linux-next: build failure after merge of the y2038 tree

Message ID 20181217201146.170ca9b4@canb.auug.org.au
State New
Headers show
Series
  • linux-next: build failure after merge of the y2038 tree
Related show

Commit Message

Stephen Rothwell Dec. 17, 2018, 9:11 a.m.
Hi Arnd,

After merging the y2038 tree, today's linux-next build (x86_64
allmodconfig) failed like this:

fs/cifs/dfs_cache.c: In function 'cache_entry_expired':
fs/cifs/dfs_cache.c:106:7: error: implicit declaration of function 'current_kernel_time64'; did you mean 'core_kernel_text'? [-Werror=implicit-function-declaration]
  ts = current_kernel_time64();
       ^~~~~~~~~~~~~~~~~~~~~
       core_kernel_text
fs/cifs/dfs_cache.c:106:5: error: incompatible types when assigning to type 'struct timespec64' from type 'int'
  ts = current_kernel_time64();
     ^
fs/cifs/dfs_cache.c: In function 'get_expire_time':
fs/cifs/dfs_cache.c:342:24: error: incompatible type for argument 1 of 'timespec64_add'
  return timespec64_add(current_kernel_time64(), ts);
                        ^~~~~~~~~~~~~~~~~~~~~~~
In file included from include/linux/restart_block.h:10,
                 from include/linux/thread_info.h:13,
                 from arch/x86/include/asm/preempt.h:7,
                 from include/linux/preempt.h:78,
                 from include/linux/rcupdate.h:40,
                 from fs/cifs/dfs_cache.c:8:
include/linux/time64.h:66:66: note: expected 'struct timespec64' but argument is of type 'int'
 static inline struct timespec64 timespec64_add(struct timespec64 lhs,
                                                ~~~~~~~~~~~~~~~~~~^~~
fs/cifs/dfs_cache.c:343:1: warning: control reaches end of non-void function [-Wreturn-type]
 }
 ^

Caused by commit

  ccea641b6742 ("timekeeping: remove obsolete time accessors")

interacting with commit

  34a44fb160f9 ("cifs: Add DFS cache routines")

from the cifs tree.

I have applied the following merge fix patch (better versions welcome):

From: Stephen Rothwell <sfr@canb.auug.org.au>
Date: Mon, 17 Dec 2018 20:03:28 +1100
Subject: [PATCH] cifs: update for current_kernel_time64() removal

Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au>
---
 fs/cifs/dfs_cache.c | 6 ++++--
 1 file changed, 4 insertions(+), 2 deletions(-)

Comments

Arnd Bergmann Dec. 17, 2018, 12:49 p.m. | #1
On Mon, Dec 17, 2018 at 10:11 AM Stephen Rothwell <sfr@canb.auug.org.au> wrote:

> I have applied the following merge fix patch (better versions welcome):
>
> From: Stephen Rothwell <sfr@canb.auug.org.au>
> Date: Mon, 17 Dec 2018 20:03:28 +1100
> Subject: [PATCH] cifs: update for current_kernel_time64() removal
>
> Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au>
> ---

Your patch looks correct, and the conflict should be easy to
resolve by Steve merging this into his tree, as
ktime_get_coarse_real_ts64() is readily available in mainline
kernels.

Acked-by: Arnd Bergmann <arnd@arndb.de>

>  fs/cifs/dfs_cache.c | 6 ++++--
>  1 file changed, 4 insertions(+), 2 deletions(-)
>
> diff --git a/fs/cifs/dfs_cache.c b/fs/cifs/dfs_cache.c
> index 70f9c9e2175c..d20cc94d7abd 100644
> --- a/fs/cifs/dfs_cache.c
> +++ b/fs/cifs/dfs_cache.c
> @@ -103,7 +103,7 @@ static inline bool cache_entry_expired(const struct dfs_cache_entry *ce)
>  {
>         struct timespec64 ts;
>
> -       ts = current_kernel_time64();
> +       ktime_get_coarse_real_ts64(&ts);
>         return timespec64_compare(&ts, &ce->ce_etime) >= 0;
>  }
>
> @@ -338,8 +338,10 @@ static inline struct timespec64 get_expire_time(int ttl)
>                 .tv_sec = ttl,
>                 .tv_nsec = 0,
>         };
> +       struct timespec64 now;
>
> -       return timespec64_add(current_kernel_time64(), ts);
> +       ktime_get_coarse_real_ts64(&now);
> +       return timespec64_add(now, ts);
>  }

In case efficiency is a concern: using ktime_t with
ktime_get_coarse_real() may be much faster if we decide to
abandon the coarse-grained timespec64 accessors in the future.

Also, I wonder if the expiration here has to use CLOCK_REALTIME,
since that is affected by leap second adjustment and
clock_settime().

In other modules, we usually concluded that it should be either
CLOCK_MONOTONIC or CLOCK_BOOTTIME, depending on whether
you want the expiration timer to keep ticking during suspend.

        Arnd
Steve French Dec. 17, 2018, 10:25 p.m. | #2
Pushed to cifs-2.6.git for-next (after minor commit description text
cleanup and adding acked-by and reviewed-by)

Paulo/Aurelien,
Let me know your thoughts about Arnd's additional comments which were
interesting.

On Mon, Dec 17, 2018 at 6:49 AM Arnd Bergmann <arnd@arndb.de> wrote:
>
> On Mon, Dec 17, 2018 at 10:11 AM Stephen Rothwell <sfr@canb.auug.org.au> wrote:
>
> > I have applied the following merge fix patch (better versions welcome):
> >
> > From: Stephen Rothwell <sfr@canb.auug.org.au>
> > Date: Mon, 17 Dec 2018 20:03:28 +1100
> > Subject: [PATCH] cifs: update for current_kernel_time64() removal
> >
> > Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au>
> > ---
>
> Your patch looks correct, and the conflict should be easy to
> resolve by Steve merging this into his tree, as
> ktime_get_coarse_real_ts64() is readily available in mainline
> kernels.
>
> Acked-by: Arnd Bergmann <arnd@arndb.de>
>
> >  fs/cifs/dfs_cache.c | 6 ++++--
> >  1 file changed, 4 insertions(+), 2 deletions(-)
> >
> > diff --git a/fs/cifs/dfs_cache.c b/fs/cifs/dfs_cache.c
> > index 70f9c9e2175c..d20cc94d7abd 100644
> > --- a/fs/cifs/dfs_cache.c
> > +++ b/fs/cifs/dfs_cache.c
> > @@ -103,7 +103,7 @@ static inline bool cache_entry_expired(const struct dfs_cache_entry *ce)
> >  {
> >         struct timespec64 ts;
> >
> > -       ts = current_kernel_time64();
> > +       ktime_get_coarse_real_ts64(&ts);
> >         return timespec64_compare(&ts, &ce->ce_etime) >= 0;
> >  }
> >
> > @@ -338,8 +338,10 @@ static inline struct timespec64 get_expire_time(int ttl)
> >                 .tv_sec = ttl,
> >                 .tv_nsec = 0,
> >         };
> > +       struct timespec64 now;
> >
> > -       return timespec64_add(current_kernel_time64(), ts);
> > +       ktime_get_coarse_real_ts64(&now);
> > +       return timespec64_add(now, ts);
> >  }
>
> In case efficiency is a concern: using ktime_t with
> ktime_get_coarse_real() may be much faster if we decide to
> abandon the coarse-grained timespec64 accessors in the future.
>
> Also, I wonder if the expiration here has to use CLOCK_REALTIME,
> since that is affected by leap second adjustment and
> clock_settime().
>
> In other modules, we usually concluded that it should be either
> CLOCK_MONOTONIC or CLOCK_BOOTTIME, depending on whether
> you want the expiration timer to keep ticking during suspend.
>
>         Arnd

Patch

diff --git a/fs/cifs/dfs_cache.c b/fs/cifs/dfs_cache.c
index 70f9c9e2175c..d20cc94d7abd 100644
--- a/fs/cifs/dfs_cache.c
+++ b/fs/cifs/dfs_cache.c
@@ -103,7 +103,7 @@  static inline bool cache_entry_expired(const struct dfs_cache_entry *ce)
 {
 	struct timespec64 ts;
 
-	ts = current_kernel_time64();
+	ktime_get_coarse_real_ts64(&ts);
 	return timespec64_compare(&ts, &ce->ce_etime) >= 0;
 }
 
@@ -338,8 +338,10 @@  static inline struct timespec64 get_expire_time(int ttl)
 		.tv_sec = ttl,
 		.tv_nsec = 0,
 	};
+	struct timespec64 now;
 
-	return timespec64_add(current_kernel_time64(), ts);
+	ktime_get_coarse_real_ts64(&now);
+	return timespec64_add(now, ts);
 }
 
 /* Allocate a new DFS target */