Fix sparse warnings:
fs/lockd/clntproc.c:57:6: warning: symbol 'nlmclnt_put_lockowner' was not declared. Should it be static?
fs/lockd/svclock.c:409:35: warning: symbol 'nlmsvc_lock_ops' was not declared. Should it be static?
Reported-by: Hulk Robot <[email protected]>
Signed-off-by: YueHaibing <[email protected]>
---
fs/lockd/clntproc.c | 2 +-
fs/lockd/svclock.c | 2 +-
2 files changed, 2 insertions(+), 2 deletions(-)
diff --git a/fs/lockd/clntproc.c b/fs/lockd/clntproc.c
index 0ff8ad4..b11f2af 100644
--- a/fs/lockd/clntproc.c
+++ b/fs/lockd/clntproc.c
@@ -54,7 +54,7 @@ nlmclnt_get_lockowner(struct nlm_lockowner *lockowner)
return lockowner;
}
-void nlmclnt_put_lockowner(struct nlm_lockowner *lockowner)
+static void nlmclnt_put_lockowner(struct nlm_lockowner *lockowner)
{
if (!refcount_dec_and_lock(&lockowner->count, &lockowner->host->h_lock))
return;
diff --git a/fs/lockd/svclock.c b/fs/lockd/svclock.c
index 5f9f19b..61d3cc2 100644
--- a/fs/lockd/svclock.c
+++ b/fs/lockd/svclock.c
@@ -406,7 +406,7 @@ static void nlmsvc_locks_release_private(struct file_lock *fl)
nlmsvc_put_lockowner((struct nlm_lockowner *)fl->fl_owner);
}
-const struct file_lock_operations nlmsvc_lock_ops = {
+static const struct file_lock_operations nlmsvc_lock_ops = {
.fl_copy_lock = nlmsvc_locks_copy_lock,
.fl_release_private = nlmsvc_locks_release_private,
};
--
2.7.4
On Tue, May 28, 2019 at 06:49:13AM -0400, Benjamin Coddington wrote:
> Maintainers, what's the best thing to do here: fold these into
> another patch version and post it (add attribution)? Add it as
> another patch at the end of the series?
Either would be fine. Yeah, if it was folded in then we'd add a line
like
[[email protected]: make symbols static to fix sparse warnings]
But I'll probably just add it on to the end for now. No need for you to
do anything.
> I have learned my lesson: add sparse to my workflow.
I dunno, I wonder if we're better off just leaving it to this CI bot.
It seems like a more efficient use of time overall than making every
contributor run it.
--b.
> Ben
>
> On 28 May 2019, at 5:06, YueHaibing wrote:
>
> >Fix sparse warnings:
> >
> >fs/lockd/clntproc.c:57:6: warning: symbol 'nlmclnt_put_lockowner'
> >was not declared. Should it be static?
> >fs/lockd/svclock.c:409:35: warning: symbol 'nlmsvc_lock_ops' was
> >not declared. Should it be static?
> >
> >Reported-by: Hulk Robot <[email protected]>
> >Signed-off-by: YueHaibing <[email protected]>
> >---
> > fs/lockd/clntproc.c | 2 +-
> > fs/lockd/svclock.c | 2 +-
> > 2 files changed, 2 insertions(+), 2 deletions(-)
> >
> >diff --git a/fs/lockd/clntproc.c b/fs/lockd/clntproc.c
> >index 0ff8ad4..b11f2af 100644
> >--- a/fs/lockd/clntproc.c
> >+++ b/fs/lockd/clntproc.c
> >@@ -54,7 +54,7 @@ nlmclnt_get_lockowner(struct nlm_lockowner
> >*lockowner)
> > return lockowner;
> > }
> >
> >-void nlmclnt_put_lockowner(struct nlm_lockowner *lockowner)
> >+static void nlmclnt_put_lockowner(struct nlm_lockowner *lockowner)
> > {
> > if (!refcount_dec_and_lock(&lockowner->count,
> >&lockowner->host->h_lock))
> > return;
> >diff --git a/fs/lockd/svclock.c b/fs/lockd/svclock.c
> >index 5f9f19b..61d3cc2 100644
> >--- a/fs/lockd/svclock.c
> >+++ b/fs/lockd/svclock.c
> >@@ -406,7 +406,7 @@ static void
> >nlmsvc_locks_release_private(struct file_lock *fl)
> > nlmsvc_put_lockowner((struct nlm_lockowner *)fl->fl_owner);
> > }
> >
> >-const struct file_lock_operations nlmsvc_lock_ops = {
> >+static const struct file_lock_operations nlmsvc_lock_ops = {
> > .fl_copy_lock = nlmsvc_locks_copy_lock,
> > .fl_release_private = nlmsvc_locks_release_private,
> > };
> >--
> >2.7.4
> On May 28, 2019, at 11:13 AM, J. Bruce Fields <[email protected]> wrote:
>
> On Tue, May 28, 2019 at 06:49:13AM -0400, Benjamin Coddington wrote:
>> Maintainers, what's the best thing to do here: fold these into
>> another patch version and post it (add attribution)? Add it as
>> another patch at the end of the series?
>
> Either would be fine. Yeah, if it was folded in then we'd add a line
> like
>
> [[email protected]: make symbols static to fix sparse warnings]
>
> But I'll probably just add it on to the end for now. No need for you to
> do anything.
>
>> I have learned my lesson: add sparse to my workflow.
>
> I dunno, I wonder if we're better off just leaving it to this CI bot.
> It seems like a more efficient use of time overall than making every
> contributor run it.
Occasionally sparse can catch a real problem that breaks bisectability.
Better to do this kind of checking early, and ensure that you test those
sparse-fixed bits.
--
Chuck Lever