[PATCH 0/3] perf tools: Fixes for some compile errors.

[PATCH 0/3] perf tools: Fixes for some compile errors.

[Feng Tang]

Hi Arnaldo and all,

I met 3 compiling problems on my x86_32 machine, and these are the fixes,
patch 1/3 only applies for perf/core branch of your git, while patch 2
applies to Linus' tree also. please review.

Thanks,
Feng

---
Feng Tang (3):
  perf tools: Fix a compiling error in trace-event-perl.c for 32 bits
    machine
  Perf tools: Fix a compiling error in util/map.c
  perf tools: Fix a misuse of for_each_set_bit() in session.c

 tools/perf/util/map.c                              |    5 ++---
 .../perf/util/scripting-engines/trace-event-perl.c |    2 +-
 tools/perf/util/session.c                          |    9 ++++-----
 3 files changed, 7 insertions(+), 9 deletions(-)

[PATCH 1/3] perf tools: Fix a compiling error in trace-event-perl.c for 32 bits machine

[Feng Tang]

On my x86_32 mahcine, there is a compile error:

	CC util/scripting-engines/trace-event-perl.o
	cc1: warnings being treated as errors
	util/scripting-engines/trace-event-perl.c: In function ‘perl_process_tracepoint’:
	util/scripting-engines/trace-event-perl.c:285: error: format ‘%d’ expects type ‘int’, but argument 2 has type ‘__u64’
	make: *** [util/scripting-engines/trace-event-perl.o] Error 1

Fix it by using the "%lld" for __u64.

Signed-off-by: Feng Tang <feng.tang@intel.com>
---
 .../perf/util/scripting-engines/trace-event-perl.c |    2 +-
 1 files changed, 1 insertions(+), 1 deletions(-)

diff --git a/tools/perf/util/scripting-engines/trace-event-perl.c b/tools/perf/util/scripting-engines/trace-event-perl.c
index d280010..1cde6aa 100644
--- a/tools/perf/util/scripting-engines/trace-event-perl.c
+++ b/tools/perf/util/scripting-engines/trace-event-perl.c
@@ -282,7 +282,7 @@ static void perl_process_tracepoint(union perf_event *perf_event __unused,
 
 	event = find_cache_event(evsel);
 	if (!event)
-		die("ug! no event found for type %d", evsel->attr.config);
+		die("ug! no event found for type %lld", evsel->attr.config);
 
 	pid = raw_field_value(event, "common_pid", data);
 
-- 
1.7.1

Re: [PATCH 1/3] perf tools: Fix a compiling error in trace-event-perl.c for 32 bits machine

[Arnaldo Carvalho de Melo]

Em Mon, Aug 27, 2012 at 03:38:25PM +0800, Feng Tang escreveu:
> On my x86_32 mahcine, there is a compile error:
> 
> 	CC util/scripting-engines/trace-event-perl.o
> 	cc1: warnings being treated as errors
> 	util/scripting-engines/trace-event-perl.c: In function ‘perl_process_tracepoint’:
> 	util/scripting-engines/trace-event-perl.c:285: error: format ‘%d’ expects type ‘int’, but argument 2 has type ‘__u64’
> 	make: *** [util/scripting-engines/trace-event-perl.o] Error 1
> 
> Fix it by using the "%lld" for __u64.
> 
> Signed-off-by: Feng Tang <feng.tang@intel.com>
> ---
>  .../perf/util/scripting-engines/trace-event-perl.c |    2 +-
>  1 files changed, 1 insertions(+), 1 deletions(-)
> 
> diff --git a/tools/perf/util/scripting-engines/trace-event-perl.c b/tools/perf/util/scripting-engines/trace-event-perl.c
> index d280010..1cde6aa 100644
> --- a/tools/perf/util/scripting-engines/trace-event-perl.c
> +++ b/tools/perf/util/scripting-engines/trace-event-perl.c
> @@ -282,7 +282,7 @@ static void perl_process_tracepoint(union perf_event *perf_event __unused,
>  
>  	event = find_cache_event(evsel);
>  	if (!event)
> -		die("ug! no event found for type %d", evsel->attr.config);
> +		die("ug! no event found for type %lld", evsel->attr.config);

Please use PRIu64 instead, like in:

grep PRIu64 tools/perf/*/*.c
  
>  	pid = raw_field_value(event, "common_pid", data);
>  
> -- 
> 1.7.1

[PATCH v2 1/3] perf tools: Fix a compiling error in trace-event-perl.c for 32 bits machine

[Feng Tang]

>From 8e2fe258906fc83869be10a1b4a40ef9acc4a2a6 Mon Sep 17 00:00:00 2001
From: Feng Tang <feng.tang@intel.com>
Date: Mon, 27 Aug 2012 14:01:33 +0800
Subject: [PATCH 1/3] perf tools: Fix a compiling error in trace-event-perl.c for 32 bits machine

On my x86_32 mahcine, there is a compile error:

	CC util/scripting-engines/trace-event-perl.o
	cc1: warnings being treated as errors
	util/scripting-engines/trace-event-perl.c: In function ‘perl_process_tracepoint’:
	util/scripting-engines/trace-event-perl.c:285: error: format ‘%d’ expects type ‘int’, but argument 2 has type ‘__u64’
	make: *** [util/scripting-engines/trace-event-perl.o] Error 1

Fix it by using the "%PRIu64" for __u64.

v2: use PRIu64 as suggested by Arnaldo.

Signed-off-by: Feng Tang <feng.tang@intel.com>
---
 .../perf/util/scripting-engines/trace-event-perl.c |    2 +-
 1 files changed, 1 insertions(+), 1 deletions(-)

diff --git a/tools/perf/util/scripting-engines/trace-event-perl.c b/tools/perf/util/scripting-engines/trace-event-perl.c
index d280010..09db03f 100644
--- a/tools/perf/util/scripting-engines/trace-event-perl.c
+++ b/tools/perf/util/scripting-engines/trace-event-perl.c
@@ -282,7 +282,7 @@ static void perl_process_tracepoint(union perf_event *perf_event __unused,
 
 	event = find_cache_event(evsel);
 	if (!event)
-		die("ug! no event found for type %d", evsel->attr.config);
+		die("ug! no event found for type %" PRIu64, evsel->attr.config);
 
 	pid = raw_field_value(event, "common_pid", data);
 
-- 
1.7.1

[tip:perf/core] perf tools: Fix a compiling error in trace-event-perl.c for 32 bits machine

[tip-bot for Feng Tang]

Commit-ID:  b1ab1bd1921536c2a97adb888effeff4370a3246
Gitweb:     http://git.kernel.org/tip/b1ab1bd1921536c2a97adb888effeff4370a3246
Author:     Feng Tang <feng.tang@intel.com>
AuthorDate: Thu, 20 Sep 2012 08:30:21 -0300
Committer:  Arnaldo Carvalho de Melo <acme@redhat.com>
CommitDate: Thu, 20 Sep 2012 08:30:21 -0300

perf tools: Fix a compiling error in trace-event-perl.c for 32 bits machine

On my x86_32 mahcine, there is a compile error:

        CC util/scripting-engines/trace-event-perl.o
        cc1: warnings being treated as errors
        util/scripting-engines/trace-event-perl.c: In function
	perl_process_tracepoint:
        util/scripting-engines/trace-event-perl.c:285: error: format
	expects type 'int', but argument 2 has type '__u64'
        make: *** [util/scripting-engines/trace-event-perl.o] Error 1

Fix it by using the "%PRIu64" for __u64.

v2: use PRIu64 as suggested by Arnaldo.

Signed-off-by: Feng Tang <feng.tang@intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/r/20120828101730.6b2fd97e@feng-i7
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
---
 .../perf/util/scripting-engines/trace-event-perl.c |    2 +-
 1 files changed, 1 insertions(+), 1 deletions(-)

diff --git a/tools/perf/util/scripting-engines/trace-event-perl.c b/tools/perf/util/scripting-engines/trace-event-perl.c
index ffde3e4..f80605e 100644
--- a/tools/perf/util/scripting-engines/trace-event-perl.c
+++ b/tools/perf/util/scripting-engines/trace-event-perl.c
@@ -282,7 +282,7 @@ static void perl_process_tracepoint(union perf_event *perf_event __maybe_unused,
 
 	event = find_cache_event(evsel);
 	if (!event)
-		die("ug! no event found for type %d", evsel->attr.config);
+		die("ug! no event found for type %" PRIu64, evsel->attr.config);
 
 	pid = raw_field_value(event, "common_pid", data);
 

[PATCH 2/3] Perf tools: Fix a compiling error in util/map.c

[Feng Tang]

This patch fix a compile warning taken as error:

	CC util/map.o
	cc1: warnings being treated as errors
	util/map.c: In function ‘map__fprintf_dsoname’:
	util/map.c:240: error: ‘dsoname’ may be used uninitialized in this function
	make: *** [util/map.o] Error 1

Signed-off-by: Feng Tang <feng.tang@intel.com>
---
 tools/perf/util/map.c |    5 ++---
 1 files changed, 2 insertions(+), 3 deletions(-)

diff --git a/tools/perf/util/map.c b/tools/perf/util/map.c
index 7d37159..5f90dda 100644
--- a/tools/perf/util/map.c
+++ b/tools/perf/util/map.c
@@ -237,15 +237,14 @@ size_t map__fprintf(struct map *self, FILE *fp)
 
 size_t map__fprintf_dsoname(struct map *map, FILE *fp)
 {
-	const char *dsoname;
+	const char *dsoname = "[unknown]";
 
 	if (map && map->dso && (map->dso->name || map->dso->long_name)) {
 		if (symbol_conf.show_kernel_path && map->dso->long_name)
 			dsoname = map->dso->long_name;
 		else if (map->dso->name)
 			dsoname = map->dso->name;
-	} else
-		dsoname = "[unknown]";
+	}
 
 	return fprintf(fp, "%s", dsoname);
 }
-- 
1.7.1

[tip:perf/core] perf tools: Fix a compiling error in util/map.c

[tip-bot for Feng Tang]

Commit-ID:  8f28f19a87cb48d13570ba774a3e85776eb36bb4
Gitweb:     http://git.kernel.org/tip/8f28f19a87cb48d13570ba774a3e85776eb36bb4
Author:     Feng Tang <feng.tang@intel.com>
AuthorDate: Mon, 27 Aug 2012 15:38:26 +0800
Committer:  Arnaldo Carvalho de Melo <acme@redhat.com>
CommitDate: Thu, 20 Sep 2012 08:36:34 -0300

perf tools: Fix a compiling error in util/map.c

This patch fix a compile warning taken as error:

	CC util/map.o
	cc1: warnings being treated as errors
	util/map.c: In function ‘map__fprintf_dsoname’:
	util/map.c:240: error: ‘dsoname’ may be used uninitialized in this function
	make: *** [util/map.o] Error 1

Signed-off-by: Feng Tang <feng.tang@intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/r/1346053107-11946-3-git-send-email-feng.tang@intel.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
---
 tools/perf/util/map.c |    5 ++---
 1 files changed, 2 insertions(+), 3 deletions(-)

diff --git a/tools/perf/util/map.c b/tools/perf/util/map.c
index b442ee4..ead5316 100644
--- a/tools/perf/util/map.c
+++ b/tools/perf/util/map.c
@@ -243,15 +243,14 @@ size_t map__fprintf(struct map *self, FILE *fp)
 
 size_t map__fprintf_dsoname(struct map *map, FILE *fp)
 {
-	const char *dsoname;
+	const char *dsoname = "[unknown]";
 
 	if (map && map->dso && (map->dso->name || map->dso->long_name)) {
 		if (symbol_conf.show_kernel_path && map->dso->long_name)
 			dsoname = map->dso->long_name;
 		else if (map->dso->name)
 			dsoname = map->dso->name;
-	} else
-		dsoname = "[unknown]";
+	}
 
 	return fprintf(fp, "%s", dsoname);
 }

[PATCH 3/3] perf tools: Fix a misuse of for_each_set_bit() in session.c

[Feng Tang]

In regs_dump__printf() it use for_each_set_bit() for bit ops by
casting a (u64 *) to a (unsigned long *), this works for 64 bits
machine, but will fail on 32 bits ones.

Fix it by using the raw bit comparing method.

Signed-off-by: Feng Tang <feng.tang@intel.com>
---
 tools/perf/util/session.c |    9 ++++-----
 1 files changed, 4 insertions(+), 5 deletions(-)

diff --git a/tools/perf/util/session.c b/tools/perf/util/session.c
index f7bb7ae..afcea6c 100644
--- a/tools/perf/util/session.c
+++ b/tools/perf/util/session.c
@@ -904,11 +904,10 @@ static void regs_dump__printf(u64 mask, u64 *regs)
 {
 	unsigned rid, i = 0;
 
-	for_each_set_bit(rid, (unsigned long *) &mask, sizeof(mask) * 8) {
-		u64 val = regs[i++];
-
-		printf(".... %-5s 0x%" PRIx64 "\n",
-		       perf_reg_name(rid), val);
+	for (rid = 0; rid < 64; rid++) {
+		if (mask & (1 << rid))
+			printf(".... %-5s 0x%" PRIx64 "\n",
+				perf_reg_name(rid), regs[i++]);
 	}
 }
 
-- 
1.7.1

Re: [PATCH 3/3] perf tools: Fix a misuse of for_each_set_bit() in session.c

[Namhyung Kim]

Hi,

On Mon, 27 Aug 2012 15:38:27 +0800, Feng Tang wrote:
> In regs_dump__printf() it use for_each_set_bit() for bit ops by
> casting a (u64 *) to a (unsigned long *), this works for 64 bits
> machine, but will fail on 32 bits ones.
>
> Fix it by using the raw bit comparing method.

Did it really cause a build failure or a program error?  If not, it
looks better to keep using for_each_set_bit() interface.  How about
casting the @mask to (void *) ?

Thanks,
Namhyung


>
> Signed-off-by: Feng Tang <feng.tang@intel.com>
> ---
>  tools/perf/util/session.c |    9 ++++-----
>  1 files changed, 4 insertions(+), 5 deletions(-)
>
> diff --git a/tools/perf/util/session.c b/tools/perf/util/session.c
> index f7bb7ae..afcea6c 100644
> --- a/tools/perf/util/session.c
> +++ b/tools/perf/util/session.c
> @@ -904,11 +904,10 @@ static void regs_dump__printf(u64 mask, u64 *regs)
>  {
>  	unsigned rid, i = 0;
>  
> -	for_each_set_bit(rid, (unsigned long *) &mask, sizeof(mask) * 8) {
> -		u64 val = regs[i++];
> -
> -		printf(".... %-5s 0x%" PRIx64 "\n",
> -		       perf_reg_name(rid), val);
> +	for (rid = 0; rid < 64; rid++) {
> +		if (mask & (1 << rid))
> +			printf(".... %-5s 0x%" PRIx64 "\n",
> +				perf_reg_name(rid), regs[i++]);
>  	}
>  }

Re: [PATCH 3/3] perf tools: Fix a misuse of for_each_set_bit() in session.c

[Feng Tang]

Hi,

On Mon, 27 Aug 2012 16:59:07 +0900
Namhyung Kim <namhyung@kernel.org> wrote:

> Hi,
> 
> On Mon, 27 Aug 2012 15:38:27 +0800, Feng Tang wrote:
> > In regs_dump__printf() it use for_each_set_bit() for bit ops by
> > casting a (u64 *) to a (unsigned long *), this works for 64 bits
> > machine, but will fail on 32 bits ones.
> >
> > Fix it by using the raw bit comparing method.
> 
> Did it really cause a build failure or a program error?  If not, it
> looks better to keep using for_each_set_bit() interface.  How about
> casting the @mask to (void *) ?

It's a compile error:

cc1: warnings being treated as errors
util/session.c: In function ‘perf_session_deliver_event’:
util/include/linux/bitops.h:104: error: dereferencing pointer ‘p’ does break strict-aliasing rules
util/include/linux/bitops.h:96: error: dereferencing pointer ‘p’ does break strict-aliasing rules
util/session.c:907: note: initialized from here
util/include/linux/bitops.h:96: note: initialized from here
make: *** [util/session.o] Error 1
make: *** Waiting for unfinished jobs....

Casting the @mask to (void *) still sees the same error

Thanks,
Feng

Re: [PATCH 3/3] perf tools: Fix a misuse of for_each_set_bit() in session.c

[Akemi Yagi]

On Mon, 27 Aug 2012 16:18:15 +0800, Feng Tang wrote:

> It's a compile error:
> 
> cc1: warnings being treated as errors util/session.c: In function
> ‘perf_session_deliver_event’: util/include/linux/bitops.h:104: error:
> dereferencing pointer ‘p’ does break strict-aliasing rules
> util/include/linux/bitops.h:96: error: dereferencing pointer ‘p’ does
> break strict-aliasing rules util/session.c:907: note: initialized from
> here util/include/linux/bitops.h:96: note: initialized from here make:
> *** [util/session.o] Error 1
> make: *** Waiting for unfinished jobs....
> 
> Casting the @mask to (void *) still sees the same error
> 
> Thanks,
> Feng

We (ELRepo) are seeing the same error while building kernel 3.7 for RHEL 
6.3 on a 32-bit system. The perf package would build fine if we run 
rpmbuild with the 'i686' target specified on a 64-bit system.

http://elrepo.org/bugs/view.php?id=335

Akemi