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

Groups > linux.kernel > #1261388 > unrolled thread

[PATCH 3/5] perf tools: Parsing libbpf return value using err.h

Back to article view | Back to linux.kernel

This discussion starts older than the indexed window; earlier articles aren't shown. The article labeled Started by below is the oldest one visible, not the original post.

Contents

  [PATCH 3/5] perf tools: Parsing libbpf return value using err.h Wang Nan <wangnan0@huawei.com> - 2015-11-03 12:10 +0100
    Re: [PATCH 3/5] perf tools: Parsing libbpf return value using err.h Arnaldo Carvalho de Melo <acme@kernel.org> - 2015-11-03 16:20 +0100
      Re: [PATCH 3/5] perf tools: Parsing libbpf return value using err.h Arnaldo Carvalho de Melo <acme@kernel.org> - 2015-11-03 16:30 +0100

#1261388 — [PATCH 3/5] perf tools: Parsing libbpf return value using err.h

In following patches libbpf would encode error code using return
pointer instead of returnning a NULL pointer to indicate error. This
patch makes a preperation to enable perf correctly receive error code
from libbpf.

Signed-off-by: Wang Nan <wangnan0@huawei.com>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
---
 tools/perf/tests/llvm.c      | 2 +-
 tools/perf/util/bpf-loader.c | 8 ++++----
 2 files changed, 5 insertions(+), 5 deletions(-)

diff --git a/tools/perf/tests/llvm.c b/tools/perf/tests/llvm.c
index 512d362..38e0d9a 100644
--- a/tools/perf/tests/llvm.c
+++ b/tools/perf/tests/llvm.c
@@ -27,7 +27,7 @@ static int test__bpf_parsing(void *obj_buf, size_t obj_buf_sz)
 	struct bpf_object *obj;
 
 	obj = bpf_object__open_buffer(obj_buf, obj_buf_sz, NULL);
-	if (!obj)
+	if (IS_ERR(obj) || !obj)
 		return -1;
 	bpf_object__close(obj);
 	return 0;
diff --git a/tools/perf/util/bpf-loader.c b/tools/perf/util/bpf-loader.c
index 0c5d174..dd6fa27 100644
--- a/tools/perf/util/bpf-loader.c
+++ b/tools/perf/util/bpf-loader.c
@@ -59,9 +59,9 @@ struct bpf_object *bpf__prepare_load(const char *filename, bool source)
 	} else
 		obj = bpf_object__open(filename);
 
-	if (!obj) {
+	if (IS_ERR(obj) || !obj) {
 		pr_debug("bpf: failed to load %s\n", filename);
-		return ERR_PTR(-EINVAL);
+		return !obj ? ERR_PTR(-EINVAL) : obj;
 	}
 
 	return obj;
@@ -96,9 +96,9 @@ config_bpf_program(struct bpf_program *prog)
 	int err;
 
 	config_str = bpf_program__title(prog, false);
-	if (!config_str) {
+	if (IS_ERR(config_str) || !config_str) {
 		pr_debug("bpf: unable to get title for program\n");
-		return -EINVAL;
+		return !config_str ? -EINVAL : PTR_ERR(config_str);
 	}
 
 	priv = calloc(sizeof(*priv), 1);
-- 
1.8.3.4

--
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]

#1261571

Em Tue, Nov 03, 2015 at 10:44:44AM +0000, Wang Nan escreveu:
> In following patches libbpf would encode error code using return
> pointer instead of returnning a NULL pointer to indicate error. This
> patch makes a preperation to enable perf correctly receive error code
> from libbpf.
> 
> Signed-off-by: Wang Nan <wangnan0@huawei.com>
> Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
> ---
>  tools/perf/tests/llvm.c      | 2 +-
>  tools/perf/util/bpf-loader.c | 8 ++++----
>  2 files changed, 5 insertions(+), 5 deletions(-)
> 
> diff --git a/tools/perf/tests/llvm.c b/tools/perf/tests/llvm.c
> index 512d362..38e0d9a 100644
> --- a/tools/perf/tests/llvm.c
> +++ b/tools/perf/tests/llvm.c
> @@ -27,7 +27,7 @@ static int test__bpf_parsing(void *obj_buf, size_t obj_buf_sz)
>  	struct bpf_object *obj;
>  
>  	obj = bpf_object__open_buffer(obj_buf, obj_buf_sz, NULL);
> -	if (!obj)
> +	if (IS_ERR(obj) || !obj)

Well, since we've adopted IS_ERR() from the kernel, we better try to
follow how it is used there, no?

Since you move to use ERR_PTR(), you probably will never return NULL,
right? So whay the (|| !obj) part?

The kernel has an IS_ERR_OR_NULL() interface tho, trying to figure out
when that would be appropriate...

- Arnaldo

>  		return -1;
>  	bpf_object__close(obj);
>  	return 0;
> diff --git a/tools/perf/util/bpf-loader.c b/tools/perf/util/bpf-loader.c
> index 0c5d174..dd6fa27 100644
> --- a/tools/perf/util/bpf-loader.c
> +++ b/tools/perf/util/bpf-loader.c
> @@ -59,9 +59,9 @@ struct bpf_object *bpf__prepare_load(const char *filename, bool source)
>  	} else
>  		obj = bpf_object__open(filename);
>  
> -	if (!obj) {
> +	if (IS_ERR(obj) || !obj) {
>  		pr_debug("bpf: failed to load %s\n", filename);
> -		return ERR_PTR(-EINVAL);
> +		return !obj ? ERR_PTR(-EINVAL) : obj;
>  	}
>  
>  	return obj;
> @@ -96,9 +96,9 @@ config_bpf_program(struct bpf_program *prog)
>  	int err;
>  
>  	config_str = bpf_program__title(prog, false);
> -	if (!config_str) {
> +	if (IS_ERR(config_str) || !config_str) {
>  		pr_debug("bpf: unable to get title for program\n");
> -		return -EINVAL;
> +		return !config_str ? -EINVAL : PTR_ERR(config_str);
>  	}
>  
>  	priv = calloc(sizeof(*priv), 1);
> -- 
> 1.8.3.4
--
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] | [next] | [standalone]

#1261596

Em Tue, Nov 03, 2015 at 12:17:16PM -0300, Arnaldo Carvalho de Melo escreveu:
> Em Tue, Nov 03, 2015 at 10:44:44AM +0000, Wang Nan escreveu:
> > +++ b/tools/perf/tests/llvm.c
> > @@ -27,7 +27,7 @@ static int test__bpf_parsing(void *obj_buf, size_t obj_buf_sz)
> >  	struct bpf_object *obj;

> >  	obj = bpf_object__open_buffer(obj_buf, obj_buf_sz, NULL);
> > -	if (!obj)
> > +	if (IS_ERR(obj) || !obj)
 
> Well, since we've adopted IS_ERR() from the kernel, we better try to
> follow how it is used there, no?
 
> Since you move to use ERR_PTR(), you probably will never return NULL,
> right? So whay the (|| !obj) part?
 
> The kernel has an IS_ERR_OR_NULL() interface tho, trying to figure out
> when that would be appropriate...

I think you should change the error reporting convention _and_ the users
at the same time, i.e. please merge this patch with the next and use
plain:

	if (IS_ERR(obj))

Instead, ok?

- 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