Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]


Groups > linux.kernel > #1241895 > unrolled thread

[PATCH] perf: Correctly identify anon_hugepage when generating map

Started byYannick Brosseau <scientist@fb.com>
First post2015-10-08 02:40 +0200
Last post2015-10-13 17:00 +0200
Articles 2 — 2 participants

Back to article view | Back to linux.kernel


Contents

  [PATCH] perf: Correctly identify anon_hugepage when generating map Yannick Brosseau <scientist@fb.com> - 2015-10-08 02:40 +0200
    Re: [PATCH] perf: Correctly identify anon_hugepage when generating  map Arnaldo Carvalho de Melo <acme@kernel.org> - 2015-10-13 17:00 +0200

#1241895 — [PATCH] perf: Correctly identify anon_hugepage when generating map

FromYannick Brosseau <scientist@fb.com>
Date2015-10-08 02:40 +0200
Subject[PATCH] perf: Correctly identify anon_hugepage when generating map
Message-ID<qh7pv-4eQ-21@gated-at.bofh.it>
When parsing /proc/xxx/maps, the sscanf in perf_event__synthesize_mmap_events
truncate the map name at the space in "/anon_hugepage (deleted)".
is_anon_memory then only receive the string "/anon_hugepage" and do not detect it.
We change is_anon_memory to only compare the first part of the string
effectively ignoring if the (deleted) part is there or not.

Signed-off-by: Yannick Brosseau <scientist@fb.com>
---
 tools/perf/util/map.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/tools/perf/util/map.c b/tools/perf/util/map.c
index b1c475d..9f7face 100644
--- a/tools/perf/util/map.c
+++ b/tools/perf/util/map.c
@@ -26,8 +26,8 @@ const char *map_type__name[MAP__NR_TYPES] = {
 static inline int is_anon_memory(const char *filename)
 {
 	return !strcmp(filename, "//anon") ||
-	       !strcmp(filename, "/dev/zero (deleted)") ||
-	       !strcmp(filename, "/anon_hugepage (deleted)");
+	       !strncmp(filename, "/dev/zero", sizeof("/dev/zero")) ||
+	       !strncmp(filename, "/anon_hugepage", sizeof("/anon_hugepage"));
 }
 
 static inline int is_no_dso_memory(const char *filename)
-- 
2.6.0

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

[toc] | [next] | [standalone]


#1245819 — Re: [PATCH] perf: Correctly identify anon_hugepage when generating map

FromArnaldo Carvalho de Melo <acme@kernel.org>
Date2015-10-13 17:00 +0200
SubjectRe: [PATCH] perf: Correctly identify anon_hugepage when generating map
Message-ID<qj9dx-1jL-45@gated-at.bofh.it>
In reply to#1241895
Em Wed, Oct 07, 2015 at 05:31:46PM -0700, Yannick Brosseau escreveu:
> When parsing /proc/xxx/maps, the sscanf in perf_event__synthesize_mmap_events
> truncate the map name at the space in "/anon_hugepage (deleted)".
> is_anon_memory then only receive the string "/anon_hugepage" and do not detect it.
> We change is_anon_memory to only compare the first part of the string
> effectively ignoring if the (deleted) part is there or not.
> 
> +++ b/tools/perf/util/map.c
> @@ -26,8 +26,8 @@ const char *map_type__name[MAP__NR_TYPES] = {
>  static inline int is_anon_memory(const char *filename)
>  {
>  	return !strcmp(filename, "//anon") ||
> -	       !strcmp(filename, "/dev/zero (deleted)") ||
> -	       !strcmp(filename, "/anon_hugepage (deleted)");
> +	       !strncmp(filename, "/dev/zero", sizeof("/dev/zero")) ||
> +	       !strncmp(filename, "/anon_hugepage", sizeof("/anon_hugepage"));

Have you tested this?

[acme@zoo ~]$ cat strncmp_sizeof.c
#include <string.h>
#include <stdio.h>

int main(void)
{
#define pattern "/anon_hugepage"
	size_t sz = sizeof(pattern);

	return printf("sizeof=%zd\nstrncmp(sz    )=%d\nstrncmp(sz - 1)=%d\n", sz,
			strncmp("/anon_hugepage (deleted)", pattern, sz),
			strncmp("/anon_hugepage (deleted)", pattern, sz - 1));
}
[acme@zoo ~]$ make strncmp_sizeof
cc     strncmp_sizeof.c   -o strncmp_sizeof
[acme@zoo ~]$ ./strncmp_sizeof
sizeof=15
strncmp(sz    )=32
strncmp(sz - 1)=0
[acme@zoo ~]$

Hint, you're not considering the '\0'.

- Arnaldo
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

[toc] | [prev] | [standalone]


Back to top | Article view | linux.kernel


csiph-web