert.richter@amd.com>
AuthorDate: Tue, 7 Aug 2012 19:43:16 +0200
Committer:  Arnaldo Carvalho de Melo <acme@redhat.com>
CommitDate: Wed, 8 Aug 2012 12:42:21 -0300

perf list: Document precise event sampling for AMD IBS

Updating man perf-list.

Signed-off-by: Robert Richter <robert.richter@amd.com>
Cc: Ingo Molnar <mingo@kernel.org>
Link: http://lkml.kernel.org/r/1344361396-7237-7-git-send-email-robert.richter@amd.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
---
 tools/perf/Documentation/perf-list.txt |   42 ++++++++++++++++++++++---------
 1 files changed, 30 insertions(+), 12 deletions(-)

diff --git a/tools/perf/Documentation/perf-list.txt b/tools/perf/Documentation/perf-list.txt
index 232be51..d1e39dc 100644
--- a/tools/perf/Documentation/perf-list.txt
+++ b/tools/perf/Documentation/perf-list.txt
@@ -20,20 +20,38 @@ EVENT MODIFIERS
 ---------------
 
 Events can optionally have a modifer by appending a colon and one or
-more modifiers.  Modifiers allow the user to restrict when events are
-counted with 'u' for user-space, 'k' for kernel, 'h' for hypervisor.
-Additional modifiers are 'G' for guest counting (in KVM guests) and 'H'
-for host counting (not in KVM guests).
+more modifiers. Modifiers allow the user to restrict the events to be
+counted. The following modifiers exist:
+
+ u - user-space counting
+ k - kernel counting
+ h - hypervisor counting
+ G - guest counting (in KVM guests)
+ H - host counting (not in KVM guests)
+ p - precise level
 
 The 'p' modifier can be used for specifying how precise the instruction
-address should be. The 'p' modifier is currently only implemented for
-Intel PEBS and can be specified multiple times:
-  0 - SAMPLE_IP can have arbitrary skid
-  1 - SAMPLE_IP must have constant skid
-  2 - SAMPLE_IP requested to have 0 skid
-  3 - SAMPLE_IP must have 0 skid
+address should be. The 'p' modifier can be specified multiple times:
+
+ 0 - SAMPLE_IP can have arbitrary skid
+ 1 - SAMPLE_IP must have constant skid
+ 2 - SAMPLE_IP requested to have 0 skid
+ 3 - SAMPLE_IP must have 0 skid
+
+For Intel systems precise event sampling is implemented with PEBS
+which supports up to precise-level 2.
+
+On AMD systems it is implemented using IBS (up to precise-level 2).
+The precise modifier works with event types 0x76 (cpu-cycles, CPU
+clocks not halted) and 0xC1 (micro-ops retired). Both events map to
+IBS execution sampling (IBS op) with the IBS Op Counter Control bit
+(IbsOpCntCtl) set respectively (see AMD64 Architecture Programmer’s
+Manual Volume 2: System Programming, 13.3 Instruction-Based
+Sampling). Examples to use IBS:
 
-The PEBS implementation now supports up to 2.
+ perf record -a -e cpu-cycles:p ...    # use ibs op counting cycles
+ perf record -a -e r076:p ...          # same as -e cpu-cycles:p
+ perf record -a -e r0C1:p ...          # use ibs op counting micro-ops
 
 RAW HARDWARE EVENT DESCRIPTOR
 -----------------------------
@@ -97,4 +115,4 @@ SEE ALSO
 linkperf:perf-stat[1], linkperf:perf-top[1],
 linkperf:perf-record[1],
 http://www.intel.com/Assets/PDF/manual/253669.pdf[Intel® 64 and IA-32 Architectures Software Developer's Manual Volume 3B: System Programming Guide],
-http://support.amd.com/us/Processor_TechDocs/24593.pdf[AMD64 Architecture Programmer’s Manual Volume 2: System Programming]
+http://support.amd.com/us/Processor_TechDocs/24593_APM_v2.pdf[AMD64 Architecture Programmer’s Manual Volume 2: System Programming]