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

Groups > linux.kernel > #1639183 > unrolled thread

[PATCH] FS: Fixing return type of unsigned_offsets

Started byPushkar Jambhlekar <pushkar.iit@gmail.com>
First post2017-05-11 06:30 +0200
Last post2017-05-11 06:50 +0200
Articles 7 — 4 participants

Back to article view | Back to linux.kernel

Contents

  [PATCH] FS: Fixing return type of unsigned_offsets Pushkar Jambhlekar <pushkar.iit@gmail.com> - 2017-05-11 06:30 +0200
    Re: [PATCH] FS: Fixing return type of unsigned_offsets Pushkar Jambhlekar <pushkar.iit@gmail.com> - 2017-05-11 06:50 +0200
      Re: [PATCH] FS: Fixing return type of unsigned_offsets Joe Perches <joe@perches.com> - 2017-05-11 07:00 +0200
        Re: [PATCH] FS: Fixing return type of unsigned_offsets Pushkar Jambhlekar <pushkar.iit@gmail.com> - 2017-05-11 07:10 +0200
          Re: [PATCH] FS: Fixing return type of unsigned_offsets Al Viro <viro@ZenIV.linux.org.uk> - 2017-05-11 08:10 +0200
            Re: [PATCH] FS: Fixing return type of unsigned_offsets Luc Van Oostenryck <luc.vanoostenryck@gmail.com> - 2017-05-11 22:10 +0200
    Re: [PATCH] FS: Fixing return type of unsigned_offsets Joe Perches <joe@perches.com> - 2017-05-11 06:50 +0200

#1639183 — [PATCH] FS: Fixing return type of unsigned_offsets

FromPushkar Jambhlekar <pushkar.iit@gmail.com>
Date2017-05-11 06:30 +0200
Subject[PATCH] FS: Fixing return type of unsigned_offsets
Message-ID<tFNTH-1w1-1@gated-at.bofh.it>
Fixing Sparse warning. It should return bool, instead it returns
int. 

Signed-off-by: Pushkar Jambhlekar <pushkar.iit@gmail.com>
---
 fs/read_write.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/fs/read_write.c b/fs/read_write.c
index 47c1d44..d672830 100644
--- a/fs/read_write.c
+++ b/fs/read_write.c
@@ -33,9 +33,9 @@ const struct file_operations generic_ro_fops = {
 
 EXPORT_SYMBOL(generic_ro_fops);
 
-static inline int unsigned_offsets(struct file *file)
+static inline bool unsigned_offsets(struct file *file)
 {
-	return file->f_mode & FMODE_UNSIGNED_OFFSET;
+	return !!(file->f_mode & FMODE_UNSIGNED_OFFSET);
 }
 
 /**
-- 
2.7.4

[toc] | [next] | [standalone]

#1639187

FromPushkar Jambhlekar <pushkar.iit@gmail.com>
Date2017-05-11 06:50 +0200
Message-ID<tFOd4-1Ci-5@gated-at.bofh.it>
In reply to#1639183
Should I change my implementation, i.e. remove '!!'?

On Thu, May 11, 2017 at 10:09 AM, Joe Perches <joe@perches.com> wrote:
> On Thu, 2017-05-11 at 09:57 +0530, Pushkar Jambhlekar wrote:
>> Fixing Sparse warning. It should return bool, instead it returns
>> int.
> []
>> diff --git a/fs/read_write.c b/fs/read_write.c
> []
>> @@ -33,9 +33,9 @@ const struct file_operations generic_ro_fops = {
>>
>>  EXPORT_SYMBOL(generic_ro_fops);
>>
>> -static inline int unsigned_offsets(struct file *file)
>> +static inline bool unsigned_offsets(struct file *file)
>>  {
>> -     return file->f_mode & FMODE_UNSIGNED_OFFSET;
>> +     return !!(file->f_mode & FMODE_UNSIGNED_OFFSET);
>
> trivia: the !! isn't necessary as by definition
> all non-zero assigns to bool are converted to 1.
>



-- 
Jambhlekar Pushkar Arun

[toc] | [prev] | [next] | [standalone]

#1639191

FromJoe Perches <joe@perches.com>
Date2017-05-11 07:00 +0200
Message-ID<tFOmJ-1Gx-7@gated-at.bofh.it>
In reply to#1639187
On Thu, 2017-05-11 at 10:13 +0530, Pushkar Jambhlekar wrote:
> Should I change my implementation, i.e. remove '!!'?

That'd be up to Al.

At least one implementation using similar bit comparisons
in fs/*.c does not use !!

fs/locks.c:static bool lease_breaking(struct file_lock *fl)
fs/locks.c-{
fs/locks.c-     return fl->fl_flags & (FL_UNLOCK_PENDING | FL_DOWNGRADE_PENDING)
fs/locks.c-}

I didn't look very hard.

[toc] | [prev] | [next] | [standalone]

#1639193

FromPushkar Jambhlekar <pushkar.iit@gmail.com>
Date2017-05-11 07:10 +0200
Message-ID<tFOwp-22k-3@gated-at.bofh.it>
In reply to#1639191
If I remove '!!', sparse flags warning:

fs/read_write.c:38:29: warning: incorrect type in return expression
(different base types)
fs/read_write.c:38:29:    expected bool
fs/read_write.c:38:29:    got restricted fmode_t

It means explicit conversion is needed.

On Thu, May 11, 2017 at 10:25 AM, Joe Perches <joe@perches.com> wrote:
> On Thu, 2017-05-11 at 10:13 +0530, Pushkar Jambhlekar wrote:
>> Should I change my implementation, i.e. remove '!!'?
>
> That'd be up to Al.
>
> At least one implementation using similar bit comparisons
> in fs/*.c does not use !!
>
> fs/locks.c:static bool lease_breaking(struct file_lock *fl)
> fs/locks.c-{
> fs/locks.c-     return fl->fl_flags & (FL_UNLOCK_PENDING | FL_DOWNGRADE_PENDING)
> fs/locks.c-}
>
> I didn't look very hard.



-- 
Jambhlekar Pushkar Arun

[toc] | [prev] | [next] | [standalone]

#1639201

FromAl Viro <viro@ZenIV.linux.org.uk>
Date2017-05-11 08:10 +0200
Message-ID<tFPst-2B9-1@gated-at.bofh.it>
In reply to#1639193
On Thu, May 11, 2017 at 10:34:02AM +0530, Pushkar Jambhlekar wrote:
> If I remove '!!', sparse flags warning:
> 
> fs/read_write.c:38:29: warning: incorrect type in return expression
> (different base types)
> fs/read_write.c:38:29:    expected bool
> fs/read_write.c:38:29:    got restricted fmode_t
> 
> It means explicit conversion is needed.

FVO"needed" equal to "needed to make sparse STFU"?  If anything, that's
sparse being wrong - evaluate.c:check_assignment_type() should do
		if (t == &ctype_bool) {
			if (is_fouled_type(s))
				warning((*rp)->pos, "%s degrades to integer",
					show_typename(s->ctype.base_type));
			goto Cast;
                }
right after
                } else if (!(sclass & TYPE_RESTRICT))
                        goto Cast;

[toc] | [prev] | [next] | [standalone]

#1639993

FromLuc Van Oostenryck <luc.vanoostenryck@gmail.com>
Date2017-05-11 22:10 +0200
Message-ID<tG2zn-2xc-1@gated-at.bofh.it>
In reply to#1639201
On Thu, May 11, 2017 at 8:07 AM, Al Viro <viro@zeniv.linux.org.uk> wrote:
> FVO"needed" equal to "needed to make sparse STFU"?  If anything, that's
> sparse being wrong - evaluate.c:check_assignment_type() should do
>                 if (t == &ctype_bool) {
>                         if (is_fouled_type(s))
>                                 warning((*rp)->pos, "%s degrades to integer",
>                                         show_typename(s->ctype.base_type));
>                         goto Cast;
>                 }
> right after
>                 } else if (!(sclass & TYPE_RESTRICT))
>                         goto Cast;

What about an explicit cast of restricted types to bool?
I think we would want the equivalent of this patch for those too.

-- Luc

[toc] | [prev] | [next] | [standalone]

#1639189

FromJoe Perches <joe@perches.com>
Date2017-05-11 06:50 +0200
Message-ID<tFOd4-1Ci-7@gated-at.bofh.it>
In reply to#1639183
On Thu, 2017-05-11 at 09:57 +0530, Pushkar Jambhlekar wrote:
> Fixing Sparse warning. It should return bool, instead it returns
> int.
[]
> diff --git a/fs/read_write.c b/fs/read_write.c
[]
> @@ -33,9 +33,9 @@ const struct file_operations generic_ro_fops = {
>  
>  EXPORT_SYMBOL(generic_ro_fops);
>  
> -static inline int unsigned_offsets(struct file *file)
> +static inline bool unsigned_offsets(struct file *file)
>  {
> -	return file->f_mode & FMODE_UNSIGNED_OFFSET;
> +	return !!(file->f_mode & FMODE_UNSIGNED_OFFSET);

trivia: the !! isn't necessary as by definition
all non-zero assigns to bool are converted to 1.

[toc] | [prev] | [standalone]

Back to top | Article view | linux.kernel

csiph-web